The Oxford Handbook of Elizabeth Anscombe 9780190887353, 0190887354

Citation preview

T h e Ox f o r d H a n d b o o k o f

ELIZABETH A N S C OM B E

The Oxford Handbook of

ELIZABETH ANSCOMBE Edited by

ROGER TEICHMANN

Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and certain other countries. Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America. © Oxford University Press 2022 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by license, or under terms agreed with the appropriate reproduction rights organization. Inquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above. You must not circulate this work in any other form and you must impose this same condition on any acquirer. CIP data is on file at the Library of Congress ISBN 978–​0–​19–​088735–​3 DOI: 10.1093/oxfordhb/9780190887353.001.0001 1 3 5 7 9 8 6 4 2 Printed by Integrated Books International, United States of America

Contents

Title Abbreviations  List of Contributors 

ix xiii

Introduction  Roger Teichmann

1

PA RT I :   I N T E N T ION 1. On Anscombe on Practical Knowledge and Practical Truth  Lucy Campbell

7

2. Intention in Action  Charles F. Capps

33

3. Intention, Knowledge, and Responsibility  Rémi Clot-​Goudard

53

4. Practical Knowledge and Testimony  Johannes Roessler

72

PA RT I I :   E T H IC A L T H E ORY 5. Anscombe’s Three Theses after Sixty Years: Modern Moral Philosophy, Polemic, and ‘Modern Moral Philosophy’  Sophie Grace Chappell

91

6. Practical Truth, Ethical Naturalism, and the Constitution of Agency in Anscombe’s Ethics  John Hacker-​Wright

118

7. Criterialism  Gavin Lawrence

138

8. Anscombe on Double Effect and Intended Consequences  Cyrille Michon

173

vi   Contents

9. Anscombe on Ought  Anselm Winfried Müller

196

PA RT I I I :   H UM A N L I F E 10. Justice and Murder: The Backstory to Anscombe’s ‘Modern Moral Philosophy’  John Berkman

225

11. Anscombe on Euthanasia as Murder  David Albert Jones

271

12. The Knowledge of Human Dignity  Micah Lott

292

13. Life and Other Basic Rights in Anscombe  Katharina Nieswandt

308

14. Anscombe and Sexual Ethics  Duncan Richter

324

15. Linguistic Idealism and Human Essence  Rachael Wiseman

344

PA RT I V:   T H E F I R S T P E R S ON 16. The First Person, Self-​Consciousness, and Action  Valérie Aucouturier

361

17. Anscombe and Self-​Consciousness  Adrian Haddock

383

18. The First Person and ‘The First Person’  Harold Noonan

397

PA RT V:   A N S C OM B E ON / ​A N D OT H E R P H I L O S OP H E R S 19. Anscombe’s Wittgenstein  Joel Backström

415

20. Anscombe and Aquinas  John Haldane

442

Contents   vii

21. Ethics and Action Theory: An Unhappy Divorce  Constantine Sandis

469

22. Anscombe and Wittgenstein on Knowledge ‘without Observation’  Harold Teichman

490

Index 

509

Title Abbreviations

ACTP

‘Analysis Competition—​Tenth Problem’, in CP2.

AIDE

‘Action, Intention and “Double Effect” ’ , in GG1.

AM

‘Authority in Morals’, in CP3.

APSM ‘Analytical Philosophy and the Spirituality of Man’, in GG1. BF

‘On Brute Facts’, in CP3.

CA

‘The Causation of Action’, in GG1.

CC

‘Contraception and Chastity’, in GG2.

CCVM ‘Contraception, Chastity and the Vocation of Marriage’, in in GG2. CD

‘Causality and Determination’, in CP2.

CNFP

‘On Contraception and Natural Family Planning’ in GG2.

CP1

Collected Philosophical Papers, Vol. 1: From Parmenides to Wittgenstein (Oxford: Blackwell, 1981).

CP2

Collected Philosophical Papers, Vol. 2: Metaphysics and the Philosophy of Mind (Oxford: Blackwell, 1981).

CP3

Collected Philosophical Papers, Vol. 3: Ethics, Religion and Politics (Oxford: Blackwell, 1981).

DHB

‘Dignity of the Human Being’, in GG1.

FP

‘The First Person’, in CP2.

FWW

‘On the Form of Wittgenstein’s Writing’, in GG3.

GB

‘Good and Bad Human Action’, in GG1.

GG1

Human Life, Action and Ethics: Essays by G. E. M. Anscombe, edited by Luke Gormally and Mary Geach (Exeter: Imprint Academic, 2005).

GG2

Faith in a Hard Ground: Essays by G. E. M. Anscombe, edited by Luke Gormally and Mary Geach (Exeter: Imprint Academic, 2008).

GG3

From Plato to Wittgenstein: Essays by G. E. M. Anscombe, edited by Luke Gormally and Mary Geach (Exeter: Imprint Academic, 2011).

GG4

Logic, Truth and Meaning: Essays by G. E. M. Anscombe, edited by Luke Gormally and Mary Geach (Exeter: Imprint Academic, 2015).

GSE

‘Grammar, Structure and Essence’, in GG4.

GW

‘Glanville Williams The Sanctity of Life and the Criminal Law: A Review’, in GG1.

x   Title Abbreviations HE

‘Human Essence’, in GG1.

HJC

‘Hume and Julius Caesar’, in CP1.

HM

‘Hume on Miracles’, in GG2.

HV ‘On Humanae Vitae’, in GG2. I

Intention, 2nd edition (Oxford: Blackwell, 1963 and Cambridge, MA: Harvard University Press, 2000—​identical in format).

IS

‘The Intentionality of Sensation: A Grammatical Feature’, in CP2.

IS*

‘The Immortality of the Soul’, in GG2.

IWT

An Introduction to Wittgenstein’s Tractatus (London: Hutchinson, 1959).

JPW

‘The Justice of the Present War Examined’, in CP3.

JPW-WSI Second part of JPW. KE

‘Knowledge and Essence’, in GG4.

KRHL

‘Knowledge and Reverence for Human Life’, in GG1.

KRPL

‘Kripke on Rules and Private Language’, in GG4.

LW

‘Ludwig Wittgenstein’, in GG3.

M

‘Morality’, in GG2.

MEC

‘Memory, ‘Experience’ and Causation’, in CP2.

METC

‘The Moral Environment of the Child’, in GG2.

MME

‘Murder and the Morality of Euthanasia’, in GG1.

MMP

‘Modern Moral Philosophy’, in CP3 (and GG1).

MOC

‘Must One Obey One’s Conscience?’, in GG1.

MT

‘Making True’, in GG4.

OMPCY

‘Does Oxford Moral Philosophy Corrupt Youth?’, in GG1.

OT

‘On Transubstantiation’, in CP3.

P

‘Pretending’, in CP2.

PJ

‘On Promising and Its Justice, and Whether It Need Be Respected in Foro Interno’, in CP3.

PI

‘Practical Inference’, in GG1.

PM

‘Prophecy and Miracles’, in GG2.

PMC

‘Parmenides, Mystery and Contradiction’, in CP1.

POD

‘Private Ostensive Definition’, in GG4.

PPDM

‘Prolegomenon to a Pursuit of the Definition of Murder’, in GG1.

PSP

‘Paganism, Superstition and Philosophy’, in GG2.

PT

‘Practical Truth’, in GG1.

QLI

‘The Question of Linguistic Idealism’, in CP1.

RP

‘The Reality of the Past’, in CP2.

Title Abbreviations   xi RRP

‘Rules, Rights and Promises’, in CP3.

SAS

‘On the Source of the Authority of the State’, in CP3.

ST

‘The Simplicity of the Tractatus’, in GG3.

TAA

‘Thought and Action in Aristotle: What Is “Practical Truth”?’, in CP1.

TBC

‘Times, Beginnings and Causes’, in CP2.

TD

‘Mr Truman’s Degree’, in CP3.

TKEA

‘The Two Kinds of Error in Action’, in CP3.

TL

‘A Theory of Language?’, in GG3.

UD

‘ “Under a Description” ’ , in CP2.

UP

‘Understanding Proofs: Meno, 85d9–​86c2, Continued’, in CP1.

WM

‘War and Murder’, in CP3.

WRPL ‘Wittgenstein on Rules and Private Language’, in GG3. WTC

‘Wittgenstein’s ‘Two Cuts’ in the History of Philosophy’, in GG3.

WWC

‘Was Wittgenstein a Conventionalist?’, in GG3.

WWP

‘Wittgenstein: Whose Philosopher?’, in GG3.

YCSC

‘You Can Have Sex without Children: Christianity and the New Offer’, in CP3.

List of Contributors

Valérie Aucouturier is a professor of contemporary philosophy at Université Saint-​ Louis–​Brussels and a member of the Centre Prospéro. Langage, Image, Connaissance. Her research work lies at the crossroads of several fields of philosophy: philosophy of mind and of psychology, philosophy of action, and philosophy of language. She is the author of L’intention en action (Vrin, 2017), Qu’est-​ce que l’intentionalité? (Vrin, 2012), Elizabeth Anscombe: L’esprit en pratique (CNRS Editions, 2012), and the editor of a number of journals and books, including Lectures contemporaines de Elizabeth Anscombe (Klesis 35, 2016) and (with Marc Pavlopoulos) Agir et penser: Essais sur la philosophie d’Elizabeth Anscombe (Publications de la Sorbonne, 2015). Joel Backström is an associate professor (title of docent) at the University of Helsinki. He is the author of The Fear of Openness: An Essay on Friendship and the Roots of Morality (ÅAUP, 2007) and co-​editor of Moral Foundations of Philosophy of Mind (Palgrave, 2019). He has published articles and book chapters on ethics and the philosophy of mind, on Wittgenstein, and on the philosophical dimensions of Freud’s thought, and contributed to two earlier Oxford Handbooks, on Wittgenstein (2011) and on Philosophy and Psychoanalysis (2018). He is the translator into Swedish of Anscombe’s seminal paper ‘Modern Moral Philosophy’ (translation published in 2001). John Berkman is a professor of moral theology at Regis College, University of Toronto, at the Graduate Centre for Theological Studies in the Toronto School of Theology, and a Fellow of the International Society for Science and Religion. He previously taught at the Dominican School of Philosophy & Theology in Berkeley, California, and was director of the Division of Moral Theology/​Ethics at The Catholic University of America. He teaches and writes in the areas of Thomistic ethics, healthcare ethics, and animal ethics. He recently published ‘The Influence of Victor White and the Blackfriars Dominicans on a young Elizabeth Anscombe: An Essay accompanying the Republication of G.E.M. Anscombe’s ‘I am Sadly Theoretical: It is the Effect of Being at Oxford’ (1938)’ in New Blackfriars in September 2021, and is editing The Oxford Handbook of Theological Bioethics. Lucy Campbell is an assistant professor in the Department of Philosophy at the University of Warwick. Prior to this she held a Leverhulme Early Career Fellowship, also at Warwick (2018–​2021), and positions at the Universities of Edinburgh and Oxford. She completed her PhD, ‘Action, Intention, and Knowledge’, in Cambridge in 2015. Her research interests are in the philosophy of mind and action, in epistemology, and in the

xiv   List of Contributors intersection of these areas, with a special focus on our knowledge of our own minds and intentional actions. Charles F. Capps earned a JD and a PhD in Philosophy in 2020 from the University of Chicago, writing his dissertation on accomplice liability. He is currently serving a two-​ year term as a judicial law clerk for the Hon. Raymond W. Gruender on the U.S. Court of Appeals for the Eighth Circuit. The views he expresses herein are his alone and do not represent the views of any judge on the court. Sophie Grace Chappell is a professor of philosophy at the Open University. She has held visiting appointments in the Universities of Edinburgh, Glasgow, British Columbia, Stirling, Reykjavik, and Oslo. She has published over a hundred articles on ethics, moral psychology, epistemology, ancient philosophy, and philosophy of religion. Her books include Aristotle and Augustine on Freedom (Macmillan, 1995), Understanding Human Goods (Edinburgh University Press, 2003), The Inescapable Self: An Introduction to Philosophy (Orion, 2005), Reading Plato’s Theaetetus (Hackett, 2005), Ethics and Experience (Acumen, 2009), and Knowing What to Do: Imagination, Virtue, and Platonism in Ethics (Oxford University Press, 2014). She has also edited or co-​edited five collections of essays in ethics. Her main current research is about epiphanies, immediate and revelatory encounters with value, and their place in our experience and our philosophical ethics; she will be publishing Epiphanies with OUP in 2022. She is a member of the Scottish Mountaineering Club, an active poet and translator of the classical Greek dramatists, and an untalented but keen cyclist and pianist. She is the UK’s first openly transgender philosophy academic, having transitioned in 2014, and campaigns on feminist and transgender issues. She lives with her family in Dundee. Rémi Clot-​Goudard is an associate professor in contemporary philosophy at Université Grenoble Alpes (France). His work deals mainly with topics of philosophy of action, language, and mind, with a special interest in Anscombe’s philosophy. He is the author of L’explication ordinaire des actions humaines (Ithaque, 2015), the editor of L’explication de l’action: Perspectives contemporaines (Recherches sur la philosophie et le langage, Vrin, 2014), and co-​editor of Abduction (Recherches sur la philosophie et le langage, Vrin, 2018). John Hacker-​Wright is an associate professor of philosophy at the University of Guelph. He is the author of Philippa Foot’s Metaethics (Cambridge University Press, 2021), Philippa Foot’s Moral Thought (Bloomsbury, 2013), and many articles on virtue ethics and moral psychology. Adrian Haddock is a senior lecturer in philosophy at the University of Stirling. Between 2017 and 2019, he was a senior research fellow in the Forschungskolleg Analytic German Idealism at the University of Leipzig. His work centres on the idea of subjectivity, and on its significance for understanding some of the fundamental concerns of philosophy. He has written on action, perception, knowledge, and language. He recently edited (with Rachael Wiseman) a collection of essays on the philosophy of G. E. M. Anscombe, The

List of Contributors    xv Anscombean Mind (Routledge, 2021), and is currently in the process of completing a book manuscript, Subject and Object. John Haldane is Emeritus Professor of Moral Philosophy, University of St Andrews; a professor of virtue theory, University of Birmingham; and a professor of philosophy, Australian Catholic University. He has published widely in several areas of philosophy, including ethics, philosophy of mind, and history of philosophy. His publications include Faithful Reason (Routledge, 2004); Reasonable Faith (Routledge, 2010); Practical Philosophy (Imprint Academic, 2009) and he has edited and written on the work of Elizabeth Anscombe, on Aquinas, and on the reception of Aquinas in analytic philosophy. David Albert Jones is director of the Anscombe Bioethics Centre, Oxford; a research fellow at Blackfriars Hall, Oxford University; and a professor of bioethics at St Mary’s University, Twickenham. He is vice-​chair of the Ministry of Defence Research Ethics Committee and an examiner for the Society of Apothecaries Diploma in the Philosophy of Medicine. He is a member of the Commission of the Bishops’ Conferences of the EU Working Group on Ethics in Research and Medicine and is a corresponding member of the Pontifical Academy for Life. His publications include editing (with Chris Gastmans and Calum MacKellar) Euthanasia and Assisted Suicide: Lessons from Belgium (Cambridge University Press, 2017), (with Luke Gormally and Roger Teichmann) The Moral Philosophy of Elizabeth Anscombe (Imprint Academic, 2016), and Approaching the End: A Theological Exploration of Death and Dying (Oxford University Press, 2007). He is the author of Angels: A Very Short Introduction (Oxford University Press, 2011). Gavin Lawrence is a professor in the Philosophy Department, University of California, Los Angeles. His research is in ancient philosophy and in practical philosophy, with special interests in Wittgenstein and Marx, and much else. Micah Lott is an associate professor of philosophy at Boston College, where he teaches courses on ethics and political philosophy. Cyrille Michon is a professor of philosophy at the University of Nantes. He has published work on medieval philosophy, metaphysics, philosophy of religion, and Anscombe’s philosophy, and is the translator into French of Intention (Gallimard, 2002). Anselm Winfried Müller is a regular visiting professor at the University of Chicago (since 2011). After teaching at the University of Oxford and the Australian National University he was a professor of philosophy at Universität Trier from 1974 to 2007 and at Keimyung University from 2007 to 2010. His German publications include books on Wittgenstein’s Tractatus (Bouvier, 1967), Practical Inference (Alber, 1982), Democracy (Kohlhammer, 1996), Euthanasia (Kohlhammer, 1997), Virtue (Kohlhammer, 1998), The Ethics of Manufacturing Human Life (Kohlhammer, 2004), Agency and Education (Ontos, 2008). Other research interests and topics of English publications are the first person; Aristotelian ethics (especially teleology, practical reason, and agency),

xvi   List of Contributors normativity, emotion, causation, natural theology, and the philosophy of Elizabeth Anscombe. Katharina Nieswandt is an associate professor of philosophy at Concordia University in Montreal. Her main areas of research are metaethics and political theory. She has published in a number of peer-​reviewed journals, including The Australasian Journal of Philosophy, The Journal of Value Inquiry, and Ethical Theory and Moral Practice. Together with Ulf Hlobil, she translated Anscombe’s main papers into German for Suhrkamp. Besides her philosophical research, she currently leads a large quantitative study on the under-​representation of women in philosophy. Harold Noonan is the Professor of Mind and Cognition at the University of Nottingham. He was previously a professor of philosophy at the University of Birmingham. He was educated at the University of Cambridge, and taught at the University of Birmingham between 1979 and 2004, before which he was a research fellow at Trinity Hall College, Cambridge. His principal teaching and research interests include the philosophy of mind and the philosophy of language, identity and personal identity, philosophical logic, and the philosophies of Frege, Russell, and Hume. He has written six books. His most recent book is A Guidebook to Naming and Necessity (Routledge, 2012). The third edition of his book Personal Identity (Routledge, 2019) was published in 2019. Duncan Richter is the Charles S. Luck III ’55 Institute Professor at the Virginia Military Institute. He is the author of Anscombe’s Moral Philosophy (Lexington Books, 2011), Ethics after Anscombe (Springer, 1999), and Wittgenstein’s Tractatus: A Student’s Edition (Lexington Books, 2021). Johannes Roessler is an associate professor (reader) in the Department of Philosophy at the University of Warwick. He has published articles in journals such as Mind and Proceedings of the Aristotelian Society on issues in epistemology, the philosophy of mind and action, and the philosophy of psychology, and he has co-​edited three volumes published by Oxford University Press. He is currently working on a monograph on perceptual knowledge. Constantine Sandis is a professor of philosophy at the University of Hertfordshire, founding director of Lex Academic, and Fellow of the Royal Society of Arts. He is the author of The Things We Do and Why We Do Them (Palgrave, 2012), Character and Causation (Routledge, 2019), and Raisons et Responsabilité (Ithaque, 2021), as well as the editor of several books, including Philosophy of Action from Suarez to Anscombe (Routledge, 2019). His selected essays are forthcoming in two volumes: From Action to Ethics (Bloomsbury, 2022) and Wittgenstein on Other Minds (Anthem, 2022). Harold Teichman was born and grew up in the middle of the previous century in a déclassé stretch of New Jersey’s contribution to the endless conurbation that surrounds New York City. He was educated at Upsala College in East Orange, the University of Bern, and Stony Brook University, at which places he concentrated on physics and mathematics. After abandoning PhD study in theoretical physics he became interested

List of Contributors    xvii in general philosophy, culminating after a few years in a lifelong focus on the work of Wittgenstein, Anscombe, and Kripke. His professional life was spent in scientific publishing and what is now called FinTech. Since 2008 he has from time to time assisted the Saul Kripke Center at CUNY with the editing of Kripke’s manuscripts. His current philosophical research centres on human action, ethics, and the varieties of modality. Roger Teichmann is lecturer in philosophy at St Hilda’s College, Oxford. He is the author of a number of monographs, including The Philosophy of Elizabeth Anscombe (Oxford University Press, 2008) and Nature, Reason and the Good Life (Oxford University Press, 2011). A collection of his essays is about to be published as Logos and Life: Essays on Mind, Action, Language and Ethics (Anthem Press, 2002). Rachael Wiseman is a senior lecturer in philosophy at the University of Liverpool. She works on the philosophy of mind, action and ethics, and the history of analytic philosophy. She is the author of the Routledge Philosophy Guidebook to Anscombe’s Intention (2015) and, with Clare Mac Cumhaill, Metaphysical Animals (Chatto & Windus, 2022), a joint intellectual biography of the Wartime Quartet of Anscombe, Foot, Midgley, and Murdoch.

Introdu c t i on Roger Teichmann

One of the most important philosophers of the twentieth century, Elizabeth Anscombe embodied a number of different things—​or if you prefer, had a variety of personae. Just as an action can be intentional under several descriptions (see Anscombe’s Intention), so an individual can be significant under several descriptions. The descriptions or personae I have in mind include: student/​friend/​translator of Wittgenstein, Roman Catholic, woman philosopher, protestor and activist, and (an) instigator of ‘virtue ethics’. People can be drawn to Anscombe for various reasons, corresponding to these different personae. An ‘Anscombean’ might be a Catholic bioethicist, a scholar of Wittgenstein, an action theorist, a philosophical feminist—​or all of these at once. Such variety in unity is also found in Anscombe’s philosophical output, which ranges wide. She wrote and lectured on ethics, politics, mind, action, language, metaphysics, ancient and mediaeval philosophy, Wittgenstein, and more. It is hard to think of a more versatile modern philosopher. And there is a connection, I think, between this versatility and the variety of personae I have referred to. Neither in the case of the personae nor in that of the philosophical topics do we have a mere medley; connections bind the items on the list. Anscombe was hardly a Jack (or Jill) of all trades. It took a powerful mind to be able to encompass those personae and those philosophical interests, and one quality of mind which was evident to all who knew her was certainly power. (Her eldest son used to recall his mother’s mind as having been like a steel rat trap; he was obviously thinking not just of the power of steel but of the capacity to stop error in its tracks.) The twenty-​two chapters of this Oxford Handbook exhibit some of the variety of Anscombe’s philosophy, the topics covered ranging from intentional action through ethics and the first person to Anscombe’s relationship with other thinkers. The affiliations and nationalities of the authors are likewise various: we have philosophers from the UK, the USA, Canada, France, Germany, and Finland. Since her death in 2001 Anscombe’s reputation has grown and grown, and her thought and work are studied in many countries, of which the list I’ve just given is only a subset.

2   Roger Teichmann In her lifetime Anscombe published both books and articles. Three volumes of articles were brought out in 1981.1 Since 2001 Anscombe’s daughter and literary executor, Mary Geach, along with Mary’s husband, Luke Gormally, have edited four further volumes of papers, many of these not previously published.2 There is a considerable Nachlass in addition to the papers in these seven volumes; the full archive is now housed at the University of Pennsylvania, and the process of digitising selections from the contents is underway. A few of the chapters in this Handbook make use of posthumously published, not to mention unpublished, material, and there will be some interesting surprises here for Anscombe scholars. Notable is the in-​depth archival research that has been carried out by John Berkman, whose chapter makes use of a number of hitherto unknown or unnoticed sources that shed light on Anscombe’s intellectual development. The classic texts remain classic, of course. Intention, ‘Modern Moral Philosophy’, and ‘The First Person’ still invite and receive attention, and several of the chapters of this Handbook deal with these. But there are discussions of many other articles; a glance at the Title Abbreviations will give the reader an idea of the breadth of coverage. Gertrude Elizabeth Margaret Anscombe was born in 1919, the youngest child of Alan Wells Anscombe, a schoolmaster at Dulwich College, and Gertrude Elizabeth (née Thomas), a headmistress. She was educated at Sydenham School. Reading done between the ages of twelve and fifteen led eventually to her being converted to Roman Catholicism, and it was as a result of reading a book by the nineteenth-​century Jesuit Bernard Boedder, Natural Theology, that she became embroiled in philosophy.3 Her drift to Rome sufficiently alarmed her parents that they asked an Anglican priest to come and talk to their teenage daughter in the hope that he could dissuade her. His efforts were in vain; she put him on the spot regarding the nature of the Eucharist, and his answer did not satisfy her. In 1937 Anscombe went up to St Hugh’s College, Oxford, to read Literae Humaniores (Classics), and at Oxford she was formally received into the Catholic Church. Her parents had threatened to cut off her funding if she made this move but she called their bluff and they continued to fund her. It’s clear that as a girl and as a young woman Anscombe was already a very determined character. While an undergraduate Anscombe met Peter Geach, who was reading Lit. Hum. at Balliol College and who was another Catholic convert. They married in 1941. Philippa Foot, herself an atheist, was to remark that the meeting of two people of such extraordinary harmony of personalities and interests might almost appear to be an argument for divine providence. After graduating, Anscombe was awarded a research fellowship at Newnham College, Cambridge, where she attended the lectures of Ludwig Wittgenstein, subsequently becoming a personal friend of his. She was to be his main English translator and one of three literary executors appointed by Wittgenstein himself, along with Rush Rhees and G. H. von Wright.

1 

CP1, CP2, CP3. GG1, GG2, GG3, GG4. 3  Cf. CP3, Introduction, vii. 2 

introduction   3 In 1946 she returned to Oxford to take up a research fellowship at Somerville College. She was later made a lecturer and much later (in 1964) a tutorial fellow. At Somerville she got to know first Philippa Foot and then Iris Murdoch and Mary Scrutton (later Midgley), and it was these four philosophers (especially Anscombe and Foot) who were to bring about something of a revolution in English-​speaking moral philosophy, questioning the ‘fact/​value distinction’ and putting the virtues and vices firmly on the map—​amongst other things. The persona of Anscombe’s that I earlier dubbed ‘protestor and activist’ first made an appearance just after the start of the Second World War when she was still an undergraduate. Anscombe and a friend, Norman Daniel, published a pamphlet, The Justice of the Present War Examined (1939), in which they argued that the war in which Great Britain was engaged was not a just one, in particular because of an evident willingness on the part of the British government to target civilians. Similar considerations lay behind Anscombe’s most famous protest, issued in 1956 against the proposal by Oxford University to honour Harry Truman, the former US president, with a degree. Anscombe publicly opposed the proposal, again on the grounds of a government’s targeting and killing innocent civilians: Truman had ordered the dropping of atom bombs on Hiroshima and Nagasaki. Anscombe’s action was widely reported in the national and international press. Anscombe was a person of extraordinary industry and stamina. Through the 1950s and 1960s she was writing philosophy, translating Wittgenstein into English, supervising students, lecturing, travelling, and all this while bearing and rearing several children; she and Peter Geach had seven children in all, four girls and three boys. In 1970 Anscombe took up a chair at Cambridge University, holding the same position that Wittgenstein had held. She was a Fellow of New Hall (now Murray Edwards College). She retired in 1986 but carried on thinking and writing for as long as she was able. I spoke earlier of the connections binding together the various philosophical topics on which Anscombe wrote. These thematic connections exist within and between the chapters of this Handbook—​the division of the book into five discrete parts is to some extent a matter of convenience. Ring-​fencing parts of the conceptual landscape is often misguided. That was true of the claim that the ‘factual’ and the ‘evaluative’ were separated by a logical chasm, and something similar seems to be true of the claim that ‘meta-​ethics’ and ‘normative ethics’ can go on independently of one another. But the subterranean strands are more pervasive than these examples indicate. Mind, epistemology, language, and ethics overlap and intertwine, and this is something that comes out quite strongly in the chapters that follow. Consider the concept of practical knowledge (knowledge of what one is doing). Anscombe famously investigates this concept, notably in Intention. A number of the chapters continue the investigation, focusing variously on acting well (Campbell, Clot-​Goudard, Hacker-​Wright), on observational and non-​observational knowledge (Roessler, Teichman), and on the first person and self-​consciousness (Aucouturier, Haddock). Or consider how Anscombe’s critique of modern moral philosophy blends linguistic, psychological, and more clearly ethical considerations. That doing ethics

4   Roger Teichmann requires having an adequate understanding of language and its role in human life is an explicit or implicit theme of more than one chapter (Lawrence, Wiseman, Backström). Finally, one could mention Anscombe’s later discussions of ‘mystical value’ and of the Thomist concept of ‘connatural knowledge’. These discussions point to issues concerning religious and non-​religious outlooks (Lott, Richter); they also pose apparent problems for ethical naturalism of the sort Anscombe herself earlier seems to have espoused (Müller); and they inform Anscombe’s own conception of the nature of murder, a topic addressed by four of the contributors (Jones, Lott, Nieswandt, Berkman). Meeting other academic philosophers at social occasions often involves being asked the question ‘What do you work on?’ The presupposition is liable to be that one must specialise in something. That may be true within natural science, where people contribute their findings and hypotheses and theories to a general pool, and in that sense collaborate in one big Research Programme. But taking things on authority is essential for science to work in this way (the entomologist trusts the endocrinologist and vice versa), and such taking on authority is not possible for philosophers. Nor is the idea of a research programme at home in philosophy, however useful it is when talking to funding bodies. Of course we are all of us busy, so at any given time can hardly be expected to be delving into action theory, philosophy of maths, metaphysics of time, scepticism, and the Neoplatonists. But never to move off your favourite patch risks not doing justice even to that patch. This was something Anscombe was clearly aware of.4 And it helps to explain the simultaneous breadth and depth of her thought. The chapters in this Handbook give a flavour of that breadth and depth, as well as embodying original philosophical thought: Anscombe’s ideas are expounded, elaborated, extended, questioned, and on occasion opposed. I am very grateful to all of the contributors to this volume, for their hard work and for their patience, but above all for their chapters, which together constitute a fitting tribute to one of the greatest of twentieth-​century philosophers. It has been a pleasure reading all this material and as editor entering into philosophical dialogue with the authors. I am also extremely grateful to my OUP editors, Lucy Randall, Hannah Doyle, and Brent Matheny. They have been unremittingly positive, helpful, and patient.

4 

As was her mentor Wittgenstein, who when asked for a thematic title by which to advertise his philosophy lectures in Cambridge offered ‘Philosophy’.

Pa rt I

I N T E N T ION

Chapter 1

On Ansc ombe on Pr actical Know l e d g e and Practic a l T ru t h Lucy Campbell

1.1 Introduction This chapter considers the relationship between two notions in Elizabeth Anscombe’s philosophy: practical knowledge, and practical truth.1 To my knowledge, Anscombe does not herself explicitly relate these notions, and the question of how they are related is not straightforward. Knowledge is rather generally knowledge of truth. So, given a category of practical knowledge and a category of practical truth, one would be forgiven for thinking that the latter is the object of the former—​that practical knowledge is simply, or essentially, or always, knowledge of practical truth. Call this the Simple View of the relationship between practical knowledge and practical truth. But the Simple View can’t be right, because as Anscombe introduces the two notions, they differ in scope. A person has practical knowledge of whatever she is intentionally doing, but not all of our intentional actions produce practical truth; practical truth is manifested only by those intentional actions which aim at and succeed in being cases of acting or living well. So the cases in which a person has practical

1 

For discussion of the Aristotelian background to Anscombe’s discussion in ‘Thought and Action in Aristotle’, thanks to Ursula Coope. Thanks also to Karen Margrethe Nielsen and Simon Shogry, who let me sit in on their lectures on the Nicomachean Ethics in Oxford in Hilary Term 2019. For comments on earlier versions of the chapter itself—​not all of which I have been able to deal with in the final version for reasons of space—​thanks to Chris Campbell, Casey Doyle, Alexander Greenberg, Nathan Hauthaler, Ben Sorgiovanni, Roger Teichmann, and audiences in Southampton, Oxford, Warwick, Toronto, Stanford, and Austin. This chapter was completed during a Leverhulme Early Career Fellowship (2018–​21; grant ref. ECF-​2018-​075), and I gratefully acknowledge the Leverhulme Trust for this support.

8   Lucy Campbell knowledge seem to outstrip the cases in which there is any practical truth on the scene to be known, and the Simple View must be false. My discussion of how to understand the relationship between practical knowledge and practical truth takes the following route. In sections 1.2 and 1.3 I explain how Anscombe introduces each notion in turn, emphasising that they both invite the label ‘practical’ due to having a special practical form or structure. I then turn to considering how the two phenomena are related. In section 1.4 I reject a view on which practical knowledge and practical truth are only ever accidentally related. The reasons for doing so favour accepting at least that when someone’s action manifests practical truth, this is the object of her practical knowledge, and essentially so. I accept this in section 1.5, but I also argue there that this cannot be the whole story about the relationship. In section 1.6, I offer what I think is a more philosophically satisfying view, which requires distinguishing within both the category of practical knowledge and the category of practical truth, a thinner purely action-​theoretical phenomenon and a thicker, specifically ethical, phenomenon. Doing so makes available what I call the Two-​Level Simple View, on which practical knowledge in the thin action-​theoretical sense is in all cases and essentially knowledge of practical truth in the thin action-​theoretical sense, and practical knowledge in the thick ethical sense is in all cases and essentially knowledge of practical truth in the thick ethical sense. In making these distinctions I go beyond Anscombe’s explicit discussion, and in the remainder of the chapter I articulate and respond to the objection that I do so in a way which is inconsistent with what Anscombe actually says. For there are certain passages in which Anscombe seems to reject what I have identified as a purely action-​theoretical notion of practical truth, something the Two-​Level Simple View requires that we accept. I respond to this worry in sections 1.7 to 1.9, arguing that in fact Anscombe commits herself to this notion, at least implicitly. I conclude in section 1.10.

1.2  Practical Knowledge: A Special Form of Knowledge Anscombe develops her conception of practical knowledge in Intention. As she points out, failing to know that one is doing something under a certain description is—​rather generally—​sufficient for one’s not acting intentionally under that description (I, 11).2 If, having driven to Hastings, I claim not to have known that I was doing so, that is enough (assuming I’m speaking truly) to show I have not driven to Hastings intentionally.

2  Davidson (2001, 50) is often said to have refuted this claim: ‘A man may be making ten carbon copies as he writes, and this may be intentional; yet he may not know that he is; all he knows is that he is trying.’ But Davidson’s conclusion can be resisted. For two different ways of doing so, see Thompson (2011) and Wolfson (2012).

On Anscombe on Practical Knowledge and Practical Truth    9 By contraposition, doing something intentionally under a certain description requires knowing that one is doing it. Practical knowledge, in the sense in which Anscombe is interested in it in Intention, is precisely this knowledge, which both concerns and is essential to action which is intentional.3 Practical knowledge, whose object is one’s own intentional action, contrasts with ‘theoretical’ (‘contemplative’ [I, 57]; ‘speculative’ [I, 87]) knowledge, which can take in any old fact.4 My theoretical knowledge has such diverse objects as the fact that dogs bark, that today is a Tuesday, that there is more than one person in the world, that Pythagoras’ theorem is true, that the pangolin is endangered, and so on. Amongst the objects of my theoretical knowledge there are also certain facts about particular actions. I have theoretical knowledge, for example, of my own non-​intentional actions (as when I notice that I have just accidentally stepped on the cat’s tail) and of the intentional actions of others (as when I know that Alexander is eating Chris’s chips). Because theoretical knowledge sometimes concerns action, practical knowledge cannot be ‘practical’ simply in virtue of its content, simply because it is knowledge of action. Anscombe is clear that practical and theoretical knowledge differ formally. More specifically, practical knowledge is ‘undermined by a mistake in performance’ (I, 57), while theoretical knowledge is ‘undermined by a mistaken judgment’ (I, 57); and practical knowledge is ‘the cause of what it understands’ (I, 87), while theoretical knowledge is ‘derived from the objects known’ (I, 87). What does Anscombe mean by these claims? ‘Undermined by a mistake in performance’.  Practical knowledge is such as to be ‘undermined by a mistake in performance’ in the sense that if what one thinks one is doing comes apart from what one is doing, so that one’s knowledge is undermined, the explanation of what has gone wrong might be simply that one hasn’t done what one was supposed to do, not—​as in theoretical knowledge—​that one has failed to ascertain what is going on.5 3  The capacity for practical knowledge of what one is doing is at the same time the capacity to know why one is doing it and what one will do in the future, when one has decided to do it. Plausibly, it is this same capacity by which we know how to do things (I, 88), where knowledge how to φ is manifested in φ-​ing intentionally. I leave these aspects of the capacity for practical knowledge to one side for reasons of space (but see n. 25). 4  This might seem to say that any item of knowledge that p will be either practical or theoretical. In my view there are forms of knowledge (e.g., knowledge of our own mental states) which fall into neither category. Whether Anscombe would agree is unclear to me—​she is usually taken to deny that ordinarily speaking we know about our states of mind [I, 13–​14], but I think the situation is more complicated than this allows for. For an argument that she views our knowledge of the position of our limbs as neither practical nor theoretical, see McDowell [2011, esp. §VII]. 5  Anscombe needn’t deny that I also judge wrongly here (I, 88–​89). The important idea is that what a failure of practical knowledge is, is a mistake in performance. It is not, as is a failure of theoretical knowledge, a failure in judgement. Anscombe can (and does) accept that failures in theoretical knowledge can give rise to failures in practical knowledge. Our example is a case in point: Alexander’s failure of practical knowledge is a failure in performance: he’s not eating the chips he means to be eating and so fails to know what he is doing. But we can imagine that the reason why he fails in his performance is that he wrongly thinks that these chips—​which are in fact mine—​are Chris’s. This is a failure in theoretical knowledge, and it is because of this theoretical failure that his action goes awry and he lacks practical knowledge. In central cases of successful intentional action, practical and theoretical

10   Lucy Campbell So, if Alexander means to be eating Chris’s chips, but he’s actually eating my chips, then he’s not acting with knowledge, because he’s making a mistake. And the mistake he’s making is in what he’s doing: he’s eating the wrong chips. By contrast, if I fail to know what Alexander is doing—​if I think (e.g.) that he’s eating my chips but in fact he’s eating Chris’s chips—​then my mistake is with my judgement: I believe the wrong thing about whose chips Alexander is eating. ‘The cause of what it understands’.  Anscombe says that it is the agent’s knowledge of what he is doing that gives the descriptions under which what is going on is the execution of an intention. (I, 87)

Since executing an intention just is acting intentionally, this means that it is a person’s knowledge of what they are doing which makes their action intentional under that description—​w hich makes it the case that they are doing that thing intentionally. Alexander is eating Chris’s chips intentionally only if he knows that he is doing so, where this knowledge is practical knowledge—​knowledge which (inter alia, and as just noted) would be undermined by a mistaken performance rather than by a mistaken judgement. The ‘what it understands’ of Alexander’s practical knowledge is that he is eating Chris’s chips, where ‘eating Chris’s chips’ is a description with a certain form, viz., that peculiar to specifically intentional action.6 Alexander’s practical knowledge that he is eating Chris’s chips is thus the ‘cause of what it understands’ (in a formal causal sense): if it weren’t for his having this knowledge, ‘eating Chris’s chips’ would not be a description of something he is doing intentionally. Contrast my theoretical knowledge that Alexander is eating Chris’s chips. Although on the face of it this has a very similar content to Alexander’s practical knowledge—​that he (Alexander) is eating Chris’s chips—​it bears a very different relationship to its object. My theoretical knowledge is ‘derived from the object known’ and is completely accidental to that object. Alexander would still be eating Chris’s chips, and doing so intentionally, whether or not I knew he was doing so. More generally, practical knowledge is essential and internal to its object; theoretical knowledge is accidental and external to its object.

knowledge will both play a role. But this does not undermine the claim that there really are two forms of knowledge here, or diminish the theoretical usefulness of distinguishing them. (See also I, 88–​89.) 6 

We could put this point by saying that Alexander knows that he is intentionally eating Chris’s chips, but this would be misleading if it were taken to suggest that Alexander knows a disjunctive proposition: that he is eating Chris’s chips and that his so doing is intentional. Putting things in this way risks suggesting that he might instead have had practical knowledge that (in this case) he was eating Chris’s chips, while lacking practical (or any) knowledge that he was doing so intentionally. This is not possible since practical knowledge cannot but be of action qua intentional. The convoluted way of describing the content of Alexander’s knowledge in the main text is designed to avoid confusion on this point.

On Anscombe on Practical Knowledge and Practical Truth    11 Knowledge ‘in intention’.  As well as saying that practical knowledge is undermined by a mistake in performance, whereas theoretical knowledge is undermined by a mistake in judgement, and that practical knowledge is ‘the cause of what it understands’, whereas theoretical knowledge is ‘derived from the objects known’, Anscombe also characterises practical knowledge’s special practical form by saying that practical knowledge is knowledge ‘in intention’ (I, 57). I think that what she means by this is that it’s simply by or in having and executing intentions that we know what we are intentionally doing. Alexander has practical knowledge that he’s eating Chris’s chips simply because he is intentionally eating Chris’s chips.7 The special ‘practical’ form of practical knowledge can now be understood in terms of its being embodied in the execution of intention. This makes practical knowledge an epistemic aspect of intentional action itself rather than (like theoretical knowledge) a separable stance which one may or may not take on what one is intentionally doing.8 Understanding Anscombe’s characterisation of practical knowledge as knowledge ‘in intention’ along these lines explains why it should be both ‘the cause of what it understands’ and ‘undermined by a mistake in performance’. If practical knowledge is embodied in the execution of intention (is knowledge ‘in intention’), then of course it would be internal and essential to one’s intentional action (would be ‘the cause of what it understands’), because executing an intention to do something just is doing that thing intentionally. And if practical knowledge is embodied in the execution of intention (is knowledge ‘in intention’), then of course it would be ‘undermined by a mistake in performance’, because a mistake in performance is precisely a failure to execute one’s intention.9 7 

Earlier I said that it is because he has practical knowledge that he is eating Chris’s chips that Alexander counts as doing so intentionally. Here I am saying that it is because he is executing his intention to eat Chris’s chips—​and so because he is intentionally eating Chris’s chips—​that Alexander has practical knowledge that he is doing so. Which is it to be? The answer is: both. All that is required for practical knowledge that one is φ-​ing, on the view I am ascribing to Anscombe, is that one is executing one’s intention to φ. This covers the second point. But what is ‘executing’ an intention, as opposed, for example, to merely doing what one intends but not doing so intentionally (as in Davidson’s [2004, p. 106] case of intending to meet his daughter at a certain restaurant on her birthday, going there to make a reservation, and—​not realising that today is her birthday—​meeting her there on her birthday, but not intentionally)? Anscombe’s answer is: practical knowledge. ‘Without [practical knowledge] what happens [e.g., ‘meeting my daughter on her birthday’] does not come under the description—​execution of intentions—​whose characteristics we have been investigating’ (I, 88). 8  Anscombe also characterises practical knowledge as ungrounded in observation or evidence of one’s acting in the way in question, and her distinction between practical and theoretical knowledge is often introduced primarily in these terms. I avoid this way of characterising the distinction because on my understanding, practical knowledge’s non-​observational and non-​inferential character is itself to be explained by its being ‘the cause of what it understands’, ‘undermined by a mistake in performance’, and ultimately, by its being knowledge ‘in intention’. That this is so is shown by the fact that Anscombe countenances varieties of non-​observational and non-​inferential knowledge which are not practical, such as one’s knowledge of the position of one’s limbs (I, 13). 9  There is vastly more to say about why, so understood, practical knowledge should be a genuine form of knowledge: why should simply executing an intention to φ meet the general requirements on knowing that one is φ-​ing? I don’t think Anscombe herself offers an answer to this question, but I have tried to provide one of my own elsewhere (Campbell 2016, ch. 6; 2018a).

12   Lucy Campbell In summary, practical knowledge is knowledge that one is φ-​ing when one is φ-​ing intentionally. It is ‘practical’ not only in the sense that it concerns action, but also formally speaking. A person has practical knowledge that she is φ-​ing simply in or by executing her intention to φ, and in this sense her knowledge is knowledge ‘in intention’.10 A person’s practical knowledge that she is φ-​ing is not a separately describable epistemic stance she may or may not take on what she is intentionally doing, but an epistemic aspect of her intentional action itself.

1.3  Practical Truth: A Special Form of Truth Anscombe’s main discussions of practical truth are found in ‘Thought and Action in Aristotle’ (first published 1965) and in ‘Practical Truth’ (first published in 1999). The notion of practical truth derives from Aristotle, who mentions it just once, in a discussion of rational choice (prohairesis), in Nicomachean Ethics VI.ii.11 Anscombe 10  The reader might worry here that I have left out perhaps the most central aspect of Anscombe’s discussion in Intention: her focus on the agent’s ability to answer questions about her own actions, in particular questions about what she is doing and why (where answering a question about why one is doing something will in many cases be saying what further one is doing in doing it). Being able to answer these questions, one might think, is something over and above the fact that she is executing her intention. Given this, how can I be right to say that for Anscombe, someone has practical knowledge that she is φ-​ing simply in or by executing her intention? But although I haven’t focused on it here, this aspect of Anscombe’s picture is part of my interpretation. This becomes clear when we bear in mind that humans can express their intentions, and that we do so, canonically, precisely by saying what we are doing, what we are going to do, and why (I, §2 et passim). Answering questions about what one is doing and why, when these questions have the special sense Anscombe details in Intention, just is expressing one’s intentions. (This shows up in the fact that an answer to the special ‘What?’ or ‘Why?’ is automatically open to further questions of the same sort—​and not to requests for evidence to support one’s answer.) So being able to answer Anscombe’s special questions asked in relation to one’s action isn’t anything over and above having intentions (and so a fortiori over and above executing intentions). Thanks to Roger Teichmann for pressing the worry I respond to here (see also Campbell 2018b). 11  The key passage reads as follows: ‘What affirmation and negation are in judgement, pursuit and avoidance are in desire. So, since moral virtue is a disposition of one’s choice, while choice is deliberated wanting, these things show that the judgement must be true and the wanting right, if the choice is to be sound, and the one must say and the other pursue the same thing (76). This, then, is thought, and truth, of a practical sort [practical truth (77)]; in the case of thought that is theoretical, and not practical nor productive, ‘well’ and ‘badly’ consist in the true or the false (this is, after all, the function of any faculty of thought), but that of a faculty of practical thought is truth in agreement with right desire (77)’ (Nicomachean Ethics 1139a 21–​31). Anscombe translates only parts of this passage; her bits are in bold, with page numbers for those passages (from TAA) in parentheses. The other bits are from the Broadie and Rowe translation (Aristotle 2002), which seemed to me to fit best with the translations Anscombe does provide. Anscombe’s main presentation of the idea of practical truth, found in TAA, is developed through a critical interpretation of Aristotle in relation to this and connected notions, with a focus on the quoted passage. In the main text I have chosen to abstract away from Anscombe’s interpretative concerns in

On Anscombe on Practical Knowledge and Practical Truth    13 clearly thinks that the notion is a useful one, and she adopts it into her philosophy of action and ethics. Like practical knowledge, practical truth is practical in form, and not only in content. It is truth that is ‘brought about—​i.e. made true—​by action’ (TAA, 77); truth in which ‘the description of what [one] does is made true by [one’s] doing it’ (ibid.), and where this includes ‘the truth of the description ‘doing well’’ (PT, 149). Practical truth is truth which is embodied in actions which aim at and succeed in being cases of acting, or doing well, and thereby are contributions to more generally living well. This is all rather abstract, so let’s clarify the idea by considering: Virtuous Alexander Alexander owes me £100. Justice requires him to pay me back, and he recognises this. He thinks he can pay me back by writing me a cheque. So he writes me a cheque, thus successfully paying me back—​and thus doing what justice requires.

The first thing to say is that in acting as he does, Alexander makes true a number of action-​descriptions. He makes true the descriptions ‘writing a cheque’, ‘paying Campbell back’, and ‘doing what is just’. This is not yet enough for Alexander to manifest practical truth. In addition, he must, in acting as he does, be both aiming at and succeeding in acting well and contributing to living well.12 Let’s imagine that these conditions are met, by stipulating (a) that Alexander wants to do what is just in virtue of viewing doing what is just as conducive to living well and (b) that doing what is just is conducive to living well.13

order not to overly complicate my presentation of her account. Nor will I comment on the accuracy of her interpretation of Aristotle. (I am not qualified to do so, but see Graeser 2005; Pakaluk 2010; Olfert 2014, 2017; Broadie 2016; Charles 2018.) Having said all that, I here attempt a (doubtless hubristically brief) gloss on the Aristotle passage, largely based on Anscombe’s interpretation in TAA. ‘Choice’ (prohairesis) is a (quasi-​)technical term for the upshot of a special form of practical reasoning, practical reasoning which is aimed at securing the object of boulesis, a specifically rational species of desire. Boulesis is desire for φ-​ing which represents φ-​ing as ‘conducive to or part of ‘doing well’’ (TAA, 75). Boulesis can thus be correct or incorrect: a boulesis for φ-​ing will be correct insofar as φ-​ing is in fact (conducive to or part of) doing well, and incorrect if not. (Boulesis thus contrasts with other forms of desire [orexis]; epithumia and thumos, which are arational and cannot be correct or incorrect.) A ‘sound’ choice/​prohairesis is a choice to do something, ψ, in order to φ, where ψ-​ing is in fact a way of φ-​ing, and φ-​ing is the object of right or correct boulesis. Executing a sound ‘choice’/​prohairesis thus secures ‘truth in agreement with right desire’ (desire in the sense of boulesis). This is practical truth. A ‘choice’ will be unsound if either (a) it is in service of an object of incorrect boulesis or (b) it is a choice to ψ in service of φ-​ing, where ψ-​ing is not in fact a way of φ-​ing—​even if the boulesis for φ-​ing is correct (i.e., even if φ-​ing is part of or conducive to living well). Executing an unsound choice/​prohairesis delivers practical falsity. 12  He will be in a position to do so only if he has some kind of ethical character (TAA, 69–​72). I leave this aspect to one side. 13  In the language introduced in n. 11, we are imagining, that is, that his desire for acting justly is a correct boulesis.

14   Lucy Campbell Because Alexander acts justly with a view to acting well, and because acting justly would in fact be acting well, this means that in acting as he does, Alexander not only makes true the descriptions ‘writing a cheque’, ‘paying Campbell back’, and ‘doing what justice requires’, but through making these descriptions true, he also makes true the description ‘acting well’.14 Practical truth as so far described has two important features. First, it is an essentially ethical phenomenon: it is manifested only in actions which aim at and succeed in being cases of acting or living well. And second, it is an essentially complex phenomenon: it is manifested in actions which make true a complex of action-​descriptions, related as means to ends, where the ultimate end—​and so the ultimate description made true—​is ‘acting well’, ‘living well’, or etc.15 In what sense is practical truth practical in form? Anscombe is clear that practical truth is to be distinguished from ordinary truth—​what we can call ‘theoretical truth’—​ in relation to its bearer. We can think of theoretical truth as attaching in the first instance to propositions or statements, whereas practical truth attaches in the first instance to actions: The notion of truth or falsehood in action would quite generally be countered by the objection that ‘true’ and ‘false’ are senseless predicates as applied to what is done. If I am right there is philosophy to the contrary in Aristotle. (Anscombe 1965, 157–​58; italics in original. See also TAA, 77)16

But an initial reaction to Anscombe’s claim to have described a formally distinctive kind of truth which attaches to action rather than to propositions or bits of language is likely to be puzzlement. Because, as we have just seen, practical truth is the truth of a complex of descriptions of what one is doing, where these are made true by acting. Isn’t the view therefore one on which practical truth does attach in the first instance to bits of language, and so not to actions? And isn’t it thus one on which practical truth is ‘practical’ only because it has a distinctive kind of practical content, and not in virtue of having a special practical form? Understanding Anscombe’s response to this concern will help us understand the structure both of practical truth and of action itself. The passage just quoted continues as follows:

14 

If these conditions are met, then Alexander not only does what is in fact just, but further acts justly. For discussion see (TAA, 69–​72). 15  I gloss over deep and important complexities here by using this disjunctive formulation, for living well is much more demanding than acting well. Living well would require acting well at least typically, over a variety of circumstances, and over a sustained period of time. Acting well might be possible more sporadically. Acting well is required for living well, but not vice versa. Thanks to Roger Teichmann for helping me see the complexities here. Sadly there is not space to iron them out in the present discussion. 16  This quotation and the next are both taken from the version of ‘Thought and Action in Aristotle’ published in R. Bambrough, ed., New Essays on Plato and Aristotle (1965). The wording departs slightly from that in TAA, for which I provide the equivalent page numbers. The differences do not affect the substance of my interpretation.

On Anscombe on Practical Knowledge and Practical Truth    15 And if, as I should maintain, the idea of the description under which what is done is done is integral to the notion of action, then these predicates [‘true’ and ‘false’] apply to actions strictly and properly, and not merely by an extension and in a way that ought to be explained away. (Anscombe 1965, 158; italics in original. See also TAA, 77)

The idea here is that, yes, practical truth is the truth of action-​descriptions, but, action-​ descriptions are integral to actions per se, so that practical truth applies ‘strictly and properly’ to action, because it applies to action-​ descriptions, and because these descriptions are themselves integral to action. In what sense is ‘the idea of the description under which what is done is done’ integral to the notion of action? A first thought is that it is only in relation to a given action-​description that any of the key action-​theoretical concepts have application. We have already seen this to be the case in relation to actions understood as intentional or as objects of practical knowledge, but it is also true in relation to our responsibility for our actions (PT, 142), to whether an action is done for a reason (I, 11), to whether it is consented to (TKEA, 3), and so on. Nathan might be sawing a plank and thereby sawing Barry’s plank, but know only that he is doing the former. Lois Lane might consent to being kissed by Superman, but not to being kissed by that dweeb Clark Kent, and so on. More fundamentally, the possibilities for so much as identifying actions presupposes the use of some description or other of what is being done, at least implicitly. I cannot indicate which action I am talking about by simply pointing at the person acting and saying ‘That action’. In so doing I might be pointing at Nathan’s sawing a plank, at his sawing one of Barry’s planks, at his making a squeaky noise with the saw, etc., not all of which need even be intentional. The word ‘that’ along with my pointing gesture simply can’t distinguish between these various things Nathan is doing. I need to make use of a description to do that. That is not to say that I need to utter a description; which one I have in mind will often be clear from the context. When a parent shouts ‘Stop doing that!’, it is not a matter of indifference, of the various things the child is up to, which it is to stop doing. If the child keeps punching while redirecting this activity away from his baby sister and towards the sofa cushions, then he may well have obeyed his parent’s order just as well as if he ceased punching altogether. The point about the identity of actions does not only apply to actions which are intentional. After falling down the stairs I say dryly ‘I didn’t want to do that’, and it’s clear to everyone present that I’m not referring to the clattering noise I made on my way down. In a given case the description one uses to identify an action might not need to be uttered explicitly, but this does not mean that we can make sense of ‘an action’ independently of any description. Anscombe makes this point too: I have on occasion stared dumbly when asked: ‘If one action can have many descriptions, what is the action, which has all these descriptions?’ The question seemed to be supposed to mean something, but I could not get hold of it. . . . The

16   Lucy Campbell proper answer to ‘What is the action, which has all these descriptions?’ is to give one of the descriptions. Any one, it does not matter which; or perhaps it would be best to offer a choice, saying ‘Take whichever you prefer’. (UD, 209)

So the description under which what is done is done is integral to action in being internal to the very identity of an action. And this means that although practical truth is the truth of a (complex of) description(s) of an action, which one makes true in acting, it is also (and thereby) truth which applies ‘strictly and properly to action’. Clarifications.  I will make three clarifications before turning to consider how practical truth and practical knowledge relate to one another. First, we do not need to assume that someone manifesting practical truth will consciously consider what living well requires, or ‘run through’ the relevant stretch of reasoning in her head prior to or while acting.17 It requires only that were her reasoning to be set out in full (‘as of course it hardly would be in real life, since one does not need to advert to the obvious’ [TAA, 72]), this would make clear that the agent viewed her action as contributing to living well, and that she did it at least in part for this reason. Second, the claim that practical truth has an importantly ethical dimension should not be misunderstood as the claim that it is manifested in action which is narrowly morally good, right, or obligatory. To suggest so would conflict with Anscombe’s well-​known thesis that modern concepts of morality, as a special sui generis kind of demandingness, ‘ought to be jettisoned’ from ethical discourse (MMP, 26).18 Rather, the idea is that practical truth is embodied in actions which manifest the practical excellences, understood as making genuine contributions to one’s life’s going well. I have given an example of someone manifesting practical truth by acting justly, but justice is not the only good that a good life will contain. Actions can count as cases of doing or living well in virtue also of being kind, healthy, funny, stylish, brave, in being done out of friendship, and in many other ways. Bonum est multiplex. One will manifest practical truth by acting in these ways as long as in so doing one is aiming at and succeeding in acting or living well, and not also in some other way acting badly. (Being brave in the course of robbing a bank would not be acting well, for example; see PI, 140.) Finally, what amounts to acting or doing or living well will be highly situation-​ dependent. It is a familiar fact that although lying, for example, would often be acting badly, there are situations (with a friend hiding in the attic and an axe murderer at the door) in which it is precisely the thing—​perhaps indeed the only thing—​to do.19 The 17 

‘It has an absurd appearance when practical reasonings . . . are set out in full. . . . It is not clear from [Aristotle’s account of practical reasoning] whether he thinks a premise must be before the mind (‘contemplated’) in order to be ‘used’, nor is it of much interest to settle whether he thinks so or not’. (I, 79). 18  An excellent recent discussion of exactly what Anscombe is advising here, and her reasons for advising it, can be found in Doyle (2018). 19  This relates to Anscombe’s idea in Intention that when the ‘major’ premise of a practical syllogism is understood as universally quantified (‘It is necessary for all men over 60 to eat any food containing vitamin X that they ever come across’), it will generally be ‘an insane one’ (I, 61); ‘I think it is even safe to say that (except in, say, doing arithmetic or dancing, i.e., in skills or arts—​what Aristotle would call τέχναι) there is no general positive rule of the form ‘Always do X’ or ‘Doing X is always

On Anscombe on Practical Knowledge and Practical Truth    17 capacity for acting well requires a kind of sensitivity to the facts and demands of one’s circumstances rather than a strict adherence to general and context-​independent rules. And possessing the various virtues or excellences itself involves understanding how to exercise them in particular cases. This relates to the previous point: the fact that action takes place in specific circumstances means that doing one thing is at the same time doing a number of other things. Acting well requires understanding one’s circumstances in such a way that one is able, in acting well in one respect (e.g., telling a hilarious anecdote), to avoid acting badly in others (e.g., bullying one’s colleague) (for discussion see also Torralba 2016, 67–​68). Summing up.  In this and the previous section I have explained the notions of practical knowledge and practical truth as Anscombe introduces them. Both ideas are hard to understand, and neither lends itself to being sketched as briefly as space allows here. But hopefully the important points for what follows are clear enough. These are, first, that practical knowledge and practical truth both have a special ‘practical’ form, as well as being ‘practical’ in terms of their contents, and second, that this special practical form is in both cases a matter of the respective phenomenon’s being—​in its own way—​ embodied in action. Practical knowledge is embodied in the execution of intention, and practical truth is the truth of a complex of action-​descriptions, where these are made true by acting and are integral to the action itself.

1.4  No Interesting Relationship between Practical Knowledge and Practical Truth? Our question is: how do practical knowledge and practical truth relate to each other? An initial suggestion was mentioned in my introduction, viz.: The Simple View Practical knowledge is, essentially and in all cases, knowledge of practical truth.20

good–​required–​convenient–​, a useful–​suitable–​etc.–​thing’ (where the ‘X’ describes some specific action) which a sane person will accept as a starting-​point for reasoning out what to do in a particular case’ (I, 62). Anscombe is here talking about practical reasoning in general rather than ‘choice’ or prohairesis in particular (see n. 11), but the point applies in both cases. 20  José Torralba (2016) seems to suggest a view like this when he says that ‘if there are two distinct forms of knowledge (theoretical and practical), it can be assumed that each must have its corresponding form of truth’ (52) and that ‘practical knowledge is of practical truth (the good action)’ (68). He is clear that practical knowledge is essential to intentional action (63) and recognises that not all intentional action will be good (e.g., 70, 72 n64). This shows that he recognises that not all intentional action will manifest practical truth, but what he does not seem to recognise is the tension between these three

18   Lucy Campbell Practical truth is embodied in action, and practical knowledge is knowledge of action, so the Simple View holds that practical knowledge is at the same time knowledge of action and of the truth embodied in it. It should now be clear that the Simple View must be false. For we have seen that a person has practical knowledge of all of her intentional actions, whereas practical truth is manifested only in actions which aim at and succeed in being cases of living well. This mismatch in the respective scopes of practical knowledge and practical truth gives rise to various kinds of counterexample to the Simple View, but perhaps the most obvious is that of vicious action.21 The torturer is presumably not manifesting practical truth as she waterboards her victim—​she is making true various descriptions, but ‘acting well’ is not amongst them. Yet she is waterboarding her victim intentionally, and has practical knowledge that she is doing so. It might be suggested at this point that practical knowledge and practical truth are simply not related—​at least not in any interesting way. On this—​I’ll call it an Accidentalist—​account, practical knowledge and practical truth share the label ‘practical’, but we should not conclude from this that they are essentially connected to one another. Practical knowledge may be knowledge of practical truth in a given case (i.e., when the intentional action known aims at and succeeds in being a case of acting or living well), but the ‘practicality’ of the one is independent of the ‘practicality’ of the other. ‘Practical knowledge’ is to ‘practical truth’, on this Accidentalist account, as ‘political hire’ is to ‘political philosopher’. The faculty’s new hire might be both a political hire (he’s been hired to placate the dean) and a political philosopher (he works on Rawls), but where this is the case it is not because of any essential or general relationship between political hires and political philosophers, any essentially related sense in which they are both ‘political’. But the knowledge and truth we are concerned with do look ‘practical’ in related—​ indeed essentially related—​senses of ‘practical’. For we saw earlier that both practical knowledge and practical truth are ‘practical’ in form and not just in relation to their content; this special practical form in each case is a matter of the phenomenon’s ideas (i.e., the problem at the heart of the Simple View). Torralba’s focus is on the essentiality of an understanding of practical truth to an understanding of human action in general, and not on the relationship between practical knowledge and practical truth. For this reason his discussion does not tell us how to understand the object of practical knowledge when what is known falls short of manifesting practical truth. Does this practical knowledge have merely theoretical truth as its content? The account I give later is aimed at avoiding having to say such things. 21 

Any intentional action which fails either to aim at or to succeed in being a case of living well will provide an example. Actions which aim at living well but fail manifest practical falsity. Some actions, such as those done ‘for no particular reason’ (I, 25), or from akrasia, do not even aim at being cases of living well, and so on my interpretation manifest neither practical truth nor practical falsity. (NB: Anscombe has also been interpreted as holding that any action which does not manifest practical truth will manifest practical falsity [e.g., by Torralba 2016]). Some actions which fail to manifest practical truth (and practical falsity) may still count as ‘good’ in a very minimal way—​simply in not being bad (see, e.g., PI, 140). But an action’s being good in this thin sense falls short of its manifesting practical truth.

On Anscombe on Practical Knowledge and Practical Truth    19 being embodied in rather than merely concerning action. Practical truth was said to be manifested in one’s making true a (set of) descriptions of what one is doing, by doing that (set of) thing(s). And practical knowledge was knowledge embodied simply in the execution of intention (knowledge ‘in intention’). We can now add that executing an intention is making true a description of what one is doing, by doing it. Executing an intention is making true the description provided by the content of one’s intention. Like practical truth, then, practical knowledge is embodied in the making-​true of a (set of) description(s) of action in action, or by acting. Of course practical truth as described earlier is not the truth of any old set of action-​ descriptions, but must include the truth of a description like ‘acting well’ or ‘living well’. Still, when practical truth is manifested in such a case, the relationship between this practical truth and practical knowledge looks much stronger than the Accidentalist account of it would allow. So the practical knowledge–​practical truth relationship looks unlike the political hire–​political philosopher relationship, and I think the Accidentalist conception of it ought to be rejected.

1.5  An Essential but Merely Conditional Relationship? The reasons for rejecting the Accidentalist account suggest a stronger conception of the relationship between practical knowledge and practical truth: The Conditional Essentialist View If a person’s action manifests practical truth, this is, essentially and in all cases, the object of her practical knowledge.

Consider Virtuous Alexander again. He makes true the complex of descriptions ‘writing a cheque’, ‘paying Campbell back’, ‘doing what is just’, and ‘acting well’, and in so doing manifests practical truth. The Conditional Essentialist view holds that this practical truth is, essentially, the object of Alexander’s practical knowledge. And this looks right. For in doing all of these things, Alexander is executing his intentions: he is executing an intention to write a cheque, to pay Campbell back, to do what is just, and, in so doing, to act well. And practical knowledge, as we saw in section 1.3, is embodied in the execution of intentions.22 22  Another way of making this point is to say that Alexander acts on the basis of practical reasoning about how to act well in the circumstances. This is in fact more in keeping with Aristotle’s presentation (see again n. 11) and Anscombe’s development of it in her work on practical truth and elsewhere—​ although for Anscombe by contrast with Aristotle it is important not to think of ‘practical reasoning’ as naming a mental process the agent engages in prior to action, but as another way of talking about the practical syllogism, an abstract ‘order’ which characterises the structure of one’s action, and whose

20   Lucy Campbell The Conditional Essentialist view captures the idea that despite having different domains (intentional and virtuous action, respectively), when they do coincide—​when one’s intentional action is virtuous—​practical knowledge and practical truth are related essentially. This is precisely what the Accidentalist view just canvassed failed to account for. Another way of thinking about the current suggestion is as linking the capacities for practical knowledge and for manifesting practical truth. The Conditional Essentialist view allows us to say that perfect exercises of the capacity for practical rationality will be ones which manifest practical truth and which embody practical knowledge, the former essentially the object of the latter.23 This gives us a very plausible picture on which a person cannot unknowingly act well or virtuously. (One can do what is in fact just without recognising that one is doing so, but one would not thereby be acting justly.) So far, so good. Nevertheless, it is not satisfying to think of the Conditional Essentialist view as giving us the whole story about the relationship between practical knowledge and practical truth. Recall that the practicality of practical truth was supposed to reside in the fact that practical truth applies strictly and properly to action, understood as the idea that in manifesting practical truth, one makes true a description of what one is doing, by doing that thing, where the description is integral to the action. And we find this structure—​ truth of a description of an action which is integral to the action itself and which one makes true by acting—​in all intentional action, not just in cases in which one of the descriptions made true is ‘acting well’ or ‘living well’. Acting intentionally is itself a matter of making true a description of what one is doing, which description is integral to the action itself. To illustrate, consider a variation of Virtuous Alexander’s case: Vicious Alexander Alexander owes me £100. Justice requires that he pay me back, and he recognises this. He thinks he can pay me back by writing me a cheque. So he writes me a cheque, thus successfully paying me back, thus doing what is just. Unlike Virtuous Alexander, Vicious Alexander wants to do what justice requires not because doing so would be acting well, but simply because he wants his neighbour’s wife to think him a good chap. His real goal is to embark on an affair with her.

Stipulate that (contra Vicious Alexander) pursuing illicit affairs is not living well, so that unlike Virtuous Alexander, Vicious Alexander does not manifest practical truth.

contents can be elicited with her special question ‘Why?’ (I, 57–​58). To avoid including too many moving parts in my exposition, I have avoided explaining the structure we find in practical truth in terms of the idea of practical reasoning, sticking instead to talking about a complex of action-​descriptions, but I hope it is not too much of a leap to see how practical reasoning fits into the picture. For an exploration of the connection between practical truth and practical inference, see Hurley (2019, unpublished manuscript). 23 

Something like this idea is implicit in Elliott (2016).

On Anscombe on Practical Knowledge and Practical Truth    21 There is nevertheless in one sense a high degree of similarity between the two Alexanders’ cases. Both Alexanders make true the descriptions ‘writing a cheque’, ‘paying Campbell back’, and ‘doing what is just’ (although only Virtuous Alexander, in so doing, acts justly—​see again n. 14). They differ in that in doing these things, Virtuous Alexander also makes true the description ‘acting well’, whereas Vicious Alexander does not. Importantly, the mere fact that Vicious Alexander doesn’t make true the description ‘acting well’ doesn’t change the metaphysics of the relationships between his action and those descriptions which he does make true, nor between these and his practical knowledge. And the idea that practical truth is a special practical form of truth was supposed to be the idea that it is truth one creates when one makes descriptions true by acting. Vicious Alexander does this no less than Virtuous Alexander does. Again, both Vicious and Virtuous Alexander have practical knowledge of what they are doing under those descriptions they make true in acting as they do, and they would also seem both to know what they are intentionally doing simply by making true the descriptions under which they are intentionally acting. (This is another way of thinking about the idea that practical knowledge is ‘the cause of what it understands’.) The Conditional Essentialist view fails to capture the idea that there is something ‘practical’ in form about the truth which Vicious Alexander knows, in knowing what he is doing, even though unlike in Virtuous Alexander’s case, he does not know ‘practical truth’ in Aristotle’s specifically ethical sense. But how can we capture this idea, given that Vicious Alexander doesn’t manifest practical truth in Aristotle’s sense? This is our puzzle.

1.6  The Two-​L evel Simple View In response to our puzzle, I propose first that we distinguish within the category of practical truth a thin purely action-​t heoretical phenomenon and a thicker ethical phenomenon. I will label these, respectively, practical truth(A) and practical truth(E). Practical Truth(A) A person manifests practical truth(A) when she makes true a description or set of descriptions of her action, where these descriptions are determined by the content of the intention(s) (and/​or practical reasoning) which she executes in acting, and so which are integral to her action. Practical Truth(E) A person manifests practical truth(E) when she manifests practical truth(A) and when the members of the set of descriptions she thereby makes true are related as means to

22   Lucy Campbell ends, with the ultimate description she makes true being ‘living well’ (or ‘acting well’, ‘doing well’, etc.).

Aristotle’s notion of practical truth is clearly the thick ethical notion (see again n. 11). That Anscombe’s main focus should also be this notion is no surprise, because her discussion of practical truth initially takes place in the context of a focused interpretation of Aristotle, and then more generally in her ethical work. But as I argued in the previous section, the object of practical knowledge seems to be truth of a special practical form (truth of an action-​description made true by and integral to action) even when one is not aiming at and succeeding in acting or living well—​so even when one fails to manifest practical truth(E). And I think that this justifies distinguishing a thinner, purely action-​theoretical notion: practical truth(A). I also propose, second, to draw a parallel distinction within the category of practical knowledge, as follows: Practical Knowledge(A) A person has practical knowledge(A) when they know what they are doing and why. Practical Knowledge(E) A person has practical knowledge(E) when they know what they are doing and why, and when in knowing this, they know that they are doing or acting well, in the sense of contributing to living a good life (to living well), and so know what living a good life involves.

Just as it should not surprise us that Anscombe’s focus in relation to practical truth is on the thicker ethical phenomenon, it also should not surprise us that in her work on practical knowledge, she concentrates on the thinner, purely action-​theoretical notion. For whereas her discussion of practical truth occurs in her ethical works, her discussion of practical knowledge occurs in her action-​theoretical work, primarily in Intention, where she is explicitly mindful of avoiding discussing ethical matters (I, §§39, 41; see also MMP, 26). The justification for distinguishing an additional thicker ethical form of practical knowledge stems from the fact that when a person’s intentional action manifests practical truth(E), their practical knowledge of what they are doing will itself be knowledge of the good (and not just knowledge that they are φ-​ing, where φ-​ing happens in fact to be good). Consider Virtuous Alexander’s practical knowledge. Virtuous Alexander knows not only that he is writing a cheque, paying Campbell back, and doing what is just, but also that he is, in doing all these things, acting well, in the sense of making a contribution to living a good life, to living well. He knows this because he is executing intentions to do all of these things, and practical knowledge is manifested in the execution of intention. Virtuous Alexander thus has, in addition to knowledge of what he is doing and why, knowledge of what living well involves, and of how to act so as to contribute to living well in the specific circumstances in which he finds himself. His knowledge includes

On Anscombe on Practical Knowledge and Practical Truth    23 knowledge of what living well—​the good life—​involves.24 And if anything is a specifically ethical form of knowledge, it is this.25 Having distinguished a thin action-​theoretical and a thick ethical version of both practical knowledge and practical truth, we are finally in a position to better understand the relationship between the two phenomena. The Simple View ran into trouble because Anscombe’s explicit discussion makes available only a thin notion of practical knowledge and a thick notion of practical truth, meaning that it had to be understood as the view that practical knowledge(A) is essentially and in all cases knowledge of practical truth(E). Such a view is demonstrably false, as attention to (e.g.) vicious action shows. But with our new distinctions in place, we can now see a way of maintaining a kind of simple view, although admittedly a complex one. This is what I call the Two-​Level Simple View: The Two-​Level Simple View Practical knowledge(A) is essentially and in all cases knowledge of practical truth(A), and practical knowledge(E) is essentially and in all cases knowledge of practical truth(E).

According to the Two-​Level Simple View, a person manifests practical truth in at least the thin action-​theoretical sense I have outlined in relation to each and every intention she executes, so that practical knowledge (in any form) does essentially take practical truth (in some form) as its object. When a person manifests practical truth in the thicker ethical sense, her practical knowledge is likewise of a specifically ethical kind, being not only knowledge of what she is in fact doing, but knowledge of this as in some way good, and so also knowledge of the good. So far, so neat. But by introducing a thin action-​theoretical notion of practical truth and a thick ethical notion of practical knowledge, I have been going beyond Anscombe’s explicit characterisations of both practical knowledge and practical truth. If my suggestion is going to hold water from an interpretative perspective, I will need to make it plausible that the distinctions I have introduced are suggested by or implicit in Anscombe’s actual discussion. I turn to this now. 24  There are difficult issues here to do with how much one must know about what living well involves in order to have practical knowledge(E), and what this knowledge looks like. It would be too demanding to suggest that one must have a perfect understanding of the good life in order to act well. Plausibly this ‘understanding’ of what living well involves can be both partial and inexplicit. It is a form of understanding which is evident in the lives of those who are good at living (or good at certain aspects of living) even if (if this is possible) they spend no time thinking about what is involved in (these aspects of) living well. See also the following note. 25  Earlier (n. 3) I noted the connection between practical knowledge in the sense of knowledge how to do things and practical knowledge in the sense of knowledge that one is doing what one is intentionally doing. Anscombe makes this connection in relation to practical knowledge(A) in Intention (I, 88). In ‘Knowledge and Reverence for Human Life’, she draws what seems to be an analogous connection between what I am calling practical knowledge(E) and what looks like a specifically ethical form of knowledge-​how, what she calls ‘connatural knowledge’ [KRHL, 57, et passim]. (Thanks to John Hacker-​ Wright for pointing me in this direction.) Further development of the ideas in this chapter would bring in an understanding of knowledge-​how as it appears in its purely action-​theoretical guise (what we might call knowledge-​how(A)), and knowledge-​how as it appears in its thicker ethical guise (what we might call knowledge-​how(E), and what Anscombe calls ‘connatural knowledge’).

24   Lucy Campbell

1.7  A Difficult Passage The problem is that there are certain passages in Anscombe’s work in which she seems explicitly hostile to the suggestion I have made, in particular to the idea of practical truth(A). There is more than one passage which can be read in this way,26 but I think the most difficult one for the Two-​Level Simple View is the following, from the end of ‘Practical Truth’: [It] may have seemed that I must attribute ‘making true’ to any cause. If a branch falls and breaks a tea-​pot, the falling branch has made it true that the tea-​pot is broken. If a dog bites my hand and it bleeds, the dog has made it true that my hand is bleeding. One might indeed say these things, but they would be trivial and pointless. Practical truth is truth created by action in a sense in which neither branches nor dogs nor children are capable of action. It might be called ‘praxistic truth’ in order to emphasise that it is truth brought about by a praxis resulting from deliberation—​i.e., by an action (in fulfilment of a choice) which satisfies the description ‘doing well’. That is a final description of what every praxis—​every ‘action’ in this limited sense—​aims at being. (PT, 149, my italics)

Call this The Difficult Passage. In it, Anscombe complains about a notion of practical truth which is so thin as to be attributable to ‘any cause’: such a notion would, she says, be ‘trivial and pointless’. My notion of practical truth(A) is not attributable to just any cause. It is manifested in intentional action, so it would not be brought about by the falling of a branch onto a teapot.27 However, Anscombe also seems to deny that the action of children could 26 

Other passages which initially look difficult for my Two-​Level Simple View turn out to be relatively easy to deal with. Consider, for example: ‘The man who forms and executes an evil ‘choice’ will also make true some description of what he does. He will secure, say, if he is competent, that such and such a man has his eyes put out or his hands cut off, that being his judgement of what it is just to do. But his description ‘justice performed’ of what he has done will be a lie. He, then, will have produced practical falsehood’ (TAA, 77). I must hold that all intentional actions—​even vicious ones—​produce practical truth, although Anscombe here appears to say that such actions produce not practical truth but practical falsehood. But in fact this is easily countered in the context of the Two-​Level Simple View by pointing out that the practical falsity such a man produces does not conflict with his also producing practical truth. In, e.g., cutting off the other man’s hands, he produces practical falsity(E), but this is consistent with his also producing practical truth(A). Nothing Anscombe says in this passage conflicts with this suggestion. Similar readings can be given of other passages which might initially appear problematic for the Two-​Level Simple View. The passage I consider in the main text cannot be dealt with so easily, which is why I make it my focus. 27  Does the biting dog manifest practical truth ? No, since it is not plausible to think of a dog (A) as intentionally drawing blood. (The dog doesn’t specifically care about Anscombe’s hand bleeding.) But perhaps a dog can do other kinds of things intentionally—​chasing squirrels, for example—​and more generally perhaps animals are happily thought of as acting intentionally under certain simple descriptions. Anscombe herself seems to commit herself to the possibility of intention and intentional action by animals (I, 5; UD, 209). If we accept that animals can act intentionally, then we also have to accept—​on the account I am offering here—​that they can manifest practical truth(A). I won’t consider here whether this is a problem for Anscombe, as I am interpreting her.

On Anscombe on Practical Knowledge and Practical Truth    25 manifest practical truth—​and implies that a sense of ‘practical truth’ which applied (inter alia) to the actions of children would also be ‘trivial and pointless’. Yet practical truth(A) can be manifested in the actions of children, because children are capable of acting intentionally. Again, Anscombe seems to say that practical truth is brought about by—​not any old intentional action but—​praxis, understood as action which aims at doing well. In so saying she again seems to reject what I have called practical truth(A), by identifying (in the terms we now have available to us) practical truth per se with practical truth(E). Doesn’t this rule out my Two-​Level Simple View as a possible interpretation of Anscombe, given that this depends on taking seriously the phenomenon of practical truth(A)?

1.8  Anscombe’s Commitment to Practical Truth(A) I don’t think it does. For although she seems to reject it in The Difficult Passage, Anscombe herself elsewhere seems committed to what I have called practical truth(A). I’ll make two points in defence of this claim. First, Anscombe seems to make use of the notion of practical truth(A) in various places; second, we can derive from her discussion an explanation of the point of the notion of practical truth(E) which suggests by analogy that there would also be a point to the notion of practical truth(A). This gives us a reason, derived from what Anscombe actually says, to deny that practical truth(A) would be ‘trivial and pointless’—​notwithstanding what she says in The Difficult Passage. I return to the question of how to interpret The Difficult Passage itself after making these points, in section 1.9. Relying on practical truth(A).  Starting with the first point, there are various places where Anscombe makes use of practical truth(A), or otherwise says things which would seem to commit her to it. Explaining the dialectical context of many of these examples would require more space than I have here, but luckily there are a couple of cases which require minimal scene-​setting, so I focus on these. In ‘Practical Truth’—​the paper from which The Difficult Passage comes—​Anscombe introduces the notion of practical truth(E) as follows: [Practical truth(E)] is truth that one produces in acting according to choice and decision. There is often resistance to the idea that one can produce truth, i.e., make something true. But can it not be that one brings it about that p? And are not p and the truth of p equivalent? I.e.,

It is true that p ≡ p. If, then, one can bring it about that p, then one can make it true that p. (PT, 147–​148)

She then gives an example in which she makes it true that she is signing a contract. This is something which of course might aim at and succeed in being a case of acting well, but it

26   Lucy Campbell certainly needn’t.28 On the contrary, she here introduces the notion of practical truth(E), which is manifested only in actions which aim at and succeed in being cases of acting well, by reference to what I have called practical truth(A), a kind of truth one brings about by acting, but which need not be a case of acting well. We find something similar in a different, earlier paper. In ‘Practical Inference’, Anscombe discusses a kind of ‘truth which you make true by acting’ (PI, 137n14), but which falls short of practical truth in Aristotle’s sense: For some reason, people find this idea [of truth which you make true by acting] very difficult. In lecturing I have sometimes tried to get it across by saying: ‘I am about to make it true that I am on this table’. I then climb on the table. Whether I have made it true that my hearers understand, I do not know. (ibid.)

As with the example of signing a contract, there is no explicit suggestion in how she sets up the example that in climbing on the table she is aiming at and succeeding in acting or living well. Still, it might be pointed out here that although she doesn’t say so, getting on the table is aimed at acting well in this case—​because it is aimed at getting her hearer to understand the point she is making, and so at contributing to her communicative endeavours, where communicating with one’s fellow humans is presumably an important feature of a life lived well, and especially for a philosopher. This might be right, but Anscombe is clear in the passage that she is able to make it true that she is on the table consistently with failing to make it true that her hearer understands her point. This would mean that she can create a form of practical truth consistently with failing to make true the description ‘acting well’ (which description she would make true in the present case only if she succeeded in making true the description ‘making my hearer understand’). Again, it looks as if Anscombe admits a conception of making true and so of practical truth which attaches to intentional action per se, and is not restricted to action which both aims at and succeeds in being a case of living well. The context of the passage just quoted gives us further reason to insist that it is practical truth(A) that Anscombe must have in mind. Aristotle describes practical truth as ‘truth in agreement with right desire’; it is truth in which one brings it about that things are the way one wants—​and ought to want—​them to be (see again n. 11). In the relevant part of ‘Practical Inference’, Anscombe describes a less committal kind of truth, ‘truth in agreement with desire’ (PI, 137, 138)—​without the qualification that the desire must be ‘right’. Anscombe’s complaint in this passage is against a conception of rational agency which recognises truth in action which is only of this latter, less committal kind. (She attributes such a view to von Wright 1972.) But in making this complaint, she commits herself to there being such a thing as this less committal kind of truth in action. The case of making it true that she is on this table by climbing on it is just such a case of truth in agreement with desire. She clearly argues here against a view on which all practical truth 28 

Earlier in ‘Practical Truth’ Anscombe gives an example in which signing a contract would be living well (PT, 142).

On Anscombe on Practical Knowledge and Practical Truth    27 is practical truth(A), but in the course of making this argument she commits herself to some practical truth being practical truth(A). An opponent might object at this point, by arguing that I am conflating two ideas which ought to be kept apart: practical truth and making true. Not all cases of truth which you make true by acting are cases of practical truth, the opponent might suggest. We get practical truth only when one of the things one makes true is the description ‘living well’. On this interpretation, Anscombe holds that making true comes in an action-​theoretical as well as an ethical form, but that practical truth itself only comes in an ethical form. But this interpretation doesn’t seem right, because Anscombe appears to treat making true as simply bringing about practical truth. The Difficult Passage is itself an example of this tendency. It starts with the worry that Anscombe might seem to have to attribute ‘making true’ to any cause, and it is this idea which she rejects as ‘trivial and pointless’. Anscombe explains the problem about ‘attributing ‘making true’ to any cause’ by emphasising that practical truth is of a special ethical kind (that it is ‘praxistic truth’, etc.). There would simply be no worry about ‘making true’—​even any prima facie worry—​ unless it was to be understood as bringing about practical truth. Motivating practical truth(A). This leads us to the second reason to think that Anscombe is committed to practical truth(A). Anscombe’s complaint that too thin a notion of practical truth would be ‘trivial and pointless’ suggests that there ought to be something to say about what the point of the thicker notion of practical truth(E) would be. I think we can get a sense of what the point of the notion of practical truth(E) might be by looking at an earlier passage in ‘Practical Truth’. Anscombe quotes Aristotle as saying: ‘For doing well is the end, and the desire in decision is for that. So decision is desiring thought or thinking desire—​and the cause of this kind is man’. (PT, 144–​5, quoting Aristotle Nicomachean Ethics 1139b 4–​5)

She continues: Note the importance of this final sentence of the passage. There is this special kind of cause operating in the world, and it is man. (PT, 145)

I want to suggest that at least part of the point of the notion of practical truth(E) and the attendant idea of making true is to capture a sense in which human agents are a special kind of cause; practical truths are truths or facts which are—​which could only be—​ manifested or precipitated by the ‘special kind of cause’ which is man. In particular, in manifesting practical truth(E), a person is the cause of goodness or virtue or excellence being instantiated in the world. Plausibly, the point of the notion of practical truth(E) is to capture the fact that man is a special kind of cause which instantiates such ethical value in the world, where the instantiation of ethical value in the world would otherwise lack an explanation. A person is a ‘special kind of cause’ in relation to such facts in that she is the principle which explains their obtaining.

28   Lucy Campbell But if this is an accurate gloss on Anscombe’s views about the point of the notion of practical truth(E), then it gives us an argument by analogy for the conclusion that there is also a point to the thinner notion of practical truth(A). For humans are also a special kind of cause in relation to their intentional actions per se, whether or not these are cases of acting or living well.29 Man is a special kind of cause in relation not only to facts which instantiate ethical value, but also in relation to what we might call mental value, intentionality, or (in a certain sense)30 rationality. The point of the notion of practical truth(A) would be to pick out a set of facts brought about by a special kind of causality—​by precisely that causality involved in acting intentionally. We can further illustrate the point by returning to The Difficult Passage. Anscombe says there that it would be ‘trivial and pointless’ to say that a falling branch ‘makes it true’ that a teapot is broken, or manifests practical truth in breaking the teapot. If the suggestion about the point of talk of ‘making true’ is right, then she’s right, because in so saying, we wouldn’t be picking out any special kind of causality, or pointing to the teapot’s breaking as explicable simply by reference to the branch. The branch does not relate to the teapot’s breaking in the same way that I relate to the teapot breaking if I smash it to bits with a hammer. In such a case, I am the cause of the teapot’s breaking in a sense in which the branch is not. I am responsible for the teapot’s breaking in this case in a way in which the branch could not be responsible for the teapot’s breaking (see also PT, 141). Yes, the branch broke the teapot by falling on it, but in so saying, we are not saying that the teapot’s breaking was down to the branch, in the way that the teapot’s breaking is down to me when I smash it with a hammer. Unlike the branch, I am a cause—​an explanatory principle—​in relation to my intentional actions in a way which makes essential reference to my mindedness, thereby explaining how I am the starting point of some chain of effects-​cum-​causes, rather than simply one effect-​cum-​cause amongst many in a chain. I am a special kind of cause, a kind which a branch could not be. My actions are causings which are not also effects of causings of the same kind. We cannot say this about the branch’s falling. For this reason, we can say that the teapot’s breaking (in one case) was ‘down to me’ by being down to my action, in a sense in which it was not (in the other case) down to the branch or to the branch’s ‘action’. The teapot was broken by the branch’s falling, but the branch did not do the breaking.31 29  I do not mean that one exerts one’s causal influence over one’s intentional actions, but that acting intentionally is at least ordinarily exerting one’s causal influence. 30  That is, the sense of ‘rational’ in which it is opposed to ‘arational’ as opposed to ‘irrational’. The limit case of instantiating rationality, intentionality, or ‘mental value’ is one in which someone acts intentionally but for no particular reason (I, §17). 31  Thomas Pink (2018) expresses essentially the same idea in a different context (in explicating Suarez’s philosophy of action). Ordinary causes like branches and bricks do not ‘determine things for themselves’: ‘When it hits the window, [a]‌brick does not determine for itself that the window breaks. The brick’s causation of the window’s breaking seems dictated by brick’s circumstances [sic] and its own given nature. That the window breaks is necessitated by the brick’s mere possession of a power to break it when hurled against it. That is why we do not think of a brick as being in control of alternative possible outcomes. It is not up to the brick whether the window breaks or not’ (11).

On Anscombe on Practical Knowledge and Practical Truth    29 Importantly, all this is so despite the fact that I don’t, in smashing the teapot with a hammer, manifest practical truth(E), but only what I have called practical truth(A), for we can easily stipulate that the teapot-​smashing case neither is nor aims at being a case of acting or living well. (I might answer the question why I smashed the teapot with a hammer by saying ‘I just thought I would’, or by showing that my action was akratic.) For this reason, we seem to have a use for a notion of ‘making true’—​which is allied to a sense in which humans are a special kind of mind-​involving cause or explanatory principle—​which applies in cases where the set of the descriptions made true does not include ‘acting well’ (or some other such description). In summary, human action is itself a special kind of causality, quite independently of whether, in a given case, the action being engaged in is a case of acting or living well. If the point of talking about making true and practical truth is to capture the sense in which man is a special kind of cause, then this is as much an argument for adopting a thin action-​theoretical notion of practical truth as it is for adopting Aristotle’s thick ethical notion. The point of the thin action-​theoretical notion of ‘practical truth’ which I have distinguished is to underscore just this kind of causality.

1.9  Interpreting the Difficult Passage I have been arguing that Anscombe is committed to practical truth(A). But it also looked as if she rejects the notion in The Difficult Passage. So where does this leave us? As far as I can see, there are three interpretative options for The Difficult Passage, none of which is completely comfortable. First, we might read the passage as it appears, as rejecting practical truth(A). This has the benefit of being a straightforward reading of the passage, but the drawback of rendering Anscombe inconsistent, since, as I have just been arguing, she is herself committed to practical truth(A). Second, we could read Anscombe as rejecting only a very thin notion of practical truth, on which branches would count as making things true. This is supported by my suggestion that the point of the notion of (of any notion of) practical truth is bound up with the idea of a special kind of causality which explains the instantiation of a special form of value in the world, since branches (etc.) are not causes in this sense. And this interpretation is consistent with (what I have argued is) Anscombe’s commitment to practical truth(A), because practical truth(A) cannot be manifested by branches. There are two difficulties for this interpretation. First, Anscombe also denies that dogs can make things true, but in other places she seems to commit herself to the claim that animals are capable of intentional action, which would commit her to the view that animals are capable of manifesting practical truth(A) (see again n. 27). Second, recall that she continues by saying, ‘Practical truth is truth created by action in a sense in which neither branches nor dogs nor children are capable of action’ (my italics). Regardless of what we think about the possibility of dogs acting intentionally, surely we can’t deny that

30   Lucy Campbell children can act intentionally. Given this, it is very hard to view Anscombe as rejecting only an extremely thin notion of practical truth rather than the more robust notion of practical truth(A).32 A final option is to read Anscombe’s claims about practical truth in The Difficult Passage as restricted to practical truth(E), so that she’s really complaining about the suggestion that practical truth(E) might be applied to dogs and branches and children. This would be consistent with her accepting practical truth(A) in other areas of her discussion, which applies to intentional action very generally, including the intentional actions of children (and perhaps dogs), but not to the ‘action’ of falling branches. But this interpretation, like the previous one, suffers from the problem that it is very difficult to read The Difficult Passage in this way. Anscombe simply doesn’t sound like she is saying this, but seems instead to be identifying practical truth (per se) with what I have demarcated as practical truth(E). It is hard to get away from the fact that she seems in this passage to reject practical truth(A) as ‘trivial and pointless’. In summary, given her apparent commitment to practical truth(A) (as argued in the previous section), I am not sure that there is any happy interpretation of The Difficult Passage. But if this is right, then we can’t be expected to give it too much weight in an interpretation of the relationship between the notions of practical knowledge and practical truth in Anscombe’s philosophy in general. If so, and we can take The Difficult Passage with a pinch of salt, then we should not view it as providing a particularly serious problem for the Two-​Level Simple View of the relationship between practical knowledge and practical truth.

1.10 Conclusion Anscombe talks about practical knowledge, and she talks about practical truth, but she doesn’t explicitly tell us how the two relate. Given that practical knowledge is introduced as taking in intentional action, whereas practical truth is introduced as manifested only in certain intentional actions—​those which aim and succeed in being cases of acting well—​we didn’t seem able to accept a simple version of the Simple View, on which practical knowledge is, essentially, and in all cases, knowledge of practical truth. I have argued that we can nevertheless accept a complex Simple View—​what I have called the Two-​Level Simple View. This requires distinguishing a thin purely action-​ theoretical and a thicker ethical notion, of practical knowledge on the one hand, and

32 

Some have suggested to me in conversation that Anscombe might mean ‘infants’ here rather than ‘children’. This would help because there is no problem with denying that infants can act intentionally or manifest practical truth(A). But Anscombe doesn’t say ‘infants’, she says ‘children’ (not only here but also in other places too, where she makes the same point [e.g. PT, 143]). At any rate, the rest of The Difficult Passage seems to positively identify practical truth per se with practical truth(E), which would independently mean that she rejects practical truth(A).

On Anscombe on Practical Knowledge and Practical Truth    31 practical truth on the other. On the face of it, viewing Anscombe as accepting practical truth(A) seemed difficult to square with what she actually says, in particular in The Difficult Passage from the end of ‘Practical Truth’. But I have argued that notwithstanding what she says there, Anscombe herself is committed to accepting practical truth(A), so that we should take The Difficult Passage with a pinch of salt. The Two-​Level Simple View thus remains a live option for linking these two important but hard to understand notions in Anscombe’s philosophy.

References Anscombe, G. E. M. 1965. ‘Thought and Action in Aristotle: What Is ‘Practical Truth’?’ In New Essays on Plato and Aristotle, edited by R. Bambrough, 143–​158. London: Routledge & Kegan Paul. Aristotle. 2002. Nicomachean Ethics. Edited by Christopher Rowe and Sarah Broadie. Translated by Christopher Rowe. Oxford: Oxford University Press. Broadie, Sarah. 2016. ‘Practical Truth in Aristotle’. American Catholic Philosophical Quarterly 90 (2): 281–​298. Campbell, Lucy. 2016. ‘Action, Intention, and Knowledge’. Doctoral thesis, University of Cambridge. https://​www.rep​osit​ory.cam.ac.uk/​han​dle/​1810/​253​765. Campbell, Lucy. 2018a. ‘An Epistemology for Practical Knowledge’. Canadian Journal of Philosophy 48 (2): 159–​177. Campbell, Lucy. 2018b. ‘Two Notions of Intentional Action? Solving a Puzzle in Anscombe’s Intention’. British Journal for the History of Philosophy 26(3): 578–​602. Charles, David. 2018. ‘Practical Truth: An Interpretation of Parts of NE VI’. In Virtue, Happiness, Knowledge: Themes from the Work of Gail Fine and Terence Irwin, edited by David O. Brink, Susan Suavé Meyer, and Christopher Sheilds, 149–​168. Oxford: Oxford University Press. Davidson, Donald. 2001. ‘Agency’. In Essays on Actions and Events. 2nd ed., 43–​61. Oxford: Clarendon Press. Davidson, Donald. 2004. ‘Problems in the Explanation of Action’. In Problems of Rationality, 101–​116. Oxford: Oxford University Press. Doyle, James. 2018. No Morality No Self: Anscombe’s Radical Skepticism. Cambridge, MA: Harvard University Press. Elliott, Jay R. 2016. ‘Anscombe on Practical Truth’. Klesis 35 (Lectures contemporaines de Elisabeth Anscombe): 108–​125. Graeser, Andreas. 2005. ‘Aristotle on Practical Truth: Coherence vs. Correspondence’. Bochumer Philosophisches Jahrbuch Fur Antike Und Mittelalter 9 (1): 191–​200. Hurley, Paul. ‘Anscombe’s Account of Practical Truth via Practical Soundness, with Detours through Sellars, Korsgaard, and Davidson’ (Claremont, CA, 2019). McDowell, John. 2011. ‘Anscombe on Bodily Self-​ Knowledge’. In Essays on Anscombe’s Intention, edited by Anton Ford, Jennifer Hornsby, and Frederick Stoutland, 128–​146. Cambridge, MA: Harvard University Press. Olfert, C. M. M. 2014. ‘Aristotle’s Conception of Practical Truth’. Journal of the History of Philosophy 52 (2): 205–​231. Olfert, C. M. M. 2017. Aristotle on Practical Truth. Oxford: Oxford University Press. Pakaluk, Michael. 2010. ‘The Great Question of Practical Truth, and a Diminutive Answer’. Acta Philosophica 1 (19): 145–​160.

32   Lucy Campbell Pink, Thomas. 2018. ‘Agents, Objects, and Their Powers in Suarez and Hobbes’. In Philosophy of Action from Suarez to Anscombe, edited by Constantine Sandis, 9–​30. London: Routledge. Thompson, Michael. 2011. ‘Anscombe’s Intention and Practical Knowledge’. In Essays on Anscombe’s Intention, edited by Anton Ford, Jennifer Hornsby, and Frederick Stoutland, 198–​210. Cambridge, MA: Harvard University Press. Torralba, José M. 2016. ‘On Morally Neutral Actions, and the Relevance of Practical Truth for Action Theory’. In The Moral Philosophy of Elizabeth Anscombe, edited by Luke Gormally, David Albert Jones, and Roger Teichmann, 51–​74. St Andrews Studies in Philosophy and Public Affairs. Exeter: Imprint Academic. von Wright, G. H. 1972. ‘On So-​Called Practical Inference’. Acta Sociologica, Problems in the Philosophy of Social Science 15 (1): 39–​53. Wolfson, Ben. 2012. ‘Agential Knowledge, Action, and Process’. Theoria 78: 326–​357.

Chapter 2

In tention in Ac t i on Charles F. Capps

2.1  Introduction Elizabeth Anscombe opens her book Intention by noting a threefold division in the topic of intention.* First, there is what she calls ‘expression’ of intention for the future—​ a statement of one’s intention to do something. Second, there is intentionally acting—​ doing something intentionally. Third, there is intention in action—​an intention with which something is done. Observing that the word ‘intention’ as it appears in these three guises is not equivocal, Anscombe sets out to uncover the unity that holds them together.1 Her approach to the topic of intention has been tremendously influential: Kieran Setiya opens the entry on intention in the Stanford Encyclopedia of Philosophy by remarking that ‘[p]‌hilosophical perplexity about intention begins with its appearance in three guises: intention for the future . . . the intention with which someone acts . . . and intentional action’.2 ‘The principal task of the philosophy of intention’, Setiya continues, ‘is to uncover and describe the unity of these three forms’.3 The standard approach to unifying Anscombe’s three divisions is to treat intention for the future, construed broadly to include the intention to do something immediately as well as the intention to do something later, as fundamental, explaining intentionally acting and intention in action in terms of it. Call this approach ‘IF Priority’. IF Priority *  I am grateful to Matthias Haase, Christian Kietzmann, Anselm Müller, Roger Teichmann, and the participants in the Practical Philosophy Workshop at the University of Chicago for their comments on drafts of this paper. 1 

I, 1. Kieran Setiya, ‘Intention’, in The Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta, Fall 2018, https://​plato.stanf​ord.edu/​entr​ies/​intent​ion/​. Unfortunately, terms like ‘intentional action’ and ‘intentionally acting’ are ambiguous between ‘doing something intentionally’ (which entails doing what one intends) and ‘acting with intentions’ (which does not). When I use ‘intentionally acting’, I mean ‘doing something intentionally’. 3  Setiya, ‘Intention’. 2 

34   Charles F. Capps is defended in some form or another by Michael Bratman, Alan Donagan, Donald Davidson, and most other writers in the mainstream of philosophy of action since the middle of the twentieth century.4 Anscombe herself rejected IF Priority. After spending §§2–​4 of Intention attempting to understand intention for the future without first elucidating intentionally acting and/​or intention in action, she concludes that the attempt is futile. Though tempting, it is a mistake to think that the key to understanding intention is ‘investigating something . . . in the sphere of the mind’ that merely ‘issues in actions’.5 Anscombe proceeds to spend most of the book (§§5–​49) elucidating intentionally acting and intention in action, returning to intention for the future only at the end (§§50–​52). Today, most who follow Anscombe in rejecting IF Priority subscribe to what I will call ‘IA Priority’: the approach to unifying Anscombe’s three divisions that treats intentionally acting as fundamental and explains intention for the future and intention in action in terms of it. According to the standard and, I take it, most plausible version of IA Priority, an agent intends to ϕ, or is acting with the intention of ϕ-​ing, in virtue of being in the process of ϕ-​ing intentionally. Eric Marcus, Michael Thompson, and Richard Moran and Martin Stone all defend a view along these lines.6 Yet although defenders of IA Priority often cite Anscombe as an ally, there is evidence that she rejected their view as well. To accommodate cases in which one acts with the intention of ϕ-​ing but does not manage to ϕ intentionally, defenders of IA Priority point out that one can be in the process of doing something even if it will never be the case that one did it: one might stop doing it or be interrupted.7 And to accommodate cases in which one is acting with the intention of ϕ-​ing but seems not yet to be in the process of ϕ-​ing, defenders of IA Priority point out that one can be in the process of doing something even if one is not far enough along for the fact that one is in the process of doing it to be evident in the progress that one has made.8 Yet immediately after making these two points herself, Anscombe declares that nonetheless there are cases in which one ‘manifestly is not’ in the process of doing what one is acting with the intention of doing.9

4 

See Michael Bratman, Intention, Plans, and Practical Reason (Cambridge, MA: Harvard University Press, 1987), 1–​11; Alan Donagan, Choice: The Essential Element in Human Action (London: Routledge, 1987), 94–​112; Donald Davidson, ‘Intending’, in Essays on Action and Events (Oxford: Oxford University Press, 2001), 83–​102. 5  I, 9. 6  See Eric Marcus, Rational Causation (Cambridge, MA: Harvard University Press, 2012); Michael Thompson, Life and Action: Elementary Structures of Practice and Practical Thought (Cambridge, MA: Harvard University Press, 2008), 85–​146; Richard Moran and Martin J. Stone, ‘Anscombe on Expression of Intention’, in New Essays on the Explanation of Action, ed. Constantine Sandis (Hampshire: Palgrave Macmillan, 2008), 132–​68. Although I am following Setiya in interpreting Thompson as defending IA Priority (see Setiya, ‘Intention’), one could read Life and Action, 112 as an endorsement of the view that intention in action is fundamental. Nothing in my argument turns on which interpretation captures Thompson’s view. 7  See, e.g., Thompson, Life and Action, 124. 8  See, e.g., Thompson, Life and Action, 140–​141. 9  I, 39.

Intention in Action    35 One of Anscombe’s examples is a case in which a man is operating a water pump with the intention of ‘replenishing [a cistern’s] water-​supply’, but replenishing the cistern’s water supply ‘is not happening because, as we can see but [the agent] cannot, the water is pouring out of a hole in a pipe on the way to the cistern’.10 If Anscombe rejected both IF Priority and IA Priority, then what approach to unifying her three divisions of the topic of intention did she endorse? Unfortunately, she never tells us explicitly. However, I believe that she supplies us with the materials to construct a version of the relatively neglected approach that I will call ‘IW Priority’—​the approach that treats the intention with which something is done as fundamental and explains intention for the future and intentionally acting in terms of it.11 In what follows, I will attempt to sketch this alternative and show that, whether or not it represents Anscombe’s considered view, it is at least plausible.12 I will divide my argument into three sections, corresponding to the three elements that constitute a complete theory of IW Priority: an account of intention for the future in terms of intention in action, an account of intentionally acting in terms of intention in action, and an account of intention in action not in terms either of intention for the future or of intentionally acting. Most of the work will go into developing an account of intention in action in section 2.2. I will develop accounts of intentionally acting and intention for the future in sections 2.3 and 2.4, respectively.

2.2  Intention in Action Anscombe famously observed that the practical syllogism ‘describes an order which is there whenever actions are done with intentions’.13 When an agent acts with intentions, her action manifests endorsement of an argument for performing the action. Of course, the agent need not be occurrently aware of endorsing such an argument. Usually, she is merely conscious of doing so in the sense that, if asked what her reasons are for performing the action, she can answer without having to observe herself. Still, the contents 10  I, 39. Additional evidence that Anscombe rejected IA Priority is available. For example, she denies that ‘a distinct . . . state’ of ‘having a certain intention’ must pre-​exist ‘the doing of what the intention was an intention to do’ but allows that it can (CA, 101). And she distinguishes within cases in which an agent is doing X with the intention of doing Y between (i) those in which the agent ‘is doing Y in, and at the same time as, doing X’ and (ii) those in which the agent ‘is not yet doing Y but only doing X with a view to doing Y, as when a man takes his gun down with a view to shooting rabbits’ (I, 31). Although Anscombe recognises that this distinction admits of borderline cases (see I, 39–​40), she evidently thinks that there are clear cases on each side. 11  Amongst the few defenders of IW Priority are Kim Frost, ‘On the Very Idea of Direction of Fit’, Philosophical Review 123, no. 4 (October 2014): 480 and George M. Wilson, The Intentionality of Human Action (Stanford, CA: Stanford University Press, 1989), 129–​130. 12  Of course, IF Priority, IA Priority, and IW Priority do not exhaust the possibilities. There are ways to unite Anscombe’s three divisions other than analysing two of them in terms of the third. 13  I, 80.

36   Charles F. Capps of the intentions with which she is acting can be read off the contents of the practical syllogism that, however implicitly, she has endorsed. This relation between the practical syllogism and intention in action cannot be a coincidence. The most natural and, I think, plausible way to account for it is to say that an agent acts with certain intentions in virtue of acting on her endorsement of a practical syllogism with certain contents. In this section, I will defend a version of this view that is compatible with IW Priority. In section 2.2.1, I will argue that endorsing a practical syllogism gives rise in the first instance to intentions in action rather than intentions for the future. This will pave the way for an account of intention in action in section 2.2.2 that places practical reasoning, but not intention for the future, prior to intention in action. In section 2.2.3, I will defend this account against an objection.

2.2.1  Practical Judgement It is not obviously compatible with IW Priority to say that an agent acts with certain intentions in virtue of acting on her endorsement of a practical syllogism with certain contents. According to a widespread conception of practical reasoning, when an agent acts on her endorsement of a practical syllogism, her action does not begin until after she has endorsed the syllogism. As Joseph Raz puts it, ‘the reasoning is over before the act is performed’.14 But if it is practical reasoning before action that generates the intentions with which an agent acts, then presumably this practical reasoning had generated those intentions before the agent acted and would have done so even if the agent had never gotten around to acting. This implies that the intentions generated by practical reasoning are, in the first instance, intentions for the (perhaps immediate) future. They may also be intentions with which the agent acts, if the agent gets around to acting. But they are independently identifiable as intentions for the future regardless. Therefore, if it is practical reasoning before action that generates the intentions with which action is performed, then it is not the case that intention for the future is to be explained in terms of intention in action. And this is incompatible with IW Priority. To make space for an account of intention in action in terms of practical reasoning that is compatible with IW Priority, I will deny that an agent who acts on her endorsement of a practical syllogism begins to do so only after having endorsed the syllogism. When an agent acts on her endorsement of a practical syllogism, I will argue, the

14 

Joseph Raz, From Normativity to Responsibility (Oxford: Oxford University Press, 2011), 133. See also Julian Fink, ‘What Is (Correct) Practical Reasoning?’, Acta Analytica 28, no. 4 (December 2013): 472–​ 473; Sarah Paul, ‘The Conclusion of Practical Reasoning: The Shadow between Idea and Act’, Canadian Journal of Philosophy 43, no. 3 (September 2013): 288; Bertram Enç, How We Act: Causes, Reasons, and Intentions (Oxford: Oxford University Press, 2003), 179–​183; John Broome, ‘Practical Reasoning’, in Reason and Nature: Essays in the Theory of Rationality, ed. José Luis Bermúdez and Alan Millar (Oxford: Clarendon Press, 2002), 85–​86.

Intention in Action    37 judgement that constitutes this endorsement is the kind of judgement that, if sound, yields the kind of nonobservational knowledge of action that Anscombe calls ‘practical knowledge’.15 I will call this kind of judgement ‘practical judgement’. Practical judgement, I will argue, does not precede but rather is contemporaneous with action. For my argument to succeed, I must show two things about practical judgement. First, I must show that practical judgement constitutes endorsement not merely of the conclusion of a practical syllogism but of a practical syllogism in its entirety, that is, affirmation of the conclusion on the basis of the premises. I will refer to this as the claim that practical judgement is ‘reasoned’. Because the intentions with which an agent acts cannot be read off the conclusion alone of the practical syllogism that she has endorsed, practical judgement must be reasoned if it is to have the right content to ground intention in action. Second, I must show that practical judgement does not precede the action that is the object of that judgement. To show this, I will argue that an agent’s practical judgement really is of her action in the sense that it is a singular judgement about that action as a particular rather than a general judgement that some action with such-​and-​ such properties is occurring or will occur. I will refer to this as the claim that practical judgement is ‘singular’.16 If practical judgement is singular, then it cannot precede the particular action that is its object. I will present my primary argument that practical judgement is reasoned and singular in section 2.2.1.1. The argument is based on what I will call the ‘chimerical’17 nature of practical judgement—​the fact that it has both descriptive and normative content. In section 2.2.1.2, I will offer a second argument for the claim that practical judgement is singular. In section 2.2.1.3, I will offer a second argument for the claim that practical judgement is reasoned.

2.2.1.1  The Chimerical Nature of Practical Judgement My primary argument that practical judgement is reasoned and singular is based on the chimerical nature of practical judgement. The first step in the argument is to show that practical judgement is chimerical in the sense that it has both descriptive and normative content. The second step is to show that the best explanation for the chimerical nature of practical judgement is that practical judgement is reasoned and singular. Philosophers disagree about whether the content of the judgement that yields practical knowledge is descriptive or normative. According to ‘descriptivists’ such as John McDowell, a practical judgement is a purely descriptive judgement about what one

15 

I, 57–​89. The claim that practical judgement is singular is plausible, of course, only if actions are particulars—​and not just upon reaching completion. I do not think that this counts against the claim that practical knowledge is singular. See Carl Ginet, On Action (New York: Cambridge University Press, 1990), 139; UD, 213. 17  I borrow this term from Anselm Winfried Müller, ‘Is Practical Truth a Chimera? Questions for Anscombe’, lecture delivered at ‘Practical Truth: Reflections on the Aristotelian Tradition’, University of South Carolina, Columbia, April 2017, https://​www.yout​ube.com/​watch?v=​5nNH​HiLB​sZM. 16 

38   Charles F. Capps is doing.18 According to ‘normativists’ such as Eric Marcus, a practical judgement is a purely normative judgement about what to do.19 Descriptivists and normativists alike cite Anscombe for support.20 And not without reason, for Anscombe appears to be of two minds on the question. On the one hand, she introduces practical knowledge as the nonobservational knowledge that an agent ordinarily has of what she is doing intentionally.21 On the other hand, she characterises the reasoning that is supposed to lead to practical knowledge as reasoning about ‘what to do’.22 To make matters worse, when considering a practical syllogism that concludes with the agent eating some ‘pigs’ tripes’, Anscombe proposes three apparently equally acceptable ‘verbali[sations]’ of this conclusion: ‘So I’ll have some’; ‘So I’d better have some’; and ‘So it’d be a good thing for me to have some’.23 The first looks like a descriptive judgement about what the agent will do; the second and third look like normative judgements about what to do. Although descriptivists and normativists each attempt to accommodate the passages in Anscombe that seem to conflict with their theory, these attempts are unconvincing, substantively as well as exegetically. Start with the descriptivists. Although McDowell addresses ‘the evaluative elements’ in Anscombe’s second and third ‘verbali[sations]’ of the conclusion of the pigs’ tripes inference,24 he does not address Anscombe’s recognition that practical reasoning is reasoning about what to do. He evidently agrees, however, because he himself characterises practical reasoning as ‘address[ing] the question what to do’.25 But if practical reasoning addresses the question what to do, then descriptivism is false. For reasoning issues in a judgement in answer to the question that the reasoning addressed. Practical judgement, then, is judgement about what to do.26 Normativists fare little better. Marcus attempts to reconcile Anscombe’s characterisation of practical knowledge as knowledge what one is doing by identifying endorsing the conclusion of a practical syllogism—​judging that an action of a certain 18 

John McDowell, ‘What Is the Content of an Intention in Action?’, Ratio 23, no. 4 (December 2010): 415–​432. 19  Eric Marcus, ‘Practical Knowledge as Knowledge of a Normative Judgment’, Manuscrito 41, no. 4 (October–​December 2018): 319–​347. 20  See, e.g., Marcus, ‘Practical Knowledge’, 329–​330; McDowell, ‘Intention in Action’, 423–​424. 21  I, 49–​57. 22  I, 60. See also TAA, 72; PI, 146. 23  I, 60–​61. Similarly, in PI, 127, Anscombe offers ‘Fiat p!’ and ‘I’ll make p true’ as interchangeable formulations of the conclusion of a practical syllogism. See also TAA, 72, 74. 24  McDowell, ‘Intention in Action’, 422–​423. 25  McDowell, ‘Intention in Action’, 431n9. 26  The descriptivist might respond that theoretical reasoning addresses the question what to believe, yet no one thinks that the conclusion of a theoretical inference must be normative. But the general form of the conclusion of a theoretical inference is a judgement that p; only in special cases will p take the form ‘I should believe that q’. The two forms are not equivalent: to conclude that q is not ipso facto to conclude that one should believe that q. It is intelligible, even if mistaken, to think that although q is true, one should not believe it. Strictly speaking, then, theoretical reasoning addresses the question what is true, not the question what it would be good to believe.

Intention in Action    39 type is to be done—​with intentionally doing an action of that type. Marcus then points out that, normally, one knows without observation what one is judging. Thus, he can accommodate Anscombe’s claim that, normally, one knows without observation what one is doing intentionally.27 The problem is that Marcus has changed the subject. The challenge is not to show how the normativist can accommodate the claim that, normally, one knows without observation what one is doing intentionally. Rather, the challenge is to show how the normativist can accommodate the identification of this knowledge with Anscombean practical knowledge. Marcus dodges the challenge by using the term ‘practical knowledge’ to refer to knowledge of a judgement reached by practical reasoning rather than the knowledge at which such a judgement aims (hence the title of his paper: ‘Practical Knowledge as Knowledge of a Normative Judgment’). When Anscombe speaks of ‘practical knowledge’, however, she presumably has in mind the knowledge at which a judgement reached by practical reasoning aims. In any case, when Anscombe speaks of an agent’s practical knowledge that she is ϕ-​ ing, she cannot have in mind knowledge of a judgement that ϕ-​ing is to be done. For the fact that one is judging that ϕ-​ing is to be done does not entail that one is actually ϕ-​ing—​unless, following Marcus, we identify judging that ϕ-​ing is to be done with intending to ϕ and insist that intending to ϕ entails being in the process of ϕ-​ing intentionally. But Anscombe denies that practical knowledge is knowledge only of ‘intention’ as opposed to ‘what happens’, and she rejects the ‘false avenue of escape’ that one is doing ‘in the intentional sense’ whatever one intends to be doing.28 Anscombe is right. When someone opening a window intentionally is asked what she is doing, she does not have to resort to observation to respond. As Anscombe puts it, the person does not have to say, ‘Let me see, what is this body bringing about? Ah yes! the opening of the window’.29 And this is not because the question is naturally interpreted as equivalent to ‘What are you intending to do?’ Everyone agrees that the agent can answer that question without resort to observation. But presumably it is ‘What are you actually getting done?’ that is being asked. Anscombe is right to point out that the agent does not have to resort to observation to answer this question either. Practical knowledge extends beyond intention to what actually happens. Practical judgement, then, is neither purely descriptive nor purely normative. The reason that descriptivists and normativists alike find support in Anscombe is that Anscombe is sensitive to the fact that practical judgement is chimeral: it is about both what one is doing and what to do. What form could the content of practical judgement take if it is chimerical? Although simply conjoining descriptivist and normativist content—​‘I’m ϕ-​ing and ϕ-​ing is to be done’—​secures the entailment that practical judgement is chimerical, it fails to capture 27 

Marcus, ‘Practical Knowledge’, 323–​324; see also Marcus, Rational Causation, 79. I, 51–​52. Indeed, it is doubtful that Anscombe would say that one ‘knows’ what one judges at all. More likely she would say that one merely ‘can say’ what one judges (I, 13–​14, emphasis omitted). 29  I, 51. 28 

40   Charles F. Capps the full truth in normativism. Specifically, it fails to do justice to the normativist insight that practical reasoning concludes in a judgement about what to do, not a judgement about the conjunction of what one is doing and what to do. The most plausible way to capture the full truth in normativism as well as descriptivism, I submit, is to affirm that practical judgement is reasoned and singular. If we affirm that practical judgement is singular, then we can admit descriptive content into practical judgement without compromising the normativist insight. Instead of saying with Marcus that practical reasoning concludes in a judgement of the form ‘ϕ-​ing is to be done’, we can say that practical reasoning concludes in a judgement of the form ‘this ϕ-​ing is to be done’. Marcus’s formulation entails nothing about what is actually happening. The singular formulation, on the other hand, entails that there is a ϕ-​ing occurring to be the object of the demonstrative reference. So far, so good. But this is not yet enough to accommodate the full truth in descriptivism. It is easy to imagine cases in which an agent has practical knowledge of what she is doing under at least two descriptions, ‘ϕ-​ing’ and ‘ψ-​ing’, where ϕ-​ing is the agent’s means to ψ-​ing. For example, suppose that Alice raises her hand in order to vote in the affirmative. The descriptivist will rightly insist that Alice has practical knowledge both that she is raising her hand and that she is voting in the affirmative. Yet if the content of practical judgement is limited to the conclusion of the agent’s practical reasoning, then affirming that practical judgement is singular entitles us to say, at most, that Alice has practical knowledge that she is raising her hand. For practical reasoning proceeds from ends to means. Thus, the conclusion of an agent’s practical reasoning is that her immediate means is to be done; mention of what are the agent’s ends relative to this means is confined to the premises. Here is where the claim that practical judgement is reasoned comes in. If practical judgement is reasoned, then it is an endorsement not of the conclusion alone of a practical syllogism, but of the conclusion on the basis of the premises. Now, the premises of Alice’s practical reasoning take her from the premise that voting in the affirmative is to be done to the conclusion that the particular action that she is performing is to be done by explaining how this action constitutes voting in the affirmative. Therefore, if practical judgement is reasoned, then the content of Alice’s practical judgement is something like ‘this action is to be done because it constitutes raising my hand and thus voting in the affirmative, and voting in the affirmative is to be done’.30 Now we can accommodate the

30 

A complete account of the content of practical judgement would also include any ‘backward-​ looking’ reason why the agent treats her ultimate objective as to be done. Here, I leave backward-​ looking reasons aside because an agent’s backward-​looking reason for pursuing her objective does not correspond to a further intention with which she is acting. See Roger Teichmann, The Philosophy of Elizabeth Anscombe (Oxford: Oxford University Press, 2008), 31–​33; Anselm Winfried Müller, ‘Backward-​Looking Rationality and the Unity of Practical Reason’, in Essays on Anscombe’s Intention, ed. Anton Ford, Jennifer Hornsby, and Frederick Stoutland (Cambridge, MA: Harvard University Press, 2011), 259–​261; I, 21–​23.

Intention in Action    41 full truth in descriptivism as well as normativism: Alice’s practical reasoning issues in a judgement both about what is to be done and about what she is doing.

2.2.1.2  Analogy to Visual Knowledge of Particulars My second argument for the claim that practical judgement is singular proceeds as follows. Suppose that Bob is learning a piano piece which requires him to depress Middle C twice, a sixteenth of a beat apart, with his right thumb. Unbeknownst to Bob, a neuroscientist has installed an implant in his nervous system that interferes with the efferent nerves in his right arm. The implant terminates the second impulse before it reaches Bob’s hand, and it slows the first impulse such that Bob’s thumb depresses Middle C precisely one sixteenth of a beat later than intended. The result is that although Bob’s attempt to play the second note at t2 fails altogether, his attempt to play the first note at t1 results in a depression of Middle C at t2 that is qualitatively indistinguishable from a successful attempt to play the second note. It is difficult to see how someone who thinks that practical judgement is general rather than singular can account for this case. I take it that Bob’s practical judgement that he is playing the second note is false. But it is difficult to see how this judgement could be both general and false. For example, one might propose that the content of the judgement is something like ‘There is an event E such that E is a depression of this key, E occurs at t2, and I am the agent of E’. But there is such an event: the action that Bob intended to be a depression of Middle C at t1. This event is a depression of Middle C, it occurs at t2, and Bob is the agent of it.31 Therefore, if the content of Bob’s practical judgement were the existentially quantified proposition that there is an event E such that . . . then that judgement would be true. Philosophers of perception have used analogous examples to motivate the view that visual judgement is singular.32 According to the standard version of this view, visual judgement is singular in virtue of employing indexical concepts, such that the content of the judgement of one who, say, sees a sleeping cat is ‘[t]‌hat cat is asleep’ rather than ‘a cat that is in a certain region . . . is asleep’.33 I believe that we should adopt an analogous account of practical judgement. At least in paradigm cases of physical actions, an agent’s practical judgement latches onto the particular action that she is performing by means of indexical concepts, just as an observer’s visual judgement latches onto what she is observing by means of indexical concepts. 31  We could even modify the case so that Bob’s attempt to play the second note causes the implant not to terminate the first impulse altogether, in case someone proposed to build it into the content of the judgement that yields practical knowledge that that very judgement is the cause of an action with the properties that the judgement identifies. See, e.g., Gilbert Harman, ‘Practical Reasoning’, Review of Metaphysics 29, no. 3 (March 1976): 445. 32  See, e.g., Matthew Soteriou, ‘The Particularity of Visual Perception’, European Journal of Philosophy 8, no. 2 (August 2000): 179–​181. 33  John McDowell, ‘The Content of Perceptual Experience’, Philosophical Quarterly 44, no. 175 (April 1994): 192; see also Soteriou, ‘Visual Perception’, 184.

42   Charles F. Capps

2.2.1.3  Acting on Grounds upon the Instant Finally, here is my second argument for the claim that practical judgement is reasoned. The principal source of resistance to this claim, I think, is a conception of reasoning, including practical reasoning, as a mental process in which one ‘step[s]‌’34 from premise to premise and finally to the conclusion, endorsing each in turn in a discrete judgement. No doubt the convention of representing an inference graphically, premise by premise, with the conclusion on its own line at the bottom, aids and abets this conception. But the conception is, in any case, wrong. That it is wrong as a conception of practical reasoning is evident from the fact that, as Anscombe puts it, ‘one can act on grounds upon the instant’35—​that is, act with intentions, without deliberating first. When an agent acts on grounds upon the instant, she endorses a practical syllogism not in a process but all at once. But if endorsing a practical syllogism happens all at once, then presumably it consists in a single judgement encompassing both the premises and the conclusion. At least when an agent acts on grounds upon the instant, then, her practical judgement is reasoned. And there is no reason to think that matters are different in cases in which the agent deliberated before acting.36

2.2.2  An Account of Intention in Action I can now present an account of intention in action in terms of practical reasoning that is compatible with IW Priority. If practical judgement is reasoned, then it has the right content to ground intention in action. And if practical judgement is singular, then it does not precede the agent’s action; hence, the intentions that it grounds are in the first instance intentions in action rather than intentions for the future. The account of intention in action that I propose on behalf of the defender of IW Priority, then, is (IW)   An agent performs V with the intention of ϕ-​ing in virtue of representing V in her practical judgment as to be done because it (partially or wholly) constitutes ϕ-​ing.

On this view, Alice acts with the intentions of raising her hand and voting in the affirmative in virtue of representing her action in her practical judgement as to be done because it (wholly) constitutes raising her hand and thus voting in the affirmative.37 34 

PI, 132. PI, 113. See also Candace Vogler, ‘Anscombe on Practical Inference’, in Varieties of Practical Reasoning, ed. Elijah Millgram (Cambridge, MA: MIT University Press, 2001), 447–​449. 36  As Anscombe puts it, ‘choice of means is [always] choice of them as means. Therefore . . . the decisions we make must contain willings of ends’ (PT, 154, emphasis in original). 37  Note that (IW), although an analysis of intention in action, is not an analysis of intentional action (here by ‘intentional action’ I mean ‘the kind of action that bears intentions’ rather than ‘doing something intentionally’; see note 2); nor is it an analysis of practical judgement. (IW) does not entail that something is an intentional action in virtue of being an instance of ‘action’ in some more generic sense that is the object of a practical judgement. For an argument against the possibility of such an analysis, see Anton Ford, ‘Action and Generality’, in Essays on Anscombe’s Intention, ed. Anton Ford, Jennifer Hornsby, 35 

Intention in Action    43

2.2.3  Intended Consequences Consider the following objection to (IW). According to (IW), every object of an intention in action is another action ϕ-​ing such that, in her practical judgement, the agent represents her action as partially or wholly constituting ϕ-​ing. But not all intentions in action are like this. Sometimes the object of an intention in action is a consequence. For example, Alice might pass Bob the hammer with the intention that Bob pounds in a nail; in this case, the object of Alice’s intention—​Bob’s pounding in a nail—​is not something that Alice represents her own action as even partially constituting. Therefore, (IW) is false. In response, the defender of (IW) can insist that in cases in which the object of an intention in action seems to be a consequence, it is really the action of bringing about that consequence. For an action to constitute ‘bringing it about’ that p, as I am using the phrase, is just for the action to be metaphysically explanatory of the obtaining of p. The explanation for the fact that p that the action provides may be only partial. For example, the fact that Alice passed the hammer to Bob provides only a partial explanation for the fact that Bob pounded in the nail; it was up to Bob what to do with the hammer once Alice passed it to him. Still, the fact that Alice passed the hammer to Bob explains why Bob was equipped to pound in the nail in the first place. This is sufficient for Alice’s passing the hammer to Bob to constitute ‘bringing it about’ that Bob pounds in the nail in the sense that I have stipulated. I submit that in any case in which the object of an agent’s intention in action seems to be a consequence, the agent represents her action in her practical judgement as to be done because it partially or wholly constitutes ‘bringing about’ the consequence in this sense. While offering this response, the defender of (IW) can—​and, I think, should—​ concede an element of truth in the objection. It is true that when an agent acts with the intention of bringing about a consequence, the consequence is something that the agent’s action does not constitute, partially or wholly. This is true even if the consequence consists in the agent herself doing something. For example, suppose that Alice signs up for a race, with the intention of bringing it about that she runs in it. Although signing up partially constitutes bringing it about that she runs in the race, it does not even partially constitute running in the race. Furthermore, it is true that, in ordinary English, the grammatical object of ‘intention with which’ is often a phrase that refers to and Frederick Stoutland (Cambridge, MA: Harvard University Press, 2011), 76–​104. Nor does (IW) entail that something is a practical judgement in virtue of being a ‘judgement’ in some more generic sense that has content of a certain form. For an argument against the possibility of such an analysis with respect to practical reasoning generally, see Anselm Winfried Müller, ‘How Theoretical Is Practical Reason?’, in Intention and Intentionality: Essays in Honour of G. E. M. Anscombe, ed. Cora Diamond and Jenny Teichman (Ithaca, NY: Cornell University Press, 1979), 91–​108. Finally, (IW) certainly is not an analysis of both intentional action and practical judgement. It does not entail that something is an intentional action in virtue of being an action (in a generic sense) that is the object of a singular judgement (in a generic sense) that has content of a certain form. For a counterexample to such an analysis, see Wilson, Intentionality, 220.

44   Charles F. Capps a consequence that the agent is bringing about rather than to the agent’s bringing about of that consequence. This phrase might be an independent clause introduced by ‘that’ (‘Alice passed the hammer with the intention that Bob pounds in the nail’) or, if the consequence consists in the agent herself doing something, a gerund phrase introduced by ‘of ’ (‘Alice signed up with the intention of running in the race’). The defender of (IW) is free to accommodate ordinary language by extending the notion of intention in action to cover, as secondary cases, intentions of consequences that the agent does not represent her action as even partially constituting. The defender of (IW) merely has to insist that the primary cases of intention in action are those whose objects are actions that the agent represents what she is doing as partially or wholly constituting.

2.3  Intentionally Acting Let us turn to intentionally acting. Although I will defend an account of intentionally acting in terms of intention in action, I do not want to say that an agent ϕ-​s intentionally simply in virtue of acting with the intention of ϕ-​ing. Even those who claim that acting with the intention of ϕ-​ing entails being in the process of ϕ-​ing intentionally should not say this. For even they should agree that the entailment does not hold in the perfective aspect. No one thinks that having acted with the intention of ϕ-​ing entails having ϕ-​ed intentionally. For example, consider a pedestrian who steps into the street with the intention of crossing it but is run over by a bus before she reaches the other side. Although she was crossing the street intentionally, it is not the case that she crossed the street intentionally. Yet by the time she had completed her first step (if not earlier), she had acted with the intention of crossing the street. Therefore, it is not the case that an agent ϕ-​s intentionally simply in virtue of acting with the intention of ϕ-​ing. Nor will it do to say that an agent ϕ-​s intentionally in virtue of acting with the intention of ϕ-​ing and actually ϕ-​ing. Alfred Mele imagines a philosopher at a conference who swipes at his glass of water in order to knock it over, with the intention of flustering his commentator. Although he misses entirely, his sudden movement so startles his commentator that she becomes flustered and loses her train of thought. The philosopher acts with the intention of flustering his commentator and actually flusters his commentator. Yet the philosopher does not fluster his commentator intentionally.38 The defender of IW Priority can handle both the pedestrian and Mele’s philosopher by insisting, with Anscombe, that it is only in virtue of an agent’s practical knowledge that she is ϕ-​ing that ‘what happens’ counts as ‘the execution of an intention’ to ϕ.39 Instead of 38 

Alfred R. Mele, Springs of Action: Understanding Intentional Behavior (New York: Oxford University Press, 1992), 182. As Anscombe puts it, it is only ‘in ordinary circumstances’ that, if someone is doing what she is acting with the intention to do, then she is doing it intentionally; ‘[t]‌he qualification is necessary because an intended effect just occasionally comes about by accident’ (I, 39). 39  I, 87–​88.

Intention in Action    45 saying that an agent ϕ-​s intentionally in virtue of the fact that she acts with the intention of ϕ-​ing and actually ϕ-​s, the defender of IW Priority can say that (IA)   An agent ϕ-​s intentionally in virtue of the fact that she acts with the intention of ϕ-​ing and the practical judgement underlying this intention gives her practical knowledge that she ϕ-​s,

where the simple-​present placeholders ‘ϕ-​s intentionally’, ‘acts’, etc. are linked in tense and aspect. Thus, according to (IA), being in the process of ϕ-​ing intentionally requires being in the process of acting with the intention of ϕ-​ing and having practical knowledge that one is in the process of ϕ-​ing; likewise, having ϕ-​ed intentionally requires having acted with the intention of ϕ-​ing and having had practical knowledge that one ϕ-​ed.40 (IA) implies, correctly, that the pedestrian was crossing the street intentionally but never crossed it intentionally. When the pedestrian was stepping into the street, the practical judgement underlying her intention to cross the street gave her knowledge that her action constituted one part of crossing the street, namely, stepping into the street. Thus, she had practical knowledge that she was in the process of crossing the street. Thanks to the bus, however, she never crossed the street; a fortiori, she never had practical knowledge that she crossed the street. According to (IA), then, the pedestrian was crossing the street intentionally but never crossed it intentionally. (IA) also implies, correctly, that Mele’s philosopher neither was flustering nor did fluster his commentator intentionally. Underlying the philosopher’s intention to fluster his commentator was his practical judgement that his action constituted flustering his commentator, which was based on the false premise that his action constituted knocking over the glass. Therefore, any belief that he was flustering or did fluster his commentator that the philosopher acquired through the exercise of his practical judgement was true only by luck and did not constitute knowledge. According to (IA), then, the philosopher neither was flustering nor did fluster his commentator intentionally. While maintaining that (IA) captures the primary cases of intentionally acting, the defender of (IA) can—​and, I think, should—​admit, as borderline or secondary cases, cases in which the agent harbours some doubt that she is actually in the process of doing what she is acting with the intention of doing. Davidson offers the example of a man who tries to produce ten carbon copies at once and succeeds.41 The defender of (IA) can allow the man to be described as ‘intentionally producing ten carbon copies’. She merely has to insist that the primary cases of intentionally acting are those in which the practical judgement underlying the agent’s intention gives her knowledge of what she is doing; hence, the more serious the agent’s doubts that she is actually in the process of 40  Although practical knowledge of what one has done is an interesting topic, I do not have the space to explore it here. See Matthias Haase, ‘Knowing What I Have Done’, Manuscrito 41, no. 4 (October–​ December 2018): 195–​253. 41  Davidson, ‘Intending’, 92.

46   Charles F. Capps doing what she is acting with the intention of doing, the more of a stretch it is to describe her as doing it ‘intentionally’.42

2.4  Intention for the Future Finally, consider intention for the future. In section 2.4.1, I will consider the strategy of exploiting the elasticity of the progressive aspect to explain intention for the future in terms of intentionally acting. Although this strategy is compatible with IW Priority, I will identify a problem with it in section 2.4.2. In section 2.4.3, I will present an alternative account of intention for the future in terms of intention in action that avoids this problem. Finally, in section 2.4.4, I will respond to an objection.

2.4.1  The Progressive Aspect One way for the defender of IW Priority to explain intention for the future in terms of intention in action is to adapt a strategy popular amongst defenders of IA Priority for explaining intention for the future in terms of intentionally acting.43 The key to this strategy is the elasticity of the progressive aspect. Imagine some house builders on a lunch break. It seems that they are building a house, even though no progress is presently being made. But if they are building a house during a hiatus in the middle of the process, then why deny that they are building a house during a ‘hiatus’ at the beginning of the process, from whenever they acquired the intention to build a house to whenever they finally broke ground?44 Perhaps, then, (IF)   An agent intends to ϕ in the future in virtue of the fact that she is ϕ-​ing intentionally but has not yet made substantial progress toward having ϕ-​ed.

(IF) suits the needs of the defender of IW Priority who endorses (IA). If intention for the future is to be explained in terms of intentionally acting, and intentionally acting is to be explained not in terms of intention for the future but in terms of intention in action, then intention for the future is ultimately to be explained in terms of intention in action.

42 

(IA) also entails that the primary cases of intentionally acting are cases in which the agent is acting with the intention of doing what she is doing intentionally. For cases that challenge this entailment, see Bratman, Intention, Plans, and Practical Reason, 113–​115; Matthew Hanser, ‘Intention and Teleology’, Mind 107, no. 426 (April 1998): 388–​398. Although I believe that the defender of (IA) can account for these cases, I do not have the space to address them here. 43  See, e.g., Marcus, Rational Causation, 88–​90; Thompson, Life and Action, 138–​142; Moran and Stone, ‘Expression of Intention’, 142–​148. 44  See Thompson, Life and Action, 140–​141.

Intention in Action    47 In fact, however, the defender of IW Priority can improve on (IF). To account for cases in which an agent intends to ϕ in the future but ‘manifestly is not’45 in the process of ϕ-​ing intentionally, the defender of IW Priority can replace (IF) with (IF')   An agent intends to ϕ in the future in virtue of the fact that she is acting with the intention of ϕ-​ing but has not yet made substantial progress toward having ϕ-​ed.

For example, suppose that Ahab sets sail with the intention of killing Moby Dick. Unbeknownst to Ahab, the white whale that he is tracking is not Moby Dick; Moby Dick was blown to smithereens last year when a nuclear bomb was tested over the ocean. I take it that (IF') provides a more plausible account than (IF) of Ahab’s intention to kill Moby Dick in the future because Ahab is acting with the intention of killing Moby Dick but is not in the process of killing Moby Dick intentionally.

2.4.2  ‘Broad’ and ‘Narrow’ Uses of the Progressive The problem with (IF) and (IF') is that the distinction between ‘broad’ and ‘narrow’ uses of the progressive aspect tracks a distinction in reality.46 There is a sense in which Alice ‘was running in the race’ as soon as she acquired the intention to do so. But there is also a sense in which obviously she was not; it is not as if she should be disqualified for an egregious false start because she signed up a month in advance. Furthermore, the claim that Alice ‘was running in the race’ in the latter, narrow sense did not gradually become more plausible as the start of the race neared. On the contrary, it remained clearly false right up until the gun fired. Admittedly, the borderline between when Alice was and was not racing, within the narrow use of the progressive, admits of some vagueness. (Was it when the gun fired or a split-​second later when Alice started moving?) But there are moments clearly on one side or the other. As a result, the defender of (IF) faces a dilemma. If she compresses the action of intentionally running in the race to fit the bounds of the narrow use of the progressive, then she cannot stretch intentionally acting to cover intention for the future. For Alice was not yet running in the race, within the narrow use of the progressive, when she signed up for the race, yet she did intend to run in the race in the future. But if the defender of (IF) stretches the action of intentionally running in the race to fit the bounds of the broad use of the progressive, then she lacks the resources to explain the transition that occurs when the race begins. Unfortunately, the defender of (IF') faces a similar dilemma. According to (IW), one performs an action with the intention of ϕ-​ing only if one represents that action in one’s 45 

I, 39. See Anthony Galton, The Logic of Aspect (Cambridge: Cambridge University Press, 1984), 89–​149. On the importance of this distinction in the criminal law, see Sarah Paul, ‘Embarking on a Crime’, Social, Political, & Legal Philosophy 3 (January 2014): 101–​124. 46 

48   Charles F. Capps practical judgement as to be done because it partially or wholly constitutes ϕ-​ing. If the defender of (IF') compresses Alice’s practical representation of herself as running in the race to fit the bounds of the narrow use of the progressive, then she cannot stretch acting with an intention to cover intention for the future. For when Alice was signing up, she did not represent any action that she was then performing as even partially constituting running in the race, within the narrow use of the progressive, yet she did intend to run in the race in the future. But if the defender of (IF') stretches Alice’s practical representation of herself as running in the race to fit the bounds of the broad use of the progressive, then she lacks the resources to explain the transition that occurs when the race begins.

2.4.3  An Alternative Account I think that the defender of IW Priority can do better than (IF'). To accommodate the distinction between the broad and narrow uses of the progressive, she should concede that an action performed with the intention of ϕ-​ing does not begin until the agent, in her practical judgement, represents herself as ϕ-​ing, within the narrow use of the progressive. Then, to account for intention for the future, the defender of IW Priority should say that (IF′)   An agent intends to ϕ in the future in virtue of acting with the intention of bringing it about that she ϕ-​s in the future.

On this view, when she signed up for the race, Alice intended to run in the race in the future in virtue the fact that, in her practical judgement, she represented something that she was then doing as partially constituting not running in the race but bringing it about that she runs in the race.

2.4.4  ‘Pure’ Intention Consider the following objection to (IF’). It is true that acting with the intention of bringing it about that one ϕ-​s covers some stretches of intending to ϕ in the future. For example, when Alice signed up for the race, she did so with the intention of bringing it about that she runs in the race in the future. As long as we are using the progressive aspect narrowly, however, acting with the intention of bringing it about that one ϕ-​s cannot cover the entirety of intending to ϕ in the future. For there are stretches of what Davidson calls ‘pure intention’: periods during which one intends to ϕ in the future but is not doing anything at all, within the narrow use of the progressive, with the intention of bringing it about that one ϕ-​s.47 For example, immediately after signing up, Alice was

47 

Davidson, ‘Intending’, 88.

Intention in Action    49 not doing anything with the intention of bringing it about that she runs in the race. But she still intended to run in the race in the future. In response to this objection, the defender of IW Priority can deny that there is any such thing as a truly ‘pure’ intention. If an agent intends to ϕ in the future, then there must be something that she is doing with the intention of bringing it about that she ϕ-​s. It may be as subtle as waiting for the opportune moment, keeping a lookout for opportunities to ϕ, deliberating about how to go about ϕ-​ing, or even deliberating about how to do something else in a way that will not get in the way of ϕ-​ing.48 For example, after she signed up for the race, what Alice was doing with the intention of bringing it about that she runs in the race in the future was waiting until the designated start time. Although some might doubt that waiting is the sort of thing that one can do with an intention, I think that this scepticism is misplaced. It seems to me that Alice’s waiting to start running until the race begins is a clear case of acting with an intention. After all, waiting until the right moment is no less crucial a part of bringing it about that she runs in the race than signing up; starting too soon will result in disqualification.49 Or, consider the famous marshmallow experiment: surely, the children who managed to wait for fifteen minutes before eating their first marshmallow did so with the intention of bringing it about that they receive a second.50 In reply, the sceptic might cite (IW) for the proposition that waiting can be done with an intention only if the agent takes waiting to contribute to bringing it about that she does what she intends to do. Then the sceptic could argue that it is possible to have an intention to do something in the future even if one could just as well do it now. For example, Bob might intend to go shopping in an hour even though the store is open now. Because waiting to go shopping in no way contributes to bringing it about that he goes

48 

See Frost, ‘Direction of Fit’, 479–​480. Note that, on this account, Alice does not treat waiting as to be done because it partially constitutes an action that includes running in the race as another part. Alice treats waiting as to be done because it partially constitutes bringing it about that she runs the race, and (as we noted in section 2.2.3) the bringing about of a consequence is not even partially constituted by the consequence, even when the consequence is something else that the agent who brings it about does. Thus, it is not as if Alice treats running in the race as to be done on the ground that it too partially constitutes bringing it about that she runs in the race. So although I agree with much of what Kim Frost says about intention for the future, I think that he errs to the extent that, following Thompson, he proposes to reduce ‘preparatory’ means to ‘compositional’ means, arguing that ‘one action is [only ever] rationalized in terms of another’ that ‘encompasses’ it (Frost, ‘Direction of Fit’, 479–​480). 50  Rarely, of course, does an agent deliberate all the way to waiting as a means to achieving her end. For example, if Alice decided in advance to vote in the affirmative, then presumably she did not proceed in deliberation to the further decision to wait until the vote was called. All this shows, however, is that deliberation typically concludes when the agent settles on a means that she can implement without further thinking about how to implement it. The practical reasoning embodied in action is richer. As Anscombe observes, it is ‘very rare for a person to go through all the steps’ of a practical syllogism in deliberation; hence, if the practical syllogism ‘were supposed to describe actual mental processes, it would be in general quite absurd. The interest of the account is that it describes an order which is there whenever actions are done with intentions’ (I, 79–​80). 49 

50   Charles F. Capps shopping, it seems that Bob cannot be waiting with the intention of bringing it about that he goes shopping. His intention to go shopping thus seems to be ‘pure’. It is not so clear, however, that one really can have an intention to do something in the future if, even after accounting for one’s other ends, one could just as well do it now. If Bob intends to go shopping in an hour, then it is presumably because he has something else to do in the meantime—​say, take a nap. Because he cannot nap and shop simultaneously, he will have to wait to go shopping. Often, the only way for an agent to bring it about that she does something is to wait for a moment when her other ends do not screen out doing it as a possibility. The point is not specific to waiting. Suppose that Carla wants to practice the saxophone but knows that playing inside at this hour would disturb her housemates. So she carries her saxophone to the garage and practices there. In abstraction from Carla’s other ends, walking to the garage in no way contributes to bringing it about that she practices the saxophone. Given that her end of not disturbing her housemates screens out practicing inside as a possibility, however, walking to the garage does contribute to bringing it about that she practices the saxophone. It is no mystery, then, that walking to the garage is something that Carla does with the intention of practicing the saxophone—​ or rather, cleaving to the primary sense of ‘intention with which’ captured by (IW), something that Carla does with the intention of bringing it about that she practices the saxophone. I conclude that (IF’) is a plausible account of intention for the future. It is easy to overlook the fact that waiting is something that we often do intentionally because we are skilled at directing our attention elsewhere while waiting. This skill is amplified, enormously, by tools and conventions. Alarms, to-​do lists, calendars, and other devices for keeping track of time allow us to divert our attention from waiting, making it possible both to remain poised for a very long time to do something (attend a wedding next year, say) and to remain poised at once to do a very large number of things. Nonetheless, such episodes of waiting are things that we do with intentions—​rather like putting one foot in front of another is something that one does with the intention of walking, even if one’s attention is absorbed in a conversation. Intention for the future is ultimately grounded in intention in action.

2.5  Conclusion Anscombe’s tripartite division in the topic of intention set the agenda for late twentieth-​ and early twenty-​first-​century philosophy of action. The standard approach to unifying Anscombe’s three divisions is to treat intention for the future as fundamental. Amongst those who dissent from this approach, most attempt to unify Anscombe’s three divisions by treating intentionally acting as fundamental. Yet there is evidence that Anscombe rejected both approaches. This raises the possibility that her own approach, however implicit, was to treat intention in action as fundamental. I have attempted to show that

Intention in Action    51 such an approach is, in any case, plausible. At the very least, it deserves more serious consideration than philosophers have given it to date.

Bibliography Anscombe, G. E. M. ‘The Causation of Action’. In Human Life, Action and Ethics: Essays by G. E. M. Anscombe, edited by Mary Geach and Luke Gormally, 89–​108. Exeter: Imprint Academic, 2005. Anscombe, G. E. M. Intention. 2nd edition. Cambridge, MA: Harvard University Press, 1963. Anscombe, G. E. M. ‘Practical Inference’. In Human Life, Action and Ethics: Essays by G. E. M. Anscombe, edited by Mary Geach and Luke Gormally, 109–​147. Exeter: Imprint Academic, 2005. Anscombe, G. E. M. ‘Practical Truth’. In Geach and Gormally, Human Life, Action and Ethics: Essays by G. E. M. Anscombe, edited by Mary Geach and Luke Gormally, 149–​158. Exeter: Imprint Academic, 2005. Anscombe, G. E. M. ‘Thought and Action in Aristotle: What Is Practical Truth?’ In Collected Papers of G. E. M. Anscombe, Volume 1: From Parmenides to Wittgenstein, 66–​77. Oxford: Blackwell, 1981. Anscombe, G. E. M. ‘Under a Description’. Noûs 13, no. 2 (May 1979): 219–​233. Bratman, Michael E. Intention, Plans, and Practical Reason. Cambridge, MA: Harvard University Press, 1987. Broome, John. ‘Practical Reasoning’. In Reason and Nature: Essays on the Theory of Rationality, edited by José Luis Bermúdez and Alan Millar, 85–​111. Oxford: Clarendon Press, 2002. Davidson, Donald. ‘Intending’. In Essays on Actions and Events, 83–​102. Oxford: Oxford University Press, 2001. Donagan, Alan. Choice: The Essential Element in Human Action. London: Routledge, 1987. Enç, Bertram. How We Act: Causes, Reasons, and Intentions. Oxford: Oxford University Press, 2003. Fink, Julian. ‘What Is (Correct) Practical Reasoning?’ Acta Analytica 28, no. 4 (December 2013): 471–​482. Ford, Anton. ‘Action and Generality’. In Essays on Anscombe’s Intention, edited by Anton Ford, Jennifer Hornsby, and Frederick Stoutland, 76–​104. Cambridge, MA: Harvard University Press, 2011. Frost, Kim. ‘On the Very Idea of Direction of Fit’. Philosophical Review 123, no. 4 (October 2014): 429–​484. Galton, Anthony. The Logic of Aspect. Cambridge: Cambridge University Press, 1984. Ginet, Carl. On Action. New York: Cambridge University Press, 1990. Haase, Matthias. ‘Knowing What I Have Done’. Manuscrito 41, no. 4 (October–​December 2018): 197–​253. Hanser, Matthew. ‘Intention and Teleology’. Mind 107, no. 426 (April 1998): 381–​401. Harman, Gilbert. ‘Practical Reasoning’. Review of Metaphysics 29, no. 3 (March 1976): 431–​463. Marcus, Eric. ‘Practical Knowledge as Knowledge of a Normative Judgment’. Manuscrito 41, no. 4 (October–​December 2018): 319–​347. Marcus, Eric. Rational Causation. Cambridge, MA: Harvard University Press, 2012. McDowell, John. ‘The Content of Perceptual Experience’. Philosophical Quarterly 44, no. 175 (April 1994): 190–​205.

52   Charles F. Capps McDowell, John. ‘What Is the Content of an Intention in Action?’ Ratio 23, no. 4 (December 2010): 415–​432. Mele, Alfred R. Springs of Action: Understanding Intentional Behavior. New York: Oxford University Press, 1992. Moran, Richard, and Martin J. Stone. ‘Anscombe on Expression of Intention’. In New Essays on the Explanation of Action, edited by Constantine Sandis, 132–​168. Hampshire: Palgrave Macmillan, 2008. Müller, Anselm Winfried. ‘Backward-​Looking Rationality and the Unity of Practical Reason’. In Essays on Anscombe’s Intention, edited by Anton Ford, Jennifer Hornsby, and Frederick Stoutland, 242–​269. Cambridge, MA: Harvard University Press, 2011. Müller, Anselm Winfried. ‘How Theoretical Is Practical Reason?’ In Intention and Intentionality: Essays in Honour of G. E. M. Anscombe, edited by Cora Diamond and Jenny Teichman, 91–​108. Ithaca, NY: Cornell University Press, 1979. Müller, Anselm Winfried. ‘Is Practical Truth a Chimera? Questions for Anscombe’. Lecture delivered at ‘Practical Truth: Reflections on the Aristotelian Tradition’, University of South Carolina, Columbia, April 2017. https://​www.yout​ube.com/​watch?v=​5nNH​HiLB​sZM. Paul, Sarah. ‘The Conclusion of Practical Reasoning: The Shadow between Idea and Act’. Canadian Journal of Philosophy 43, no. 3 (September 2013): 287–​302. Paul, Sarah. ‘Embarking on a Crime’. Social, Political, & Legal Philosophy 3 (January 2014): 101–​124. Raz, Joseph. From Normativity to Responsibility. Oxford: Oxford University Press, 2011. Setiya, Kieran. ‘Intention’. In The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta. Fall 2018. https://​plato.stanf​ord.edu/​entr​ies/​intent​ion/​. Soteriou, Matthew. ‘The Particularity of Visual Perception’. European Journal of Philosophy 8, no. 2 (August 2000): 173–​189. Teichmann, Roger. The Philosophy of Elizabeth Anscombe. Oxford: Oxford University Press, 2008. Thompson, Michael. Life and Action: Elementary Structures of Practice and Practical Thought. Cambridge, MA: Harvard University Press, 2008. Vogler, Candace. ‘Anscombe on Practical Inference’. In Varieties of Practical Reasoning, edited by Elijah Millgram, 437–​464. Cambridge, MA: MIT University Press, 2001. Wilson, George M. The Intentionality of Human Action. Stanford, CA: Stanford University Press, 1989.

Chapter 3

Intention, Know l e d g e , and Resp onsi bi l i t y Rémi Clot-​G oudard

3.1  Introduction Responsibility, as are many other related concepts (innocence and guilt, praise and blame, reward and punishment), is a fundamental topic for ethics, philosophy of law, and philosophy of religion. The core conceptual question, however, pertains to the philosophy of action (which Anscombe did much to bring to the fore). It is a question of elucidating the conditions under which it is sensible and right to impute some particular action to an agent and to hold him responsible for it and for some or all of its consequences. Obviously, such an investigation must also consider the conditions under which an agent may be excused or exonerated from what is imputed to him, and the reasons or motives that may effectively discharge him partially or totally of guilt or even of having committed the alleged act.1 Aristotle famously held that we are first and foremost responsible for our voluntary actions, and that we should define ‘the voluntary’ (to hekousion) as ‘[1]‌that of which the origin is in oneself, [2] when one knows the particular factors that constitute the location of action [praxis]’.2 In contemporary debates, the stress has often been put on the first condition, understood as a condition of control: what type or degree of control must an agent have over his action to be rightly held responsible for it? This question gave rise to numerous and intense discussions about the possibility and reality of free will. The second part of Aristotle’s definition introduces what may be called an epistemic condition of responsibility: to be rightly held responsible for something he does and for what 1  See also John Austin, ‘A Plea for Excuses’, in Philosophical Papers, ed. J. O. Urmson and G. J. Warnock, 3rd ed. (Oxford: Oxford University Press, 1961), 175–​204. 2 Aristotle, Nichomachean Ethics, trans. Christopher Rowe (Oxford: Oxford University Press, 2002), III.1, 1111a22–​24, 125.

54   Rémi Clot-Goudard follows from it, the agent must know what he is doing. Whereas many philosophers have recently put emphasis on the importance of understanding this condition, few have perceived how relevant Anscombe’s work is to clarifying the point at issue. From the very beginning of her published work3 to late papers written after she retired, it was a constant concern of Anscombe to try to elucidate difficult questions about the extent to which human agents can be held responsible for what they do. As Rachael Wiseman convincingly put it, one cannot even expect properly to understand her major book Intention unless one can see it ‘as the product of the ethical debates in which Anscombe was engaged between 1956 and 1958’, through which she ‘came to realise that moral philosophy had lost sight of the distinctive use of the question ‘What is she doing?’ to mark out the class of intentional actions’, which ‘is essential to identifying the nature and quality of an act, a category without which moral philosophy cannot proceed’.4 What prompted Anscombe’s reflection was the diagnosis that the role and importance of our concept of intention in our common practice of identifying actions had been blurred in philosophical debates about responsibility, and that a grammatical clarification of it was sorely needed. In this chapter, I would like to suggest that the effort to elucidate the meaning of the epistemic condition of responsibility is a constant concern in Anscombe’s writings, from Intention to her later papers. As such, it may be taken as a common thread relating many different papers, which can then be read as a single many-​faceted effort to build a complex and subtle answer to a central question: To what degree is the knowledge an agent has of what he is doing necessary for our imputing some action to him, and hence for holding him responsible for it? In other words, what role does an agent’s knowledge play in the imputability of an action, and why? In what follows, I will focus on retracing the main features of Anscombe’s view about the epistemic condition of responsibility by establishing the following points: 1. Responsibility in the case of human agents has to be distinguished from mere physical causality; human agents are rational causes. Ascribing an action to a human agent implies taking his will into account. 2. Agent knowledge (in the sense of knowing what is happening insofar as one is doing it) is a necessary and sufficient condition for imputing intentional actions to someone; ignorance may be grounds for qualifying what an agent does as involuntary. 3. However, the agent’s ignorance of what he is doing under such-​and-​such description cannot always be put forward as a ground for exoneration: the epistemic condition of responsibility functions differently according to logical differences between types of action predicates; ignorance can be voluntary and hence blameworthy, when the agent could and should have known what he is ignorant of. 3 

See, e.g., G. E. M. Anscombe, ‘War and Murder’, in CP3, 51–​61; Mr Truman’s Degree, in CP3, 62–​7 1. Rachael Wiseman, ‘The Intended and Unintended Consequences of Intention’, American Catholic Philosophical Quarterly 90, no. 2 (Spring 2016): 2. 4 

Intention, Knowledge, and Responsibility    55 4. Elucidating this concept of voluntary ignorance brings to light a fundamental obligation to know which takes its meaning from and has its roots in Anscombe’s Thomist background.5

3.2  Responsibility and Causality A broken glass lies on the terrace floor: who, or what, is responsible for this? The playing child, a gust of wind, the dog that ran against the table leg? To wonder who or what is responsible here implies asking who or what has brought about such-​and-​ such state of affairs. Ascribing responsibility, in the widest sense of this word, involves identifying the origin of this state by going back to its cause—​after which one may take certain measures (lecturing the child, clearing the table, calming the dog to prevent further damage, etc.). But however natural this move may seem in the context of an investigation of what happened, the responsibility of human agents does not amount to the mere fact of their having causally contributed to the production of something. Any endeavour to analyse responsibility only in terms of mere physical causality without taking into account what the agent thought of his own action would fail, because causality in itself does not provide a satisfactory criterion for imputing actions and is not enough to account for certain fundamental distinctions we need when appraising someone’s responsibility. This will become apparent when we consider several questions that are of interest when such an appraisal is required: 1. To what extent is an agent implicated in what he does? This question raises the problem of the content of the agent’s will. I may be the one who caused the breaking of the glass on your terrace floor, but there are different possible cases here: I may have clumsily let it fall from my hand, say, or I may have intentionally dropped it in order to take my revenge for something you did (the glass is a favourite possession of yours). The resulting state of affairs is the same: your glass is broken, and I am the cause of it. However, there is an important distinction between the cases: the first is an involuntary act and the second an intentional act. Obviously, knowing that the broken glass is an effect of which I am the cause is not enough for you to know how to describe what I did. Physical causation alone cannot illuminate the agent’s will, because doing something does not consist only in being the cause of some state of affairs. 2. Who is the real source of action? Considering causal relations alone does not help us to account for differences between degrees of agency: it does not allow us to 5  Though fully relevant to the object of this paper, Anscombe’s views on the doctrine of ‘double effect’ are not considered here for reasons of length, but the reader will find a complete discussion of them in this volume’s c­ hapter 8 by Cyrille Michon.

56   Rémi Clot-Goudard justify imputations of responsibility to agents who did not take part in the physical realisation of a plan, that is, to the ‘brains’ behind a plan as well as to their subordinates. This is the case when, for instance, a Mafia godfather orders one of his hitmen to get rid of a nuisance: we would not accept that he be discharged from all guilt on the ground that he did not kill the victim with his own hands. In a quite different context, it was an analogous question which so exercised Anscombe in the debate concerning President Truman’s honorary degree at Oxford. Having protested against conferring such an honour on Mr. Truman by arguing that a man who had ordered the bombing of civilians as a war act had actually committed mass murder and should be treated accordingly, Anscombe later recalled what the censor of St Catherine’s said, namely that ‘you can’t make a man responsible just because “his is the signature at the foot of the order.” ’6 However, it would not be enough here to say that, after all, having signed the order implies being the cause of the bombing and that this relation of causation is what responsibility amounts to. If the only things to consider were relations of causation, on what grounds could we distinguish between the case where someone has been given a command that he executes (e.g., the hitman who is told by the godfather to push someone over a cliff and who acts accordingly) and the case when a natural cause makes us do something (e.g., a gust of wind makes me stumble as we are admiring the landscape at the top of a cliff, and in falling, I push you over)? Taking the physical causation point of view cannot allow us to track down the relation between an action and the order of which it is the execution. Actually, the very idea of responsibility would then be doomed to collapse and to dissolve in the network of causes and effects, as Anscombe suggested in one critical review of common moral ideas of her time:7 she mentions ‘a gentle, tolerant, and civilised idea of responsibility for things once they have been done’, typical of ‘the best liberal minds’, according to which ‘responsibility is causality . . . for to hold someone in good standing responsible for what he did is to ascribe the whole causality of it as an event to him’. But such an ascription is ‘unfair’, in this view, because an event is always a conjunction of a huge number of causal factors, and that eventually leads one to think that ‘an agent is himself the victim of causality’. 3. When can we blame the agent for not having done anything? We sometimes blame people for not having prevented some course of events, e.g., for having let the zucchini gratin burn in the oven. Here the agent is taken to be causally responsible for something, not in the positive sense that he actively contributed to spoiling the gratin but in the negative sense that, had he turned off the oven, the gratin would not have burned. Now, accepting that agents can be responsible for things in this negative sense raises a very difficult question. There is an indefinitely large number of things happening because of the fact that we refrain from acting or fail to act;

6  7 

Anscombe, TD, 66. G. E. M. Anscombe, ‘Does Oxford Moral Philosophy Corrupt the Youth?’, in GG1, 164–​165.

Intention, Knowledge, and Responsibility    57 should we be held responsible for all of these consequences? If there is a conceptual difference between omitting to do or refraining from doing on the one hand and not doing on the other hand, how is it to be accounted for?8 Obviously, taking relations of causation as our only criteria will not be enough. Human action cannot be accounted for only in terms of physical causality. Conflating responsibility and physical causality would actually amount to removing all relevant distinction between natural and rational agents. In this regard, it is significant that in order to appraise the responsibility of a human agent, we need to take into account what he could sincerely say of his own action, what he was up to, how he wanted to attain his goal. It is certainly true to say that human agents are causes—​provided that we add: they belong to a kind of cause which thinks what it’s doing and what it’s producing, this thinking being an essential part of the doing. The task for philosophy here is to elucidate this rational causality of human agents, agents able to act for reasons—​what we may call the Aristotelian question, because it finds an echo in the following passage of the Nichomachean Ethics (VI, 3, 1139b4–​6): ‘[D]‌oing well is an end, and desire is for the end. Hence decision is either intelligence qualified by desire or desire qualified by thought, and human beings are originators of this sort’ (my italics).9 It is to deal with this question that Anscombe proposes to revive Aquinas’s distinction between actus humanus and actus hominis.10 The actus hominis, or ‘act of a human being’, refers to any event in the description of which the agent figures as a mere subject of attribution, the consideration of his thoughts being in no way essential to the truth of the description. Classical examples are idly stroking one’s beard, falling over, etc. On the other hand, the actus humanus, or ‘human action’, is what is done by the agent ‘under the command of reason’. Whatever may be the correct understanding of this last expression, acknowledging the necessity of such a distinction is what lies behind the idea that responsibility comes in degrees—​not in the sense of degrees of proximity (or distance) of consequences from an imputed action, nor in the sense of a quantum or proportion of share in the causation of something, but in the sense of what we may call the degree of formal causality, this being a function of the degree to which the agent has engaged himself in the rational conception of what he did. Hence the three degrees of responsibility distinguished by Anscombe in MME:11 1. To be cause of something, regardless of any use of rational powers. 2. To be cause of something, this requiring the use of rational powers: rational agents are those who can be asked to account for their actions and deeds (logon didonai),

8  Cf. G. E. M. Anscombe, ‘Murder and the Morality of Euthanasia’, in GG1, 261–​277. This distinction, its meaning and importance for ethics, is at the core of debates concerning ‘negative responsibility’ (see, e.g., works by Jonathan Glover, Peter Singer, Bernard Williams). 9 Aristotle, Nichomachean Ethics, 178. 10  Cf. G. E. M. Anscombe, ‘Action, Intention and “Double Effect,” ’ in GG1, 207–​208. 11  Cf. Anscombe, MME, 261–​262.

58   Rémi Clot-Goudard who can answer for what they did (or did not); responsibility is here equivalent to accountability. 3. To have intentionally done something illicit: here, responsibility amounts to guilt; what is at issue then is the content of the agent’s will, which gives its nature and character to the act. Where might the temptation to reduce responsibility to causation (and then to drop any consideration of intention, focusing only on observable consequences) come from? One motive may be the idea that intentions are inscrutable to observers. Behind this suspicion, already expressed by Kant, lies a quite natural, Cartesian way of viewing intention as a purely internal phenomenon, a mental state or event which only the agent whose intention it is may directly know or have access to, while remaining hidden or discoverable only by indirect means to any other agent.12 Anscombe’s reaction is to show that this is bad philosophy of psychology, or more exactly a bad grammatical account of the concept of intention. It is true that it may sometimes be the case that only the agent can say what is or is not his intention,13 and this can certainly induce the picture of intention as an inner phenomenon. But such a case does not show that an agent’s intention is always hidden to an observer, as can be seen from the fact that we are able to give descriptions of what people do that are also descriptions of their intentions.14

3.3  Intentional Action and Agent’s Knowledge Anscombe’s monograph Intention can be read as a long and complex answer to the Aristotelian question. Her move is to suggest that the conditions for imputing an action to someone will be elucidated by a synoptic view of the functioning of action descriptions in order to grasp the centrality of the concept of intention. One of her main general goals is to relate the description of intentional actions to agent’s knowledge: ‘it is the agent’s knowledge of what he is doing that gives the descriptions under which what is going on is the execution of an intention’.15 How should we understand ‘agent’s knowledge’ here? In this section, I will confine myself to sketching the fundamental features of Anscombe’s account.

12 

Cf. G. E. M. Anscombe, ‘ “Glanville Williams” The Sanctity of Life and The Criminal Law: A Review’, in GG1, 247. 13  Cf. I, §27, p. 48. 14  I, §4, p. 7. 15  I, §48, p. 87.

Intention, Knowledge, and Responsibility    59

3.3.1  There is always an indefinite plurality of true descriptions of what an agent is doing. The famous example of the pumping man, given in Intention §23, is meant to point to a general fact, which in turn raises important conceptual questions. Anscombe wants us to observe that in answer to the question ‘What is the man doing?’, a plurality of true descriptions can be given: he is moving his arm up and down, he is producing a clicking noise, he is making the pump work, he is replenishing the water supply, he is poisoning the inhabitants . . . By ‘description’ here, we are to understand a predicate by means of which one can refer to what is going on where the subject of the predicate is the agent, in a somewhat Russellian way analogous to the way in which you may designate an individual by resorting to a definite description.16 The conceptual questions raised by this phenomenon include the following: Amongst the set of true descriptions that apply to what the man is doing, which ones are those under which what he is doing is indeed the execution of one of his intentions? And which of these latter are referring to one and the same action, and why?17 Anscombe’s move from here consists in sketching the functioning of a complex language-​game, in which an observer asks the agent why he is doing A, where ‘A’ is a true description of what the agent does. When this question asks for reasons for acting (and not for proofs or causes), its applicability, i.e., the fact that the agent gives a positive answer to it (or would give one, if he was asked) or does not reject the question is the criterion marking the fact that the description A is a description of an intentional action, i.e., an action which is the execution of one of the agent’s intentions. In substance, it is a question of allowing the idea that intentional actions are those for which the agent may be meaningfully asked to account, by giving reasons which purport to exhibit their goodness.

3.3.2  The descriptions under which what the agent is doing is intentional are those under which the agent knows what he is doing. It may happen that the agent rejects the question ‘Why are you doing A?’ on the grounds that he did not know that he was doing A. For instance, the agent may reject the question ‘Why are you sawing Smith’s plank?’ on the grounds that he was unaware that this plank belonged to Smith. Indeed, an agent knows what he is doing only under some of the descriptions which are true of what he does. Though all are co-​referential, any one cannot be substituted for any other in the context ‘He knew he was doing X’. The agent’s

16  17 

Cf. G. E. M. Anscombe, ‘ “Under a Description,” ’ in CP2, 208–​219. Cf. I, §23, p. 37.

60   Rémi Clot-Goudard rejecting the question ‘Why are you doing A?’, provided that it be sincere, shows that his action was not intentional on his part under that description. Conversely, this shows that an agent may be correctly said to do A intentionally only if he knows what he is doing under this very description A. An action is then intentional under some description: this famous Anscombean point means that for an action A to be intentional, it is necessary that the agents knows what he is doing under this description A.18

3.3.3  However, ‘knowing what one is doing’ can have different meanings. We have to distinguish two possible meanings of ‘knowing what one is doing’. I may know what I am doing because I have observed it or because someone has told me that I was doing such-​and-​such, as when an observer asks me a question and thereby informs me that I was doing such-​and-​such. ‘Why are you making all this dust?—​Oh, sorry, I was not aware of it’: in that case, which for Anscombe counts as a rejection of the applicability of the question ‘Why?’, the description does not give an intentional action of mine. The knowledge I got in such a way about my own doings is only contingently related to the fact that I am the agent: I come to know what I do as any observer would. But agent knowledge, strictly so called, has a striking feature, illustrated by the following sort of case. While closing my eyes, I am writing ‘I am a fool’ on the blackboard; I know that this is what I am doing even if I am not looking at my hand holding the piece of chalk and moving it on the surface.19 In such a case, my knowledge of what I am doing does not rely on any kind of observation. I know what is going on under the description A inasmuch as I am the agent performing the action described as A. How should we account for this striking feature? Anscombe thinks that it may be elucidated by appeal to the idea that agent knowledge is practical knowledge, which she defines, following Aquinas’s account, as ‘the cause of what it understands, unlike “speculative” knowledge, which “is derived from the objects known” ’.20 The contrast between speculative (or theoretical) and practical knowledge is at the root of the tale of the two lists developed in Intention, §32. In this parable, one man is doing some shopping, following a list of items to buy, while a detective tailing him writes down on a second list what the man actually buys. Though both lists refer to the content of the man’s basket, their uses differ in that the detective’s list is meant to reflect the items actually bought,

18 

It should be noted here that such a condition is certainly a necessary but not a sufficient one. I may know, e.g., that I am making a lot of noise in sanding the parquet, without necessarily making such noise intentionally: it is not intentional on my part, though it may be said to be voluntary on my part (inasmuch as it is a side effect of my action to which I consent; but if I was constantly complaining about this awful noise, making such a noise may be said to be an involuntary act on my part). 19  This does not mean that observation does not play any role in the performing of intentional actions; but that observation is not the source of the knowledge of what I am doing (I, §48, pp. 87–​89). 20  I, §48, p. 87.

Intention, Knowledge, and Responsibility    61 whereas the man’s list is what the content of the basket has to reflect: it is guiding the man’s action. Hence the difference between two kinds of mistake: if there is any discrepancy in the detective’s record, the error lies in the list, which has to be corrected; but if there is any discrepancy between the man’s list and the items bought, the error lies in what he bought, not in the list, and this is what he would have to correct by another action. The difference between the two lists is then analogous to a difference between two ways of comparing what is thought and what is the case, and this is how we are to understand the contrast between ‘theoretical’ and ‘practical’. While theoretical knowledge is meant to reflect the way things are, practical knowledge is the knowledge of what is the case when what is the case conforms to what the agent intends to do. But the practical knowledge demonstrated by an agent when he says truly what he is doing is more than the expression of what is guiding his action: saying that ‘practical knowledge is ‘the cause of what it understands’ . . . means more than that practical knowledge is observed to be a necessary condition of the production of various results; or that an idea of doing such-​and-​such in such-​and-​such a way is such a condition. It means that without it, what happens does not come under the description—​execution of intentions—​whose characteristics we have been investigating’.21 In other words, while articulating his practical knowledge of what he is doing, the agent gives descriptions which are the standards in the light of which what he is doing can be counted as intentional action on his part.22 The mysterious non-​observational nature of practical knowledge vanishes once it has been related to the logic of practical reasoning, i.e., deliberation. Deliberation is a kind of reasoning in which an agent asks himself what he should do in the circumstances he finds himself in. It is a kind of first-​person reasoning, oriented towards an end actively desired by the agent, whose function is to elaborate a course of action by identifying regressively the many steps by which the agent may attain the desired end, till the agent discovers a first step which lies directly within his power. The rational order thus conceived has its analogy in the range of descriptions of the corresponding action. It is indeed characteristic of intentional actions that they exhibit a complex rational order mirrored in the logical relations holding between descriptions A, B, C, D of one and the same act, such that the agent can say ‘I’m doing A in order to do B, I’m doing B in order to do C . . .’; conversely, such a means-​end order opens up the possibility of the question ‘How do you do D?’23 Anscombe’s idea is that, when such a piece of practical reasoning 21 

I, §48, pp. 87–​88. Cf. Vincent Descombes, ‘Comment savoir ce que je fais?’, Philosophie 1, no. 76 (December 2002): 15–​ 32; Richard Moran, ‘Anscombe on ‘Practical Knowledge’, in Agency and Action, ed. John Hyman and Helen Steward, Royal Institute of Philosophy Supplement, vol. 55 (Cambridge: Cambridge University Press, , 2004), 43–​68. 23  It must be noted here that intentional actions are not only those the descriptions of which can be ordered along such a structure ABCD. The applicability of the ‘Why?’ question in the relevant sense, which Anscombe takes to be the criterion for descriptions of intentional actions, can also elicit answers in terms of interpretative or backward-​looking motives that have nothing to do with further intentions (I, §13, pp. 20–​21). The intentionalness of an action does not depend on the existence of an answer to the question ‘Why?’ of the form ‘in order to. . . .’ But Anscombe’s point is to argue that the logical possibility 22 

62   Rémi Clot-Goudard is sound and when the agent does not commit any mistake in the execution while acting upon it, then he is able to say what will happen and indeed what is happening. This ability to say truly what is happening is what is called practical knowledge. It is a form of knowledge proper to rational causes as human beings are.24 Imputability of an intentional action A to an agent thus requires that he knows what he is doing under the description A, and that such knowledge is practical knowledge. Conversely, it gives the rationale of the fact that sincere ignorance on the agent’s part that he is doing A is a ground for not calling his action intentional. Ignorance often renders an action involuntary, and this may give a ground for discharging the agent of any responsibility; but it is not always the case. When and to what extent does ignorance exonerate the agent? And when it does not, why doesn’t it?

3.4  When Ignorance Exonerates or Excuses As Anscombe writes, ‘we can often say that an action was either intentional or involuntary’:25 when being guilty of having performed A requires an intention on the part of the agent, ignorance, if successfully established, may exonerate the agent from the action A, as it is incompatible with having done A intentionally. In such cases, his action under the description A can be called an involuntary one: this is not something the agent wanted to do. But it would be wrong to stick to this disjunction without further specification. In fact, the epistemic condition on responsibility does not always function in one and the same way: there are numerous different cases according to different kinds of action descriptions, which embody different conditions of imputability. Moreover, an agent’s ignorance can be about different things. In line with the account of agent knowledge adumbrated in Intention, these topics are discussed at length by Anscombe in her paper ‘Two Kinds of Error in Action’, where she first introduces the general principle ‘error destroys action’ before considering its application in moral philosophy. The general point is that it is necessary to distinguish between different kinds of action descriptions or predicates, according to the way the epistemic condition on responsibility is embedded in their application: ‘There are some descriptions “X” of of such an answer is crucial to the very existence of the form of description she is investigating (cf. I, §20, p. 33). 24  From a Thomistic point of view, practical knowledge is the kind of science God has of the world as His Creator: see Aquinas, Summa Theologica, I, Iae, q. 14, art. 8, as well as Prologue, Summa Theologica, Ia IIae, and Peter Geach’s comments in Logic Matters (Oxford: Blackwell, 1972), 324–​327. 25  Cf. G. E. M. Anscombe, ‘On Being in Good Faith’, in GG2, 105. (In Intention, §7, pp. 12–​13, Anscombe points out the complex use of the term ‘involuntary’, which ‘neither means simply non-​ voluntary, nor has an unproblematic sense of its own’.)

Intention, Knowledge, and Responsibility    63 things done that cannot hold unless the subject knows he is doing X; for example, making a contract. There are some that can hold though the subject does not know he is doing X; for example, signing a transfer. But there are, further, some interesting intermediate cases’.26 Some descriptions of what an agent is doing cannot be true unless he knows what he is doing under such-​and-​such description. Anscombe often gives marriage as a paradigmatic example of that sort: ‘to marry someone’ is a description that we may call intrinsically intentional, in that it cannot be truly said of someone that he or she is marrying unless he or she intends to do it.27 If the agent is ignorant of a certain circumstance of the situation, e.g., if he believes falsely that he is taking part in a mere rehearsal, his ignorance establishes the lack of intent and consequently destroys the action under this description.28 The intent is formally embedded in the description conceived as a type; there is logically no room for saying ‘he did it involuntarily’. By contrast, some descriptions X allow for their being truly applied even if the agent does not know himself to be doing X, such as ‘offending someone’.29 The other ‘intermediate cases’ Anscombe has in mind obviously refer to descriptions displaying an illicit character, as is shown by her examples: ‘adultery, murder, bigamy’.30 The general question here is to know whether an agent’s ignorance is a sufficient ground not to impute him the alleged action or, if it is not, whether he should be held guilty for it. In other words, how does ignorance affect the content of an agent’s will? Adultery, along with bigamy, is one type of intermediate case. The question is: is it true to say of a man that he committed adultery if he has had sexual intercourse with a woman he thought was not married, but who actually was? The case is intermediate because we hesitate between answering ‘He committed adultery, but involuntarily’ and ‘He did not commit it’. In such cases, there is room for a distinction between formal and material ascription of the act. Because of his ignorance, the agent did not formally commit adultery: he did not intend it, and one cannot ascribe him the will to do something illicit. But in spite of his ignorance, we can nonetheless say that he has materially committed it: he did what adultery consists in (having intercourse with a married person), though he 26 

TKEA, 4. Cf. I, §48, p. 87: ‘here it might strike someone as curious that in general special proof of intention is not required; it is special proof of lack of it (because one of the parties did not know the nature of the ceremony, for example) that would invalidate a marriage’. 28  TKEA, 4. This is possible because concepts X such as ‘marrying’ are actually complex predicates which can be decomposed into more basic ones, in the form ‘doing ABC to do X’. To marry is thus to say ‘yes’ when such-​and-​such person asks such-​and-​such a question in such-​and-​such circumstances, etc. 29  This has to do with what Anscombe writes in Intention §47, p. 85, where she argues that a large number of action descriptions which are not intrinsically intentional would not exist, had we not in our conceptual apparatus the form of description ‘intentional actions’. Descriptions like ‘offending someone’ pertain to this larger class: their dependency on the form of description delineated by Anscombe consists in that they could not be used to describe unintentional actions if there were no cases in which they are used to describe intentional actions, cases which are constitutive of their sense. 30  TKEA, 4. 27 

64   Rémi Clot-Goudard had no intention to do it.31 The result of such an account is that those particular actions and their upshot can be imputed to their agent, though he may be discharged of guilt because of his ignorance, which then is not exonerating but functions as an excuse.32 However, this excuse cannot always be effective: Anscombe points out that for the agent not to be guilty of adultery, he must sincerely and reasonably think that his partner is not married, and his sincerity could be doubted if, say, the woman wore a wedding ring. Ignorance does not excuse the agent when he did not enquire into his partner’s marital status where it would have been reasonable to expect that he does so. In other words, lack of intent and ignorance do not always block the imputation of an act and the guilt related to it (see later discussion). Murder, to which Anscombe dedicates a lot of attention in her papers, is another type of intermediate case. The concept of murder embeds a formal condition of intentionalness, as did the concept of marriage, but its functioning is slightly different from that of such a description as ‘marrying’. On the one hand, for it to be true that an agent commits a murder, it is not necessary that he thinks of himself as committing one. But on the other hand, committing a murder is deliberately doing such-​and-​such a thing in such-​and-​such circumstances, and it is necessary that the agent knows himself to be doing such-​and-​such a thing, etc. for it to be true to impute to him the intention to commit murder, e.g., putting poison in his wife’s soup to have her swallow it. Hence a contrast with the former case, in which it was not necessary for the agent to know that he was doing XYZ in circumstance C, when A consists in doing XYZ in circumstance C, for it to be true to say of him that he did A. But what if the agent said that he did not know that doing XYZ in circumstance C would count, say, as an act of murder? It would then be a case of so-​called ignorance which Anscombe labels ignorance of law, by contrast with ignorance of facts, a distinction she illustrates by the example of theft. If you think that in such-​and-​such circumstance, you have the right to take into your possession a certain good, but are mistaken about the circumstances actually holding, then it is a case of ignorance of facts (e.g., you take my pen, thinking it has been abandoned on the table, whereas it is mine). However, as Anscombe remarks, such ignorance can fully exonerate you only insofar as it is non-​blameworthy; in some cases, you could be blamed for your lack of

31  The distinction is analogous to the scholastic one between formal and material objects, as in the example of the hunter who shot his father dead, thinking he was a stag: the formal object of his shooting was a stag (he intended to shoot at a stag), but the material object of his shooting was his father. See TKEA, 5. 32  Cf. AIDE, 212–​213. Most of the time, there is probably no particular need to distinguish between exoneration and excuse, but we may want sometimes to do this in order to take into account the difference between two kinds of case. When the action an agent is intentionally engaged in is not specifically wicked but happens to have a bad upshot which is due to ‘sheer bad luck’ or is ‘utterly accidental’, then this bad effect does not make the action a bad one, and ignorance is exonerating; the agent is ‘in the clear’. On the other hand, there are mixed cases in which the action, though not wicked, has something of a bad result which ‘need[s]‌excuse and therefore pardon’—​typically, when the bad result could have been avoided had the agent paid more attention to some circumstances.

Intention, Knowledge, and Responsibility    65 knowledge. But if you are aware of the circumstances holding, while thinking wrongly that in such-​and-​such circumstance, you have the right to take into your possession a certain good, then you are in ignorance of the law (e.g., you think that a house which is unoccupied for a year is an abandoned house and that you are allowed to take possession of it; though you may be right about the fact that it is unoccupied, you are wrong about the principle). Now, can ignorance of law be a ground for exoneration? In legal matters, as Anscombe points out,33 it is not an accepted excuse. Though judicial practice does take into account such ignorance to some degree, the general principle ignorantia juris non excusat holds nonetheless: it is the very condition of the exercise of judicial powers.34 In moral philosophy, however, it seems that such a defence should be allowed under certain conditions. It would probably seem preposterous in the poisoned soup example. But Anscombe appeals to the situation of a public executioner who is ordered to put to death a convicted person he knows to be innocent, without being able to prove publicly his innocence; would, then, the fulfilling of his terrible task count as committing injustice and murder? Whatever may be the right answer to this particular question (highly debated amongst theologians), the difficulty of arriving at it should be acknowledged as an exonerating motive: if law is hard to know with certainty, then one should not be blamed for ignorance of it. But wherever the law is easy to know, then ignorance of it cannot be rightly invoked to exonerate oneself from an illicit act. In the final analysis, it appears that both ignorance of facts and ignorance of law can exonerate an agent, provided that such ignorance does not result from a blameworthy failure on the agent’s part. An act done in ignorance may be blameworthy nonetheless. What is the rationale for this paradoxical possibility? Anscombe’s view, as we will see, consists in introducing the difference between voluntariness and intentionalness: ‘one cannot intentionally, but can voluntarily, do something without knowing one is doing it’;35 it can be said of the ignorant agent that he committed voluntarily such-​and-​such an act if he is guilty of his ignorance, i.e., if it is voluntary on his part, if one can impute it to his will. By putting forward this condition, Anscombe suggests that there is in fact a general obligation to know whether doing XYZ in circumstance C is an instance of such-​and-​such an illicit act. Where does this obligation come from, and how should we account for it?

33 

TKEA, 6–​7. The rationale of this principle and of its fair application may of course need further elaboration. For instance, the principle may hold only if the law is made a matter of public knowledge; that is, if it is promulgated. Besides, the proliferation of laws and codes in modern democracies, along with their growing technicity, may be questioned as they may be an obstacle to the autonomy of judgement of modern citizens. As interesting as they are, however, these questions lie outside the scope of the present essay. 35  TKEA, 7n2. 34 

66   Rémi Clot-Goudard

3.5  Blameworthy Ignorance and the Obligation to Know On this topic, Anscombe defends a view which has its roots in Aquinas: ignorance can be voluntary when it can be imputed to the agent’s will.36 How should we understand this? Actually, the expression ‘voluntary ignorance’ may signify different things. To see both options, we have to distinguish between ‘wanting not to do something’ and ‘not wanting to do something’, that is, lacking the intention to do (or not to do) something. In the first case, ‘voluntary ignorance’ can mean (but does not necessarily mean) that ignorance results from a decision not to know something, from a positive act of will aiming at ignorance, as when someone avoids asking a certain question because he knows that he would then have to take the answer into consideration and he does not want that. But ignorance can be voluntary in a second sense, when it is the mere negative result of a lack or failure of action, i.e., when it stems from the agent’s not having done something that he should and could have done. This possibility is built by analogy with the pilot of a ship who is responsible for a shipwreck because he did not do something which, as a pilot, he should and could have done—​because, say, he stayed in his cabin playing cards. In such a case, he can be blamed for not having acted, and the bad effect which results from this omission can be rightly imputed to him. The same goes for the lack of action which leads to an agent’s ignorance. Ignorance may then be voluntary, not only when it results from a decision on the agent’s part but also when it was necessary and possible for the agent to know something in order to do well.37 Let us consider each of these two conditions in turn.

3.5.1  Necessity: ‘He Ought to Know’ The existence of such a necessity is the first condition through which an omission can be considered as voluntary and its consequences imputed to the agent. But what kind of duty or necessity is here at work? According to Anscombe, it is a matter of practical necessity, that is, the kind of necessity which Aristotle defines in Metaphysics Delta as follows: ‘necessary’ in that sense is that without which one will not obtain a certain good or avoid a certain evil.38 Thus, ‘the pilot must navigate to preserve the ship; 36 

Cf. TKEA, 8; ‘Sin’, in GG1, 132, where Anscombe cites Aquinas, Summa Theologica, Ia IIae, q. 76, art. 3. 37  Doing well is intended here to be the same notion as Aristotle’s eupraxia, which he took to be the formal goal of every action. It is basically the same as acting well. On the two conditions for ignorance being voluntary, see also ‘On Being in Good Faith’, in GG2, 109; ‘Sin’, in GG2, 136: ‘This “could and ought to have” is thus a required condition for the truth of a charge against someone that he is guilty of the ill consequences of his not acting. By extension, we can apply it in our account of blameworthy ignorance’. 38 Aristotle, Metaphysics, Book Delta, 1115a20–​26. See also Anscombe, RRP, 100.

Intention, Knowledge, and Responsibility    67 the cook must put salt in the potatoes to cook them well’.39 The strength of the modal ‘must’ in those examples is relative to a certain end; as we can see, actions that are practically necessary to obtain such-​and-​such an end (aimed at under the guise of the good) may be contingent matters in themselves: the cook can add salt or not, the pilot navigate or not. In a series of lectures given in 1989, Anscombe appeals more explicitly to Aquinas40 to explain her position on the practical nature of blameworthy ignorance and of the obligation to know. She recalls first the distinction between nescience and ignorance: where ‘nescience’ refers to the simple fact of not knowing, ‘ignorance’ is not knowing what one needs to know in living, including fundamental truths which any human being is born able to learn. Amongst these, according to Aquinas, ‘there are things without the knowledge of which [the agent] cannot engage in such action as is due from him’, which encompass ‘the things that concern [his] condition or [his] job’, but also ‘things of faith and the universal precepts of the law’ that concern the very condition of human beings qua rational beings tending towards the good.41 Thus, ignorance conceived as not knowing what one is required to know in order to do well can then appear, in Aquinas’s perspective, as the result of a blameworthy inaction. As Anscombe explains, ‘the cause of ignorance is the non-​application of one’s mind to knowing and this itself, this non-​application of one’s mind to knowing what one ought to know, is a sin of omission’.42 In other words, ignorance is culpable when it is due to negligence. As we can see, Anscombe following Aquinas relies particularly on the analogy, implied in the examples of the pilot and the cook, between holding a certain position or status with associate duties and the general condition of human beings qua rational agents. But if it is easy to concede that acting and fulfilling one’s role or status implies carrying out certain duties, how are we to understand the view that the very condition of human beings implies something analogous? This makes sense against the background of Thomist ideas: human beings qua rational beings tend towards the good and doing well, which includes the flourishing of the human form of life; but doing well requires a sufficient knowledge of right and wrong, of ‘universal precepts of law’ which any human being is alleged to be born able to learn.43 Anscombe’s account of blameworthy ignorance appears thus to be committed to a strong conception of ethics which includes a normative view of human nature and the existence of moral truths (and errors).44

39 

TKEA, 9. Cf. Aquinas, Summa Theologica Ia IIae, q. 76, art. 2, cited in ‘Sin’, 131. 41  Cf. ‘Sin’, 136: ‘For the conduct of life, a lack of control by right reason is either blameworthy or only blameless because one is blamelessly in a gravely disordered state approximating to insanity’. 42  ‘Sin’, 132. 43  This has to do with ‘natural law’, as Aquinas understood it that can be described as a natural disposition to seek one’s good which stirs human reason; cf. Summa Theologica Ia IIae, q. 94, art. 1–​2. 44  On this, see, e.g., TKEA, 7. 40 

68   Rémi Clot-Goudard

3.5.2  Possibility: ‘He Could Know’ If negligence is not applying one’s mind to what one ought to know, it is culpable insofar as applying and not applying one’s mind are two opposites which equally lie within the will’s power. Just as a pilot of a ship who would not be physically able to steer it because of a sudden heart attack, say, could be exonerated from the loss of the ship, an agent who was not in a condition to know what he ought to know would be exonerated from the culpability for his resulting ignorance. It would not count as negligence on his part. However, Anscombe remarks that, though the agent’s condition may be a relevant feature to take into account, the most interesting sense of possibility involved here (‘he could have known’) concerns ‘the knowability of the thing to someone in his position’, that is, ‘given what he already knows’.45 The relevant question is often ‘Was this piece of knowledge accessible to him?’ rather than ‘Was he in such a condition that he could know it?’ The agent may not be guilty of negligence when it is difficult to know what the law says (be it positive or natural law) about the point at issue.46 On the other hand, wherever the law is easy to know, the agent will not be allowed to give as an excuse that he did not know it or that he did not think of it. Moreover, Anscombe’s idea of knowability is related to some degree of rational development of human agents: she holds that no one is subject to obligations unless he already knows a lot of things, and in the first place has mastered a language (hence the fact that children are not submitted to the same obligations as adults, and this would hold of persons suffering from strong cognitive disabilities). Does this mean that the contingent condition of the human agent may never function as an exonerating reason? Though Anscombe acknowledges this possibility, she shows herself very critical of what she perceives as an abuse of the Thomist notion of invincible ignorance, i.e., the ignorance that cannot be undone by any effort or applying of one’s mind (studium).47 Knowing is often a matter of checking something. But what if the pilot of the wrecked ship said, ‘It did not occur to me to check the sonar’? It would not exonerate him, for even if it was true, one would like to answer back, ‘It should have occurred to you’. Now, this reply may sound strange to our ears: how could I want that the idea that I should check the sonar occur to me?48 In fact, the apparent difficulty can be lifted by interpreting the answer ‘It should have occurred . . .’ as an accusation of casualness: it means that the agent does not pay enough attention to certain important things.

45 

Cf. ‘On Being in Good Faith’, in GG2, 111. Cf. the example of the public executioner in TKEA, 7. 47  Cf. Aquinas, Summa Theologica Ia IIae, q. 76, art. 2, cited in ‘Sin’, 132. 48  Cf. St Alphonsus’s reflections about Suarez, mentioned by Anscombe in ‘Sin’, 132: if no idea comes to my mind which arouses the will to know, how could I be responsible for my ignorance, given that it is not in my power then to move towards this knowledge? These reflections are a source of paradox: if to do something voluntarily, I must have an idea of it, and if to arouse in me the idea of doing something, I must have the idea of arousing it, how could I ever do anything voluntarily? For I would need to have the idea of having the idea, etc. But this, of course, is absurd. 46 

Intention, Knowledge, and Responsibility    69 As Roger Teichmann convincingly points it out,49 an agent who says ‘It did not occur to me’ by way of excuse tells us something about the general orientation of his will. What is at stake is intellectual virtue (or vice): in order to act well, one has to get used to making enough space for some things and none for others in one’s practical considerations and deliberations, when one is engaged in such-​and-​such a course of action.

3.6  Conclusion During a seminar she gave at Johns Hopkins in 1987, Anscombe remarked that Moore was wrong to say that Socrates is responsible for a better state of the world taken as a whole. Besides the fact that an expression such as ‘a better state of the world taken as a whole’ may not have any coherent and useful meaning, Socrates is not responsible for the world: he can be responsible only for what depends on him, that is, for the contingent futures that are in his power.50 A limitless responsibility is a nonsense. As Rachael Wiseman noted,51 Anscombe aims at articulating a ‘practicable’ view of responsibility towards the future (limited to what one intends or voluntarily brings about) and a ‘severe’ view of responsibility towards the past (according to which one is entirely responsible for what one has done, intentionally or voluntarily). Anscombe’s conception of responsibility is nonetheless a strong one: an agent need not always fully know what he is doing for a guilty act to be imputed to him, because he may be responsible for his ignorance. In her eyes, the knowledge an agent has of his own action qua agent is that through which his action counts as intentional; this practical knowledge is the rationale for what can be imputed to the agent as intentional actions. However, the voluntary is a larger domain than the intentional, and the agent is responsible for what is voluntary on his part. Moreover, it is the law (in its moral or legal sense), and not what the agent thinks to be the law, that gives the standard by which the agent’s voluntary action will be evaluated as just or unjust.52 By insisting on the importance of the law, Anscombe intends to resist subjectivism, defined here as the idea that I am justified if I have acted upon my conscience. By stressing the fact that ignorance and error can be culpable, she rejects the view that having acted upon one’s beliefs is enough to discharge the agent

49  Roger Teichmann, ‘The Voluntary and the Involuntary: Themes from Anscombe’, American Catholic Philosophical Quarterly 88, no. 3 (Summer 2014): 485–​486. 50  I owe this anecdote to Vincent Descombes, who was then teaching at Johns Hopkins University and attended the seminar. 51  Wiseman, ‘The Intended and Unintended Consequences of Intention,’ 17. 52  See ‘On Being in Good Faith’. See also the criticism against the modern interpretation of mens rea in TKEA, 6: the modern doctrine says that a guilty mind is a mind conscious of wrongdoing, where the old doctrine—​to which Anscombe adheres—​said that if the law forbids to kill this man or to-​kill-​a-​man-​ in-​such-​and-​such-​circumstances, and if one shows that the agent wanted to kill-​a-​man-​in-​such-​and-​ such-​circumstances, then his guilty intent has been proved: indeed, the agent wants to do what the law forbids to do.

70   Rémi Clot-Goudard from all guilt and to exonerate him from the consequences of his act. In fact, sincerity (or bona fides, good faith) should not be conceived as ‘a secret colour of one’s thoughts’ but rather as ‘the purity of one’s intentions in thinking these thoughts’;53 by this, she means that we should think this or that thought because, for all we know, it seems to be the relevant truth of the matter at hand, ‘as opposed to its appearing to be the relevant truth on the matter in hand because it is a convenient or otherwise tempting thought, i.e. because there is an inclination to think it’.54 As Perceval guiltily failed a first time in the Quest because he did not ask the Wounded King what was the wonder which he saw thrice, an agent may fail to know what he should have known, and this can be imputed to him. Thus, doing well binds us to a practical obligation of knowledge: when we don’t try to know what we ought to and could know, this appears as reproachable negligence, against which rises the need for the intellectual virtue of phronesis.55

Bibliography Anscombe, G. E. M. ‘Action, Intention, and “Double Effect”’ (1982). In Human Life, Action and Ethics: Essays by G. E. M. Anscombe, edited by Luke Gormally and Mary Geach, 207–​226. Exeter: Imprint Academic, 2007. Anscombe, G. E. M. ‘Does Oxford Moral Philosophy Corrupt the Youth?’ (1957). In Human Life, Action and Ethics: Essays by G. E. M. Anscombe, edited by Luke Gormally and Mary Geach, 161–​167. Exeter: Imprint Academic, 2005. Anscombe, G. E. M. ‘Glanville Williams The Sanctity of Life and The Criminal Law: A Review’. In Human Life, Action and Ethics: Essays by G. E. M. Anscombe, edited by Luke Gormally and Mary Geach, 243–​248. Exeter: Imprint Academic, 2005. Anscombe, G. E. M. Intention. 2nd edition. 1957; Cambridge, MA: Harvard University Press, 2000. Anscombe, G. E. M. Mr Truman’s Degree (1957). In Ethics, Religion and Politics: Collected Philosophical Papers, vol. 3, 62–​7 1. Oxford: Blackwell, 1981. Anscombe, G. E. M. ‘Murder and the Morality of Euthanasia’ (1982). In Human Life, Action and Ethics: Essays by G. E. M. Anscombe, edited by Luke Gormally and Mary Geach, 261–​277. Exeter: Imprint Academic, 2005. Anscombe, G. E. M. ‘On Being in Good Faith’. In Faith in a Hard Ground, edited by Luke Gormally and Mary Geach, 101–​112. Exeter: Imprint Academic, 2008. Anscombe, G. E. M. ‘Rules, Rights, and Promises’ (1978). In Ethics, Religion and Politics: Collected Philosophical Papers, vol. 3, 97–​103. Oxford: Blackwell, 1981. Anscombe, G. E. M. ‘Sin’ (1989). In Faith in a Hard Ground, edited by Luke Gormally and Mary Geach, 117–​156. Exeter: Imprint Academic, 2008. Anscombe, G. E. M. ‘The Two Kinds of Error in Action’ (1963). In Ethics, Religion and Politics: Collected Philosophical Papers, vol. 3, 3–​9. Oxford: Blackwell, 1981. 53 

‘On Being in Good Faith’, in GG2, 103. ‘On Being in Good Faith’, in GG2, 103. 55  Thanks to Vincent Descombes, Jean-​Yves Goffi, Constantine Sandis, Severin Schroeder, and Roger Teichmann for their helpful comments on an earlier version of this chapter. All errors are mine, though I committed them in ignorance. 54 

Intention, Knowledge, and Responsibility    71 Anscombe, G. E. M. ‘ “Under a Description” ’ (1979). In Metaphysics and the Philosophy of Mind: Collected Philosophical Papers, vol. 2, 208–​219. Oxford: Blackwell, 1981. Anscombe, G. E. M. ‘War and Murder’ (1961). In Ethics, Religion and Politics: Collected Philosophical Papers, vol. 3, 51–​61. Oxford: Blackwell, 1981. Aristotle. Metaphysics. Translated by W. D. Ross. In The Complete Works of Aristotle, edited by Jonathan Barnes, vol. 2, 1552–​1728. Princeton, NJ: Princeton University Press, 1995. Aristotle. Nichomachean Ethics. Translated by Christopher Rowe. Oxford: Oxford University Press, 2002. Austin, John. ‘A Plea for Excuses’. In Philosophical Papers, edited by J. O. Urmson and G. J. Warnock, 3rd ed., 175–​204. Oxford: Oxford University Press, 1961. Descombes, Vincent. ‘Comment savoir ce que je fais?’ Philosophie 1, no. 76 (December 2002): 15–​32. Geach, Peter. Logic Matters. Oxford: Blackwell, 1972. Moran, Richard. ‘Anscombe on “Practical Knowledge.”’ In Agency and Action, edited by John Hyman and Helen Steward, 43–​68. Royal Institute of Philosophy Supplement, vol. 55. Cambridge: Cambridge University Press, 2004. Teichmann, Roger. ‘The Voluntary and the Involuntary: Themes from Anscombe’. American Catholic Philosophical Quarterly 88, no. 3 (Summer 2014): 465–​486. Wiseman, Rachael. ‘The Intended and Unintended Consequences of Intention’. American Catholic Philosophical Quarterly 90, no. 2 (Spring 2016): 1–​21.

Chapter 4

Pr actical K now l e d g e and Testi mony Johannes Roessler

4.1  Introduction Early on in Intention, Anscombe writes that ‘the indicative (descriptive, informatory) character is not the distinctive mark of ‘predictions’ as opposed to ‘expressions of intention,’ as we might at first sight have been tempted to think’ (I, 3). The context of this remark is a somewhat unusual example, of a speech act used not just to express an intention and to inform an audience but also to give an order (‘Nurse will take you to the operating theatre,’ said by a doctor to a patient in the presence of a nurse). But the point itself has a wider significance: what Anscombe calls expressions of intention are routinely used to share knowledge of what one is, or will be, doing. Asked how he knows he’s going to be taken to the operating theatre, the patient ‘would say that the doctor told him’ (I, 3). The text of Intention is peppered with conversations in which agents tell an audience what they are doing, or why. The prominence of such exchanges is of course in keeping with Anscombe’s central thesis, that intentional actions are actions to which the (reason-​seeking, second-​person) ‘question ‘Why?’ is given application.’ Evidently Anscombe thought that what she called practical knowledge could be shared with others.1 In this respect, such knowledge would seem to be no different from any other knowledge. Yet the purported special features of practical knowledge may make a difference to what is involved in sharing it. I think it is a key commitment of Anscombe’s view that they do make such a difference. There is a sense in which some of her central notions—​ including ‘expression of intentions’ and ‘practical knowledge’—​are introduced, in the first place, as part of an analysis of how we understand agents’ knowledge of their

1 

For an illuminating discussion of this point, see Longworth (2019).

Practical Knowledge and Testimony    73 intentional actions from what might be called the ‘participants’ perspective’: the perspective of participants in the practice of sharing such knowledge.2 My aim here is to set out this suggestion in more detail and make a case for it. I also want to consider the suggestion’s bearing on one of the hard problems in Anscombe exegesis. In a gnomic passage, Anscombe blames modern philosophy’s ‘incorrigibly contemplative conception of knowledge’ (I, 57) for having obscured the possibility of practical knowledge. Just what that conception involves, and what is wrong with it, has been the subject of sustained debate and disagreements. Reflection on the participants’ perspective, I will be suggesting, can help to make progress with these issues. The root problem with the contemplative conception is not that it fails to heed certain insights of ‘ancient and medieval philosophers’ but that, contrary to its dominant self-​image, it is revisionary: it is at odds with the conception of agents’ knowledge that we actually have and that informs our practice of sharing such knowledge.

4.2  The Previous Owner Condition There are two practices that matter for my argument: the practice of sharing knowledge through testimony and the practice of questioning or challenging claims to knowledge. The relationship between them is a delicate matter, and I cannot attempt to offer a fully adequate discussion of it here. In thinking about it, we naturally encounter two kinds of consideration. On the one hand, reflection on the act of telling suggests that in telling something to someone one is liable to be held answerable in certain ways. Full participation in the practice of telling (as speaker or audience) requires appreciating that an act of telling can be open to certain kinds of normative questions, aimed in various ways at probing the credentials of the speaker’s claim to knowledge. On the other hand, it seems to be a compelling thought (which has loomed large in recent work on testimony) that, at least in central cases, the possibility of testimonial knowledge depends on the audience trusting the speaker’s assurance (Faulkner 2011; McMyler 2011; Moran 2018). The two perspectives may seem to pull in opposite directions. The importance of holding a speaker answerable may seem to place the onus on the audience. Responsible uptake may seem to require subjecting the speaker’s claim to critical scrutiny, where this involves making up one’s own mind about the justice of her claim to knowledge. Reflection on the importance of trust may seem to suggest that testimonial knowledge, in core cases, involves basic, irreducible forms of epistemic dependence on the speaker’s assurance. I will skirt the difficult issues arising from this apparent tension by proposing a maxim that, it seems to me, is independently plausible and weak enough to command

2 

I borrow the term from Richard Moran’s discussion of the ‘social acts’ involved in communication and testimony; see his 2018 passim.

74   Johannes Roessler assent from both of the potentially conflicting perspectives. Typically, when A comes to know that p as a result of B telling her that p, it will be reasonable for A to think that B knows that p. There are unusual cases, of course, as when I come to know that p because, as I recognise, your assertion that p reflects a combination of insincerity and incompetence (you are mistaken about whether p and also wish to mislead me). Again, a one-​year-​old is presumably able to learn from others without being able to think of her informants as knowing the things they are telling her. And occasionally someone’s testimony may sensibly be treated simply as adding to our stock of evidence for and against a certain view. Still, ordinarily we think of testimony as a matter of sharing knowledge, and so in receiving testimony we take it that the knowledge we acquire had a ‘previous owner.’ Importantly, if the transmission of knowledge works properly, it must be reasonable for us to credit our informant with knowledge. For ease of reference, I will call this as the ‘Previous Owner’ condition. It is not easy to say what is involved in meeting this condition, but it is worth stressing that the condition itself is not particularly demanding. One thing that is not required, for example, is that A should have independent reason to think that B knows whether p. It seems perfectly possible (and perhaps routine) for A reasonably to credit B with knowledge whether p simply by relying on B’s very testimony that p. That is not to say that the Previous Owner condition is idle or trivial. One way to see this is to consider a situation in which A has independent grounds for doubting B’s claim to knowledge that p, perhaps, though not necessarily, because she has reason to doubt that p is true. This would normally be enough to block the transmission of knowledge. The appropriate way for A to react to B’s testimony, in this situation, would not be to accept her testimony but to probe its credentials. Unsurprising responses would include ‘How do you know?,’ ‘Are you sure?,’ and ‘Why do you believe that p?’ In different ways, all of these questions can be used to ask for reassurance. For current purposes, the first question—​’How do you know?’—​is of particular interest. A good answer would ordinarily be expected to do two things: it would provide relevant information about how you came to know that p, and it would dispel possible doubts as to whether your attitude to p—​the attitude you express by telling us that p—​is indeed knowledge, rather than mere belief or conjecture. The answer would help to explain your possession of knowledge in a way that would simultaneously establish the credentials of your claim to knowledge. Thus, the participants’ perspective combines two concerns: a concern with understanding your epistemic position and a concern with the correctness of your claim to knowledge.

4.3  Sharing Practical Knowledge To turn now to the case of sharing knowledge of one’s intentional actions: suppose B is telling A ‘I am buying butter’ or ‘I am going to buy butter,’ and as a result A, believing B, comes to know that B is buying or going to buy butter. There is a well-​known list of

Practical Knowledge and Testimony    75 features that Anscombe suggests mark out B’s statement: it is an ‘expression of intention’ and so conveys a distinctive kind of knowledge, characterised as ‘non-​observational,’ ‘practical’ (as opposed to ‘contemplative’ and ‘speculative’), ‘knowledge in intention,’ and so on. How much, if any, of this material should be expected to be available to A and B? Are the various features on the list just intended to provide a philosopher’s theory of B’s knowledge, or are they, or some of them, thought to articulate the way we think about knowledge of intentional actions from the participants’ perspective? I think there can be no doubt that at least the first item on the list—​the idea that B’s statement is an ‘expression of intention’—​is supposed to be commonly recognised. A salient feature of expressions of intention is that they are statements that are ‘justified, if at all, by a reason for acting, as opposed to a reason for thinking them true’ (I, 15). (Anscombe uses ‘predictions’ instead of ‘statements,’ but her point can be extended to expressions of intentions that do not concern the future but inform one’s current activities, standardly expressed in English by the use of the present progressive.) Anscombe’s point here is not just that as a matter of fact, when it comes to ‘expressions of intentions,’ people tend to justify these, if at all, by giving one sort of reason rather than another. To call a statement an ‘expression of intention’ is to imply that it is correctly or properly justified in a particular way and that we would ordinarily justify it in that way (if we do) because we appreciate that this is the appropriate sort of justification to ask for and offer. Anscombe’s examples bring this out. A request for evidence in response to the statement ‘I am going to take a walk’ (‘What makes you think so?’) would not just be unusual but in some sense out of place, and it would be out of place not because it would be tactless or conversationally inappropriate (as it would be in response to ‘I am going to be sick’) but because it would reflect a misapprehension of what you were doing in telling us ‘I am going for a walk,’ viz. expressing an intention. Consider now the way A would make his own knowledge that B is buying butter intelligible to himself and others, by reflecting ‘B told me.’ Presumably what this means is that A takes B to have shared with him her own knowledge, in the way we commonly share knowledge through communication: roughly speaking, by addressing assertions to each other, recognised as amounting to claims to knowledge. This raises a further question about the participants’ perspective: how does A understand B’s epistemic position? And how does the fact that B’s statement is (also) an expression of intention bear on this? We can begin to see the difficulties here by noting that neither of two familiar models seems to be germane to the present case. I will label them the evidential model and the model of first-​person authority. The evidential model says that it is reasonable to credit an informant with knowledge of some subject matter insofar as it is reasonable to assume that they have at their disposal relevant evidence or epistemic reasons, or at least were in possession of such evidence when they acquired the piece of knowledge. Might this be how A understands B’s epistemic position vis-​à-​vis her buying butter? The trouble is that if A made sense of B’s knowledge in terms of evidence available to B, then it should be a legitimate move on A’s part to ask for B’s evidence. Questions such as ‘What makes you think you are buying butter?’ should be intelligible and in certain situations apt. Yet to ask such questions would reveal a failure to recognise what B was doing in telling A she

76   Johannes Roessler was buying butter, viz. expressing her intention. B’s statement is one that is ‘justified, if at all, by a reason for acting, as opposed to a reason for thinking [it] true’ (I, 15). In the light of the failure of the evidential model, it is natural to think that A’s willingness to accept B’s claim to knowledge is surely connected with the first-​person character of B’s statement. It is a familiar suggestion that first-​person self-​ascriptions of attitudes or experiences are ordinarily granted a distinctive authority, often characterised by reference to (a) the idea that, barring insincerity, such statements are false only in exceptional cases; (b) the idea that special resources are needed to make sense of mistaken belief (roughly, mistakes would be indicative of irrationality or confusion rather than merely reflecting, say, insufficient evidence or perceptual error); and (c) the impropriety of second-​person questions aimed at probing or challenging the credentials of such statements. Should we think of A’s acquiescing in B’s claim to knowledge as a matter of acknowledging her ‘first-​person authority’? This model seems more promising.3 The failure to execute intentions, Anscombe writes, ‘is necessarily the rare exception’ (I, 87). Furthermore, B’s statement is naturally viewed as immune to lines of questioning that would be appropriate in the case of claims to empirical knowledge (not only ‘Why do you believe?’ but also ‘How do you know?’—​I shall return to this). Yet the model of first-​person authority quickly breaks down under closer scrutiny. Even if failures to execute intentions are relatively exceptional (and note that Anscombe goes on to qualify her point by saying that ‘what is necessarily the rare exception is for a man’s performance in its more immediate descriptions not to be what he supposes’ [I, 87; my emphasis]), they are, at least in the case of less immediate descriptions, a familiar phenomenon and, importantly, a phenomenon that is often susceptible of a ready explanation. For example, owing to inattentiveness one may press button B when one means, and takes oneself, to be pressing button A (see I, 57). This, of course, is what Anscombe calls ‘Theophrastus’s point’: claims to knowledge of what one is doing are exposed to a distinctive epistemic risk, of being wrong owing to an ‘error in performance.’

4.4  Two Kinds of Practical Errors These problems with the model of first-​person authority relate to (a) and (b), the general presumption of correctness and the difficulty of making sense of error. One might try to salvage the model by tweaking the two conditions, perhaps adding clauses about the possibility of ‘errors in performance.’ But I think this would not be a profitable line to pursue. There is a deeper problem with the model, to do with (c). Understandably, ‘Theophrastus’s point’ has always occupied a central position in commentary on 3  Falvey (2000) argues that on Anscombe’s view, agents have ‘a general warrant to present their expression of what they intend to be doing as descriptions of what they are doing’ (37) and refers to agents’ ‘first-​person authority’ with respect to expressions of intention (38).

Practical Knowledge and Testimony    77 Anscombe’s Intention. Yet Anscombe makes it clear that an ‘error in performance’ is not the only error that may falsify claims to knowledge of what one is intentionally doing. Such claims may also reflect a distinctive kind of ‘error of judgement.’ Consideration of this second kind of error, I suggest, not only brings out what is wrong with the model of first-​person authority; it also suggests the beginnings of an alternative, more promising model. In her discussion of the case of ‘a man going round a town with a shopping list in his hand,’ Anscombe writes: [T]‌he discrepancy [between the contents of the shopping list and what the man actually buys] might arise because some of the things were not to be had and if one might have known they were not to be had, we might speak of a mistake (an error of judgement) in constructing the list. If I go out in Oxford with a shopping list including ‘tackle for catching sharks,’ no one will think of it as a mistake in performance that I fail to come back with it. (I, 56)

I first want to suggest that the passage speaks against two not uncommon assumptions in the literature on Intention: • Some expositions of Anscombe’s account of practical knowledge give the impression that the risk of an ‘error of judgment’ applies only to claims to ‘speculative,’ not claims to ‘practical’ knowledge. For example, Rosalind Hursthouse (2000, 103) takes ‘practical knowledge’ (or claims to such knowledge) to be marked by this fact: ‘But when I am in error, the mistake lies in the performance, not in a judgment about what I am doing.’ Yet in connection with the shark example, Anscombe speaks of an error of judgement ‘in constructing the list,’ where making the list is characterised as an ‘expression of intention’ (I, 56). In these circumstances, the statement ‘I am/​will be buying tackle for catching sharks’ would fail to express knowledge of what one is doing not because of an error of performance but because of an error of judgement.4 • A connected assumption is that by ‘judgement’ Anscombe means the act of judging or affirming or saying something. An ‘error of judgement’ would thus be an act of falsely or unjustifiedly judging that something is so. The assumption seems to be implicit in Hursthouse’s phrase ‘a judgment about what I am doing.’ It also informs McDowell’s (2010, 429) gloss on the ‘Theophrastus point’ (the point that ‘when I say ‘Now I press button A—​pressing button B,’ the error is not of judgement but of performance’): in this sort of case, McDowell suggests, ‘the primary defect is in what one is doing,’ though there is a ‘derivative defect not in what one is doing but in what one says.’ (The context suggests that the ‘derivative defect’ is supposed to be an example of an ‘error of judgement.’) But the assumption delivers a strained reading 4  The point is emphasised by Roger Teichmann (2008, 23) in his discussion of ‘Theophrastus’s principle.’

78   Johannes Roessler of ‘a mistake (an error of judgement) in constructing the list.’ ‘An error of judgement,’ it seems to me, is more naturally heard as a defective exercise of judgement—​ where ‘judgment’ is not an act but a capacity (the ‘power of judgement’). Of course, exercising one’s judgement may involve acts of judging that something is so, but it is not obvious that it has to involve such acts, and in any case it cannot be reduced to a series of judgings in the ‘act’ sense.5 The upshot is that we should resist the temptation to align the distinction performance/​ judgement with the distinction practical/​theoretical. What we have in cases in which ‘one might have known’ that an item on the shopping list was not to be had is precisely an error of practical judgement. There is something wrong here, not with the execution of the intention but with the intention itself: it reflects a wrongheaded choice of a means to achieve a given end, a failure properly to exercise one’s capacity for practical (specifically, calculative) reasoning. Someone who commits this sort of error will end up not knowing what she is doing. She will be apt to tell us ‘I am buying butter’ (or ‘I am going to buy butter’) when she is not, owing to her miscalculation. What undermines her claim to knowledge is not her lack of success (the statement ‘I am buying butter’ may be true even if she never lays her hands on butter—​compare ‘She was buying butter when the bomb went off ’) but her practical inability: ‘she was buying butter’ is arguably incompatible with ‘there was no butter to be had.’6 Correlatively, there are familiar questions A may direct to B to make sure no such error is operative: ‘How do you know they are selling butter?’ or ‘What makes you think the shop is open on a Sunday?’ The propriety of these sorts of question provides a basic objection to the model of first-​person authority: contrary to (c), the credentials of claims to practical knowledge are clearly open to intelligible questioning. True, the challenge these questions present to B’s claim to knowledge that she is/​will be buying butter is in a sense indirect. Their direct target is B’s entitlement to certain assumptions informing her practical reasoning. It is significant that the question is ‘How do you know they are selling butter?,’ not ‘How do you know you are buying butter?’ Still, there is an intelligible connection between B’s entitlement to assume that she will be able to buy butter and her entitlement to express her intention to do so by making an assertion as to what she is or will be doing. Suppose B has nothing reassuring to say in response to A’s question. This would not just mean that something has gone wrong in an area of B’s thinking somehow adjacent to her intention. It would mean that there is something wrong with the statement she is making in expressing her intention, and indeed with the intention itself.

5 

See also Anscombe’s remark on the status of Aristotle’s account of practical reasoning: if the account ‘were supposed to describe actual mental processes, it would in general be quite absurd. The interest of the account is that it describes an order which is there whenever actions are done with intentions’ (I, 80). When actions are done with intentions, it is natural to suppose, the capacity of practical reason/​practical judgement is exercised, but this may or may not involve acts of judging or saying something. 6  For discussion of the truth conditions of attributions of activities in the present or past progressive, see Falvey (2000), Thompson (2011), Wolfson (2012).

Practical Knowledge and Testimony    79 I said that reflection on the ‘error of judgement’ displayed in the shark case simultaneously puts pressure on the first-​person authority model and points in the direction of an alternative account of the participants’ perspective on practical knowledge. The direction I have in mind is this: if claims to practical knowledge are liable to be compromised by two distinctive sorts of error (errors of ‘judgement’ and of ‘performance’)—​ compromised, that is to say, by the agent’s failure properly to exercise two requisite practical capacities (to form realistic intentions and to execute them competently)—​then a recognition that no such error is being committed—​that the agent is properly exercising both kinds of capacities—​should enable us to see that and why her claim to such knowledge is correct. To put the point another way, A’s grasp of what would be good questions to ask in order to probe B’s entitlement to her claim to knowledge should be expected to reflect some understanding of the nature of B’s entitlement to that claim—​an understanding that may make it reasonable for A to credit B with knowledge.

4.5  Practical Reason as a Capacity for Knowledge I started from the observation that the participants’ perspective plays a significant role in Anscombe’s account of expressions of intention. That it would be inappropriate to ask B for evidence in support of her view that she is buying butter is supposed to be apparent to A and B, not just to philosophers who work on the nature of intentional action. I further suggested that participants appreciate not only the impropriety of certain questions but also the propriety of others, and I mooted the idea that this might show some understanding of B’s epistemic position. We seem to be pre-​theoretically familiar with a distinction between intentions that are, as we might provisionally put it, conducive to knowledge and intentions that are not—​for example, the intention to buy butter vs. the intention to buy tackle for catching sharks (in Oxford). Our grasp of that distinction, it is natural to assume, makes it reasonable for us to think that the agent knows what she is doing in one case but not the other. But how is our understanding of the epistemic role of intentions here to be articulated? The trouble is that an intention to do something and knowledge that one is/​will be doing it seem to be quite different sorts of states or attitudes. How might the former make possession of the latter intelligible? One familiar idea is that reflection on one’s intention to buy butter provides a basis for inferring (and so coming to know) that one will buy butter (see, e.g., Paul 2009). This, though, is hard to reconcile with Anscombe’s analysis of the participants’ perspective. It would suggest that a request for evidence in response to B’s statement ‘I am going to buy butter’ would be unexceptionable. On the other hand, one might suggest that it is the agent’s intentionally acting that establishes a link between her intention and her knowledge: intentions make our actions intelligible, and ‘practical’ knowledge is ‘but an aspect of intentional actions’ (Haddock 2011, 165).

80   Johannes Roessler That way of connecting intention and action, however, would suggest that the canonical way to establish whether B knows she is buying butter is to satisfy oneself that she is (intentionally) buying butter. (If she is, then she must know she is, given that this sort of knowledge is ‘but an aspect’ of the activity.) Yet this seems to get things backward. When we learn from B what she is doing, we come to know that she is (intentionally) buying butter in a way that depends on its being reasonable for us to credit her with knowledge that she is buying butter. My suggestion is that the link between intention and knowledge is provided by the distinctive kind of statement around which Anscombe’s discussion revolves. What makes such statements special is that they simultaneously perform two roles: they are used to express intentions but also purport to be statements of fact. As Stuart Hampshire (1965) puts it, a statement of this kind has a ‘double aspect.’ The distinctness of the two aspects can be brought out by noting that there are ways of engaging with the statement that selectively target just one aspect. For example, ‘Aren’t you supposed to be on a diet?’ challenges the intention expressed by ‘I am buying butter’ without questioning the statement of fact. But that is not to say that the two aspects are unconnected. For, as we have seen, there are also ways of probing the statement that simultaneously implicate both aspects. ‘Are you sure the shop is open?’ hints at a possible defect in the intention that would simultaneously undermine the statement of fact. How is this possible? How is the intention supposed to be related to the credentials for the statement of fact? Anscombe’s answer to this turns on her view of the nature of practical reasoning. (‘The notion of ‘practical knowledge’ can only be understood if we first understand ‘practical reasoning’’ [I, 57].) The question I just raised—​how is it possible for an intention to bear on the credentials of a statement of fact—​is naturally heard as insinuating that surely the putative bearing is puzzling, in view of X; where X may be, say, the difference between mental states with different ‘directions of fit’ (intentions supposedly aim for the world to fit with them, statements of facts to fit the world), or the difference between thinking concerned with what would be desirable/​what one has reason to do vs. thinking concerned with what is the case/​what one has reason to believe. But I think on Anscombe’s view, there is no such puzzle here, and the intimation of a puzzle reflects a flawed conception of practical reasoning. Not only is there nothing paradoxical about the idea that practical reasoning can warrant a statement of fact; on the contrary, it would be paradoxical to suggest that one’s reasoning could be genuinely practical if it did not warrant (or at least aim to warrant) a statement of fact. To bring this out, consider a view of practical reasoning on which B’s reasoning about how to obtain butter cannot licence the statement ‘I am going to buy butter’ but only weaker statements, to the effect, say, that she has most reason to, or ought to, buy butter.7 I think from Anscombe’s perspective, the trouble with any such weakening manoeuvre is this. If B is not entitled to affirm that she will buy butter, it is (or at least should be)

7 

For example, on Davidson’s account, practical reasoning licenses only evaluative or normative judgements. For objections to that view from an Anscombean perspective, see McDowell (2010).

Practical Knowledge and Testimony    81 an open question for her whether she will, which in turn means it should be an open question whether she has identified a way by which she will achieve her objective, to secure butter. And that amounts to the admission that her exercise of calculative practical judgement has not been fully successful; given her uncertainty over the factual question of whether she will buy butter, she should surely think of an alternative plan, in case she will not buy butter. The point is that ignorance about the factual question has an immediate practical significance. Calculative practical reasoning is not simply reasoning about practical matters but reasoning ‘towards action’: if successful, it is the first step towards achievement of the very goal that informs the reasoning.8 In effect, the weakening manoeuvre, by denying that practical reasoning ever entitles us to regard the factual question of what we will be doing to be settled, denies that we ever successfully exercise practical judgement. In the light of this, the questions to which B lays herself open in telling A that she is going to buy butter may be glossed as follows. B’s practical reasoning will warrant her double-​aspect statement only if she is entitled to rely on the premise that going to the shop will enable her to obtain butter. Plausibly, she is entitled to do so only if she knows the shop is open. If A has qualms on that front, it will not be reasonable for him to think B knows what she is doing. On the other hand, if B can produce a satisfactory account of how she knows the shop is open, this will provide reassurance that she (also, and connectedly) knows what she is doing. Admittedly, this way of putting things may be somewhat stronger than anything Anscombe says. As she characterises the ‘error of judgement’ in the shark case, the agent is intending to buy certain things even though she ‘might have known they were not to be had’ (I, 56). That does not imply that a flawless exercise of practical judgement would be informed by knowledge of the premises relied on. Still, I take it such a view is independently plausible,9 and it is certainly consistent with Anscombe’s account. ‘Errors of practical judgement’ may then be seen as falling into two kinds. One may have failed properly to utilise relevant considerations that were broadly speaking available to one (including things ‘one might have known’), and thus be guilty of reasoning poorly. Or one’s reasoning may have been ill-​informed through no fault of one’s own. Either way, one’s expression of intention would not amount to a justified statement of fact. I suggested that the participants’ perspective combines two concerns: a concern with understanding the informant’s epistemic position and a concern with the correctness of her claim to knowledge. The role sound calculative judgement plays in warranting a double-​aspect statement speaks to both concerns. If B’s intention reflects sound judgement, she is entitled to express that intention by saying something that is, amongst other things, a statement of fact—​a statement purporting to express knowledge. To see how this bears on A’s understanding of B’s epistemic position, it is instructive to reflect on two further elements of that understanding.

8 

9 

For a detailed exposition of this view of calculative practical reason, see Vogler (2002). For different but converging perspectives on this, see Hyman (1999); Hawthorne and Stanley (2008).

82   Johannes Roessler First, even an intention that is sound from the calculative point of view may fail to get executed, owing to an ‘error in performance.’ Of course, such an error may be easily corrected—​in Adrian Haddock’s (2011, 169) phrase, it may be a mere ‘hiccup or glitch.’ But if you tell us you are pressing button 1, yet out of carelessness or clumsiness press button 2, then—​supposing these buttons can be pressed only once—​your statement will be falsified by your error of performance. That is one reason why, in the case of intentions expressed by the use of the present progressive, it is important to keep an eye on ‘the material one is working on’ (I, 89). Sound practical judgement warrants a statement of fact, conditional on the absence of errors in performance (more serious than mere hiccups). Second, suppose an agent’s well-​formed intention and skilful performance are such as to warrant the double-​aspect statement ‘I am doing x.’ Is this enough to credit her with knowledge of doing x? Or is such knowledge to be attributed to her only if she actually reflects, in speech or thought, that she is doing x? I think neither suggestion is quite right. Consider someone who absent-​mindedly performs some routine task. Does she know what she is doing? Such cases elicit a range of intuitions, but I assume that at least sometimes, the natural verdict will be that she does. Her situation is not like that of someone who can easily learn something, though she does not yet know it; it is like that of someone who knows something, though she is not currently thinking about it. A related case is what (in ‘On Being in Good Faith’) Anscombe calls ‘knowledge without realisation.’10 In certain circumstances it may be correct to say that a reckless driver knows he is putting pedestrians at risk, without realising it. He may be quite correct to say ‘I did not think of that’ or ‘I did not think of it like that,’ but these statements ‘do not disprove knowledge’ (GG2, 105). On the other hand, suppose an agent is disabled from realising what she is doing. Suppose there are factors, such as motivation, or emotions, or a less than rational state of consciousness, that prevent her from thinking, or at least thinking clearly, about her action and its objective. She may thus, at least temporarily, be unable to express her intention in acting, even to herself. In such cases, it seems natural to say that the agent is not (or at least not properly) aware of what she is (intentionally) doing, as Anscombe seems to acknowledge.11 It is here that we arguably find room, within an Anscombean approach to intentional action, for the possibility of intentional actions that fail to satisfy what is sometimes called the ‘cognition condition’ on intentional action.12 In summary, while there can be practical knowledge without reflection/​realisation, possession of practical knowledge plausibly requires a (non-​incapacitated) ability to express the relevant intention. Putting all of this together, consider the following conditions: (i) B’s intention is informed by sound calculative practical judgement. (ii) B makes no (irredeemable) error in performance. (iii) B is able to express her intention by making a ‘double-​aspect’ statement. 10 

I am grateful to John Schwenkler for drawing my attention to this essay. ‘It is clear that, for any deed X, you cannot have intentionally done X unless you know you are doing X, except in a psychoanalytical sense in which there can be unconscious intentions’ (GG2, 104). 12  See Small (2012) for a recent discussion and defence of that condition. 11 

Practical Knowledge and Testimony    83 I have suggested that (i) and (ii) play a significant role in guiding A’s response to B’s act of telling him what she is or will be doing. It is not that A should be expected to have some independent way of ascertaining whether these conditions are satisfied. He may reasonably trust B’s testimony. Still, if there are grounds for suspicion, it may be reasonable for him to request reassurance, and the request will reflect his understanding of the connection between the two conditions and B’s claim to knowledge of what she is doing. The connection has both a normative and an explanatory dimension. The satisfaction of the two conditions would mean that B’s claim is unobjectionable and B may reasonably be credited with knowledge of what she is doing. The satisfaction of the two conditions also makes B’s possession of knowledge intelligible, at least by the lights of the participants’ perspective. The relevant kind of intelligibility is of course quite different from an account of how someone knows something by reference to her exploiting a way of finding out about the relevant fact. It must be different, since a way of finding out could warrant only a statement of fact, not a ‘double-​aspect’ statement. Rather, we might think of B’s sound practical judgement and dexterous performance as enabling conditions of her knowing what she is doing. If these conditions are satisfied, and if B is not disabled from expressing her intention, she will be in a position knowledgeably to reflect on what she is or will be doing, by expressing her intention.

4.6  What Is Wrong with the Contemplative Conception? I want to end by considering how my discussion of the participants’ perspective bears on one of Anscombe’s central doctrines, that ‘practical knowledge’ confutes modern philosophy’s ‘incorrigibly contemplative conception of knowledge.’ According to the ‘contemplative conception,’ ‘[k]‌nowledge must be something that is judged as such by being in accordance with the facts. The facts, reality, are prior, and dictate what is to be said, if it is knowledge’ (I, 57). Anscombe’s denial that the conception applies to knowledge of our intentional actions is often construed as a metaphysical claim. Practical knowledge, it is suggested, is not ‘a reality distinct from what is known’ (McDowell 2010, 432). It is ‘an aspect of the actuality of its objects’ (Haddock 2011, 163). On the face of it, though, the ‘contemplative conception’s’ central claim is a normative one—​a claim about how knowledge is to be ‘judged.’ Judging knowledge ‘as such,’ I take it, is a matter of judging whether a purported piece of knowledge really is knowledge, that is, whether a claim to knowledge is correct. But what does it mean to judge knowledge as such ‘by being in accordance with the facts’? Is it not obvious that the correctness of a claim to knowledge that p depends on whether it is a fact that p? I think the second clause of the contemplative conception makes it clear what Anscombe has in mind. The contemplative conception maintains not just that knowledge is factive

84   Johannes Roessler (something Anscombe does not deny) but that the facts enjoy a certain explanatory priority. The canonical way to judge a purported expression of knowledge that p is to determine whether in saying ‘p’ the subject is appropriately responsive or receptive to the ‘dictate’ of the facts. If S knows that p, there must be a way in which she is receptive to the fact that p. Only if there is some such way can there be a good reason for crediting S with knowledge. It is not entirely clear, of course, why Anscombe associates this view with ‘modern philosophy’ as a whole. (Kant springs to mind as a counterexample.) I think the best historical fit may be with a view commonly known as ‘Oxford realism.’ In Kant’s Theory of Knowledge (1909), H. A. Prichard wrote. ‘If there is to be knowledge, there must first be something to be known. In other words, knowledge is essentially discovery, or finding what already is.’13 A more recent statement of the view may be the suggestion that claims to knowledge are as such open to the question ‘How do you know?’ (Williamson 2000), supposing that ‘How do you know?’ asks for a way in which you were able to find out. In any case, the ‘contemplative conception’ is a thesis not primarily about the metaphysics of knowledge but about the kind of account of how we come to have knowledge that can properly underwrite claims to knowledge and so give us a reason to think that what we have is indeed knowledge. Of course, there are good reasons for knowledge attributions that shed no light on how the agent knows what she knows. A good reason to think S knows that p may be that everyone knows that p. The contemplative conception will also allow that forgetting how one discovered that p does not necessarily compromise one’s claim to knowledge that p. Still, on the contemplative conception, if you claim to know that p, it will always (in principle) be a good question how you discovered that p, and a successful defence of your claim will typically turn on the answer to that question. It is this condition that, so Anscombe insists, ‘practical knowledge’ fails to satisfy. We can divide Anscombe’s diagnosis into two parts. The first part is an account of the ways we ordinarily engage with claims to practical knowledge, the upshot of which is that the request for a way in which you were able to find out that you are buying butter would be in conflict with our conception of the kind of knowledge you are sharing with us in telling us ‘I am buying butter.’ Briefly, the correct way to test the epistemic credentials of your statement would be to ask questions that probe the soundness of the calculative judgement embodied in the intention expressed by your statement; it would not be to enquire into how you were able to find out what you are doing. The trouble with such an enquiry may be put in terms of an incompatibility between a practical and a theoretical ‘stance’ on the question whether one is (or will be) doing x. If you treat this as a matter to be determined by expressing your intention, you cannot simultaneously treat it as calling for the deployment of some way of finding out whether you will do x (see Moran 2001).

13 

Quoted after Marion (2000, 308), which provides an illuminating discussion of Oxford realism.

Practical Knowledge and Testimony    85 A contemplative theorist might respond to this by conceding that her view is in a sense revisionary. Perhaps the way we ordinarily make sense of our knowledge of what we are intentionally doing does not invoke any way of finding out. But one might argue that from the perspective of a philosophical understanding of such knowledge we can and should nevertheless insist on the indispensability of that sort of explanation, and we should try to construct one.14 A common way to try to incorporate knowledge of our intentional actions into a ‘contemplative’ epistemology is to invoke evidence or inferences that (a) are supposed to make such knowledge philosophically intelligible as the product of some appropriate way of finding out, but (b) whose operation is supposed to be remote from the agent’s awareness.15 In the light of (b) it is thought to be unsurprising that we ordinarily take such knowledge to be groundless. But acknowledging that point, it is argued, is compatible with insisting that only suitable evidence can make our knowledge (philosophically) intelligible, and claims to knowledge (philosophically) defensible. I think Anscombe’s reaction—​this is what I call the second part of her diagnosis—​ would be that the strategy of insulating a ‘contemplative’ account from our ordinary practice faces a significant hurdle. The strategy would be committed to denying that the way we ordinarily make sense of practical knowledge, in the light of soundly reasoned intentions and our ability to express them in a certain way, provides a fully satisfactory reason to think that we do know what we are or will be doing. It is dissatisfaction with our naïve picture of practical knowledge that motivates the quest for some way in which we might be seen (from the vantage point of a philosophical theory) to discover what we are doing, by being sensitive to suitable evidence. The question is whether rejection or suspension of our naïve picture is compatible with acknowledging that the activities under consideration—​the activities that form the subject matter of the knowledge we are trying to understand—​are intentional activities. (At this point, a broadly metaphysical thesis about the relation between intentional action and practical knowledge comes into play—​but I think it is a fairly weak one.) Briefly, the hurdle facing the revisionary strategy arises from internal connections that are central to Anscombe’s analysis, between intentional action, knowledge of what one is doing, and the participants’ perspective on knowledge of what one is doing. The connection between the first two is familiar: intentional actions are open to second-​ person ‘reason-​seeking’ questions, questions whose applicability presupposes that the agent knows without observation what she is doing.16 The connection with the participants’ perspective may be less obvious, but I think it is implicit in the qualification of the required knowledge as non-​observational. The idea is not (just) that intentional action involves knowledge that, from the perspective of epistemological theorising, can be seen to be non-​observational. One way to bring this out is to consider a case in which A and B treat B’s knowledge that she is buying butter as observational. They regard ‘How 14 

See Velleman (1989) for a response along these lines. For critical discussion, see Roessler (2013). For different versions of this move, see Velleman (1989) and Paul (2009). 16  At least in the case of agents endowed with language: see Anscombe’s brief discussion of our disposition to describe some species of non-​human animals as acting intentionally (I, 86–​89). 15 

86   Johannes Roessler do you know you are buying butter?’ as a fitting question, and B’s reply is, say, ‘I found some butter in my shopping basket.’ On Anscombe’s view, this would be enough to show that the activity under consideration is not intentional under that description. ‘Without observation’ is meant to capture the way A and B need to think about B’s knowledge in sharing it, if it is to be the kind of knowledge we have of what we are intentionally doing. Acting intentionally, as we ordinarily conceive it, then, is inseparable from the capacity to know what one is doing‚ as we ordinarily conceive it. If that analysis is correct, it raises the question whether we could coherently embrace the ‘contemplative conception.’17 The question merits more detailed investigation than I can offer here. In particular, the possibility that I think would deserve further examination is that the revisionary strategy would commit us, as philosophers, to rejecting or suspending explanations and validations of knowledge we are committed to accepting, as reflective agents and participants in the practice of sharing knowledge of what we are doing.18

References Anscombe, Elizabeth. 1957. Intention. Oxford: Blackwell. Anscombe, Elizabeth. 2017. ‘On Being in Good Faith.’ In Faith in a Hard Ground: Essays on Religion, Philosophy and Ethics by G. E. M. Anscombe. Edited by M. Geach and L. Gormally. Exeter: Imprint Academic. Falvey, Kevin. 2000. ‘Knowledge in Intention.’ Philosophical Studies 99: 21–​44. Faulkner, Paul. 2011. Knowledge on Trust. Oxford: Oxford University Press. Haddock, Adrian. 2011. ‘‘The Knowledge That a Man Has of His Intentional Actions.’’ In Essays on Anscombe’s Intention, edited by A. Ford, J. Hornsby, and F. Stoutland, 147–​169. Cambridge, MA: Harvard University Press. Hampshire, Stuart. 1965. Freedom of the Individual. London: Chatto and Windus. Hawthorne, John, and Jason Stanley. 2008. ‘Knowledge and Action.’ Journal of Philosophy 105(10): 571–​590. Hursthouse, Rosalind. 2000. ‘Intention.’ Royal Institute of Philosophy Supplements 46: 83–​105. Hyman, John. 1999. ‘How Knowledge Works.’ Philosophical Quarterly 49: 433–​451. Longworth, Guy. 2019. ‘Sharing Non-​observational Knowledge.’ Inquiry, November 15, 2019. https://​www.tand​fonl​ine.com/​doi/​abs/​10.1080/​00201​74X.2019.1680​430. Marion, Mathieu. 2000. ‘Oxford Realism: Knowledge and Perception I.’ British Journal for the History of Philosophy 8(2): 299–​338. McDowell, John. 2010. ‘What Is the Content of an Intention in Action?’ Ratio 23(4): 415–​432. McMyler, Benjamin. 2011. Testimony, Trust, and Authority. Oxford: Oxford University Press. 17 

This way of putting the question is intended to be reminiscent of the sort of dialectic Barry Stroud (2011) pursues in Engagement and Metaphysical Dissatisfaction. See in particular his suggestion that it is hard to find ‘enough distance [enough for the purpose of detached metaphysical theorising] between our conception of the world and the world it is meant to be a conception of,’ given our ‘unavoidable immersion’ in that conception (145). 18  A previous version of this chapter was presented at a workshop on practical knowledge at Fribourg in 2019. I would like to thank the participants for discussion. I am particularly grateful to the editor for his helpful comments.

Practical Knowledge and Testimony    87 Moran, Richard. 2001. Authority and Estrangement. Princeton, NJ: Princeton University Press. Moran, Richard. 2018. The Exchange of Words. Oxford: Oxford University Press. Paul, Sarah. 2009. ‘How We Know What We’re Doing.’ Philosophers’ Imprint 9: 1–​24. Roessler, Johannes. 2013. ‘The Epistemic Role of Intentions.’ Proceedings of the Aristotelian Society 113(1): 41–​56. Small, Will. 2012. ‘Practical Knowledge and the Structure of Action.’ In Rethinking Epistemology, edited by G. Abel and J. Conant, vol. 2, 133–​227. Berlin: De Gruyter. Stroud, Barry. 2011. Engagement and Metaphysical Dissatisfaction. Oxford: Oxford University Press. Teichmann, Roger. 2008. The Philosophy of Elizabeth Anscombe. Oxford: Oxford University Press. Thompson, Michael. 2011. ‘Anscombe’s Intention and Practical Knowledge.’ In Essays on Anscombe’s Intention, edited by A. Ford, J. Hornsby, and F. Stoutland, 198–​210. Cambridge, MA: Harvard University Press. Velleman, David. 1989. Practical Reflection. Princeton, NJ: Princeton University Press. Vogler, Candice. 2002. Reasonably Vicious. Cambridge, MA: Harvard University Press. Williamson, Timothy. 2000. Knowledge and Its Limits. Oxford: Oxford University Press. Wolfson, Ben. 2012. ‘Agential Knowledge, Action and Process.’ Theory 78: 326–​357.

Pa rt I I

E T H IC A L T H E ORY

Chapter 5

An sc om be’s Th re e T h e se s a fter Sixt y Ye a rs Modern Moral Philosophy, Polemic, and ‘Modern Moral Philosophy’ Sophie Grace Chappell

5.1 Anscombe famously begins her 1958 polemic ‘Modern Moral Philosophy’ (MMP) by presenting—​her word—​three theses. MMP1: We should stop doing moral philosophy, ‘at any rate until we have an adequate philosophy of psychology, in which we are conspicuously lacking’. MMP2: We ought, ‘if this is psychologically possible’, to jettison all the following concepts (at least these five, and by implication, all others sufficiently closely associated with them): moral obligation, moral duty, the morally right, the morally wrong, and the moral ought. MMP3: ‘The differences between the well-​known English writers on moral philosophy from Sidgwick to the present day are of little importance. (MMP, 1)

My question in this essay is how MMP1–​3 have held up in the sixty years since Anscombe published ‘Modern Moral Philosophy’.1 One upshot will be exegetical: it is MMP3 that

1  Thanks for comments and criticisms to Constantine Sandis and Roger Teichmann, and to the audience at the Iris Murdoch on Moral Perception conference in The Queen’s College, Oxford, April 2018, on whom I first tried out the material on salience in this chapter. Also to my Open University colleagues at our research conference at Chicheley Hall, Buckinghamshire, in May 2018. This chapter was written during my tenure of a Leverhulme Major Research Fellowship, 2017–​2020; I am happy to acknowledge the Leverhulme Trust’s invaluable support.

92   Sophie Grace Chappell bears the main emphasis in MMP. So in this discussion MMP3 will come first and get the lion’s share of space; I won’t even get to MMP1 and MMP2 till section 5.7. Contrary to what is probably the majority reading of MMP (see, e.g., Crisp 2004), I see MMP1 and MMP2 as little more than auxiliary theses, and MMP3 as—​to echo the essay’s very last sentence—​her main complaint.2 The main upshot of my enquiry will be an answer to the question how MMP1–​3 have held up since 1958. My answer will be that they have held up remarkably well. That is not quite the same as saying that they (still) look plausible or right. MMP was first published sixty years ago, and even an exegetical claim like MMP3 can have a best-​before date. MMP3 is a claim about similarities; similarity claims depend for their plausibility on what features of the comparanda are salient; what is salient depends on time and place and audience. So MMP3 might have been ‘truer’—​it might have been a more penetrating, apt, and revealing observation—​in 1958 in Oxford than it is now in, say, Dundee or New York or Addis Ababa. As for MMP2 and MMP1, they are recommendations. So they might be wiser counsel to Anscombe’s own contemporaries in 1958 than to us today in 2018. In fact, I suggest, we need Anscombe’s advice less now because it worked then; though it is still salutary to be reminded of it. Another reason why there is a significant difference between saying that MMP1–​3 are simply right, and saying more guardedly that they have held up well, is that all three are offered in the course of what I have already called a polemic. Meaning what?

5.2 ‘Polemic’ and ‘polemical’ are often used dismissively: ‘X is being polemical’ is said to imply ‘X is not doing philosophy properly’ and/​or ‘We needn’t listen to X’s argument, if any’.3 And often enough, quite fairly. But I do not here use ‘polemic’ in this dismissive and exclusionary way. I want at least to consider the possibility that writing that is admittedly polemical might be other philosophically interesting things as well as polemical; that at least sometimes polemic might be philosophically fruitful; and so that polemic might be consistent with ‘doing philosophy properly’. For my purposes here, polemic is a philosophical discourse driven not by curiosity but by anger—​originally the anger of war: Greek polemos. A polemic is focused not so much on dialogical exploration of positions that, for all we can tell at the outset, may well be right, as on diagnosis, denunciation, and destruction of positions that we know in advance are fundamentally wrong. For my purposes here the polar opposite of a polemic, though not the only alternative to it, is a peirastic, a trying-​out, a tentative exploration. 2  Roger Crisp, ‘Does Modern Moral Philosophy Rest on a Mistake?,’ Royal Institute of Philosophy Supplement 54 (2004): 75–​93. 3  Cristina Chimisso reminds me that in recent French philosophy, e.g., in Bachelard, polémique can signify a philosophy of rational resistance, guerrilla warfare against a dominating ideology.

Anscombe’s Three Theses after Sixty Years    93 So far as I know, ‘peirastic’ as I am using it here, in contrast and opposition to ‘polemic’, is—​as an English word—​a neologism of mine. I borrow the word from the Hellenistic-​period editors of Plato, who classified his works in a number of categories. They counted, e.g., Meno, Theaetetus, Ion, Charmides, and Euthyphro as peirastic dialogues. Amongst their other classifications were ethical, to do with character and/​ or action (Apology, Crito, Phaedo, Phaedrus), apotreptic, ‘dissuasive’ (Gorgias, Hippias Major, Euthydemus), endeictic, ‘accusatory’ (Protagoras), political (Laws, Republic), logical (Statesman, Sophist, Cratylus, Parmenides) and maeutic, to do with midwifery (Alcibiades I, Lysis, Laches—​it is interesting that Theaetetus is not in this last category).4 This evidence suggests that the ancient Platonic editors were open to a philosophically important possibility. This is that different works of philosophy, even when they are by the same author and mostly share the same ostensible form, may be importantly different kinds of speech-​act and intended to be taken in different ways by their readers. So with ‘peirastic’ and ‘polemic’. These, along with numerous other possibilities, including those listed by the ancient Platonists, are two distinct and different styles or modes of philosophical writing, two different possible tones of voice, in which an author might address her readership. There is no need to be rigidly schematic with such classifications (as the ancient editors were rather prone to be) or to suppose that polemic and peirastic are mutually exclusive—​any more than a tone of voice cannot be both sarcastic and playful or resigned and angry. Polemic is no less an exact craft than peirastic—​it takes considerable skill to do either right. Polemical skill is evinced, for example, in some of the best pages of Plato, Augustine, Voltaire, Carlyle, Kierkegaard, Nietzsche, Russell, and Simone de Beauvoir. But polemic, unlike peirastic, does lend itself readily to hyperbole, to overstatement of the literal truth for the sake of some deeper-​than-​literal truth that is got at either in the way that caricature and satire get at the truth (by telling exaggeration of what is really there, in order that what is really there may stand out from its background), or in the way that studiedly provocative statements get at the truth (by eliciting a reaction that, when dialectically combined with the original provocation, helps us towards a deeper understanding). As a method of philosophy, polemic’s loose way with the truth is one of its main features, and also one of its main faults. (It has plenty of other faults; more about them shortly.) Anscombe is in general most fastidious about the truth. Yet it seems obvious (argument coming in a moment) that there is deliberate hyperbole in at least much of the text of MMP. So we make better sense of at least much of MMP by reading it as polemic rather than as some more down-​to-​earth and factual kind of speech-​act, such as the straightforward (and perhaps rather humdrum) assertion characteristic 4 

Some dialogues apparently remain unclassified in this scheme: Menexenus, Symposium, Timaeus, Critias. NB: no Platonic dialogue is classified by the ancient editors as polemic. ‘Polemic,’ according to the OED, comes from church controversies in Milton’s time, where its natural opposite is ‘eirenic.’ Nor did the ancient editors classify any of Plato’s works as aporetic—​that is a nineteenth-​century Plato scholars’ word. (Thanks to Stephen Halliwell.)

94   Sophie Grace Chappell of most philosophy journal articles in 2018. MMP is in this respect like various other publications of Anscombe’s that we certainly do better to take not as straightforward assertion but as sarcasm or irony or satire. One case in point, close to MMP in theme and time and tone, is ‘Does Oxford Philosophy Corrupt Youth?’5 Consider also Anscombe and Joe Feldman, ‘On the Nature of Justice in a Trial’,6 a piece which Mary Geach describes as ‘heavily satirical’;7 also, the withering brevity of Anscombe’s ‘A Note on Mr Bennett’.8 One small example of polemical hyperbole in MMP is Anscombe’s single sentence (p. 3) about Bishop Butler:9 ‘Butler exalts conscience, but appears ignorant that a man’s conscience can tell him to do the vilest things’. This is so quick, rude, and high-​handed that it is funny—​at any rate it makes me chuckle—​and (as I shall argue in a moment) it is pretty clearly deliberate hyperbole. Anscombe is being rude and high-​handed about Butler because he is not her main concern. For a start she is more interested in Hume, though Hume is not her main target either, and she is rude about him too. She also wants to warn us, first, that Butler’s view of ethics has a flaw that she takes to be critically important (its dangerous openness, as she sees it, to ‘vile’ actions); and second, that she will be prepared to give equally summary treatment to the main targets of her polemic when she gets to them, and is challenging her readers to find a reason why such swift brutality will not be entirely deserved. So Anscombe’s sixteen-​word takedown of Butler makes a kind of sense read as hyperbole, as deliberately provocative over-​statement. But Anscombe cannot seriously and literally think that there is no more than her one sentence to be said about Butler—​not even in the way of objections to Butler, not even in the way of points about Butler that are relevant to getting clear about modern obligation-​talk. For in fact—​and here we really might wish that Anscombe had gone into the matter more peirastically—​Butler is an extremely interesting forerunner of such talk. He sees the demands of conscience as having something very like a categorical status, which, in his Pauline words (Romans 2:14), makes us a ‘law unto ourselves’.10 Again, Anscombe’s more specific charge is that Butler tells us to follow conscience even though conscience can be wildly mistaken. But first, so does Aquinas.11 And second, Butler in the Sermons runs a systematic analogy between conscience in the individual and authority in the state. The question ‘But what is 5 

G. E. M. Anscombe, ‘Does Oxford Philosophy Corrupt Youth?’, The Listener, February 14, 1957, 266–​271. 6  G. E. M. Anscombe and Joe Feldman, ‘On the Nature of Justice in a Trial’, Analysis 35, no. 2 (1972): 33–​36. 7  Mary Geach, l.c. xvii. 8  G. E. M. Anscombe, ‘A Note on Mr Bennett’, Analysis 26, no. 6 (1966): 208. 9  As she would certainly not have been prepared to call him. Even in these ecumenical days, the online Catholic Encyclopaedia still takes the decidedly hard-​line view that the archbishopric of Canterbury has been vacant since the death of Cardinal Pole in 1538: https://​www.newadv​ent.org/​cat​ hen/​032​99b.htm. No doubt Anscombe, as a traditionalist Catholic writing in the 1950s, regarded Joseph Butler’s ostensible occupancy of the diocese of Durham as an equal imposture. 10  See Butler, Sermon 3. 11  Thomas Aquinas, Summa Theologica 1a2ae.19.5–​6.

Anscombe’s Three Theses after Sixty Years    95 the voice of conscience, here and now?’ is as perennially apt, and as audible to Aquinas and Butler alike, as the question ‘But what is political authority, here and now?’ And no form of the view that either conscience or political authority12 is infallible commits us to any form of the view that either is automatically identifiable.13 The case of Butler, then, is a small example of polemic, and polemical hyperbole, in MMP. What is a large example? My answer is: the whole thing. The entire essay is shot through with polemic, and with polemical hyperbole. It is not for nothing that Anscombe begins MMP by saying that, in it, she presents three theses. ‘Presents’ is her word, not ‘argues for’. That is not a sign that there is no argument in MMP. But it is a sign of what often predominates, namely polemic, and often polemical hyperbole. Not that every single claim that Anscombe makes in MMP is polemic, and therefore cannot be assessed for sober literal truth or argumentative form; indeed, by the end of this chapter, I will have rather moved away from considering its polemical aspects and will be focusing much more on its serious historical and factual claims. As already pointed out, writing that is polemical can be other things as well, including assertoric (and even peirastic). But a polemical tone is pervasive, and hyperbole is frequent, in MMP. It is crucial to bear in mind this stylistic feature when attempting to assess the content of MMP. We academic philosophers today are much too apt to treat every and any philosophical text alike: with plodding literalism, as a list of asserted propositions that stand, or so one hopes, in various kinds of relevance and inferential relations to each other, and that might be minimally represented as a formal proof in logical notation, with the rule of inference from preceding steps specified, for each move, in brackets in the margin. Despite the enormous sociological, institutional, and structural pressures upon us to model our practice upon the practice of our university colleagues who work in the scientific and mathematical disciplines, that is not the only or necessarily the best way to read or write even texts in maths and natural science, let alone any other kind of text. In the humanities in general and in ethical philosophy in particular, precisely because these enquiries do not or at least should not aim at impersonal neutrality or objectivity, the distinctive voice, tone, style, and standpoint of the individual researcher is an essential part of what she has to say. And what she has to say may not always be straightforward factual assertion or parade-​ground-​logical argument. It can be various other things too. One of them, as in MMP, is polemic. The dangers of polemical philosophy are self-​evident and uncontroversial. There is a danger of straw-​manning, a danger that an uncharitable interpretation of one’s interlocutor’s position will be an inaccurate one. There is a danger of ad hominem irrelevance, of ‘going for the player not the ball’. There is a general danger of being, as they say, ‘carried away by one’s own emotions’, that is, of being persuaded not by arguments but by irrelevant emotional effects, and/​or of producing such malpersuasion in one’s hearers. 12 

Or, come to that, papal authority. On the infallibility of political authority as such, see the marvellous—​and decidedly polemical—​ opening chapter of Hilaire Belloc, The French Revolution (1911). 13 

96   Sophie Grace Chappell And there are specific dangers in the particular kinds of emotions that polemic characteristically produces, such as anger, scorn, abhorrence, and contempt, none of which is a passion that tends readily towards sweetness, and perhaps not much towards light either. (A little polemic can be good entertainment, but doing philosophy would not be much fun if polemic was the normal mode of philosophical exchange, as it is of Twitter exchange.) Again, there is a danger of abusive power-​relations between yourself and your interlocutor and/​or audience, for example, those involved in browbeating, bullying, and outgrouping. Most people reading this will, unfortunately, almost certainly have been in seminars where they have seen well-​paid, job-​secure senior philosophers (often male ones) using their status, and perceptions of their status, to deflect or deflate questions or objections put to them by badly paid, job-​insecure junior philosophers (often female ones). Certainly, then, polemic has its dangers. Yet perhaps there is a danger, too, in thinking that just every interlocutor can be appropriately dealt with by non-​polemical philosophy. Maybe sometimes polemic is deserved. Whether or not that is right, Anscombe clearly thought so.

5.3 I have said that polemic is philosophical discourse driven, not by curiosity, as peirastic is, but by anger. We have it on Anscombe’s daughter’s authority that Anscombe wrote MMP in a white heat of passion. While Philippa Foot was on research leave at Berkeley in California—​perhaps in Trinity Term 1957?—​her Somerville colleague Anscombe took over Foot’s undergraduate tutorial teaching in moral philosophy. So ‘my mother settled down to read the standard modern ethicists and was appalled’.14 What were the features that appalled Anscombe? I believe they were exactly those mentioned in MMP1–​3. Most appalling to her of all, perhaps, was the similarity noted in MMP3: The overall similarity is made clear if you consider that every one of the best known English moral philosophers [since Sidgwick]15 has put out a philosophy according to which, e.g., it is not possible to hold that it cannot be right to kill the innocent as a means to any end whatsoever and that anyone who thinks otherwise is in error. . . .16 Now this is a significant thing: for it means that all these philosophies are quite incompatible with the Hebrew-​Christian ethic. . . . The prohibition of certain things 14 

Mary Geach, introduction to HLAE, xvii. That this restriction is intended becomes clear on the next page of MMP. 16  In the passage omitted here, Anscombe notes that R. M. Hare is, strictly, an exception from her generalisation. In Moral Thinking and elsewhere Hare says explicitly that no choice of principles is as such either correct or erroneous, though some choices Hare calls ‘fanatical’ (Freedom and Reason, 105)—​ a term that he means pejoratively, though it was a term of praise amongst the Nazis. 15 

Anscombe’s Three Theses after Sixty Years    97 simply in virtue of their description as such-​and-​such identifiable kinds of action, regardless of any further consequences, is certainly not the whole of the Hebrew-​ Christian ethic; but it is a noteworthy feature of it; and if every academic philosopher since Sidgwick has written in such a way as to exclude this ethic, it would argue a certain provinciality of mind not to see this incompatibility as the most important fact about these philosophers, and the differences between them as somewhat trifling by comparison. (MMP, 17–​18)

‘Provinciality’ is a notable word here; no doubt, for Anscombe, the opposite of ‘provinciality’ is ‘catholicity’. Seen from her catholic, and indeed Catholic, viewpoint, the English moral philosophers from Sidgwick on are ‘all the same’ because they all deny or disregard a thesis central to the Judaeo-Christian ethic: the thesis of moral absolutes, the thesis that some action-​types are forbidden just as such—​where forbidden means literally forbidden, forbidden by divine command. Four questions arise here. First, does Anscombe’s claim really apply to all ‘of the best known English moral philosophers since Sidgwick’? Not literally all: Anscombe herself was well enough known to be one exception, and Geach another. Iris Murdoch, Philippa Foot, Bernard Williams, Stuart Hampshire, and Isaiah Berlin were all well-​known moral philosophers, English at least by adoption, whose desks were within twenty minutes’ walk of Anscombe’s desk when she wrote these words. So her sweeping ‘all’ looks like hyperbole. We know who she has in her sights: Hare, Nowell-​Smith, G. E. Moore, Harold Prichard, Green, Rashdall, Spencer, Ross, Carritt, Ewing, Broad, and the like. But what she actually says is too strong. Second, is Anscombe right about ‘the Hebrew-​Christian ethic’? Does it, indeed, affirm moral absolutes based, as she clarifies elsewhere, on divine commands? Plenty of moral theologians thought otherwise (see Pink 2004: 183 for a reminder that there were mediaeval opponents of divine command ethics, as well as proponents); it is a significant exegetical task in biblical theology to prove the point either way. Certainly some arguments against absolutism rest merely on misunderstandings, e.g., of what ‘innocent’ means in ‘No Intentional Killing of the Innocent’. (Here ‘innocent’ does not, as sometimes, mean either ‘young and naïve’ or ‘would be acquitted in a court of law’. It means what its Latin original in-​nocens means: ‘a non-​aggressor’. In this sense enemy soldiers are, pace, e.g., Crisp [2004: 76], paradigm non-​innocents.) But there are good arguments too for a non-​absolutist version of the Hebrew-​Christian ethic, starting, e.g., from Rabbi Hillel’s or Jesus’s ‘summary of the law’ (Matthew 22:37–​40), which makes two and only two principles foundational: loving God and loving your neighbour. For Hillel or Jesus, if there are absolute prohibitions, this will require further argument, perhaps, e.g., to show that observing such principles is (partly) constitutive of loving God and your neighbour. I think further argument to exactly that effect can be given. The point remains that the absolutism of the Judaeo-​Christian ethic is not as completely obvious and indisputable as Anscombe makes it look. Third, what would modern moral philosophers look like if they were not, in Anscombe’s sense, ‘all the same’? Well, obviously, they wouldn’t all reject moral absolutes

98   Sophie Grace Chappell (either as a matter of course or after careful thought); at least some of them (besides Anscombe herself) would defend the existence of at least some moral absolutes. They might then engage in debates about which the moral absolutes were and why, and where the boundaries of those absolutes lay and why; such debates would, I think, necessarily get them into the territory that Anscombe labels ‘philosophy of psychology’, into questions about the nature of the virtues and of the good for human beings, and so (I will suggest in section 5.7) into ways of circumventing the ban on ‘moral philosophy’ that Anscombe enounces in MMP1. Debates between them might, in fact, look rather like the kind of debates that late antiquity saw, between Christians, Jews, Stoics, and others—​though, we might hope, with less polemic than those debates (and indeed less polemic than Anscombe). Fourth, why do ‘the modern moralists’ reject moral absolutes? Because they are all, in Anscombe’s sense of a term that she apparently17 invents right here, consequentialists: they all follow Sidgwick in believing that there is ‘no distinction between foreseen and intended consequences, so far as responsibility is concerned’.18 This view about the wide scope of responsibility is what Anscombe primarily means by ‘consequentialism’ (and so where it needs a label, I shall call this view ‘Anscombe-​ consequentialism’). The word is Anscombe’s, but it has since standardly been given a rather different sense from Anscombe’s, to mean ‘someone who holds that any action is right iff it produces the best consequences’. Anscombe thinks, rightly, that this consequentialism (I’ll just call it ‘consequentialism’, or if its difference from Anscombe-​consequentialism needs special emphasis, ‘standard consequentialism’) comes rather naturally in the wake of Anscombe-​consequentialism. She may also think, wrongly, that standard consequentialism is logically entailed by Anscombe-​ consequentialism: ‘once [someone] has started to look at the matter in this light [viz., the light of Sidgwick’s view about responsibility], the only reasonable thing for him to consider will be the consequences and not the intrinsic badness of this or that action’ (MMP, 19, my italics). Now certainly this accurately tracks the dialectic as it actually developed between Sidgwick and Moore, and certainly standard consequentialism is one reasonable direction for someone to go, starting from Anscombe-​consequentialism, in developing a view about the moral assessment of possible actions. But there are others. Here is one 17 

But see now Berkman. Perhaps Sidgwick is merely following a path already sketched out by John Stuart Mill. Certainly Utilitarianism espouses what we now call rule utilitarianism rather than act, but on the other hand the System of Logic (bk. 3, ch. 5) defends a universal view of causation that, at the very least, goes naturally with a universal view of responsibility. ‘[C]‌auses of all descriptions [prevent] the effects of other causes by virtue (for the most part) of the same laws according to which they produce their own[. This] enables us, by establishing the general axiom that all causes are liable to be counteracted in their effects by one another, to dispense with the notion of negative conditions entirely, and limit the notion of cause to the assemblage of the positive conditions of the phenomenon.’ My italics mark Mill’s erasure of negative conditions as a distinct category, which seems like a significant step towards the denial of any distinct category of negative responsibility. 18 

Anscombe’s Three Theses after Sixty Years    99 way to hold that there is ‘no distinction between foreseen and intended consequences, so far as responsibility is concerned’, without assessing actions as the consequentialist does, only with respect to their consequences: you might think that both intended and merely foreseen consequences matter morally to the same degree, N, but that N is low, and that there is some other consideration that where present always trumps consequences of either kind. N might even be zero—​though the resulting view of the moral assessment of possible actions is a pretty crazy one. The trumping consideration might even be the intrinsic goodness or badness of the actions proposed—​and it is not obvious this time that the resulting view is crazy at all. So Anscombe-​consequentialism does not entail standard consequentialism: it is neither necessary nor sufficient for it, though the move from the one to the other is natural, and Anscombe is surely right that it was made. For Anscombe-​consequentialists (says Anscombe), if someone is given a choice between doing something disgraceful and being imprisoned with the side effect that it will then be impossible for him to look after his children, then since this latter consequence is just as much his responsibility as any other, he has to take it just as much into account as any other. So, very quickly—​maybe, as Anscombe perhaps suggests, inevitably—​he ends up considering only consequences, and the only question for deliberation is the question of how these consequences are to be evaluated. But this move to standard consequentialism is, in Anscombe’s view, morally disastrous: By [Anscombe-​consequentialism], the kind of consideration that would formerly have been regarded as a temptation, the kind of consideration urged on men by wives and flattering friends, was given a status by moral philosophers in their theories. (MMP, 20)

At first sight this looks polemical indeed; actually it looks like mere old-​fashioned sexist harrumphing. But if you compare Crito 44e and Phaedo 60a, to which I suspect Anscombe is alluding, then you may be inclined, at any rate, to delete the ‘mere’. Anscombe means perhaps that it would have been weakness and/​or a yielding to flattery in Socrates to consider the side effects, for his family and friends, of his own preparedness to face execution on a monstrously unjust but legally proper conviction, or at any rate to consider them as of equal importance with the intended effects or character of his action. Considering that sort of thought, in that sort of way, is, Anscombe says, a temptation. As she sees it, the trouble with modern moral philosophy’s pervasive Anscombe-​consequentialism is that it leads us straight into temptation. Hence her final verdict on Anscombe-​consequentialism, in which she enounces her notorious refusal to argue with its adherents: [I]‌f someone really thinks, in advance, that it is open to question whether such an action as procuring the judicial execution of the innocent should be quite excluded from consideration—​I do not want to argue with him: he shows a corrupt mind. (MMP, 29)

100   Sophie Grace Chappell

5.4 How might Anscombe justify this refusal to engage? Well, first, here too there is a whiff of polemical hyperbole. Despite appearances, Anscombe cannot mean a literal, general refusal when she says ‘I do not want to argue’. For she did in fact continue to argue—​albeit often polemically—​with people who ‘really think’ this way; witness her further debates with Jonathan Bennett, John Harris, Glanville Williams, and Nowell-​Smith and Hare themselves. Also, there is a clear sense in which she herself thinks, in advance, that it is ‘open to question’ whether actions of types absolutely forbidden by the Judaeo-​Christian ethic ‘should be quite excluded from consideration’. She must think so; for she herself raises that question. Of course her own answer to it is emphatically that such action-​types should be excluded. And of course she really does think that philosophers ‘show a corrupt mind’ if they propose a model for practical deliberation in which the absolutely forbidden action-​types are allowed within deliberation’s scope ‘in advance’, i.e., from the beginning and by the model’s design. (As opposed to such action-​types’ popping up in deliberation in real time, in which case the agent is, as Anscombe says, ‘just normally tempted’.) But MMP itself opens up the question of how a model for practical deliberation should be designed. If that question is open, then so is the question of what ‘should be quite excluded from consideration’. Anscombe’s position has the paradoxical feature of forcing her to give consideration to the question what things should never be given consideration, which is itself, in a sense (not of course the most direct one), a way of considering the things that should never be given consideration. What, anyway, is so bad, so ‘corrupt-​minded’, about designing a model of deliberation so that it allows absolutely forbidden action-​types within deliberation’s scope? Why does Anscombe (at least ostensibly) refuse even to engage with those who offer such models? It seems that a consequentialist might retort to her that, in the end, it doesn’t matter if we do deliberate over bad action-​types, provided we don’t actually choose them. If they are really as bad as she thinks, then they won’t ever win out in our deliberations. So (it might be said) we should reserve our denunciations for those for whom these bad alternatives are the output of their deliberations, not one input. Really? Consider here a well-​known remark of Bernard Williams’s, which Williams himself used to say was suggested to him by Peter Goldie. This too has the mildly paradoxical feature of being deliberative words about what Williams calls ‘deliberative silence’: An effective way for actions to be ruled out is that they never come into thought at all, and this is often the best way. One does not feel easy with the man who in the course of a discussion of how to deal with political or business rivals says, ‘Of course, we could have them killed, but we should lay that aside right from the beginning’. It should never have come into his hands to be laid aside. It is characteristic of [the] morality [system] that it tends to overlook the possibility that some concerns are best embodied in this way, in deliberative silence.19 19 

Bernard Williams, Ethics and the Limits of Philosophy, 206. Cf. John McDowell, ‘Virtue and Reason,’ and Jonathan Dancy, Moral Reasons (OUP, 1993), on ‘silencing.’

Anscombe’s Three Theses after Sixty Years    101 Suppose the best thing that you and I could possibly do this afternoon is go for a walk in the park, and you ask me ‘What shall we do this afternoon?’ ‘Well’, I musingly reply, ‘we could burn some puppies to death; or we could drown Grandma in the creek; or we could go into town with dirty needles and infect strangers with AIDS; or we could blow ourselves up committing a terrorist atrocity; or we could go for a walk in the park’. Being good at calculating consequences, we see straight off that going for a walk in the park is our best option from these alternatives. So we do that, and (by hypothesis) we thereby maximise the good. By the argument just given, there is absolutely nothing wrong with our deliberation. By that argument, in fact, there would be absolutely nothing wrong with me as a deliberator if stuff like this happened all the time. Suppose I always consider puppy-​ burning as one of my alternatives, in every piece of deliberation that I ever do. If no puppies, matches, and petrol are readily available, I of course think about how to procure some; when I am making what for most people is a two-​way choice in the Breakfast Preserves aisle, the option of nipping out of the supermarket to torch a few baby spaniels is always there for me, alongside jam and marmalade, as a third possibility. And suppose that I share all my deliberating with you. Is it not even slightly likely that you will start to think that I am not quite a normal deliberator? Is there not even a small chance that you will be at first uneasy in my company—​wondering if I am labouring a bad-​taste running joke or something—​then increasingly angry with me, and finally appalled at what you have come to think is a none too well buried sadistic obsession of mine?20 The consequentialist may say ‘Relax, everything’s fine. She chooses jam and the walk in the park and the rest of it. She never actually burns any puppies. And puppy-​ burning is a really bad option, so pretty certainly she never will’. Is that how a normal person will respond? I don’t think so. What a normal person will say, if I keep banging on about burning puppies sufficiently insistently—​and even if I never actually burn any—​is ‘Enough. This is disgusting. I don’t wish to spend time with her. She shows a sick mind’. We need not agree with Anscombe’s own specific list of absolutely forbidden options, which includes ‘sodomy’, whatever that is,21 to get the deeper point that our character is shown in what alternatives we take seriously, and that even to deliberate about some options is indeed, as she says, evidence of a corrupt mind. It is not actually terribly nice to be made to consider the option of puppy-​burning, and I apologise to my readers for forcing it on them. It is, alas, obvious that I could easily have picked

20 

Dear reader: Don’t worry. The author is exceedingly fond of dogs, including infant ones, particularly spaniels, and except in philosophical thought-​experiments, she never ever sets even small numbers of them even slightly on fire. Promise. 21  By ‘sodomy’ Anscombe appears to mean not what the English law used to mean by it, i.e., what it also called ‘buggery,’ i.e., anal sex. Rather she apparently means what her church’s tradition also calls ‘unnatural sexual relations,’ of which it charmingly takes any kind of homosexual activity and bestiality to be the main forms.

102   Sophie Grace Chappell much more horrible possibilities; indeed when I started writing this I did, then toned them down to the relatively innocuous, but still extremely nasty, example of puppy-​ burning, because the even worse examples left a seriously bad taste in my mouth. As they always do. (Following on from this experience, I resolve henceforth to stop bringing lethal threats to miners into trolleyology. Miners are humans. All this banging on about how to kill them [or not] is at least distressing and distasteful, and if Anscombe is right, corrupting too. So from now on, at least when I discuss it, every being at lethal risk from that runaway underground train is going to be either an adorable puppy or a cute kitten or perhaps a big-​eyed baby seal. And what are all these puppies, kittens, and above all seal cubs doing down a mine? That is just the kind of question that consequentialists never ask. So neither will I.) The puppy-​burning example does its work—​and then, thank goodness, we can stop thinking about it—​if it gets the point home that there are some action-​types for which we have pretty much the same reaction as Anscombe. She thinks that it ‘shows a corrupt mind’ ever to deliberate over certain action-​types, such as treachery and murder and apostasy and ‘sodomy’; we think that someone would be ‘sick’ if he always deliberated over certain action-​types, such as puppy-​burning. The two thoughts are not identical, but they are spectrum-​related. And consequentialism has the remarkable feature that, at least in its basic form, it cannot make good sense of either thought. It is because consequentialism has this sort of effect on our practical reasoning, and on our account of practical reasoning, that Anscombe thinks that it deserves polemical denunciation. And on that, I suggest, she is close to right. In the age of the internet troll it should come as no surprise to us to hear that there really are some moral positions to which the appropriate response is not argument but simple disengagement (and perhaps denunciation too). But this is anything but a new thought. It does not take Dorothy Parker to teach us that there are books that deserve not to be read, nor even to be tossed lightly aside, but to be hurled with the greatest possible force.22 No theme is more salient in Plato’s dialogues than the theme of the sophist, the charlatan, the pretend or sham philosopher, the person whose argumentative purposes are not those of good faith, but venal or studiedly perverse or both, and whose presence therefore constantly raises the question when and how we should engage with him, and indeed whether we should engage with him at all. As Rachel Barney has recently reminded us,23 Aristotle knew about trolls too. He is after all the author of a 22  For some time, internet searches enabled me to retain my fervent hope that the book of which Dorothy Parker said this was Atlas Shrugged. But more skilful investigations, for which I thank Michael Morris, made it clear that Parker never said this. The case is an example of what we may call Attribution Magnetism: the real author of the bon mot was someone much less well-​known called Sid Ziff: https://​ quotei​nves​tiga​tor.com/​2013/​03/​26/​great-​force/​. 23  Rachel Barney, ‘[Aristotle], On Trolling,’ Cambridge University Press, May 3, 2016, https://​www. cambri​dge.org/​core/​journ​als/​jour​nal-​of-​the-​ameri​can-​philos​ophi​cal-​asso​ciat​ion/​arti​cle/​aristo​tle-​on-​ troll​ing/​540BB​557C​8218​6C33​BFFB​61E3​5A0B​5B6/​core-​rea​der.

Anscombe’s Three Theses after Sixty Years    103 book with a title that we could almost translate as ‘On the Charlatans’ ‘Killer Arguments’ (De sophistarum elenchis). Throughout the Organon and elsewhere, he is deeply concerned with the examination of bad and perverse arguments, just as such. Indeed, at Topics 105a3–​8, he is explicit that there are times when even a philosopher ought to stop arguing: It is not necessary to look every problem over, or every thesis, but only the ones that might be puzzles for someone who stands in need of argument rather than of castigation or perception. Those who are puzzled whether or not we should honour the gods or love our parents need castigation; those who are puzzled whether or not snow is white need perception.

Perhaps the verbal form of ‘castigation’ (kolasis) is, precisely, polemic.

5.5 Sometimes we see consequentialists involved in casuistry, where the challenge to absolutists like Anscombe is ‘Come up with an action-​type that is so bad that it would never be right to do something of that kind, no matter how bad the consequences that threatened’. Then the absolutists nominate puppy-​burning (or whatever), and much ingenuity is deployed in inventing counterexamples, and the result is a kind of obsessive focusing of our deliberative attention on some truly terrible action-​type. On this casuistical game Anscombe comments, in a tone of ironic mildness but with an edge of acerbity to it, that the point of considering hypothetical situations, sometimes very improbable ones, seems to be to elicit from yourself or someone else a hypothetical decision to do something of a bad kind. I don’t doubt that this has the effect of predisposing people—​who will never get into the situations for which they have made hypothetical choices—​to consent to similar bad actions, or to praise and flatter those who do them, so long as their crowd does so too, when the desperate circumstances imagined don’t hold at all. (MMP, 22–​23)

At other times consequentialists are not open to the charge that Anscombe here presses about this sort of casuistry, that by making us focus on truly horrible alternatives and encouraging us to foster conditional dispositions about them, they are (directly and immediately) corrupting our characters. At other times consequentialists do not focus on any alternatives, because they try to include all alternatives. Here the charge against consequentialism—​which I think can be recovered from MMP, though it is certainly not in the foreground—​is not that it gives us corrupted characters, but that it gives us no characters at all.

104   Sophie Grace Chappell What I have in mind here is an argument against consequentialism, in the modern sense, not in Anscombe’s, that I have been trying to get consequentialists to listen to for getting on for twenty years.24 (Without much success.) Consequentialism’s basic model for practical deliberation tells each agent to look at all the options or alternatives available to her at any given time and choose the best option from this option-​range. When applied successfully the model is supposed to be universal—​the agent really will deliberate over all her options—​and also objective—​the option-​range that she deliberates over really will include all and only those options that are actually there in the world. For consequentialism to work it needs to be both universal and objective; otherwise, consequentialist deliberation could hit on the best option only by chance. The trouble for these two aspirations is that they are in conflict with a third aspiration which is at least as non-​negotiable as they are, namely, completability. For anything to be recognisable as a model of deliberation at all, and for anyone who operates according to the model ever to hit on any options at all, it must depict deliberation as typically completable in a finite time. But if consequentialist deliberation really is universal and objective, then it won’t be completable; and if it really is completable, then it won’t be universal and objective. The requirement of objectivity, in the consequentialist model, is the requirement that the world should just give us our option-​range—​that we should be able to read off our option-​range, immediately, simply from the way things are. The requirement of universality is the requirement that the world should give us our entire option-​range—​that the option-​range that we thus read immediately off the way things are should be complete. It is usually supposed—​explicitly or implicitly—​that it is the great strength of consequentialism to be objective and universal in this way: consequentialism is supposed to have the advantage over every other approach to ethics, that it gets us immediately to the way things are in the world. The trouble with this is obvious. ‘The world’ ‘gives us’ no option-​range whatsoever. Or at any rate, it gives us no finite option-​range. The world is a blooming, buzzing confusion of indefinitely many indefinite, and indefinitely individuated, overlapping possibilities. What can it mean to have in your option-​range all the options that there are? If it means anything at all, it apparently means having an infinitely large option-​range. But then, if deliberation means considering all of these infinitely many options, I can never complete the deliberation whereby I am supposed to choose between them. In fact, I can’t even start it; I can’t even set up the option-​range that I should consider, never mind complete its consideration. Consequentialist deliberation has the form of a disjunctive syllogism: ‘Do A or B or C; but not B; and not C; so do A’. The option-​range is represented by the disjunction in the first premise. But for consequentialist deliberation to be objective and universal, this disjunction has to mention every option that I actually have; whereas for consequentialist deliberation to be completable, it has to be finitely long. But it can’t be both.

24 

For my original presentation of it see ‘Option Ranges,’ JAP 2001.

Anscombe’s Three Theses after Sixty Years    105 Even if this deliberative paralysis could be overcome—​if there was some way of individuating and regimenting the options available to me at any time so that they form a countably infinite array rather than just an uncountably indefinite one—​it would remain true that the person who chooses, at each choice-​node, amongst all the options that are available to her was a person, not (as with the casuistry) of bad character but of no character. Nothing would be in principle ruled out for her (not even puppy-​ burning). As people used to say, she would have no backbone; in her there would be no resistances. Except of course a resistance to acting less than maximisingly. If that is indeed the one moral absolute that consequentialists actually do recognise, as Crisp (2004: 76n5) suggests, then the alternatives for consequentialists would appear to be either never acting or constantly violating their own absolute. For maximising is what the consequentialist is committed to doing. But in practice, maximising is never going to happen.

5.6 ‘The world’ ‘gives us’ no option ranges, but practical wisdom might. What we need is a doctrine of salience: an account of what the option-​ranges are that good deliberators will characteristically recognise. As you may put the question if you’re an Aristotelian or a virtue-​ethicist: What will the phronimos have in her option-​ranges? Or if you’re a consequentialist (it might be suggested): What are the option-​ranges that it maximises the good for agents to have? The main trouble for the consequentialist here is that her form of the question prompts a regress up the orders. To be consistent with her own maximising principles she has to be asking, not what option-​ranges it is good for deliberators to have, but what option-​ranges it is optimal for them to have. But then she is confronted with another disjunctive syllogism, this time beginning, ‘Deliberators should recognise option-​ranges of form A, or of form B, or of form C . . .’ And here too, for the argument to be even set up, the first premise of the syllogism must be at once objective and universal, but also complete. For reasons already given, it is guaranteed not to be both. Of course she can still postpone dialectical checkmate by moving up another order. But by now, a sense of futility should be dawning. Alternatively she can prevent the regress by giving up on maximising at some stage—​ and no doubt the earlier the better. Then she can at least ask the question, albeit in the non-​consequentialist form ‘What are the option-​ranges that it is good for agents to have?’ However, the problem for her then is that this non-​consequentialist question looks likely to have a non-​consequentialist answer. In general (the answer is likely to be), it is good for agents to deliberate as virtuous agents do in fact deliberate—​which means: in ways which sometimes advert to consequences, but are overall thoroughgoingly non-​ consequentialist. So in particular with option-​ranges, it seems very likely to turn out that there is no distinctively consequentialist way of thinking about them; at this point

106   Sophie Grace Chappell as at others, what consequentialism tells us to do is stop reasoning like consequentialists, and think, instead, in roughly the way that virtue ethics recommends. This is the familiar phenomenon that is usually called the self-​effacement of consequentialism and is often recognised with a rather puzzling complacency by consequentialists themselves, as if it were a minor wrinkle in the consequentialist position. Such complacency seems entirely misplaced; it seems to rest on a failure to take the phenomenon seriously. If we do take self-​effacement seriously, a better name for it will be ‘self-​abolition’. If what consequentialism itself tells us is that there is no room in our deliberation for consequentialism, then we have to take that as meaning literally no room for it: no room at all. And that means giving up on consequentialism altogether and everywhere, right back to the train of thought’s beginning in a question that turns out, by its end, to have been one that we should never have asked in the first place. Or not, at any rate, for any other purpose than to watch it get reduced to absurdity.

5.7 This is where the dialectic takes us if we consider consequentialism in the second of the two ways that I offered in section 5.5, as offering us, not a game played with specific casuistical cases but a general model for practical deliberation. I have said something, in brief outline, about why I think that general model must fail. Whether or not that is right, notice that this discussion is one that we have been led into by the dialectic of Anscombe’s own argument. Though she does not follow this dialectic out in MMP, maybe Anscombe herself would endorse it. (One remark she does explicitly make that is clearly relevant to this line of thought is ‘An action will fall under a number of descriptions if it falls under any’ [MMP, 5]; so of course a decision about how to count a given action as an option in an option-​range will depend on describing it a particular way.) But note that it seems fair to describe this dialectic as a dialectic in moral philosophy; and recall MMP1: MMP1: We should stop doing moral philosophy, ‘at any rate until we have an adequate philosophy of psychology, in which we are conspicuously lacking’.

If Anscombe would be willing to endorse this dialectic or something like it, how do we square this willingness with MMP1? Two answers suggest themselves. One is that Anscombe would regard the dialectic as a contribution not to moral philosophy but to ‘philosophy of psychology’. The other is that MMP1 too is not to be taken too literally: it too should be understood to have a polemical force, as a kind of shock tactic. Or indeed as having a pedagogical force: ‘Suppose—​at least for now—​that you forget about that method completely. Just try and solve the problem this way’. This is the kind of thing that we teachers say to our students all the time; what Anscombe says in MMP1 perhaps has a similar force.

Anscombe’s Three Theses after Sixty Years    107 There is something, I think, in both these answers. A trawl through MMP quickly produces a whole cluster of questions that Anscombe would like to consider and thinks are very much open for consideration by moral philosophers (or philosophers who would otherwise be moral philosophers were it not for Anscombe’s ban). A plausible surmise as to why is that they all come for her under the heading of ‘philosophy of psychology’ rather than ‘moral philosophy’ in the sense that MMP1 tells us to stop doing. Besides the question what intention is, and how to tell an intended consequence from a merely foreseen one (obviously enough), the cluster includes the questions how to individuate action-​types and how to specify morally relevant descriptions; the nature of virtue, and of specific virtues like justice; what counts as a paradigm, and what counts as a borderline, case of injustice; what counts as ‘intrinsically unjust’, and what as ‘unjust given the circumstances’; what pleasure is; and, with particular urgency, how it is that ‘a just man is a good man’. When Anscombe recommends, in MMP1, that we stop doing moral philosophy, she is not at all recommending that we stop these enquiries. On the contrary, since these are the kinds of things that she has in mind under the heading ‘philosophy of psychology’, these are the very enquiries that she thinks we should pursue.

5.8 Is there, then, anything that Anscombe literally and non-​polemically means the moral philosophers of 1958 should just stop doing—​not just for a moment, for the purposes of pedagogy or heuristics or as argumentative shock tactics, but altogether? Yes, there is. Consider MMP2: MMP2: We ought,25 ‘if this is psychologically possible’, to jettison all the following concepts (at least these five, and by implication, all others sufficiently closely associated with them): moral obligation, moral duty, the morally right, the morally wrong, and the moral ought.

What Anscombe thinks we should stop doing, altogether and indeed permanently, is what I shall call ought-​first moral philosophy. She thinks we (or the philosophers of 1958) should stop doing first-​order normative ethics that starts from ‘linguistic analysis’ of the word ‘ought’,26 from its meaning and alleged logical properties, or from particular uses of that word, to issue ‘universal prescriptions’, and that only gets to ‘philosophy of psychology’, to problems like those raised in the list of questions in my last full 25  This is of course deliberate: Anscombe introduces her proposed ban on the special moral ought using the word ‘ought’ in a non-​special sense, to underline the fact that she recognises an ordinary Ought that is, as she herself says, ‘indispensable’. 26  In what follows I designate the word by ‘ought’, the concept by ‘Ought’.

108   Sophie Grace Chappell paragraph, from and after this linguistic analysis of the word ‘ought’. Her target, in one word, is Hare. Though I do not seriously mean that she is thinking only of him—​she explicitly says otherwise—​certainly Richard Hare’s views in The Language of Morals, Moral Thinking, and Freedom and Reason are a perfect fit for the position that Anscombe is targeting in MMP. Hare’s project, which was still in progress when Anscombe wrote—​in 1958 he hadn’t yet published Moral Thinking or Freedom and Reason—​is to divide possible utterances along two axes, one way into descriptions and prescriptions, the other way into particular and universal. Then his interest is in identifying the features that any modal auxiliary, like, in English, ‘must’ or ‘needs’ or ‘should’ or ‘might’, has to have in order to count as what he calls a universal prescription. He claims that it is a matter of the meaning of ‘ought’—​at least when used in its full-​strength sense, i.e., morally—​to express an overriding universal prescription. This gives Ought, in its moral sense, an analytical centrality and primacy in Hare’s moral philosophy. In casuistry, it means, for instance, that we can deduce what we are saying about any other like cases, if we say that ‘X ought to phi in C’. And in the philosophy of psychology, it means that we can give an explanation of the phenomenon (or alleged phenomenon) of akrasia, of deliberately doing what you know you ought not to do, that begins from considerations about what it is for the akratic person to assert ‘I ought not to do this’.

5.9 What’s so bad about ought-​first moral philosophy? First, at the level of philosophy of psychology, Anscombe sees ought-​first moral philosophy as a hopelessly back-​to-​front research program. To attempt to shed light on, for instance, the phenomenon of akrasia by analysing the ways in which an Ought can be less than what Hare calls ‘full strength’ (see Freedom and Reason, ch. 5) is to fail to shed any light on anything substantive. The substantive question is not the extent to which anyone has or has not actually and fully sincerely issued some overriding universal prescription. The substantive question is why, for her, there is a gap and indeed an actual conflict between her motivations and her moral beliefs, and what kind of psychological and motivational states she is likely to be in when such gaps and conflicts occur. Ought-​first moral philosophy fails to shed any light on this. We cannot address that gap, or that conflict, by advancing a theory like Hare’s of the meaning of ‘ought’. Second, Anscombe sees ought-​first moral philosophy as not only mistaken about first-​order ethics but also—​to use her word for it again—​corrupting. It is not, of course, that the ought-​first programme in itself directly entails particular first-​order moral views that she sees as wicked; Hare remains enough of a Kantian to be offering a theory in which the formal structure comes first, and the substantive commitments follow on within the framework laid out by specifying the formal structure. (‘Ethics, the study of the logical properties of the moral words, remains morally neutral [its conclusions

Anscombe’s Three Theses after Sixty Years    109 neither are substantial moral judgements nor entail them, even in conjunction with factual premisses]’.)27 But, also like Kant, no substantive commitments follow until the agent actually wills or chooses or desires something. What Hare’s theory says is that if phi-​ing is ‘your principle’ in case A (if, that is, you make a moral commitment to phi-​ing in case A; if you affirm the principle that one ought to phi in A)—​then that commits you not only to phi-​ing in A but also to endorsing and recommending phi-​ing in any and every A-​like case. So with ethical commitments, universalisability follows on prescription. But where do the prescriptions come from? Anscombe’s answer to this is, I think, likely to be the same as her answer to the question how ‘the consequentialist’ deals with borderline cases: The consequentialist, in order to be imagining borderline cases at all, has of course to assume some law or standard relative to which this is a borderline case. Where then does he get the standard from? In practice the answer invariably is: from the standards current in his society or his circle. And it has in fact been the mark of all these philosophers that they have been extremely conventional; they have nothing in them by which to revolt against the conventional standards of their sort of people; it is impossible they should be profound. But the chance that a whole range of conventional standards will be decent is low. (MMP, 22)

Just likewise, I suggest, Anscombe’s answer to the question ‘Where then does Hare get the prescriptions that he wants to submit to his universalisability test?’ will be ‘From the standards current in his society or his circle’. This too, according to her, will have the effect of making Hare’s moral philosophy entirely conventional. As she argues in ‘Does Oxford Philosophy Corrupt Youth?’, it will be exactly as corrupt as the society in which it appears, neither more or less. But given Anscombe’s views about the morality of her society (in 1958), this from her is clearly nothing like praise. Anscombe’s third reason for rejecting Hare’s style of ought-​first moral philosophy is that she thinks it takes no proper account of the history of our moral concepts. This brings us to MMP2, her famous claim that we should ‘jettison’ the special moral Ought and related concepts. I discuss this in sections 5.10 and 5.11.

5.10 If someone professes to be expounding Aristotle and talks in the modern fashion about ‘moral’ such-​and-​such, he must be very imperceptive if he does not constantly feel like someone whose jaws have somehow got out of alignment: the teeth don’t come together in a proper bite. (MMP, 3)

27 Hare, Freedom and Reason, 97.

110   Sophie Grace Chappell One of the oddest features of Anscombe’s famous essay is that in these words she seems—​ you might have thought, against all reasonable odds—​to be alluding to Nietzsche’s polemical work Twilight of the Idols: ‘How much has conscience had to chew on in the past! And what excellent teeth it had! And today—​what is lacking?’ A dentist’s question. (‘Maxims and Arrows’, 29)

The likelihood that this is not a mere coincidence of words, and that Anscombe was deliberately alluding to the Nietzsche passage, is increased by the striking fact that her strategy against ought-​first moral philosophy is a genealogical one. Nietzsche said that the philosophy department of some leading university [should] offer a series of prizes for essays on the evolution of moral ideas. . . . I would propose the following question, which deserves the attention of philologists, historians, and philosophers alike: What light does the science of linguistics, especially the study of etymology, throw on the evolution of moral ideas? (On the Genealogy of Morals, Essay 1, endnote and section 4)

He also said, more pithily, that ‘all concepts in which an entire process is semiotically concentrated defy definition; only something which has no history can be defined’ (Essay 2, section 13). Anscombe’s thesis about Ought, and in particular the special moral Ought, is proposed in precisely this Nietzschean spirit of etymological enquiry which, it is at least possible, will prove to be a debunking enquiry. She takes Ought to be a reservoir of hidden historical insights—​insights which may have many things to tell ‘us moderns’ about our values and our value systems, but not necessarily things that are agreeable to our ears or our theoretical assumptions. Because the concept Ought is historically conditioned in this way, it is not something that can be analysed as if the word ‘ought’ were a modal auxiliary that stands for a timeless, un-​nuanced, un-​shadowed concept of non-​negotiable, universally applicable practical-​requiredness (in short, as a ‘thin concept’ rather than as what Anscombe, in effect, argues it is—​a ‘thick concept’).28 Yet that is precisely how Hare wants to treat it. Hare tells us that Ought simply marks an overriding universalised prescription. (At least in its central and fully moral usage. But on his picture, fallings-​off from the central case happen by degrees of loss of overridingness or universality or prescriptiveness, not by historical nuancing.) In saying this, he claims to be excavating the true semantic content of the notion. Yet Hare’s ambition is in this way to explain a meaning: ‘The function of moral philosophy (or at any rate the hope with which I study it) is that of helping us to think better about moral questions by exposing the logical structure of the language in which this thought is expressed’ (Freedom and Reason p.v.). So, we might reasonably protest, it is extraordinary how little use his account of Ought makes of the known facts of the history of ‘ought’. For the purposes of linguistic analysis, both the word and the 28 

Cf. my ‘There Are No Thin Concepts,’ in Kirchin 2015.

Anscombe’s Three Theses after Sixty Years    111 concept, we might have thought, must be worth tracing back to their actual historical origins. If we do that, we will of course find evidence of change in the word: (contemporary) English has not existed for ever, and once upon a time ‘ought’ was āhte, the Anglo-​Saxon past-​tense first-​and third-​person singular of āgan, ‘to owe’.29 But surely such evidence of change in the word is also at least likely to be evidence of change in the concept? No ordinary contemporary speaker of English naturally hears the connection that āhte made between ‘I ought to phi’ and ‘I owed to phi’. But even this small fact about the history of the word tells us that there was a time when speakers of (what was then) English did hear that connection. Indeed it tells us, presumably, that they were able at some point to endow the word with its practical or moral sense because they heard this connection. ‘I owed to phi’ is semantically incomplete, and therefore dubious English. We can rephrase it in rather more filled-​out and grammatical English as ‘It (then) came to be my debt that I should now phi’. But this filled-​out rephrasing makes it look like a reference to someone who is my creditor is written into the structure of āhte, as is a reference to a time when that someone became my creditor. That is, it makes it look like once upon a time ‘ought’ or its ancestral was naturally heard as a word for a four-​place relation involving an agent, a deed, someone to whom the agent’s deed is owed, and a time at which that debt was incurred. Obviously nothing like that is written into our Ought: the modern English Ought is (at any rate grammatically) two-​place, involving only an agent and a deed. And so two questions become pressing. The first question is: My debt to whom? And the second question is: My debt incurred when? That Ought, and/​or its ancestral āhte, should raise these questions; that it should once have had this four-​place relational structure rather than today’s two-​place structure—​if this is not a significant point about the history of Ought, and hence about what Ought has meant in our society even if it no longer does today, I find it hard to imagine what might be. And yet the point finds no place in Hare’s analysis of the word and the concept of Ought. Of course there can be all sorts of debts, incurred at all sorts of past times, to all sorts of creditors. In that variety, we might say, lie the roots of the variety of the possible uses of ‘ought’, some of them entirely mundane and down-​to-​earth—​and indisputably not specifically moral—​that both Hare and Anscombe so stress, and which Anscombe also insists are practically indispensable. (As already noted [note 24], she is by no means calling for us to jettison every Ought, just the specially moral one.) But when we read Anscombe’s insistence upon the divine-​command background to what has become our Ought, we may remember that English was the language of a Catholic nation for well over nine hundred years—​if we count from St Augustine’s landing in ad 597 to

29 

‘Ought’ can still be the past tense of ‘owe’ in Shakespeare (Henry IV Part I [1597?], Act 3, Scene 3):

PRINCE HAL.   Thou say’st true, hostess; and he [Falstaff] slanders thee most grossly. HOSTESS.   So he doth you, my lord; and said this other day you ought him a thousand pound.

112   Sophie Grace Chappell the completion of the Henrician reformation by about 1540. Given this background, we may well suggest that if English has, historically, had any strong sense of a specially moral sense of Ought, then that sense will have been focused upon one particular kind of past debt, and one particular creditor: the past debt of obedience to God’s revealed commands that, according to Catholic Christianity, we have owed to God ever since the giving of the Law. (Whenever theologians take that to be: on Sinai if we mean the revealed law; perhaps in Eden if we mean the natural law.) It is in the light of this history that Anscombe insists that our conceptual grip upon Ought, as a specially binding word of command, stands or falls with our conceptual grip upon the specifically Christian and Catholic tradition of God as our moral legislator and creditor. What is taken away from Ought, when a God to whom we have owed a debt of obedience since long ago is taken away from our ethical consciousness, is precisely the third and fourth places in the four-​place debt relation just sketched. Without that God as creditor, and without the giving of his law at some particular time, it no longer makes sense—​or at least, it no longer makes the same sense—​to say, in what we now call moral contexts, that ‘I ought to phi’. For in those moral contexts, what this originally meant was ‘Since Eden (or Sinai) I have been in debt to God that I should phi’. And when we said ‘I ought to phi’ in non-​moral contexts, it meant ‘Since some time I have been in debt to someone else that I should phi’: since the start of my apprenticeship, to a master craftsman, perhaps; or since the beginning of our friendship, to a friend; or since I swore fealty to a guild, to that guild. But if a culture ceases to believe in God, or at any rate in a creditor God as I have called him, for our moral debt to be owed to, then in the case of the moral ought all we have left is ‘the interesting situation’, as Anscombe herself calls it, of the notion of a debt that is a debt yet to no creditor. (Or, as we might perhaps more grandly come to call it, a categorical or unconditional debt—​a debt that we might take to be all the grander for having no particular beginning and for being owed to no one.) ‘It is as if ’, says Anscombe (MMP, 11), ‘the notion ‘criminal’ were to remain when criminal law and criminal courts had been abolished and forgotten’. Or in MacIntyre’s alternative dystopia—​one which, alas, looks significantly closer to hand in 2018 than it did in 1981, when he wrote After Virtue—​it is as if ‘a Know-​Nothing political movement’ had taken power and successfully abolished ‘science teaching in schools and universities, imprisoning and executing the remaining scientists’. After this crash ‘all they possess are fragments, and knowledge of experiments detached from any knowledge of the theoretical context which gave them significance . . . those contexts which would be needed to make sense of what they are doing have been lost, perhaps irretrievably’ (1). My own debt metaphor is different from Anscombe’s legal metaphor and MacIntyre’s science metaphor. But all three metaphors make the same point: that because of the history of England, we now use the English word ‘ought’ in ways divorced and estranged from its historical origins, and that there is something seriously parochial about treating Ought in isolation from this history, and ‘moral’ and ‘morality’ and ‘right’ and ‘wrong’ and ‘obligation’ and ‘duty’ as if they were no more than adjectival and verb-​nominal correlatives of ‘ought’.

Anscombe’s Three Theses after Sixty Years    113 It is also her reference to this history that makes it almost trivially correct for Anscombe to draw attention to the fact that the Greeks, and in particular Aristotle, had no such concept as our special moral Ought, and so lacked also the concepts involved in our using the moral terms as if they were in this way merely grammatical inflections of ‘ought’. Her point here is not in the least (pace Crisp 2004, 84) that Aristotle and his contemporaries lacked a vocabulary for expressing practical necessity in general. Of course they didn’t; no form of life remotely like ours would be at all likely to lack that. But then, nothing is achieved against Anscombe by instancing, as Crisp goes on to instance, the words with which the Greeks expressed their concept of practical necessity, such as dei and khrê; to do this simply misses her point. Anscombe’s thesis is not that the Greeks lacked a concept of practical necessity simpliciter, but that they lacked a notion of practical necessity that is divorced, as our ‘special moral Ought’ and our notion of morality in general are now divorced, from its original explanatory background in Catholic-​ Christian divine-​command theory. To say that the Greeks of Aristotle’s time did not have that is hardly contestable. Without access to a time-​machine, how could they? The difference between our special moral Ought and anything remotely like a sense of moral obligation that the Greeks of Aristotle’s day might have felt is at least as great as the difference between the single uniquely morally authoritative God of Sinai, with the enormously detailed body of commands that the Bible depicts him as promulgating, and the many contesting and conflicting pagan Greek gods, with the handful of whimsical, haphazard, sometimes merely spiteful, often downright disastrous wishes that they sporadically impose on those who take the dire and presumptuous risk of seeking them out. Of course you can speak, if you really insist, on divine commands in both contexts. But it would seem, to put it mildly, a little tin-​eared to miss the colossal differences between the ethical worlds of Moses, Abraham, and Jacob on the one hand, and Pentheus, Io, and Hippolytus on the other. As a consequence of Ought’s estrangement from its origins in a picture of a debt relation with a particular foundation in divine history, our modern predicament is (so Anscombe concludes) now this. We have a notion of moral obligation as all-​pervasive, universal, overriding, and yet rootless—​lacking any kind of clear foundation in any deeper ethical notion or any wider cosmological picture, and so lacking also in anything like a justification or rationale. ‘You ought because you ought because you ought’; to think otherwise is, so Prichard famously and so erroneously suggested, to rest moral philosophy on a ‘mistake’. Crisp (2004, 86n33) repeats Prichard’s suggestion when he wonders why Anscombe ‘never considers the suggestion that morality might be self-​ standing, requiring no justification from elsewhere’. But of course Anscombe does consider this suggestion. It is one of the central theses of MMP that any such conception is unintelligible and reduces ‘ought’ to a word of mere ‘mesmeric force’, ‘apt to have a strong psychological force, but which no longer signifies a real concept at all’ (MMP, 14). If her argument works, what Anscombe has shown is, amongst other interesting results, that our special moral Ought is not the only possible form that a conception of practical necessity may take. It is a very historically specific version of practical necessity. Given its estrangedness from its own origins in the notion that the obligation-​bearer

114   Sophie Grace Chappell stands now under a debt incurred in the past to God, it is of dubious intelligibility. The remarkable mesmerism of the special moral Ought can easily inveigle us into thinking that the word actually means something clear—​and also, into thinking that it is a notion that all societies and all ethical philosophers whatever must really have had, however ill they unfortunately expressed it. (Hare certainly had this thought, and so did Adkins in Merit and Responsibility. The best retort to it I know is still Robin Collingwood’s trireme [1939, 64; Knowing What to Do, vi].) Both thoughts are importantly false, and pedagogically speaking, the best remedy to both is to take Anscombe’s advice in MMP2 not as mere polemic but actually to follow it. All of the time when discussing the ethical thought of Aristotle and the other ancient Greeks, and most of the time when thinking about ethical matters for ourselves, we indeed do best simply to jettison the entire vocabularies of ‘ought’ and ‘moral’. For us at any rate, and for now at any rate, these words are infected with misconception and false history or (ahistory). We should just drop them.

5.11 Section 5.10 has presented a defence of MMP2. There are some obvious ways a counter-​ attack might go. Here, to close both the discussion of MMP2 and the chapter itself, are three. First, a critic might say that Anscombe and Hare (and their respective allies) are simply working on different projects. Anscombe wants to insist on the historical genealogy of ‘ought’. Hare wants to transcend that history by proposing an account of moral obligation that will be true in any society no matter what its history, and will therefore give us the resources to overcome the relativism of our various pasts. Nothing wrong with either project, but they aren’t in conflict either. The trouble with this criticism is simply that—​as Anscombe points out—​there is no such generic or universal notion of moral obligation. As I said at the end of section 5.10, the insistence that there must be ‘really’ is just more evidence of the special moral Ought’s ‘mesmeric power’. There is a relevant generic notion, of course, but it is not any kind of moral obligation; it is practical necessity. The idea that some things must be done, by some agents, for some reasons, is certainly, as I have already pointed out, not one that could be dispensable for any beings remotely like us, living any way remotely like ours. But moral obligation is just one historical formation of this notion of practical necessity. If Anscombe is right, it is, at least as we have it, a deformation of practical necessity. It follows that there is no such thing as a generic or universally true account of moral obligation. We might well have—​no doubt we should have—​an ecumenical and cosmopolitan moral impulse towards bringing everyone in the world within the scope and jurisdiction of one and the same set of basic ethical values. But if that is our aim, then our concept of moral obligation is no place to start. We should start instead with the

Anscombe’s Three Theses after Sixty Years    115 notion of practical necessity and seek an account of this that can command global agreement. Now the whole point of the notion of practical necessity is that it rests upon and interconnects explanatorily with other notions, to do with human well-​being, what it is good for humans to be, how human beings may or should or do choose, be motivated, and intend—​in short, with the notions central to Anscombe’s own ‘philosophy of psychology’. So what we have here is actually an argument for replacing Hare’s programme with Anscombe’s. I turn to a second criticism of MMP2, which presses the following question. Anscombe’s approach implies a charge of parochialism against analyses of ‘ought’ that ignore the word’s Anglo-​Saxon history. But she herself focuses on the history of England in a way that, for a start, looks like it does not even keep in contact with the rather different histories of Ireland and Scotland—​never mind anywhere further afield with an even more different history, such as America. Nor, similarly, does she look at modern languages other than English. But, we might think, some of these—​French, Spanish, and Italian, for example—​have the interesting feature that they can serve as controls. They are the languages of firmly Catholic cultures where there has never been a Protestant reformation. So if Anscombe is right, the divine-​command-​based ‘ought’ that she describes should still be flourishing in these countries. Is it? Obviously this is a big question, a thorough answer to which would involve a lot of European history and sociology. And obviously it can be said that the control is far from pure. In France, for instance, there was no successful Protestant reformation. But pretty much at the same time, and for reasons closely associated with the success of Protestantism elsewhere, there was in France a massive breakdown of the Catholic-​ Christian worldview, and it was replaced—​at least on the commanding heights of French society—​by the Enlightenment-​rationalist outlook of the philosophes. In fact, if the key feature of the Catholic worldview for Anscombe’s purposes is the centrality to it of divine command, we might think that something far less like a Catholic morality was present in Voltaire’s rationalist Paris than in Calvin’s authoritarian Geneva—​or Carlyle’s Calvinist Ecclefechan. Still, with those caveats, I don’t see why Anscombe’s response to the challenge should not be a moderately bullish one. It is hardly an unfamiliar thought that the ethical cultures of European Catholic societies are different from those of Protestant ones, or at least have been until the recent past, when secularisation and convergence have set in. And it might be right to find some of the reasons why in the sort of factor that Anscombe appeals to. What is certainly true is that every national language in Catholic southern Europe has the notion of debt or something like it built historically into its moral vocabulary.30 30  The same is true, though the in-​building is less obvious, in German and some other Germanic languages. German sollen, ‘to be obliged to,’ cf. English ‘shall,’ is connected via proto-​German with Schuld, ‘guilt’ or ‘debt.’ The same root is there in Swedish, Icelandic, Norwegian, and Anglo-​Saxon skuld, whence our ‘should’; this too seems originally to have been the past-​tense form of ‘to owe’—​cf. German sollte. However, in ordinary Norwegian, Danish, and Swedish today, this root has been replaced by a root that appears to suggest a quite different metaphor, the metaphor of a moral obligation as a burden: ‘I

116   Sophie Grace Chappell This is almost exclusively thanks to the Latin debeo: the Italian and French for ‘I ought’ are dovrei and je devrais, respectively (both subjunctives), and the Spanish, Portuguese, and Catalan even preserve a past-​tense form (yo debería, eu deveria, jo hauria; this last, interestingly, is a past tense of ‘to have’). So we can add all this too to the evidence base for the thesis that ought-​talk in a whole variety of languages expresses or used to express a four-​place relation. It expresses a present debt incurred at some past time by someone, to someone else. And at the focus of the concept there is the special case of that relation where the debt is to God, and the past time is Eden or Sinai, and the content of the debt is the practical necessity of keeping God’s commands. The third and final objection to Anscombe’s theses about Ought fastens on this notion of divine command. ‘Even if Anscombe is right about all this’, the objection goes, ‘her theses are no use to us. Maybe ‘ought’ does have buried in its history a reference to Catholic divine-​command ethics; but that just reinforces the necessity, for us, of turning quite away from any ethical outlook even remotely like Anscombe’s own Catholicism. If she is right about the history of Ought, then no wonder we can’t copy her in staying on the traditional side of the massive gulf that separates moderns from her own beloved medievals; no wonder we will have, instead, to be secular virtue ethicists like Philippa Foot. For if we know anything in ethics today, we know that divine-​command ethics is a busted flush. It’s an ethics of threat, and that means barely an ethics at all—​and anyway, what about the Euthyphro dilemma?’ In a sense, of course, none of this is actually an objection to anything Anscombe says. She herself agrees that a divine command ethics is likely to be uninhabitable by most modern sensibilities, and recommends a revived Aristotelian eudaimonism, or something like it, as an alternative for secular people. It is also true, no doubt, that divine-​ command ethics often has been presented as a system of threats: that is certainly how Anscombe’s husband’s work on the subject strikes at least one reader.31 Still, I will close by noting three counters to this sort of response. First, there is actually no reason why divine command ethics must be taken as a system of heavenly threats. With Jesus’s summary of the law in mind again, we might say that it can be seen instead as a matter of learning to respond to heavenly love. Second, on the Euthyphro dilemma: this is typically presented as the question whether we should do what is right because God commands it, or do what God commands because what he commands is right. I am at a loss to see why that is so

ought to go’ in these three languages is, respectively, jeg burde gå, jeg burde gå, and jag borde gå. One is at any rate free, I think, to see this linguistic replacement as culturally significant and having to do with distinctively Protestant notions of being burdened by one’s (in the first instance) individual conscience. Meanwhile Dutch and Frisian use, for ‘ought’, words like moeten, from the Indo-​European root that gives German magen and Macht and English ‘must’ and ‘may’ and ‘might’—​a root that has the peculiarity that it seems to have begun meaning a power or freedom to do something, and ended meaning an unfreedom not to do it. 31 

P. T. Geach, The Virtues and ‘Divine Commands and Moral Requirements,’ Philosophical Quarterly 30, no. 119 (1980): 180–​181.

Anscombe’s Three Theses after Sixty Years    117 commonly regarded as such a killer question for theological ethics. It is hardly a strike against the foundations of geometry to ask whether a certain figure is equiangular because it is equilateral, or equilateral because it is equiangular. Necessary properties come together, and if there are explanatory relations between them, they typically run in both directions. If God is God, then his commanding certain things, and the rightness of those things commanded, are both necessary properties of his. To insist that only one of these properties can be explanatorily fundamental, and then brow-​clutch about which it is, is not to be hot on the track of a refutation of divine command ethics. It is merely (if I may borrow some Anscombean acerbity) to make a display of our own logical incompetence.32 To leave the last word to Anscombe herself, she also wrote a fine little essay on Plato’s Euthyphro—​less well-​known than it deserves to be—​in which she exposes another possible way of showing why the supposed dilemma does not hold at all generally. It can happen, she points out, that someone in authority over me wants to test and exercise my obedience. Such a superior may give me an order simply to see me obey it. When this happens, my doing the required act is obedience because it is what my order-​ giver has ordered me to do, but it is also what my order-​giver has ordered me to do because my doing it is obedience. Moreover, my action in obeying is the action that pleases the order-​giver because it is obedience, but my action is also obedience because it is the action that pleases the order-​giver. Both explanatory relations hold; if there is a circularity here, it is not in any sense a vicious one. So now on piety, Anscombe suggests, ‘piety is always what obedience is sometimes’: Obedience is sometimes what the order-​giver is after. But piety is always what the god is pleased with. Then just as, when obedience is what the order-​giver is after, it may not matter what the particular content of the order is, so the pious act may always in itself be indifferent. In order for there to be a pious act at all, there will have to be something or other specified to be the vehicle of the piety, something for the gods to be pleased at one’s doing. But it will be substantively pointless. (Gormally 2011, 16)

Here too, then—​at least if this prejudice seeks to use Plato’s Euthyphro as its authority—​ the common prejudice that divine command ethics is unsustainable is itself rendered unsustainable. No doubt Anscombe would want, in the end, to point us towards a kind of ethics that combines the Aristotelian naturalism that she recommends to secular people, with the divine command ethics that she herself draws out of her Catholic heritage, and of which she here presents one fragment of a defence. What the result of such a combination might be is, no doubt, a story for another occasion, but we can at least put a name to it. The name is Thomism.

32 

See further my ‘ “On Hearing God Speak: Socrates” daimonion and Euthyphro’s “Dilemma,” ’ European Journal for the Philosophy of Religion 2010.

Chapter 6

Pr actical Tru t h , Et h i c a l Natu ralism , a nd t h e C onstitu tion of Ag e nc y in Ansc omb e ’ s Et h i c s John Hacker-​W right

6.1  Introduction In a list of twenty opinions common amongst Anglo-​American philosophers that are inimical to the Christian religion, Anscombe includes this: ‘Ethics is formally independent of the facts of human life, and, for example, human physiology’ (GG2, 67). Not only is that belief contrary to Christianity, it is also, on Anscombe’s view, ‘a philosophical error and can be argued to be such on purely philosophical grounds’ (GG2, 68). Here we find a characteristic of Anscombe’s views on ethics generally: there is a complementarity between a purely philosophical approach and an approach informed by faith. In her treatment of the concept of sin, she offers two definitions: according to the first, a sin is a behaviour against right reason; according to the second, it is a behaviour against divine law (GG2, 117). Initially, the first definition might seem at odds with her view that ethics must take account of the facts of human life. Yet the definition must be read along with her rejection of the autonomy of ethics (GG2, 67). For her, what accords with right reason is thoroughly shaped by the facts of human life. Further, for Anscombe, these two definitions of sin are connected. If human beings were different in any of various ways, for example if we were born able to talk, never slept, sprang back to life when killed if properly kept, then what accords with right reason would be different. Moral virtues and vices would likewise be different, and so would the divine law, for it would command what is in accordance with that different nature. Hence, Anscombe adheres to a ‘law conception of ethics’ in the sense that she gives this term in ‘Modern Moral Philosophy’, where it is defined as the view that ‘what is needed for

Practical Truth, Ethical Naturalism, and the Constitution of Agency 119 conformity with the virtues failure in which is the mark of being bad qua man . . . is required by divine law’ (CP3, 30). On her view, then, those who do not embrace theism can have sound ethical views: they are views about what makes one good or bad qua human being. Anscombe follows Aristotle and Aquinas in taking human beings to be themselves the starting points of a certain sort of action; that is to say, we characteristically initiate change on the basis of appetites that are shaped by a normative conception of the good. Human actions can be defective, exhibiting vice, or not, exhibiting virtue depending on whether our reasoning and appetites operate as they should. As Anscombe unequivocally puts it, ‘ “That was a morally good action” is equivalent to . . . “That was a good human action,” and “That was a morally bad action” is equivalent to . . . “That was a bad human action” ’ (GG1, 203). Ethical norms can thereby be grounded on the kind of thing that we are, which can be grasped independently of affirming or denying any views on God. Still, the rejection of theism is not without consequences for our ethical views, according to Anscombe. These consequences go unrecognised by modern moral philosophers who advocate notions of moral obligation and moral right and wrong. For the two definitions of sin just given, though complementary, do not have the same sense. The second definition, sin as contrary to divine law, carries a conception of obedience, which is not connected to the notion of sin as behaviour against reason. Hence, if we restrict ourselves to the first definition, we must give up some of our ethical concepts that are associated with notions of obedience to law, which includes all concepts related to ‘morality’ as modern moral philosophers think of it.1 Hence, Anscombe outlines, though she does not herself subscribe to, a version of ethical naturalism. Specifically, she outlines an Aristotelian version of ethical naturalism on which norms of human conduct are grounded on what makes us good qua human. This is a conception of ethics centred on the human good, without many of the concepts that are central to modern moral philosophy. It is Anscombe’s conception of ethics as grounded in the human life form, without appeal to divine law, that I mean to take up and interpret in this chapter. A central question for this sort of Aristotelian ethical naturalism is: How does our kind membership generate norms that bind us? Why should I care about the sort of thing that I am when I deliberate about how to act? John McDowell brought out the problem here with the famous rational wolf thought experiment in ‘Two Sorts of Naturalism’. If a wolf suddenly develops rationality, it may question natural norms like ‘wolves hunt in packs’. The deliberative question a rational agent faces is ‘What should I do?’, to which the appeal to what members of the species do seems irrelevant.2 Likewise with human beings, I may question whether, for example, ‘humans make moral codes

1  For more on the critique of morality in Anscombe’s ethics, see my ‘Virtue Ethics without Right Action: Anscombe, Foot, and Contemporary Virtue Ethics’, Journal of Value Inquiry 44, no. 2 (June 2010): 209–​224 and James Doyle, No Morality, No Self (Cambridge, MA: Harvard University Press, 2018), 31–​51. 2  John McDowell, ‘Two Sorts of Naturalism’, in Mind, Value, and Reality (Cambridge, MA: Harvard University Press, 1998), 172.

120   John Hacker-Wright and recognize rights’ carries any weight with me: why should I do what humans do? One can say that a human who fails to recognise rights is a bad human, but why should I care about that? This way of construing naturalism assumes that it is offered in a certain realist register; there are moral facts that are at bottom descriptive natural facts about what makes a human good. One might then argue that what makes a human practically rational is recognising those facts as providing reasons. But it is not obvious why it should be important to me to be practically rational if that boils down to the descriptive claim that this just is how humans reason when they reason well. Here again, we might question why we should care; how do these facts apply to me in my attempt to figure out how best to live?3 Yet there is another way of construing naturalism, which is to take it as a version of constitutivism. Constitutivism is the view that the norms of human conduct are constitutive norms of action; by adhering to these norms we make ourselves able to act, and so constitute ourselves as human agents. If we fail altogether to adhere to those, we fail to constitute ourselves as human agents or at least constitute ourselves defectively. Christine Korsgaard, defending a Kantian version of constitutivism, argues that we must adhere to formal principles of willing, the categorical and hypothetical imperatives, in order to constitute ourselves as autonomous and efficacious agents.4 Since autonomy and efficacy are constitutive of action as such, these principles are constitutive principles of action, and through acting on them, we make ourselves into agents; conversely, failing to act on them we, by degrees, fail to be agents. An Aristotelian version of this view would argue that moral virtues are constitutive principles of action because they enable us to act towards our naturally given aim: happiness. Vices are defects of our agency inasmuch as they prevent us from attaining what we all aim at in acting. Vicious agents think they are aiming at happiness but are not, in fact, so that their agency is undermined by the false conception at which they aim. The correct normative conception of how to live is available to us all, and we can take it up or reject it, and so it is up to us whether we develop the virtues or vices, whether we constitute ourselves well or badly. This is roughly the view that Anscombe presents, or so I shall argue here. Constitutivism can be understood as a form of ethical naturalism inasmuch as we are a kind of thing, an organism, that faces this task of self-​constitution, which we can do more or less well. It pertains to our nature as an organism to act, and therefore our constitutive norms as a form of life include norms for action.5 This view lends itself to 3 

For further criticism of the recognitional realist version of naturalism, see Jennifer A. Frey, ‘How to Be an Ethical Naturalist’, in Philippa Foot on Goodness and Virtue, ed. John Hacker-​Wright (Cham: Palgrave-​Macmillan, 2018), 47–​84. Frey ascribes the view to Foot. 4  Christine Korsgaard, Self-​Constitution (Oxford: Oxford University Press, 2009), 81–​108. 5  For more on the connection between ethical naturalism and constitutivism, see my ‘Ethical Naturalism and the Constitution of Agency’, Journal of Value Inquiry 46, no. 1 (March 2012): 13–​23 and Jennifer A. Frey, ‘Happiness as the Constitutive Principle of Action in Thomas Aquinas’, Philosophical Explorations 22, no. 2 (April 2019): 208–​221.

Practical Truth, Ethical Naturalism, and the Constitution of Agency 121 an Aristotelian interpretation inasmuch as Aristotle himself agrees with the idea that it is in our nature to act, as when he says, for example, ‘a human is, alone among animals, a starting point of certain actions’ (Eudemian Ethics 1122b19) and ‘decision is desiring thought or thinking desire—​and the cause of this kind is man’ (Nicomachean Ethics 1139b5, cited and translated by Anscombe at GG1, 153). Constitutivism is an appealing position in metaethics for three main reasons. First, it shows how some relevant set of norms is practically relevant to us: the norms are internal to any undertaking. They are norms that we must take up as we deliberate and engage in action in order to act, rather than norms we learn about through observation and then subsequently apply in acting. Inasmuch as we care about anything enough to act, the norms that are constitutive of action cannot be indifferent to us. Second, it makes our responsibility for adhering to those norms transparent; it isn’t as though the norms are objects of observational knowledge we might have failed to grasp or forgotten to apply. Rather, they are norms that are within our reach if action is within our reach. Third, the content of those norms is tied directly to the conditions of human action. Constitutivists argue that the principles of human action that we must take up as we act are norms with moral content. Though the ethical naturalism that Anscombe describes is close to that of her fully subscribed neo-​Aristotelian colleague Philippa Foot, there are some significant divergences between their views. Foot’s Natural Goodness is very programmatic and can be read as adhering to a recognitional realist construal of naturalism as opposed to a constitutivist reading. Anscombe, through focusing on action, develops notions of practical knowledge and practical truth that are distinctively constitutivist in tenor, though this taxonomy of views was not around when she wrote. Without reference to Anscombe, Christine Korsgaard raises this complaint about realism: For much of the twentieth century, just as for the three centuries or so that preceded it, philosophers remained in thrall to the view that the function of all human concepts, and perhaps of all conceptual inquiry is to describe the world.6

It is striking to note the parallel with Anscombe’s complaint in Intention: Can it be that there is something that modern philosophy has blankly misunderstood: namely what ancient and medieval philosophers meant by practical knowledge? Certainly in modern philosophy we have an incorrigibly contemplative conception of knowledge. Knowledge must be something that is judged as such by being in accordance with the facts. The facts, reality, are prior and dictate what is to be said, if it is knowledge. And this is the explanation of the utter darkness in which we found ourselves. (I, 57)

6 

Christine Korsgaard, ‘Realism and Constructivism in Twentieth-​Century Moral Philosophy’, in The Constitution of Agency (Oxford: Oxford University Press, 2008), 325.

122   John Hacker-Wright Through emphasising the distinctiveness and importance of practical knowledge, Anscombe delivers the resources for developing a constitutivist version of Aristotelian naturalism. Here, I will show that Anscombe has resources for responding to objections to the very possibility of Aristotelian constitutivism that have recently been set forth by Korsgaard. Anscombe, with her understanding of practical or praxistic truth and connatural knowledge, points to an approach to constitutivism informed by Aristotle and Aquinas that offers a route to respond to the obstacles to neo-​Aristotelian constitutivism presented by Korsgaard. The resulting constitutivist view offers a compelling interpretation of the naturalistic aspect of Anscombe’s ethical views.

6.2 Korsgaard on Aristotelian Constitutivism In ‘Constitutivism and the Virtues’, Korsgaard discusses the possibility of an Aristotelian approach to constitutivism and shows that it faces three important obstacles, one of which she believes to be insurmountable.7 First, she raises a concern about Foot’s specific version of ethical naturalism; according to Korsgaard, Foot’s view appears to render the appeal to virtue otiose. That is because Foot, in Natural Goodness, appears to think of virtue simply as a matter of taking a certain set of considerations as reasons for action, or as she puts it: [T]‌he description ‘just’, as applied to a man or a woman, speaks of how it is with him or her in respect of the acceptance of a certain group of considerations as reasons for action. If justice is a virtue, this is what the virtue of justice rectifies, that is, makes good. . . . Similarly, if charity is a virtue, this is because it makes its possessor’s action good in the area of aims such as the relief of poverty.8

Korsgaard here raises the question of how virtue, for Foot, is supposed to ‘make good’ someone’s reasoning. On Korsgaard’s interpretation, Foot takes the virtues to be instrumental to achieving the human good, which can be grasped through generic Aristotelian categoricals that describe our life form, such as ‘Humans establish rules of conduct and recognize rights’. Since this generic statement characterises our life form, and justice enables us to realise something that characterises our form of life, justice is a virtue. For Korsgaard, this view focuses on the third-​person perspective: we are saying of someone that she acts from the right reasons; someone counts as virtuous when they take certain considerations as reasons for action.

7  Christine Korsgaard, ‘Constitutivism and the Virtues’, Philosophical Explorations 22, no. 2 (April 2019): 98–​116. 8  Philippa Foot, Natural Goodness (Oxford: Oxford University Press, 2001), 12.

Practical Truth, Ethical Naturalism, and the Constitution of Agency 123 Korsgaard contrasts this view of virtue with Aristotle’s. According to Korsgaard, Aristotle takes virtue to be constitutive of our cognitive grasp of the good. Without the virtues we cannot grasp what is good or bad in particular situations. That is because the virtues are states of the appetites that direct them to pursue or avoid certain things. We perceive the world in light of these appetites; in light of them certain things show up as good and to-​be-​ pursued, others as bad and to-​be-​avoided; if we are virtuous, those perceptions reflect our ‘normative conception of the world’. Our perceptions, informed by our appetites, prompt us to certain actions. This properly Aristotelian account of the role of the virtues, according to Korsgaard, makes them constitutive of an effective will, enabling one to perceive and respond to the world one encounters in accordance with a normative conception of the world. Foot cannot say this, Korsgaard thinks, because she does not explicitly endorse the moral psychology behind Aristotle’s account. If virtue is simply a matter of registering certain considerations as reasons, it does not seem necessary to the proper constitution of the will as it is for Aristotle. Character is, on Korsgaard’s reading of Foot, not playing the sort of role that would make it partly constitutive of the will, as it is for Aristotle. The second issue is the worry that the virtues, even as Aristotle construes them, do not meet what Korsgaard calls the ‘constitution requirement’. Any candidate for a constitutive norm must be genuinely constitutive of the sort of thing it is. Yet the moral virtues do not meet this requirement on the Aristotelian account because vice also seems to be constitutive of the will. That is, it seems that the vicious agent also makes choices and, if clever, can effectively pursue them. As Korsgaard notes, there are passages in Aristotle in which he claims that a vicious agent is at variance with himself, yet there are other passages in which he claims that the vicious agent is unlike an akratic agent in being without regret about his vicious actions. If an Aristotelian takes the vicious agent to be capable of fully integrated choice, it seems she cannot be a constitutivist, since the virtues will not be exclusively constitutive of the will. A third issue is that Aristotelian constitutivists have a problem with moral responsibility. Korsgaard argues that the Aristotelian constitutivist cannot meet what she calls the ‘self-​constitution requirement’. On Korsgaard’s account, we are fully responsible for an action only when we are the agents of an action. That means we are efficacious in bringing the act about, rather than the act being the upshot of a chain of prior causes. Further, we are responsible for actions only when we efficaciously act in such a way that the action reflects our normative commitments. Korsgaard thinks that Aristotle can capture the latter: on his account, according to Korsgaard, we are responsible for actions that emerge from states of character that reflect our normative conception of the world. But what isn’t clear about Aristotle’s view, according to Korsgaard, is that we are responsible for our characters. After all, the vicious person thinks he is doing what he ought to do. As Korsgaard points out, ‘the unjust or self-​indulgent person would not only have to know that the states of his character are produced by his actions, but also that the states he was producing were the wrong ones’.9

9 

Korsgaard, ‘Constitutivism and the Virtues’, 110.

124   John Hacker-Wright Korsgaard believes that Aristotle and Aristotelians more generally cannot meet the self-​constitution requirement because they rely on a naturalistic conception of agency according to which actions emerge from a certain kind of mental cause. In the case of the Aristotelian naturalistic account, that cause is a state of character. Since we are not responsible for our states of character, we cannot fully constitute ourselves. By contrast, Korsgaard advocates what she calls a normative conception of agency, on which it is the form of certain principles that we follow that makes us into agents.10 When we determine our actions in accordance with the categorical and hypothetical imperatives, we form our actions in such a way that we make ourselves into agents. Because Aristotle and his modern followers advert to character to explain action, and that character can itself be caused by one’s environment rather than the self, they cannot account for moral responsibility. Korsgaard concludes that there are insuperable obstacles to the idea of an Aristotelian constitutivism. Though Aristotle can possibly address the first two issues she raises, he falters on the third, and according to Korsgaard, there is no evident way to make good on that while remaining consistently Aristotelian. Yet I contend that the place to look for answers to these objections is in Anscombe’s account, since it highlights the importance of the practical point of view to a properly formulated Aristotelian naturalism.

6.3  Anscombe’s Moral Psychology of the Virtues Whether or not Korsgaard is right about Foot’s account of virtue, Anscombe clearly and explicitly adheres to an Aristotelian moral psychology of virtue. This is perhaps clearest in Anscombe’s brief but illuminating treatment of connatural knowledge. Connatural knowledge is knowledge that someone has because she possesses a certain virtue (GG1, 60). This concept shows up in Aquinas, who says, ‘[A]‌bout matters of chastity, a man who has learnt the science of morals judges rightly through inquiry by reason, while he who has the habit of chastity judges rightly of such matters by a kind of connaturality’ (Summa Theologica II-​II, q. 45, a. 2).11 The ‘kind of connaturality’ in question is the connaturality of second nature. The unqualified sense of connaturality applies to things in their natural inclinations, as to move towards the centre is connatural to heavy bodies, on Aquinas’s view (Summa Theologica I-​II, q. 26, a. 1).12 The ‘kind 10 

See also Christine Korsgaard, ‘The Normative Conception of Agency’, in Rational and Social Agency: The Philosophy of Michael Bratman, ed. Manuel Vargas and Gideon Yaffe (Oxford: Oxford University Press, 2014), 190–​214. 11  Anscombe does not cite this passage, though it is clearly the one that informs her discussion. Instead she says, ‘I haven’t been able to find St. Thomas giving the term the application I have been describing. Indeed, I don’t know the source of that application’ (GG1, 60). 12  See Taki Suto, ‘Virtue and Knowledge: Connatural Knowledge According to Thomas Aquinas’, Review of Metaphysics 58, no. 1 (September 2004): 68.

Practical Truth, Ethical Naturalism, and the Constitution of Agency 125 of connaturality’ mentioned in the previous passage is a connaturality that comes from an acquired habitus. Through that connaturality, the person with a virtue acquires a capacity to recognise ‘what action will accord with and what ones will be contrary to the virtue’ (GG1, 60), and this happens through a ‘sort of perception’ rather than through the formulation of explicit judgements. As Anscombe argues, ‘the knowledge that it is wickedness to do what the Nestlé Company is reported as doing—​to persuade poor African women to feed babies their powdered milk instead of suckling them . . . belongs with a just way of looking at things’ (GG1, 62). Anscombe here outlines a distinctive practical perspective on the world that comes with having a will that is inclined to certain ends. Virtue here is not merely gestural; that is, it does not merely point to otherwise specifiable principles of the agent that lead her to act well. Here moral virtue is something necessary to the goodness of the good will; indeed, it is necessary to the will as a capacity for going for the good in action, as without virtue, we cannot correctly identify an action as good and act on it under that apprehension. To see this, it is crucial to grasp the role of inclination and character on Anscombe’s account. Moral virtue is for her, as for Aristotle and Aquinas, a good quality of our appetites or inclinations. So to see the importance of moral virtue for her, we must see the importance of inclinations and character. First, it is necessary that an action identified as a good action be one I am inclined to do. We can be inclined mistakenly, or upset through weakness of will, but to take something to be good is not merely an intellectual labelling; it implies a practical stance of pursuit. Hence, we need our inclinations to be engaged in the pursuit of anything that is taken to be good in this sense. In part for this reason Anscombe thinks we need the virtues not simply to correct ways in which we tend to be diverted from an end that could be achieved anyway, for example, to shore up our ability to withstand fear or restrain our tendency to take too much for ourselves. The virtues are necessary if we are to be oriented in our inclinations towards good actions. Second, we need character (virtue or vice) in order to possess a normative conception of how to live. Anscombe argues that our ability to go for anything as a ‘constitutive means toward the good way of going on’ requires moral character (CP1, 71). That is because going for something in that way requires more than a passing inclination or having the thought ‘I am hereby doing well’, or even to have an inclination or such a thought while acting on one occasion. Rather, the thoughts and inclinations need to inspire habitual action to be credible as being a part of one’s conception of living well. Virtue will be what enables us to attain the correct conception of going on. Although vicious agents have a conception of living well to which they take themselves to be adhering, as I will show, Anscombe holds that vice impedes one’s agency. Third, we need character to bring reason and appetite together; we could imagine that we continually have the thought ‘I am doing well’ while having a response that is simply a natural reaction, and this would seem to meet the second criterion just given. But through character the appetites take on a quality that reflects what we affirm, not simply our natural responses. This passage from Aquinas’s Commentary on Aristotle’s

126   John Hacker-Wright Nicomachean Ethics, a comment on book III, ­chapter 5, helps to bring this view out clearly. He states: The appetites are inclined to an object in two ways: one, by reason of a passion of the soul, the other by reason of habitus. Under the impulse of passion it happens that a thing is judged good as it is at present. Thus to one who is afraid of drowning it appears good at the moment to throw his merchandise overboard; as fornication to one filled with lust. But the judgment by which a man accounts a thing good in itself and absolutely, arises from the inclination of habitus. . . . [Aristotle] therefore says that since a man in some measure is the cause of his own evil habitus by reason of his continual sinning . . . it follows that he himself is the cause of the imaginative reaction that follows such a habitus, i.e., of the appearance by which this thing seems to be good in itself.13

Here, the object in question is an action; we can be drawn to an action at a particular moment through the mere impulse of a passion, but that falls short of the identification of the object as good in itself. The judgement or the connatural knowledge that an action is just or unjust is not a mere reaction, but rather an appetitive response that is informed by our prior choices and commitments. It is a voluntary response that is the outgrowth of appetites that are qualified by my normative conception of doing well, shaped by my prior choices. This is what it is to take an action to be good or bad in itself. A habitus relates me to particular occasions for action as good or bad because the habitus qualifies my appetites; it shapes them in response to ends that I have undertaken. With a nod to McDowell, we might say the habitus remakes some of our passive capacities, our sense appetites, so that they reflect our agency. The perceptual role of the virtues is not something incidental, but partly constitutive of having a will, since through them we relate to our perceived environment in a way that we otherwise could not: not through mere natural reactions but as a spontaneous and willed pursuit or avoidance. When Anscombe takes on Aquinas’s notion of connatural knowledge, she highlights the importance of inclination to shaping the virtuous agent’s perspective on the world. A generous agent perceives that an act is mean through being inclined against the action: ‘that inclination itself is a sort of perception of the meanness of acting even without the judgment being formulated. . . . If the judgment does get formulated, the formulation is an expression of what was already expressed in the rejection’ (GG1, 60). The inclination and the practical judgement possess a sort of equivalence in this formulation; it is an equivalence of content. Yet the inclination has a certain priority. First, it gives the knowledge specificity that it would likely lack without being driven by inclination, for the knowledge arises from ‘experience of life, of suffering, and above all of moral practice’. Hence, Anscombe says, ‘the knowledge itself, if highly theoretical, is not likely to be anything but rough and crude’ (GG1, 64). Second, the inclination is what makes my

13 

Thomas Aquinas, Commentary on Aristotle’s Nicomachean Ethics, trans. C. J. Litzinger (Notre Dame, IN: Dumb Ox Books, 1964), 229.

Practical Truth, Ethical Naturalism, and the Constitution of Agency 127 judgement a practical judgement and makes the knowledge I thereby formulate the distinctive sort of non-​indifferent knowledge at issue here. Further, just as what is wanted plays an essential role in making our thoughts about the world practical, character, as habitual desire, is necessary to make our practical thinking an expression of a normative view of the world. It is only through habitual desire that what is wanted can present itself as more than something contingently and fleetingly wanted; only the habitually desired, that is, what is desired as an upshot of formed character, reflects my will, my spontaneity. This addresses the worry that virtue might be gestural and otiose, as Korsgaard worries it becomes in Foot’s account. For Anscombe, at any rate, it certainly is not. Yet here one might press Korsgaard’s second worry. Why is it virtue alone that can constitute the will? Could not vice just as well constitute the will, reflecting my malign spontaneity just as virtues reflect the just spontaneity of the virtuous agent? Anscombe certainly recognises this issue. In her treatment of connatural knowledge, she acknowledges that the clever villain can know the things known by the virtuous agent ‘out of a sharpness of intelligence’, but confesses that she is unsure what to say about him (GG1, 64). Further, she acknowledges that the vicious agent is capable of acting efficaciously on habitual bad desires; he chooses his actions in the robust sense of prohairesis in Aristotle, which involves intellect and character and goes for actions as constitutive of means of the good way of going on. The akratic agent does not choose in this sense because what he aims at, even if he deliberates to get it, does not issue from his normative conception of how to live. But that is not the case for the wicked man, who is, it seems, fully on board with his own wicked acts: they represent his conception of living well. This raises the worry that any habitus (or hexis) would constitute the will, not merely virtue. Korsgaard takes it that Aristotle can avoid this only if we go with one of two apparently contradictory accounts he gives of the vicious agent. Specifically, we must adopt the account according to which the vicious agent is not unified in his appetites. Aristotle says in book 9 of the Nicomachean Ethics that ‘base people . . . are at odds with themselves and, having an appetite for one set of things wish for another, the way people who lack self-​control do’ (9.4 1166b18–​24). Elsewhere, Aristotle claims that ‘vice escapes its possessor’s notice’ (7.8 1150b35–​46), suggesting that the vicious agent is unselfconscious in his vice, and therefore unified in his pursuit of his evil ends. These certainly are inconsistent depictions of vice in Aristotle, but it is not clear that the resolution of Korsgaard’s second worry turns on adopting the first of these accounts. According to Korsgaard, the second account depicts the vicious person as having intact agency, efficaciously and autonomously going for an incorrect conception of the good, and so it seems that the principles of his intact agency are the vices. Yet Korsgaard is incorrect in thinking that conflicting appetites are the only way for one’s agency to be rendered defective through vice. Rather, as Anscombe shows us, vicious agents’ agency is defective because in acting viciously, they act under a false conception of what they are doing, and indeed, they are incapable of making true something that they are aiming to bring about: namely, that in acting as they do, they are acting well. Hence, according to Anscombe, the vicious eo ipso lack something necessary for non-​ defective human agency.

128   John Hacker-Wright

6.4  Practical Truth and Vice In the Nicomachean Ethics, Aristotle introduces a notion of ‘practical truth’, defining it as ‘truth in agreement with right desire’ (6.2 1139a30). In ‘Thought and Action in Aristotle’ and ‘Practical Truth’, Anscombe interprets this rather mysterious formula. It is useful, Anscombe suggests, to contrast this formula with ‘truth in agreement with desire’: that latter formula simply means ‘things being as a desirer wants them to be’ (GG1, 153). The formula that Aristotle gives seems to suggest an external standard being brought to desire, to determine that some people are desiring the wrong things. But then is the suggestion that here is some practical falsehood if things are as a vicious desirer wants them to be? The answer is, in a sense, yes, but because the vicious desirer cannot bring it about that things are as she wants them to be; the vicious agent is not capable of making things be as she wants them to be. Of course, if a vicious agent is clever, he can bring it about that some things are realised as he wants them to be, but there is more to ‘things being in accordance with choice’ than this. When a mature agent deliberates and acts, he is doing so with an implicit belief that he is acting well, and ‘doing well’ is what the vicious agent cannot bring about because of his misguided aims. His actions will therefore necessarily be practically false. It may nevertheless be unclear how an action can be false; Anscombe explains this view using some key notions from her theory of action.14 First, Anscombe takes it as obvious and uncontroversial that an action can be variously described, and that it is voluntary or intentional under some descriptions and not under others. For example, an action may be intentional under the description ‘putting the book on the table’ while not intentional under the description ‘putting a book on a puddle of ink’ (CP2, 208). If someone asks me why I am putting the book on a puddle of ink, the answer is presumably that I didn’t realise I was doing that, and hence that was not part of my intention. On the other hand, if I am putting the book on the table intentionally, this will be something I know that I am doing. This knowledge is practical knowledge, which is the cause of what it understands (I, §48, invoking Aquinas’s formulation of practical knowledge from Summa Theologica I-​II, q. 3, a. 5). As Rachael Wiseman aptly points out, this is not a definition or explanation of practical knowledge.15 The formula makes sense only in the context of Anscombe’s Aristotelian account of practical reasoning. On that account, the paradigmatic case of intentional action is one in which I can give a reason that addresses the question ‘Why?’ addressed to my action. If I am pumping some water from a cistern, I can answer that I am replenishing the water supply in the house: there is a practical, calculative order to my actions; I am doing y in doing x. There may be things I am doing for no particular reason, but then the

14  My concern here will not be whether Anscombe’s views on practical truth accurately interpret Aristotle but, rather, with the importance of her conception of practical truth for her ethical views. 15  Rachael Wiseman, Routledge Guidebook to Anscombe’s Intention (London: Routledge, 2016), 174.

Practical Truth, Ethical Naturalism, and the Constitution of Agency 129 question still has an application: I still know what I am doing, even if I am doing it for no particular reason. Although I can go wrong in my performance or be mistaken about what I am doing (say, because there is, unbeknownst to me, a hole in the pipe through which I am pumping the water), my practical knowledge is the cause of the execution of my intentions. That is to say, it is because I know that I am pumping water to refill the water supply that I am doing so, when I am. Another way of putting this is that my knowledge that I am doing something under a certain description makes it true that I am doing that, when I am. There is a sense of ‘making true’ that applies distinctively to action: by signing a contract I make it true that the contract is signed, and this is distinct from any sense in which it might be said that a dog makes it true that my hand is bleeding from having bitten it or a falling branch makes it true that my teapot is broken (GG1, 157). This way of speaking, according to Anscombe, would be trivial and pointless. The notion of practical truth marks a distinct kind of cause, one that we ourselves are, inasmuch as we have causal powers that rest in ‘desiring thought or thinking desire’, as Aristotle describes it. The notion of making true has a point in this domain because it is integral to our agency, and there are interesting ways in which we can fail to make true the description under which we suppose ourselves to be acting. Practical truth, as Anscombe interprets it from Aristotle, applies to truth ‘that one produces in acting according to choice and decision’ (GG1, 155). Hence, she is speaking about actions that are deliberately undertaken as constitutive means to living well, actions that are seen as part of living well, or praxeis. These actions are a subset of human actions, which includes the class of all voluntary actions. Human acts comprise a class that is itself a subset of acts of a human being. The later class includes involuntary actions, such as tripping on the carpet and flying across the room. Voluntary actions, for Anscombe, include omissions, which are not done intentionally. In omission, I fail to act in a way that I could have acted. As Anscombe puts it, ‘Consent may reside in not taking care where you could have and, in the nature of the case as available to your understanding, you needed to’ (GG1, 209). In voluntary acts of this sort, there is a failure to undertake a deliberate action that is necessary for me to live well. On Anscombe’s view, ‘all human action is moral action. It is all either good or bad. (It may be both, GG1, 209)’. The view is that whenever we act voluntarily, there is a question of whether we have acted well or not, a matter of whether our reason and inclinations are oriented towards what is genuinely good, towards a correct conception of living well qua human being. Every human action can be given a moral description; there are of course many morally neutral ways of describing an action, but that does not completely describe the action as under the command of reason. Only a moral description completely describes the action and classifies it as good or bad. On Anscombe’s view, an action is good if in no way bad, since it is not in conflict with living well qua human. Actions such as walking around the block or nibbling a mushroom can be good actions provided they would not be chosen if they were in conflict with other things that one ought to be doing; they are then episodes of praxis that are chosen as part of living well. These actions Anscombe describes as ‘innocent’ (GG1, 205).

130   John Hacker-Wright A virtuous agent will have insight into the moral nature of his actions through connatural knowledge; the things he is inclined to do will be things that are a part of his conception of living well, which is a sound conception. In acting on his virtuous inclinations, then, he will know he is acting well and thereby make true that he is so living. And this is Anscombe’s conception of Aristotle’s ‘truth in agreement with right desire’. But the vicious agent cannot bring about practical truth in that sense. As she puts it: The man who forms and executes an evil ‘choice’ will also make true some description of what he does. He will secure, say, if he is competent that such and such a man has his eyes put out or his hands cut off, that being his judgment of what it is just to do. But his description of ‘justice performed’ of what he has done will be a lie. He, then, will have produced practical falsehood. (CP1, 77; see also GG1, 155)

The vicious agent cannot make true that he is acting justly and doing well in acting as he does. The vicious agent’s actions (praxeis) are practically false because they do not exhibit acting well, as he, at least implicitly, claims that they do. And it is part of ‘things being in accordance with our choice’ as Anscombe understands it that we make it true that we act under the description acting well when we are acting as we see to be acting well. On this account, then, being unable to make true that we are acting well is a significant limitation on our agency: as a vicious agent, I cannot fully make things be the way I want them to be. This, then, is a way that the vicious agent’s agency is compromised by vice even though there is no disharmony in the vicious agent’s appetites as he experiences them. As Anscombe points out in her treatment of connatural knowledge, a clever agent may be aware of which actions are mean and unjust, achieving through a sharpness of intelligence what the rest of us achieve through experience of life and moral practice. Anscombe says she is uncertain of what to say about such cases, but surely what is to be said here is that the application of intelligence in this case is to anticipate the behaviour of those who are guided by justice and generosity so as to carry out his own nefarious ends more effectively. His knowledge of virtue is therefore not simply indifferent knowledge; it advances his own aims to know which things are just, but not as it advances the aims of a just person. Perhaps, like Callicles, he thinks those who follow what is just are weak, and his own conception of living well preys on the vulnerabilities of those who follow conventional justice. Hence his knowledge of justice serves him as a means to unjust ends. He is in no way moved to follow what is just, and he does not through acting on his flawed conception of living well make it true that he acts well. Another case that Anscombe mentions but does not address is that of the wicked man who keeps himself physically fit so as to pursue his evil purposes. Is he acting well in so doing? It is clear from what Anscombe says elsewhere that this action inherits badness from the further evil ends for which it is done; although it is not per se a bad type of action to intend, the further intention with which it is performed makes that particular act of exercise bad. But what of a villainous agent who packs a nutritious lunch for his

Practical Truth, Ethical Naturalism, and the Constitution of Agency 131 child and sends her off to school?16 In this case, the action is not instrumentally related to the villain’s vicious ends. It is also clearly not an end that in any obvious way makes his conception of living well flawed; indeed, if anything, we are inclined to see this episode as redeeming. Yet the question arises how this episode fits within the villainous agent’s life as a whole. It might seem that the villainous agent could at least make it true that in fixing his child’s lunch he acts well. Yet the action is arguably at odds with other things he is doing in his life that damage the moral environment of his child. His criminal actions actively damage the social order that his child is growing up within. Since his aims fail to harmonise, he is not acting fully rationally, even as he packs this lunch, and hence he is not acting well and cannot make it true that he does. Two further points are worth bearing in mind about this account. It is not the case that a virtuous agent can always make true that she is acting well. After all, she may find herself situated in a tragic dilemma, in which case whatever she does is so horrible that it cannot count as good, and as a result her life is marred.17 It would go beyond the scope of this chapter to fully explore such cases, but it is important to note that tragic dilemmas can arise through no fault of the agent. In such cases, the world may be set against our constituting ourselves as agents and undermine our possibilities for living well. In other cases that don’t rise to the level of tragic dilemmas, we might find ourselves uncertain about what to do. For example, punishing children for misbehaving often seems to present such uncertainty: will it help the child to see what was wrong in what they did or only alienate them and make them more intransigent in their wrongdoing? It may be that further deliberation or experience would yield the insight about what is required, but it seems at least possible that there will be areas of uncertainty in the lives of even the ideal phronimos. About this point, the correct thing to say is that the connatural knowledge of the virtuous agent renders an action practically true when one has such knowledge, but even the virtuous may on occasion lack such knowledge. In such a case, reason cannot fully determine an action. We may, in such a case, make something about our action true, and we have practical knowledge that we are doing that, but our action doesn’t come completely under the command of reason because we are uncertain about whether it is indeed good. In such a case, we may simply lack the insight to constitute ourselves fully: we simply do not know how best to live. But here we get to Korsgaard’s third challenge, which she thinks the most insurmountable for the putative Aristotelian constitutivist. It seems that Aristotelian constitutivism cannot meet what she calls the self-​constitution requirement. The worry is this: if we are not fully responsible for our character, then we cannot be morally responsible for our actions. And the answer that I have given here to the second challenge—​that vicious agents’ agency is defective through being unable to make true that their actions are instances of living well—​seems to undercut the possibility of addressing the third

16  17 

A case raised by Sergio Tenenbaum in discussion. See Rosalind Hursthouse, On Virtue Ethics (Oxford: Oxford University Press, 1999), 71–​77.

132   John Hacker-Wright challenge. For it seems they may act in the belief that they are acting well even if the virtuous agent knows better.

6.5  Self-​C onstitution and ‘The Moral Environment of the Child’ We have just seen that the completion of my agency depends on my ability to make true a moral action description that applies to what I am doing. In particular, I have to be able to make true the description of my action as an instance of acting well. If, through vice, I cannot do that, then my agency is damaged: there isn’t a true description of my action according to which I am acting well because I am acting on a flawed conception of how to live. The challenge that a would-​be Aristotelian constitutivist faces from Korsgaard is to address how we can be responsible for our character. It seems that through a poor upbringing, we might end up with a flawed conception of how to live and be unaware of it as such. If the knowledge that I am not acting well is unavailable to me, which seems to follow from the previous section’s argument, then surely we cannot be responsible for the possession of that flawed character or the actions that issue from it. In ‘The Moral Environment of the Child’ Anscombe addresses some features of her view of moral development that provide a reason to think this objection can be answered by an Aristotelian constitutivist. There she writes: Moral action descriptions are not natural event descriptions. But it is part of the natural history of mankind that the human young acquire concepts corresponding to them, or in some cases, at least concepts in which they are rooted, as adultery is in marriage, or stealing in that of property. . . . Quite generally: to grow up as a child of normal intelligence in a human society is eo ipso to be equipped with a range of concepts which form the raw material of moral action descriptions, and in many cases to acquire these as well, at least in rough inchoate form. (GG2, 225)

The focus here is on the acquisition of the concepts that provide the possibility for moral action descriptions. Since, as we have seen, getting an action under a true moral action description is essential for making one’s actions responsive to reason, and the latter is essential to human agency, it is not surprising that Anscombe would consider this to be part of the ‘natural history of mankind’.18 She emphasises in this essay that even an upbringing that fails to impart ‘moral convictions and sentiments’ will impart these concepts. A human environment just is a moral environment, on her account, and for

18  With this phrase Anscombe no doubt alludes to Wittgenstein’s Philosophical Investigations (§§25, 415), which highlights the centrality of practices such as ‘[g]‌iving orders, asking questions, telling stories, having a chat’ to our natural history.

Practical Truth, Ethical Naturalism, and the Constitution of Agency 133 that reason ‘even if it is a bad [environment] it will be full of matter that provides the raw material for the ethical’ (GG2, 233). If true, this yields the conclusion that moral action descriptions are within reach even for someone who is badly brought up; there is the possibility that anyone can be brought to understand moral action descriptions. Yet this might seem small consolation given what she says about connatural knowledge. Since this knowledge is dependent on possessing moral virtue, she seems to adhere to Aristotle’s view that ‘virtue makes the aim right’ (Nicomachean Ethics 6.13 1144a7–​9), in which case, it is doubtful that she can answer Korsgaard’s challenge since the possession of connatural knowledge still hinges on whether one has acquired good character in one’s upbringing. As noted earlier, in her treatment of connatural knowledge Anscombe treats the idea of a villain who discerns what is just and generous through the application of cleverness. Taking her use of ‘cleverness’ here in an Aristotelian sense, the idea seems to be that the villain is applying intellect alone to discern what for virtuous folks is a matter of inclination. The mechanism of this mystifies Anscombe apparently because of the crucial role that the inclinations play in discerning just from unjust; the intellect cannot precisely fit under concepts what the inclinations can very well grasp. As she puts it, ‘the knowledge itself, if highly theoretical, is not likely to be anything but rough and crude’ (GG1, 64). Yves Simon gives the example of an experienced businessman who rejects a business proposal that appears good and honest, explaining to a less experienced partner ‘sometimes you can smell things you cannot see’.19 The metaphor of smell implies there are distinctions between cases that are difficult or even impossible to articulate but that inclination can correctly pick out. On such a view, moral virtue would seem to be required to make the finer distinctions, and hence it would be perplexing that there can be clever villains who can make out the distinctions by sheer force of intellect. Yet, if I am correct in suggesting that the villainous are not indifferent to the distinctions, but rather have a different affective relation to them than do the virtuous, then it is not the case that they are discerning by sheer force of intellect. In ‘Authority in Morals’, Anscombe seems to recognise that villainous people often have partial moral knowledge that might well be grounded in inclination: We sometimes imagine someone with a terribly bad upbringing, who is taught all sorts of misbehavior as right, and taught to despise much that is good and we think: what about such a person? But people of the most horrible principles know quite well how to cry out against injustice and lying and treachery, say, when their enemies are guilty of them. So they in fact know quite a lot. (CP3, 45)

From this passage it seems clear that those with a bad upbringing can recognise and are not indifferent to injustices perpetrated against themselves. Hence, it would seem to follow that they have some knowledge that is also grounded in inclination; it is simply that this knowledge is incomplete. 19 

Yves Simon, Practical Knowledge (New York: Fordham University Press, 1991), 19.

134   John Hacker-Wright This idea further suggests that the vicious are inculcated into some use of concepts like justice and honesty that are not simply indifferent, and also not simply a matter of learning to exploit those who apply virtue terms sincerely; the vicious also apply those terms in propria voce and sincerely, yet incorrectly, failing to recognise the injustices they perpetrate in the name of, say, their own pleasure; they may at the same time care very much about genuine injustices of others towards them. In other words, it may be correct to say of the natural history of human beings not just that we are inculcated into the use of concepts that figure in moral action descriptions but that we also learn to care about justice in some form. We don’t necessarily learn about the proper application of this concept, of course, but the ‘ethical raw materials’ that we get from the human environment are more than intellectual, even if they are incomplete and twisted. This further suggests that the villainous are not precluded from coming to grasp the proper application of such terms through their upbringing. Rather, it is a matter of coming to exercise their rational powers to correct and perfect their conception of justice, which requires an application of the will. Although still somewhat sketchy, the line of thinking I am tracing in various of Anscombe’s papers suggests a Wittgensteinian neo-​Aristotelian response to a view that is grounded in a supposedly corrective interpretation of Aristotle. That view is that, despite Aristotle’s claims to the contrary, a bad upbringing must be morally catastrophic, given Aristotle’s views of habituation, choice, and vice. On his views, properly understood, one with such an upbringing cannot be held responsible for their defective character. This is the view Korsgaard is advancing in her scepticism about the possibility of Aristotelian constitutivism. Terence Irwin likewise argues that someone strongly conditioned to be a Nazi may be held responsible for speeding or stealing from fellow Nazis, but not responsible for his Nazi values or actions that befit a Nazi, since these are not susceptible to deliberative argument.20 An Anscombean response to this view starts with questioning this view of habituation. One is not inculcated into a specific sort of life as the only possible available conception of living well, even by a bad upbringing. Or we might say, if there were such an upbringing, it would be one that somehow foreclosed the possibility of practical thought of the sort that is essential to human agency. Though there are environments that may bring that about, what happens in them surely does not count as upbringing. Presumably, the Nazi agent is capable of thinking that he is living well through a relatively specific set of values, and if so, he is capable of considering other possible courses of values, including non-​Nazi specifications of living well, even though he may be strongly inclined to dismiss them. As Anscombe states in ‘Practical Inference’, ‘It is human to have generic ends’ (GG1, 142). Generic ends are those that can be specified in various ways; the Nazi takes his racist mode of life to be a relatively specific way of fulfilling the highly generic end of living well. Of course, the Nazi is wrong: the Nazi way of life is not a specification of living well. Still, to have generic ends is to have ends that are otherwise specifiable, and the Nazi has those.

20 

T. H. Irwin, ‘Reason and Responsibility in Aristotle’, in Essays on Aristotle’s Ethics, ed. Amélie Oksenberg Rorty (Berkeley: University of California Press, 1980), 140.

Practical Truth, Ethical Naturalism, and the Constitution of Agency 135 I think such thoughts lie behind Anscombe’s claim that connatural knowledge is ‘not unavailable to us who are not virtuous’ (GG1, 66). Rather, it is there, though ‘it may be there as a mere glimmer whose sign is the understanding of human language with all its multifarious action and motive descriptions, its machinery for accusing others and excusing oneself ’ (GG1, 62). Hence, we return to the idea from ‘The Moral Environment of the Child’, and we see now that Anscombe is committed to the idea that there is some training of the desires and affects in the acquisition of moral descriptions, though that may well fall short of a specific moral training in ‘moral convictions and sentiments’ (GG2, 226). The availability of connatural knowledge means that the vicious can get this knowledge through an application of the will. And conversely, they may actively keep that knowledge at bay through ‘affected’ or ‘politic’ ignorance, which is a matter of avoiding that which would require one to reconsider plans and commitments. Hence, Anscombe rejects the idea some find in Aristotle that bad habituation is morally catastrophic. She believes that the vicious can reconstitute themselves. Hence, the sort of failure of self-​ constitution that vice is, for Anscombe, one that prevents the vicious from making true that they act well in acting as they do. But they are not prevented from acting some other way and eventually making it true that in doing that they acted well. It would, of course, require a reconsideration of what it is to act well and habitual commitment to that view to bring that about. But it is at least possible that this could happen, on Anscombe’s view. Of course, there is clearly some truth to the catastrophic account: since the commitment to a certain view of doing well is a matter of habitual action, there is no way for someone to change that view at the time of a single action (CP1, 71). At that point, the ignorance a vicious person exhibits is ‘invincible’ in that he cannot overcome it at the time of action, but it is nevertheless not excusable, since he ought to have known better (GG2, 106). At the time of action, what might amount to a revision in his vicious normative conception can count only as a surprising inclination accompanied by thoughts like ‘Now I see! I’ve gotten it wrong all along!’ If it is indeed followed by a revision in habitual action (assuming the occasions are provided), then the first action will count as a revision to his normative conception of how to live. Surely, this will not be an easy change to make, if one has been strongly committed to a different way of living. Realistically, such a change might involve the sort of backsliding and doubt that so often accompanies major life changes. So the catastrophic account gets it right that it is not up to the agent to change on a dime. But if the Anscombean account I am expounding here is correct, it does not imply a lack of responsibility on the part of the agent, for he could have adopted a different conception of how to live earlier. The limit case, perhaps, is someone who is raised on Nazi views until he begins to reason for himself. He first begins to reflectively apply his conception of living well in deliberately chosen intentional actions, starting out with that flawed conception, which he immediately calls into question in favour of a sound normative conception. He then has to prove his new commitments through ensuing actions that continue his new, correct conception, before he can claim responsibility for the goodness of those first actions, even if he cannot reject responsibility should he fail to carry through with his new conception. The initial action would be a

136   John Hacker-Wright fledging action that registers as bad to him in his backsliding into Nazism and to us as a disappointing fluke.

6.6  Conclusion I have outlined an interpretation of some of Anscombe’s ethical writings on which I have taken her to be an Aristotelian constitutivist. Against Korsgaard, I have shown that there are resources in Anscombe’s thinking for a plausible Aristotelian constitutivism. In particular I have attempted to show that Anscombe can take on the challenge of showing that we are responsible for our character in a way that goes beyond what we find in Aristotle. Knowledge of how to live is, on Anscombe’s view, available to us all, even if some of us take pains to suppress and distance ourselves from such knowledge. Hence, we are fully responsible for making ourselves into the sort of person we are, for constituting ourselves well or badly. Anscombe’s extensive exploration of action bears fruit for ethics in this regard. We can see that, contra Korsgaard, an Aristotelian constitutivist need not be committed to a ‘naturalistic conception of action’. Korsgaard takes Aristotle to be committed to the view that we act only when our actions are caused by our state of character: the action then expresses our normative conception of the world. This is obviously not Anscombe’s conception of action: when I act intentionally I have practical knowledge of what I am doing, but this is related as a formal cause to what I am doing. When I have connatural knowledge, my practical knowledge also contains awareness that in so acting, I am acting well, and given that this knowledge is available to us all, we must act on this knowledge so as to constitute ourselves as agents who are capable of practical truth. This, it seems to me, reflects a normative conception of action along the lines that Korsgaard herself is offering. Hence, there are at least two conceptions of constitutivism that are capable of meeting the self-​constitution requirement, and not only the Kantian conception, as Korsgaard claims. Of course, there will be differences between these two versions of constitutivism, as the Aristotelian moral psychology differs in many ways from the Kantian moral psychology. My only goal here has been to show that Anscombe has resources to meet some prominent objections against the very idea of Aristotelian constitutivism.21

Bibliography Aquinas, Thomas. Commentary on Aristotle’s Nicomachean Ethics. Translated by C. J. Litzinger. Notre Dame, IN: Dumb Ox Books, 1964. 21 

I would like to thank Jennifer Frey, Matthias Haase, Anselm Müller, Evgenia Mylonaki, Candace Vogler, and the audience at the 100 Years of Anscombe Conference in Athens, Greece.

Practical Truth, Ethical Naturalism, and the Constitution of Agency 137 Aristotle. Eudemian Ethics. Translated by Brad Inwood and Raphael Woolf. Cambridge: Cambridge University Press, 2013. Aristotle. Nicomachean Ethics. Translated by Roger Crisp. Cambridge: Cambridge University Press, 2004. Doyle, James. No Morality, No Self. Cambridge, MA: Harvard University Press, 2018. Foot, Philippa. Natural Goodness. Oxford: Oxford University Press, 2001. Frey, Jennifer A. ‘Happiness as the Constitutive Principle of Action in Thomas Aquinas’. Philosophical Explorations 22, no. 2 (April 2019): 208–​221. Frey, Jennifer A. ‘How to Be an Ethical Naturalist’. In Philippa Foot on Goodness and Virtue, edited by John Hacker-​Wright, 47–​84. Cham: Palgrave-​Macmillan, 2018. Hacker-​Wright, John. ‘Ethical Naturalism and the Constitution of Agency’. Journal of Value Inquiry 46, no. 1 (March 2012): 13–​23. Hacker-​Wright, John. ‘Virtue Ethics without Right Action: Anscombe, Foot, and Contemporary Virtue Ethics’. Journal of Value Inquiry 44, no. 2 (June 2010): 209–​224. Hursthouse, Rosalind. On Virtue Ethics. Oxford: Oxford University Press, 1999. Irwin, T. H. ‘Reason and Responsibility in Aristotle’. In Essays on Aristotle’s Ethics, edited by Amélie Oksenberg Rorty, 117–​155. Berkeley: University of California Press, 1980. Korsgaard, Christine. ‘Constitutivism and the Virtues’. Philosophical Explorations 22, no. 2 (April 2019): 98–​116. Korsgaard, Christine. ‘The Normative Conception of Agency’. In Rational and Social Agency: The Philosophy of Michael Bratman, edited by Manuel Vargas and Gideon Yaffe, 190–​214. Oxford: Oxford University Press, 2014. Korsgaard, Christine. ‘Realism and Constructivism in Twentieth-​Century Moral Philosophy’. In The Constitution of Agency, 302–​326. Oxford: Oxford University Press, 2008. Korsgaard, Christine. Self-​Constitution. Oxford: Oxford University Press, 2009. McDowell, John. ‘Two Sorts of Naturalism’. In Mind, Value, and Reality, 167–197. Cambridge, MA: Harvard University Press, 1998. Simon, Yves. Practical Knowledge. New York: Fordham University Press, 1991. Suto, Taki. ‘Virtue and Knowledge: Connatural Knowledge According to Thomas Aquinas’. Review of Metaphysics 58, no. 1 (September 2004): 61–​79. Wiseman, Rachael. Routledge Guidebook to Anscombe’s Intention. London: Routledge, 2016.

Chapter 7

Criteria l i sm Gavin Lawrence

To show the fly the way out of the fly-​bottle. —​Wittgenstein, Philosophical Investigations (1953), 309 But we told ourselves that this was only a quite special case of deriving; deriving in a quite special garb, which had to be stripped from it if we wanted to see the essence of deriving. So we stripped those particular coverings off; but then deriving itself disappeared.—​In order to find the real artichoke, we divested it of its leaves. —​Wittgenstein (1953), 164a; (versus Tractatus 4.002, on logic lying behind, at the core of, everyday language) Only in the flow (Fluß) of life do words have their meaning. —​Wittgenstein (Last Writings Vol. 1. 913; cf. Zettel 135; ‘im Strome,’ ‘im sprachlichen Verkehr,’ Malcolm (1958/​1966) 93; 90)

7.1  Introduction It is a joy, and a privilege, to write on Anscombe, and to do so in connection with her brief, enduringly brilliant, paper ‘On Brute Facts.’1 It is part of her fertile development of later Wittgenstein’s focus on meaning, in the varied range of uses that naturally makes intention prominent—​as in ‘what do you

1 

I am especially grateful to the editor Roger Teichmann for many insightful comments; and also to my colleagues Barbara Herman, Calvin Normore, and Andrew Flynn.

Criterialism   139 mean by that?,’ whether of doing or saying (PI 693)2—​and that rejects philosophical conceptions of it as a primitive psychological state, an accompanying feeling or ‘inside’ private act, one supposedly needed to give life, form and drive, to ‘outside’ performance, without which that appears shapeless or mere parroting (AWL 49; cf. BF paragraph 4; I §§3, 27). The approach invites, or rather commands, us to see language as part and parcel of overall human behaviour (PI 23b, 7d, RPP I.151; LA I.5; cf. TLP 4.002(b)), and this in all sorts of ways: e.g., ‘Block!,’ I bring it; ‘Fire!,’ I run—​or pull the trigger; you ask a question and I shrug; I smile, you laugh, she says ‘How funny!’ Thus Wittgenstein’s emphases, for example: (a) On how our language-​games are multi-​form, evincing a diversity of use both ‘external’ and ‘internal’—​in the sense that describing, far from being the only game, is one amongst many, and moreover is itself multiform, as equally with others, including explaining, justifying, giving a reason (Mannigfaltigkeit, PI 23, 24, cf. 108a.). (b) On how we explain meaning by articulating the grammar of their ‘moves’, how our ways of talking about something, along with our behaviour generally in its regard, compose and articulate, what it is, viz. what questions and inferences, what consequences, actions and reactions, are in place; whether surprise, expectation, interruption, et al., are so; what doubts are contextually intelligible in what varying circumstances, and what mistakes (OC 154–​155); how qualifiable, and how not; whether admitting of degree, of countability, of distinguishing simple from more complex cases (LW I.967); what count as easy, hard, or grey cases; what happened then, before, and after (e.g., PI 179; PI Part II.ix; QLI; cf. LSD 20; RPP I.151); the occasions and contexts of use (‘Three pence!’: LA I.5.n2). Matters both general and particular: e.g., the applicability of questions and other behaviours to general categories (What? When? Where? How?) and to particularities (Cuttable? Scoopable?). In short, grammar as essence (PI 371, 373): ‘grammar’ understood as giving definition to, contouring, giving form to, what we talk about, not language on one side, world on other (cf. PI 370). Our world is shaped, articulated, by our grammar—​grammar as not only language, but our whole current form of life, our culture, wherein language is embedded and contextualised (RFM 335). And so, in the case, for example, of water, as something with a nature to be scientifically investigated, progressively, experimentally: what a nature is in such a use, within the grammatical variety of ‘What is it?’ questions, and equally of ‘scientific’; or with intentional action, delineating the relevant, in-​forming, sense of ‘Why?’ (I §§5–​6, 9–​16, 17–​18; etc.). Our language-​games, like our multifarious tools, say, for farming, while fashioned to serve our needs and interests, are equally shaped by the world as we find it and what it takes to get a handle, a grip, on it. (Not unlike the hand itself.) (c) On how individuals learn these moves, as with games, and our tools, linguistic, attitudinal, and other, for teaching them—​stopping and forcing modals, reinforcing encouragement, tone, hand, etc. (e.g., PI 143b, 145, 53b). (d) On how we frame games, create rules of many kinds, 2 

‘PI’: This and other abbreviations of works by Wittgenstein will be found explained in the References. Other abbreviations in this chapter are to works by Anscombe and are explained in the Title Abbreviations at the start of this volume.

140   Gavin Lawrence and other normative notions, that Anscombe characterises as ‘naturally unintelligible.’3 And (e) even on how basic language itself grew by being inserted into, exploiting the interstices, the ‘ecological’ niches, of even more primitive human behaviour and natural reaction (Z 545): think how an involuntary cry of death-​pain resonates as an alarm to others, inviting—​creating the space for—​being mimicked and diversely exploited, now as warning of danger, now as deception, an intentional or pre-​intentional use of sound. These games, further behaviours, made available by, and exploiting, such more basic, contingent, natural facts of life, then in turn become themselves part of a richer background, thus creating an exponential proliferation of opportunity for further exploration and exploitation, as also for internal variety.4 (Think too of other sounds and gestures: of the hunter’s imitation of birdsong, the drum sound of feet on earth, the storyteller saying ‘The rain fell—​pitter-​patter,’ their fingers twinkling downward.)5 The thrust lies against much modern semantics, or to sidelining it and its, at times obscure, promissory purposes. Similarly, economic theories of rationality, while possibly useful for specific technical, scientific purposes, positing, say, ‘purely’ self-​interested agents, are thus tightened for a purpose and are evidently false as general accounts of human practical rationality—​just as also, for Marxists, the basic tenets of political economy, assumed as its archai, are questionable as the basis for larger, humanly normative, economic understanding. Theoretic, ‘pure’ semantics seems equally questionable, certainly fine for some specialised technical

3 

‘After all, once this transformation has taken place, the following is true: in such a case you are told “you can’t” do something you plainly can, as comes out in the fact you sometimes do. At the beginning, the adults will physically stop the child from doing what they say he “can’t” do. But gradually the child learns. With one set of circumstances this business is part of the build-​up of the concept of a rule; with another, of a piece of etiquette; with another of a promise; in another, of an act of sacrilege or impiety; with another of a right. It is part of human intelligence to be able to learn the response to stopping modals without which they wouldn’t exist as linguistic devices and without which these things: rules, etiquettes, rights, infringements, promises, pieties and impieties would not exist either’ (RRP, 101), and, with rules, comes all language (97). Anscombe here builds from Hume’s account of promising. Yet there is also natural normativity. Not all nature, especially organic nature, turns out ‘naturally intelligible’, understood as flat or non-​normative, in Hume’s sense of mere regularity between independently identifiable items: think of the grammar of natural processes (e.g., Geach 1975). Specifying suitably independent relata is problematic for empiricism (e.g., behaviour descriptions are either too near or too far). 4  As this example suggests, much basic life is environmentally communal, turning on similar reactions natural across species (flight before fire, gathering around fallen prey, or the tree in fruit): much behaviour—​doing and reacting—​is of its nature inter-​specific (predating and its evasion). The behaviour of other species, as of individuals of the same species, is itself part of the environment. There is advantage, let alone necessity, in having such a more generic, wider than species-​specific, awareness of the environment. (For example, among uses of alarm calls, [a]‌signalling to stealth predators they have been spotted, ‘perception advertisement’; [b] ‘eavesdropping’ between species; and [c] deceiving, as with fork-​tailed drongos exploiting pied babblers [Ridley et al., 2007]. See Zuberbühler (2009), (a) 279, 284–​285, 287; (b) 278, 296, 300–​304; (c) 289, 308–​309; and references there. Criteria for some predicates (‘pain’) are not obviously, or obviously not, anthropocentric extensions, but rather grounded more primitively in a wider world of information, behaviour, and interaction. 5  Quite normal in telling children stories; see also, e.g., the griot/​griotte tradition in West Africa.

Criterialism   141 end, e.g., articulating a formal system, a calculus, for some machine purpose, but its import and power to illuminate ‘the entirely fluid use of words’ (AWL 48), the ever evolving workings and creative purposes of human natural language, its aspirations to constitute a complete, ‘rigorous,’ explanatory core of what it is to understand or master a language or to precisify it or even underpin its intelligibility, are presumed rather than justified (PI 23, 81, 89ff.; cf. 65, 109; cf. Carnap 1963, 28–​29; Strawson 1963).6 Each writes its preferred basic assumptions into defining rationality, economy, or meaning, and then entraps us by trading on the ambiguous interface between investigating a topic restrictively defined from, and for, a specific technical perspective and the unrestricted topic now supposedly rendered intelligible by this theoretical perspective, now finally put on a proper ‘scientific’ footing where before there had been no, or only ‘folkloristic,’ imprecise, apprehension (cf. TLP 3.323–​325). An idealisation, or ‘purity,’ for a certain purpose transmutes into an absolute, an unconditional, Ideal—​like an ideal ‘mechanical jurisprudence’ (Hart 1961, 124–​125, 135), the a priori order of the world (the order of possibilities) captured in a perfect, complete, utterly exact, logical language (PI 97, 101–​ 108): the God of language, Logic (theos ēn ho Logos), operating, like Him, outside or pre-​ Life, behind every sentence of every language, constituting a single, every-​which-​way determinate, static, crystalline, structure (PI 107–​108; the universal form of the proposition, 65, 92–​93, 108, 134–​135, etc.; cf. AWL §15, pp. 67–​68), ‘a regimentation of language’ (PI 130), a machine, a pianola (PI 157c). Theology semanticised, semantics theologised. In its construction and delineation, its ‘in-​forming’ of the say-​able, it exactly determines the thinkable and the be-​able (PI 97a; cf. TLP 4.114–​115). Doesn’t there have to be clear logical form and structure hidden behind the messy ‘imprecision’ of colloquial, as-​she-​ is-​spoke, language (Umgangssprache, TLP 3.4–​4.002 vs. PI 120)? Yet this is an illusion with fugitive content. The illusion set aside, our stance is not thusly theoretic, not one of super-​concepts (PI 97), super-​science over data. Language-​mastery, like rational action which it exhibits and in which it is embedded, is constitutive of human creativity, of dynamic life. The priority of Life and Praxis over the partial perspectives of theory and techne: viewed orthogonally, no need here for an orchestrating scaffolding behind. The world is shaped, structured, not by Logic sublime but by that now ‘rotated . . . around our real need as pivot,’ our evolving diversity of purpose: hidden Logic reconceived as open Grammar (PI 108a, 97): as Life, upfront, if infinitely complexifiable. Human

6  More recent logic and formal semantics attempt to develop further powerful tools to get a theoretic handle on natural, life-​embedded, language and reasoning. For example, dynamic semantics; non-​ monotonic, ‘defeasibility’, logic (e.g., Horty 2012); developments in ‘super Semantics’ that seek to acknowledge and accommodate the multi-​modality apparent in both language-​internal and language-​ external directions, i.e., exploring how to integrate spoken language, signs, gestures, body language, iconic resources generally, in order ‘to have a complete theory of meaning in a natural language’; and again how to extend its applicability to other areas, viz. pictures (static and dynamic), music, dance, and animal communication and meaning (Schlenker 2018). To what extent such attempts can be successful, and to what ends, lie outside our direct consideration. (While notions of completeness, and of ‘complete theory’, need to be given sense, attempts at formalisation reveal, at least in negative silhouette, something of the complexity of natural language, and human life.)

142   Gavin Lawrence reflectiveness, ordinary, yet extraordinary and creative, is part of that, producing industry and art, and attentive grammatical description and clarification, like Aristotle’s work on prepositions, or Anscombe’s on ‘Why?’7

7.2  Two Fly-​B ottles TLP, resonating with a Kantian demarcation between natural causality and freedom, extruded value from the effable world of a priori factual possibilities. Anscombe innovatively showed how it never left the world, world once restored, orthogonalised, by later Wittgenstein; showed how to avoid the traps ensnaring modern moral philosophy: ‘the’ Fact-​Value distinction, and the Practicality of (Moral) Evaluation. Our focus here is the first, the Fact-​Value distinction. Crudely, it seemed that evaluation, and, in particular, moral evaluation, had, if objective, then to be super-​objective, a matter of truth in some sui generis realm of reality outside ordinary natural factuality, a commitment with further roots in a certain Platonism and theism, in the model offered by geometry, and curiously bolstered by Hume’s strictures against its being any already intelligible ‘relation of ideas’ or ‘matter of fact,’ and by Moore’s allegation of naturalistic (or metaphysical) fallacy; alternatively, if not super-​objective, then subjective, a matter not of fact but of human taste, attitude, and reaction. Each option had apparent pros and cons. However, the problematic nub was insufficiently recognised: bad philosophy of language. ‘Good’ and ‘bad,’ for instance, didn’t work as either side supposed. These were forced against their grammar, and so nonsensical. This, where acknowledged, led to apparently ad hoc semantic responses: one side posited some special ‘moral’ use, with a mysteriously peculiar grammar;8 the other posited such alternatives as error theories, or, given the multi-​functionality of language, non-​descriptive grammatical roles, expressivist or prescriptivist, etc. Viewed orthogonally, both sides fundamentally accepted a Fact-​Value distinction, for both opposed Values to ordinary, ‘natural’ (sic), Facts, the one making them non-​factual, the other factual in a non-​natural realm. Yet much the same is true of those who wished to reduce moral terms to ordinarily objective ‘natural’ ones. These, like Materialists in reducing, or rejecting, Mind in favour just of Body, are still party to the distinction, and in their reduction embrace a scientistic view of reality. The point of these further ‘evaluative’ predicates, their distinctive augmentation of that reduced reality, vanishes (cf. Geach 1956/​1967, 37).9 (The moves of chess, e.g.,

7 

There is a ubiquitous scientising use of ‘theory’ that misrepresents as explanatory what is simply clarificatory (‘shoots through bedrock’). Nietzsche (1887, 3.25), more deeply suspicious, senses in science a hankering for absolutes and purity, more refurbishment of, than antagonism to, the ascetic ideal. 8  Cf. Geach 1956/​1967, 135–​136; Foot 1958/​1978, 100. (For an analogous, peculiarly absolute, ‘detached’ use of ‘ought’ and ‘should’, see Prichard 1912/​1949, 7–​8, discussed later.) 9  On such ‘naturalism’ see Prior (1949, 8–​10) and Ayer (1936/​1942, 103, 104–​105) on reducing moral judgments to a subclass of psychological or sociological judgments.

Criterialism   143 castling, are more than the movements of pieces of wood or ivory on a board.) Many distinctions can be made with such terms (as bodily versus mental health), so with talk of fact and value: just not ‘the’ distinction. Yet the Fact-​Value distinction, far from acknowledged as nonsensical, appears inescapable: for doesn’t it exhaust the possibilities? Evaluations are either factual or not; if factual, they either reduce to natural ones, or they do not; if not, they are then factual about a non-​natural realm. Where else to go? (cf. PI 352). Nonetheless, viewed orthogonally, what seemed central is insignificant, indeed non-​significant: ‘the first step is the one that altogether escapes notice’ (PI 308), as one might begin to suspect by wondering what is meant by ‘natural.’10 So put, this is no addition to that taxonomy, but a rejection of it. A second trap is the Practicality Problem. ‘Surely evaluative judgments have in some sense to be action-​guiding, motivating or reason-​giving, and moral ones authoritatively and universally so.’ But how? What, if any, is the connection with practical reason, with a rational ‘ought’ (against one claimed, e.g., as the expression of quiet desire)? We come across something of the same range of options. Some ‘naturalisms,’ as with Hume, effectually reduce or abolish practical reason, suspecting something abstrusely metaphysical in any such practically rational ‘ought,’ however narrowly circumscribed (e.g., Stevenson 1937). Other options are those of Foot’s Fork: ‘either some consideration is a reason for an agent iff it suitably serves some end they care about; or else, disconnected from any such ground of ‘interests,’ it has an ‘automatic’ reason-​giving force,’ i.e., the categorical thrust of a rationally pure ‘ought.’ The latter horn is Super-​Objective, the agent intuiting a distinctive property, or relation, which either itself is, or else imposes, a pure or unconditional rational obligation (unconditionality temptingly provides a sense for ‘moral’); and—​just as Moore had us intuit the special non-​natural property named by ‘good’11—​ mysteriously mesmeric and prima facie lacking sense.12 The former, End-​relative, horn, 10 

Cf. MMP in CP3, 27 on the naturalistic fallacy. Taking an orthogonal perspective is a kind of triangulating out, denying sense to an assumption common to both sides (here the distinction of fact and value). This often involves countering a Philosophical tendency to generalisation by, e.g., looking instead to intermediate cases (cf. PI 122). If ‘natural’ predicates here are supposed those of psychology and sociology, what range is that? What of ‘natural’ facts outside such sciences, e.g., being a six-​foot-​tall bruiser? And if ‘natural’ thus expands, what of thick predicates, ‘rude’ or ‘cruel’? And if those, what of thinner and yet thinner ones? And of distinguishing ‘thick’ and ‘thin’ anyway? 11  Mackie (1977, 32) suggests Moore’s characterisation of the property, or object, named by ‘good’ as non-​natural ‘leaves room’ for its having an action-​guiding, will-​entangling, force that merely ‘natural’ properties lack. Maybe any ‘oughtness’ of ends is identified with, or intuitively follows, their intrinsic goodness. (Moore [1903, 115–​118] views actions instrumentally, their ‘oughtness’, like their goodness, a matter of their consequences for intrinsic goods.) 12  For example, notably Prichard (1912/​1949). Samuel Clarke (1706/​1969, §§225–​226) introduces ‘fitness’ as the relation which imposes rational obligation on us. God’s will always does determine itself by fitness, viz. chooses ‘to act only what is agreeable to justice, equity, goodness and truth, in order to the welfare of the whole universe’. By contrast, our will ought to—​the obligation registers the potential of human will to fail so to determine itself (as with Kant). The relation imposing the obligation is accessed, as for God’s will, by that translucent quasi-​mathematical intuitional rationality (§226), quite separate both from God’s willing us or commanding us so to act, and from self-​interested, prudential, considerations (§225). Fitness might thus be considered the ruling principle, or norm, of practical reason: to see it is to appreciate its rational force.

144   Gavin Lawrence by contrast, embraces a subjectivity of reason-​hood, a rational ‘ought’ conditional on the satisfying of an agent’s contingent desires, the object of desire creating the norm for rational agency.13 On this, any pure, unconditional, categorical, rational necessity, or authority, is abandoned, and readily so as something ill-​begotten. Universality, however, is still securable, if only now a contingent one, either by positing some universally held non-​moral end (say, prudential self-​interest) to which moral action is by happenstance the best or necessary means, or by positing a contingently universal acceptance of moral ends. (The first is problematic over the tight-​corner problem, over offering reasons of the wrong kind, and over its moral-​prudential distinction; the second seemingly implausible in universally ascribing moral ends, resource-​less in its need to prioritise moral over other ends, and revisionary in the contingency it gives such moral ends—​and generally in its excluding the rational evaluation of ends: any ordinary claim, e.g., that justice is an end an agent ought to have, is senseless and counts as mere ‘bluff ’ or wishfulness, an idea that, so generalised, creaks.)14 Both Super-​Objectivist and Subjectivist accounts of practical reason ally in one thing: the ‘Why?’ question asking for reasons must, if successful, come to an end. But this truism is then falsely taken to have certain implications. This is because both view it through the classic lens of a ‘mediating’ conception of reasoning, on which its function is to supply a medium, or middle term. If so, then for the question ‘Why?’ to come to an end, reason must ultimately drop out on pain of an infinite regress.15 The giving of reasons can be stopped only by reaching a special point, special in that no reason can now be given, no ‘Why?’ asked, something for which there is no reason, no ‘medium’; something ‘im-​mediate.’ This has severe grammatical consequences. When reasoning, so conceived, gives out, the options left are Hume’s triad: on the one side, those still of Reason, or Rationality, but now in a different, wider, quasi-​perceptual sense, viz. the unmediated intuiting either (Option 1) of a property, of final value, goodness 13  In such a vein, one may allow ‘oughts’ also relative to technical, or more broadly, systemic, dimensional ends, and then regard prudence and morality also as such systems (cf. Foot 1972/​1978, nn8 and 15). If so, one can grant a certain objectivity to what morality requires (the dissolution of the first trap), to what one ought morally to do, while yet maintaining a subjectivity of reason over whether one ought to do what one morally ought, this practical ‘ought’ depending on happening to care about the system’s end (Foot 1972/​1978, nn8 and 15). Gripped at the time by a broadly instrumental neo-​Humean conception of practical reason, Foot failed to envisage the possible role of the good as the formal object of practical reason (cf. McDowell 1978; Lawrence 1995). 14  Retreats could concede first, with Hume, that there is no per se priority to sympathy and moral ends; and then further, with Foot (1972/​1978), that moral ends are not plausibly universal either. An alternative is to posit that such ends would be universally held (and primary?) under some condition, e.g., of full information or after psychoanalysis’. But this needs argument and erroneously denies any rational understanding of worth. Regarding bluffs etc., there are such cases, as when my claim that ‘you ought . . .’ is code for ‘it would suit me if you did.” But this generalisation lacks plausibility (see n22). 15  Talk of ‘reason’ here means reasoning; however, Hume will talk of Reason, in the theoretic sphere, as including Intuition of im-​mediate first principles as well as demonstrations from them (the Aristotelian and scholastic conception of a science) and, in the empirical and practical sphere, as covering perception and cause and effect: i.e., generally the cognitive apprehension of im-​mediate empirical contingent truths.

Criterialism   145 or moral goodness, belonging to an object or (Option 2) of a relation, say, of moral fitness or of moral obligation, between two objects (the doing of a favour and an expression of gratitude); on the other, (Option 3), a special reactive feeling, desire, or attitude, unmediatedly responding to the object or the (supposed) total situational array. The alternatives appear a world apart,16 and far from seeming nonsensical, again seem inescapable. Reasoning can work out only what things are good in that they lead to other good objects, an instrumental, or more broadly subservient, ground of value; but there have to be objects of final value, and this value cannot lie in leading to other objects, nor then be calculated by reasoning. Where then?—​either in the objects themselves, a strongly intrinsic value or goodness, whether apparently contingent property in individual objects or relation necessarily binding them, either way needing to be rationally intuited, or else in the subject, a matter of their reaction or attitude to the objects.17 Where else to go? Yet again, viewed orthogonally, the apparently significant difference—​cognitive intuition or non-​cognitive reaction?—​disappears in mistaken grammar. Briefly: (a) Humeanism is guilty of a mistaken assimilation of value to desire (‘good’ doesn’t work like that); and again of making all desires per se reason-​giving, however bad and corrupt, however trivial, however bizarre (pica). The Super-​Objective alternative remains mysterious both in itself and in the way it grabs the will (both aspects critiqued by Hume). (b) Moreover their shared regress, demonstrating that reasoning must drop out, assumes that all reasons, all answers to ‘Why?,’ are further reasons; in this it is a petitio principii, and the claim itself anyway false. (c) Fundamentally, our two traps are guilty of parallel grammatical misunderstandings. As with predicates like ‘good’ and ‘bad,’ so too with the directive, norm-​stating, auxiliary verbs ‘ought’ and ‘should,’ it is part of their very grammar that the question ‘Why?’ asking for reasons is always in place: their fundamental grammar is ‘good to φ: R,’ ‘ought to φ: R,’ where R is a reason. 16  If founded on Reason, we are dealing with truths, either (Option 1) matters of fact, contingent and a posteriori, or (Option 2) relations of ideas, necessary and a priori. If not founded on Reason, then (Option 3) it is founded on sentiment, on the reaction of some special pleasure or pain, approval or disapproval; it is thus assimilated at a general level to matters of taste. Moore, Prichard, and Ayer along with Stevenson, respectively, illustrate modern versions of these. For the last, see Carnap (1963, 81–​82) reporting his own view and those of the Vienna Circle, 1926–​1935. (Ayer visited the Circle in 1933.) 17  (i) The contingency makes its presence in the object alarmingly tenuous—​mightn’t its attachment to the object and its natural properties simply flicker on and off in the same object or be inconsistent across similar objects? It needs sturdier anchoring, say, in ‘supervenience’! Both rationalist options can seem distorted projections of criterialism, trying to do its work, once a semantic move between predicates is misconceived as a metaphysical gap needing to be filled. (ii) Absolute ‘ought-​ness’, moral obligation, or rightness, may be claimed as a special object, whether of intuition or of feeling, that is separate from final value, from the good, now conceived as delimited, e.g., to self-​interest and/​or to beneficence (as against justice). Thus Prichard (1912/​1949) contrasts obligation, the sphere of morality, with intrinsic goodness (5–​7) and links end or purpose with goodness and desire, not obligation: desire and a sense of obligation (or equivalently ‘duty’ [6, 12], a sense or recognition of rightness [6, 7]) are rather ‘co-​ordinate forms or species of motive’ (10). Yet confusingly he talks also of ‘morality and virtue as independent, though related, species of goodness’ (11), that ‘in each case the act possesses an intrinsic goodness’ (12).

146   Gavin Lawrence The Practicality Problem I take dissolved by the traditional conception of practical reason: of course reasons, an account or logos, can be given why certain things are, and others not, good or worthwhile, best or wisest, as ends to pursue. It is what Plato attempts in the Republic, in answering Thrasymachus: as with health, so with justice—​ you make out the rational worthwhileness of a final end by describing, by giving an account, positively, of what life is like with it, and then, negatively, of what life is like if it is withdrawn or absent. As against a desire-​satisfaction system, whose formal end is just that, traditional practical reason is conceived as a good-​delivery system, its formal end—​the end that constitutes practical reasoning or deliberating as the activity that it is—​being to answer and work out the practicable good, i.e., what is best and wisest for the agent to do, which, if rational on this conception, and not impeded (e.g., by depression or self-​doubt), they will do. These ‘dissolutions’ of the two traps work in tandem: the former secures an ordinary factuality about human goods and bads, about what is wise or worth pursuing, what not; the latter an ordinary factuality about what reasons there are for the agent in terms of the first set of facts, these being what the agent needs correctly to register and navigate in order to act and live successfully. Connected though these are as two pieces of a puzzle, our focus here is Anscombe’s dissolution of the first trap, leaving the second for further discussion.18

7.3  The Contrast with Hume 7.3.1  Hume and the Flip-​Flop ‘On Brute Facts’ invites us to look at one strand in Hume’s remark on the ‘is-​ought’ transition through the lens of ‘is-​owes.’ As noted, Hume takes the moral distinction—​the ‘judgement’ of things as good or bad, in a certain restricted, ‘moral’ sense—​to have three options for its ‘foundation.’ In the First Appendix to the Second Enquiry, he offers five arguments against the two options of Reason, i.e., of rational intuition. The second argument initially assumes a parallel between geometrical and moral reasoning,19 and then counters that. It runs:

18  For the traditional conception, see, e.g., Lawrence (2004, 2018). As I understand it, early on Anscombe (BF, MPP), Foot (1958/​1978, 1958–​1959/​1978), and Geach (1956) all appreciated the dissolution of the first trap, but regarding the second, Foot was trapped in a neo-​Humean conception; Geach appreciated that he couldn’t yet see the way out; Anscombe came nearest (in I), but I suspect either failed to appreciate or disagreed with aspects of the traditional conception (e.g., Lawrence 2004). 19  Hume regards relations-​of-​ideas theorists as his main opponents (hence the initial focus on the alleged parallel between geometrical and moral reasoning). Rightly so in his contemporary scene. The idea of morality done more geometrico is common to Spinoza and Locke; Samuel Clarke (1706/​1969) presents a version (see n12). Hume apparently cedes this Aristotelian-​scholastic model of science,

Criterialism   147 1. We make our moral judgements only when we suppose ourselves fully informed of all the relevant relations of ideas and all the matters of fact in the situation, i.e., the total array (e.g., who started attacking whom and why, etc.), the full ‘complication of circumstance’ (Hume 1777/​1902, §237). 2. Discovering these—​relations of ideas and facts—​is the work of Reason, in its two operations. 3. So moral judgement occurs only when we take the work of Reason to be done. 4. So this judgement cannot itself be a work of Reason, and so, as the only option left, must be some non-​rational, or non-​cognitive, reaction to the total array that Reason provided. (Moral judgement is not then analogous to geometrical reasoning, but rather to judgements of beauty, as Hume urges in the complementary third argument.) The following Flip-​Flop schema illustrates this:

A J Array (Moral) Judgment 1.  She cut his leg off she acted badly; did wrong 2.  1, but she was a surgeon and the leg was gangrenous. she acted well 3.  2, but it could have been saved by penicillin. she acted badly 4.  3, but the penicillin was in short supply and needed to save babies. she acted well 5.  4, but she didn’t know it was in short supply. she acted badly

and so on. (We could equally put on the judgement side ‘she did what was unjust/​just,’ or, more strongly, ‘she acted justly/​unjustly’; or again ‘she did what she should not/​ought not’ and ‘she did what she should/​ought.’)20

viz. intuition plus demonstration, for mathematical sciences. He objects to its application to morality, arguing (i) that any candidate ‘immediate’ relation would equally apply to the inanimate, and (ii) that its requisite practicality demands an a priori causal connection to the will, something impossible on his account of causality (1739 and 1740/​2000, 3.1.1.21–​23). We may rather query the model’s commitment to an actus intellectus, a special faculty of nous, a suspiciously convenient ‘solution’ to the model’s problem of cognising immediate first principles. (‘Beam me up, Scotty’) 20  ‘Ought to/​should φ’: I take these locutions to imply a kind of cost to be incurred by not complying. On the traditional conception, the cost is that of not doing what (the agent supposes) is best for them to do, what reason, as they see it, demands of them: and so a failure to φ is a failure to be fully rational, rational tout court, not in some particular dimension—​a failure to act and live successfully tout court. (A necessary condition of so doing is that the agent so values it, though there is more complexity here.)

148   Gavin Lawrence The example, if artificial, suffices. For Hume, the Flip-​Flop schema encapsulates that:

(i) Moral judgements change—​flip-​flop—​from one evaluation to its opposite in response to changes in the different supposed total arrays of relations of ideas and matters of fact. (ii) The A-​J connection between array and judgement is a natural one. It is a matter of how we are set up by nature (or moulded by social training) to react to certain different inputs, something for empirical investigation: of human psychology. Hume provides us with an empirical theory of the general natural properties that elicit these ‘judgements’ in humans. With Martians, or indeed with different human tribes, the connection could be set up quite differently, indeed with completely opposite reactions; it is simply a matter of a ‘blind’ natural causal connection, of stimulus-​response.21 The facts and the reactions (‘the valuations’) are logically independent. (This sort of account invites an evolutionary story that seeks to explain current A-​J pairings in terms of their survival value.) (iii) There is no need, nor proper room, for a ‘so.’ If present, it may mark a causal-​explanatory connection, a pattern of how we, or a particular agent, are set up to react. It is not a genuine rational, ‘logos-​governed,’ inferential transition, representing the array as the justificatory grounds for the judgement. Alternatively, it may be viewed as our attempt to aggrandize our reactions by having them masquerade as indeed ‘reasonable,’ as ‘rationally necessary,’ both to ourselves and others, when really they are just a blind natural setup, causal with no inferential or justificatory force: their rational guise mere illusion, a ‘bluff.’ In this direction those discounting any such rational transition have some obligation to describe the content of such a bluff. 22

21 

Hume takes it as unlikely, even ‘impossible’, that nature could set up entirely individual connections anywhere, and so too between sentiments and eliciting particulars (the claim needs more disambiguation). So we should seek for general eliciting features (e.g., 1739 and 1740/​2000, 2.1.3.5–​7; 2.1.4.1; 2.2.7.3; 3.1.2.6). He offers a tripartite empirical ethical theory (1777/​1902, app. 1, §239). (P1) identify the moral sentiment(s); (P2) gather empirical data about what worldly particulars elicit these sentiments; (P3) theorise over this data about the general shared eliciting properties. He theorises there are four, and only four, such: pleasurable to self, and to others; useful to self, and to others (e.g., §§217–​219; compare Plato, Gorgias 474d–​e). 22  Mackie (1977) pursues both lines. He takes the ‘because’ as causal, arguing that the best explanation of why people approve of monogamy is because they participate in that way of life rather than the other way around (36). He seems either ignorant that the opponent, opting for the latter, isn’t in their ‘because’ offering such an efficient causal explanation, or else takes them to be offering an illusory rationalisation, a ‘bluff ’ (in the manner of Anscombe, Foot, Williams). Such allegations need care given that the content of the illusion, the sense of ‘reason’ at issue in rationalisation, hasn’t been granted intelligibility: insufficient probing here emerges in ambivalence over characterising these ‘bluffs’ as nonsensical or as false.

Criterialism   149

7.3.2  Anscombe and the Flip-​Flop Following Anscombe, the Flip-​Flop Schema reads very differently. These different arrays are brute relative to their corresponding judgements. What the flipping and flopping show is the presence of criteria to which our judgements of acting well and acting badly are differentially sensitive and by which they are ruled or governed. The changes are principled, ‘rule-​governed.’ The ‘so’ indeed marks an inferential rational connection: conversely, the A contains the ‘Why?’ or ‘How so?’ when asked of the J, i.e., provides its logos or account, its justification (cf. BF, points 5 and 6). It is not that the judgement can’t be challenged, but some special story—​further fact—​would be needed to defeat it, or bring it into question (we can read the series of arrays as a series of defeats of the previous claims by the inclusion of a further consideration whose inclusion requires a change in evaluation on pain of unintelligibility or a yet further addition). Each added defeating fact can be viewed as a ‘special story,’ perhaps better ‘a recontextualisation,’ relative to the preceding claim. As with Foot’s (1958/​1978) example of ‘rude,’ the ‘act well/​badly’ is rule-​governed. To play the game of ‘rude’ assessment, you must abide by its rules—​the criteria for its application; it is a public word, and it is not for you to set the rules, to advance or to dismiss as relevant whatever you choose. It is not ‘a private enterprise’ (Foot 1958/​1978, 108). You cannot deem any behaviour you like rude, or discount it as not rude, and still be understood, still be viewed as possessing mastery of that predicate, as competent to play that game: the game sets its rules. And what goes for ‘rude’ goes for ‘good’ and ‘bad.’ You can’t call anything you like a good or bad car, nor a good or bad human, nor a good or bad action, and be understood. (Not a claim grounding in cultural relativism, nor in social practice, but, generally, in successful working, and, for humans, in facts of our nature and life. If some society holds child labour just, then that predicate is an entry point to an ongoing, sometimes difficult, critical conversation, in a common project: holding it just, they engage in a certain human terrain, and their responses must stay principled and relevant, lest their ‘just’ fall out of our understanding. Cf., e.g., Lawrence 2018, 2020.) The relevant criteria are defeasibly sufficient: that is, the J holds if the circumstances are normal. Equally some set of these in a range of considerations is necessary.23 So also with pain. (a) If I am grimacing and holding an injured body part, that is defeasibly sufficient for my being in pain, defeasible in that my ‘grimace’ could be my unusually crooked smile, or a nervous twitch, etc. But absent such defeating ‘special stories,’ I am in pain, given the rules of the ‘pain’ game. Equally, (b), in the absence of special stories, ‘I’m in pain!’ is not intelligible if there I am laughing, dancing on the table tops, 23  No particular condition need be necessary; even if a single thread does run the length of the rope, that need not be so everywhere and would not constitute the essence (cf. AWL 45; cf. PI 67b). The level of specificity of the condition’s description introduces some complexity. It might be artificially concocted (PI 67c). Again, while one may think that, e.g., ‘showing disrespect’ is something common to all things rude, this seems not so much criterion as alternative, perhaps illuminating specification of our concern in the predicate’s game (see also point 6 in BF, 24–​25).

150   Gavin Lawrence denying I even have any inclination or disposition to hold a body part, grimace, cry, whimper, etc. I am not intelligible even to myself. Some ‘pain’ criterion would be contextually necessary; some counter-​criterion, like laughing and dancing, defeasibly sufficient to defeat or counter a pain-​claim. The connection is linguistic or conceptual, not psychological. (This doesn’t rule out strange, challenging, and penumbral cases.) Super-​Objectivists and Subjectivists alike demand that reason assemble a totality of natural facts and relations prior to (moral) evaluation, and construe the latter as a subsequent task for a further psychological faculty to address, be it Reason-​as-​Intuition or reactive Feeling.24 This demand for a complete ‘complication of circumstance,’ a total array, is appealing, in registering sensitivity to (i) the impropriety of sidelining unwelcome counter-​evidence and (ii) the possibility of further, undermining factors. Yet, in this demand, central conceptual problems (cf. section 7.2) emerge again. First, ‘[o]‌ne cannot mention all the things that were not the case, which would have made a difference if they had been’ (BF, point 6). Second, the required totality can hardly be the (nonsensical) infinite totality of all facts and relations, but the totality of relevant ones. But what principle of relevance can reason employ to capture the relevant ‘complication of circumstance’? Mightn’t any extra fact, however curious or incurious, present or past—​Tutankhamun’s having nipped out for a beer behind a pyramid—​change any insight or reaction?25 In advance there is no available content restriction. Is a restriction then merely empirical, forthcoming from our (individual?) experience? or given by yet another mysterious act of Intuition or reactive Feeling? (cf. Mackie’s [1977, 41, 49] worry about ‘links’.)26 Third, notoriously, since the gap is not open to reasoning, any disagreement, of intuition or reaction, is rationally irresolvable. All disputants can do is first check they agree on the relevant totality of the array, and then view it again, in the hope that this time it will elicit the same Intuition or Feeling in both. Failure now will indicate a kind of psychological blindness, or radical divergence, of ‘perception’ or ‘feeling.’27 By contrast, for Anscombe, the principle of relevance is linguistic or grammatical (established by the rules of the game).28 And what matters is not a totality, even of relevant facts, but sufficiency plus normalcy, i.e., the absence of reason to suppose

24  For the Subjectivist demand, following Hume, Ayer (1936/​1946, 110–​112); for the Super-​Objectivist, Prichard (1912/​1949, 7–​9, 16–​17); cf. Moore (1903, viii, x–​xi); 1912, ch. 7: the totality of consequences). 25  Moreover, what justifies their assumption that the relation between fact and evaluation is constant? (cf. n17)? Geometrical relations may be eternal, but why, God’s providence apart, that of moral fitness? Or Moore’s non-​natural property be fitful and inconsistent, wandering over the natural world like a cloud (supervenience added later, an ad hoc measure to secure consistency)? And moral feelings variable, as with what one feels like eating for dinner? Is cruelty on or off tonight? Some more staid than others? 26  There is pressure to construe relevance as causal, not rational—​embodying, as Hume saw, no suitable content constraint nor any a priori connection with the will: the pressure is a product of having so delimited a conception of rational connection that subjective reaction is left as the only alternative. (The potential regress, R1 to F1; R2 to F2, where F2 =​(R1 to F1); R3 to . . .: the least of its problems.) 27  Thus Foot’s (1958/​1978) focus on moral argument. ‘Horizontalists’, in allowing ‘descriptive meaning’ as a necessary layer of meaning along with ‘evaluative meaning’, can evade the relevance difficulty. 28  Cf. Foot (1958/​1978): the public rules set both what is relevant and irrelevant.

Criterialism   151 defeating conditions present.29 What counts as sufficient will vary from context to context. Moreover the notion of a relevant totality in many cases makes no obvious sense. There seems no more sense to a total or complete description of the justice or generosity of an action than of a painting or a piece of music, or of an elbow or a mouse. At a certain point, depending on the purpose, one cries ‘Enough! Enough!’ There is no sense to it even as an ‘unattainable idealisation’ (often an ignis fatuus beckoning to a Philosophical bog). Of course one can give a sense to ‘complete for this purpose,’ or stipulate a finite list of items to be checked off. Second, as Anscombe points out, the brute relative relation iterates: we can have a series of facts with other facts brute relative to them, a series of stacked predicates:30

so so so so so so

He left the potatoes on the doorstep for me I was supplied by him I owe him I am discharging my debt by paying him I am acting honestly I am acting virtuously I am acting well or successfully: living a good life.

At the end we are still dealing with a factual claim in the very same ordinary sense as at the start—​simply a higher level predicate with more complex criteria, a more sophisticated game, ordinarily reflecting more complex interests and concerns.31 There is in both the brute relative relation and in the stacking of predicates a complex dependence on background, general and particular circumstances, and wider contexts. Just as sentences have sense or point only in the context of a language, so too this—​the language and its sentences—​have the sense they do only in the form of life of which they are a part, embedded in living (PI 23b, 25; section 7.1). This is au fond a matter of the facts that lie in the background and render activities, their descriptions, available and pointful—​of use—​in the first place: equally these (e.g., facts about our world, about human nature, the diversity of language-​games and practices, abilities, needs, and social history) shape or contour the activity and its description. The activity of going into a shop to buy a 100 gram slice of organic Cheshire cheese relies on a huge ‘support group’ of multifarious interlocking facts: facts about human intentional action, about the general behaviour of cheese (PI 142), Cheshire’s slice-​ability rather than gouge-​ ability; facts about contemporary and historical conventions—​practices, customs, and institutions: the presence of (cheese) shops at this historical time and place, buying and 29 

Admittedly mistakes can be made here. But it is also a failure of rationality to keep going back to check that the oven is off. Yet if surreptitious criminal oven-​lighting were to become prevalent . . . Looked at the other way, the conversation remains naturally open to further refinement and elaboration. (‘He was rude!’ ‘Yes indeed, but you don’t know the half of it!’) 30  See BF (23, 24, paragraphs 7 and 11); MMP (in CP3, 28–​29, paragraph 11). 31  ‘But of course, if this is a valid point, it holds equally for the description of a set of events as: the grocer’s supplying me with potatoes’ (BF, 23).

152   Gavin Lawrence selling, currency, weighing and measuring, export and import, wholesale and retail; the making of cheese, and of local varieties; of the English division of shires; the idea of organic food and of standards and certification for that . . . Our initial list will (endlessly) proliferate as we consider their dependencies in turn on still others. Without these it wouldn’t be available: try supposing them different and the shape will change, and may eventually lose its point and become unavailable (PI 142, 23a, Pt. II.xii). Anscombe clearly articulates these points: that of the life-​background and its practices without which the description has no use—​no life nor meaning; that the existence of this is not itself asserted in the description (cf. PI 179b; RFM VI.11.3); that the assertion does not claim that the contextual circumstances are normal, i.e., that no defeating conditions are present—​the assertion is simply made in the background context of no obvious reason to suppose anything untoward. (Of course it can be made also in some qualified form.) Anscombe refers to the notion of a background ‘institution’ in BF, such as currency (cf. PI 584a, ‘die Institution des Geldes’), but clearly understands ‘institution’ to cover a wide range of background practices and ‘normal procedures’ (BF, point 2; in PI 199b, ‘Gepflogenheiten (Gebrauche, Institutionen),’ the bracketed notions seem glosses rather than different instances; cf. PI 337, 380), ones including speaking a language (PI 540). Which background factors are more salient obviously differs, depending on which are more or less immediate parts of the sense-​informing, sense-​sustaining frame in the context. These create areas of stability and normalcy, and themselves depend on and interlock with further frames; they also sustain the sense of abnormal and exception. If they change, ‘if exception became rule and rule the exception’ (PI 142, 227e; OC 63), then our practices change or fade away: if counterfeit currency is known or believed to outnumber the genuine, or if pretense becomes the norm (an epidemic of actors) . . . Or again, if their own sustaining background changes, if coins start to melt like chocolates in the pocket . . . Sustaining all are ‘the very general facts of nature,’ which likewise, if imagined different, certain concepts would no longer be available and would lose their point (PI Pt. II xii); ‘the more abnormal the case, the more doubtful it becomes what we are to say’ (PI 142). ‘Doubtful’ about what we can now intelligibly say, as our games lose their grip and traction. One becomes dumbfounded. (In On Certainty Wittgenstein pursues a particular interest in the most general, global facts of nature, the existence of the external world, its long past, induction, that humans have heads and parents . . . backgroundings that are, at the time, like the bed of the stream as against its flow, the road as against its traffic: our best world picture [Weldbilt], scaffolding, inherited background, framework, substratum, anchor: the things which stand fast [feststehen] and within which we proceed or go on [handeln], accepting them without question of, or sense to, their justification, and which, as a kind of global limit, frame the intelligibility of the things which we can doubt or find mistaken. Unless there is a very special context, not easy to imagine, there is no sense to a doubt, ‘Do I have a head?,’ nor to any reassurance that could be looked for.)32 32  ‘Bit by bit there forms a system of what is believed, and in that system some things stand unshakeably fast and some are more or less liable to shift. What stands fast does so, not because it is intrinsically obvious or convincing; it is rather held fast by what lies around it’ (OC 144; cf. n35 later).

Criterialism   153 This position I call criterialism. Its central struts are: C1, the relation of being brute relative to C2, its iteration, the stacking of predicates consequent on that C3, the contextual backgrounding that underpins both the bruteness and the stacking. We could say the particular situations, or contexts, shape and control the criteria. The approach reveals that evaluation of all sorts is simply an ordinary part of the realm of ordinary facts, ordinary objectivity, ordinary truth. Conversely, just as all predicates can be viewed as factual, so all can be viewed as ‘evaluations,’ as measures, their criteria as it were ‘calibrations’ of their application, criteria of their success (whole or partial) at being a this or a that, at being F or G: all descriptive evaluations, from ‘table’ to ‘just.’ It is simply that some predicates tend to gather more complexity around them, given their role in our lives: ‘acting honestly’ more than ‘leaving groceries on a doorstep.’ It is a grammatical misunderstanding to view ‘evaluations’ either as truths in some shadowy and mysterious Super-​Objective realm or as Subjective reactions. In sum, no sense has been given to supposing Facts and Values logically independent: the attempt founders from the start in grammatical misunderstanding.33

7.3.3  Summation I equate Anscombe’s ‘brute relative to’ with ‘criterial’ and take her choice as an avoidance of jargon that we might presume already to understand. She characterises the brute relative relation, or transition, between descriptions in six points (BF, 24–​5, A1–​A6). We touched on these earlier. (Roughly, our C3 collects A2, A3, and A5; C2 is clearly advocated but not taken itself to illuminate bruteness; C1 relies on characterising bruteness in terms of its defeasible sufficiency, A4, and non-​delimitable range of necessary descriptions, A1, and its relation to entailment, A6. We return to A1 and A4 later.) Anscombe’s position is explicitly modest. The A1–​A6 relations are neither claimed to be all nor always to hold (BF, 24). Her focus is on bruteness providing some illumination, not on the limits of its illumination. Thus it leaves aside also what counts as description, just as we leave aside, e.g., whether it is helpful to talk of criteria in connection with, say, ‘perhaps’. ‘Limit’ here is something which, as the Big Typescript (BT 312) remarks, philosophers are inclined to think you can see beyond, limit as fence: the ‘limits of human understanding’ as though there were a larger understanding lying beyond. (More needs saying about its interpretation by, e.g., Strawson [1985, 14–​21] and Searle [2011].) There are fluidities between what is at one moment taken as grammar, at another a move within it. A different, phenomenological concern with background, with ‘the natural facts of (social) life’, one appearing more empirical, hypothetical, than grammatical, runs from Husserl via Alfred Schütz (in the Vienna Circle of von Mises), to the ethnomethodology of Harold Garfinkel (e.g., 1964/​1967). 33  Two lines of thought attempt to accommodate this: ‘horizontalists’ distinguish two layers of meaning, ‘descriptive’ and ‘evaluative’, while ‘verticalists’ distinguish thick and thin evaluative predicates (Lawrence 2018, n40).

154   Gavin Lawrence In sum, in BF Anscombe uses ‘is/​owes’ to introduce and illuminate an interesting kind of transition, neither contingent nor logical entailment (cf. Geach 1956/​1967, 38–​ 39): brute relative to. It shows how to block the temptation to suppose ‘the’ fact-​value distinction intelligible and lets the stacking of predicates go all the way up to ‘bad person’ and ‘bad act.’ Anscombe in showing how there is no ‘gap’ in the ‘is-​owes’ case, thereby articulates a rationale implicitly to apply to ‘is-​ought,’ at least taken as the ordinary ‘ought’ of practical rationality. However, just as some tried to make ground by distinguishing thick and thin predicates (cruel versus good), so some may see a gap between evaluative predicates (‘just,’ ‘cruel,’ ‘good’ . . .) and ‘ought,’ which seems of a different ilk, and then find that gap in Anscombe’s argumentation. The Traditional conception I believe solves this (whether its judgements about what it is best or wisest to do are understood as brute relative to, or to entail, what an agent rationally ought to do; cf. A6). In MMP this turns out but one of three richly challenging points Anscombe develops from considering Hume’s problem-​setting remarks on the ‘gap’ transition from ‘is’ to ‘ought’ (in CP3, 31, 32). A second enquires into the transition from ‘is’ to ‘needs,’ with its relation to a thing’s good and bad, and Aristotelian necessity (what something must have in order to flourish) (31, 41). Grant these two, there is then a third: the objector says, ‘Perhaps you have made your point about a transition from ‘is’ to ‘owes’ and from ‘is’ to ‘needs’: but only at the cost of showing ‘owes’ and ‘needs’ sentences express a kind of truths, a kind of facts. And it remains impossible to infer ‘morally ought’ from ‘is’’ (32). Here Anscombe agrees, but on grounds that ‘morally ought’ ‘no longer signifies a real concept at all’: divorced from its backgrounding in a divine command conception of ethics, it now has no content. Philosophers’ mistake then is to ‘try and find an alternative (very fishy) content and to retain the psychological force of the term’ (32). One may worry about this diagnosis; however, our focus is simply the first.

7.4  Further Aspects of Criterialism So much for the basics. The position and its Wittgensteinianism invite further filling out. C4: Defeasibility: the ‘incomplete’ specifiability of defeating conditions. Where criteria are defeasible, there is no possibility of a complete list of the defeating conditions: what are ‘normal circumstances’ can be indicated only by some (paradigmatic) examples of defeating conditions, i.e., with an open ‘and so on’ (PI 208e–​f ), and the extent of their range indicated by offering ‘diverse’ examples (MMP, in CP3, paragraph 11; cf. OC 26–​27). (Anscombe offers a quick argument for this: defeasibility, or exceptionality, iterates; the special stories can themselves have special stories that defeat their defeating-​ness, and so on.)34

34 

BF, paragraph 4; MMP (in CP3, 28–​29, paragraph 11). Cf. PI 84a; RPP (1980, I §137 init.).

Criterialism   155 Moreover, these defeating conditions, or ‘special stories,’ ‘do not come into consideration without a reason’; that is, they mark real, not merely logically possible, occasions for doubt over whether the context is normal or needs recontextualising (BF, paragraphs 4 and 10; cf. PI 84, 87). (There are criteria here too. If I suggest something you accept is, or may be, after all a hoax, there are criteria for being, and also for suspecting, a hoax.) Such reasons thus enable one to ‘say how the situation is deviant from the usual ones’ (Z, 118). (What mode, or modes, can defeat take? Clearly normal criteria sometimes turn out not present after all: they were ‘as if ’ present, as in play or non-​culpable ignorance [the agent didn’t know that tu or du presumed on intimacy]. But does defeasibility always involve such ‘de-​clutching’ [a disjunctive view]? Or can they be present as criteria but outweighed by other criteria? Or does that render them merely prima facie or pro tanto criteria or reasons—​what would have functioned as criteria in normal contexts? Matters for another time.) C5: The ‘incomplete’ specifiability of criteria themselves, XYZ, etc. Equally, at least for many, perhaps all, concepts or descriptions, there is no possibility of, nor sense to a complete list of their criteria, or sets of criteria (cf. Z, 118). These too can be indicated by paradigmatic examples, core judgements, and again ‘diverse’ ones help to indicate the range (BF, A1, 24).35 (Again the criteria will themselves in turn have criteria, whose specifiability is, in turn, ‘incomplete.’) So, first, the ‘incompleteness’ of both C4 and C5 is of the kind ‘and so on ad infinitum’ (PI 208f). It is not a second best, in the absence of something more adequate, viz. a ‘complete’ definition (PI 71b, 75). ‘Completeness’ here makes no sense, unless we give it one—​and any such ‘completeness’ could not be the ‘ideal,’ absolute one, only one tightened relative to a particular purpose (cf. PI 47; ‘there is no such thing as a completed grammar’: 1932–​1933, AWL, 21). So I scare-​quote ‘incomplete’ because, like ‘complete,’ taken absolutely, without a specifying context, there is no really intelligible contrast here, any more than between ‘simple’ and ‘complex’ (PI 47) or ‘precise’ and ‘vague,’ ‘exact’ and ‘rough,’ etc. Second, the role of paradigmatic examples in both C4 and C5 is connected with the remark ‘If language is to be a means of communication there must be agreement not only in definitions but also (queer as this may sound) in judgments’ (PI 242), itself connected with meaning as use, and with our learning uses, training in judgement (cf. PI 226e; OC 139). Agreement in core judgements is also defeasibly sufficient to indicate mastery. (To play chess we must agree about what counts as a move, and as, e.g., a knight’s move, etc.; one might master this without an explicitly formulated rule, one’s moves showing one’s mastery (PI 31b), the understanding of such a rule requiring agreement about correct instances.) 35  As in teaching ‘blue’ one would point to, ostend, a range (‘that too is blue’), to avoid being misunderstood over-​specifically (cf. PI 28). The range is also kept in place and determined by what surrounds it, the other colours, and where one will qualify and judge ‘it is greenish blue’ or ‘purplish blue’, etc.

156   Gavin Lawrence Third, both points also allow for grey, or penumbral cases (Grenzfalle). There are not absolutely sharp edges, but uses trail off from core cases into ones where we explore qualifications (‘greenish blue’; ‘slightly irritable’) to penumbral ones where we become increasingly uncertain what to say.36 (We return to this.) In sum, C4 and C5 combine to constitute an ‘incompleteness’ of specification, an apparent ‘leakage,’ in both directions, internal, C4, and external, C5: one can completely specify neither the ‘abnormal’ conditions where any (set of) criteria fail to operate, nor the criteria themselves. (The point goes beyond C1, which is compatible with a complex determinacy, of a fixed number of defeating, and of sufficiency, conditions.) This is unsurprising: to think otherwise reveals the grip of an illusory picture of the workings of language (and life). C6: Teaching such predicates. This ‘incompleteness’ is manifest in the teaching of descriptions, predicates, or concepts. ‘But if a person has not yet got the concepts, I shall teach him to use the words by means of examples and by practice—​And when I do this I do not communicate less to him than I know myself ’ (PI 208a; cf. 209b, 210–​213, 71b). ‘I do it, he does it after me; and I influence him by expressions of agreement, rejection, expectation, encouragement. I let him go his way, or hold him back; and so on’ (PI 208c; cf. b–​d, 145b). (a) This kind of training reaches beyond the examples given: the ‘and so on’ is not an abbreviation for a determinate list, but one ad infinitum (PI 208e–​g). The whole notion of giving, and of understanding, something as an example or an instance is to understand it as a rule, projecting a rule . . . and this is tied to, or underpinned by, such basic natural facts of human life as imitation, and response and sensitivity to pattern and same again (as in the hand-​stacking game described by Anscombe in RRP, 101). (b) Moreover it is no more than I know myself (cf. PI 69), and all this—​my examples, the practice, the corrections, etc.—​is perfectly adequate: no second best, in lieu of that perfectly complete definition. The temptation is to suppose there must be such a definition that, like a ‘rigid mechanism,’ already and perfectly underlies and determines the predicate’s application. In one direction we suppose this is hidden, a rigid rule that as teacher and competent user I must have, but for some reason cannot articulate—​for doesn’t it run away when I try? (PI 89c)—​leaving me able only, as make do, to give you hints, orient you, in the hope you will guess at and find the same hidden rule in yourself. That is just another ‘beetle’ (cf. PI 293b, 304). Alternatively I view the hidden essence comprising my mastery of the predicate as a yet unformulated definition, which, were another to formulate it, I would

36 

(a) ‘[A]‌nd while it may lead you to stretch a point on the circumference, it will not permit you to destroy the centre’ (MMP, in CP3, 36). (b) Under certain drugs, patients may still avow feeling pain but, somehow disassociated from it, not evince typical pain criteria. We may struggle to understand this, as may the patient. Much depends on what they find natural to say, and on what, if any, consequences are at stake. Moving into the penumbra we start not knowing what to say (PI 142, 80). (c) Some cases overlap, clearly sharing some criteria but resisting others; while happy to point out similarities and differences (they are distant cousins in family resemblance), we reject a ‘yes or no?’, excluded middle question, unless it is important to decide it for a particular purpose, and even then may offer a qualification.

Criterialism   157 recognise as the very thing I had been unable to (PI 75)! But that couldn’t do the job: for any proffered formal definition is equally mis-​interpretable (PI 71b, 34): it doesn’t contain its applications wrapped up in itself;37 perhaps we feel it ‘forces’ or ‘necessitates’ a particular application on us, which ‘force’ we then falsely misplace and picture as coming from inside it, our (‘inarticulate yet real’) grasp or understanding of it (PI 146; cf. 140a, 152), rather than simply from our training in the practice. (A resonance with Hume.) In short, it can’t do ‘the job,’ for the job isn’t itself intelligible. Feeling a space, a need, we posit something, a beetle, for it. It can’t fill it, because the space, the need, wasn’t really there to begin with. That was our misapprehension. (Ostensive definitions can work, but require a place already prepared (PI 28–​31); presumably the same is true of formal ones, as of a bishop in chess defined by the rules of its possible moves.) C7: The variety of criteria C5 is unsurprising also because the criteria, and counter-​ criteria, themselves are of multiple intersecting kinds, language being part and parcel of human life. Language both adds further aspects to that (cf. PI 25, 23–​24, 27a)—​with their own complexity of grammar, e.g., of the tenses and moods, of the difference between predicative and attributive adjectives, etc.—​and is itself deeply interwoven into the larger behavioural weave of life (cf. section 7.1): and this not only general but individual (the particular way someone shows they are amused or annoyed, or are rude). Thus tone of voice, body language, gesture, and ostension (PI 208d, 43b), emotional reaction (surprise, expectation, etc.); samples, photographs (PI 79b); surrounding circumstance—​what went on before and after and at the time; are just as much criteria as, for example, what questions are, and what are not, in place to ask. (Consider for instance what it means that someone is speaking ironically, or in love). ‘Criterion’ as used by Wittgenstein here is itself a word of natural language, and as such a family resemblance term, like ‘derive.’ It could be a ‘technical term,’ tightened for a particular purpose, but is not so here.38 Given all this, ‘life-​game’ might be as appropriate as ‘language-​game.’ C8: Family resemblances: Local variation; Change; Tension; Networks The kind of ‘imprecision,’ ‘inexactness,’ and ‘incompleteness’ just noted allows for a number of things we find, and would expect to find, present in natural language as a tool for human use (PI 569). Broadly speaking we can regard these as forms of family resemblance.39 (i) Local variation in the criteria across communities—​perhaps influenced by local circumstances and needs (an analogue to biological sub-​species). (ii) Change over time. As circumstances, needs, and outlooks change, we find changes in the criteria or in the prominence of some over others, as old concepts are pressed into new service, metaphors become literal, etc. (iii) Tensions and divisions within the use. A variety of uses may develop within a single community. This may be indicated by different

37  ‘We have the wrong idea that the use of a word is like pulling a thread from a bobbin: it is all there and needs only to be unwound’ (Lecture V, 1934, in AWL, 83). 38  Cf. Glock (1996, 93). Wittgenstein’s earlier uses involve contrasts with hypothesis and with symptom. (Allowing that very contingent things can function as criteria (cf. PI 79b, 62) may lie in part behind the wider use, and the relative disinterest here in contrasting criterion and symptom, pace PI 354.) 39  Some connection of criteria with family resemblance is noted by others: e.g., Lycan (1971, 121).

158   Gavin Lawrence metaphorical extensions being found natural. The concept may simply broaden and become more accommodating, perhaps with a range of different nuances in different contexts; or have two related uses that can on the whole co-​habit; or become ambiguous; or split into two uses, the predicate seeming now to its users homonymous. (iv) Networks, elaborations, and refinement. Human interest may also drive further elaboration, a proliferation of distinctions, with degrees and modifications, there being as much interest in the refining of distinctions as in their application—​in developing more sophisticated ‘games,’ networks which can register and capture more nuanced information, and which in their networking are tools of exploration: our interest more in the using them, open-​endedly, to explore the subtleties of character than in their typography, more Henry James than Theophrastus.40 Take ‘rude’ as an example. At its root lies the sense of ‘rough, raw, unworked’ (e.g., of broken, as yet unshaped stone). This develops over time in different directions. Thus at one, ‘gross-​grained’ level we find, e.g., Roget’s rough division into the subheadings of ‘violent, shapeless, ignorant, inelegant, ugly, vulgar, uncivilized, uncivil, disrespect, health’—​and one can see the varying connections of each with the Ur-​root, and also some crisscrossing. Then take one of these, say, the disrespectful: we can uncover a further ‘finer-​grained’ network, different sub-​groupings of criteria, which can sustain their own, more aspectual, often crisscrossing predicates. Thus: (a) impertinent, pert, malapert, impudent, forward; (b) uncivil, uncultured, uncouth; discourteous, impolite, ill-​mannered; coarse, ill-​bred; unmannerly; indecorous; ungracious; ungentlemanly; (c) insulting, offensive, insolent; giving affront; (d) slighting; derisive; derogatory, disparaging; catty; (e) tactless; insensitive; undiplomatic; (f) churlish; boorish; brusque; blunt; crass; oafish; (g) cavalier; dismissive; off-​hand; disregarding; supercilious; disdainful; and so on. Some may be more on the periphery of rudeness and disrespect (cf. ‘tactless’; ‘coarse’; ‘insufferable’), or count as such only in certain contexts (‘blunt to the point of rudeness,’ ‘too cheeky by half,’ ‘out of line’; sardonic, ironic, sarcastic; arrogant; 40 

Gallie (1956) examines concepts where (i) various parties have an as yet dominating interest in, or are under pressure to preserve, a common concept or standard use of predicate, (ii) but are competing to secure acceptance of a certain set as its central (dominating) criteria. There are a number of unclarities. (A) If ‘essentially contested’ is understood as a contest over essence, then this looks like a Platonist error by the parties concerned, to which the ‘artichoke’ is a response: there is no ‘real’ answer (as there is to what water is, but not derivation). (B) Perhaps it means certain predicates naturally lend themselves to dispute over their central criteria. The various candidates are in effect family resemblances, and the dispute looks to be a dispute involving an interest in claiming one as the real ‘family face’, a family argument. If it is not mistake (A), then this looks like tightening for a specific purpose: in the immediate, for control over the predicate. Why that might matter is presumably various (something Nietzsche would see as a power play over language control, synchronically and diachronically). There may be enough shared content, agreement about rough criteria, in the predicate to make control of it seem ‘worth fighting over’—​the accolade of being ‘champions’, although such rough agreement may also be shared only partially and overlapping-​ly between parties, making assimilation to a concept/​conception distinction non-​viable. Alternatively there may be nothing more shared than the ‘name’ and a certain history, but it is regarded, like a flag, sentimentally or as having symbolic value. Which predicates are more liable to this at any juncture is presumably contextual. Yet at least of equal interest as contestation is the rational exploration and refinement of criteria.

Criterialism   159 excessively polite or punctilious), while others are more paradigmatic. And again we may take one of these and uncover a still ‘finer-​grained’ network of discriminations already marked by predicates, or spaces, niches, into which they may be inserted should needs and pressures lead that way, spaces already explorable by qualifications of many kinds. In short, we find a whole terrain or network, a shifting delta-​structure of aspects that differentiate and also crisscross, both ‘internally’ and with neighbouring networks.41 Sometimes our interest in these may be as supports in a claim that someone, or some behaviour, is rude; at others in exploring the form the rudeness takes and what more fine-​grainedly that reveals.42 Why bother even to think there must be a common hidden essence, although there may be an organising interest or concern (or set thereof)? The demand, the ‘must’ is a product of Philosophical illusion, for an illusory purpose.

7.5  Criterialism and Philosophy: Alleged Defects of Natural Language: ‘Imprecision,’ ‘Inexactness,’ ‘Incompleteness’ Since Wittgenstein the notion of criterion, and his use of it, have generated a rich and complex literature, too rich to pursue here. His own use has been interpreted and criticised in various ways, notably on the very ground he himself was principally concerned to address: the ‘leakage.’

7.5.1 The Objections of Frege and Plato Doesn’t the very open-​endedness of Criterialism’s C5 and C4, its incompleteness and defeasibility, make our natural talk inoperable? If the meaning of predicates is not ‘fixed,’ not securely bounded on all sides, then surely speakers don’t know what they are talking

41  We can crudely distinguish ‘internal’ and ‘external’ networking (itself an iterable distinction). Thus ‘rude’ networks also with other predicates of character and behavioural assessment, etc. It is an old city, ‘rudeness’ a roughly drawn area which borders on others—​has roads into them—​and internally has its own grey edges and more central points, both shifting over time; thoroughfares, and side streets; even smaller streets and alleyways; houses with histories and extensions. The ‘city’ is not something ‘complete’, neither externally (PI 18, versus TLP 4.001), nor in its internal divisions and complexity (PI 23a): the street map grows in both ways. We can tighten it, make more precise divisions for, say, voting, tax, or transport reasons, or indeed divide it differently for these different purposes (PI 17). 42  Only in thrall to a model of scientific use and ideal precision would one be tempted to compare natural language to “a crude, primitive pocketknife” (Carnap 1963, 938): conscious of its plasticity of use, but seeing in that only bluntness, inexactness, and not its infinite delicacy of refinement and probing.

160   Gavin Lawrence about! don’t know whether it is F or not F: the world slips through our language; our language loses traction with the world. Frege compares a concept to an area and says that an area with vague boundaries cannot be called an area at all. This presumably means that we cannot do anything with it. (PI 71b)43 But if the concept ‘game’ is un-​circumscribed like that, you don’t really know what you mean by a ‘game.’ (PI 70a) [A]‌nd are we to say that we do not really attach any meaning to this word [sc. ‘chair’], because we are not equipped with rules for every possible application of it? (PI 80) But then the use of the word is unregulated, the ‘game’ we play with it is unregulated. (PI 68b; cf. 100)

And so no real game at all. So the objector. This ‘no fixed boundary’ objection has a more directly Platonic expression. (1) Asked why this is a bank, the answer is ‘because it is the side of river,’ and why that, ‘because it is a financial institution’—​the difference in explanans apparently registering a difference in explanandum. (2) Asked why this, that, or the other is beautiful, we get a range of ordinary replies: ‘because of its colour,’ ‘because it is elevating,’ ‘because of the bone structure’ . . . Yet in this case we take ourselves to be talking of a single thing, the beauty common to these various things, and not something different in each case. But with its different explanations, the case looks exactly like (1)! (3) The Same Explanatory Grounding Principle demands that if the many beautifuls are to be a genuine class, a genuine cut across reality, there must be a single explanation in each case that grounds the class, is the principle of its unity, and determines in advance what is, and is not, beautiful. So there is a special, distinctly Philosophical or Scientific task: to uncover and define that principle, the hidden essence of Beauty itself, the single correct explanation of the beauty that all these things have in common, if we are not in fact to be talking of different things in each case, or else talking in dream-​speak of a fugitive beauty ever escaping the clutch of our apprehension, a vague recollection: we seem to see it only for it to disappear as we try to explain it (cf. PI 89c). Really to grasp, to understand, what we are saying we need that distinctly Philosophical understanding. Whether we talk of objects or concepts, we have the same demand for that X which all xs have in common (PI 66, 71b, 72–​73); for that apparently hidden essence (PI 71b, 75, 43  Frege (1903/​1970, vol. ii, §56, p. 159) writes, ‘The definition of a concept (of a possible predicate) must be complete; it must unambiguously determine, as regards any object, whether or not it falls under the concept (whether or not the predicate is truly assertible of it). . . . To a concept without sharp boundary there would correspond an area that had not a sharp boundary-​line all round, but in places just vaguely faded away into the background. This would not really be an area at all; and likewise a concept that is not sharply defined is wrongly termed a concept at all’. (1903/​1970), Vol. ii, §56, 159. Cf. also TLP 4.112, 116. Frege ties the point to the law of excluded middle: ‘The law of the excluded middle is really just another form of the requirement that the concept should have a sharp boundary’. For more on this aspect, see section 7.6.

Criterialism   161 91, 102), that principle, that already contains in advance (PI 80, 190–​191) the determination for every particular case whether it is, or is not, F;—​be it Forms or rules as rails (PI 218–​219a; cf. 228), allowing no loopholes (PI 80) and no grey cases: the idea of Beauty in its exact perfection (cf. ‘clear and distinct ideas,’ self-​interpreting, translucent, as evidently carrying their application, their standard, in themselves). In analysis this is typically conceived of as a set of necessary and sufficient conditions. Criterialism’s double ‘leakage’ apparently undercuts both: no particular condition seems necessary, and any normally sufficient are defeasible (C4 and C5).

7.5.2 The Double Whammy In response, Wittgenstein, as often, deploys a double strategy. (A) What appear as defects in natural language, as its falling short of an ideal language’s absolute standards, are in fact no defects at all: everything ‘is in order as it is’ (PI 98; BB 27).44 Our ordinary predicates indeed aren’t absolutely, or completely, bounded, yet they still work and have traction. They are as exact as they need to be—​or as inexact, if we knew what ideal exactness was; certainly, if need be, we can tighten them for a particular purpose, drawing a more definite boundary where there is not one ordinarily and was none before. But this is not an absolute one (PI 69, 76). They have their senses in many particular language-​games, senses that typically don’t have some single thing in common, no hidden essence, but rather a family of related senses.45 It is not just ‘game’ and a few others that are family resemblance concepts, but most, if not all, natural language concepts (nb. ‘derive’). If you except scientific ones, then you are likely not thinking of spanakopita triangles, nor of the fluctuations of scientific definitions over time (79e).46 The Platonist ideal is likely part of the illusion everywhere. But set this question aside. 44 

What TLP 5.5563 (already referred to in PI 97), claims of the perfectly exact, ‘logical’, structure behind everyday language, the Investigations claims of that ‘inexact’ ordinary language, the recontextualisation of the remark thus a ‘continuity with a difference’: virtually the inverse. No wonder the temptation to publish the works side by side. The remark here does not imply ordinary reform of language is impossible, ‘to prevent misunderstandings in practice’ (PI 132). Thus we might choose to differentiate homophones in their spelling, not allowing them to be homographs (‘intention’ and ‘intension’) or use capitonyms to differentiate polish/​Polish, smith/​Smith, march/​March, etc. 45  Cf. Hume’s (1739 and 1740/​2000, 3.1.2.4) ‘distant resemblance’ of pleasures, albeit connected to abstract ideas. (For ‘exact’: PI 69, 71, 88, 91; complete/​incomplete: PI 86c, 87, 75; ideal: PI 81, 98, 100–​ 101, 103, 105; hidden essence: PI 90–​92, 97b, 102; rules for every possible application: PI 80; what is common: PI 65, 66, 72.) 46  Family resemblance concepts are the norm rather than the exception, pace, e.g., Prior (1949, 11). Asked to remove the tables from the house, do I include the coffee table, the sideboard? a wooden surface rigidly suspended from the ceiling or cantilevered from the wall (am I to saw it off?)? the one in the doll’s house, a life-​size one built of Lego? the Robert Therrien? You may think the use of ‘triangle’ above metaphorical, but why suppose the concept geometrically centred rather than that as tightened for a geometrical purpose? Admittedly there are scientifically centred concepts, but even these may fluctuate, creating a family resemblance within that range of use.

162   Gavin Lawrence (B) Yet not merely is the ideal absolute standard, that measure which threatened to ‘reveal’ the inadequacy of natural talk, not such a threat: it itself makes no sense—​’chimaeras of false idealization’ (LW II, 48). There is no sense to that standard of exactness, of complete boundedness; of an ideal or perfect language as one whose concepts already determine for any and all possible cases whether or not they apply, the ‘Language of the Excluded Middle.’ Rather such an ideal is situated outside all particular language games, it is, as it were, outside life: logic sublimed: no use, no traction, and no sense.47 The very idea of a game ‘everywhere bounded by rules,’ that in containing answers to every possible doubt makes it immune to all doubt, is fugitive (PI 84, 87).48 It is mistaken to think it attainable through further rules that remove doubts in applying the first rules; that invokes an infinite regress, and if, to stop that, you are trained adequately to use the later rules otherwise than by appealing to still further rules, then why not with the first! (PI 85–​86, 201b). And such is not a training in an—​impossible—​immunity. In sum, no single definition has been given of ideal exactness (PI 69, 88e; cf. 47), no definition could be given, none is needed, and no candidate ‘complete definition’ would be invulnerable to misinterpretation (PI 71b). The picture that a word lacks meaning if we ‘are not equipped with rules for every possible application of it’ (PI 80), i.e., securing the elimination of all possible doubt in its applicability, with rules as rails already bound into the predicate or concept, tempts us, driven in part by the Law of Excluded Middle, but pressed in detail makes no sense (cf. PI 352). We are seduced by an ideal space that we think is, must be the deep structure, the greatest generality, of reality and life, a logic sublimed, a hidden crystalline structure that in determining the possibilities of the say-​able, determines the possibilities of the be-​able (PI 97, 108a; cf. section 7.1). But that, while seeming utterly indispensable to life, its very core, is outside life, an illusion (PI 101–​103, 110–​115). The points echo the earlier critique of ‘simple’ and ‘complex’ in connection with names and logical analysis (PI 47).49 These, taken in a ‘philosophical,’ seemingly necessary sense, lack any. In sum, we cannot give rules so ideally strict: we don’t have them, not even lurking as unformulated ones awaiting release by someone more articulate (PI 75). We can admittedly articulate a boundary, give a stricter rule, directed at a specific purpose (PI 68a), but, that tightening aside—​and it is not a complete, not an absolute one (not that we understand what that would be)—​our knowledge, our mastery of a predicate consists in our grasp of, our training in an open-​ended list of criteria: Isn’t my knowledge, my concept of a game, completely expressed in the explanations I could give? That is, in my describing examples of various kinds of game; shewing 47  The idea of language as all names, mastery of whose reference is pre-​required before using them (PI 26–​27), or of ‘naming’, as the infallible ‘associating’ of word with thing (PI 38, 39, 116) is one subliming of logic, as is also that of the single essence, the hidden general form, of the proposition, to which all must conform if they are to say anything. 48  Cf. PI 68b: ‘[B]‌ut no more are there any rules for how high one throws the ball in tennis, or how hard; yet tennis is a game for all that and has rules too’. 49  ‘It makes no sense at all to speak absolutely of the “simple parts of a chair” ’ (PI 47a). Analogous points can be made about Aristotle’s metaphysical project of the absolutely Perfect Substance as the measure of success in substance: e.g., Lawrence (2005).

Criterialism   163 how all sorts of other games can be constructed on the analogy of these; saying that I should scarcely include this or this among games; and so on. (PI 75)

For mastery, there is no more to know than this sort of thing (the artichoke leaves), yet it is enough: nothing is missing.

7.5.3 Defeasibility and Qualification Discussion of criteria often focuses on the feature of defeasibility. But situations and circumstances can fail to be normal in two ways: ones where their abnormality defeats, maybe by declutching, the usual criteria, and ones where it calls rather for qualification. Here too the types of qualification are ‘incomplete,’ although core examples can be given to indicate the range.50 Its ‘incompleteness’ also stretches (‘leaks’) in both directions, external and internal: both in varying modes of qualification (degree, approximation, manner, intensity, aspect . . . types of adverbs) and in internal progression of qualification (‘he was rude’; ‘well slightly’; ‘more than slightly’; ‘yet he has a heart of gold’; ‘perhaps it touched a sore point’; etc.) The qualification can develop and be a (forking) path of exploration, of further contrasts and nuances, and, in its complexification, develop into a virtual mise-​en-​scène. We are here in the central terrain of Aristotle’s ubiquitous deployment of the haplōs-​ pōs distinction: of the truth conditions of something said or viewed unqualifiedly, and of where failure to qualify would mislead into falsity.51 (‘Was he rude?’ ‘Well not exactly, just a bit cheeky.’) This is often heavily contextual on the type of situation and interest at issue. In natural language, vagueness can appear as itself a qualification (‘his behavior was vaguely insulting’; ‘vaguely bald’; ‘borderline blue’; ‘tall-​ish’)52 and may be introduced for a variety of reasons. It can equally on occasion be what is problematic and prompt a call for further specification, or precisification, to satisfy an interest: ‘I want the blue.’ ‘Which?’ ‘Any will do’ versus ‘The cerulean’ versus ‘Take this to the paint shop and have them mix a match.’ It may prove difficult to provide: there is a place for the

50  One can contrast epistemically sourced qualifications from criterially sourced ones, as for instance when I am unsure whether you are faking pain, whether the criteria really are present. But this is somewhat delicate as, say, epistemic unsureties about rudeness can lead one to qualify and contour more sensitively and informatively what one is facing. 51  For example, Aristotle asks whether (in Greek) ‘NN is akratic’ (full stop) says NN is akratic about everything, or only over the physical appetites. It is the latter. So, wishing to say NN is akratic about anger, you must add the qualification ‘about anger’, otherwise you will be saying something else. It is not that ‘akratic’ doesn’t apply, but here requires that added qualification. (Context might do it.) 52  ‘Don’t these in turn have their own vague borderlines?’ Maybe in certain contexts, but not necessarily. A demand for further precisification may be met with incomprehension: ‘But is the tea really hot-​ish?’

164   Gavin Lawrence imponderable and the gestural (‘It has a bit of the smell of Mexican marigold’). But there is also Vagueness as a logical problem, thrown into prominence by a commitment to sharp conceptual boundaries and the Law of Excluded Middle, and belonging in that neck of the woods. As such it is of little concern to living language, where qualification is a powerful tool, one that makes available an infinitely delicate palette.

7.5.4 Lack of Fixity The lack of fixity is notable if, like Plato, you look ‘for definitions corresponding to our concepts in aesthetics and ethics’ (PI 77). Perhaps in answer to a query about N being a good candidate, I say, ‘N is just,’ and am then asked ‘Why?’ or ‘How so?’ I may have something definite in mind as a criterion for N’s justice, and, if of a certain kind, or magnitude, it may be enough to support such a judgement, sensitive as that is to temporal extension—​and not, for example, provoke a response like ‘Yes, but that was only once.’ Equally, however, I may have nothing very definite ‘in mind,’ indeed nothing (cf. PI 179c); yet when asked, am prepared to come out with all sorts of criteria, or, surprised to have my judgement called (cf. RPP I. 245), not be prepared and left scrambling. I may, when asked, look assessively at my interlocutors and select criteria I think will help or hold water with them, even if not ones that count most with myself. And if, say, the first were shot down (‘Actually it wasn’t N who did that’), I am ready with plenty of other ‘props’ (Stützen) in the wings. Of course if these are defeated one after another, and counter-​criteria produced, I will begin to wonder: ‘It seems I don’t know N as well as I thought’ or even ‘I wonder whether we are talking about the same person’ or. . . .53 This lack of rigid limits or boundaries—​the fluidity—​in the supporting criteria, here for ‘evaluative’ concepts, but equally elsewhere, Wittgenstein finds also in our use of proper names, e.g., ‘Moses’ (PI 79). Typically we use a name without an already fixed meaning.54 ‘But how can this be?’ Won’t it now appear ‘that I’m using a word whose 53  Criteria may be vague: ‘I just feel it in my bones’ (even ‘Trust me!’); a track record of ‘good bone feelings’, may be enough, depending, e.g., on the contextual seriousness of the point, and on evidently long acquaintance with N. Such vague bases of character assessment acknowledge the place for ‘gut feelings’ or ‘a good nose for . . .’ (PI Pt. II.xi, p. 228e), for imponderables (unwägbare Evidenz), things not easily, if at all, articulable, as in a hitchhiker’s immediate judgment about whether to accept a ride, where they may say nothing beyond ‘It didn’t feel right’ or ‘Something about his eyes’; or an experienced fireman’s sense of a floor suddenly going ‘spongy’, losing structural strength (Klein 1998/​1999, 15); or chicken-​sexing, where the backdrop for ‘Trust me, I know’ is simply past success, a knowledge of one’s ability to do something, albeit without knowing how. Often it involves long experience, as of facial expressions, particularly in areas especially of the non-​conceptual (not non-​conceptualisable). See esp. PI 223–​229, esp. 227g; connected also to appreciation and connoisseurship (LW I.925). 54  Cf. PI 99. For a particular purpose, we can somewhat fix it or tighten it (cf. PI 68), make it more ‘exact’, albeit not ‘absolutely’. (Cf. PI 84a and 86c, 87, 88 with 47; much danger lies in ‘exactness’, e.g., PI 88, 91b, 100, etc.). ‘For the purposes of this conversation I am going to use “Moses” to mean the person who led. . . .’ This is in effect how Wittgenstein understands Russell’s suggestion about ordinary proper names (PI 79a); it thereby changes—‘tightens’—​the sense of ‘Moses did not exist’, which otherwise would beg for clarification.

Criterialism   165 meaning I don’t know?’ (PI 79d).55 Well, its lack of any such fixity evidently doesn’t impair my use of it nor others’ understanding of it! My ability to answer ‘Who do you mean?’ with criteria of all sorts. I might have had none in mind; not have known which I would choose—​and while thinking, suddenly see N through the window, or in a photograph, perhaps never seen before: ‘That’s N!’56 And if one fails, I just choose another . . . I may have little idea what range, what ‘cluster’ of descriptions, and of other things, I have as supports to draw on. ‘Just look!’ (cf. PI 68, 40, 71, 77, 81).57 Just as if asked ‘By ‘N’ do you mean the φ-​er?,’ for many φs, I may say ‘yes,’ even for φs I might never have thought of myself. ‘The sign-​post is in order—​if, under normal circumstances, it fulfills its purpose’ (PI 87c). Theory tempts us to a false fixed ideal of exactly ‘what gets said’: natural language is ‘fixed’ enough (‘Stand there!’). In seeking a ‘theory’ of how names, as with other predicates, work in language, we look for something fixed that for natural purposes isn’t fixed, doesn’t need to be fixed, and can’t as such be ‘fixed,’ albeit tightened relative to a particular purpose, a tightening that may then somewhat reconfigure the game played with it and the sense and truths depending from it, etc. (PI 79a). We can distinguish these three aspects. The multiplicity, or diversity, of a particular game. This occurs both externally and internally. Thus we talk of justice in regard to acts, practices, character, institutions, and again adverbially of manners (‘justly’); the criteria, while related, differ; each comes with its surround of a space of relevant reasons and with a surround of what is intelligibly arguable. Yet also, internal to each, the shape of this surround too diversifies with particular contexts, and the standards and concerns normally present in those. (And all come with a further surround of how intelligibly they can be recontextualised. It turns out not a one-​off problem of one child bullying another, but something endemic to the school: the parameters of consideration and argument shift.) The qualifiability of predicates. As earlier, qualification is another aspect, working in a number of ways, whether to avoid misleadingness or for refinement and deepening appreciation and understanding. Third, the penumbra of predicates. Change 55 

Cf. also, e.g., PI 68b, 70, 82. Hallett (1977, 161) refers to some Geach ‘lecture notes’ where Geach correctly says that these sections are concerned with the rejection of ‘rigid criteria’ for the use of a word, yet claims Wittgenstein remained ‘a disguised descriptionist’. Rather I take Wittgenstein to focus on our actual use of ‘Whom do you mean by “N”?’, in answer to which a vast range of criteria is possible, as with the photograph. (Is that a “description”? Is directing a police artist’s sketch ‘visual describing’? etc. What traction is at issue?) 57  ‘Who is N?’ ‘Someone I knew long ago’; ‘A man you don’t know’; ‘A man I once read about, but can’t remember where—​and maybe I haven’t got the name quite right—​anyway this guy. . . .’ This may be enough (and nothing need be untoward in indefinite descriptions). Perhaps your doubt was whether I was updating you on information about a mutual acquaintance and was supposing N someone you knew. You can register N in your universe to the extent needed; you may even have ‘clues’ of how to find out more, although you may not, and that not matter at all (‘Someone my father once read about in a book’). The sort of things we may normally be expected to know to underwrite a successful use of a name, absence of which might defeat it, varies from context to context, with the language-​game (as with exclamations: PI 27). ‘I wonder whether N had a happy life’ said standing before an old gravestone marked ‘N’. But maybe this is defeated—​can’t be ‘played’—​if the tribe’s custom is to outnumber real gravestones with fake ones, ‘to confuse the evil spirits’; or can be played only if qualified, ‘whether N, if he/​she was a real person . . .’ 56 

166   Gavin Lawrence the context(s) sufficiently from the normal ones where the application has its home(s), which give the concept its life and meaning, and it drifts towards unintelligibility. This itself appears in different ways: it may enlarge a use (like metaphor); or be recoverable with further explanation; or be a simple, correctible, misuse; it may be a (comic) playing with language (a child’s iterated ‘Why?’); or be Philosophical, in taking the superficial form of grammaticality to underwrite generalised intelligibility—​’the aura of sense travels with it’ (PI 117, 594a), away from those particular contexts that gave it life and meaning/​use.

7.6  The Point(s) of ‘Language-​Games’/​L ife-​Games So far criterialism, regarding a predicate’s application, is modelled on the rules of a game and on an understanding of such rules, or criteria, as open-​ended (‘and so on ad inf.’).58 Yet still under-​discussed is the equally important aspect of the point of these predicates and of the language-​games that employ them, of why we have the predicates we do. This aspect is evident in the focus on meaning as use, the point of a tool.59 En passant: just as a word is a word within a sentence, so a sentence is so within a language (cf. PI 49); so too our predicates are pieces within a larger game (rules for a pawn, within rules for chess), games that are flexible at every level: We see that what we call ‘sentence’ and ‘language’ has not the formal unity that I imagined but is a family of structures more or less related to one another. (PI 108a, cf. 18, 23a, 199c)

So the point(s) at issue in a language-​game, in applying a predicate (sentence, language, table, cruel), can come into nearer and further focus, narrower and wider contexts, and take on different nuances, colour, and resonance—​import and pointedness—​from its settings in and interactions with different ranges of other predicates and behaviours. This holism and contextualism, language’s delta-​structure,60 fends off the resurgence of a tendency to atomism, now with practice or ‘game,’ as before with ‘name.’ (A) The Multiplicity of Language-​Games. Clearly we employ the predicates we do—​ maintain, develop, invent, change, or lose them as may be—​because they speak to our 58 

Wittgenstein rejects the view he shared earlier, of language as a matter of operating calculi with fixed or definite rules (PI 81, cf. 227e). 59  PI 43, cf. 30a, 120f, 138, etc. A foretaste, in connection with mathematical propositions, occurs in TLP 6.211: ‘(In philosophy the question “What do we actually use this word or this proposition for?” repeatedly leads to valuable insights.)’ 60  Cf. Lawrence (2018, 240–​241). Cf. PI 122a, ‘seeing connexions’. Language is multifunctional; its many tributaries intersect, intertwine, flow back, the channels and sandbanks shift, etc.

Criterialism   167 needs, interests, concerns, and problems; and as these change humanly, locally, or temporally, so do our predicates. They are our tools for our human purposes and adapted to them. Language doesn’t have a single central function—​description (PI 23d, 304 vs. TLP 5.471–​472): clearly our language-​games have a multiplicity, a variety of different uses, of points or functions (PI 23, 25, 27a), just like games. And there is a multiplicity to what we call ‘description’: it has multiple functions or uses (PI 24b, 290b, 291). (‘Describe him to me!’ Would a series of headshots from different angles count as an answer? This depends on the context, on what is at issue.) So, as ever, both external and internal variation. These language-​games constitute and articulate our form of life at its particular historical juncture (PI 23b; cf. LA I.5, I.26). We have focused on the application conditions, the criteria, for predicates. But we need also to take account of the point of their uses: the import and resonance of their application. Our interest, say, in an action’s being kind voices our immediate concern with meeting another human’s need, yet, so doing, resonates further, in ever widening contexts: it is the sort of action by people we want to associate with, whom we want to live amongst; it expresses and reinforces a communal ideal, and informs our sense of our species being. ‘What belongs to a language-​game is a whole culture’ (LA I.26). (One thinks of Herder.) Generally the different levels of predicates and their various interactions allow for a multiplicity of mutual illumination (as goes on in playing a game). It is important not to be misled here. We are not, in our talk of criterialism, talking about mere concepts or words, but about what is going on in real human life at a certain historical juncture.61 Die Umgebung einer Handlungsweise: the surround of a way of acting or going on (GB 16e; and its whole context). All too gesturally, I draw attention to three other aspects, all somewhat working together. (B) The ‘Upwards Perspective’: From Need and Interest to Predicate. There are upward-​looking pressures to increase, and fill out, what we are finding deficient in our current expressive power. (i) One pressure is for enrichment. After a hike, I report seeing some birds and flowers, even butterflies, and their various doings, singing and blooming: ‘How beautiful!’ If, however, I have some knowledge, the whole experience is richer: some species unexpected, some on a migration path, earlier than usual, the alarm call of a green woodpecker, the starlings were anting, etc.; likewise curious absences and malfunctions. I notice what is going is on around me in a wholly different register: I understand it or begin to, aware of its complexity. Admittedly an observer, say, a hunter-​gatherer, long acquainted with the terrain and its seasons might have some of such understanding, even without an identifying vocabulary (or with one slanted to more specialised tasks), but its presence is an aid and stimulus to observation and engagement, and positions one more easily to pursue further

61  For example, LA I.5 and n2: ‘When we build houses, we talk and write. When I take a bus, I say to the conductor: “Threepenny.” We are concentrating not just on the word or sentence in which it is used . . . but on the occasion on which it is said.’ Cf. ‘The question is: “In what sort of context does it occur?” ’ (PI Pt.II. ix, p. 188e).

168   Gavin Lawrence knowledge.62 So with human behaviour and practical life generally. Our vocabulary, forged from human experience, enables us to frame the possibilities of what is going on more complexly, to register limits in our appreciation, and to contour a more sensitive and informed response (including knowing what further information to seek). Sophistication builds on sophistication (as with ‘rude’ earlier). Even with no further additions we have a linguistic palette rich enough to explore an infinity of hues, with the possibilities of telling conjunctions (‘vulnerability and entitlement’), compounds (‘the howdumbdeado’nicht’), portmanteau words, modifications (adverbs of manner and degree), qualifications (‘blunt to the point of rudeness’), metaphors and similes, repetition, all the many figures of speech, etc., not to mention wider behaviour, of emphasis, tone, gesture, body-​language, mime, etc.63 Moreover, enrichment goes in both directions: an articulation of what is in a sense already there, albeit its wider settings more easily achieved and appreciated through the lens of the predicates; and this reciprocally widens the possibilities of what can be there—​of intentional action, of action-​ descriptions, enabling things of more resonance to be done. A kind of cantilevering.64 (ii) Then there is pressure for releasing expressions, pressure from interests and concerns, perhaps overlapping, that make certain factors salient to us, make us look out for and register certain similarities across a range of situations. We give each other a look, a wink, an ‘ooh’ or ‘aah’ of common understanding, although as yet lacking a fuller linguistic articulation. This grouping of situational factors, of course open-​ended, creates the niche, the cradle, for a possible predicate: its place is prepared and invites filling with a new coinage, a compound, an old word in a new place, a foreign import, etc.: a word or phrase that releases the tension.65 (There is an analogue with the shift Aristotle sees from experience to full skill: experience aligns and groups together certain similar experiences but doesn’t yet articulate a principle that binds them together, but presses us to articulate this.) There is also an explanatory element here in practical philosophy. ‘What’s wrong with sticking your tongue out at someone?’ ‘It’s rude—​it shows disrespect.’ A certain action-​description is set in a wider context, an augmented reality. We have achieved a further articulation of what matters to us: a further elucidation of the point at issue, which makes it part of a larger map. The predicate clarifies by elucidating 62  Cf., e.g., Tinbergen (1958/​1974). The remarkable painter Joan Eardley offers us an aesthetic analogue: ‘When I’m painting . . . I hardly ever move from one spot. . . . Once I start in a place I find I don’t want to move because I am trying to do something and you are never really satisfied with what you are doing and so you keep on trying and the more you try the more you think of new ways to do this particular subject so you just go on and on’. 63  ‘[I]‌t would be possible to imagine people who as it were thought much more definitely than we, and used different words where we use only one’ (PI Pt.II. ix, p. 188e). There are various ways to explore and increase richness. 64  How the presence of language enlarges our scope of perception and action is a complex topic: e.g., a challenge to unpack ‘That’s a Monet sunset’. 65  I am thinking in the first instance of words for which there is already a grammatical profile (cf. PI 30, 29, 31c–​d). More is required for conceptual innovation (cf. PI 208). Yet presumably there is no hard and fast line between these, albeit clear cases at either end. Different languages will reach for different techniques (ancient Greek, as a building-​block language, naturally leans to compounds).

Criterialism   169 the point at issue, not just ‘naming’ it but setting it a wider context of concerns, as ‘rude’ does in the wider field of our needs to respect each other. (C) The ‘Downward Perspective’: Predicates as Criteria of Relevance. Our needs, interests, and concerns are focal points—​themselves various, and, even internally, are groupings of various factors, naturally allied and bound (tightening and loosening, incorporating new, losing old, as may be; cf. PI 23). Where central, they are likely to be marked by predicates: our concerns with justice, with kindness, with love, with our responsibilities. As one heuristic, the Stoics invited the agent to consider what behaviour was appropriate by considering the ‘names,’ the ‘masks’ or guises (prosōpa) that belonged to them (‘human’; ‘father’; ‘mother’; ‘senator’; ‘teacher’; ‘doctor’; etc.), names as norm-​setting measures (cf. the Confucian rectification of names). We can extend this to action-​descriptions. Aiming to act well, there are the criteria of its success, the virtues: wishing to do what is wise, just, kind, or generous, we are to consider things under a certain heading, from a certain perspective. These foci of concern work downward by providing needed criteria of relevance—​directing our attention, sorting out the salient and germane, the criterial factors of success and of failure, from the plethora of circumstantial factors (cf. note 28). Not just anything is relevant. Yes, we have some ideas, some paradigmatic, some rougher, of what to look for, but there is also a sense of exploration, of imagination, of being struck anew—​a creativity as alive in one’s reflection as in one’s actions. We bring these concerns to bear, as a lens or framework through which we assess situations, where our reflection takes the form of an open-​ended and imaginative exercise: factors not initially held relevant now strike us anew. (D) The Game Itself and Its Elaboration. As already noted in C8 in section 7.4, there are games where our interest is often, even typically, not so much in the application of predicates as in their further discrimination, the subtlety and elaboration of their criteria, the interplay between them, a sense of open exploration, of greater resources, and finer sensitivities. The development of a tool set enabling us to explore the terrain, what is at issue, more subtly. Often our concern is not so much, say, with whether Iago is wicked or not, but in the appreciation and deeper understanding of the form of his wickedness, the leaves of that artichoke: more injured pride than jealousy, or jealousy tinged with envy? An insecurity of identity manifesting in a needed self-​regard as secret puppet-​master? . . .66 A game may have an apparently focal predicate, the topic it seems to be all about—​say, in aesthetics, the judgement whether something is beautiful or not (such the implication of Hume’s third argument in Appendix 1). However, like ‘chess,’ it seems more mere name of the game than either its point or a move in it. The judgement is tangential, if not irrelevant, to appreciation, hardly appearing in our aesthetic conversations, whether as initial explanandum or targeted conclusion.67 Indeed there 66  ‘But he must be jealous or not jealous’. The claim here lacks interest, even sense. What are you after? Yes, there are elements of jealousy here. ‘So he must be a bit jealous or not a bit jealous’. What? Do you have an eccentric interest in the Law of the Excluded Middle? The insistence seems misplaced in this conversation, where our interest is in exploration, not rigidifying conceptual boundaries (cf. n36). 67  LA I and II, e.g., I.5, 8, 17–​18, 20, 26, 35; II.2. Think too of the use of language in developing and maintaining a friendship and of appreciation as a matter of developing a practically rational engagement

170   Gavin Lawrence may be no such central predicate to the game, or overlapping games, as in literary criticism. Often our interest in character and action is of this wider sort. It comes in also in our agential deliberation, in our interest, say, not merely in doing what is generous but in being thoughtful about the present one gives, a present that shows one’s appreciation and sensitivity to the needs of another, beyond perhaps what they realise of themselves. As with literary criticism, there is no limit to this.68

References Ambrose, A., ed. (1979). Wittgenstein’s Lectures: Cambridge 1932–​1935 (Blackwell: Oxford). Ayer, A. J. (1936/​1942). Language, Truth and Logic (Victor Gollantz: London). Carnap, R. (1963). ‘Intellectual Autobiography’. In Schilpp, P. A., ed. The Philosophy of Rudolf Carnap (OpenCourt: La Salle, Illinois), 3–​84. Clarke, S. (1706/​1969). A Discourse of Natural Religion. In Raphael, D. D., ed., British Moralists 1650–​1800, vol. 1 (Oxford University Press: Oxford). Cross, A. (2017). ‘Art Criticism as Practical Reasoning’. British Journal of Aesthetics 57, no. 3, 299–​317. Eardley, J. (1960). Taped interview, Joan Eardley Archive, Scottish National Gallery of Modern Art, Edinburgh. GMA/​A09. Foot, P. (1958/​1978). ‘ “Moral Arguments.”’ (reprinted in Foot, P. (1978), Virtues and Vices) (Blackwell: Oxford). Foot, P. (1958–​-​1959/​1978). ‘ “Moral Beliefs.”’, ibid. In Virtues and Vices (Blackwell: Oxford). Foot, P. (1963/​1978). ‘ “Hume on Moral Judgement.”’ ibid. In Virtues and Vices (Blackwell: Oxford). Foot, P. (1972/​1978). ‘Morality as a System of Hypothetical Imperatives’. In Virtues and Vices (Blackwell: Oxford), with added footnote. Foot, P. (1978). Virtues and Vices (Blackwell: Oxford). Frege, G. (1903/​1960). Translations from the Philosophical Writings of Gottlob Frege. Edited by Geach, P. and Black, M. 2nd edition (Blackwell: Oxford). Gallie, W. B. (1956). ‘Essentially Contested Concepts’. Proceedings of the Aristotelian Society 56, 167–​198. Garfinkel, H. (1964/​1967). ‘Studies of the Routine Grounds of Everyday Activities’. Social Problems 11, no. 3, 225–​250. Revised as ch. 2 in Studies in Ethnomethodology (Hoboken: Prentice Hall, NJ). Geach, P. (1956/​1967). ‘Good and Evil’. Analysis 17, 33–​42. Reprinted in Foot, P., ed. Theories of Ethics (Oxford University Press: Oxford). Geach, P. (1975). ‘Teleological Explanation’. In Körner, S., ed., Explanation (Blackwell: Oxford), 76–​93. Glock, H.-​J. (1996). A Wittgenstein Dictionary (Blackwell: Oxford).

rather than a theoretical evaluation. (See Cross 2017 for some interesting thoughts on this.) Perhaps the expression has a variety of uses, all other than the illusory one we suppose at its very heart: when transported and at a loss for words, we murmur or exclaim ‘How beautiful!’; or it marks a kind of formal success, often needless to say, like ‘Checkmate!’; or registers the speaker as not philistine but among appreciators, though not constituting their appreciation; etc. That such uses are wholly or largely extrinsic to appreciation curiously echoes Hume’s stance, though ‘without his animus’. 68 

Cf. Foot (1958/​1978, 109).

Criterialism   171 Hallett, G. (1977). A Companion to Wittgenstein’s ‘Philosophical Investigations.’ (Cornell University Press: Ithaca, NY). Hart, H. L. A. (1961). The Concept of Law (Oxford University Press: Oxford). Horty, J. F. (2012). Reasons as Defaults (Oxford University Press: Oxford). Hume, D. (1739 and 1740/​2000). The Treatise of Human Nature. Edited by Norton, D. F. and M. J. (Oxford University Press: Oxford). Hume, D. (1777/​1902). An Enquiry Concerning the Principle of Morals. Edited by Selby-​Bigge, L. A. (Clarendon Press: Oxford). Klein, G. (1998/​1999). Sources of Power (MIT Press: Cambridge, MA). Lawrence, G. (1995). ‘The Rationality of Morality’. In Hursthouse, R., Lawrence, G., and Quinn, W., eds., Virtues and Reasons: Essays in Honour of Philippa Foot (Oxford University Press: Oxford), 89–​147. Lawrence, G. (2004). ‘Reason, Intention and Choice: An Essay in Practical Philosophy’. In O’Hear, A., ed., Philosophy, Supplement No. 54: Modern Moral Philosophy (Cambridge University Press: Cambridge), 265–​300. Lawrence, G. (2005). ‘Snakes in Paradise: Problems in the Ideal Life in NE 10’. Southern Journal of Philosophy 43, supplement, 126–​165. Lawrence, G. (2018). ‘The Deep and the Shallow’. In Hacker-​Wright, J., ed., Philippa Foot on Goodness and Virtue (Palgrave Macmillan: Cham, Switzerland), 187–​255. Lawrence, G. (2020). ‘Operating under the Idea of the Ideal: Aristotle’s Naturalism’. In Hähnel, M., ed., Aristotelian Naturalism: A Research Companion Book Series (Springer: Cham, Switzerland), 11–​34. Lycan, W. G. (1971). ‘Noninductive Evidence: Recent Work on Wittgenstein’s “Criteria.” ’ American Philosophical Quarterly 8, no. 2, 109–​125. Mackie, J. L. (1973). Ethics (Penguin: Harmondsworth). Malcolm, N. (1958/​1966). Ludwig Wittgenstein: A Memoir. Biographical Sketch by G. H. von Wright (Oxford University Press: Oxford). McDowell, J. (1978). ‘Are Moral Requirements Hypothetical Imperatives?’ Aristotelian Society, Supplementary Volume 52, no. 1, 13–​29. Moore, G. E. (1903). Principia Ethica (Cambridge University Press: Cambridge). Moore, G. E. (1912). Ethics (Oxford University Press: London). Nietzsche, F. (1887/​1997). On The Genealogy of Morality (Cambridge University Press: Cambridge). Prichard, H. A. (1912/​1949). ‘Does Moral Philosophy Rest on a Mistake?’ Mind 21, 21–​37. Reprinted in Moral Obligation (Oxford University Press: Oxford). Prior, A. (1949). Logic and the Basis of Ethics (Oxford University Press: Oxford). Ridley, A. R., Child, M. F., and Bell, M. B. V. (2007). ‘Interspecific Audience Effects on the Alarm-​Calling Behavior of a Kleptoparasitic Bird’. Biology Letters 3, 589–​591. Schlenker, P. (2018). ‘What Is Super Semantics?’ Philosophical Perspectives 32, no. 1, 365–​453. Searle, J. (2011). ‘Wittgenstein and the Background’. American Philosophical Quarterly 48, no. 2, 119–​128. Stevenson, C. L. (1937). ‘The Emotive Meaning of Ethical Terms’. Mind 46, 14–​31. Strawson, P. F. (1963). ‘Carnap’s View on Constructed Systems versus Natural Languages in Analytic Philosophy’. In Schilpp, P. A., ed. The Philosophy of Rudolf Carnap (Open Court: La Salle, Illinois), 503–​518. Strawson, P. F. (1985). Skepticism and Naturalism: Some Varieties (Methuen: London). Tinbergen, N. (1958/​1974). Curious Naturalists, revised edition (Penguin Education: Harmondsworth).

172   Gavin Lawrence Wittgenstein, L.[AWL] Wittgenstein’s Lectures, Cambridge 1932–​ 35, from the notes of A. Ambrose and M. MacDonald, ed. A. Ambrose (Oxford: Blackwell, 1979). [BB] The Blue and Brown Books (Oxford: Blackwell, 1958). [BT] The Big Typescript, tr. C. G. Luckhardt and M. A. E. Aue (Oxford: Blackwell, 2004). [GB] ‘Remarks on Frazer’s “The Golden Bough,” ’ ed. R. Rhees (Synthese, 17, 1967), 233–​253. [LA] Lectures and Conversations on Aesthetics, Psychology and Religious Beliefs, ed. C. Barrett (Oxford: Blackwell, 1970). [LSD] ‘The Language of Sense Data and Private Experience –​Notes taken by R. Rhees of Wittgenstein’s Lectures, 1936’, in Philosophical Occasions, ed. J. Klagge and A. Nordman (Indianapolis: Hackett, 1993). [LW I] Last Writings on the Philosophy of Psychology, Volume 1, ed. G. H. von Wright and H. Nyman, tr. C. G. Luckhardt and M. A. E. Aue (Oxford: Blackwell, 1982). [LW II] Last Writings on the Philosophy of Psychology, Volume 2, ed. G. H. von Wright and H. Nyman, tr. C. G. Luckhardt and M. A. E. Aue (Oxford: Blackwell, 1992). [OC] On Certainty, 2nd edition, ed. G. E. M. Anscombe and G. H. von Wright, tr. D. Paul and G. E. M. Anscombe (Oxford: Blackwell, 1969). [PI] Philosophical Investigations, 2nd edition, ed. G. E. M. Anscombe and R. Rhees, tr. G. E. M. Anscombe (Oxford: Blackwell, 1958). [RFM] Remarks on the Foundations of Mathematics, 2nd edition, ed. G. H. von Wright, R. Rhees, and G. E. M. Anscombe, tr. G. E. M. Anscombe (Oxford: Blackwell, 1978). [RPP I] Remarks on the Philosophy of Psychology, Volume 1, ed. G. E. M. Anscombe and G. H. von Wright, tr. G. E. M. Anscombe (Oxford: Blackwell, 1980). [TLP] Tractatus Logico-​Philosophicus, tr. D. F. Pears and B. F. McGuiness (London: Routledge and Kegan Paul, 1961). [Z] Zettel, ed. G. E. M. Anscombe and G. H. von Wright, tr. G. E. M. Anscombe (Oxford: Blackwell, 1967). Zuberbühler, K. (2009). ‘Survivor Signals: The Biology and Psychology of Animal Alarm Calling’. Advances in the Study of Behavior 40, 277–​322.

Chapter 8

An sc om be on D ou bl e E f fect and I nt e nde d C onsequ e nc e s Cyrille Michon

8.1  Introduction The traditional doctrine of double effect (DDE) was an important source of inspiration for Elizabeth Anscombe’s moral philosophy and philosophy of action. Rooted in a famous text by Aquinas on self-​defence,1 the doctrine draws a conceptual distinction between the intended consequences of an action and those that are only foreseen, and grounds on this distinction a moral distinction. This last is often understood as the permissibility to bring about a certain state of affairs (e.g., killing an innocent) knowingly while it would be impermissible to bring it about intentionally, even only as a means to a good end. Thus, the doctrine connects the analysis of action to the question of moral duties and responsibility, in particular with reference to absolute prohibitions (things one can never do). It is no surprise that, at the time she wrote Intention (1957) and ‘Modern Moral Philosophy’ (1958), Anscombe discussed the DDE, even if she does not mention it explicitly in those texts. But she does in other texts of the same period, which are in line with these two ‘highlights.’2 In fact she already referred to the doctrine, in a quite precise way, showing good knowledge of Aquinas’s position, in her first published paper, almost twenty years before (1939), saying it required a ‘philosophical elucidation

1  Summa theologiae II-​II, q. 64, a. 7 is the source of the notion of duplex effectus. Though subject to different interpretations, the traditional understanding of it is that a private person may resist an attacker with proportionate violence, even while foreseeing his possible death, but may not intend to kill the attacker. The bad outcome should be outside the intention (praeter intentionem). People in charge of the community have a right to kill intentionally, though not for personal reasons (revenge). 2  TD, 61–​70; WM, 51–​61; GW, GG1, 243–​248; and later see ‘Simony in Africa,’ GG2, 239–​246.

174   Cyrille Michon of the notion of intentional action,’ and thus already setting the stage for her future work on action.3 More than twenty years later (1982), she devoted her Aquinas lecture to both the philosophy of intentional action and the DDE.4 There, she seemed to adopt a very cautious attitude towards the DDE and a quite inclusive notion of what is an intended consequence as opposed to a merely foreseen one, arguing that some consequences of an action should be held as intended even if they did not belong to the precise aim of the agent: it would be hypocritical to deny that one intended killing civilians by bombing an area including a city, even if the purpose was not to kill civilians but only to destroy military targets. The consequence is simply ‘too close’ to be considered merely foreseen, and the DDE should rather draw a line between more and less distant consequences. Such a criticism of the abuses of double effect has itself been criticised as a departure from a sound understanding of intention and of her own analysis. This is the aspect of Anscombe’s conception of the DDE that this chapter is mostly concerned with. In the section 8.2, I present the doctrine and Anscombe’s general understanding of it, of its indispensability for ethics, and of the possible abuses of it. In section 8.3, I expose the ‘Closeness Problem,’ which arises from the kind of abuse just mentioned, and Anscombe’s answer to it in her 1982 lecture, insisting on the distance and accidentalness of merely foreseen consequences. In section 8.4, I deal with the accusation of inconsistency between her moral philosophy and the analysis of action she developed in Intention and argue that she has made room, in a specific way, for a distinction between two scopes of intention. In section 8.5, I suggest that this distinction can be generalised and that we then have two conceptual distinctions: between narrowly and broadly intended consequences and between broadly intended and merely foreseen ones. Finally, section 8.6 argues that both metaphysical distinctions can ground moral ones, and thus reshape the debate over the DDE.

8.2  The Doctrine of Double Effect in Anscombe In her unpublished review of Glanville Williams, Anscombe summarises the DDE thus: The doctrine states that in considering an intentional action we must distinguish between what is intended and what can be foreseen but is not intended. What is intentional will be the end (supposing it to be achieved) and the means chosen for it; and in order for an action to be sound, both end and chosen means must be legitimate. Things which can be foreseen as side effects of what is done, which would not 3  Elizabeth Anscombe and Norman Daniel, ‘The Justice of the Present War Examined: A Criticism Based on Traditional Catholic Principles and Natural Reason,’ 1939. This pamphlet (JPW) is reprinted in CP3, 72–​81. 4  AIDE, 196–​215.

Anscombe on Double Effect and Intended Consequences    175 themselves be permissible either as ends or as means, will not vitiate the action if it is itself necessary; and may not vitiate it if it is good and the side effects not disproportionately evil; a standard illustration would be that a woman could take quinine (if it were the only remedy) to cure malaria, though it might also cause abortion, but could not take a remedy with that sort of effect to relieve a headache.5

Another common illustration, under the supposition that war may be fought for a just cause, and that there is a just cause, is the opposition between Strategic Bombing, intending to destroy a military target while causing some civilian deaths as foreseen side effects, and Terrorist Bombing, aiming at causing civilian deaths in order to improve military prospects by demoralising the enemy. One could also speak of collateral versus focal damages, or of accidental versus instrumental harms. The DDE was traditionally invoked as giving permissibility conditions: the first kind of damages or harms could be permitted in certain circumstances, while the second ones would be prohibited whatever the circumstances. But the principle that roots the DDE, the distinction between foreseen and intended consequences, can also be invoked more generally to say that the first ones would be easier to justify than the second ones. Finally, the DDE bears traditionally on acts (type), but many philosophers, even amongst those who criticise the doctrine (like Tim Scanlon or Judith Jarvis Thompson), would accept it as an indication for the moral evaluation of agents.6 Anscombe seems to have constantly understood the doctrine as bearing on acts and as delivering only permissibility conditions. She refers to it, in a traditional way, as an accompaniment to and softening of her absolutism: though there are things one can never do, whatever the reasons and whatever the circumstances, there are cases in which one can bring about the same bad effect if that effect is not intended, neither as an end nor as a means, but is only a side effect. The doctrine is both important and delicate and requires a good philosophy of action, as Anscombe already said in JPW. Let us set aside her judgement that the war was not 5  GW, GGI, 247. The traditional presentation of the DDE mentions three or four conditions: 1. the action must be good in itself or morally neutral; 2. the good effect must not be achieved by means of the bad; 3. the reason for the action must be the right effect and not the wrong one; 4. the bad effect must be proportionate to the good effect (in the sense that it would be a greater evil to avoid it without producing the good effect). See Jean-​Pierre Gury, Compendium Theologiae Moralis (Regensburg: Georgii Josephi Manz, 1874) and more recently T. Cavanaugh, Double-​Effect Reasoning, Doing Good and Avoiding Evil, Oxford Studies in Theological Studies (Oxford: Clarendon Press, 2006), ch. 1, who quotes Gury (25). Cavanaugh elaborates on the last condition (31–​37). 6  See T. M. Scanlon, Moral Dimensions: Permissibility, Meaning, Blame (Cambridge: Basic Books, 2008), ch. 1. For a general overview of the current debate, see William FitzPatrick, ‘The Doctrine of Double Effect: Intention and Permissibility’, Philosophy Compass 7 (2012), 183–​196 and Alison McIntyre’s entry ‘Doctrine of Double Effect’, Stanford Encyclopedia of Philosophy, September 7, 2011, http://​plato.stanf​ord.edu/​archi​ves/​fall2​011/​entr​ies/​dou​ble-​eff​ect/​, as well as her criticism in ‘Doing Away with Double Effect’, Ethics 111, no. 2 (2001), 219–​255. A recent discussion is in D. K. Nelkin and S. C. Rickless, ‘So Close, yet So Far: Why Solutions to the Closeness Problem for the Doctrine of Double Effect Fall Short’, Noûs 49, no. 2 (2015), 376–​409 and ‘Three Cheers for Double Effect’, Philosophy and Phenomenological Research 89, no. 1 (2014), 125–​158.

176   Cyrille Michon fought for a just cause by her country and her evaluation of the international situation, which she herself recognised to be an underestimation of Hitler’s intentions and of the danger for Europe. She did not deny that war in general could be justified by a just cause, nor that there was a just cause for that war, and she mentioned the DDE, with explicit reference to Aquinas, to explain that justified acts of warfare might include bombings with the side effect of killing civilian people. The DDE allowed it even if one held as an absolute principle the prohibition against intentionally targeting civilian populations. But Anscombe already expected that the war would lead to bombardments of cities, the killing of whose civilians could not be considered a side effect, and so to violations of this absolute prohibition. Her fears were to be more than confirmed by the (terrorist) bombings over Nagasaki and Hiroshima. Those events motivated her protest against Truman’s Honoris Causa doctorate in 1956. And the justifications given in defence of Truman’s degree prompted from Anscombe a severe criticism, which is in part a criticism of wrong applications of the DDE. One could not argue, as some did, that the president did not really intend to kill innocent people but just to put some ink on some paper; nor that he intended only to stop the war and limit the number of victims: that might have been his end, but he achieved it by bombing cities and so by killing non-​harming (innocent) people, these outcomes being aimed at as means, and therefore counting as intentional. This controversy is the background of Intention, which elaborates on the crucial distinction between foreseen and intended consequences, even though Anscombe does not mention the DDE. Nor does she in MMP, but she does target those ethical views that follow Sidgwick in obliterating the distinction between intention and foresight and fall under the same criticism of ‘consequentialism.’ The idea that moral evaluations must be decided only on the basis of expected consequences is incompatible with the ethical tradition (Judaeo-​ Christian, but not only Judaeo-​Christian) according to which there are actions one should never do, such as intentionally killing the innocent.7 As she said in GW: [O]‌nly if we make this distinction between the intentional and the foreseen is it possible to avoid a course of reasoning which will justify anything at all, however atrocious. And we all do make such a distinction: for example, a man who goes to prison rather than do something disgraceful may thereby knowingly deprive his family of support, but is not held to be therefore guilty of this as if he had aimed at it or chosen it as a means to some end of his.8

Without absolute prohibitions and with a consequentialist approach, a judge might be justified in condemning an innocent in order to avoid a riot with many foreseen victims. But, it might be claimed, absolute prohibitions could lead someone to the dilemma

7  Anscombe makes the point (in TD and WM) that pacifism, in making the same assimilation of the intended (killing) to the foreseen, and rejecting both, ends up with the legitimisation of intentional killing of the innocent in extreme cases. 8  GW, GGI, 247; same example in MMP.

Anscombe on Double Effect and Intended Consequences    177 of doing something sinful whatever the choice. Here the dilemma is avoided because only the intentional deprivation of support would be a sin, not its occurrence as a side effect of avoiding a sin. And more generally, without the DDE, the doctrine of absolute prohibitions could not resist the objection that such prohibitions are unrealistic. It would lead to abstaining from building roads or cars, flying planes, practicing dangerous surgery, and so on, since one knows that the habitual performance of such actions will necessarily bring about the death of innocent people.9 The DDE is then important to protect the doctrine of absolute prohibitions from too easy criticism, but it is also suspect for it can be wrongly applied and abused. ln the same text on Williams, Anscombe forged an epigram that the corruption of non-​Catholic moral thought has consisted in the denial of this doctrine, and the corruption of Catholic thought in the abuse of it.10 The first corruption leads to consequentialism and the abandonment of absolute prohibitions, which are a ‘bedrock for ethics.’ The second corruption maintains verbally absolute principles, but in fact leads also to a form of consequentialism, since it becomes very hard to see how any bad effect need be intended. I will turn now to this second form of corruption. In different texts, and the idea is at the heart of Intention, Anscombe explains that the intellectual abuses in question are linked to a false, Cartesian conception of intention as a mental act, an interior little speech, that one could direct at will, and hence to the erroneous moral theory of the ‘direction of intention.’11 But while what we intend depends on us, it is not up to us, when we bring about a state of affairs for the sake of some goal, whether or not we intend it. To say that one intended the end but not the means is a gross hypocrisy. Thus, in her first paper, she said: It has been argued that it is justifiable to attack civilians because their death is an example of ‘double effect.’ But this is no example of double effect, which is

9 

As she would later say, ‘One cannot say that no action may be done which foreseeably or probably leads to some death, or that all such actions are murderous. Why, the very begetting of a child would be murder at that rate—​for the child will surely die. Or if that seems too zany an example: you can’t build roads and fast vehicles, you can’t have various sports and races, you can’t have ships voyaging over the seas, without its being predictable that there will be deaths resulting. And much that is done in medicine and surgery is done knowing it involves the risk of death—​painkilling drugs which may kill the patient before his disease does, and high-​risk surgery’ (AIDE, 208–​209). 10  GW, GGI, 247; WM, 58. 11  See I, §25; GW, GGI, 247. Generally one refers to Pascal’s battle against the Jesuits in the Provinciales (esp. Letter VII) and their authorisation of a form of mental reservation which would not be a lie, but would nonetheless be an intentional disguise of truth (‘She is not here’ adding mentally ‘in this room,’ though she is in the house). Or the most famous example of the interdiction of duelling which would still allow one to reach the location where the duel should take place, take a sword, encounter one’s enemy, and be then in a state of legitimate defence. This whole doctrine of ‘direction of intention’ is a target of her philosophy of intention that has immediate moral relevance (see also ‘Simony in Africa’ in GG2, 23*). See T. Cavanaugh, ‘Abuses of Double Effect: Anscombe’s Principle of Side-​Effects, and a (Sound) Account of Duplex Effectus’, in Double-​Effect: Theoretical and Practical Challenges, ed. John O’Callaghan and Craig Iffland (Notre Dame, IN: University of Notre Dame Press, forthcoming).

178   Cyrille Michon exemplified when an action designed to produce one effect produces another as well by accident. (JPW, 78)

A second, more subtle hypocrisy, and abuse of the DDE, according to Anscombe, is what would later be called area bombing: It is a different thing, while making one group of persons a target, to kill others by accident, and to make a group of persons a target, in order, by attacking them all, to attack some members of the group who are persons who may legitimately be attacked. The first case involves no sin; the second involves murder and is not an example of double effect. (JPW, 79)

That seems clear enough: one cannot target a whole in order to reach a part without targeting the whole being a means to it. But, it might be objected, cannot one bomb an area in order to reach some part of it, without aiming at killing the other people in the area, even though it is certain they will be killed? Their death is not a means to the end. Anscombe’s reason seems to be that the action which is a means to the agent’s end is eo ipso a killing of innocent persons.12 This result is too close to what the agent explicitly intended (i.e., bombing the area) to be considered as a side effect, exterior to the means. While the first corruption of thought consists in making out that one can drop the means out of the scope of the agent’s intention (which would be limited to the end), the second form of corruption consists in making out that one can drop from the scope of the agent’s intention certain effects or aspects of the means (the bad aspects), even though they are essential to the action. This explanation is not given explicitly in JPW, but it is in Anscombe’s Aquinas lecture (AIDE). There is then a certain consistency in Anscombe’s treatment of this second kind of corruption—​the ‘abuse’ of the DDE. Which is not to say that this moral verdict is in line with her philosophy of action in Intention, nor that it is correct. I will come to this, but before I do so we have to examine the paradox that emerged from the subtle kind of abuse of the DDE, the so-​called Closeness Problem, and look at Anscombe’s detailed later answer to that problem.

12  In a manuscript note published by L. Gormally, ‘Intention and Side-​Effect: John Finnis and Elizabeth Anscombe’, in Reason, Morality and Law: The Philosophy of John Finnis, ed. J. Keown and R. George (Oxford: Oxford University Press, 2013), 93–​108, Anscombe makes the same point about the shooting of an arrow which is not eo ipso the transfixing of a man, even if there is a risk of transfixing a man. But, in certain circumstances, if the purpose in the shooting could be attained only by transfixing a man (means), then one could not argue that one did not intend to damage the person, on the basis that one would have avoided the damage of the transfixing if it were possible. Transfixing is eo ipso damaging the person. The analogy serves to illustrate the case of abortion by craniotomy in which the killing could not be considered a side effect, but the means itself.

Anscombe on Double Effect and Intended Consequences    179

8.3  The Closeness Problem and the Intentionalness of an Action In AIDE Anscombe retracts two aspects of her preceding defence of the DDE. She now considers that the doctrine as such is a failure because it is a ‘package deal,’ giving a kind of recipe for the permissibility of actions bringing about bad effects that it would be intrinsically unjust to bring about intentionally. She has two worries: one is its condition as a recipe for giving moral cover to what would otherwise be vicious acts, and the other is the importance given to the consequentialist condition of the doctrine on the balance of good and evil.13 I must confess that I do not fully understand why this attention to consequences should be such a threat. Maybe she thinks it creates a consequentialist atmosphere that would undermine the real point of the doctrine. But it is clear that one has to take consequences into account, and it is clear that this condition comes after the condition concerning the intended versus foreseen distinction; it does not trump it. On the contrary: according to the DDE, the intended/​foreseen condition must be first applied, and it should trump the mere consideration of consequences. Now if we drop the balance condition, we no longer have a set of jointly sufficient conditions for permissibility or excusability, but only the necessary condition that Anscombe calls the ‘Principle of Side Effects’ (PSE), according to which ‘the prohibition on murder does not cover all bringing about of deaths which are not intended.’14 Of course, she agrees that something must be added to the PSE in order to tell when it is permissible to bring about the unfortunate foreseen result. And it seems difficult not to introduce some form of balance of good and evil here. But it might not be automatically sufficient to declare the action permissible once the other conditions are satisfied. The PSE does not, by itself, give licence to perform the action, but it makes room for its permissibility. It is a necessary condition for exemption of guilt, and the refusal of the ‘package deal’ is the denial that one can formulate a set of conditions that would be sufficient in all cases. The second retraction is more exegetical: Anscombe now thinks that Aquinas’s original source for the doctrine, in his text on self-​defence, does not present the doctrine of double effect. I guess this is because Aquinas’s reference to a proportionate self-​defence is not a reference to the balance between good and evil consequences, but only to the measure of violence necessary to repel the aggression. She does not deny that the text relies on the crucial distinction between intended and foreseen consequences, but she refers, without comment, to another text by Aquinas (Summa Theologica I-​II, q. 20, a. 5), 13 

‘[I]‌f we adopt that one principle, of the balance of good over evil in the expected upshot, then it becomes obscure why we could not do this where the causation of death was perfectly intentional’ (AIDE, 213). 14  AIDE, 214. R. Teichmann formulates the PSE thus: ‘There are some things which, if aimed at either as ends or as means, necessarily render the action in question a bad action; but which, if brought about as a foreseen side effect of an action, do not necessarily render that action a bad action’ (The Philosophy of Elizabeth Anscombe [Oxford: Oxford University Press, 2008], 113; see the discussion on 112–​120).

180   Cyrille Michon which shows that an essential (per se) consequence of an action always adds to its goodness or badness, even if it is not foreseen (praecogitatus), but that this is not true of accidental (per accidens) consequences.15 More important is her discussion of what has come to be called the Closeness Problem. The cases she had already considered of acts, like area bombing, whose consequences are too close to the intended means to be taken as not intended, had been turned into an objection to the DDE by Philippa Foot in her famous paper ‘The Problem of Abortion and the Doctrine of Double Effect’ (1967).16 Foot constructed a scenario (I will call it Pot-​Holer) that put the doctrine in trouble: the spelunkers in the cave have no other way of getting out, as the water rises and threatens to drown them all, than to blow up a fat man (the pot-​holer in Anscombe’s version) who is stuck in the only exit. How could one say that in blowing him up the others would not kill him intentionally? But at the same time, the death of the innocent pot-​holer is not a means towards the end of getting out. It is a physically necessary consequence of the means, which is just to reduce the man’s volume, in order to make room for the others to get out. But the consequence seems too close to the action’s intended result to be considered differently from an intended effect. Either the conceptual distinction between foreseen and intended effects is absurd in this case, or this distinction, which is metaphysically grounded, is morally irrelevant. Either way that is simply too bad for the doctrine. The closeness objection ruins the principle of double effect by showing its inconsistency. In her answer to Foot, in AIDE, Anscombe makes two moves concerning this scenario. One is to compare it with another scenario, in which a close rock situated above the man’s head can be moved, so that the man will be killed anyway, his head being crushed.17 She admits, without discussion, that in Pot-​Holer, blowing up the man himself is an intentional killing, and so a murder. She also admits one might find a difference between Pot-​Holer and Close Rock. (Some people in the cave might refuse the first action as murder, but see the second as not murderous.) But if the rock is close enough, the difference vanishes. And she argues that even if the killing of the fat man is not a means, and not ‘part of the aim,’ it is nonetheless intentional, and it would be absurd to apply the DDE/​PSE: it would come down to the mere choice of a description under which one acts, to a ‘direction of intention.’18 So one should regard moving the rock in this situation as a murderous act. 15 

Since the text, as understood by Anscombe, does not bear on foreseen consequences, I do not see how it can be relevant to the DDE. 16  Philippa Foot, ‘The Problem of Abortion and the Doctrine of Double Effect,’ Oxford Review 5 (1967): 5–​15, reprinted in Virtues and Vices and Other Essays in Moral Philosophy (1978; Oxford: Blackwell, 2003), 19–​32. 17  Anscombe’s version: ‘Imagine a pot-​holer stuck with people behind him and water rising to drown them. And imagine two cases: in one, he can be blown up; in the other, a rock can be moved to open another escape route, but it will crush his head. He will be killed by it’ (AIDE, 210). 18  ‘At this point the Doctrine of Double Effect helps itself to an absurd device, of choosing a description under which the action is intentional, and giving the action under that description as the

Anscombe on Double Effect and Intended Consequences    181 Second move: suppose the rock is at some distance but will probably kill the man if moved. The uncertainty and the remoteness give a principle to draw a line between this case and the preceding ones, since there is room to say that one does not intend to kill the man, and so, using the principle of side effects, one might move the distant rock without being guilty of murder. Or at least, one might see oneself as not incurring guilt, though a very strict moral agent might also see himself as incurring moral guilt in that case. As such, the response is not convincing. First, both features, uncertainty and remoteness, might be undermined by a sorites argument. Concerning uncertainty: it is obvious that to bring about a bad consequence with full certainty is worse than taking the risk of bringing it about. But what is the right measure of uncertainty? And, in any case, how could you apply the principle of side effects to the cases of shutting doors to avoid sinking or to avoid a fire, when it is certain that people on the other side of the doors will die? Concerning physical distance: how distant or remote must the rock be? And if one thinks more of causal than of spatial distance: how many other relevant conditions must be in place for the effect? True, vagueness could be admitted in the application of the concept of ‘murder’ as well as in the measure of sufficient distance. And one might see oneself as allowed or forbidden to X because one sees the distance between X and Y as sufficient or insufficient.19 Second, there is no choice of a particular description of the action, and so no ‘direction of intention.’ The agents do not restrict their intention to the end (escaping) at the expense of the means (blowing up the man or moving the rock), nor do they choose the description that corresponds to their means, to what is useful for the end: this is not dependent on them (as the choice of the means is). If they denied that the death of the man was intended, that would be only because that death is not useful in itself: it does not belong to their purpose, though it is of course very close.20 Third, Anscombe admits without discussion that blowing up the man is an intentional killing, and asserts the same of Close Rock, a situation she compares to moving a knife in a certain region of space where there is a certain living human body, or cutting a rope with a climber suspended. And she adds: ‘Nonsense,’ we want to say, ‘doing that is doing this, and so closely that you can’t pretend only the first gives you a description under which the act is intentional.’ (AIDE, 222)

intentional act. ‘I am moving what blocks that egress’, or ‘I am removing a rock which is in the way’. The suggestion is that that is all I am doing as a means to my end’ (AIDE, 210). 19 

See section 8.5, note 37. Thanks to Roger Teichmann for pointing out that ‘absolutely prohibited’ applies to act-​descriptions which themselves may be sorites-​vague. 20  I do not distinguish between an intentional killing and an action in which the death is intended and where the death occurs. This is not to say that the death is the agent’s final end; it might be a means, but then it is chosen and so intended (as an intermediary end).

182   Cyrille Michon We might agree that it would be nonsense to make a moral difference (or an important one), but still claim that in the proper sense of ‘intending,’ the killing was not intentional (the death was not intended). In Intention Anscombe had shown that we cannot automatically substitute one description of an action for another, even when both are recognised by the agent, in order to give the content of her intention. (One may intend it under one description but not under the other).21 When the death of the person is not a means that is intended in order to reach a certain end, we should not say that the person killed intentionally, even though she might be guilty of murder (the killing is still voluntary), and even if such a murder were worse than some intentional killings of the innocent. In AIDE, Anscombe argues that what is true of the intention with which one acts—​ intention of the end—​is not true of the action’s intentionalness, meaning by this fact that the action which is a means to the end is intentional.22 In a sense, that guards us from the first, gross abuse of the DDE, which would restrict the intention of the agent to the intention of the end. But Anscombe is now arguing that an action could be intentional under a description mentioning an aspect that is not properly intended (e.g., killing the fat man), even as a means, in the sense that the purpose of the means is not to bring about that aspect. One would have thought that the means being chosen, its achievement, which is needed for or useful to the end, was the object of a subordinate intention, bearing on the result that is represented as useful for the end. But Anscombe says that the intentionalness of the action does not depend only on the intention the agent ‘has in mind’; it depends also on the ‘facts of the matter.’ She says: Circumstances, and the immediate facts about the means you are choosing to your ends, dictate what descriptions of your intention you must admit. (AIDE, 222)

So it is not only that one cannot deny one’s intending the action that is a means to one’s end: neither can one deny that one’s intention is in part determined by the ‘immediate facts about the means.’ While the intention with which is only in the agent’s eye, or first person, the intentionalness of the action also depends on third-​person factors, and so 21  The context of ‘to intend’ or of ‘intention that’ is intentional, opaque: two coextensive descriptions cannot necessarily be substituted one for the other salva veritate; the identity of reference is not sufficient, an identity of sense would be needed. 22  She first introduces the distinction between intentionalness of an action and its purpose, saying, ‘[I]‌t is hardly noticed that intention may relate to the intentionalness of the particular act that is done, as well as to the purpose for which it is done,’ and it is Cartesian prejudice to confuse them. But there are two kinds of purposes, that of the end and that of the means. Of course, an action may receive a moral evaluation distinct from the evaluation of its end (what it is done for), in virtue of its intentionalness. But this, which characterises the action done for the end, supposes an intention or purpose to achieve what is needed by and useful for the end. So the intentionalness of the action performed as means supposes not only an intention of the end (intention with which the action is performed) but an intention or purpose bearing on the means itself (intention with which simpler elements are brought about as means of the means, e.g., the corporeal movements needed to walk which is a means to go to a meeting). What is now at stake is whether one can separate the purpose of the action as a means from its intentionalness.

Anscombe on Double Effect and Intended Consequences    183 not only on what the agent aims at as a means.23 The aspects of an action that are thus covered by its intentionalness but are not necessarily aimed at are those that are sufficiently close to the intended result.24 But, again, how are we to determine which aspects or results are close enough? The criterion of what is intended as a means versus what is foreseen is clear, and congenial to Anscombe’s philosophy of action and intention. The criterion of distance is not only vague; it seems to be of another kind. Thus John Finnis commented on that text: I think this attempt to distinguish the intended from the unintended by reference to sheer physical ‘immediacy’ of cause and effect is unsound, a confusion of categories, an elision of human behaviour with human action. I know of no argument that Anscombe has brought against her own analysis, twenty-​five years earlier in her book Intention, of the intentions of the man who pumps poisoned water into a house.25

Finnis opposes Anscombe to herself, but he also criticises severely the present account according to which consequences can be intended though not under the description under which they can be considered as means or as useful to the agent.26 Finnis might have a point in maintaining the distinction between what is and what is not useful for the end, and so between what is and what is not aimed at as a means, even for close effects, but Anscombe could also be right if that distinction had no moral impact. In order to defend the DDE, she nonetheless needs to defend a distinction between two kinds of consequences that are not intended as means: those that are merely foreseen and those she says are still intended even if not as ‘part of the aim’. The distance/​closeness criterion is too vague, but it is completed with the distinction between what is essential to an action, by opposition to what is accidental to it, and due to the circumstances. Applied to consequences, the idea is that some circumstances depend on the nature of the action only, others on the action plus the circumstances. One might thus distinguish Pot-​Holer and Close Rock, arguing that it is a circumstance of the action of moving a rock that a man is under the rock, but not a circumstance of blowing up the man that the man is under the dynamite. That is why a consequence might be certain and physically close to the action, but nonetheless foreseen and not intended: it depends on the circumstances. The distance would then make a difference between Close Rock 23  The current debate over the first-​person and the third-​person accounts of intention opposes the so-​ called New Natural Law Theorists (Boyle, Finnis, Grisez, Tollefsen) to those who would call themselves more Aristotelian or Thomist, like O’Brien, Frey, and Flannery. In fact, it divides (not surprisingly) also interpreters of Aquinas. 24  See Gormally, ‘Intention and Side-​Effect,’ and the response by J. Finnis, in the same volume (see note 12), ‘Reflections and Responses’, 480–​485. 25  J. Finnis, ‘Intentions and Side-​Effects’, in Collected Essays, vol. 2, ‘Intention and Identity’ (Oxford: Oxford University Press, 2011), 192. 26  He also criticises Anscombe’s assertion that to deny that those consequences close to the action’s result are intended would be a form of direction of intention: it does not depend on one’s choice whether a certain state of affairs is or is not a means to a certain end.

184   Cyrille Michon and Distant Rock, the first being a case of murder but not the second. But what is the criterion of distinction between what is essential and what is accidental to an action?27 And does it deliver the moral verdict expected? Before examining those two points, it seems appropriate to come back to Intention and to the passage referred to by Finnis.

8.4  Narrow and Broad Intention In Intention Anscombe does not mention the DDE, but she clearly grounds the distinction between intended and merely foreseen consequences of an action, opposing what is intentional to what is voluntary. She draws the distinction in §49: Something is voluntary though not intentional if it is the antecedently known concomitant result of one’s intentional action, so that one could have prevented it if one would have given up the action; but it is not intentional: one rejects the question ‘Why?’ in its connexion. (I, 89)

An action is intentional under a description, that is, the agent must not only know the description in question, but the description should also correspond to the ‘Why?’ question asking for a reason for acting. If I saw through a plank, knowing that it is Smith’s, but not because it is Smith’s, then I saw through Smith’s plank voluntarily, but not intentionally. If I do not care about the plank being Smith’s, I reject the question ‘Why are you sawing through Smith’s plank?’ as having no point. But I cannot deny that I knew, accepted, and so voluntarily did.28 The ‘Why?’ question has already been discussed at some length before the passage I’ve quoted, in order to introduce the concept of intention without begging the question by presupposing the concept of a reason for acting. When the answer to the ‘Why?’ question refers to a future state of affairs that is the agent’s end, it gives the intention with which the action is performed (an end, ultimate or intermediate). Though this is only one way to answer the ‘Why?’ question which shows the action to be intentional, Anscombe argues for the conceptual link between the notion of an intentional action

27  Remoteness and uncertainty are indications of distance and of accidentalness. But we have agreed that some consequences could be certain and physically close without being intended. And the distinction essential/​accidental has been viewed as a way to complete the criterion of distance, so we cannot have it rely on this criterion. 28  Not everything that is voluntary must be thus foreknown and accepted: some voluntary movements are not (I, §49). Nor is it obvious that everything that is foreknown in an action or as a consequence of it should be said to be voluntary. Amongst foreknown aspects or consequences of the action, only those that have some salience are apt for the ascription of responsibility. (Thanks to J. Frey for pointing this out to me.)

Anscombe on Double Effect and Intended Consequences    185 and that of the intention with which (I, §20).29 And in any case, actions about which questions concerning DDE are raised are those that are for the sake of some end. They are means to such an end, and correspond thus to the question ‘How?’ asked about the end reached (how did you do X?) or intended (how will you/​I do X?). Any description of the action that mentions its end or presents it as a means to the end is a description under which it is intentional. ‘Why are you pumping the water? To replenish the water supply.’ ‘How did you replenish the water supply? By pumping the water.’ Both descriptions, ‘pumping water’ and ‘replenishing the water supply,’ are descriptions under which the action is intentional. One describes it as a means, the other through its end. We can say that both descriptions or conceptions enter the practical reasoning of the agent who aims at replenishing the water supply by pumping the water. This short summary is in line with those philosophers, like John Finnis, who consider that intention bears only on what are the agent’s purposes, or what belongs to the agent’s plan of action, either as an end or as a means. They are chosen for themselves or for the sake of something else, and they are chosen under the description under which they are considered either as good in themselves or as good instrumentally. All other results or concomitants of the action (the shadow projected on the rock, the noise or the rhythm of the pumping) are not properly objects of intention, even when they are known or foreknown: they do not belong to the plan; they are neither aimed at as an end nor as a means; they are not useful. They are accidental to the intention, and, so it seems, to the action. But if they are known and so accepted by the agent, they can be said to be brought about voluntarily, and one is responsible for them and would need a justification for allowing them or bringing them about. We could formulate the DDE by saying that sometimes (in particular circumstances) something that it is always prohibited to bring about intentionally (e.g., killing an innocent) can be brought about voluntarily. Of course, this is true only sometimes or even rarely. Most often such acts by which we voluntarily bring about a bad consequence are evil acts, and forbidden. The domain of the voluntary is the domain of moral accountability, which is then broader than the domain of the intentional. But, whereas there are no exceptions to absolute prohibitions when the consequence is intended, there might be exceptions or excuses when it is only foreseen and so accepted and voluntarily brought about. Now, that Anscombe adopts such a first-​person approach to the object of intention seems to be confirmed by her analysis of the case where the man at the pump is informed about the poison laid in water. If he has been hired just to pump water, only aims at earning his pay as an end, and at pumping water as the means to that end, then his learning later on that the water is poisoned does not enter his practical reasoning. To the question ‘Why did you replenish the house water-​supply with poisoned water?,’ his

29  The other ways to answer the ‘Why?’ question are the mention of a backward-​looking motive (‘he killed my father’) or of an interpretative motive (‘out of anger’). See I, §16 for a summary. And Teichmann 2008: 44–​46 for a commentary on §20.

186   Cyrille Michon answer is a rejection of the question in the form ‘I didn’t care about that. I wanted my pay and just did my usual job.’ Anscombe comments: In that case, although he knows concerning an intentional act of his—​for it, namely replenishing the house water-​supply, is intentional by our criteria—​that it is also an act of replenishing the house water-​supply with poisoned water, it would be incorrect, by our criteria, to say that his act of replenishing the house supply with poisoned water was intentional. And I do not doubt the correctness of the conclusion; it seems to shew that our criteria are rather good. (I, 42)

Anscombe’s criteria here seem to involve the first-​person account of intention. The intended consequences are those that are aimed at under the descriptions that belong to the agent’s plan or purposes and to which the ‘Why?’ question applies; the other aspects (pumping poisoned water) might be foreknown and voluntarily accepted, but they are not intended.30 One might resist the conclusion that Anscombe’s analysis of this case vindicates the first-​person account of intention. For in the same section (§25), she also considers the case of the man being informed that the water is poisoned before being hired, and being hired, presumably, in order that he pump poisoned water.31 The man could not give the same answer as before. But Anscombe goes farther: even if the man’s end was still only to earn his pay, and he did not care about poisoning the house, nonetheless the means to earn his pay would be, in that case, to pump poisoned water and so to poison the house: [T]‌he commission by the acceptance and performance of which he gets the money is—​however implicit this is allowed to be—​to pump poisoned water. (I, 44)

In such a case, according to Anscombe, we cannot accept a denial (either from the agent or from an observer) that the action was intentional under the description ‘pumping poisoned water.’ Here there is a constraint, not only on the truthfulness of what the agent can say, but on what he really intends. So that while we can find cases where ‘only the man himself can say whether he had a certain intention or not’; they are further limited by this: he cannot profess not to have had the intention of doing the thing that was a means to an end of his. (I, 44)

The limits are constraints on the intention. It is true that the agent’s personal purpose does not include poisoning the water supply, and so it is accidental to his purpose that 30 

But Anscombe says, ‘[I]‌f what he said was true, that will not absolve him from guilt of murder!’ So, though the core notion of murder is that of intentional killing, a murder can be just a voluntary non-​ intentional killing. See note 37. 31  Between the two cases there is a long discussion, prompted by the first case, about the criterial basis for the attribution of the intention: should we go by the agent’s thoughts (and by what the agent says), or are there some external indications that the agent had such an intention?

Anscombe on Double Effect and Intended Consequences    187 the water is poisoned. But it is not accidental to the bigger plan to which he commits himself: the plan of the one who hired him. This plan includes explicitly the pumping of poisoned water, and since the man knows it, and is hired in order to realise this plan, we could say that he adopts the hirer’s plan, purposes, and intention. We might also say that he performs an action that is a means to two ends: his personal end of earning his pay, and the hirer’s end to poison the water supply. He has adopted the hirer’s end, so he intentionally pumps poisoned water. I suggest, then, that we introduce a distinction between the agent’s proper intention and the intention he accepts though it is not primarily his, in terms of narrow and broad intention, or scopes of intention. The broad intention in the example is in fact another agent’s proper (or narrow) intention, which becomes the first agent’s own (broad intention) by way of the hiring, and of his accepting the terms of the hiring. The important point is that a broadly intended consequence is not only foreseen. Though it remains accidental to the narrow intention of the agent, there is a point in saying that it is not accidental to the action but essential to it. And the reason is that it is the object of the broad intention which determines the whole nature of the action. We might have found here a way to vindicate Anscombe’s view in response to Finnis’s worry. We now have a criterion for saying that a particular consequence is essential to the action though not part of the first-​person account of the agent’s intention or purpose. In the present case, the criterion is given by the existence of another intention (the hirer’s) that determines the nature of the action that the agent performs knowingly. But this is a very particular case, of subordinate action, an agent’s entering the plan of another agent. It remains to be seen whether we can find a similar way of attaching some consequences essentially to an action, when the consequence is not narrowly intended, for more ordinary cases, without subordination. And it also remains to be seen how we should understand the doctrine of double effect if we accept that there is a sense of intending and intention (i.e., broad intention) that goes further than the agent’s proper purposes. Should the doctrine be understood on the basis of the opposition between mere foresight and intention (narrow and broad) or between narrow intention and what is not narrowly intended (merely foreseen or broadly intended)? I take each question in turn in the two following sections.

8.5  Broad Intentions: Social Practices, Natural Effects, and Constitution Apart from actions that are subordinate to another agent’s intention, how could we determine that some consequences are essential to an action, and so broadly intended? It should be noticed first that a sort of subordination of one’s action to an intention or teleology embedded in the action is at play when the action is an instance of a social practice, recognised as such by the agent. (In fact, hiring someone is a practice and implies

188   Cyrille Michon that the person adopts the explicit end of the job for which he or she is hired.) If I buy your car, I intend, in virtue of the practices of buying, selling, and private property, that you cease to have rights on this car, even though this is not something that I am looking for. It might just be a question of helping you, e.g., by paying the insurance policy and letting you use the car as you like. I would authorise you to use it, but you would cease to have juridical rights over it, and this is something that I broadly intend in virtue of the practice I am knowingly participating in.32 Other cases of practices, more at stake in DDE discussions, are medical or military practices. In the example given by Anscombe of an abortive medicine, the narrow intention is not to abort, but to cure the cancer, and the practice embeds only that curative intention. It just happens that the woman is pregnant, so that the medicine is abortive. Hence the abortion is only foreseen, even if certain; it is not intended, even broadly (though of course it is possible for the medicine to be taken with the intention to abort). The aim of the practice, here, confirms that of the agent’s narrow intention. Consider now the case of lethal weapons: one might use a grenade to defend oneself, knowing that it is lethal, but without narrowly intending the victim’s death—​it was the only way to repel the attacker. But it is embedded in the action of throwing a grenade that it will kill the targeted person: grenades are made for that. It is then not accidental to such an act, even performed in a context of self-​defence, that it will kill the targeted enemy: one cannot use such a weapon without at least broadly intending to kill. On Aquinas’s terms, a private person would not be allowed such an act of self-​defence.33 Whereas he would accept the act of repelling the assailant in a staircase, and maybe of repelling him over a cliff, where death is certain. In those cases, the act of repelling would not be in itself an act of killing, if it is accidental to the action that there is a cliff and a great void. The death of the assailant would be merely foreseen even if certain. True, another agent might narrowly intend to kill the assailant by repelling him into the void, thus doing the same thing in one sense, but with a different narrow intention.34 So some actions have an internal teleology independently of the agent’s proper intention, because of the social practices that they exemplify. The embedded teleology or intention can either yield the conclusion that, though certain, the consequence was not broadly intended but only accidental (abortion as a consequence of taking a cancer medicine), or it can yield the conclusion on the contrary that the consequence was broadly intended and not accidental (use of a lethal weapon). Other actions (e.g., 32 

In Anscombe’s terminology we could say that the fact of the agent’s proper intention and action is brute relative to the fact of her action as exemplifying a practice. Cf. BF. 33  At least according to a plausible (and strict) interpretation of Aquinas. 34  Matthew O’Brien has emphasised the importance of social practices in his account of intention in his dissertation, ‘Practical Necessity: A Study in Ethics, Law and Human Action’ (available online) and in two papers: ‘Elizabeth Anscombe and the New Natural Lawyers on Intentional Action,’ National Catholic Bioethics Quarterly 1 (2013), 47–​56, and with R. Koons, ‘Objects of Intention: A Critique of New Natural Law Theory’, American Catholic Philosophical Quarterly 86, no. 4 (2012), 655–​703. See the criticisms by C. Tollefsen, ‘Response to Robert Koons and Matthew O’Brien “Objects of Intention: A Critique of New Natural Law Theory,”’ American Catholic Philosophical Quarterly 87, no. 4 (2013), 751–​778.

Anscombe on Double Effect and Intended Consequences    189 Cliff) do not exemplify a practice, and one cannot attribute a broad intention to bring about a consequence that is not narrowly intended merely on the grounds that the consequence is certain. Problematic cases are those for which the action does not exemplify an existing practice with a clear teleology, and yet the result seems too close to count as merely foreseen. So it is with Pot-​Holer, which is analogous to a therapeutic abortion by craniotomy,35 and to Area Bombing. It is clear that Terrorist Bombing aims at killing civilians: that is the object of the agent’s intention. And it is clear that Strategic Bombing may have civilian deaths only as side effects (they are neither narrowly nor broadly intended). But one can bomb a city without intending narrowly to kill the civilians, as with Area Bombing, where one intends to destroy military targets not caring about the citizens, just as when an Arsonist burns down a house without narrowly intending to kill the inhabitants. And Anscombe seems to claim that in such cases the homicide is intentional, and is murder.36 But without reference to a higher intention or to a social practice, how to draw the line between what is broadly intended and what is merely foreseen? To answer that the death is not accidental but essential to the action is not an explanation; it is just another formulation of the idea of broad intent. And to draw the distinction between essential and accidental as a distinction between epistemic or causal distance and immediacy cannot work. Anscombe herself admits of cases where the agent is certain about the foreseen consequence and does nonetheless not intend it, as shutting doors in a sinking submarine when people are on the other side. As for the causal relation, Finnis is right when he says that nothing is closer to the action of pumping water than the consequence of pumping poisoned water, if the water is poisoned. Turning to the Cave cases: not only is there no principled distinction between Close Rock and Distant Rock, but we can also ask why the closeness between the action and its consequence in Pot-​Holer would be greater than in Close Rock. Intuitively we are reluctant, like Anscombe, to deny that in Pot-​Holer the agent performs an intentional killing. But we agree with Finnis, and maybe also with Anscombe, that the death of the pot-​holer is not ‘part of the aim’ of the agent, and so is not narrowly intended. If the killing is intentional (intentionalness of the action) and the death is broadly intended, and so essential to the action, we need a criterion other than epistemic or causal closeness. The embedded intention or teleology coming from another’s plan or from a social practice can serve as a model for thinking of a natural equivalent. We could say that the

35  The Catholic magisterium authorised some forms of therapeutic abortion, such as hysterectomy in the case of uterine cancer, because the practice exists whether or not the uterus is diseased, but it condemned the craniotomy (for an extra-​uterine pregnancy), because the action-​type could not be performed without killing a baby. 36  It is intentional in the sense of the intentionalness of the action, because Anscombe is clear that there is no purpose to kill, neither as an end nor as a means: ‘The arsonist who burns down a house, not caring that there are people there, is as much a murderer if they are burned to death by his action, as if he had aimed to kill them. This action falls squarely within a penumbra surrounding the hard-​core part of the concept of murder, which contains only intentional killing. The penumbra is fuzzy at the outer edges—​that is, there are borderline cases. But that fact does not mean that an absolute prohibition on murder makes no sense’ (AIDE, 249).

190   Cyrille Michon intentional killing, in the case of the man at the pump hired to pump poisoned water, is constituted by his accepting the job, or supervenes on his doing the job of pumping the water. In a similar way, my buying your car constitutes the loss of your rights over it, which supervenes on the selling, and my intentionally depriving you of your rights likewise supervenes on that selling. The relation of constitution (and supervenience) is the product of the narrowly intended action (under the description that is the proper aim of the agent) and of the existence of a higher intention or of a social practice: their conjunction entails the broadly intended action.37 These are cases of intentional or conventional constitution. But we can also think of relations of natural supervenience and natural constitution, as opposed to mere causation. Let us consider lethal weapons: the killing of the attacker is constituted by the act of repelling him when it is done with a grenade, but caused by the act of repelling him only when the attacker is pushed down the staircase. True, in Grenade we refer to the embedded teleology of lethal weapons meant for killing. And so we still refer to a relation of intentional or conventional constitution. But in Pot-​Holer the dynamite is not a weapon, and we cannot refer to the practice of blasting people, nor to a higher plan which would include explicitly the death of the pot-​holer. But we might argue that the act of blowing the man up in Pot-​Holer naturally constitutes his death, while the moving of the rock only causes it. We are speaking here of natural, not conventional, constitution; with that in mind, I am borrowing from William FitzPatrick the opposition between mere causation and constitution which I understand as follows:38 an effect is constituted by a state of affairs if the mere description of this state of affairs allows us to infer, on the basis of our knowledge of the laws of nature, the production of the effect in question. It is merely caused by it if, in addition to the state of affairs in question and the laws of nature, other aspects of the situation, the circumstances, must be mentioned in order to infer its production. In this sense, the blasting of the fat man constitutes the killing of him, while moving the rock only causes it. Similarly, an area bombing, even if intended to destroy only military objectives, constitutes the destruction of civilian lives, whereas the strategic bombing of a military objective only causes it, because this consequence depends on the circumstance that civilians are in the vicinity of the military objective (while the presence of civilians in a city is not a mere circumstance). We see then that the relationship of constitution is not that of immediacy or physical closeness (between the intended effect and the foreseen effect) but that it depends on what is considered as a circumstance. But why should we say that the man’s situation close to the rock is a circumstance in Rock while his situation stuck in the exit is not one in Pot-​Holer? Of course, the man’s situation is accidental to the intention to move the rock, and it is essential to the intention to blast the man. But what about the description ‘blasting the body that is an obstacle to escape’? Is it not then accidental and a circumstance that the obstacle happens to be a 37  Once again, the action considered only as determined by the agent’s narrow intention is brute relative to the action considered as determined by the higher intention or social practice. 38  See in particular William FitzPatrick, ‘The Intentional/​Foreseen Distinction and the Problem of “Closeness,” ’ Philosophical Studies 128 (2006), 585–​617.

Anscombe on Double Effect and Intended Consequences    191 man? And what about Rope Climber: that the man will fall into the void is accidental to the cutting of the rope, but it is essential to the cutting of the rope over the void. We could either say that the rope climber’s fall to his death is constituted by the cutting of the rope over the void and the laws of nature, or that it is merely caused by the cutting of the rope, since one must add to the description of the action the circumstance that the rope climber is over the void. The distinctions between what is essential and what is accidental, between constitution and mere causation, and the notion of a circumstance of the action (or of the intention) seem to depend on the description under which the action is presented. I do not think this relativity to descriptions is fatal to the theory. The reason is that we are trying to ground a distinction between broad intention and mere foresight that is relative to agents. We are then more interested by what the agent sees as a circumstance of her action than by the objectivity of the circumstance. Such a relativity to the agent’s point of view does not count as arbitrary because the agent chooses neither the description under which her action is intentional nor the descriptions under which she knows what action she is performing and with what consequences. What is important is whether or not the agent sees herself as the only cause of the result.39 It is quite obvious that in Pot-​Holer the spelunker sees himself as the only cause (and his action as constituting the fat man’s killing), and possible that in Distant Rock he does not, and sees the presence of the man under the rock as a circumstance (and so his action as merely causing the death). In the cases of Close Rock and Rope Climber the question is debatable. But the fact that the distinction between essence and circumstance is fuzzy is not a problem. As Samuel Johnson, quoted by Anscombe, says, twilight is no objection to the distinction between day and night. So the metaphysical distinction between broad intention and mere foresight can be maintained and explained by the distinction between an effect that is constituted by a narrowly intended action (broad intention) and an effect that is merely caused by a narrowly intended action (mere foresight). But this does not undermine the previous

39 

It has been objected to me, by Anselm Müller and Candace Vogler, that one might be responsible for one’s beliefs and for the way one sees things. I agree that it can be so in cases which are of immediate moral relevance. (I might be responsible and blameworthy for some racist beliefs.) But this is less obvious when the belief bears on what is or is not a circumstance of an action. Now, as Roger Teichmann has pointed out to me, the pumper hired to pump poisoned water might certainly think that he is not the ‘sole cause’ of the poisoning, being only ‘a cog in a machine.’ If he were not blameworthy for this thought, it should come as an excuse, thus undermining the point made earlier, according to which the broad intention to poison the house was imposed on him by the way of ‘conventional constitution’ of the intention. To maintain that verdict, one should argue that such a belief would be blameworthy. I am tempted to follow that line. In my argument, I only suggest that if one could see oneself as not the only cause, at a purely causal level, that belief might make a difference to one’s accountability. But to deny having the broad intention to poison the house in the case of the hired pumper would not be a purely causal belief, but one involving also a judgement of moral responsibility (‘I’m not morally responsible for that’), and it is that moral judgement that bears the possible blameworthiness: it would be due to bad faith or to another fault of character. So, in a derivative way, one might be blameworthy for one’s belief about one’s causal role in a certain upshot.

192   Cyrille Michon distinction between narrow and broad intention. Both are good metaphysical (conceptual) distinctions and can be used in tandem. What remains to be seen is if and how the metaphysical distinctions ground moral distinctions.

8.6  Double Effect: The Metaphysical and the Moral Distinctions According to the preceding analysis, if the foreseen effect (e.g., death) is only caused by the intended action (blasting of the rock), it is because circumstances known to the agent (presence of the man), but for which she is not responsible, causally contribute with her action to produce the effect in question. The agent therefore does not see herself as the only cause responsible for this effect. She is certainly causally responsible for it, but only partially. This may of course be sufficient for moral responsibility, but it is also understandable that it contributes to diminish moral responsibility, or that it even leads to exemption in some circumstances. To be sure, the agent does not share her causal responsibility with another moral agent, but with factual circumstances. It may nevertheless be sufficient to say that the agent is less involved in producing an effect only caused by her than in producing an effect constituted by her intentional action. The agent could always hope that an obstacle might interfere between the intended result and the foreseen one, and she could regret not having within her reach the means to prevent it (the means of blocking the rock, alerting the civilians in the targeted factory, etc.), for the intended and foreseen results are naturally separable. The difference in commitment for the agent, between the intended outcome and the merely foreseen one, provides the basis of a difference in moral evaluation. And a greater distance between action and outcome brings more circumstances into play and thus further reduces the agent’s own causal share. On the other hand, if the effect is constituted by the action, then the agent sees herself as the only cause of that effect, as having to bear full causal responsibility, which necessarily puts her moral responsibility at a level at least as high as if she were only a partial cause. As Anscombe noted, the spelunkers might feel they would be murderers if they blew up the fat man, but not if they blew up the rock at some distance even though they were certain it would kill the man (and they might be ambivalent if the rock were close to the man). The metaphysical distinction between an accidental consequence that is merely foreseen and an essential one that is broadly intended can ground a moral difference in moral responsibility. Anscombe is thus vindicated. Let us now reconsider the first distinction between narrowly intended and broadly intended consequences, the latter being understood as the consequences constituted by the narrowly intended ones. Pot-​Holer illustrates the case of a broadly intended death, since the man’s being blown to pieces (narrowly intended state of affairs) constitutes his dying (broadly intended one). What would a narrowly intended death and narrowly intentional (instrumental) killing be in a situation otherwise similar? Let Nazis be the

Anscombe on Double Effect and Intended Consequences    193 following one: the spelunkers are in the cave, but cannot exit because a group of Nazi terrorists threatens them with lethal weapons. They tell the spelunkers that they will be able to escape only if they kill the fat man. In that case, the killing would be instrumental. They would have to narrowly intend his death and kill the fat man in order to exit. If they did not, even if they shot at him and by a miracle he survived, they would not escape and they would be killed by the Nazis. Is it not plausible that a speleologist might agree to blow up his companion to free the exit but not agree to kill him under the threat of a terrorist? He would thereby show that he sees the killing in Nazis as going against the prohibition of murder, but not in Pot-​Holer, or at least that he sees the first killing as more abhorrent and harder to justify than the second one. The metaphysical difference between narrow and broad intention seems then to form the basis for a moral difference, and the reason is that the agent’s commitment is less in the production of a broadly intended consequence than in the production of a narrowly intended one. This may be questioned, and it may be argued that there is no shared causality this time, to mark a lesser commitment with potentially lesser responsibility. The agent sees herself as the only cause of the effect she causes, both if it is the narrowly intended result of her action and if it is an effect constituted by this intended result, and so broadly intended. The two this time are inseparable. Perhaps the agent could consider that it is the laws of nature, or nature itself, or God that contributes with her to produce the bad effect. That if she had supernatural powers, she could prevent it without giving up her plan for the narrowly intended result. The argument may be tenuous. It seems to me, however, that it is not a bad one. But here is another way to highlight the contrast: when foreseen effects (not narrowly intended) are at stake, the difference between certainty and uncertainty, the epistemic distance, makes a difference; it might be permitted to take the risk of killing someone (without intending it) where it would be forbidden to act with certainty that a death would result. But in the case of a narrowly intended effect, the uncertainty of its production does not modify the moral evaluation of the action. A criminal shot will be just as criminal and blameworthy when performed by a bad gunman as when it is by a good one, even if its chance of hitting the target is lower. In other words, the certainty test makes no moral difference to the assessment of the (narrowly) intended effect, whereas it makes a moral difference to the assessment of the foreseen and even broadly intended effect. It probably is the nature of the agent’s commitment that explains this asymmetry. When the effect is certain, the agent is the only responsible cause, whether the effect is narrowly or broadly intended (and constituted by the intended effect), but the narrow intention increases her commitment and therefore her moral responsibility. Finnis is thus vindicated. In defending the relevance of two distinctions and three categories—​narrowly intended, broadly intended, and merely foreseen consequences—​I have distinguished my position from Anscombe’s, but also from Finnis’s, while agreeing with them in opposing the consequentialism that denies these distinctions. The pairing of the two extremes (narrowly intended vs. merely foreseen) constitutes the paradigm of acts with double effect, such as the pair of strategic and terrorist bombardments. In such cases we have an opposition between intended (in the narrow sense) and foreseen, but what

194   Cyrille Michon is foreseen is also caused only by what is intended and depends on circumstances that are additional to the intended effect. Anscombe considers that the distance created by the circumstances is crucial and that without it the DDE would be sophistic. And she assimilates narrow and broad intention in opposition to foresight. Finnis considers that it is the narrow intention that is crucial, and he assimilates broad intention with foresight. I agreed with Finnis against Anscombe: the physical closeness does not obliterate the conceptual (metaphysical) distinction between narrow and broad intention. But I followed Anscombe, against Finnis, by giving importance to the conventional and to physical distance when these are understood in terms of constitution vs. mere causation. And I argued that both conceptual distinctions could be the basis for a moral difference. A possible explanation of Anscombe’s neglecting the distinction between narrow and broad intention might be rooted in her moral absolutism. The main interest of the doctrine of double effect was to explain why acts with effects that it would be absolutely prohibited to cause intentionally could sometimes be permitted. Anscombe admits that a killing that is not intentional (neither narrowly nor broadly) but only voluntary can be, and is often, murder. But, following the DDE, she considers that it is not always so, and this is why the distinction between intended and foreseen is important. She might admit a conceptual difference between narrow and broad intention, but that distinction, according to her, could not allow us in any case to cross the border of an absolute prohibition: area bombing is always forbidden, just as terror bombing is; therapeutic abortion by craniotomy is always forbidden, just as non-​therapeutic abortion is; it is forbidden to cause the fat man’s death in Pot-​Holer, just as it is in Nazis. Even if she is right, this moral stance on absolute prohibition is compatible with a difference in degrees of responsibility, based on the difference in causal involvement of the agent.40 And one could be an absolutist and disagree with Anscombe on which distinction is fundamental for the DDE (or the PSE). Thus, the New Natural Law Theorists, with John Finnis, seem to ground the DDE and the limits of absolute principles on the distinction between narrow intention and foresight (including broad intention), allowing all forms of therapeutic abortions if any is allowed, or area bombing when performed with a purely strategic intent (but not nuclear deterrence, which presupposes the intention of killing civilians). What I have argued is that they should accept a moral distinction between bad consequences that are merely foreseen and those that are broadly intended. One can rank the different commitments in the bringing about of a bad consequence, with their potentially differing levels of responsibility, thus: (i) Narrowly intended as an end (sadistic murder or bombing). (ii) Narrowly intended as a means (terrorist bombing, Nazis). (iii) Broadly intended, and constituted by a narrowly intended consequence (area bombing, pot-​holer, craniotomy). 40  But it might very well be, as pointed out to me by Candace Vogler, that Anscombe does not really make room for those degrees of responsibility: her point is to draw the line of permissibility and impermissibility; the rest might be reserved to God’s eye.

Anscombe on Double Effect and Intended Consequences    195 (iv) Merely foreseen, and merely caused by a narrowly intended consequence, but certain (strategic bombing, cliff, rope climber, rock, hysterectomy)—​with degrees. (v) Merely foreseen and uncertain (staircase)—​with degrees. (vi) Unforeseen effects that might be due to negligence (shooting error)—​with degrees. It is one thing to accept the conceptual and moral distinctions, and another to say where the line is separating the always prohibited from the sometimes permitted. Anscombe draws the line of absolute prohibitions between (iii) and (iv), Finnis and the New Natural Law Theorists between (ii) and (iii). A more laxist view on absolute prohibitions (consequentialist) would draw it between (i) and (ii), and a very strict one (maybe Aquinas on self-​defence) between (iv) and (v). Although this analysis seems to depart from Anscombe’s remarks on the doctrine of double effect, it relies largely on her philosophy of action. As such, it seems to me that it owes her a lot, and remains faithful to her intention.41

41 

I wish to thank Roger Teichmann for many helpful comments and suggestions on previous versions of this chapter.

Chapter 9

Ansc ombe on Ou ght Anselm Winfried Müller

9.1  Introduction Anscombe’s essay ‘Modern Moral Philosophy’ (MMP, in CP3, 26–​42) is notorious for rejecting ‘the notion ‘morally ought’ [ . . . ] It would be most reasonable to drop it. It has no reasonable sense outside a law conception of ethics’ (MMP: 32). Does she have good reasons for this rejection? What, if anything, does she think should take the place of talk about ‘moral obligation’? Or can one justify it independently of belief in a divine lawgiver? To answer these (and some other) questions, I’ll proceed by examining various alternative ways of understanding ought that Anscombe’s work suggests, beginning (in section 9.2) with the ‘very fishy’ law conception she attacks (MMP, 32), and finishing (in section 9.6) with a Natural Law account in the spirit of Aquinas. On the way, I am going to consider the chances of understanding moral requirements in terms of Aristotelian necessity, ultimate purpose, and dictate of conscience (sections 9.3–​9.5). There is no space to discuss those parts of Anscombe’s work that explain how ‘rules, rights, and promises’ give rise to an ought by relying on conventions without which human life would be impossible or ghastly.1 I’ll only observe here that, surely, the same sort of overall practical necessity attaches to one’s justice in matters of rights or promises as does to one’s kindness, say, or one’s readiness to help. Hence Anscombe’s brilliant observations on the very special constitution of a promise to Φ as a reason to Φ can have no bearing on the question: What is the moral quality of a practical ought that goes, not only with the Aristotelian necessity of keeping promises, but equally with all other requirements of good conduct? 1 

The relevant essays are BF, in CP3, 22–​25; PJ, in CP3, 10–​21; SAS, in CP3, 130–​155; RRP, in CP3, 97–​ 103; QLI, in CP1, 112–​133.

Anscombe on Ought   197

9.2  ‘Mesmeric Force’—​O ught by Chimerical Law MMP is deprecatory. With respect to our topic it complains that law conceptions of ethics, as proposed by her colleagues, are untenable because no credible lawgiver is being assumed (section 9.2.1), and that accordingly these philosophers use expressions such as ‘morally ought/​obliged’ without meaning (section 9.2.2). I’ll briefly clarify these points, leaving for later the question whether Anscombe is right, for the reasons she advances, to ban the language of law, obligation, and ‘moral ought’ from ethics.

9.2.1  Law without Legislator Of the famous three theses defended in ‘Modern Moral Philosophy,’ the second says that the concepts of obligation, and duty—​moral obligation and moral duty, that is to say—​and of what is morally right and wrong, and of the moral sense of ‘ought,’ ought to be jettisoned if this is psychologically possible; because they are survivals, or derivatives from survivals, from an earlier conception of ethics which no longer generally survives, and are only harmful without it. (MMP, 26)

This earlier conception is ‘the divine law conception of ethics,’ which holds that what is needed for conformity with the virtues failure in which is the mark of being bad qua man (and not merely, say, qua craftsman or logician)—​that what is needed for this, is required by divine law. (MMP, 30)

If you reject the idea of God as legislator, can you nevertheless retain a law conception of ethics? Anscombe briefly examines candidates for alternative sources of legislation. The dictates of conscience as well as ‘the ‘norms’ of a society’ are ruled out because they will often allow or even decree atrocities (MMP, 37). The Kantian idea of ‘legislating for oneself ’ is incoherent and ‘absurd,’ inter alia, because ‘the concept of legislation requires superior power in the legislator’ (MMP, 27; cf. 37). Deference to the universe ‘might lead one to eat the weaker according to the laws of nature.’ Contract is given a chance; but ‘you cannot be in a contract without having contracted’ (MMP, 37). It is more plausible to ‘look for ‘norms’ in human virtues’; but here ‘‘norm’ has ceased to be roughly equivalent to ‘law,’’ and we are ‘nearer to an Aristotelian than a law conception of ethics’ (MMP, 38).

198   Anselm Winfried Müller

9.2.2  Norm without Content The philosophers who treat moral norms as laws without a lawgiver have deprived themselves of the means to determine their content. On the one hand, they have abandoned the idea that this content can be known because the norms are promulgated as laws by proper authority. On the other, since law is as such a purely formal notion, it does not itself supply us with a criterion by which to tell a valid moral norm from a merely apparent one. And they cannot allow facts to provide us with relevant criteria. In consequence, the question whether one ought, tout court, or morally, to act a certain way has lost its meaning. For there is no criterion by which to answer it after the notion of the moral has been severed from its reference to the ‘human passions’ (MMP, 29; cf. 26–​27). We cannot, then, explain the ‘moral ought’ in terms of the ethical virtues. Instead it is now ‘equated [ . . . ] with ‘is obliged,’ or ‘is bound,’ or ‘is required to,’ in the sense in which one can be obliged or bound by law, or something can be required by law’ (MMP, 29–​30) Hence, absent the assumption of a lawgiver, that ought has ‘become a word of mere mesmeric force,’ ‘containing no intelligible thought: a word retaining the suggestion of force, and apt to have a strong psychological effect, but which no longer signifies a real concept at all’ (MMP, 32). Anscombe herself of course accepts a divine law conception of ethics. But her advice to those who don’t is to remember that ‘you can do ethics without [the expression ‘morally ought’], as is shown by the example of Aristotle’ (MMP, 32–​33). We’ll examine this suggestion in the next section.

9.3  Requirements of Nature—​O ught by Natural (Aristotelian) Necessity The second thesis of MMP leaves us with the question: What account is to be given of the word ought that we actually use in judgements that we call moral? What can be said about the grounds on which a way of behaving can be said to be what one ought, or ought not, to do? Anscombe answers this question by pointing to two patterns of connection by which that use of ought is tied to content: a pattern of constitution (section 9.3.1) and a pattern of requirement (section 9.3.2). In virtue of such a pattern the ‘moral ought’ turns out to be an ordinary kind of ought: for something to be a way one ‘morally ought to’ act is nothing other than for it to be required for a specific kind of goodness. How required, and for what kind of goodness? Anscombe answers this question in more than one way (sections 9.3.3–​9.3.6). But unfortunately she also gives us reason to call these answers into question (section 9.3.7).

Anscombe on Ought   199

9.3.1  More and Less Brute: Relations of Constitution The pseudo-​sense of ought that Anscombe attacks goes with non-​descriptivism and the Humean denial that ought to be can be derived from is (MMP, 31). To realise the error of these views we have to study ‘the relation between brute and less brute facts’ (MMP, 31). This is a relation of constitution, explained more fully elsewhere.2 Here it will suffice to rehearse the example she uses in MMP: A grocer ‘had potatoes carted to my house and they were left there.’ These ‘are brute facts relative to ‘he supplied me with potatoes’’ (MMP, 28). But the fact thus constituted is brute in relation to further facts. Thus, in combination with the grocer’s having sent me a bill, his having supplied me with potatoes amounts to this: ‘I owe the grocer such-​and-​such a sum.’ This in turn ‘would be one of a set of facts which would be ‘brute’ in relation to the description ‘I am a bilker.’ Bilking is of course a species of ‘dishonesty’ or ‘injustice.’’ Hence one can ‘conceive ‘bilking,’ ‘injustice’ and ‘dishonesty’ in a merely factual way’ (MMP, 29). For whatever ultimately constitutes injustice, say, turns out to be ‘merely factual.’ What matters in the present context is this: We are made to see, first, that the practice of a virtue such as justice, or the vice opposed to it, ultimately consists in doings that can be ascribed on the basis of uncontroversial factual evidence; second, that from such ascriptions one can, ceteris paribus but equally uncontroversially, derive ascriptions of good and bad conduct. One might think that this already gets us to an acceptable understanding of the ‘moral ought.’ Isn’t acting justly/​unjustly brute in relation to acting in a way one ought/​ought not to act? Can’t we, therefore, go from ‘I did what in the circumstances constituted injustice’ to ‘I ought not to have done that’? We can. Ought has legitimately been derived from is. But the derivation involves a step—​viz. the last one—​that needs separate explanation and justification. In what sense does just conduct constitute a way one ought to act? What does the ought in question mean? And what is the rationale of applying it to requirements of virtue?

9.3.2  Aristotelian Necessity: Relations of Requirement To answer these questions we have to attend to the actual use of the ‘ordinary (and quite indispensable) terms ‘should,’ ‘needs,’ ‘ought,’ ‘must.’’ They ‘relate to good and bad: e.g. machinery needs oil, or should or ought to be oiled, in that running without oil is bad for it, or it runs badly without oil. According to this conception, of course, ‘should’ and ‘ought’ are not used in a special ‘moral’ sense when one says that a man should not bilk.’ In saying such a thing, one is using ‘should,’ or ‘ought to,’ in its ordinary sense, applying it 2  Cf. BF, where Anscombe discusses details, including defeasibility of the implication by which a less brute type of fact is constitutive of a more brute one. Cf. also ‘The Moral Environment of the Child’ (METC), in GG2, 224–​233.

200   Anselm Winfried Müller to a particular kind of topic, ‘a moral subject-​matter: namely that of human passions and (non-​technical) actions’’ (MMP, 29). What both man and machinery need to/​ought to/​should have or do is something necessary for them—​necessary in the sense of ‘that without which some good will not be attained or some evil avoided.’3 This relation between what is needed by a certain kind of thing and the thing itself, or between how something ought to operate and what good is served by such operation, has come to be called Aristotelian necessity. It is by Aristotelian necessity that a man ought, i.e., needs, to practice justice and the other virtues—​much as a machine ought/​needs to be oiled, and a plant ought/​needs to get water. So much for Anscombe’s explanation of ought as applied to human conduct. It is the ordinary ought. One’s behaviour ought to constitute the practice of virtue because the practice of virtue is necessary, a human need. And if Φ-​ing is a case of bilking and hence unjust, it follows that one ought not to Φ, because it is necessary not to act unjustly.4 But necessary, or needed, for what? What is the good that cannot be attained without virtue, what the bad that cannot be avoided by acting unjustly, cowardly, intemperately, etc.? Why, e.g., should you refrain from bilking: to get the grocer to supply potatoes next time as well? What do you need the practice of justice for: to get on well in your society? or to contribute to its functioning? Do you need virtue in order to be a human being? to be a happy one? or a good one?

9.3.3  Virtues of the Good Man The last answer is implied when Anscombe tells us that what you ‘morally ought’ not to do by way of ‘a thought or a consented-​to passion or an action or an omission in thought or action, is something contrary to one of the virtues the lack of which shows a man to be bad qua man (and not merely, say, qua craftsman or logician)’ (MMP, 30). This type of virtue—​I’ll write virtue for it—​will no doubt be more or less coextensive with the virtues commonly so called. But the quotation specifies it only as a quality that one needs in order to be a good man. It suggests various lines of enquiry. To begin with, we shall want to know in what sense one cannot be a good man without being virtuous. Does it mean a conceptual connection, e.g.: to be a good man is to have a good will, and the virtues are by definition (the) virtues of the will? In this case, their ‘necessity’ seems to be a logical truth rather than an Aristotelian requirement. Alternatively, we may conceive of virtue as serving, and even constituting, the 3 

RRP, 100. Cf. similar passages in CP3: 9, 15, 18–​20, 139. For Aristotle’s explanation of this meaning of necessity, cf. Metaphysics 1015a22–​23. 4 The general moral requirement to act well does not always translate into here-​and-​now necessities. This is important in itself (e.g., as an argument against utilitarianism) but of no relevance to present concerns. So I’ll use the phrase ‘ought to Φ’ wherever Φ-​ing constitutes virtuous conduct.

Anscombe on Ought   201 goodness of a good man, but not connect the concepts logically. On this understanding one cannot be a good man without being virtuous in the same way as nothing can be a rock without consisting of atoms. If being a good man is taken to require virtue in this sort of way, there will be a genuine question how the virtues serve, viz. constitute, a man’s goodness.5 But perhaps this careful differentiation between alternative readings of Anscombe’s remark is really a fatal simplification? It is indeed the case that oil can be said to be required for a machine by Aristotelian necessity only because being oiled does not come into the notion of the good running of the machine that is served by it (cf. MMP, 29). But consider a different analogy: the concept of a kind of organism may refer us to concepts of its organs and functions, but because of the mutual teleologies that connect these with each other and with the whole organism, we can still distinguish between the various ways in which they serve the life of the whole. In a similar way, the virtues might serve a man’s goodness, even though this goodness cannot be identified independently of the constituent virtues. In any case, however, two questions remain. First: Which virtues are virtues? How are we to settle whether, in order to be a good man, one needs to be, e.g., just? And, second, the question to be considered in the next subsection: Why should I aim at being a good man? Is this goodness in turn required by Aristotelian necessity for some more ‘architectonic’ or final good, a ‘human telos,’ as we might call it?6

9.3.4  Necessary for an Individual Human Telos Anscombe seems to run these two questions together when she says that one can ‘reasonably’ ask ‘whether one might ever need to commit injustice’ (MMP, 41). But she does not ignore the distinction. For, on the one hand, she admits that ‘in present-​day philosophy an explanation is required how an unjust man is a bad man, or an unjust action a bad one [ . . . ] the proof that an unjust man is a bad man would require a positive account of justice as a ‘virtue’’ (MMP, 29). On the other hand—​and most importantly—​a later passage implies that even being a good man need not count as an ultimate telos: one need not identify it with flourishing even if one takes it to be an essential constituent of it: One man—​a philosopher—​may say that since justice is a virtue, and injustice a vice [ . . . ] an act of injustice will tend to make a man bad; and essentially the flourishing of a man qua man consists in his being good (e.g. in virtues) [ . . . ] and even if, as it must be admitted may happen, he flourishes less, or not at all, in inessentials, by avoiding injustice, his life is spoiled in essentials by not avoiding injustice—​so he still needs to perform only just actions. (MMP, 41) 5 

6 

Cf. GG1, 201, 203–​204, 209–​215. Cf. ‘Practical Inference’ (PI), in GG1, 109–​147. I’ll return to this question in section 9.4.

202   Anselm Winfried Müller No doubt the philosopher who is being imagined to hold these views has Anscombe’s sympathy. But she hesitates to agree. Mainly because the concept of human flourishing is ‘doubtful’ (MMP, 41). Given the two open questions I have articulated at the end of section 9.3.3, the idea of replacing the empty ‘moral ought’ by an ordinary ought explained in terms of Aristotelian necessity is exploratory rather than explanatory. Anscombe does not identify any human telos that requires one to act, e.g., justly, courageously, or gratefully. This observation is no objection to the reductive project. The intention of MMP is professedly destructive. Still, what are the prospects of convincingly analysing the moral ought as one case of Aristotelian necessity amongst others? We may discern (1) problems concerning the kind(s) of necessity, or causality, that connect the human telos with being a good man, and being a good man with the virtues; (2) problems raised by the task of specifying the human telos; (3) and the task of settling which virtues are virtues and, more specifically, why it is these—​the virtues of character, or virtues tout court—​that are generally thought to make a man good and enable him to contribute to the common good as well as attain the human telos. These problems must be shown at least to have an accessible solution if we are to be convinced that an account in terms of Aristotelian necessity will provide us with an understanding of ought, as applied to matters of morality. The third of these problems includes the question: How can virtue and being a good man, or acting well, be related teleologically to both a man’s human telos and the welfare of his wider community? This is the topic of the next two subsections.

9.3.5  Necessary for the Common Good A decade after MMP, Anscombe’s almost equally famous paper ‘On Promising’ relates the Aristotelian necessity of acting well ‘to the common good’ (PJ, 19). You need to act justly etc. not only to be a good man but also in order for your community to do well: for grocers to be assured that customers will pay their bills, for people not to be prevented from exercising their rights, for important joint projects to succeed, etc. Good conduct has a collective as well as an individual telos. We may note that the two necessities here distinguished differ from each other in two important ways. First, there is the difference between a telos whose achievement can be predicated of the acting individual and a telos realised in a state of affairs that only accidentally involves the agent. And second, that individual telos seems to be, or essentially involve, the agent’s good character, whereas achievement of the collective telos consists in, or essentially comprises, ‘extra-​moral’ goods. It is this latter teleology of good conduct that Anscombe is considering when she writes, ‘[G]‌etting one another to do things without the application of physical force is a necessity

Anscombe on Ought   203 for human life.’7 PJ treats the promising practice as a paradigmatic case of Aristotelian necessity created by a collective telos. One needs to keep one’s promises because the whole practice ‘is part and parcel of an enormous amount of human activity and hence of human good; of the supplying both of human needs and of human wants’ (PJ, 18) Similarly, as a form of sexual relationship, monogamy is ‘beneficial’ and ‘advantageous.’8 More generally: since any human community needs certain virtues to get on well, this Aristotelian necessity translates into another one by which its members need them in order to contribute.

9.3.6  Communal vs. Individual Goods and Constitutive vs. Functional Necessities Nevertheless, the two telê we are considering—​common good and human telos—​give rise to different types of requirement. What you need in order to be a good man are the virtues that constitute your goodness. And in order to have these you must act accordingly, namely, do whatever justice, temperance, and so on require. Every occasion counts for the constitution of that goodness (though we do not withhold the ascription of virtue because of occasional failings). It is different with the requirements of that component of the community’s (and the agent’s) well-​being, or welfare, that is effected by virtue. Here the necessity attaches to the virtues because their practice is useful, not in every instance but by and large. Thus, where acting virtuously consists in Φ-​ing, so that you ought to Φ, one way of explaining this ought is to say: your Φ-​ing is necessary for the constitution of your good character; another is to say: your Φ-​ing, though not necessarily causing any good, is an integral part of a practice that by and large causes good. Well, what’s wrong with that? We seem to have two explanations in terms of Aristotelian necessity. So much the better for the naturalisation of the ‘moral ought’! But can the two explanations coexist? (‘Two alibis are better than one!’) If the very meaning of the relevant ought is to be explained, two distinct accounts are one too many. Two different Aristotelian necessities define two different oughts. In no part of her work does Anscombe explicitly reject one of the two explanations of ought in favour of the other. Nor can we say that the two kinds of Aristotelian necessity appealed to come to the same thing. For we do not want to say either (1) that the value of a man’s goodness derives from its social usefulness or (2) that the common good is a telos subordinate to the agent’s being a good man. Indeed, the duality of explanation raises the suspicion that neither of them is appropriate. 7  PJ, 18. It is likely to be this functional, or instrumental, teleology also that is on her mind when she later speaks of reason’s ability to ‘work out’ what is ‘good and bad according to [ . . . ] nature’s needs’ (‘Sin,’ in GG2, 117–​156, here: 123). 8  ‘Good and Bad Human Action’ (GB), in GG1, 195–​206, here: 196.

204   Anselm Winfried Müller But in a way the whole issue is made obsolete by the fact that Anscombe herself came to think that some kinds of virtuous practice are not required by Aristotelian necessity—​not, at least, by being instrumental to the common good.

9.3.7  Mystical Value She now argues that several components of undoubtedly virtuous conduct can be said to ‘serve’ the quality of human life only in the degenerate sense of being constitutive of a good character. They are not required by Aristotelian necessity; their value is ‘mystical.’ Anscombe’s examples include aspects and corollaries of chastity (sexual shame and the appreciation of virginity and celibacy cannot be explained exhaustively by a need they serve), the ‘respect due to a man’s dead body’ (why should we not ‘put [it] out for the collectors of refuse to pick up’?), and the prohibition against murder (it serves our comfort: we feel safe; but this is not why one ought not to kill the innocent).9 We might add principled veracity (the brazen lie, which nobody believes, may not even be intended to deceive). In all such cases, the point of having and practicing virtue is ‘supra-​utilitarian and hence mystical.’ Anscombe’s notion of the mystical is less than clear. And her examples may be debatable. But it seems undeniable that they (and others) point to a dimension of value that cannot be reduced to instrumentality, functionality, utility. Anscombe associates it with the recognition of man’s dignity as spirit.10 But the question how non-​functional value might be associated with the human telos, or human dignity, remains deeply mysterious, as is shown by an example I take from Philippa Foot (20013, 116): Thoughts about cruelty to animals, or about the wanton destruction of useful or beautifully ingenious living things, belong within the usual distinctions of virtues and vices. Goethe told his secretary Eckermann of a certain Englishman who, owning an aviary, was so struck one day by the beautiful appearance of a dead bird that he straight away had the rest killed and stuffed. Hardly a crime! And yet there was something wrong with that man.

Does the introduction of ‘mystical value’ exclude a naturalist account of ought? Not necessarily. For although the mystical value of virtuous dispositions and actions does not consist in the gratification of natural needs, it is, according to Anscombe, in the nature of human reason to be realised in such dispositions and actions, and the human dignity recognised through them is the dignity of man’s spiritual nature.

9  ‘Contraception and Chastity’ (CC), in GG2, 170–​191, here: 187. Cf. GG1, 68–​73, 260, 267, 269–​270; GG2, 172–​173, 209, 212. 10  ‘The Dignity of the Human Being,’ in GG1, 67–​73, esp. 72. Cf. GG1, 260, 267; GG2: 187; Coope 2015; Müller 2016.

Anscombe on Ought   205 What is excluded, however, by the admission of mystical value is a common, Anscombe-​inspired type of comprehensive functionalist explanation of the necessity to act well. Possibly all virtue, mystical or not, is required (by conceptual rather than Aristotelian necessity!) for being a good man, and being a good man is required (by conceptual or Aristotelian necessity?) for attaining the human telos. But at least the commonly invoked Aristotelian necessity by which virtue seems to be required for the common good is no longer available as an account of the requirement to act well, once the existence of mystical virtue is recognised. This result has two important implications. First, we cannot aim at, let alone achieve, a materially adequate characterisation of the virtues by invoking their functional necessity. We can indeed formally characterise them as virtues, or qualities, that one needs to be a good man. But this is not a criterion. Hence, if there is mystical virtue, a unitary account of what is a virtue—​an account we might have hoped to be given in terms of Aristotelian necessity—​is not in sight. We have not been shown a way of telling virtues from other virtues. And, second, Anscombe’s rejection of the ‘autonomy’ of ethics may be in jeopardy. She holds that ethics is not ‘independent of the facts of human life.’11 I take this to imply that, given sufficient intelligence, one can ground moral judgements in the knowledge of such facts by ‘work[ing] out’ what is ‘good and bad according to [ . . . ] nature’s needs’ (‘Sin,’ in GG2, 123). Mystical evaluations would not satisfy this condition.12

9.4  Moral Rationality—​O ught by Purpose Anscombe famously says: [J]‌ust as man has so many teeth [ . . . ] so perhaps the species man, regarded not just biologically, but from the point of view of the activity of thought and choice in regard 11 

‘Twenty Opinions Common among Modern Anglo-​American Philosophers,’ in GG2, 67. Cf. ‘Authority in Morals’ (AM), in CP3, 43–​50, esp. 49–​50. 12  Allow me to add that, pace Anscombe, the idea of such a grounding of moral judgements is dubious anyway—​quite independently of the recognition of mystical value. When someone says ‘This machine ought to be oiled,’ we understand them, ceteris paribus, as basing this judgement on the opinion that, on account of given spatial and causal conditions, the machine will not operate in accordance with its purpose unless it is oiled—​even if neither we nor they know anything about these conditions. Similarly, when someone says ‘That plant ought to get some water,’ we take the ought to be justified by a pattern of operation that includes absorption of water and keeps the plant alive and healthy. Now, do we understand ‘These people ought to keep their promises’ in an analogous way? Anscombe is of course right to draw attention to the functionality of the promising practice. But even though the ‘mechanism’ by which promise-​keeping has a function in human life is more perspicuous here than with, say, forgiveness or humility, the presumption of such functionality does not seem to enter our understanding of that ought any more than it does in cases of a mystical ought, i.e., when we treat as virtuous something that cannot reasonably be presumed to be needed for some good and serve it by a ‘mechanism,’ however subtle.

206   Anselm Winfried Müller to the various departments of life [ . . . ] ‘has’ such-​and-​such virtues: and this ‘man’ with the complete set of virtues is the ‘norm,’ as ‘man’ with, e.g., a complete set of teeth is a norm. (MMP, 38)

Now compare this exchange: 1. A: You ought to have thirty-​two teeth, and you don’t. B: Well, too bad! Wouldn’t it be nice if I did? with this one: 2. A: You ought to pay your debts, and you don’t. B: Well, too bad! Wouldn’t it be nice if I did? The comparison suggests three differences between the two uses of ought. First, even if in both exchanges the ought is rooted in Aristotelian necessity, the good that you cannot attain without thirty-​two teeth is not relevant to your being a good man—​and, presumably, to the human telos—​as is the good that you cannot attain without justice. Second, the ought in (2), but not in (1), implies a reason to act. You may indeed find you ought to see the dentist in order to keep as many teeth as possible. But this ought goes beyond ought to have thirty-​two teeth. (The latter is just as ‘inert’ as ought to be born with ten fingers.) Third, the cynicism exhibited by B’s reply in (2) brings to mind that ought (2) has practical force, while ought (1) does not. This force no doubt presupposes the second difference. But it adds to it. For B may recognise an Aristotelian necessity to act on considerations of justice without being moved by them. The first of these differences is of course deep and fundamental. Anscombe speaks to it in what she says about the spirituality (rather than just rationality!) of man. In this chapter I cannot pursue the questions it raises (cf., however, Müller 2016). Leaving the challenges posed by the third difference to later sections, I’ll now consider the relevance of Anscombe’s work to the second. How does she think one can rationally be affected by moral requirements (section 9.4.1)? And where do reasons for acting as one ought to act come to an end (section 9.4.2)?

9.4.1  Morality in the Service of a ‘Main Purpose’ Can the Aristotelian necessity of a way of acting be rationally ignored? Does the fact that you cannot be a good man without justice give you a reason to act justly? Anscombe treats the question whether ‘to commit injustice . . . [might not] be the best thing to do’ as intelligible (MMP, 41). And we read, ‘‘[B]‌ilking’ is of course a species of ‘dishonesty’ or ‘injustice.’ (Naturally the consideration will not have any effect on my actions unless

Anscombe on Ought   207 I want to commit or avoid acts of injustice.)’ (MMP, 29). Presumably, the consideration that injustice prevents me from flourishing will likewise fail to influence my actions unless I want to flourish. The idea that moral necessities are not effective in people independently of what they want is resumed in the concluding pages of PJ. These pages form a somewhat puzzling part of the text (PJ, 19–​21). But the pieces of the puzzle will more or less fall into place, I think, on an interpretation that reconstructs the overall picture from the following materials. First, there is no non-​derivative, bare requirement. Justice, for instance, is of course required simply as a constituent of acting well by conceptual necessity. But for something to count as acting well, it has to play a ‘role in generally promoting [ . . . ] people’s good’ (PJ, 20): if one is required to act well, one is so by some form of natural necessity—​whether this is taken to be conditional on ‘the good of the agent [or] the common good’ (PJ, 19). Second, in order to become reality, a necessity to Φ must be perceived as, or matched by, something like a reasoned project. This perception, and the agent’s consequent Φ-​ ing, can be viewed as the conclusion of a practical inference. The inference has premises that exhibit Φ-​ing as required for some purpose actually intended by the agent. It is practical in virtue of a characteristic ‘principle of inference’ (PJ, 20), which, absent countervailing purposes, takes him from the truth of the premises to Φ-​ing.13 Third, three types of agent make their appearance in Anscombe’s text. They differ in their attitude to the necessity of paying debts (and other moral requirements) in accordance with their respective motivating aspirations: N1 has a ‘main purpose,’ MP, perhaps a ‘substantive aim the attainment of which is conceived to be, and really is, happiness’ (PJ, 20). He believes that this purpose ‘can be served only by acting well, as such’ (PJ, 19; cf. PI, 145–​147), that to act well he has to act justly, and to act justly he has to pay his debts. The practical inference takes N1 from these premises to paying his debts. Now, according to Anscombe, his principle of inference is the necessity to act well (PJ, 19–​20). But this is compatible with my saying that it is from MP that the principle takes N1 to paying his debts. For MP supplies him with a dominant purpose, so that there can be no countervailing consideration that might defeat the requirement, arising from MP, to act well. Moreover, since MP is a ‘purpose that can be served only by acting well, as such[!]‌’ (PJ, 19), N1 does not need to ask himself whether what is here and now constitutive of acting well, viz. paying these particular debts, is also required by MP. In this sense, for N1, the principle ‘Go to Φ-​ing from Acting well requires Φ-​ing’ follows from the principle ‘Go to Φ-​ing from MP requires Φ-​ing.’ N2 differs from N1 in treating the necessity of acting well as an ‘axiom’ rather than a principle of inference (PJ, 20–​21). N2 seems to come in three variants. The first makes ‘morality itself [his] substantive aim.’ The legitimate question ‘Why act well?’ has no place in his practical thinking; so from the perspective of this question,

13 Cf. Intention, §§33–​41.

208   Anselm Winfried Müller the requirement to act well is introduced as an axiom that restricts the range of possible inferences from N2’s other aims. Another instance of N2 would be a man who assumes that acting well is required by Aristotelian necessity for a ‘substantive’ good, say, X. Since, however, he does not know what X might be (if he did, he could make X his main purpose), he cannot infer the need to pay his debts on the basis of any more remote and basic end than acting well. Instead he, too, will in all his practical inferences add, as an ‘extra premise,’ the ‘axiom’ that (by whatever inscrutable necessity)14 he must act well. Like this man, the third version of N2 fails to ‘make that [his] good which nevertheless is [his] good.’ But he does have some main purpose. This purpose, however, is such as not by itself to require good conduct. So he, too, treats the requirement to act well as an axiom, apt to restrict the possibilities of pursuing that main purpose, but incapable of demonstration by Aristotelian necessity or anything else. N3 is a more straightforward type. His main purpose does not require good conduct as such, and he won’t let his aspirations be marred by treating the moral axiom as a restraining consideration. ‘When he employs it at all it will be because he has satisfied himself that in this case his main purpose cannot be attained without doing what is good or avoiding what is bad’ (PJ, 20). Indeed, even in such a case he will, by paying his debt, be acting well materially rather than formally. This interpretation leaves us with questions that I cannot go into here, for instance: If acting well ‘is an essential means[!]‌’ for the attainment of one’s good (PJ, 21), can it then still consist in practicing virtue for virtue’s sake?15 How can anyone use ‘It is necessary to act well’ as an axiom if the idea of a bare requirement is unintelligible? ‘No bare requirement!’ is a consequence of the ambition to understand the practical ought quite generally, and therefore in its application to acting well, in terms of Aristotelian necessity. If this necessity is not relative to a common good, it is a requirement of the agent’s own telos. This telos is set by his human nature. But he cannot realise it in acting unless he conceives of, treats, is guided by it as reason-​supplying dominant purpose. This is what Anscombe assumes in the text I have just interpreted, and also, it seems, when she says that only ‘if man has a last end which governs all [ . . . ] can that illusory ‘moral ought’ be exorcised’ (PI, 147).16

14 

For ‘the necessity of which the axiom speaks is that necessity of that without which good cannot come about to which Aristotle drew our attention’ (PJ, 20). 15  ‘Virtuous acts must be chosen for themselves’ (GB, 198). 16  Is this view internalist? Well, in PJ we are told that ‘if someone does genuinely take a proof that without doing X he cannot act well as a proof that he must do X, then this shows [ . . . ] that he has a purpose that can be served only by acting well, as such’ (PJ, 19). It follows that you cannot reasonably treat ‘To act well, I must do X’ as giving you a reason to do X, unless you pursue some purpose that requires you to act well. This implication amounts to internalism unless there is an unconditional requirement on you to have such a purpose—​a requirement not ultimately conditional on the pursuit of a purpose. But whence that requirement?

Anscombe on Ought   209

9.4.2  Whence the Human Telos? Axioms apart, then, if you are to base your good conduct—​the paying of your debts, etc.—​on adequate reasons, there must be an end that is both, objectively, identical with the human telos and, subjectively, what you pursue as a main purpose. Because of this second condition, ‘it is intelligible for a man to say he sees no necessity to act well [ . . . ] except as it serves his purposes’ (PJ, 19; cf. MMP, 29). The second condition does not indeed enter the meaning of the relevant ought: one ought to pay one’s debts, whatever one’s main purpose. Nevertheless, the ideal of moral rationality demands that one ‘make that one’s good which is one’s good’: ‘How can one instruct an archer to aim at an unseen target?’ (AM, 49–​50). But how is he to see it? How is one to know what the human telos consists in? Indeed, how can there be such knowledge? For, if the ought that applies to good conduct expresses a necessity that derives from a human telos, what are we to say about the ought that applies to the adoption of that telos itself as one’s main purpose? By what sort of requirement is it that the human telos is an end that you ‘ought [!]‌to have’ (PI, 145)? If it is Aristotelian necessity, it is not practicable until you have made the (further) end, or good, which requires that telos, your main purpose. But can there be such a more remote end, and what could it be? If, on the other hand, Anscombe means that the human telos itself is such as to require you to make it your main purpose, how can it at the same time coincide, in all its practical requirements, with your main purpose and require that you make it that? And, in either case, how is coincidence of your main purpose with an ultimate telos to be attained? Anscombe compares the end that someone has adopted as his main purpose with the end that a doctor ‘aims at’ in treating his patients (PJ, 19). And here the question ‘Ought he to aim at this?’ might be answered by answering the question whether, for him, it was required by such-​and-​such more comprehensive end that he ought to have. But is there room for an analogous question concerning a man’s main purpose? By what standard is it to be answered, given that the goodness of conduct (which is the obvious candidate) is itself taken to be a requirement of the right kind of main purpose? And yet the question ‘Ought a man to make such-​and-​such his overall purpose?’ does seem to make sense! Moreover, we have not yet heard how theoretical reflections on the teleology of (relatively) brute movements, actions, good conduct, and whatever end it serves, come to have any impact on how to act. Anscombe does not explicitly distinguish between X’s practically effective consciousness of a necessity to Φ from the merely theoretical knowledge that even a Martian could have of the necessity for (a human) X to Φ. But the question how the fact that ‘such-​and-​such ‘ought’ to be or ‘is needed’ is supposed to have an influence on your actions’ is of course on her mind, and the idea of a specifically practical necessity is present in her writings.17 To this topic we must now turn. 17  Cf. Anscombe’s actual use of phrases such as ‘accepting the necessity’ (PJ, 18), ‘recognition’ (PJ, 19), ‘accept what one says and act accordingly’ (AM, 43; cf. 44, 47, 49), ‘recognized ideal’ (‘You Can Have Sex without Children: Christianity and the New Offer’ [YCSC], in CP3, 82–​96, here: 94).

210   Anselm Winfried Müller

9.5  Consciousness as of Obligation—​O ught by Dictate of Conscience An ought-​thought of yours which arises from a main purpose that you think you cannot attain without acting well is a practical one. But we may still ask whether you can somehow derive it from a corresponding theoretical thought that represents an Aristotelian necessity, or whether something else, perhaps your intention of the purpose, is required as a source of that ought’s practicality (section 9.5.1). The discussion of this question leads to the idea that we cannot, after all, do without the ideas of a law-​like moral ought, and that the consciousness of moral obligation is at least one source of the unity of the notion of morality (section 9.5.2). This idea seems to be acknowledged by the significance Anscombe ascribes to conscience (section 9.5.3).

9.5.1  Practicality Sometimes Anscombe clearly draws or implies a distinction that amounts to a distinction between theoretical and practical ought. Thus, ‘teaching morals will be, not getting the pupil to think something, not giving him a statement to believe, but getting him to act’ (AM, 47; cf. METC, 226, 229–​230). The ought to Φ that, in this context, a parent would both use and want his child to accept is an ought that, ceteris paribus, implies the speaker’s readiness to Φ. It is an ought that expresses not, or not only, belief but something Anscombe herself calls recognition, or acceptance (cf. note 17). There is knowledge how to act that is ‘productive of action.’18 And there is of course the search for such knowledge by means of a kind of inference whose conclusion is manifested in conduct. We cannot, it seems, reduce the practical use of ‘I ought to Φ’ to one that expresses belief in the existence of certain needs that result from human nature. Such a belief can be held by anyone, including scoundrels and anthropologically minded Martians. They might agree that virtue serves some human good—​without being willing, or in a position, to pursue it. If there is nonetheless a conceptual connection between the Aristotelian necessity by which X, qua human, needs to Φ and X’s practical acceptance that he ought to Φ, this will be a philosophical insight, not a premise from which X himself is likely to arrive at Φ-​ing. I admit that the theoretical/​practical distinction raises problems. In particular, can we ascribe a systematic difference of significance and use to a sentence that we do not want to classify as ambiguous? How can ‘I ought to Φ,’ when uttered by itself, have a 18 

‘On Contraception and Natural Family Planning’ (CNFP), in GG2, 199–​205, here: 200.

Anscombe on Ought   211 special practical import, given that it must have the same meaning in ‘If I ought to Ψ, I ought to Φ’? How, on the other hand, can judgements of the form ‘I/​You ought to Φ’ have ‘the slightest influence on your actions’ (MMP, 30) if their significance is exhausted by the explanation that one cannot attain such-​and-​such a good unless one Φ-​s.? For an explanation of this form is (1) not what we learn when learning to use ought in relation to virtuous conduct; (2) likely to become dubious where turned into a reason for Φ-​ing; (3) neither likely nor really suitable to move to action even if the place of ‘such-​and-​such’ is taken by the kind of description a competent naturalistic ethicist might give of the human telos. So much for a distinction between theoretical and practical ought that is not made explicit but acknowledged by Anscombe. As a way of characterising the practical recognition of requirements of virtue, however, the distinction is insufficiently specific. It does not separate what is accepted as required from whatever else is wanted or wished. A fortiori, it leaves the nature of specifically moral requirements unexplained.

9.5.2  Unifying Moral Experience What, then, gives the notion of moral requirement its unity? So far, I have talked about our topic in terms of (1) virtue and good conduct, (2) moral subject matter, and (3) Aristotelian necessity. Now, (2) has been explained in terms of ‘human passions and (non-​technical) actions’ (MMP, 29)—​which does not tell us much unless we know the dimensions of evaluation in which passions and actions constitute moral subject matter (as opposed to, say, matter for etiquette, or law). These dimensions, however, refer us to (1): they are the ones determined by the virtues (some of which, by the way, such as justice and veracity, do not as such regulate passions). What, however, counts as a virtue? This concept receives at most a formal unity by being confined to the collection of those qualities that are constitutive of the good man—​a notion that has not been clarified (cf. section 9.3.3). The possibility of relying on (3) and appealing to a conception of the common good, in relation to which the virtues exhibit Aristotelian necessity, is precluded (inter alia) by Anscombe’s own admission, or insistence, that the value of virtuous conduct is partly mystical. So what is morality, as generally conceived? An essential part of the answer seems to be: We connect the idea of morality, at least, inter alia, with a specific, unifying way of being conscious of different requirements, a characteristic experience that can be variously described as a feeling of obligation, as conscience, as a sense of right and wrong, etc. Philosophy has to concede to this experience a place in the constitution of the notion of morality.19

19 

‘Moral experience’ of course takes different forms. The dictates of conscience, in particular, can relate to what is required in a general way, or absolutely and always, or here and now.

212   Anselm Winfried Müller It should, more particularly, include a bit of phenomenology and acknowledge that ideas of lawlike obligation, requirement, and duty, of the illicit, the permitted, etc., as associated with the notions of acting well and virtue, are spontaneous and indispensably meaning-​constitutive metaphors.20 Pace Anscombe, they are not just trappings left over from a discarded theory of morality: they articulate a type of experience.21 If this is right, the expression ‘moral obligation’ is capable of a meaning that is more respectable than the ‘mere mesmeric force’ (MMP, 32) left lingering substanceless like the grin of the Cheshire Cat. And Anscombe’s advice, addressed to her colleagues, ‘to discard the notion ‘morally ought’’ (MMP, 41) was rash even at the time. I am not here arguing that moral consciousness is strictly the experience of laws, or obligations. (What could this be like?) It is, rather, an experience that finds spontaneous expression in words and phrases taken over from the language of law or claim or injunction. Some such experience seems to be an indispensable backdrop to the unity of our notion of morality—​a unity that determines the extension of the concept of virtue and that does not, on Anscombe’s own showing, derive from criteria supplied by Aristotelian necessities alone. You experience a moral ought as unconditional and as conditionalising all other oughts and woulds. But not as an arbitrary taboo, as beyond your power (or anybody’s) to create or remove. Nor as constituting, as it were, a foreign body in your mind, but rather as inseparable from what you want to be. As a ‘claim’ that may precede, accompany, follow action. The practical import of ought is reflected in moral experience ideally (i.e., in accordance with its teleology) as commitment, as readiness to comply; or, otherwise, in a sense of deficiency. What you practically think you ought to do differs from what you merely intend to do in that, failing to do it, you do not just regret or feel disappointed or try again: you typically have (at least traces of) a ‘bad conscience,’ feel guilt and remorse, perhaps repent, confess, apologise, and try to improve.22 Conscience involves the disposition to represent (especially on pertinent occasions) in judgement both the claim and the recognition of its authority.

20  The transfer of legal language to dictates of conscience calls to mind Wittgenstein’s (1953: 216; cf. 197, 204, 215) remarks on spontaneous secondary uses of expressions. Roger Teichmann kindly draws my attention to Cora Diamond’s (1991: 237) suggestion that the expression one ought absolutely is ‘forced on us’ as a ‘secondary sense’ extension of an ordinary, relative ought. 21  Indeed, philosophical and theological law conceptions may themselves reflect a defining aspect of moral experience—​though only at a remove, in that they use the vocabulary of law to interpret, rather than express, the experience. There are of course other interpretations, too, such as Freudian ones, and it would be the task of a ‘phenomenology of conscience’ to show their inadequacy. 22  To describe the experience in more detail, one would also have to attend to the circumstances, behavioural and other, that go with such phrases as ‘conscience tells me,’ ‘I must /​cannot . . .,’ etc. There are, admittedly, transitional attitudes between remorse, on the one hand, and the cynical ‘fig for what I ought to do’ on the other. So, on the psychological level, the distinction between merely theoretical agreement that one ought to Φ—​including perhaps a sense of undischarged obligation (cf. section 9.6.1)—​and (practical) acknowledgement that one ought to Φ is not sharp.

Anscombe on Ought   213

9.5.3  The Significance of Conscience Anscombe herself is more impressed with the fatal fallibility of conscience than with its authority (MMP, 27–​28, 37). She repeatedly attacks ‘the idea that one’s own personal conscience is necessarily the supreme arbiter in matters of right and wrong’ (AM, 46). And in a manuscript with the title ‘Must One Obey One’s Conscience?’ (GG1, 237–​241), she follows Aquinas in holding that, in general, both following and acting against an erring conscience are wrong. On the other hand, although the dictates of conscience are fallible, ‘with doing [as opposed to unreflective believing], however obedient one is, one can hardly escape being one’s own pilot’ (AM, 48). So Anscombe does admit a ‘sense in which it is indeed impossible to take anything but one’s conscience as arbiter of right and wrong’ (AM, 47).23 The voice of conscience, one might say, can be influenced and tuned, but not superseded, by advice or theoretical judgements of absolute Aristotelian necessity or the like. When talking about conscience, Anscombe also allows herself to use the language of obligation and obedience. In ‘Must One Obey [!]‌One’s Conscience?’ she writes that ‘conscience is judgement about what is all right, or wrong, or obligatory [!] to do’ (GG1, 239; cf. AM, 46). The paper does not seem to presuppose a divine lawgiver, and yet it recognises a quality of moral consciousness—​and thereby a force of ought—​that one associates with the idea, not of natural needs or the like, but of law.24 Two major ethical theories can be viewed as interpretations of this force of ought: Natural Law and Kantian Autonomy. As we have seen, Anscombe rejects the latter because there is no such thing as legislating for oneself. The former she finds abandoned by contemporary philosophy. But she herself professes to accept it in a form, more or less, that it has assumed in the work of Aquinas.

23  Unfortunately, she obscures this sense by comparing conscience with memory: ‘I can make no judgement about the past without some reliance on my own memory. But only a fool thinks that his own memory is the last word, so far as he is concerned, about what happened. [ . . . ] Similarly, in practical matters, a man must put some reliance on his own conscience [. . . But] only a foolish person thinks that his own conscience is the last word, so far as he is concerned, about what to do’ (CP3, 46). I do not deny the analogy. But there is a very pertinent difference: if you correct your memory judgement by putting more trust in traces, or another’s memory, or a public record, the result is not itself a memory judgement. Memory is not the ultimate ‘arbiter’ of the past: your reasons to trust another source need not involve memory. If, on the other hand, you correct a verdict of your conscience by relying on ‘another man’s moral counsel’ or ‘some public source of moral teaching’ (CP3, 46–​47), the result is again a judgement of conscience. Even if you take divine revelation as a source of moral knowledge, should not your conscience play an essential role in accepting it as really coming from God? 24  This accords with the distinction of practical judgements from ‘judgements in the field of theoretical morals’ and others that are ‘idle, without consequences’ (AM, in CP3, 48). Anscombe admits that ‘if the divine law obliges not to commit injustice by forbidding injustice, it really does add something to the description ‘unjust’ to say there is an obligation not to do it’ (CP3, 41; cf. GG2, 123). In the present section I have argued that the consciousness as of obligation already does that.

214   Anselm Winfried Müller

9.6  Underived Practical Knowledge—​O ught by Natural Law I have argued that one may reasonably agree that morality is experienced as law-​like without presupposing the existence of a divine lawgiver. One can then ask whether that experience is best interpreted by maintaining that God supplies us with the consciousness of requirements that do in fact represent a law which he creates as a constituent of human nature. Anscombe does not explicitly ask this question. But some of her writings clearly show her accepting Natural Law as a philosophical conception of moral goodness and of the knowledge of the corresponding requirements. I’ll sketch the doctrine she finds in Aquinas, which seems at last to give sense and unity to the idea of a moral ought (section 9.6.1). It raises questions (section 9.6.2), but answers others that were not dealt with satisfactorily by the provisional explorations in earlier sections (section 9.6.3). Nevertheless, there are prima facie discrepancies between the reliance of Natural Law on underived moral knowledge and Anscombe’s insistence that human reason be able to account for moral norms (section 9.6.4). So I conclude by ventilating the question whether she could unify her observations on ought by a modified conception of natural law (section 9.6.5).

9.6.1  Synderesis Taking her bearings from Aquinas, Anscombe says, ‘‘Natural law’ is simply a way of speaking about the whole of morality, used by Catholic thinkers’ (CC, 179). It is ‘promulgated to every grown man in his knowledge of good and evil’ (MMP, 37). And, as she notes already in a youthful text from 1939, ‘it binds because it is the law of his nature.’25 What Aquinas ascribes to every human being is synderesis, the knowledge of generic moral principles, a disposition to judge correctly concerning basic moral requirements.26 This knowledge is practical in the special sense explained in

25  The Justice of the Present War Examined (JPW), in CP3, 72–​81, here: 72–​73. The natural law can be viewed as a reflection, in the mind of man, of the eternal law, which is a ‘dictate of reason’ in the mind of his creator. ‘Aquinas called it ‘the participation in the eternal law of the rational creature’’ (JPW, 73). The reference is to Summa Theologica I-​II q 91 a 2 ad 3; cf. I q 22 a 1–​2, I-​II q 91 a 1–​2. 26  Summa Theologica I-​II q 90 a 4 ad 1 (promulgatio legis naturae est ex hoc ipso quod Deus eam mentibus hominum inseruit naturaliter cognoscendam). The natural law tells you that, e.g., you must avoid ignorance, thieving, and unjust harming; that you must keep contracts and not imprudently give offense. It even prescribes the immunity of ambassadors (Aquinas 1934: 1018–​1019; Summa Theologica I q 94 a 2).

Anscombe on Ought   215 section 9.5.1. It is natural in that God promulgates the natural law by creating man equipped with that knowledge. It is therefore underived, not the result of discursive thinking.27 For the same reason it is universal. It is indeed actualised paradigmatically in the way it is applied in good people’s judgements of conscience about what they ought to do here and now. But even in evil people its ‘murmurings’ are never totally stifled (Summa Theologica I q 79 a 12 c). An unmediated, non-​theoretical consciousness of basic and generic moral requirements, however much obscured by circumstances or bad habituation, is present in the vicious as well as the virtuous. In Anscombe’s words, ‘People of the most horrible principles know quite well how to cry out against injustice and lying and treachery, say, when their enemies are guilty of them. So they in fact know quite a lot’ (AM, 45).28 Here, then—​in the fact that everyone is supplied with practical knowledge of the principles of natural law—​we finally seem to have the source of an ought that can legitimately be called a moral one. Anscombe is convinced that that knowledge is indeed available to all of us. But she does not do much to elaborate or defend this conviction and to win her colleagues over to it. In the following three subsections I wish to indicate that Natural Law can stand up well to objections and that it throws light on a number of problems raised in previous sections, but that some of Anscombe’s views on the role of reason seem to be in tension with it. I’ll begin by answering some critical questions about it.

9.6.2  Doubts about Natural Law 1. Are we to believe in innate moral principles? Anscombe would certainly agree with Aquinas that even the ‘first principle of practical reason’ that ‘good is to be done and evil avoided’ cannot be known as long as the concepts involved are not (yet) available.29 So knowledge of principles is not strictly inborn. 27 

Summa Theologica I q 79 a 12 c. Cf. also I-​II q 90 a 4 ad 1; q 91 a 2 c; ad 3; q 94 a 1 ad 2; a 2; a 6; II-​II q 47 a 6 c; ad 1. While human nature supplies you with the practical consciousness of generic principles that articulate what virtuous conduct as such consists in and with the ability to reason from there, the rest is up to you: you have to develop this ability into wisdom, and the natural inclination to implement those principles into reliable ethical virtue; and on this basis to end up doing what actual occasions require you to do. 28  But might not their ‘knowledge’ be merely theoretical, like the Martians’, and their outcry correspondingly hypocritical or cynical? It might. But they could be blaming the wrongdoer. And blame for Φ-​ing does not only presuppose, in the Φ-​ing wrongdoer, voluntariness and cognitive access to the requirement not to Φ. (You do not blame a vase for being ugly, or people of the past for eating foods we now think harmful.) No, it also presupposes (rudimentary) acknowledgement of the badness of Φ-​ing in the blamer. (What sort of assumption about a Martian would allow you to say that he or she (?) is blaming you for deceiving them—​not just bellowing, or else observing that your behaviour does not accord with Aristotelian necessity?) 29  Summa Theologica I q 94 a 2; I-​II q 51 a 1 c; Aquinas 1934: 1018–​1019.

216   Anselm Winfried Müller 2. How, then, can synderetic knowledge be a natural endowment? An animal’s nature is not restricted to things essential to its initial realisation. And human nature, in particular, is not restricted to what (if anything) is characteristic of men independently of their social and rational form of life. Hence one tends to acquire relevant concepts and patterns of practical thinking, to which one is disposed by human nature, in the course of a normal upbringing (cf. Summa Theologica I-​II q 51 a 1 c). ‘Natural virtues’ and ‘inclinations’ dispose one to become virtuous (I-​II q 63 a 1 c). And, when defending ‘authority in morals,’ Anscombe draws attention to ‘the great assumption [ . . . ] that no one who is taught at all can fail to be taught a great deal that is true and that to a great extent verum index sui et falsi’ (AM, 45). But nature does not preclude impairment of rational and ethical any more than physical growth. 3. Doesn’t this mean that one may, in spite of synderesis, come to hold inadequate moral principles? Well, to some extent, the possibility of deficiency in one’s natural endowment might explain why moral beliefs can be uncertain and a matter of dispute. The main reason, however, why one’s moral convictions are inadequate will be failure to cultivate ‘phronetic’ knowledge (as actualised in judgements of practical wisdom) and to give shape to one’s character in accordance with virtue. 4. Where does Natural Law draw the line between synderetic and phronetic knowledge? I find no answer to this question in Anscombe.30 Aquinas does, to match the difference between (acquired) wisdom and (natural) synderesis, distinguish particular from universal principles (Summa Theologica I-​II q 58 a 5 c)—​but by what criterion? Like the particular ones, the universal principles in question are already material rather than formal. A serious defence of Natural Law will have to answer this question, too. Meanwhile let us observe that it does answer questions raised in the course of our examination of suggested ways of understanding the ought of virtue.

9.6.3  Virtues of Natural Law 1. It is trivially true that, as conceived of by Anscombe, Natural Law is not open to the objections raised against the ideas of moral obligation etc. discussed in section 9.2. But it is important here to keep two things apart: the agent’s moral consciousness or experience as of a law, on the one hand; and the philosopher’s interpretation of it as knowledge of natural law, on the other. It may, in particular, be plausible to maintain that ‘it is not possible to have such a conception [that moral requirements are laws] unless you believe in God as a law-​giver’ (MMP, 30). But, given that the notion of a moral law does imply that of a legislator, why shouldn’t a philosopher be led to believe that there is a legislating God

30  While, to the best of my knowledge, she makes no use of the notion of synderesis, she might have raised the question in this form: What are the starting points of wise deliberation, and where does wisdom get them from?

Anscombe on Ought   217 by an antecedent certainty that there is such a law, rather than vice versa?31 And, above all, agents’ consciousness of a law-​like ought does not commit them to belief in God. 2. Anscombe’s objection that the so-​called moral law lacks content cannot be levelled against Natural Law, since the content of synderetic knowledge is revealed in the consciousness of specific moral obligations. 3. In section 9.3.3 we confronted Anscombe with the question: What is the good that gives rise to Aristotelian necessities and cannot be attained without virtue? This is now a question for further investigation, once it is agreed that moral requirements are experienced as law-​like. Of course, the fact that promise-​keeping, compassionate helping, perseverance are useful will not escape the notice of an ordinary agent, and the prospect of good effects plays an important role in wise deliberation. But functionality is not a primary component in a general account of what we mean by the relevant ought. 4. It follows that one’s compliance with the requirements of virtue does not depend on one’s pursuit of an end in the sense of either the common good or the human telos. 5. Natural Law leaves room for mystical virtue. The shaping of the human mind by virtue, whether of instrumental or mystical value, is as such conformity with synderesis, hence with human nature. The (mutual) teleologies of the virtues within a life of this nature are, once more, a further question. 6. Synderesis determines the scope of the moral ought and hence of virtuous conduct. The question which virtues are virtues is answered on this basis. 7. Since there is no appeal to a standard beyond the authority of synderesis, the question whether one must obey God makes sense at most on a conception whereby God is not the origin of the natural law. 8. Natural Law gives an account of moral knowledge. We can know how to act in given situations on the basis of synderesis. Synderetic knowledge how to act is not based on anything, but it is, ceteris paribus, available to anyone. 9. It is, above all, not based on theoretical knowledge.

9.6.4  Incompatibilities The features and implications I have enumerated make Natural Law attractive. But is it Anscombe’s view? The idea of natural law is present in her ethical thinking throughout her writing life (from JPW to ‘Sin’ in GG2). And she explicitly refers it to the work of Aquinas. But not everything she says about it reflects his conception. Thus she says that ‘making use of the gifts of reason and revelation,’ one can ‘discover [!]‌by reason’ what the natural law consists in (YCSC, 83). And that, according to Natural 31  Or does it, after all, make sense for someone to express their experience of moral obligation by saying they are (not just: feel as if) under an absolutely binding law, and leave it open whether there is any legislator? Thomas Pink (2016: 117) observes that even philosophers have held the view that the existence of natural law does not entail the existence of God.

218   Anselm Winfried Müller Law, ‘the general precepts of morality are laws promulgated by God our Creator in the enlightened human understanding when it is thinking in general terms about what are good and what are bad actions. That is to say, the discoveries of reflexion and reasoning [!] when we think straight about these things are God’s legislation to us (whether we realise this or not)’ (CC, 179). ‘These things’ are something ‘reason could work out [!]‌without belief in divine commandments’ (GG2, 123)—​presumably by attending to requirements of a common good, ascertaining the existence and nature of a human telos, and studying what it involves or requires.32 But if so, one’s knowledge how one ought to act is not underived, or ‘indemonstrable,’ in the way that knowledge of natural law is according to Aquinas.33 Nor is that Anscombian understanding of natural law immune to objections. For, first, it does not answer the question how the knowledge of moral norms, as derived from the study of Aristotelian necessities, can be practical.34 Second, knowledge of human telos, of the common good, and of pertinent Aristotelian necessities, even if within reach for a brilliant philosopher, is not the kind of knowledge that is accessible to a moral agent as such.35 Neither, third, does it seem to be required and actually employed by a virtuous person. Even if we assume, with Anscombe, that nothing can be ‘good or bad not for any reason, not because of any facts,’ even if we reject the possibility that ‘this is to be believed, and could [!]‌not be known or inferred from anything else’ (AM, 49), it does not follow from this that knowledge that one ought to Φ must be inferred from ‘facts’ that require one to Φ. Moreover, if some requirements of virtue are mystical, it seems that they have to be known independently of any understanding of a (functional) rationale.

32 

Anscombe recommends ‘as a first principle [ . . . ]: Aim at what human life is for attaining.’ That ‘there is such a true end [ . . . ] can well be argued’ (‘Sin,’ in GG2, 120). If our nature, and hence the human telos, were created differently, ‘the divine law would be different [ . . . ] and it would command and forbid the things that would then be good and bad according to that nature’s needs.’ By consideration of Aristotelian necessities, reason can then ‘work out’ from man’s actual nature and telos how natural law in fact tells us to act (‘Sin,’ 122–​123). Cf. also AM, 49–​50, where we read ‘that you cannot comply with a law whose point you don’t grasp’ and ‘How can one instruct an archer to aim at an unseen target?’ 33  For him, synderesis is ‘instilled in men’s minds’ (Summa Theologica I-​II q 90 a 4 ad 1) as knowledge, ‘quasi indemonstrable’ (Aquinas 1934: 1018–​1019), of first principles (Summa Theologica I-​II q 94 a 1 ad 2) ‘from which reason proceeds’ (I-​II q 58 a 5 c; cf. I q 79 a 12 c). Cf. also I q 94 a 2; II-​II q 47 a 6 c; II-​II q 47 a 15 c. I-​II q 51 a 1 c says in effect that, while the conception of a principle has to await the formation of relevant concepts, its content is supplied by the natural constitution of human reason. 34  Interestingly, Anscombe does not treat the unmistakably practical judgements of conscience as representing, ceteris paribus, immediate knowledge of natural law, whereas for Aquinas such knowledge (viz. synderesis) can, qua giving rise to judgements of conscience, be said to prompt good, and object to bad, behaviour (Summa Theologica I q 79 a 12 c). 35  ‘A thoroughly honest person [ . . . ] may not be able to give much explanation why such-​and-​such an act is dishonest, may not be armed with a theoretical account; he merely has a good sense of what is honest and what is not. But reason demands that there be an account; and those who don’t live rightly [ . . . ] badly need an explanation’ (CNFP, 200). Do they? And if they do, how do they get it?

Anscombe on Ought   219

9.6.5  A ‘Syndromic’ Conception? But perhaps these objections are overstated, and there is a variant of Natural Law that takes account of them. If there is such a variant, it might permit an understanding of ought towards which the various things that Anscombe says about moral necessity converge. Such a variant natural law conception will be ‘syndromic’ in that it runs together aspects of thinking that we have so far kept separate. Five such separations are relevant. 1. Practical and motivating vs. theoretical and explanatory. Syndromic Natural Law treats practical knowledge as able to incorporate into virtuous motivation explanatory teleological insight. The idea is, roughly, this: if you have a conscience at all—​if you are practically conscious of unconditional requirements—​the theoretical knowledge that, e.g., some ultimate telos requires humans to Φ will make it rational for you to ‘absorb’ this requirement into your practical understanding of how you ought to act. 2. Ordinary vs. philosophical. Such appropriation of theoretical understanding is relevant if it extends to contents that I have so far treated as irrelevant and hardly accessible to the agent as such. True, in spite of consideration (1), your reason, and motive, e.g., to keep your promise consists in your having given it. But this does not exclude that you have an idea, albeit vague, of (a) the common good, or/​and something you ought and want to be, and (b) the significance for this of promise-​keeping in general; and that ideas (a) and (b) play a role in your knowledge that you ought to keep that particular promise. (‘Where would we be if we did not keep our promises?’ ‘What if everybody did that?’ ‘Would you like to be treated like that?’ ‘Do you want to be a bilker?’ ‘How could I still look myself in the face?’) Do these considerations help with mystical value? Perhaps they do, if we refrain from restricting the ‘significance’ of virtuous conduct to Aristotelian necessity in the sense of functional value. If, e.g., we want to deny, with Anscombe, that the treatment of a human corpse can be ‘good or bad not for any reason, not because of any facts,’ or that some moral norm ‘is to be believed, and could not be known or inferred from anything else’ (AM, 49), we might view as relevant facts the compatibility or incompatibility of that treatment with felt respect for the deceased. 3. Common good vs. human telos. Three questions regarding the teleologies of virtuous conduct face Syndromic Natural Law: (a) What is the human telos, and, if it is happiness, what does it consist in? (b) What is the common good? (c) If both of these are served by virtuous conduct, how do the two objective ends relate to each other? Without at least a beginning of answers to (a) and (b), the project of providing moral obligation with a basis in human nature and to tell virtues from other qualities (as assumed in (2)) does not get off the ground at all. But (c), too, needs an answer. For, if good man (or acting as one ought) is a teleological notion, it must not be explained in terms of two unrelated telê. 4. Underived vs. reasoned. Inspired by Aquinas’s concept of true connatural judgement (e.g., Summa Theologica I-​II q 58 a 5 c), Anscombe came to believe that a kind of underived moral knowledge is found in good people—​t heir virtues give

220   Anselm Winfried Müller them knowledge of what it is virtuous to do—​and that ‘it is not unavailable to us who are not virtuous,’36 even if we ‘badly need an explanation’ (CNFP, 200). Here there may be further room for ‘cooperation’ between immediate and ‘worked-​out’ knowledge. Moreover, the knowledge one acquires by being brought up in the usual way inevitably comprises general practical knowledge of how to act that is underived in that one has not thought it out for oneself (AM, 45). And upbringing ‘supplies the raw material, the matter of which the ethical is the form’ (METC, 225)—​once more a way the child acquires, by being prompted rather than by reasoning, what is by and large knowledge.37 No sharp dividing line seems to run between either case of ‘underived’ knowledge and the kind of moral knowledge that an agent might himself derive from more general moral knowledge or else from factual considerations.38 5. Individual vs. communal. ‘It is in the nature of man to know moral principles.’ This statement of Natural Law might be taken in a sense that does not require the knowledge involved to be in each head. It is in our nature to be taught and to consult, to counsel and to persuade. Syndromic Natural Law casts doubt on the significance of the five dichotomies I have mentioned. Why should divine promulgation of norms take the form exclusively of unmediated certainties, a sort of rational instinct? Why should it not work through the media of spontaneous cognitive response to experienced ‘material,’ upbringing, authority, counsel, shared tradition, reflection and investigation, teleological argument, theoretical as well as practical inference, and, perhaps, conversion?39 I do not know whether a consistent account along these lines is achievable, let alone true. But I do think that Anscombe’s understanding of moral knowledge must be something like Syndromic Natural Law, if the various things she says regarding our awareness of the requirements of virtue are to fit into a consistent conception of basic moral knowledge. Absent such convergence, Anscombe cannot have a unified account of ought, for an unqualified practical ought-​judgement implies a claim to moral knowledge. Even in this case, however, we cannot but admire the boldness and circumspection, thoroughness and productivity with which she has explored the most central aspects of the topic. 36 

‘Knowledge and Reverence for Human Life’, in GG1, 59–​66, here: 59. Note that the provision both of suggestive material and of explicit moral teaching is part of the natural form of human life, so that the resulting knowledge has a claim to representing natural law. 38  As we have seen, for Anscombe (though, I suppose, not for Aquinas) even knowledge arrived at by explicit ‘reflexion and reasoning’ (CC, 179) can be part of natural law, because the use of reason is natural to man. 39  She even identifies as Natural Law the view ‘that when human beings by their own reason have been able to think things out, their realization was itself a promulgation to them of the divine laws.’ But she adds, somewhat surprisingly, ‘I am not sure how important this belief is’ (‘Sin,’ in GG2, 123; cf. AM, 46–​47). 37 

Anscombe on Ought   221

References Aquinas, Thomas. 1934. In decem libros ethicorum Aristotelis ad Nicomachum expositio. Turin: Marietti. Coope, Christopher M. 2015. ‘Making Morality Intelligible’. Philosophy 90: 403–​455. Diamond, Cora. 1991. ‘Secondary Sense’. In Cora Diamond, ed., The Realistic Spirit. Cambridge, MA: MIT Press, 225–​241. Foot, Philippa. 2001. Natural Goodness. Oxford: Clarendon. Müller, Anselm W. 2016. ‘The Spiritual Nature of Man’. In Luke Gormally, David A. Jones, and Roger Teichmann, eds., The Moral Philosophy of Elizabeth Anscombe. Exeter: Imprint Academic, 11–​37. Pink, Thomas. 2016. ‘Anscombe, Williams and the Positivization of Moral Obligation’. In Luke Gormally, David A. Jones, and Roger Teichmann, eds., The Moral Philosophy of Elizabeth Anscombe. Exeter: Imprint Academic, 98–​118. Wittgenstein, Ludwig. 1953. Philosophical Investigations. 2nd edition. Oxford: Blackwell.

Pa rt I I I

H UM A N L I F E

Chapter 10

J u stice and Mu rde r The Backstory to Anscombe’s ‘Modern Moral Philosophy’ John Berkman

10.1  Introduction This chapter has a simple thesis: Anscombe’s work in moral philosophy, as represented by ‘Modern Moral Philosophy’ (MMP) and other publications up through the mid-​ 1960s, as well as her work on action-​descriptions and the philosophy of mind, were driven by her concern to recover the absolute moral prohibition on murder, and the virtue of justice as the appropriate basis for it.1 In other words, MMP was not written as per se a contribution to ‘meta-​ethics’ but used moral theory to expose the basis of her society’s and her colleagues’ failings concerning the moral question of murder.2 MMP was one prominent element in what would prove to be an ongoing project (one that Anscombe would never complete) to recover the concept of murder, a ‘moral action

1  My heartfelt thanks to Anthony Kenny for numerous conversations and insights which inspired this paper. Mary Geach and Luke Gormally generously shared otherwise unknowable details about Anscombe, and Mary Geach granted permission to publish materials from the Anscombe archive. My thanks to Peter Conradi, Arthur Gibson, John Haldane, Fergus Kerr, James Klagge, Alasdair MacIntyre, and Rachael Wiseman for sharing their knowledge of Anscombe; to Bethany Antos, Christian Erbacher, Richard Finn OP, Valerie Flessati, Ronald Hustwit, Stephen Leach, and Kate O’Donnell for assistance accessing archival sources; and to Daniel Cheely for financial support. Benjamin Lipscomb graciously shared materials and drafts of his forthcoming The Women Are Up to Something: How Elizabeth Anscombe, Philippa Foot, Mary Midgley, and Iris Murdoch Revolutionized Ethics (Oxford: Oxford University Press, 2022). My thanks for critical comments on drafts from Robyn Boeré, John Finnis, John Hayes, Anselm Müller, John Rist, and especially Luke Gormally and Roger Teichmann. This essay is dedicated to the memory of Joseph M. Boyle Jr. 2  In 1981, Anscombe writes, ‘[M]‌y interest in moral philosophy has been in particular moral questions [rather] than in what is now called ‘meta-​ethics’’ (CP3, viii).

226   John Berkman concept of great complexity.’3 But that puts the matter too abstractly. Anscombe’s passion was to expose and critique the assumptions and arguments of those who thought massacring innocent people could be morally justified, either as a serious instrument of statecraft or merely as an exercise of clever philosophers proposing serious moral conclusions from fantastic examples. While MMP was rightly seen as revolutionary, most of the claims in MMP were restatements or expansions of points Anscombe had made in earlier (somewhat obscure) publications.4 Even her celebrated invention of the term ‘consequentialist’ was not original to MMP. I point this out not to downplay MMP’s monumental significance but to indicate that to be rightly understood, MMP must be placed within the context of an ongoing project that was an obsession of Anscombe’s. What led Anscombe to write MMP, Intention, and dozens of other essays, pamphlets, lectures, talks, and letters to editors, was to help educate people so they would not be fooled when political leaders, moral degenerates, or Oxford moral philosophers attempted to justify this or that particular murder or massacre.5 My goal here is to show Anscombe’s ongoing and overriding commitment to the project of recovering the concept of murder. I pursue this goal in large part by locating many of Anscombe’s early works in the context of her larger life commitments. To keep this chapter a manageable size, I will not be able to provide much in way of the details of, or evaluate the relative strength and weakness of, the many arguments of Anscombe I present. Commentators on MMP typically presume that the three theses that begin MMP constitute its point. I believe this is mistaken for two reasons. First, these theses appear to be added after the fact, as if an editor instructed Anscombe to summarise her arguments. I posit that the original version of MMP began at the second paragraph, with its marvellous opening line: ‘Anyone who has read Aristotle’s Ethics and has also read modern moral philosophy must have been struck by the great contrasts between them.’6 Second, elsewhere in the essay Anscombe explicitly states the essay’s main point—​her

3 

CP3, vii–​viii. Rachael Wiseman’s ‘The Intended and Unintended Consequences of Intention’, ACPQ 90, no. 2 (2016), 207–​227, and her blog Women in Parenthesis, http://​www.wom​enin​pare​nthe​sis.co.uk first alerted me to the 1957 exchange between Anscombe, Hare, and Nowell-​Smith. 5  With regard to Intention, a discussion of which is unfortunately beyond the scope of this chapter, Mary Geach says that ‘the [Truman] episode made her give the course of lectures that became the book Intention. If people were capable of excusing Truman by saying he had only signed his name on a piece of paper, there were things that she knew that they did not’ (GG1, xiii–​xiv). 6  Anselm Müller’s description of how Anscombe typically began essays supports this view. As Müller puts it, ‘Anscombe has some very characteristic ways of introducing a topic. . . . [W]‌here there is no need to clarify the context, she plunges medias in res with the very first sentence of most of her articles: a striking observation, an intriguing question, an unexpected paradox, reference to a deceptively harmless well-​known fact, a puzzling—​or innocent-​looking—​quotation. No preliminaries, no excuses, no explanations or justifications.’ Anselm Müller, ‘The Personality of Elizabeth Anscombe,’ unpublished manuscript, 39. Shared by the author on January 12, 2021. 4 

Justice and Murder    227 ‘complaint.’7 She keeps her main point under wraps until MMP’s final paragraph and buries it in a 217-​word sentence: [M]‌odern moral philosophy . . . construct[s] systems according to which . . . it is left open to debate whether such a procedure as the judicial punishment of the innocent may not in some circumstances be the ‘right’ one to adopt . . . they teach a philosophy according to which the particular consequences of such an action could ‘morally’ be taken into account by a man who was debating what to do; and if they were such as to conflict with his ‘ends’, it might be a step in his moral education . . . to decide: in such-​and-​such circumstances one ought to procure the judicial condemnation of the innocent. And that is my complaint.8

Anscombe’s main point in MMP is that the prevailing Oxford moral philosophies coming after Henry Sidgwick9 are what she labels ‘consequentialist’, meaning that their practitioners ‘construct systems according to which it is left open to debate whether [murdering] the innocent may not in some circumstances be the right one to adopt’.10 While Anscombe’s three theses do not constitute her complaint (i.e., her main point), they do outline in a nutshell her fall-​redemption narrative for moral philosophy, the potential saviour of modern moral philosophy being what she calls the philosophy of psychology. MMP breaks new ground beyond Anscombe’s earlier publications in developing the historical narrative of moral philosophy’s fall, highlighting in particular the work of Henry Sidgwick in determining the character of modern (i.e., Oxford) moral philosophy, and in expanding her discussion of the means by which modern moral philosophy might be redeemed. While there are various digressions in the article, the three theses, if one rearranges their order, are the keys to her narrative, and each receives about one-​third of the paper. A: ‘Concepts of . . . moral obligation and moral duty . . . of morally right and wrong, and of the moral sense of “ought,” ought to be jettisoned’,11 because having lost the notion of a divine or other giver of a moral law, the way Oxford moral philosophers use the term ‘moral’ has ‘no discernible content’.12 The opening pages of MMP elaborate this thesis, narrating the devolution of the ‘moral’ ought into incoherence, Hume being singled out for showing that ‘no content could be found in the notion “moral ought.” ’13 Anscombe 7 

Thanks to Roger Teichmann for suggesting I speak of her complaint as her main thesis. MMP, 42. To understand this quote in its context, one should read Anscombe’s use of ‘punishment’ and ‘condemnation’ as meaning ‘execution.’ Elsewhere in the essay Anscombe explicitly refers to her concern as with the ‘judicial execution of the innocent.’ Which fits Anscombe’s primary concern with ‘murder,’ the killing of the innocent as a means to one’s ends. 9  Sidgwick (1832–​1900) himself does not qualify as an ‘Oxford moral philosopher,’ having spent his entire career from undergraduate onwards at Trinity College, Cambridge. My thanks to Nicholas Denyer of Trinity College, Cambridge for reminding me of Sidgwick’s milieu. 10  MMP, 42. 11  MMP, 26. This is Anscombe’s second thesis. 12  MMP, 26, 41. 13  MMP, 32. 8 

228   John Berkman suggests that previously belief in a divine lawgiver had given meaning to the ‘moral’ ought and duty.14 But having rejected a law-​based conception of morality, no other adequate basis for retaining the ‘moral ought’ has been found.15 B: ‘The differences between the well-​ known English writers on moral philosophy since Sidgwick to the present day are of little importance’,16 because all of them are consequentialists, i.e., philosophers who refuse to rule out any class of actions (e.g., mass murder, ‘roasting babies alive’, torture, etc.) as absolutely wrong. Rather, ‘in such-​and-​such circumstances, you ought to do’ such things.17 Prior to Sidgwick, moral philosophers agreed that some acts were simply wrong, no one appealing to (typically fantastical) circumstances to justify heinous acts of murder, extortion, etc.18 Sidgwick advanced the thesis that ‘it does not make any difference to a man’s responsibility for an effect of his action which he can foresee, that he does not intend it.’19 Anscombe says this novel claim about the nature of moral responsibility has been accepted by all subsequent prevailing Oxford moral philosophies, and by accepting it, they have no basis for ruling out that in certain circumstances one ‘ought’ to do what natural justice considers intrinsically wicked, acting so you ‘may be a virtuous character’, and doing what ‘the good man’ would do.20 C: ‘It is not profitable . . . to do moral philosophy . . . until we have an adequate philosophy of psychology’. ‘What we need is an account of human nature, human action, what a virtue is, and most importantly human flourishing’.21 Here lies the redemption of moral philosophy, for only by recovering such accounts is there an adequate basis for the view, universally held from Plato to Mill, that certain kinds of action like murder are simply ruled out. Although Anscombe calls for a recovery of the virtues, her interest in MMP is with the recovery of one particular virtue, that of justice. In fact, her interest in justice seems to be the basis for reintroducing the notion of the virtues!22 Anscombe refers to justice over seventy times in MMP, and although ‘truthfulness’ is referenced as a virtue, there are no references to cardinal virtues such as courage and temperance. The final four pages of MMP are devoted to sketching the basics of an Aristotelian-​Thomistic

14 

MMP, 30. MMP, 37–​38. 16  MMP, 26. This is Anscombe’s third thesis. 17  MMP, 26, 42. 18  For Mill, ‘acts of murder . . . could [not] be otherwise described. . . . [If a] proposed action is of such a kind as to fall under some one principle [e.g., against murder] established on grounds of utility, one must go by that. . . . Mill assumes . . . there is no question of calculating the consequences of an action such as murder’ (MMP, 27, 33). 19  This is Anscombe’s corrected version of Sidgwick’s thesis, the original thesis stating ‘that it does not make any difference to a man’s responsibility for something that he foresaw, that he felt no desire for it, either as an end or as a means to an end’ (MMP, 35). 20  Anscombe further claims that Sidgwick’s viewpoint is necessarily shallow because it lacks a law or standard (except perhaps the standard of one’s circle of friends) by which one might even recognise ‘borderline cases’ (MMP, 36). 21  MMP, 26, 41. This is Anscombe’s first thesis. 22  MMP, 30. 15 

Justice and Murder    229 ethic of the virtue of justice.23 Her main concern, as already noted, it to recover the concept of murder understood in the light of the virtue of justice. Anscombe’s preoccupation with the concept of murder begins in 1956, sparked by her practical protest against a particular injustice, the proposal by Oxford University to give an honorary degree to Harry Truman. The way Anscombe sees it, since Truman was notorious to the world as a mass murderer, he should be ineligible for such an honour. As an elector in the vote to ratify or reject the giving of this honour, Anscombe considers it incumbent upon her not to cooperate in the evil of praising and flattering Truman by ratifying the Hebdomadal Council’s proposal to honour him.24 In writing up her pamphlet Mr. Truman’s Degree (TD) immediately after making her protest, Anscombe briefly conjectures a connection between a fundamental failing of the university’s leaders (and the leaders of British society more generally) to obey the requirements of natural justice, and the character of prevailing Oxford moral philosophies. Ten months later, after a rancorous exchange with the Oxford moral philosophers R. M. Hare and Patrick Nowell-​Smith, Anscombe considered her hypothesis confirmed, namely, that the prevailing Oxford moral philosophies were incapable of pronouncing mass murder, or any other class of action, as ‘beyond the pale’, unworthy of consideration. While these moral philosophers were typically very conventional, and might well pronounce certain actions as wrong, such pronouncements amounted to little more than the claims of the emotivists (which they rejected), because nothing in their moral philosophies allowed them to rule out as inadmissible any class of actions, whether it be torturing or slaughtering the innocent, judicially condemning the innocent to death, or any other horrible or despicable action. Starting with TD and continuing with ‘Does Oxford Moral Philosophy Corrupt Youth?’, Intention, MMP, and then in countless publications over the next decades, Anscombe would attend to various aspects of ethics and the philosophy of psychology necessary to recover the concept of murder. The more she wrote on the topic of murder, the more she realised that ‘murder is a moral action concept of great complexity’. Innumerable issues had to be addressed, such as ‘whether its definition includes its wrongfulness, and also whether and why capital punishment and some other deliberate killings under the authority of the state do not fall under the concept “murder.” ’25 The traditional absolute rejection of murder had been based on a conception of morality related to characteristics of persons. Formerly, as Anscombe noted, if a person was 23 

MMP, 38–​42. The Hebdomadal Council, composed of eighteen members elected by Congregation, was the chief policy-​forming committee of the university. Chaired by the vice-​chancellor, it met weekly during term. When Anscombe made her protest, one of the Council members was Janet Vaughan, the principal of Somerville College, who had originally hired Anscombe at Somerville and who went to great lengths throughout the 1950s to find funding to keep Anscombe employed at Somerville in a series of irregular appointments. See Vera Brittain, The Women at Oxford: A Fragment of History (London: George G. Harrap, 1960, 13); Maureen Owen, ‘Dame Janet Vaughan, D.B.E. 18 October 1899–​9 January 1993’, Biographical Memoirs of Fellows of the Royal Society 41, (1995), 493. 25  CP3, vii–​viii. 24 

230   John Berkman unjust, lacked the virtue of justice, one could conclude the person was bad qua human being. However, with the prevailing Oxford moral philosophies since Sidgwick, the formerly universally shared conception of the inherent wrongfulness of murder as injustice had disappeared. For W. D. Ross and R. M. Hare, the prevailing moral philosophers since Sidgwick, no kind of action was in itself inherently wrong. This essay proceeds in six sections. In the next two sections, I present Anscombe’s background, how commitments she made early in her life shaped her conviction that the killing of innocent persons was absolutely wrong. Her protest against Truman in 1956 would not have surprised those who knew her well, even though Anscombe did not have a permanent academic appointment in Oxford.26 While she took a courageous stand in objecting to Truman’s degree, her stand was neither novel nor unexpected in light of her background. I shall argue that her opposition to Truman was not even the most courageous moral stand that she had taken up to that point in her life. These sections look at relevant biographical information, some of it hitherto unknown. In the fourth and fifth sections, I discuss the various elements of Anscombe’s protracted campaign against Truman’s degree, which was more systematic, extensive, and strategic than has been recognised. As Anscombe made abundantly clear, the reason for her protest and her pamphlet was because in honouring Truman, her colleagues were showing that they had ‘lost the concept of ‘murder’,27 in that they refused to even acknowledge that Mr. Truman, in directly targeting and killing over 100,000 innocent women, children, elderly, and disabled, was a mass murderer. If Oxford was going to honour Truman, they should at least acknowledge they were honouring a mass murderer. One part of Anscombe’s campaign against Truman, her pamphlet Mr. Truman’s Degree, includes her first characterisation of ‘Oxford moral philosophy’, and this is the subject of the fifth section. While her first characterisation is only one paragraph long, it includes the main features of her later characterisations. Here, for the first time, Anscombe claims that the key feature which characterises the predominant moral philosophies in Oxford since World War I is their denial that any class of actions can be morally ruled out, the logical upshot of this denial being their willingness to ‘morally justify’ the mass murder of the innocent. Furthermore, A. E. Harvey of the talks department at BBC Radio found Anscombe’s brief but pointed characterisation of Oxford moral philosophy in TD ‘suggestive’, and his fascination with that one paragraph in Anscombe’s pamphlet would lead directly to her further three characterisations of Oxford moral philosophy, culminating in her treatment in MMP.

26 

Although Anscombe arrived at Somerville College, Oxford simultaneously with Foot in 1946, Foot received her ‘fellowship’ (the British equivalent of tenure) at Somerville in 1950, whereas Anscombe would remain on a series of makeshift appointments, receiving her fellowship only in 1964, at age forty-​ four, having been at Somerville for eighteen years. Then as now, a protest attracting worldwide attention consisting in negative publicity for Oxford University was not an advised career move for an ‘untenured’ faculty member. 27  TD, 69.

Justice and Murder    231 In the sixth and seventh sections, I examine her next two characterisations of Oxford moral philosophy, first in her February 1957 BBC address ‘Oxford Moral Philosophy: Does It Corrupt the Youth?’, and second, in an extensive and heated exchange with Hare and Nowell-​Smith in the pages of The Listener that followed upon her BBC talk. There Anscombe expands on how the predominant moral philosophies of Ross and Hare replace natural justice and the prohibition of murder with consequentialism.

10.2  A Youth Yearning for Social Justice 10.2.1  Embracing Catholicism’s Economic and Political Ideals Raised by an atheist father and a nominally Anglican mother, Anscombe encountered Catholicism at the age of twelve. Influences on her spiritual journey included the history of the English recusants, Bernard Boedder’s Natural Theology, Proust, Bernard Shaw (mostly the prefaces), and numerous works of Chesterton, especially his The Everlasting Man. Convinced of its truth by age fifteen, Anscombe refers to converting to Catholicism in the summer of 1934,28 although she would not be formally received into the Catholic Church until after her nineteenth birthday. At the time of her birth on March 18, 1919, Anscombe’s family was in Limerick, Ireland, her father having been posted there eight months earlier to provide ‘security duties’ for the British Empire in a predominantly Catholic country that was by then agitating for independence. Limerick was a Sinn Fein stronghold.29 Three weeks after Elizabeth’s birth, an attempt to rescue a wounded Irish Republican Army volunteer from a local hospital led to the killing of a police officer. The city was put into lockdown: a curfew was imposed and entry to the city tightly regulated. This in turn provoked a general strike in the city. At the same time, the second wave of the 1918–​1919 influenza epidemic was spreading in Limerick.30 In April 1919, Anscombe’s father, who had sustained severe injuries in France in 1917, was adjudicated fit for ‘sedentary duties’

28 

Elizabeth Anscombe, ‘I Am Sadly Theoretical: It Is the Effect of Being at Oxford’, Catholic Herald, July 1938. Republished in New Blackfriars 102, no. 1101 (2021), 724–​727. See also Amelie Rorty, The Many Faces of Philosophy (Oxford: Oxford University Press, 2004), 498–​502. 29  Robert Graves, Good-​bye to All That: An Autobiography (New York: Cape, 1930), ch. 26. Thanks to Prof. John Hayes for this reference. 30  For details on the pandemic, see Tom Toomey, ‘Limerick and the 1918–​1919 Spanish Influenza Epidemic’, Old Limerick Journal, Winter 2011, 24–​29. Thanks to Dr. Matthew Potter, director of the Limerick Museum, for this source.

232   John Berkman only, and at the end of May 1919 the family, including ten-​week-​old Elizabeth, returned to England.31 At her parents’ request, Elizabeth was baptised in the Church of Ireland (Anglican) Cathedral of St. Mary’s in Limerick on May 7, 1919; the rite was performed by the dean of the Cathedral.32 Although the Protestant community in Limerick was a minority, it was disproportionately represented in the merchant and professional classes. The Irish independence movement exacerbated long-​standing friction between the economically dominant Protestant minority and the Catholic, mainly working-​class majority. In contrast, as a middle-​class family, the Anscombes returned to an England where not only the social and political elites but the middle classes as well were almost exclusively Protestant. While England had a small and ancient Catholic aristocracy, there was practically no Catholic middle class in England, the vast majority of English Catholics being working class (and largely of Irish descent) and concentrated in the North.33 That Anscombe’s parents expressed their opposition to her interest in Catholicism by sending an Anglican priest to convince her to remain an Anglican is telling. For a middle-​class English family in 1935 London, ‘religious’ concerns with Catholicism were not likely to be clearly distinguished from political and class issues. In choosing to become a Catholic in England, Anscombe would be choosing to identify not only with a people of a different religious faith but also with a community marked by a lesser social and economic status, and with differing political ideals.34 In light of this, Anscombe’s parents’ opposition to her adolescent conversion to Catholicism should come as no surprise. As we shall see, as a young convert Anscombe wholeheartedly embraced Catholicism’s social, economic, and political ideals.

10.2.2 The Blackfriars Dominicans and Justice When Anscombe went up to St. Hugh’s College, Oxford in October 1937, she immediately sought out friends in Oxford’s small Catholic community.35 Besides the Old Palace, 31 

Captain Anscombe spent 1920 teaching at an army college in Cologne, Germany, and was decommissioned in January 1921. 32  My thanks to Prof. John Hayes for suggesting nuances to my account of the political situation in Limerick at that time. For more on the details of Anscombe’s brief time in Ireland, see John Hayes, ‘G. E. M. Anscombe—​Irish-​Born Philosopher’, History Ireland 28, no. 5 (2020), 42–​44. 33  As Alan White, OP puts it, ‘Before the Second World War English Catholicism was essentially a working class phenomenon. It had small upper class and aristocratic constituencies, but was lacking in a substantial middle class’. Alan White, ‘A History of Blackfriars and New Blackfriars’, New Blackfriars 77, no. 906, (1996), 324. 34  Jay Corrin notes that prior to World War II, in the UK only a few hundred Catholics received a degree in higher education each year. J. P. Corrin, ‘The English Catholic New Left: Battling the Religious Establishment and the Politics of the Cold War’, Social Sciences 7, no. 60 (2018),3. 35  In 1937 there were about 50 female and 160 male Catholic students at Oxford. See Walter Drumm, The Old Palace: A History of the Oxford University Catholic Chaplaincy (Dublin: Veritas Publications, 1991), 83, and Evelyn Waugh, Monsignor Ronald Knox (Toronto: Little, Brown and Company, 1959), 213.

Justice and Murder    233 Oxford’s Catholic Chaplaincy, which at that time ministered only to male students, two such communities were facilitated by Jesuits and Dominicans. The Jesuits operated St. Aloysius, the only Catholic parish serving central Oxford. The Dominicans had recently returned to Oxford—​having been kicked out by Henry VIII—​and re-​established Blackfriars.36 St. Aloysius and Blackfriars were located near to each other, both being on Anscombe’s way from St. Hugh’s into central Oxford. Speaking of his experience from the late 1940s, Brian Wicker wonderfully contrasts the Campion Hall Jesuits and the Blackfriars Dominicans. To feel the difference between . . . Campion Hall and . . . Blackfriars was to experience an almost tangible shock. Campion Hall was like something out of Henry James: a rich tapestry of dark browns and reds, complex and subtle in its visual as well as its intellectual organization. Its inhabitants wore their black suits and soutanes as a uniform: one had to dig quite deep to discover the person within. . . . Blackfriars, by contrast, was light and airy; a set of variations upon the theme of whiteness and transparency. The habits were worn in such a way that they expressed, rather than disguised, the individuality of the person inside: some were immaculate and pure as cream from a Jersey cow, others looked like tattered table-​cloths hanging on the shapeless frames of scarecrows or tramps. If you felt like an angel, then in Blackfriars you would look like one. If you felt like a down-​at-​heel beggar, then that is how you dressed too. And nobody minded—​or so it seemed to me. Yet clearly, the Dominicans could not have survived if they had been merely a collection of eccentrics: what was hidden was not the individual personality (that was on show all right) but the community organization of the ‘brethren’. Not the fruit, but the family tree, was the thing that must have been there, out of sight of the mere visitor, yet alive and growing all the same.37

The ‘organising’ prior of the merry band of eccentrics at Blackfriars was Fr. Hilary Carpenter, who was also editor of Blackfriars when Anscombe arrived in the fall of 1937. Anscombe chose Blackfriars for the instruction necessary to be formally received into the Church. In the late 1930s, Blackfriars-​style Catholicism, as expressed in its journal Blackfriars, singularly integrated Catholic spiritual, political, economic, and social teaching in intellectually sophisticated, unconventional, and non-​nationalistic ways. Its attitudes were in striking contrast to the politically engaged Catholic periodicals sponsored by the Jesuits and the English bishops.38 Anscombe’s nascent Catholicism was very much the 1930s Blackfriars variety.

36 

The Dominicans had had a presence in Oxford from 1221 until 1538, when Henry VIII dissolved all of the Catholic religious orders. 37  Brian Wicker, ‘Making Peace at Spode’, New Blackfriars 62, no. 733 (1981), 311–​320. 38  I am here referring specifically to The Tablet (owned by the British bishops until 1936 and then edited by Douglas Woodruff), and The Month (Jesuit). Other periodicals holding to the dominant Catholic political perspective at the time would be the Catholic Herald, The Dublin Review, The Universe, Colosseum, and Stella Maris (Campion House, Osterley). Those similar to Blackfriars included The Sower, The Cross and the Plough, and The Catholic Worker (London edition).

234   John Berkman

10.2.3 A Teenage Apologia for Justice In April 1938, Anscombe was received into the Catholic Church. She had just turned nineteen. Two months later, she wrote and got published a vivid picture of her Catholic outlook and commitments. Her brief teenage apologia was entitled ‘I Am Sadly Theoretical: It Is the Effect of Being at Oxford.’39 It already articulated some of the key themes that Anscombe would make famous twenty years later in ‘Modern Moral Philosophy.’ In this teenage essay, Anscombe first emphasises the need for what Bernard Williams later calls ‘thick description’ for imparting morality when doing ethics. As Anscombe puts it, ‘vaguely theorizing benevolence’ does not suffice. Nor does merely recovering the language of ‘justice’ and the other virtues per se so long as the use of that language remains abstract. Second, the need for exemplification shows the priority of witness, since we learn justice and the other virtues only by seeing them exemplified in the lives of just and virtuous individuals. ‘ “The just man justices,” yes, but what is justice in this thing or that?’40 Justice requires a series of exemplifications of a rich conception of justice. Third, the witness must be of something morally compelling, which for Anscombe is Catholic social doctrine in action. ‘We must be the first to accept the natural moral law, to deal justly, suppress usury, underselling, unjust prices and wages, to respect and increase the human dignity of the poor by restoring to them greater control over their own lives.’ In ‘I Am Sadly Theoretical’ Anscombe also shows a remarkable knowledge of and commitment to Catholic social doctrine. As far as she was concerned, what prevented wholesale conversion to Catholicism in England was the failure of regular Catholics to live the faith, in particular in the marketplace. Sadly, English society

39 

In the April 28, 1938, edition of the Catholic Herald, the foremost British Catholic weekly proposed a new feature titled ‘Under Twenty-​Five’. All Catholics of this age group were invited to contribute an article to the Herald, and they would be paid ‘the usual rates’. What did the Herald want? ‘What do they feel that life has in store for them these young men and women who have just left school? What are the things that give them hope? The things that make them despair? Catholic authorities on social questions, educational and otherwise would give a lot to know the answers. Detailed information on this matter should interest young and old alike. . . . From the highest born to the lowest; there will be no passing over of the office boy and girl earning 10s., nor will those be excluded from giving their opinions whom luck or outside influence has favoured.’ Potential authors were instructed to read at least one of the first two articles, each by a young journalist. These articles appeared beginning in May 1938 and continued through the end of that year. Anscombe was the tenth contributor to the series. 40  Here Anscombe alludes to the second stanza of Gerard Manley Hopkins’s ‘As Kingfishers Catch Fire’.

I say móre: the just man justices; Keeps grace: thát keeps all his goings graces; Acts in God’s eye what in God’s eye he is—​ Chríst—​for Christ plays in ten thousand places, Lovely in limbs, and lovely in eyes not his To the Father through the features of men’s faces.

Thanks to Benjamin Lipscomb for alerting me to this allusion.

Justice and Murder    235 could not identify Catholic employers as those who were sure to pay just wages, nor Catholic shopkeepers as those who could be relied on to charge fair prices. In her apologia, Anscombe goes so far as to suggest that the appropriate remedy for Catholic employers who maliciously (as opposed to ignorantly) oppress their employees is excommunication.

10.2.4 The Catholic Social Guild and Economic Justice While Anscombe was still an undergraduate, two other Oxford Catholic communities influenced her social justice commitments. Besides Blackfriars, there was the Catholic Social Guild (CSG) and the PAX Society, both of which had strong links with Blackfriars. Anscombe had become acquainted with the CSG by her first year at Oxford. The mission of the CSG was to educate working-​class Catholics about Catholic social teachings, especially as embodied in Rerum Novarum and other papal social encyclicals. The CSG was led by Fr. Leo O’Hea, SJ. Ironically, but unsurprisingly, O’Hea’s fidelity to Rerum Novarum and other papal teaching periodically got him in trouble with the English Catholic hierarchy. In 1931 O’Hea wrote: For though no Catholic society in Great Britain has had fuller approval from the Hierarchy, none, perhaps has met with greater criticism—​proof enough that the Guild deals with, and has some little influence upon the actualities of modern life. It is criticized for what it is said to teach and for what it does not teach.41

No wonder that O’Hea, famous as a lecturer, published little. In her apologia, Fr. O’Hea is the only individual mentioned, and the CSG is the only organisation referenced. Anscombe quotes from a lecture or conversation with O’Hea: To quote Fr. OHea, S.J.: ‘It wouldn’t occur to an employer to mention wages in the confessional—​he’d talk of distraction in his prayers rather, for wages go by the market price of labour.’42

Conversation between Anscombe and O’Hea was manageable, since O’Hea was principal of the Catholic Workers College in Oxford, located a few blocks from Anscombe’s college.43 She concluded her apologia by asking what ‘we as young Catholics are supposed to do?’

41  Leo O’Hea, ‘The Work of the Catholic Social Guild and Catholic Workers’ College’, in Il XL Anniversario Della Enciclica ‘Rerum Novarum’ (Milano: Società Editrice, 1931), 426. 42  Anscombe, ‘I Am Sadly Theoretical’, 725–​726. 43  The Catholic Workers College (out of which O’Hea also directed the CSG) had opened in 1922 at 3 Walton Well Road. It would be headquartered there until the Catholic Workers College (renamed Plater College) moved to Boars Hill in 1954.

236   John Berkman We have the Catholic Social Guild—​it is our duty to use it both to find out and to propagate the exact implications in detailed practice of papal social doctrine. In this we must avoid two things: mere vaguely theorising benevolence, and the claim that some private theory is papal doctrine, and that the church is committed to some particular political system. But once the truth is discovered it must be so loudly proclaimed, enforced, and practiced, that the outside world is left in no doubt that there is a concrete Christian claim and practice of social justice. (Anscombe, ‘I Am Sadly Theoretical’, 726–​727)

In Anscombe’s social justice–​oriented vision of Catholicism, she is committed to seeing working-​class Catholics formed in Catholic social teaching. At nineteen she was already eager to overcome vague notions of benevolence or justice, calling herself to decisive action: I have lived all my life among non-​Catholics and anti-​Catholics . . . I have little experience except of argument to report, and now long for something more decisive. . . . [T]‌here is enormous scope for activity in the future, and my chief ambition is to be doing something, or to be seeing something done, about it. (Anscombe, ‘I Am Sadly Theoretical’, 727)

In less than a year Anscombe would choose to do something beyond vague appeals to justice, something truly radical and almost unheard of for a Catholic. Something that English Catholic priests and the English Catholic hierarchy regularly condemned as wrong.

10.3  Anscombe’s Conscientious Objection to Injustice 10.3.1  The PAX Society and Justice in War From 1936 onwards, Blackfriars was filled with articles about the justice of war. In contrast to most of the English Catholic press, Blackfriars included a range of perspectives, emphasising the scholastic tradition of just war, a tradition which did not presume that one’s own nation’s proposed wars were necessarily to be deemed ‘just’.44 Blackfriars’s 44 

During World War II, one of the tasks of (the very patriotic) Cardinal Hinsley, Archbishop of Westminster, was to find chaplains to serve the British armed forces during the war. Peter Hebblethwaite writes, ‘[Hinsley] saw to it that the Forces were provided with chaplains . . . and while he had full confidence in the Jesuits and Benedictines, he said that “if I tried the [Dominican] Friars I might have got Pacifists or [Spanish] Republicans or other odd fish.” ’ Peter Hebblethwaite, ‘Into the Mainstream with Cardinal Hinsley,’ New Blackfriars 68, no. 807 (1987), 352–​353, quoting Thomas Moloney, Westminster, Whitehall, and the Vatican: The Role of Cardinal Hinsley 1935–​1943 (Tunbridge Wells, Kent: Burns &

Justice and Murder    237 commitment to mining the best of the Catholic tradition’s teaching on war had been severely tested by the Spanish Civil War, from which Blackfriars emerged as a rare intellectually significant Catholic periodical that for the most part refused to take sides.45 In 1936, a small Catholic peace organisation was founded, called the PAX Society. PAX’s early intellectual leaders were three Catholic laymen: the philosopher E. I. Watkin, the historian Donald Attwater, and the artist Eric Gill. All three had strong connections to the English Dominicans. Attwater and Gill were third-​order Dominicans, and all three regularly contributed to Blackfriars, which in turn regularly reviewed their published work. While likely only a minority of friars at Blackfriars supported that PAX perspective, it remains fair to conclude that Blackfriars Priory was the unofficial centre of the UK Catholic peace movement.46 Oates, 1985), 145. Despite Hinsley’s attitude, at least seventeen Dominicans served as military chaplains in World War II, including Hilary Carpenter, OP, the prior of Blackfriars. (My thanks to Fergus Kerr, OP, for providing a list of English Dominicans who served as military chaplains during World War II.) 45  While a September 1936 editorial in Blackfriars took a strongly anti-​Republican (i.e., anti-​leftist) stance, most editorials and articles in Blackfriars refused to take a side in the conflict. In Fighting for Franco: International Volunteers in Nationalist Spain during the Spanish Civil War: 1936–​1939 (London: Leicester University Press, 2001), Judith Keene argues that Cardinal Hinsley of Westminster was completely committed to Franco’s Nationalist side, and that most of the Catholic Press—​e.g., The Universe, Catholic Herald, Catholic Times Tablet, the Jesuit periodical The Month—​were focused on ‘Red Horrors’ perpetrated by the leftist government. Keene notes two exceptions; Blackfriars and the Catholic Social Guild’s journal Christian Democrat took neither side in the conflict (see 51–​52, 86n29). Keene could have also noted the British edition of the Catholic Worker. While Aidan Nichols claims that ‘the editorial policy of Blackfriars favoured the de jure (Republican) government,’ any perspective from a Catholic periodical less than openly for Franco would have been seen this way. Aidan Nichols, Dominican Gallery: Portrait of a Culture (Leominster: Gracewing, 1997), 48. 46  The PAX perspective maintained by Dominican friars stationed in Oxford, such as Victor White, Richard Kehoe, and Conrad Peplar, was likely a minority viewpoint at their priory in Oxford. However, sophisticated presentations of views on war and peace congenial to PAX were being published regularly in Blackfriars, especially by White and Gerald Vann, OP. Haldane suggests that it was Gerald Vann, OP, who particularly influenced Anscombe’s views on morality and war. See John Haldane, ‘Anscombe: Life, Action and Ethics in Context’, Philosophical News 18, no. 1 (2019), 45–​75. During the pre-​war period Vann published prolifically on war from a PAX-​ friendly perspective, in Blackfriars and numerous other academic and popular periodicals. However, unlike Victor White, Vann was not teaching in Oxford but at the Dominican preparatory school Laxton Hall in Northhamptonshire, a three-​hour drive north of Oxford. The ideological nature of Cardinal Hinsley’s aversion to the PAX viewpoint was revealed when Hinsley denied a nihil obstat to Vann’s 1939 book, Morality and War. To deny a nihil obstat for anything other than theological error (e.g., for a political viewpoint) would be an abuse of that power. Yet Vann’s book had already been approved by Dominican censors. Vann published some twenty or so books in his lifetime, and this was the only one denied a nihil obstat. Once World War II began, the Dominican provincial apparently took steps to counter and/​or silence the PAX perspective in Blackfriars. In November 1939, White’s monthly ‘Extracts and Comments’, a regular feature of Blackfriars for more than five years under White’s pseudonym ‘Penguin’, suddenly ceased. In March 1940, Blackfriars Prior and Provincial Bernard Delany, OP, wrote to Vincent MacNabb, OP, asking for a judgement on the orthodoxy of Gerald Vann’s January 1940 article ‘Patriotism and the Life of the State.’ After February 1940, neither Vann nor White had an article in Blackfriars for the next eight months, during which time Blackfriars published numerous articles on the war, including some commissioned from non-​Dominicans and military chaplains. These articles took a much more

238   John Berkman The mission of PAX was, first, to give potential or actual Catholic conscientious objectors (COs) moral support and, second, to give ‘practical support and professional help’ if these COs got into trouble.47 PAX provided this support based on its understanding of war in light of Catholic social teaching. Thus, a secondary purpose of PAX was to educate Catholic clergy and laity about war in light of natural reason and traditional morality. Drawing upon the just war tradition of Augustine, Aquinas, and Vitoria, Pax summarised its views as follows: On grounds of Catholic teaching, PAX stands for: refusal to take part in [unjust] war; the right to refuse all military service [i.e., to conscientious objection]; disarmament—​unilateral if necessary; social justice for its own sake and as an essential condition of peace.48

PAX members often referred to themselves as ‘just war pacifists’49 but vehemently distinguished themselves from mainstream ‘absolutist’ pacifist groups, in that PAX refused to reject all war outright.50 One particular view about war that set PAX apart from mainstream Catholic views was its refusal to grant to the state the final word on the justice of a war. If a Catholic discerned that his or her nation was embarking on a war that was unjust according to just war criteria, then morally speaking, that individual must refuse to take part.51 PAX’s view was that historically, the rise of nation-​states and their ‘state-​sovereignty’ claims to be final arbiters of the justice of their wars had debased just war theory, rendering it impotent. Opposing nations could both claim justice was on their side, without being subject to a higher moral authority. PAX challenged this dominant nationalistic view of just war, advocating two central principles: 1. While the use of force was permissible under some circumstances, wars ‘between nations for national ends’ were morally unjustifiable because the evils of them far outweighed any legitimate gains.

favourable stance on Britain’s war policy. In October 1940, Fr. Vincent MacNabb was commissioned to write an article on obedience in Blackfriars. Vann was also asked to contribute an article, which appeared following MacNabb’s article in that issue. My thanks to Fr. Richard Finn for providing me a copy of MacNabb’s March 5, 1940, letter to Delany, which can be found in the Dominican archives at Douai Abbey. For more on the early influences on Anscombe, see John Berkman, ‘The Influence of Victor White and the Blackfriars Dominicans on a young Elizabeth Anscombe’, New Blackfriars 102, no. 1101 (2021), 706–​723. 47 

The original PAX pamphlet, published in 1936. This summary of PAX’s aims was published on PAX’s stationery in 1938. See Valerie Flessati, PAX: The history of a Catholic peace society in Britain 1936–​1971 (doctoral dissertation, University of Bradford, 1991), 49. 49 Flessati, PAX, 5. 50  For example, because the UK’s Council of Christian Pacifist Groups was based on ‘absolutist’ pacifist conviction, PAX would not affiliate with the Council’ (Flessati, PAX, 27). 51 Flessati, PAX, 16. 48 

Justice and Murder    239 2. Those who recognize that national wars are unjust have the duty to refuse to take part in them and to do whatever they can to ameliorate the causes of such wars.52 The PAX Society understood its view to be consonant with the twentieth-​century papal social teachings which challenged this nationalistic view of war.53 However, these papal teachings were typically ignored or interpreted minimalistically by national Church hierarchies. In seeking to recover the classic just war tradition by appealing to papal social teaching, PAX was effectively setting itself up in opposition to the English Catholic Church hierarchy. For at that time, the ‘standard’ Catholic view was still that citizens could not make individual judgements about the justice of a war since they did not know all the circumstances. In the words of Canon Mahoney, English Cardinal Hinsley of Westminster’s theological consultant on the morality of war, if the state ‘calls upon its citizens to defend it, the pacifist attitude may not be adopted by any Catholic, or any other Christian, or indeed by any reasonable man, since the rational presumption is, unless the contrary is quite evident, that the State has come to its decision on just and sufficient grounds’.54 Since Catholicism had no tradition of conscientious objection, the PAX perspective was considered highly suspect. The kind of view enunciated by Mahoney was the view of the typical Catholic priest. In other words, a good Catholic could not be a CO, and Catholics who thought that they should be COs were either naïve or held a heretical theology.55 There were also reasons peculiar to the English context that encouraged the Catholic hierarchy to be reflexively supportive of Britain in any war. In the 1930s there were still numerous anti-​Catholic organisations fomenting prejudice that, e.g., Catholics were not trustworthy British subjects since their primary loyalty was to a foreign head of state.56 Like American Catholics, English Catholics were inherently suspect in a Protestant 52 

Quoted from the original PAX leaflet produced in 1936. See Flessati, PAX, 523. In 1940, the PAX Society issued the pamphlet War, Conscience, and the Rule of Christ, which consisted entirely of excerpts from the encyclicals of Leo XIII, Benedict XV, Pius XI, and Pius XII. 54  E. J. Mahoney, ‘Notes on Recent Work: Moral Theology,’ Clergy Review, January 1941, 68. Mahoney was quoting Henry Davis, S.J., a prominent English moral theologian. 55  For example, tribunal judges for cases of COs regularly noted that of all the various Christians groups, Catholic COs seemed to have the most difficulty getting their priests to speak on their behalf before the tribunal. Catholic priests were the most likely to refuse to offer assistance. If they did offer assistance, it would most often be to acknowledge the sincerity of the Catholic CO, while noting the CO was mistaken with regard to Catholic teaching. In addition, the major scholarly works on conscientious objection in Britain published after World War II typically made almost no reference to Catholicism. For example, Denis Hayes’s 1949 Challenge of Conscience makes no mention of Catholic COs. Those mentioned in the book who are likely Catholic (i.e., Irish COs) seem to have been COs on political rather than religious grounds. In fact, the typical narrative of these ‘ethnic’ Catholics who became COs is that they left the Catholic Church over the issue. 56  One of these organisations was known as CIVIC, the Council for Investigation of Vatican Influence and Censorship, whose goal was to show that Catholics were predisposed to follow the commands of Rome and not those of British civil authorities. See Joan Keating, ‘Discrediting the ‘Catholic State’: British Catholics and the Fall of France,’ in Frank Tallett and Nicholas Atkin, eds., Catholicism in Britain and France since 1789 (London: The Hambleton Press, 1996), 27. 53 

240   John Berkman country, only more so because of the particular history of English recusants and the (sometimes justified) accusations of disloyalty or even treason levelled against them. Cardinal Hinsley, the de facto spokesperson for English Catholicism, saw a golden opportunity for Catholics to again demonstrate their loyalty to Crown and Country: ‘No matter how great our hatred of war, we cannot stand idly by and allow our neighbour to be ruthlessly enslaved or done to death. We must have confidence in our King and his counsellors, our lawful rulers.’57 Accordingly, PAX’s message was not welcomed by the English Catholic hierarchy. Soon after PAX’s founding, any priests who had joined were ordered to resign. Despite this ban Victor White and Gerald Vann, the most prolific contributors to the journal Blackfriars during the pre-​war period, assured PAX of their willingness to offer informal assistance.58 The highest-​profile work of PAX in 1938 and 1939 was a series of pamphlets authored by eminent theologians, all of which appealed in varying degrees to recent papal teachings. In 1938 Nicholas Berdyaev and Luigi Sturzo contributed pamphlets on justice in international law, and Gill a pamphlet on wealth and greed as the major cause of war. The following year Attwater, Watkin, and Gerald Vann contributed pamphlets.59 By the spring of 1939, once her Honour Moderations examinations were done, Anscombe’s concern for justice naturally turned to thinking about the impending war. For one thing, her fiancé, Peter Geach, would be having to make a decision about participation in any forthcoming war. One thing that is clear is that neither Anscombe nor Geach thought that Britain would fight what the Catholic tradition considered a ‘just war’, and everyone who understood this was morally required to become a CO. Geach chose to become a CO.60 57  Thomas Moloney, Westminster, Whitehall, and the Vatican: The Role of Cardinal Hinsley 1935–​1943, 134. Nichols adds that unlike other British bishops, who always spoke in their capacity as shepherds to their flocks, ‘Cardinal Hinsley often spoke as a national leader first . . . not as a bishop teaching his own flock but as a Catholic Englishman testifying to his sense of spiritual communion with Englishmen outside the fold’ (Dominican Gallery, 44). 58  Most of Victor White’s contributions to Blackfriars in the 1930s were under the pseudonym ‘Penguin,’ black and white being the colours of the Dominican tunic and cloak. As Penguin, White contributed to Blackfriars on topical—​often political—​questions almost every month from January 1934 to October 1939. For White as Penguin, see Adrian Cunningham, ‘Victor White: A Memoir,’ in A. C. Lammers and A. Cunningham, eds., The Jung-​White Letters (London: Routledge, 2007), 311. For Vann’s offer of assistance to PAX, see Flessati, PAX, 48–​49, 52. 59 Flessati, PAX, 47. The pamphlets were Nicholas Berdyaev’s War and the Christian Conscience (1938); Luigi Sturzo’s Morality and Politics (1938); Eric Gill’s And Who Wants Peace? (1938); E. I. Watkin’s The Crime of Conscription (1939); Donald Attwater’s Bombs, Babies, and Beatitudes (1939); and Gerald Vann’s Common Sense, Christianity and War (1939). 60  Geach and Anscombe met in June 1938 and were engaged by the end of 1938. See Luke Gormally, Moral Truth and Moral Tradition (Dublin: Four Court Press, 1994), 2. While a Balliol College undergraduate, Peter Geach held radical political views. From January to June 1937 Geach was regularly in the international press as the leader of the Oxford Jacobites. With Edward VIII having just abdicated, Geach ‘denounced a certain George Windsor as a pretender to the throne’ and declared that Prince Ruprecht of Bavaria was the rightful king of England. More than once during those months Geach publicly proclaimed before a crowd, ‘Whereas, it is manifest from the late unhappy events that the House

Justice and Murder    241 Coincidently or not, for each of the next three terms, from April 1939 until March 1940, Anscombe would be tutored by either Victor White or Donald MacKinnon.61 White and MacKinnon happened to be the two most formidable advocates of the PAX viewpoint then teaching in Oxford.62 During that period Anscombe became a member of PAX, became deeply involved with the Oxford PAX chapter, and co-​authored with fellow PAX member Norman Daniel The Justice of the Present War Examined, a pamphlet they prepared for publication by the Oxford chapter of the PAX Society.63 It is not surprising that Anscombe, a fired-​up convert keen to ‘take practical action’ should expend her energies on the PAX Society. of Windsor cannot defend the dignity of the Crown and the liberties of the people against the common enemies of King and People: and whereas the office of kingship can in no wise claim the crown of these realms: and whereas there is none other who may rightfully be our king except only the royal Prince Ruprecht of the House of Stuart. Now, therefore, we do hereby with one voice and consent of heart and tongue publish and proclaim that the Royal Prince Ruprecht is our most gracious Sovereign Lord Ruprecht, by the grace of God of England, Scotland and Ireland King Defender of the Faith: to whom we do acknowledge our absolute allegiance; and we do beseech Almighty God to grant that by His Majesty’s speedy and glorious restoration a true king may rule once more among a free people. Given in His Majesty’s loyal city of Oxford. . . . God save the King.’ Geach added, ‘The crown would be offered Ruprecht only in event of a successful Jacobite rebellion, for which I pray all the time. . . . I’m afraid there will have to be some violence—​though a lot can be done by propaganda. If the King cannot be restored except by bloodshed, blood must be shed.’ While no doubt Geach’s being an Oxford undergraduate allowed the authorities to largely dismiss his words as a prank, Geach’s pronouncements over a five-​ month period were clearly more than that. See, e.g., The Scotsman, February 1, 1937, 18; the March 7, 1937, editions of the New York Herald Tribune (A4), Boston Globe (A2), and Hartford Courant (A20); Chicago Tribune, May 9, 1937, 3; and Vancouver Sun, June 8, 1937, 2. 61 

That Anscombe devoted the lion’s share of her efforts that year to her tutorials with Victor White and Donald MacKinnon (as opposed to tutorials with her other tutors during that period) is plain from a comparison of the tutorial reports sent to her college at the end of each term. Victor White’s summary: ‘I have been completely satisfied with Miss Anscombe’s work for me. She has evidently worked hard, & I find her quick & intelligent.’ Donald MacKinnon’s summary evaluation was that ‘[Anscombe] has an unusually clear and definite grasp of her own ideas. She seems to give every prospect of doing first class work in this subject.’ In contrast to the evaluations of White and MacKinnon, her three other tutors during that period said ‘she is really more than a little lazy’; ‘Miss Anscombe . . . has only produced one essay, which . . . was rather slight. . . . Miss Anscombe might fail to cover enough of the necessary ground in Greats’; and ‘Of Miss Anscombe I hardly know what to say. On the two occasions when I have seen her, she has . . . been singularly reluctant to express any definite view.’ With acknowledgement to the St. Hugh’s College archives. 62  For Donald MacKinnon’s advocacy of the PAX perspective with regard to war, see André Muller, ‘Donald M. MacKinnon: The True Service of the Particular, 1913–​1959’ (doctoral dissertation, University of Otago, Dunedin, 2010). 63  Anscombe and Daniel became dues-​paying members of the PAX Society on December 2, 1939, along with sixteen other Oxford students. The March 1940 issue of the PAX Bulletin said, ‘The Oxford branch of PAX is shortly bringing out a pamphlet, The Justice of the Present War Examined,’ so it was published at the earliest in late March. The next issue of the PAX Bulletin, in May 1940, recommended Anscombe and Daniel’s recently published pamphlet. By then Anscombe claims to have withdrawn it from publication. However, considering the December 1940 issue of the PAX Bulletin says that their pamphlet is ‘particularly recommended to COs preparing statements,’ one has to wonder how diligent Anscombe was in withdrawing it. (PAX membership records and references to the PAX Bulletin are courtesy of the PAX Society archives.)

242   John Berkman As a teenage convert devoted to the Catholic faith, including its hierarchical structure and social justice teachings, for Anscombe to commend conscientious objection (and Geach to become a CO)64 and embrace a marginal position on a most controversial issue shows not only the influence of Blackfriars, the CSG, and PAX but also her incredible self-​confidence and strength of character.65 For while England was unusually accommodating to COs, ‘conchies’ in England during World War II were still social pariahs, fired from government and private-​sector occupations, regularly assailed in the British press, and periodically assaulted in rural areas.66

10.3.2 The Argument of The Justice of the Present War Examined In early April (or possibly late March) of 1940 Anscombe and Daniel published The Justice of the Present War Examined: A Criticism Based on Traditional Catholic Principles and on Natural Reason.67 In the preface they emphasised that the pamphlet was a ‘team effort’, that it represented the discerned conclusions of members of the Oxford PAX Society: This pamphlet tries to delineate what the attitude and behaviour of the Christian should be in the present situation. It presents the results achieved in a series of open discussions held at Oxford both before and after the declaration of war. . . . The application of these principles is extended to a condemnation of the war. . . . Those concerned are grateful to the association ‘PAX’ to which many of them belong, and which has constantly upheld an interpretation of the teaching of the Church similar to that expounded here.

64  Luke Gormally tells the story, ‘When Peter appeared before a tribunal judge to plead conscientious objection he was asked was there any kind of war he believed met the just war criteria. His answer I am told was ‘the Crusades’ (I don’t know whether he specified a particular crusade).’ Email to the author, August 24, 2020. 65  She did not likely get support from her military veteran father, nor her older twin brothers, one of whom joined the RAF, the other training for Anglican ordination. 66  For attitudes in Britain towards COs during the early years of World War II, see John Tulloch, ‘The Return of the ‘Conchie’: Newspaper Representations of Conscientious Objectors and Pacifists in World War II,’ in S. Gibson et al., eds., Representations of Peace and Conflict (London: Palgrave Macmillan, 2012), 45–​66. 67  While the pamphlet was published either in late March or early April 1940, it is harder to indicate precisely when Anscombe and Daniel wrote it. However, at the very least, a major part of the pamphlet must have been written between December 1939 and March 1940. For Anscombe and Daniel’s pamphlet drew extensively on Pope Pius XII’s Summi Pontificatus, quoting from it seven times, and three of the quotations were taken from Ronald Knox’s idiosyncratic translation. While Summi Pontificatus was published in October 1939, the Knox translation, titled ‘Darkness over the Earth,’ was not published until late November 1939.

Justice and Murder    243 Their two-​part pamphlet adhered to an orthodox PAX viewpoint: part 1, titled ‘War and the Moral Law’, explained how in light of the classic just war criteria the British were fighting an unjust war; part 2, titled ‘War, Society, and the Individual’, argued why Catholics must then be COs. In the first part Anscombe and Daniel argue that the stated policy of the British government had led the government to violate and/​or intend to violate three just war criteria. The most serious violation is ‘unjust means’, i.e., the British government’s stated direct intent under certain circumstances to murder innocent civilians.68 For example, Britain’s and France’s naval blockade, which included all food products, was intended to starve the entire German population.69 Furthermore, Britain has declared its intent, under certain conditions, to directly target non-​combatants (i.e., innocent civilians) in bombing campaigns. While some sinful government actions would not render a war unjust, the British government’s stated policy of unjust deliberate killing is murder and this is a great sin which ‘cries to heaven for vengeance’; if therefore, the warring state intends, under any circumstances, to commit it as a means of prosecuting the war, then the war is made wicked.70

In the second part of their pamphlet Anscombe and Daniel emphasise that each and every Catholic must determine in conscience if the war is unjust. Since no state can require that ‘we write a blank cheque on our consciences’, if a Catholic determines a war is unjust, he or she sins in participating in the war.71 A government that demands that its citizens sin is a state asking to be worshipped.72 Catholics must defy a state which demands its citizens sin in the certainty of its own righteousness.73 68  The three conditions that Anscombe and Daniel say are violated are that the British government does not have upright intentions as to (a) the means and (b) the ends in fighting the war and that (c) the good effects will not outweigh the probable evil of the war (JPW, 73). That they are primarily concerned with the violation of ‘just means’ is clear not only from the emphatic language used but also by the fact that whereas in the original pamphlet the discussion of ‘just aims’ (74–​75) and ‘good vs. bad effects’ (79–​ 80) merit less than three pages of discussion each, they devote almost ten pages to a consideration of ‘just means’ (75–​79). 69  The British government was already ‘blockading Germany with the intent to starve the population. The present war is therefore wrong on account of means’ (JPW, 79). 70  JPW, 79. This condition of ‘just means’ plays two distinct roles in just war theory. In some cases, it renders a war intrinsically unjust. In other cases, one can balance the likely good effects of the war with the likely bad effects. However, such balancing cannot be employed to justify the deliberate killing of non-​combatants. ‘If the choice lies between our total destruction and the commission of sin, then we must choose to be destroyed’ (JPW, 79). 71  JPW, 73; JPW-​WSI, 34. In CP3, Anscombe republished only the first half of their 1940 pamphlet The Justice of the Present War Examined, even though the entire pamphlet was jointly authored by Anscombe and Daniel. References to the second half of the pamphlet, which was subtitled The War, Society, and the Individual and has never been republished, will henceforth be cited as JPW-​WSI with page references to the original pamphlet. 72  They appeal to the Fribourg conventus, which instructs Catholics to reject the modern doctrine of state sovereignty, as it is in harmony with neither reason nor the doctrine of the Church (JPW-​WSI, 28). 73  JPW-​WSI, 31.

244   John Berkman JPW draws on Aquinas’s account of human nature, justice, and actions.74 Human beings are to realise their potential by flourishing according to their nature. The natural law is in turn a reflection of divine law. Revelation guides human reason towards doing the right thing.75 Just relationships in a society and between individuals, as social animals, make possible and constitute a central part of the flourishing of human beings.76 The justice of acts by governments and/​or individuals are evaluated by their intentions. Intentions are evaluated in terms of ‘the nature of the completed acts themselves’, their motives, and their effects.77 Anscombe and Daniel draw on this account of human actions to refute specious claims to justify killing innocent civilians, especially ‘Catholic’ appeals to the principle of double effect to justify such attacks. They argue that neither blockades (‘which are intended to starve the entire population, non-​combatants together with combatants’) nor the direct intention to bomb civilians could ever be justified by the principle of double effect.78 As for the government apologists who claim it impossible to morally distinguish (non-​combatant) civilians from combatants: To this there is only one reply. The civilian population behind an army does not fulfill the conditions which make it right to kill a man in war. Civilians are not committing wrong acts against those who are defending or restoring rights. They are maintaining the economic and social strength of a nation, and that is not wrong, even though that strength is being used by their government as the essential backing of an army unjustly fighting in the field.79

Finally, while one of the conditions for a just war was that the good effects must outweigh the bad effects, they replied: While it is true that one must balance the likely good effects of the war with the likely bad effects, the ‘balancing’ cannot justify intentionally unjust acts. ‘If the choice lies between our total destruction and the commission of sin, then we must choose to be destroyed.’80

74  Anscombe uses traditional Catholic sources and pontifical documents to make her argument. She cites John Eppstein’s Catholic Tradition and the Law of Nations (London: Burns Oates & Washbourne Ltd., 1935), a five-​hundred-​page compilation of authoritative Catholic Church statements on issues of international order, and particularly on war, along with numerous references to Aquinas’s Summa Theologiae, and an extended reference to Pope Pius XII’s first encyclical, ‘Darkness over the Earth’ (Summi Pontificatus), translated by Ronald Knox and published by the Catholic Social Guild and the CTS in late November 1939. 75  JPW, 72. 76  JPW, 73. 77  These criteria for evaluating human acts correspond to Aquinas’s criteria of evaluating acts in relation to their objects, ends, and circumstances. See Aquinas, Summa Theologiae, I-​II. 78  JPW, 78–​79. 79  JPW, 77–​78. 80  JPW, 79.

Justice and Murder    245 As noted earlier, Anscombe and Daniel’s advocacy of the PAX view regarding the duty of individuals to exercise their conscience with regard to whether they can fight in a war was perhaps the most significant break between PAX’s stance and the dominant Catholic perspective on war at that time.81 Contrary to a typically argued view, they insisted that individuals must make their own judgement and were not morally permitted to unequivocally trust the state, and then blame it if it was either proposing to and/​or engaging in an unjust war.82 Anscombe and Daniel were aware of the difficulty of the path of the CO, noting how ‘grievous it may seem to stand apart from our fellow-​countrymen.’83 Their stance was notably courageous, since at the time they published their pamphlet, the Nazis had captured Denmark and Norway and were preparing to overrun Belgium, the Netherlands, and France.84 But they concluded with a hopeful note, saying that in resisting war there is plenty to do, to direct our will to proper activity in our society.85 The proper activity is seeking peace in all that we do, to seek the union of all people’s wills under God.

10.3.3 The Publication of The Justice of the Present War Examined The PAX Society, while not formally a Catholic organisation, had always sought Catholic ecclesiastical approval (i.e., a nihil obstat) for its publications. Once the war began, realising such approval would be denied, PAX had to decide whether to publish Anscombe and Daniel’s pamphlet without it. Unresolved at their March 1940 meeting, at the April 5th meeting PAX decided to cease publications under its banner.86 For this reason Anscombe and Daniel published JPW only under their own names. While JPW was thus not formally a PAX pamphlet, in their preface Anscombe and Daniel were effusive in expressing gratitude to the PAX Society, which ‘constantly upheld an interpretation of the teaching of the Church similar to that expounded here.’

81 

The last three sections (12–​14) of the pamphlet are titled ‘On the Duty to Resist the Present War,’ ‘On the Individual’s Duty to Society,’ and ‘The Individual Problem of Conscience’ (JPW-​WSI, 29–​37). 82  JPW-​WSI, 30. 83  JPW, 73. 84  By that point of the war (April 1940), the rate of conscientious objection to military service in Britain, even amongst the religious traditions favourable to it, was almost nil. See Denis Hayes, Challenge of Conscience (London: George Allen & Unwin, 1949), 382. 85  JPW-​WSI, 32. 86  Another pamphlet under consideration for publication at the same time as Anscombe and Daniel’s was Mark Fitzroy’s War, Conscience, and the Rule of Christ: Compiled from the Encyclicals of Leo XIII, Benedict XV, Pius XI, and Pius XII. Fitzroy’s pamphlet consisted entirely of quotations from various popes on war and conscience. It was eventually published as a PAX pamphlet, but its publication was delayed for some two years.

246   John Berkman After the April 5 meeting, Anscombe and Daniel moved expeditiously to publish their pamphlet, making it available for sale in bookstores and newsagents in Oxford and London in April 1940. Word of their pamphlet soon got to the archbishop of Birmingham, who in trying to find the authors tellingly chose to write to the prior of Blackfriars. Archbishop’s House, Edgbaston, Birmingham, 15 1st May 1940

My dear Father, Do you know the two young people Elizabeth Anscombe and Norman Daniel who have produced a pamphlet on the present war and are trying to make folk pay 6d for it? I’m told that they are Catholic, and there are lots of references to St. Thomas in the pamphlet. But they have had the pamphlet published without submitting it to ecclesiastical authority, or at least without having official approval of authority, and I don’t know whether this is done through ignorance or by design. If you know them you may be able to tell me whether they are acting innocently, or are deliberately taking a line opposed to that of the Hierarchy of this country. I’m just recovering now from my duodenal ulcer. What about your election? A blessing to you. Yours devotedly in Christ, ✝​Thomas, Abp. of Birmingham87

The bishop obviously got a response to his liking, as his follow-​up letter indicates. Archbishop’s House, Edgbaston, Birmingham, 15 3rd May 1940

Dear Father Prior, (And I am glad for Blackfriars’ sake that you still are Prior.)88 I thought you would be the right man to give me a sensible answer to those questions. I did not ask you whether your Fathers were connected in any way with the pamphlet because I did not want to insult you, and I think I know Dominicans better than that. But all the same I am grateful for your vigorous clearing of the air. . . . God bless you and at Blackfriars. Yours devotedly in Christ, ✝​Thomas, Abp. of Birmingham

The fact that these letters to Prior Hilary Carpenter, OP wound up in the possession of Anscombe and in her archives is some indication that the relationship between the archbishop and Blackfriars was not as sympatico as it appears on the surface. Word was 87  Box 14, File 588, Collegium Institute Anscombe Archive at the University of Pennsylvania, Kislak Center for Special Collections, Rare Books and Manuscripts. 88  According to the Catholic Herald on May 12, 1940, Carpenter was re-​elected between April 30 and May 2, so that may account for the bishop referring to Carpenter as ‘Father’ in the first of the two letters.

Justice and Murder    247 communicated to Anscombe and Daniel to withdraw their pamphlet, which Anscombe says they dutifully did.89 Anscombe was left with sufficient copies such that, fifteen years later, she was able to give them away to those who requested.90

10.4  Conscientiously Objecting to the Injustice of Murder in Wartime: Mr. Truman’s Degree 10.4.1  Anscombe’s Attitude to ‘Oxford Moral Philosophy’ As a student and then protégée and friend of Wittgenstein from 1942 to 1951, Anscombe published very little, focusing on learning from Wittgenstein how to think philosophically and engaging in endless conversations, both with him and innumerable other philosophers.91 In addition, at least from 1948 onwards Anscombe was focused on learning from Wittgenstein his Austrian German and how to translate his Philosophical Investigations. In the five years after Wittgenstein’s death in April 1951, Anscombe’s labours focused her work as one of his literary executors, sorting through Wittgenstein’s Nachlass, and editing, translating, and publishing his Philosophical Investigations and what would be called Remarks on the Foundations of Mathematics.92 In the 1940s, besides being mentored by Wittgenstein, Anscombe engaged in innumerable philosophical conversations both in Cambridge and in Oxford. From the mid-​ 1940s onwards she attended the meetings of the Moral Sciences Club in Cambridge, the Jowett Society and Socratic Club in Oxford, and courses on university lecture lists, such as the class offered by J. L. Austin and Isaiah Berlin on perception, titled ‘Things.’93 Furthermore, from the time she returned to Oxford to take up a research fellowship at Somerville College in 1946, she was in conversations about the problems with Oxford moral philosophy with old undergraduate friends; first with Philippa Foot, and shortly thereafter with Mary Scrutton (Midgley) and Iris Murdoch. In the late 1940s, they discussed extensively what was wrong with Oxford moral philosophy and how they 89  Since the Christmas 1940 issue of the PAX Bulletin was still recommending JPW, noting that it is ‘particularly recommended to COs preparing statements,’ one has to wonder to what extent Anscombe and Daniel ‘withdrew’ their pamphlet from circulation. 90  Anscombe, letter to Charles Thompson, June 2, 1956. Courtesy of PAX Archives. 91  For the evidence that Anscombe first met (and attended the lectures) of Wittgenstein in 1942, see James Klagge, ‘The Wittgenstein Lectures Revisited,’ Nordic Wittgenstein Review 8, nos. 1–​2 (2019), 48, 51. 92  For the story of Anscombe’s work on editing Wittgenstein’s Nachlass, see Christian Erbacher, Wittgenstein’s Heirs and Editors (Cambridge: Cambridge University Press, 2020). 93  Anscombe’s participation in Austin and Berlin’s seminar, offered in Michaelmas 1946, as well as her general disdain for Austin, is recounted in Mary Warnock, A Memoir (London: Duckworth, 2000), 68–​73.

248   John Berkman might change it.94 As Midgley put it, ‘[T]‌his was when we all hammered out our various thoughts on that topic, a lot of which we later published.’95 Their objections to Oxford moral philosophy were many and various. Murdoch excelled at placing these philosophers in context: What were their goals? What were they afraid of? According to Midgley, the primary fears of Oxford philosophers were to be seen as either weak or judgemental. So their work became narrowly focused and conventional, eschewing practical moral judgements, i.e., ‘morality’.96 Though these philosophers restricted their focus to ‘ethics’, i.e., to the linguistic and conceptual analysis of a narrow and reductionistic range of moral notions, their substantive (but hidden) moral viewpoints would be exposed and challenged in succeeding years by Murdoch, Foot, and Anscombe. While Murdoch’s and Foot’s criticisms of and attitudes towards Oxford moral philosophy were pointed but polite, Anscombe’s attitude was scathing. In the genteel culture of Oxford moral philosophy, Anscombe was the bull in the china shop. Her abrasiveness was to some degree simply a part of her persona, reinforced by her interactions with Wittgenstein.97 But Anscombe’s anger towards Oxford moral philosophy also had a moral basis. Oxford moral philosophy was frivolous. In 1947, Murdoch wrote with admiration and awe of Anscombe’s moral seriousness and ‘ruthless authenticity.’98 Ethics ought to clarify, rather than obfuscate, serious questions about morality. Anscombe’s scathing attitude to prevailing Oxford moral philosophy was already fully formed by 1950, as described by O. K. Bouwsma after one of their conversations:99 What interests me is her attitude towards philosophers at Oxford. . . . She thought I would find the atmosphere depressing. Cheap. Cheap. There aren’t any problems.

94 

The only year that all four of them were living in Oxford was 1948–​1949, Murdoch arriving in 1948 and Midgley departing in 1949. In the late 1940s, Anscombe was living only part time in Oxford. Her husband and two children were living in Cambridge, where the ‘family home’ would be until after Wittgenstein’s death in 1951. 95  Mary Midgley, Owl of Minerva (Abingdon: Routledge, 2005), 147. 96  For further insight on 1940–​1950s Oxford moral philosophy, see Midgley, Owl of Minerva, 124ff., 156ff. 97  Midgley thinks that Anscombe’s moral seriousness was given an ‘edge’ as she came under Wittgenstein’s influence (Owl of Minerva, 115). Wittgenstein was a tortured soul with little tolerance for philosophers whom he considered frivolous, which seemed to be practically all of them in Cambridge and Oxford. According to Midgley, who knew Anscombe through the 1940s, Anscombe seemed to increasingly share Wittgenstein’s disdain for most Oxbridge philosophers the more she studied under him. 98  Murdoch wrote this in her journal in July 1947, and in 1958 also recalled how much the moral seriousness of Anscombe and Yorick Smythies had meant to her in the late 1940s. See Peter Conradi, Iris: A Life (New York: W.W. Norton, 2000), 273, 284. 99  Beginning in the fall of 1950, O. K. Bouwsma was in Oxford for the academic year to give the John Locke Lectures. Bouwsma had been a student of Moore, and it was the significance of Bouwsma’s writing on Moore that had led Ryle to make Bouwsma the first non-​British John Locke lecturer. Unbeknownst to Ryle, Bouwsma’s interests had moved from Moore to Wittgenstein. With that shared interest, Bouwsma and Anscombe often met for tea and conversation in 1950–​1951.

Justice and Murder    249 And these people are all above it. They have easy dispositions for all difficulties . . . a certain shallowness which goes with their cleverness. It’s as though they feel they have the key to all mysteries and feel now a delight in their unmasking of the emptiness of all pursuits. Wittgenstein talks about [some] men as serious and deep. Perhaps it’s just that these [clever but shallow] men strike Miss Anscombe as like magicians who with a certain trickery and sleight of hand expose the poor ninny philosophers whom they seize upon. The ninny philosophers may not have had the benefit of borrowed cleverness, but they were very earnest, they had problems to which they gave their lives and hard labor. These people have nothing to do but debunk. . . . It isn’t then that these people are mistaken in what they say. It is that they have nothing but this show they put on. What a clever boy I am! Wittgenstein talked too about his own work. ‘[Philosophy]’s not important . . . I don’t recommend it. It’s for people who cannot leave it alone.’ So this is not important. What is important must fall outside. And suppose there is nothing outside! Poor souls! Very well, these other philosophers made mistakes, in earnest, but what now are you doing in earnest? There you are crowing over the mistakes of earnest men. So you will never make an important mistake, for nothing is important to you. Wonderful! Crow!100

These ‘crowing’ philosophers claimed that questions of personal ‘morals’ were independent of their theorising about ‘ethics.’ This allowed for ‘plausible deniability’ with regard to any particular view on ‘morals’ that seemed to follow from their theorising. Unfortunately for them, Anscombe recognised that their meta-​ethical claims in fact required particular moral viewpoints, and when the time came, Anscombe was determined to smoke them out. But why did Anscombe not ignore the frivolity of Oxford moral philosophy and simply get on with her work? She chose not to ignore it for the same reason she would choose not to ignore Mr. Truman’s proposed honorary degree. During March 1956, the Hebdomadal Council announced its proposed list of honorary degree candidates, one of whom was former US president Harry Truman.101 One of its many responsibilities was putting forward nominations for honorary degrees to be voted on by Congregation. An unusual feature of Oxford honorary degrees at the time was that the Congregation, the sovereign body of the university which acts as its parliament, was required to vote to ratify all honorary degree recipients after the proposed recipients were announced. In 1956, the vote to approve honorary degree recipients was scheduled for May 1, only seven weeks before the June 20 Encaenia, the annual event at which the majority of Oxford honorary degrees were awarded.

100  I am most grateful to Prof. Ronald Hustwit for sharing these notes with me. These are notes made by Bouwsma of his conversation with Anscombe on September 25, 1950. This comes from a notebook from which were also drawn the material for the book O. K. Bouwsma, Wittgenstein Conversations 1949–​ 1951, ed. J. L. Craft and Ronald E. Hustwit, Indianapolis (Hackett, 1986). These notes can be found in the Bouwsma archives of the Ransom Center at the University of Texas. 101  The Hebdomadal Council was Oxford’s chief executive body from 1854 to 2000.

250   John Berkman This announcement appalled Geach and Anscombe, as they considered Truman notorious as a mass murderer. Geach urged Anscombe to protest the proposal to honour Truman.102 The obstacle facing Anscombe was that much of her society did not see Truman as a mass murderer, much less a notorious one. So from April through June 1956, Anscombe prepared and executed three local protests, adroitly campaigning against Britain’s ‘loss of the conception of murder.’103 Her aim was to persuade her Oxford colleagues to recognise Truman as a mass murderer, and as a result refuse to cooperate in ‘the evil of praising and flattering’ him either by voting for him for the degree or attending the degree ceremony. While Anscombe’s campaign apparently failed in its immediate aims, her protests had side effects of far greater significance than she could have possibly imagined. For it turned out that the image of a youthful female Oxford don making a solitary protest before such an august body as the Ancient House of Congregation at Oxford University captured the imagination of the press worldwide. Her protests brought far greater attention to the murderous injustice of Truman’s order to drop atomic bombs on Hiroshima and Nagasaki than anything Anscombe could have hoped to attract through her philosophical writings. While Anscombe’s May 1, 1956, call for a non placet (it does not please us) vote from Congregation was almost unanimously voted down, her speech made her world-​famous, more so than she would ever be again. Her protest and portions of her speech were picked up in an untold number of newspapers around the world, large and small. Afterwards, Anscombe received dozens of letters from around the world.104 Furthermore, her pamphlet, written and published privately for her Oxford colleagues, would, despite its limited availability, eventually influence some of the most significant philosophical work of the following generation on the morality of war (e.g., Walzer and Nagel). Arguments embodied in each of her three local protests would all feature in her developed account of the injustice of murder in war.

10.4.2 Anscombe’s Protest #1: The Speech In late April, Anscombe notified the Council of her intention to vote against honouring Truman at the May 1 Congregation meeting.105 When the vice-​chancellor read the names to receive honorary degrees, 102  Mary Geach indicates that it was Peter Geach who first suggested that Anscombe protest the degree. 103  TD, 69. 104  Box W3, File 394, Collegium Institute Anscombe Archive at the University of Pennsylvania, Kislak Center for Special Collections, Rare Books and Manuscripts. There are approaching one hundred letters in the archives, and many of these letters included the newspaper clipping about her speech from their local newspapers. In addition, Anscombe received some number of letters which are not in the archives. There is evidence in Anscombe’s archives and elsewhere that she dutifully responded to many or even most of these letters. 105  Times of London, Daily Telegraph, Manchester Guardian, New York Times, and other major papers reported on that meeting of Convocation. What tipped the press off to be present is unclear.

Justice and Murder    251 Miss Anscombe, a slight figure in cap and gown, rose and said, so quietly that she could hardly be heard, ‘I have come here on purpose to vote against the proposed degree honoris causa to Mr. Truman.’106

Anscombe noted that Truman not only acknowledged but boasted that he was responsible for the obliteration of Nagasaki and Hiroshima. She rejected the claim that Truman acted courageously, since he had nothing to lose in his decision.107 However, ‘I should like to think that he had one thing to lose, and that that one thing was the chance of an honorary degree from Oxford.’108 She summarised her opposition with a question that would be printed in hundreds of newspapers around the world: ‘If you do give this honour, what Nero, what Genghis Khan, what Hitler or what Stalin will not be honoured in the future?’ Having received advance notice, Vice-​Chancellor A. H. Smith of New College tasked historian Alan Bullock, known for his recent biography of Hitler, with responding on the Council’s behalf. Bullock initially argued that responsibility for the bombings could not be placed at Truman’s feet: A great many people were involved in the responsibility of the manufacture and delivery of the bomb, and we can’t select one man as being solely responsible even if his was the signature at the bottom of the order for the bomb to be dropped.109

Anscombe found Bullock’s attempt to diminish Truman’s responsibility odd, coming from an expert on World War II–​era Germany. As she put it, his argument would not have impressed the judges at Nuremburg.110 Bullock was apparently not impressed by his own argument, as he quickly pivoted to a utilitarian calculation of Truman’s positives and negatives. While the Hebdomadal Council did not necessarily agree ‘with every action [Truman] has taken in his life’, Truman had kept the US from its post–​World War I isolationist stance, instead bankrolling Europe after the war. On Bullock’s utilitarian calculation, the good of Britain’s being financially benefited after the war outweighed the mass murder of 100,000+​innocent civilians. Bullock’s final verdict was ‘Mr. Truman has done some good.’111 106 

Oxford Mail, May 1, 1956. In TD, Anscombe wavers on this point: ‘Light has come to me lately: [courage] is an acknowledgement of the truth. Mr. Truman was brave because, and only because, what he did was so bad. But I think the judgment unsound. Given the right circumstances . . . a quite mediocre person can do spectacularly wicked things without thereby becoming impressive’ (TD, 64). 108  Oxford Mail, May 1, 1956. 109  Oxford Mail, May 1, 1956. Nine months later, in OMPCY, Bullock’s arguments will be grist for Anscombe’s mill when she develops her critique of how Oxford moral philosophers understand ‘responsibility’ in a very different way before a person’s act from their understanding of ‘responsibility’ after a person’s action. In Anscombe’s view, they manipulated the notion of ‘responsibility’ depending on how things turned out. 110  TD, 65. 111  TD, 66. 107 

252   John Berkman

10.4.3 Anscombe’s Protest #2: The Bet Two weeks after Anscombe’s speech, which had gotten into the international press, was not yet the talk of Oxford, at least to judge by Alasdair MacIntyre’s unawareness of it when he encountered Anscombe at a meeting of the Socratic Club on May 14. On that evening he presented a paper on miracles, with Bernard Williams responding. MacIntyre notes that while Anscombe participated in the discussion, he had heard nothing of her speech in Congregation.112 However, that same evening, Arthur Goodhart, master of University College and former holder of the Oxford Chair in Jurisprudence (1931–​1951), spoke at the English-​Speaking Union, vigorously defending the proposed honorary degree.113 Goodhart assailed Anscombe’s intellect, that her likening Truman to Genghis Khan, Hitler, and Stalin ‘showed confusion of thought’, for ‘the bomb on Hiroshima saved British as well as American lives’.114 Goodhart followed up his speech with an article in the Oxford Mail, in which he sought to define when it was legal to kill civilians in war-​time. Anscombe replied to Goodhart in the Oxford Mail in a way that generated maximum publicity. Slyly, she began by noting his expertise in international law. Then she complemented Goodhart’s civic engagement as president of the Pedestrians’ Association,115 praising him for his opposition to ‘lethal carelessness on the roads.’ However, Anscombe then questioned Goodhart’s consistency: Why did Goodhart not express equal concern for the lethal carelessness exhibited in the atomic bombing of Nagasaki and Hiroshima?116 Anscombe then proposed to bet Goodhart £100 that he could not justify the bombing of Hiroshima under international law.117 To win the bet, Goodhart had to answer one of two questions: 1. Was there any part of international law in existence in 1945 stating that it was permissible to make civilian populations the direct object, the intended target, of attack in cases of ‘necessity’?

112 

Alasdair MacIntyre has confirmed Anscombe’s participation at the May 14, 1956, meeting of the Socratic Club. Although Anscombe participated in the discussion, he had not yet been introduced to her, and had as of yet not heard about Anscombe’s objection to Truman. Email to author, December 16, 2020. 113  Goodhart (1891–​1978) resigned from his professorship when he was elected master of University College in 1951, and H. L. A. Hart was elected to his chair in 1952 at age forty-​four. 114  Goodhart’s speech was on May 14. ‘Oxford Degree for Mr. Truman: Reply to Critic,’ Daily Telegraph, May 15, 1956. 115  An association formed to advocate the protection of pedestrians from motor vehicles. 116  Anselm Müller notes, ‘A strong inclination towards sarcasm is conspicuous in Anscombe’s talks and writings, too, esp. in early comments on fashionable stances on questions of morality’ (The Personality of Elizabeth Anscombe, 13n11; see also Müller, ‘The Personality of Elizabeth Anscombe’). 117  This was no small sum, being about 12 percent (about £2,500 in 2020 funds) of Anscombe’s salary at the time.

Justice and Murder    253 2. Was there any part of international law in existence in 1945 condemning such attacks only in cases where they were ‘unnecessary’?118 Anscombe suggested the bet be adjudicated by Humphrey Waldock, the Chichele Professor of Public International Law at Oxford.119 She duly deposited the £100 with the editor of the Oxford Mail.120 In making this public bet, Anscombe was continuing her quest to show that according to international law, Truman had committed mass murder. This second protest aimed to persuade her Oxford colleagues to boycott the Encaenia. Goodhart did not attempt to win the bet and claim the £100.121

10.4.4 Anscombe’s Protest #3: The Pamphlet Mr. Truman’s Degree After delivering her protest speech at Congregation, Anscombe wrote and published TD in a few weeks, making it publicly available by the end of May 1956.122 The speed at which she produced her pamphlet, as well as its arguments, attest to her continued effort to persuade her Oxford colleagues to boycott the Encaenia on June 20. Her pamphlet concluded with practical advice for the cowardly—​t ake a ‘sick day’: It is possible still to withdraw from this shameful business to some slight degree; it is possible not to go to Encaenia; if it should be embarrassing to someone who would normally go to plead other business, he could take to his bed. I, indeed, should fear to go, in case God’s patience suddenly ends.123

118 

Anscombe would donate £100 to Goodhart’s Pedestrian Association if he could demonstrate an affirmative answer to the first question, and £50 if he could prove only the second point. 119  Sir Humphrey Waldock (1904–​1981) was Chichelle Professor of Public International Law at Oxford from 1947 to 1972. He had been the undersecretary to the British Admiralty in World War II and, after retiring from Oxford, was elected in 1973 to the International Court of Justice, which was the chief judicial arm of the United Nations. He was president of that Court in 1981, when he died of a heart attack. Waldock had formerly been president of both the European Commission of Human Rights and the European Court of Human Rights. ‘Sir Humphrey Waldock, 77: Head of International Court,’ New York Times, August 18, 1981, B19. 120  As of June 1, 1959, Goodhart had made no effort to collect it. See Ann Chesney, ‘Don, Philosopher, and Happy Mother of Six,’ Manchester Guardian, June 1, 1959, 3. 121  Chesney, ‘Don, Philosopher, and Happy Mother of Six,’ 3. 122  On June 2, 1956, the Manchester Guardian announced the pamphlet for sale, summarising its arguments. That same day, Anscombe mailed a copy of her newly published pamphlet to Charles Thompson, editor of the PAX Society newsletter. Anscombe was responding to Thompson’s request, after hearing of her protest in the press, that she contribute a piece for the Pax Bulletin (courtesy of the Pax archives). 123  TD, 71.

254   John Berkman

10.4.5 Anscombe Protest #4: Staying Away from the Encaenia Having kept her objections to Truman continuously in the news cycle, Anscombe essentially guaranteed that Truman would be queried about her protests upon his arrival in the UK. When that time came, Truman likely feigned ignorance about Anscombe’s protests.124 Newspapers worldwide covered the Encaenia. Their write-​ups all referenced Anscombe’s protest, and their explanations of her objections occupied as much as 25 percent of the coverage in some newspapers. While (as far as is known) Anscombe failed in her efforts to keep colleagues from attending the Encaenia, she undoubtedly succeeded beyond her wildest dreams in bringing worldwide attention to Truman’s murderous injustice and helping Britain recover from its ‘loss of the conception of murder.’125 The newspapers even mentioned Anscombe’s boycott, noting that she ‘marked her displeasure by staying away from the Sheldonian Theatre. She would spend the day working as usual, she said.’126

10.4.6 Anscombe’s Goal: Recovering the Concept of Murder Anscombe’s primary concern in her protests was clear: ‘The loss of the conception of murder . . . is my chief interest in this pamphlet.’127 The opening paragraphs of TD expressed this concern in asking how, between 1939 and 1945, the US lost its prohibition of killing civil populations in the conduct of war.128 While pagans and totalitarians

124  ‘Truman Not Told of Objections to Oxford Degree,’ Oxford Mail, June 18, 1956. In his dealings with the university in 1955–​1956, Truman was less than honourable. In November 1955 he wrote the Chancellor Lord Halifax, accepting the invitation to receive the honorary degree, making a ‘firm commitment’ to attend, and promising to say nothing until it was officially made public. Within weeks, Truman gave public interviews bragging about the offer but saying that he had not yet agreed to accept ‘the beefeater’s hat degree’ (Oxford Mail, November 21 and 26, 1955). Halifax commented, ‘I think Truman’s speech is absolutely monstrous’ (letter to the Registrar, November 24, 1955). Truman’s antics continued into April, saying that the university ‘had been asking him for the past two years to accept’ and he ‘hoped’ to do so (Oxford Mail, April 4, 1956). After the Encaenia, Truman absconded with the ceremonial cap and gown which he had been loaned for the occasion. So Truman being untruthful with regard to knowing about Anscombe’s protest prior to receiving his degree would not have been out of character. Archives of the Proceedings of the Hebdomadal Council, Bodleian Library, Oxford. 125  TD, 69. 126  See ‘Mr Truman Borrows a “Floppy Hat” for His Degree: Lord Halifax on the “Truest of Allies,” ’ Manchester Guardian, June 21, 1956, 5. 127  TD, 69. 128  Anscombe’s examples of the loss of natural justice are, first, the change in viewpoint of the US president between 1939 and 1945 with regard to killing civilians and, second, General Eisenhower’s obliviousness to the moral basis of prisoner care, that it is based on defencelessness, not virtue or nationality.

Justice and Murder    255 like Hitler and Stalin might be expected to do such things, how could the self-​professed Baptist Truman ‘exult’ in massacring civil populations? In 1940 in JPW Anscombe and Daniel had presumed a shared moral currency in English society, and thus critiqued direct attacks on civilian populations by appeal to traditional notions of natural justice, the Law of Nations, and just war theory. But in TD, Anscombe realises that ‘only fragments of [such shared assumptions] are left’, and so avoided direct appeals to such notions.129 Although in TD Anscombe restricts her focus to the loss of moral vocabulary concerning murder in war, this loss of the conception of murder is only one element of a broader societal loss of moral vocabulary. Anscombe will later address this more substantive loss of moral vocabulary—​such as the language of moral duty and the moral ‘ought’—​in MMP. Perceiving this ‘fragmentation’ of societal moral vocabulary, Anscombe instead begins her argument by appealing to basic moral phenomenology, such as the sense of outrage people feel when the enemy violates certain moral norms in war. In so doing, Anscombe engages in moral catechesis, gradually reintroducing the language necessary to account for this sense of outrage and in so doing work towards recovering the concept of murder. Towards that end, Anscombe starts with the basic distinction between legitimate and illegitimate killing in war, a standard maintained in all previous international laws and agreements: [W]‌ith Hiroshima and Nagasaki we are not confronted with a borderline case. In the bombing of these cities it was certainly decided to kill the innocent as a means to an end. And a very large number of them, all at once, without warning, without the interstices of escape or the chance to take shelter.130

Anscombe sadly recognises that her society’s military and political leaders not only do not see ‘innocent civilians’ as a moral given; these leaders throw up multiple objections which she has to refute. Her society’s lack of respect for the notion of ‘innocents’ is further evident in its tendency to avoid the language of murder altogether, speaking only of ‘killing’ in a way that assumes all killing is morally equivalent: ‘Soldiers, women, babies, the disabled or insane—​it was all the same’.131 For Anscombe, this in a way had to be expected, since the definition of murder—​‘to choose to kill the innocent as a means to your ends’132—​requires an ability to distinguish between those who are innocent and those who are not.

129 

TD, 67. Anscombe only covertly appeals to the traditional just war criteria. Her attack on ‘unconditional surrender’ fits the criteria of ‘just cause’. Her claim that private persons may not intentionally kill is an appeal to ‘just authority’. Her defence of ‘innocents’ appeals to ‘non-​combatant immunity’. 130  TD, 64. 131  TD, 64, 66. 132  Anscombe’s definition does not encompass all acts of murder, but acts covered by this definition are all acts of murder.

256   John Berkman So, morally speaking, Anscombe has to start from scratch, explaining the notion of ‘innocents’ in the context of a war, and why soldiers, even conscripted soldiers, are not innocent in the morally relevant sense. The innocent in a war are those who neither fight nor supply those fighting with ‘the means of fighting’. ‘Innocent’ is not a trait of the individual; it entails having the characteristic of ‘not harming’. Thus, soldiers who surrender become innocents and cannot be attacked or mistreated. Without naming it, Anscombe also draws on the principle of double effect: while innocents may not be made the object of any attack, if some innocents are killed inadvertently in the course of scrupulously careful attacks intended to hit military targets only, that is not murder.133 At the end of TD, Anscombe puzzles at the action of the Hebdomadal Council and casts about for some light as to why so many Oxford colleagues are willing to honour Truman. She considers the two dominant Oxford moral philosophies over the previous forty years and finds that ‘both these philosophies contain a repudiation of the idea that any class of actions, such as murder, may be morally excluded’.134 This is Anscombe’s first characterisation of Oxford moral philosophy, and while it is brief—​only one paragraph—​here she already anticipates much of the argument of MMP, and thus what she says in this paragraph is worthy of analysis. After TD, Anscombe will characterise Oxford moral philosophy twice more before her fourth and final characterisation in MMP, which will constitute the majority of MMP. While her critique of Oxford moral philosophy expands, all four characterisations are firmly focused on one central point: Oxford moral philosophy was constitutively unable to rule out and thus necessarily open to the view that some acts of (mass) murder are good and/​or right acts. Ross had already acknowledged this in print, and Hare would later do so.135

10.5  Anscombe’s Query: Does Oxford Moral Philosophy Justify Murder? In TD, the unnamed representatives of the two ‘prevailing moral philosophies’ are W. D. Ross and R. M. Hare.136 Ross’s moral philosophy comes in for the greatest scorn, Anscombe judging it to pretentiously feign moral seriousness. While maintaining the 133 

TD, 66. TD, 71. 135  See notes 137 and 171 in this chapter. 136  Anscombe would name both in later publications. In September 1955, nine months prior to writing TD, she heard her husband, Peter Geach, read his paper ‘Good and Evil’ to the Philosophical Enquiry Group at Spode House (The Tablet, September 17, 1955, 286). In his paper, Geach had similarly named Ross and Hare as the representatives of the two prevailing Oxford moral philosophies of the previous forty years. However, Geach would not publish his paper for another sixteen months. Peter Geach, ‘Good and Evil’, Analysis, December 1956. Before Geach finally published his paper, he personally invited Hare’s response. In reply, Hare acknowledged that he (the ‘Oxford moralist’) was the object of Geach’s attack in the article. R. M. Hare, ‘Geach: Good and Evil’, Analysis, March 1957. 134 

Justice and Murder    257 language of ‘duty’, Ross evacuates any recognisable meaning such that it retains only ‘mesmerising force’. Ross allows for only prima facie duties, which he acknowledges ‘are not actually duties’. His notion of a prima facie duty is merely something which one must ‘consider’ in one’s decision-​making process. He abandons the notion of duty because he thinks duties conflict in a way which requires us to weigh them against each other. On Ross’s theory, ‘morally good actions’ are subjective: they are good if the agent acts according to these prima facie duties. In contrast, ‘morally right actions’ are objective: they are good if they produce the ‘maximum possible good.’ The agent is to determine the right act by her ‘moral sense’. Anscombe notes that on Ross’s theory Himmler’s efforts at exterminating Jews is to be considered ‘morally good’, as long as Himmler acts from the ‘motive of duty’, which has ‘supreme value’.137 Similarly, Truman’s decision to massacre civilians should be called the right action if it secures the greatest advantage. While Ross’s theory acknowledges a ‘prima facie’ duty not to kill the innocent, ‘most thoughtful people’ will consider the prima facie duty to secure some societal advantage to outweigh the prima facie duty of justice.138 Furthermore, since Ross’s theory equates ‘outweighing’ with ‘being more obligatory,’ massacring innocent people would on some occasions be not only morally right but morally obligatory. After discussing Ross’s older prevailing moral philosophy, Anscombe turns to the currently prevailing Oxford moral philosophy—​Geach calls them ‘the Oxford moralists’—​ and her description perfectly matches ­chapter 4 of Hare’s Language of Morals. For the ‘Oxford moralists’ such as Hare, it is impossible to have general moral laws, because moral decisions often involve the modification of principles.139 So a principle like ‘it is wrong to poison’ is merely a rule of thumb, which an experienced person learns when to modify or break. Furthermore, the ultimate justification of the principles a person has depends on the ‘way of life’ the person chooses.140 The distance of Hare’s view from the

137  For Ross, a ‘prima facie duty’ is the characteristic an act has ‘which would be a duty proper if it were not at the same time [an act] of another kind which is morally significant’; a right act is obligatory, and is the act which is productive of the maximum possible good; the morally good act must be done from certain motives and ‘the motive of sense of duty is of “supreme worth,” “the supreme moral value,” and “the morally best motive.” ’ Furthermore, ‘whatever be its consequences,’ the morally good act ‘is good in itself . . . even when it is not a right act.’ All of these quotes can be found in W. D. Ross, ‘What Makes Right Acts Right’, in The Right and the Good (Oxford: Oxford University Press, 1930), 19; W. D. Ross, ‘The Nature of Morally Good Actions’, Proceedings of the Aristotelian Society 29 (1928–​1929), 251, 252, 260, 264; and W. D. Ross, ‘The Basis of Objective Judgments in Ethics’, International Journal of Ethics, 37, no. 2 (January 1927), 113–127. 138  ‘The interests of the society may . . . make it right to punish an innocent man ‘that the whole nation perish not.’ The prima facie duty of consulting the general interest has proved more obligatory than the perfectly distinct prima facie duty of justice. This is, I believe, how most thoughtful people feel about . . . the invasion of the rights of others’. W. D. Ross, ‘The Ethics of Punishment’, Journal of Philosophical Studies 4, no. 14 (April 1929), 209. 139  R. M. Hare, The Language of Morals (Oxford: Clarendon Press, 1952), 62–​64. 140 Hare, Language of Morals, 73.

258   John Berkman perspective of natural justice, or the view that killing innocent persons is unconditionally wrong can be seen in the following passage: If asked to justify as completely as possible any decision, we have to bring in both effects—​to give content to the decision—​and principles, and the effect in general of observing those principles. Thus a complete justification of a decision would consist of a complete account of its effect, together with a complete account of the principles which it observed, and the effects of observing those principles. . . . [I]‌f pressed to justify a decision completely, we have to give a complete specification of the way of life of which it is a part.141

As for the significance of these prevailing Oxford moral philosophies, Anscombe was unresolved as to whether they were influential or merely symptomatic of contemporary British culture. Evidently, a number of readers of Anscombe’s one-​paragraph summary of the prevailing Oxford moral philosophies took an interest. One of those readers was A. E. Harvey of the BBC. On July 18, only seven weeks after the publication of her pamphlet, Anscombe got a letter from Harvey at the BBC Talks Department asking her to ‘develop the theme of the relevance of Oxford philosophy to situations such as the one which inspired your pamphlet’.142 Which is how, seven months later, on January 27, 1957, Anscombe came to record her wireless talk ‘Oxford Moral Philosophy: Does It Corrupt the Youth?’

10.6  How Oxford Moral Philosophers Justify Murder 10.6.1  Satirising Oxford Moral Philosophy: Why Justifying Murder Does Not Corrupt the Youth Eight months after the end of the Truman affair, Anscombe’s radio broadcast presented her with a second opportunity to characterise Oxford moral philosophy.143 Anscombe

141 Hare, Language of Morals, 69. 142 

See Benjamin Lipscomb, ‘The Women Are Up to Something’, Royal Institute of Philosophy Supplement 87, no. 40 (2020), 22. Lipscomb references the BBC Written Archives Centre, RCONT3, A. E. Harvey to G. E. M. Anscombe, July 18, 1956. 143  Anscombe’s talk was broadcast on February 5, 1957, on the Third Programme of the BBC and appeared in print in revised form in The Listener on February 14. The Listener printed transcripts of BBC broadcasts, advertised upcoming broadcasts, and printed letters to the editor. The Third Programme was the third of three BBC radio channels at the time. It had a much smaller listening audience than the first and second programmes, known as the Home Service and the Light Programme. It was the ‘intellectual’

Justice and Murder    259 recognised that if one wants to accuse moral philosophers of justifying murder, one had best not do it directly. So she took the highly unusual step of presenting her philosophical viewpoint on her BBC broadcast in the form of a satire, it being so unusual that her BBC producer T. S. Gregory was unaware that she was employing that genre, thinking her presentation was done in earnest.144 He misinterpreted her talk even though Anscombe stated in the talk itself that the ‘direct teaching of ideas is not, nowadays, the best way of setting about changing people.’145 This failure to understand Anscombe’s choice of genre is arguably Anscombe’s fault, that she failed to present a sufficiently grotesque and absurd picture of Oxford moral philosophy such that the satire was plainly evident to all.146 On the other hand, Anscombe’s choice to make her satire more subtle arguably made the satire ultimately more powerful. Furthermore, Anscombe’s sweet, angelic voice was hardly the expected vehicle for a biting satire. Even the Oxford moral philosopher Patrick Nowell-​Smith admitted to not understanding her talk until he read the printed version. Anscombe was, as usual, extremely trenchant, packing a full-​blown critique of Oxford moral philosophy into a mere thirteen paragraphs. She began her talk by noting that the moral philosophy she was critiquing was one associated with ‘linguistic analysis.’ While its practitioners might not like being lumped together, ‘their work looks roughly alike from the outside.’ Having been instructed to consider ‘especially up-​to-​date Oxford moral philosophy’ the object of her attack was again R. M. Hare.147 While Hare was not attacked by name, the views critiqued were recognisably Hare’s, and he well knew it, as evidenced by his letters to The Listener that followed her talk. The ostensible purpose of Anscombe’s BBC talk is to answer the question she left unresolved at the end of Mr. Truman’s Degree: Is Oxford moral philosophy influential on, or merely symptomatic of, the ‘highest and best ideals of society at large’? She settles on the latter view. Oxford moral philosophers are not corrupting the youth, since they proffer nothing new, being as they are ‘perfectly in tune’ with the ideals of one sort of conventional English morality. Oxford moral philosophers merely have on offer ‘the philosophy of the flattery of [the spirit of the time].’148 Anscombe’s sardonic indictment of Oxford moral philosophy becomes even more biting (than the critique that they have nothing original to say) when she claims that one of the ‘highest ideals’ of British society that Oxford moral philosophers parrot is the justification of mass murder. The bulk of her radio broadcast presents the two-​step ‘method’ by which Oxford moral philosophers go about justifying murder: first, Oxford moral philosophy sets channel featuring relatively highbrow content. It was seen as an attempt to bring elite culture to the masses. However, most of the masses were not listening. 144 

See Lipscomb, ‘The Women Are Up to Something’, 23n43. Anscombe’s choice of the genre of satire went along with her suspicion (OMPCY, 162). 146  As Northrop Frye notes, ‘Two things, then, are essential to satire; one is wit or humour founded on fantasy or a sense of the grotesque or absurd, the other is an object of attack’. Northrop Frye, The Anatomy of Criticism: Four Essays (1957; Toronto: University of Toronto Press, 2006), 209. 147  OMPCY, 161. 148  OMPCY, 167. 145 

260   John Berkman aside the language and practices of natural justice, rejecting the view that morality is inherent in human actions and character; second, Oxford moral philosophy evaluates human actions purely instrumentally, the ‘rightness’ of actions based on their ‘managerial’ prowess.149 Their favourite moral criteria are appeals to ‘responsibility’ and ‘decreasing suffering’, using that language in oleaginous ways, along with banal or fantastic examples and mad calculations regarding the ‘total situation’ in the future. This is their means of ‘managing’ society (and justifying mass murder).

10.6.2 Step #1: Oxford Moral Philosophy’s Repudiation of Natural Justice Anscombe characterised the Oxford moral philosophers’ rejection of ‘natural justice’ as an ‘anti-​Platonic’ conception of justice, although only in one respect: on Plato’s view a just society required just people, just as a healthy society required healthy people. For the Oxford moral philosophers, since ethics was public policy, calculating and promoting the general good, it did not require just individuals. In the dystopian society of the Oxford moral philosophers, judges would take no concern for Aristotle’s ‘archaic and metaphysical’ conception of natural justice, blissfully ignoring the requirement of natural justice to give everyone their due. For the Oxford moral philosophers, nothing per se is owed in justice to any particular individual. This Anscombe ironically called ‘the great insight of modern times’ that injustice may be nobody’s fault. Having left behind the conception of natural justice, Oxford moral philosophers would skilfully deploy three ‘weasel words’ to justify injustice: responsibility, suffering, and ‘way of life.’

10.6.3 Step #2: Appeal to ‘Responsibility’, Avoiding ‘Cruelty and Suffering’, or ‘Way of Life’ When Oxford moral philosophers wish to justify injustice—​e.g., killing innocent persons—​they claim it is one’s responsibility to do so. Why? They have calculated that the total situation will be worse if you don’t kill these innocent persons. The mantra is ‘You are answerable for the future if you can make it better.’ Thus ‘responsibility’, a high-​ sounding ideal, not only justifies doing something inherently wrong but also entails an obligation to act unjustly to bring about the best consequences: ‘For example, it was

149 

‘You can learn at Oxford that . . . justice . . . is a term like “well-​arranged.” . . . Injustice may be nobody’s fault, and that what is required is good arrangement. With this goes a preference for policy—​ which is the effort to calculate and promote the general good over archaic and metaphysical notions of justice’ (OMPCY, 163–​164). Anscombe will imply that rhetoric about ‘the best interests of all’ is used by those with power to justify injustices against the poor, widows, parents, and criminals.

Justice and Murder    261 right to massacre the Japanese because it was (or at least was thought) productive of a better total state of affairs than not doing so would have been’.150 Drawing implicitly on the example of Bullock’s attempt to justify Truman, Anscombe exposes the Oxford moralists’ inconsistent appeal to ‘responsibility.’ To persuade a person to do an unjust action, prior to the action Oxford moralists emphasise the person’s agency, that the individual is obliged to do the unjust action because their action will make a great difference to the future state of affairs. However, after the unjust action—​such as massacring innocent civilians—​Oxford moralists then downplay the agent’s responsibility, have ‘a gentle, tolerant, and civilized idea of responsibility’ where all those who are in any way causally connected to the unjust action all share in the responsibility, and thus no one individual’s action makes them really responsible. Here Oxford moralists remind us of the antics of small children and nationalistic newscasters. If you break a child’s toy, the child says ‘You broke my toy.’ But if the child breaks the toy, agency disappears, as in ‘My toy broke.’ Similarly, on newscasts in the United States on December 7, you’ll hear ‘On this day in 1941 the Japanese bombed Pearl Harbor.’ However, on August 6 you’ll hear ‘On this day in 1945 an atomic bomb was dropped on Hiroshima.’ While humorous on one level, Anscombe is speaking from the experience of hearing Bullock and Oxford moralists argue in precisely this way to justify Hiroshima. An alternative means of justifying unjust actions is to appeal, not to maximising happiness but to avoiding cruelty and suffering. When the Oxford moralists want to justify an unjust act, they appeal to the (always) horrendous suffering of persons if the unjust action is not done. The Oxford moral philosophers typically invent fantastic examples to justify the doing of unjust acts: ‘If anyone should try saying that some action was bad, a case (however fantastic) is at once imagined in which a consequence of doing that action is that some horrible suffering is averted, and that settles the question.’151 Anscombe was familiar with these kinds of moves, as she noted in TD: I must be saying ‘You may not do evil that good may come,’ which is a disagreeably high-​minded doctrine. . . . Come now: if you had to choose between boiling one baby and letting some frightful disaster befall a thousand people—​or a million people, if a thousand is not enough—​what would you do? Are you going to strike an attitude and say ‘You may not do evil that good may come’? (People who never hear such arguments will hardly believe they take place, and will pass this rapidly by.)152

Anscombe considered the Oxford moralists’ constant appeal to fantastic examples as part of their means of doing away with justice as central for morality. Anscombe also notes how this doctrine of ‘suffering’ undermines the distinction between legitimate killing in war and murder, for with this view ‘what is dreadful in war is 150 

OMPCY, 164. OMPCY, 165–​166. 152  TD, 65. 151 

262   John Berkman purely the ‘use of force’, aggression, the amount of suffering; who, for example, is made the object of attack, with what justification, does not make much difference’.153 The final means of justifying injustice, according to Anscombe, is to appeal to our ‘way of life’. Anscombe recognises that such appeals to ‘our standards’ and ‘what we say’ is the in-​group jargon of the powerful who presume themselves to be the good and benevolent, who impose their ‘standards’ more often than not for their own benefit. A classic example of a colonialist attitude applied to morality.

10.7  Detective Anscombe Extracts Confessions to Justifying Murder 10.7.1  Getting Oxford Moral Philosophers to take Anscombe Seriously Anscombe’s BBC broadcast would of course not go unnoticed. For one thing, Hare had been sent an advance copy of it, in part because a response from an Oxford moralist was originally part of the suggested format.154 The two letters to the editor of The Listener that pleased Anscombe the most were from the Oxford moralists Nowell-​Smith and Hare. Their initial letters paternalistically chided Anscombe for unclarity and had the flavour of annoyed and weary schoolmasters trying to help the dim-​witted student understand the lesson. At Oxford Anscombe neither taught nor wrote on ethics, so they presumed her criticisms were ‘out of line’ and required correction.155 However, they gravely underestimated Anscombe, and she wound up doing the ‘schooling’. Anscombe’s exchanges with Nowell-​Smith and Hare have a similar trajectory. Her initial response to each of them essentially ignores their challenges, and instead responds very briefly, poking and provoking them further. Anscombe seems to wants to draw out Nowell-​Smith and Hare further, have them put their ‘moral’ cards on the table, and

153 

OMPCY, 166. My thanks to Benjamin Lipscomb for bringing to my attention the background to Anscombe’s BBC address in 1957, as well as the fact that she prepared a second talk for them. 155  For example, the Greats undergraduate Miranda Villiers was tutored in ‘Logic and Aristotle’ by Anscombe from Michaelmas (October) 1956 to Hilary (March) 1958, focusing on epistemology. She did moral philosophy with Philippa Foot and political philosophy with Mary Warnock. According to Villiers, Anscombe never discussed morals with them: ‘[T]‌he only even vaguely ethical pronouncement I can remember was a statement, in connection with goodness knows what, that she always paid bills exactly three weeks after they were presented: a period which seemed to her to represent a nice balance between contempt for Mammon and consideration for the shopkeeper’. Miranda Villiers, Somerville 1954–​1958, ‘Miss Anscombe Was My Tutor’, unpublished memoir, Somerville College archive, courtesy of the Fellows of Somerville College. 154 

Justice and Murder    263 apparently she thinks the best way to do that is to enrage them. And enrage them she does. In their second responses, both assail Anscombe, but in doing so expose their views more fully. Anscombe then takes full advantage in her further replies. I first analyse her exchanges with them individually, and then summarise how Anscombe lumps Nowell-​Smith and Hare together in her conclusion that they are both what she labels ‘consequentialists’.

10.7.2 Anscombe’s Exchange with Nowell-​Smith Nowell-​Smith’s initial (February 21) letter deigns to initiate Anscombe into the complexities of Oxford moral philosophy: The general burden of [Anscombe’s] criticism . . . is that people tend to judge all actions by their consequences rather than by their ‘nature and quality.’ . . . Miss Anscombe seems to be (though I can scarce believe that she is) ignorant of the difficulties involved in drawing a distinction between an act and its consequences.156

The irony of Nowell-​Smith’s letter is his apparent ignorance that when he writes his letter, his Oxford colleague Anscombe is midway through her course of lectures in Oxford that result in her book Intention. Considering that Intention is being advertised for sale less than six months after Nowell-​Smith’s letter, it is possible that Anscombe has already sent her completed manuscript to Blackwell’s.157 Nowell-​Smith presents Anscombe with the false dilemma fallacy, trying to get her to choose one of only two options: For example, was Mr. Truman’s ‘act’ the signing of an order, the killing of a number of Japanese, or the saving of a number of Japanese and other lives? If it was the first only, Miss Anscombe has, on her own principles, as little right to condemn it as Mr. Truman’s supporters have to defend it, since both judgements turn on its consequences. But if the killing is to be included in the nature and quality of Mr. Truman’s act, why not the saving of lives? . . . [I]‌t is not an easy matter to say where and how [the line] is to be drawn. It is with the elucidation of just such difficulties that moral philosophy is concerned.158

In response to his long letter, Anscombe is infuriatingly brief. She first mocks Nowell-​ Smith’s ‘colossal difficulty’ with discerning the character of an act, introducing the example of an action that includes (a) sending chocolates through the mail, (b) poisoning your aunt, and (c) collecting an inheritance. She then notes that she’s read his 156  Patrick Nowell-​Smith, ‘Oxford Moral Philosophy’, letter to the editor, The Listener, February 21, 1957. 157  Blackwell’s was advertising Intention in the August 16, 1957, issue of the Times Literary Supplement. 158  Nowell-​Smith, ‘Oxford Moral Philosophy’.

264   John Berkman book and finds his condemnation of consequentialism unpersuasive.159 Finally, she bemoans the fact that the problem with moralists like Nowell-​Smith and Hare is not that they say that it is good, e.g., to commit fraud in the courts or to evict widows from their homes; rather, the problem is that their moral philosophies justify such acts; i.e., those holding their moral philosophies have no firm basis to condemn such acts as bad as such.160 If it had indeed been her intent to enrage, Anscombe is successful. In Nowell-​ Smith’s second response, he fulminates against Anscombe, accusing her of injustice in failing to take his Hiroshima example with proper seriousness, of innuendo in misleading people about his philosophical and moral views, of travesty in substituting a ‘fantastic hypothetical case’ of poisoning by chocolates for his difficult example about the bombing of Hiroshima, and sarcasm in claiming he would have ‘colossal difficulty’ about what a ‘good man’ is to do in the example of the poisoned chocolates.161 Now that Nowell-​Smith has expressed his views sufficiently, Anscombe’s tactics change. Her second response is systematic and comprehensive. She replies that far from a ‘travesty,’ her chocolate poisoning example perfectly parallels Nowell-​Smith’s Hiroshima bombing example, except that it is superior in not getting some facts wrong.162 See Table 10.1. In both examples (b) is an act of murder, as a means to the end of (c). How could Nowell-​Smith find her chocolates example obvious, but the Hiroshima example difficult? Anscombe calls Nowell-​Smith a ‘child of his times’ in that he cannot see the parallel, and ‘how his philosophy helps him not to!’ What

Table 10.1  Nowell-​Smith’s Example

Anscombe’s Example

(a) signing of an order

(a) sending chocolates through the mail

(b) killing of a number of innocent Japanese

(b) poisoning your aunt

(c) saving of a number of Japanese and other lives?a

(c) collecting an inheritance

a Nowell-​Smith, ‘Oxford Moral Philosophy’, February 21, 1957.

159  Nowell-​Smith’s and Hare’s moral philosophies justify intrinsically wrong actions in slightly different ways. Nowell-​Smith’s Ethics (308) has one ‘frame a moral principle under which [one] ‘manages’’ to bring the action, whereas Hare’s Language of Morals instructs one to formulate ‘a new ‘decision of principle’’ which constitutes an ‘advance in [one’s] moral thinking’ (MMP, 42). 160  Elizabeth Anscombe, ‘Oxford Moral Philosophy’, letter to the editor, The Listener, February 28, 1957. 161  Patrick Nowell-​Smith, ‘Oxford Moral Philosophy’, letter to the editor, The Listener, March 14, 1957. 162  According to Anscombe, the problem with Nowell-​Smith’s example is that his ‘saving of a number of Japanese and other lives’ was not factual. It was well-​known that the Japanese wanted to surrender.

Justice and Murder    265 infuriates Anscombe is how Nowell-​Smith and other Oxford moralists obfuscate the obvious: The suggestion that no one can treat ‘Do no murder’ as an intelligible commandment in a broadcast without a preliminary exposé of the philosophical problems of defining an action seems to me in a high degree comic.163

10.7.3 Anscombe’s Exchange with Hare In Hare’s initial (February 21) letter to the editor, amongst other things, he feigns incredulity and attempts to reduce Anscombe’s argument to a reductio ad absurdum: She thinks it wrong to judge acts by the foreseen consequences of committing them; for example . . . that the person who ordered the atom bomb to be dropped on Hiroshima had no duty to consider whether anybody would be killed by the explosion. Yet in a recent pamphlet Anscombe accused Mr. Truman of being a murderer.164

Although Hare raises many challenges, Anscombe goes on the attack rather than answering. As with Nowell-​Smith, she makes fun of Hare, feigning that Hare has questioned her sexual morals.165 Although Anscombe then makes light of this supposed insult, she apparently succeeds in threatening Hare’s sense of propriety as an English gentleman. After softening him up, she invents a new term to describe Hare’s moral theory: Mr. Hare is openly a consequentialist. It should be stated clearly that this means, and he has implicitly stated, that there is no sort of action whatever of which it is correct to say ‘One doesn’t have to consider whether to do this or not, in any circumstances; it is simply excluded’.166

In TD and her BBC lecture, Anscombe emphasised that the defining characteristic of Oxford moral philosophy was its rejection of ‘the idea that any class of actions, such as murder, may be absolutely excluded’.167 Hare, along with Ross, had been the unnamed target. Here, in her February 28, 1957, letter to the editor, Anscombe for the first time uses the term ‘consequentialist’, names it as THE defining feature of Oxford moral philosophy, and applies it to Hare. And in this response, as with her response to Nowell-​Smith, 163 

Elizabeth Anscombe, ‘Oxford Moral Philosophy’, letter to the editor, The Listener, March 21, 1957, 479. 164  R. M. Hare, ‘Oxford Moral Philosophy’, letter to the editor, The Listener, February 21, 1957. 165  ‘I won’t sue Mr. Hare for suggesting I give lecherous talks on the wireless: I realize the cause is just a classical education’ (Anscombe, ‘Oxford Moral Philosophy’). 166  Anscombe, ‘Oxford Moral Philosophy’. 167  TD, 71.

266   John Berkman her focus remains on the fact that their moral theories refuse to rule out the moral legitimacy of murder. Anscombe successfully puts Hare on the defensive. In his second letter Hare begins by explaining how Anscombe has misunderstood his use of ‘lubricity’.168 He isn’t questioning her morals. He then gives an earnest answer to her challenge. With regard to the morality of the hydrogen bombs, Hare says it depends on whether it brings about the best consequences. Hare then incredulously asks, ‘What else could possibly be the right answer to that question? Seek to produce the worst consequences’? Finally, he expresses exasperation that Anscombe has not answered his questions, but merely called me a ‘consequentialist’. She does not explain what ‘consequentialism’ is. If it is the opinion that we are morally responsible for what we willingly and wittingly bring about, is it so damnable a heresy?169

In her earlier letter of February 28, Anscombe had invented the term ‘consequentialist’, defined it as the approach to ethics which refuses to rule out tout court the moral rightness of mass murder or any other sort of action, and applied the term to Hare, all in two sentences. Hare, never having heard the term before, either did not understand or refused to acknowledge her definition. For Anscombe, consequentialism as she had defined it, not how Hare misunderstood it, was indeed a most damnable of moral heresies.

10.7.4 Summary of Anscombe’s Exchange with Nowell-​Smith and Hare By the end of her exchange with Nowell-​Smith and Hare, Anscombe considers herself to have confirmed them as holding the views which she outlined in her radio broadcast. She has gotten Nowell-​Smith to reveal that he thinks that even though Truman gave the order to drop the bomb, ‘it was not altogether obvious’ that we can attribute to him the killing of Japanese. It is bad enough that an Oxford historian like Bullock makes such an absurd claim, but here is an Oxford moral philosopher doing the same. This despite the fact that, as Anscombe never ceases to point out, Truman himself had proudly proclaimed his full responsibility for dropping the 168  Hare had begun his first (February 21) letter, ‘Owing to the lubricity of [Anscombe’s] style’. One meaning of ‘lubricity’ is something that is lascivious, intended to arouse sexual desire. Hare’s second (March 28) letter begins, ‘Lubricity: slipperiness, shiftiness, oiliness’ . . . OED . . . so there was no call to take the word sensu obsceno’. 169  R. M. Hare, ‘Oxford Moral Philosophy’, letter to the editor, The Listener, March 28, 1957. Contrary to the common claim that Anscombe first used the term ‘consequentialist’ in ‘Modern Moral Philosophy’, her letter of February 28, 1957, in The Listener—​published almost a year prior to MMP—​contains her first use of ‘consequentialist’. It is notable that it is used specifically to characterise the moral perspective of R. M. Hare.

Justice and Murder    267 bombs on Hiroshima and Nagasaki. Furthermore, Truman had done so in all the major English newspapers only a few months earlier when he came to Oxford to collect his honorary degree. As for Hare, he is not denying that Truman is responsible for murdering, that is, purposefully and deliberately choosing to instantly obliterate 100,000+​innocent civilians, mostly women and children, as a means to his ends. Hare is arguing that it can be and in fact was morally right for Truman to engage in such mass murder. Of course Hare avoids calling this ‘murder’, but if pressed that murder is the killing of innocents, Hare will defend murder ‘in some instances’.170 This is what leads Anscombe to speak of Hare as openly a consequentialist. In the midst of her response, Anscombe also notices how Nowell-​Smith and Hare manipulate the notion of ‘moral principles’. She says that Nowell-​Smith speaks of employing a large stock of principles and then choosing the one that lets him do what he wants to do, whereas Hare constantly modifies his principles, depending on what produces the effects he likes. Anscombe recognises that Nowell-​Smith’s and Hare’s appeal to ‘principle’ derives from a law conception of morality. Like Kant, they reject divine law and follow a ‘Kantian law within’. Anscombe characterises Kant’s notion of giving oneself the moral law as ‘bosh, and incoherent bosh at that’. Against it, she argues that any form of law, including human legislation, has to be a communal enterprise. The notion of ‘legislating for oneself ’ is incoherent bosh because one can at any moment modify one’s personal ‘law’, and the vote is always 1–​0 in favour of the change!171 As Anscombe memorably puts it, Kant’s ‘legislating for oneself ’ is like Don Quixote’s sidekick Sancho Panza giving himself whacks on his own bottom! As Anscombe sees that Ross’s prima facie duties are not actually duties, so Hare’s and Nowell-​Smith’s ‘principles’ are not actually principles, and their ‘moral’ ought is vacuous, or at best useful for tyrants to tyrannise and moral scofflaws to think of themselves as morally superior rather than immoral. She will repeat and expand this rejection of the ‘moral ought’ of the Oxford moral philosophers in MMP.

10.8 Conclusion Anscombe notes in her March 21 response to Nowell-​Smith and Hare that they had given her excellent material for another broadcast talk! She wasn’t kidding. She is so pleased by the outrage her talk generates that even before Hare and Nowell-​Smith send their first replies Anscombe drafts another broadcast talk, titled ‘Principles’, and submits 170 In Language of Morals, Hare avoids almost completely the terms ‘murder’ and ‘kill’. In later

works he will write of situations in which murder is the right act. See, e.g., R. M. Hare, Moral Thinking (Oxford: Oxford University Press, 1981), 135. 171  ‘The phrase “legislating for oneself ” is an absurd one; whatever a man does “for himself ” is not legislating’. Anscombe, ‘Oxford Moral Philosophy’, letter to the editor, The Listener, April 4, 1957.

268   John Berkman it to the BBC.172 However, behind the scenes Hare and Nowell-​Smith are making a ruckus, objecting to her giving any more talks. So Anscombe is being quite literal when she writes that ‘their boisterous complaints had scared off the BBC’.173 So in April 1957, her exchange with Nowell-​Smith and Hare in The Listener ended, Anscombe starts work on a more systematic response to their moral philosophies. She is invited to address the Voltaire Society, a quirky dining society in Oxford where she can expect the attendance of some friends and at least a few sympathetic colleagues, though certainly not an entirely friendly audience.174 As she prepares her address Hare’s defence of consequentialism and Nowell-​Smith ‘colossal’ difficulty with action-​descriptions and intention are still on her mind. As with her previous discussions of the Oxford moralists, foremost on her mind as she writes about ‘modern moral philosophy’ is how it goes about legitimating murder. Rather than satire, ‘Modern Moral Philosophy’ takes on the genre of complaint, and in this paper Hare and Nowell-​Smith will be explicitly named as the objects of her complaint. In this chapter I have sought to clearly delineate the background leading up to ‘Modern Moral Philosophy’ in terms of Anscombe’s character, her moral concerns, and her prior writings. Placing MMP in these contexts, I have argued that the driving rationale that led her to write MMP was her objection to Oxford moral philosophers’ inability to rule out ‘murder’ as always wrong. It is important to note that Anscombe’s definition of murder is purely descriptive: ‘to choose to kill the innocent as a means to your ends’. She attacks Nowell-​Smith because his philosophy tells ‘him that ‘murder’ is a term partly ‘evaluative’ (he probably thinks it means ‘reprehensible killing’) and the same for ‘innocent’, and so they can’t be ‘inferred’ from a mere ‘description’ and similar stuff ’.175 While I have not been able to do the relevant analysis in the confines of this chapter, Anscombe’s Intention is another work that arose from her protests against Truman and the arguments of his defenders about whether Truman’s signing an order made him responsible for the act of murdering innocent Japanese.176 Intention is centred 172 

Anscombe’s letter to the BBC enclosing a draft of ‘Principles’ is postmarked February 20, only six days after her first talk is printed in The Listener. 173  Anscombe concludes her March 21 letter, ‘I should have liked to entertain the public with all this enriched with further instructive quotations; but the Third Programme has been scared off ’. Apparently this was due to pressure brought by Hare and others in Oxford. Hare’s complaints about Anscombe came not only after her BBC talk. Prior to Anscombe’s recording it, Hare had been shown a manuscript of the talk and tried to get the BBC to scuttle it. After Anscombe’s talk was aired on February 5, Hare’s initial reaction was to ignore it so as not to give it further publicity. Something evidently changed his mind, for which moral philosophy can be thankful, since in doing so Hare inspired Anscombe to invent the term ‘consequentialist’. 174  For details of the Voltaire Society and Anscombe’s relation to it, see John Haldane, ‘Anscombe: Life, Action and Ethics in Context’, Philosophical News 18 (2020), 26–​27. 175  Elizabeth Anscombe, letter to the editor, The Listener, March 21, 1957, 478. 176  Mary Geach believes this is the case, saying that the Truman episode ‘made her give her course of lectures which became the book Intention’ (GG1, xiv). Further evidence that Anscombe wrote Intention after her protest against Truman is mixed. An argument against it is the introduction to Intention itself, where Anscombe says the research that resulted in Intention was supported by the Mary Somerville fellowship, which had ended in 1952, and her Rockefeller Foundation grant, which began in 1951. On

Justice and Murder    269 around an extensive example of appropriate act-​descriptions with regard to a complex poisoning plot, reminiscent of Anscombe’s example to Nowell-​Smith of poisoning one’s aunt. Anscombe’s preoccupation with the concept of murder would continue long beyond MMP, in fact for the rest of her life. The same year as she published MMP Anscombe wrote a critique of Glanville Williams’s understanding of murder in his book The Sanctity of Life and the Criminal Law, and in the early months of 1958 wrote three articles or letters to the editor all focusing on the question of murder in relation to the use of nuclear weapons. Anscombe’s ‘War and Murder’ may well have been written in 1958 or 1959.177 In the early 1960s she would take up the question of murder in essays like ‘Authority in Morals’ (1961), ‘Two Kinds of Error in Action’ (1963), ‘Thought and Action in Aristotle’ (1965), and ‘Who Is Wronged: Philippa Foot on Double Effect’ (1967). After she took up her professorship at Cambridge in 1970, Anscombe ran a seminar, ‘On Killing Human Beings’, for at least four years. Results of work in this period included her articles ‘On the Nature of Justice in a Trial’ (1972), ‘On the Source of the Authority of the State’ (1978), and ‘Prolegomenon to the Pursuit of the Definition of Murder’ (1979), as well as numerous essays that remained unpublished.178 For Anscombe, the root failure of Oxford moral philosophy was to cede priority to a moral theory over fundamental and pre-​philosophical moral convictions, such as that killing innocent people is inherently wrong. Following the wisdom of great philosophical traditions, and also that of religious traditions such as Judaism and Christianity, for Anscombe the prohibition against killing innocent people as a requirement of natural justice must be a presupposition of legitimate theorising about morality. As Anscombe

the other hand, in her report to the Rockefeller Foundation at the end of 1954, besides six possible Wittgenstein editing and translation projects, Anscombe lists her own ongoing projects (e.g., her book on Wittgenstein’s Tractatus and work on Aristotle) but mentions no project on intention or even in the philosophy of psychology. Furthermore, the fact that she says ‘the greater part of [Intention] was delivered as a course of lectures in the Hilary Term [January–​March] of 1957’, and that this was her first series of university lectures in the philosophy of psychology is the strongest evidence that Intention was written in the second half of 1956. In addition, Anthony Kenny, who first got to know Anscombe in the fall of 1957, recited to me evidence that Anscombe at that time was well-​known for writing lectures and articles at the last minute. 177  In writing various letters to the editor in early 1958, Anscombe found herself arguing in defence of letters by Walter Stein. Stein would edit Nuclear Weapons and Christian Conscience in which Anscombe’s ‘War and Murder’ appeared in 1961. However, the book had an extended gestation: at least one author (Roy Shaw) had to pull out of the project, and there were repeated failed efforts to receive an imprimatur. Herbert McCabe was outraged by its rejection on political and not theological grounds. At least one other essay was submitted by 1958 (Roger Smith’s), so it is certainly possible that Anscombe contributed her essay at that time. 178  Other Anscombe essays that discuss murder, and this is not by any means a complete list, include ‘Morality’ (1982), ‘Two Moral Theologians’ (1980s), ‘Action, Intention, and Double Effect’ (1982), ‘The Dignity of the Human Being’ (1980s), ‘Christians and Nuclear Weapons Designed for the Destruction of Cities’ (198), ‘Murder and the Morality of Euthanasia’ (1994), ‘Sins of Omission: Non-​Treatment of Controls in Clinical Trials’ (1983).

270   John Berkman wryly notes, the idea that she should have to defend the prohibition of murder in light of Nowell-​Smith’s moral theory was in a high degree comic. She would also have added that it showed Nowell-​Smith’s moral degeneracy. For if recognising the inherent wrongfulness of killing innocent people is not a prerequisite for granting legitimacy to an individual’s moral thinking, what Nero, what Genghis Khan, what Hitler, or what Stalin could we legitimise next?

Chapter 11

An sc ombe on E u t ha nasia as Murde r David Albert Jones

11.1  Introduction Elizabeth Anscombe was opposed in principle to the legalisation of euthanasia, even if this were, if it could be, restricted to voluntary euthanasia of a person whose condition ‘is so irremediably wretched that it is fortunate for him “to die.” ’1 A key argument that she deploys in several places, worked out in most detail in a paper titled ‘Murder and the Morality of Euthanasia,’2 is that the prohibition of euthanasia is implied by the prohibition of murder. The structure of her argument seems to be as follows:

1. 2. 3. 4. 5.

1 

The ‘hard core of the concept of murder’ is ‘the intentional killing of the innocent.’3 Euthanasia is ‘one kind of [intentionally] killing innocent people.’4 Therefore, euthanasia is murder. ‘There is an absolute prohibition on murder.’5 Therefore, there is an absolute prohibition on euthanasia.

MME, 268. MME was produced for a working party on euthanasia convened by the Linacre Centre for Healthcare Ethics (renamed in 2010 the Anscombe Bioethics Centre). The report of the working party was published in 1982 and is reprinted in Luke Gormally (ed.), Euthanasia, Clinical Practice and the Law (London: Linacre Centre, 1994), 1–​107. Chapter 3 of that report was largely based on this paper, with some contributions by others, but the version published in GG1 is the original version of which Anscombe was the sole author. 3  MME, 262. The phrase ‘kill[ing] the innocent’ recurs elsewhere in Anscombe’s work; see MMP, 33–​ 34; AIDE, 225. 4  MME, 268. 5  MME, 264; see also MME, 274–​276; AIDE, 221–​225. 2 

272   David Albert Jones Anscombe does not state explicitly that ‘euthanasia is murder’ in this paper or elsewhere (3), but the proposition follows directly from how she characterises the paradigm case or ‘core of the concept’ (1) and the nature of euthanasia (2), and it is the middle term between the prohibition of murder (4) and her conclusion that euthanasia is or ought to be prohibited (5). It is also evident in the title of the paper and in the structure of the text, for the discussion of euthanasia comes after a more general discussion of murder and is followed by a discussion of killing by omission. This last section of the paper begins, ‘[T]‌o complete our account of murder, we need to consider omissions.’6 The discussion of euthanasia is thus part of Anscombe’s ‘account of murder.’ As expressed in the form just outlined, this argument is valid. If the premises (1), (2), and (4) could be established, it would constitute a strong argument against legalisation of euthanasia. However, all three premises are contested.

11.2  Defining Murder In the Nicomachean Ethics, Aristotle asserts that, while virtue characteristically is a mean between vices, some actions and passions ‘have names that directly connote depravity.’7 He gives as examples of actions adultery, theft, and murder. Jonathan Barnes, amongst others, interprets this passage as equivalent to saying that murder ‘is wrong by definition.’8 If the wrongfulness of murder is included in its definition, this would establish one premise in Anscombe’s argument (4), but at the cost of undermining the first premise (1). For on Barnes’s account the question of whether an act falls within the concept of murder cannot be established without first demonstrating that it is a wrongful killing. Thus euthanasia cannot be accounted as murder unless it were first shown to be wrongful killing, which is precisely the question at issue. If murder is wrong ‘by definition,’ in the sense implied by Barnes, then Anscombe’s argument against euthanasia is either question-​begging or trivial. On Barnes’s reading, Aristotle’s assertion is not a substantive judgement about the morality of an action but is an analytic judgement about how the word ‘murder’ is used. However, later in the same passage, Aristotle states that actions with these names are wrong not by way of exceeding nor by way of falling short of a mean. For some actions there is no ‘mean.’ Thomas Aquinas in his commentary on this passage says that such actions are wrong in themselves and not (only) by excess or defect: secundum se sunt mala; et non solum superabundantia ipsorum vel defectus.9 Aquinas takes Aristotle to be saying that actions that are bad secundum se cannot be rendered good by moderating or 6 

MME, 271.

7 Aristotle, Nicomachean Ethics, trans. J. A. K. Thomson (Harmondsworth: Penguin Classics, 1976),

II.6. 1107a. 8  Jonathan Barnes, ‘Introduction to the Ethics of Aristotle’, in Aristotle, Ethics, 23. 9  Thomas Aquinas, Commentary on Aristotle’s Ethics, Book II, Lecture 7. 329.

Anscombe on Euthanasia as Murder    273 tempering when or where or how frequently or in what way they are performed.10 It is not only that certain names are used to signify actions as being bad but, also, that there are certain actions that are bad in themselves. There are actions that lack some necessary feature or include some contradictory element in relation to the telos of the action and the kind of action that it is, and there are names that pick out these actions. Anscombe, in an unpublished lecture, argues that seeking to include wrongness in the definition of murder would render the condemnation of murder vacuous: If that should be true, ‘Thou shalt do no murder’ only means ‘Thou shalt not kill human beings in cases where thou shalt not.’ Then the commandment, however venerable, will be no sort of contribution to an argument that something is damnable because it is murder.11

Even if murder were always wrong, this need not imply that the wrongness of murder is part of its definition. The same can be stated in relation to adultery. The term ‘adultery’ refers to voluntary sexual acts between a married person and someone who is not that person’s spouse. It can be characterised in this way by reference to intentional actions abstracting from the question of whether it is unjust or unchaste of itself or whether it is harmful to others or wrong for some other reason. Adultery represents a kind of betrayal and typically causes other harms, and hence the word carries with it negative connotations. For this reason, those who seek to excuse or justify extramarital sexual relations typically avoid the word. Nevertheless, the question ‘Is adultery ever justified?’ remains a meaningful one. On Anscombe’s view, the assertion ‘adultery is wrong in general but there may be extraordinary circumstances that justify it’12 is false. It can only be false and non-​trivial because adultery is not ‘wrong by definition’. Anscombe makes a good case for saying that the wrongness of murder (whether it be regarded as always wrong or simply wrong prima facie and for the most part) should not be included in a definition of murder. Rather, paradigm cases of murder can be identified by reference to intentional human action prior to the moral evaluation of this kind of action. Hence, she seeks to identify the hard core of the concept of murder (premise 1) rather than defining murder as wrongful killing. A problem remains, however, as Anscombe acknowledges, that murder is ‘a complex and high-​level action description.’13 It includes a central case or cases and a penumbra 10 

Anscombe uses the phrase ‘intrinsically evil/​wrongful/​unjust’ (for example, MME, 266; MMP, 38–​ 39; AIDE, 222), which derives from the early modern terminology intrinsece malum, but there is a clear overlap between this concept and Thomas’s use of malum secundum se. 11  Elizabeth Anscombe, ‘Killing and Murder: Introductory,’ Anscombe Interim Archive, File 511, p. 2, quoted in Luke Gormally, ‘On Killing Human Beings,’ in The Moral Philosophy of Elizabeth Anscombe, ed. Luke Gormally, David Albert Jones, and Roger Teichmann, St. Andrews Studies in Philosophy and Public Affairs (Exeter: Imprint Academic, 2016), 134n4. 12  Mark D. White, ‘Is Adultery Ever Justified?,’ Psychology Today, July 3, 2012, https://​www.psyc​holo​ gyto​day.com/​gb/​artic​les/​201​207/​is-​adult​ery-​ever-​justif​i ed. 13  PPDM, 26; see also MME, 261; AIDE, 219.

274   David Albert Jones of cases around this but such that some still fall squarely within the concept of murder broadly construed. Nevertheless, this ‘penumbra is fuzzy at the outer edges,’14 so that ‘there are borderline cases’ even given a broad conception of the term. In the penumbra, but still acts of murder, in Anscombe’s opinion, are actions where the death is not intended but where it occurs in the context of some other vicious intention, such as the action of an arsonist or a terrorist who is indifferent to the number of people killed. Anscombe regards some cases of omission of life-​sustaining treatment, where there is foresight but no intention of hastening death, as lying on the fuzzy border. The wrongfulness of ‘letting die’ will depend on the ‘particular moral necessity of saving life.’15 There are such moral necessities especially for someone with skill and knowledge and a role dedicated to healing. Nevertheless, positive obligations typically admit of limits or exceptions, and that is why the decision not to save a life that could be saved ‘is a deep and important question of medical ethics.’16 It is a borderline case. ‘Nor need the borderline be between murder and manslaughter; it may be between murder and innocence.’17 What does not sit on the border, on Anscombe’s account, is intentional killing of the innocent, whether by action or by omission. This is always murder.

11.3  Why Only the Innocent? The term in Anscombe’s description of the paradigm case of murder that is most in need of further explanation is ‘innocent.’ Why should the word ‘murder’ be restricted to the killing of human beings who are innocent? If the term referred to innocence of all moral fault, then there are few if any adults of sound mind who would qualify. However, if ‘innocent’ in this definition relates specifically to actions that deserve death, then the definition becomes ‘murder is the intentional killing of people who do not deserve to be killed.’ It needs to be explained why this definition does not suffer analogous problems of circularity to those that Anscombe identified in the definition of murder as ‘wrongful killing.’ It is clear from her discussion that Anscombe has a narrower definition of ‘innocent’ in mind. Her specific concern is to make space for the possibility of justified intentional killing in war, policing, and ‘perhaps’ capital punishment. The term ‘innocent’ has long been used in a technical sense in these contexts: in a military context, to refer to non-​ combatants, those not engaged in waging war; in a civilian context, to non-​aggressors, those not engaged in unjust acts of violence.18 In a criminal context, ‘innocent’ simply means not guilty of the specific crime with which one has been charged.19 14 

AIDE, 222; see also MME, 262, 274. MME, 283. 16  Elizabeth Anscombe, ‘Commentary on John Harris’s “Ethical Problems in the Management of Some Severely Handicapped Children,” ’ in GG1, 280. 17  MME, 286. 18  WM, 53. 19  A circumstance invoked repeatedly in MMP, 34, 38, 40. 15 

Anscombe on Euthanasia as Murder    275 An alternative way that Anscombe marks the distinction she seeks to make is by use of the term ‘private.’ Thus, ‘you don’t have to know what some private person killed his uncle for in order to know he committed murder . . . (I say ‘some private person’ to exclude the cases of capital punishment, cops and robbers, and warfare all in a lump).’20 A ‘private’ action, on this account, is ‘not an action belonging to the exercise of civil authority.’21 This helps clarify what is meant by ‘innocent’ in the sense used in Anscombe’s definition of murder. All human beings are innocent except those who are engaged in unjust acts of violence of a kind that is justly resisted by civil authority. The terms ‘innocent’ and ‘private’ have different meanings, but they mark the same distinction in Anscombe’s thought because the use of lethal violence against unjust aggression, i.e., against the non-​innocent, is reserved to public authorities, i.e., by the non-​ private. For the private person all human beings are innocent. Similarly, when a public officer of the state kills an innocent person (i.e., a non-​combatant or non-​aggressor), he or she is not acting with the authority of the state and the action ‘is necessarily a private one.’22 A soldier who intentionally kills a civilian commits unlawful killing, and this is not a justified act of war but a crime, as it would be for any private person. Similarly, a police officer acts unlawfully if the use of lethal force is disproportionate or is motivated solely by personal enmity. Anscombe departs from other contemporary philosophers and theologians23 in reserving to civil authorities the intentional use of lethal force. She does not regard self-​ defence as a justification for intentional killing. Only defence of the order of justice by officers of civil authority justifies intentional killing. On this view, which Anscombe also finds in her reading of Thomas Aquinas,24 self-​defence allows only for the use of weapons that typically are not lethal. It does not permit a private person to use means that typically are lethal (for example, shooting someone in the head) or that aim at death (for example, smothering an abusive partner in his sleep), even if the threat to one’s own life is real. She acknowledges that ‘[this] doctrine is severe.’25 Anscombe’s writings on the right of the state to use violence tend to conflate actions of the police with those of the military. In practice, these differ in important characteristics so that serious reason is needed before imposing martial law on the civilian population or bringing in the army to quell civil unrest. Nevertheless, she has a reasonable case that if the police are to resist wrongdoers who are willing to use lethal force, then the police

20 

AIDE, 219; see also PPDE, 256, 258; MME, 263; WM, 54. AIDE, 218; see also MME, 266. 22  MME, 266. 23  For example, Timothy Chappell, ‘What Have I Done?’, Diametros 38 (2013): 109n37; Jeff McMahan, ‘Innocence, Self-​Defense and Killing in War’, Journal of Political Philosophy 2, no. 3 (1994): 193–​221. Germain Grisez agrees with Aquinas and Anscombe that the tradition rules out intentional killing in self-​defence but differs from Anscombe (and, arguably, from Aquinas) in not regarding the intentional use of (typically or even inevitably) lethal means as intentional killing. See Germain Grisez, Living a Christian Life (Quincy, IL: Franciscan Press, 1993), 473. 24  AIDE, 228–​229. 25  AIDE, 228. 21 

276   David Albert Jones also will need to be able to meet force with force. This is at least analogous to the use of military force against an external enemy. In both cases, the practice of violence by and for the state is justified only when necessary to resist violent injustice. Both practices provide the occasion for excessive, indiscriminate, or otherwise unjust actions, but these abuses do not undermine the possibility of the just use of lethal force: For the foundation [of the right of the state to use violence] is the human need of protection against unjust attack. The fact that the attack is unjust secures the justice of supplying the need by the use and threat of violence. The need creates the first task of government, which cannot be supplied in a large society without laws and an administration of justice.26

Anscombe is more cautious in relation to capital punishment, as shown, for example, by the way she qualifies statements on this topic with ‘perhaps’27 and ‘it may be.’28 She is clear that capital punishment need not imply a disregard for the dignity of human nature. Indeed, she argues that to say that someone deserves to die is to regard him or her as having a dignity not possessed, as far as we know, by any other animal: To regard someone as deserving death is very definitely regarding him, not just as a human being but as endued with a dignity belonging to human beings, as having free will and as answerable for his actions.29

Nevertheless, exonerating capital punishment of the charge of violating human dignity is not equivalent to justifying the practice. Anscombe acknowledges that someone ‘may have reason against it,’ and she regards some ‘grisly forms’30 of execution (including the gas chamber and the electric chair) as contrary to human dignity. She also notes that a murderer who kills out of vengeance may also regard his victim as ‘deserving death’31 and therefore not violate human dignity directly, but, in her view, any killing by a private person falls under the prohibition of murder. What is necessary to justify the practice of capital punishment is a demonstration that the state needs to have this power if it is to protect people from unjust attack. This necessity is evident when it comes to the police having access to lethal weapons to confront criminals with lethal weapons. It is not evident in relation to capital punishment. Nevertheless, even though Anscombe does not make the case for instituting capital punishment, she is surely correct to exclude capital punishment from the core meaning of murder. If the legal system is otherwise just, even those who oppose the death penalty do not thereby account the executioner a murderer. 26 

MME, 265; see also WM, 51–​52. MME, 267. 28  MME, 261. 29  DHB, 79. 30  DHB, 80. 31  DHB, 79–​80. 27 

Anscombe on Euthanasia as Murder    277

11.4  Is Euthanasia Murder? Anscombe has identified key elements that characterise the hard core or central case of murder. There are murders outside this central case, such as murder where death was not intended, but these are not relevant for her argument. Essential to her argument is the claim that all actions that fall within this core of the concept can thereby be classified rightly and fittingly as murder. The first problem with describing the core of the concept of murder as ‘the intentional killing of the innocent’ (premise 1) is that it does not include the further qualification ‘by another.’ As it stands, the formulation applies to cases of suicide, except perhaps where the intentional self-​killing is a form of capital punishment (such as the act of Socrates in drinking hemlock). All other cases of suicide would then be examples of ‘self-​murder,’ the one person both murderer and murdered. This view is not novel and has had its defenders. Before the Latinism ‘suicide’ was coined in the seventeenth century and popularised in the nineteenth as part of a move to see the phenomenon in more clinical terms, the phrase ‘self-​murder’ was common in English.32 The thought underlining this language may be traced back at least to the fourth century ce. Augustine of Hippo interpreted the commandment ‘Thou shalt not kill’ as prohibiting any intentional killing of a human being unless authorised directly by God (roughly equivalent to Anscombe’s sense of ‘private’ killing) and thus as applying both to one’s neighbour and to oneself, ‘for to kill oneself is to kill a human being.’33 It is noticeable, however, that Anscombe does not use the term ‘self-​murder,’ nor does she overtly describe suicide as murder. Indeed, earlier in the paper she states that, according to one conception of murder, ‘there is an absolute prohibition on murder’ and that ‘up to now (forgetting about abortion, which has never been legally classified with murder) [this has] been the stance of English law.’34 Anscombe adverts to the fact that even before the Abortion Act of 1967, English law had never classified the crime of abortion as murder, but she neglects to mention ‘the stance of English law’ in relation to suicide. At the time Anscombe was writing suicide was not subject to any legal sanction, and prior to the Suicide Act of 1961, while a felony (under the heading felo de se) it was not generally35 classified with murder. 32 

Alexander Murray, Suicide in the Middle Ages, volume 2, The Curse on Self-​Murder (Oxford: Oxford University Press, 1998); Brian Barraclough and Daphne Shepherd, ‘A Necessary Neologism: The Origin and Uses of Suicide’, Suicide and Life-​Threatening Behavior 24, no. 2 (1994): 113–​126; Andreas Bähr, ‘Between “Self-​Murder” and “Suicide”: The Modern Etymology of Self-​Killing’, Journal of Social History 46, no. 3 (2013): 620–​632. 33 Augustine, City of God, trans. Henry Bettenson (London: Penguin, 1984), I.20, p. 32. 34  MME, 264. 35  It is necessary to say ‘generally’ as some earlier legal scholars in the English common law tradition had invoked the idea of self-​murder, but this was not reflected in Roman law nor consistently applied in the English common law.

278   David Albert Jones The murderer commits a grave injustice against the victim. Murder is wrong principally because the victim is wronged, though utilitarians ‘are here compelled to talk only of the disturbing effect of murder on the living.’36 Murder is therefore ‘a topic which has to do with equality in a deeper way than any equality of shares in material advantages.’37 However, justice concerns relations with others. In the case of suicide, no injustice is committed by the perpetrator against the victim, for they are one and the same.38 One cannot be unequal with, or unjust to, or do wrong to oneself. The victim is not wronged, and if the act is wrong this will be for other reasons, including but not limited to the secondary concerns Anscombe mentions, that is, ‘the disturbing effect . . . on the living.’ In practice, the victim of suicide is typically impulsive, and the stressors that make suicide attractive simultaneously undermine the voluntariness of the act. But even if it is considered, in abstract, as a kind of voluntary human act, it is clearly a very different kind of act than murder. It is not an injustice to the victim, is not illegal, and, even when it was illegal, it was not classified as murder. Hence, Anscombe’s description of the paradigm case of murder captures at least one kind of act, suicide, that is not fittingly described as murder. Hence, the argument, as expressed here, is not sound. Even if euthanasia falls within the core concept of murder as Anscombe characterises this, it does not follow thereby that euthanasia is murder. The first part of the argument fails, but the question remains whether, nevertheless, euthanasia may rightly and fittingly be described as a form of murder. In euthanasia, unlike suicide, there are two people, one of whom does the killing and the other of whom is killed. The question therefore arises whether the person killed is wronged as the victim of murder is wronged. It may be noted that while suicide is not and generally was not classified legally as murder, mercy killing is classified as murder, and it happens periodically that carers are prosecuted for murder (or for attempted murder where causality is in doubt or manslaughter where intention is in doubt) for ending the lives of their charges.39 Neither the motivation of the carer nor the presence of an overt request and subsequent consent by the victim is a defence against a charge of murder. The legal status quo, not only in the United Kingdom but in many parts of the world, accords with Anscombe’s view of the matter, and hence her argument merits further consideration. The use of the term ‘innocent’ in Anscombe’s definition of the core concept of murder and, what amounts to the same thing, the way Anscombe describes murder as intentional killing by a ‘private’ person, aptly describes the actions of an individual carer who decides to end the life of the person for whom he or she is caring: the individual mercy killer. However, an advocate of legalising euthanasia might object that, when euthanasia 36 

MME, 266. PPDM, 254; see also PPDM, 253–​254; DHB, 78–​79. 38  So that the intention to kill oneself is not reducible to the intention to kill someone and being the person that one intends to kill. See Michael Wreen, ‘The Definition of Suicide’, Social Theory and Practice 14, no. 1 (Spring 1988): 1–​23. 39  David Sapsted, ‘“Mercy Killing” Husband Convicted of Murder’, Telegraph, May 10, 2007; Sandra Laville, ‘Mercy Killing Mother Cleared of Murder after Helping Seriously Ill Daughter Die,’ Guardian, January 25, 2010. 37 

Anscombe on Euthanasia as Murder    279 is legalised and brought under a structure of regulation and accountability, it ceases to be a ‘private’ act. The doctor acts under the direction of civil authority and, in this sense, as a public officer of the state when acting in accordance with euthanasia law. Be this at it may,40 this sense of ‘public’ civil authority is not what Anscombe means by her contrast of public and private. Her concern is specifically with the authority of the state to use lethal force to constrain unjust violence. Euthanasia cannot be justified on this basis, and advocates of euthanasia would hardly wish to defend it on this basis. Where euthanasia is advocated or established, it is not regarded as a justifiable state-​ imposed harm but as a state-​regulated benefit, with the involvement of the state needed only to guard against abuse. Hence, the key ethical and political question is not whether euthanasia is a harmful action that could nevertheless be authorised or tolerated by the state, but whether euthanasia is a harmful action (either because harmful to the person euthanised or because harmful to others). The classic definition of murder in English law, set down by Coke in the eighteenth century. restricts murder to lethal acts committed ‘with malice aforethought’.41 This phrase implies premeditation and distinguishes murder, in its core meaning, from acts done rashly in the heat of the moment. This is a more restricted definition than Anscombe’s, as clearly it is possible to kill an innocent person intentionally without having previously thought of doing so, for example, if the killing were opportunistic. Nevertheless, euthanasia is premeditated, and so adding ‘aforethought’ to the definition would make no difference to whether euthanasia fulfils the definition of murder. Where Coke’s further specification of the act of murder is relevant to euthanasia is in relation to ‘malice’. Murder, in the central case, involves malice, that is, the intention to do harm or injury or at least the intention callously to do what is recognised as harmful or injurious. This central case is well brought out by Anscombe when she states that human dignity is violated ‘every time people are killed for others’ convenience as the Nazis killed mental defectives . . . or when anyone murders his fellow human for advantage to himself ’.42 Murder is where people ‘choose to kill the innocent as a means to their ends’.43 Killing another person for one’s own convenience or as a means to one’s ends fails to respect the most fundamental kind of human equality, equality in dignity.44 In contrast, the mercy killer or assistant in suicide may well be treating the person killed according to that person’s stated ends and not as a means to the mercy-​killer’s ends or the 40 

That there is a distinction here in practical terms should not be taken for granted. The impression that regulation and reporting requirements for euthanasia or assisted suicide are an effective means of accountability is largely illusory. See, for example, John Keown, The Law and Ethics of Medicine: Essays on the Inviolability of Human Life (Oxford: Oxford University Press, 2012), ch. 9. 41  Sir Edward Coke, Institutes of the Laws of England (London: Printed [by Adam Islip] for the Societe of Stationers, 1628). See Louis Blom-​Cooper and Terence Morris, With Malice Aforethought: A Study of the Crime and Punishment for Homicide (Oxford: Hart, 2004). 42  DHB, 78. 43  TD, 64. 44  PPDM, 254.

280   David Albert Jones suicide-​assistant’s advantage or convenience. There may be no malevolence nor any callous indifference. If the victim is suffering and requests to be killed, then the person who accedes to the request to hasten death may have no malice towards the victim. The killer may act out of benevolence. This is not to say that the judgement is correct; benevolence (wishing well) is not beneficence (doing good). Misplaced benevolence can lead to objective harm. Nevertheless, the motivation of the act, even if it causes harm, clearly distinguishes the act of mercy killing from the central case of murder. There seems to be a logical problem with construing death as a benefit, for the dead are not present to enjoy the benefit. Nevertheless, in the context of discussing omissions, Anscombe acknowledges that ‘it is possible for someone with the greatest respect for life to think it does not seem a misfortune for the patient to die soon’.45 She warns that this ‘is a deep matter which he cannot know’46 and seeks to limit this to cases where treatment is causing severe burdens and especially where the patient ‘is dying anyway’.47 Nevertheless, she acknowledges that there are cases where action in the knowledge that someone else’s death will be hastened need not be a failure to honour the deeper sense of human equality, the sense based on the common dignity of human nature. If this can be true of foreseen deaths, the question arises whether it might not also be true of some intended deaths. This is not to doubt that euthanasia and assisted suicide are both species of killing. Some advocates of ‘assisted dying’ reject the description of euthanasia as mercy killing or reject the term ‘physician-​assisted suicide’ because the person wishes only ‘to take control of their impending death,’ knowing that ‘their condition is terminal.’48 This argument gets its plausibility only from the fact that the means of death, a lethal dose of poison, is subtle in its mode of action and bears a superficial resemblance to the use of analgesia or of sedatives to relieve symptoms such as pain or agitation. If life is ended by some other means, such as being smothered with a pillow,49 then it is obvious that this is an act of killing, albeit an act ostensibly aiming at the person’s good. Similarly, someone with a terminal illness who is expected to die of that cause may instead choose to die by shooting himself. Far from excluding the possibility of suicide, ‘chronic and terminal illnesses can be a significant risk factor in suicide.’50 If someone requests and receives ‘assisted dying,’ though the person may be ‘dying anyway,’ in the sense that the person would otherwise have died of some other cause, the actual cause of death is the lethal dose. This is either administered by another, in the case of euthanasia, or self-​administered, in the case of assisted suicide. The objection to calling such actions 45 

MME, 273. MME, 273. 47  MME, 273. 48  Sarah Wootton, ‘Neither Euthanasia nor Suicide, but End-​of-​Life Choice’, Guardian, June 22, 2009. 49  As Ann Wickett alleged was done to her mother by right-​to-​die activist Derek Humphrey after her mother survived an attempt at assisted suicide. 50  Loise Bazalgette, William Bradley, and Jenny Ousbey, The Truth about Suicide (London: Demos, 2011), 9. 46 

Anscombe on Euthanasia as Murder    281 ‘killing’ or ‘suicide’ derives from the further connotations or emotional impact of these words despite, or perhaps because of, the terms themselves being accurate. Euthanasia and assisted suicide involve the intentional killing of the innocent, but they lack the element of malevolence that is characteristic of murder in the central case. Furthermore, while mercy killing is murder in English law, the act of acceding to a request to be killed by a dying patient bears much more similarity to assisting suicide in such circumstances than it does to the killing of another for one’s own advantage. Assisting suicide may well be an injustice and is a serious crime in English law, but significantly, it is distinct from the crime of murder. In English law, euthanasia is classified as murder, but viewed as a human action it falls within what Anscombe calls the penumbra of the concept of murder, and arguably it falls on or near the border between murder and assisting suicide. Thus administering poison at the person’s request (euthanasia) seems closer to supplying the person with poison to self-​administer (assisted suicide) than it does to poisoning not on request but for the poisoner’s ends (murder in the paradigm case). The final premise of Anscombe’s argument (4) is that ‘there is an absolute prohibition on murder.’ This follows easily if murder is thought of as ‘wrong by definition,’ but such a move is problematic for reasons outlined earlier. If murder is defined by reference to intentional human acts, for example, as ‘the intentional killing of the innocent,’ then it is difficult to know how to establish that murder is prohibited absolutely. Anscombe states that ‘the prohibition is so basic that it is difficult to answer the question as to why murder is intrinsically wrongful.’51 There are of course many extrinsic reasons for prohibiting murder, at least in general and for the most part: murder undermines the security and order of society; it causes fear and pain; it harms not only the person killed but a wider circle of people who were close to the victim; it is unfair and contradicts the golden rule; it cuts life short and deprives someone of positive experiences; it frustrates the hopes and plans of the person and contradicts his or her will. That murder is seriously wrong, at least in general and for the most part, is accepted very widely because murder is wrong for so many reasons. It is an archetypal injustice. On the other hand, many of these extrinsic reasons admit of exceptions. That murder is wrong also for intrinsic reasons, secundum se, as Aquinas maintained, and without exception, and irrespective of fairness or voluntariness on the part of the victim, is something that is much harder to demonstrate. In the context of the present argument, the difficulty in demonstrating that the prohibition of murder is absolute adds further uncertainty to the question of whether to classify euthanasia as murder. Inasmuch as euthanasia falls within the penumbra of the concept of murder, at or near the borderline with assisting suicide, the classification of euthanasia as murder is likely to be accepted only by someone already convinced of the wrongness of euthanasia (and of assisting suicide) secundum se. Indeed, even such a person may still doubt the aptness of describing euthanasia as ‘murder,’ just as a person

51 

MME, 266.

282   David Albert Jones who regards suicide as wrong secundum se may nevertheless reject the description ‘self-​ murder.’ For the person not already convinced of the wrongness of euthanasia secundum se, the description of voluntary euthanasia as ‘murder’ will seem both implausible and question-​begging. The argument presented at the beginning of this chapter, while valid, is not sound. It is not fitting to designate suicide as (self-​)murder, and hence the description provided is inadequate. Furthermore, even if the description of the paradigm case were adequate, Anscombe herself acknowledges that the argument also requires acceptance of a premise that is so basic that it is difficult to demonstrate: the absolute prohibition of murder. Finally, even if an adequate definition of murder could be provided and the absoluteness of the prohibition could be demonstrated, it remains that euthanasia, while classified as murder for some purposes, seems to lie on what Anscombe calls the fuzzy border of the penumbra of the concept. Someone not yet convinced that euthanasia is wrong secundum se may for that very reason doubt that euthanasia should be placed squarely within the scope of the concept. The attempt to move from the classification of euthanasia as murder to the prohibition of euthanasia seems thus to argue ignotum per ignotius. It is perhaps significant that, despite its importance for her argument, Anscombe nowhere asserts, in so many words, that suicide is self-​murder or that euthanasia is murder. The middle term is a weak link in the argument. If euthanasia lies at or near the border between murder and assisting suicide, an evaluation of the act secundum se will require some consideration of suicide. Indeed, this may offer a better route to show why the action is not to be encouraged or assisted than does the comparison with murder in the central case. Anscombe explores the commonality between euthanasia and suicide, and some of her remarks help disclose the deeper reasons for objecting to the practice. However, in this context she also develops a further argument that, whether sound or unsound, is problematic at a rhetorical or political level: the argument that euthanasia and assisting suicide should be prohibited because they are ‘irreligious.’

11.5  Euthanasia, Suicide, and a Religious View of Life Anscombe argues that ‘the considered recommendation of suicide and killing in face of suffering . . . is irreligious, in a sense in which the contrasting religious attitude—​one of respect before the mystery of human life—​is not necessarily connected only with some one particular religious system.’52 Euthanasia, like suicide, is contrary to ‘acceptance of life—​which is God’s gift—​and of death, as it comes from him’ for ‘acceptance of life and 52 

MME, 269.

Anscombe on Euthanasia as Murder    283 death is what justice is in circumstances of unavoidable dying: it is accord with God’s will.’53 Human death is ‘awesome,’54 and human life is always to be treated with a respect as ‘a sign and acknowledgement of what it is for.’55 And what is human life for? Human beings ‘are made by God in God’s likeness, to know and love God.’56 The person who recommends suicide as an escape from suffering regards life ‘as a good or bad hotel, which must not be too bad to be worth staying in [but] to the man of religious feeling, the claim lacks reverence and insight.’57 While Anscombe does not regard this religious perspective as being ‘connected only with some one particular religious system,’ her reasoning in fact relies on a particular conception of God, or rather, of the world in relation to God. She asserts that life is ‘God’s gift’; that human beings are ‘made by God’; that human beings are made ‘in God’s likeness’; and that human beings are fulfilled by coming ‘to know and love God.’ She asserts that, in at least some circumstances, the virtue of justice involves accepting ‘God’s will.’ These theological claims are common to Jews and Christians, and to an extent also to Muslims, but they go beyond what is explicit in the writings of Plato or Aristotle, and if present in the religions of India, they are expressed in quite different terms. Anscombe begins, then, not only with theological premises but with quite distinctive theological premises, associated primarily with Judaeo-​Christian tradition, and her cursory remarks provide only the basic outlines of an argument. If there is merit in her argument it will only be for those who accept these premises. Anscombe does not rely only on the fact that the intentional killing of the innocent is prohibited by a religious commandment, though this is something she asserts repeatedly.58 The contribution of her theological reflections lies more in giving explanatory force to the prohibition. The creation of human beings by God and in the image of God provides a basis for the claim that every human being possesses a unique dignity.59 The doctrines of creation and redemption reveal why acceptance of God’s will, in life and in death, is not an arbitrary imposition but is a fundamental aspect of the human situation. Suicide is typically not a free or voluntary act, but inasmuch as it is such an act, it represents not only the rejection of God’s gift of life but also the denial of death as something that comes from God. On this basis, it can be said that everyone dies either by martyrdom or suicide,60 meaning that everyone dies either with a disposition to accept death from God or with a disposition not to accept death as a given but to make death an act of self-​assertion. 53 

MME, 270. MME, 270. 55  MME, 270. 56  MME, 270. 57  MME, 270, citing David Hume’s Dialogues Concerning Natural Religion, Part XI (London, 1779). 58  For example, MME, 265; MMP, 34; WM, 61. 59  See DHB; Gormally, ‘On Killing Human Beings’. 60  On this claim, defended both by Karl Rahner and by Herbert McCabe, see David Albert Jones, Approaching the End: A Theological Exploration of Death and Dying (Oxford: Oxford University Press, 2007), ch. 6. 54 

284   David Albert Jones Anscombe’s argument, though expressed in a cursory fashion, remains relevant for those who accept its premises. It is far from redundant. For while Anscombe asserts that Christians ‘would not be recommending suicide in any form,’61 opinion polls consistently show some level of support for ‘assisted dying’ amongst practising Christians.62 Furthermore, even in the time of Voltaire there were some theologians he characterised as ‘apostles of suicide,’ and there are some contemporary theologians for whom this description is equally apt.63 The problem with Anscombe’s argument is twofold. In the first place it is persuasive only to someone who accepts her theological premises. Anscombe claims that ‘it is not necessary or good, for those who do not have a religious attitude to life, to seek to destroy it where its principles inspire our treatment of people.’64 However, while an atheist may think that religion has social utility in general, in relation to any particular issue religion can be regarded as beneficial only where it supports behaviour that the atheist has independent reason to regard as useful, right, or just. If the atheist has no good reason to think some behaviour is justified, then the fact that it is supported by people with a ‘religious view of life’ will not supply this reason. It is more likely to be thought an imposition preventing people from realising other goods, such as choice or autonomy. A second problem with Anscombe’s turn to theology at this point in her argument is that it gives the impression that the ‘irreligious’ person has no reason to rule out suicide in these circumstances. In relation to abortion Anscombe was rightly critical of a caricature of those who object to the practice as doing so for invented reasons ‘belonging to a very special religious position.’65 She asserted in reply that ‘the question whether what was inside a pregnant woman was a small human being was the same sort of question as whether what was inside a pregnant mare was a little horse, or what was inside a pregnant cat was a little kitten.’66 It is clear that an atheist can recognise abortion as homicide and can object to the practice on this basis. In relation to the prohibition against killing the innocent, what is needed is a demonstration by appeal to reason and to truths about human nature, of the absoluteness of the prohibition, but Anscombe does not provide this. Though she does not use the language of ‘the sanctity of life,’ her defence of the prohibition on theological grounds suffers from

61 

MME, 269. See, for example, Ben Clements, ‘Religion and Attitudes towards Euthanasia in Britain: Evidence from Opinion Polls and Social Surveys’, British Religion in Numbers, January 31, 2014, http://​www.brin. ac.uk/​relig​ion-​and-​attitu​des-​towa​rds-​eut​hana​sia-​in-​brit​ain-​evide​nce-​from-​opin​ion-​polls-​and-​soc​ial-​ surv​eys/​. 63  Voltaire, ‘On Suicide, and the Abbe St. Cyran’s Book Legitimating Suicide’, in Voltaire’s Philosophical Dictionary, ed. and trans. William F. Fleming (New York: E. R. DuMont, 1901), 3:19–​33; see David Albert Jones, ‘Apostles of Suicide: Theological Precedent for Christian Support of “Assisted Dying,” ’ Studies in Christian Ethics 29, no. 3 (2016): 331–​338. 64  MME, 271. 65  KRHL, 76. 66  KRHL, 77. 62 

Anscombe on Euthanasia as Murder    285 the same problem that bedevils that language.67 It suggests that the upholding of the prohibition of intentional killing by those who do not subscribe to any religious faith is ‘a legacy of their religious heritage’68 and that it lacks rational justification independent of religious considerations.

11.6  Suicide Prevention and the Logical Slippery Slope Anscombe would have done better to allude to reasons that an atheist or agnostic might have for objecting to suicide, and hence to encouraging or assisting suicide, and surely there are many such arguments.69 While suicide has its philosophical advocates, these are few, and Anscombe is right both to take Hume as characteristic and to take his conception of suicide as trivialising human life and human death. Being alive is not just visiting a hotel. One does not need a religious sense to recognise this. In contrast, suicide has been criticised on deeper and more varied grounds by thinkers both unworldly and worldly, from Plato and Aristotle to Kant and Heidegger. Neither Schopenhauer nor Camus, both of whom regarded suicide as a serious philosophical problem and both of whom rejected previous philosophical accounts of self-​killing, found in suicide an answer to the suffering and absurdity of life. Wittgenstein, who frequently contemplated suicide and three of whose four brothers died by their own hand, regarded it as a threat to the very basis of morality: ‘If suicide is allowed, then everything is allowed. If anything is not allowed, then suicide is not allowed.’70 This modern existential concern perhaps finds a precursor in the thought of Thomas Aquinas, who regarded killing oneself as wrong, first, because it violated the most fundamental of principles, shared with all substances, that of the good of existing.71 While ‘the personal desperation which may lead to suicide . . . may excite our pity,’72 there is something either shallow or disturbing about actually encouraging suicide in others. 67  See David Albert Jones, ‘An Unholy Mess: Why “the Sanctity of Life Principle” Should Be Jettisoned’, New Bioethics 22, no. 3 (2016): 185–​201. 68  Glanville Williams, The Sanctity of Life and the Criminal Law (London: Faber and Faber, 1958), 31; see John Keown and David Albert Jones, ‘Surveying the Foundations of Medical Law: A Reassessment of Glanville Williams’s The Sanctity of Life and the Criminal Law’, Medical Law Review 16, no. 1 (2008): 85–​ 126; GW. 69  See, for example, Jennifer Michael Hecht, Stay: A History of Suicide and the Philosophies against It (New Haven, CT: Yale University Press, 2013). 70  Ludwig Wittgenstein, Notebooks 1914–​1916, ed. Georg Henrik von Wright and Elizabeth Anscombe, trans. Elizabeth Anscombe (Oxford: Basil Blackwell, 1961), 91e. 71  Thomas Aquinas, Summa Theologiae 2a2ae Q. 64, art.6; also 1a2ae Q.94 art.2. 72  MME, 269.

286   David Albert Jones In practice, suicide is typically less of an action than a reaction, often impulsive, to circumstances that seem overwhelming. Nevertheless, while modern Western society tends to view the person who dies by suicide more as a victim than a perpetrator, suicide is widely regarded as a harm. The rhetoric of suicide prevention is by its nature universal: ‘every suicide is a tragedy.’73 Every suicide cuts short a life. The imperative to prevent suicide is an expression of solidarity, a determination to help people to value their lives by expressing the value that they have to others. While the tragedy of suicide is more evident in young and physically healthy people, suicide is a greater threat to the old and those with chronic sickness or disability, especially if they are socially isolated. The disability activist Alison Davis, who for many years lived with a desire for death, commented, ‘My greatest piece of good fortune was that I had friends who did not share my view that my life had no value.’74 Similarly, the bioethicist Nicholas Tonti-​Filippini, who lived with chronic illness for many years, warned that ‘the fear of being a burden is a major risk to the survival of those who are chronically ill.’75 Without the protection of the legal prohibition on encouraging or assisting suicide, the lives of disabled people are made disturbingly ‘contingent upon maintaining a desire to continue in the face of being classified as a burden to others.’76 It is noticeable that over the years since physician-​ assisted suicide was legalised in Oregon there has been a steady rise in the number of deaths and a steady rise in the proportion of these that cite, as a reason, the fear of becoming a burden to family, friends, or caregivers. This proportion has risen from 36.7 percent in the first five years to 63.6 percent in the most recent year audited (2017).77 Anscombe noticed that whereas ‘most propaganda for euthanasia assumes it should be voluntary . . . it needs pointing out that they would still think it was wronging him, but for the accompanying judgement that his condition is so irremediably wretched that it is fortunate for him ‘to die.’’78 However, if the case for euthanasia relies on the judgement that people’s lives are ‘useless or burdensome to themselves or the world,’79 73  World Health Organization, Preventing Suicide: A Global Imperative (Geneva: World Health Organization, 2014), 2. 74  Alison Davis, ‘Why Euthanasia/​Assisted Suicide Would Have Robbed Me of the Best Years of My Life’, Catholic Medical Quarterly 63 (2013): 5–​7, quoted in David Albert Jones, ‘Assisted Dying and Suicide Prevention’, Journal of Disability & Religion 22, no. 3 (2018): 298–​316. 75  Nicholas Tonti-​Filippini, About Bioethics, volume 2, Caring for People Who Are Sick or Dying (Ballan: Conor Court, 2012), 112. 76  Tonti-​Filippini, About Bioethics, 112. For an extended philosophical exploration of this argument, see David Velleman, ‘Against the Right to Die’, Journal of Medicine and Philosophy 1, no. 6 (December 1992): 665–​681; Xavier Symons and Reginald Chua, ‘“Alive by Default”: An Exploration of Velleman’s Unfair Burdens Argument against State Sanctioned Euthanasia,’ Bioethics 34, no. 3 (2020): 288–​294. See also Kevin Fitzpatrick and David Albert Jones, ‘A Life Worth Living? Disabled People and Euthanasia in Belgium’, in Assisted Suicide and Euthanasia: Lessons from Belgium, ed. David Albert Jones, Calum MacKellar, and Chris Gastmans (Cambridge: Cambridge University Press, 2017), 133–​149. 77  Oregon Public Health Division, Center for Health Statistics, Death with Dignity Act Annual Report Year 21: 2018, 2019, https://​www.ore​gon.gov/​oha/​PH/​PROVI​DERP​ARTN​ERRE​SOUR​CES/​EVA​LUAT​ IONR​ESEA​RCH/​DEAT​HWIT​HDIG​NITY​ACT/​Docume​nts/​yea​r21.pdf. 78  MME, 268. 79  MME, 269.

Anscombe on Euthanasia as Murder    287 then this judgement can equally be made ‘of people whose mental capacity is gone or much diminished.’80 Anscombe thus sketches out what John Keown would call the ‘logical slippery slope’ from voluntary euthanasia to non-​voluntary euthanasia.81 Since Anscombe wrote, this a priori argument has been buttressed with empirical evidence. In the Netherlands and Belgium hundreds of patients are killed every year by what is euphemistically called ‘life ending without explicit request.’82 It might seem that this move could be resisted if the judgement that life is useless or burdensome were reserved to the person himself or herself. However, in the first place this underestimates the degree to which the judgement of self-​worth is framed by the legislative context, so that after legalisation people who are eligible are forced to have to justify their continued existence. In the second place, if the judgement of value becomes entirely subjective, then the practice slips in the opposite direction, towards granting euthanasia or assisted suicide for less and less objective reason. The culmination of this is the facilitation of suicide in the case of a young, physically healthy person with depression. Such cases provide the most evident examples where there is a duty to seek to prevent suicide, and yet in such cases suicide can be and occasionally is facilitated by medical professionals in the Netherlands and Belgium.83 Anscombe argues that, while it is unjust and gravely wrong to aim to kill by omission, the decision to kill by positive action involves a greater clarity about one’s aim. She therefore judges that ‘the contempt for human lives is therefore greater with positive action.’84 For analogous reasons the aim of bringing about death is clearer in the case of euthanasia than assisting suicide. It is indeed possible for a doctor to prescribe painkillers knowing that there is a significant risk that the patient will use them for suicide, without the doctor wishing or intending that they are used for this purpose.85 On this basis it 80  MME, 269. Here Anscombe anticipates the argument of Mary Warnock: ‘If you’re demented, you’re wasting people’s lives—​your family’s lives—​and you’re wasting the resources of the National Health Services.” Quoted in Martin Beckford, ‘Baroness Warnock: Dementia Sufferers May Have a “Duty to Die,”’ Telegraph, September 18, 2008. 81  See John Keown, Euthanasia, Ethics and Public Policy: An Argument against Legalisation, 2nd edition (Cambridge: Cambridge University Press, 2018); David Albert Jones, ‘Is There a Logical Slippery Slope from Voluntary to Non-​voluntary Euthanasia?’, Kennedy Institute of Ethics Journal 21, no. 4 (2011): 379–​404. 82  David Albert Jones, ‘Euthanasia and Assisted Suicide in Belgium: Bringing an End to Interminable Discussion’, in Assisted Suicide and Euthanasia: Lessons from Belgium, ed. David Albert Jones, Calum MacKellar, and Chris Gastmans (Cambridge: Cambridge University Press, 2017), 254. 83  Jones, ‘Euthanasia and Assisted Suicide in Belgium’. 84  Anscombe, ‘Commentary on John Harris’, 282. 85  No less a thinker than Daniel Sulmasy seems to suggest that the physician is necessarily complicit in assisting suicide in such circumstances if he or she knows that the patient has the ‘possible intention’ of attempting suicide: ‘A physician might write a prescription for an opioid analgesic to treat pain, and the patient might surreptitiously stockpile the pills and take them in a suicide attempt. This is always a possibility with any drug that is used clinically, whether it is an opioid analgesic or digitalis. If one recognises that this is a possibility, but has no indication that this is the patient’s intention, one is not an accomplice in the suicide. However, if the patient clearly signals such a possible intention, one is an accomplice if the patient commits suicide’. Daniel Sulmasy, ‘The Rule of Double Effect: Clearing Up the Double Talk’, Archives of Internal Medicine 159 (March 22, 1999): 546. This seems unduly rigorist as a

288   David Albert Jones seems that the act of euthanasia shows a greater contempt for human life than the act of assisted suicide. Furthermore, voluntary euthanasia more readily leads to or encourages non-​voluntary euthanasia, as in both cases it is the doctor who is the principal actor. Anscombe points out that, if the rationale for euthanasia is supposed to be the dignity of human freedom and self-​determination, ‘it is inconsonant with this to ask someone else to do so grave a thing.’86 It is a curious fact that in all jurisdictions that permit both euthanasia and assisted suicide the overwhelming majority are by euthanasia rather than assisted suicide.87 Furthermore, the overall number of assisted deaths is considerably higher than in jurisdictions that permit assisted suicide but not euthanasia. There seems to be a contradiction between a law that ostensibly rests on the wish of patients to take control of the process of dying and the actual practice where most patients prefer to cede control to the doctor. This difference in rates of assisted dying also raises the question of whether requests for euthanasia may often be somewhat ambivalent, given that where people have to do the deed themselves, fewer people do so.88 These considerations seem to show that voluntary euthanasia (in Canada and the Low Countries) shows a greater contempt for human life than assisting suicide (in some states of the USA and Switzerland), and non-​voluntary euthanasia shows still greater contempt. Where voluntary euthanasia is legal there is always the potential for imperfectly voluntary or non-​ voluntary or involuntary euthanasia. Being killed without request, on the basis of a judgement that life ‘is useless or burdensome to themselves or the world,’89 is far closer to the central case of murder, and the more so the more the judgement concerns the burden to others. As the sceptical philosopher Bernard Williams wryly remarked, ‘‘He would be better off dead’ can be said for many dubious reasons: the most dubious is that we would be better off if he were dead.’90 This is not, however, to imagine that assisting suicide is benign or harmless. Suicide is not contrary to justice, but assisting suicide is unjust because it is a failure to extend the imperative of suicide prevention to people who are chronically sick or disabled. Anscombe states that murder is ‘a topic which has to do with equality in a deeper way than any equality of shares in material advantages.’91 The same is true of suicide prevention.

patient may signal this as a possible intention (i.e., a possibility the patient is seriously considering), but a physician may still prescribe the analgesia as the best way to treat the pain, hoping and intending that the patient use the drugs for the reason prescribed. 86 

MME, 269. The most recent figures for the Netherlands show 216 assisted suicides out of 5,875 assisted deaths (i.e., 3.7 percent); in Belgium 52 cases out of 6,880 (0.75 percent); and in Canada 6 out of 5,085 (0.1 percent). 88  Claudia Gamondi et al., ‘Legalisation of Assisted Suicide: A Safeguard to Euthanasia?’, The Lancet 384, no. 9938 (2014): 127. 89  MME, 269. 90  Bernard Williams, Ethics and the Limits of Philosophy (Cambridge, MA: Harvard University Press, 1985), 42. 91  PPDM, 254. 87 

Anscombe on Euthanasia as Murder    289

11.7  Conclusion Anscombe’s attempt to use the concept of murder as the basis for the prohibition of murder is, at one level, both unconvincing and unnecessary. Key elements of the core case of murder are missing from the act of suicide, and euthanasia seems to fall on the fuzzy border between murder and assisting suicide. Furthermore, Anscombe’s theological defence of the prohibition of suicide, while valid, rests on premises that are not widely accepted outside the Judaeo-​Christian tradition. Nevertheless, while the question of suicide is a deep philosophical problem, there are good reasons to consider the act a form of self-​ harm and to consider suicide prevention a duty of human solidarity. Anscombe’s reflection on the relation of human equality to the prohibition of murder can also be applied to the prevention of suicide. The social virtue of solidarity is no less important in relation to the disabled and chronically sick than in relation to the young and physically healthy. Anscombe’s insight that euthanasia relies not only on voluntariness but also on a negative view of the worthwhileness of life provides the basis for a different kind of argument: the logical slippery slope. Furthermore, her argument about the inconsonance of ceding an act to others that is ostensibly justified on the basis of autonomy, shows why euthanasia is a still more dangerous practice than assisting suicide. Euthanasia is killing by another, and voluntary euthanasia both logically and in practice leads to non-​ voluntary euthanasia. While voluntary euthanasia may not fall within the core concept of ‘murder,’ non-​voluntary euthanasia falls at least within the penumbra of the concept. Hence, Anscombe’s initial argument is justified, though via a less direct route. Legalising euthanasia logically and in practice leads to murder, in the sense of killing someone without their request or consent, and this is a good reason to maintain its prohibition.

Bibliography Anscombe, Elizabeth. ‘Commentary on John Harris’s “Ethical Problems in the Management of Some Severely Handicapped Children.” ’ In GG1, 2005. Aristotle. Nichomachean Ethics. Trans. J. A. K. Thomson. Harmondsworth: Penguin Classics, 1976. Augustine. City of God. Trans. Henry Bettenson. London: Penguin, 1984. Bähr, Andreas. ‘Between “Self-​Murder” and “Suicide”: The Modern Etymology of Self-​Killing’. Journal of Social History 46, no. 3 (2013): 620–​632. Barnes, J. ‘Introduction to The Ethics of Aristotle.’ In Aristotle, Ethics, p. 23. Barraclough, Brian, and Daphne Shepherd. ‘A Necessary Neologism: The Origin and Uses of Suicide.’ Suicide and Life-​Threatening Behavior 24, no. 2 (1994): 113–​126. Bazalgette, Louise, William Bradley, and Jenny Ousbey. The Truth about Suicide. London: Demos, 2011. Beckford, Martin. ‘Baroness Warnock: Dementia Sufferers May Have a “Duty to Die.” ’ Telegraph, September 18, 2008. Blom-​Cooper, Louis, and Terence Morris. With Malice Aforethought: A Study of the Crime and Punishment for Homicide. Oxford: Hart, 2004.

290   David Albert Jones Chappell, Timothy. ‘What Have I Done?’ Diametros 38 (2013): 86–​112. Clements, Ben. ‘Religion and Attitudes towards Euthanasia in Britain: Evidence from Opinion Polls and Social Surveys’. British Religion in Numbers, January 31, 2014. http://​www.brin. ac.uk/​relig​ion-​and-​attitu​des-​towa​rds-​eut​hana​sia-​in-​brit​ain-​evide​nce-​from-​opin​ion-​polls-​ and-​soc​ial-​surv​eys/​. Coke, Edward. Institutes of the Laws of England. London: Printed [by Adam Islip] for the Societe of Stationers, 1628. Davis, Alison. ‘Why Euthanasia/​Assisted Suicide Would Have Robbed Me of the Best Years of My Life’. Catholic Medical Quarterly 63 (2013): 5–​7. Fitzpatrick, Kevin, and David Albert Jones. ‘A Life Worth Living? Disabled People and Euthanasia in Belgium.’ In Assisted Suicide and Euthanasia: Lessons from Belgium, edited by David Albert Jones, Calum MacKellar, and Chris Gastmans. Cambridge: Cambridge University Press, 2017, 133–​149. Fleming, William F. (ed. and trans.). Voltaire’s Philosophical Dictionary. New York: E. R. DuMont, 1901. Gamondi, Claudia, et al. ‘Legalisation of Assisted Suicide: A Safeguard to Euthanasia?’ The Lancet 384, no. 9938 (2014): 127. Gormally, Luke (ed.). Euthanasia, Clinical Practice and the Law. London: Linacre Centre, 1994. Gormally, Luke. ‘On Killing Human Beings.’ In The Moral Philosophy of Elizabeth Anscombe, edited by Luke Gormally, David Albert Jones, and Roger Teichmann, St. Andrews Studies in Philosophy and Public Affairs (Exeter: Imprint Academic, 2016), 134n4. Gormally, Luke, David Albert Jones, and Roger Teichmann. The Moral Philosophy of Elizabeth Anscombe. St. Andrews Studies in Philosophy and Public Affairs. Exeter: Imprint Academic, 2016. Grisez, Germain. Living a Christian Life. Quincy, IL: Franciscan Press, 1993. Hecht, Jennifer Michael. Stay: A History of Suicide and the Philosophies against It. New Haven, CT: Yale University Press, 2013. Hume, David. Dialogues Concerning Natural Religion. London, 1779. Jones, David Albert. ‘Apostles of Suicide: Theological Precedent for Christian Support of “Assisted Dying.” ’ Studies in Christian Ethics 29, no. 3 (2016): 331–​338. Jones, David Albert. Approaching the End: A Theological Exploration of Death and Dying. Oxford: Oxford University Press, 2007. Jones, David Albert. ‘Assisted Dying and Suicide Prevention’. Journal of Disability & Religion 22, no. 3 (2018): 298–​316. Jones, David Albert. ‘Euthanasia and Assisted Suicide in Belgium: Bringing an End to Interminable Discussion’. In Assisted Suicide and Euthanasia: Lessons from Belgium, edited by David Albert Jones, Calum MacKellar, and Chris Gastmans. Cambridge: Cambridge University Press, 2017, 235–​257. Jones, David Albert. ‘Is There a Logical Slippery Slope from Voluntary to Non-​voluntary Euthanasia?’ Kennedy Institute of Ethics Journal 21, no. 4 (2011): 379–​404. Jones, David Albert. ‘An Unholy Mess: Why “the Sanctity of Life Principle” Should Be Jettisoned’. New Bioethics 22, no. 3 (2016): 185–​201. Keown, John. Euthanasia, Ethics and Public Policy: An Argument against Legalisation, 2nd edition. Cambridge: Cambridge University Press, 2018. Keown, John. The Law and Ethics of Medicine: Essays on the Inviolability of Human Life. Oxford: Oxford University Press, 2012. Keown, John, and David Albert Jones. ‘Surveying the Foundations of Medical Law: A Reassessment of Glanville Williams’s The Sanctity of Life and the Criminal Law’. Medical Law Review 16, no. 1 (2008): 85–​126.

Anscombe on Euthanasia as Murder    291 Laville, Sandra. ‘Mercy Killing Mother Cleared of Murder after Helping Seriously Ill Daughter Die’. Guardian, January 25, 2010. McMahan, Jeff. ‘Innocence, Self-​Defense and Killing in War’. Journal of Political Philosophy 2, no. 3 (1994): 193–​221. Murray, Alexander. Suicide in the Middle Ages. Volume 2, The Curse on Self-​Murder. Oxford: Oxford University Press, 1998. Oregon Public Health Division, Center for Health Statistics. Death with Dignity Act Annual Report Year 21: 2018. 2019. https://​www.ore​gon.gov/​oha/​PH/​PROVI​DERP​ARTN​ERRE​ SOUR​CES/​EVA​LUAT​IONR​ESEA​RCH/​DEAT​HWIT​HDIG​NITY​ACT/​Docume​nts/​yea​ r21.pdf. Sapsted, David. ‘‘Mercy Killing’ Husband Convicted of Murder’. Telegraph, May 10, 2007. Sulmasy, Daniel. ‘The Rule of Double Effect: Clearing Up the Double Talk’. Archives of Internal Medicine 159 (March 22, 1999): 545–​550. Symons, Xavier, and Reginald Chua. ‘ “Alive by Default”: An Exploration of Velleman’s Unfair Burdens Argument against State Sanctioned Euthanasia’. Bioethics 34, no. 3 (2020): 288–​294. Tonti-​Filippini, Nicholas. About Bioethics. Volume 2, Caring for People Who Are Sick or Dying. Ballan: Conor Court, 2012. Velleman, David. ‘Against the Right to Die’. Journal of Medicine and Philosophy 17, no. 6 (December 1992): 665–​681. White, Mark D. ‘Is Adultery Ever Justified?’ Psychology Today, July 3, 2012. https://​www.psyc​ holo​gyto​day.com/​gb/​artic​les/​201​207/​is-​adult​ery-​ever-​justif​i ed. Williams, Bernard. Ethics and the Limits of Philosophy. Cambridge, MA: Harvard University Press, 1985. Williams, Glanville. The Sanctity of Life and the Criminal Law. London: Faber and Faber, 1958. Wittgenstein, Ludwig. Notebooks 1914–​1916. Ed. Georg Henrik von Wright and Elizabeth Anscombe. Trans. Elizabeth Anscombe. Oxford: Basil Blackwell, 1961. Wootton, Sarah. ‘Neither Euthanasia nor Suicide, but End-​of-​Life Choice’. Guardian, June 22, 2009. World Health Organization. Preventing Suicide: A Global Imperative. Geneva: World Health Organization, 2014. Wreen, Michael. ‘The Definition of Suicide.’ Social Theory and Practice 14, no. 1 (Spring 1988): 1–​23.

Chapter 12

The Knowl e d g e of Hum an Di g ni t y Micah Lott

12.1  Introduction In this chapter, I bring together a number of Anscombe’s ideas concerning our knowledge of human dignity. In section 12.2, I focus on her account of the connatural knowledge of dignity. In section 12.3, I turn to Anscombe’s claim that human dignity is the dignity of human nature, and to her understanding of our nature as ‘spirit.’ In section 12.4, I look at her notion that human beings have ‘mystical value’ and the role of theology and religious feeling in understanding human dignity. In section 12.5, I consider the idea that human dignity calls for respect of a specifically directed sort: respect, and respectful treatment, that is owed to every human being. While this is an attractive thought, it is not one that Anscombe discusses in detail. I explore the notion of directed respect and its connection to human rights and natural justice. There appears to be conflict between an ideal of natural justice and Anscombe’s conventionalism about rights and duties. But this conflict, I argue, may be merely apparent.

12.2  Connatural Knowledge, Justice, and Reverence for Human Life At the beginning of her essay ‘Knowledge and Reverence for Human Life,’ Elizabeth Anscombe distinguishes between two types of knowledge: knowledge of the dignity of human nature and knowledge of indifferent truth. The latter includes ‘all that is more often called knowledge: mathematics and the natural sciences, logic and psychology, history and the things that have happened within people’s personal memories, and so

The Knowledge of Human Dignity     293 on.’1 The truths of mathematics, natural science, etc. are indifferent insofar as they do not provide you with ends for acting, or judge for or against your purposes. At the end of the essay, Anscombe says that no truth is indifferent if it affects what is to be done. And of course the truths of mathematics, natural science, etc. might affect what is to be done, in the sense that they are relevant to figuring out the means to your ends. In what sense, then, are these indifferent truths? They are indifferent on their own, because their relevance to action depends upon whatever ends you happen to have. They do not, in themselves, dictate what to do or how to live. This suggests that knowledge of the dignity of human nature is knowledge of an intrinsically or essentially non-​indifferent truth—​i.e., knowledge of what your ends should be, of good versus bad purposes. But in her essay, Anscombe does not turn directly to the topic of non-​indifferent truth. Rather she approaches that topic by first considering a different notion: connaturality. For she is interested in the connatural knowledge of the dignity of human nature, which is ‘the most important sort of knowledge of it.’2 The general idea of connatural knowledge is that some things are readily known by beings with a certain nature because of a similarity between the nature of the knowers and the things known. For example, knowledge of material substances is connatural to human beings because we ourselves are material beings (an example that Anscombe finds in Aquinas). The specific sort of connatural knowledge that Anscombe focuses on is the connatural knowledge that is proper to someone with a certain moral virtue. This is an instance of the more general idea of connaturality insofar as virtue is like a second nature, and it is because the virtuous person has such a (second) nature that she has connatural knowledge of the things pertaining to virtue. That is, she knows about things that are ‘alike’ to her, and she knows on account of this likeness—​e.g., about just things because she is just, about generous things because she is generous. What, then, does the connatural knowledge of the morally virtuous consist in? Such knowledge is ‘a capacity to recognize what action will accord with and what ones will be contrary to virtue.’3 Moreover, this recognition does not come from cleverness, but through the inclinations of the virtuous person. For example, a generous person will quickly reject—​or not even consider—​those possible actions that would be ungenerous or mean, because her generosity inclines her against them: ‘The one with connatural knowledge is inclined against the action and that inclination itself is a sort of perception of the meanness of acting even without the judgments being formulated.’4 And because such connatural knowledge comes through inclination, it is no accident when a person with that knowledge acts in accordance with these perceptions. However, while the virtuous person’s connatural knowledge comes through inclination, such knowledge involves more than good inclinations or right ends. For 1 

KRHL, 59. KRHL, 61. 3  KRHL, 60. 4  KRHL, 60. 2 

294   Micah Lott the virtuous person knows in particular what to do. She is able to deliberate and act well. And for this she requires information about the world, or what Anscombe calls knowledge of indifferent truth. Put another way, the connatural knowledge of the morally virtuous is fully present only in those with practical wisdom, and practical wisdom ‘has two sides to it, the side that connects up with information about what has happened, with scientific information, with knowledge of procedures and so on, and the side that connects up with good inclination, the inclination towards good ends.’5 Thus some of what the virtuous know, and that shapes their actions, is the sort of thing that could be found in books, or acquired from simple observation of a particular situation (‘The recognition that someone is hungry or cold is the merest observation.’)6 But other things they know—​concerning what is good and worthwhile in human life—​cannot be acquired in those ways. This knowledge arises chiefly from good upbringing and moral practice. How, then, does the connatural knowledge of the virtuous relate to the knowledge of the dignity of human nature? Is all connatural moral knowledge also knowledge of human dignity? Perhaps. But Anscombe seems to associate the knowledge of human dignity especially with the virtue of justice. The knowledge of human dignity is knowledge of the great worth of human beings, and of how humans should be respected, in light of what a human being is. It is knowledge that human life merits reverence, and that it is a grave mistake to treat humans in ways that are callous or cavalier. The just have such knowledge connaturally. This knowledge shows itself in the just person’s connaturally making certain judgements. These might be judgements about general truths—​e.g., ‘that a human being is of more worth than many sparrows.’7 Or they might be judgements about particular situations—​e.g., ‘that it is wickedness to do what the Nestlé company is reported as doing—​to persuade poor African women to feed babies their powdered milk, instead of suckling them; to value money making far above respecting the lives of human infants.’8 Moreover, the just person’s connatural knowledge of human dignity is something that can be manifested in her behaviour, even when not explicitly formulated. This happens when a just person, acting from a perceptual capacity shaped by virtuous inclinations, registers such-​ and-​such an action as to be done, or to be avoided, because of what respect for human dignity requires.

5 

KRHL, 63–​64. KRHL, 65. To this example, one might object that the recognition is not ‘the merest observation’ in any sense that contrasts with having good ends or good character. For making such an observation, or failing to make it, may well be morally significant—​e.g., the sort of observation it is characteristic of the compassionate person to make, and that the callous person typically fails to make, and that shapes their differing responses to those around them. Anscombe is sensitive to these points; see the full passage in KRHL. For more on the role of ‘plain worldly knowledge’ in practical wisdom, see Rosalind Hursthouse, ‘Practical Wisdom: A Mundane Account,’ Proceedings of the Aristotelian Society 106 (2006), 285–​309. 7  KRHL, 62. Anscombe is here quoting the words of Jesus in Matthew 10:31. 8  KRHL, 62. 6 

The Knowledge of Human Dignity     295

12.3  ‘Man Is Spirit’: The Dignity of Human Nature Knowledge of human dignity is the knowledge that human beings are of great value and worthy of respect. If there is such a knowledge, then it must be true that human beings are of great value and worthy of respect. So we can ask: What explains human dignity? What is it about human beings that gives them special value and makes them worthy of respect? Anscombe is clear that human dignity is a matter of the dignity of human nature. Individual human beings are the bearers of dignity because of the kind of thing that a human being is. Moreover, because all human beings are equal in nature, they are also equal in human dignity: ‘There is just one impregnable equality of all human beings. It lies in the value and dignity of being a human being.’9 For Anscombe, it is crucial that human beings do not have dignity on account of their being ‘persons,’ where personhood is understood as a quality, or set of qualities, that a particular human being might fail to possess because of illness or injury or developmental stage. Individual human beings are indeed persons, and we might even say that this is the basis of their dignity. But for Anscombe that is true only if we understand personhood as something that belongs to all human beings in virtue of their nature—​i.e., because each is an individual member of the species ‘human being.’ As she says, ‘A human being is a person because the kind to which he belongs is characterised by rational nature. . . . One is a person just by being of this kind, and that does indeed import a tremendous dignity.’10 If we grant that individual humans have tremendous dignity in virtue of their nature, we might still wonder what about human nature forms the basis of human dignity. Some of Anscombe’s most interesting comments on this issue come in her discussions of murder. Someone who kills another for gain, as a means to his own purposes, is a paradigm case of someone who fails to respect human dignity. Anscombe contrasts such a killer with someone who kills from vengeance. And she claims, perhaps surprisingly, that while killing from vengeance may not be acceptable, it is not eo ipso a violation of human dignity. The reason is that the vengeful killer sees his victim as deserving to die. And in that way, the vengeful killer treats his victim not as a mere thing but as a being who is accountable, in a way that stones or trees or other animals are not: ‘To regard someone as deserving death is very definitely regarding him, not just as a human being but as endued with a dignity belonging to human beings, as having free

9 

DHB, 67. MME, 268. For Anscombe, the ethical implications of a proper understanding of personhood are tremendous. This is especially so in cases of abortion and euthanasia, where people might grant that what is being killed is a human being but insist that it is not a person. That claim depends on a sense of personhood that Anscombe flatly rejects. As she says, ‘You cannot be killing a human being and not be killing a person’ (MME, 268). 10 

296   Micah Lott will and as answerable for his actions.’11 The dignity of human nature, then, seems to reside especially in our being creatures whose form of life is characterised by freedom and answerability—​and, presumably, by the whole complex of rationality and language that goes along with that. This is ‘the truth that man is spirit. He moves in the categories of innocence and answerability and desert—​one of the many signs of a leap to another kind of existence from the life of other animals.’12 As this remark shows, Anscombe does not think of our spiritual nature as involving an immaterial substance of the Cartesian sort.13 Rather our spiritual nature is our species’ characteristic orientation towards the good: the human capacity to form an understanding of the good, to guide ourselves according to the standard of goodness, and to hold ourselves and others accountable for living and acting well.14 And not only this. For Anscombe also understands our spiritual nature to include our characteristic orientation towards the truth: the human capacity to form an understanding of reality as something of which we are a part, and to make judgements that are self-​consciously subject to the standard of truthfulness.15 As Anselm Müller says, ‘[I]‌t seems right to say that the basis of our dignity is a nature that directs us towards the ends of truth and goodness by providing us with a consciousness of being subject to those ends as norms. This consciousness cannot find expression in the manifestation of a purely animal form of life; it characterizes the spiritual nature of man.’16 Anscombe’s remarks about freedom and answerability shed light on an intriguing passage in KRHL, where she suggests that some knowledge of human dignity comes through the learning of language. Speaking of the connatural knowledge of the just, she says: 11 

DHB, 68. Anscombe is clear that she is not defending the imagined vengeful murderer or saying that he has a right to kill, but merely contrasting him with the one who kills for gain. Of course, one might argue that vengeful killing involves some violation of human dignity, even if not as egregious as that involved in killing for gain. For instance, one might argue that proper respect for humans qua free and answerable beings requires holding others accountable for their actions in ways that are public, transparent, even-​handed, etc.—​and that vengeful killing is incompatible with this. 12  MME, 267. 13  For Anscombe’s criticisms of the Cartesian conception of the soul’s immateriality, see APSM, 3–​16. 14  Cf. Candace Vogler’s discussion of a non-​theological answer to the question why a human being has dignity that requires reverence—​an answer ‘that can be sketched in terms of the fact that she and I are alike in that we both are members of the species of creature that faces the challenge of the ethical’. Candace Vogler, ‘Why Human Beings Matter: Anscombe on the Nature and Point of Human Life’, in G. E. M. Anscombe and Human Dignity, ed. John Mizzoni (Aston, PA: Neumann University Press, 2016). 15  This helps to explain why attacks on mental and intellectual development can be assaults on human dignity, as when a state makes it a crime to teach a class of people to read. See DHB, 70. 16  Anselm Müller, ‘The Spiritual Nature of Man’, in The Moral Philosophy of Elizabeth Anscombe, ed. Luke Gormally, David Albert Jones, and Roger Teichmann (Exeter: Imprint Academic, 2016), 32. In an early, long-​unpublished paper, probably from the 1950s, Anscombe gives a somewhat different, and more limited, account of the human spirituality: ‘I put it forward that the spirituality of the human soul is its capacity to get a conception of the eternal, and to be concerned with the eternal as an objective, and perhaps also as something that can be leant on and feared’. Elizabeth Anscombe, ‘The Immortality of the Soul’, in Faith in a Hard Ground, ed. Mary Geach and Luke Gormally (Exeter: Imprint Academic, 2008), 74.

The Knowledge of Human Dignity     297 The spirit of such knowledge is what is called a gift of the Holy Ghost; the light of it a light that is there to enlighten everyone who comes into the world. I do not mean that everyone actually has this light on in his mind, for it may have been extinguished or never allowed to come on. It may be there as a mere glimmer, whose sign is the understanding of human language with all its multifarious action and motive descriptions, its machinery for accusing others and excusing oneself.17

This passage hints at the following line of thought. Human language brings into view what a human being is, and in particular it brings into view our nature as free and answerable. For in learning a language that describes human action and motivation, we come to grasp ourselves and others as acting for reasons, as capable of deciding what to do and as answerable for our decisions. And we are initiated into linguistically infused practices in which others hold us accountable and we hold them accountable. Such practices manifest our nature as ‘spirit,’ both in the sense that our nature makes such practices possible and in the sense that these practices disclose or reveal that nature. Which is to say, they reveal that in which the dignity of human nature especially resides. Thus to learn a language, and to participate in linguistically infused practices of answerability, brings with it a glimmer of that knowledge that shines brightly in the just—​i.e., a knowledge of the worth of human beings, and the treatment they deserve, qua bearers of a nature that is characterised by freedom and answerability. So human beings characteristically acquire at least some knowledge of human dignity by living as human beings.18

12.4  Mystical Value and Religious Outlooks 12.4.1  The Mystical Value of Human Beings At the end of ‘Prolegomenon to the Pursuit of a Definition of Murder,’ Anscombe refers to human beings as having ‘mystical value,’ and she suggests that this is part of what a sufficient consideration of murder must take into account.19 This suggests that human

17 

KRHL, 62. In relation to this, consider Anscombe’s comment, ‘No one could have the concepts corresponding to the words used in the [ten] commandments, if he had not lived in an environment in which he learns the inwardness of all sorts of ways of going on; he must live a specifically human life with human practices’ (METC, 225). For a related line of thought, consider Cora Diamond’s discussion of what goes into determining our concept of ‘human being’ in her essay “Eating Meat and Eating People’, in The Realistic Spirit (Cambridge, MA: MIT Press, 1991), 319–​334. See especially 323–​324. 19  PPDM, 260. 18 

298   Micah Lott dignity, which is grievously violated in murder, has some connection to mystical value. Elsewhere Anscombe speaks about ‘mystical perception’; as examples of this, she gives (a) the perception of the dishonour done to the body by treating sexual acts as the casual satisfaction of desire and (b) the knowledge that a dead human body deserves respect and should not be put out with the garbage. Although mystical, such perceptions are not rare. Indeed, they are ‘as common as humanity.’20 What, then, does Anscombe mean by mystical value, and what does mystical perception have to do with our knowledge of dignity? Anscombe connects the idea of mystical perception to a distinction between two types of virtue: those that are utilitarian in character, and those that are ‘supra-​utilitarian and hence mystical.’21 The former include virtues like temperance towards food and drink and honesty about property. They are fundamentally utilitarian in character insofar as their point is the promotion of the material well-​ordering of human life. And hence such virtues ‘have a purely utilitarian justification.’22 In contrast, virtues like reverence for life or chastity, although they may be useful for our material well-​being, do not have a purely utilitarian justification. The material well-​ordering of human life is not the only, or chief, point of these virtues. In what, then, does the goodness or value of such mystical virtues consist? Curiously, Anscombe says very little about this. But her idea seems to be that the mystical virtues are those that embody the proper recognition, or mystical perception, of the mystical value of human beings. And the mystical value of human beings is a matter of their great intrinsic worth, which calls for respect and reverence. This is not instrumental value. It is not a matter of the usefulness of one thing for another, or the way something serves our purposes. Rather it the sort of value that calls for us to have certain purposes or puts a limit on our purposes, insofar as those purposes are, or are not, compatible with respect for the worth of human beings. If this is correct, then the mystical value of human beings is another way of expressing the idea of human dignity. And the goodness or value of the mystical virtues lies in the way that they enable and manifest the proper appreciation of human dignity rather than in their usefulness for attaining something that humans characteristically need or desire, such as safety or physical comfort. Furthermore, what you grasp via mystical perception—​the dignity of human nature—​is directly relevant to how you should act. Mystical perception is thus the perception of an intrinsically or essentially non-​ indifferent truth. With respect to Anscombe’s distinction between the utilitarian and supra-​utilitarian virtues, there are two things to note. First, we might accept the essence of Anscombe’s distinction but reject her suggestion that it divides the virtues into two types. Rather than sorting virtues into types, it seems better to think of two aspects of many, perhaps

20 

CC, 187. CC, 187. 22  CC, 188. 21 

The Knowledge of Human Dignity     299 all virtues. Anscombe already concedes that supra-​utilitarian virtues can also be useful for our material well-​being. And it seems equally true that honesty about property, which Anscombe classifies as a utilitarian virtue, connects with values that go beyond material well-​ordering. To respect a person’s property can be a way of respecting her. And a justification of our practices around property will not only include considerations about which behaviours are useful for our material well-​being. It will also include considerations about what counts as respectful treatment of other human beings, given the kind of social and linguistic animals that we are. Indeed, one might argue that even utilitarian justifications acquire their full support, or normative force, from supra-​utilitarian considerations. After all, it is plausible to think that the material well-​ ordering of human life is so important precisely because human beings themselves are of great worth. The second thing to note about Anscombe’s distinction is that it raises the question of how claims about human dignity might be justified or supported with arguments. I mean both the general claim that human beings have dignity and also claims about which attitudes and actions are required by respect for human dignity. Now, when Anscombe speaks about a utilitarian justification for a virtue, I assume she has in mind the following sort of argument: Given the kind of creatures that human beings are, including our characteristic needs and capacities, such-​and-​such way of behaving is something upon which human good depends—​i.e., it plays a role in the characteristic life of human beings and is thereby important for securing our well-​ being. And thus such-​and-​such way of behaving is good, or excellent, in a human being. Instead of calling such arguments utilitarian, we might call them functional, since they operate by identifying the function of the virtues in human life, the part they play.23 As Anscombe points out, even when arguments like this are successful, they are not adequate to establish the mystical value of human beings. It is one thing to say that, given our needs and natural repertoire, human beings cannot get on well without certain habits of thought, feeling, and action. It is another thing to say that human nature is of great value and merits respect and reverence. If someone doubts that human beings have dignity or that dignity requires such-​and-​such behaviour, it seems misguided to argue that behaving as if human beings have dignity is important for the material well-​ordering of human life. And thus reflecting on Anscombe’s category of the supra-​ utilitarian and mystical leads us to ask: Is there any way to give arguments for the reality of human dignity or to respond to a dignity sceptic?

23  For Anscombe’s articulation of this sort of argument, see SAS, 130–​155. Drawing on Anscombe’s work, Philippa Foot endorses such functional arguments in her account of virtue as a type of natural goodness. See her Natural Goodness (Oxford: Oxford University Press, 2001). For an argument about the limits of functional arguments, see Micah Lott, ‘Justice, Function, and Human Form’, in Normativität des Lebens—​Normativität der Vernunft?, ed. Martin Hähnel and Markus Rothhaar (Berlin: De Gruyter, 2015), 75–​92.

300   Micah Lott As I understand Anscombe, her answer is: Not really. What there is, instead, is the glimmer of light that characteristically comes from learning a language and living as a human being—​’a specifically human life with human practices’24—​in which human dignity is grasped, though perhaps dimly. And if all goes well, there is also good upbringing and moral education, through which the knowledge of dignity may be imparted, especially as the connatural knowledge of the virtuous. Moreover, if you have religious faith, this may provide you a particular basis for accepting claims about human dignity, as is true for Anscombe herself.

12.4.2  Theological Outlooks and Religious Attitudes So far we have encountered two ideas that are likely to have religious overtones for many people: spiritual nature and the mystical. In each case, we have seen that what Anscombe means by these terms is not as distinctly religious as one might have thought. For instance, when Anscombe speaks about mystical perception she does not have in mind ecstatic states of union with the divine. However, she does think that there are distinctly religious and theological explanations of human dignity. This is unsurprising, since religious outlooks typically include a distinctive understanding of what a human being is. The Christian view of human beings, which Anscombe endorses, understands human beings as created ‘by God, in God’s likeness, to know and love God.’25 Human life is God’s gift, and the proper end of a human being is communion with God. Thus every human is of great worth, and all human lives worthy of respect: What people are for is, we believe, like guided-​missiles, to home in on God, God who is the one truth it is infinitely worth knowing. . . . It’s this incredible potentiality, this incredible possibility, of the knowledge of God of such a kind as even to be sharing in his nature, which Christianity holds out to people; and because of this every life, right up to the last, must be treated as precious.26

The Christian view of human beings is one development of a religious attitude towards life. One might have a religious attitude towards life in a general sense, without giving it a specific theological or religious formulation. Such an attitude involves ‘a respect before the mystery of human life.’27 The person of religious feeling regards human life and death as awesome and not to be treated in a cavalier way. An aspect of this feeling is ‘a certain fear before the idea of ever destroying a human life.’28

24 

METC, 225. MME, 270. 26  CC, 173. 27  MME, 269. 28  MME, 270. 25 

The Knowledge of Human Dignity     301

12.5  Directed Duties, Human Rights, and Natural Justice 12.5.1  Human Value and Intrinsically Relational Normativity It is clear that Anscombe holds that human dignity is the inherent value possessed equally by all human beings, and that generally speaking the proper response to this value is respect or reverence—​and not, say, ‘promoting’ this value by bringing more humans into existence to increase the overall ‘amount’ of dignity. I now wish to consider what Anscombe might say about the further idea that this respect is something owed to every human being. That is, the inherent value of human beings calls for respect, and respectful treatment, of a specifically directed sort.29 As Talbot Brewer says, ‘[I]‌t is true of human beings not only that they ought to be treated as bearers of irreplaceable value, but that they are due such treatment. If we did not so treat them, we would not merely do the wrong thing; we would wrong them.’30 We can put this point in terms of directed duties. A directed duty is a duty that one party owes to another party. In this way, directed duties manifest a distinctive type of deontic normativity, which we might call relational or bipolar normativity.31 This type of normativity is expressed in judgements like ‘X has a duty to Y to Φ’ or ‘X wronged Y by Φ-​ing.’ The distinctive feature of such claims is that they join two parties as ends, or poles, of a single normative relation. The normativity here is thus intrinsically relational. In contrast, we can express a non-​relational or monadic type of deontic normativity in judgements like ‘X has a duty to Φ’ or ‘X did wrong by Φ-​ing’ or ‘It is permissible for X to Φ.’ I may have a duty to Φ without owing that duty to anyone.32 The content of such a monadic duty might refer to another person, insofar as Φ-​ing might consist in treating someone a certain way. But if the duty is merely monadic rather than directed, then

29  For helpful discussion, see Ariel Zylberman, ‘Human Dignity’, Philosophy Compass 11, no. 4 (2016), 201–​210. 30  Talbot Brewer, ‘Acknowledging Others’, in Oxford Studies in Normative Ethics, ed. Mark Timmons (Oxford: Oxford University Press, 2018), 9. In explaining the idea of irreplaceable value, Brewer points to the way that the loss of one human life cannot be compensated for by the creation or preservation of other human lives. Human beings are not fungible, like money. 31  For a helpful discussion of relational vs. monadic normativity, see Michael Thompson, ‘What Is It to Wrong Someone? A Puzzle about Justice’, in Reason and Value: Themes from the Moral Philosophy of Joseph Raz, ed. R. Jay Wallace, Philip Pettit, Samuel Scheffler, and Michael Smith (Oxford: Clarendon Press, 2004), 333–​384. For discussion of different attempts to explain a duty’s direction, and the challenges facing those explanations, see Simon May, ‘Directed Duties’, Philosophy Compass 10, no. 8 (2015), 523–​532. 32  Such are the non-​directed duties in Joel Feinberg’s imaginary Nowheresville, in his ‘The Nature and Value of Rights’, Journal of Value Inquiry 4, no. 4 (1970), 243–​260.

302   Micah Lott my failure to Φ would not amount to wronging that person, even if it would amount to my doing wrong. The person affected by my failure would be, in Michael Thompson’s phrase, ‘the occasion, not the victim, of my fall.’33 The idea we are now considering is that the requirement of respect for human dignity is not like that. Rather, as Ariel Zylberman says, ‘human dignity is thought to generate a directed duty of respect, such that disrespect is not simply the wrong thing to do, but is a wrong to the person.’34 Now, in paradigm cases of directed duties (and perhaps in all cases), a duty owed by one party corresponds to a right possessed by the party to whom the duty is owed. My duty to you not to strike you corresponds to your right against me not to be struck. Indeed, these appear to be two aspects of the same normative relation. To focus on my duty or your right is to view a single relation from either end. Thus if dignity generates a directed duty of respect, then it seems equally to generate a right to be respected, and rights to respectful treatment. And indeed, in legal and political contexts, human dignity and human rights are frequently connected.35 For instance, the Basic Law for the Federal Republic of Germany, first adopted in 1949, states the following in Article I: ‘(1) Human dignity shall be inviolable. To respect and protect it shall be the duty of all state authority. (2) The German people therefore acknowledge inviolable and inalienable human rights as the basis of every community, of peace and of justice in the world.’36 Philosophers, too, have argued that human dignity is the ground of human rights, or an essential part of that ground.37 Some have suggested that dignity itself should be explained in terms of rights—​i.e., that the essence of human dignity is to be the bearer of certain fundamental rights.38 The basic idea I wish to focus on is this: Human dignity grounds (or is constituted by) normative relations amongst human beings, in which every person has a duty of respect towards every other person, and every person has a right to respect from every other human being. In this way, human dignity is the ground 33 

Thompson, ‘What Is It to Wrong Someone?’, 340. Ariel Zylberman, ‘The Relational Structure of Human Dignity’, Australasian Journal of Philosophy 96, no. 4 (2018), 739. 35  For different philosophical accounts of dignity and human rights, see John Tasioulas, ‘Human Dignity and the Foundations of Human Rights’, in Understanding Human Dignity, ed. Christopher McCrudden (Oxford: Oxford University Press, 2013), 291–​312; Stephen Darwall, ‘Equal Dignity and Rights’, in Dignity: A History, ed. Remy Debes (Oxford: Oxford University Press, 2017), 181–​201. 36  In a similar fashion, c ­ hapter 2 of the Constitution of the Republic of South Africa states, ‘Everyone has inherent dignity and the right to have their dignity respected and protected’. 37  Tasioulas offers a nuanced account of how human dignity, understood as ‘an intrinsically valuable status that merits our respect’, can be combined with universal human interests to ground human rights norms. Like Anscombe, Tasioulas understands human dignity to be grounded in human nature and to inhere in every human being ‘from the moment of their coming into existence as an individual human being until their death’ (‘Human Dignity’, 305). For scepticism that Tasioulas’s account of dignity succeeds in explaining directedness, see Zylberman, ‘Relational Structure’. 38  Joel Feinberg, for instance, writes that ‘respect for persons (this is an intriguing idea) may simply be respect for their rights, so that there cannot be the one without the other; and what is called “human dignity” may simply be the recognizable capacity to assert claims. To respect a person, then, or to think of him as possessed of human dignity, simply is to think of him as a potential maker of claims’ (‘Nature and Value of Rights’, 252). 34 

The Knowledge of Human Dignity     303 (or embodiment) of an order of natural justice—​of relations of duty and right that hold amongst all human beings as such. To speak about dignity is to speak about the normative status that every human being has within this order, a status which must be explicated in terms of essentially relational deontic norms. What would Anscombe say about dignity understood in this way? How might her views either support or critique such a conception of natural justice? And what does this conception of human dignity imply about our knowledge of human dignity?

12.5.2  Human Rights, Social Practices, and Dignity At first glance, Anscombe might not seem interested in any connection between dignity and human rights. She never mentions human rights in some of her most extended discussions of human dignity, such as KRHL and DHB.39 But upon further inspection, some of her statements suggest that she accepts, or is at least open to, a natural justice view. Recall that when the just person (qua just) recognises what is to be done or avoided, this manifests her connatural knowledge of the dignity of human nature. Surely the just person’s capacity to recognise what to do is guided by a sensitivity to relational deontic norms, by her awareness of what is owed to people and of people’s rights. And that suggests that human dignity itself is the kind of thing that should be understood in terms of such relational normativity. In addition, Anscombe stresses that ‘someone who is murdered suffers a great wrong’ and that if someone is wronged, then ‘he has a right which is violated.’40 Whatever might be the effects of murder on others, ‘the wrong done in murder is done first and foremost to the victim.’41 And Anscombe rejects the view that murder is only a legal concept, without application outside of what given laws allow or forbid.42 Taken together, these remarks suggest a picture of natural justice in which all humans have a moral status that can be spelled out in terms of directed duties and corresponding rights—​at the very least, a duty not to murder and the right not to be murdered! But matters are more complicated. For one thing, Anscombe holds that the right which is violated in murder is not the basis of the wrongfulness of murder, but the other way around: ‘The prohibition on murder is indeed a great charter of right to all of us, but it is the prohibition that comes first and not the right.’43 Against this, we might ask: Why suppose that either the prohibition or the right is prior, normatively speaking? Rather, on the sort of view of natural justice sketched earlier, the prohibition and the right arrive 39 

Anscombe does mention rights in DHB, and she is clearly aware of the distinction between (a) rights as legally defined and (b) rights that exist apart from any particular legal recognition. But she does not use the term ‘human rights’. Nor does she offer an account of dignity as the ground of human rights or seek to explain the idea of dignity in terms of rights. 40  MME, 266. 41  CC, 187. 42  TKEA, 9. 43  MME, 266.

304   Micah Lott on the scene together. For the prohibition on murder has a directed quality—​it concerns what is owed to others—​and thus it is an essentially relational norm, with one party’s right already built in as one pole of the relation.44 However, even if Anscombe were to accept this point, there remains a key strand in her thinking that seems incompatible with a natural justice view. This is her conventionalism about human rights and their corresponding duties. This is laid out most clearly in her essay ‘On the Source of the Authority of the State.’ In that essay, Anscombe argues that the concept of rights arises from a particular type of linguistic practice that employs modal terms like ‘cannot’ and ‘must not’ to prevent action, even when the action is physically possible. Anscombe labels these ‘stopping modals.’ One use of stopping modals involves a special mention of a person—​e.g., ‘ “You can’t sit there, it’s John’s place,” “You can’t eat that, it’s for N,” “You can’t do that, it’s for N to do.” ’45 And this is the origin of the concept of a right: ‘We have here a very special use of the name of a person, or a very special way of relating something to a person, which explains (not is explained by) the general term “right.” . . . The general term “right” is constructed because, as it were, our language feels the need for it. As, for example, a general term “relation” was invented’.46 With regard to stopping-​modals, Anscombe argues that we cannot understand the relevant sense of ‘necessary’ or ‘impossible’ apart from the practices in which people treat certain considerations as reasons for doing and refraining from certain things.47 In that way, the necessity of rights and duties is internal to our practices, and all rights and duties are conventional: ‘The existence of such a thing as rights consists in the regular existence of certain proceedings, certain reactions, an integral part of which is the use of certain linguistic forms.’48 How, then, might we criticise or justify a given practice and its corresponding conception of rights? Is there a way to show that ‘what is counted as a right is no right, and something not counted as a right is after all a right?’49 This is a question about ‘the rights of man’—​i.e., human rights—​and Anscombe calls this ‘the grave question.’50 Her answer is that we can support or criticise a given right-​and-​duty-​defining practice by appeal to

44 

Anscombe gives two reasons for thinking that the wrongfulness of murder is the basis of the right not to be murdered: ‘(a) there is not a simple right to life, but rather a right not to be murdered, and (b) if there were a certain right to life upon which the wrongfulness of murder is based, it would be difficult to see why it should not be waivable’ (MME, 266). As far as I can tell, neither of these considerations tells against the type of natural justice view I have in mind. 45  SAS, 139. 46  SAS, 142. 47  As Anscombe says, ‘Here is (at least in embryo) the idea of “a right.” The reasons why it is at once so clear and so inexplicable are these: it is clear because we have learned to respond to these stopping cannots, to comply with them, to issue them ourselves, to infringe them. It is inexplicable because, look as we may, we cannot find an interpretation of this “cannot,” just as we couldn’t find any interpretation of the peculiar “necessity” (called “obligation”) generated by promising. The truth is: there is no interpretation to give, in any of these cases’ (SAS, 140). 48  SAS, 142. 49  SAS, 141. 50  SAS, 141.

The Knowledge of Human Dignity     305 what is required for the realisation of human good. In brief, if one can show that human good depends upon a particular practice, and if that practice requires the recognition of certain rights and duties, then that would provide rational support for embracing that practice, and hence for the recognition of those rights and duties. Likewise, one could critique a given practice by showing that it was inimical to human good. The key to this type of argument is that it combines two sorts of necessity. There is the necessity that is internal to a given practice, the necessity of the ‘must’ and ‘cannot’ which explains the concept of a right. And there is the Aristotelian necessity of what must exist if human beings are to realise their good and to avoid harm. This necessity is external to our practices. It is by bringing these two sorts of necessity together that Anscombe aims to accept conventionalism about rights and duties while avoiding a relativism in which there is no possible rational evaluation of what people take to be rights and duties. In a phrase: Aristotelian realism about human good combined with conventionalism (but not relativism) about human rights.51 Thus it appears that Anscombe rejects a natural justice view. For given her conventionalism about duties and rights, it looks like relational deontic norms do not go ‘all the way down’ to the relations amongst human beings as such. Rather all duties and rights exist only on the level of actual human practices, and thus when relational deontic norms bind persons, they bind them qua members of a given practice, not qua human being.52 A view like this faces several challenges. One challenge says the view gives the wrong kind of reason for respecting others and their rights.53 A second challenge says the view has objectionable moral implications. For imagine an encounter between two humans from distant cultures who do not share any right-​and-​duty-​defining social practices. Conventionalism seems to imply that upon meeting these humans have no duties towards one another, not even the duty not to murder one another. But that seems wrong. As Thompson says, ‘In seeing the practice-​outsider as making as good an example of a murderable as any practice-​insider, we plainly put ourselves into dikaiological connection with her and commit ourselves to the falsehood of any such account.’54 51  Call such an argument a crossing modals argument. It is a particular instance of the kind of functional arguments we considered in section 12.4. Like other functionalist arguments, a crossing modals argument appeals to (1) what is necessary because human good depends upon it and (2) what is necessary because it is required to have what is identified in (1). What is special about a crossing modals argument is that what is required for human good is a practice that is constituted by the participants’ recognition of another sort of necessity—​the ‘must’ and ‘cannot’ of stopping modals. For discussion of Anscombe’s account of stopping modals, and her crossing modals argument, see Roger Teichmann, The Philosophy of Elizabeth Anscombe (Oxford: Oxford University Press, 2008), 94–​102. 52  For such an interpretation of Anscombe’s position, see Katharina Nieswandt, ‘Anscombe on the Sources of Normativity’, Journal of Value Inquiry 51 (2017), 141–​163. 53  For versions of this challenge, see Darwall, ‘Equal Rights and Dignity’; Zylberman, ‘Relational Structure’. 54  Thompson, ‘What Is It to Wrong Someone?’, 376. Thompson presents this as a problem for Hume. But on the interpretation of Anscombe I am considering here, her view of fundamental duties and rights amounts to the Humean view as Thompson describes it. For further discussion of this issue, see Matthias Haase, ‘Life and Recognition: Michael Thompson’s Practical Naturalism’, in Aristotelian Naturalism: A Research Companion, ed. Martin Hähnel (Cham: Springer Nature Switzerland, 2020), 259–​261.

306   Micah Lott Perhaps, however, Anscombe is not committed to such complete conventionalism about rights and duties after all. Perhaps the position she spells out in SAS is actually consistent with a natural justice view. We might distinguish between (a) the social practices that are the necessary condition for acquiring the ideas of ‘right’ and ‘duty to’ and which embody different substantive conceptions of rights and duties, and (b) a genuine normative order governing all humans as such, expressible in terms of relational deontic norms, and with which particular right-​and-​duty-​defining practices put their bearers in touch—​when all goes well. And drawing on a framework of Aristotelian naturalism, we might say that, while it often does not go well, it is no accident when it does so. For it belongs to ‘the human’ to develop a conception of the order of natural justice that governs human beings. It is thus characteristic of human beings (in the sense of being naturally good in them) when their social practices embody a correct, rather than distorted, conception of fundamental human rights and duties. And this remains true even if correct conceptions are comparatively rare, statistically speaking. Such a view faces many challenges, which I will not explore here. But it represents one possible way to develop an Anscombean conception of dignity in conjunction with the attractive thought—​not much explored by Anscombe herself—​that the value of human beings calls for respect, and respectful treatment, of a specifically directed sort.

Bibliography Anscombe, Elizabeth. ‘Analytic Philosophy and the Spirituality of Man’. In Human Life, Action and Ethics, edited by Mary Geach and Luke Gormally (Exeter: Imprint Academic, 2005), 4–​16. Anscombe, Elizabeth. ‘Contraception and Chastity’. In Faith in a Hard Ground, edited by Mary Geach and Luke Gormally (Exeter: Imprint Academic, 2008), 170–​191. Anscombe, Elizabeth. ‘The Dignity of the Human Being’. In Human Life, Action and Ethics, edited by Mary Geach and Luke Gormally (Exeter: Imprint Academic, 2005), 67–​73. Anscombe, Elizabeth. ‘The Immortality of the Soul’. In Faith in a Hard Ground, edited by Mary Geach and Luke Gormally (Exeter: Imprint Academic, 2008), 69–​83. Anscombe, Elizabeth. ‘Knowledge and Reverence for Human Life’. In Human Life, Action and Ethics, edited by Mary Geach and Luke Gormally (Exeter: Imprint Academic, 2005), 59–​66. Anscombe, Elizabeth. ‘The Moral Environment of the Child’. In Faith in a Hard Ground, edited by Mary Geach and Luke Gormally (Exeter: Imprint Academic, 2008), 224–​233. Anscombe, Elizabeth. ‘Murder and the Morality of Euthanasia’. In Human Life, Action and Ethics, edited by Mary Geach and Luke Gormally (Exeter: Imprint Academic, 2005), 261–​277. Anscombe, Elizabeth. ‘On the Source of the Authority of the State’. In Ethics, Religion, and Politics: Collected Philosophical Papers, vol. 3 (Oxford: Blackwell, 1981), 130–​155. Anscombe, Elizabeth. ‘Prolegomenon to the Pursuit of a Definition of Murder’. In Human Life, Action, and Ethics, edited by Mary Geach and Luke Gormally (Exeter: Imprint Academic, 2005), 253–​260. Anscombe, Elizabeth. ‘Two Kinds of Error in Action’. In Ethics, Religion, and Politics: Collected Philosophical Papers, vol. 3 (Oxford: Blackwell, 1981), 3–​9.

The Knowledge of Human Dignity     307 Brewer, Talbot. ‘Acknowledging Others’. In Oxford Studies in Normative Ethics, edited by Mark Timmons (Oxford: Oxford University Press, 2018), 9–​31. Darwall, Stephen. ‘Equal Dignity and Rights’. In Dignity: A History, edited by Remy Debes (Oxford: Oxford University Press, 2017), 181–​201. Diamond, Cora. The Realistic Spirit (Cambridge, MA: MIT Press, 1991). Feinberg, Joel. ‘The Nature and Value of Rights’. Journal of Value Inquiry 4, no. 4 (1970), 243–​260. Foot, Philippa. Natural Goodness (Oxford: Oxford University Press, 2001). Haase, Mathias. ‘Life and Recognition: Michael Thompson’s Practical Naturalism’. In Aristotelian Naturalism: A Research Companion, edited by Martin Hähnel (Cham: Springer Nature Switzerland, 2020), 247–​263. Hursthouse, Rosalind. ‘Practical Wisdom: A Mundane Account’. Proceedings of the Aristotelian Society 106 (2006), 285–​309. Lott, Micah. ‘Justice, Function, and Human Form’. In Normativität des Lebens—​Normativität der Vernunft?, edited by Martin Hähnel and Markus Rothhaar (Berlin: De Gruyter, 2015), 75–​92. May, Simon. ‘Directed Duties’. Philosophy Compass 10, no. 8 (2015), 523–​532. Müller, Anselm. ‘The Spiritual Nature of Man’. In The Moral Philosophy of Elizabeth Anscombe, edited by Luke Gormally, David Albert Jones, and Robert Teichmann (Exeter: Imprint Academic, 2016), 10–​32. Müller, Anselm. ‘ “Why Should I?” Can Foot Convince the Sceptic?’ In Philippa Foot on Goodness and Virtue, edited by John Hacker-​ Wright (Cham, Switzerland: Palgrave Macmillan, 2018), 151–​186. Nieswandt, Katharina. ‘Anscombe on the Sources of Normativity’. Journal of Value Inquiry 15 (2017), 141–​163. Tasioulas, John. ‘Human Dignity and the Foundations of Human Rights’. In Understanding Human Dignity, edited by Christopher McCrudden (Oxford: Oxford University Press, 2013), 291–​312. Teichmann, Roger. The Philosophy of Elizabeth Anscombe (Oxford: Oxford University Press, 2008). Thompson, Michael. ‘What Is It to Wrong Someone? A Puzzle about Justice.’ In Reasons and Value: Themes from the Moral Philosophy of Joseph Raz, edited by R. Jay Wallace, Philip Pettit, Samuel Scheffler, and Michael Smith (Oxford: Oxford University Press, 2004), 333–​384. Vogler, Candace. ‘Why Human Beings Matter: Anscombe on the Nature and Point of Human Life.’ In G. E. M. Anscombe and Human Dignity, edited by John Mizzoni (Aston, PA: Neumann University Press, 2016). Zylberman, Ariel. ‘Human Dignity.’ Philosophy Compass 11, no. 4 (2016), 201–​210. Zylberman, Ariel. ‘The Relational Structure of Human Dignity.’ Australasian Journal of Philosophy 96, no. 4 (2018), 738–​752.

Chapter 13

Life and Ot h e r Basi c Rights in A ns c ombe Katharina Nieswandt

13.1  Rights As Moves within Social Practices On my reading, Elizabeth Anscombe holds the following view on rights:1 Claim 1. A right is a bundle of possible and impossible moves within a social practice. Some of these are moves by the right-​holder, some are moves by others. Claim 2. A social practice is a widely played game that serves a telos. To some extent, this telos fixes the rules of the game. For many such games, participation is not optional.

Let me flag that both claims constitute substantial interpretations. Anscombe nowhere explicitly offers an account of rights; these are a recurring topic in several of her papers.2 These claims are my attempt to combine her scattered discussions into a coherent and plausible account. This metaethical account can be of interest even to readers who do not share Anscombe’s views within normative ethics.

13.2  Practice-​Internal Modality: Rules It is easiest to explain the two claims by means of an example. Imagine you had a visitor from a culture where private property is unknown, and you had to explain to her what it means to 1  For a detailed discussion, please see Katharina Nieswandt, ‘Anscombe on the Sources of Normativity’, Journal of Value Inquiry 51, no. 1 (2016): 141–​163. 2  Most important, RRP, SAS, and PJ.

Life and Other Basic Rights in Anscombe    309 say that the apples on the tree in the next garden belong to your neighbour Adam. Your visitor desires to pick these apples; it is easy to climb over the fence between your garden and Adam’s, but nevertheless the visitor cannot pick these apples, and you have to explain to her the meaning of this ‘cannot.’ This leads us to Claim 1 because you might say ‘We cannot pick these apples (prohibition) without asking Adam (permission).’ Or ‘If we eat them, Adam could demand that we compensate him (command).’ Or ‘Adam is the only one allowed to pick these (privilege).’ What you would be doing here is to give an extensional definition of the concept ‘property right’ (or of a certain kind of property right). You would provide a list of statements about possible actions on your/​your visitor’s part and on Adam’s part. Anscombe calls these statements ‘stopping modals’ if they express prohibitions and ‘forcing modals’ if they express commands.3 Stopping/​forcing modals are usually formulated using modal auxiliary verbs (‘must’, ‘can’t’, etc.), but we can also use the adjectives ‘possible’ and ‘necessary’ (e.g., ‘It is not possible for you to pick these apples’ or ‘It is necessary for you not to pick these apples’). Following Anscombe, this extensional definition is the only definition one can give of a specific right without already invoking the general concept ‘right.’4 What it means for Adam to own these apples just is that others cannot take these apples and that others who do take these apples need to compensate Adam and that he can pick these apples and etc. This leads us to Claim 2 because the type of modality invoked here needs explanation. In what sense is it necessary for the visitor not to pick the apples; in what sense is it impossible for her to pick them? As Anscombe points out, ‘in such a case [where] you are told you ‘can’t’ do something you plainly can, as comes out in the fact that you sometimes do.’5 If you believe that the prohibitions, commands, privileges, etc. around Adam’s apples are justified, then you regard this case as one of moral modality. If your modal statements refer only to the social fact that the respective behaviours are officially forbidden, required, etc. in your society, then you regard it as legal modality. Both moral and legal modality fall into a category that we could name ‘normative modalities.’ Anscombe provides the following illustration: ‘If I say ‘You can’t wear that!’ and it’s not, for example, that you are too fat to get it on, that’s what I call a stopping modal.’6 It is physically possible for you to put on these clothes. What makes it impossible is that these clothes violate some norm. It might be a norm of beauty (aesthetic modality). It might also be respect: while you do look fantastic in these clothes, they would be inappropriate for the funeral to which you are going. Other examples of normative modalities are grammatical modality and mathematical modality. 3 

One could think that a third category were necessary to express privileges, ‘permitting modals’. As Roger Teichmann, The Philosophy of Elizabeth Anscombe (Oxford: Oxford University Press, 2008), 97 points out, however, privileges can be defined through negated stopping modals, such as ‘It is not the case that Adam mustn’t pick these apples’. 4  See her rendering of ‘Hume’s Circle’, RRP, 99–​101. 5  RRP, 101. 6  RRP, 101.

310   Katharina Nieswandt Normative modality blocks actions in that it makes them incorrect. Physical modality, on the other hand, blocks actions in that it makes them practically impossible. Your visitor cannot pick the apples on the top branches because these are out of reach, and she cannot pick more than five apples in ten seconds if picking one apple takes at least two seconds. It is at this point of the explanation that we need the concept ‘practice.’ For the action to be incorrect means for it to violate some norm. What provides the norm? What endows it with force? Immanuel Kant, e.g., would say that it is a norm of practical reasoning: your visitor cannot want to pick the apple without violating the norm of consistency for her actions. Anscombe offers a different solution. For her, such a norm is some rule of a widely played game—​in this case the property game. Your visitor is physically able to pick the lower-​hanging apples, but that would be theft; i.e., it would violate the norms either of morality or the local law or both. In sum, for Anscombe to say that someone has a right is to say that certain actions are (in)correct within a corresponding practice.

13.3 Practice-​Grounding Modality: Goods I defined a social practice as a widely played game that serves a telos. (Note that this definition is much narrower than many current uses.) We already discussed the example of private property; some other examples are contractual affairs (such as promising and marriage), elections, and money. Here is not the place for a comprehensive discussion of what a social practice is. I intend to highlight only three features of practices that are relevant to our topic. The telos of the practice often makes it necessary: 1. to have either this or a similar practice. 2. for the practice to (not) have certain rules. 3. that every member of the relevant group participate in the practice. It is necessary to have some practice of private property (if it is), e.g., because without it ‘some good will not be attained or some evil avoided.’7 Philippa Foot defends a similar view when she calls the practice of promising an ‘Aristotelian necessity.’8 A prominent historical defender of this kind of view on property is David Hume, who regards this practice (which he calls a ‘convention’) as a precondition for human beings to live together in larger, more anonymous groups, in other words, to leave the state of nature: 7 

8 

RRP, 100, citing Aristotle, Meta 1015a22–​23. Philippa Foot, Natural Goodness (Oxford: Oxford University Press, 2001), 15–​17 and 43–​46.

Life and Other Basic Rights in Anscombe    311 [T]‌he principal disturbance in society arises from those goods, which we call external, and from their looseness and easy transition from one person to another; [in order to be able to live together, human beings] . . . must seek for a remedy. This can be done after no other manner, than by a convention enter’d into by all the members of the society to bestow stability on the possession of those external goods, and leave every one in the peaceable enjoyment of what he may acquire by his fortune and industry. By this means, every one knows what he may safely possess; and the passions are restrain’d in their partial and contradictory motions. . . . [I]t is by that means [that] we maintain society.9

Various property practices are conceivable that could achieve this aim. This leads us to the second feature of practices, viz. that the practice’s telos often makes it necessary for the practice to have certain rules but not others. If, e.g., Hume should be correct in that the telos of the practice of property is to create ‘stability of possessions’ and to ‘restrain the passions’, and that this can be achieved only if everyone is left ‘in the peaceable enjoyment of what he may acquire by his fortune and industry’, then any actual property practice must be evaluated on the basis of whether it fulfils these criteria. To give an example: If R, a rule for allocation in property practice P, says that everyone gets to keep the proceeds of their personal labour, then P could be expected to achieve its telos. If R says that the proceeds are largely allocated to the owner of the means of production or to whichever citizen wore the most colourful shirt the previous Tuesday, then P cannot achieve its telos. This ties in with the third feature listed earlier, viz. that practices often require all group members to participate. This is certainly true of the practice of private property. If individual group members could exclude themselves, then the participants would not be secure in their belongings. I began my explanation of the three features of practices with the assumption that it is necessary for human beings to have a practice of property. To what kind of modality am I referring here? It cannot be a practice-​internal necessity. That is, it would not be against the rules of some further practice not to have a practice of property. Therefore, not having this practice is not ‘incorrect’; it does not violate any rules. According to Anscombe and Foot, the modality is that of realising certain goods. If, e.g., it is good to live in larger social groups and, ultimately, to form a state, and if Hume is right about the enabling role of property in this, then it is necessary to have the practice of property in order to realise the good of political community. Following Geach,10 Anscombe understands ‘good’ to mean either ‘good as’ or ‘good for.’ That is, it either is good as an individual of a certain kind or good for something else. To what does ‘good’ refer here? Anscombe’s answer (as well as that of Geach, Foot, Hursthouse, and other neo-​Aristotelians) is: it refers to the good life for human 9  David Hume, A Treatise of Human Nature: Being an Attempt to Introduce the Experimental Method of Reasoning into Moral Subjects, ed. David Norton and Mary Norton (Oxford: Oxford University Press, 2000), Sec. 3.2.2, §9. 10  Peter Geach, ‘Good and Evil’, Analysis 17, no. 2 (1956): 33–​42.

312   Katharina Nieswandt beings (i.e., to human ‘well-​being’ or ‘flourishing’). Applied to our example of property, philosophers have claimed both: it is good as a human being to live in a political community. We are political animals, so political community is a constitutive aspect of our species-​specific flourishing (as Plato and Aristotle claim). And it usually is good for a human being to do so due to the instrumental advantages of group life (which Hume has in mind when he advocates property). In sum, to say that a practice is necessary means that it is the only way to achieve an important human good. Human nature determines what things constitute important goods for human beings, and thereby it sets the norm for adopting (or rejecting) practices.

13.4 The Justification of Rights versus Practices What is the upshot of these general considerations about moves and practices for our topic of rights? Claim 1, that a right is a bundle of moves that can be made or blocked within a practice, has important implications for the justification of rights and hence for what rights exist.11 Anscombe applies Wittgenstein’s metaphysical perspective, according to which we must think of normative demands in terms of games and rules, to the topic of rights. She therefore also draws a ‘distinction between justifying a practice and justifying a particular action falling under it.’12To use a stopping modal such as ‘You can’t take these apples!’ is to make a move within the property game, according to Anscombe. The justification for this move will consist in an explication of the rules; the justification of the property game as a whole is of no consideration at this point. It would at best be superfluous, but often false to mention it. Let me illustrate this, using our earlier example.13 If the visitor demanded a justification of your stopping modal ‘You can’t take these apples’, then it would be correct if you replied, ‘Because these apples are Adam’s property’. To do this would mean to supply the ‘logos’ of the stopping modal, as Anscombe calls it. You mention a fact that sounds like an independent reason: because of the fact that Adam owns these apples, a second fact obtains, viz. that you cannot take them. A closer look reveals, however, that what you do here simply is to provide the umbrella category for the second fact. The fact that the visitor (just like you and like all people 11 

For rights, justification and existence are tied together. Katharina Nieswandt, ‘Do Rights Exist by Convention or by Nature?’, Topoi 35, no. 1 (2016): 313–​325. If it is not justified, e.g., that governments pre-​ emptively monitor all citizens, then governments have no right to do so; i.e., then a right of governments to pre-​emptively monitor all citizens does not exist. 12  John Rawls, ‘Two Concepts of Rules’, Philosophical Review 64, no. 1 (1955): 3. 13  For a detailed discussion, please see Katharina Nieswandt, ‘What Is Conventionalism about Moral Rights and Duties?’, Australasian Journal of Philosophy 97, no. 1 (2018): 15–​28.

Life and Other Basic Rights in Anscombe    313 except Adam) cannot take these apples is simply one element in the bundle of facts that together constitute Adam’s ownership. You essentially say, ‘You cannot take these apples because you cannot take these applesproperty right.’ Or ‘You cannot take these apples because the rules of a certain practice forbid it and, f.y.i., that practice is private property.’14 As Anscombe points out, this justification is vacuous or circular. (She takes it to spell out ‘Hume’s Circle.’) For what you essentially say is ‘It is impossible for you to take the apples, and the sense in which it is impossible is that it is incorrect, and what ‘incorrect’ means is that it violates the rules of some game.’ In other words, you cannot take these apples because the rules forbid it. Nevertheless, Anscombe argues, it is the correct and complete answer. It would be a category mistake to continue at this point with a justification of the game. For example, it would be false to say ‘You cannot take these apples because the rules of the practice of private property forbid it and because this practice is necessary in order to enable us to live in larger groups and hence you should support this practice’, just as you should support Oxfam. This is most obvious for leisure games: it might be that the justification for playing poker is to win money, but the justification for why you now put out the small bid is that you sit left of the dealer—​you would have to make this move even if it lost you money. You cannot reject this duty under which the rules put you by saying ‘But look, one plays this game to win money, and this move will lose me money’.15 The justification of the practice is (usually) of no interest where a justification for an individual move within the practice is demanded. On my interpretation, Anscombe takes a social-​constructivist position on rights and a naturalist position on practices. Human nature necessitates (if it does) that groups of human beings have some practice of property, e.g., and it sets the boundaries for which rules are acceptable. Once this practice is in place, individual human beings can have property rights to individual objects, which simply means that certain moves within the practice are open to them but are blocked for others (privilege) or that certain corresponding moves are required (command) or forbidden (prohibition) for others.

13.5 Two Problems for Social Constructivism For property rights, this hybrid of social constructivism and Aristotelian naturalism is not implausible. Many readers who do not share Anscombe’s philosophical leanings would still agree that property is a human invention, that it serves important social purposes, that different property practices are appropriate for different societies, and that 14 

See Anscombe on ‘logoi’, esp. RRP, 101–​102. This example is modelled on one by Anselm Müller (in Was taugt die Tugend? Elemente einer Ethik des guten Lebens (Stuttgart: Kohlhamme—​Verlag, 1998), 91). 15 

314   Katharina Nieswandt human nature sets some boundaries here. (If you personally believe that property rights are natural rights, use another example, such as marriage or elections.) For certain basic rights, however, Anscombe’s hybrid appears less appealing. What social practice is necessary, e.g., in order for me to have a right not to be murdered? In fact, it is common in contemporary human rights literature to define human rights as those rights that do not presuppose any social practices, i.e., as natural rights.16 Two problems arise for social constructivism about basic rights: Metaphysical Problem. There are rights for which there seems to be no corresponding practice. Epistemic Problem. That these rights exist seems more certain than most things we could mention about a corresponding practice and its justification. An easy way to avoid these problems would be to sort the relevant moral demands into another category than rights. Perhaps it is incorrect to understand murder as a rights violation? Anscombe, however, explicitly says that there is a right not to be murdered: ‘[S]‌omeone who is murdered suffers a great wrong. . . . If someone is wronged, he has a right which is violated’.17 At the same time, she rejects the very idea of natural rights: ‘A right is not a natural phenomenon. . . . It is in this respect like a rule and a promise: that ‘natural unintelligibility’ which Hume attributed to promises is found in all three things.’18 On my reading, Anscombe avoids the two problems because the good protected by the right to life (and perhaps by other rights) plays a transcendental role in her metaethics. Before we proceed, it is important to flag that Anscombe defines the right to life very narrowly: ‘[T]‌here is not a simple right to life, but rather a right not to be murdered.’19 This definition excludes a right to be brought into existence (i.e., a deontic analogue of Parfit’s ‘repugnant conclusion’),20 and it excludes requirements such as that of abolishing all cars, given that cars potentially kill people.

13.6 Basic Rights For Anscombe, basic rights are not metaphysically different from other rights. My summary of this will be brief because I have already developed these ideas elsewhere.21 Take

16 

In an overview article, John Tasioulas calls this the ‘orthodox conception’ of human rights: ‘Are Human Rights Essentially Triggers for Intervention?’, Philosophy Compass 4, no. 6 (2009): 938. 17  MME, 266. 18  SAS, 138, referring to Hume, Treatise, Sec. 3.2.1 and 3.2.5. 19  MME, 266. 20  Derek Parfit, Reasons and Persons (Oxford: Clarendon Press, 1984), ch. 17. 21  Nieswandt, ‘Do Rights Exist’.

Life and Other Basic Rights in Anscombe    315 the following example by Anscombe: the right of parents to tell their children what to do (on certain matters and within certain boundaries) and to not be interfered with in this domain. ‘[T]‌hose who have and carry out the task of bringing up children quite generally perform a necessary task. It cannot be done without the children’s obedience. So those people have a right to such obedience,’ Anscombe argues, and outsiders may not prevent them from imposing their decisions.22 In examples like these, we have two kinds of necessity, a ‘crossing of modals’:23 1. It is necessary that those who have the role of bringing up children tell these children what to do (Aristotelian necessity). 2. ‘If someone has a role or function which he ‘must’ perform, or anything that he ‘has’ to do, then you ‘cannot’ impede him.’24 3. Therefore, it is necessary for you not to impede those who have the role of bringing up children when they tell these children what to do (obligation and corresponding right). There is some good to be obtained or some evil to be averted here (1), and in conjunction with a version of the principle that ought-​implies-​can (2), we deduce the necessity to grant a certain right to some people (3). Parents, kindergarten teachers, etc. perform this necessary task in our society; hence they must be granted the privileges necessary to carry out the task. One can always raise worries and further questions regarding any argument of this form, such as: Is the task really necessary? Are those tasked with it the best to carry it out? ‘But in form it is sound enough, if I am right about the relation between ‘a right’ and certain modals,’ Anscombe concludes.25 This yields the following metaphysical picture: A good to be achieved, such as the upbringing of children, can give rise to rights. This happens in the following way: Human societies can adopt a practice that enables people to realise this good. The shape of this practice can vary significantly for different times and places, but any human society that is to realise the relevant good needs some such practice. For the example of upbringing, contemporary Western societies by default give the necessary privileges to the biological parents and, to a lesser extent, educational professionals. The boundaries of these privileges vary even amongst Western societies—​some allow home-​schooling, e.g.; others demand that this part of a child’s upbringing be carried out by professionals. Other societies had and have very different arrangements. Different circumstances (such as the form of the economy, the state of technology, the size of the population, etc.) can make different practices appropriate here; i.e., there is considerable room for 22 

SAS, 145. SAS, 144. 24  SAS, 144. 25  SAS, 145. 23 

316   Katharina Nieswandt legitimate diversity. Still, we can objectively judge certain practices to be outside the realm of what is legitimate or to be better than others because the telos to be achieved provides an external norm for their evaluation. We can also criticise societies for making it impossible for their members to realise a certain good if these societies fail to adopt any practice at all to further this good. Thus, one could think that material security in old age is an important good and criticise societies that do not make any provisions for the elderly, such as a pension system. Anscombe’s view of basic rights can hence be summarised as follows: Certain things are goods for human beings; some of these are basic goods. Goods indirectly give rise to rights in that they make the adoption of certain practices necessary (and preclude that of others), and within these practices, we assign rights. Some of these rights can be summarised under the umbrella term ‘basic rights’ or ‘human rights’ due to their close connection with goods that are indispensable for human flourishing. But their metaphysical structure is not different from less important and very obviously conventional rights, such as my right to seat B-​35 in the theatre, for which I purchased the ticket.

13.7 The Transcendental Role of the Good Life Epistemically, however, the right to life differs from trivial rights—​and the same might be true of other basic rights. While ‘[t]‌he prohibition on murder is indeed a great charter of right to all of us, it is the prohibition that comes first and not the right,’ Anscombe claims.26 Within Anscombe’s view of practical reasoning, an absolute prohibition must be understood as the exclusion of a sound practical inference. There is no valid pattern of inference and set of true premises available that would conclude in, e.g., the murder of an innocent person.27 Why think that an act of murder can never be the conclusion of a sound practical inference? Anscombe endorses ‘the great Aristotelian parallel,’28 according to which the aim of practical reasoning is the good, just as the aim of theoretical reasoning is the truth. ‘[E]‌very action and rational choice is thought to aim at some good,’ ultimately that of ‘living well and acting well.’29 Like Aristotle, Anscombe understands this good as the good life (or ‘well-​being’ or ‘flourishing’) of human individuals.30 To forbid theft, 26 

MME, 266. This formulation was suggested to me by Ulf Hlobil. 28  PI, 157. 29 Aristotle, The Nichomachean Ethics, trans. and ed. Roger Crisp (Cambridge: Cambridge University Press, 2000), 1094a1–​2 and 1095a21–​22, respectively. 30  Very importantly, the claim that the good life is the ‘aim’ of practical reasoning does not mean that every single action must ultimately be justified by appeal to the good life. It would be false, e.g., to think 27 

Life and Other Basic Rights in Anscombe    317 e.g., perhaps even to define it, we need to tell a story like Hume’s, Anscombe believes, which explains what good property rights serve and which thus explains how the practice of property makes human life better. Consider ‘the character of rational argument to shew that it is wrong to steal, or commit adultery, or that we ought to keep a rule of the road where there is traffic, or a close season for game or fish where stocks must be replenished. The arguments are of the form ‘Obedience to this law is needed for human good.’ The unit whose good the argument seeks is the human individual, considered generally.’31 To murder someone is the ultimate rejection of this good; it ‘is to destroy that being which is the point of those [i.e., moral] considerations.’32 A right not to be murdered can be assigned within various social practices—​any justified system of law, e.g., would contain it. But it is our knowledge that there can be no sound practical inference concluding in murder which serves to justify the adoption of such practices and hence of the relevant right, not vice versa. Anscombe, I submit, makes a transcendental argument here: All practical reasoning is directed at the good life; therefore, it is in principle excluded that there could be practically rational (and hence justified) actions aimed at destroying this life. A necessary precondition for the possibility of there being such a thing as practical rationality is that such rationality aim at the good of the beings in question (us); so it can’t conform to practical rationality to destroy such beings.33 This is why ‘[t]‌he prohibition [on murder] is so basic that it is difficult to answer the question as to why murder is intrinsically wrongful.’34 The transcendental argument for the right to life that I reconstructed in the previous paragraphs is modelled on a transcendental argument that Anscombe explicitly makes—​for the boundaries of the right of the state to exercise violence. In her paper ‘On the Source of the Authority of the State,’ Anscombe argues that, as for the authority of parents, the source of the state’s authority is a necessary task. In the case of the state, there are two such tasks: protection from violence35 and, associated with that, the ‘administration of justice.’36 A major claim in her paper is that it is surprisingly difficult to define (and in practice to distinguish) between a legitimate state and ‘a place that was rather smoothly run by the Mafia.’37 There is, however, ‘one consideration here which has something like the position of absolute zero or the velocity of light in current physics. It cannot possibly be an exercise of civic authority deliberately to kill or mutilate innocent subjects. . . . [Because] authority in the command of violence (which was what we first that the reason why X must keep her promise to Y is the amount of flourishing to be secured for Y or for their shared community. Practice-​internal actions, such as the keeping or breaking of a promise, are prime examples of actions that must not be justified by appeal to human flourishing. (See Section 13.4 above.) 31 

MME, 266–​267, emphasis added. MME, 267. 33  This formulation was suggested to me by Roger Teichmann. 34  MME, 266. 35  SAS, 135. 36  SAS, 154. 37  SAS, 133. 32 

318   Katharina Nieswandt saw as distinguishing government from a Mafia in control of a place) is based on its performance of a task which is a general human need. A way of treating someone which puts him outside the class of those for whom the task is performed puts him outside the class of those subject to the authority.’38 That is, in the tradition of Augustine and Aquinas,39 Anscombe argues that the privileges of the state—​to exercise violence, to demand money of its citizens, etc.—​are justified by its task, and, appealing to Hobbes,40 that the main task here is protection. This justification entails that the state loses its privileges if it refuses to fulfil the tasks that ground them. (The same would apply to parents or teachers who do not act for the good of the children they are bringing up.) Anscombe calls this ‘consideration zero.’41 A state that murders innocent citizens thereby annihilates its own authority. Regarding the right to life, I submit, Anscombe makes a similar move. A system of ethics that allows murder of the innocent annihilates that which has to serve as the grounds of all practical reasoning, viz. the good life of the human individual. There might be more actions for which this holds—​consider torture or rape. Murder, however, is the most fundamental and obvious example of an action that can never be part of a good life (or of the exercise of state authority). If we give up the aim of individual human flourishing, Anscombe seems to think, then we give up the framework within which our arguments take place; in other words, we give up our basis for evaluating actions.

13.8 Alternative Conceptions of the Good Not everyone would accept Anscombe’s claim that it can never be rational to murder an innocent person. There are at least four alternative conceptions of the good at which practical reasoning is directed. First, one could think that this good, Primitive Hedonism, consists in maximal satisfaction of the agent’s egoistic desires. On this conception, an agent needs to consider only her own individual good life, not that of other people, and this good life is defined in a way that is compatible with being a murderer: satisfaction of my desires might require that I murder my rich aunt, so murder could easily be rational.

38 

SAS, 155, emphasis added. For a modern rendering, see Jacques Maritain, The Rights of Man and Natural Law (London: Centenary Press, 1944). 40  SAS, 137, citing Leviathan, ch. 17, §9. 41  SAS, 145. 39 

Life and Other Basic Rights in Anscombe    319 Others think that the aim of practical reasoning is not the good life of each individual at all but either Consequentialism, the best possible state of the world, which is determined as an aggregate of the goodness of the lives of individuals. or Totalitarianism, the good of a political community, which is metaphysically distinct from the aggregate of the goodness of the lives of its members. or Nietzscheanism, the good life of some superior individuals, whereas the lives of all other individuals are of no consideration. Let me elaborate on the example of consequentialism. Anscombe famously accused consequentialists of showing ‘a corrupt mind,’42 given that (certain) consequentialists require the murder of the innocent if this saves a greater number. I believe that she intended this accusation in the literal sense of rumpere, ‘(to) destroy,’ ‘(to) rupture.’ The consequentialist tries to prove certain actions as good through a move that gives up the very aim of human action and hence the very standard by which their goodness is measured. Consequentialists, however, would likely reply that they simply hold a different standard of the goodness of actions. Their criterion is the aggregate of all individual lives, and that might require the sacrifice of some individuals.

13.9 Anscombe’s Concept of Dignity Anscombe does not think that the consequentialist aggregation is possible. With Geach, she holds that ‘best (state of the) world’ is not a well-​defined term.43 More important, however, she seems to believe that such aggregation is incompatible with the idea of human dignity. ‘[T]‌o kill an innocent person because it seems a good idea that he should die’44 means ‘not respecting in [t]his victim the dignity of a human being at all.’45 Her view here has some overlap with Kant’s demand to always (also) treat others as ends in themselves and with the standard accusation against consequentialism that it ignores 42 

MMP, 40. Geach, ‘Good and Evil’, 41–​42. 44  MME, 266. 45  DHB, 68. 43 

320   Katharina Nieswandt the separateness of persons. The very concept of dignity or of a person, Anscombe seems to think, implies that we regard that being’s flourishing as the source of legitimate demands on us, some of which are absolute. Those who believe in human dignity would hence have to reject any of the four views just listed. In this last section, I want to show that Anscombe’s definition of ‘dignity’ is plausible and that, consequently, her transcendental view of the right to life (which might extend to other basic rights) is plausible. Admittedly, adherents of the four views will not be moved by a successful demonstration that their view is incompatible with human dignity. However, a lot of public discourse today shares Anscombe’s commitment that grave violations of dignity are a non-​negotiable reason to rule out actions. Therefore, the justification that Anscombe provides for this commitment can still be of interest to many readers. Her view is quite similar to a widespread approach in the contemporary human rights literature.46 Just as Aristotle holds that the good life grounds all ‘action and rational choice,’ many contemporary political theorists hold that dignity grounds all human rights catalogues, such as the UN’s Universal Declaration of Human Rights (henceforth UDHR).47 There is no universally accepted definition of ‘dignity,’ but on most conceptions, including the UN’s, respect for dignity entails a strong demand to consider the well-​being (or ‘good life’) of each human individual for their own sake, independent of this person’s character, their social contributions, social status, wealth, or similar considerations. To give an example, one of the rights prescribed in the UDHR is that to education: ‘Education shall be directed to the full development of the human personality’; it ‘shall be free’ and ‘be equally accessible to all on the basis of merit.’48 This right makes sense if our target is a good life (or at least an equal chance at a good life) for each and every single individual. If our target were maximal economic growth, then it could make more sense to adopt a caste system instead, where some carry out menial tasks and receive no education because all educational resources are shifted to the brightest in each generation. Adopting an educational caste system, however, is out of the question for any serious supporter of the UDHR. Why? Because it would seriously neglect the lives of those who were sorted into the lower caste, and concern for each and every 46 

E.g., James Griffin’s On Human Rights (Oxford: Oxford University Press, 2008). Jeremy Waldron’s ‘Is Dignity the Foundation of Human Rights?’, in Philosophical Foundations of Human Rights, ed. Rowan Cruft, Matthew Liao, and Renzo Massimo (Oxford: Oxford University Press, 2015) provides an overview. 47  UN General Assembly, Universal Declaration of Human Rights, 217 (III) A. (Paris, 1948). Anscombe would probably have worried about the fact that the addressee of ‘shall’ here and in the formulation of all other demands which the declaration puts forth is left vague. (The Declaration’s preamble characterises the list to follow as ‘a common standard of achievement for all peoples and all nations, to the end that every individual and every organ of society, keeping this Declaration constantly in mind, shall strive by teaching and education to promote respect for these rights and freedoms and by progressive measures, national and international, to secure their universal and effective recognition and observance, both among the peoples of Member States themselves and among the peoples of territories under their jurisdiction’.) 48  UDHR, Art. 26.

Life and Other Basic Rights in Anscombe    321 individual’s good life was part of what it meant to respect their dignity and hence was the justification for the whole Declaration in the first place. The good life of the individual is the reason why education ‘shall be directed to the full development of the human personality.’ We would need to substitute this Declaration for an entirely different justificatory framework—​a Universal Declaration of Economic Growth—​in order to make a case for the caste system. Similarly, Anscombe must think that we need an entirely different system of ethics than the ones within which most people currently argue in order to coherently claim that, for instance, murdering one innocent person to save multiple others is rational. We would need a framework that does not contain the demand to consider each individual’s good life for their own sake. This effectively means that we need a framework that does not draw on dignity, given that, on the common understanding, consideration of the individual’s good life forms an essential part of what it means to respect their dignity. Anscombe clarifies her conception of dignity only in broad outline. She seems to hold that membership in the human species bestows dignity. She explicitly opposes conceptions such as Peter Singer’s,49 which make rights conditional upon actualised abilities and hence deny basic rights to infants, to the severely impaired, etc.: It is a mere trick to draw on the weight that this word ‘person’ has . . . if you then go on to explain the word so that it is rather like the word ‘magnet.’ A piece of iron gets magnetised and so becomes a magnet; later it may get demagnetised and stops being a magnet though it is still the same piece of iron. If indeed you explain the word ‘person’ as meaning someone, e.g., who can talk (has self-​consciousness) and lead a social life (have inter-​personal relations) you may say that someone can be the same human being but no longer a person.50

As Singer’s proposal shows, it is possible to define ‘dignity’ without the element of unconditional consideration for each individual’s good life: it could be that only the good life of beings who display certain abilities needs consideration. Kantians offer still another alternative; they define ‘dignity’ through the concept of consistency: since I value valuing in my own person, Christine Korsgaard argues, I must value it in all other persons.51 Anscombe’s conception of dignity, however, as centrally comprising concern for a good life for each individual, is well-​aligned with common conceptions of dignity, such as the conception behind the UDHR. If you subscribe to this kind of conception, then you must accept her claim that murdering an innocent person can never be justified.

49 

Peter Singer, ‘All Animals Are Equal’, in Animal Rights and Human Obligations, ed. Tom Regan and Peter Singer (Englewood Cliffs, NJ: Prentice Hall, 1989), 148–​162. 50  MME, 268. 51  See particularly Christine Korsgaard, Self-​Constitution: Agency, Identity, and Integrity (Oxford: Oxford University Press, 2009).

322   Katharina Nieswandt

13.10 Summary and Evaluation Anscombe’s metaphysical position on rights is that rights consist in allowed and forbidden moves within a social practice. There can be no right outside of a practice, just as there can be no move outside of a game. The right to life—​understood as a right not to be murdered—​is not metaphysically different. But there is an epistemic difference: to question this right is to question the point of moral deliberations. The good life of the individual is that at which practical thought is ultimately directed, Anscombe holds, just as theoretical thought is directed at the truth. To give up this orientation hence means to give up practical reasoning. Anscombe’s construction might imply that the same holds for other basic rights, such as the right not to be tortured or the right not be raped. She did not, however, develop her ideas on basic rights and on human dignity in sufficient detail. Anscombe regularly draws on the concept ‘dignity’ when she discusses the right to life. Her view of basic rights has structural similarities with the currently dominant paradigm of human rights, where the dignity of the human individual is the grounding telos, too. To question whether everyone’s dignity must always be respected is to question this human rights framework (and perhaps the very idea of human rights). The definition of dignity implied in, e.g., the UDHR centrally includes that which Anscombe understands by respect for dignity: concern for the good life of each human individual.

Bibliography Anscombe, Elizabeth. ‘The Dignity of the Human Being.’ In Human Life, Action and Ethics, edited by Luke Gormally and Mary Geach, 67–​73. Exeter: Imprint Academic, 2005. Anscombe, Elizabeth. ‘Modern Moral Philosophy.’ In The Collected Philosophical Papers of G. E. M. Anscombe, volume 3, 26–​42. Oxford: Blackwell, 1981. Anscombe, Elizabeth. ‘Murder and the Morality of Euthanasia.’ In Human Life, Action and Ethics, edited by Luke Gormally and Mary Geach, 261–​277. Exeter: Imprint Academic, 2005. Anscombe, Elizabeth. ‘On Promising and Its Justice and Whether It Need Be Respected in Foro Interno.’ In The Collected Philosophical Papers of G. E. M. Anscombe, volume 3, 10–​21. Oxford: Blackwell, 1981. Anscombe, Elizabeth. ‘On the Source of the Authority of the State.’ In The Collected Philosophical Papers of G. E. M. Anscombe, volume 3, 130–​155. Oxford: Blackwell, 1981. Anscombe, Elizabeth. ‘Practical Inference.’ In Human Life, Action and Ethics, edited by Luke Gormally and Mary Geach, 109–​147. Exeter: Imprint Academic, 2005. Anscombe, Elizabeth. ‘Rules, Rights and Promises.’ In The Collected Philosophical Papers of G. E. M. Anscombe, volume 3, 97–​103. Oxford: Blackwell, 1981. Aristotle. The Nichomachean Ethics. Translated and edited by Roger Crisp. Cambridge: Cambridge University Press, 2000. Foot, Philippa. Natural Goodness. Oxford: Oxford University Press, 2001. Geach, Peter. ‘Good and Evil.’ Analysis 17, no. 2 (1956): 33–​42.

Life and Other Basic Rights in Anscombe    323 Griffin, James. On Human Rights. Oxford: Oxford University Press, 2008. Hume, David. A Treatise of Human Nature: Being an Attempt to Introduce the Experimental Method of Reasoning into Moral Subjects. Edited by David Norton and Mary Norton. Oxford: Oxford University Press, 2000. Korsgaard, Christine. Self-​Constitution: Agency, Identity, and Integrity. John Locke Lectures 2002. Oxford: Oxford University Press, 2009. Maritain, Jacques. The Rights of Man and Natural Law. London: Centenary Press, 1944. Müller, Anselm. Was taugt die Tugend? Elemente einer Ethik des guten Lebens. Stuttgart: Kohlhammer-​Verlag, 1998. Nieswandt, Katharina. ‘Anscombe on the Sources of Normativity.’ Journal of Value Inquiry 51, no. 1 (2016): 141–​163. Nieswandt, Katharina. ‘Do Rights Exist by Convention or by Nature?’ Topoi 35, no. 1 (2016): 313–​325. Nieswandt, Katharina. ‘What Is Conventionalism about Moral Rights and Duties?’ Australasian Journal of Philosophy 97, no. 1 (2018): 15–​28. Parfit, Derek. Reasons and Persons. Oxford: Clarendon Press, 1984. Rawls, John. ‘Two Concepts of Rules.’ Philosophical Review 64, no. 1 (1955): 3–​32. Singer, Peter. ‘All Animals Are Equal.’ In Animal Rights and Human Obligations, edited by Tom Regan and Peter Singer, 148–​162. Englewood Cliffs, NJ: Prentice Hall, 1989. Tasioulas, John. ‘Are Human Rights Essentially Triggers for Intervention?’ Philosophy Compass 4, no. 6 (2009): 938–​950. Teichmann, Roger. The Philosophy of Elizabeth Anscombe. Oxford: Oxford University Press, 2008. UN General Assembly. Universal Declaration of Human Rights. 217 (III) A. Paris, 1948. Waldron, Jeremy. ‘Is Dignity the Foundation of Human Rights?’ In Philosophical Foundations of Human Rights, edited by Rowan Cruft, Matthew Liao, and Renzo Massimo, 117–​137. Oxford: Oxford University Press, 2015.

Chapter 14

Ansc ombe a nd Sexual Et h i c s Duncan Richter

Anscombe’s views on sexual ethics are widely known and have been widely criticised.1 My goal in this chapter is to think about what she might have said instead, if we take some of these criticisms seriously. I will not attempt to discuss every aspect of sexual ethics but will focus particularly on homosexuality and contraception. Those who do not take criticisms of Anscombe’s views on these issues seriously, or who take them seriously but end up rejecting them, will see no reason to seek alternatives to the views Anscombe defends. But it seems worthwhile, for those of us who do not accept Anscombe’s conservative beliefs, to ask what else she might have believed instead. So far as Anscombe was Catholic there is not much room for her to have thought anything else than she did. But her faith was not blind, and she offered reasoned defences of her sexual morals. To the extent that her defence of these positions fails, then, she ought to have had other options available to her. In what follows I will try to explain what these are. This will require me first to review her positions, the defence she offers of them, and the criticisms that have been aimed at them. I will then evaluate what I take to be the strongest of these criticisms and consider what plausible positions are left to those who would follow Anscombe as far as reason allows.

Anscombe on Sexual Ethics Because sex, or specifically ‘genital union,’ as Anscombe calls it, is inherently procreative, she believes that there are rules of conduct that are peculiar to sex, derived from the 1  An earlier version of this chapter was presented at the von Wright-​Wittgenstein Research Seminar at the University of Helsinki on March 14, 2019. I am grateful to the audience members and to Roger Teichmann for their helpful comments.

Anscombe and Sexual Ethics    325 nature of sex itself.2 She therefore believes in sexual ethics as its own area of ethics, with its own peculiar rules. This might seem doubtful to, for instance, utilitarians, who might expect general rules, or even just one general moral principle, to cover sexual behaviour. But it is not utterly counter-​intuitive. If we think about adultery, paedophilia, and rape, then we might agree that there is something special about the ethics of sexual activity. Rape is a terrible crime, not only because of the violence, or the threat of violence, involved in it, and not only because the will of the rapist is imposed on the victim against his or her will. The fact that it is sexual activity that is forced on the victim is part of what makes rape so bad. At least, those who think of it as a terrible crime, I believe, see it this way. Something similar could be said about the sexual abuse of children. This too might involve violence, but it need not. A child might not understand well enough what is going on to resist. This makes the act an abuse of trust or of innocence, but the abuse seems especially bad precisely because sex is involved. The generally accepted badness of adultery also brings out the widespread and deeply felt conviction that sexual activity has a special, albeit hard to specify, importance. Adultery is surely not bad only because it is something that people vow not to commit when they get married. The vow is made because adultery is already regarded as something very much to be avoided. Based on an idea of sex as having particular importance, Anscombe argues against the use of contraception. In ‘You Can Have Sex without Children,’ she writes, ‘[Y]‌ou turn copulation into a wrong and shameful act if before or during or after the act you do something that you suppose destroys the possibility of conception and do this in order to destroy that possibility.’3 The act of copulation is, she says, intrinsically (even if not usually) generative. It is therefore good, in the absence of defeating circumstances, since generation is good. Considered as intentional actions, sex with a condom and sex using a contraceptive pill are equally bad, in her view. This is precisely because they are intentionally non-​generative. The objection to Anscombe’s view will be that there is nothing wrong with such sex, other things being equal. Her reply to that will be that there is something wrong, because of the intrinsically generative nature of the sexual act. Contraceptive sex not only changes the nature of the act (which, simply qua change, need not be bad) but intentionally undoes the very special good of this kind of act. The act through which the miracle of life occurs is deliberately made into a different, non-​miraculous kind of act. Engaging in that kind of act constitutes disrespect for the value of life, is a failure of due appreciation. This, I think, is at least an intelligible view. It is not, however, the only such view that one can take of the matter. One might fully intend to have lots of children and yet judge that now is not the time to risk pregnancy, and so use some method of contraception. Anscombe accepts this as long as the method used is a matter of timing rather than, say, use of a condom. But it is not clearly irrational to think that it does not, and perhaps cannot, matter what method of contraception is used. Anscombe might argue that it is important because the rhythm method merely

2  3 

See CCVM, 211. YCSC, 84.

326   Duncan Richter takes advantage of certain facts about nature without actively trying to work against nature, while contraceptive pills, condoms, and other artificial devices involve going out of one’s way to prevent something that might otherwise naturally occur. But the importance of this distinction is not so obvious that reason itself demands that we value it. It is intelligible to believe that one should accept whatever life gives, or takes away, and that therefore even the rhythm method of contraception is unacceptably wilful. It is intelligible to share Anscombe’s view of the matter. And it is intelligible that one would regard any method of contraception, other things being equal, as both all right and morally on a par with any other. So Anscombe’s view, while understandable, is far from being obviously the only reasonable or rational view one can take. Anscombe might reply that her view can be correct without being obviously correct. We might have to investigate more thoroughly or think more deeply in order to see which of these views is best. I wonder, however, whether there is any way to settle such disagreements. Gareth Moore, who shares many of Anscombe’s beliefs, sees a limited place for contraception, as long as it is within the context of a non-​casual relationship. This place is limited because: In so far as the use of contraception—​or of natural family planning—​diminishes the sense of life as a gift and encourages us to see the world in terms of convenience and manageability, it is incompatible with Christianity.4

On the other hand, this does not mean that we must never use contraception, according to Moore. Imagine, for instance, ‘a devout, poverty-​stricken mother of ten.’5 She might well believe that now is not a good time to have another child, and hence use contraception. Moore’s point is to suggest that some use of contraception might be compatible with Christianity. My point is rather that what constitutes disrespect or, on the contrary, shows due appreciation is not objectively, rationally decidable. It is not so hopelessly arbitrary as to be not worth thinking about, but matters such as what unduly diminishes the sense of life as a gift and what unduly encourages seeing the world in terms of convenience seem inherently subjective. One might object that I have misleadingly inserted the word ‘unduly’ into words taken from Moore, but Moore does not mean that everything that, for instance, diminishes the sense of life as a gift is incompatible with Christianity. Any burden might do that, just as any convenience might encourage seeing the world in terms of convenience. Further work would be needed if it were to be demonstrated that Anscombe’s view is demonstrably the best to take. It is true that refraining from having sex seems quite different from having sex with contraception, but whether this is morally significant remains to be shown. So we should look at what else Anscombe says on the subject. Here is another of her arguments against contraception:6 1. Sex with contraception embodies the intention to use one another’s bodies to achieve orgasm and not to engage in a generative act. 4 

Gareth Moore, The Body in Context: Sex and Catholicism (London: Continuum, 2001), 179.

6 

See YCSC, 96 for the passage I am paraphrasing here.

5 Moore, The Body in Context, 180.

Anscombe and Sexual Ethics    327 2. If this is all right then it looks as though it would be all right if a couple unable to engage in generative sex because of injury caused by an accident were to engage in mutual masturbation instead. 3. But this would clearly not be all right. 4. Therefore contraceptive sex must not be all right either. The problem with this is that step 3 seems false. The word ‘clearly’ does not belong, and the very idea that the situation in 2 would be at all wrong (although the injury is clearly unfortunate) seems mistaken. Anscombe’s argument could, therefore, be taken as a defence of contraception (of the form modus tollens), even though she very clearly does not mean it this way. Some of the strongest objections to Anscombe’s arguments on sexual ethics come from Jenny Teichman.7 She points out that, according to Anscombe, ‘[i]‌n safe period sex a non-​condemnable intention (to avoid generation) is a further intention, and in contraception a non-​condemnable intention (to avoid generation) is embodied in the act (that is, makes it what it is). But how can an act which is otherwise all right be condemned on account of its embodying, as part of its intrinsic character, an intention which is also all right?’8 It does seem as though the reason why contraceptive sex is supposed to be wrong (i.e., not all right) has not been fully explained. Nevertheless, some additional explanation could be offered. Both the intention to give a small gift and the act of giving a small gift are all right, but deliberately turning the giving of a large gift into the giving of a small gift might not be all right in certain circumstances. For instance, the change might disappoint a recipient who had been led to expect a larger gift, or the change might be motivated by spite. Similarly, it might be thought, the deliberate turning of a potentially generative act into a non-​generative act might not be all right, even though non-​generative acts are otherwise all right. The cases are not parallel, though. If contraceptive sex is bad it is not bad because God was anticipating something more and has now been let down. If God expects more of us it is not in this sense that he does so. And if the reason for using contraception were spite, then the act would not embody an intention that is all right. So Teichman’s objection that Anscombe’s argument seems to need something more stands, and it is hard to see how this extra something could be provided. Another reductio from Anscombe can be found in ‘Contraception and Chastity,’ where she writes: If contraceptive intercourse is permissible, then what objection could there be after all to mutual masturbation, or copulation in vase indebito, sodomy, buggery, when normal copulation is impossible or inadvisable (or in any case,

7  Jenny Teichman, ‘Intention and Sex’, in Intention and Intentionality: Essays in Honour of G. E. M. Anscombe, ed. Cora Diamond and Jenny Teichman (Brighton: Harvester Press, 1979), 147–​161. 8  Teichman, ‘Intention and Sex’, 157.

328   Duncan Richter according to taste)? It can’t be the mere pattern of bodily behaviour in which the stimulation is procured that makes all the difference! But if such things are all right, it becomes perfectly impossible to see anything wrong with homosexual intercourse, for example.9

The idea seems to be that homosexual intercourse is so obviously bad that contraceptive intercourse must be so too. But neither is obviously bad. One better point is that the mere pattern of bodily behaviour cannot be the important thing. It might not be obvious that this is the case, but I think that it is the case. And Anscombe’s line of thought might be worth following at least a little farther. She has in mind someone who does not object to (a) vaginal sex with contraception but does object to such acts as (b) mutual masturbation and (c) anal sex. How might such a combination of views be justified? Well, one might say, a is not the same thing as either b or c, so there is no contradiction in being against one or two but not the others. Anscombe’s response is to accept that there is no outright logical problem, but to dismiss this set of views as implausible. If b and c are to be opposed, then this ought, she seems to think, to be on the basis of some more general principle. And what might that be if not one that would equally rule out a because it too is inherently non-​generative? Her point is a good one, but she does not seem to consider the possibility that someone might accept a, b, and c. Which means that she offers no argument here against that combination of views. Elsewhere she writes, ‘[T]‌hough I know Catholics who solemnly defend homosexual activity, I don’t know any who make propaganda for bestiality, group-​sex or paedophilia. No doubt, however, all that will come as the world grows more accepting of these things’.10 This is a relatively common view today, that homosexuality belongs with such acts as bestiality and paedophilia.11 The essential feature of ‘these things’ is that they are unnatural, in a specific sense of that word, that is, they are sexual acts inapt for the bringing of new human life into the best kind of circumstances for the raising of a child. Group sex, of course, might produce a child, but not in circumstances that Anscombe, or Aquinas, would think good for the child.12 Homosexuality (as a kind of act, not an orientation or identity), bestiality, and paedophilia will not produce children, and are therefore forms of what Aquinas calls unnatural vice. To her considerable credit, Anscombe is much more consistent on this than seem many people today who lump ‘these things’ together. Typically (I am speaking of the most visible ones, not what might be a silent majority of sincere religious believers) they do not speak out against masturbation, for instance, while clearly she does. Nevertheless,

9  G. E. M. Anscombe, ‘Contraception and Chastity’, in GG2, 170–​191, 183. Hereafter this paper is referred to as CC. 10  HV, 197. 11  The view is common enough that the Southern Poverty Law Center includes it on a list of anti-​gay myths (https://​www.splcen​ter.org/​fight​ing-​hate/​intel​lige​nce-​rep​ort/​2011/​10-​anti-​gay-​myths-​debun​ked). 12  See Aquinas, Summa Theologiae II-​II, Q154, A2.

Anscombe and Sexual Ethics    329 there are aspects of homosexuality that make it very different from both bestiality and paedophilia. Consent is chief amongst these.13 Homosexual sex can be consensual, while bestiality cannot be. Paedophilia cannot be meaningfully consensual (that is, a child can appear to consent to acts it does not understand, but this lack of understanding makes the ‘consent’ void). Group sex can also be consensual, but arguably it cannot be loving in the way that homosexual sex can be. (See later discussion for more on this.) As Anscombe sees it, the ethics of contraception is related to the ethics of homosexuality and same-​sex marriage. Here she is on same-​sex marriage: If you can turn intercourse into something other than the reproductive type of act (I don’t mean of course that every act is reproductive any more than every acorn leads to an oak tree but it’s the reproductive type of act) then why, if you can change it, should it be restricted to the married? Restricted, that is, to partners bound in a formal, legal union whose fundamental purpose is the bringing up of children? For if that is not its fundamental purpose there is no reason why for example ‘marriage’ should have to be between people of opposite sexes. But then, of course, it becomes unclear why you should have a ceremony, why you should have a formality at all.14

Her argument seems to be something like this:

1. 2. 3. 4. 5.

Marriage properly involves formality and ceremony. This goes with marriage’s fundamental purpose being the bringing up of children. Children are produced sexually. Therefore sex ought to be of the generative type and occur only within marriage. Same-​sex marriage makes no sense.

Premise 1 is plausible, but not undeniably true. Nor is its precise meaning very clear: how formal and ceremonial must a wedding be? Why? Precise answers to these last two questions probably do not matter from Anscombe’s point of view, at least as far as the larger point about sexual ethics goes. It would be enough for the present argument if some formality and ceremony were called for. Let us grant that it is. Premise 2 is also debatable. I married my wife unsure whether we would ever have, or want to have, children. Was the wedding therefore misguided? Should it have involved as little ceremony as it did, or less? Reasonable people might disagree about this. It does not seem unreasonable to believe that less is at stake in a marriage that will

13 

Harm is another that comes to mind, but one can be wronged without being harmed in a straightforward sense, as has been brought out by Christopher Miles Coope in his ‘Making Morality Intelligible’, Philosophy 90 (2015), 403–​455. Coope allows for an extended notion of harm based on what Ronald Dworkin says about treating someone with indignity in Life’s Dominion: An Argument about Abortion, Euthanasia and Individual Freedom (New York: HarperCollins, 1993) (see Coope, 446n90), so we need not deny that child abuse is always harmful, but I will stick to the less controversial matter of consent here. 14  CC, 172.

330   Duncan Richter involve no raising of children. Nor, however, would it seem unreasonable for someone to object strongly that their marriage was every bit as important as someone else’s even though they had no intention of raising children. A loving commitment to live together through sickness and health is quite valuable enough to justify considerable ceremony. Anscombe might agree with this point, though, but still want to distinguish ceremony of this kind from that associated with marriage proper, which she sees as being essentially connected to bringing up children. Premise 3 is true, although in vitro fertilisation is not obviously worse. The conclusion 4 looks like a non sequitur, although this might be my fault rather than Anscombe’s. It might seem to follow from premises 3 and 2, but these could also seem to justify non-​married couples’ engaging in non-​generative kinds of sexual activity. If generative sex belongs in marriage and non-​generative sex outside it, then same-​sex marriage might make less sense than the traditional kind. But this rests in part on premise 2, which is highly questionable. A same-​sex marriage with the intention of raising children, either adopted children or children from previous relationships, might call for ceremony and celebration (going by premises 1 and 2). Which would then call 5 into question. If the purpose of a formal and ceremonial wedding is to mark and celebrate a lifelong commitment of two (or more) people to love each other, then there is no evident reason at all why these people should not be of the same sex. Unless, perhaps, real love is thought to be impossible in those circumstances. It seems common to have doubts about the possibility of true love in a polyamorous relationship, although I have no idea whether such doubts are justified nor how to assess their reasonableness, but it seems old fashioned and implausible to doubt that same-​sex couples can really love each other. Anscombe says, ‘It is, I believe, universal to regard marriage as having a sort of honourableness and dignity about it. This is obviously connected with its role in reproducing and rearing children.’15 I think she is right that the dignity of marriage is connected with its role in producing and raising children. But it is not, I believe, connected only with this. There is also a seriousness about marriage that is lacking, or might be thought to be lacking, when two people merely live together with no public, formal announcement of their lifelong commitment to one another. She also brings up the idea that marriage could be ‘a pact of mutual complicity in no matter what sexual activity upon one another’s bodies’.16 This is not, however, something that she recommends. As long as the pact is genuinely consensual one might see nothing wrong with such complicity, and indeed perhaps there really would be nothing wrong with it, but I think that Anscombe’s objections would have to do both with her ideas about the proper nature of sex (that it is, and ought to be, open to the possibility of generation) and with ideas about dignity. The mere facts that I give consent and I am not hurt are not, arguably, enough to justify any sexual act whatsoever in which my body is 15  G. E. M. Anscombe, ‘Address to the Clergy: On Contraception and Natural Family Planning’, in GG2, 199–​205, 203. 16  CC, 185.

Anscombe and Sexual Ethics    331 involved. Kant would agree with Anscombe on this, for reasons that are likely to be familiar to readers of this chapter. Marriage is not about all manner of sexual activity, according to Anscombe, but it is about sex, just as sex is about marriage: Humanly speaking, the good and the point of a sexual act is: marriage. Sexual acts that are not true marriage acts either are mere lasciviousness, or an Ersatz, an attempt to achieve that special unitedness which only a real commitment, marriage, can promise. For we don’t invent marriage, as we may invent the terms of an association or club, any more than we invent human language. It is part of the creation of humanity and if we’re lucky we find it available to us and can enter into it. If we are very unlucky we may live in a society that has wrecked or deformed this human thing.

This—​that the good and the point of a sexual act is marriage—​is why only what is capable of being a marriage act is natural sex.17 It is crucial to her argument that ‘we don’t invent marriage,’ but she gives us no reason to believe this claim. It is a very strong and rather mysterious claim: we no more invent marriage than we invent language. But language is in some sense invented. It is surely a human creation, albeit no one can identify the person or persons who invented it, and imagining its being invented is just about impossible. If we think of the invention of language as being the invention of meaning, which is how I imagine the Wittgensteinian Anscombe would think of it, then no (meaningful) thought or sentence could ever introduce language. The first thought ever entertained is unthinkable (qua first thought, that is). Something similar could indeed be said about marriage. The first marriage, if we can make sense of these words, would not be part of any institution, and hence not of the institution of marriage. To say this is not to deny that there was a time when this institution did not exist. Of course there was. Presumably marriage, like language, evolved gradually. But still, it seems that half-​sense and half-​marriage are inconceivable. Something either has meaning, or is a marriage, or it does, or is, not. Something marriage-​like is conceivable, certainly, and we might disagree about exactly what to count as a marriage. Common-​law marriage, same-​sex marriage, and open marriage, for instance, are similar to paradigm cases of marriage yet different enough for it to be intelligible for some people not to count them as real marriages. It is harder, I think, to think of cases that are meaning-​like without unambiguously having, or lacking, meaning. Nevertheless, both meaning and marriage (however we define it) seem to rely on conventions, so that it is hard to conceive of either without a more or less established practice. So how was there ever a first to have meaning or to be a real marriage? And how could a human act ever introduce such a thing into the world? I do not mean that God must have miraculously granted us these gifts, as if that would clear up all the mystery. I mean that we are close

17 

CC, 185.

332   Duncan Richter to the limits of what we can understand here, unless, perhaps, we adopt a very different way of thinking.18 Marriage ‘is part of the creation of humanity’, Anscombe says. This could mean that God’s creation of marriage is part of His creation of human beings, or it could mean that marriage is part of what humanity itself has created. (And, of course, if one believes that humanity created marriage one might still believe that this creation was guided by God.) What seems most important is less who created marriage as what role it now plays in being human. Marriage in a broad, not specifically Christian sense is indeed one of the most fundamental and widespread institutions, along with funeral rites and various forms of celebration of the birth of a child. Arguably taking marriage seriously is as important to being human, to living a recognisably human life, as is taking the respectful disposal of dead bodies seriously. I do not know how to determine whether this is the case or not, but it seems at least a reasonable suggestion. If it is right, then the most liberal attitude towards sex would perhaps indeed be mere lasciviousness, while the liberal but less casual attitude would support the ersatz ‘attempt to achieve that special unitedness which only a real commitment, marriage, can promise’. Anscombe’s view is conservative, but many people with liberal attitudes would probably agree with this assessment. Sex outside marriage, they might say, is all right, but really belongs in relationships that are something like trial runs for marriage. Relationships that do not become marriages, and one-​night stands, are mistakes of a kind. Perhaps not best thought of as ‘morally wrong’, but not the goal either. They might be more or less harmless missteps, but they are still a kind of ersatz. I do not mean to say that this view is right, but I think it is very close to the mainstream view reflected and promoted in popular media. And although it usually seems gentler, more forgiving, than Anscombe’s view, it is recognisably close to her view in some important respects. Indeed an objection to Anscombe’s view might be made on just this basis. So far as the sexual acts of which she disapproves are (merely) ersatz they are not seriously objectionable. If they are also supposed to embody contempt for human dignity, then this would need to be shown. Perhaps, then, Anscombe is entitled to nothing more than the gentler, and more popular, view that I have just sketched. If we are to take marriage seriously, then we need to be realistic about it and recognise that not every marriage is always blissfully happy. Anscombe has this to say about unhappy marriages: [I]‌n an irremediably unhappy marriage one ought still to love the other . . . though [one] does not feel the sort of affection which cannot be commanded, and which is simply good fortune.

18  On inventing things such as marriage, in ST (210) Anscombe writes that the ‘question about the origin of language’ is one that ‘knowledgeable intelligent people have banned as a topic of investigation. Wittgenstein once said to me ‘Why shouldn’t men have been created ploughing and sowing?’ It would seem to me difficult to give any reason why not. And equally if you add speaking to the list’.

Anscombe and Sexual Ethics    333 But for such couples what application is there for the frequent description of sexual intercourse as ‘justified not merely for procreation but by its part in married life, as an expression of mutual love, tenderness, affection, and respect’? Are we to infer that people who are unlucky in their married life ought to abstain from intercourse?19

Shortly after this she continues thus: Clearly there are many marriages which are imperfectly happy by reason of uncongeniality but are sustained in being by habit and loyalty to the marriage bond; sexual intercourse plays a significant part in sustaining such marriages. Teaching about marriage ought absolutely not to be irrelevant to the unhappy, and flattering to the lucky. Thus the old vindication of intercourse as ‘rendering the marriage debt,’ which many find repellent nowadays, is more realistic than they are; it makes no assumption as to the state of the affections.

If sex and love go together, are we to infer that a couple that falls out of love should stop having sex? Not necessarily. Perhaps they could behave in a kind and loving way without feeling much warmth for one another. This kind of action might help feelings of affection grow. And sex might follow, or be part of, this recovery of love. We might wonder, though, whether it is really its procreative potential, and this alone, that gives sex its special importance. Anscombe rejects ‘the view which associates genital union with love’.20 She does so on the ground that more than two people might love one another, and she rejects sex involving more than two people. I will not discuss the ethics of group sex here, but I do want to discuss the connection between sex and love. The vague specialness of sex that I mentioned earlier seems to have to do with its very personal nature, its intimacy. Sexual betrayal is a deep kind of betrayal, sexual abuse a deep form of abuse. Deep in the sense that we tend to care about it a lot, and that it seems to take root in us, being exactly the kind of thing that might take years of therapy to get over (in a case of rape or incest, for instance) or that might create feelings that are hard to express or to live with. Which is not to say, of course, that everyone is affected or reacts in the same way to such experiences. Perhaps some people can more or less shrug them off. But it also makes sense, is intelligible, if adultery or sexual assault leads to suicide or murder. Without thinking that such a response is right, we can understand it, because we understand that sexual infidelity and sex crimes are the kinds of act that people sometimes do respond to in such ways. That we care so much about sex because it is how we procreate is a plausible hypothesis, but it seems to me that it is nothing more than this. People do not become blasé about marital infidelity, for instance, if they know that their partner cannot impregnate or get pregnant. Perhaps they would feel worse if their unfaithful partner actually did impregnate, or get pregnant by, someone else. There does not seem to be anything

19 

20 

YCSC, 92. CCVM, 211.

334   Duncan Richter inevitable about this, however, and part of the reason why one might feel worse in this kind of case is the possible fear that pregnancy outside the relationship makes it more likely that the relationship will end. Concern about sexual infidelity does not seem to be essentially about reproduction, even if psychological accounts of what causes such concern refer centrally to reproduction. Nor is most people’s concern about infidelity primarily about the disrespect shown for the procreative act or God’s plan. Anscombe knows that we do not care only about the danger of pregnancy in sexual ethics, of course, but her view regarding what makes unethical sexual behaviour bad has centrally to do with the nature of sex as generative, as she sees it.

Common Attitudes towards Sexual Ethics Three things are worth distinguishing here: what it is, if anything, about sexual acts that makes ethically bad ones bad (in a way that is specific to sexual ethics); what it is that we care about in people’s sexual behaviour; and what might cause us so to care. An evolutionary psychologist might speculate that we care about who has sex with whom because sex determines whose genes survive and whose do not. That is plausible, but not unquestionably true. What we care about, however, is clearly not just this. Those who disapprove of masturbation are concerned about the abuse of human sexuality, not (primarily) missed opportunities for pregnancy. Those who disapprove of the sexual abuse of children are concerned (primarily) about the treatment of children, not either sexuality as such nor possible pregnancy. What makes unethical sexual acts unethical in some people’s eyes is very much the kind of thing that Anscombe is concerned about, not surprisingly, given that her thinking is in line with an ancient tradition. But as others see it, such acts as masturbation and the use of contraceptives are really not bad at all. One version of this kind of view would regard sex as simply no big deal, so that what is bad about adultery, for instance, is all about betrayal of trust and not at all about sex. Another version of this more liberal view, however, accepts that sex is an especially or unusually significant kind of act, and yet still sees at least some casual sexual acts, such as masturbation, as completely unproblematic. It is worth considering how coherent this kind of view is. In order to do that, it will help if we first spell out what it is that we care about with regard to other people’s sexual behaviour and why this might be. Sex is not only potentially procreative; it is also an especially intimate act. Two people cannot be physically closer than when they are having sex with each other, and this closeness is often part of the goal. Physical intimacy is, at least ideally, part of the point of getting into bed with someone else. The goal is not simply to achieve orgasm by whatever means possible, using another’s body as a sex toy for one’s own pleasure. One’s own pleasure is not (always) the only goal, and it matters that one is interacting with a human being—​not just a body and not just a simulation, however popular sex dolls might be

Anscombe and Sexual Ethics    335 or become. Emotions also run high in a non-​accidental way during sex, and these are typically, or ideally, shared. So there is (often) a strong, even if not always very deep, emotional as well as physical connection. This (both the physical and the emotional intimacy, together with the intensity of the experience), it seems to me, is at least roughly what many people care about in sex. It is what makes sex outside an established relationship such a betrayal. It is (part of) what makes rape such a horrible way to force one’s will on another. But neither homosexuality nor contraception affects the special intimacy of sex between two people. The degree of intimacy of an act is not affected by the sex of those engaging in it, nor by what, if anything, they do to avoid pregnancy. This intimacy makes casual sex not so much wrong as seemingly paradoxical. The air of paradox, the apparent self-​contradiction of unserious inter-​personal intensity, goes away if we think of sex as involving, or aiming at, only physical intimacy and not emotional intimacy or something like love. Or if we think of sex as using someone else’s body for purposes of (possibly mutual) masturbation. Or if we think, or perhaps find, that even strong emotional intimacy, or feelings thereof, might not last long enough to be taken seriously. It is hard to imagine someone achieving sufficient distance from such feelings (feelings of deep friendship or love) to have an attitude towards them and yet having this attitude be one of unseriousness while the (ex hypothesi strong) feelings are present and alive. That is, successfully dismissing as unserious what feel like serious emotions while one is still having those very emotions sounds difficult and unlikely. But before and after the activity that generates them one might be casual about these feelings, perhaps even while looking forward to generating similar feelings again. Which might be because the feelings just are not all that strong, or because they are regarded as an accidental side effect of the desired physical pleasure, or because the feelings, however strong they might be in the moment, are, after all, only temporary. Some people will regard those with this kind of relaxed attitude towards sex as relatively unfeeling (and perhaps to be pitied) or as shallow (and perhaps to be condemned), but there is nothing inevitable or mandatory about such opinions. I have been contrasting two kinds of liberal views of sex. One, while permissive with regard to such things as contraception and masturbation, takes sex with another person to be the kind of thing about which one cannot really be casual, at least not without some bad faith. The other is more liberal still. Anscombe’s position, by contrast, is more conservative. This is not only because of what she sees as the natural purpose of sex. Holding one’s breath (as long as one does not thereby die) and eating casually are all right, according to Anscombe, even though breathing and eating are essential to life. Casual sex, however, and kinds of sexual activity that are not apt for generation, according to her, are not. The reason why has to do with shame. There is, she says, ‘a deep association between sex and shame.’21 Bernard Williams lends support to this view: The basic experience connected with shame is that of being seen, inappropriately, by the wrong people, in the wrong condition. It is straightforwardly connected with

21 

YCSC, 87–​88. Compare CC, 186.

336   Duncan Richter nakedness, particularly in sexual connections. The word aidoia, a derivative of aidōs, ‘shame,’ is a standard Greek word for the genitals, and similar terms are found in other languages.22

If we are not convinced, however, that sex is naturally, or inevitably, or properly associated with shame, then Anscombe offers little to persuade us otherwise: ‘This is a mysterious matter: I assume that it stems from the fall. Shame is a matter of nature, not of culture or personal fault: but surely, of flawed nature.’23 She does not mean, or claim, that all sex is shameful. Sex, even sex for pleasure, can be virtuous (being intrinsically generative, it is good). But other kinds of sex, for instance, homosexual sex and masturbation, are allegedly associated with ‘rewardless trouble of spirit’.24 She also says, ‘Sexual acts are not sacred actions. But the perception of the dishonour done to the body in treating them as the casual satisfaction of desire is certainly a mystical perception’.25 Without rejecting the notion of mystical perception one can still acknowledge that what is so perceived is not perceived universally, and when one perceives it, or claims to do so, but others do not then an impasse is reached. Anscombe also provides no evidence that homosexual sex and masturbation cause trouble of spirit without reward, and this is surely a debatable claim. Anscombe’s argument is based on various considerations, including what naturally belongs with shame, what naturally belongs with ceremony and formality, and what human beings need. These are all somewhat vague ideas, though, and we are likely to disagree about whether anything is naturally shameful, say, as well as about which things these might be. It is not a failing on Anscombe’s part to care about such matters. Indeed, it is good that she takes such a refreshingly sophisticated view of human psychology (rather than seeing it as being all about maximising utility, for instance) and that she takes psychology, or perhaps simply the things we care about, so seriously. But it does mean that there is plenty of room to disagree with her on any particular conclusion drawn from such debatable and hazy premises.

More Liberal Catholic Attitudes One might wonder, however, how much we can disagree with Anscombe if we take seriously her particular philosophical and religious commitments. Or, rather than asking simply whether she could or could not accept this or that belief or practice, we might wonder how great the distance is between Anscombe and those who do accept this or that belief or practice, and what reason there is for whatever distance we find. In terms 22 

Bernard Williams, Shame and Necessity (Berkeley: University of California Press, 1993), 78. YCSC, 88. 24  CC, 186. 25  CC, 187. 23 

Anscombe and Sexual Ethics    337 of homosexuality and Catholicism, we might turn to the work of James Martin, SJ.26 Martin does not argue that the church is wrong to reject homosexual acts, but he does adopt a rather different attitude than is evident in Anscombe’s writing. While she talks about ‘buggery,’ for instance, he points out that the Catechism of the Catholic Church says that ‘men and women with deep-​seated homosexual tendencies’ should be accepted with ‘respect, compassion, and sensitivity’.27 Anscombe might point out, as Martin does, that the catechism calls such an inclination ‘objectively disordered’ and says that homosexual acts are contrary to the natural law. They close the sexual act to the gift of life. They do not proceed from a genuine affective and sexual complementarity. Under no circumstances can they be approved.28

Martin calls such language as ‘objectively disordered’ ‘unnecessarily cruel’ and reports that some people within the church have wondered whether it should be changed or even abandoned.29 He quotes Cardinal Christoph Schönborn, the archbishop of Vienna, who in 2015 said of a friend’s same-​sex union: One shares one’s life, one shares the joys and sufferings, one helps one another. We must recognize that this person has made an important step for his own good and for the good of others, even though, of course, this is not a situation that the church can consider regular.30

To this Martin adds: ‘Before all else’, the pope says, LGBT people should be treated with dignity. That’s an immense statement, and by the way, nowhere does he mention anything about an ‘objective disorder’.31

Martin’s evident desire to avoid cruelty and to respect the dignity of all human beings is somewhat different from Anscombe’s style, although she certainly would not support cruelty or disrespect. Her primary concern is somewhat different from Martin’s, though. She is defending a traditional view that she sees as being attacked and abandoned, while he is addressing, perhaps amongst others, supporters of that view and calling for at least

26 

James Martin, Building a Bridge: How the Catholic Church and the LGBT Community Can Enter into a Relationship of Respect, Compassion, and Sensitivity (New York: HarperOne, 2018). 27  No. 2358. The latter expression is quoted in Martin, Building a Bridge on 29 and 74—​see also the subtitle of the book. 28  Quoted in Martin, Building a Bridge, 73–​74. 29  See Martin, Building a Bridge, 75. 30 Martin, Building a Bridge, 66. 31 Martin, Building a Bridge, 102.

338   Duncan Richter a change in style, in how it is presented and defended. There need not be a difference in substance between Anscombe’s ethics and Martin’s. With regard to the meaning of sex and marriage, Pope Francis has written of tenderness as a virtue belonging to marriage.32 He also talks of the ‘unitive meaning’ of marriage, as well as ‘its call to grow in love and its ideal of mutual assistance.’33 Again, this is compatible with the position defended by Anscombe, even if the emphasis and presentation differ from hers. Like Anscombe, Pope Francis has something to say about same-​sex unions as well: We need to acknowledge the great variety of family situations that can offer a certain stability, but de facto or same-​sex unions, for example, may not simply be equated with marriage. No union that is temporary or closed to the transmission of life can ensure the future of society.34

Ensuring the future of society does not appear to have been one of Anscombe’s primary concerns, although it is hardly plausible that she was indifferent to it. So we might add this to the list of reasons that she may have had for opposing same-​sex marriage. On the other hand, Pope Francis rejects here the simple equation of same-​sex unions with marriage. This is not a rejection of same-​sex unions simpliciter. The refusal to count same-​sex unions as just the same, or just as valuable, as any other marriages will not be accepted by every proponent of same-​sex unions, but it is far more liberal than Anscombe’s position appears to be. It also seems to be more or less compatible with the arguments she presents. According to Pope Francis, the good of the spouses ‘includes unity, openness to life, fidelity, indissolubility and, within Christian marriage, mutual support on the path towards complete friendship with the Lord.’35 Much of this is possible within a same-​sex union. Openness to life is the most obvious exception, but then pregnancy is not a real possibility for every heterosexual couple. And yet, the pope quotes the Catechism of the Catholic Church No. 1654: ‘[S]‌pouses to whom God has not granted children can have a conjugal life full of meaning, in both human and Christian terms.’36 It is true that he suggests that children benefit from being raised by a mother and father who love one another and bring both feminine and masculine characteristics to the family, but there is no requirement that women be especially feminine or men masculine.37 One person might be feminine in some ways and masculine in others, and any two people, whatever their sex, are likely to bring different characteristics to a family.

32 Francis, Amor Laetitia: On Love in the Family (Huntington, IN: Our Sunday Visitor, 2016), 24. 33 Francis, Amor Laetitia, 30.

34 Francis, Amor Laetitia, 42. 35 Francis, Amor Laetitia, 58.

36 Francis, Amor Laetitia, 60. 37 

See Francis, Amor Laetitia, 112–​113.

Anscombe and Sexual Ethics    339 Furthermore, the pope himself sees that masculinity and femininity are vague, perhaps even fluid categories: It is true that we cannot separate the masculine and the feminine from God’s work of creation, which is prior to all our decisions and experiences, and where biological elements exist which are impossible to ignore. But it is also true that masculinity and femininity are not rigid categories.38

In other words, even if we accept that the distinction between masculine and feminine is not (at all?) a human creation and ought to be accepted as part of God’s work, it is still possible to accept a degree of flexibility with regard to what is masculine and what is feminine. So far as there is truth in the idea that families benefit from a mix of masculine and feminine characteristics—​and there is much to discuss here—​such a mix might be provided by at least some, and perhaps even all, same-​sex couples. Not because all loving couples include one relatively masculine partner and one relatively feminine one, but because few individuals and even fewer couples are likely to have exclusively masculine or exclusively feminine characteristics.

Non-​C atholic Alternatives As well as asking what else Anscombe might have thought on this topic without giving up some of her most fundamental commitments, it is worth asking what those of us who do not share those commitments might nevertheless learn from her thinking. In an interview with Silver Bronzo, Cora Diamond comments on Anscombe’s work on sexual ethics in a way that is relevant to this.39 Diamond rejects the idea that Anscombe’s work in this area can easily be divided into the part aimed at philosophers in general and the part aimed specifically at Catholics. Diamond notes that ‘when she is talking about contraception, she is arguing that the view she recommends is not something that has to be derived from revealed truth. She argued that her position was backed up by rational considerations’.40 Diamond compares Anscombe’s denial that having sex is ‘like picking a mushroom and casually eating it’ with D. H. Lawrence’s views on ‘cocktaily sex.’41 Anscombe and Lawrence, she says, take the discussion of sexual ethics ‘in a direction that needs to be explored.’42 Diamond adds Stanley Hauerwas’s name to this list of people who have something interesting to say, who keep open ‘questions of specifically

38 Francis, Amor Laetitia, 180. 39 

Cora Diamond, ‘Philosophy in a Realistic Spirit: An Interview by Silver Bronzo’, Iride 26, no. 69 (May–​August 2013), 239–​280. 40  Diamond, ‘Philosophy in a Realistic Spirit’, 278. 41  Diamond, ‘Philosophy in a Realistic Spirit’, 279. 42  Diamond, ‘Philosophy in a Realistic Spirit’, 279.

340   Duncan Richter sexual morality’ as something that ought to be seen as an issue.43 That is, all three authors treat sex as an area or kind of human activity with its own ethics rather than as just another field to which a general ethic (e.g., respect autonomy, avoid harm, and so on) is to be applied. The essay by Hauerwas that Diamond refers to is ‘Sex and Politics: Bertrand Russell and “Human Sexuality.” ’44 As the paper’s title suggests, Hauerwas emphasises the political aspect of sex. Not in the sense that he has much to say about feminism, but he questions the liberal assumption that sex is a private matter. On the contrary, he argues, ‘Marriage is, as Russell argues, a biological institution to beget and rear children for the ends of particular communities.’45 Sex outside marriage, insofar as this bears on the begetting and rearing of children within marriage, is equally a political issue on this view. And it is hard to see how sex outside marriage could have nothing to do with sex in marriage. Love comes into marriage, according to Hauerwas, not because this is the nature of marriage but because Christianity commands us to love one another.46 As far as sexual ethics goes, it is not clear what follows from this. This is not a criticism of Hauerwas but a simple paraphrase. He ends his essay with the assertion that ‘we have no idea what or how sexual activity should be embodied in our character until we know how marriage should be shaped and sustained,’ and he does not claim to have the answer to that question.47 For our purposes, the relevance of all this is primarily that Hauerwas regards sex as both weighty and mysterious, and not as something we can understand or deal with well if we treat it simply as another mode of activity in which private individuals may or may not choose to engage. Lawrence’s views, which Diamond groups with Hauerwas’s, are expressed in his essay ‘A Propos of Lady Chatterley’s Lover.’ Here he writes: A young girl and a young boy is a tormented tangle, a seething confusion of sexual feelings and sexual thoughts which only the years will disentangle. Years of honest thoughts of sex, and years of struggling action in sex will bring us at last where we want to get, to our real and accomplished chastity, our completeness, when our sexual act and our sexual thought are in harmony, and the one does not interfere with the other.48

Lawrence rejects both the puritanical refusal to think about sex and the ‘modern young jazzy and high-​brow’ tendency to think nothing of it, as if it were ‘a slightly nasty

43 

Diamond, ‘Philosophy in a Realistic Spirit’, 279. Stanley Hauerwas, ‘Sex and Politics: Bertrand Russell and “Human Sexuality,”’ Christian Century 95 (1978), 417–​422. 45  Hauerwas, ‘Sex and Politics’, 420. 46  See Hauerwas, ‘Sex and Politics’, 420. 47  Hauerwas, ‘Sex and Politics’, 422. 48  D. H. Lawrence, ‘A Propos of Lady Chatterley’s Lover’, in Phoenix II: Uncollected, Unpublished, and Other Prose Works, ed. Warren Roberts and Harry T. Moore (New York: Viking Press, 1968), 487–​ 515, 490. 44 

Anscombe and Sexual Ethics    341 toy’ or a cocktail.49 In contrast, he praises the Catholic Church and its appreciation of the significance of sex: ‘To the priest, sex is the clue to marriage and marriage is the clue to the daily life of the people and the Church is the clue to the greater life.’50 Echoing some of the political concerns of Bertrand Russell and Stanley Hauerwas, Lawrence writes: Christianity brought marriage into the world: marriage as we know it. Christianity established the little autonomy of the family within the greater rule of the State. Christianity made marriage in some respects inviolate, not to be violated by the State. It is marriage, perhaps, which has given man the best of his freedom, given him his little kingdom of his own within the big kingdom of the State, given him his foothold of independence on which to stand and resist an unjust State. Man and wife, a king and queen with one or two subjects, and a few square yards of territory of their own: this, really, is marriage. It is a true freedom because it is a true fulfilment, for man, woman, and children. Do we, then, want to break marriage? If we do break it, it means we all fall to a far greater extent under the direct sway of the State.51

Marriage, he argues, is the best way for two souls to develop in unison, rejecting both total self-​interest and complete altruism, and offering development, fulfilment, sustenance, and nourishment until the very end of life.52 The sharing of life’s changes and experiences with a partner is essential to full development of the soul in this life, Lawrence believes, which makes something like marriage essential, but he adds that marriage must be sexual, and that ‘this is the meaning of the sexual act’.53 With this in mind he rejects ‘cocktaily’ sex, masturbation, and homosexuality, the last being merely ‘secondary’ to heterosexual sex, while the first two are positively harmful to society, as he sees it.54 It is hard to know exactly how to evaluate this view, partly because Lawrence does not offer very much in the way of argument for it. The idea that something like marriage has great value is plausible. The idea that marriage ought to be sexual is also plausible. So too is the idea that this is where sex belongs in human life, or perhaps, at least, where it most belongs. But, aside from the fact that plausibility is not the same as established truth, one might surely wonder whether we should add a qualifier such as ‘for the most part’ or ‘ideally, but not necessarily’ to these claims. Might there not be people who are better off alone or in a non-​sexual relationship? The great strength of Lawrence’s view, it seems to me, is that it rejects a casual attitude towards sex that is hard to square with

49 

Lawrence, ‘A Propos of Lady Chatterley’s Lover’, 491. Lawrence, ‘A Propos of Lady Chatterley’s Lover’, 499. 51  Lawrence, ‘A Propos of Lady Chatterley’s Lover,’ 502. 52  See Lawrence, ‘A Propos of Lady Chatterley’s Lover’, 503. 53  Lawrence, ‘A Propos of Lady Chatterley’s Lover’, 506. 54  See Lawrence, ‘A Propos of Lady Chatterley’s Lover,’ 492 for the word ‘cocktail’. See 508 for his views on masturbation and homosexuality, as well as ‘modern sex-​activity’. 50 

342   Duncan Richter both the physical connection between sex and babies (which, should this need saying, we value tremendously) and with the fact of human psychology that we care very much about sex, in the sense that we tend to think about it a lot and in the sense that we tend to care about it deeply. The greatest weakness of his view, as I see it, is the rejection of homosexual sex as secondary. Such sex lacks the connection to babies, true, but there is surely no reason to believe that either the reality or the possibility of babies is essential to the full development of the soul. Keeping in mind Anscombe’s example of the couple that engages in mutual masturbation because of injuries that make other forms of sex impossible for them, we might also question Lawrence’s blanket rejection of masturbation. It seems to be only cocktaily homosexuality and cocktaily masturbation that he has a strong argument (I do not say watertight case) against.

Conclusion What Lawrence offers, then, is a reasonably plausible outline of a sexual ethic, some parts of which are more plausible than others. It is defended on grounds other than those offered by Anscombe in defence of her view, but it is not a wholly different view than the one she favours. If we are not prepared to go as far from Christianity as Lawrence, then we might prefer the rhetorically gentler version of Catholic teaching offered by Pope Francis and Father Martin. Or, of course, we might prefer something much more liberal or, like Anscombe’s view, more conservative. My own view is that none of these positions can be demonstrated uncontroversially to be right or wrong, and, because of this, all are worth thinking about. I have especially been concerned to bring out possible shortcomings in Anscombe’s position and to look for possible alternatives that are not too distant from hers. I do not claim at all to have shown that any of these is right, but I hope I have clarified how they differ from Anscombe’s and some of what there is to be said for them.

Bibliography Anscombe, G. E. M. ‘Address to\ the Clergy: On Contraception and Natural Family Planning’. In Faith in a Hard Ground: Essays by G. E. M. Anscombe, edited by Luke Gormally and Mary Geach (Exeter: Imprint Academic, 2008), 199–​205. Aquinas, Thomas. Summa Theologiae II-​II. Q154. A2. Coope, Christopher Miles. ‘Making Morality Intelligible’. Philosophy 90 (2015), 403–​455. Diamond, Cora. ‘Philosophy in a Realistic Spirit: An Interview by Silver Bronzo’. Iride 26, no. 69 (May–​August 2013), 239–​280. Dworkin, Ronald. Life’s Dominion: An Argument about Abortion, Euthanasia and Individual Freedom. New York: HarperCollins, 1993. Francis. Amor Laetitia: On Love in the Family. Huntington, IN: Our Sunday Visitor, 2016.

Anscombe and Sexual Ethics    343 Hauerwas, Stanley. ‘Sex and Politics: Bertrand Russell and “Human Sexuality”’. Christian Century 95 (1978), 417–​422. Lawrence, D. H. ‘A Propos of Lady Chatterley’s Lover’. In Phoenix II: Uncollected, Unpublished, and Other Prose Works, ed. Warren Roberts and Harry T. Moore. New York: Viking Press, 1968, 487–​515. Martin, James. Building a Bridge: How the Catholic Church and the LGBT Community Can Enter into a Relationship of Respect, Compassion, and Sensitivity. New York: HarperOne, 2018. Moore, Gareth. The Body in Context: Sex and Catholicism. London: Continuum, 2001. Patterson, Charlotte J. Lesbian and Gay Parenting. Washington, DC: American Psychological Association, 2005. Teichman, Jenny. ‘Intention and Sex’. In Intention and Intentionality: Essays in Honour of G. E. M. Anscombe, ed. Cora Diamond and Jenny Teichman. Brighton: Harvester Press, 1979, 147–​161. Williams, Bernard. Shame and Necessity. Berkeley: University of California Press, 1993.

Chapter 15

Linguistic I de a l i sm and Hum an E s se nc e Rachael Wiseman

15.1  Introduction In ‘The Question of Linguistic Idealism’ Anscombe writes: [T]‌here are, of course, a great many things whose existence does depend on human linguistic practice. The dependence is in many cases an unproblematic and trivial fact. But in others it is not trivial—​it touches the nerve of great philosophical problems. The cases I have in mind are three: namely rules, rights and promises. With each of these there is associated a certain use of modal notions.1

In this remark Anscombe connects the question of linguistic idealism with the subject matter of ethics. This chapter is an attempt to say something about that connection, and by doing so to give a novel perspective on the task that Anscombe sets us in ‘Modern Moral Philosophy’.2 In particular, I want to focus on the second of Anscombe’s three theses: [T]‌he concepts of obligation and duty—​moral obligation and moral duty, that is to say—​and of what is morally right and wrong, and of the moral sense of ‘ought,’ ought to be jettisoned if it is psychologically possible.3

The proposal I want to explore is, in a nutshell, the following. When Anscombe says in ‘Modern Moral Philosophy’ that we should jettison talk of ‘moral obligation’, ‘moral

1 

QLI, 118. MMP, 26. 3  MMP, 26. 2 

Linguistic Idealism and Human Essence     345 duty’, and ‘moral ought’, she is thinking in part about the ‘use of modal notions’ that is characteristic of a statement of a rule, right, or promise. She thinks that this use expresses a kind of necessity that depends for its existence on human linguistic practice. The character of this dependence makes the obligations and duties poor candidates for the role we want them to perform in moral philosophy when we speak of moral obligation and moral duty. Anscombe suggests that we can instead derive ethical norms from the concept human, a concept that is expressed by and in human linguistic practice, but in such a way that the objects that fall under it (viz. humans) do not depend for their existence on that practice. This is the right place to seek the sort of necessity that we want in ethics because these norms do not depend for their existence on human linguistic practice. As such, though the linguistic modal notions that express them are, trivially, dependent on human linguistic practice, the normativity they express is not; as such, a moral philosophy that makes central such modal notions delivers realism in ethics. In section 15.2 I describe the form of linguistic idealism that is contained in Wittgenstein’s later work and adopted by Anscombe—​Anscombe calls it ‘a partial idealism’.4 The aim in this section is to sketch a philosophical and methodological background against which to situate that second thesis from ‘Modern Moral Philosophy’. In section 15.3 we come to the quotation with which I began, and the idea that the use of modal notions in the expression of rules, rights, and promises expresses a species of necessity that is created by, rather than merely expressed in, human linguistic practice. The aim will be to show the way—​or at least one way—​in which this ‘touches the nerve of great philosophical problems’. Section 15.4 outlines the sort of ethics that would not be dependent on human linguistic practice in that way and which Anscombe’s partial idealism can accommodate. The claim is that Anscombe is a (partial) linguistic idealist but a moral realist. She achieves this by orienting her moral philosophy towards a set of facts that relate to what is of serious importance to humans as such. These are facts about which she is a realist—​humans do not depend for their existence on human linguistic practice—​and so this is where she looks for the sort of norms that can sustain moral realism. The outcome is the position that Anscombe recommends in ‘Modern Moral Philosophy’ when she says that we should look for the norms of ethics where we find statements such as ‘that a man ought to have so many teeth’.5

15.2  Wittgenstein and the Question of Linguistic Idealism Empirical idealism, or empirical pragmatism, is the view, or cluster of views, according to which truth or reality is in some way created or constructed by the human mind. 4  5 

QLI, 118. MMP, 38.

346   Rachael Wiseman Cabbages and horses and pain and the colour red exist only in virtue of us. According to these views, it is a mistake to think that sameness and difference are features of the world independent of human thought. Rather, it is our concepts, ideas, our forms of thought that pattern the world and divide it up into things that belong together and things that do not. The generality that belongs to thought does not have its basis in anything beyond thought itself. Linguistic idealism is a version of idealism that takes an explicitly linguistic form. It says that ultimately it is human language that imposes this structure, that ‘essence is created by grammar’, in the sense that (and this is Anscombe’s formulation) what a thing is (the ‘essential’) is ‘the mark of a concept, not the property of an object’.6 If one thinks, as the later Wittgenstein did, that we should understand concepts in terms of human linguistic practice, then if one is an idealist, one will be a linguistic idealist. The question of whether linguistic idealism is true need not of course invite a straight yes or no answer. One might be an idealist about one area of reality, one set of objects, and a realist about another. For example, one might think that horses would exist even if there had been no human concept horse, but that the colour red would not without the human concept red. That it is a truth independent of human language that water boils at 100 degrees centigrade but not that murder is wrong. Even a quite committed realist will want to recognise that some things are trivially dependent for their existence on human linguistic practice—​words, sentences, and predicate expressions, for example. The question of whether Wittgenstein, in his later thought, was a linguistic idealist is one that naturally occurs in light of passages such as the following: If anyone believes that certain concepts are absolutely the right ones, and that having different concepts would mean not realizing something that we realize—​then let him imagine certain very general facts of nature to be different from what we are used to, and the formation of concepts different from the usual ones will become intelligible to him.7

Wittgenstein reminds us, often, that we can imagine linguistic practices, and hence concepts, that are quite different from ours. He writes, for example, of a group of people who ‘piled . . . timber in heaps of arbitrary, varying height and then sold it at a price proportionate to the area covered by the piles’. He says that they would not ‘mean the same by ‘a lot of wood’ and ‘a little wood’ as we do’.8 The possibility of different linguistic practices should not alone make a quarrel with the empirical realist—​she ought to accept that our concepts are not the only possible ones. But Wittgenstein also says that it wouldn’t follow from the fact our concepts differ from those of another linguistic

6 

QLI, 113. Ludwig Wittgenstein, Philosophical Investigations, trans. G. E. M. Anscombe (Oxford: Blackwell, 1953), §23. 8  Ludwig Wittgenstein, Remarks on the Foundations of Mathematics, trans. G E. M. Anscombe, ed. G. H. von Wright and Rush Rees (Oxford: Blackwell, 1956), 149. 7 

Linguistic Idealism and Human Essence     347 community that either they or we are missing something. He disbars us from saying, straight off, of the wood sellers that they are making a mistake and that we realise something they do not. And this does seem to rule out a realist attitude. To be a realist is to think that reality exerts a normative constraint on language, one that provides a criterion against which certain concepts can be said to ‘fit’. A realist would want to say: ‘The wood sellers’ concept wood is not merely different from ours; it is deficient. Their concept, unlike ours, fails to express something essential about wood itself: namely that wood does not change in quantity by being stacked in different ways. What we mean by ‘a lot of wood’ and ‘a little wood’—​that part of the grammar of our language—​expresses something about the essence of wood that the wood sellers miss, namely that wood is a solid, rather than a gaseous, substance’. But Wittgenstein does seem to deny, or at least demur from saying, that the wood sellers’ concept is wrong. This is what we would expect from a linguistic idealist, for whom ‘the “essential” is “the mark of a concept, not the property of an object.” ’ Much excellent work has been done by scholars of Wittgenstein to show that though he does indeed rule out any variety of realism that begins from the idea of our language fitting or failing to fit the world, his denial does not, in the end, amount to a wholesale commitment to linguistic idealism.9 I am going to piggyback on that work, in concert with Anscombe’s, in the following sketch of Wittgenstein’s ‘partial linguistic idealism’. What Wittgenstein does reject is a variety of realism that Anscombe calls ‘stupid empiricist realism’.10 According to stupid empiricist realism, our concepts are formed on the basis of an abstraction from experience, and so something that can better or worse fit the data from which the abstraction is made.11 But though empiricist realism is ‘stupid’, idealism is nevertheless ‘false’, and Wittgenstein rejects it too. It is false, and obviously so thinks Anscombe, that cabbages and horses and pain depend for their existence on human thought. The difficulty—​the ‘enormous difficulty,’ Anscombe says—​is ‘to steer in the narrow channel here: to avoid the falsehood of idealism and the stupidities of empiricist realism’.12 I think the difficulty of finding a third way is obvious enough, but let me just draw your attention to one particular difficulty around which numerous others might be thought to cluster. If we abandon the idea that our language fits or fails to fit ‘the given’ we encounter in experience, something that Wittgenstein clearly does encourage us to give up, then we seem to have lost the idea of the world exercising a normative constraint on thought. In the absence of that constraint, any attempt to maintain a realistic, 9  I am thinking in particular of work by Ilham Dilman, David Bloor, Cora Diamond, and Roger Teichmann. 10  QLI, 115. 11  See, e.g., Peter Geach, Mental Acts: Their Contents and Their Objects, Studies in Philosophical Psychology (London: Routledge & Kegan Paul, 1957) for a reconstruction of Wittgenstein’s argument against Abstractionism. 12  QLI, 115. But see Richard Gaskin, Language and World: A Defence of Linguistic Idealism (London: Routledge, 2020) for a subtle and compelling version of full-​blooded linguistic idealism in a Wittgensteinian vein.

348   Rachael Wiseman or small-​‘r’-​realist, stance—​where small-​‘r’-​realism insists only that there are many things whose existence is not dependent on human thought—​seems bound to issue in a devastating sort of scepticism. If there are objects beyond language, but those objects exercise no normative constraint on thought, then we seem helpless to resist the idea of a conceptual scheme frictionlessly spinning in the void (to borrow from John McDowell). Linguistic idealism gives thought a grip, but only at the cost of relinquishing even small-​‘r’-​realism. To bring into view the narrow course that Wittgenstein steers, Anscombe reminds us that there is more on the scene, in Wittgenstein’s later work, than a set of linguistically expressed concepts and a world that they may or may not fit. Wittgenstein’s later work does not lock us into a picture of an abstract symbolism and a world that must share an isomorphic structure with it if the former is to be capable of representing the latter. Rather we have a concrete conception of language as a human tool, the logic—​ grammar—​of which is shaped by and shapes the lives and actions of the humans who fashioned that tool for their own employment and to serve their collective ends. Let me say a little bit about this familiar idea. Concepts, for the later Wittgenstein, are human capacities that have their home in human life, where ‘human life’ is not the lives of a particular set of historically situated human beings—​we who are now alive, say—​but the human form of life. Wittgenstein sometimes speaks of the ‘history of mankind’ in this context, but we mustn’t take that to mean any temporally extended collection of individual humans. Rather, to speak of the human form of life is to speak of the patterns and structures into which individual humans are born and live and within which thought and action are intelligible. This is not, of course, to suggest that human concepts, human life, does not change, or that there are not concepts—​linguistic practices—​that are specific to humans at a particular historical or geographical moment. But Wittgenstein does, it seems to me (and this is something Anscombe points out),13 assume that there are structures to human life that are connected to ‘very general facts of nature’—​such as that humans are mammals, that human infants are dependent on adult care for many years—​and are as such ahistorical. Michael Thompson and Philippa Foot remind us that descriptions of the human form of life that connect to these ‘very general facts of nature’ have a peculiar timelessness. ‘The human infant loses its milk teeth’; ‘Humans are bipedal’.14 When Wittgenstein speaks of ‘human linguistic practice’ he means the whole of human language-​using life. So, in seeking to find a narrow course between ‘stupid empiricist realism’ and ‘false linguistic idealism’, we can follow him and look away from the dual relation between language (ways of operating with words) and reality (facts) and turn instead to the nexus of human linguistic practice, reality, and life. He looks to the whole of human life into which language is interwoven and the whole of human language into which human life is interwoven. Anscombe’s delightful illustration 13 

QLI, 113. Philippa Foot, Natural Goodness (Oxford: Oxford University Press, 2001); Michael Thompson, ‘Apprehending Human Form’, Royal Institute of Philosophy Supplements 54 (March 2004): 47–​74. 14 

Linguistic Idealism and Human Essence     349 of this—​in ‘The Question of Linguistic Idealism’—​concerns the concept length. She reminds us that the linguistic practice that characterises this concept is not merely its production of the word ‘length’ on ‘suitable occasions’ but includes activities of measuring, of weighing, of giving and receiving and putting into special places, of moving about in a huge variety of ways, of consulting tables and calendars and signs and acting in a way which is connected with that consultation.15

A person who has the concept length will—​normally—​use the word ‘length’ only on suitable occasions. But this is not all that possessing the concept amounts to. To possess the concept length is to participate in, or be able to participate in, these practices, practices that can be mastered by an individual who is aphasiac or mute. Here, the ‘general facts of nature’ against which this practice emerges and continues to be sustained include such things as that objects do not spontaneously change their length, that measuring rods do not turn to jelly, that human memory is more reliable than that of a goldfish. Compare, from Anscombe’s ‘Practical Inference’: [I]‌f things were quite different from what they actually are—​if there were for instance no characteristic expressions of pain, of fear, of joy; if rule became exception and exception rule; or if both became phenomena of roughly equal frequency—​this would make our normal language-​games lose their point.—​The procedure of putting a lump of cheese on a balance and fixing the price by the turn of the scale would lose its point if it frequently happened for such lumps to suddenly grow or shrink for no obvious reason.16

Wittgenstein’s interweaving of language, life, and world-​involving activity is what creates a space to resist the stupid empiricist realist demand that our concepts must ‘fit’ the world our language represents. Rather our concepts will, and do, serve our collective needs and interests, needs and interests that exist and arise for living creatures like us inhabiting a natural world that contains many things that do not depend for us on their existence or essence. Our concept length—​the activities into which the use of the word ‘length’ are interwoven—​is a reflection of our needs, our nature, and our habitat. The wood sellers lack our concept wood. The concept they possess is superficially similar but fundamentally different, fundamentally different because it belongs in a form of life that is radically unfamiliar. Their activities with wood—​including their systems of commerce and practices of building, construction, carpentry, and forestry (if they have such things)—​will be profoundly different from ours, and these differences will, as Wittgenstein writes, express and give rise to different interests and needs. The question of whether their concept wood is ‘the right one’, if we do raise it, will be raised with respect to those interests and to the form of life in which they are embedded. Perhaps

15 

16 

QLI, 117. PI, 142.

350   Rachael Wiseman their climate or habitat is so unlike ours that they do not build houses or shelters, and so the exchange of these items has for them to have a wholly ritualistic function, connected to myths that would shed light on the importance of surface area rather than weight. Perhaps their cognitive faculties are different, in ways that would render our concept unusable. We can accept this and still say: wood does not change its mass or quantity depending on how it is stacked. If the foregoing is something like right, then this is the key to avoiding the ‘stupidity of empirical realism’ without embracing the ‘falsehood of idealism’. Empirical realism is ‘stupid’, I take it, because it is a shallow and superficial picture of how language and the world relate. It is a picture we find ourselves drawn towards, but like all those pictures that represent the ‘first move’ in philosophy, it is one that we will soon become dissatisfied with. This dissatisfaction may lead us to say what we know to be false: that were it not for human linguistic practice ‘there could have been no such event as that a wolf killed three deer in seven days’—​that is Anscombe’s example.17 Wittgenstein’s organic and concrete conception of language means that abandoning the idea that language must ‘fit’ reality does not shuttle us back onto the counter-​view that reality must ‘fit’ language. Rather, the question of a ‘fit’ between language and reality can be eschewed and replaced by a question about the suitability of our concepts—​concepts we can describe by describing our linguistic practices—​to serve the needs and direct the attention of creatures of our kind living in the natural world in which we find ourselves. I will come back to the possibility of critique contained in this question. What is attractive about the proposal just sketched is that it provides a framework within which to resist linguistic idealism in the case of a great many concepts about which philosophers might otherwise find themselves rather exposed. The description ‘a wolf killed three deer in seven days’ employs concepts of number and time. We might add ‘a hungry grey wolf killed three deer in seven days, causing a good deal of pain’. Here we bring in number and colour and sensation. Bearing in mind that our possession of those concepts is a matter of our having certain linguistic practices, practices that serve our needs and express our interests, we can say that were we not to have those practices we may not have those concepts, or we may have had different ones. And yet this realisation does not touch the fact that a hungry grey wolf killed three deer in seven days, causing a good deal of pain. That said, it is hard to see what it would be for us not to have those concepts, or ones very close to them. Anscombe asks, ‘Could we . . . imagine ‘general facts of nature’ so different that people did not have the concept ‘horse’—​although there were horses around?’18 Such a thing, she suggests, is possibly unimaginable. But this is not to say that the existence of horses compels us to form our concept horse. The constraint that reality exercises on thought is not the ‘stupid’ constraint imagined by an abstractionist empiricism, whereby an experience dictates to its subject ‘what is to be bound up together with

17 

18 

QLI, 116. QLI, 114.

Linguistic Idealism and Human Essence     351 it’,19 but the constraints that the world places on lifeforms that inhabit it, and that arise from the natural interests and instincts of that lifeform. Constraints like: if you can’t recognise a hungry wolf, you won’t last long. Or our instinctual interest in other living things. In a passage that Anscombe quotes in Intention, Wittgenstein writes, ‘Concepts lead us to make investigations, are the expression of our interest, and direct our interest’.20 We perhaps cannot imagine creatures like us but who lacked a special interest in the nature and classification of other living things.

15.3 Rules, Rights, and Promises Let’s now turn to those things that—​according to Anscombe—​have an essence that is not merely expressed in human linguistic practice but is rather created by it. Anscombe notes that the existence of such things is in general ‘an unproblematic and trivial fact’. It is a trivial fact that there would be no sentences were it not for human linguistic practice. Marks on a page could exist, of course, but those marks are sentences only in the context of those practices. Less trivially: multinational corporations, theatres, schools, buses, vacuum cleaners, and garden centres would not exist were it not for human linguistic practice. Entities with this sort of complexity require for their existence a level of human cooperation and (broadly) scientific knowledge that would not be possible without communicative linguistic practice. Here, though, the ‘could not’ in were it not for these linguistic practices there could not have been these things does not express a logical impossibility but an empirical one. For us, these are things that depend for their existence on human linguistic practice. But we can (just about) imagine a community of non-​linguistic creatures whose life contained things that we would be inclined to name as such—​Anscombe experiments with and soon abandons this sort of imaginative exercise.21 But aside from these prosaic cases, Anscombe draws our attention to some remarkable cases, cases which, she says, ‘touch the nerve of great philosophical problems’. These cases concern rules, rights, and promises—​here, she thinks, the ‘could not’ expresses a logical impossibility. Anscombe says that the ‘musts’ of rules, rights, and promises are amongst those things that are created, rather than merely expressed, by our linguistic practices. Rules of conduct, rights, and promises are associated with a ‘certain use of modal notions,’ such as ‘You mustn’t . . .’; ‘You can’t . . .’; ‘You ought . . .’; etc. Modal notions, we will see, have a number of uses, but the ‘certain use’ that is associated with rules, rights, and promises is connected with the existence of a prohibition. The use refers to a particular species of necessity, namely, the necessity of not doing such-​and-​such because it is forbidden. Anscombe asks: Is it the case, that is, that prohibitions owe their existence to 19 

QLI, 114.

21 

QLI, 117.

20 Wittgenstein, Philosophical Investigations, §570, quoted in I, 84.

352   Rachael Wiseman human linguistic practice, specifically, to the linguistic practice of using modal notions in a particular way? Anscombe’s answer to this question is yes—​and it is this yes here, I want to suggest, that explains her resistance to any attempt to ground moral philosophy in rules, rights, and promises. Rules of conduct are descriptions of actions cast in a particular mood: ‘All a’s will/​ will not do such-​and-​such’. Suppose we ask ‘What is the difference between a rule and an observed regularity?’ We might reply as follows: ‘The meaning of a rule is that one must act in accordance with it, whereas the meaning of a regularity is that one will act in accord’. This answer will be understood and may be helpful to someone who is learning English as a second language or to a child of a certain age. But this answer is not a philosophical account of the difference, one that tells us what it is for a description of an action to be cast, specifically, in the mandatory future. It does not, as Wittgenstein and Anscombe would put it, tell us the grammar of the expression ‘Candidates will not write in red ink’ when it is used to express a rule rather than a regularity. This point, you may have noted, is a mirror of the one Anscombe makes at the start of Intention. There she says that though the distinction between expression of intention and prediction is usually intuitively obvious, we have not yet a philosophical understanding of the difference. She suggests, recall, that the difference is that though both are descriptions of an action cast in the future tense, the former is justified by a ‘reason for action’ and the latter by evidence—​it is this suggestion that causes her to embark on her investigation into the question ‘Why?’ With Intention in mind, we might look at the expression ‘Candidates will not write in red ink’ and offer the following suggestion: what makes this an expression of a rule rather than a prediction based on an observed regularity is that the former, and not the latter, is justified by reason for action rather than empirical evidence. The question ‘Why?’, we might say, will have its epistemic application if the expression states a regularity, and its practical application if it states a rule. But a little reflection shows that this is not true. To justify a rule by giving a reason for action—​to answer the practical question ‘Why?’—​ would be to justify the existence of the rule itself. This would be to give reasons as to why writing in red ink was something a candidate ought not to go in for. For example, if the exam scripts are to be scanned by a machine that does not recognise red ink, then this gives candidates a practical reason not to write in red ink, and it may as such explain the existence of the rule. But a rule is not a rule in virtue of having such a justification—​rules may be arbitrary, spurious, stupid, or malicious. The rule ‘Candidates will not write in red ink’ may have been imposed by a facetious administrator who was wholly ignorant of the machine’s capabilities. In that case candidates would have a practical reason to act in accord with the rule (to not write in red ink), even though there was, in fact, no reason for the prohibition itself. What we want to focus on in our account is the character of the prohibition expressed in the stopping modal ‘You must not . . . ’. The misfire helps to bring into focus the feature we are after. With a rule of conduct, it is the existence of the rule itself, and not any

Linguistic Idealism and Human Essence     353 justification for refraining from or performing the action it describes, that generates the prohibition. I am prohibited from writing in red ink because there exists the rule Do not write in red ink; the existence of the prohibition makes no contact with practical reasons such as if you write in red ink, the machine shan’t read your script. This is why Anscombe says that the phrase ‘You can’t X because that’s the rule’ does not give a reason why one ought not to X, but rather classifies the expression ‘You can’t do X’ as the expression of a rule.22 It is the self-​referentiality of rules, Anscombe thinks, that shows that the prohibitions that we find in rule-​governed practices are created, rather than expressed, by those practices in a way that buses and vacuum cleaners are not. It is the rule itself that brings into existence the prohibition on (or permissibility of) doing such-​and-​ such, and it does so regardless of whether there are good reasons for or against doing such-​and-​such. The separation I have described is indeed crucial if rules, rights, and promises are not to be routinely over-​ridden by practical considerations. It also explains why the practical question ‘Why shouldn’t I do such-​and-​such?’ is both always intelligible and always inappropriate in the face a prohibition. Even in matters of life and death, in the presence of a rule that forbids murder or permits life-​saving treatment, one can always intelligibly ask for a reason for acting—​one that goes, as it were, around the rule and straight to the action itself. If the response to this question is ‘Because it’s a rule,’ this is not an answer to the practical ‘Why?’ question ‘What’s the good of not X-​ing?’ It is rather a reminder that the proposed action, X-​ing, is prohibited by a rule and that, as such, the goodness or not of X-​ing is one that is irrelevant here and now. This is what we do, as Wittgenstein would say. So, the practical question—​a reason-​ giving answer—​can remain unanswered without this rendering my action irrational. And if the practical reasons weigh heavily in favour of doing that which the rule says in prohibited, we may in the end go against the rule while remaining cognizant of its legitimacy. What sustains the prohibition (or permission) that the rule, right, or promise creates—​the necessity of doing or not doing a certain thing—​is not, then, the existence of widely shared practical reasons for doing or not doing that thing, but our continued practice of treating a use of ‘You mustn’t X’ as the expression of a rule, right, or promise—​a practice that is not threatened by the occasional rule-​breaker or defaulter. We do, though, have numerous strategies for protesting that the words ‘You mustn’t X’ is not, or ought not, to be treated as the expression of rule, strategies that do seek to threaten the rule’s standing as such: it was not issued by a legitimate authority; it is part of a practice that is harmful or pointless; I’m not playing. It is in this sense, then, that the use of modal notions in the expression of rules, rights, and promises brings into existence the prohibition that the rule expresses. And it is for this reason that we cannot be realists about the species of necessity associated with this use. 22 

RRP, esp. 332ff.

354   Rachael Wiseman

15.4  Not Forbidden but Serious Anscombe thinks that a moral philosopher who focused on a class of rules that have moral subject matter and tried to bolster those rules so that their normative force was not dependent on merely human linguistic practice would be on a fool’s errand. Without an appeal to something outside the rule—​e.g., the authority of Him who commands it—​which would give a practical reason to follow the rule, the best that can be mustered is a hybrid of regularity and convention. This is why moral philosophers would do better to banish rules of conduct from their mind when they begin their ethical enquiries. Where next, then? I said at the start that I wanted to bring a new perspective on the task that Anscombe sets us in ‘Modern Moral Philosophy’. I hope what I have done is sketch one way to approach her thought that moral philosophers should begin their task by jettisoning ‘the concepts of obligation and duty . . . and of what is morally right and wrong, and of the moral sense of “ought.” ’ It was also my hope to gain some clarity about the task that remains for us, and though what I say now is going to be brief, I think that our discussion of linguistic idealism has made clear the materials we have to work with when we are attempting to do ethics. When we rejected the picture of a ‘fit’ between language and reality in favour of the idea that our concepts fit and reflect our needs and interests, we brought onto the scene a set of norms that relate to human nature. The background framework of partial linguistic idealism, then, orientates us towards the question of what humans need. It is here that we can reconnect the practice of laying down and acting in accordance with rules with human life. This is to bring into view a use of modal notions that is associated with a species of necessity that is independent of the human linguistic practice to which that use belongs. To see the sort of norms that are available to us, consider the use of modal notions that describe the goal-​oriented activity of a living thing. Suppose a wolf is stalking a deer. The wolf tracks the deer into a forest, and just as she is closing in for the kill, something unexpected happens. Instead of staying downwind of the deer, she moves upwind. The wolf, we say, has made a mistake. Given what she was attempting—​hunting the deer—​she has done something she ought not to have done. Here, then, is a use of a modal notion that goes with a species of necessity about which we can be realists: wolves must not step upwind of their prey. Now even were it the case that a rule for wolves that captured that norm—​Don’t step upwind of a deer!—​would have to be a linguistic instrument, we should recognise that the use of the modal notion here is quite different from that we considered previously: the norm it expresses would have existed even if the linguistic instrument did not. The norm that governs the wolf ’s action is not the deontic modal Stepping upwind of the deer is forbidden!, but is rather a norm internal to the activity hunting a deer. As such, the necessity that the rule expresses is not dependent on human linguistic practice. A wolf who could ask ‘Why shouldn’t I?’ would be given a practical reason (an answer to the question ‘What’s the good of staying downwind?’)

Linguistic Idealism and Human Essence     355 and not an answer that assigned the expression ‘Don’t step upwind of the deer’ to the logos-​type rule. The same goes for all the norms that are internal to our practices—​linguistic or non-​ linguistic. If a person is seeking to act in accord with a description—​be it hunting or dancing, or reciting the Lord’s Prayer, building a house, or digging a trench—​the action description provides a normative standard against which their action can be measured. Here, then, is what you might call ‘naturally intelligible normativity’. The fact that some of these descriptions are of action-​types that depend for their existence on human linguistic practice (e.g., reciting the Lord’s Prayer) does not mean that the modal notions that are in play when someone who acts under that description should or should not do such-​and-​such are created by human linguistic practice. In fact, what we have here are the everyday oughts, shoulds, and musts that Anscombe speaks of: a pregnant woman ought to watch her weight; a ballerina ought to point her toes; a wolf must stay downwind of a deer. We often want to formulate these as rules—​for example, if we are attempting to train someone in the activity in question—​but the necessity that the rule expresses belongs to the pattern that is laid down by the action-​description, and is not created by the existence of a linguistic practice that treats a linguistic expression as a rule. We may now insist, having articulated the rule, that in addition to X being something one ought not to do by the standard laid down by the practice, X is forbidden. Such a move might be wise if the practical reasons exert a weak psychological force or if they are difficult to grasp. For example, ‘Children must brush their teeth before bed’ or ‘Students must not consult Wikipedia’. Such a prohibition will be an ‘idle cog,’ as Wittgenstein would put it, unless the rule itself is backed up by an authority or one we have an independent—​practical—​ reason to follow. It may be helpful in connection to this point to recall Wittgenstein’s observation (in the Philosophical Investigations) that we can speak of an activity as rule-​governed when a rule is ‘an aid in teaching,’ when a rule is ‘an instrument of the game’, or when a rule is ‘something that can be read off the practice’.23 The fact that we can speak of rule-​governed activity in the first and last case—​we read the rules off the practice and use those rules as aids for teaching—​should not mask the fact that there is a ‘use of modal notions’ that is quite different from the ‘special use’ that concerns what is permitted or forbidden. One anxiety about the picture I am trying to recommend is that for all its norms—​ its shoulds and oughts and musts—​there is nothing explicitly moral in view. It can seem that we have described a world in which humans, like wolves, participate in activities that they can be said to do well or badly—​by the standard laid down by the activity itself—​but that the sense of an action being good or bad, well or badly done, lacks ethical significance. Doesn’t this mean there is no ethical stance—​just the practical?

23 Wittgenstein, Philosophical Investigations, §54.

356   Rachael Wiseman Consider again the context of deliberation in which I find myself asking the question of practical rationality in the face of rules that tells me ‘Such-​and-​such is forbidden!’ Suppose I am about to lie to my friend, and I am stopped in my tracks by the rule I have learnt: ‘Lying is forbidden!’ However, I can still ask the practical ‘Why?’ question: Why shouldn’t I, in this particular situation, lie? We can characterise that deliberation as one in which I weigh my own individual pleasure against the ‘right’—​and so as a dilemma of self-​interest versus morality. But we can also, from the position we have now reached, cast it as a question about the patterns and structures in which we act and the paths that our concepts lay down—​about the concepts lying and fidelity and the place those concepts have in the context of human life and of my human life. Perhaps the lie would be kind. My deliberation is then about the way in which I balance and rank kindness and fidelity in the context of friendship, or this friendship in particular. A serious deliberation of this kind would be a reflection on the concepts of kindness, truthfulness, and friendship, as well as a reflection on the situation in which I am now placed. It is in the context of the practical question that the topic of practical wisdom and soundness of judgement have a bearing. Philippa Foot gives a lovely example of this when she imagines deliberating about whether to break a lightly made and inconsequential promise because the opportunity has arisen to go to an ‘especially marvellous circus’.24 Here we might wonder whether keeping this promise would be foolish or fastidious or culpably austere. Our concepts work well when the patterns they lay down reflect our interests and our needs and direct our attention in ways that serve us well. There need be nothing illicit in the question ‘What are the limits of this practice?’ It does not imply the attitude of the amoralist but may be asked from a stance of serious enquiry into human life. Reflect again on the wood sellers. We saw that the worry about rejecting stupid empiricist realism was that we lose the possibility of saying ‘These concepts are the right ones’ or ‘These people don’t recognise something that we do’. But we also said, then, that concepts are a reflection of and shape our interests and needs. Concepts lead us to make investigations, are the expression of our interest, and direct our interest. A criticism of the wood cutters’ concept would be one that showed that their lives were in some way thwarted by the concept they were working with. That their practices with wood were, as Mary Midgley puts it, responsible for their lives going badly in certain respects. We might say it would be better for them if they did not measure wood in that way, but rather in this way, and in so saying we would be measuring their linguistic practice against a standard provided by their interests and needs. We might say of a community who lived in the forest and lacked the concept wolf that their lives would go better if their attention were directed in that way! They’d get eaten less often, for one! This, I am suggesting, is exactly the sort of pragmatist critique that is in view when we ask ‘Why should I?’ in the face of a norm that is conventionally followed. This question asks whether our

24 

Philippa Foot, ‘Rationality and Goodness’, in Modern Moral Philosophy, ed. Anthony O’Hear, Royal Institute of Philosophy Supplement 54 (Cambridge: Cambridge University Press, 2004), 1–​14.

Linguistic Idealism and Human Essence     357 current practices are thwarted, lacking, misdirected by the patterns our concepts lay down, or perhaps about the way in which a particular pattern has come to dominate or obscure others. But the position does not collapse into idealist pragmatism because it is underpinned by a realism that insists that there are general facts of nature, including facts about human beings, that exist independently of our interests, concepts, and linguistic practices. Wittgenstein writes, ‘At the end of my lecture on ethics I spoke in the first person. I believe that is quite essential’.25 It is essential, I am suggesting, because the ethical stance is the stance from which I undertake practical deliberation, deliberation that deepens my understanding of human essence and of the norms and patterns that shape our lives together. This deepening understanding is ethical thinking of the sort that Murdoch tells us about. This is why, I think, Mary Midgley and Philippa Foot tell us that ‘moral’ is the superlative of ‘serious’.26 A final remark. How, you might ask, is all this to be squared with Anscombe’s famous moral absolutism? Doesn’t that require the very use of stopping modals that is associated with such-​and-​such being not just something one ought not, practically, to do, but it’s being something that one is forbidden from doing? I think this appearance is in part valid and in part a mistake. Anscombe’s view, I think, is that amongst the prohibitions that we ‘read off the practice’ are a small number that will belong to any serious reflection on human nature, such that a human form of life that did not contain such norms would be manifestly and objectively bad. Norms like ‘Do not lie’ and ‘Don’t be cruel’ do not belong to this special set: these are rules about which the deliberative stance is neither puzzling nor suspect. But Anscombe does think that there are some norms that, as it were, connect directly to human essence. The fact that a human must not murder or torture or enslave, for example, is on her view a fact we can ‘read off ’ human essence. So the mistake in the objection is to think that Anscombe’s absolutism requires, as the ‘modern moral philosopher’ would have it, rules with a special kind of compelling and binding force. This mistake is connected with Williams’s complaint that Anscombe’s ethics contains a version of the Nietzschean nihilism—​that her morality depends on God as legislator. What I think is a legitimate concern is whether there are rules derivable from human essence that have the status of absolute truths. Whether there are such truths will depend on what a human is, and on whether dignity and intrinsic value belong to our essence.27 And there is a question here about whether Anscombe’s Catholic metaphysics plays a role. 25 

Friedrich Waismann, Wittgenstein and the Vienna Circle: Conversations, ed. Brian McGuiness (Oxford: Blackwell, 1979), 117–​118. 26  Mary Midgley, ‘Is “Moral” a Dirty Word?’, Philosophy 47, no. 181 (July 1972): 206–​228, 224; Philippa Foot, “When Is a Principle a Moral Principle?,” Proceedings of the Aristotelian Society, suppl. 28, no. 1 (1954): 95–​134. 27  Thanks to Andy Hamilton and Clare Mac Cumhaill for extremely helpful comments on earlier drafts of this paper and numerous brilliant and illuminating philosophical conversations. Also to audiences at University of Liverpool’s workshop ‘Linguistic Idealism’; University of Pennsylvania’s 2019 Anscombe Archive Conference on Mind & Action; Oxford University’s Jowett Society. Special thanks to Roger Teichmann.

358   Rachael Wiseman

Bibliography Foot, Philippa. Natural Goodness. Oxford: Oxford University Press, 2001. Foot, Philippa. ‘Rationality and Goodness’. In Modern Moral Philosophy, Anthony O’Hear (ed.), Royal Institute of Philosophy Supplement 54. Cambridge: Cambridge University Press, 2004: 1–​14. Foot, Philippa. ‘When Is a Principle a Moral Principle?’ Proceedings of the Aristotelian Society, suppl. 28, no. 1 (1954): 95–​134. Gaskin, Richard. Language and World: A Defence of Linguistic Idealism. London: Routledge, 2020. Geach, Peter. Mental Acts: Their Content and Their Objects. Studies in Philosophical Psychology. London: Routledge & Kegan Paul, 1957. Midgley, Mary. ‘Is ‘Moral’ a Dirty Word?’ Philosophy 47, no. 181 (July 1972): 206–​228. Thompson, Michael. ‘Apprehending Human Form’. Royal Institute of Philosophy Supplements 54 (March 2004): 47–​74. Waismann, Friedrich. Wittgenstein and the Vienna Circle: Conversations. Brian McGuiness (ed.). Oxford: Blackwell, 1983. Wittgenstein, Ludwig. Philosophical Investigations. G. E. M. Anscombe (trans.) Oxford: Blackwell, 1953. Wittgenstein, Ludwig. Remarks on the Foundations of Mathematics. G E. M. Anscombe (trans.). G. H. von Wright and Rush Rees (eds.). Oxford: Blackwell, 1956.

Pa rt I V

T H E F I R ST P E R S ON

Chapter 16

The First Pe rs on, Se lf-​C onsci ou sne s s , and Act i on Valérie Aucouturier

16.1  Introduction In her 1975 paper ‘The First Person’, Elizabeth Anscombe chooses a special type of example to illustrate the claim, which will be further developed in this chapter, that I-​ thoughts are ‘subjectless’, namely that first-​person sentences do not predicate something of an identified subject:1 The reason why I take only thoughts of actions, postures, movements and intended actions is that only those thoughts both are unmediated, non-​observational, and also descriptions (e.g. ‘standing’) which are directly verifiable or falsifiable about the person of E.A. Anyone, including myself, can look and see whether that person is standing. (FP, 35, my emphasis)

Any reader familiar with Anscombe’s Intention will recognise in this choice some features of what she there calls ‘non-​observational knowledge’, a key concept for the understanding of practical knowledge and a kind of knowledge characteristic of both the knowledge of ‘the position of one’s limbs’ and of one’s own intentional actions (I, §8).2 In this chapter I would like to explore the reasons behind this choice of ‘Anscombianly-​ preferred examples’ of ‘actions, postures, movements and intended actions’, as opposed to what Anscombe herself calls ‘Cartesianly-​preferred examples’, which are ‘far 1  Thanks to Roger Teichmann for his precious feedback on crucial issues of this paper. All the mistakes remain mine. 2  See R. Wiseman, ‘What Am I and What Am I Doing’, Journal of Philosophy 10 (2017): 536–​550.

362   Valérie Aucouturier removed in their descriptions from the descriptions of the proceedings, etc., of a person in which they might be verified’ (FP, 35). My hypothesis is that this choice of example to characterise ‘self-​consciousness’—​understood as ‘something manifested by the use of ‘I’, granted that ‘I’ is ‘not a name’ everyone uses ‘to speak only of himself ’3 (FP, 24–​25)—​ may shed light on the role of ‘self-​consciousness’ in intentional action and more specifically in practical knowledge. In other words, rather than reading ‘The First Person’ as a paper on the mere logic of the first-​person pronoun which defends the provocative thesis that ‘I’ is non-​referential,4 I shall read it as an extension of Intention focused on the philosophical concept of ‘self-​consciousness’ which is key to our understanding of practical knowledge and intentional action. However, as we shall see in section 16.2, in Intention practical knowledge (and the knowledge of what one is doing in general) is mainly considered on the basis of examples expressed in the third person (including the case of animals). But, since such knowledge soon appears closely tied to how the agent herself conceives of her own action, it is obviously related to what we may call a first-​person point of view. And the difficulty lies in this particular feature of practical knowledge.5 Therefore, in section 16.3, I will turn to an investigation of the way Anscombe’s understanding of the first person can help us clarify the role of self-​consciousness in action. To do so, I will distinguish three kinds of examples (including Anscombianly-​preferred ones) of uses of ‘I’ which manifest self-​consciousness in various ways. I will then conclude with some remarks on the special role of self-​consciousness in intentional action as distinguished from its role in self-​knowledge.

16.2  Knowing What One Is Doing 16.2.1  From a Third-​Person Point of View In order to understand the sense in which self-​consciousness is involved in human action we need to understand what Anscombe calls ‘practical knowledge’, and more specifically the sense in which practical knowledge is related to an agent’s point of view. Readers have sometimes felt bewildered that Anscombe’s enquiry into intention did not start with the consideration of what people have in mind, with ‘an introspective explanation of intention’ (I, iii), but rather with a consideration of ‘what a man actually does’

3 

See Ludwig Wittgenstein, Philosophical Investigations (Oxford: Blackwell, 2009), §§404–​410. This is how the paper is usually read, with a few exceptions, such as Vincent Descombes, who, to my knowledge, was one of the first readers to suggest that this reading was misled. See Vincent Descombes, Le parler de soi (Paris: Gallimard, 2014), 252–​294. 5  Which is illustrated, for instance, by the long discussions of §25 about the agent’s authority regarding her own intentional actions. 4 

The First Person, Self-Consciousness, and Action    363 (I, §4).6 Although there is no room to develop this point here, their bewilderment was partly unjustified, since a close reading of Intention’s first paragraphs shows how the mere consideration of intention as ‘something whose existence is purely in the sphere of the mind’ (I, §4) leads to philosophical dead ends; in other words, it leaves us with nothing philosophically interesting to say. Indeed, such considerations do not provide any sensible way of distinguishing between an expression of intention and any other kind of prediction or estimate for the future, between an intentional action and a non-​ intentional action.7 But this is not our present topic. So let us simply take note of the fact that all Anscombe’s examples of intentions and intentional actions are stated in the third person, as are her investigations of how a certain sense of the ‘Why?’ question helps to elucidate the concept of intention. A paradigmatic example is the consideration of the pumping man at §23 of Intention: Now we ask: What is this man doing? What is the description of his action? First, of course, any description of what is going on, with him as subject, which is in fact true. . . . However, our enquiries into the question ‘Why?’ enable us to narrow down our consideration of descriptions of what he is doing to a range covering all and only his intentional actions. ‘He is X-​ing’ is a description of an intentional action if (a) it is true and (b) there is such a thing as an answer in the range I have defined to the question ‘Why are you X-​ing?’ (I, §23)

The ‘range’ in question is defined as excluding that the agent ‘notices’ that he is X-​ing or needs to infer that he is X-​ing. As Anscombe already granted at this point of the book, an agent has ‘non-​observational knowledge’ of her own intentional actions and need not enquire into her reasons for acting. Now, I would like to pay close attention to the role of the ‘subject’ of the action, referred to in this quote, ‘any description of what is going on, with him as subject’, and distinguish it from the ‘agent’ of an intentional action proper. (Although the terminology here is rather superficial, because one could be said to be the agent of an involuntary act.) My hypothesis is that, in the quote, the ‘subject’ must be understood as the grammatical subject of an action sentence or any sentence containing an action verb. While in the terminology I chose here, the ‘agent’ will be not just the grammatical subject of any action sentence but will also be the subject of a sentence giving a description under which her action is intentional—​i.e., an expression of practical knowledge. Now, my thesis will be that this difference—​which, as Anscombe emphasises, has (notably but not only) to do with the difference between non-​observational knowledge and knowledge from observation or evidence—​is related to the uses of ‘I’. But before saying more, it is worth noting that this claim is less than obvious since Anscombe herself has no particular preference for examples in the first-​person singular. For instance, in the

6  See, for example, Michael Bratman, Intention, Plan and Practical Reason (Cambridge, MA: Harvard University Press, 1987), 3. 7  See I, §§1–​4 and Valérie Aucouturier, L’intention en action (Paris: Vrin, 2018), 76–​79.

364   Valérie Aucouturier quote, ‘I’ does not appear but rather ‘what he—​this man—​is doing’, and Anscombe does not start her enquiry into the question ‘Why?’ in the first person but in the third person (occasionally using, as we shall see, examples from the animal realm). That is to say, the role of the first-​person point of view in the characterisation of intentional action will appear from the perspective of a question raised from either a second-​person (‘Why are you X-​ing?’) or a third-​person (‘Why is she X-​ing?’) point of view.

16.2.2  Practical Knowledge As Intention proceeds, this non-​observational knowledge of one’s own intentional action gets clarified through the concept of practical knowledge. In short, practical knowledge is a key feature of intentional action. It is an agent’s knowledge of what she is doing—​of ‘what happens’,8 granted that what happens is an intentional action (in a nutshell: an action for which I can straightforwardly give reasons to act without proceeding to any sort of enquiry). Hence practical knowledge is knowledge of a particular object, namely my own intentional action as opposed to other things I might (knowingly9 or not) be doing and as opposed to any object of ‘contemplative’ or ‘observational’ knowledge. Indeed, practical knowledge is practical—​as opposed to theoretical—​in that it is knowledge of an action (but more specifically of one’s own intentional action). This is one feature of practical knowledge which distinguishes it from theoretical knowledge. One other, already suggested, important distinctive feature is its non-​observational character: I do not know what I am doing as a result of noticing it or of any kind of enquiry. But this characterisation is not enough to mark the specificity of practical knowledge. For there are other things I know non-​observationally, according to Anscombe: the position of my limbs (I, §8), a mental cause (e.g., that the leap and loud bark of the crocodile [§8] or the vision of something [§9] made me jump), etc. What all these cases have in common and what allows us to talk about ‘knowledge’ is that ‘there is a possibility of being right or wrong’; ‘a contrast exists between “he knows” and “he (merely) thinks he knows” ’ (I, §8); or, in the words of the quotation that opens this chapter, they ‘are directly verifiable or falsifiable’ about the person concerned. This excludes, for instance, that the concept of knowledge should apply to my feelings, emotions, or—​to some extent—​to mere expressions of intentions (see I, §§4, 8, 25), which are all cases where being mistaken or being right or wrong is not even a logical possibility10 or ‘a move in the game’, as Wittgenstein put it.11 8 ‘I do what happens. That is to say, when the description of what happens is the very thing which I should say I was doing, then there is no distinction between my doing and the thing’s happening’ (I, §29). 9  For there are things I may knowingly be doing while not intending them (e.g., using my shoe soles). Anscombe would rather call them ‘voluntary’ than ‘intentional’ (I, §§23, 49). 10  See later discussion for what is involved in the idea of a ‘logical’ possibility or impossibility. 11  ‘We distinguish in chess between good and bad moves, and we call it a mistake if we expose the queen to a bishop. But it is no mistake to promote a pawn to a king’. Ludwig Wittgenstein, The Blue and Brown Books (Oxford: Basil Blackwell, 1958), 67.

The First Person, Self-Consciousness, and Action    365 Further, what distinguishes practical knowledge from the non-​observational knowledge of other things I might be doing (or knowing) is that it is knowledge of an intentional action, which ties it to the concept of ‘practical reasoning’: ‘The mark of practical reasoning is that the thing wanted is at a distance from the immediate action, and the immediate action is calculated as the way of getting or doing or securing the thing wanted’ (I, §41). In other words, it is not just a matter of knowing what I am doing, but of knowing what further I am doing in doing something: practical knowledge implies that I can at least provide two related descriptions of what I am doing intentionally.12 It is not my purpose, though, to elaborate on the characterisation of practical knowledge as being practical (i.e., an articulation of means towards ends) here, since I would like to focus on another aspect of it, namely the fact that practical knowledge is first personal.

16.2.3  Non-​Language-​Users Interestingly enough, the importance of the agent’s first-​personal point of view is emphasised, in Intention, not only through a question addressed in the third person but also through the case of non-​language-​users, who, by definition, are not ‘I-​users’:13 Since I have defined intentional action in terms of language—​the special question ‘Why?’—​it may seem surprising that I should introduce intention-​dependent concepts with special reference to their application to animals, which have no language. Still, we certainly ascribe intentions to animals. The reason is precisely that we describe what they do in a manner perfectly characteristic of the use of intention concepts: we describe what further they are doing in doing something (the latter description being more immediate, nearer to the mere physical): the cat is stalking a bird in crouching and slinking along with its eyes fixed on the bird and its whiskers twitching. The enlarged description of what the cat is doing is not all that characterizes it as an intention (for enlarged descriptions are possibly of any event that has describable effects), but to this is added the cat’s perception of the bird, and what it does if it catches it. The two features, knowledge and enlarged description, are quite characteristic of description of intention in acting. Just as we naturally say ‘The cat thinks there is a mouse coming’, so we also naturally ask: Why is the cat crouching and slinking like that? And give the answer: It’s stalking a bird; see, its eye is fixed on it. We do this, though the cat can utter no thoughts, and cannot give

12 

Indeed Anscombe distinguishes cases of acting intentionally and knowing what you’re doing (by being in a position of giving reasons to act) from cases of proper practical knowledge (which are for their part related to practical reasoning). 13  To speak of non-​language-​users is, to some extent, more accurate than to speak of animals, for some animals (e.g., pets) may learn a few conventional signs, while some human beings are not (or not yet) language-​users. The point is to highlight the fact that being a language-​user is a key feature here. But of course, to do well, we would have to say more about what being a language-​user entails. Obviously ‘conventional meaning’ (I, §3) is an important feature for Anscombe.

366   Valérie Aucouturier expression to any knowledge of its own action, or to any intention either. (I, §47, my emphasis)

At the very beginning of Intention (§2) Anscombe already drew our attention to the fact that animals have intentions but do not (verbally) express them, mostly in order to remark that the recognition of an intention does not require that we ‘consider something internal’: an observation against already mentioned ‘introspective explanations of intention’. So non-​speaking animals (some of them) do act intentionally, and some conception of the knowledge of what they are doing is involved in this characterisation. Even more so as they can make practical mistakes (in their grasp of the circumstances of the action, for example), just like any human agent. See for instance this passage in ‘Under a Description’: [I]‌t sounds as if the agent had a thought about a description. But now let’s suppose that a bird is landing on a twig so as to peck at bird-​seed, but also that the twig is smeared with bird-​lime. The bird wanted to land on the twig all right, but it did not want to land on a twig smeared with bird-​lime. If it landed on the twig in order to peck at the bird-​seed, can’t we say it took landing on the twig to be a way of getting into a position to peck at the bird seed? . . . This way of talking does not presuppose that the bird has any thoughts about descriptions. . . . [S]omeone who says the bird’s action was intentional (or voluntary) under one description, not under the other, need not enter into that dispute at all. . . . That the bird is not a language-​user has no bearing on this. (UD, 209–​210)

So the idea of their action being intentional under a description and not under another applies as easily and straightforwardly to non-​language-​users as to language-​users. And it would be wrong to suppose that attribution of intentions to animals is merely anthropomorphic or metaphorical, that our concepts would work entirely differently when applied to non-​language-​users; and it would be wrong to suppose that one would need to be a language-​user (and still less to ‘mentally contemplate’ a description of one’s action) in order to have intentions. If not all of them (especially those involving specific human activities, e.g., telephoning), intentional verbs do apply to animals while they do not apply non-​metaphorically to non-​living beings. There are no reasons to suppose that there would be a gap in our uses of such verbs depending on the subject to which they apply. Still, there is, according to Anscombe, an anthropological difference.14 And this difference is related to expressions (of intention) understood as manifesting a linguistic capacity (‘if we will allow certain bodily movements with a conventional meaning to be included in language’ [I, §2]). Now the question is: how does language make a difference 14  This is right at least to the extent that Anscombe is interested in human action, and the case of non-​ language-​users plays a strategic role here. About the concept of human action, see Valérie Aucouturier, ‘Human Action’, in A. Haddock and R. Wiseman (eds.), The Anscombean Mind (Oxford: Routledge, 2022), 333–​354.

The First Person, Self-Consciousness, and Action    367 here? Surely not because some description occurs in the agent’s head. If the bird can indeed miscalculate the circumstances, it does not give expression to any knowledge of the logical possibility of a mistake. In other words, it does not give expression to practical knowledge insofar as practical knowledge involves the logical possibility of a mistake (and more specifically of a mistake in performance):15 the expression of practical knowledge is an expression of knowledge in that it is liable to be corrected. The bird’s behaviour is not, in this sense, the expression of any kind of practical knowledge (although it involves some knowledge of the circumstances). It does not follow from these remarks that practical knowledge is specifically human but only that its expression is. And expression is an important way of making or claiming a potential contrast between what one intends or intended to do and what one actually did (especially when it is not behaviourally observable). Now, the importance of linguistic expression relates to the point I wish to discuss in what follows. Indeed, although Anscombe wishes to silence this aspect in Intention—​most likely in order to avoid the traps of ‘psychologism’ (§2)—​it is clear that practical knowledge is a knowledge exclusively in the first person and that the expression of practical knowledge can occur only in the first person. This remark leads straightforwardly to the point that understanding this specificity implies understanding the role of the first person. However, I do not wish to treat the use of ‘I’ as a decisive criterion or as the unique criterion marking a difference between language-​and non-​language-​users. What I want to highlight is that it makes a difference to the nature of the action considered that it is done by an I-​user, i.e. (if we follow Anscombe on this point), a self-​conscious agent.

16.3  Self-​C onsciousness and the First Person 16.3.1  ‘Uses’ of ‘I’ It is therefore time to turn to the characterisation of self-​consciousness understood as ‘something manifested by the use of ‘I’, granted that ‘I’ is ‘not a name’ everyone uses ‘to speak only of himself ’ (FP, 24–​25). How can this understanding of self-​consciousness shed light on the concept of ‘knowledge without observation’ and more specifically on the knowledge of one’s own doings and of ‘practical knowledge’? Remember that ‘knowledge without observation’ is understood as a generic feature of various cases16 which are meant to be distinguished. All these cases have in common 15 See Intention, §2 (‘I write something other than I think I am writing’); §32 (‘I say to myself: “Now I press Button A”—​pressing Button B’), where ‘what you did was a mistake, because it was not in accordance with what you said’. 16  From the position of one’s limbs, involuntary actions, to voluntary and intentional actions.

368   Valérie Aucouturier that they relate to a kind of knowledge expressed in the first person. Most of them are considered in the context of the elucidation of the ‘Why?’ question and linked to answers the agent herself could give to the question. Therefore, we need to account for their first-​personal character. In order to do so, I will turn to Wittgenstein’s distinction between ‘two uses of ‘I’’ in The Blue Book.17 For various reasons, which will soon appear in what follows, I will not fully endorse Wittgenstein’s analysis of the two uses or kinds of example he addresses. However, I will show that he was right to distinguish them and should at least have distinguished a third kind, namely the one covering what I earlier called ‘Anscombianly-​ preferred examples’. I will argue that the consideration of these various kinds of examples can help us understand various modalities of self-​consciousness, which are contained not entirely in the mere use of the first-​person pronoun but rather are contained in the practice which consists in speaking in the first person, for oneself, on one’s behalf, and about oneself.

16.3.2  ‘Object’: The First Kind of Example Let us have a look at the first kind of example (of ‘uses of “I” ’) considered by Wittgenstein: Examples of the first kind of use are these: ‘My arm is broken’, ‘I have grown six inches’, ‘I have a bump on my forehead’, ‘The wind blows my hair about.’. . . The cases of the first category involve the recognition of a particular person, and there is in these cases the possibility of an error, or as I should rather put it: The possibility of an error has been provided for. . . . It is possible that, say in an accident, I should feel a pain in my arm, see a broken arm at my side, and think it is mine, when really it is my neighbour’s. And I could, looking into a mirror, mistake a bump on his forehead for one on mine.18

Although Wittgenstein legitimately identified here a specific category of examples that ought to be distinguished from another that he calls the ‘use as subject’, two difficulties should be pointed out: Wittgenstein should not have talked about ‘the recognition of a particular person’, and he should not have talked about a use of ‘I’ as ‘object’. Indeed, I could, looking into a mirror or a photograph, mistake a bump on someone else’s forehead for one on mine. This kind of mistake, Wittgenstein suggests, would involve the ‘recognition’ or rather the misrecognition ‘of a particular person’, that is, as Anscombe would say, of the ‘object that one is . . . the human animal that one is’ (FP, 34, my emphasis). However, Wittgenstein’s characterisation was, at the very least, poorly worded. Indeed, the mistake or misrecognition at stake here has nothing to do with ‘tak[ing] the wrong object to be the object [one] means by “I” ’ (FP, 30, my emphasis). The mistake is 17 Wittgenstein, Blue and Brown Books, 67.

18 Wittgenstein, Blue and Brown Books, 67, my emphasis.

The First Person, Self-Consciousness, and Action    369 rather parallel to the one illustrated by Anscombe’s example: ‘The bishop may take the lady’s knee for his, but could he take the lady herself to be himself?’ (FP, 30). In other words, I can take someone else’s bump to be mine, but that would not count as a case of misrecognition, that is, of taking myself to be someone else. I would not thereby be talking about someone else than me in saying ‘I have a bump’;19 I would merely mistakenly think that someone else’s bump is mine. In fact, Wittgenstein should not have talked about ‘the recognition of a particular person’ in this first category of examples, since no recognition of a particular person takes place here. And he should not have talked about the use of ‘I’ as ‘object’, since ‘I’ does not play the role of pointing to any kind of identifiable object in the considered examples. Rather, to address the specificity of these examples, Wittgenstein should have drawn our attention to the distinction between two kinds of mistakes: one which is logically possible and the other which is logically excluded. The difference between these two kinds of mistake will show that self-​ascription of any corrigible property does not work the same way as third-​personal ascription, although it is at least partly based on the same kind of activity. Indeed, when I say Sacha has a bump on his forehead, I identify a person to whom I ascribe such a property. And I can make two different mistakes here: I can wrongly think that the person I am talking about (when I say ‘Sacha’) who has a bump on his forehead is Sacha, or I can wrongly think that Sacha has a bump on his forehead. In other words, I can misidentify or misrecognise the person I am talking about (first kind of mistake) or I can misattribute some property to someone (second kind of mistake). And of course, I can make both mistakes simultaneously: I can wrongly think that someone I mistakenly take to be Sacha has a bump on his forehead. But when it comes to me and my alleged bump, although my mistake may rest on the fact that, for instance, I took someone else in the mirror to be me, still my mistake lies in the object or in the ascription itself; I did not take the person I am talking about (namely me) for another. That is to say, I did not mistakenly refer to someone else in saying ‘I’. I simply said something wrong, and I said this about me (not about someone else). In other words, in spite of their syntactic (or surface grammar) resemblance with predicative sentences in the third person, I-​sentences and self-​ascriptions are not predicative attributions where a name or a proper name is ‘combined with other signs . . . and the result of this combination is an independent semantic unity which is liable to be true or false’.20 I-​sentences ‘do not involve the connection of what is understood by a predicate with a distinctly conceived subject’ and are therefore, in Anscombe’s sense, ‘subjectless’ (FP, 36). I-​sentences are ‘subjectless’ in the sense that they do not provide any identification principle since the very possibility of misidentification (or, for that matter, 19 

Actually, as we shall see, this also does not entail that I necessarily pick out the right object in saying ‘I’: ‘The suggestion of getting the object right collapses into absurdity when we work it out and try to describe how getting hold of the wrong object may be excluded’ (FP, 31). 20 Descombes, Le parler de soi, 110. Similarly, ‘no one thinks that “it is raining” contains a referring expression, “it” ’ (FP: 30).

370   Valérie Aucouturier successful identification) has not been provided for. To refuse this claim would be to give in to the temptation of denying the specificity of the first person and to treat I-​sentences as mere parallels to third-​person or predicative sentences.21 The reason why Wittgenstein was wrong in treating these examples as uses of ‘I’ ‘as object’, and as cases of misrecognition of an object which ‘I’ is supposed to stand for (‘a particular person’), is that it tends to suggest that using ‘I’’ would involve some kind of identification. As if I needed to rightly identify the person I am talking about in saying ‘I’ or to ‘look for a subject’ (FP, 36). But the truth is that I-​sentences do not identify anyone or anything, and the recognition at stake in these examples matters only for the second kind of mistake: misascription is not misidentification. To take someone else to be myself is not yet to take myself to be someone else.22 It does not consist in taking ‘I’ for another. I cannot by mistake misuse ‘I’ to mean someone else than me, even though I can say a whole bunch of wrong things about me, including that I am Napoleon or the Queen of England or a great philosopher. From a charitable reading of Wittgenstein’s ‘recognition of a particular person’, one could say that it points to the fact that these categories of I-​sentences are made on the basis of the observation of someone, who should normally be me, in the sense of the ‘object’ or the ‘human animal’ that I am. Therefore, they can be mistaken without questioning the fact that, as first-​person sentences, they are indeed about me. And we understand why it was misleading to talk about a ‘use of ‘I’ as object’ where ‘I’ is not the object of the possible mistake or misrecognition, but rather what is said of me, the object of my statement. If I-​sentences are ‘subjectless’, they are not necessarily objectless in the sense that they may say something (possibly wrong) about me.

16.3.3  ‘Subject’: The Second Category of Examples But in order to say more about the logical impossibility of the misidentification of the speaker who says ‘I’, we need to turn to the second category of examples distinguished by Wittgenstein from the ‘object’ use:

21  The logician’s (and our) temptation to treat first-​personal sentences as predicative sentences stems from a desire to treat them in isolation from a practice, a context in which they are uttered, which is dialogic. Taken in isolation from any discursive circumstances, I-​sentences do not say anything of anyone since they do not say, by themselves, whom they are talking about. See Descombes, Le parler de soi, 75–​77. Of course the objection is not directed at Wittgenstein himself, who, presumably, treats only a certain kind of first-​person sentence as parallel to third-​person ones and who eventually gave up the distinction between uses of ‘I’ for this very reason. 22  Of course, it may happen that I take myself to be someone else in a case where, for instance, I see myself in the mirror (or in a photograph) and think it is my sister, and say ‘You look sad’ or ‘My sister looks sad’. But in that case, I would not be using the first person anyway. (Thanks to Roger Teichmann for attracting my attention to this case.)

The First Person, Self-Consciousness, and Action    371 Examples of the second kind are: ‘I see so-​and-​so’, ‘I hear so-​and-​so’, ‘I try to lift my arm’, ‘I think it will rain’, ‘I have toothache’. . . . [T]‌here is no question of recognising a person when I say I have a toothache. To ask ‘are you sure it’s you who have pain?’ would be nonsensical. Now when in this case no error is possible, it is because the move which we might be inclined to think of as an error, a ‘bad move’, is no move of the game at all. . . . To say, ‘I have pain’ is no more a statement about a particular person than moaning is.23

Interestingly, these examples correspond to the categories of what Anscombe calls ‘Cartesianly-​preferred examples’ (FP, 35). They attracted Wittgenstein’s attention because they exhibit a genuine specificity of ‘I’ sentences by pointing to the heterogeneous class of what are sometimes called ‘psychological concepts’ or ‘verbs’, as illustrated in the analysis of Moore’s paradox.24 In a nutshell, these examples exhibit what is sometimes called the first-​/t​ hird-​person asymmetry of the indicative present, the neglect of which happens to be at the source of many philosophical traps or temptations (mainly solipsism): I do not say that I am in pain or that I believe it will rain on the basis of some corrigible observation any more than I can moan by mistake, while I can of course be mistaken when I say someone else is in pain or believes it will rain. However, Wittgenstein’s way of addressing the asymmetry in The Blue Book is, though not wrong, at least incomplete. Unlike with the use as object, he says, ‘there is no question of recognising a person’, here and ‘no error is possible’ in the sense that the possibility of a mistake is not merely empirically excluded (as if it happened to be that I never misrecognise myself), but it is logically excluded (as it is logically excluded in chess ‘to promote a pawn to a king’). But, from what we just said, misrecognition should also have been excluded from the first range of cases ‘as object’. So the essential (or grammatical) difference between the object-​use and the subject-​use must lie somewhere else. It lies in the second kind of mistake: the one I called ‘misascription’. Indeed, in these examples, it is logically excluded that I pick the wrong person in saying ‘I’ (misrecognition) as it is for all ‘normal’ uses of ‘I’. But, while it would be possible for me to take someone else’s broken arm or the lady’s knee to be mine or to wrongly think I am V.A., I could not wrongly think I am in pain.25 Misascription is not a move in the game at all. But here again, Wittgenstein’s wording is misleading. He specifically insists on the fact that I cannot doubt I am the one in pain, while insisting on the fact that I do not identify

23 Wittgenstein, Blue and Brown Books, 67, my emphasis. 24 

See Wittgenstein, Philosophical Investigations, II, x; Mary McGinn, ‘Wittgenstein and Moore’s Paradox’, in R. Heinrich et al. (eds.), Image and Imagining in Philosophy, Science and the Art, vol. 1 (Frankfurt: Ontos Verlag, 2011), 59–​72; Descombes, Le parler de soi, III; Jocelyn Benoist, Les limites de l’intentionalité (Paris: Vrin, 2005), ch. 8. 25  Of course, there is a sense in which I might wrongly think I see or hear something, if there is nothing to be seen or heard. But if I maintained that I see or hear whatever I claim to be seeing or hearing, no one could deny the fact that I, at least, have the sensation of seeing or hearing something. This is the reason why Wittgenstein places these examples in the ‘use as subject’. See also Anscombe, IS.

372   Valérie Aucouturier anyone in saying I am in pain. It is ambiguous because, as we saw, neither was I pointing or identifying anyone when I said I had a bump on my forehead any more than, ‘when someone points to the sun with his hand, he is pointing both to the sun and himself because it is he who points’.26 Of course, my statement was based on the observation of a person, a human being (namely me). But the difference lies in the fact that in the case of pain (and of psychological concepts in general) I do not infer my statement from any (self-​)observation. Which is why saying where I am wounded and where I feel pain have two distinct grammars. When I say I am in pain, I am not telling you this particular body is in pain.27 My pain is not, properly speaking, an object of knowledge; I have it. These examples play a crucial role in our understanding of subjectivity and of the first-​/​third-​person asymmetry that distinguishes first-​personal discourse and authority. There is a whole range of cases in which I cannot misascribe some attribute or be wrong about myself, because in these cases I am not an object of knowledge for myself. These examples, for that matter, make more acute the necessity to renounce the subject/​object dichotomy in order to understand the grammar of first-​person sentences. There is a sense in which the typology of mistakes (misidentification and misascription) that I characterised earlier is plainly inappropriate here, and its function is purely didactic in that it helps to bring out the contrast between ordinary predicative statements (in the third person) and first-​person sentences. In the ‘subject’ examples, to speak for oneself manifests the subject’s authority regarding what she says: I could be insincere, but I could not be plainly wrong. Not that I would always rightly identify what I feel or think or believe, etc., but because recognition is here irrelevant. Just to avoid a rather frequent misunderstanding, note that it is important that the specificity of this kind of example depends on a grammatical feature of first-​person sentences and not on an alleged empirical fact (e.g., that I would never, in fact, be mistaken about what I believe or whether I am in pain). The point is grammatical or logical in the sense that, should someone (including myself) question whether I am really in pain or whether I really do believe such-​and-​such or intend to do such-​and-​such, etc., ultimately I will always have the last word. For that matter, I am the only one in a position to settle the potential (though usually inappropriate) question. This is the grammatical point.

16.3.4  Beyond the Subject and the Object In short, Wittgenstein was wrong in distinguishing the ‘object’ and ‘subject’ kinds of example in terms of the recognition of a particular person. In doing so, he kept thinking the difference between the two cases through the old subject/​object distinction by suggesting 26 Wittgenstein, Blue and Brown Books, 67.

27  Wittgenstein even imagines a case where I would feel pain in another person’s body (Blue and Brown Books, 49) to insist on the contingency of the fact that my pain is in my body and to insist on the fact that my pain is mine as soon as I am the one who expresses it.

The First Person, Self-Consciousness, and Action    373 that, in the one case, ‘I’ is a subject and person misidentification is impossible, and in the other case, ‘I’ is an object (‘for herself ’) and person misidentification is possible. Wittgenstein thus misled his readers into understanding the contrast between uses of ‘I’ as object and as subject in terms of the possibility or impossibility of (mis)recognition of a particular person, which led directly to the ‘guaranteed reference thesis’ or the ‘immunity to error through misidentification’ claim.28 According to this claim,29 the difference between the two cases could be thought in terms of the possibility of misidentification in the ‘object’ case and the impossibility of such an error or the immunity to such an error in the ‘subject’ case. The first use would allow the possibility of misrecognising the person I am talking about in the first person, while the other would exclude it. But we saw this characterisation is at best misplaced, because in fact, as Anscombe argues in FP, the very concepts of identification or recognition do not apply to first-​person sentences, and the kind of mistake which is allowed in the first case and impossible in the second cannot be thought in terms of immunity or non-​immunity to error through misidentification. Indeed, in both cases, the question ‘Who?’ which would aim at identifying the speaker (or writer) of an I-​sentence or an I-​thought is no move in the game at all: If you are a speaker who says ‘I’, you do not find out what is saying ‘I’. You do not for example look to see what apparatus the noise comes out of and assume that that is the sayer. If that were in question, you could doubt whether anything was saying ‘I’. (FP, 29)

To say that the question ‘Who’? is excluded in this sense from I-​sentences does not entail that we need not check who is speaking because we already know it for sure. It rather means that the very question is logically excluded from the position of the speaker (not of the reader or hearer, of course!). And here, ‘logically excluded’ does not merely mean that it would be ‘highly improbable’ or impossible, but that it is, as a ‘grammatical fact’, ‘literally unconceivable’30 for me to wonder who I am talking about when I say ‘I’. Speaking of a guaranteed reference or of an immunity to error through misidentification suggests that we always pick the right object when we say ‘I’. But it is not that I always identify the right person when I say ‘I’; it is not that the ‘Who’? question always receives the right answer. The question simply is logically excluded. No identification takes place, which leaves no room for misidentification either. The question of knowing whom I ascribe such an attribute to when I speak in the first person is simply void: Getting hold of the wrong object is excluded, and that makes us think that getting hold of the right object is guaranteed. But the reason is that there is no getting hold of an object at all. (FP, 3)

28  Cf. Sidney Shoemaker, ‘Self-​Reference and Self-​Awareness’, Journal of Philosophy 65, no. 19 (October 3, 1968): 555–​567. 29  Which is not, as such, endorsed by Wittgenstein. 30 Descombes, Le parler de soi, 135.

374   Valérie Aucouturier As we saw, the difference between the two cases does not depend on any recognition of a person at all, but on the fact that in the ‘subject’ use what is said about me is non-​ contingently ascribed to me, while in the ‘object’ use what is said about me is contingently ascribed to me. In other words, in the second case, whether the thing is said by me or someone else does not really make a difference: the activity I deploy to say what I say about me is akin to the one I deploy to say something about someone else.31 The difference between the two cases rests in what is said of me and in the fact that, in the one case (object), I can say wrong things about me while, in the other (subject), I cannot be wrong (but possibly lying). The first case contains the possibility of a mistake (notably based on misobservation), while the second excludes it. I can wrongly think my arm is broken when it is someone else’s, while I cannot mistakenly think I am in pain. Hence the difference does not really concern the function of ‘I’ in itself but some specificities of a certain kind of example of first-​person statements. And this difference is not fully contained in the use of ‘I’ but rather in what is said of me or by me in the first person, that is, in the spoken speech-​act. However, Wittgenstein was right to highlight the specificity of first-​ personal statements (or rather speech-​acts in general) in what he clumsily calls ‘the use as subject’, since they reveal some peculiarities of first-​personal discourse, namely the first-​/​third-​ person asymmetry.32

16.3.5  ‘Agent or Patient’: The Third Category of Examples It is now time to turn to the third category of examples considered by Anscombe (the ones I called ‘Anscombianly-​preferred examples’), that is, examples of ‘actions, postures, movements and intended actions’ such as ‘I am standing’, ‘I jumped’, etc. (FP, 35). The reason why they are of interest is that they are ‘unmediated, non-​observational, and also descriptions . . . which are directly verifiable or falsifiable about the person’

31 

Vincent Descombes, Exercices d’humanité (Paris: Les Petits Platons, 2013), 116: ‘Est-​ce que je me connais par une activité transitive? Oui, si l’activité que je déploie pour me connaître est du même genre que celle que je déploie pour connaître quelqu’un d’autre. . . . Dès lors, la biographie que je pourrais produire par une activité purement réfléchie devrait, elle aussi, respecter cette condition d’une coïncidence contingente entre le sujet et l’objet. Elle n’aura pas le caractère d’une auto-​biographie, elle aurait pu être écrite par quelqu’un d’autre.’ 32  Which constitutes the grammatical ground of many mistakes in philosophy (mainly solipsism) and reveals the grammatical truth in which these mistakes take root. First-​person authority as manifested by this special asymmetry of so-​called psychological concepts also has important consequences regarding the possibility of speaking of oneself in the past in a ‘psychological’ mode, as a self-​examination (cf. Descombes, Le parler de soi). For though this kind of self-​examination is corrigible (see the analysis of Moore’s paradox in Wittgenstein), we would not be able to do it if first-​person statements did not exhibit this specificity.

The First Person, Self-Consciousness, and Action    375 (FP, 35). In other words, they combine some features of both the ‘object’ and ‘subject’ examples. Indeed, unlike the first category of examples and like the second, they are not based on any kind of ‘(self-​)observation’: ‘observation does not show me which body is the one. Nothing shows me that’ (FP, 34). I do not usually notice I am standing as I notice I have a bump on my forehead. (Of course, in some circumstances, I may simply notice I am standing, for instance, in a photograph. This would be an example of the ‘object’ kind.) On the other hand, unlike the ‘subject’ kind of example and like the ‘object’ one, I can be wrong, error is possible: though my statement or thought that I am standing is not based on any evidence or observation, it can be verified or falsified on the basis of observation: ‘Anyone, including myself, can look and see whether that person is standing’ (FP, 35). Which is of course not the case when I am in pain. Just as in the two other cases, it is not a matter of recognising or considering who is (say) standing, but of what the speaker is doing. But, interestingly, though corrigible, the ‘What?’ question is not answered on the basis of self-​observation. The interest of these examples is twofold: they exhibit the specificity of ‘I-​thoughts’ in relation to the ‘subject’ (or ‘Cartesianly-​preferred’) examples (namely self-​consciousness—​our next topic), while not giving in to the solipsist temptation, since they are ‘directly verifiable or falsifiable’: These I-​thoughts are examples of reflective consciousness of states, actions, motions, etc., not of an object I mean by ‘I’, but of this body. These I-​thoughts (allow me to pause and think some!) . . . are unmediated conceptions (knowledge or belief, true or false) of states, motions, etc. of this object here, about which I can find out (if I don’t know it) that it is E.A. About which I did learn that it is a human being. (FP, 34)

Contrary to the ‘object use’, these ‘unmediated conceptions’, though they are about E.A. or about ‘this body’ or this ‘human being’ (and can be corrected by looking at E.A. or at ‘this body’ or what she is doing) are not based on the observation of E.A. or of this body. In conclusion, I may speak in the first person to talk about my body, the object or the human animal that I am; I may speak in the first person to talk about my state of mind, my feelings, my impressions, my sensations, etc., to make a promise, to bet, to make a commitment, etc.; I may speak in the first person to describe or state what I am or was doing.33 All of these cases exhibit, in a more or less explicit way, what Anscombe 33  Actually, a real investigation of the first person would have to distinguish, within the three uses or kinds of example I described here, various cases and explore the differences it makes that they are expressed in the first person. For instance, I do not say anything here about committing sentences (or what Austin at some point calls ‘performative utterances’) which can be committing only insofar as I utter them in my own name: ‘If I utter the word “He bets”, I only state that he utters (or rather has uttered) the words “I bet”: I do not perform his act of betting, which only he can perform. . . . Similarly, an anxious parent when his child has been asked to do something may say “he promises, don’t you Willy?” but little Willy must still himself say “I promise” if he is really to have promised.’ John L. Austin, How to Do Things with Words (Oxford: Oxford University Press, 1975), 63.

376   Valérie Aucouturier calls ‘self-​consciousness’ in that none of these cases requires that I look for the subject I am talking about in saying ‘I’. Now, there are differences in the degrees to which self-​ consciousness is manifested in these various cases, and what I would like to explore next and finally is this and the consequences it has for our understanding of the relation between self-​consciousness and intentional action.

16.4  Self-​C onsciousness and Action 16.4.1  Self-​Consciousness and Self-​Knowledge Of course, none of what I have said excludes my taking myself as an object of consideration or knowledge. Actually, such an attitude towards oneself would amount to what Anscombe calls ‘self-​knowledge’: Self-​knowledge is knowledge of the object that one is, of the human animal that one is. (FP, 34)

In other words, the claim that ‘I-​sentences’ are ‘subjectless’, i.e., that they do not consist in identifying a subject, does not forbid that we talk about ourselves and that we do so in the first person. But it points to the difference between speaking about oneself in the first person and speaking about oneself (or someone else) in the third (or second) person.34 I could indeed refer to myself, for instance, by talking about me in the third person (FP, 27), using my own name. In this case, I could indeed mistakenly speak of myself (without knowing it) as in Anscombe’s example: ‘When John Smith spoke of John Horatio Auberon Smith (named in a will perhaps) he was speaking of himself, but he did not know this’. If so, then ‘speaking of ’ or ‘referring to’ oneself is compatible with not knowing that the object one speaks of is oneself. (FP, 27)

Such a case, while being an example of speaking of oneself, does not display what Anscombe calls ‘self-​consciousness’ because ‘self-​consciousness’ is manifested in the act of speaking of oneself in the first person. To that extent, ‘self-​knowledge’ (or the ‘object’ examples) displays self-​consciousness and should be distinguished from mere ‘self-​ reference’. Indeed, if I say something about myself using my own name, I could both fail to refer to myself (just like John Smith does) and say wrong things about me (or whomever I am talking—​or failing to talk—​about). But, as we saw, if I say something about me

34 

See Anthony Kenny, The Metaphysics of Mind (Oxford: Oxford University Press, 1992), 88; Descombes, Le parler de soi, 18.

The First Person, Self-Consciousness, and Action    377 in the first person, I can say something wrong but cannot, in speaking in the first person, fail to speak on my own behalf. I can only exhibit self-​knowledge or self-​misknowledge. Following a distinction suggested by Descombes, we can state the difference between speaking of oneself in the third person (without self-​consciousness and/​or self-​ knowledge) and speaking of oneself in the first person by distinguishing, on the one hand, ‘self-​reference’ or the ‘designation of oneself by the speaker’ and, on the other, ‘self-​ consciousness’ or ‘self-​designation’. In the first case (self-​reference), to designate oneself (e.g., by naming oneself) is ‘a reflective act and so a transitive act’. While, in the second case, ‘self-​designation’ is ‘not a transitive act and does not refer to anyone identified as that person in particular among other persons’.35 In the first case, I could say someone (in the mirror, say, or in a photograph) has a bump on her forehead and not know that it is me. In the second case, I could not do that, though I could mistakenly think I have a bump on my forehead, in which case my belief about myself would be erroneous. I would exhibit a lack of self-​knowledge but not a lack of self-​consciousness. Which means that the kind of judgement that expresses self-​knowledge is a kind of judgement someone else could make about me, a kind of judgement on which we could agree or disagree, even if it concerns my own state of mind or character. Indeed, Anscombe continues the earlier quote as follows: Self-​knowledge is knowledge of the object that one is, of the human animal that one is. ‘Introspection’ is but one contributory method. It is a rather doubtful one, as it may consist rather in the elaboration of a self-​image than in noting facts about oneself. (FP, 34)

So introspection is not excluded from her account as a kind of self-​observation, but it ought not to be conflated with ‘subject’ examples. What it ‘may consist’ in is ‘rather the elaboration of a self-​image’, a way of narrating oneself, than a description based on self-​observation. Now, what Wittgenstein saw is that, in ‘Cartesianly-​preferred examples’, not only could I not lack self-​consciousness, but what I say about me cannot be mistaken as self-​knowledge could be (in which case, of course, it would not be knowledge at all). This is why we should be reluctant to talk about self-​knowledge regarding Cartesianly-​ preferred examples.36 Indeed, while I could wonder whether a certain bump is mine, 35  ‘Person’ being here understood not as a ‘Lockean self ’ but as a ‘flesh and blood human being’. Vincent Descombes and Charles Larmore, Dernières Nouvelles du Moi (Paris: PUF, 2009), 151–​152. 36  Something specific ought to be said about first-​personal thoughts and utterances regarding my past ‘psychological’ attitudes (beliefs, intentions, thoughts, etc.) or what Descombes calls l’examen de conscience (self-​examination). To some extent they are a variety of self-​knowledge, since I can say I used to believe or think something I no longer do (‘I thought it was raining, but it was not’), while I cannot do so in the present tense (cf. Descombes, Le parler de soi, 328; Wittgenstein, Philosophical Investigations, II, x). But to some extent they are closer to ‘Anscombianly-​preferred examples’ in that I possess a kind of authority concerning my past feelings, thoughts, beliefs, intentions, etc. that I do not possess concerning, for instance, my (former) physical appearance. See Richard Moran, Authority and Alienation (Princeton, NJ: Princeton University Press, 2001).

378   Valérie Aucouturier I could not wonder whether a certain pain (or feeling, or sensation, or intention, etc.) is mine. In this distinction, we see a good conceptual reason to distinguish between self-​ consciousness as it is exhibited in uses of ‘I’ in general and self-​knowledge as it is exhibited in a certain kind of example of first-​personal utterances or thoughts37 concerning ‘the object that one is’, ‘the human animal that one is’.

16.4.2  ‘Unmediated Agent-​or-​Patient Conceptions of Actions, Happenings and States’ Now, the reason why Anscombe prefers examples of ‘unmediated agent-​or-​patient conceptions of actions, happenings and states’ (FP, 36) to display the grammatical features of self-​consciousness is that they are neither entirely behavioural (like the object use, which would be the one preferred by the logician) nor entirely authoritative (like the Cartesianly-​preferred examples), but they are a bit of both. They are counterexamples both to the Cartesian tendency to emphasise the ‘private’ character of self-​consciousness (as if only I could tell whether I am self-​conscious) and to the logician’s tendency to look for cases where first-​person sentences state something about the speaker which can be verified. In fact, self-​consciousness is not fully contained or expressed in one of these uses: it is consciousness neither of a private Ego nor of the object (the human animal) that I am, as a preference for either kind of example would tend to display. This is the reason to prefer an intermediary kind of example, where self-​consciousness, and more obviously the lack thereof, can be displayed. This is why Anscombe’s use of William James’s pathological example of Baldy’s lack of self-​consciousness is relevant: We were driving . . . in a wagonette; the door flew open and X, alias ‘Baldy’, fell out on the road. We pulled up at once, and then he said ‘Did anyone fall out?’ or ‘Who fell out?’—​I don’t exactly remember the words. When told that Baldy fell out he said ‘Did Baldy fall out? Poor Baldy!’38

If we were to consider a Cartesianly-​preferred example here, we would push at the limits of meaning rather than attend the manifestation of a lack of self-​consciousness. Indeed, if Baldy had said ‘Someone feels pain, but whose pain is it?’, that would sound extremely strange and we would have a hard time trying to make sense of what he says because we would be unable to suppose that he is talking about some pain that he feels, since, to paraphrase Wittgenstein, ‘to ask “are you sure that it’s you who have pains?” would be nonsensical’.39 On the other hand, if we favour a more ‘behaviourist’ example, if Baldy had said ‘Did anyone break his arm?’ or ‘Who broke his arm?’ when he had broken his 37  To that extent, as I said, self-​knowledge implies self-​consciousness in that it is not knowledge about anyone (perhaps happening to be me), but about me. 38  William James, Principles of Psychology, vol. 2 (London: MacMillan, 1901), 273n, quoted in FP, 36. 39  Wittgenstein 1959: 67.

The First Person, Self-Consciousness, and Action    379 own arm, that would not sound particularly strange since, as we saw, there is an element of contingency or a possibility of error in self-​attribution. But here, we have the example of something happening to Baldy, of which Baldy is conscious, but which he fails to recognise as happening to him. In her comment on the passage, Anscombe insists that this example sheds light on the difference between being merely conscious and being self-​conscious. The strangeness of the case rests on the fact that Baldy has some consciousness of a happening but fails to recognise it is happening to him. He speaks as if the fall were someone else’s. He is treating himself as a third person. But the strangeness does not rest on the fact that he needs to observe what happens, since the case suggests that he is conscious of the happening (the falling). The strangeness rests on the fact that in spite of his consciousness of what happens, he needs to enquire whom it is happening to; he is ‘looking for a subject’ where he should not have to. Of course, the sense in which he should not have to look for a subject is the same here as the sense in which one should not have to look for a subject of one’s own pain or one’s self-​ knowledge: it is logically excluded. The strangeness of the case—​its pathological structure—​ rests on the fact that we normally have ‘unmediated agent-​or-​patient-​conceptions of [our own] actions, happenings and states’; that is, we do not need to enquire who they are happening to, whose actions they are, whose states they are. Another hint regarding the reasons why Anscombe prefers this kind of example is very likely that the logical structure of action sentences should not be treated as predicative. Of course, it is always possible (and not forbidden) to analyse an action sentence as predicating something of a subject, such that in a sentence like ‘The cat is chasing a mouse’, we would treat ‘is chasing a mouse’ as a predicate of ‘The cat’ (or ‘chasing’ as a relational predicate of ‘the cat’ and ‘the mouse’). But in doing so, we would simply miss the specificity of the non-​predicative structure of action sentences, which rather relate one (or more) agent (and/​or patient) to a certain happening or action (in given circumstances). Indeed, action sentences state or describe what happens or happened, and then the question may arise, in the active form, ‘Who did, who is responsible, who is the cause of what happens?’, and in the passive form, ‘To whom did it happen?’ This is what Descombes calls the ‘narrative [as opposed to predicative] structure’ of action sentences.40 Action sentences provide what Anscombe calls a description of ‘what happens’.41 Depending on the description considered, one can say that what happens was intentional, unintentional, voluntary, involuntary, etc.42 For instance, if ‘I was not aware’ of what happened or that I was the cause of it (I, §6) or if I knew what happened only on the ground of observation (I, §8), then what happened was not intentional. I can, for instance, saw a plank not knowing it is Smith’s. Then my action of sawing Smith’s plank was not intentional (under this description), though my action of sawing a plank was.43 Or I may betray a secret without knowing it was a 40 Descombes, Le parler de soi, 116. 41 

I, §29.

43 

I, §6.

42 Descombes, Le parler de soi, 117; Aristotle, Nicomachean Ethics, 1111a1; I, §6.

380   Valérie Aucouturier secret; then, though I intentionally spread some information, I did not intentionally betray a secret.44 The point is that my ignorance or knowledge of some circumstances of what happens may change my relation to what happens (notably in terms of responsibility). I may be an agent of what happens (intentional or not, voluntary or not). I may, on the other hand, be subjected to or undergo (knowingly or not) what happens, namely be a patient. (Or again, what happens may be indifferent to me; I may have nothing to do with it.) In any case, what happens is not something which is predicated (in the sense of the logical structure of predicative sentences) of me (or anyone else, for that matter). This is yet another reason for preferring the ‘agent-​or-​ patient’ kind of examples of I-​utterances to the ‘object’ or ‘subject’ ones, which can too easily be analysed as exhibiting a kind of predicative structure (though we saw they do not either). To say but one last thing about Baldy, based on these last remarks: what is surprising in Baldy’s case is precisely that he is aware of something happening to him, but unaware that it is to him that it is happening. Here is the abnormality of the case, revealing his strange lack of self-​consciousness. Not that we are necessarily aware of things happening to us (or even, as we saw, of things we are doing), but when we are, we normally do not need to ask or wonder to whom these things are happening, and this is the mark of self-​consciousness in action and (so to speak) passion.

16.5  Conclusion: Back to Practical Knowledge I have not in this essay said everything there is to say about the relation between speaking in the first person and being a self-​conscious agent, as exemplified in a variety of examples. But, by way of a conclusion, I would like to make a few remarks about the consequences these considerations may have on our understanding of what Anscombe calls ‘practical knowledge’ (understood, remember, as the knowledge I have of my own intentional actions). First, being self-​conscious is an essential feature of practical knowledge: By the knowledge a man has of his intentional actions I mean the knowledge that one denies having if when asked e.g. ‘Why are you ringing that bell?’ one replies ‘Good heavens! I didn’t know I was ringing it!’45

I cannot fail to know I am the agent of my own intentional action, or rather if the question ‘Who does it?’ is relevant to me, what happens cannot be my own intentional 44 Descombes, Le parler de soi, 117; Aristotle, Nicomachean Ethics, 1111a10. 45 

I, §28.

The First Person, Self-Consciousness, and Action    381 action.46 But, of course, self-​consciousness is not a sufficient characterisation of practical knowledge, since it concerns, more broadly, speaking of oneself. However, in combining features from the ‘subject’ examples (the unmediated and non-​observational aspect) and from the ‘object’ examples (the possibility of mistake), the knowledge of one’s own intentional actions, postures, etc. reveals both its ‘psychological’ aspect—​it shares with Cartesianly preferred examples the kind of authority derived from unmediatedness—​and its ‘predictive’47 or epistemic dimension of knowledge of ‘what happens’. This also shows that practical knowledge ought not to be assimilated to ‘self-​ knowledge’ (the ‘object’ examples), which requires a form of self-​observation and thus of contingency in one’s ‘relation’ to oneself. Indeed, the objects of practical knowledge are my own intentional actions, while the objects of self-​knowledge are my proceedings, my qualms, my characteristics as I observe them. This double kinship with the ‘subject’ and ‘object’ examples sheds light on the ambiguity of practical knowledge which is partly at the root of misunderstandings amongst readers and commentators who are tempted either to make Anscombe claim what she does not (or sustain a paradoxical thesis), namely that the authority of practical knowledge stems from a kind of incorrigibility, or to see her as missing the first-​person authority in action by giving the final word to observation, as if practical knowledge existed only in virtue of a certain verification that what I intend does indeed match what I do. Both interpretations are simply incompatible with Anscombe’s views. The first interpretation, by which practical knowledge is treated as incorrigible, is incompatible with the idea that knowledge should in principle be liable to mistake and thus corrigible. If practical knowledge was the mere ‘knowledge’ of my intentions and could not be questioned by the fact that I do not do what I think I am doing, it would be no knowledge at all. The second temptation, to claim that practical knowledge can be acknowledged only in the coincidence between my intentions and my actions, is incompatible with the non-​observational dimension of practical knowledge and the exclusion of a detour by self-​observation. However, though wrong, this double temptation is grounded on the ‘intermediary’ character of these ‘agent-​or-​patient’ examples. And this intermediary character can be illustrated by another distinction Anscombe borrows from the scholastics, between an ‘intentional’ (or ‘formal’) and a ‘material’ point of view on the action: You shoot, as you think, at a stag; what you thought was a stag turns out to be a man. In one sense—​the formal—​your object was a stag; in the other—​the material—​it was a man.48

46 Descombes, Le parler de soi, 121; Aristotle, Nicomachean Ethics, 1111a10. 47 

See I, §3. TKEA, 5. See also a formulation of the distinction in the third person in IS, 9: ‘A man aims at a stag; but the thing he took for a stag was his father, and he shoots his father’. 48 

382   Valérie Aucouturier It may happen, as is the case here, that there is a discrepancy between what you intend and what you actually do. A blind person might sign a contract believing she is signing a petition. It would make sense, then, to distinguish between the intentional point of view by which what you were doing was A—​’aiming at a stag’ or signing a petition—​and the material point of view of what you actually did—​shooting a man or signing a contract. In such cases the agent’s practical knowledge would be flawed, wrong. However, and this is the difficult feature of practical knowledge, there still is a sense in which what the agent was doing was, to his knowledge, aiming at a stag or signing a petition. And such descriptions are relevant to what happened. On the other hand, if everything goes right (and this is the central case of intentional action), the intentional and material points of view (descriptions) are equivalent in the sense that it makes no sense to distinguish them. Then ‘I do what happens’, and have practical knowledge of what happens. Of course, the ‘material’ description of what happens then is not ‘any description of what is going on, with [the agent] as subject, which is in fact true’ (I, §23). It is only insofar as the material description is somehow dependent on the intentional description that we can speak of practical knowledge, i.e., when these two descriptions are in fact one and the same, coming, on the one hand, from the agent’s point of view and, on the other, from an observer. This grammatical ambivalence of I-​thoughts about intentional actions makes more visible both the authority of practical knowledge in its relation to self-​consciousness and its epistemological status as knowledge proper. And in doing justice to this grammatical ambivalence we can avoid taking practical knowledge to be a species of self-​knowledge (understood as a certain observational and thus contingent relation to oneself and one’s own actions).

Chapter 17

Ansc omb e a nd Self-​C onsci ou sne s s Adrian Haddock

1. A central insight of G. E. M. Anscombe’s essay ‘The First Person’1 is that self-​ consciousness is not consciousness of oneself.2 2. Anscombe remarks that, ‘when we speak of self-​consciousness [we] mean something manifested by the use of “I” ’ (25). She goes on to remark that ‘‘I” is neither a name, nor another kind of expression whose logical role is to make a reference, at all’ (32). As she sees it, the idea that self-​consciousness is consciousness of oneself, and the idea that ‘I’ is an expression whose logical role is to make a reference, are two aspects of the same idea. 3. To bring out the difference between self-​consciousness and consciousness of oneself, Anscombe invites us to imagine a society of persons, each member of which uses an expression that refers, on his lips, to himself.3 She gives this expression as ‘A’. Consider an arbitrary member of this society, and call him ‘Smith’. On Smith’s lips, ‘A’ refers to himself, and as such manifests consciousness of himself. But it does not manifest self-​ consciousness, on account of how its reference is fixed. To bring this out, suppose that Smith fixes its reference by focusing on someone given to his senses, and saying: (1) ‘A’ refers to this person.

In Smith’s utterance of (1), ‘this person’ refers to a certain person through their being given to Smith’s senses in a certain way at the time of his utterance. Anscombe notes that ‘for each person there is one person of whom he has characteristically limited and also 1 

FP, in CP2; all references in parentheses are to this essay. Sebastian Rödl’s wonderful book Self-​Consciousness (Cambridge, MA: Harvard University Press, 2007) begins as follows: ‘She who is self-​conscious has the power to think of and refer to herself ’ (1); the ensuing discussion rests on the assumption that self-​consciousness is a power to think of and refer to oneself. Anscombe denies this assumption. 3  In this example, Anscombe is using ‘person’ in what she later describes as ‘the sense in which it occurs in “offenses against the person”. . . . “A person” [in this sense] is a living human body’ (33). 2 

384   Adrian Haddock characteristically privileged views: except in mirrors he never sees the whole person, and can only get rather special views of [the person] he does see’ (24)—​namely, himself. In Smith’s utterance of (1), ‘this person’ refers to the one person of whom Smith has these views, through his having these views of this person. And because it is through this utterance that Smith fixes the reference of ‘A,’ so—​on Smith’s lips—​’A’ equally refers to this person. On Smith’s lips, then, ‘A’ refers to himself, and as such manifests consciousness of himself. But this consciousness is not self-​consciousness, because it is manifested by a word whose reference was fixed in this manner.



4. In the imagined society, it is not just Smith who fixes the reference of ‘A’ by an utterance of (1), in which the demonstrative refers to the person the speaker sees through the speaker’s having these ‘characteristically limited and characteristically privileged views’ of this person (‘except in mirrors he never sees the whole person, and can only get rather special views of [the person] he does see’). Everyone in this society fixes the reference of ‘A’ in the same manner. And so, on the lips of everyone in this society, ‘A’ refers to himself, and as such manifests consciousness of himself. But in no case is this self-​consciousness, because of how the reference of ‘A’ is, in every case, fixed. 5. Why is this not self-​consciousness? When Anscombe first presents her example, she writes as if this is obvious: [W]‌hen we speak of self-​consciousness, we don’t mean [this]. We mean something manifested by the use of ‘I’ as opposed to ‘A’. (25)



But she goes on to give a reason, which has become familiar in the literature, and which brings out a crucial difference between ‘I’ and ‘A’. 6. Consider a member of the society we have described, and consider an assertion, on their lips, of a sentence in which ‘A’ is held together with a predicate ‘F’. The fact that the reference of ‘A’ is fixed in the manner we described provides for the possibility of this assertion being based on an assertion that contains a demonstrative to which the conception associated with ‘person’ is internal. More fully, it provides for the possibility of this assertion being based on two further assertions containing this kind of demonstrative—​an assertion of (*) ‘This person is F,’ and an assertion of (**) ‘A is this person’. In so doing, it provides for the possibility of this assertion being in error in two different ways: either by the person referred to by the demonstrative not being F (that is, by an assertion of [*]‌being false) or by this person not being the referent of ‘A’ (that is, by an assertion of [**] being false). The second of these errors is possible because it is possible that the person referred to by the demonstrative is not the speaker, that is, not the person referred to by the demonstrative in the utterance of (1) that fixed the reference of ‘A’. Whereas the demonstrative in the reference-​fixing utterance refers in a manner

Anscombe and Self-Consciousness    385 that guarantees that its referent is the ‘one person [of whom the speaker] has [the] characteristically limited and also characteristically privileged views’—​namely, himself—​the manner that the demonstrative in an assertion of (*) or (**) refers does not guarantee this. The first of these errors is an error of predication, and the second is an error of identification. In virtue of how its reference was fixed, it is internal to ‘A’ that any subject-​predicate assertion with it as subject can be based on assertions of (*) and (**). And, as such, it is internal to ‘A’ that no subject-​ predicate assertion with it as subject is immune to an error of identification. This is the reason why ‘A’ is distinct from ‘I’—​because it is not internal to ‘I’ that no assertion with it as subject is immune to this kind of error. Wittgenstein gives ‘I have toothache,’ ‘I see red,’ and ‘I think it will rain’ as clear examples of assertions with ‘I’ as subject which are immune to an error of this kind.4 7. The upshot is that, even though, on everyone’s lips in the imagined society, ‘A’ manifests consciousness of himself, in no case is this self-​consciousness, just because of this difference between ‘A’ and ‘I’. 8. It does not follow that ‘I’ is not a referring term, merely that, if it is such a term, then it is not one whose reference is fixed in the manner of the reference of ‘A’. But if ‘I’ is a referring term, what kind of referring term could it be? It could not be one that is like ‘A’ in that its reference is fixed by the utterance of a demonstrative, but unlike ‘A’ in that the demonstrative whose utterance fixes its reference does not refer to its referent in a manner that guarantees that its referent is the speaker; for if it was such a term, then the reflexive character that must characterise ‘I’, insofar as it is a referring term, would be missed (if ‘I’ is such a term then, on anyone’s lips, it refers to himself). However, there is a kind of referring expression whose reference need not be fixed by an utterance of a demonstrative—​namely, a demonstrative. Might it be that ‘I’ is a word of this kind—​specifically, a word of this kind that refers in a manner that carries the present guarantee of reflexive reference, perhaps through its speaker’s having the ‘characteristically limited and . . . characteristically privileged views’ of its referent that Anscombe describes? 9. The reference of a demonstrative is fixed simply through the obtaining of a suitable nexus—​specifically, simply through something that meets the conception internal to the demonstrative being given to the senses of its speaker in some manner at the time of its utterance. As its reference is fixed in this way, the reason for thinking that no assertion with ‘A’ as subject is immune to error through misidentification does not carry over to assertions whose subjects are demonstratives. It is because ‘A’ is a term whose reference is fixed through the utterance of a demonstrative that the possibility of assertions with ‘A’ as subject being based on assertions of the sentence ‘A is this person’ is provided for,





4 

Ludwig Wittgenstein, The Blue and Brown Books (Oxford: Basil Blackwell, 1975), 66–​67.

386   Adrian Haddock and with it the possibility of their being subject to identification-​error. It might seem that this difference is enough to immunise assertions with demonstratives as subject against this kind of error. And it might seem that, insofar as assertions with demonstratives as subject are immune to this kind of error, there is no obstacle to understanding ‘I’ as a demonstrative expression of some kind. 10. But there is a fundamental difference between ‘I’ and any demonstrative. Demonstratives are vulnerable to reference failure: there may be nothing meeting the conception internal to a demonstrative that is given to the senses of its speaker at the time of its utterance. In the assertion ‘This parcel of ashes is all that is left of poor Jones,’ the demonstrative may fail to refer because no parcel of ashes is given to the senses of its speaker at the time of the assertion. A more extreme possibility is needed to illustrate reference failure in the case of ‘this person,’ such as its speaker suffering from some kind of sensory deprivation (by being immersed inside Anscombe’s famous sensory deprivation tank, for example). But this is enough to establish the contrast with ‘I,’ because in its case the possibility of reference failure is equally precluded. If ‘I’ is a referring expression, then it is guaranteed to refer in every one of its occurrences. 11. There have been attempts to question this.5 But its ground lies in a feature of ‘I’ that we exploited in our discussion of the imagined society. If ‘I’ is a referring expression, then it refers, in every one of its occurrences, to its speaker. In this respect, ‘I’ matches ‘A’; the crucial difference between ‘I’ and ‘A’ is that the reference of ‘I’ cannot be fixed through the utterance of a demonstrative. (It is because the reference of ‘A’ is fixed in this manner that ‘A’ is a name of some kind.) Whatever kind of expression ‘I’ might be, it is not a referring expression that refers in the manner either of a demonstrative or of a name whose reference is fixed by a demonstrative, such as ‘A’. 12. The orthodoxy in the contemporary literature seems to be that nothing positive can be said about how the reference of ‘I’ is fixed. However, it is widely believed that something positive can be said about the sense of ‘I’—​namely, that it is reflexive in character. Both the accounts of ‘I’ that we have considered take it for granted that ‘I’ is a device of reflexive reference, and as such a device with a reflexive sense. As Anscombe puts it, the accounts take it for granted that, if ‘I’ is a referring term, then it ‘expresses a way its object is reached by [its user], what Frege called an ‘Art des Gegebenseins’ [one such that] the only object reached in that way by anyone is identical to himself ’ (23). This way of being given an object is the sense, or (equivalently) the consciousness that the use of ‘I’ expresses, or manifests, insofar as ‘I’ is a referring term. It is consciousness that is reflexive in character, in that it concerns the very speaker on whose lips it is manifested. For Anscombe, the idea that ‘I’ is a referring term, and the idea that ‘I’ expresses 5 

See Gareth Evans, The Varieties of Reference, ed. John McDowell (Oxford: Clarendon Press, 1982), ch. 7, §6.

Anscombe and Self-Consciousness    387 such a consciousness (or sense) are the same idea. This is not a substantive assumption. If someone wants to claim that ‘I’ can be a referring term even if it does not express any sense at all, Anscombe will not demur; she will simply say that whether ‘I’ is a referring term, in this sense, is not her topic. Her topic is consciousness, and specifically whether the consciousness manifested by the use of ‘I’—​self-​consciousness—​is consciousness of oneself. On Anscombe’s lips, the remark that ‘I’ is not a referring term is simply a way of giving voice to the idea that self-​consciousness is not such consciousness. 13. The specific explanations of ‘I’ that we have considered—​which try to understand ‘I’ as a kind of name (namely, ‘A’), or as a kind of demonstrative—​seek to understand what this reflexive way of being given an object comes to. The sheer idea that ‘I’ expresses this reflexive ‘Art des Gegebenseins’ does not explain how the reference of ‘I’ is fixed in such a way as to guarantee that it does express this; the explanations that we have considered seek to fill this gap. And it is a gap that needs to be filled, if the claim that ‘I’ expresses this reflexive sense is to be justified, on account of a feature of this sense that Anscombe brings out in the early pages of her essay—​namely, the impossibility of saying what ‘I’ refers to in a manner that expresses or manifests this sense. Self-​consciousness is something manifested by the use of ‘I’; but if self-​consciousness is the reflexive consciousness that we are currently trying to understand it to be, then it cannot be manifested by any explanation that says what ‘I’ refers to. This might lead us to claim that ‘I’ is a device of (so-​called) ‘direct reference’: a referring term all right, but not one that expresses any ‘Art des Gegebenseins’.6 This claim does not stand opposed to Anscombe’s remark that ‘I’ is not a referring term—​on the contrary, given what her remark comes to, it is one way of making it. Her topic is the consciousness manifested by the use of ‘I’ and whether this is consciousness of oneself; to be justified in thinking that it is such consciousness, it is not enough to say that ‘I’ is a term that conforms to a certain reflexive reference rule and then to propound this rule—​for that does not settle whether ‘I’ manifests such consciousness. To settle this, a further explanation is needed, which explains how the reference of ‘I’ is fixed, in such a way as to guarantee that it does manifest consciousness of this form. 14. Why does Anscombe think there is no way to say what ‘I’ refers to that succeeds in expressing the ‘reflexive’ sense that it would express, if self-​consciousness were consciousness of oneself? Here it helps to consider a moment in Anscombe’s essay that might seem puzzling. Directly after her remark that ‘‘I’ is neither a name nor another kind of expression whose logical role is to make a reference, at all,’ she writes: Of course, we must accept the rule ‘If X asserts something with ‘I’ as subject, his assertion will be true if and only if what he asserts is true of X’. But if 6 The locus classicus here is David Kaplan, ‘Demonstratives’, in Joseph Almog, John Perry, and Howard

Wettstein (eds.) Themes from Kaplan (New York: Oxford University Press, 1989), 481–564.

388   Adrian Haddock someone thinks that is a sufficient account of ‘I’, we must say: ‘No, it is not’, for it does not make any difference between ‘I’ and ‘A’. The truth condition of the whole sentence does not determine the meaning of the items within the sentence. (32–​33)

The fact that ‘A’ equally conforms to the rule shows that the rule is not a sufficient account of ‘I’. However, the idea that the rule does not determine ‘I’ as a referring expression might seem wrong. It does not follow, from the fact that an assertion of a sentence with ‘I’ as subject has the same truth condition as an assertion of a sentence with ‘A’ as subject, that in each assertion ‘I’ and ‘A’ are the same in sense: we know that the sense of ‘A’ is that of a name whose reference is fixed by a demonstrative, and we know that this cannot be the sense of ‘I’. But we might think that it does follow, from the sheer fact that assertions with ‘I’ as subject have truth conditions determined by this rule, that ‘I’ is a referring expression of some kind. Anscombe denies this, however, on the grounds that the truth of the rule entails nothing about the consciousness manifested by ‘I’, and specifically nothing about whether this is consciousness of oneself. Assertions with ‘I’ as subject can conform to the rule, and as such can be assigned truth conditions in accordance with the rule, without on this account being assertions in which ‘I’ functions as a term that expresses a reflexive ‘Art des Gegebenseins’. 15. As the rule is a generalisation, we can consider instances, such as: (2) If Smith asserts something with ‘I’ as subject, his assertion will be true if and only if what he asserts is true of Smith.

One feature of both (2) and the rule is that neither uses the word they concern—​namely, ‘I’—​to refer to anyone. In this respect, they contrast with the clauses of a meaning theory, as the work of Donald Davidson and others has taught us to think of these.7 Such a theory might contain as clauses: (3) ‘Hesperus’ refers to Hesperus. (4) ‘Phosphorus’ refers to Phosphorus.

Each clause not merely refers to a certain word—​on the left-​hand side of ‘refers to’—​ and to a certain heavenly body—​on the right—​and says that a certain relation holds between this word (specifically, this name) and this heavenly body; each clause refers, on the right, to the heavenly body by using the very name that it refers to on the left. John McDowell puts this by saying that the clauses

7 

See the editors’ superb introduction to Gareth Evans and John McDowell (eds.), Truth and Meaning: Essays in Semantics (Oxford: Clarendon Press, 1976).

Anscombe and Self-Consciousness    389 specify the object presented by the names [in a manner that is] constrained to present it in the ways in which the respective names present it. [And they] meet this constraint—​surely infallibly—​by actually using the respective names.8

For this reason, each clause may be said not merely to say that a certain word refers to a certain object but—​in saying this—​to express or show the sense of the word, through putting it to its referential use. For present purposes, we need not be concerned with the details of meaning theories, as Davidson and McDowell understand these, but merely with this idea of saying what a word refers to in a way that shows its sense. Anscombe makes essential use of this idea, as we shall see. 16. In contrast to the clauses we have just considered, neither the rule nor its instances show the sense of the word they concern, because they do not put the word to its referential use.9 We can see this by trying to supplement the rule with a further rule that seeks to explain the referential contribution that ‘I’ makes to the determination of the truth conditions that the earlier rule specifies. This further rule would say something like: (5) In using ‘I’, X refers to himself.

This rule speaks, not of what an expression (here, ‘I’) refers to on the lips of a speaker, but of what a speaker refers to in using this expression. But for present purposes, this is no more than a verbal difference: the idea of what an expression refers to, on a speaker’s lips, and the idea of what the speaker refers to, in using the expression, are the same idea. The crucial point is that, as (5) is a generalisation, we can consider its instances, for example: (6) In using ‘I’, Smith refers to himself.

And now, Anscombe’s objection to (6) runs as follows:10 If the reflexive pronoun [in (6)] is the ordinary one, then it specifies for us who frame or hear the sentence an object . . . namely the object designated by . . . ‘Smith’. [But] as Frege held, there is no path back from reference to sense; any object has many ways of being specified, and in this case, through the peculiarity of the construction, we have succeeded in specifying an object (through [‘Smith’]) without specifying

8 

John McDowell, ‘On the Sense and Reference of a Proper Name’ (1977), in his Meaning, Knowledge, and Reality (Cambridge, MA: Harvard University Press, 1998), 181. 9  Compare here A. N. Prior’s terrific, but effectively unknown paper, ‘I’, in B. Y. Khanbhai, R. S. Katz, and R. A. Pineau (eds.), The Jowett Papers, 1968–​69 (Oxford: Basil Blackwell, 1970), 1–10. (For some excellent criticism of the details of Prior’s position, see the response by Anselm Müller, in the same volume.) 10  Here I have abstracted away from certain complexities of Anscombe’s own discussion to bring out what I take to be the fundamental point.

390   Adrian Haddock any conception under which Smith’s mind is supposed to latch onto it. (22–​23, her emphasis)11

Her italicisation of ‘Smith’ implies that, even though in using (6) we fail to specify any conception under which Smith’s mind latches onto the object, we do specify a conception under which our mind latches onto it. ‘Conception’ here is another word for ‘sense’.12 And we specify the way in which our mind presents the object that we say Smith uses ‘I’ to refer to in that we show its way of presenting it, through using the word ‘Smith’. But we do not show the way in which Smith’s mind supposedly presents this object. And that is because, in using (6), we do not refer to this object through using the word that (6) says that Smith uses in referring to it. No one can, in using (6), refer to this object through using this word (not even Smith). The form of (6) as an instance of (5) ensures that it cannot, in saying what it does, show the sense of ‘I’ on Smith’s lips. 18. It is widely believed that, directly after the passage just cited, Anscombe gives an argument that Gareth Evans summarises as follows: [Anscombe] argues that either [(6)] is simply equivalent to [‘In using ‘I,’ Smith refers to Smith’], in which case [(5)] is wrong; or else it means something which can be elucidated only by reference to the first-​person pronoun, in which case [(5)] is useless.13

But Evans’s response to the argument that he attributes to Anscombe shows that he does not understand what she is saying in the remarks of hers that directly follow the passage just cited. Anscombe does think that from (6) we can infer: (7) In using ‘I,’ Smith refers to Smith.

11 

In his ‘Frege’s Theory of Predication: An Elaboration and Defense, with Some New Applications’, Philosophical Review 103, no. 4 (1994): 599–​637, Ian Rumfitt writes that, in this passage, Anscombe ‘evidently presupposes that the semantic function of the ordinary reflexive pronoun . . . is to make a designation; in Anscombe’s own words in the passage quoted, such a pronoun “specifies for us who frame [a sentence containing it—​IR] an object” ’ (632)—​for example, in the case of (6), the object designated by ‘Smith’. But in order to specify this object for us, the pronoun in (6) need not designate Smith; it need only indicate that Smith is not only the subject who refers but also the object he refers to. In (5), in which no individual referring subject is mentioned, the function of the pronoun is evidently not to refer but merely to indicate the sameness of the referring subject and the referred to object; it would be strange to think that, in instantiating (5) with ‘Smith’ (6) changes the semantic function of the pronoun to that of a device of singular reference. That the pronoun indicates this sameness in (6) is not something that Rumfitt would deny. And that it indicates this without referring to Smith is a point that Rumfitt insists upon. The sheer fact that Anscombe says that this pronoun ‘specifies for us . . . an object’ is no reason to think that she denies this. 12  This is not quite the significance that the word ‘conception’ bore in FP, §6, or that which it bears at other points in her paper (for example, in the discussion of proper names on 26); there it signifies the idea of what something is (a so-​called sortal concept). 13  Evans, op. cit. The Varieties of Reference, p. 258.

Anscombe and Self-Consciousness    391 And although she never says anything to imply that (6) and (7) are ‘simply equivalent,’ it is instructive to ask why Evans thinks that (5) would be wrong if they were. It might seem that he thinks this because he assumes that the office of (6) is to say what ‘I’ refers to, on Smith’s lips, and in so doing to show its sense, and he sees that on no one’s lips can the sense of ‘I’ be that of ‘Smith’. However, the way Evans understands the second limb of the argument that he attributes to Anscombe stands in tension with this assumption. It is tempting to replace the ordinary reflexive in (6) with a special reflexive that registers the wrongness of (7). We begin with: (8) In using ‘I,’ X refers to himself*.

Here we indicate the specialness of the reflexive by marking it with a star. And we explain its specialness by saying that it signifies that the speaker refers to himself (ordinary reflexive) in the specific way associated with ‘I’ on his lips. To consider an instance: (9) In using ‘I,’ Smith refers to himself*.

And now we explain the significance of the reflexive, in this instance, by saying that (9) is simply equivalent to: (10) In using ‘I,’ Smith refers to himself (ordinary reflexive) in the way associated with ‘I’ on his lips.

According to Evans, Anscombe thinks that (9) is ‘useless’ because it can be elucidated only ‘by reference to the first person,’ that is, only by (10). That might seem to match Anscombe’s own point in the passage cited earlier, for (10) does not, on the right-​hand side of ‘refers to,’ use ‘I’ to refer to Smith, and so does not show the sense that it bears on his lips, but refers to ‘I’ instead. However, Evans’s response to the difficulty that he takes Anscombe to raise shows that he thinks the problem with (10) is not that ‘I’ is referred to rather than used to refer to Smith, but that ‘I’ shows up on the right-​hand side at all, for he goes on to explain the special reflexive in a different way, which neither refers to the first person on its right-​hand side nor uses it to refer to anyone, but rather says that the truth of (9) consists in the fact that ‘I,’ on Smith’s lips, is associated with a reflexive way of referring. He holds that (9) is simply equivalent to: (11) In using ‘I,’ Smith refers to himself (ordinary reflexive) in a reflexive way which is such that the only one that anyone refers to in this way is identical to himself.

And (11) no more shows the sense of ‘I’ on Smith’s lips than (7) or (10) does. It cannot be, then, that Evans thinks that the office of an explanation of ‘I’ as a referring term is to show its sense, and, as such, it cannot be that the reason why Evans thinks that (7) is

392   Adrian Haddock wrong is that it fails to do this. But then what is wrong with (7), by Evans’s lights? I do not think this question has an answer. 19. The argument that Evans attributes to Anscombe has come to be known as her ‘circularity argument’;14 Evans’s understanding of the argument, borne out by his response, makes that title apt. But the title carries an implication that is quite contrary to Anscombe’s intentions. For Anscombe, the problem with (10) is not that it is circular but that it is not circular,15 in the manner of (3) and (4). Their circularity consists—​specifically—​in their putting to referential use on the right-​hand side the very word that they mention on the left, and—​as such—​in their showing the sense of the word that they concern. 20. As we have seen, the fact that ‘I’ conforms to a reflexive rule that permits the assignment of truth conditions to subject-​predicate assertions with ‘I’ as subject does not fix it that ‘I’ has the sense of a referring expression; and explaining ‘I’ as a referring expression by means of a reflexive generalisation, such as (5), equally fails to fix it that ‘I’ expresses the reflexive way of being given an object that it needs to express if it is to be counted a bona fide referring term. That is the argument Anscombe gives, directly after the passage we cited. As she puts it, ‘the explanation of the word ‘I’ as ‘the word which each of us uses to speak of himself ’ is . . . no explanation if that reflexive has in turn to be explained in terms of ‘I’. . . . We seem to need a sense to be specified for . . . ‘I’ [and, to] repeat the Frege point: we haven’t got this sense just by being told which object a man will be speaking of . . . when he says “I” ’ (23). The question is: what would it be for us to ‘have’ this sense? Anscombe seems to recognise only two possibilities: it would be for us to have an explanation that says what the word refers to, and in so doing expresses it sense; or, it would be for us to have an explanation that says how the reference of the word is fixed. And in the case of ‘I,’ no explanation of either kind has so far been forthcoming. This may lead us to try to assimilate ‘I’ to a referring expression of a different kind, specifically, to a definite description; for, if we could say ‘‘I” refers to such-​and-​such’—​where ‘such-​and-​such’ is replaced by the description to which it is being assimilated—​then we would have an explanation which at once says what ‘I’ refers to, and in so doing expresses its sense through using this description. Alternatively, it might lead us to look again at the case 14 

See, in particular, Rumfitt, ‘Frege’s Theory of Predication’; James Doyle, No Morality, No Self: Anscombe’s Radical Skepticism (Cambridge, MA: Harvard University Press, 2018). Rumfitt defends an explanation of ‘I’ of the same general form as Evans’s, whereas Doyle tries to criticise Rumfitt’s explanation by defending what he calls Anscombe’s ‘circularity argument’. It is no surprise that Doyle cannot sustain this criticism and is led to conclude that there ‘does not seem to be any important consideration yet adduced [by Anscombe or Rumfitt] that would decide the issue [between them] either way’ (130). There could not be any consideration that would decide ‘the issue’ between them, because there is no issue: Anscombe does not advance a ‘circularity argument’, and Rumfitt does not engage with her actual argument, because he defends an explanation of the same form as Evans’s. 15  I am grateful to Sebastian Rödl for this way of putting things.

Anscombe and Self-Consciousness    393 against ‘I’ as a demonstrative, for it might seem that the earlier case against that idea was not conclusive; specifically, we might wonder why there could not be a kind of demonstrative that is immune to reference failure, on account of the fact that its referent is something whose givenness to consciousness is guaranteed by the sheer activity of using ‘I,’ or, more generally, by the sheer activity of thinking what is expressed in using ‘I’—​’I-​thinking,’ as we might call it. Perhaps ‘I’ is a kind of demonstrative whose referent consists in nothing more than this activity? 21. Let us consider this second suggestion first. It would be to treat the referent of ‘I’ as a Cartesian Ego. Or, rather, a stretch of one.16 People have sometimes queried how Descartes could conclude to his RES cogitans. But this is to forget that Descartes declares its essence to be nothing but thinking. The thinking that thinks this thought—​this is what is guaranteed by ‘cogito’. (31)17

However, this position, according to Anscombe, has the intolerable difficulty of requiring an identification of the same referent in different ‘I’-​thoughts. (This led Russell at one point to speak of ‘short-​term selves’.) (31)

The idea of the activity of I-​thinking is the idea, not of a countable particular, but of the activity of I-​thinking, a mass: something that can be carved into countable particulars, but not itself such a particular. This explains why Anscombe speaks of ‘I’ as referring to a stretch of such activity. Specifically, it would refer to that stretch of I-​thinking which is constituted by the time at which ‘I’ occurs. As we might put it: ‘I’ refers to ‘this I (this stretch of I-​thinking)—​the one given now’. But then a stretch of I-​thinking given at one time will be distinct from a stretch of I-​thinking given later: at each succeeding time, a new stretch will come into being. This is the fundamental reason why it is tempting to speak of ‘short-​term selves’. That temptation might initially arise as a response to what seems an epistemic difficulty: ‘How do I know that this I (this stretch of I-​thinking) is the same I (the same stretch of I-​thinking) as the one I referred to earlier?’ But the difficulty is at a deeper level. If a stretch of I-​thinking is constituted by the time at which it is given, then the very idea of an I persisting over time comes to nothing—​a stretch referred to at one time is, as such, a different stretch to the one referred to at a different time. However, as Anscombe sees, even this does not appreciate the ultimate difficulty. Ultimately, the very idea of a stretch of I-​thinking cannot be made out. She asks, ‘How do I know that ‘I’ is not ten thinkers thinking in unison? Or perhaps not quite succeeding?’ (31). These questions reflect the tendency of even this difficulty to come into focus as epistemic in

16  ‘ “One” ’ is infelicitous; ‘such’ would be better the stretch at issue is a stretch of the activity of I-​ thinking—​not a Cartesian Ego, but that out of which any such Ego is carved. 17  At stake here is the very idea of a manifold of thinkers, an idea she returned to in her ‘Has Mankind One Soul? An Angel Distributed over Many Bodies’ (1985), in GG1.

394   Adrian Haddock character. But, as with the earlier difficulty, this difficulty is at a deeper level. Any time admits of an indefinite number of divisions, and the mere idea of I-​thinking supplies no resources to specify which time, from among this indefinite multiplicity, constitutes a stretch. The very idea of singular reference to a Cartesian Ego falls apart when we try to think it through. And it is worth noting a corollary. Even though we have couched this criticism as a criticism of ‘I’ as a special kind of demonstrative, it equally works against an account which understands the sense of ‘I’ as given by a (putative) definite description of the sort that Anscombe employs when she speaks of ‘the thinking that thinks this thought,’ where ‘this thought’ purports to refer to (as it was put earlier) ‘this stretch of I-​thinking’. The moral of our discussion is that the putative demonstrative that this putative definite description employs is not a real demonstrative at all. There is no prospect of assimilating ‘I’, either to such a putative demonstrative or to such a putative description. 22. Each of the positions that we have so far considered assumes that ‘I’ expresses a way in which something is given to consciousness. And the upshot of our discussion is that this cannot be right. According to Anscombe, abandoning this assumption is ‘the principal root of the philosophic idea of ‘the subject’: that ‘I’ does not stand for any object, not for anything presented’.18 I do not think that Anscombe takes issue with this idea. She does, however, consider the attempt to express the idea by speaking of ‘the subject’ or ‘the I’ and saying things like ‘The I [is] subject, not object, and hence invisible’. She calls the attempt to try to express the idea in this way ‘raving’, on account of the fact that, because it continues to speak of ‘the subject’, in contrast to ‘anything presented,’ it is ‘difficult to avoid the idea that [it] has introduced an extraordinary sort of object: something that is indeed not presented, because it is what presentations are made to’. Once that idea has been introduced, it is hard to resist the conclusion that it is this extraordinary object that is somehow given to consciousness in the use of ‘I’. And it is against this background that Anscombe makes perhaps the central remark of her essay: So long as that is the assumption [i.e., so long as there is the assumption that ‘I’ is a referring term] you will get the deep division between those whose considerations show that they have not perceived the difficulty—​for them ‘I’ is in principle no different from my ‘A’; and those who do—​or would—​perceive the difference and are led to rave in consequence. (32)

Those who have not perceived the difficulty might say that ‘I’ is different from ‘A,’ in that it does not reach its object through observation (for example); but they do not think the devices are different ‘in principle’ because they take it for granted that ‘I’ expresses some sort of way in which an object is given to consciousness. Those who have perceived the difficulty, by contrast, appreciate this difference, for they see that ‘I’ does not ‘stand 18 

G. E. M. Anscombe, ‘The Subjectivity of Sensation’ (1976), in CP2.

Anscombe and Self-Consciousness    395 for . . . anything presented’. And yet, even though they see this, they may be apt to express their insight by using forms of words that it is hard to understand without assuming that an object of some extraordinary kind must be somehow given to consciousness, in the use of ‘I’. Although they perceive the difference, their perception has not yet achieved full clarity. 23. We seem fated to oscillate between these positions. That is the problem. And this is the solution: ‘I’ is neither a name nor another kind of expression whose logical role is to make a reference, at all’. (32)

Only by achieving a firm grip on this solution—​only by ridding ourselves of the idea that ‘I’ expresses any kind of way of being given any kind of object—​can we acknowledge the specialness of ‘I’, but without falling into the raving she describes. 24. That is her solution. And I think it is undeniable. If ‘I’ is a referring term, in the sense of a term that expresses some kind of ‘Art des Gegebenseins,’ then it can be neither a demonstrative nor a name whose reference is fixed by a demonstrative. If we say that it is a referring term of a sui generis kind, which is sufficiently explained by a reflexive generalisation such as (5), then we have the problem that neither our generalisation nor any of its instances fixes it that the term expresses any way of being given an object: our putative explanation does not settle it that ‘I’ is a referring term, in Anscombe’s sense. This can make it seem that, if ‘I’ is such a term, then it must be one whose sense is given by a demonstrative, or a definite description, of a special kind—​one that involves the givenness to consciousness of the very activity of thinking that is involved either in using the term or (more generally) in thinking what is expressed by its use. But this suggestion collapses when we try to think it through. And with that, we arrive at the insight that can lead us to ‘rave’. The conclusion now seems plain: we must adopt Anscombe’s solution, on pain of raving. But her solution is radical. And to start to see just how radical, consider the utterance of ‘I am in pain’. In this utterance, ‘I’ does not express any way of being given an object, and, as such, the utterance does not belong to objective discourse, in the sense of discourse that bears the form of reference, and predication, in that it is composed of items—​be they utterances (assertions, for example) or sentences—​that determine truth conditions that bear this form. And the same goes for every utterance with ‘I’ as subject—​every ‘I-​utterance’. 25. To say that an I-​utterance does not determine a truth condition that bears this form is not to say that a truth condition of this form cannot be imposed onto the utterance ‘from the outside’. This is what the rule ‘If X asserts something with ‘I’ as subject, his assertion will be true if and only if what he asserts is true of X’ enables us to do. The truth-​condition that we can specify through the sentence that is used on the right-​hand side of an instance of this rule bears the form of reference and predication insofar as this sentence unites a term that expresses

396   Adrian Haddock a way of being given an object with a predicate. But insofar as it does, this sentence cannot—​given Anscombe’s conclusion—​be a sentence with ‘I’ as subject. And just for this reason, it cannot itself express the sense of the utterance that the instance concerns. (The instance might be, for example, ‘Smith’s assertion of ‘I am F’ is true if, and only is, Smith is F’.) The truth conditions, specified by the sentences that are used on the right-​hand side of instances of the rule, are imposed onto the I-​utterances that the instances concern, rather than being truth conditions that the utterances themselves determine. 26. Anscombe’s fundamental insight might be put by saying that there is a radical difference between objective discourse and subjective discourse, in the sense of discourse composed of utterances whose subject-​term is ‘I’. Cases of objective discourse determine truth conditions that bear the form of reference and predication, and as such are capable of stating facts bearing this form; but the semantic character of the first person fixes it that I-​utterances cannot state facts that bear this form. 27. That is a merely negative idea, of course. The form of subjective discourse is not that of reference and predication, in that the utterances it comprises do not determine truth conditions bearing this form. What is the form of such discourse? Anscombe does not answer this question. But if it is asked, the answer is that we do not know. Or, put differently: if we do know, we do not yet have any clear and distinct idea about how to manifest our knowledge clearly and distinctly. 28. The question of the form of subjective discourse strikes me as being of the most fundamental importance. At the time of writing, however, it is not one that most contemporary philosophers recognise as a question. To the extent that they think anything on the matter, they tend unhesitatingly to assume that utterances with ‘I’ as subject belong to objective discourse. But perhaps that will change—​as an increasing number of people come to read Anscombe with the kind of intensity and sympathy that befits a great philosopher. Cora Diamond once wrote, of Anscombe’s work in general, that it shows ‘how much there is that had not been seen before’.19 Nowhere is this clearer than in her penetrating and pathbreaking writings on the consciousness that is manifested in the use of ‘I’.

19 

Cora Diamond, ‘Preface’, in C. Diamond and J. Teichman (eds.), Intention and Intentionality: Essays in Honour of G. E. M. Anscombe (London: Harvester, 1979), xiii.

Chapter 18

T he First Pe rs on a nd ‘The First Pe rs on’ Harold Noonan

18.1 In her famous (or infamous) paper ‘The First Person’ Anscombe argues that ‘I’ is not a referring expression: ‘I’ is neither a name nor any other kind of expression whose logical role is to make a reference, at all.1 She does so by arguing that if ‘I’ were a referring expression it would have to be one whose reference was a stretch of a Cartesian Ego—​ an intolerable conclusion. But she also argues in addition that the first-​person pronoun is not a referring expression because it manifestly does not function as one, so anyone tempted to grasp the Cartesian nettle cannot feel complacent. Her position is obviously influenced by Wittgenstein, but it appears that she goes further than he does, at least in his published writings, and she does not appeal explicitly to the distinction he makes, which seems highly relevant, between the use of ‘I’ as object and the use of ‘I’ as subject. (It does appear that Wittgenstein thinks that in its use ‘as object’ ‘I’ does have the role of referring, but in its use ‘as subject’ this is not so.) Her no-​reference thesis has met with general incredulity. Two examples: one from van Inwagen: Professor Anscombe’s position is that it is not the function of the word ‘I’ to refer. The word is thus unlike ‘the present King of France’, which is in the denoting business but is a failure at it; rather the word, despite the fact that it can be the subject of a verb . . . is not in the denoting business at all . . . for Anscombe the word ‘I’ refers to nothing in a way like the way in which ‘if ’ and ‘however’ refer to nothing.2 1 

FP, 32. Peter van Inwagen, ‘ “I Am Elizabeth Anscombe” Is Not an Identity Proposition’, Metaphysica 2, no. 1 (2001): 6. 2 

398   Harold Noonan Strawson writes: It is simply an error to separate, as Miss Anscombe implicitly does, the semantics of the use of ‘I’ sentences from the question whether the use of ‘I’ is referential, and the question of what reference it makes. What McDowell generously describes as ‘the beginning of wisdom’ on this question is also its end; though it would be better to replace the lofty word ‘wisdom’ with something more modest like ‘plain sense’ (except that they often in philosophy come to much the same thing). Seriously to question whether, in any standard use of ‘I’ a person is referring to him or herself is as futile as seriously to question whether in any standard use of ‘now’ as a temporal adverb a person is referring to a (more or less extended) present.3

Most who discuss Anscombe’s paper interpret her as denying that ‘I’ is a referring expression and reject her thesis, though a few (Wiseman,4 Doyle5) dissent. I shall examine Anscombe’s argument. I shall not dissent from the majority view that Anscombe maintains the no-​reference thesis, and I shall conclude, also with the majority, that she is wrong to do so. ‘I’ is a referring expression and should be grouped specifically with the pure or automatic indexicals, including ‘here’ and ‘now.’ But it is a consequence that self-​reference (i.e., the self-​conscious and successful use of ‘I’) need not involve ‘the connection of what is understood by a predicate with a distinctly conceived subject.’6 That is, in intending to refer to himself (to use ‘I’ in accordance with its customary meaning) a speaker need not form an intention to refer to the such-​and-​such, when ‘such-​and-​such’ provides an identification of the speaker, which singles him out from everything else. It is a further consequence that ‘I’ is not guaranteed a reference and that a thinker of an ‘I’-​thought need not be the reference of the thought even if there is one. In arguing these points, I am following Evans7 and will appeal to work by Snowdon8 and Lewis.9 To a considerable extent I think this vindicates Anscombe.

18.2 I begin with a discussion of referring expressions. Anscombe’s no-​reference thesis is that ‘I’ is not a name or ‘any other expression’ which has a referring role. In her first paragraph she writes that ‘in these writers [Descartes and 3 

Peter Strawson, ‘Reply to John McDowell’, in The Philosophy of P. F. Strawson, Lewis Hahn (ed.) (LaSalle, IL: Open Court, 1998), 149. 4  Rachael Wiseman, ‘What Am I and What Am I Doing?’, Journal of Philosophy 114, no. 10 (2017): 536–​550. 5  James Doyle, No Morality, No Self: Anscombe’s Radical Skepticism (Cambridge, MA: Harvard University Press, 2018). 6  FP, 36. 7  Gareth Evans, The Varieties of Reference (Oxford: Oxford University Press, 1982). 8  Paul Snowdon, Persons, Animals, Ourselves (Oxford: Oxford University Press, 2014). 9  David Lewis, ‘Survival and Identity’, in The Identities of Persons, Amelie Oksenberg Rorty (ed.) (Berkeley: University of California Press, 1976), 17–​40.

The First Person and ‘The First Person’     399 St. Augustine] there is an assumption that when one says ‘I’ one is naming something such that the knowledge of its existence, which is knowledge of itself as thinking in all the various modes, determines what it is that is known to exist.’10 The first category of referring expression to which Anscombe denies ‘I’ belonging to is thus that of proper names. Of course, since Kripke’s revolution proper names have been a subject of huge controversy. But Anscombe is assuming a Fregean viewpoint. She writes, ‘If ‘I’ expresses a way its object is reached by him [its user], what Frege calls an ‘Art des Gebenseins,’ we want to know what that way is. . . . [T]‌his is to treat ‘I’ as a sort of proper name.’11 Again: If . . . ‘I’ is supposed to stand for its object as a proper name does, we need an account of a certain kind. The use of a name of an object is connected with a conception of that object . . . as the conception of a city is [related] to the names ‘London’ and ‘Chicago,’ that of a river to ‘Thames’ and ‘Nile,’ that of a man to ‘John’ and ‘Pat . . . .’ That is why some philosophers have elaborated the notion of ‘selves.’ . . . And just as we must be continuing our reference to the same city if we continue to use ‘London’ with the same reference, so we must each of us be continuing our reference to the same self (or ‘person’) if we continue to use ‘I’ with the same reference.12

On the account from Kripke a competent user of a proper name need have no knowledge at all of its bearer, certainly no identifying knowledge, no way of reaching it and it alone. By contrast, Anscombe’s position is that a competent use of a name requires identifying knowledge of its bearer (though perhaps this can differ from user to user), including knowledge of the sort of thing it is. So if ‘I’ is a name, then if a user of it knows anything he can express in the form ‘I am such-​and such’ (e.g., ‘I am a thinker’) he must know something non-​trivial he can express in the form ‘I am the such-​and-​such.’ A substantial part of Anscombe’s argument is that if one can refer to an object, oneself, using ‘I’ at all, one must be able to do so even when one has no such knowledge—​ unless one is a stretch of a Cartesian Ego. So either ‘I’ is not a name, or we are stretches of Cartesian Egos. Of course, this leaves room for a Kripkean to accept Anscombe’s main claim about the possible use of ‘I’ in the absence of identifying knowledge while holding that ‘I’ is nonetheless a referring expression and indeed a proper name, linked to its bearer by a causal-​ historical connection. But this is hardly an attractive view. There was no initial baptism in which I ‘fixed the reference’ of my use of ‘I’ by a reference-​fixing description with subsequent uses causally linked to that baptismal event. Even if the Kripkean view of proper names is correct, then, it does not provide a plausible alternative to Anscombe’s no-​reference thesis about the first person. The second category of referring expression to which Anscombe draws attention is that of definite descriptions. Her dismissal of the suggestion that ‘I’ belongs to that

10 

FP, 21. FP, 13. 12  FP, 26. 11 

400   Harold Noonan category is blunt. The only serious candidate for such an account is ‘the sayer of this,’ where ‘sayer’ implies ‘thinker.’ But ‘[h]‌ow could one justify the assumption that there is just one thinking which is the thinking of this thought . . . just one thinker?’13 The third category of referring expression Anscombe explicitly considers is that of demonstratives. Again, as in the case of proper names, she insists that the use of a demonstrative requires a conception of its object, an answer to the question ‘This what?,’ and that given that ‘I’ can have reference even in cases of sensory deprivation (and total amnesia), as her opponents assume, its reference, if it is assimilated to demonstratives, must be a stretch of a Cartesian Ego, since nothing else can be ‘present to me’ in such a situation. So, notable by their omission from Anscombe’s list are the pure or automatic indexicals, which are neither names nor descriptions, nor even demonstratives, since their use requires no accompanying demonstration. This seems to be a huge omission, since this is the obvious grouping: ‘I’ goes with ‘here’ and ‘now.’ But Anscombe’s first argument in her paper has, if sound, the consequence that this grouping is incorrect, and I think is intended to do so. It is, I believe, unsound. Nevertheless, I shall argue, as indicated earlier, that if we endorse this grouping we should accept that ‘I’, though it can have reference, is not guaranteed to do so, which is Anscombe’s main point. In some circumstances (as in the case of ‘here’ and ‘now’) there will be too many eligible referents for a token utterance of ‘I’ and no identifying intention to single just one out. So though an ‘I’-​utterance can have a unique reference it need not always do so, and its use by a thinker need not involve a reference to that thinker; neither of these things has to be the case for a linguistically competent and successful use of ‘I’—​just as an utterance of ‘here’ within a region need not have a reference at all, and if it has it need not be to that region.

18.3 The first argument Anscombe gives for her no-​referring thesis is that the tempting idea that ‘I’ can be explained as ‘the word each one uses in speaking of himself ’ faces a dilemma: either circular or incorrect. It is circular if the reflexive pronoun in the dictum is ‘a special one which can only be explained in terms of the first person’—​the indirect reflexive. It is incorrect if it is the ordinary reflexive. The ordinary reflexive occurs in ‘Cato killed himself,’ which is equivalent to ‘Cato killed Cato.’ It also occurs, Anscombe notes, in ‘When John Smith spoke of John Horatio Auberon Smith (named in a will perhaps) he was speaking of himself, but he did not know this.’ When ‘himself ’ is the ordinary reflexive ‘John Smith was speaking of himself ’ follows from ‘John Smith was speaking of John Smith,’ which in turn follows from ‘John Smith was speaking of JHAS and John Smith =​JHAS.’ But in these circumstances

13 

FP, 3.

The First Person and ‘The First Person’     401 John Smith does not know that he himself is JHAS; that is, he does not know what he would express by ‘I am JHAS.’ This last occurrence of ‘himself ’ is the indirect reflexive. It identifies for us what John Smith does not know: it is ‘I am JHAS.’ When the reflexive in the report is the ordinary one, by contrast, as in ‘John Smith (reading the will out loud) spoke of himself but did not know this,’ we are not told under what conception the mind of the subject of the report latches on to its object (himself).14 So, Anscombe argues, if the reflexive in the dictum ‘‘I’ is the word each one uses to speak of himself ’ is the indirect reflexive, which can be explained only in terms of the first person, we have only a viciously circular account of ‘I,’ not an explanation of its meaning. That is, the dictum so read does not tell us how its meaning differs from that of other referring expressions. What of the other horn of the dilemma? Anscombe says that if in ‘It’s the word each one uses in speaking of himself ’ ‘himself ’ is the ordinary reflexive, this cannot explain what ‘I’ means, thought of as a referring expression. And, if so, she adds, it will be no use to expand the dictum to ‘It is the word each one uses when knowingly and intentionally speaking of himself.’ ‘For did not Smith knowingly and intentionally speak of Smith? Was not the person he intended to speak of—​Smith? and so was not the person he intended to speak of—​himself?’15 The addition is completely convincing. But the initial argument for this horn of the dilemma (that is, that if ‘himself ’ in the dictum is the ordinary reflexive it cannot explain what ‘I’ means thought of as a referring expression) is open to challenge. First, we must note that the dictum is anyway false, as stated. Everyone speaks of himself using other words. Some people do so quite a lot—​De Gaulle, the Queen, Margaret Thatcher—​and some people could conceivably avoid ‘I’ or any equivalent construction completely. So the dictum is better expressed: Whenever someone uses ‘I’ he speaks of himself. If ‘himself ’ here is the ordinary reflexive this is equivalent to: For any person x, when x uses ‘I,’ x refers to x. In this no reflexive pronoun occurs. But this does seem to distinguish the meaning of ‘I’ from that of any other pronoun, name, or description. Comparable dicta: ‘for any place p, an utterance of ‘here’ at that place refers to p’ and ‘for any time t, an utterance of ‘now’ at t refers to t.’ This is why it seems correct to group ‘I’ with ‘here’ and ‘now.’

18.4 Having satisfied herself that ‘I’ cannot be explained either by way of the ordinary or by way of the indirect reflexive, and thus, I think, in her own mind having ruled out its grouping with the pure indexicals, Anscombe next turns to the proposal that it is a sort of proper name, one that everyone has but uses only of himself.

14  15 

FP, 23. FP, 22.

402   Harold Noonan She first considers the suggestion that her thesis that ‘I’ is not a proper name ‘seems to reduce to the triviality that we perhaps would not call a word a proper name if everyone had it and used it only to speak of himself.’16 That is, that ‘I’ is ‘only not called a proper name because everyone uses it only to refer himself.’ The purpose of her parable of the ‘A’-​users is to establish that this is not so. ‘A’ is supposed to be an example of an expression which each one uses to refer to himself and no one else which is clearly a proper name.17 I think that Anscombe’s use of this parable does establish this point. But there is a variation of the parable (which I introduce later) in which ‘A,’ while not a proper name, is a referring expression. The actual use of ‘I’ is closer to that of ‘A’ in this variation than in Anscombe’s original. Yet in this variation (as in the original) an ‘A’-​user is not guaranteed by a linguistically competent and successful use of ‘A’ to refer to himself. This variation illustrates how an expression analogous to ‘I’ may not have a guaranteed reference and yet be a referring expression (i.e., one capable of reference)—​the option in the case of ‘I’ dismissed by Anscombe and most of her opponents and which I will argue for. Anscombe’s parable runs as follows: Imagine a society in which everyone is labelled with two names. One appears on their backs and the top of their chests, and these names, which their bearers cannot see, are various, ‘B’ to ‘Z’ let us say. The other, ‘A,’ is stamped on the inside of their wrists, and is the same for everyone. In making reports on people’s actions everyone uses the names on their chests or backs if he can see these names or is used to seeing them. Everyone also learns to respond to utterance of the name on his own chest and back in the sort of way and circumstances in which we tend to respond to utterance of our names. Reports on one’s own actions, which one gives straight off from observation, are made using the name on the wrist. Such reports are made, not on the basis of observation alone, but also on that of inference and testimony or other information. B, for example, derives conclusions expressed by sentences with ‘A’ as subject, from other people’s statements using ‘B’ as subject. (FP, 24)

Anscombe goes on: Thus for each person there is one person of whom he has characteristically limited and also characteristically privileged views: except in mirrors he never sees the whole person, and can only get rather special views of what he does see. Some of these are specially good, others specially bad. Of course, a man B may sometimes make a mistake through seeing the name ‘A’ on the wrist of another, and not realising it is the wrist of a man whose other name is after all not inaccessible to B in the special way in which his own name (‘B’) is. (FP, 24)

16 

17 

FP, 23–​24. FP, 24.

The First Person and ‘The First Person’     403 It seems clear that ‘A’ in this society is, as Anscombe says, a name, albeit one which is the same for everyone but used by each person to refer only to himself (that is to say, there may be occasions when B, say, uses ‘A’ to speak of someone other than B, but such uses of ‘A’ will be mistakes, misuses of ‘A’). For it is evidently Anscombe’s intention that the sense of ‘A’ in B’s mouth should be such that ‘B’ can correctly use ‘A’ only to speak of something (i) of which he gets the special view she speaks of and (ii) of which the sense of the name ‘B’ is a mode of presentation. This explains why she thinks the use of ‘A’ involves reidentification.18 The idea is that for some reason ‘A’-​users do not, or perhaps cannot, use their public names to refer to themselves. Each must substitute for a statement he would make using his public name an equivalent statement with ‘A’ replacing it. So B can infer from other statements about him a conclusion he expresses with ‘A’ as subject, and he can also make assertions with ‘A’ as subject on the basis of observation of features of that person of whom he has a characteristically special view. But such assertions may be mistaken if, as can happen, another person happens to be one of which he has on this occasion a special view. But now consider my variation of the fantasy. Consider first why Brown’s using ‘I’ to refer to Smith is wrong. Is it because Brown is then referring to someone other than Brown with ‘I’? No. Telling him that will not necessarily enlighten him. For Brown might not know that he is named ‘Brown’ and still be able to use ‘I’ correctly. Brown has learned to respond to utterances of the name ‘Brown’ in the appropriate way, but it is no part of his understanding of ‘I’ that he recognise that in his mouth it is equivalent to ‘Brown.’ Let us imagine, then, that it is no part of the understanding of ‘A’ that it is required of B that he assent to ‘B is A.’ What is required of B to show a grasp of the meaning of ‘A’ is merely that he recognise he must use ‘A’ only to refer to that item of which he gets the limited and privileged views Anscombe talks about. Now that item is B himself. So it looks as if ‘A,’ in this variation of Anscombe’s story, is an expression which is necessarily used by each person to refer to himself, while yet being such that for any member of the society when ‘N’ is that person’s public name, ‘A is N’ may be something he does not know. However, this can be maintained only if it is really the case that, for example, B would be necessarily using ‘A’ incorrectly if he were to use it to refer to C. But this is not so. For that B is the object of which B gets a special view is not a necessary fact, but a consequence of the nature of B’s perceptual apparatus. If this were tampered with so that the view B previously got of B he now got of C—​and this is especially easy to imagine if we think, as Anscombe suggests, of the ‘A’-​users as machines rather than people—​there would be no mistake in B’s using ‘A’ to refer to C. In so using it, in fact, he would be using it in just the way he was obliged to in order to use it correctly, given its meaning. (We can, of course, speak of the meaning of ‘A,’ and its sense in someone’s mouth, in this variant of Anscombe’s story.) This becomes obvious once it is realised that there

18 

FP, 27.

404   Harold Noonan would be no way of explaining to B the mistake he was making in using ‘A.’ Before B was tampered with, if he asserted ‘F(A)’ as a result of misidentifying someone else as ‘A,’ one could show him that that person was not the one of which he had a special view, hence not the one of which the sense of ‘A’ as used by him was a mode of presentation. But after B has been tampered with, if he asserts ‘F(A)’ as a result of observing that F(C), what can one say to him? To tell him that he is B is no help since he no longer believes that he is B; i.e., he no longer assents to ‘A is B.’ To tell him ‘You are not C’ (given that in this society ‘You’ is governed by the rule that ‘You are F’ addressed to X is true if and only if X can correctly assert ‘A is F’) will not help since if he checks (i.e., checks whether or not he should assent to ‘A is C’), he will find that he should. And there is just nothing else to say to him. Philosophers, attempting to explain ‘I’ as a sort of referring expression, sometimes suggest that it is equivalent to ‘this self.’ But the idea is not that any legitimate use of ‘this self ’ is equivalent to a use of ‘I.’ Rather it is that ‘I’ is correctly used if it is used in the way ‘this self ’ would be used if it were used by any self only to refer to that self presented to it in inner perception.19 They then face the task of explaining why no self can have an inner perception of any other self—​since, according to their account of the meaning of ‘I,’ any self of which a self has an inner perception, whether or not it is itself, is a self it can correctly refer to with ‘I.’ The defender of the position that ‘A’ can be used correctly in our variant of Anscombe’s story only if it is used to refer to its user faces the analogous task of explaining why it is impossible for C to be so presented to B that B would be correct in using ‘A’ to refer to C—​except that in this case it clearly is possible. So ‘A’ as used in my variation of Anscombe’s parable is not a proper name. Unlike ‘A’ in the original parable, reidentification is not involved in its use. In order to determine whether a mistake has been made in using ‘A’ no recourse is needed to what is the case at a time other than the time of utterance of ‘A.’ If B should say ‘A has a broken leg,’ he can be shown to be mistaken by showing that the body of which he is capable of a special view has no broken leg. By contrast, since the reference of a proper name is that object which at a certain time fulfilled a certain condition, to establish whether a mistake has been made in using a proper name may require recourse to what is the case at a time other than the time of utterance of the name. So ‘A’ in my variation of Anscombe’s story is not a proper name, since its use does not involve reidentification, yet it is a referring expression though it does not have a guaranteed reference. The same, I will argue, is true of ‘I.’

19 

As Anscombe points out, the intelligible use of demonstrative pronouns does not require the presence of a referent but only the presence of something to latch on to. I may utter ‘these ashes’, meaning the ones in the urn, though I do not see the ashes but only the urn. The ashes, if they exist, are my referent, but what my utterance latches on to is the urn. In the same way I could use ‘this self ’ to refer to the self connected to a particular body, even if that self was not an object of perception for me, but only the body. In the use of ‘this self ’ which is imagined by philosophers to be equivalent to ‘I’, however, the referent and what the utterance latches on to have to be identical; i.e., the referent has to be present to consciousness—​and present, moreover, in a certain special way.

The First Person and ‘The First Person’     405

18.5 After considering the idea that ‘I’ is a proper name Anscombe turns next to the idea that it is a sort of demonstrative expression (like ‘A’ in the variant of the parable). But at this point she appeals to the ‘guaranteed reference,’ which ‘I’ is supposed to have to argue that if this assimilation is made the referent of ‘I’ must be a Cartesian Ego. A consequence of the ‘guaranteed reference’ of ‘I’ she is supposing is that I can only ever use ‘I’ correctly to refer to myself. Hence that I must always use it to refer to the same thing. In this respect ‘I’ is unlike the variant use of ‘A’ I have explained and like the use of a proper name. But, if so, how can it function as a demonstrative? How can it be thus both like ‘A’ in the variant story and like a proper name? It seems, Anscombe says, that this reference could be sure-​fire only if the referent of ‘I’ were both freshly defined with each use and also remained in view so that nothing else was ever taken to be ‘I’. But consider how this could be. It has to be maintained that the ‘inner sense’ by which I can perceive myself qua thinking thing is necessarily restricted in its scope to myself—​ or to my own thoughts (if I am to be thought of as latching on to them primarily and think of myself via the demonstrative-​including-​description ‘the thinker of these thoughts’). But how might such a restriction in range be understood? It is tempting to think that one cannot be acquainted with another’s self or thoughts in the way he is because there is a sort of barrier, opaque to one’s inner sense, behind which he and his thoughts lie. But we cannot take this thought seriously. What cannot be seen because it lies behind a barrier could be seen if the barrier was down. But we do not want to allow the possibility of any circumstances in which we could know the thoughts of another in the way that he knows them. Another idea that comes to mind is that we should think of the impossibility of knowing another’s thought in the way he does by analogy with the impossibility of seeing sound or hearing colours. According to this idea, the reason my inner sense cannot be extended in its range beyond my own thoughts is that my own thoughts constitute the entire class of its proper objects. But this entails that there are as many kinds of inner sense as there are individual thinkers, and that my thoughts, your thoughts, and a third person’s thoughts have no more in common than a colour, a sound, and a smell. Why, then, are they all called ‘thoughts’? And how can this be so if we are all the same kind of thinking thing? One possibility remains, suggested by Anscombe’s reference to ‘an imaginative tour de force on the part of Locke’: might not the thinking substance which thought the thought ‘I did it’—​the genuine thought of agent memory—​nonetheless be a different thinking substance from the one that could have had the thought ‘I am doing it’ when the act was done? ‘Thus he detached the identity of the self or “person” from the identity even of the thinking being which does the actual thinking of the I-​thoughts’.20 According to 20 

FP, 26.

406   Harold Noonan Locke, in this circumstance, though one thinking substance recalls what another did, this makes the two thinking substances one person. Locke was thinking of diachronic identity only, but his idea may be applied to synchronic identity too. One might maintain that this is the reason one person cannot perceive by inner sense another person or his thoughts: any other thinking substance to whose thoughts a thinking substance has access by inner sense thereby counts as the same person as the given thinking substance—​whether or not it is the same substance. (Perhaps, as Anscombe puts it, ‘I am ten thinkers thinking in unison, or perhaps not quite succeeding. That might account for the confusion of thought which I sometimes feel.’)21 However, this proposal can help to explain how my inner sense is necessarily restricted in its range to my own thoughts only if my own thoughts =​the thoughts of whatever is the same person as me. But this is so only if what I am is essentially a person. So the proposal does not explain how my inner sense can be necessarily restricted in its range to my own thoughts if what I am essentially is a thinking substance—​or a human being. All it comes to is a stipulation that if I perceive by inner sense the thoughts of another thinking substance or human being he must be called the same person as me. But this plainly gets us nowhere. So we come to Anscombe’s conclusion, which is not yet that ‘I’ is not a referring expression, but that if it is, I am not a human being or a thinking substance but ‘a Cartesian Ego. . . . Or rather, a stretch of one. People have sometimes queried how Descartes could conclude to his RES cogitans. But this is to forget that Descartes declares its essence to be nothing but thinking. The thinking that thinks this thought—​that is what is guaranteed by “cogito.” ’22 Anscombe arrives at this conclusion on the basis of her famous Tank Argument: Let us suppose that [what ‘I’ stands for is some other object than a stretch of a Cartesian Ego]. A plausible one would be this body. And now imagine that I get into a state of ‘sensory deprivation.’ Sight is cut off, and I am locally anaesthetized everywhere, perhaps floated in a tank of tepid water; I am unable to speak or to touch any part of my body with any other. Now I tell myself, ‘I won’t let this happen again!’ If the object meant by ‘I’ is this body, this human being, then in these circumstances it won’t be present to my senses; and how else can it be ‘present to’ me? But have I lost what I meant by ‘I’? Is that not present to me? Am I reduced to, as it were, ‘referring in absence’? I have not lost my ‘self-​consciousness’; nor can what I mean by ‘I’ be an object no longer present to me. This seems both right in itself, and will be required by the ‘guaranteed reference’ we are considering. (FP, 31)

For good measure, we can also assume that Anscombe has lost all knowledge of what she has done. In this situation she can have no way of identifying herself—​if she is a human being, and indeed if she is not. No object, not even substantial Cartesian Ego, is present to her. Yet she can still, as she says, think first-​personal thoughts. She has not lost what she means by ‘I’.

21 

22 

FP, 31. FP, 31.

The First Person and ‘The First Person’     407 Her argument is, I think, completely persuasive if directed against the thesis that ‘I’ is analogous to a demonstrative (or ‘A’, as used in my variation of her parable). But it is completely unpersuasive if we take it as directed against the thesis that ‘I’ is any sort of referring expression, since the underlying presupposition is that ‘I’ can be a referring expression only if it is a proper name, description, or demonstrative. The possibility that ‘I’ is to be grouped with the pure indexicals is not considered. But in order to use ‘here’ I need have no information from my environment. So if ‘I’ is grouped with the pure indexicals, Anscombe’s argument falls flat. Of course, I think that this would not disturb her since I think that the first argument in the paper, the circularity argument, is meant to rule out the possibility that ‘I’ is a pure indexical. But I have claimed that this argument is unpersuasive. So we are back to the beginning. For all Anscombe has said, we can say that what distinguishes ‘I’ from other referring expressions is that it is governed by the rule that for any x, if x uses ‘I’, x refers to x.

18.6 But the game is not yet over. For if we accept the assimilation of ‘I’ to the pure indexicals, we have no reason to insist that ‘I’ has a guaranteed reference. In fact, we need not accept that any expression can, in virtue of its meaning, be such that if x uses it x must refer to x, irrespective of any knowledge x has other than of its meaning. Actually, this point is evident in the cases of ‘here’ and ‘now.’ Where is here? What place? Places are, as Gareth Evans once said,23 just too thick on the ground for there to be an answer unless when I wrote that question I had some identifying description in mind. (I didn’t.) The same is true of ‘now.’ Perhaps ‘I’ should be compared instead with ‘today,’ since as uttered now by me it does pick out a particular period, if we understand it as the period between the previous and the next period of darkness. But ‘today’ is not guaranteed a reference either. It will not necessarily have one uttered at daybreak or sunset—​utterances take time. These remarks will seem obvious and irrelevant since persons are not divisible or composable like places and times. But this misses the main point. Situations are possible in which persons overlap, and in these situations a thinker of an ‘I’-​thought is not guaranteed a reference to itself unless we insist on describing them in ways that have no justification except that they ensure a guaranteed reference for any I-​user’s use of ‘I.’ In this way, I think, Anscombe is substantially vindicated. The first case I have in mind is appealed to by Snowdon.24 People think, and they do so in virtue of their brain activity; damage to the brain will destroy the capacity for

23 Evans, The Varieties of Reference, 169.

24 Snowdon, Persons, Animals, Ourselves, 247.

408   Harold Noonan thought, not so for damage to many other and larger organs. This is an empirical fact, which might not have been so. People can also be severely mutilated, whittled down to not much more than a brain, and continue to be conscious thinking things. Again, this might not have been so. Because of these facts it seems not unacceptable to say that brains think and that they think the thoughts of the people whose brains they are. Some philosophers are reluctant to say that brains think, but there seems to be nothing categorically absurd in saying that my brain is thinking. But if it is, when I am thinking about myself my brain is thinking in a first-​person manner. (How else?) But that does not mean that it is thinking false thoughts about itself. Its thoughts will track the conditions of the animal whose brain it is. It will truly think ‘I am standing up’ if and only if the animal whose brain it is is standing up. It will justifiably think ‘I am sixteen stone’ if and only if the evidence presented to it is that the animal whose brain it is is sixteen stone. So there are two thinkers, but not two objects of first-​person reference. We do not have to say that the brain is thinking falsely that it is a person or an animal. Its assent to ‘I am an animal’ is correct because its reference is to the animal, not itself. Of course, this line of thought can be resisted. But to do so seems a kind of conventionalist sulk. To insist that brains cannot think, given what we know about them and their importance to our mental lives, does not seem to be an insistence on a matter of fact. But if this way of speaking is allowed, it is hard to dismiss this line of thought. For how could two such relevantly indistinguishable thinkers as the person and his brain differ in what they are thinking of? How, as it were, could they direct their attention differently? Another situation which it does not seem wrong to describe as one in which a thinker of an ‘I’-​thought is not the reference of that thought is the symmetrical fission case, familiar from the literature on personal identity. In this fission case, when two hemispheres from one brain are transplanted, according to the ‘no rival candidate’ or ‘best candidate’ neo-​Lockean accounts of Shoemaker25 and Parfit,26 the original person ceases to exist and two new people come into existence, though if either side of the story, as it were, had been the whole story (i.e., only one hemisphere had been successfully transplanted) no one would have ceased to exist. If we reject these accounts because of this, which is to embrace the thought that whether later x is identical with earlier y cannot depend only on facts about x and y (which is a rough statement of ‘the Only x and y principle’; see Noonan),27 we must, with Lewis,28 describe fission in terms of multiple occupancy. Before the fission two persons, two conscious beings, Lefty and Righty, are coincident. They cease to be coincident with the fission but continue to exist and continue to be conscious. Now in this fission case so described, utterances of ‘I’ before the transplant are not guaranteed a determinate singular reference. What is Lefty referring to when he says 25 

Sydney Shoemaker, ‘Persons and Their Pasts’, American Philosophical Quarterly 7 (1970): 269–​285. Derek Parfit, ‘Personal Identity’, Philosophical Review 80 (1971): 3–​27. 27  Harold Noonan, Personal Identity, 2nd edition (London: Routledge, 2018). 28  Lewis, ‘Survival and Identity’. 26 

The First Person and ‘The First Person’     409 (simultaneously with Righty), ‘I am hungry’? He cannot be speaking of himself alone, since he has no way of uniquely identifying himself, no way of directing attention to himself alone. Either he fails to refer or, as Lewis says, he thinks a plural thought, with the content ‘We both . . .’ or ‘At least one of us. . . .’ If he does not know what is going to happen plausibly, he fails to refer.29 The case is essentially no different from the following. I use ‘Tom,’ as I think, to refer to one of my acquaintances. In fact, two identical twins, Dick and Harry, have been fooling me. Hence, if it is known that fission is going to take place, assuming the multiple occupancy story, there is no intelligible doubt that Lefty and Righty can express by uttering ‘I wonder whether I will go left or right’. Of course, it may be objected that the description of the fission case as involving multiple occupancy is just wrong. But this does not seem right. Even if it cannot be shown to be superior to the ‘best candidate’ story, it is hardly to be considered a matter of fact that it is just incorrect. But then it is not just (flatly) correct to say that any utterance of ‘I’ has a guaranteed reference. Just as ‘here’ has no determinate reference when uttered (as it must be) within a multiplicity of overlapping places, so ‘I’ may have no determinate reference when it is uttered by a multiplicity of overlapping persons. Following on from this we can consider the whole cerebrum transplant case. Those who are happy to say that in fission, as a result of the separating chains of psychological continuity, two people (like two intersecting roads) are initially present, can also say, if they wish, that in this case also two people are present, one of whom persists by psychological continuity, the other by bodily continuity—​these are different kinds of continuity, either sufficient for persistence, as Nozick says.30 But since a whole cerebrum transplant is as good as everyday persistence, they can say, given the asymmetry absent from the fission case, the person who persists by psychological continuity (there is only one) can make a determinate singular reference to himself before the transplant. Consequently, they can say, given this asymmetry, it can be said that both thinkers make determinate singular references with ‘I’, though not singular references to distinct individuals (since each thinks all and only what the other thinks). So there is just one reference, and one of the two thinkers is not speaking about himself when he says ‘I’. Again, to emphasise, the point is not that this is the correct description of the case. It is simply that there is nothing to show it wrong. So, once again, insisting on a description that does not involve multiple occupancy just because that guarantees that the reference of any ‘I’-​thought is determinately the single thinker of that thought seems like a conventionalist sulk.

29 

A single token of ‘I’ may be produced by two speakers with different intentions so that two singular references take place. Compare the case, described by Mark Johnston in ‘Hylomorphism’, Journal of Philosophy 103 (2006): 652–​698, in which two people with differing intentions together create the ambiguous road sign ‘Begin Highway’—​one as a name of a highway (‘the [Menachem] Begin Highway’), the other as an instruction (‘Start highway’). This makes sense because we can imagine the sign constructed by a single person with both intentions—​who gets paid twice. But in the fission case Lefty and Righty do not have different reference-​determining intentions before the fission. 30  Robert Nozick, Philosophical Explanations (Cambridge, MA: Harvard University Press, 1981), ch. 2.

410   Harold Noonan My final case is one based on the medical literature, not science fiction. This is the possible case of conjoined twins sharing a cerebrum but not a brainstem nor any other vital organs involved in the life processes thought to individuate organisms.31 It is surely not unacceptable to describe this as a case of conjoined twinning (cephalopagus). As Campbell and McMahan write: That there would be two organisms would be suggested by the fact that they would be separable, perhaps even with technologies that already exist or will exist soon. They might, for example, be separated asymmetrically, with one taking the cranium, the cerebrum, and one each of the cerebella and brain stems, and the other taking only a cerebellum and brain stem and thus requiring an artificial cranium to house them. This would result in two self-​sustaining organisms: one relevantly like a normal person, the other, without a cerebrum, relevantly like a patient in a persistent vegetative state whose cerebrum has been destroyed but who could remain biologically alive with little external support other than hydration and nutrition. Alternatively, these hypothetical cephalopagus twins might be divided symmetrically, with each taking . . . one cerebral hemisphere. Each would be like a patient who has received hemispherectomy.32

It does not seem incorrect to say that in this case there are two thinkers, each of which is thinking ‘I’-​thoughts, but it does not seem plausible to say that either thinker will be able to make singular reference to itself or know that it was one rather than the other. Campbell and McMahan prefer to describe the situation as one in which there are three individuals: two non-​thinking organisms and one person, who is made up of the matter in the consciousness-​generating area of the brain (which goes with the view that, strictly speaking, we are proper parts of organisms—​functional brains, not organisms, in fact, are thinkers). Another view is that of the animalist. There are two thinkers, since two animals, each of which, accepting the guaranteed reference for ‘I’, is thinking a singular thought about itself when they simultaneously think an ‘I’-​thought (like, they must say, pace Lewis, Lefty and Righty in the fission cases), which may be a true thought about only one of them. Given these alternatives, as I said, it does not seem implausible to say that ‘I’ in this case, like ‘here’ in almost every situation, lacks a determinate singular reference.

18.7 I conclude that the thesis against which Anscombe’s paper is directed, that ‘I’ is a device of guaranteed self-​reference, is not the truism most of her opponents assume. ‘I’ is 31 

W. Metz, Ultrasound in Obstetrics and Gynaecology (New York: Thieme Medical Publishers, 2001), 289–​290; see also Tim Campbell and Jeff McMahan, ‘Animalism and the Varieties of Conjoined Twinning’, in Animalism, Stephen Blatti and Paul Snowdon (eds.) (Oxford: Oxford University Press, 2016). 32  Campbell and McMahan, ‘Animalism and the Varieties of Conjoined Twinning’, 248.

The First Person and ‘The First Person’     411 a referring expression in that it can be used to make a singular reference, and is sometimes a device of self-​reference. But it is not guaranteed to be either of these things. Its linguistically competent and faultless use by a thinker may involve a singular reference to another thinker, or no determinate reference at all. The successful use of ‘I’ need not involve the connection of what is understood by a predicate with a distinctly conceived subject.33

Bibliography Campbell, Tim, and McMahan, Jeff. ‘Animalism and the Varieties of Conjoined Twinning’. In Animalism, Stephan Blatti and Paul Snowdon (eds.), 229–​252. Oxford: Oxford University Press, 2016. Doyle, James. No Morality, No Self: Anscombe’s Radical Skepticism. Cambridge, MA: Harvard University Press, 2018. Evans, Gareth. The Varieties of Reference. Oxford: Oxford University Press, 1982. Johnston, Mark. ‘Hylomorphism’. Journal of Philosophy 103 (2006): 652–​698. Lewis, David. ‘Survival and Identity’. In The Identities of Persons, Amelie Oksenberg Rorty (ed.), 17–​40. Berkeley: University of California Press, 1976. Metz, W. Ultrasound in Obstetrics and Gynaecology. New York: Thieme Medical Publishers, 2001. Noonan, Harold. Personal Identity. 2nd edition. London: Routledge, 2018. Nozick, Robert. Philosophical Explanations. Cambridge, MA: Harvard University Press, 1981. Parfit, Derek. ‘Personal Identity’. Philosophical Review 80 (1971): 3–​27. Shoemaker, Sydney. ‘Persons and Their Pasts’. American Philosophical Quarterly 7 (1970): 269–​285. Snowdon, Paul. Persons, Animals, Ourselves. Oxford: Oxford University Press, 2014. Strawson, Peter. ‘Reply to John McDowell’. In The Philosophy of P. F. Strawson, Lewis Hahn Rorty (ed.), 146–​150. LaSalle, IL: Open Court, 1998. van Inwagen, Peter. ‘ “I Am Elizabeth Anscombe” Is Not an Identity Proposition.’ Metaphysica 1, no. 2 (2001): 5–​8. Wiseman, Rachael. ‘What Am I and What Am I Doing?’ Journal of Philosophy 114, no. 10 (2017): 536–​550.

33 

FP, 36.

Pa rt V

A N S C OM B E ON /​A N D OT H E R P H I L O S OP H E R S

Chapter 19

Ansc ombe’s Wi t tg e nst e i n Joel Backström

19.1  Introduction Elizabeth Anscombe first met Ludwig Wittgenstein when attending his lectures at Cambridge in 1942. The teacher-​student relationship evolved into a close friendship lasting until Wittgenstein’s death in 1951, with Wittgenstein for a time living in an upstairs room in Anscombe’s house.1 As one of three literary executors named in Wittgenstein’s will, Anscombe played a key role in presenting his thought to the world by bringing out editions of his unpublished writings, starting with the Philosophical Investigations in 1953. Furthermore, as the translator of many of these works, originally written in German, Anscombe gave Wittgenstein his English writer’s voice. When reading Wittgenstein in English we are, very often, reading Anscombe. This chapter doesn’t discuss these aspects of the Anscombe-​Wittgenstein relationship, however.2 Instead, it examines two closely intertwined themes: Anscombe’s understanding of Wittgenstein’s thinking and his influence on hers. I begin by sketching Anscombe’s general sense of Wittgenstein’s thought and of his influence on her and her view of his philosophical ‘method’, which she doesn’t want to call quite that (sections 19.2 and 19.3). Then I trace his pervasive influence in her writings on intention, the first person, and certain problems about meaning, viewing these as unified by a Wittgensteinian critique of the subject-​object paradigm of thought (sections 19.4 and 19.5). After showing the connection between Anscombe’s seminal paper ‘Modern Moral Philosophy’ and Wittgenstein’s diagnosis of metaphysical illusions, I close by discussing Anscombe’s and Wittgenstein’s respective approaches to the question of

1  On the personal relationship, see Monk (1991) (see ‘Anscombe’ in the index); cf. also Geach (1988); Leach (2020, 5–​6, 9–​10). 2  On Anscombe’s work as literary executor, see Backström (forthcoming), Erbacher (2016); Erbacher and Krebs (2015); Gibson (2019).

416   Joel Backström truth in religion, which may in some respects not be as far apart as might initially appear (sections 19.6 and 19.7).

19.2  A Veiled Face According to Anthony Kenny (2019, 19), Anscombe once said, ‘I don’t have a single idea in my head that wasn’t put there by Wittgenstein’. A striking statement, but hardly surprising. Any philosophically inclined person who spent years talking with someone like Wittgenstein would unavoidably have their thinking deeply and indelibly formed by his. Anscombe’s daughter Mary Geach reports her mother saying that, really, neither she nor her contemporaries ‘who called themselves philosophers were . . . philosophers’; when asked who was, ‘she named Wittgenstein’ (GG3, xiii). Anscombe herself writes of an initial phase of ‘besotted reverence’, when ‘almost anything that Wittgenstein said sounded true and important to me, except for some things about religion’: But it is not entirely true, since I argued with him a good deal in philosophy. Sometimes I plain did not understand what he was saying, and then I more or less had to leave it alone; sometimes I understood his sentences but could not understand the reasons for them and that was where I had the most instructive arguments of my life. But once I understood I was inclined to be convinced. (AW, 35)

This kind of ‘overwhelming influence . . . which makes something seem convincing because it is part of X’s teaching’ is a bad thing, Anscombe says, for then one isn’t really thinking for oneself, and she ‘soon began to struggle against it’, telling herself, ‘I had no right to think something which impressed me in this way; I had got to be able to think whatever I could really claim to think in, so to speak, a dull way, and with reasons which I could give and which ought to be visible as having point and force to minds that had not suffered that great impression’ (AW, 35–​36). This may explain Anscombe’s free use of Wittgensteinian ideas without explicit acknowledgement of their source; taking her debt to Wittgenstein for granted, she tries to understand the ideas themselves, make their significance clear by putting them to work. As we’ll see, Wittgenstein’s influence is indeed evident across her writings, in matters great and small—​and also in the way she writes about other philosophers. Volume 1 of her Collected Philosophical Papers is subtitled From Parmenides to Wittgenstein, but the order of development in her own thinking appears to have been the reverse. ‘It is a very striking fact about Wittgenstein’s thought’, she wrote in a letter to G. H. von Wright, ‘that he reverts to problems of Greek philosophy, and one of the things for which I am grateful to him is that he has caused me to read Plato and Parmenides with more understanding’ (LvW, 1948). As she told Mary Geach, the past masters had seemed to her like ‘beautiful statues’, but the vitality of Wittgenstein’s thought ‘had brought them alive for her’ (GG3, xiii).

Anscombe’s Wittgenstein   417 While Anscombe had learned from Wittgenstein, he remained an enigma to her. In a letter to Paul Engelmann in 1958, she writes: I must confess that I feel deeply suspicious of anyone’s claim to have understood Wittgenstein. That is perhaps because, although I had a very strong and deep affection for him, and, I suppose, knew him well, I am very sure that I did not understand him. It is difficult, I think, not to give a version of his attitudes, for example, which one can enter into oneself, and then the account is really of oneself: is for example infected with one’s own mediocrity or ordinariness or lack of complexity. (Engelmann 1968, xiv)

Anscombe often expresses this sense of not really having understood Wittgenstein the man or his thinking. For example, about Philosophical Investigations §108, where Wittgenstein discusses the central difference between the Tractatus and his later understanding of philosophy, and says the task is ‘turning our whole examination round . . . about the fixed point of our real need’, Anscombe asks, ‘But what . . . did he mean by ‘the fixed point of our real need’? I do not know, and I suspect that without understanding this we shall at best have a poor understanding of that book [the Investigations]’ (KE, 211; cf. ST, 179). Speaking of the contrast between the impression made by the Tractatus and the Investigations, she says: [F]‌rom the Tractatus as it were a face looks out very clearly. With Philosophical Investigations I at least have the impression of a veiled face, or one which does not appear strongly. Strange contrast! In the Tractatus it is said that what is shown cannot be said—​there is something that our sentences try to speak out but gulp on. Yet the Tractatus seems to succeed in saying what, according to it, ‘can’t be said’. The Investigations insists that it is an error to think that there is something that can’t be said, and yet it seems often to be nearly revealing something which yet does not come into view. (FWW, 192)

Anscombe says that she once heard someone ask Wittgenstein what it all came to, what was, so to speak, the upshot of the philosophy he was teaching in the 1940s. He did not answer. I am disposed to think that there wasn’t an answer he could give. That, namely, he did not think out a total position as in writing his first book; that, rather, he was constantly enquiring; some things he was pretty sure of, but much was in a state of enquiry. I therefore deprecate attempts to expound Wittgenstein’s thought as a finished thing. . . . Predictions of ‘what Wittgenstein would say’ about some question one thought of were never correct. (LW, 169)

This passage need not imply that Anscombe thought Wittgenstein should have laid out a ‘total position’ or that he was trying but failed to do so. In fact, he disputes the very idea of taking ‘positions’, ‘total’ or ‘local’, in philosophy (cf. PI, §§118, 126–​129), as also the idea

418   Joel Backström that philosophy has discreetly demarcated ‘topics’, treatable in isolation.3 In philosophy, one must learn to find one’s way through a whole landscape of life and confusion; ‘the very nature of the investigation’, Wittgenstein says, ‘compels us to travel over a wide field of thought criss-​cross in every direction. . . . The same or almost the same points [are] always being approached afresh from different directions’ (PI, preface). Thus, ‘[p]‌roblems are solved (difficulties eliminated), not a single problem’ (PI, §133). The individual problems aren’t of interest just in themselves, however, but because ‘each one casts light on the correct treatment of all’ (Z, §465); the aim is to ‘see connexions’ between them and to get an ever clearer overview of the whole (PI, §122). Rather than searching for a ‘position’ to occupy, then, Wittgenstein seems to show us ways of moving about thinkingly in the landscape of our life and understanding. In his classes, Anscombe says, he ‘sometimes said he was . . . giving examples of ‘five-​finger exercises’ in thinking’ (LW, 169).

19.3  Against Method Anscombe’s frank acknowledgement of not really understanding Wittgenstein makes one suddenly aware of the strangeness of the self-​certainty amongst many of his commentators, who tend to convey the impression that they understand him, or at least what they present as the essence of his teaching, perfectly. Yet such claims to understand are often misunderstandings—​as one can say even without any claim to a perfect positive understanding of one’s own. In Anscombe’s view, the logical positivists certainly misunderstood the Tractatus, which they thought they followed. In her 1959 book An Introduction to Wittgenstein’s Tractatus, she tried to show that ‘empiricist or idealist preconceptions’, and the focus on epistemological questions characteristic of modern philosophy generally, were ‘a thorough impediment to the understanding’ of that book, which is rather concerned with the problems of meaning that exercised Frege and, before him, Plato and other ancient philosophers (IWT, 12–​13/​GG4, 4–​ 5).4 Anscombe thought the positivists’ misunderstanding of early Wittgenstein was paralleled in ‘the indirect influence of Wittgenstein’s later work . . . through derivative schools such as the Oxford “linguistic school,” ’ which misunderstood it.5 Both schools

3  Anscombe, however, seems at times to have been drawn to a ‘separate topics’ approach to philosophy, including in her work as editor of Wittgenstein, most notably in her view of the material she edited as On Certainty in 1969 as ‘a single treatise on a single topic’ (LvW, September 13, 1967); a view contested by her fellow literary executor Rush Rhees. For further discussion, see Backström (forthcoming), and cf. Anscombe’s apparently somewhat different later assessment of On Certainty, quoted in section 19.7. 4  Space doesn’t permit discussing the details of Anscombe’s hugely influential but also criticised Tractatus interpretation. This is the aspect of her Wittgenstein understanding that has been most discussed in the literature; indeed, one might almost say, the only aspect that has been discussed in any depth. Diamond (2019) is a recent book-​length study. 5  Anscombe to Chadbourne Gilpatric, October 18, 1955, quoted in Leach (2020, 14).

Anscombe’s Wittgenstein   419 tended to believe that Wittgenstein had provided a simple method for making philosophy as hitherto known obsolete: the ‘verification principle’ in the positivists’ case; the idea that ‘meaning is use’ and that philosophical difficulties can be cleared up by simply describing the ordinary use of linguistic expressions in the case of ordinary language philosophers. The Tractatus, especially TLP 6.53, suggested to the positivists a ‘quick and easy way of dealing with ‘metaphysical’ propositions’ by asking ‘what sense-​observations would verify and what falsify them? If none, then they are senseless’ (IWT, 150/​GG4, 125). The problem with this reading of the Tractatus, Anscombe observes, is that there’s ‘nothing about sensible verification’ in the book, and no ‘suggestion of a general method for criticizing sentences’ in the verificationist way (IWT, 150/​GG4, 125).6 ‘The general method that Wittgenstein does suggest’ for such criticism is, she says, ‘that of “shewing that a man has supplied no meaning [or perhaps: “no reference”] for certain signs in his sentences” [TLP, 6.53]’, and such criticism ‘could never be of any very simple general form; each criticism would be ad hoc, and fall within the subject-​matter with which the sentence professed to deal’ (IWT, 151/​GG4, 125–​126). Anscombe illustrates what this might mean with an example ‘from Wittgenstein’s later way of discussing problems’, an example Tom Stoppard later used in his play Jumpers: [Wittgenstein] once greeted me with the question: ‘Why do people say that it was natural to think that the sun went round the earth rather than that the earth turned on its axis?’ I replied: ‘I suppose, because it looked as if the sun went round the earth’. ‘Well’, he asked, ‘what would it have looked like if it had looked as if the earth turned on its axis?’ (IWT, 151/​GG4, 126)

This exemplifies Wittgenstein’s genius for staging surprise attacks that, through a sudden reversal of perspective, reveal his interlocutor’s preconceptions and blind spots; his comments could, as Anscombe puts it, ‘open the ground up before one’s feet, quite suddenly’ (AW, 6). It also illustrates how easily we confuse our everyday ways of speaking with a quasi-​theory, which may then need correction by scientific theory. To say that the sun moves across the sky is unproblematic as long as we speak non-​theoretically, but this doesn’t licence the move to a geocentric cosmology, for if we view both the earth and the sun as ‘heavenly bodies’ related to each other in a system of such bodies, the sun’s ‘movement’ could equally be an appearance caused by the earth’s movement as vice versa. More generally, the example illustrates how philosophical difficulties arise not from explicit theoretical commitments but, as Wittgenstein says, from our having committed ourselves without realising it to ‘a particular way of looking at the matter’ (PI, §308), which then leads us into bewildering dead ends—​or again, as in this case, makes what isn’t clear at all seem plain and obvious. 6  Anscombe records that, ‘hearing people speak of “the verification principle”, [Wittgenstein] exclaimed “Who is supposed to have invented that?” and when it was indicated that it was attributed to him, he said “What? Me?” ’ (KE, 208). Cf. the statements by Wittgenstein quoted in Klagge (2019, 17–​18).

420   Joel Backström In 1953, during the heyday of ‘linguistic analysis’, Anscombe remarked on how citing the supposed ‘ordinary use’ of expressions, a habit people had picked up from later Wittgenstein, played a rhetorical role similar to references to ‘verification’ for the positivists. She thought this confused: [T]‌he objection ‘but the word is not used like that’ . . . which has acquired such dogmatic force in philosophical discussion . . . ought . . . never to be used unless it is clearly shewn how, for the particular problem in hand, some particular feature of a use is essential to a meaning. . . . Wittgenstein . . . is always saying ‘look at the use!’ But it is in fact a difficult thing to do; and it is very easy to think that one is doing it by attending to unimportant features: a procedure which he once compared to describing a naval lieutenant in terms of the stripes on [his uniform.] (BBC, 235)7

Against both positivism and ordinary language philosophy, Anscombe insists that Wittgenstein didn’t provide easy solutions for philosophical perplexity. On the contrary, he ‘made one feel that it was very difficult to look in the right direction’; he ‘often seemed to understand one’s philosophical thoughts and problems better than one did oneself. One would say what one thought, then he would amplify it, make it seem more convincing, carry it deeper—​and then undo it’ (AW, 13, 7). While Wittgenstein’s contribution does lie in his way of doing philosophy, no real ‘method’ can be extracted from his writings or teaching; It is quite commonly said that Wittgenstein taught, not a doctrine, but a method of doing philosophy. This is unluckily in an important sense not true. If someone teaches a method, it ought to be possible for someone else both to learn and to teach it: it ought to be quite clear what the procedure is, what moves you make at what points, and so on. Now it is possible perhaps to list a number of tricks which Wittgenstein used and, therefore, taught. E.g., that of asking ‘As opposed to what?’ or ‘and what would it be like if it were not so?’ when a proposition is advanced. Or that of asking ‘What is the picture that is being used here?’ Or that of taking a solution in a peculiarly literal, empirical sense. Or that of inventing different cases which shew quite clearly that the implication of a term or the meaning of a statement is not determined in advance in some way in which one thought it was. Or that of asking ‘To whom is this said, and in what circumstances?’ or ‘What kind of proposition is this?’ Or that of assuming a criterion of identity of a kind suggested by a philosophic thesis and deducing absurd consequences. Or that of asking what would shew that a word had a certain sense if you had to learn the language it belonged to without interpreters. I could go on. But none of these moves, nor all of them taken together, guarantee that anybody will find a solution to any problem at all. There is no method taught by which you know when such and such a move will be fruitful, carry you 7  This relates to Anscombe’s impatience with J. L. Austin’s brand of linguistic analysis, evident in her response (P) to his piece on pretending. Kenny (2019, 17) says that Anscombe and Austin ‘detested each other cordially’; ‘she . . . would say scornfully that she expected any day that some student would report . . . that Austin had discovered a difference between “enough” and “sufficient.” ’

Anscombe’s Wittgenstein   421 deeper, cause you really to touch the nerve of the problem under consideration. These tricks can be played with complete superficiality. Nothing takes the place of having ideas, of being capable of observation and insight; and Wittgenstein did not I think teach a method of attaining these. When he makes one of these moves it has great point; but he does not teach you when it will have point to make a given move. (BBC, 234)8

When put plainly like this, the point is obvious. Of course philosophical problems cannot be clarified by applying set formulas, as it were feeding ‘the problem’ into a machine—​ and there is a problem to clarify only insofar as someone has a problem, as machines don’t. And yet, the point is constantly missed, as illustrated by the positivists and the ordinary language philosophers. One could also say: the reason why there cannot be a method in philosophy is that, as we saw, philosophical difficulties are created by commitments to ways of thinking of which we aren’t aware and that, Wittgenstein suggests, we don’t even ‘want to see’, which is why ‘what is most obvious may become the hardest of all to understand’ (CV, 25, transl. modified). Methods can only solve problems you’re willing to admit you have, but in philosophy, the hard thing is precisely admitting this: ‘It is not a difficulty for the intellect but one for the will that has to be overcome’ (CV, 25). We’ll return to this crucial point.

19.4  The Subject-​Object Paradigm Let’s now turn to Wittgenstein’s substantial influence on Anscombe’s thinking. In order not to get lost in a mass of details, I will consider this under one particular aspect: a reorientation of the approach to philosophical problems away from the dominant subject-​ object paradigm of philosophical thought, and towards a focus on what happens between human beings. I see that reorientation as the most radical aspect of Wittgenstein’s later work, although it isn’t part of standard interpretations, or anyway isn’t given the sense I’ll try to indicate, and Anscombe never explicitly formulates it as I will. Nevertheless, I’ll try to show that it unifies central themes in her writings. The dominant conception Wittgenstein criticises sees questions in terms of a subject’s relation to an object of some kind, and so effectively models our relation to ourselves and to each other on our relation to things. Thus, it populates the world with thing-​like mental, logical, and other ‘entities’, the presence of and ‘facts’ about which supposedly account for the most basic dimensions of human life. In fact, the subject-​object framing makes these dimensions disappear from view or appear only in incomprehensibly paradoxical form. Anscombe’s writings on personal identity and intention and on private ostensive definition and certain problems about meaning—​three central topics of hers that I’ll focus on—​show just 8 

All quotation marks within the quote have been changed to double; in the original they are first double, then single, for no obvious reason.

422   Joel Backström this, or at least that’s one way of seeing them as belonging together. Hopefully, looking at Anscombe’s thinking in this light also helps get this theme in Wittgenstein into clearer focus. Subject-​object thinking dominates philosophical debates about ‘the self ’. ‘The subject’ itself is taken to be some kind of object—​’Being able to mean ‘I’ is . . . explained as having the right sort of thing to call ‘I’’ (FP, 25)—​and then failure to find this imagined object gives rise to scepticism about personal identity. As Anscombe notes in ‘The First Person’, the debate is ‘self-​perpetuating . . . so long as we adhere to the initial assumption . . . that ‘I’ is a referring expression’ (FP, 32). That assumption is shared both by those who ‘have not perceived the difficulty’ with whatever is proposed as the supposed self-​object referred to (a body, a Cartesian ego, some properties of a brain, etc.), and those who have, ‘and are led to rave in consequence’ (FP, 32). The debate is dissolved, Anscombe says, by the realisation that when I say that I’m doing, feeling, or thinking this or that, ‘I’ is not used ‘to make a reference, at all’ (FP, 32). I’m not somehow identifying my ‘I’ (or ‘self ’) as the object referred to; if there were such an act of identification, I might mistakenly refer to your ‘self ’ instead of mine—​but that’s nonsense. As Anscombe notes, the very fact that ‘getting hold of the wrong object is excluded . . . makes us think that getting hold of the right object is guaranteed’ by the extraordinary nature of the object (a Cartesian ego, say), when really ‘there is no getting hold of an object at all’ (FP, 32). Anscombe doesn’t mention this in her paper, but her basic argument can be found in Wittgenstein, perhaps best expressed in the Blue Book (BB, 66–​70; cf. PI, §§404–​411; WL46/​7, 47). Obviously, one can mistake oneself for another or another for oneself in a certain sense. Thus, I might mistakenly think a childhood photograph of you shows me as a child. But in making the mistaken identification, pointing to the photograph and saying ‘That’s me!’, I don’t first have to identify myself as the speaker, I simply speak. As Wittgenstein put it, ‘The man who cries out with pain, or says that he has pain, doesn’t choose the mouth which says it’ (BB, 67–​68; cf. Anscombe’s FP, 29). This reveals the radical difference between the two senses of self-​relation in play. The relation to myself manifest in my speaking, which is eo ipso a relation to others, those I address in speaking, is not the relation to any object, not to my ‘self ’, my ‘mind’, or my ‘body’, conceived ‘objectively’. Moving my tongue to speak isn’t like moving an object; intending to move my tongue, I cannot inadvertently move yours, as I might inadvertently move your car instead of mine. Anscombe puts the point by noting that while ‘This is my body’ might be said to mean that ‘My idea that I am standing up [say] is verified by this body, if it is standing up’, and while observation of my body may, in special circumstances—​I may be under the influence of hypnotic suggestion, say—​reveal that I was mistaken to think I was standing up, ‘observation does not show me which body is the one [is mine]. Nothing shows me that’ (FP, 34, emphasis added). For how do I know that that very body is the one for me to observe, the one whose position in this case falsifies my sense that I am standing up? And, again, do I first, in order to make the observation, have to decide which eyes (namely: mine) to direct at that body? When I act, whether physically or in thinking something to myself, my basic relation to myself, including to my body, is not that to an object at all.

Anscombe’s Wittgenstein   423 This brings us to the topic of intention. In a 1954 letter to von Wright, Anscombe acknowledges Wittgenstein as the source of the most basic idea of her best known book, Intention (1957). She writes that reflecting on ‘the discussion of ‘Tun’ and ‘wollen’’ (doing and willing) in §§611ff. of Philosophical Investigations, ‘together oddly enough with’ her reading Aristotle’s Ethics, ‘has got me on to a . . . very interesting line about ‘how does one know what one is doing?’—​when e.g. one is writing a letter’; ‘I want to say that I know what I do as I know what I say. That is, not from the facts at all’ (LvW, just before Easter 1954). The ‘connexion with Aristotle’, she explains, was through his notion of practical reasoning, whose conclusion ‘isn’t made true by the considerations leading to it’ but ‘by my doing whatever it is’, which means that ‘the truth in question’ is not ‘something shewn to the intelligence that arrives at it’ (LvW, just before Easter 1954). We don’t conclude that we’re doing X from observation of any kind, introspective or otherwise, nor from an acceptance of the premises of our deliberative argument. Rather, insofar as we deliberate about what to do—​in contrast to simply acting, without deliberation, as we often do—​our resolving to act is itself the ‘conclusion’. And, of course, even theoretical reasoning is a practical activity, a mode of action; one isn’t observing conclusions somehow forming, but draws them.9 Saying this would be trivial were it not, again, for the domination of subject-​object thinking and the ‘utter darkness’ into which we’re plunged by what Anscombe calls its ‘incorrigibly contemplative conception of knowledge’ (I, 57). This presents an intention as a kind of quasi-​object—​a mental or brain state—​from recognising or registering which I know what I intend. Here, as in the case of the self, the object-​conception engenders the absurd consequence, pointed out by Wittgenstein, Anscombe notes (I, 6), that it would be possible to mistake one’s intention for some other, in the way one can make a mistake in executing it. I can mistakenly pick up your book instead of mine, but I cannot mistakenly think I intend to pick up mine when I really intend to pick up yours.10 Following Wittgenstein, Anscombe reorients the discussion of intention. One cannot see an intention or intentional action by looking ‘into the contents of [the agent’s] mind’, at ‘the movements of muscles or molecules’, or at anything else that ‘can be determined about the man by himself at the moment [of action]’ (I, 9, 29). If contemptuous words go through my mind as I greet someone, for example, this doesn’t by itself show that my greeting is secretly contemptuous, unless I mean those words, silently direct them at the other as signs of contempt (I, 48–​49). In intending, as in remembering, imagining, etc., at issue is not what passively, object-​like, is in my mind, but what I actively, with 9 In Intention, Anscombe discusses these points, e.g., at pp. 49–​54 (non-​observational knowledge) and 56–​89 (practical knowledge/​reasoning). 10  Wittgenstein brings out the absurdity in the ‘observational’ conception of knowledge of one’s intention (e.g., in imagining) by asking: ‘Would you say: “I see a man with white hair etc., I suppose I’m imagining N but perhaps it’s only someone who looks very much like him”?’ (PO, 455). Of course, one can be confused or self-​deceived about one’s intentions, but what one then self-​deceptively refuses to acknowledge, or is simply confused about, is something one knows without observation, with the mediation of no thing. And being confused is different from making a mistaken identification; it’s more like a general breakdown of orientation (cf. the discussion of poor confused ‘Baldy’ in FP, 36).

424   Joel Backström understanding, do. If there is something in my mind, say, something that somehow ‘feels like’ a memory-​image, I still have to judge that things were as the image represents them for this to be remembering. But, Anscombe asks, echoing Wittgenstein, ‘if one can judge “Things were as this represents them,” why cannot one simply judge “Things were thus and so”—​for example, “Jones was there”?’ (MEC, 126) Thus, there’s no need for an image or ‘[some other] intermediary to have, as it were, the peculiar colour of memory’, and since, even were it there, it wouldn’t make a memory, the ‘object-​like’ intermediary ‘drops out of the analysis, and the causality . . . is between the original witnessing of the event and the present thought . . . of [the person who] knows that such-​and-​such occurred’; this, Anscombe says, ‘is an original phenomenon of causality: one of its types—​whether or not anyone has yet classified it as such’ (MEC, 126–​127; cf. Wittgenstein’s remarks in, e.g., Z, §610; PI, §213). However, the dominance of subject-​object thinking makes this causality—​a human being who tells us what she remembers, with no thing intervening or deciding this for her—​appear impossible, and the fact that someone remembers something immediately gets reinterpreted as a matter of finding ‘a definite thing that happens’ somehow in her; whether in her mind or brain, doesn’t matter (MEC, 128, emphasis added). My action in forming an intention, remembering, saying something, etc. doesn’t consist in some ‘purely interior’ act of meaning or intending, to which my ‘outer’ behaviour would be only contingently ‘attached’, an idea whose confusion Anscombe, again following Wittgenstein, ably demonstrates, and to which we’ll return (cf. I, 28–​29, 42–​44, 47–​49). It is rather a matter of how I concretely relate myself to others in the situation where I act. This is the import of Anscombe’s oft-​quoted positive characterisation of intentional actions as those to which ‘a certain sense of the question ‘Why?’ is given application’; the sense, namely, ‘in which the answer, if positive, gives a reason for acting’ (I, 9). This means that intentions and intentional actions exist only in the context of our living with, wondering about, objecting to, challenging and explaining ourselves to each other. This isn’t to deny the existence or importance of ‘inner’ occurrences, such as mental images or words that one thinks of without uttering, but to say what—​namely, their setting within these interpersonal relationships—​gives these inner occurrences, or for that matter any ‘outward’ performances, their significance (cf. I, 48–​49). We don’t have intentions, and then also happen to ask each other about what they are; rather, to have intentions in the full human sense is to participate in this life of mutual exchange and questioning.11 In Anscombe’s words, while it is true that we have a ‘special interest in human actions’, one can characterise what that interest is only by giving ‘a type of description that would not exist if our question ‘Why?’ did not’ (I, 83; cf. I, 34). Crucially, we question and answer each other not just out of curiosity, or simply for purposes of explanation or prediction, but out of an essentially moral concern; a

11 

Cf. Teichmann (2015). On Anscombe’s view of animal intentions, see I, 86–​87; CP2, 91; UD, 209–​210.

Anscombe’s Wittgenstein   425 concern with ‘good and evil’ (I, 21–​24), with our responsibility for our actions, to each other. ‘All human action is moral action’, Anscombe says, hence ‘moral’ does not stand for an extra ingredient which some . . . actions have and some do not’ (AIDE, 209). There are morally indifferent action-​descriptions—​if we’re told only that someone sat down, say, this doesn’t suggest anything for good or ill—​but considering the particular action in concreto means seeing it in an interpersonal context of moral concern, where what it meant or implied for good or ill is revealed (AIDE, 210, 213–​214). For example, the one who sat down may have done so when she should have (concretely) stood up to someone, so hers was an act of cowardice or resignation. And in cases where the sitting down really was indifferent, that too can be seen only in the light of moral concern, which reveals acts as significant or indifferent. Anscombe underlines the essentially interpersonal character of this concern: ‘the single human individual’ in relation to whom one acts, e.g., who suffers the consequences of one’s action, ‘occupies a pre-​eminent position in questions about good and bad action’ (AIDE, 218).12 Anscombe’s remark that ‘moral’ doesn’t denote ‘an extra ingredient’ of (some) actions, but is rather the very element in which action has its being, makes explicit the basic insight that orients Wittgenstein’s discussion of our, to use an inadequate term, ‘psychological concepts’—​and in fact, it seems to me, his whole later philosophy. Although he doesn’t name it explicitly using the concept ‘moral’, this essentially moral orientation is the very reorientation away from subject-​object thinking and towards the interpersonal dimension. To see how, let’s briefly consider Wittgenstein’s way of discussing pain, which aims to show that our recognition (understanding) of pain cannot be conceived as, basically, an epistemological matter in which we determine whether a mental ‘thing’, pain, is present in the other, with ‘moral concerns’ entering only secondarily, in our caring (or not) about the pain we’ve registered. Wittgenstein turns things around: the ‘most primitive form’ of recognition of the other’s pain is ‘a reaction to somebody’s cries . . . a reaction of sympathy. . . . We comfort him, try to help him’ (PO, 381). As he says, if someone has a pain in his hand, ‘one does not comfort the hand, but the sufferer: one looks into his face’ (PI, §286). To focus, in one’s theoretical search for the essence of pain, on ‘the pain’, that ‘object’ one imagines is there to be registered in the sufferer’s mind or brain, is like focusing on the bruised hand, disregarding the person whose hand it is. Bandaging the hand or dampening the pain by administering painkillers is indeed of interest to us—​but only given that we care about and pity the sufferer. This is moral concern insofar as it is precisely a caring-​for-​the-​other. This caring concern creates a ‘field’ of possible, morally charged responses: compassion, cruelty, negligence, etc. In the absence of that

12 

In light of the ineliminably moral character of ascriptions of intention and other mental states, Anscombe’s call, in other places, for developing an adequate ‘philosophy of psychology’ (which for her centrally includes ‘theory of action’) before moving to ethics (I, 78; MMP, 26, 38), seems clearly untenable. That is, it can amount only to an arbitrarily drawn line between philosophical reflections, such as her remarks on taking pleasure in cruelty, or on the need to see what one intends as in some sense good (I, 73, 75), that, while manifestations of moral understanding, are not called ‘ethics’, and others that are (cf. I, 75–​76).

426   Joel Backström concern, why would the other things (bandaging, etc.) be of any interest at all—​or if they were, what would show this to be an interest in pain? Now Wittgenstein formulates an apparent objection to his way of discussing pain: ‘Yes, but there is something there all the same accompanying my cry of pain. . . . And this something is what is important’ (PI, §296). In response, he agrees that there’s no greater difference than that between being in pain and just pretending, behaving as if one were, and so naturally he doesn’t want to say that there’s ‘nothing there’, no pain. So he says the pain ‘is not a something, but not a nothing either’; the point being that we cannot escape such paradoxical formulations until we ‘make a radical break’ with the idea that we always think and speak about objects of some kind or other (PI, §304). We don’t just refer to objects. More basically, we respond to each other in the various modes of moral concern, whether in comforting the sufferer or asking the one who made her suffer, ‘Why?’13

19.5  Closing the Myth-​M aking Factory The protest ‘Yes, but there is something there all the same’ (PI, §296) voices the most primitive conviction of subject-​object thinking—​although this doesn’t yet tell us why that conviction arises, what is tempting about it. The conviction doesn’t just concern pain; at stake is the feeling that, as Anscombe says, my experience ‘gives me, shows me, an object’, and that [s]‌uch . . . objects are my sole direct cognitive connection with the world. If it is a pain, for example, it acquaints me indirectly with my body; it itself is an object of direct acquaintance. ‘There is something there!’ I want to say. Such objects are not independent of the experiencing subject, but given that they are being experienced, they exist . . . [and] all my knowledge of empirical reality depends on my having these objects. (POD, 249)

One of the first great philosophical lessons Anscombe learned from Wittgenstein—​ supposing it makes sense to think of its being permanently learned—​was to see through this idea, or fantasy. She was captive to it in the form of phenomenalism, which she ‘hated . . . [but] felt trapped by’; she couldn’t see how to get, say, from seeing a yellow expanse (which seemed undeniable) to knowledge that it was a pack of cigarettes—​’I couldn’t see my way out of it but I didn’t believe it. It was no good pointing to difficulties about it. . . . The strength, the central nerve of it remained alive and raged achingly. It was only in Wittgenstein’s classes in 1944 that I saw the nerve being extracted, the central 13  For more on the essentially ethical, interpersonal or ‘I-​you’-​perspective arguably at the core of Wittgenstein’s later philosophy, and of human understanding as such, see Nykänen (2018) and Backström (2019a).

Anscombe’s Wittgenstein   427 thought ‘I have got this, and I define ‘yellow’ (say) as this’ being effectively attacked’ (CP2, viii; cf. POD, 244). This myth of private ostensive definition can be ‘read’ from the subjective or the objective side, either ‘I know what I mean’ or ‘the object tells me what it is’; the amalgamation of these two sides of the subject-​object perspective creates the paradoxical idea of the private, i.e., subjective object. What Wittgenstein shows is that the myth is indeed a myth, a confusion.14 As Anscombe puts it, it can seem to me that I know what my object of experience is simply by privately attending to and ‘naming’ it (say) ‘A’, only because I ‘already know what an A is, or what sort of thing is to be called an “A” ’ (POD, 256). For there’s no such thing as attending to a mere, undefined ‘this’; you may attend to a sound, say, and name it, but then you have already decided that you’re attending to a sound, and you can call that sound ‘A’ only if you have ‘a technique of using “A,” ’ a way of determining what ‘that’ sound is, what else, if anything, is to be called the same sound: [But h]ow can this that I am attending to tell me all of that or determine any of it? . . . This? What? There has to be an answer to that, and that is why the answers to the questions are already determined . . . in my approach to the object; and that is why it is a fraud to think that the object, uncharacterised by such determination, can fix them. I already know. So I am not fixing what I mean by ‘A’. Or if I am fixing it, it’s just the last point that I am fixing. ‘A’, say, is to be a name of a piece in a game—​that I know, and now I say it is this piece. (POD, 255)

How is it that I already know these things? By having learned a language, not a private language only I can speak, whatever that would mean, but the language I learned in growing up, surrounded by people who spoke to me and in whose conversation and life I joined. The drama of meaning isn’t played out between me, the subject, and ‘my’ object, as in philosophising we’re apt to imagine (cf. PI, §38), but between us, as we talk to each other about various objects. So, are we now to say, as Wittgenstein commentators often imply, that this explains everything? That I couldn’t define things, fix the meaning of words for myself, but others could do it for me, and when I had got my fixes from them, from ‘the community’, I could then go on applying that knowledge to further cases? But how would others be any more able than I to fix meanings? Could my parents just point and say ‘This is yellow’? How would I know what they meant, any more than I would know what I myself meant if I just ‘pointed in my mind’? Would I not here, again, need to know what was pointed to; a colour, not the object so coloured, for example? Indeed I would, and so a community of language users seems just as unable as the solitary individual to fix the meaning (reference) of words. No wonder, then, that the ‘ancients and medievals had a problem whether and how teaching is possible. . . . Is it possible for one human being to teach another?’ (WRPL, 240). Wittgenstein, Anscombe says, made an important ‘contribution to that discussion’ by emphasising that ‘the teacher cannot succeed in teaching unless the pupil has certain 14 

Anscombe details where and how in POD, 236–​248.

428   Joel Backström reactions which he is not obliged to have and which the teacher can’t teach him; is responsive in certain ways in which he does not have to be’ (WRPL, 240). Consider the slave boy to whom Socrates teaches geometry by question-​and-​answer in the famous scene in Plato’s Meno (82b–​85b), a scene Anscombe ‘became convinced . . . was no fiction’ when she repeated Socrates’s experiment with a child who, to her ‘astonishment and pleasure . . . answered just as the slave did’ (HE, 33).15 That children respond as they do to these questions is, she says, ‘something that cannot be taught; it is a prerequisite to teaching’ (WRPL, 240). Teaching is possible only against this background: ‘You can do something which is called teaching; but he only ends up knowing, if he has had these reactions, not those, in the course of the teaching. However, what is called teaching does end up, if successful, with the pupil able to do what in a certain sense he cannot be taught’ (WRPL, 240).16 Furthermore, most of what we learn in connection with being introduced into a life permeated and transformed by language isn’t explicitly taught us even in the apparently paradoxical sense that geometry is taught. For example, ‘First I thought I would tell him and then I thought I would not’ is an everyday thing to say—​but now Anscombe asks, ‘[H]‌ow does one learn to say such things? Perhaps I can show you how to saw a plank; I cannot show you the way to have a thought like that—​so how do you learn?’ (WWP, 215). That isn’t a rhetorical question, with an answer too obvious to state. But neither is it a question expecting an answer; in which direction would we even start looking for one? So, really, we cannot say how or why we learn, but we do in fact learn, and teach; our words do have meaning. Wittgenstein, Anscombe says, doesn’t try to explain how this can be. There’s ‘no theory of language in Wittgenstein’, no attempt to show ‘how noises are significant speech’, to somehow construct meaning out of, or derive it from, ‘external relations’ between bits of sound, behaviour, etc., starting as it were from ‘words without faces’ (TL, 203, 193, 200). Instead, Wittgenstein shows that attempted derivations don’t work, for any arrangement of such bits—​say, the ink marks that you’re looking at on this page—​could be taken in multiple ways, and so they don’t by themselves amount to any meaning at all. But when you, being literate in English, read them as letters of words saying this-​or-​that, they do say that; ‘what one actually does . . . that is what fixes the meaning’ (QLI, 122, first emphasis added; cf. Wittgenstein, PI, §§454, 139–​141). We understand each other and so understand each others’ words; that’s a fact, but our attempts to explain the fact, or again to justify it, to find a necessity in it, fail.17 Anscombe suggests that this is what is ‘most difficult to accept’ in Wittgenstein’s philosophy (RP, 118). Philosophy, she says, ‘is to a great extent a huge factory for the manufacture of necessities’, that is, of fantasies of ‘necessary explanations, necessary

15 

Anscombe often returns to this example: see GSE, 217–​218; HE, 33–​35; LW, 163–​164; and UP. Cf. Anscombe’s parallel remarks against ‘method’ in philosophy, mentioned earlier. 17  Those who assume that we need and must demand a ‘grounding’ that ‘justifies’ our meaning-​claims will think Wittgenstein a sceptic about meaning, because he shows our proposed ‘grounds’ don’t actually ground anything. But as Anscombe underlines, what we find in Wittgenstein is really ‘a rejection of skepticism’, a rejection of the very demand for ‘grounding’ (KRPL, 261). 16 

Anscombe’s Wittgenstein   429 connections’, and Wittgenstein ‘arouses a certain hatred among us’ insofar as ‘[h]‌e’s out to deprive us of our factory jobs’ (WTC, 184). Anscombe is in the factory-​closing business, too. In a discussion clearly Wittgensteinian in inspiration, but containing many original arguments, she argues for the illusoriness of the pervasive assumption that ‘causality is some kind of necessary connection’, so that if A caused B in one case, an A must cause B in every case, unless other causes intervene (this is the subject of CD; cf. Wittgenstein, PO, 370–​411). And the idea of ‘logical necessity’ is just as illusory, she notes; it isn’t necessitation at all, in the sense of a compulsion to think anything (cf. her PI, 130). To be sure, there is ‘the logical must: you ‘can’t’ have this and that . . . you must grant this in face of that’; these concrete ‘musts’ and ‘cant’s’ are ‘the more basic expressions in logical thinking’ on which our abstract, metaphorical notions of ‘laws of thought’ are built, just as ‘You can’t move your king’ is ‘the more basic expression for one learning chess, since it lies at the bottom of his learning the concept of the game and its rules’ (QLI, 121). However, these ‘musts’ don’t compel like a power ‘from outside’; rather, one might say, ‘logic’ is the form or articulation of what our understanding reveals to us. The slave boy isn’t compelled to anything through Socrates’s questioning; this is a free moment in his slave life. Socrates helps him see how the geometrical figures are related, but it is he who sees; ‘it was not because he was told, but because he himself perceived that it was true, that he knew’ (UP, 36). There was no necessity that he should see—​he might not have understood—​but he didn’t have a choice about what to think once he did understand, since ‘I cannot understand a proof and yet not know that it is correct’ (UP, 39). This seems paradoxical only because we model understanding on a situation of being constrained or left free by another, from outside, which is the wrong picture here. In terms of the critique of the subject-​ object paradigm, one could say that Wittgenstein is concerned to show that there’s no objective necessity—​stemming from the side of the objects we deal with—​that forces us, as it were independently of ourselves, to understand them as we do. And often, when we think something can be understood only one way—​where this isn’t in fact an expression of understanding but a blocking of it, a refusal to consider further—​Wittgenstein comes up with another possibility, as in his question about the sun’s movement. Hence, his Shakespearean motto, ‘I’ll teach you differences’, or, less solemnly, ‘You’d be surprised’ (Drury 1984, 157). However, he rejects subjectivism along with objectivism—​they are two sides of a false coin—​and if we’re tempted to imagine that we can understand things however we like, he reminds us of how our knowing and understanding anything at all depends on the grace of nature, and ourselves, not starting to behave in crazy ways, lumps of cheese or other everyday objects ‘suddenly [starting to] grow or shrink for no obvious reason’, for example (PI, §142). And if that were to happen, how would you know whether it was you or the world that had gone crazy? In this sense, Anscombe emphasises (QLI, 131–​133), Wittgenstein is not a ‘linguistic idealist’ but rather tries to articulate, as he put it, a ‘realism without empiricism’ (RFM, VI:23; quoted in QLI, 133). ‘The object’, e.g., the ‘sense-​datum’, doesn’t simply come with its own meaning written on it, as naïve empiricism supposes, and yet ‘the subject’, whether individual or communal, cannot just decree or arbitrarily change the meaning of things either. Rather, how far, and in what way, you and I can understand

430   Joel Backström things, and each other, is for us to find out, ever anew, by engaging in dialogue about those things. And here, a central concern of Wittgenstein’s is to show how, in that dialogue, actually making sense of a pretended possibility is often much more difficult than we imagine. This is important, because metaphysics in the critical-​diagnostic sense he gives the term is the creation of illusory possibilities no less than of illusory necessities. A striking example is analysed in Anscombe’s paper ‘Modern Moral Philosophy’, to which we now turn.

19.6  The Fake Sense of Corrupt Minds Unlike the rest of her philosophy, Anscombe’s writings on ethics and religion may appear (and partly are, I think) unrelated or even alien to Wittgenstein’s way of thinking.18 Nonetheless, her best known, and most controversial, thesis in ethics—​ that the characteristically modern ideas of specifically moral obligations, wrongs, etc. are ‘survivals’ from a long-​since-​abandoned Christian ‘law conception of ethics’ and should be ‘jettisoned’ because they are ‘only harmful without it’ (MMP, 26, 30) is clearly, even if she doesn’t advertise this, a homologue of Wittgenstein’s basic analysis of metaphysical confusion.19 Paraphrasing something Anscombe says about Kripke, one might say she is the mother of the thesis, although it was ‘begotten in [her] by Wittgenstein’ (cf. KRPL, 263). What Anscombe rejects as confused are not everyday moral responses, say, protesting against something as unjust or greedy (MMP, 33), but a specific way of conceiving ‘morality’. She claims that while belief in God-​given moral commandments has all but disappeared, the special associations, the ‘special emphasis and . . . feeling’ of that moral ‘ought’ whose original sense was ‘required by divine law’, remain (MMP, 30). Having been ‘cut off from the family of concepts from which it sprang’, the word ‘ought’, used in this supposedly ‘moral’ sense, ‘no longer signifies any real concept at all’; it has only ‘mesmeric force’ (MMP, 32). We have here ‘the survival of a concept outside the framework of thought that made it a really intelligible one’; ‘all the atmosphere of the term is retained while its substance is guaranteed quite null’ (MMP, 31, 40–​41).

18  The focus of this chapter as a whole is on what Anscombe and Wittgenstein share, rather than on what may divide them—​as such an equally interesting focus for investigation. While I say something about the tensions between their views on religion in the next section, I won’t discuss such tensions in the case of ethics, where I concentrate on one main, generally overlooked Wittgensteinian borrowing by Anscombe. To broaden the focus would take us too far afield, partly because just getting clear about Wittgenstein’s ‘view on ethics’ would demand a long discussion. As I have suggested, ‘ethics’ doesn’t refer to a clearly bounded ‘topic’ but rather to sustained reflection on the basic moral and interpersonal dimension of human life and understanding. (For some discussion of this, in relation to Wittgenstein, see Backström [2018]; cf. Nykänen [2018].) 19  Commentators generally overlook or ignore this homology; Conant (1996) implicitly recognises it, however.

Anscombe’s Wittgenstein   431 Now, this is precisely how Wittgenstein characterises the spinning of metaphysical illusions through our transferring words from their ‘original home’ in various everyday settings to wholly different, ‘sublimated’ contexts, where we imagine, wrongly, that they still make sense, ‘[a]‌s if the sense were an atmosphere accompanying the word, which it carried with it into every kind of application’ (PI, §§116–​117; cf. PI, §§344–​352; OC, §10). Wittgenstein’s endeavour to ‘bring words back from their metaphysical to their everyday use’ (PI, §116) doesn’t mean simply reverting to everyday ways of talking—​that would clarify nothing (BB, 58–​59)—​but showing how our metaphysical transferring manoeuvres create, as Anscombe says, only an ‘appearance of meaning’ (RP, 114–​115). The problem is that the metaphysical uses either cut away the basis of the everyday use they wish to retain, or remain dependent on the everyday use they wish to repudiate; either way, they ‘sin against the grammar of their terms’ (HE, 28; cf. Wittgenstein, RPP-​1, §548). Thus, in speaking of ‘moral’ demands in the specific way Anscombe is criticising, we appear to be saying something gravely consequential because of the lingering association with God’s law—​remember, this isn’t an abstract idea but the historical reality of the biblical tradition—​even as we not only reject any belief in divine law but allow so-​ called moral principles that are absolutely opposed to that law’s actual commandments to take its place (MMP, 40–​42).20 Anscombe’s ‘complaint’ (MMP, 42) against the characteristically modern moral outlook she criticises is precisely that it defends, or deems in principle defensible, substantial ‘moral judgements’ such as condoning the murder of innocents as a means to our ends, that are ‘quite incompatible with the Hebrew-​Christian ethic’, for which actions such as murder were ‘forbidden whatever consequences threaten’; this, she thinks, is ‘the most important fact’ about the modern outlook (MMP, 34). It doesn’t merely keep the idea of the moral law, but without God; rather, it rejects the actual moral judgements that law entailed. Now, Anscombe suggests that we use the mesmerism, the mere formalism of the ‘moral ought’, precisely in order to get away from the moral substance of the Hebrew-​Christian ethic, especially its refusal to ‘flatter avarice and power’ as worldly ethical codes do, and its prohibition against ‘the shedding of innocent blood’ (PM, 36). In her view, our willingness to embrace murderous policies such as bombing whole cities out of existence, or, again, matter-​of-​course abortions on a mass scale—​to name two modern evils of grave concern to her21—​shows that we’re losing our sense of ‘the dignity of human nature’, accepting ever more easily that ‘human beings can . . . be killed so that others can have the life they think they want’ (HV, 198). This pervasive consequentialist sensibility reveals, Anscombe thought, that our minds are becoming ‘corrupt’; we’re not simply tempted by evil in particular cases but are losing our sense of the very distinction

20  It might be objected that speaking of divine law is hardly an ‘everyday use’ of words, but itself an extension and transformation of the everyday notion of human law. Indeed, but the point is that modern moral philosophers bank on the divine law conception while ostensibly repudiating it, in the same way as ‘metaphysics’ generally (ab)uses everyday language. I actually think a divine law conception of ethics has grave problems of its own, but this isn’t the place to discuss them. 21  See CP3, vii–​ix, 51–​81; GG1, 243–​283; GG2, 214–​223, 234–​238.

432   Joel Backström between good and evil, and so of anything being a temptation, not just an ‘option’, even if perhaps an ultimately undesirable one (MMP, 40, 36–​37).22 Anscombe’s claim is that this (substantial) moral corruption and the (formal) incoherence or emptiness of the contemporary moral ‘ought’ are, as Olli Lagerspetz (2006, 446) puts it, ‘two sides of a coin’. For, as Wittgenstein pointed out, if we’re to speak sensibly ‘there must be agreement not only in definitions but also . . . in judgments’ (PI, §241), and if we can take paradigms of evil, such as murdering innocents, and call them ‘morally good’, our words have lost all meaning, have become mere formalist vehicles of metaphysical illusion, however solemnly we declare them.23 I don’t know what Wittgenstein would have thought of Anscombe’s analysis of the ‘moral ought’, but however that may be, her analysis illustrates how ‘metaphysics’ in his critical-​diagnostic sense isn’t confined to seminar rooms. Morally and existentially false attitudes, which may grip individuals or whole cultures, generally include upholding illusions of meaning of the same kind as those in metaphysics; one might call this everyday metaphysics. This is connected to the crucial point—​sometimes clearly articulated, but often left implicit, by Wittgenstein (cf. Backström 2013)—​that, while metaphysical claims fail to make coherent sense, and in this sense are empty, ‘idling’ speech (PI, §132), they have important if ultimately destructive functions in creating comforting or exciting, apparently self-​justifying illusions that allow us to indulge various fateful tendencies. The illusions are created by systematically (which doesn’t necessarily or even typically mean consciously) abusing words, phrases, and idioms that in other contexts have a good use, by transferring them to quite different contexts where they serve to mischaracterise things and put them in a false light, as in presenting greed as ‘economic necessity’ or the wish to meddle in people’s lives as ‘moral concern’. Here, it is crucial to create an atmosphere, characteristic tones, and

22  The term ‘consequentialism’ was coined by Anscombe to refer precisely to this kind of view, where no ‘option’, however evidently evil, is excluded in advance, since everything depends on the consequences, whether intended or merely foreseen (MMP, 36; on the development of Anscombe’s critique of consequentialism and the substantial moral concerns motivating it, see John Berkman’s contribution to this volume). Anscombe doesn’t claim, of course, that this is the only form of moral corruption, or that, except for this modern aberration, moral life has tended to be uncorrupted. She is simply diagnosing the specific features of what she takes to be a dominant contemporary form of moral corruption. To mention a closely related issue, while greed has myriad manifestations and can, it seems, be found in all societies, the modern (capitalist) idea and practice of ‘profit maximisation’ in effect makes greed into the sole, driving principle of economic life, while simultaneously undermining the very possibility of seeing anything at all as greedy; it’s all just plain, profit-​maximising economic rationality. 23  The fact that we see the evil in killing innocents, say, isn’t merely an effect of our history with the Hebrew-​Christian law conception but part of human moral understanding as such (which doesn’t prevent people killing innocents and making up spurious justifications for it, of course). As Anscombe says, while ‘the motives, spirit, meaning and purpose of the moral life of Christians depends on revelation . . . the content of the moral law, i.e. the actions which are good and just, is not essentially a matter of revelation’ (AM, 50). Does she prove this? No, but how could one? That would presuppose a reference to something clearer, more certain, than the evil of killing innocents, say, by means of which its evil would be proven—​and what could that be?

Anscombe’s Wittgenstein   433 postures of seriousness, concern, etc. around these false ways of speaking (to others or to oneself in thought).24 Now, insofar as the spirit of a time is characterised by the dominance of collective atmospheres, seeking philosophical understanding involves struggling to consciously feel and see through the fog, rather than unthinkingly breathe it—​in and out, circulating the same bad air—​as Anscombe thought contemporary moral philosophy did; ‘conceived perfectly in the spirit of the time [it] might be called the philosophy of the flattery of that spirit’ (OMPCY, 167). As the characterisation ‘flattery’ indicates, there’s not just passive failure to think here, but an active endeavour to ‘belong’ in the collective movement, the dominant ways of thinking and speaking and feeling of the times, what Christianity calls worldliness. Wittgenstein explicitly says what the flattery quote from Anscombe suggests: that the deepest difficulty of philosophy is this struggle against one’s own wish to ‘live with the herd that has created this language [of the times] as its proper expression’ (PO, 185; cf. Backström 2011, 739–​744). And this difficulty, we should note, need not only be there in connection with obviously morally and existentially charged questions. For example, Anscombe characterises the false, deterministic conception of causality she criticises as ‘a bit of Weltanschauung: it helps form a cast of mind which is characteristic of our whole culture’ (CD, 133). She reports once remarking to Wittgenstein, when they were discussing some topic or other, ‘But of course everybody understands that in such and such a way’—​to which he replied, ‘If everybody understands a thing in a certain way that is sure to be false!’ and then quoted Nietzsche: ‘If something true receives public acknowledgement, ask in the interests of what lie this has happened’ (AW, 47; cf. Backström 2019b).

19.7  Unfinished Business As the previous discussion indicates, a Wittgensteinian conception of philosophy as descriptive and clarificatory rather than justifying or explanatory (PI, §§122–​133) doesn’t imply the impossibility of critiquing our ways of thinking, even in quite general and pervasive terms. A widespread misunderstanding that Anscombe wants to correct is precisely ‘the impression that Wittgenstein came to think that concepts were uncriticisable. This tribe has these, another perhaps other ones. Neither is right, or wrong’ (ST, 179). This idea, she says, would make Wittgenstein into ‘a trivialiser’, which he was not (ST, 179). Rather, he was a problematiser of simplistic, misleading ideas about what

24 

As Anscombe notes (P, 93), hypocrisy, the pretence, to others and to oneself, to be concerned where, or in a way that, one really isn’t, ‘carries with it an implicit demand for respect for an atmosphere evoked by the pretender, which surrounds not the reality, but the idea of such things as being principled, or cultured, or saintly’—​and she aptly characterises cynicism not as freedom from hypocrisy but as conscious ‘pretence of hypocrisy’.

434   Joel Backström criticising concepts or forms of life involves, and Anscombe follows him in this. To see how, let’s consider her writings on religion. Wittgenstein said, enigmatically, ‘I am not a religious man but I cannot help seeing every problem from a religious point of view’ (Drury 1984, 79). He told Anscombe that, while he ‘could not understand the idea of loving God’, he also thought he ‘could not be great friends with someone wholly lacking in something religious’ (AW, 2–​3). Of Anscombe and another devout Catholic friend, however, he said that, while he found the ‘symbolisms of Catholicism . . . wonderful beyond words’, he ‘could not possibly bring [him]self to believe all the things that they believe’ (Drury 1984, 102; Malcolm 2001, 60). Anscombe, for her part, told a friend, ‘On the topic of religion, Wittgenstein is sheer poison’ (Kenny 2019, 20). When one reads her writings on Wittgenstein and religion, however, a closer, if also undecided, relationship emerges, and it isn’t Wittgenstein’s own thought but a certain influence he had on others that she thinks poisonous—​but then his influence on other philosophers quite generally was deemed by her ‘great and . . . bad’ (BBC, 234, emphasis added).25 As Anscombe notes, Wittgenstein detested ‘rationality, or would-​be rationality, in religion’; he rejected the idea of natural theology, that is, of ‘attempted reasoning from the objects of the world to something outside the world’, and generally didn’t want ‘the jagged edges and spikes’ of faith smoothed out (QLI, 122–​123).26 Now Anscombe draws a parallel to a strand in Catholic thinking: In the Catholic faith, certain beliefs (such as the Trinity, the Incarnation, the Eucharist) are called ‘mysteries’; this means at the very least that it is neither possible to demonstrate them nor possible to show once and for all that they are not contradictory and absurd. On the other hand contradiction and absurdity [are] not embraced; ‘This can be disproved, but I still believe it’ is not an attitude of faith at all. So ostensible proofs of absurdity are assumed to be rebuttable, each one in turn. Now this process Wittgenstein himself once described: ‘You can ward off each attack as it comes’ (Personal conversation). But the attitude of one who does that . . . is not that of willingness to profess contradiction. On the contrary. On the other hand, religious mysteries are not a theory, a product of reasoning; their source is quite other. Wittgenstein’s attitude to the whole of religion in a way assimilates it to the mysteries: thus he detested natural theology. But . . . what part of this was philosophical (and therefore something which, if right others ought to see) and what part personal, it is difficult to say. (QLI, 122–​123)

25  Wittgenstein himself agreed. Anscombe reports him telling her, ‘I never produce fire but only superficiality and sham superficiality [sic!] in people’, and when John Wisdom wrote him, saying, ‘You have changed philosophy’, he showed her the letter and said ‘I wish I hadn’t’ (AW, 17; cf. Wittgenstein, CV, 70–​7 1). 26  Anscombe says, for instance, that Wittgenstein ‘did not despise belief in miracles; only “rational” belief in miracles. He would not despise the man whose taking something as a miracle was a sort of . . . confession wrung from him. What he could not stand was the suggestion that a reasonable man could see as a matter of reasonable consideration that such and such miracles happened. He thought this a stupid misunderstanding’ (AW, 12).

Anscombe’s Wittgenstein   435 Anscombe’s writings on religion, that is, primarily, on Catholic Christianity, do initially make a rather rationalist impression, as compared to Wittgenstein’s remarks on the matter. (He once told her, ‘You may be able to fill reasoning with religious content; I never could’ [AW, 3].) But a closer look reveals Anscombe’s ‘rationalism’ to be decidedly anti-​theoretical in its attempt to make sense, as far as she could, but no further, of the articles of faith. In discussing the belief she shared in the survival of the soul after death and the resurrection, she thus says that, while the practice of praying for the dead makes it meaningful to speak of the existence of the dead between death and the resurrection, ‘the good of philosophy will be to cure us of the impulse to try and have a theory of the existence of the separated soul’ (IS*, 78). In her writings about belief in miracles and prophecy (PM; HM), she similarly attempts to detach these beliefs from the ‘theories’, metaphysical in roughly Wittgenstein’s critical sense, that have been thought necessary to ‘justify’ them. And in a paper on transubstantiation, Anscombe says that ‘the mystery of the faith . . . is the same for the simple and the learned. . . . [B]‌oth believe the same, and what is grasped by the simple is not better understood by the learned: their service is to clear away the rubbish which the human reason so often throws in the way to create obstacles’ (OT, 112). In her view, it seems, this ‘rubbish’ consists as much in proposed theoretical ‘justifications’ of faith as in criticisms purporting to show its absurdity. As in philosophy generally, the urge for metaphysical justifications and scepticism mutually live off each other. Anscombe also objected to the general, vague respect for religion as a phenomenon typical of educated, liberal Westerners, who think that ‘[a]‌ll religions are the same really: they are a lot of different paths to the same end’—​to which she drily responds, ‘[N]ice religions, you mean, don’t you?’ (PSP, 60). She thought some followers of Wittgenstein—​’I can’t speak for Wittgenstein [himself]’—​exemplify a philosophically sophisticated version of this attitude (PSP, 58). Insisting on the unlikeness of religious language to the truth-​claims of science, and on the personal character of religious belief, the upshot of their view is that there’s ‘no such thing as a religion’s being true’; religious belief is ‘better compared to somebody’s being in love than to his believing anything true or false’ (PSP, 58). She characterises this view as an expression of . . . the heart of actual historical and present paganism: namely having and respecting the various worships of many gods and hating the exclusiveness of the true religion. That exclusiveness branded the ancient Jews as atheists, enemies of the gods. For our philosophic pagans there is no such thing as the true religion or the true god; the many religions can perhaps be like many pearls on a string. That one string which each religion may be hung on, is something rich and significant in the depths of the self. All peoples have gods and it is contemptible to be scornful of them for this: what matters is whether there is this depth (of religion) in a man’s heart. But the question is: Why? Why should it matter? (PSP, 60)

This is indeed the question. Otherwise put: what does it mean for me to ‘respect’ someone’s belief if I don’t think there’s any truth in it? Suppose a philosopher is, in

436   Joel Backström Anscombe’s words, ‘so very modern’ that he doesn’t think he can ‘judge a savage’s religion of placating spirits to be a delusion’; confronted with Christians praying for the dead he will equally not say that they are wrong, ‘nor yet that he disbelieves their presuppositions about the existence of those spirits, but only that he does not do any of these things’ (IS*, 79, emphasis added). What will this ‘tolerantly respectful’ philosopher say to someone like Anscombe, who actually believes that hers is the true religion? Will not his ‘respect’ suddenly appear patronisingly disrespectful, like assuring someone who tells you they just witnessed a murder that you ‘respect’ their viewpoint and the ‘deep seriousness’ with which they obviously treat the matter—​but then doing nothing towards investigating whether what they say is true, whether someone really was murdered? This is, Anscombe says, a falsely aestheticised, ‘sentimental’ attitude, ‘like someone who thought he was watching a play when some real action was going on’ (IS*, 74). Crucially, assuming the ‘respectful’ attitude also serves—​and this may be its secret motivation—​to protect oneself from the challenge of actually confronting and engaging with the claims of the believer, thus raising the question of what their possible truth, or a rejection of them as false, might amount to for oneself. For these beliefs have effectively been written off as just expressions of personal inclination, of a ‘way of life’—​a notion whose use, Anscombe notes, ‘carries . . . connotations both of satisfactoriness and arbitrariness, nor does it lack the upward-​looking glance’ (OMPCY, 166).27 Nota bene, two believers may equally protect each other from challenge by affirming their ‘common faith’, whose actual meaning, what they supposedly share, is never questioned. So ‘believing’ a religion’s creed or not need not be a decisive difference. That is, what believing (or not) comes to is an open question, to be asked ever again, if the faith is to be real faith, or its rejection not unthinking or self-​deceived. Anscombe’s objection to the ‘Wittgensteinian’ pagan’s denial that there can be truth in religion doesn’t imply that it is clear what ‘truth in religion’ means; the problem is rather that ‘tolerant’ pagans of this sort assume it is clear what ‘truth’ must mean, and so confidently deny that there can be any in religion. By contrast, Anscombe learned from Wittgenstein that one cannot assume one knows what exactly ‘truth’, ‘disagreement’, ‘rejection’, etc. mean in various contexts; one must try to find out. As she says, ‘Anyone saying something religious is—​or ought to be (?)—​doing something religious in saying [it]’, and Wittgenstein’s question is ‘What is he doing, saying that?’ (AW, 14) Wittgenstein, she says, ‘is not moved by the question: Is it true? because to say p is true is to say p and so the original question remains: what is he doing in saying that?’ (AW, 14). So what can truth mean here? Anscombe insists that religious belief belongs in, although it isn’t simply reducible to, a landscape of interpersonal moral concern. For

27  Wittgenstein’s own, very different attitude to someone (I don’t say: to anyone, a priori) who believed their religion true in a sense that makes other religions false (there are other senses of true), might perhaps be gleaned from this recollection of Anscombe’s: ‘He more than once ticked me off for being “cock-​sure.” He used to blow me up for any manifestation of this.—​I remember once saying something like “What people have had such a history as the Jews!” and he said at once, with irritation, “What do you know of the histories of all the peoples there have been on the earth?” ’ (AW, 11).

Anscombe’s Wittgenstein   437 example, although ‘the conception of an immaterial substance . . . is a delusive one’, we can sensibly speak of ‘the spirituality of the soul’ (IS*, 71). What makes ‘a non-​superstitious and non-​fabulous belief in spirits’ possible, and what provides ‘the reason for speaking of the spirituality of the soul’ at all, is not, she says, ‘a quasi-​physical common property, but that human beings are in for a final orientation towards or away from the good’; without this, belief in the immortality of the soul and the resurrection ‘would have no significance’ (IS*, 82–​83; cf. Wittgenstein’s remarks on the meaning of belief in the resurrection, CV, 38–​39). As Anscombe also notes, the ‘mystical’ isn’t necessarily ‘out of the ordinary’; thus, ‘the feeling for the respect due to a man’s dead body: the knowledge that a dead body isn’t something to be put out for the collectors of refuse to pick up . . . is mystical; though it’s as common as humanity’ (CC, 187). None of this, of course, proves the truth of any particular religious belief; rather, these comments indicate, as it were, the kind of terrain on which religious beliefs might be made sense of, or again meaningfully criticised. A central problematic here, connected to the discussion of the previous section, concerns the relationship between religious belief and collective belonging: the role of saying ‘I believe X’ in the sense of ‘. . . as one of us who believe X’. Anscombe spoke with conviction as a Catholic to Catholics, often to criticise what she saw as corrupt tendencies in the Church to which she belonged (as attested by many papers in CG2). Wittgenstein, by contrast, seems to have been wary of spiritual or doctrinal community even if combined with such readiness for self-​criticism. He says that the philosopher isn’t—​that is, must constantly fight against the temptation to become—​‘a citizen of any community of ideas’ (Z, §455). In discussing Roman Catholicism with Drury (1984, 102) Wittgenstein objected to the dogmatic ‘narrowness’ he sensed even in otherwise humble and impressive priests—​‘I like to feel free to discuss anything with anyone I am with’—​and when Drury said he thought he ‘could be happy working as a priest among people whom I felt shared the same beliefs as I have’, Wittgenstein replied, ‘Oh, don’t depend on circumstances. Make sure that your religion is a matter between you and God only’. So, insofar as a confession of faith implies a conscious or unconscious limiting of one’s freedom of thought and response to keep one within the community of believers, Wittgenstein would reject it. This isn’t to say that a confession of faith must imply this (although it clearly often does so, in fact), but that whether it does is crucial for the nature of the confession. It is equally clear that refusing overtly confessing to any faith in no way protects one from bondage to the collective spirit of one’s place and time. The struggle to free oneself from bondage to this most powerful ‘community of ideas’ never ends. Belonging in the same family with moral responses, religious truth-​claims are not, basically, empirical claims that might be scientifically tested; as Anscombe says, ‘Science can correct only scientific error’ (QLI, 125; cf. WWC 228–​230).28 More generally,

28 

Anscombe (PM, 20–​25) and Wittgenstein (CV, 37–​38) seem to take quite different attitudes to the question of the historical truth of the Gospel narratives, however.

438   Joel Backström Wittgenstein underlines that one can sensibly speak of mistakes only where there are, in Anscombe’s words, ‘unsatisfied criteria of correctness, criteria which correspond to the intention of the speaker’ (QLI, 124). Crucially, however, one might want to reject or question what someone says, or some of one’s own earlier claims, for many reasons other than their being a mistake, seeing them, say, as flattery of the spirit of the times, religious or secular, as superstition, madness, or some otherwise problematic response to the world (cf. QLI, 124; WWC, 222, 226). Anscombe emphasises that Wittgenstein’s discussions, in his last writings, of the different possibilities here are more radical and heterogeneous than is usually realised. ‘[W]‌e should not regard the struggling investigations of On Certainty as all saying the same thing’; for instance, ‘a world-​picture is not the same thing as a religious belief, even though to believe is not in either case to surmise’, and we ‘cannot get [Wittgenstein] right, but only commit frightful confusions, by making assimilations’ (QLI, 130). In general, before one pronounces an apparently strange belief false or deluded or senseless—​and one’s own different belief true and sensible—​one had better be clear what kind of sense the believer tried but in one’s view failed to make: ‘It is like finding nothing in an author; if one is clear that one understands him, and judges that there is nothing in him, that is all right; but if one merely finds nothing in him (without being able to see what is supposed to be in him) then one cannot say “Never, in my right mind, will I admire him” ’ (IS*, 81; cf. WWC, 223). Anscombe’s sensitive discussion of the case of someone who has the idea that devils plague him (IS*, 80–​81) brings out how difficult it may be to know how to respond even to beliefs that initially seem obviously deluded. The difficulties here are essentially moral, existential, and spiritual, because they concern the way one is able and willing to relate to another person, the one who has the initially strange beliefs, and as Anscombe rightly notes, over the years, Wittgenstein’s interest moved from revealing the nonsense in metaphysical pseudo-​claims to the kind of ‘not making sense’ that is manifest in our difficulties in understanding each other (I, 27). And here, Anscombe insists, Wittgenstein’s attitude, while enquiring and concerned to understand rather than judgingly dismiss the other—​or, rather, precisely because that was his attitude—​is not the tolerant pagan’s, who tries to avoid facing conflicts by bracketing the question of truth. On Anscombe’s view, Wittgenstein doesn’t espouse the ‘cultural relativism’ of incommensurability; we may have, between two people, ‘a ‘disagreement in the language they use’—​but then it really is a disagreement’ (QLI, 131, emphasis added). Viewed from the subject-​object perspective, a disagreement can be genuinely decided, the truth of the matter determined, only by determining what is ‘objectively’ the case, whether by finding out the facts or by inspecting ‘the argument’, conceived as a kind of logical ‘object’ that has certain features that make it valid or not. If there’s no objective necessity to see the situation this way or the other, one is subjectively licensed to see it as one pleases. That is, either the truth of the matter will be determined as it were independently of us who disagree, or it won’t be determined at all. Now, this framing leaves out the most crucial dimension of moral understanding: the task of asking oneself, in conscience and in the absence of any objective proof, how one actually understands the situation—​which isn’t at all the same as how one might wish to see it—​and how far one

Anscombe’s Wittgenstein   439 can actually understand the other and the other can understand oneself, which may be much further than one wants to admit, in one’s wish to uphold a self-​servingly false view of oneself as different from and/​or misunderstood and unfairly criticised by the other. As Nykänen (2019) suggests, the subject-​object perspective may be so tempting in life and in philosophy precisely because it denies or obfuscates this daunting task. However that may be, Anscombe certainly didn’t think Wittgenstein had said the last word on our difficulties of understanding: ‘The case of conflict remains unfinished business’ (QLI, 133).

References Writings by Anscombe cited by abbreviations used only in this chapter AW—​Anecdotes about Wittgenstein. (Unpublished.) Typescript made by Luke Gormally of two Anscombe notebooks deposited at the Anscombe Archives at University of Pennsylvania. BBC—​‘ “Ludwig Wittgenstein”—​A BBC Radio Talk by Elizabeth Anscombe in May 1953’. Edited by C. Erbacher, J. Jung, and A. dos Santos Reis. Nordic Wittgenstein Review 8 (1–​2), 2019, 225–​240. LvW—​Letters to G. H. von Wright. (Unpublished.) Correspondence deposited at The National Library of Finland, Helsinki. Note: Permission to quote from Anscombe’s unpublished letters and ‘Anecdotes’ was kindly provided by Mary Geach, © M C Gormally.

Abbreviations of Wittgenstein’s writings BB—​The Blue and The Brown Books. 2nd ed. Oxford: Basil Blackwell, 1969. CV—​Culture and Value. Edited by G. H. von Wright in collaboration with Heikki Nyman. Translated by Peter Winch. Revised edition of the text by Alois Pichler. Oxford: Basil Blackwell, 1998. OC—​On Certainty: Reprinted with Corrections and Indices. Edited by G. E. M. Anscombe and G. H. von Wright. Translated by Danis Paul and G. E. M. Anscombe. Oxford: Basil Blackwell, 1974. PI—​Philosophical Investigations. Edited by G. E. M. Anscombe and R. Rhees. Translated by G. E. M. Anscombe. 2nd ed. Oxford: Basil Blackwell, 1958. PO—​Philosophical Occasions 1912–​1951. Edited by J. C. Klagge and A. Nordmann. Indianapolis, IN: Hackett, 1999. RFM—​Remarks on the Foundations of Mathematics. Edited by G. H. von Wright, R. Rhees, and G. E. M. Anscombe. Translated by G. E. M. Anscombe. 3rd ed. Oxford: Basil Blackwell, 1978. RPP-​1—​Remarks on the Philosophy of Psychology. Vol. I. Edited by G. E. M. Anscombe and G. H. von Wright. Translated by G. E. M. Anscombe. Oxford: Basil Blackwell, 1980. TLP—​Tractatus logico-​philosophicus. Translated by D. F. Pears and B. F. McGuinness. London: Routledge & Kegan Paul, 1961. WL46/​7—​Wittgenstein’s Lectures on Philosophical Psychology 1946–​1947. Edited by P. T. Geach. Chicago: University of Chicago Press, 1988. Z—​Zettel. Edited by G. E. M. Anscombe and G. H. von Wright. Translated by G. E. M. Anscombe. 2nd ed. Oxford: Basil Blackwell, 1981.

440   Joel Backström Works by others Backström, J. (2011). ‘Wittgenstein and the Moral Dimension of Philosophical Problems’. In O. Kuusela and M. McGinn (eds.), The Oxford Handbook of Wittgenstein. Oxford: Oxford University Press, 729–​752. Backström, J. (2013). ‘Wittgenstein, Follower of Freud’. In Y. Gustafsson, C. Kronqvist, and H. Nykänen (eds.), Ethics and the Philosophy of Culture: Wittgensteinian Approaches. Newcastle upon Tyne: Cambridge Scholars Publishing, 212–​244. Backström, J. (2018). ‘From Nonsense to Openness: Wittgenstein on Moral Sense’. In E. Dain and R. Agam-​Segal (eds.), Wittgenstein’s Moral Thought. New York: Routledge, 247–​275. Backström, J. (2019a). ‘Philosophy of Mind and/​ as the Repression of Interpersonal Understanding’. In J. Backström, H. Nykänen, N. Toivakainen, and T. Wallgren (eds.), Moral Foundations of Philosophy of Mind. Cham, Switzerland: Palgrave Macmillan, 231–​266. Backström, J. (2019b). ‘Pre-​truth Life in Post-​truth Times’. Nordic Wittgenstein Review, Special Issue 2019: Post-​Truth, 97–​130. Backström, J. (forthcoming). ‘Naked Please! Elizabeth Anscombe as Editor and Translator of Wittgenstein’. In T. Wallgren (ed.), The Creation of Wittgenstein. London: Bloomsbury. Conant, J. (1996). ‘Reply: Nietzsche, Kierkegaard and Anscombe on Moral Unintelligibility’. In D. Z. Phillips (ed.), Religion and Morality. Houndmills: Macmillan, 250–​298. Diamond, C. (2019). Reading Wittgenstein with Anscombe, Going On to Ethics. Cambridge, MA: Harvard University Press. Drury, M. O’C. (1984). ‘Some Notes of Conversations with Wittgenstein’ and ‘Conversations with Wittgenstein’. In R. Rhees (ed.), Recollections of Wittgenstein. Oxford: Oxford University Press, 76–​171. Engelmann, P. (1968). Letters from Ludwig Wittgenstein with A Memoir. New York: Horizon Press. Erbacher, C. (2016). ‘Wittgenstein and His Literary Executors: Rush Rhees, Georg Henrik von Wright and Elizabeth Anscombe as Students, Colleagues and Friends of Ludwig Wittgenstein’. Journal for the History of Analytical Philosophy 4 (3), 1–​39. Erbacher, C. and Krebs, S. (2015). ‘The First Nine Months of Editing Wittgenstein: Letters from G. E. M. Anscombe and Rush Rhees to G. H. von Wright’. Nordic Wittgenstein Review 4 (1), 195–​231. Geach, P. (1988). ‘Editor’s Preface’. In P. T. Geach (ed.), L. Wittgenstein, Wittgenstein’s Lectures on Philosophical Psychology 1946–​1947. Hemel Hempstead: Harvester Wheatsheaf, xi–​xv. Gibson, A. (2019). ‘Anscombe, Cambridge and the Challenge of Wittgenstein’. In J. Haldande (ed.), The Life and Philosophy of Elizabeth Anscombe. Exeter: Imprint Academic, 23–​41. Kenny, A. (2019). ‘Elizabeth Anscombe at Oxford’. In J. Haldande (ed.), The Life and Philosophy of Elizabeth Anscombe. Exeter: Imprint Academic, 12–​22. Klagge, J. (2019). ‘The Wittgenstein Lectures, Revisited’. Nordic Wittgenstein Review 8 (1–​2), 11–​82. Lagerspetz, Olli. (2006). ‘Anscombe on the Moral Ought and Moral Corruption’. Philosophical Papers 35 (3), 435–​455. Leach, S. (2020). ‘Chadbourne Gilpatric and Ludwig Wittgenstein: A Fateful Meeting’. Nordic Wittgenstein Review 9, DOI 10.15845/​nwr.v0i0.3568, 1–​21. Malcolm, N. (2001). Ludwig Wittgenstein: A Memoir. 2nd ed. Oxford: Oxford University Press. Monk, R. (1991). Ludwig Wittgenstein: The Duty of Genius. London: Vintage. Nykänen, H. (2018). ‘Wittgenstein’s Radical Ethics’. European Journal of Psychoanalysis, https://​ www.jour​nal-​psy​choa​naly​sis.eu/​wittge​nste​ins-​radi​cal-​eth​ics.

Anscombe’s Wittgenstein   441 Nykänen, H. (2019). ‘This Thing with Philosophy’. In J. Backström, H. Nykänen, N. Toivakainen, and T. Wallgren (eds.), Moral Foundations of Philosophy of Mind. Cham, Switzerland: Palgrave Macmillan, 329–​362. Teichmann, R. (2015). ‘Why “Why?”? Action, Reasons and Language’. Philosophical Investigations 38 (1–​2), 115–​132.

Chapter 20

An sc om be an d Aqu i nas John Haldane

I am glad you find St Thomas’ moral philosophy useful. I do too. I notice that you’ve looked him up especially, or perhaps attended to him especially, in connection with questions that had arisen in your own mind; and that is a natural way to get a lot out of an author. You know something of what a question is about, not just by reading him, and so in some way you notice more of the detail of what he says—​I am inclined to think that if you go on thinking hard, you will find more passages come alive for you.1

20.1  Introduction The steady growth of interest in the thought of Elizabeth Anscombe over the past two decades has been focused mainly on the areas of ethics and action represented primarily by ‘Modern Moral Philosophy’ (1958) and Intention (1957). This interest has had three broad effects. First, as scholars have studied these works they have become aware of elements within and relationships between them that had previously been unnoticed or gone uncommented upon. Second, they have paid closer attention than previous readers to related writings: some from the same period, such as Mr Truman’s Degree (1956), ‘Does Oxford Moral Philosophy Corrupt the Youth?’ (1957), ‘On Brute Facts’ (1958), ‘War and Murder’ (1961), and ‘Two Kinds of Error in Action’ (1963), and other, later ones in which Anscombe revisits themes in these early writings. Third, as readers have sought to understand the direction of approach and what is being proposed in Anscombe’s generally intricately woven analyses and arguments (on other matters as 1 

Elizabeth Anscombe in a letter to the present author, March 10, 1983.

Anscombe and Aquinas    443 well as on ethics and action) they have looked to find clues in sources and influences that might help make things clearer. One such is Aristotle, another is Wittgenstein, though her relation to these figures is rather different. From the first, she takes certain ideas, including (a) the hylomorphic (form/​matter) constitution of substances including animals, and amongst these human beings; (b) notions of material, formal, and final (in addition to efficient) causality; (c) the categorial classification of predicates; (d) the primacy of dispositions in ethical life; and (e) the idea of practical rationality. From the second, she derives a method of enquiry; an acute sensitivity to differences in thought and language, and to the role of context in constituting meaning, in both semantic and actional senses; and a suspicion of broad generalisation and of theoretical systematisation. More recently, however, there has been a growing sense of Anscombe’s indebtedness to Thomas Aquinas, though quite what the specifics and the extent of that debt are has remained unclear for three reasons. First, because of most readers’ general unfamiliarity with Aquinas’s writings. Second, because of ignorance about important aspects of Anscombe’s intellectual biography and writings. Third, because of her inexplicitness, particularly in works intended for academic readers, regarding his influence on her thought. In her introduction to From Plato to Wittgenstein, the third of a series of four posthumous collections of Anscombe’s writings, her daughter Mary Geach notes: Anscombe drew upon [Aquinas’s] thought to an unknowable extent: she said to me that it aroused prejudice in people to tell them that a thought came from him: to my sister she said that to ascribe a thought to him made people boringly ignore the philosophical interest of it, whether they were for Aquinas or against him. . . . She once called Aquinas ‘a strikingly good philosopher’. . . . However, she seems from her remains not to have done much expounding of Aquinas. (GG3, xix) And in her introduction to the previous volume, Faith in a Hard Ground, Geach writes: She devised a method which she recommended to me, of mining Aquinas for helpful philosophical points: this was to prospect for philosophically usable bits in the Summa theologiae by considering to what Catholic doctrine her particular philosophical problem was relevant. (GG2, xiv)

Related to this is the comment quoted at the outset, but in that case and in others what Anscombe turned to in Aquinas might be directly philosophical discussions rather than theological ones from which the philosophy had to be extracted. While the extent to which she drew from St. Thomas may be uncertain, it is possible to get a good sense of her familiarity and involvement with his thought by taking note of her engagement with others appreciative of and knowledgeable about his writings; cataloguing her direct and indirect references to his work (not just to Summa Theologiae [hereafter ST] but to lesser known works such as De Malo and De Veritate); identifying ideas that look to be applications of ones to be found in his work; and taking some

444   John Haldane measure of the extent to which she was in broad agreement with or dissented from his positions on major philosophical issues. In the next section, ‘Getting Hooked on Philosophy, without Knowing It’, I provide an account of the beginnings of Anscombe’s encounter with the thought of Aquinas initially as mediated by others, starting with her early reading before getting to Oxford University, developing through her first years there and her engagement to Peter Geach. This expands upon her own brief account and explains features of the Oxford Catholic philosophical culture of the 1930s as represented by members of two Catholic congregations: the Order of Preachers (OP) and the Society of Jesus (SJ), generally known as the Dominicans and the Jesuits, respectively. Thereafter, I explore a series of themes and sets of influences and interlocutors under the section titles ‘Wittgenstein, Lewis, and Philosophical Psychology’; ‘The Philosophical Inquiry Group, Anthony Kenny, and Natural Theology’; and ‘The Return to Ethics and Action’, before concluding with ‘An Admirer but Not a Disciple’, in which I consider the question of how far she agrees (and disagrees) with Aquinas on what one would think of as central issues for philosophers interested in and perhaps committed to Catholic teachings. In an appendix, I list published writings in which Aquinas is cited or mentioned without citation, providing references to the Summa and other writings, either those given by Anscombe herself or ones which I have determined by context and content, together with a brief summary of unpublished material relating to Aquinas.

20.2  Getting Hooked on Philosophy, without Knowing It Anscombe has long been acknowledged as an important figure in twentieth-​century philosophy, both as the translator of major works of Wittgenstein and as a singular thinker in her own right, distinguished by her boldness, insight, and creativity.2 There has been little serious or informed attention given, however, to her intellectual biography, with interest being drawn to anecdotes illustrating her eccentricities and more recently to the celebration of her as an independently minded, courageous, and creative woman philosopher. The first tends to misplace her distinctness and the second to see it as an instance of a broader tendency to rebel against a male-​ dominated profession. What these perspectives miss are her early and enduring intellectual, moral, and spiritual attachments and their development in like-​minded

2  I discuss these matters further, distinguishing between her approach to autonomously philosophical issues and religious ones, in J. Haldane, ‘A Philosopher of Singular Style and Multiple Modes,’ in A. O’Hear, ed., A Centenary Celebration: Anscombe, Foot, Midgely and Murdoch (Cambridge: Cambridge University Press, 2020), 31–​60.

Anscombe and Aquinas    445 company outwith the mainstream of British culture. From very early on Anscombe was an ‘outsider’ who saw her work as a philosopher as that of serving the need to free herself and others from aspects of modern thought, and her allies in this were thinkers whose time, formation, or independence of mind placed them also outside prevailing philosophy and ethics. It is important, then, to begin with an account of her early self-​formation in part to note early but recurrent areas of interest (including ones unconnected to ethics and action) and to identify an aspect of Anscombe’s character and education that is related to her intense concern with injustice and ‘spirited’ denunciations of the morality of her times. Given the importance of her associations particularly with fellow Roman Catholics, I give some attention to describing and explaining features of the Catholic intellectual culture before, during, and after the Second World War, particularly as these relate to invocations of the thought of Aquinas. These matters were important aspects of her life as a thinker (and doer) but remain little known even amongst her admirers. In her introduction to Metaphysics and the Philosophy of Mind (the second volume of collected papers published in her lifetime) Anscombe writes: My first strenuous interest was in the topic of causality. I didn’t know that what I was interested in belonged to philosophy. As a result of my teen-​age conversion to the Catholic Church—​itself the fruit of reading done from twelve to fifteen—​I read a work called Natural Theology by a nineteenth century Jesuit. I read it with great appetite and found it all convincing except for two things. One was the doctrine of scientia media, according to which God knew what anybody would have done if, e.g., he hadn’t died when he did. . . . I found I could not believe this doctrine. . . . I did not know at the time that the matter was one of fierce dispute between the Jesuits and the Dominicans, who took rather my own line about it. So when I was being instructed a couple of years later by a Dominican at Oxford, Fr Richard Kehoe . . . I told him that I couldn’t see how that stuff could be true. He was obviously amused. . . . But it was the other stumbling block that got me into philosophy. The book contained an argument for the existence of a First Cause, and as a preliminary to this offered a proof of some ‘principle of causality’ according to which anything that comes about must have a cause. The proof had the fault of proceeding from a barely concealed assumption of its own conclusion. . . . In all this time I had no philosophical teaching about the matter; even my last attempt [at avoiding begging the question] was made before I started reading Greats at Oxford. It was not until then that I read Hume and the discussion in Aquinas, where he says that it isn’t part of the concept of being to include any relation to a cause but I could not understand the grounds of his further claim that it is part of the concept coming into being [ST, Ia. q.44, art. 1, ad 1]. The other central philosophical topic which I got hooked on [without] even realizing that it was philosophy, was perception. I read a book by Fr Martin D’Arcy S.J., called The Nature of Belief and got just that out of it [i.e., that it was an issue debated by philosophers]. (CP2, vii–​viii)

Anscombe first learned about Roman Catholicism at the age of twelve through reading a book about the lives of priests during the Elizabethan period of persecution, many of

446   John Haldane whom were Jesuits.3 The basic chronology of her path to Rome was initially made public in the first article she ever published, which was in June 1938. Michael de la Bédoyère, the ex-​Jesuit editor of the English weekly The Catholic Herald, sought contributions to a series entitled ‘Under Twenty-​Five’, in which young people were invited to write about their faith, how they saw the condition of the world, and their hopes and fears for the future. Anscombe’s article, ‘I Am Sadly Theoretical’, states, ‘Converted in 1935, recently entered the Church. Went up to Oxford (St Hugh’s) in 1937 to read Greats’.4 The conversion was around the age of sixteen, and she began instruction by Fr. Kehoe in her first year at Oxford (1937–​1938), when she was eighteen. The following April she was received into the church by Kehoe but at St. Aloysius, the (then) Jesuit parish, rather than at Blackfriars, the nearby Dominican priory. Fr. D’Arcy, whose book The Nature of Belief had been published in 1931, was then master of Campion Hall, the Jesuit house of study in Oxford, and had previously taught at Stonyhurst College in Lancashire, where de la Bédoyère had been a pupil of his. Lord Longford (Frank Pakenham), who had been a Fellow at Christ Church, also read The Nature of Belief and was later converted by D’Arcy, describing him as ‘a unique figure in the Oxford I knew as an undergraduate and a don. . . . [N]‌o individual had so great an influence on such a wide circle of intellectuals and others’.5 D’Arcy was on good terms with A. J. Ayer and Isaiah Berlin, who as students had attended his classes, and with Donald MacKinnon (later tutor to Philippa Foot, Mary Midgley, and Iris Murdoch); he had in common with Berlin and MacKinnon the distinction of having won the John Locke Scholarship. Ayer later wrote, ‘Among the very few classes I attended for pleasure was one [on Thomas Aquinas] given by Father D’Arcy’.6 He was also famous for making Catholic converts of current and former undergraduates, including, in the years shortly prior to Anscombe’s arrival, Anne Freemantle (Jackson), Robert Speaight, and Evelyn Waugh. Additionally, D’Arcy was a prolific academic and apologetic writer. Between 1927 and 1932 he contributed three papers to the Aristotelian Society, including on Aquinas’s account of cognition, and on the priority of the good over the right, as well as producing one of the first English-​language books on the thought of Aquinas.7 In 1932 he also began to give BBC radio talks (in 1933 on St. Thomas Aquinas and in 1937 several titled ‘Christian Morals’) and in the 1950s appeared in BBC ‘Brains Trust’ debates with Ayer and Bertrand Russell. The unnamed author of Natural Theology, meanwhile, was the Dutch Jesuit Bernard Boedder, whose book had been published in the Stonyhurst Philosophical Series produced out of St. Mary’s Hall, the Jesuit seminary adjacent to Stonyhurst College.8 3  The book is likely to have been English Catholics in the Reign of Queen Elizabeth (London: Longmans, Green, 1920) by J. H. Pollen, SJ, vice-​postulator for the beatification of the English Catholic martyrs. 4  G. E. M. Anscombe, ‘I Am Sadly Theoretical’, Catholic Herald, July 8, 1938, 7. 5  Lord Longford, Avowed Intent (London: Little, Brown, 1994), 80. 6  A. J. Ayer, Part of My Life (London: Harper Collins, 1977), 87. 7  M. D’Arcy, ‘Knowledge according to Aquinas’, Proceedings of the Aristotelian Society [PAS] 28, 1928, 177–​202; ‘The Good and the Right’ PAS 32, 1932, 171–​206; St. Thomas Aquinas (London: Benn, 1930). 8  Bernard Boedder, Natural Theology, 2nd edition (London: Longmans, Green, 1902).

Anscombe and Aquinas    447 Also, in use amongst Catholic students in Oxford at the time of Anscombe’s arrival was A Primer of Moral Philosophy by Henry Keane, SJ, D’Arcy’s predecessor but one as master of Campion Hall, presenting natural law ethics with reference to Aquinas’s Summa Theologiae (Pars Secunda).9 Initially, then, Anscombe’s glimpses of Catholicism, philosophy, and Aquinas were through a Jesuit rather than a Dominican lens, though her reminiscence quoted earlier suggests she may then have been unaware of the philosophical significance of that distinction. Also, at that point anyone of her social background interested in Catholicism who read the quality press and listened to BBC radio would have encountered D’Arcy, who, in the words of a biographer, had become ‘the foremost English apologist for Roman Catholicism’.10 It is almost certain, therefore, that by the time she got to Oxford she was acquainted with something of his apologetic output. Taking account of all of this one might have expected that once there she would have gravitated towards D’Arcy and Campion Hall, but she did not. Why? The answer reveals aspects of her character and outlook then and thereafter. Pope Leo XIII is principally associated with two encyclicals: Aeterni Patris, ‘On the Restoration of Christian Philosophy’ (1879), and Rerum Novarum, ‘On Capital and Labour’ (1891). In the former he powerfully commended and even commanded the study and promotion of the thought of Aquinas, thereby inaugurating the neo-​Thomist revival.11 This was taken up by the Jesuits along with others, but while there was a pre-​existing tradition amongst them of studying Aquinas, they tended to see him less as an overarching authority than as a resource, as were Duns Scotus and their own Francisco Suarez. Hence, they did not have the undivided loyalty to St. Thomas nor the Aquinian sensibility of his Dominican confreres. For the latter the situation was different. While the Jesuits had been in Oxford since 1875, the Dominican studium was established there in 1929, only fifty years after Aeterni Patris, fewer than forty after Rerum Novarum, and less than a decade before Anscombe’s arrival in the university. The ‘newcomers’12 were enthusiastic about pursuing 9 

Interestingly, in light of Anscombe’s later work on action and ethics it contains short chapters such as ‘Human Acts’ setting out and discussing conditions of knowledge, voluntariness, and freedom, and ‘Human Acts and the Good for Man’, including a presentation of ‘the principle of double effect’ and identifying and examining ‘the determinants of morality’: ‘These Determinants are three: (1) The end which the agent has in view. (2) The means he takes to secure his end. (3) The circumstances in which he acts. For a good action all these three must be good. . . . If any one of them is evil, the act will be bad. . . . No act is in the concrete indifferent. . . . All acts [such as walking, eating, and a host of others] share in the good or evil of the end for which they are done. In practice there is no such thing as an act which is neither good or bad’. See H. Keane, A Primer of Moral Philosophy, 2nd edition (Oxford: Catholic Social Guild, 1931), 35–​41. Anscombe defends this Thomistic thesis in ‘Good and Bad Human Action’, GG1, 204–​206. For a fuller discussion and defence, see J. Haldane, ‘Is Every Action Morally Significant?’, Philosophy 86 (337) July 2011, 375–​404. 10  E. Yarnold, ‘Martin D’Arcy’, in D. Cannadine, ed., Oxford Dictionary of National Biography (Oxford: Oxford University Press, 2004). 11  ‘We exhort you, venerable brethren, in all earnestness to restore the golden wisdom of St. Thomas, and to spread it far and wide for the defence and beauty of the Catholic faith, for the good of society, and for the advantage of all the sciences’. Aeterni Patris (Vatican: Libreria Editrice Vaticana, 1879), para. 31. 12  The Dominicans first arrived in Oxford in 1221 and were expelled in 1538.

448   John Haldane and promoting Aquinas and applying his mode of thought to contemporary moral and social issues and keen to support students of whatever sort and background. The Dominicans are a mendicant order with a history of radicalism in the service of the poor. The Jesuits in that period, many of whom were Oxford graduates and themselves converts, were to the established and privileged Oxford world the culturally and socially acceptable face of Catholicism. D’Arcy’s relentless acquisition of religious artworks and artifacts (‘objets D’Arcy’) and the high-​table and common room hospitality appealed to some, but to others it seemed indulgent and disagreeable. There was not only a lack of contact between the two Catholic halls and a degree of long-​standing theological rivalry as in the matter mentioned by Anscombe of God’s knowledge of future contingents, but some divergence on issues of social justice. The only person and group named in her 1938 article are Fr. (Leo) O’Hea and the Catholic Social Guild. O’Hea was a Jesuit and a former contemporary of D’Arcy at Stonyhurst, but his charism and energies were differently directed in Oxford, where he ran the Catholic Workers College. Anscombe titles the section in which she quotes O’Hea approvingly on the sinfulness of injustice ‘Deeds Speak Louder’ and begins it, ‘[W]‌e can do less to convert people now than at any time by merely writing and talking’. Whether intended as such, that remark may have been read at Campion Hall as an implied rebuke to the D’Arcy style of apostolate. At any rate her outlook, combined with the outspoken expression of it, would have been more welcome at Blackfriars, where a ditty circulated amongst the Dominicans regarding the imagined D’Arcy call to the jeunesse dorée: Are you rich and nobly born? Is your soul by troubles torn? Come, and I shall heal them all: Martin D’Arcy, Campion Hall.13

If reading Jesuit authors in her mid-​teens brought Catholicism and philosophy forth in her, it was the Dominicans of Oxford and Peter Geach who developed her awareness of Aquinas and the relevance of his thought to her ethical and philosophical concerns. Geach, who was at Balliol College and three years ahead of her, had also been instructed and received into the Catholic Church by Fr. Kehoe in 1938. They did not meet, however, until early June, after their respective receptions. Writing five decades later, Geach dated his own reading of Aquinas to the time of his religious conversion: I knew that I must now completely rethink my philosophical position. So I began reading Aquinas: since I was a classical scholar, it was no great toil to read his lucid 13  There are several slightly different versions of this in circulation. In After the Victorians: The World Our Parents Knew (London: Hutchinson, 2005) A. N. Wilson quotes something close to this but erroneously attributes it to John Betjeman. Meanwhile H. J. A. Sire, Father Martin D’Arcy (Leominster: Gracewing, 1997), 70–​7 1, citing Alan Pryce Jones, The Bonus of Laughter (London: Faber, 1987), offers ‘Are you rich and nobly born? /​Is your soul with sorrow torn? /​Come and we shall find a way /​I’m Father D’Arcy, sir, SJ’.

Anscombe and Aquinas    449 prose in the original; I helped myself out as regards the subject matter with the faithful Dominican translation. What then counted as Thomist philosophy was in a bad way. . . . [P]‌rovidentially I did not approach Aquinas with such blind guides. I have kept on reading Aquinas ever since; I hope I have continued to learn from him. Two pieces of my work—​’Form and Existence’ in God and the Soul and ‘Aquinas’ in Three Philosophers—​have been directly concerned with Aquinas: his influence has been constant.14

During the following academic year Geach held a research studentship and spent much of the time reading through the Summa Theologiae and thereafter carried a pocket edition with him. He also became engaged to Anscombe and many years later described their intellectual relationship in those early years up to the point of her meeting Wittgenstein in 1942: The bond between us two in the first instance had comparatively little to do with shared philosophical thought: it was of course immensely important that both of us were Catholics and recent converts. Elizabeth did not take her degree till three years after we met, and we did not get married till then. For those years, and for some time after our marriage, Elizabeth had a lot of philosophical teaching from me; I could see she was good at the subject, but her real development was to come only under the powerful stimulus of Wittgenstein’s lectures and her personal conversations with him.15

The ‘faithful Dominican translation’ of the Summa to which Geach refers is that which was published in several volumes beginning in 1912.16 In the late 1950s the Dominicans decided to produce a new translation which would, in the words of the general editor (Thomas Gilby, OP), ‘render the thought of St Thomas into the freedom of another idiom without circumlocution or paraphrase’.17 While these later volumes have the merit of presenting the Latin text in parallel, the translations are often loose and the earlier work corresponds more accurately to Aquinas’s precise prose. This is why Geach writes (in the 1980s) of the ‘faithful’ translation. It was that earlier edition that he preferred to consult and an abridgement of which he carried. Anscombe was of the same mind. In our correspondence she wrote, ‘I don’t altogether like the translations you quote, and am not sure whether by the ‘Blackfriars’ edition you mean the newish one . . . or the old ‘English Dominican Fathers’ translation. It looks like the former. If so, you can probably make out by examining the Latin how much is slant from the translators’. When quoting him, however, she often used her own rendering, and her

14  P. Geach, ‘A Philosophical Autobiography’, in H. A. Lewis, ed., Peter Geach: Philosophical Encounters (Dordrecht: Kluwer, 1991), 8. 15  Geach, ‘A Philosophical Autobiography’, 10–​11. 16  The Summa Theologiae of St. Thomas Aquinas, 2nd and revised edition, literally translated by Fathers of the English Dominican Province (London: Washbourne, 1912–​). 17  St Thomas Aquinas, Summa Theologiae, Vol. 1 (London: Eyre & Spottiswoode, 1964), vii.

450   John Haldane papers include several undated translations (in her own hand) of parts of extracts from the Summa, including the whole of the four articles of Question 11 of the Prima Pars (First Part) on ‘The Unity of God’. In 1939, however, Anscombe’s focus was not on Aquinas’s philosophical theology but on aspects of his moral thought and the tradition it shaped as it might be applied to the circumstances and likely conduct of the anticipated conflict with Germany, and here the contemporary Dominican influence was important. Shortly after Chamberlain’s declaration of war she and fellow student Norman Daniel self-​ published the pamphlet The Justice of the Present War Examined: A Criticism Based on Traditional Catholic Principles and on Natural Reason. Part 1, ‘The War and the Moral Law’, is by Anscombe. It quotes or cites Aquinas four times on the nature of the natural (moral) law and its application to the conditions of just killing. The style of her referencing, though accurate, is idiosyncratic and inconsistent, suggesting unfamiliarity with standard conventions. What is more significant and not hitherto noted, however, is the resemblance of the structure of her presentation and her arguments to those advanced in the preceding two years in talks and publications by English Dominicans (including in their then monthly review Blackfriars), specifically Conrad Pepler, OP, Gerald Vann, OP, and Victor White, OP, who themselves invoked Aquinas and the Catholic Just War tradition deriving from him. There can be little doubt that what Anscombe wrote was influenced by discussions going on at Blackfriars Oxford, where Vann and White were teaching. Also, I think it likely that Vann’s 1939 Thomistic book, Morals Makyth Man, had some influence on her famous 1958 essay, ‘Modern Moral Philosophy’.18

20.3  Wittgenstein, Lewis, and Philosophical Psychology Geach mentions the powerful effect of her encounter and then sustained study with Wittgenstein (between 1942 and his death in 1951). In this period her mind was drawn away from the ethical and political preoccupations of her undergraduate years, and Wittgenstein’s hostility to philosophers’ opining on such issues may have been further reason to refrain from writing on them. That, however, does not diminish the likelihood that she thought about the relation of Aquinas’s methods and ideas to those of Wittgenstein. On his bookshelf throughout the period when Anscombe and Geach knew him up to the day of his death (alongside Augustine’s Confessions, Frege’s Grundlagen, and William James’s Principles of Psychology) were two volumes of Aquinas’s Summa Theologiae in Latin and German containing questions 1 to 26 of the Prima Pars, which 18 

I examine in some detail the matters discussed in this paragraph in ‘Anscombe: Life, Action and Ethics in Context’, Philosophical News 18, 2019, 45–​75.

Anscombe and Aquinas    451 are those concerning the existence and nature of God as objects of natural reason.19 In May 1953, the month that her translation of the Philosophische Untersuchungen (Philosophical Investigations) was published, Anscombe recorded a BBC radio talk (broadcast in July) in which she said, ‘[Wittgenstein had] also a certain limited liking for Aquinas: that is limited to some of the questions that Aquinas asked’.20 This is echoed by Rush Rhees, who along with Anscombe and Georg von Wright was an executor of Wittgenstein’s Nachlass: ‘The only remark of Wittgenstein’s about Aquinas that I can remember was that he found him extremely good in his formulation of questions but less satisfactory in his discussion of them’.21 According to Maurice Drury (another former student and lifelong friend), however, Wittgenstein had said to him that ‘Aquinas is really wonderful when he starts drawing distinctions’.22 Between 1939 and 1953 Anscombe published only two pieces: her ‘Reply to C. S. Lewis’s Argument That ‘Naturalism’ Is Self-​Refuting’ (1948)23 and ‘The Reality of the Past’ (1952).24 The latter is confessedly Wittgensteinian in character: ‘Everywhere in this paper I have imitated his ideas and methods of discussion’ (CP2, 114). Although the former is not so obviously Wittgensteinian in manner, it reflects his attention to distinctions between (concepts of) causal and rational features and relations. At one point she states, ‘[S]‌uch an argument [as Lewis’s] is based on a confusion between the concepts of cause and reason, which arises because of the ambiguity of such expressions as ‘because’ and ‘explanation’’ (CP2, 227). At the outset of the essay she has a footnote reading, ‘I wish to acknowledge that I was very greatly helped in writing this paper by discussing it with Mr Y. Smythies’ (CP2, 224). Like her, Yorick Smythies was a student and friend of Wittgenstein and a Catholic convert. She also wrote to Wittgenstein the day following her refutation of Lewis. But the distinction she draws and the point about different kinds of explanation are also derivable from Aquinas not as that between causality and reason but as one within the broad category of causality between material and efficient and formal and final causes, and again between the material and the formal aspects of hylomorphic unities, primarily animate substances. At various places in the Summa Aquinas refers to the bodily (material) and the rational aspects of action and speech, not troubling about any conflict or incompatibility between these. Yet he also

19 

See Garth Hallett, A Companion to Wittgenstein’s Philosophical Investigations (Ithaca, NY: Cornell University Press, 1977), 761. I asked Anscombe what happened to these and she said that the books had gone to Rush Rhees and that when he died his widow had sold or donated most of his library. I subsequently made further inquiries which led me to think that they had been ‘binned’. 20  See Christian Erbacher, ‘“Ludwig Wittgenstein”—​A BBC Radio Talk by Elizabeth Anscombe’, Nordic Wittgenstein Review 8 (1–​2) 2019, 225–​240, 232. 21  As reported from a letter from Rhees to Garth Hallett quoted in Hallett, A Companion to Wittgenstein’s ‘Philosophical Investigations” (Ithaca, NY: Cornell University Press, 1977), 761, also documented in G. Citron, ‘Wittgenstein’s Philosophical Conversations with Rush Rhees (1939–​50): From the Notes of Rush Rhees’, Mind 124 (493) 2015, 54. 22  See G. Citron, ibid. 23 In Socratic Digest 4 (2) 1948, 7–​16, reprinted in CP2. 24  In Max Black, ed., Philosophical Analysis (New York: Ithaca Press, 1950), 36–​56, reprinted in CP2.

452   John Haldane distinguishes between different kinds of explanation allowing that otherwise aligned and cooperative causes may come apart.25 Interestingly, reviewing The Nature of Belief in the Journal of Theological Studies in 1933, F. R. Tennant charged D’Arcy’s work with eliding distinctions in terms that anticipate those of Anscombe’s 1948 critique of Lewis: [C]‌auses of believing are sometimes treated as if identical with reasons or grounds of believed propositions; and thus distinctly logical questions are confused with distinctly psychological questions, and the solution of the whole problem is made to appear somewhat easier than it really is.26

D’Arcy himself read three papers to the Socratic Club between 1943 and 1944, and given that they were orthodox co-​religionists it is possible that he and Anscombe were jointly present on some occasions, though D’Arcy was in America at the time of her exchange with Lewis. Although a small minority in the university, Roman Catholic laity and religious contributed significantly (fifty papers) to meetings of the Socratic Club from the year of its inception (1942) to the last of the listed meetings (1969). Lay philosophers presented twenty-​three papers: Anscombe read four; Geach six; Michael Dummett, Brian McGuiness, and C. C. W. Taylor two each. Six Dominicans presented fifteen papers, including four from Richard Kehoe, two each from Gerald Vann and Victor White, and one from Conrad Pepler. The Jesuits contributed eleven, including three from Thomas Corbishley, D’Arcy’s successor at Campion, and two from Frederick Copleston. Anthony Kenny, of whom more shortly, presented twice while a priest, and once subsequent to his laicisation. From 1946 to 1952 Anscombe held a research fellowship at Somerville College, then for the next six years was supported by the Rockefeller Foundation in her work editing and translating Wittgenstein. Towards the end of this second period she began to pursue her other interests, giving graduate seminars and returning to the theme of justice in the conduct of war. Out of this emerged another self-​published pamphlet, Mr Truman’s Degree (1956), Intention (1957), and the other essays in what Aquinas terms philosophia practica. In ‘The War and the Moral Law’ she had a section on ‘double effect’ cast in terms of directly intended and incidental consequences, and citing Aquinas (ST, 2a 2ae, q.64) she writes, ‘No action can be excused whose consequences involve a greater evil than the good of the action itself, whether these consequences are accidental or not’.27 25  In this connection, though not as referring to Aquinas, see Anscombe’s further thoughts, almost forty years later, on Lewis’s revised argument: ‘C. S. Lewis’s Rewrite of Chapter III of Miracles’, in R. White, J. Wolfe, and B. Wolfe, eds., C. S. Lewis and His Circle: Essays and Memoirs from the Oxford C. S. Lewis Society (Oxford: Oxford University Press, 2015). 26  F. R. Tennant, review of Martin D’Arcy, The Nature of Belief, Journal of Theological Studies 34 (135) July 1933, 287. 27  CP3, 78. In later writings Anscombe considers and endorses a course of action that would seem to violate the principle here stated (see CP3, 35), but the matter admits of further specification and refinement.

Anscombe and Aquinas    453 ‘Double effect’ does not feature in the Truman essay, but it gets attention in two pieces written a little later: a critical notice of Glanville Williams’s The Sanctity of Life and the Criminal Law (declined by the commissioning journal and published only posthumously) and ‘War and Murder’ (1961), which replicates a few passages from the former where she characterises double effect by reference to a distinction between the intended and merely foreseen effects of voluntary actions. She observes, ‘The denial of [the principle of ‘double effect’] has been the corruption of non-​Catholic thought, and its abuse the corruption of Catholic Thought’,28 and she relates the latter abuse to the adoption from the seventeenth century of ‘Cartesian psychology’, according to which ‘an intention was an interior act of the mind which could be produced at will’. The implied contrast is with earlier medieval views according to which intention is an intrinsic aspect of action, a position she argued for on her own account in Intention. Concern about the (mis)application of double effect and about the invocation of Aquinas in relation to it became more pronounced in later writings. First, she substitutes the expression ‘principle of side-​effects’, illustrated as merely stating a possibility: ‘[W]‌here you may not aim at someone’s death, causing it does not necessarily incur guilt. . . . Without such excuse [of sufficient necessities or legitimate purposes], foreseeable killing is either murder or manslaughter’.29 By the time of her address (1982) on receiving the Aquinas medal from the American Catholic Philosophical Association she was keen, first, to emphasise how limited the side-​effects principle is (causing an unintended side effect does not necessarily incur guilt) and, second, to disassociate Aquinas himself from what by then was widely presumed to be the substance of double effect justifications. She writes: I will end by protesting at the ascription of the Doctrine of Double Effect to Aquinas: the phrase ‘duplex effectus’ occurs in his discussion of killing in self-​defence (Summa Theologica 2a 2ae, q.64, a.7). [Latin text follows] . . . Those who wished to see in this text the package Doctrine of Double Effect claimed that in speaking of proportionate means St Thomas was introducing their doctrine of a proportion of good over evil in the upshot. I am not concerned to discuss St Thomas’ view of self-​ defence, only to note the false interpretation. If one wants to know St Thomas’ general opinion for responsibility for evil consequences of actions, this is not the place to look, but rather at what he has to say about the relation of an eventus sequens [consequence] to the goodness or badness of an action (1a 2ae, q.20 a.5). (GG1, 225–​226)

And in 1990, in what must be one of the shortest academic pieces ever published, she writes in response to Michael Coughlan’s ‘Using People’, ‘He assumes that ‘the principle

28 

29 

CP3, 54. ‘Murder and the Morality of Euthanasia’, in GG1, 274–​275.

454   John Haldane of double effect’ is supposed to exonerate a causer of any evils so long as they are not intended as means or end. Error’.30

20.4  The Philosophical Enquiry Group, Anthony Kenny, and Natural Theology Between invoking Aquinas’s ST 2a 2ae in 1939 and defending it in 1982 and later against misinterpretation, Anscombe had continued to engage with the Dominicans and with other Catholic academics. In 1950 she brought Conrad Pepler of the Oxford Priory, and former editor of Blackfriars, to talk with the ailing Wittgenstein, who had asked her to arrange for him to meet a ‘non-​philosophical’ priest. Later Pepler, who had instructed Smythies during his conversion, arranged for Wittgenstein to stay at the Dominican Priory in Leicester, but by then he was too ill to do so. It was Pepler who attended with Anscombe, Smythies, Drury, and Ben Richards at Wittgenstein’s deathbed and who conducted the graveside prayers the following day. The next year (1952) Pepler left Oxford to become warden of the Dominican conference centre Spode House at Hawkesyard Priory, where two years later Fr. Columba Ryan, OP, founded the Philosophical Enquiry Group, which met there annually for the next twenty years.31 Anscombe and Geach were regular attenders; others included James R. Cameron, Michael Dummett, Brian McGuiness, Anthony Kenny, and the Dominicans Cornelius Ernst and Herbert McCabe. The application of Aquinian ideas to matters of current philosophical and theological interest was a mainstay, and Anscombe presented a number of papers to meetings, including ‘Prophecy and Miracles’ (1957),32 ‘The Immortality of the Soul’ (1960s),33 ‘On Being in Good Faith’ (1960s),34 and ‘Contraception and Natural Law’ (1964).35 Aquinas is cited in the third and fourth of these (in relation to acting against, and in accord with erroneous conscience, and on the mutual obligations of spouses in respect of sex), but he is conspicuously absent from the second (for a reason I return to). It and the short essay on contraception were presented in partnership with papers on the same topics by McCabe, who, while a great admirer and expositor of Aquinas, was very taken with trends in analytic philosophy 30  G. E. M. Anscombe, ‘A Comment on Coughlan’s “Using People,” ’ Bioethics 4 (1) 1990, 62. Her next longest article, also contra-​consequentialism, runs to forty-​nine words: ‘A Note on Mr Bennett’, Analysis 26 (6) June 1966, 208. 31  On the role of this and of the Dominicans more generally in helping to bring Aquinas into the orbit of British philosophers, see J. Haldane, ‘Late Twentieth Century Reception of Aquinas in Analytical Philosophy’, in M. Levering and M. Plested, eds., The Oxford Handbook of the Reception of Aquinas (Oxford: Oxford University Press, 2021), 515–​535. 32  GG2, 20–​39. 33  GG2, 69–​83. 34  GG2, 101–​112. 35  New Blackfriars 46 (540) June 1965, 517–​521.

Anscombe and Aquinas    455 represented by the work of Gilbert Ryle, Peter Strawson, and Geach, and with what had recently become available of Wittgenstein’s ideas about thought and language through Anscombe’s edition of the Philosophical Investigations. She and Geach were prized as the leading philosophers in the Spode House group, and it was McCabe who introduced Anthony Kenny to them there. Kenny had trained for the priesthood at the English College in Rome, where he had studied at the Jesuit Gregorian University.36 Although classes there were in an eclectic, neo-​scholastic mode, he had got interested in analytic philosophy, taking advice on reading from Frederick Copleston, who was on the faculty there. Having completed his licentiate in theology Kenny proceeded to a doctorate under the supervision of another anglophone Jesuit philosopher, the Canadian Bernard Lonergan. This research was more directly on analytic philosophy, specifically ‘linguistic analysis and religious language’, and on that account he was sent to Oxford in 1957 to complete his dissertation as a visiting student. This opportunity led to his also enrolling for an Oxford doctorate, and so began a journey that took Kenny away from the priesthood but onwards into and upwards within the world of Oxford philosophy. His view of the value of Aquinas’s thought, which hitherto he had studied only as a set text (De Ente et Essentia), was transformed by observing Geach’s use of it; the year of his arrival was that of the publication of Geach’s Mental Acts as well as of Intention. He also began to learn more deeply about Wittgenstein by attending Anscombe’s graduate seminars on themes and arguments in the Investigations and in 1958 presented ‘Wittgenstein and Aquinas’ to the Socratic Club, at which he was a frequent attender.37 As Kenny’s faith in Catholicism waned, so grew his doubts about the rational credentials of theistic belief. The First Vatican Council decreed that ‘the Catholic Church holds and teaches that God, the source and end of all things, can be known with certainty from the consideration of created things, by the natural power of human reason’.38 Whether beyond the ‘thatness’ (existence) of God reason could also determine the ‘whatness’ (nature) of God is a further matter, but the first was already becoming a problem for Kenny. In 1959 he left Oxford to serve in a Liverpool parish and with doubts growing began a two-​track correspondence with Anscombe the following year, one part of which concerned the relation of faith, knowledge, and certitude, the other part theistic proofs and specifically the cosmological argument. The interest of this correspondence in the present context is threefold: it returned Anscombe to her early preoccupation with improving the Jesuit Fr. Boedder’s proof; it led her to study relevant Aquinian texts; and it induced her to publish several essays related to these concerns

36 

For an account of his life during the period covered here, see A. Kenny, A Path from Rome (Oxford: Oxford University Press, 1986), and on his encounter and dealings with Anscombe, see ‘Elizabeth Anscombe at Oxford’, in J. Haldane, ed., The Life and Philosophy of Elizabeth Anscombe (Exeter: Imprint Academic, 2019), 12–​22. 37  Published as ‘Aquinas and Wittgenstein’, Downside Review 77 (249) 1959, 217–​235. 38  Decrees of the First Vatican Council, Session 3, April 24, 1870, ch. 2, ‘On Revelation’, 1, see https://​ www.papal​ency​clic​als.net/​counc​ils/​ecu​m20.htm.

456   John Haldane and to challenges made by Kenny, though (save in one unnoted case) those essays were not explicitly linked to theistic arguments.39 Question 2 of the Prima Pars of the Summa consists of three articles, all related to the general issue of the existence of God. With characteristic brevity and directness Aquinas writes, ‘primo considerandum est an Deus sit’ (first, we consider whether God exists). He prefaces the presentation of his theistic proofs by differentiating between two kinds of arguments: those a priori which proceed from knowledge of the existence and nature of a thing to its effects, and those a posteriori which reason from effects to their cause(s). His ‘five ways’ are all instances of the latter sort, the second based on efficient causation. Aquinas writes: There is no case known (neither is it, indeed, possible) in which a thing is found to be the efficient cause of itself; for so it would be prior to itself, which is impossible. Now in efficient causes it is not possible to go on to infinity, because in all efficient causes following in order, the first is the cause of the intermediate cause, and the intermediate is the cause of the ultimate cause, whether the intermediate cause be several, or only one. Now to take away the cause is to take away the effect. Therefore, if there be no first cause among efficient causes, there will be no ultimate, nor any intermediate cause. But if in efficient causes it is possible to go on to infinity, there will be no first efficient cause, neither will there be an ultimate effect, nor any intermediate efficient causes; all of which is plainly false. Therefore, it is necessary to admit a first efficient cause, to which everyone gives the name of God.40

In Anscombe’s account of her route into philosophy she refers to the author of Natural Theology offering ‘a proof of some ‘principle of causality’ according to which anything that comes about must have a cause’. Boedder’s first statement of the principle is ‘[E]‌verything that has a beginning has a cause’ and his second is ‘[W]hatever does not exist of absolute necessity, cannot exist without a proportionate cause’41 (he takes this latter to express what is implicit in the former), from which he proceeds to the conclusion: ‘Not all beings can be effects; there must be something which is a cause without 39 

The piece in question, ‘Hume Reconsidered’, discussed later, is not listed in the bibliography of Roger Teichmann’s study The Philosophy of Elizabeth Anscombe (Oxford: Oxford University Press, 2008) or in the updated (2012) version of that produced by Luke Gormally, Christian Kietzmann, and Jose Maria Torralba, on an earlier edition of which Teichmann’s is based. The only reference antecedent to this that I know of is by Anthony Kenny in his book The Five Ways: St Thomas Aquinas’ Proofs of the Existence of God (London: Routledge & Kegan Paul, 1969), 67, where he writes, ‘The fallaciousness of Hume’s argument was pointed out by Miss Anscombe in [“Hume Reconsidered”]’. Kenny’s book derived from a lecture course given in Oxford in 1967–​1968 and bears slight traces of his correspondence with Anscombe of 1959–​1960. She did not refer to it in writings, including those published posthumously, but Geach reviewed it in Philosophical Quarterly 20 (80) July 1970, 311–​312. See also his own essay ‘Causality and Creation’, Sophia 1 (1) 1962, reprinted in P. Geach, God and the Soul (London: Routledge & Kegan Paul, 1969), which also has thematic links to the Anscombe-​Kenny correspondence of which he was aware. 40  Summa Theologiae, 1a, q.2, a.3. 41 Boedder, Natural Theology, 32–​33.

Anscombe and Aquinas    457 being the effect of another cause, and this something must be self-​existent’.42 In her correspondence with Kenny,43 Anscombe begins with something seemingly simpler, namely, the question of how the whole world comes to be, and the answer that it must be produced by something capable of producing it, adding that this something is what people mean by ‘God’. The mention of ‘a capable’ cause is not, I think, tautological but a reference to what Boedder termed ‘a proportionate’ one; while the last part echoes a refrain used by Aquinas after each of the five ways, in the case of the second, ‘quod omnes dicum Deum’ (this all speak of as God). Kenny responds by arguing that the question of the coming to be of things is ambiguous between asking of each thing what caused it, and of asking of some totality, ‘the World’, what was the cause of that. He allows that the first is straightforward and the second is intelligible if treated as equivalent to an aggregation of questions about individual things, but otherwise it has no sense. From this opening exchange there unfolds a long and interesting to and fro covering such themes as the nature of change, intentional verbs, existence, time, God’s knowledge of future contingents, explanation, and causality. In the course of this Aquinas is often referred to and quoted by each of them, with Anscombe correcting herself over an interpretation of what he says about God’s eternity, having reread the parts of the Summa and consulted the De Veritate, adding that she prefers the former to the latter because of its easier style and crisper thought. She states succinctly her interest in looking to his work: ‘I am willing to examine St Thomas in the hope of getting clear about the matter in hand. . . . ‘Perhaps there is some truth to be found here’ is my only reason for considering him at all’.44 After ranging wide, the discussion returns to the original issue of seeking causes of things and of the ‘World’. This begins to shift from the intelligibility of asking the cause of the latter to the principle that whatever comes into existence comes to be from something. Here an argument which she finds in Hume is considered, which she would make the subject of several publications: that the idea of something coming into existence without a cause is intelligible since we can without contradiction imagine cases of it.45 Anscombe considers this and the broader issues in two essays published in the mid-​1970s: ‘Whatever Has a Beginning of Existence Must Have a Cause’ and ‘Times, Beginnings and Causes’.46 Her response is twofold: first, there is a difference between (a) imagining a beginning-​of-​existence of x without imagining a cause of it and 42 Boedder, Natural Theology, 34.

43  She kept this in a notebook with his postcard replies taped in and her notes and responses recorded. It runs to more than one hundred pages. This is part of the Anscombe Papers originally gathered and catalogued by Luke Gormally and now in the ownership of the Collegium Institute in Philadelphia, currently on loan to the Kislak Center for Special Collections in the University of Pennsylvania Library; see https://​www.col​legi​umin​stit​ute.org/​ansco​mbe-​arch​ive-​info. I am grateful to Dr. Daniel Cheely, Executive Director of the Collegium Institute and his staff for providing me with scans of material from the Anscombe Archive. 44 Anscombe, Notebook, Anscombe Archive, University of Pennsylvania Library, File 559, 46. 45  See D. Hume, A Treatise of Human Nature, book I, part III, section III: ‘Why a Cause Is Always Necessary’. 46  The former originally in Analysis 34 (5) 1974, and the latter in Proceeding of the British Academy 60, 1974. The first is reprinted in CP1, the second in CP2. Two other articles that can be traced in part to the

458   John Haldane (b) imagining a beginning-​of-​existence-​of-​x-​without-​any-​cause-​of-​it; and second, that the latter (which is what is required for Hume’s argument) is not shown to be possible by imagining x appearing and nothing else, and describing this as a case of (b), for that is simply to title the image, not to show that it thereby represents a real possibility. In the second of these publications she introduces Aquinas as holding that ‘the relation to a cause does not enter into the definition of the thing that is caused’ (ST, 1a, q.44, a.1), but the purpose is to show that this is a different point and it does not help Hume’s case. What she does not mention in either of these 1970s publications is the interest of the issue of the principle of causality for proving the existence of a First Cause. However, in a review article, ‘Hume Reconsidered’, written shortly after the conclusion of the correspondence with Kenny in which she reproduces phrases and points from that, she writes: Anyone for example who has an interest in natural theology should consider Hume. For if Hume is right then there is no such science. . . . [I]‌t is not permissible to claim that one can give a cosmological argument that ought to be accepted by philosophers, if one cannot answer this well known cavil [about the principle of causality].47

And then, having made her points against Hume, she concludes: But if it is true, in whatever manner, that every beginning of existence does have a cause, then, since causes cannot be very well supposed to move in a circle, it follows that either there is an infinity of beginnings of existence produced by causes which in their turn began to exist, or there is at least one cause which had no beginning of existence.48

Anscombe does not expressly eliminate the first disjunct, as does Aquinas, though she may intend to leave that to the reader. So far as I am aware there are only two other places in her writings where she entertains what look to be (proto)theistic arguments.49 One is her speculative re-​interpretation of Anselm’s Proslogion 2 proof as something other than an ontological one.50 She argues that the traditional presentation of it as such requires a premise that is not in Anselm’s own argument but is provided by Aquinas in his rejection of it (ST, 1a, q.2, a.1 ad 2). It is uncertain how to characterise the proof as Kenny postcard correspondence are ‘Before and After’ (1964) and ‘Causality and Determination’ (1971), both reprinted in CP2. 47 

G. E. M. Anscombe, ‘Hume Reconsidered’, Blackfriars 43 (502) April 1962, 188. Anscombe, ‘Hume Reconsidered’, 189. 49  Among her papers is a one-​page partial translation and summary of Aquinas’s Tertia via. This is not developed there or elsewhere, but it is of interest in claiming that Aquinas is concerned with a totality which perhaps relates to her opening question in the Kenny correspondence, where she asks about how the ‘World’ came to be. 50  G. E. M. Anscombe, ‘Why Anselm’s Proof in the Proslogion Is Not an Ontological Argument’, Thoreau Quarterly 17, 1985, 32–​40 and ‘Russellm or Anselm’, Philosophical Quarterly 43, 1993, 500–​504. Both are reprinted in GG3. 48 

Anscombe and Aquinas    459 she understands it, though it looks to remain a conceptual a priori one. More in keeping with Aquinas’s effect-​to-​cause approach is an argument briefly stated in ‘Human Essence’, where the starting point is the existence of humans as rational beings (manifest in their use of language) and the bridging premise that what produces rational beings must itself be an intelligence.51 This might be seen as a combination of teleological and cosmological reasoning, but what is most clear is that for at least five decades from her pre-​university days through to her early years at Cambridge, Anscombe was engaged intermittently but at times intensively with the task of making good something akin to Aquinas’s Secunda via.

20.5  The Return to Ethics and Action Around the same time as the last essays on the principle of causality Anscombe began to devote her attention and energies to considering once more issues in philosophy of action and practical ethics, and Aquinas was relevant to both. Twenty years earlier in Intention she drew on the Summa in partial characterisation of practical knowledge as ‘the cause of what it understands’.52 This is a parallel application to the case of human agency (and rational agency more generally) of the principle scientia Dei causa rerum,53 and in the passage she cites (ST, 1a 2ae, q.3, a.5, obj. 1) Aquinas links the two cases, writing that ‘man is like God, by his practical intellect, which is the cause of things understood, rather than by his speculative intellect, which derives its knowledge from things’; in the prologue to the Prima Secundae he relates this to the imago Dei doctrine and extends it to freedom and direction, writing, ‘[I]‌t remains for us to treat of His image, i.e. man, inasmuch as he too is the principle of his actions, as having free-​will and control of his actions’ (ST, 1a 2ae, praefatio). While Aquinas is named and cited only once in Intention he is elsewhere in the background and sometimes close to the text. On the general matter of agents’ knowledge of the intentions with which they act, their intention in acting, as well of what they are doing intentionally, Anscombe may have been influenced by what Aquinas says in answer to the question ‘Whether intention is an act of the intellect or will?’ (ST, 1a 2ae, q.12, a.1). Favouring the latter, he considers the suggestion that it is a matter of a cognitive ‘perceptive power’ (apprehensivae potentiae) and responds, ‘Intention is called a light because it is 51 

‘Human Essence’, in GG1. Intention, §48, p. 87. 53  Those who have noted Anscombe’s reference to the Summa on this point seem to assume that the idea that God’s knowledge of things is practical (causal) originates with Aquinas, but it is already explicit in Augustine, De Trinitate, 15, 13, where he writes that ‘with regard to created beings, both spiritual and corporal, He does not know them because they are, but they are because he knows them’ (non quia sunt ideo nouit, sed ideo sunt quia nouit). For some discussion of the relation of Anscombe’s thinking about practical knowledge to ideas in Aquinas, see John Schwenkler Anscombe’s Intention: A Guide (Oxford: Oxford University Press, 2019), 6: ‘Practical Knowledge’. 52 

460   John Haldane manifest to him who intends. Wherefore works are called darkness because a man knows what he intends, but knows not what the result may be’ (ST, 1a 2ae, q.12, a.1 ad 2). Here we have the idea that while primarily a matter of the will, intention is also known directly in action, another indication of its belonging to the sphere of practical knowledge. Later, in considering constraints on possible objects of wanting, she introduces the idea, in explication of Aristotle’s practical syllogism, that an expression of wanting has to represent its object as somehow desirable (to the agent), and goes on to say, ‘The conceptual connection between ‘wanting’ (in the sense which we have isolated . . .) and ‘good’ can be compared to the conceptual connection between ‘judgement’ and ‘truth’. . . . [T]‌he notion of ‘good’ that has to be introduced in an account of wanting is not that of what is really good but of what the agent conceives to be good’.54 The corresponding points are made in several places by Aquinas most relevantly at ST 1a 2ae, q.9 a.1, where he writes, ‘The will is a rational appetite. . . . [S]o the voluntary appetite tends to a good which is apprehended. Consequently, in order that the will tend to anything, it is required not that this really be good, but that it be apprehended as good [non requiritur quod sit bonum in rei veritate, sed quod apprehendatur in ratione boni]’. And more generally he holds that the first principle of practical reason is self-​evident, being ‘founded on the notion of good, viz. that “good is that which all things seek after.” Hence this is the first precept of law, that “good is to be done and pursued, and evil is to be avoided” ’ (ST, 1a 2ae, q.94, a.2). It is a central claim of Intention that actions are intentional under descriptions and that not every true description of what is done is a description of it as intentional. This relates to the issue of intentionality, not in the specific sense of being intended but in the more general respect of being thought or done under an aspect or via a conception which spans both agency and cognition, the second of which is discussed by Anscombe in several places but most extensively in ‘The Intentionality of Sensation’ (1965), where she deploys a distinction between ‘formal’ and ‘material’ objects. This is an adaptation of the Thomistic distinction between obiectum materiale and objectum formale quo (formal object under which) derived from Aquinas, as when he writes, ‘Sacred doctrine is one science. The unity of a faculty or habit is to be gauged by its object, not indeed, in its material aspect [quidem materialiter] but as regards the precise formality under which it is an object [secundum rationem formalem obiecti]’ (ST, 1a, q.1, a.3). In explanation of her usage Anscombe writes, ‘An intentional object is given by a word or phrase which gives a description under which. It will help if we consider shooting at, aiming. A man aims at a stag [formal object]; but the thing he took for a stag was his father [material object], and he shoots his father’.55 Looking then to Aquinas we find him writing, when considering whether ignorance constitutes involuntariness, that ‘ignorance stands in [relation to the act of will] concomitantly, when there is ignorance of what is done, yet so that if it were known, it still would be done . . . as in the example, when someone wished indeed to kill an enemy, but killed him in ignorance,

54  55 

40, p. 76. CP2, 9–​10.

Anscombe and Aquinas    461 thinking to kill a stag. Such ignorance does not make an act involuntary, because the outcome of it is not against the will, but it makes an act not voluntary, because that cannot be actually willed which is unknown’ (ST, 1a, 2ae, q.6, a.8). Somewhat relatedly in discussing the question ‘[W]‌hether someone is guilty of murder through killing someone by chance?’ (ST, 2a 2ae, q.64, a.8) Aquinas introduces the idea of a cause which does not directly produce an effect but removes an obstacle to its occurrence, terming this a ‘causa removens prohibens’, a phrase that Anscombe introduces in ‘The Causation of Action’ (1983) to illustrate a contrast within the category of efficient causes between directly active ones and others that remove an obstacle to a cause taking effect.56 As the years proceeded, Anscombe’s interests turned more to practical ethics and casuistry in the sense of working through and out from morally perplexing situations, and in these she became more explicit in drawing upon Aquinas, just as she had done as an undergraduate protesting the injustice of modern warfare. I cannot discuss these here, but readers interested in exploring them should look to the references given in the appendix below. 57

20.6  An Admirer but Not a Disciple There is only one publication by Anscombe that is directly about Aquinas, and that was a 1965 review of the first few volumes of the new English Dominican translation of the Summa Theologiae. This is concerned largely with the fallacious inference that since what is known must be true, therefore only what is necessarily true can be known, and with the question of whether Aquinas commits or evades the fallacy. Anscombe finds him largely but not entirely innocent of doing so. Towards the end she turns from medieval to contemporary philosophy and suggests that it is now (1965) better placed to appreciate ideas from the former because of its own (re)discovery of the non-​contingent connection between mental acts and their objects. She writes: It seems clear that thoughts are characterized by what they are of, with no substantive being of their own; but how this is so is so intensely obscure that one surveys the obscurities of the scholastic esse intelligibile, whose actuality is the same thing as the actual occurrence of a thought of such-​and-​such, with a not totally unfavourable eye.58

In the event, ‘not totally unfavourable’ was not a case of English understatement actually meaning ‘favourable’ (and if American ‘very favourable’ and maybe even ‘enthusiastic’),

56 

Reprinted in GG1. See also Candace Vogler, ‘Aristotle, Aquinas, Anscombe, and the New Virtue Ethics’, in T. Hoffman, J. Muller, and M. Perkams, eds., Aquinas and the Nicomachean Ethics (Cambridge: Cambridge University Press, 2013), 239–​257. 58  G. E. M. Anscombe, ‘Necessity and Truth’, Times Literary Supplement, February 14, 1965, reprinted in CP1, 85. 57 

462   John Haldane for in posthumously published writing59 Anscombe explicitly expressed incomprehension at the Aquinian account of thought as involving the reception of the intelligible forms of things—​a position Peter Geach expounded and discussed approvingly in his work.60 In consequence, in another essay she also dismissed the favoured Thomistic proof of the immateriality of the intellectual soul based on this reception thesis combined with the claim that, since in receiving cognitively the form of a material object one does not become an object of that kind, intentional assimilation must therefore be immaterial. From which it is further argued that since powers express the nature of the agent, that which has the power to receive forms immaterially must itself be immaterial.61 I cited this second essay earlier as one given to the Philosophical Enquiry Group, saying that its absence of reference to Aquinas is conspicuous. That is in part because of the Dominican setting of its presentation and the fact that many and perhaps most of those present would have, if not endorsed the proof, looked upon it ‘with a not totally unfavourable eye’. But also because the paper was paired with one of the same title by Herbert McCabe, in which he presents that very Aquinian argument.62 Additionally, without it or some functional equivalent of it (which Anscombe never sought to provide) the question of the possibility of surviving death is given no positive answer. As became clearer, however, it was not only the argument that she disfavoured but its conclusion also—​at least conceived of as a piece of metaphysics. Anscombe professed the Catholic teaching that it is proper to pray for the departed souls but also to seek their intercession with God, but at the same time found its ontological presupposition impossible to make sense of and counselled remaining philosophically abstemious on the issue.63 This is significant so far as concerns her relation to Aquinas. For unlike customary Thomists she did not approach his work as a system to be adopted and applied in line with his own teachings. Instead, she saw him as a resource to be consulted, and where warranted to follow not as a magister but as the author of an insight or telling argument. In a late interview she was invited to respond to the observation that ‘Thomas Aquinas also held that the spiritual nature of the soul seemed to him to be rational grounds for affirming the possibility of the soul surviving death’.64 Her reply is discouraging to the interviewer and likely to be so to many of her Catholic and otherwise religious admirers: Probably he did. I would say that to him it was a problem and it is not clear to me that he solved the problem. I don’t think I know his writings on the subject well enough to say, but I would expect that he thought that he had solved the problem. What I do

59 

See ‘Thought and Existent Objects’, in GG4. See, for example, ‘Form and Existence’, in P. Geach, ed., God and the Soul (London: Routledge & Kegan Paul, 1969), 42–​64 and ‘On Truth as an End’, in P. Geach, ed., Truth and Hope (Notre Dame, IN: University of Notre Dame Press, 2001), 67–​78. 61  See ‘The Immortality of the Soul’, in GG2. 62  In A. Kenny, ed., Aquinas: A Collection of Critical Essays (London: Macmillan, 1969), 297–​306. 63  I discuss and argue against Anscombe’s position, and the grounds of it, in ‘Anscombe and Geach on Mind and Soul’, in Haldane, The Life and Philosophy of Elizabeth Anscombe. 64  R. A. Varghese, ed., Great Thinkers on Great Questions (Oxford: One World, 1998), 55. 60 

Anscombe and Aquinas    463 note is that for him it was a problem and that is the right approach to this matter, which should be a problem.65

Given her familiarity with and study of aspects of Aquinas’s writings as evidenced in the foregoing, including her claim to find the argument from the nature of thought to the immateriality of the soul unconvincing, it is hard to resist the conclusion that in saying ‘I don’t think I know his writings on the subject well enough’ she is either disingenuous or had deliberately refrained from reading relevant parts in the expectation that she would have to reject them. This, however, is speculation; what is evident from the preceding discussion is that she had read and thought a good deal about some of Aquinas’s ideas from her days as an undergraduate outraged at social, economic, and military injustices through to her mature but still revolutionary writings on action and causation. Unlike Peter Geach she did not invoke Aquinas widely or write at length on his ideas. While she had great respect for him, her pre-​ modern master was Aristotle, and she assumed the traditional reading of Aquinas as an Aristotelian who reiterated, amplified, and supplemented Aristotle in ways that allowed for the incorporation of theistic and specifically Christian ideas. She also tended to side with a more naturalistic Aristotelianism against Thomas’s attempts to use arguments in the De Anima to establish the existence of a separable immaterial soul. One area, however, where she certainly found him more satisfactory than Aristotle was in the analysis of actions and of the constituents and aspects of them as these contribute to their ethical quality. In the correspondence from which I quoted at the head of this essay she wrote: It is very remarkable how many things contribute to the goodness or badness of an action; evidently not just the ‘kind’ or ‘species’ of action that it is. And also, not the end only, given a good ‘species’, For you could give alms to get honoured by men; a good species of act deformed by what—​in the circumstances—​turns out to be a bad end. But even a good kind of act done with a good end in view may be rendered bad through the circumstances; as that it involved neglect of something it was necessary you should do. All these complications, and more, seem to be very well discussed by St Thomas [referring to ST, 1a 2ae, q.18, aa.1–​9]. I daresay you have read [1a 2ae] Q6 article 3, as you do refer to Q6. Possibly article 3 did not signify much to your interests. I suspect that if you consider it carefully you might de-​simplify your picture a bit. Is the voluntary failure of the pilot through whose neglect the ship sinks an act-​by-​omission? Must voluntary failure consist in conscious non-​performance? or in any act of choice or chosen act? Must it be intentional?66

65 Varghese, Great Thinkers on Great Questions, 55.

66  For discussion of these matters, see Anscombe, ‘Good and Bad Human Action’, in GG1; Haldane, ‘Is Every Action Morally Significant?’; J. Haldane, ‘Identifying Privative Causes’, Analysis 71 (4), October 2011, 611–​619.

464   John Haldane For the republication of Intention by Harvard University Press in 2000, Donald Davidson provided a remarkable, oft-​quoted endorsement: ‘Intention is the most important treatment of action since Aristotle’. High praise from high quarters but, unless one regards it as allowable hyperbole elicited by the occasion, it should be treated with caution. Davidson himself read a little of Aquinas on the topics of voluntariness (ST, 1a 2ae, q.6) and on passion as a cause of wrongdoing (ST, 1a 2ae, q.77).67 Aquinas’s moral psychology and theory of agency are, however, developed through some seventy questions, and even the sixteen restricted to the analysis of human acts contain ninety-​four articles; his is by far the clearest candidate for the award ‘the most important treatment of action since Aristotle’ (up to the point of Intention). Also, and this is the more salient point, had Davidson been aware of the range of Aquinas’s investigations (and of Anscombe’s) he might also have realised ways in which she is aligned with Aquinas not in the manner of a disciple or an adherent expositor but rather as one who was long inspired by his insights into the rational and moral structure of intention and action. That inspiration notwithstanding, however, she argued her own way (inspired also by Wittgenstein’s methods) through and out of the confusions and entanglements in thinking about action and its sources that had resulted from Cartesian, Humean, and degenerate scholastic moral psychology: independently minded, courageous, and creative.

Appendix: References to Aquinas in Anscombe’s Writings Aquinas DM =​ De Malo DV =​ De Veritate SLE =​ Sententiae libri Ethicorum (Commentary on the Nicomachean Ethics) ST =​ Summa Theologiae QQ =​ Quaestiones Quodlibetales

Anscombe 1939. ‘The Justice of the Present War Examined’ CP3, 73: ‘[The natural moral law] is the law of [man’s] nature. . . . Aquinas called it ‘the participation in the eternal law of the rational creature’’ (ST, 1a 2ae, 91, 2, ad 3). CP3, 73n2: ‘For a fuller discussion of the Natural Law, see Thomas Aquinas [ST] 1a 2ae, 91, 1a 2ae, 94, or any textbook of moral theology’.

67 

See Donald Davidson, ‘How Is Weakness of the Will Possible?’, in Essays on Actions and Events (Oxford: Oxford University Press, 2001), 21–​42.

Anscombe and Aquinas    465 CP3, 77: ‘These two points must therefore be maintained: to quote St Thomas Aquinas, ‘It is unlawful for a man to intend to kill anyone in order to defend himself, except for one with public authority; and he intending to kill a man for his own defence refers this to the general good’ (ST, 2a 2ae, 64, a7). CP3, 78: ‘[N]‌o action can be excused whose consequences involve a greater evil than the good of the action itself ’. CP3, 78n55: ‘Ctr [ST] 2a 2ae, 64, De Homicidio, on the example of killing on self-​ defence “the force used must be proportioned to the necessity.” ’ 1957. Intention p. 87: ‘[T]‌ he account given by Aquinas of the nature of practical knowledge holds: Practical knowledge is ‘the cause of what it understands’, unlike ‘speculative’ knowledge, which “is derived from the objects known” ’ (ST, 1a 2ae, q.3, a.5, obj. 1). 1958. ‘Pretending’ CP2, 88n4: ‘I owe notice of this aspect of anger to Plato, made intelligible by Aquinas. . . . “The two inclinations [the concupiscible and the irascible] do not reduce to a single principle. . . . For in general as desire burns higher anger sinks, and as anger burns higher desire sinks”’ (ST, Ia, q.81, art. 2). 1958. ‘Glanville Williams’ The Sanctity of Life’ GG1, 246: ‘Other errors [of Glanville Williams include] . . . that St Thomas Aquinas never discussed whether celibacy was a sin’ (cf. ST, 2a 2ae, q.152, art. 2). 1959? ‘The Immortality of the Soul’ GG2, 69: ‘I cannot at present accept the idea that spirituality is soulishness itself, or the character of being a rational soul, for the only well developed argument I know of to this position seems unconvincing to me. I mean the argument that thought and understanding are immaterial, since no act of a bodily organ is thinking or understanding’ (ST, 1a, q.75, a.2 and a.6). 1960? ‘On Being in Good Faith’ GG2, 104: ‘[A man] may shoot his father by mistake for a stag, but this was not his formal object . . . and so he is not guilty of parricide. The distinction between material and formal [objects of action] is so made by St Thomas’ (QQ, 3, q.12, a.2; see also ST, 1a 2ae, q.6, a.8). GG2, 106–​107: ‘St Thomas has an article in the Summa Theologiae: Whether a will that departs from an erroneous conscience is bad?’ (ST, 1a 2ae, q.19, a.5). 1965a. ‘Thought and Action in Aristotle’ CP1, 74: ‘[If by a practical syllogism] you mean a type of reasoning (St Thomas would call it ‘theoretical de practis’’ (ST, 1a, q.14, art. 16c). 1965b. ‘Necessity and Truth’ (review of early volumes of new translation of ST). CP1, 81–​85. 1968. ‘You Can Have Sex without Children’ CP3, 84–​85: ‘Following St Thomas, one might define as ‘sins against nature’ complete sexual acts which deviate from complete acts of ordinary intercourse’ (ST, 2a 2ae, q.154, a.11).

466   John Haldane CP3, 89–​90: ‘St Thomas’s doctrine is at this point faulty and confused’. CP3, 95–​96: ‘St Thomas replies: the order of nature is disturbed in a matter where what is done concerns the good of the species. . . . A parallel might be offered in the case of property’ (ST, 2a 2ae, 66, art 7). 1970. ‘Philosophers and Economists: Two Philosophers’ Objections to Usury’ GG2, 247–​249: ‘Aquinas expressed the objection [to seeking interest on a loan] in a more fundamental way [than Aristotle]’ (ST, 2a 2ae, q.78, a.1). 1974. ‘Times, Beginnings and Causes’ CP2, 152: ‘As Aquinas remarks in Summa Theologiae . . . The relation to a cause does not enter into the definition of the thing that is caused’ (ST, Ia. q.44, art. 1, ad 1). 1975. ‘Contraception and Chastity’ GG2, 175–​177: ‘Aquinas makes two contributions. . . . [H]‌e makes the remark that a man ought to pay the marriage debt if he can see his wife wants it without her having to ask him. And he ought to notice if she does want it’ (ST, Supp. Q.64, a2 corpus). ‘His second contribution was his definition of the ‘sin against nature’. . . . He defines this type of sin as a sexual act of such a kind as to intrinsically unfit for generation’ (ST, 2a 2ae, q.154, a.11). 1976. ‘On the Hatred of God’ GG2, 61: ‘[T]‌o human nature corrupted by sin the divine law repressing vices seems intolerable, and much more the punishments which are to be inflicted upon us for our offences. On account of such effects, as is said by the Angelic Doctor, God is hated by some’ (ST, 2a 2ae, q.34, a.1). 1979a. ‘Chisholm on Action’ GG1, 80: ‘[Chisholm’s account of ‘agent causality’] . . . is reminiscent of Aquinas’s account of creation as a relation which is real in the creatures but not real in God’ (ST, 1a, q.45, a.3). 1979b. ‘On Humanae Vitae’ ‘[Humanae Vitae] contains no nostalgia for the past, no lamenting in favour of times when women were thought of as obviously not equal citizens, not suitable witnesses in a law court, for example (as St Thomas remarks somewhere) let alone judges etc’. (ST, 2a, 2ae, q.70, a.3). 1980? ‘Two Moral Theologians’ GG2, 166–​167: ‘[Vermeersch] is admirable in his criticism of one who thinks to find the ‘principle of double effect’ in St Thomas’ article on this subject’ (ST, 2a 2ae, q.64, a.7). ‘St Thomas thinks that even in self-​defence one may not kill on purpose. . . . All the same, St Thomas’ rigorous doctrine has its difficulties’. 1982a. ‘Murder and the Morality of Euthanasia’ GG1, 273: ‘And, since negligence too is voluntary, so also may be the harm produced by an unintended omission, if it constitutes negligence. . . . The principle by which such harm is voluntary was stated succinctly by St Thomas: it was both possible and necessary for will (or: the agent) to act, and it (he) did not’ (ST, 1a 2ae, q.6, a.3).

Anscombe and Aquinas    467 1982. ‘Action, Intention and “Double Effect” ’ GG1, 225–​226: ‘I will end by protesting at the ascription of the Doctrine of Double Effect to Aquinas: the phrase “duplex effectus” occurs in his discussion of killing in self-​defence’ (ST, 2a 2ae, q.64, a.7). ‘If we want to know St Thomas’ general opinion on responsibility for evil consequences of actions, this is not the place to look but rather at what he says about the relation of an eventus sequens to the goodness or badness of an action’ (ST, 1a 2ae, q.20, a.5). 1985a. ‘Anselm’s Proof ’ GG3, 40: ST, 1a, q.2, a.1 ad 2. 1985b ‘Truth: Anselm or Thomas?’ GG3, 54ff.: DV, q.1, a.4, a.3, a.5 ad 2; obj. 6, obj. 7; ST, 1a, q.16, a.7, ad 4; 1a, q.16, a.8, ad.3, ad.4. 1986. ‘Knowledge and Reverence for Human Life’ GG1, 60: ‘As I was using the word [‘connatural’], it is a character of knowledge. . . . The word ‘connatural’ of course has to do with ‘nature’. So far as I have been able to notice in St Thomas, digging around with a lexicon, its principal use in him is to talk of what is readily known by beings of a certain nature’ (ST, 2a 2ae, q.45, a.2). 1989. ‘Sin’ GG2, 127–​141: ‘Voluntariness and Sins of Omission’ and ‘God’s Causality’ (DM, q.3, a.7, a.9, a.10; SLE, 3, 1, 387; ST, 1a 2ae, q.6, a.3, a.4; q.49, a.2; q.76, a.2, a.3). 1990. ‘The Early Embryo’ GG2, 216–​218: ‘Aquinas thought that the male contribution . . . causes the female-​supplied material to start living with active vegetative life’ (ST, 1a, q.118, a.1, esp. ad 4). ‘In just one case he believed that the conceptus was a complete tiny human being: this was Jesus’ (ST, 3a, q.33, a.2 ad 2). ‘What is most interesting is his thesis that in the process of production, there is a stage of living with animal life without being e.g. a man or a horse’ (ST, 1a, q.119, a.2). ‘I have referred to him partly because I think he is intrinsically worth referring to in this context; and also having read him precisely because I knew he was what is ineptly called a ‘mediate animationist’ and I was myself inclined to be what is meant by that’. GG2, 222–​223: ‘The opinion of St Thomas, that original sin as originated is in the human conceptus by inheritance from Adam . . . is not universally accepted by modern theologians; but it appears the most reasonable’ (ST, 1a, q.83, a.1) 1992. ‘Embryos and Final Causes’ GG1, 54: ‘St Thomas Aquinas thought that the normal product of conception was [first] alive with vegetative life . . . then was alive with animal life. . . . But also that the creation of the human soul was special and not a matter of development like the little animal’s becoming a horse’ (ST, 1a, q.118, a.2). 1992. ‘On Wisdom’ GG2, 260: ‘[H]‌ere was a case of the ‘greater’, as St Thomas calls them, among the Hebrews having some knowledge of the Holy Trinity in the times before the Messiah’ (Summa Contra Gentiles, IV, 2).

468   John Haldane 1993. ‘Practical Truth’ GG1, 156: ‘We—​and, according to St Thomas, God too—​can have theoretical knowledge of what our present practical knowledge is of ’ (ST, 1a, q.14, a.16). 196?. ‘Thought and Existent Objects’ GG4, 198–​199: ‘My interest is in how thought is of its objects when the objects of thought are actual existents. I have very vaguely seen—​though I would say I have not understood—​the picture presented by Aristotle and adopted, perhaps with some elaboration or addition, by Aquinas’ (ST, 1a, q.78, a.3; q.79, a.3).

Chapter 21

Et hics and Act i on T h e ory An Unhappy Divorce Constantine Sandis

21.1  Moral Philosophy until Anscombe In her seminal 1958 paper ‘Modern Moral Philosophy’ (MMP), Elizabeth Anscombe complains, inter alia, about the lack of ‘an adequate philosophy of psychology’ within moral philosophy, at least as practiced from Sidgwick onwards. While her complaint (which I review in detail in section 21.2) was not without cause, it remains a striking fact that there was considerably more moral psychology within moral theory before 1957 than there has been ever since. Moreover, much of it was conducted by a number of Anscombe’s explicit or implicit targets, particularly G. E. Moore (1903), W. D. Ross (1930), H. W. B. Joseph (1931/​1933), A. C. Ewing (1938), John Macmurray (1938), and H. A. Prichard (1945).1 Anscombe presumably didn’t find any of this work ‘adequate’, and I shall try to show why this was so. Prima facie, however, her moral psychology and philosophy of action has much in common with many of these targets, not to mention Hegel. In order to allow for sufficient depth of comparison, I shall focus my discussion on Ross, who is one of the explicit targets of MMP, but will add comparisons to other philosophers along the way. Ross defends the proto-​Anscombean view that ‘any act may be correctly described in an indefinite, and in principle infinite, number of ways’ and that what I do could, for

1 

It has also been present, not merely in Aristotle (Ethica Nicomachea) and Aquinas (Summa Theologica [ST], I–​II, 1–​21), but across the entire history of modern moral philosophy from Suarez onwards (see essays in Sandis 2019b). The ‘Morality’ heading of Hegel’s Philosophy of Right (1821:§§105–​ 141) is of particular relevance to Anscombe’s interests. For comparisons between Hegel and Anscombe, see Taylor (1979, 1983), Quante (1993), and Sandis (2010). For Aquinas and Anscombe, see Jensen (2010).

470   Constantine Sandis instance, be truthfully described as ‘fulfilling my promise’, ‘putting the book into our friend’s possession’, and ‘the packing and posting of a book’: [A]‌ny of the acts we do has countless effects. . . . Every act therefore, viewed in some aspects, will be prima facie right, and viewed in others prima facie wrong. . . . [A]ny act may be correctly described in an indefinite, and in principle infinite, number of ways. An act is the production of a change in the state of affairs. . . . I may have promised, for instance, to return a book to a friend . . . to send it by a messenger or to hand it to his servant or to send it by post. . . . [I]n each of these cases what I do directly is worthless in itself. . . . [T]his is not what we should describe, strictly, as our duty; our duty is to fulfil our promise, i.e. to put the book into our friend’s possession. . . . What I do is as truly describable in this way as by saying that it is the packing and posting of a book. . . . And if we ask ourselves whether it is right qua the packing or posting of a book, or qua the securing of my friend’s getting what I have promised to return to him, it is clear that it is in the second capacity that it is right . . . by its own nature and not because of its consequences. (Ross 1930: 41–​44)

There is much here that anticipates Anscombe. She would later even apply the same preposition (‘qua’) to distinguish her view that actions may be intentional under a description from the nonsensical claim that actions-​under-​a-​description are intentional (UD).2 So why did Anscombe dismiss Ross’s work in such strong terms? His ‘objectivism’ may have lapsed into a form of ‘consequentialism’ so anathematic to Christian morality (see section 21.2) that she didn’t want to debate its details under the guise of doing ‘moral philosophy’,3 but this is not in itself a reason to dismiss an entire method of doing moral philosophy that is uncannily similar to her own. A clue to our answer may be found in an earlier passage of The Right and the Good. Ross writes: [G]‌reat confusion . . . has been introduced into ethics by the phrase ‘a right action’ being used sometimes of the initiation of a certain change in the state of affairs irrespective of motive, and at other times of such initiation from some particular motive, such as sense of duty or benevolence. I would further suggest that additional clearness would be gained if we used ‘act’ of the thing done, the initiation of change, and ‘action’ of the doing of it, the initiating of change, from a certain motive. We should then talk of a right act but not of a right action, of a morally good action but not of a morally good act. And it may be added that the doing of a wrong act may be a morally good action; for ‘right’ and ‘wrong’ refer entirely to the thing done, ‘morally good’ and ‘morally bad’ entirely to the motive from which it is done. (Ross 1930: 6–​7; cf. Sidgwick 1874: book III, ch. 12; Macmurray 1938)

From all this he concludes, in a deliberately provocative and paradoxical manner: [N]‌othing that ought to be done is ever morally good. . . . [T]he only acts that are morally good are those that proceed from a good motive. . . . [A]ction from a good 2  For the relation between Ross’s use of the conjunction and his ethic of prima facie duties, see Ross (1930: 29). 3  Cf. Wiseman (2016b: 10–​11).

Ethics and Action Theory    471 motive is never morally obligatory. . . . [W]hat is morally good is never right. . . . That action from a good motive is never obligatory follows from the Kantian principle . . . that ‘I ought’ implies ‘I can’. . . . [H]owever carelessly I pack or dispatch the book, if it comes to hand I have done my duty, and however carefully I acted, if the book does not come to hand I have not done my duty. Of course I should deserve more praise in the second case than in the first. (Ross 1930: 132)

Anscombe has little time for this sort of distinction between what is done and the doing of it from a certain motive. She is consequently disinclined to relate the former to the right and the latter to the good, a disposition strengthened by her independent suspicion of the very distinction between good and right action (see section 21.2). It is worth recalling, at this juncture, that Anscombe’s objection was not that moral philosophers lacked a philosophy of psychology or action per se, only that they were in desperate need of one that was ‘sound’, or at least ‘adequate’. The problem, at least from Anscombe’s point of view, is that Ross puts his multiple descriptions of action to the disingenuous use of swallowing up features that would normally be regarded as their consequences, thereby presenting them as intrinsic properties that are ‘utterly independent of consequences’: That which is right is right not because it is an act, one thing, which will produce another thing, an increase in general welfare, but because it is itself the producing of an increase in the general welfare. . . . [W]‌e have to recognise the intrinsic rightness of a certain type of act, not depending on its consequences but on its own nature. (Ross 1930: 47, emphasis in original)

Anscombe may have additionally found Ross’s argument to the conclusion that an action that is good can never be right (and vice versa) to be ‘unsound’ because it fails to capture the correct relation between motive, intention, action, and duty. Whatever the explanation, she seems to have thrown the baby out with the bathwater. For there is a sound and morally important difference to be made between the things we do and our doings of them, especially in relation to questions concerning intention, foresight, consequences, and intrinsic wrongness.4 Ross’s point about the rightness of an action being divorceable from the goodness or badness of one’s performing is a sensible and important one, sharing affinities with Mill’s (1863: 18–​20) stance that ‘a right action does not necessarily indicate a virtuous character’ and that ‘actions which are blameable often proceed from qualities entitled to praise’.5 Yet Anscombe’s view that the things we do are happenings or events (I: §29; TKEA: 4; PJ: 10–​11; UD: 208–​210) with morally pertinent descriptions seems to leave no space for it. This is because it rules out the possibility that two people can do the very

4  5 

See Hornsby (1993). For contrast, see Kant (1788: 5, 147–​148).

472   Constantine Sandis same thing, even though only one of them is acting from a good motive.6 In this she goes against her teacher Wittgenstein, who cites this very possibility as an explanation for why self-​understanding can be so difficult to achieve at times: It is hard to understand yourself properly since something that you might be doing out of generosity & goodness is the same as you may be doing out of cowardice or indifference. To be sure, one may act in such & such a way from true love, but also from deceitfulness & from a cold heart too. (Wittgenstein 1977/​1998: 54 [MS 131 38: 14.8.1946], my emphasis; cf. 1953: §§ 253–​254)

While Anscombe can allow that two people can do the same thing, she understands this as there being at least one description that is applicable to both deeds (e.g., ‘drinking gin’) though many may not be (e.g., acting cowardly or bravely). But while some descriptions of the things we do have our intention built into them (e.g., to lie is, at least typically, to say something untrue with the intention of deceiving), many do not. Moreover, motives—​as opposed to intentions—​lend themselves more naturally to descriptions of our doings and not our deeds.7 To be sure, in speaking loosely we may say that someone did a cowardly thing, but all this means is that they acted in a cowardly fashion; the thing they did (drink gin) is itself neither cowardly nor brave. A year after Ross’s book was published, Joseph wrote the following by way of commentary:8 The same act, an objector might say, may surely be done from very different motives; and therefore the act must be something, irrespective of the motive. But are they really the same act? Different acts, having different motives, may work themselves into the same movements of bodies; but these are not the acts. A man who was fond of oysters might eat a plateful put before him for the sake of their flavour; a man who loathed them might do so to avoid hurting his host’s feelings; a man who loathed or was indifferent to them might do so to prevent his neighbour, whom he knew to be fond of them and he disliked, from having two portions. I think these are three different acts, one morally good or else kindly, one morally bad or spiteful, one indifferent. They are not three instances of one act, viz. eating a plateful of oysters. . . . [A]‌n act proper is not analysable into behaviour and motive; it is indivisible. (Joseph 1931/​ 1933: 45–​46)9

6  Strictly speaking, one could defend an identity theory between action-​event and thing done while still allowing for the subtler distinction between one’s doing X and the event of one’s doing X (see Sandis 2012: 33), but the stance seems eccentric. 7  It is perhaps no accident that Ross’s discussion focuses much more on motives and Anscombe’s on intention. 8  Earlier in the chapter, Joseph (1931/​1933: 37) uses the locution ‘it is said’ before discussing views now associated with Ross’s The Right and the Good, which appeared in print only after Joseph’s proofs had been corrected (Joseph 1931/​1933: v). The revised 1933 edition adds a footnote to Ross’s book on p. 19. 9  I owe this reference to Andreas Lind.

Ethics and Action Theory    473 Joseph’s view anticipates Anscombe’s insofar as they both target Ross by maintaining that ‘no act exists except in the doing of it, and in the doing of it there is a motive’.10 The Joseph-​Anscombe thesis offers a persuasive account of our doings but ignores the distinction (implicit in Wittgenstein and explicit in Ross) between the deeds we do and our doings of them from some particular motive. This is done at the cost of allowing that one may do the right thing from the wrong motive. One might object that ‘A did the same thing as B’ simply amounts to there being some description of what A did which is also true of what B did (e.g., ‘give money to X’ or ‘show off to his friends’). But it would be a category mistake to think that such descriptions apply to things done as opposed to our doings of them. If A murdered X then we can indeed describe this very thing that A did by saying (with some loss in specificity) that A killed X, for the latter is a sub-​set of the former. But if A gave money to X, we cannot truthfully describe the thing done (give money) as showing off, since one can do this very same thing without even intending to show off in the process. A particular instance of giving money may be truthfully describable as showing off, but the things we do, in the sense in which two people can do the same thing, are not instances of anything. In the case of the person whose donating a large sum of money is a case of showing off, then, there is not one single thing done that is good under one description and bad under another but, rather, one event of someone acting badly in doing two distinct things (one right, the other wrong). Hursthouse writes: [A]‌ct honestly, charitably, generously; do not act dishonestly, etc. . . . ‘[T]he adverbs connote not only doing what the virtuous agent would do, but also doing it “in the way” she would do it, which includes “for the same sort(s) of reason(s).” ’ . . . What is misleading about this phrase is that it obscures the fact that, in one way, the agent is not ‘doing the right thing’. What she is doing is, say, trying to impress the onlookers, or hurting someone’s feelings, or avoiding punishment. (Hursthouse 1999: 29n7, 125)

But while one’s act of donating to charity may also be correctly described as one’s trying to impress the onlookers, this doesn’t give us a reason to deny that in so acting a person may do (at least) two things: donate to charity and impress the onlookers, one of which is right and the other wrong. Anscombe can, of course, allow that one can donate to charity with the (bad) intention of impressing the onlookers. On her view, however, this provides a true description of what was done, thereby leaving no space for the view that one can do the right thing with a bad intention. This forces us to say that what was done was right under one description and wrong under another. But this is to confuse acting rightly with doing what happens to be the right thing to do, despite one’s nefarious motives.11 In Ross’s (1930: 53) terminology, it is to confuse such things as ‘doing what is just’ with such things acting ‘in the spirit of justice’.

10  11 

Joseph (1931/​1933: 38). See Sandis (2017, 2021).

474   Constantine Sandis Anscombe’s underlying account of action as a happening contrasts with that of Prichard, according to whom to do something is to bring about a change: It is, no doubt, not easy to say what we mean by ‘an action’ or by ‘doing something.’ Yet we have in the end to allow that we mean by it originating, causing, or bringing about the existence of something viz. some new state of an existing thing or substance, or, more shortly, causing a change of state of some existing things. . . . [B]‌y ‘moving our hand’ we mean causing a change of place in our hand; by ‘posting a letter’ we mean bringing about that a letter is in a pillar-​box. (Prichard 1932: 84–​85)

This view anticipates those of G. H. von Wright (1963) and, more recently, Maria Alvarez and John Hyman (1998). By the end of World War II, however, Prichard (1945: 272–​274) had come to embrace a volitionist account of action as a form of mental activity.12 Contra Macmurray and Ross, he now claimed that the term ‘action’ was not ambiguous at all: it referred to our mental ‘doings’ and not to their effected changes, which constitute our ‘deeds’ (Prichard 1945: 275; cf. von Wright 1963: 37ff.). Anscombe would have undoubtedly rejected Prichard’s invalid inference from the thought that we might conceivably fail to achieve anything we set out to do, to the conclusion that all we ever have a duty to do is to set ourselves (viz. will) to bring something about. Indeed, no adequate philosophy of psychology could ever allow for such an inference. But if, in uttering ‘I do what happens’, Anscombe had been running a million miles from Prichard’s volitionism, then she ended up too far in the other direction.

21.2  ‘Modern Moral Philosophy’ MMP remains as divisive today as it was when it was first published sixty years ago. Some hail it to be of huge philosophical and historical importance, not least by effectively giving birth to neo-​Aristotelian Virtue Ethics as exemplified by Philippa Foot, Rosalind Hursthouse, Alasdair MacIntyre, and John McDowell (Richter 2011; Solomon 2008: 110–​111). Others present it as a dated or otherwise irritating text containing baffling and unsubstantiated claims, the deciphering of which is not worth the candle (e.g., Blackburn 2005). One recurring complaint has been that Anscombe is unfair in dismissing the ideas of dead white men with brief statements that contain more disrespect than they do argument. Bishop Butler is ‘ignorant’ (MMP 27), Immanuel Kant ‘useless’ and ‘absurd’ (MMP 27), David Hume ‘sophistical’ (MMP 28), and J. S. Mill ‘stupid’ (MMP 33). These are the philosophers she likes. The rest of them are something much worse: ‘consequentialists.’

12 

The general shape of Prichard’s account is retained in the early work of Jennifer Hornsby (1980: 46–​ 48, 60–​63), who replaces willing with trying (retracted in Hornsby 2010).

Ethics and Action Theory    475 Anscombe uses the term in MMP as a pejorative,13 but it was quickly reclaimed as a badge of honour by all its major proponents. It is fashionable nowadays to remark that Anscombe meant something rather different by ‘consequentialism’ than we do today.14 Yet her own characterisation of it as the view that the ‘right’ action is that which produces the best possible consequences (MMP 33, quoted later) is one endorsed by most contemporary consequentialists.15 The exegetical difficulty arises because Anscombe protects the utilitarian Mill from this particular charge yet includes ‘objectivists’ such as her near-​contemporary W. D. Ross, best known for defending the view that actions can be wrong in virtue of their intrinsic value, regardless of their consequences: There is a startling change that seems to have taken place between Mill and Moore. Mill assumes . . . that there is no question of calculating the particular consequences of an action such as murder or theft. . . . In Moore and in subsequent academic moralists of England we find it taken to be pretty obvious that ‘the right action’ is the action which produces the best possible consequences (reckoning among consequences the intrinsic values ascribed to certain kinds of act by some ‘Objectivists.’ (MMP 33)16

This double move is key to understanding the last of three related theses that MMP famously sets out to defend. These have proven to be as hard to interpret as they are easy to state: (T1) It is not profitable for us at present [1958] to do moral philosophy; that should be laid aside at any rate until we have an adequate philosophy of psychology, in which we are conspicuously lacking. (T2) The concepts of obligation, and duty—​moral obligation and moral duty, that is to say—​and of what is morally right and wrong, and of the moral sense of ‘ought’, ought to be jettisoned if this is psychologically possible; because they are survivals, or derivatives from survivals, from an earlier conception of ethics which no longer survives, and are only harmful without it. (T3) The differences between the well-​known English writers on moral philosophy from Sidgwick to the present day are of little importance. (MMP 26)17

13  For an earlier coinage by her, see Berkman in this volume. Cf. Browne (1945: 508), who understands consequentialism as the view that ‘the right act produces the maximum possible good’, complaining that this is empty if unaccompanied by a specification of ‘what natural kinds of thing it is right to bring into being’. I owe this reference to Aaron Garrett. 14  See, for example, Diamond (1997), Teichmann (2008: 86), and Wiseman (2017: 18). 15  Cf. M. Geach (2008: xvii). 16  Entire papers could be written about the degree to which this passage offers plausible interpretations of either Mill or Moore. There is room for disagreement, for example, on whether murder and theft could ever fall under Mill’s (1863: 25) ‘knotty points’, the answer depending on whether he conceives of them as being unjust by definition (discussed further later in this chapter). 17  Chappell in this volume argues that Anscombe’s main complaint is T3 and that T1 and T2 are ‘little more than auxiliary theses’.

476   Constantine Sandis What makes T3 true for Anscombe is that the philosophers in question are all ‘consequentialists.’ Henry Sidgwick’s predecessor Mill is off the hook from this charge for two distinct reasons. First, he was careful (at least according to Anscombe) to distinguish intended from merely foreseen consequences of an action. To understand how this helps to avoid ‘consequentialism’, one needs the ‘adequate philosophy of psychology’ required to reveal the role played by intention in determining the nature of any given action.18 Second, Mill (1863: 27) explicitly states that utilitarianism can never conflict with justice, going out of his way to explain why his philosophy is compatible with Christian morality in particular. He thus allows that utilitarianism leads us to re-​evaluate whether acts of stealing or kidnapping must always be unjust, while rejecting the consequentialist conviction that an unjust act could ever be permissible, let alone obligatory: Have mankind been under a delusion in thinking that justice is a more sacred thing than policy, and that the latter ought only to be listened to after the former has been satisfied? While I dispute the pretensions of any theory which sets up an imaginary standard of justice not grounded on utility, I account the justice which is grounded on utility to be the chief part, and incomparably the most sacred and binding part, of all morality. Justice is a name for certain classes of moral rules which concern the essentials of human well-​being more nearly, and are therefore of more absolute obligation, than any other rules for the guidance of life. . . . [T]‌o save a life, it may not only be allowable, but a duty, to steal or take by force the necessary food or medicine, or to kidnap and compel to officiate the only qualified medical practitioner. In such cases, as we do not call anything justice which is not a virtue, we usually say, not that justice must give way to some other moral principle, but that what is just in ordinary cases is, by reason of that other principle, not just in the particular case. By this useful accommodation of language, the character of indefeasibility attributed to justice is kept up, and we are saved from the necessity of maintaining that there can be laudable injustice. (Mill 1863: 57–​58, 62)

Is Mill paying mere lip service to justice here, or does he take the thickness of the concept to entail that no plausible moral theory can be at odds with it? In not counting him as a ‘consequentialist’, Anscombe charitably opts for the latter understanding, viz. that he would give no weight to unjust actions, no matter what their effects: Mill assumes, as we saw [27], that there is no question of calculating particular consequences of an action such as murder or theft. (MMP 33)

Yet Mill’s view ultimately seems to be that any action prescribed by utilitarianism must be just by definition. If so, Anscombe would be wrong in her assessment that ‘it did not occur to him that acts of murder and theft could be otherwise described’ (MMP 27). 18  This is partly because Anscombe primarily conceives of ‘consequentialism’ as a view regarding the sphere of personal responsibility (see Frey 2019: 10–​12). For more on Mill’s anti-​consequentialist utilitarianism, see Vogler (2001).

Ethics and Action Theory    477 Indeed, we may plausibly attribute to Mill the view that some thefts are just precisely because they can be described as the taking of necessary food or medicine. On this point, the difference between Mill and someone like Aquinas is more semantic than it is moral. Unlike Mill, Aquinas maintains that all theft is unjust, but he also asserts that in cases of dire emergency it is not theft to take from another’s possessions: When a person is in extreme need of material things, and there is no way of emerging from his extremity but by taking what belongs to another, the surplus which another possesses becomes common property, and the taker is not guilty of theft. (ST, IIaIIae.66.7)19

Anscombe’s evaluation of Mill contrasts starkly with that of Ross, according to whom any ‘intrinsic’ property of action, including that of being unjust, may in principle be outweighed by sufficiently positive consequences: Oxford Objectivists of course distinguish between ‘consequences’ and ‘intrinsic values’ and so produce a misleading appearance of not being consequentialists. But they do not hold—​and Ross explicitly denies—​that the gravity of, e.g., procuring the condemnation of the innocent is such that it cannot be outweighed by, e.g., national interest. Hence their distinction is of no importance. (MMP 33n4)

So understood, Ross allows that there are times when, all things considered, we are not only permitted but morally obliged to commit acts of murder, adultery, or whatever. Anscombe rejects his thesis that there is no value so sacred that it cannot in principle be trumped by the greater good as being ‘consequentialist’, despite the fact that Ross’s (1930: 21ff.) ethic of prima facie duties explicitly allows that goods of beneficence and non-​maleficence are often outweighed by concerns of fidelity, gratitude, justice. As Christopher Coope has argued, she would have also rejected some of Hursthouse’s views on the same grounds.20 Despite her own definition, then, Anscombe cannot ultimately view ‘consequentialism’ as the simple equation of ‘rightness’ with producing the best consequences. As Mary Geach (2008: xvii) puts it, one might hold the ‘consequentialist’ view that ‘there is no act so bad [that] it might on occasion be justified by its consequences’ while rejecting the consequentialist view that ‘the right action is always that which produces the best consequences.’ In his 1956 article ‘Good and Evil’, Peter Geach denounces Ross’s moral outlook for very similar reasons. I quote at some length: I am deliberately ignoring the supposed distinction between the Right and the Good. . . . Aquinas . . . finds it sufficient to talk of good and bad human acts. When

19 

I owe this reference to Sophie Grace Chappell. Coope (2006: 46ff.). While I agree with Coope on this point, I don’t share the conception of justice he uses to frame it. 20 

478   Constantine Sandis Ross would say that there is a morally good action but not a right act, Aquinas would say that a good human intention had issued in what was, in fact, a bad action; and when Ross would say that there was a right act but not a morally good action, Aquinas would say that there was a bad human act performed in circumstances in which a similar act with a different intention would have been a good one (e.g. giving money to a beggar for the praise of men rather than for the relief of his misery). . . . [P]‌eople who think that doing right is something other than doing good will regard virtuous behaviour as consisting, not just in doing good and eschewing evil, but in doing, on every occasion, the right act for the occasion. This speciously strict doctrine leads in fact to quite laxist consequences. A man . . . if he knows that adultery is an evil act, will decide that (as Aristotle says) there can be no deliberating when or how or with whom to commit adultery. But a man who believes in discerning, on each occasion, the right act for the occasion, may well decide that on this occasion, all things considered, adultery is the right action. Sir David Ross explicitly tells us that on occasion the right act may be the judicial punishment of an innocent man ‘that the whole nation perish not’ for in this case ‘the prima facie duty of consulting the general interest has proved more obligatory than the perfectly distinct prima facie duty of respecting the rights of those who have respected the rights of others.’ (P. Geach 1956: 41–​42)21

Geach’s outing of Ross as a crypto-​consequentialist is, of course, directly linked to T3.22 His further rejection of Ross’s distinction between goodness and rightness is closely tied to Anscombe’s other two theses. In particular, his contention that we should make do without the concept of a ‘right action’ at all helps to explain why Anscombe keeps ‘the right action’ within scare quotes. It also sheds light on her middle thesis (T2) that we must try to jettison the concept of a distinctly moral obligation. Terms like ‘moral obligation’ and ‘morally right action’ ought to be jettisoned because they are survivals of an earlier, quasi-​legal conception of morality and make no sense outside of the related contexts and practices that originally gave them meaning. This is not a rejection of morality, nor even of a moral ought, but only of the distinctively secular and mesmeric ‘moral ought’ that has been detached from the aforementioned conceptions.23 James Doyle (2018) has recently offered a more radical reading of T2. According to Doyle, Anscombe’s claim is that the term ‘moral’ as ordinarily used is literally

21  To this he sneeringly adds, ‘We must charitably hope that for him the words of Caiaphas that he quotes just had the vaguely hallowed associations of a Bible text, and that he did not remember whose judicial murder was being counselled.’ 22  Geach and Anscombe would presumably be equally hostile to the moral particularist claim that there are no principles concerning right action (e.g., Dancy 2004). But particularism at the level of things done may be combined with generalism at the level of character traits (Sandis 2020, 2021; cf. Swanton 2015). Were Anscombe open to a conceptual distinction between doings and things done (see section 21.3) she could more easily allow for such a view, whose origins lie with Aristotle’s thought that the mean is grasped through perception and not by reasoning (Ethica Nicomachea 1109b; but see Price 2005). 23  Cf. Solomon (2008: 114) and Gremaschi (2017), the latter finding Anscombe’s concerns much more parochial than the former. White (1968: 77ff.) convincingly distinguishes between ‘obliged to do’ and ‘ought to do’.

Ethics and Action Theory    479 meaningless, standing for an empty pseudo-​concept that provokes a certain feeling in us but has no content whatsoever.24 Divine commands, on this understanding, neither describe nor create distinctly moral obligations, but only religious ones. Whatever the merits of the view itself (Doyle finds it more plausible than I do), we should be wary of it as an interpretation of what is going on in MMP, not least because there are plenty of later writings in which Anscombe endorses law conceptions of morality with no sign of having had the slightest change of mind. In ‘Authority in Morals’ (AM), Anscombe speaks of ‘moral conclusions’, ‘revelation of moral belief ’, of a ‘moral truth’ concerning ‘what kinds of thing ought to be done and ought not to be done’ as well as of ‘the moral law’ as a ‘range’ and of taking one’s morality from someone else, concluding that ‘the content of moral law, i.e., the actions which are good and just, is not essentially a matter of revelation.’ Similarly, in ‘The Moral Environment of the Child’ (METC 231), Anscombe states that ‘Catholic Christianity teaches a strict moral code’ and speaks, without scepticism, of ‘truth in the moral code’ and ‘our morality.’ This Christian morality is contrasted with ‘a morality which consisted solely of absolute prohibitions on fairly definitely described actions’ (presumably Kant’s). The idea of two contrasting moralities forms the core of her short essay ‘Morality’ (1982) (M 113), in which she explicitly rejects the thought that there is such a thing as morality, not because she is a sceptic about moral concepts—​she writes that ‘human beings have always had morality’ and talks of ‘that part of morality which is associated with duties to God’—​but because she finds Christian morality so distinct from the consequentialist one that they amount to two entirely different moralities: one that prohibits murder, and one that not only permits—​but can even demand—​it. As Mary Geach writes in a 2005 letter to the Times Literary Supplement,25 ‘Anscombe herself, of course, had no intention of jettisoning the concepts of moral obligation and duty, which are needed to frame her other principal claim, which is that certain things are forbidden, whatever the consequences.’ While Christian morality does indeed require us to embody certain virtues, the question of which virtues we ought to have may also be addressed from the point of view of what is good for us, qua human. In pointing this out, Anscombe is in no way abandoning deontic terminology in favour of the aretaic (see Coope 2006: 22). The deontic and the aretaic are simply two different frameworks for talking about the very same goodness. By returning to Aristotle’s talk of human excellence and virtue, MMP thus seeks to find a common language through

24 

Cf. Richter (2019). Doyle (2019) has since revised his view, but still maintains that Anscombe thought—​and was right to think—​that nothing could count as understanding the word ‘moral’. This goes against the more natural reading of Anscombe’s quasi-​Nietzschean genealogy as having been at least partly motivated by Wittgenstein’s thought that the meaning of any given term or expression is dependent upon the practices that give it life (see Sandis 2019a; cf. Frey 2018). Wittgenstein’s influence on Anscombe’s understanding of normativity is further made evident in her discussions of forcing and stopping modals (PJ, RRP, SAS). 25  M. Geach (2005a; see also 2005b).

480   Constantine Sandis which religious and secular thinkers alike (including Anscombe and her friend Philippa Foot) may converse about morality, and perhaps even reach agreement.26

21.3  Contemporary Normative Ethics Anscombe’s first thesis in MMP is that ‘[i]‌t is not profitable for us at present [1958] to do moral philosophy; that should be laid aside at any rate until we have an adequate philosophy of psychology, in which we are conspicuously lacking’ (MMP 26). By ‘philosophy of psychology’ she is not referring to the philosophy of cognitive science that currently goes under that name today,27 nor to the philosophy of mind that used to share it,28 but to an investigation into the concepts of action, character, intention, motive, desire, pleasure, and the relations between them: In present-​day philosophy an explanation is required how an unjust man is a bad man, or an unjust action a bad one. . . . [I]‌t cannot even be begun until we are equipped with a sound philosophy of psychology. . . . This part of the subject matter of ethics is, however, completely closed to us until we29 have an account of what type of characteristic a virtue is—​a problem, not of ethics, but of conceptual analysis—​and how it relates to the action in which it is instanced. . . . For this we certainly need an account at least of what a human action is at all, and how its description as ‘doing such-​and-​such’ is affected by its motive and by the intention or intentions in it; and for this an account of such concepts is required. (MMP 29)30

The blueprint for this philosophy of action had already been laid down by her the year before in her masterpiece, Intention. This book had already made good on MMP’s request for ‘an account at least of what a human action is at all, and how its description as ‘doing such-​and-​such’ is affected by its motive and by the intention or intentions in it’ (MMP 29).31 Whether Anscombe herself thought she had already provided an adequate philosophy of psychology, or merely a sketch of one, is a moot point.

26  Anscombe became increasingly pessimistic about the latter happening on any kind of wide scale (see essays in GG1). Jennifer Frey informs me that Aquinas was far more sanguine on this front. 27  See, for example, Botterill and Caruthers (1999), Bermúdez (2005), Thagard (2007), and Weiskopf and Adams (2015). 28  For example, Block (1980). 29  See Sandis (2019c). 30  Anscombe’s contention here remains unaffected by John Rawls’s (1974–​1975) famous argument for the independence of moral theory from the sorts of issues he associates with epistemology and the philosophy of mind and language. But even if Rawls’s argument could be extended to show that many issues in moral theory are independent from philosophical psychology, we should not expect a theory of right action to remain silent on the relation of action to motive, intention, and consequence. 31  For an excellent account, see Wiseman (2016a: ch. 2; 2016b: §3).

Ethics and Action Theory    481 The term ‘philosophy of psychology’ has long since been hijacked by philosophers of cognitive science to describe what they do, leaving ‘philosophical psychology’ as the term of choice for those still interested in asking philosophical questions about human psychology, and ‘moral psychology’ and ‘philosophy of action’ for the areas covering the kinds of issues that Anscombe is referring to. Indeed, both these fields owe part of their existence as we know it to Anscombe. In the aftermath of Intention and MMP, the ‘philosophy of action’ developed into a subject in its own right, albeit more closely associated with the philosophy of mind than with ethics.32 Such branching-​off comes at a heavy price, for ethics without philosophy of action is blind, and philosophy of action without ethics empty. The philosophy of mind and action during the past sixty years has thus developed alongside that of normative ethics, with very little interaction between them. This has enabled philosophers to defend theories of ‘right action’ while remaining conspicuously silent about what actions are, or how to best conceive of their relation to intentions (on the one hand) and consequences (on the other). As a result of all this, contemporary moral philosophy has been neatly divided into the following four branches, which had yet to properly separate in 1958:

a) Meta-​ethics b) Normative ethics (‘moral theory’) c) Practical or ‘applied’ ethics d) Moral psychology To be sure, theorists debate the extent to which views within (a–​d) are interrelated, but they are generally content to teach them as separate ‘modules’ and have been known to profess expertise in any one of them while claiming near-​ignorance on the remaining three. People do, of course, defend philosophical positions according to which one cannot do (c) without (b) and/​or (b) without (a), but even these are parasitic on the divisions in question. Most important, (d) is typically reserved for questions relating to agency, motivation, moral responsibility, akrasia, and so on that are thought to be largely independent of (a–​c). In complete opposition to this, MMP’s first thesis effectively tells us that one cannot do (a–​c) at all without first doing (d). While Anscombe certainly cared for ‘practical’ issues relating to everyday life as well as to medical, military, and legal policies, she did not see these issues as remotely separable from either moral philosophy or the philosophy of psychology.33 I shan’t concern myself much with (a) and (c), save to recall that the contemporary obsession of engaging in (c) by comparing intuitions about increasingly absurd trolley cases is an unintended consequence of an argument made by Foot in which she defends, 32  For a brief period, philosophy of mind was also called ‘philosophy of psychology’ (see, for example, Block 1980). 33  This is evident across all her work in ethics, but particularly so in TD (for which, see Wiseman 2016b).

482   Constantine Sandis against Hare’s consequentialism, an original combination of the doctrine of double effect and the doctrine of doing and allowing.34 Anscombe would have been much more horrified by much of what falls under either of these ‘branches’ today than anything she was objecting to in 1958. Hardly any of it can be attributed as an effect of MMP, though. This is not the case with (b) and (d), so I shall focus on these. I begin with some paradigmatic mainstream positions within normative ethics or ‘moral theory’: Act consequentialism is the claim that an act is morally right if and only if that act maximizes the good. (Sinnott-​Armstrong 2019:§1) An act is wrong if and only if it is forbidden by the code of rules whose internalization by the overwhelming majority of everyone everywhere in each new generation has maximum expected value in terms of well-​being. (Hooker 2000: 32) [A]‌n act is wrong if it would be disallowed by any principle that no one could reasonably respect. (Scanlon 1998: 197) An action is right iff it is what a virtuous agent would characteristically do in the circumstances. (Hursthouse 1999: 17, see also 28–​29) An act is wrong just when such acts are disallowed by the principles that are optimific, uniquely universally willable, and not reasonably rejectable. (Parfit 2011: 413)35 What makes an action morally right is that it originates in a person’s good will. (Sullivan 1989: 117) [D]‌eontologists think that acts are wrong because of the sorts of acts they are. (Davis 1991: 210)36

It’s hardly news that Anscombe would be particularly hostile to consequentialist theories, whose current division into several sub-​species (act and rule focusing on actual, probable, expected, or intended consequences) has partly resulted as a response to some of her arguments in MMP and elsewhere. But what about deontology and virtue ethics? Surely, as a believer in Divine Command Theory and the view that some actions are absolutely prohibited, Anscombe could have no problem with deontology? And was MMP not striving towards a kind of virtue ethics? My answer to both these questions is a negative one.

34  Foot (1967/​1978: 23ff.); for more context, see Hacker-​Wright (2013: 107–​109) and Coope (2015). Anscombe anticipates and rejects a crucial component of trolley reasoning in MMP 40. Sixty years later, philosophy’s most prominent appearance in popular culture is in the trolley-​obsessed The Good Place, in which one of central characters (Chidi Anagonye) is a ‘Professor of Ethics and Moral Philosophy’. The droll conjunction reminds me of the first time I taught ‘Ethics’ for Florida Institute of Technology’s study abroad programme at Oxford. All of the other professors introduced themselves as teaching courses X and Y to great enthusiasm from the students. But when I introduced myself as the lecturer for ‘Moral Philosophy’, I was greeted with baffled silence, until one of the students hesitantly asked ‘Do you mean “Ethics”?’, to all-​round relief. 35  The word ‘virtue’ cannot be found in any of the three volumes of Parfit’s On What Matters. 36  Andreas Lind has brought to my attention the fact that normative theories frequently conflate accounts of rightness conditions with accounts of right/​wrong-​making.

Ethics and Action Theory    483 To begin with, the very issue of what all of these theories are doing viz. attempt to provide accounts of ‘the right action’ is problematic, for two reasons. First, as we have already seen (§I), Anscombe follows Geach in being troubled by the very notion of an action being morally right or wrong, as opposed to good or bad. This worry relates to the larger question of what is meant by ‘action’ in the first place. Robert M. Adams (1979: 74) expresses a commonplace certainty when he writes that ‘[w]‌hat every competent user of ‘wrong’ must know about wrongness is, first of all, that wrongness is a property of actions (perhaps also of intentions and various attitudes, but certainly of actions).’ Accordingly, normative theorists feel licenced to remain silent on the question of what an action is. In fact, it is shocking quite how little they are prepared to say on this topic. The optimistic assumption is that one can simply ‘plug in’ one’s favourite account of action, without this affecting the plausibility of the theory in question, let alone what it would even mean for an action to be right.37 One explanation for this might be that there is nothing to puzzle over here. As Prichard notes: The question ‘What is acting or doing something?’ seems at first unreal, i.e. a question to which we already know the answer. For it looks as though everyone knows what doing something is and would be ready to offer instances. No one, for instance, would hesitate to say to another ‘You ought to go to bed’, on the ground that neither he nor the other knows the kind of thing meant by ‘going to bed.’ Yet, when we consider instances that would be offered, we do not find it easy to state the common character which we think they had which led us to call them actions. (Prichard 1945: 272)

Moral theorists might occasionally say something about whether they take actions to be events, or whether they are talking of act ‘types’ or ‘tokens’.38 By and large, however, one finds little conceptual exploration of the relation between motive, intention, action, and consequence, save perhaps on questions of merely adjacent interest to the main event, such as the ‘doctrine of double effect’ or ‘the doctrine of doing and allowing’. In the second half of the twentieth century, the prevailing view of actions was Davidson’s (Anscombe-​inspired) contention that they are a sub-​set of events.39 Yet hardly anyone seems to care about what it might mean for an event to be right or wrong (morally or otherwise) or, for that matter, morally good or bad.40 Proponents of all sides share a related tendency to draw a hard distinction between the deontic and the evaluative, focusing their interest in action on its rightness or wrongness, and reserving terms like ‘good’ and ‘bad’ for its motives and/​or consequences. From this big leap, it is but a small step to the conclusion that normative ethics is in the business of providing deontic 37 

See Sandis (2017). Schapiro (2001, 2021) and Hurley (2018) also argue that different conceptions of action render competing normative views plausible, but what they really have in mind are different theories of agency. 38  Cf. Wetzel (2006: §2.2) and Hanser (2008). 39  Davidson (1963). 40  By contrast, we know what it is for an event to be considered non-​morally good or bad for us or other creatures (e.g., for a drought to be a bad thing, and subsequent rain good).

484   Constantine Sandis accounts, leaving evaluative concerns to the ‘branch’ that is moral psychology.41 Anyone who insists otherwise is branded a virtue-​ethicist. Should we not at least rejoice in the post-​MMP ‘revival of virtue ethics’? For some the writing was on the wall from the outset: [W]‌hen the phrase ‘virtue ethics’ first came on the scene a number of people, I suspect, must have had a certain sinking feeling—​without perhaps quite realizing why. The thing, we supposed, was almost bound to go to the bad. This gloomy assessment has I think proved quite realistic. (Coope 2006: 20)

Anscombe argued that a philosophical concern with virtue should permeate moral philosophy. Instead, it has led to just one more theory, competing with deontology, consequentialism, and contractualism to provide the best account of right and wrong action.42 Julia Annas laments: Doing the right thing turns out not to be so central in an ethics in which virtue is central. A virtue ethical theory will be interested in virtuous action, but will not get much out of the notion of right action. (Annas 2011: 47)

No one has done more than Rosalind Hursthouse (1999: 26, 28) to put forward virtue ethics on the map as ‘a genuine rival to utilitarianism and deontology’, which can ‘give an account of right action in such a way as to provide action guidance’ (see also Hursthouse 1996). While she did well to emphasise the guiding power of being concerned with virtue, the offering of a normative theory of right action could not be further removed from what Anscombe was hoping to achieve with MMP.43 This may serve to explain Hursthouse’s ambivalence towards this aspect of her own work. She may boast that virtue ethics ‘is at least a possible rival to deontological and utilitarian theories’—​one that ‘can come up with an account of ’ right action—​but she tellingly adds that it does this only ‘under pressure, only in order to maintain a fruitful dialogue with the overwhelming majority of modern philosophers for whom ‘right action’ is the natural phrase’ (Hursthouse 1999: 223, 69; cf. Swanton 2003: 245). It’s as if the wimpy school nerd feels reassured to have finally been accepted by those big nasty bullies, experiencing just a shade of residual resentment. Virtue ethics thus solidifies itself as just one more position within moral theory, offering an account of right action in terms of what the virtuous agent would characteristically do in the circumstances. Such theories allow one to ask whether virtue ethics and consequentialism might be compatible. It is, after all, conceivable that the virtuous person is one

41 

Consequentialism and other mainstream normative theories involve the promotion of goodness (see Korsgaard 1993), while virtue ethics sees goodness as a (not necessarily causal) mark of right action, but neither approach offers accounts of good action. 42  Cf. Solomon (1988, 2003) and Coope (2015). 43  The point is put forth with a panoply of arguments by Coope (2006: 26–​39).

Ethics and Action Theory    485 disposed to perform whichever actions are expected to promote the greatest amount of good.44 If this is what constitutes moral philosophy, then Anscombe is not making a move within it. Her morality is not in direct competition with other normative theories, because it isn’t in the business of producing a theory of right action at all.

21.4  Conclusion Sixty years after MMP, contemporary moral philosophy is replete with consequentialist thinking obsessed with trolley cases, a re-​branded ‘philosophy of psychology’ that replaces conceptual explorations with unrefined findings from cognitive science, an experimental form of ‘moral psychology’ that culminates in the situationist scepticism about character traits, and the espousal of virtue ethics as a theory of ‘morally right action’ that may even be compatible with utilitarianism. By Anscombe’s lights, moral philosophy would seem to have been in far better shape between Moore and MMP than it is now. Whatever one’s assessment of the views of action that Anscombe sought to combat and the account that she put forward in their place, MMP seems to have inadvertently created a wedge between ethics and action theory. This has largely been the result of two consequences that Anscombe could not have easily foreseen, and most certainly didn’t intend. The first is the establishment of ‘virtue ethics’ as one more position within normative ethics, theorising that an action is right if (and only if) it is what the virtuous agent would (advise us to) do. The second is the development of ‘moral psychology’ as an independent and increasingly empirical ‘branch’ of ethics whose interest in questions of motivation, agency, and responsibility have been largely sawed-​off investigations into the good and the right. Ironically, before these interventions British moral theorists were less inclined to separate their defence of any particular account of ‘right action’ from their views in moral and philosophical psychology. Within the work of Joseph, Ross, and Prichard alone, we find a properly philosophical psychology (one that includes conceptual explorations of the relation of action to motive, intention, and the will) to be central to moral thought. To end with a ray of hope: the twenty-​first century has seen a resurgence of neo-​Anscombeans producing impressive work in moral philosophy (e.g., Coope 2006; 44 

See M. Geach (2008: xvii–​xviii). Roger Crisp (1992: 154) has argued for a ‘Utilitarianism of the Virtues’ according to which the virtuous agent lives ‘in such a way that the total amount of utility in the history of the world is brought as close as possible to the maximum’ (cf. Hursthouse 1999: 5, 7–​8). More recently, Crisp (2015, 2019: 142–​145) has defended the additional view that if we understand virtue ethics as providing an account of right and wrong action (as Hursthouse does), then it collapses into a form of deontology. He suggests, further, that this can be avoided by focusing on the question of the value of virtue, as opposed to the notion of right action (cf. Baron 1997; Singleton 1999). Anscombe’s insight, by contrast, is that we cannot even begin to answer questions concerning right action without understanding what it is to act virtuously. It would be a mistake, however, to attempt to transform such an understanding into a normative theory.

486   Constantine Sandis Solomon 2003, 2008; Thompson 2008; Brewer 2009; Vogler 2009; Teichmann 2011; Frey 2019; Hain 2019). This welcome revival of interest in her work is an opportunity for moral philosophers to work well and finally put things right.45

References Adams, Robert Merrihew (1979). ‘Divine Command Metaethics Modified Again’. Journal of Religious Ethics 7, no. 1, 66–​79. Alvarez, Maria and Hyman, John (1998). ‘Agents and Their Actions’. Philosophy 73, 218–​245. Annas, Julia (2011). Intelligent Virtue. Oxford: Oxford University Press. Aquinas, Thomas (1265–​1274). Summa Theologica, Opera Omnia. Bilingual Edition. Vols. 13–​22. Steubenville, OH: Aquinas Institute, 2013. Aristotle (c. 350 bc). Ethica Nicomachea. I. Bywater (ed.). Oxford: Clarendon Press, 1894. Baron, Marcia (1997). ‘Kantian Ethics’. In Baron, Marcia W., Pettit, Phillip, and Slote, Michael (eds.), Three Methods of Ethics. Oxford: Blackwell, 3–​91. Bermúdez, José Luis (2005). Philosophy of Psychology: A Contemporary Introduction. London: Routledge. Blackburn, Simon (2005). ‘Anscombe’s Ethics’. Letter, Times Literary Supplement, October 14. https://​www.the-​tls.co.uk/​artic​les/​priv​ate/​anscom​bes-​eth​ics-​5/​. Block, Ned (1980). Readings in Philosophy of Psychology. Vol. 1. Cambridge, MA: Harvard University Press. Botterill, George and Caruthers, Peter (1999). The Philosophy of Psychology. Cambridge: Cambridge University Press. Brewer, Talbot (2009). The Retrieval of Ethics. Oxford: Oxford University Press. Browne, S. S. S. (1945). ‘Right Acts and Moral Actions.’ Journal of Philosophy 42, no. 19 (September 13), 505–​515. Coope, Christopher (2006). ‘Modern Virtue Ethics.’ In Chappell, Timothy (ed.), Values and Virtues: Aristotelianism in Contemporary Ethics. Oxford: Oxford University Press, 20–​52. Coope, Christopher (2015). ‘Review of David Edmonds “Would You Kill the Fat Man?”’. Philosophical Quarterly 65, no. 259, 275–​313. Crisp, Roger (1992). ‘Utilitarianism and the Life of Virtue’. Philosophical Quarterly 42, no. 167, 139–​160. Crisp, Roger (2015). ‘A Third Method of Ethics?’. Philosophy and Phenomenological Research 90, no. 2, 257–​273. Crisp, Roger (2019). Sacrifice Regained: Morality and Self-​Interest in British Moral Philosophy from Hobbes to Bentham. Oxford: Oxford University Press. Dancy, Jonathan (2004). Ethics without Principles. Oxford: Oxford University Press. Davidson, Donald (1963). ‘Actions, Reasons, and Causes’. Journal of Philosophy 60, 685–​700. Davis, Nancy (Ann) (1991). ‘Contemporary Deontology’. In Singer, Peter (ed.), A Companion to Ethics. Oxford: Blackwell, 205–​218.

45 

An earlier version of this chapter was published in 2020 as ‘Modern Moral Philosophy before and after Anscombe’, in S. Miguens and Mª D. García-​Arnaldos (eds.), Reason, Reasoning, and Action (special issue of Enrahonar dedicated to Anscombe), 39–​62. Special thanks to Louise Chapman, Sophie Grace Chappell, Jennifer Frey, David Hunter, Andreas Lind, and Roger Teichmann, who have all helped to improve it, though it has not improved half as much as they would have liked.

Ethics and Action Theory    487 Diamond, Cora (1997). ‘Consequentialism in Modern Moral Philosophy and in “Modern Moral Philosophy”’. In Oderberg, David S. and Laing, Jacqueline A. (eds.), Human Lives: Critical Essays on Consequentialist Bioethics. London: Palgrave Macmillan, 13–​38. Doyle, James (2018). No Morality, No Self: Anscombe’s Radical Scepticism. Cambridge, MA: Harvard University Press. Doyle, James (2019). ‘How Should We Understand Anscombe’s Claim That “Moral” Does Not Express a Concept?’ Unpublished manuscript. Ewing, Alfred C. (1938). ‘What Is Action?’ Proceedings of the Aristotelian Society, suppl., 17, 86–​101. Foot, Phillipa (1967/​1978). ‘The Problem of Abortion and the Doctrine of Double Effect’. Oxford Review 5, 5–​15, reprinted in her Virtues and Vices (Oxford: Oxford University Press), 19–​32, to which page numbers refer. Frey, Jennifer A. (2018, September 26). ‘Review of No Morality, No Self’. Notre Dame Book Reviews, https://​ndpr.nd.edu/​news/​no-​moral​ity-​no-​self-​anscom​bes-​radi​cal-​ske​ptic​ism/​. Frey, Jennifer A. (2019). ‘Revisiting Modern Moral Philosophy’. Royal Institute of Philosophy Supplement 87, 1–​23. Geach, Mary (2005a, October 14). ‘Anscombe’s Ethics’. Letter, Times Literary Supplement, no. 5349. https://​www.the-​tls.co.uk/​artic​les/​priv​ate/​anscom​bes-​eth​ics/​. Geach, Mary (2005b, October 21). ‘Anscombe’s Ethics’. Letter, Times Literary Supplement, no. 5350. https://​www.the-​tls.co.uk/​artic​les/​priv​ate/​anscom​bes-​eth​ics-​3/​. Geach, Mary (2008). ‘Introduction’. In Geach and Gormally 2008, xiii–​xxi. Geach, Mary and Gormally, Luke (eds.) (2008). Faith in a Hard Ground: Essays on Religion, Philosophy and Ethics: Essays by G. E. M. Anscombe. Exeter: Imprint Academic. Geach, Peter (1956). ‘Good and Evil’. Analysis 17, 33–​42. Gremaschi, Sergio (2017). ‘Anscombe on the Mesmeric Force of “Ought” and a Spurious Kind of Moral Realism’. Ethics & Politics 19, no. 2, 51–​86. Hacker-​Wright, John (2013). Philippa Foot’s Moral Thought. London: Bloomsbury. Hain, Raymond (ed.) (2019). Beyond the Self: Virtue Ethics and the Problem of Culture. Waco, TX: Baylor University Press. Hanser, Matthew (2008). ‘Actions, Acting, and Acting Well’. In Shafer-​Landau, Russ (ed.), Oxford Studies in Metaethics, vol. 3. Oxford: Oxford University Press, 171–​198. Hegel, G. W. F. (1821). Elements of the Philosophy of Right. A. W. Wood (ed.). H. B. Nisbet (trans.). Cambridge: Cambridge University Press, 1991. Hooker, Brad (2002). Ideal Code, Real World. Oxford: Oxford University Press. Hornsby, Jennifer (1980). Actions. London: Routledge. Hornsby, Jennifer (1993). ‘On What’s Intentionally Done’. In Shute, Stephen, Gardner, John, and Horder, Jeremy (eds.), Action and Value in Criminal Law. Oxford: Clarendon Press, 57–​74. Hornsby, Jennifer (2010). ‘Trying to Act’. In O’Connor, Timothy and Sandis, Constantine (eds.), A Companion to the Philosophy of Action. Oxford: Wiley-​Blackwell, 18–​25. Hurley, Paul (2018). ‘Consequentialism and the Standard Story of Action’. Journal of Ethics 22, no. 1, 25–​44. Hursthouse, Rosalind (1996). ‘Normative Virtue Ethics’. In Crisp, Roger (ed.), How Should One Live? Essays on the Virtues. Oxford: Oxford University Press, 19–​33. Hursthouse, Rosalind (1999). Virtue Ethics. Oxford: Oxford University Press. Jensen, Steven J. (2010). Good and Evil Actions: A Journey through Saint Thomas Aquinas. Washington, D.C.: Catholic University of America Press. Joseph, H. W. B. (1931/​1933). Some Problems in Ethics. 2nd (corrected) impression. Oxford: Clarendon Press.

488   Constantine Sandis Kant, Immanuel (1788). Kant: Critique of Practical Reason. M. J. Gregor (ed. and trans.). Cambridge: Cambridge University Press, 1997. Korsgaard, Christine (1993). ‘The Reasons We Can Share: An Attack on the Distinction between Agent-​Relative and Agent-​Neutral Values’. Social Philosophy and Policy 10, no. 1, 24–​51. Macmurray, John (1938). ‘What Is Action?’ Proceedings of the Aristotelian Society, suppl. 17, 69–​85. Mill, John Stuart (1863). Utilitarianism. Sher, George (ed.). Indianapolis, IN: Hackett, 1979. Moore, George Edward (1903). Principia Ethica. Cambridge: Cambridge University Press. Parfit, Derek (2011). On What Matters: Vols. 1 and 2. Oxford: Oxford University Press. Price, Anthony W. (2005). ‘Was Aristotle a Particularist?’ Proceedings of the Boston Area Colloquium of Ancient Philosophy 21, 191–​212. Prichard, Harold Arthur (1932). ‘Duty and Ignorance of Fact’. In Prichard 2004, 8–​101. Prichard, Harold Arthur (1945). ‘Acting, Willing, Desiring’. In Prichard 2004, 272–​281. Prichard, Harold Arthur (2004). Moral Writings. MacAdam, Jim (ed.). Oxford: Clarendon Press. Quante, Michael (1993). Hegel’s Concept of Action. D. Moyer (trans.). Cambridge: Cambridge University Press. Rawls, John (1974–​1975). ‘The Independence of Moral Theory’. Proceedings and Addresses of the American Philosophical Association 48, 5–​22. Richter, Duncan (2011). Anscombe’s Moral Philosophy. Lanham, MD: Lexington Books. Richter, Duncan (2019). ‘Morality and Moral Law: A Reply to James Doyle’. Forma de Vida 17 (Anscombe). https://​form​adev​ida.org/​dritch​erfd​v17. Ross, W. D. (1930). The Right and the Good. Oxford: Clarendon Press. Sandis, Constantine (2010) ‘The Man Who Mistook His Handlung for a Tat: Hegel on Oedipus and Other Tragic Thebans’. Bulletin of the Hegel Society of Great Britain 62, 35–​60. Sandis, Constantine (2012). The Things We Do and Why We Do Them. London: Palgrave Macmillan. Sandis, Constantine (2017). ‘The Doing and the Deed: Action in Normative Ethics’. Royal Institute of Philosophy Supplement 80: Philosophy of Action, 105–​126. Sandis, Constantine (2019a, May 24). ‘Neo-​Anscombians’. Times Literary Supplement. https://​ www.the-​tls.co.uk/​artic​les/​priv​ate/​neo-​ansc​ombi​ans/​. Sandis, Constantine (ed.) (2019b). Philosophy of Action from Suarez to Anscombe. London: Routledge. Sandis, Constantine (2019c). ‘Who Are ‘We’ for Wittgenstein?’ In Appelqvist, Hanne (ed.), Wittgenstein and the Limits of Language. London: Routledge, 172–​196. Sandis, Constantine (2020). ‘Was Jesus a Moral Particularist?’ Philosophers’ Magazine 89, 6–​8. Sandis, Constantine (2021). ‘Virtue Ethics and Particularism’. 95th Proceedings of the Aristotelian Society, suppl. 95, no. 1, 205–​232. Scanlon, Thomas, M. (1998). What We Owe to Each Other. Cambridge, MA: Belknap Press. Schapiro, Tamar (2001). ‘Three Conceptions of Action in Moral Theory’. NOÛS 35, no. 1, 93–​117. Schapiro, Tamar (forthcoming). ‘Kant’s Approach to the Theory of Agency’. In Sylvan, K. and Chang, R. (eds.), The Routledge Handbook of Practical Reason. Sidgwick, Henry (1874). The Methods of Ethics. London: Macmillan. Singleton, Jane (1999). ‘Virtue Ethics, Kantian Ethics, and Consequentialism’. Unpublished manuscript. Sinnott-​Armstrong, Walter (2019). ‘Consequentialism’. In Zalta (ed.), The Stanford Encyclopedia of Philosophy. https://​plato.stanf​ord.edu/​archi​ves/​fall2​021/​entr​ies/​conse​quen​tial​ism.

Ethics and Action Theory    489 Solomon, David (1988). ‘Internal Objections to Virtue Ethics’. Midwest Studies in Philosophy 13, 428–​440. Solomon, David (2003). ‘Virtue Ethics: Radical or Routine?’ In Zagzebski, Linda and DePaul, Michael (eds.), Intellectual Virtue: Perspectives from Ethics and Epistemology. Oxford: Oxford University Press, 57–​80. Solomon, David (2008). ‘Elizabeth Anscombe’s ‘Modern Moral Philosophy’: Fifty Years Later’. Christian Bioethics 14, no. 2, 109–​122. Sullivan, Roger, J. (1989). Immanuel Kant’s Moral Theory. Cambridge: Cambridge University Press. Swanton, Christine (2003). Virtue Ethics: A Pluralistic View. Oxford: Oxford University Press. Swanton, Christine (2015). ‘A Particularist but Codifiable Virtue Ethics’. In Timmons, Mark (ed.), Oxford Studies in Normative Ethics, vol. 5. Oxford: Oxford University Press, 38–​63. Taylor, Charles (1979). ‘Action as Expression’. In Diamond, C. and Teichman, Jenny (eds.), Intention and Intentionality. Sussex: Harvester Press, 73–​89. Taylor, Charles (1983). ‘Hegel and the Philosophy of Action’. In Laitinen, Arto and Sandis, Constantine (eds.), Hegel on Action. London: Palgrave Macmillan, 22–​41. Teichmann, Roger (2008). The Philosophy of Elizabeth Anscombe. Oxford: Oxford University Press. Teichmann, Roger (2011). Nature, Reason, and the Good Life: Ethics for Human Beings. Oxford: Oxford University Press. Thagard, Paul (ed.) (2007). The Philosophy of Psychology and Cognitive Science. Amsterdam: North Holland. Thompson, Michael (2008). Life and Action: Elementary Structures of Practice and Practical Thought. Boston: Harvard University Press. Vogler, Candace (2001). John Stuart Mill’s Deliberative: An Essay in Moral Psychology. New York: Garland. Vogler, Candace (2009). Reasonably Vicious. Boston: Harvard University Press. von Wright, Georg Henrik (1963). Norm and Action. London: Routledge & Kegan Paul. Weiskopf, Daniel and Adams, Fred (2015). An Introduction to the Philosophy of Psychology. Cambridge: Cambridge University Press. Wetzel, Linda (2006). ‘Types and Tokens’. In Zalta, Edward N. (ed.), The Stanford Encyclopedia of Philosophy (Fall 2018 edition). https://​plato.stanf​ord.edu/​archi​ves/​fall2​018/​entr​ies/​types-​ tok​ens. White, Alan, R. (1968). ‘On Being Obliged to Act’. In G. N. A. Vesey (ed.), Royal Institute of Philosophy Lectures, vol. 1: 1966–​67. London: Macmillan, 64–​82. Wiseman, Rachael (2016a). Anscombe’s Intention. London: Routledge. Wiseman, Rachael (2016b). ‘The Intended and Unintended Consequences of Intention’. American Catholic Philosophical Quarterly 90, no. 2, 207–​227. Wiseman, Rachael (2017). ‘G. E. M. Anscombe’. In D. Pritchard (ed.), Oxford Bibliographies in Philosophy. Oxford: Oxford University Press. Wittgenstein, Ludwig (1953). Philosophical Investigations. G. E. M. Anscombe (trans.). Oxford: Blackwell. Wittgenstein, Ludwig (1977/​1998). Culture and Value. G. H. von Wright with H. Nyman (eds.). Peter Winch (trans.) 2nd ed. revised by A. Pichler. Oxford: Blackwell.

Chapter 22

Ansc ombe a nd Wit tgenst e i n on Knowled ge ‘ wi t h ou t Observat i on’ Harold Teichman

22.1 Anscombe’s invocation, in Intention, of an odd, sui generis kind of knowledge, which she takes to be fundamental for any non-​circular, practice-​based, conceptual account (like hers) of what human action is, has engendered considerable confusion and resistance, even amongst her most sympathetic readers. She employs two terms for this odd variety of knowing: knowledge without observation and practical knowledge. The concept denoted by the first term has its origin in Wittgenstein’s late writings, though the term is hers, and the concept denoted by the second was given that name by various medieval thinkers, notably Aquinas—​but the two descriptive names, in Anscombe’s account, point to the same thing seen from two different angles. I will argue that much of the confusion and resistance surrounding Anscombe’s use of these terms is owing (1) to an unshakeable attachment to some form of a ‘justified true belief ’ account1 of all knowledge and (2) to a lack of recognition of the fact that Anscombe’s whole mode of proceeding in Intention is based as much on the methods of the later Wittgenstein as on those of Aristotle. If one pays little attention to Wittgenstein’s methods and results2 and approaches Anscombe’s work from any one of 1  I use this phrase to indicate any post-​Gettier account of knowledge that requires the truth of what is known together with some kind of justification of that knowledge, counterfactual or factual. 2  Wittgenstein’s later (post–​Blue Book) ‘results’ are not philosophical theories, but rather accounts of our concepts in terms of the practices we exhibit in weaving the words expressive of those concepts into

Anscombe and Wittgenstein on Knowledge ‘without Observation’    491 the usual action-​theoretic directions available to us now—​whether it be post-​Davidson analytic action theory or some view of action refracted through a Kantian or Thomist lens, say—​then some misunderstanding is almost inevitable.3

22.2 Taking the second point first, I want initially to point out a close parallel between a key methodological statement by Wittgenstein in RPPII and the first twenty or so sections of Intention, which may help to illuminate Anscombe’s peculiar manner of proceeding in that book and set the stage for her treatment of knowledge without observation. In RPPII §63 Wittgenstein states the outline of a ‘plan’, which he will continue in §148:4 Plan for the treatment of psychological concepts. Psychological verbs characterized by the fact that the third person of the present is to be identified by observation, the first person not. Sentences in the third person of the present: information. In the first person present, expression. ((Not quite right.)) Sensations: their inner connexions and analogies. All have genuine duration. Possibility of giving the beginning and the end. Possibility of their being synchronized, of simultaneous occurrence. All have degrees and qualitative mixtures. Degree: scarcely perceptible—​unendurable. In this sense there is not a sensation of position or movement.5

This statement of a ‘plan’ is not merely aspirational on Wittgenstein’s part; it is buried deep in the middle of an ongoing, minutely focused execution of the plan itself, in the second of the two lengthy RPP typescripts. There is also a foretaste of it in RPPI §836. our activities. It will emerge, I hope, that in Intention Anscombe uses some of Wittgenstein’s results to obtain new results of the same kind. 3  The argument here is entirely exegetical, since we need a clear view of what Anscombe was saying—​which has been much contested—​before criticism can find a secure foothold. So I will not attempt to defend Anscombe’s or Wittgenstein’s views against competing accounts of intention, action, or knowledge. I shall also assume at least a glancing acquaintance with PI and Intention on the part of the reader of this chapter and will not defend my interpretation of the late Wittgenstein against, e.g., attempted ‘resolute’ accounts of his thinking. One hopes that the appropriateness of the interpretation will to some extent ‘take care of itself ’. 4  See ­chapter 2 of Joachim Schulte’s (1993) book Experience and Expression for a discussion of the plan. The details of the plan suffuse Wittgenstein’s post-​1945 notes and typescripts concerning philosophy of psychology, and are worked out in innumerable specific remarks, some of which found their way into PI-​PP. Aficionados of Wittgenstein’s term treatment (Behandlung) should note that it is not used in this passage in the sense in which a physician ‘treats’ an illness (PI, §255). I will go on to use the adjectives ‘psychological’ and ‘mental’ interchangeably in this chapter. 5  The boldface emphasis is mine and will be justified later.

492   Harold Teichman The whole patchwork of remarks on psychology in PI-​PP and the post-​1945 Nachlass testifies in a mass of detail to his devotion to such a plan of conceptual investigation. Our developed psychological concepts (expressed in psychological verbs and verb phrases) are far more subtle than what can be adumbrated in some of the simplest language-​games in the beginning of PI, such as the ‘builder’s game’ in §2, involving ‘shaped-​object’ concepts (or at least proto-​concepts; see RFM, 433) like ‘slab’ or ‘beam’, or the ‘shop for five red apples’ game sketched in §1, which may be seen as outlining very basic versions of our concepts of ‘number’, ‘standing colour’, ‘kind of object’ (as well as undermining the ‘Augustinian’ conception that language fundamentally works by ‘standing for things’). What psychological concepts might seem to ‘refer to’ usually escapes direct public ostension or determination—​slabs can publicly be contrasted with beams, colours can be pointed to (once the linguistic groundwork for such pointing is established), and counts can be agreed upon unambiguously, whereas pains, beliefs, perceptions, expectations, or thoughts are not fully public in the same way (which is not to say they are fully private either, as the private-​language arguments in PI are intended to show). So games which are essential for the acquisition and use of everyday psychological concepts will be, as it were, games on top of games. The latter, more ‘basic’ games will often be more suited to describing material features of our ordinary environment, or will be more broadly applicable and amenable to intersubjective agreement on the state of play than the full-​bore mental games, though these are not hard and fast requirements.6 It might strike us that Wittgenstein’s ‘plan’ begins with a remark that is clearly grammatical in all senses of the word, concerning first-​and third-​person present-​tense psychological utterances: ‘Psychological verbs [are] characterized by the fact that the third person of the present is to be identified by observation, the first person not.’ This is immediately expanded in a corresponding rough distinction, already familiar to readers of PI, between expression in the first person and information in the third. For example, ‘to be in pain’ is a psychological verb. If I avow I’m in pain (truly), this avowal has no observational component of any kind on my part, for Wittgenstein; if I declare ‘She’s in pain’, however, I must do so on the basis of something like my or others’ observation, testimony, or conclusions therefrom. This should be familiar Wittgenstein territory. This opening statement of a verb asymmetry is meant to indicate a quite general feature of our mental-​concept language-​games.7 Let’s call it the mental-​verbs principle. In order to move on to the characterisation of more specific mental concepts, Wittgenstein then outlines the practice of more specific games, each of which individually will respect the mental-​verbs principle. The first, more specific group of mental concepts addressed by Wittgenstein in RPPII §63 is the class of sensation-​concepts (further classes of concepts, mental images 6 

The notion of ‘more basic’ games may be left somewhat vague for our purposes. The precise scope of this criterion for the mental might be disputed, but Wittgenstein clearly meant it to apply to a wide range of prima facie mental concepts. In RPPI §836 he tentatively calls these ‘Erlebnisbegriffe’ (the quote marks are his). 7 

Anscombe and Wittgenstein on Knowledge ‘without Observation’    493 and the emotions, will appear later in this remark and RPPII §148).8 ‘Sensation’ here includes both feelings and perceptions; this usage is common to both Wittgenstein and Anscombe.9 The first general feature of games that define sensation mentioned in §63 is attribution of duration and possible simultaneity: for example, I can talk about how long a pain or a sound lasts and whether I heard a recording start playing at the same moment when I felt the spring-​like resistance as I pushed the ‘play’ button. Now what I want to highlight here, with a view to understanding Anscombe’s procedure in Intention, is how Wittgenstein begins with a general grammatical characterisation of the conceptual territory (the mental) via the mental-​verbs principle, and then continues with finer general determinations of more specific mental concepts (the immediate ones in question being sensation and imagination). In so doing he will exhibit a dependence of the ‘mental’ concept sensation, e.g., on subsidiary—​for his immediate classificational purpose—​concepts like knowledge, observation, expression, image, information (each having some mental overtones itself) and duration, simultaneity, position, degree (which are not prima facie mental concepts at all)—​all of these words listed are mentioned in the text of §63, not all of which was quoted earlier. RPPII §63 continues, in part: Place of feeling in the body: differentiates seeing and hearing from sense of pressure, temperature, taste and pain. (If sensations are characteristic of the position and movement of the limbs, at any rate their place is not the joint.) . . . Pain differentiated from other sensations by a characteristic expression. This makes it akin to joy (which is not a sense-​experience). . . . While I am looking at an object I cannot imagine it.

It is important to note that these boldface ‘subsidiary’ concepts are not meant to be reductive of the sensation-​concepts in any philosophically substantive sense; rather, groups of concepts are simply being situated by Wittgenstein relative to each other (and thus their corresponding games and grammar are so situated). One could call such a web of conceptual elucidations a reduction only if it were complete, and Wittgenstein offers no guarantee of completeness in his grammatical remarks. Here we may catch a glimpse of the layering of ‘games on top of games’ mentioned earlier: the game of talking about seeing, e.g., differs from the game of talking about feeling pain in that the latter involves referring to a position of feeling in the body while the former does not. The game of pain is played ‘on top of ’ the game of position. It also should be emphasised that these brief classifications in RPPII §63 and §148 are not meant to be the result of introspection—​rather they result from the examination of

8  These two classes certainly don’t exhaust the mental, according to Wittgenstein. See his remarks on thought, belief, hope, or expectation, for example, which are neither sensations nor emotions; these remarks litter parts of PI and the later notebooks. 9  See Anscombe (1965).

494   Harold Teichman the language-​games we use in employing the words and concepts involved. This is seen more clearly when specific instances of these concise programmatic remarks are worked out in detail throughout the late notebooks and typescripts.10 With this sketch of Wittgenstein’s plan in mind, a corresponding plan of conceptual ‘analysis’ in the first (roughly) twenty sections of Intention becomes visible. What Anscombe is trying to elucidate in that work is the group of interrelated concepts intentional, reason for action, and the (in)voluntary. She begins by declaring that these concepts can be fully characterised via a basic language-​game with a clearly grammatical flavour—​this is her homologue, for the concept of ‘intention’, of Wittgenstein’s mental-​verbs principle that spans the psychological—​the ‘Why are you φ-​ing?’ game.11 In §16 of Intention, Anscombe, after some already arduous conceptual work, summarises the results she has obtained so far, in specifying the scope of the ‘Why?’ game (note the boldface concepts in the following; again the emphasis is mine). The reader should imagine or try to ‘hear’ Anscombe’s paragraph as a correlate, for Intention, of Wittgenstein’s ‘plan’ for his voluminous unfinished notes on mental concepts in the Nachlass: Intentional actions are a sub-​class of the events in a man’s history which are known to him not just because he observes them. . . . [Such] actions, then, are the ones to which the question ‘Why?’ is given application, in a special sense which is so far explained as follows: the question has not that sense if the answer is evidence or states a cause, including a mental cause; positively the answer may (a) simply mention past history, (b) give an interpretation of the action, or (c) mention something future. In cases (b) and (c) the reason is already characterised as a reason for acting, i.e. as an answer to the question ‘Why?’ in the requisite sense, and in case (a) it is an answer to that question if the ideas of good or harm are involved in its meaning as an answer; or again if further enquiry elicits that it is connected with ‘interpretative’ motive, or intention with which. (Intention, §16)

So I wish to highlight, in the following summaries 1 and 2, the near-​homology of method between Wittgenstein in his post-​1945 notes on psychology and Anscombe in Intention. I submit that this way of reading Intention can help demystify the strangeness of Anscombe’s mode of proceeding—​ a strangeness which made

10  For example, the last line quoted, ‘While I am looking at an object I cannot imagine it’, might mistakenly be taken to be a piece of phenomenology. But see, e.g., RPPI §653: ‘[H]‌is pictures [Bilder] are “independent” of him. What does that mean? He couldn’t use thoughts to dispel them. If, e.g., I imagine [sich vorstellen] the death of my friend, I may tell myself “Don’t think about it, think of something else”; but that wouldn’t be said to me if I were seeing the event before my eyes, e.g. on a film. Then I’d reply to someone who in the assumed case said to me “Don’t think about it”: “Think about it or not, I’m seeing it”.’ The discussion here is entirely of what can be said, by another or by Wittgenstein—​no introspection or phenomenology is invoked; this is typical of most of the remarks in the RPP volumes. 11  I will ignore the past-​tense form of ‘Why?’ in this chapter. It brings in the additional concept (game) of memory, which modifies the ‘can say’ (see later discussion) nature of a present-​tense answer to the question. Anscombe’s use of the present progressive is not accidental.

Anscombe and Wittgenstein on Knowledge ‘without Observation’    495 the book almost incomprehensible to its initial reviewers and still leaves people scratching their heads. 1. Wittgenstein identifies a basic language-​ game grammatical feature that characterises a very wide range of our mental concepts: the mental verbs principle. He further qualifies what counts as an instance of these games using the concepts information and expression. He goes on to portray the subclass of mental concepts that can be called sensations using sensation-​independent12 concepts like duration, simultaneity, degree, location, and related mental concepts like imagination. 2. Anscombe identifies a basic language-​game with a grammatical flavour that characterises our concept of intention: the ‘Why are you φ-​ing?’ game. She further qualifies what counts as an instance of this game (i.e., whether the intended sense of ‘Why?’ ‘has application’) in terms of such non-​voluntary concepts as knowledge, observation, evidence, cause, past/​present/​future, interpretation, goodness, and motive. Why doesn’t Anscombe just try to explain to us, in some traditionally philosophical way, what intention (i.e., the will) is—​in the manner of Kant or Aquinas, say—​some of her readers might well be inclined to ask. I contend that this is precisely the same question as: Why doesn’t Wittgenstein do something more traditionally philosophically palatable in his own late work? (It should be noted that ‘intention’ itself is a mental concept, of course, and verbal expressions of intention themselves obey the mental-​verbs principle: Wittgenstein’s asymmetry between ‘I’m going to φ’ and ‘He’s going to φ’ holds, though Anscombe doesn’t lean on this.) A fully convincing exegesis of this near-​homology between the methods of Wittgenstein and Anscombe of course would require a good deal more elaboration than I will attempt in this space, and I fully acknowledge the presence of a clear Aristotelian influence in Anscombe’s manner of thought that is absent in Wittgenstein. But I would like to propose that the parallels between their approaches that we have briefly set out make an apt frame for our discussion of Anscombe’s ‘appropriation’ of Wittgenstein’s remarks about the sort of knowledge we have of the position and motion of our limbs: knowledge without observation.

22.3 The class of things known without observation by a human being is concisely introduced by Anscombe in the first paragraph of §8 of Intention. The paradigm example of this class is our ordinary knowledge of the position of our limbs—​what is usually called, perhaps tendentiously, kinaesthetic knowledge—​but she will, after justifying this formula as

12 

That is, concepts that may overlap with sensation concepts, but which have wide ranges of application that are not directly ‘mental’.

496   Harold Teichman applying to kinaesthetic phenomena, go on to extend this form of description to apply also to our ordinary knowledge of what we are doing intentionally. Though it probably was not apparent to her original readers, what she says in §8 about limb position and pain is clearly a distillation of what Wittgenstein wrote in various places, including PI, the two volumes of RPP, the first volume of LW, and OC,13 as the following juxtapositions indicate: Anscombe:  [Describing non-​circularly the notion of the ‘involuntary’] can be done as follows: we first point out a particular class of things which are true of a man: namely the class of things which he knows without observation. E.g. a man usually knows the position of his limbs without observation. It is without observation, because nothing shews him the position of his limbs; it is not as if he were going by a tingle in his knee, which is the sign that it is bent and not straight. (Intention, §8) Wittgenstein:  [The interlocutor suggests:] ‘My kinaesthetic sensations apprise me of the movements and positions of my limbs.’ [Wittgenstein replies:] I let my index finger make an easy pendulum movement of small amplitude. I either hardly feel it or don’t feel it at all. Perhaps a little in the tip of the finger, as a slight tension. (Nothing at all in the joint.) And this sensation apprises me of the movement?—​for I can describe the movement exactly. ‘But still, you must feel it, otherwise you wouldn’t know (without looking) how your finger is moving.’ But ‘knowing’ it only means: being able to describe it. (PI-​PP, §§56–​57) Anscombe:  [continuing §8] Where we can speak of separately describable sensations, having which is in some sense our criterion for saying something, then we can speak of observing that thing; but that is not generally so when we know the position of our limbs. Yet, without prompting, we can say it. Wittgenstein:  If I did not see that my arm has moved after being convinced I had moved it with my face turned away, I should become confused and should presumably trust my eyes. Seeing can at any rate tell me whether I have carried out my intended movement exactly, e.g. have reached the position that I wanted to reach; the feeling wasn’t able to do that. I feel that I am moving all right, and I can also judge roughly how by the feeling—​but I simply know what movement I have made, although you couldn’t speak of any sense-​datum of the movement, of any immediate inner picture of the movement. And when I say ‘I simply know . . .’ ‘knowing’ here means something like ‘being able to say’ [sagen können] and is not in turn, say, some kind of inner picture. (RPPI, §390) Anscombe:  [continuing §8] I say however that we know it and not merely can say it, because there is a possibility of right or wrong: there is point in speaking of knowledge only where a contrast exists between ‘he knows’ and ‘he (merely) thinks he knows.’

13 

See, notably, RPPI, §§382–​408, 455, 698, 754, 758, 767, 771–​773, 784–​786, 790, 794–​798, 843–​844, 948; RPPII, §63; LW1, §§386–​389 (repeated in LW2, p. 4, and PI-​PP, §§56–​59); OC, §41.

Anscombe and Wittgenstein on Knowledge ‘without Observation’    497 Wittgenstein:  For ‘I know’ seems to describe a state of affairs which guarantees what is known, guarantees it as a fact. One always forgets the expression ‘I thought I knew.’ (OC, §12) Anscombe:  [continuing §8] Thus, although there is a similarity between giving the position of one’s limbs and giving the place of one’s pain, I should wish to say that one ordinarily knows the position of one’s limbs, without observation, but not that being able to say where one feels pain is a case of something known. Wittgenstein:  ‘I know where I am feeling pain’, ‘I know that I feel it here’ is as wrong as ‘I know that I am in pain’. But ‘I know where you touched my arm’ is right. (OC, §41)

Wittgenstein, throughout his writings, calls attention to our different uses of words like ‘know’: he has no unitary account of knowledge, because we, in our usage, do not use the word in a wholly unitary fashion. In OC he discusses other senses of ‘know’ than the one he and Anscombe are interested in here, for example, the ‘I know’ of ordinary empirical knowledge or the ‘I know’ that means ‘I have learned from authority.’ I shall try to make clear the sense of ‘I know’ that’s relevant here, that is, the sense in question when Anscombe, prompted by Wittgenstein, speaks of knowledge without observation, via two connected definitions, as follows: (CS) ‘I can say X’ means by definition ‘I can, with unique authority, and without any grounds, observational or otherwise, vouch for X.’ (KWO) ‘I know X without observation’ (abbreviated as ‘I knowWO X’) means by definition ‘I can say X (as defined in CS)—​but what I (sincerely and understandably) vouch for in saying X is not indisputably true, and in exceptional cases I am subject to objective public correction of this knowledge by myself or someone else.’

Wittgenstein introduced these concepts, these fundamental usages, in connection with our talk about the position of our limbs and our pains—​or, perhaps more accurately, he observed these patterns of usage in those kinds of talk. The CS part of the usages is common to both language-​games, those to do with our limbs and those to do with our pains. Anscombe, in a stroke of brilliance, noticed that KWO applies formally to games other than limb position as well, to those involving intention, to part of what we call the involuntary, and to what she termed ‘mental’ efficient causes, such as the unexpected sight of a face at the window.14 Therefore, following Anscombe and Wittgenstein, we may think of KWO, an ‘avowal’ concept that specialises CS, as a specific logical form of ‘knowledge’ (knowledge of a kind that is clearly not ‘justified true belief ’ since it has no justification and may sometimes be untrue) having rather broad applicability to a number of our most fundamental concepts concerning human feeling and acting. What Anscombe writes about knowledge in Intention can be fully understood only if one 14 

See especially Intention, §§6–​10 and 16.

498   Harold Teichman keeps in mind, throughout the book, her Wittgensteinian leitmotif: ‘there is point in speaking of knowledge only where a contrast exists between ‘he knows’ and ‘he (merely) thinks he knows.’ The ‘pure’ case of ‘can say’ allows no possibility of error on the part of the sincere sayer under ordinary circumstances, i.e., no correction of what is thereby said by others: examples of this case are my honestly saying that I’m in pain or saying where it hurts me. A hearer’s disputing what I vouch for in this way will involve adducing grounds for dispute on her part: something like doubting my sincerity or my mental health, or expressing confusion about what I could possibly have meant—​’as, e.g., if you say that your foot, not your hand, is very sore, but it is your hand you nurse’ (Intention, 14). The ‘impure’ case of ‘can say’, i.e., the case in which we ‘knowWO’, allows for the sincere comprehensible knower sometimes being wrong and being corrected, though being wrong is never the norm here: Anscombe’s examples of this form of knowledge are knowing the position and motion of one’s limbs, knowing one’s intention (what one is intentionally doing),15 knowing a ‘mental cause’ such as a loud bang that made me jump, and knowing an involuntary movement of my body, like giving a jerk as I fall asleep.16 The ‘can’ on the right-​hand side of the CS definition is a logical (i.e., grammatical, semantic) ‘can’: it means there is a possibility of declaring the truth in this way that is licenced by the language game(s) in play. It is analogous to the ‘can’ in ‘You can move your rook in such-​and-​such a way’ or ‘A unique tangent can be found at any point on a circle’ (for Wittgenstein mathematics is rule-​governed in roughly the same sense that chess is). ‘With unique authority’ means that in any game in which ‘can say’ is valid, the person who says something in this way is the one who gets to say it, according to the rules of the game. (As noted, what is said with this unique authority may sometimes be questioned, though no other speaker gets to ‘take over’ the authority.) ‘Without any grounds’ means no results of observation, perception, feeling, fact-​checking, or logical consequence are relevant for me, the sayer, to what I truly can say in this sense: the question ‘How do I know?’ addressed to myself here makes no sense. The additional qualification, over and above what is meant in ‘I can say X’ in the KWO definition of ‘I knowWO X’, which requires my corrigibility in so saying X, means that in a logical sense KWO is stronger than CS (since the former must satisfy an additional condition), whereas in an epistemic sense KWO is weaker than CS (since the former is subject to correction and the latter is not). When Anscombe says, regarding one’s knowledge of the position of one’s limbs, ‘I call this sort of being able to say “knowledge” and not merely “being able to say” ’ (Intention §8, 14), her ‘merely’ is an indication that she’s talking about the logical sense.17

15  This application of ‘knowing without observation’ to intention references KWO, but contains a refinement that will be explained. 16  For an excellent discussion of the basics of Anscombe’s reasoning about these seemingly odd categorisations, see Wiseman (2016, 92–​97). 17  This ambiguity concerning which is ‘stronger’, ‘know WO’ or ‘can say’, may have been behind Michael Thompson’s (2011, 198) twitting Anscombe for her Wittgensteinian ‘tics’: ‘Whether my foot hurts or

Anscombe and Wittgenstein on Knowledge ‘without Observation’    499 It should be noted that it is part of the meaning of ‘vouch for’ in the definition of ‘can say’ that ‘I can say X’ entails that I’m aware of, conscious of—​or at least can come to be aware of at the time of saying what I ‘can say’—​what is thereby said, X. This doesn’t mean that I necessarily had to have a conscious thought of X ever before being asked. If asked, ‘Why are you going upstairs? Are you fetching your camera?’, for example, I may reply, ‘No, I’m looking for the camera manual, since I’m planning to do a long exposure later’, without any implication that I explicitly had the thought that I needed to go upstairs to look for the manual at some moment before, or even during, my starting up the stairs. This ‘vouch for’ part of the definition of ‘knowWO’ has the consequence that a sincere denial that I was aware of—​that is, that I knew in the sense of awareness (part of the language game of ‘Why?’ here, not a bit of introspective psychology)—​what I was interrogatively accused of doing, is a denial that I can vouch for the accuser’s description, and is thus a denial that I knewWO that I was acting under that description, which is why Anscombe says ‘[‘Why?’] is refused application by the answer ‘I was not aware I was doing that.’’ (It must be said that Wittgenstein, as far as I can tell, did not explicitly consider the ‘awareness’ sense of ‘knowledge’, though he may be touching on this sense in RPPI §564 or §735.) We may now approach the key step in Anscombe’s application of the KWO form to intention. In the cases of limb position or what the ‘mental cause’ of some involuntary response was, the knowledge without observation in question is of a state of affairs. It may be corrected by a third party who is observing me in the normal way we make public corrections concerning objective states of affairs. ‘I’m sorry, you don’t have your legs crossed: I can see them’ or ‘It wasn’t a loud bang, it was an odd surreptitious scratching that made you jump—​in fact I was recording with my phone and can play it back for you’ may serve properly as public corrections of particular descriptive errors in my vouching for my knowledgeWO. These mistaken avowals by me about my legs or a loud bang were not expressions of speculative knowledge, since they had no grounds (by definition), but they are nevertheless contradicted by my interlocutor in the same way she would contradict a grounded speculative assertion of empirical fact. The crucial move Anscombe makes at this point, in order to apply the general KWO form of knowledge to the game of my knowing what I’m doing, is to allow objective public correction only of my execution of what I’m doing, not of those descriptions that I vouch for when I honestly say I’m doing X but fail in pulling X off in some way. If the KWO form is to apply to what I do, as opposed to where my limbs are, the only correction (some form of corrigibility is needed for my ‘can say’ additionally to rise to the

whether I intend to leave Uppsala some day . . . these are supposed to be things I can say. I cannot myself be said to know them “except perhaps as a joke.” ’ Thompson has missed the point of Anscombe’s distinctions; she would indeed say he can have knownWO that he intended to leave Uppsala, though not whether his foot hurt. (There is no evidence that Anscombe made a distinction between the idioms going to φ and intend to φ, although she mostly favoured the former: ‘This consideration disinclines us to call [expression of intention] a prediction . . . even though “I intend to go for a walk but shall not go for a walk” does sound in some way contradictory’ [Intention, 5]). See also Wiseman (2016, 91–​92).

500   Harold Teichman status of a ‘knowingWO’) allowed in the ‘intentional action’ game is of how well I pull off, execute what I’ve avowed I’m doing. My avowal of the description of what I’m doing, qua description, is not subject to criticism in the doing-​something game, even if the description turns out to be strictly false owing to my botched execution of an action fitting that description. If the botched execution is pointed out to me after my avowal of Xing, I can say, ‘In any case I was Xing, although I didn’t succeed.’ We may slightly rewrite KWO, in its intentional version, to reflect this account, as (KWOI) ‘I know I’m Xing without observation’ (abbreviated as ‘I knowWO I’m Xing’) means by definition ‘I can say I’m Xing (as defined in CS)—​but what I (sincerely and understandably) vouch for in saying I’m Xing is not indisputably true in its execution, and in exceptional cases I am subject to objective public correction of what I said I was doing, in light of its faulty execution, by myself or someone else.’

It might be objected to this formulation that it sounds rather like the assertion that I ‘really ‘do’ in the intentional sense whatever I think I am doing’, which Anscombe rejects in Intention §29 as a way out of her difficulties. But KWOI, as the form of a practice, depends on there being a quite regular, quotidian identity between what I can say I’m doing and what I actually succeed in doing: whence the ‘exceptional cases’ qualification in KWOI. When in such an exceptional case it emerges that I merely thought I knewWO I was Xing, my knowledgeWO has been compromised, though the description of what I sincerely thought I was doing is not: an honest description is the basis for impugning the execution. Whether the description is properly understood and genuine, though, can be questioned; there may sometimes only be what Wittgenstein called ‘imponderable evidence’ (PI-​PP, §§355–​360) of an intention. In this situation Anscombe says, ‘[T]‌here comes a point where a man can say “This is my intention” and no one else can contribute anything to settle the matter. . . . I.e. here “knows” only means “can say” ’ (Intention §27, 48). Consider how KWOI fits with what Anscombe says about contradicting an expression of intention, an expression of knowledgeWO, in Intention §31—​here ‘contradict’ has the oldest sense in the OED, that of ‘speaking against’: ‘I’m poisoning the inhabitants’ said by her pumper is contradicted by ‘You won’t, for I’m going to stop you’, she says, not ‘You aren’t, in fact’. The contradictor thus allows the description ‘poisoning’ to stand, but predicts he will interfere with or prevent the execution of what the pumper knowsWO he is doing. This way of reading the application of the logical category knowledge without observation to intentional action is, I think, the only clear way to make charitable sense of Anscombe’s ‘dark sayings’ in Intention about execution and Theophrastus’s Principle. Kim Frost (2019, 333) has noticed that Anscombe should be seen as cashing in her early aside (Intention, 4) that ‘there are other ways of saying what is not true besides lying and being mistaken’ later in Intention, and we may read §45 in that way, where full-​blown practical knowledge and the ‘writing I am a fool on the blackboard’ case are discussed.

Anscombe and Wittgenstein on Knowledge ‘without Observation’    501 In that place, Anscombe appears to directly say, with infuriating concision, that if something went wrong with the chalk I was holding, such that the phrase ‘I am a fool’ did not legibly appear when I’d intended to write that, my knowledge of my intention in failure was the same as my knowledge would have been if I’d managed to write the phrase successfully. I want to say that that’s exactly what she did mean, regarding her (practical) knowledge (without observation) of what she had been writing. If one allows the ‘funny’ sort of knowledgeWO that I’ve described in KWOI to be what’s in play here, the paradoxical appearance of these passages evaporates. She knewWO (could say) what she was writing in either case, but stumbled in writing it in the bad case, which made her execution, though not her description, corrigible.18

22.4 Armed with the logical category KWOI and its application to intentional action, we may now address some of the daunting puzzles in Anscombe’s treatment of practical knowledge in Intention §§45–​48. She first broaches the topic of practical knowledge in §32 of that work, as something the medievals saw a need for but which modern philosophy seems to have forgotten, immediately after she describes the situation of someone who inadvertently presses the wrong button: The case that we now want to consider is that of an agent who says what he is at present doing. Now suppose what he says is not true. . . . [And what if he is] simply not doing what he says? As when I say to myself ‘Now I press Button A’—​pressing Button B—​a thing which can certainly happen. . . . And here, to use Theophrastus’ expression again, the mistake is not one of judgment but of performance. That is, we do not say: What you said was a mistake, because it was supposed to describe what you did and did not describe it, but: What you did was a mistake, because it was not in accordance with what you said. (Intention, §32, 56–​57)

Here, applying the KWOI pattern, the button-​presser has knowledgeWO of his pressing Button A, whereas observation (by another or himself) will show that he accidentally pressed Button B. The knowledgeWO of what he was doing, however, i.e., pressing Button A, on this reading remains valid after the error, and the point of this ‘funny’ concept of knowledge can perhaps be seen in its justifying the use of the adverb ‘accidentally’ here: ‘Why did you press Button B?’ ‘I was pressing Button A, but hit Button B by mistake, since I usually press Button B.’ The button-​presser’s hearer may believe this reply 18 

Anscombe is not entirely clear whether it matters whether one says in such a case ‘I thought I knew’ or ‘I knew (falsely and corrigibly)’. In her first discussion of the case of writing something incorrectly (Intention, §29, 53), she says ‘without the eyes he knows what he writes’ (this occurs in a section in which she is raising problems for herself, but she seems to assert it in her own voice). I treat ‘I knew but now am properly corrected’ as equivalent to ‘I thought I knew’ in KWOI cases throughout.

502   Harold Teichman or not, but if he does, he is acknowledging the button-​presser’s knowledgeWO of what he was doing. Note that Anscombe says nothing here or elsewhere to suggest that she is distinguishing what counts as ‘knowing without observation what one is doing’ from ‘having practical knowledge of what one is doing’, and I will thus take them to be one and the same thing. The point of having two names for the same knowledge will emerge in Intention §48. Anscombe then turns, in §47,19 to the ‘form of description of intentional actions’. Her discussion here is notoriously difficult to interpret; it might even be said to positively invite misunderstanding. It looks as if she is dividing this form into a threefold group of verb phrases exemplified in (a) ‘sliding (passively) on ice for the fun of it’, (b) ‘intruding on a gathering’, and (c) ‘telephoning one’s friend’, listed in ascending order of ‘implied intentionality’ of the description-​verbs in question. Verbs of type (a) have no implied intentional form unless complemented by an ‘in order to’ or interpretative phrase like ‘for the fun of it’; those of type (b) depend on the ‘Why?’ game for their existence as concepts, though they may be used non-​intentionally sometimes; and those of type (c) are almost invariably used to express intention. It is true that she outlines this threefold classification of the rich selection of verbs we have at our disposal that stand ready to express descriptions of intentional action. It seems to me false, however, that she intends this classification itself to be the essence of the ‘form of description of intentional actions.’ The point is that, in analogy with how (Anscombe 1971, 137) we can learn the general term ‘cause’ only after learning a host of particular verbs displaying the form of description ‘cause’—​like infect, scrape, burn—​ we can learn the ‘Why?’ game only in the course of learning a host of verbs possibly displaying the form of description ‘intentional action.’ The key sentences here are ‘In fact the term ‘intentional’ has reference to a form of description of events. What is essential to this form is displayed by the results of our enquiries into the question “Why?” ’ (Intention, 84). Looking ahead to §48, I shall take this to mean that any verb phrase φ that can truly describe an event E involving what a single human body does or undergoes, such that E (under description φ) is properly subject to Intention’s ‘Why are you φ-​ing?’ question, has the requisite form of description. The climax of this section of the book—​the place where the ‘form of description of intentional actions’ is put to work and where the phrase ‘practical knowledge’ is allowed to throw a whole new light on KWOI as so far explained—​is Anscombe’s taut working out of the meaning of a Thomistic formula granting causal power to practical knowledge: 19 

I agree with Schwenkler (2015, 2019) that the ‘maximal’ concept of practical knowledge that Anscombe begins anew with in Intention §45—​after a long digression on practical reason, its relation to intention ‘with which’, wanting, and picking alternatives—​the case of a master architect who builds his structure in his imagination in the finest detail, choosing alternatives as he goes, is inspired by Aquinas’s picture of the mind of God creating what it already fully comprehends. Since Anscombe isn’t doing theology in Intention, I will treat this idealisation of human practical knowledge as an illustrative digression; in fact in the text she immediately falls back to earth, to the homely case of writing on the blackboard, mentioned earlier.

Anscombe and Wittgenstein on Knowledge ‘without Observation’    503 [W]‌e can say that where (a) the description of an event is of a type to be formally the description of an executed intention [and] (b) the event is actually the execution of an intention (by our criteria) then the account given by Aquinas of the nature of practical knowledge holds: Practical knowledge is ‘the cause of what it understands’, unlike ‘speculative’ knowledge, which is ‘derived from the objects known.’ (Intention, 87, my boldface)

The key to interpreting the point of the conjoined conditions (a) and (b),20 which is not at all obvious from Anscombe’s text, I shall argue, is the preceding passage on the same page: [F]‌ailure to achieve what one would finally like to achieve is common; and in particular the attainment of something falling under the desirability characterization in the first premise [of the practical reasoning corresponding to a hierarchy of means to an end in action]. It often happens that people do things for pleasure and perhaps get none or little, or for health without success . . . and these failures interest us. What is necessarily the rare exception is for a man’s performance in its more immediate descriptions not to be what he supposes. (Intention, 87)

What she is describing here might be called trying to achieve a desirable goal by truly acting intentionally but not hitting that goal, while still managing intentionally to achieve some means (a ‘more immediate description’ of an action) to that goal. For example, trying to improve one’s health by a deliberate dietary change that turns out in fact to be inefficacious (according to later medical science): replacing green beans with beef, say, in an Atkins diet. If on a particular occasion, I am truly describable as deliberately eating beef for my health but I am not truly describable as improving my health (a more remote and general goal), my action of eating (an ‘event’) satisfies conditions (a) and (b) of the ‘causal’ paragraph under the ‘eating beef ’ description—​ Why am I doing it? For my health. Did I carry it out in actuality (eating beef)? Yes—​ and my practical knowledge (without observation) that I am deliberately eating beef is here the cause of what it understands: my eating beef; that I know I’m doing it brings about what I’m doing. (More on this later: I maintain that the important sense of ‘cause’ here is efficient cause.) However, the same event, under the ‘less immediate’ description ‘improving my health’, fails condition (b), since the health improvement is not actual, and we cannot invoke Thomas’s formula; my knowledge without observation that I’m improving my health fails to cause my health to improve. Another case of satisfying (a) but failing (b) is straightforward: a direct failure in execution. So I may mistake cooked ostrich for beef, and eat that when I meant to consume beef. Then the description ‘eating beef ’ was formally appropriate but did not actually occur. In that case also Thomas’s formula fails: I

20 

Setiya (2017a) clearly expresses the interpretative difficulty in making the argument in Intention §§47–​48 consistent.

504   Harold Teichman had practical knowledge, but it didn’t cause what it understands—​eating beef—​rather it caused the eating of ostrich. A third case of failing (b) would be a situation in which my car is slowly rolling backward, unbeknownst to me, the driver, while I have my foot on the brake pedal and intend to be stationary. (This has happened to me.) The description ‘rolling slowly backward in my car’ can express an intention or not: it has the form required by condition (a), but it does not describe an intentional action in fact because of my lack of awareness, and thus fails condition (b). What about condition (a), the formal condition, failing? Well, then there is no practical knowledge of what I’m doing under that description, since the question ‘Why?’ lacks application. No practical knowledge, no causation by practical knowledge. So (a) is a minimal requirement for both intentional description and Thomas’s formula. Requirement (b) serves to filter out (i) the ‘less immediate descriptions’ of my action that may be too remote to find execution in practice, and (ii) formally intentional true descriptions that are not actually intentional, and also to require the ‘more immediate descriptions’ of my action to be truly carried out without error. So being the cause of what it understands is on Anscombe’s account not the definition of practical knowledge, as it seems to have been for Aquinas. What defines it is the KWOI logical form, but Anscombe has in §48 asserted that that form leads to the truth of what Aquinas said about practical knowledge in many actual cases: those situations where we have learned to use a rather ‘immediate’ description of what we are doing, a description that has the logical form required to be in play in the ‘Why?’ language-​game, and where we have in fact successfully done what we knewWO we were doing under that description. In such cases Anscombe’s formula ‘I do what happens’ applies, and practical knowledge is the cause of what it understands: that is, it is the cause of what I’m doing under the relevant description. Now it has sometimes been asserted in the literature21 that Anscombe here is talking about formal causation (in Aristotle’s sense) only. It is trivially true, of course, that where the Thomistic formula applies to practical knowledge, it is that knowledge which confers the form on what happens: ‘it is the agent’s knowledge of what he is doing that gives the descriptions under which what is going on is the execution of an intention’ (Intention, 87). But what is much more interesting here is that what is being described is also a kind of efficient causation. This need not be contrary to our intuitions about the situation at all; after all, it would not be unnatural to say, in a case of successful intentional action, that my practical knowledge of what I’m doing, as I’m doing it, brings about what I’m doing. Or that what I’m doing derives from, arises out of, comes of (Anscombe 1971, 136) my concurrent practical knowledge. In fact, Hornsby has pointed out an implicit appeal in Intention to a form of efficient causation that escapes the Davidsonian ‘event causes event’ pattern, which she calls agent

21  Setiya (2017a, 159) explicitly makes this claim, and Moran (2004, 67) seems to as well. Schwenkler (2015) emphasises the formal claim, but his (2019, 173–​175) acknowledges a kind of efficient causality as well.

Anscombe and Wittgenstein on Knowledge ‘without Observation’    505 causation.22 I take Hornsbian agent causation to be an apt name for the efficient causation expressed in the Thomistic formula. There are several exegetical reasons favouring the ‘efficient’ conclusion. (1) Anscombe was very well aware of Aristotle’s four causes in 1957, but in the passage in question, citing a medieval Aristotelian (!), she declines to use the term formal cause. (2) Nowhere else in Intention does she speak of anything but efficient causes: seeing a face at the window that makes me jump is efficient; hearing the martial music making me pace up and down is efficient; a thought hammering away in my mind making me sign a document is efficient (all in §5). The bark of the crocodile making me jump in §8 is efficient. (3) What Anscombe says practical knowledge is ‘more than’, in the same place (Intention, 87–​88), has the form of efficient cause: it’s more than what’s ‘observed to be a necessary condition of the production of necessary results’ and more than that ‘an idea of doing such-​and-​such in such-​and-​such ways is such a [necessary] condition.’ These formulations embody the concept of necessary efficient cause that is criticised as lacking application to all efficient causes in her later paper (Anscombe 1971); she doesn’t deny these phrases’ applicability here, but declares them ‘less than’ or perhaps ‘weaker than’ what she wants to say. If she had simply wanted to point out that the Thomistic formula’s causality was formal (and not efficient), her existing text would be quite odd.23 (4) As Hornsby has pointed out,24 there is a sharp parallel between the discussion in §47 and Anscombe’s list of causative verbs in Anscombe (1971). What is that paper concerned

22 

Hornsby (2011, 105–​110ff.). Hornsby considers a person (‘Ann’) carrying—​‘carrying’ is a ‘causative verb’ for Anscombe (1975)—​a suitcase. Hornsby (2011, 107) says, ‘The causality here is internal to an event. Ann’s carrying the suitcase is the event of its being carried’. Hornsby is not asserting that the event of carrying is self-​caused; rather, she is noting the concurrency and inseparability of the carrying and its cause: that’s what makes the cause ‘internal’. Note that Ann’s KWOI, my presumptive cause, is concurrent with and inseparable from her action of carrying. ‘We might say that in such cases the person does the thing . . . non-​mediately’ (Hornsby 2011, 108). No mediation is conceptually needed between my knowledgeWO of my carrying, as I do it, and my carrying. So Hornsby’s agent causality may be seen as describing the same thing as KWOI efficiently causing ‘what it understands’. To the objection that any form of the agent’s knowledge is a mere property of the agent, and that agent causality properly speaking requires the agent, not one of her properties, to be a cause, I would reply that KWOI is not an incidental agential property, and press the question: What is an agent, for Anscombe? The paradigm of human agency (as opposed to the agency of a human actor’s physical mass or of non-​human animals) is clearly intentional action. In her essay ‘The First Person’, Anscombe (1975, 33–​35) explicitly connects what I have called KWOI with the essence of the use of ‘I’. This would apply to ‘I am φ-​ing (intentionally)’, which seems to be a fine expression of agency, even though ‘I’ doesn’t refer to any agent, she says. These matters clearly deserve further discussion; I am here merely recommending a reading along these lines as illuminative. 23  The ‘more than’ form of words might conceivably mean she intended to indicate a ‘stronger’ form of efficient causality than that expressed in necessary generalities. Or that there was an additional condition to be attached to the efficient-​causality phrases (perhaps just that of exhibiting formal causality as well). Her next sentence declares that without practical knowledge, what happens fails to be describable as execution of intentions. This would indicate a necessity for the practical knowledge, whether logical (formal cause) or ‘lacking this, no efficient cause’, or both. Interpretation is vexed, but doesn’t well support a ‘merely formal’ reading. 24  Hornsby (2011, 106).

506   Harold Teichman with? Answer: efficient causality, exclusively. The list of causative verbs found there (most of them, anyway) would fit on Anscombe’s left-​hand list in Intention §47, and so are capable of describing an event of doing that is known without observation and subject to the ‘Why?’ game: ‘[I]‌n learning to speak we learned the linguistic representation and application of a host of causal concepts. A small selection: scrape, push, wet, carry, eat, burn, knock over, keep off, squash, make (e.g. noises, paper boats), hurt’ (Anscombe 1971, 137). So, in summary, Anscombe’s concept of intention is defined by her ‘Why are you φ-​ ing?’ language-​game, which is itself partly determined by the concept of knowledge without observation, as applied to intention (what I have called the KWOI form of knowledge). This form of knowledge remains knowledge of a description even when what it knows (an action under that description) has misfired or been interrupted in execution, and the description does not hold. This form of knowledge may also be called ‘practical’, and when certain common conditions are met it is truly the efficient, as well as the formal, cause of what it understands, as Aquinas said. This is Anscombean agent causation, in which the practical knowledge of a human agent is described as causing an event directly and concurrently, without the aid of a prior Davidsonian event.

References Abbreviations of Wittgenstein’s Works PI Philosophical Investigations, revised 4th edition, Hacker and Schulte, eds. Chichester: Wiley-​ Blackwell, 2009. PI-​PP Philosophy of Psychology—​A Fragment, in PI. Previously known as Part II of PI. RFM Remarks on the Foundations of Mathematics, revised edition. Cambridge MA: The MIT Press, 1978. RPPI Remarks on the Philosophy of Psychology, volume I. Chicago: University of Chicago Press, 1980. RPPII Remarks on the Philosophy of Psychology, volume II. Chicago: University of Chicago Press, 1980. LW1 Last Writings on the Philosophy of Psychology, volume I. Chicago: University of Chicago Press, 1982. LW2 Last Writings on the Philosophy of Psychology, volume II. Oxford: Basil Blackwell, 1992. OC On Certainty. Oxford: Basil Blackwell, 1969.

Other Works Cited Anscombe, G. E. M. (1963) Intention, 2nd edition. Oxford: Basil Blackwell. Anscombe, G. E. M. (1965) ‘The Intentionality of Sensation’. In Anscombe (1981). Anscombe, G. E. M. (1971) ‘Causality and Determination’. In Anscombe (1981). Anscombe, G. E. M. (1975) ‘The First Person’. In Anscombe (1981). Anscombe, G. E. M. (1981) Collected Philosophical Papers. Volume 2. Minneapolis: University of Minnesota Press.

Anscombe and Wittgenstein on Knowledge ‘without Observation’    507 Ford, Anton, Hornsby, Jennifer and Stoutland, Frederick, eds. (2011) Essays on Anscombe’s Intention. Cambridge MA: Harvard University Press. Frost, Kim (2019) ‘A Metaphysics for Practical Knowledge’. Canadian Journal of Philosophy 49 (3), 314–​340. Hornsby, Jennifer (2011) ‘Actions in Their Circumstances’. In Ford, Hornsby, and Stoutland (2011). Hyman, John, and Steward, Helen, eds. (2004) Agency and Action –​Royal Institute of Philosophy Supplement: 55. Cambridge: Cambridge University Press. Moran, Richard (2004) ‘Anscombe on “Practical Knowledge”’. In: Hyman and Steward (2004), 43–​68. Schulte, Joachim (1993) Experience and Expression: Wittgenstein’s Philosophy of Psychology. Oxford: Clarendon Press. Schwenkler, John (2015) ‘Understanding Practical Knowledge’. Philosophers’ Imprint 15 (15), 1–​33. Schwenkler, John (2019) Anscombe’s Intention: A Guide. New York: Oxford University Press. Setiya, Keiran (2017a) ‘Anscombe on Practical Knowledge’. In Setiya (2017b). Setiya, Kieran (2017b) Practical Knowledge: Selected Essays. New York: Oxford University Press. Thompson, Michael (2011) ‘Anscombe’s Intention and Practical Knowledge’. In Ford, Hornsby, and Stoutland (2011). Wiseman, Rachael (2016) Anscombe’s Intention. Abingdon: Routledge.

Index

For the benefit of digital users, indexed terms that span two pages (e.g., 52–​53) may, on occasion, appear on only one of those pages. A abortion, 277, 284, 431–​32 absolutism. See moral absolutism accountability, 57–​58 acting well, 7–​8, 13, 14, 16–​17, 21, 22–​23, 206–​8 action-​descriptions, 13, 14, 15–​16, 21, 59–​60, 62–​63, 129, 168–​69, 184, 190–​91, 363, 424–​25, 470, 472, 502–​3, 504, 506 intentional–​material distinction, 381–​82 moral, 129, 132–​33, 424–​25 threefold verb classification, 502 actus hominis /​actus humanus, 57 Adams, Robert M, 483 adultery, 54, 59–​62, 63–​64, 273, 324–​25, 333–​35 agent knowledge, 53–​55, 58–​62, 69–​70 as practical knowledge, 60–​61 agent or patient [unmediated] conceptions. See Anscombianly-​preferred examples akrasia, 108, 123, 127 animals, 362. See also non-​language users actions of, 24 Annas, Julia, 484 Anscombe, Elizabeth archive (University of Pennsylvania), 2 BBC radio broadcast (1957), 230, 258–​60, 258–​59n.143, 267–​68 and Blackfriars Priory, 233 and Catholic Social Guild, 235–​36 conversion to Catholicism, 2, 231, 232, 445–​46 on departed souls, 435, 462 and economic justice, 235–​36 family and childhood, 231–​32 ‘I am Sadly Theoretical’ (Catholic Herald), 234–​35, 236, 445–​46

Listener exchange with Hare & Nowell-​ Smith, 229, 262–​67 and Moral Sciences Club, 247–​48 as ‘outsider’, 444–​45 and PAX Society, 236–​42, 245–​47 and Peter Geach, 2, 3, 240, 240–​41n.60, 242, 250 review of Summa Theologiae translation, 461–​62 satire in, 93–​94, 258–​60 and social justice, 231–​36 and Socratic Club, 247–​48, 252 and Somerville College, Oxford, 3, 229n.24, 230n.26, 247–​48, 452–​53 writing style and standpoint, 95, 226n.6 youth, 2, 231–​36 Anscombe and Wittgenstein, 247 An Introduction to Wittgenstein’s Tractatus, 418–​19 as literary executor, 2, 415 as translator, 2, 415 understanding of, 416–​17 view of ‘method’, 420–​21 Anscombianly-​preferred examples (agent or patient [unmediated] conceptions), 361–​62, 374–​76, 378–​80 Anselm, 458–​59 answerability. See responsibility Aquinas, 66, 67, 94, 213–​14, 238, 244, 476–​78. See also ‘Thomism’ and action, 451–​52, 459–​61 Commentary on the Nicomachean Ethics, 125–​26, 272–​73 on connaturality, 124–​25, 219–​20, 293 and consequences, 453

510   Index Aquinas (cont.) and double effect, 453 and ethics, 459–​61 influence on Anscombe, 55, 443–​44 practical vs. theoretical knowledge, 459, 503–​6 on self-​defence, 173–​74, 179–​80, 188, 195, 275, 453 on suicide, 285 Summa Theologiae, 59–​60, 443, 451–​52, 456, 460 synderesis, 214–​15 translations of, 448–​50 Aristotelian naturalism. See ethical naturalism Aristotelian necessity, 198–​200, 203–​4, 304–​5, 310, 315 Aristotle, 12–​13n.11, 14, 22, 53–​54, 109, 113, 120–​ 21, 123, 260, 272–​73, 316–​17 De sophistarum elenchis, 102–​3 Ethics, 423 influence on Anscombe, 443, 495 Metaphysics, 66–​67 Nicomachean Ethics, 12–​13, 12–​13n.11, 27, 57, 120–​21, 127, 128, 133, 272 Topics, 102–​3 and trolls, 102–​3 Attwater, Donald, 237, 240 Augustine, 93, 238, 277, 398–​99, 492 authority, 143–​44 consideration zero, 317–​18 parental (see upbringing) political, 94–​95, 279, 304, 317–​18 Ayer, A J, 446 B Barnes, Jonathan, 272–​73 Barney, Rachel, 102–​3 Bennett, Jonathan, 100 Blackfriars Dominican Priory, Oxford, 232–​33, 237, 242, 445–​46. See also Dominican Order of Preachers Blackfriars (journal), 233, 236–​37, 240, 450 Carpenter, Fr Hilary OP (Prior), 233, 246–​47 style of Catholicism, 233 and UK Catholic peace movement, 237 Boedder, Fr Bernard SJ, 2, 231, 446–​47 Natural Theology, 2, 231, 445, 446–​47, 456–​57

bombing of Hiroshima and Nagasaki, 250, 251, 252–​53, 255, 261, 264–​65, 266–​67. See also Truman Bouwsma, O K, 248–​49, 248n.99 Bratman, Michael, 33–​34 Brewer, Talbot, 301–​2 bruteness, 149, 151, 153, 199 brute-​relative relation, 151–​52, 153, 154 Bullock, Alan, 251, 261 Butler, Bishop Joseph, 94–​95, 474–​75 C Calvin, John, 115 Campion Hall Jesuit Community, Oxford, 232–​33 Camus, Albert, 285 capital punishment, 274, 276 Carlyle, Thomas, 93, 115 Cartesian Ego, 393–​94, 397, 405, 406, 422. See also ‘I’ ‘stretch of ’, 393–​94, 397, 399, 400 Cartesianly-​preferred examples, 361–​62, 371, 375, 377–​79 Cartesian psychology, 452–​53 Catholicism Catholic Herald, 445–​46 Catholic Social Guild, 448 English, 232, 239–​40 causation, 28, 54, 428–​29, 445, 451–​52, 456–​58, 504. See also First Cause argument agent causation, 504–​6 efficient causes, 456, 497–​98, 499, 503, 504–​6 formal, 504–​6 human being as cause, 27, 28 and responsibility, 55–​58 character, 123, 124, 125–​27, 133 responsibility for, 123 children, 68 abuse of (see sexual abuse of children) actions of, 24–​25 raising of (see upbringing) Clarke, Samuel, 143n.12 Closeness Problem, 179–​84. See also consequences; double effect Collingwood, Robin, 113–​14 concept horse, the, 346, 350–​51

Index   511 connatural knowledge, 23n.25, 122, 124–​25, 126–​27, 130, 133, 135, 292–​94 defined, 293 and dignity of human nature, 294 conscience, 69–​70, 94–​95, 197, 213 consciousness. See self-​consciousness consequences, 183–​84 and circumstances, 183–​84 intended, 43–​44 intended vs. foreseen, 173–​78, 179, 184, 352, 476 (see also double effect) consequentialism, 177, 179, 227, 266, 431–​32, 474–​75, 476, 477, 484–​85 Anscombe-​consequentialism, 98–​99, 474–​75 Anscombe’s first use of term, 265–​66, 266n.169 as casuistry, 103 conception of the good under, 319 self-​effacement of, 106 standard definition, 98–​99 universality, 104–​5 constitutivism, 120–​21, 122–​24, 136 contemplative knowledge. See theoretical knowledge contraception, 325–​29, 334–​35, 339–​40 control (over action), 53–​54 Coope, Christopher, 477, 484 Crisp, Roger, 97, 105, 113 criteria, 149, 153, 154–​59 incompleteness, 155 variety of, 157–​58 criterialism, 154–​66 defined, 153 leakage, 156 D Daniel, Norman, 241, 242, 450 D’Arcy, Father Martin, 445, 446, 447, 448, 452 The Nature of Belief, 445, 452 Davidson, Donald, 8n.2, 11n.7, 33–​34, 45–​46, 48–​49, 388, 483–​84, 504–​5, 506 on Intention, 464 Davis, Alison, 286 deeds /​doings distinction, 472–​74 defeasibility, 149–​50, 154, 163–​64 definite descriptions, 399–​400

deliberation. See practical reasoning demonstratives, 384–​86, 395, 400, 405–​7 reference, 40 vulnerability to reference failure, 386 deontic vs. aretaic terminology, 479–​80 deontic /​evaluative distinction, 483–​84 deontology, 482 Descartes, René, 58, 393, 398–​99, 406. See also Cartesian Ego Descombes, Vincent, 377, 379–​80 descriptivism, 37–​38 desire, 120–​21, 128, 145 habitual (see character) Diamond, Cora, 339–​40, 396 dignity, 276 Anscombe’s concept of, 319–​21 equality in, 279–​80 of human nature, 294, 295–​97, 431–​32 and human rights, 302 as mystical value, 298 and social practices, 303–​6 of spiritual nature, 204, 295–​97 theological attitudes to, 300 United Nations conception of, 320 vengeful killing and, 295–​96, 296n.11 direction of intention, 177n.11 divine command ethic, 97, 111–​12, 113, 115–​17, 478–​79, 482 divine law, 118–​19, 197, 198, 244, 430, 431, 441. See also law conception of ethics divine lawgiver, 112, 113, 227–​28, 357 doctrine of double effect. See double effect doing well. See acting well Dominican Order of Preachers, 232–​33, 237, 445, 447–​48, 450, 454. See also Blackfriars Priory Donagan, Alan, 33–​34 double effect, 244, 256, 452–​54 abuse of, 173–​74, 177, 178, 182 in Anscombe, 174–​78 area bombing example, 178, 180, 188–​89, 194 collateral vs. focal damage, 175 defined, 173–​75 instrumental vs. accidental harms, 175 metaphysical /​moral distinction, 192–​95 permissibility conditions /​excusability, 175

512   Index double effect (cont.) remoteness, 181 therapeutic abortion example, 174–​75, 188–​89, 194 uncertainty, 181 Doyle, James, 478–​79 Drury, Maurice, 437, 450–​51, 454 duty, 306, 344–​45 E emotions. See under psychological concepts empirical idealism /​pragmatism, 345–​46 English moral philosophers (since Sidgwick). See Oxford moral philosophy error, 498 in action, 62 of judgement, 76–​78, 79, 81 in performance, 76–​77, 79, 82 (see also Theophrastus’ principle) practical, 76–​79 ethical naturalism, 117, 119–​20, 121, 122 euthanasia Anscombe’s opposition to legalisation, 271 as murder, 271–​72, 277–​82 and religion, 282–​85 Euthyphro dilemma, 116–​17 Evans, Gareth, 390–​91, 407 F fact-​value distinction, 3, 142–​46, 153, 154 faith, 118–​19, 436, 437 Finnis, John, 183–​84 First Cause argument, 445, 456–​57 first person. See ‘I’ ‘The First Person’, 383, 397 Fitzpatrick, William, 190 Flip-​Flop, 146–​48, 149–​53. See also Hume, David flourishing (human). See good life Foot, Philippa, 2, 3, 97, 247–​48, 310, 348, 356, 357, 474, 479–​80 aviary example, 204 and double effect, 180 Foot’s Fork, 143–​44 Natural Goodness, 121, 122 as neo-​Aristotelian, 121, 122 ‘rude’ example, 149, 158–​59

as secular virtue ethicist, 116 trolleyology, 481–​82, 482n.34 Francis (Pope), 337–​39 Frege, Gottlob, 159–​61, 389–​90 Art des Gegebenseins, 386–​87, 388, 395, 399 Frost, Kim, 500–​1 G Gallie, W B, 158n.40 Geach, Mary, 2, 93–​94, 96, 226n.5, 443, 477, 479–​80 Geach, Peter, 2, 97, 240, 448–​49, 455, 461–​62, 483 ‘Good and Evil’, 256n.136, 477–​78 and Oxford Jacobites, 240–​41n.60 generic ends, 134 Gill, Eric, 237, 240 golden rule, the, 281 Goldie, Peter, 100 good, the, 296 common good, the, 202–​4, 205 consequentialist view of, 319 Nietzschean view of, 319 primitive hedonist view of, 318 totalitarian view of, 319 transcendental role of, 314 understood as ‘good life’, 316–​17 Goodhart, Arthur, 252–​53 good life, 7–​8, 13, 16–​17, 22–​23, 125, 129, 201–​2, 311–​12, 316–​17 transcendental role of, 316–​18 grammar, 138–​42, 145, 157, 492, 493 grenade example, 188 grocer example, 199. See also bruteness guilt, 58 H habituation, 134 habitus, 124–​25, 126 Haddock, Adrian, 82 Hampshire, Stuart, 97 ‘double aspect’, 80–​82, 83 haplos-​pos distinction, 163 happiness, 120. See also good life Hare, R M, 96n.16, 107–​9, 110–​12, 114–​15, 229, 257–​58, 265–​66, 481–​82 Harris, John, 100

Index   513 Harvey, A E (BBC), 230, 258 Hauerwas, Stanley, 339–​40 Hebrew-​Christian ethic. See Judaeo-​Christian ethic Hegel, Georg Wilhelm Friedrich, 469 Heidegger, Martin, 285 hekousion, to,. See voluntary, the Hesperus /​Phosphorus example, 388 Hinsley, Cardinal, 236–​37n.44, 239–​40 homosexuality, 327–​29, 334–​35, 336–​37, 341–​42 Hornsby, Jennifer, 504–​6 How? 184–​85 human nature, 295–​96 religious outlooks /​theological attitudes, 300 spiritual, 296 human rights. See under rights human value. See under value Hume, David, 140n.3, 146–​54, 227–​28, 310–​11, 316–​17, 474–​75. See also Flip-​Flop Hume’s circle, 313 Hume’s triad, 144–​45 ‘Hume Reconsidered’, 456n.39, 457–​58 Hursthouse, Rosalind, 77–​78, 473, 474, 477, 484 I ‘I’, 361, 367, 375–​76, 422 guaranteed reference, 373, 385, 386, 402, 404, 405, 406, 407 misidentification, 373, 384–​85 no-​reference thesis, 383, 385, 397–​401, 422 reference failure, 386 identity. See personal identity ignorance, 54, 62, 66–​67, 68–​69, 460–​61 affected /​politic, 135 blameworthiness of, 54, 64–​65, 66–​69 distinct from nescience, 67 as exoneration or excuse, 62–​65 of facts, 64–​65 invincible, 68–​69, 135–​36 of law, 64–​65 voluntary, 55, 66–​67, 68–​69 imperatives, categorical /​hypothetical, 120, 124 imputability, 54, 62 inclination, 293–​94. See also character indexicals, pure /​automatic, 398, 400, 407 ‘here’ and ‘now’, 400, 401, 407

indifferent /​non-​indifferent truth, 292–​94, 298 innate moral principles, 215 innocence, 97, 256, 274–​76 intellectual virtue, 68–​69 intention, 10, 459–​61 broad vs. narrow, 184–​87, 188–​89 Cartesian conception of, 58, 177 expression of intention, 33, 72–​73, 74–​75, 366–​67 intentionally acting, 33, 44–​46 intention for the future (see intention with which) intention with which, 33, 35–​44, 46–​50, 184–​85 personal identity and (see personal identity) ‘pure’ intention, 48–​50 tripartite division, 33, 50–​51 unification of, 33–​35, 50–​51 Intention, 54, 58, 263, 494 methodology cf. Wittgenstein, 490–​92, 493, 494–​95 introspection, 362–​63, 377, 493–​94, 499 involuntary, the, 62, 62n.25 involuntary acts, 497–​98 Irwin, Terence, 134 is-​ought /​is-​owes, 146, 154. See also fact-​value distinction J James, William Baldy example, 378–​79, 380 Jesuits, 232–​33, 445, 447–​48 Jesus Christ, summary of the law, 97, 116 Joseph, H W B, 469, 472–​73 Judaeo-​Christian ethic, 96–​97, 176, 283, 431–​32 judgement, 77–​78. See also practical judgement judicial execution of the innocent, 227n.8 Jumpers (Tom Stoppard play), 419 justice, 16, 122, 201, 225, 228–​29, 234–​35, 260 and knowledge of human dignity, 294 natural (see natural justice) and utilitarianism, 476–​77 just war tradition, 175–​76, 236–​42, 243, 244, 450 ‘The Justice of the Present War Examined’, 3, 242–​47, 450

514   Index K Kant, Immanuel, 58, 84, 143n.12, 267, 319–​20, 330–​31, 474–​75, 479 Kantian autonomy, 197, 213 norms of practical reasoning, 310 Keane, Fr Henry SJ, 446–​47 A Primer of Moral Philosophy, 446–​47, 447n.9 Kehoe, Fr Richard OP, 445–​46, 448, 452 Kenny, Anthony, 454–​59 correspondence with Anscombe, 455–​58 Keown, John, 286–​87 killing. See also murder of the innocent, 176, 274–​76 by omission, 272 vengeful, 227–​28, 276, 295–​96, 296n.11 kinaesthetic knowledge, 495–​97 knowledge, 54, 59–​62. See also practical knowledge, theoretical knowledge conditions, 82–​83 of the dignity of human nature, 292–​93, 295, 296 evidential model, 75–​76 first-​person authority model, 76–​77 of indifferent truth, 292–​93 ‘in intention’, 74–​75 as justified true belief, 490–​91 non-​observational, 74–​75, 361, 363–​65, 367–​ 68, 490, 495–​501, 502 testimonial (see testimony) without observation (see non-​observational knowledge) ‘without realisation’, 82 Korsgaard, Christine, 120, 121, 122–​24, 127, 131–​ 32, 134, 136, 321 Kripke, Saul, 399 L language, 155, 157–​58, 161, 331–​32, 427 alleged defects of, 157, 159–​66 and human life, 138–​40, 157 and intention, 366–​67 and knowledge of human dignity, 296–​97 learning of, 427 vagueness, 163–​64 language-​games, 59, 140–​42, 160, 161–​63, 492, 493, 495

diversity /​multiplicity of, 138–​40, 165–​66 as life-​games, 157, 166–​70 penumbra of predicates, 156, 165–​66 point(s) of, 166–​70 law conception of ethics, 118–​19, 197, 227–​28, 267, 430, 478–​79 law of Excluded Middle, 162, 163–​64 Lawrence, D H, 339–​42 law without lawgiver, 197 length (concept), 348–​49 Leo XIII (Pope), encyclicals, 447–​48, 447n.11 Lewis, Clive Staples, 451–​52 life as gift, 282–​83, 300, 326 linguistic idealism, 346 defined, 346 and ethics, 344, 355, 356 Listener, The, Anscombe’s exchange in, 229, 259, 262–​63 living well. See good life, the Locke, John, 405–​6 logical positivism, 418–​19 M MacIntyre, Alasdair, 112, 252, 474 MacKinnon, Donald, 241, 446 Mahoney, Canon E J, 239 Marcus, Eric, 37–​39 marriage, 63, 329–​33, 341 Martin, Fr James SJ, 336–​38 masturbation, 327–​29, 334–​35, 336, 341–​42 McCabe, Fr Herbert OP, 454–​55, 462 McDowell, John, 37–​38, 77–​78, 119–​20, 126, 348, 388, 389, 398, 474 meaning, 138–​40, 162, 421–​22, 427, 428 fixity, lack of, 164–​66 meaning theory, 388–​89 means and ends, 14, 21–​22, 40, 174–​75, 177, 178, 182–​83 Mele, Alfred 44, 45 mental concepts. See psychological concepts mental images. See under psychological concepts mental verbs. See psychological verbs Midgley, Mary, 3, 247–​48, 356–​57 Mill, John Stuart, 98n.18, 228n.18, 471–​72, 474–​75, 476–​77 mistakes, typology of, 366–​67, 368–​70, 371–​72

Index   515 modality, 309–​10, 344–​45, 351–​52, 353, 354–​55 obligation and corresponding right, 315 physical, 310 practice-​grounding (goods), 310–​12 practice-​internal (rules), 308–​10 modals, 138–​40, 309, 344–​45, 351–​53, 354–​55, 357. See also rules; rights; promises ‘crossing of ’, 315 deontic, 354–​55 forcing, 309 stopping, 304, 309, 312–​13, 352–​53, 357 ‘Modern Moral Philosophy’, 91–​92, 118–​19, 239, 268, 344–​45, 426, 430–​33, 474–​80 Anscombe-​consequentialism, 228, 475, 475n.13 Anscombe’s ‘complaint’, 91–​92, 226–​27, 431–​32, 469 pedagogical force of, 106 recovery of concept ‘murder’, 225–​26, 229, 236–​47, 250, 254 three theses, 91–​92, 109–​17, 197, 226–​29, 344, 475–​76, 478, 480 Moore, Gareth, 326 Moore, G E, 69–​70, 97, 98–​99, 142–​43, 469, 475 moral absolutism, 97–​98, 175, 194, 357 moral consciousness, 210–​13 moral obligation, 113, 114–​15, 143–​44, 196, 197, 198, 210–​12, 344–​45, 478 moral particularism, 478n.22 moral philosophy, 106–​7, 470 before Anscombe, 469–​74 English (since Sidgwick) (see Oxford moral philosophy) metaethics, 481 normative ethics (see normative ethics) ‘ought-​first’, 107–​9 in Oxford (see Oxford moral philosophy) practical /​applied ethics, 481 moral psychology, 469, 481 of the virtues (Anscombe), 124–​27 (Aristotle), 123 moral rationality, 205–​9 moral responsibility. See responsibility Moran, Richard, 34 Müller, Anselm, 226n.6, 296 murder, 64–​65, 295–​96, 297–​98, 431–​32 absolute prohibition on, 225

Anscombe’s preoccupation with, 229, 269 capital punishment as, 274, 276 definition, 272–​74 in English Law (Coke), 279 equality in dignity, 278, 279–​80, 283, 288 euthanasia as, 271–​72, 277–​82 of the innocent, 274–​76 as involving complex action-​description, 229, 273–​74 Nicomachean Ethics, 272–​73 as private action, 275, 276, 278–​79 reasons for prohibition, 281 and religion, 282–​85 as rights violation, 281–​82, 303–​4, 314 and self-​defence, 275 by the state, 225–​26 and suicide, 277–​78, 280–​81, 282–​85 suicide prevention, 285–​88 and wronging, 229–​30, 272–​73, 278, 303–​4, 314 Murdoch, Iris, 3, 97, 247–​48, 357 mystical perception, 297–​98, 329, 336 mystical value. See under value N naturalism. See ethical naturalism natural justice, 301–​6 Natural Law, 213, 214–​20 doubts about, 215–​16 incompatibilities (Anscombe–​Aquinas), 217–​18 New Natural Law Theory, 183n.23, 194, 195 synderesis, 214–​15 syndromic conception, 219–​20 virtues of, 216–​17 natural theology, 434, 454–​59 necessity, 114–​15, 143–​44, 351–​52. See also modality Aristotelian (see Aristotelian necessity) ‘fantasies of ’, 428–​29 logical necessity, 428–​29 Nietzsche, Friedrich, 93, 110, 433 Twilight of the Idols, 110 non-​language users, 365–​67 non-​observational knowledge. See under knowledge non-​voluntary concepts, 495 normative ethics, 107–​8, 480–​85

516   Index normativism, 37–​39 normativity, 355 deontic, 301–​2 norms, 121 relating to human nature, 354–​55, 357 without content, 198 Nowell-​Smith, Patrick, 97, 100, 229, 259, 263–​65 nurse example, 72 Nykänen, H, 438–​39 O obedience, 117, 213 obligation. See moral obligation O’Hea, Fr. Leo SJ, 235, 448 ‘ought’, 110–​14, 115–​16, 143–​44, 196, 198–​205, 344, 430. See also moral obligation chimerical law, 197–​98 etymology, 111, 115–​16n.30 Greek concept of practical necessity, 113 as historically conditioned, 110–​11 Natural Law account of (see Natural Law) practical /​theoretical distinction, 210–​11 subjectivist account, 69–​70, 144–​45, 150 super-​objectivist account, 144–​45, 150 Oxford linguistic school, 418–​19 ‘meaning is use’, 418–​19 ordinary language philosophy, 418–​19, 420 ‘Oxford moralists’. See Hare, R M, Nowell-​ Smith, P, Prichard, H A ‘Oxford Moral Philosophy: Does it Corrupt the Youth?’. See BBC radio broadcast Oxford moral philosophy (since Sidgwick), 96–​97, 227 Anscombe’s characterisations of, 228, 229–​30, 247–​50, 256, 258–​59 justification of injustice, 260–​62 justification of murder, 256–​60 Oxford Objectivism, 470, 475, 477 ‘Oxford realism’, 84 P paedophilia. See sexual abuse of children pain, 149–​50, 156n.36, 371–​72, 425–​26, 493, 497 ‘partial idealism’. See linguistic idealism participants’ perspective, 72–​73, 74, 75, 81, 83, 85–​86

PAX Society, 237 and conscientious objection, 238, 239 and papal social teaching, 239 peirastic, 92–​93 Pepler, Fr Conrad OP, 454 perception, 123, 126, 445 personal identity, 422 animalist view, 410 cerebral transplant case, 409 conjoined twins case, 410 and intention, 421–​22 symmetrical fission case, 408–​9 synchronic /​diachronic, 405–​6 personhood, 295 phenomenalism, 426–​27 Philosophical Enquiry Group, The, 454–​55 papers delivered by Anscombe, 454–​55 philosophy of action, 469, 480–​81 philosophy of psychology, 97–​98, 106, 107, 227, 228–​29, 425n.12, 469, 471, 480–​81 vs. philosophical psychology, 481 phronesis. See practical wisdom piety, 117 pigs’ tripes example, 38 Plato, 93, 159–​61, 260, 418–​19 classification of dialogues, 93 Crito, 99 Euthyphro, 117 Meno, 427–​28 Phaedo, 99 Republic, 146 and sham philosophers, 102–​3 polemic, 92–​96 dangers of, 95–​96 defined, 92 etymology, 92, 93n.4 hyperbole, 93–​95 (see also Butler, Bishop Joseph) ‘Modern Moral Philosophy’ as, 95 pot-​holer example, 180–​81, 183–​84, 188–​93, 194 practical error(s). See under error practical judgement, 36–​42, 78, 126–​27 chimerical nature of, 37–​41 reasoned nature of, 42 acting on grounds upon the instant, 42 singularity of, 40, 41 analogy to visual judgement, 41

Index   517 practical knowledge, 3–​4, 8–​12, 60–​62, 72–​73, 74–​76, 78, 121, 362–​67, 380–​82, 490, 500–​6 as action-​theoretical phenomenon, 8, 22 as cause of what it understands, 21, 60–​61, 128–​29, 459, 503, 504 defined, 8–​9 as ethical phenomenon, 8, 22 first person, 362, 364, 367 ‘in intention’, 11–​12 as knowledge of the good, 22–​23 as non-​observational knowledge, 38, 39, 490, 502 relationship with practical truth, 7–​8, 17–​21 self-​consciousness and, 380, 381 vs. theoretical knowledge, 9, 10, 60–​61, 503 third person, 362–​64 as undermined by mistake in performance, 9–​10, 11 practical necessity, 66–​67 practical reason, 12–​13, 12–​13n.11, 26–​27, 28, 140–​54 as capacity for knowledge, 79–​83 desire-​satisfaction vs. good-​delivery system, 146 practical reasoning, 36, 38, 40, 61–​62, 80–​81, 128–​29, 316, 356, 357, 365, 423 practical syllogism, 16–​17n.19, 35–​37, 38, 42, 460 practical truth, 12–​17, 121 as action-​theoretical phenomenon, 8, 16, 21, 22 Anscombe’s commitment to, 25–​29 as complex phenomenon, 14, 16, 19 defined, 13 as distinct from theoretical truth, 14 as ethical phenomenon, 8, 14, 16, 21–​22 making true, 18–​19, 24, 25–​27, 29, 129 praxistic truth, 24, 27 relationship with practical knowledge, 7–​8, 17–​21 and vice, 128–​32 practical wisdom, 69–​70, 131–​32, 216, 293–​94, 356 practice(s). See social practice(s) praxis, 24–​25, 53–​54 prediction(s), 72 previous owner condition, 73–​74 Prichard, H A, 84, 97, 145n.17, 469, 474, 483

principle of side-​effects, 179, 181, 453. See also double effect private ostensive definition, 421–​22, 427 prohairesis, 12–​13, 127 prohibitions, 173–​74, 176–​77, 309, 313, 316, 351–​53, 357 promising, 202–​3, 344–​45, 351–​53 proper names, 399, 401–​4 two names parable, 60–​61, 402–​4 property. See under rights psychological concepts, 492–​93, 495 duration, 492–​93 emotions, 364, 492–​93 imagination, 493 mental images, 492–​93 sensation, 492–​93 simultaneity, 492–​93 psychological verbs, 371, 491, 492, 493, 495 psychology, 491–​92 pump example, 59, 183, 185–​87, 189–​90, 363, 500 R rape, 324–​25, 334–​35 rational wolf (thought experiment), 119–​20 Raz, Joseph, 36 realism, 345, 346–​47 recognitional, 121 ‘stupid empiricist realism’, 347, 349, 350–​51 recognition, 368–​69, 370–​7 1 reference failure. See under ‘I’ referring term /​expression, 398–​400. See also definite descriptions; demonstratives; indexicals; proper names reflexives, indirect /​ordinary, 400–​1 religion, 435–​38 and collective belonging, 437 miracles, 434n.26, 435 mysteries, 434, 435 prophecy, 435 truth of, 435–​36 responsibility, 15, 53–​55, 69–​70, 123, 260–​61 Anscombe’s work on, 54 and causality, 55–​58 epistemic condition of, 53–​54, 62 (see also Aristotle) negative causal, 56–​57 three degrees of, 57–​58

518   Index reverence for human life, 294, 300 non-​theological view, 296n.14 Rhees, Rush, 2, 450–​51 rights, 119–​20, 344, 351–​53, 354, 355, 357 Anscombe’s conventionalism on, 304–​5, 306 basic, 313–​16 and dignity, 302–​6 and linguistic forms, 304 natural, 313–​14 permission, 309 privilege, 309, 313 of property, 308–​9, 310–​11, 312–​14, 316–​17 relativism, 304–​5 and social practices, 303–​6, 308, 310, 312–​13 social-​constructivist conception (see social constructivism) rope climber example, 190–​91, 195 Ross, W D, 97, 229–​30, 256–​57, 469–​72, 473, 475, 477–​78 Anscombe’s critique of, 470, 471 good /​right action distinction, 470–​72, 477–​78 The Right and the Good, 470–​7 1 rules, 156–​57, 162, 351–​53. See also modals vs. observed regularity, 352 Russell, Bertrand, 59, 93, 340, 341, 393 S same-​sex marriage, 329–​30, 338 Scanlon, Tim, 175 Schönborn, Cardinal Christoph, 337 Schopenhauer, Arthur, 285 scientia media, doctrine of, 445 self, the,. See ‘I’ self-​consciousness, 386–​87, 406 and action, 376–​80 differentiated from consciousness of oneself, 383–​84 and first person, 367–​76 as manifested by use of ‘I’, 361–​62, 384, 387 and self-​knowledge, 376–​78 self-​constitution, 120–​21, 123, 131–​36 self-​reference, 377 semantics, theoretic, 140–​42 sensory deprivation tank, 386, 406 sex group sex, 328–​29, 333 and intimacy, 334–​35

liberal Catholic attitudes, 336–​39 and love, 333 non-​Catholic views, 339–​42 and shame, 335–​36 sexual abuse of children, 324–​25, 328, 334 sexual ethics, 334–​36 Anscombean, 324–​34 common attitudes towards, 334–​36 shopping list example, 77, 78 Sidgwick, Henry, 96–​97, 98–​99, 227, 228, 469 Simon, Yves, 133 sin, 119 as behaviour against divine law, 118–​19 as behaviour against right reason, 118–​19 of omission, 67 Singer, Peter, 321 Smythies, Yorick, 451–​52, 454 social constructivism, 313–​14 metaphysical problems, 314 social practices, 187–​92, 303–​6, 308 definition, 310 features of, 310, 311 justification of, 312 Society of Jesus (SJ). See Jesuits Socrates, 69–​70, 99, 277 geometry example, 427–​29 Socratic Club, The (Oxford), 452, 455 sodomy, 101–​2, 327–​28 speculative knowledge. See theoretical knowledge St Aloysius Church, Oxford, 232–​33, 445–​46 Stone, Martin, 34 Strawson, Peter, 398 subject /​object dichotomy, 368–​74, 376, 396, 421–​26 suicide, 277–​78, 280–​81, 282–​85 suicide prevention, 285–​88 see also euthanasia synderesis, 214–​15 T teaching and learning, 427–​28 Teichman, Jenny, 327 Teichmann, Roger, 68–​69 telos, 201–​2, 209, 308, 310, 311 temptation, 99 Tennant, F R, 452

Index   519 testimony /​testimonial knowledge, 73–​74 Theophrastus’ principle, 76–​78, 500–​1 theoretical (contemplative/​speculative) knowledge, 9, 60–​61, 72–​73, 77, 83–​86, 121, 423, 499 contrast with practical knowledge, 9, 10 defined, 9 as derived from the objects known, 11 of intentional actions of others, 9 of non-​intentional actions, 9 as undermined by mistake in judgement, 9, 10 theoretical reasoning, 423 theoretical truth, 14 Thomism, 67, 117, 447–​49, 460–​61, 462, 490–​91, 502, 503–​6 Thompson, Judith Jarvis, 175 Thompson, Michael, 302, 348 three theses. See under ‘Modern Moral Philosophy’ Tonti-​Filippini, Nicholas, 286 Torralba, José, 17–​18n.20 tragic dilemmas, 131 Truman, President Harry S, honorary degree, 3, 55–​56, 176, 229, 230, 247–​56, 263 Anscombe’s bet with Goodhart, 252–​53 Anscombe’s speech to Congregation, 250–​51 and Encaenia, 249, 253, 254 Hebdomadal Council, 229n.24, 249, 251, 256 and Intention, 226n.5 ‘Mr Truman’s Degree’ (pamphlet), 229, 230, 253 recovering the concept of murder, 254–​56 U Universal Declaration of Human Rights (United Nations), 320–​21 upbringing, 131, 132, 220, 293–​94, 300, 314–​16 ‘The Moral Environment of the Child’, 132–​36, 479 utilitarianism, 476. See also consequentialism V value, 145 of human beings, 295, 298, 299, 301–​3 instrumental, 298 mystical, 204–​5, 297–​300

van Inwagen, Peter, 397 Vann, Fr Gerald OP, 237–​38n.46, 240, 450 Morals Makyth Man, 450 verificationism, 418–​19, 419n.6 vice(s), 120, 123, 127, 135 and practical truth, 128–​32 vicious action, 18, 20 virtue, 122, 125, 200, 201, 228–​29, 480 Aristotelian mean, 272–​73 supra-​utilitarian (mystical) /​utilitarian distinction, 298–​99 virtue ethics, 474, 482, 483–​85 virtues, 123, 200–​1 Anscombe’s moral psychology of the, 124–​27 of the good man, 200–​1, 202, 205 volitionism, 474 Voltaire, 93, 284 voluntary, the, 53–​54, 185, 186 voluntary acts, 129, 184–​85 W Watkin, E I, 237, 240 White, Fr. Victor OP (‘Penguin’), 237–​38n.46, 240, 241, 450 ‘Who?’, 373 ‘Why?’, 72, 138–​40, 144–​45, 184–​85, 352, 363, 367–​68, 424, 494, 495, 502 Wicker, Brian, 233 will, 55, 66, 68–​69, 123, 125, 126, 127 Williams, Bernard, 97, 100, 234, 288, 335–​36, 357 Williams, Glanville, 100, 269, 452–​53 Wiseman, Rachael, 69–​70, 226n.4 Intention, 54 Wittgenstein, Ludwig, 2, 157, 247, 249, 471–​72 concepts and human life, 346, 348–​49, 350–​51 I as subject/​object, 397, 421–​26 beyond, 372–​74 ‘object’ use of ‘I’, 368–​70 ‘subject’ use of ‘I’, 370–​72 influence on Anscombe, 226–​27, 416–​18, 421–​22, 426–​27, 428–​29, 430, 436, 443, 449, 450–​51 and linguistic idealism, 345–​51, 358, 429–​30 meaning, 138–​40

520   Index Wittgenstein, Ludwig (cont.) ‘On Certainty’, 152, 437–​38 on method, 418–​21 pain (see pain) Philosophical Investigations, 355, 415, 417, 423 plan for the treatment of psychological concepts, 491–​92, 494 and religion, 434–​35, 436–​37 rule-​governed activities, 312, 355 on self-​understanding, 471–​72 on suicide, 285

Tractatus, 417, 418–​19 use of ‘know’, 497 wood seller example, 346–​47, 349–​50, 356–​57 World War II, justice of, 3, 175–​76 ‘The Justice of the Present War Examined’, 3, 242–​47, 450 Wright, G H von, 26–​27 Z Ziff, Sid /​Dorothy Parker, 102n.22 Zylberman, Ariel, 301–​2