Free Will and the Rebel Angels in Medieval Philosophy 9781107155381, 110715538X

In this book Tobias Hoffmann studies the medieval free will debate during its liveliest period, from the 1220s to the 13

349 55 4MB

English Pages 320 [314] Year 2020

Report DMCA / Copyright

DOWNLOAD FILE

Polecaj historie

Free Will and the Rebel Angels in Medieval Philosophy
 9781107155381, 110715538X

Citation preview

F R E E W I L L A N D TH E R E BE L A N G E L S IN MEDIEVAL PHILOSOPHY

In this book Tobias Hoffmann studies the medieval free will debate during its liveliest period, from the 1220s to the 1320s, and clarifies its background in Aristotle, Augustine, and earlier medieval thinkers. Among the wide range of authors he examines are not only wellknown thinkers such as Thomas Aquinas, Duns Scotus, and William of Ockham, but also a number of authors who were just as important in their time and deserve to be rediscovered today. To shed further light on their theories of free will, Hoffmann also explores their competing philosophical explanations of the fall of the angels, that is, the hypothesis of an evil choice made by rational beings under optimal psychological conditions. As he shows, this test case imposed limits on tracing free choices to cognition. His book provides a comprehensive account of a debate that was central to medieval philosophy and continues to occupy philosophers today. tobias hoffmann is Professor of Philosophy at the Catholic University of America. He has published widely on medieval thinkers including Albert the Great, Thomas Aquinas, Henry of Ghent, and Duns Scotus, and he is the editor of Weakness of Will from Plato to the Present (2008), A Companion to Angels in Medieval Philosophy (2012), and co-editor (with Jörn Müller and Matthias Perkams) of Aquinas and the Nicomachean Ethics (Cambridge University Press, 2013).

FREE WILL AND THE REBEL ANGELS IN MEDIEVAL PHILOSOPHY TOBIAS HOFFMANN The Catholic University of America

University Printing House, Cambridge cb2 8bs, United Kingdom One Liberty Plaza, 20th Floor, New York, ny 10006, USA 477 Williamstown Road, Port Melbourne, vic 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 79 Anson Road, #06–04/06, Singapore 079906 Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning, and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781107155381 doi: 10.1017/9781316652886 © Tobias Hoffmann 2021 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2021 A catalogue record for this publication is available from the British Library. isbn 978-1-107-15538-1 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

F R E E W I L L A N D TH E R E BE L A N G E L S IN MEDIEVAL PHILOSOPHY

In this book Tobias Hoffmann studies the medieval free will debate during its liveliest period, from the 1220s to the 1320s, and clarifies its background in Aristotle, Augustine, and earlier medieval thinkers. Among the wide range of authors he examines are not only wellknown thinkers such as Thomas Aquinas, Duns Scotus, and William of Ockham, but also a number of authors who were just as important in their time and deserve to be rediscovered today. To shed further light on their theories of free will, Hoffmann also explores their competing philosophical explanations of the fall of the angels, that is, the hypothesis of an evil choice made by rational beings under optimal psychological conditions. As he shows, this test case imposed limits on tracing free choices to cognition. His book provides a comprehensive account of a debate that was central to medieval philosophy and continues to occupy philosophers today. tobias hoffmann is Professor of Philosophy at the Catholic University of America. He has published widely on medieval thinkers including Albert the Great, Thomas Aquinas, Henry of Ghent, and Duns Scotus, and he is the editor of Weakness of Will from Plato to the Present (2008), A Companion to Angels in Medieval Philosophy (2012), and co-editor (with Jörn Müller and Matthias Perkams) of Aquinas and the Nicomachean Ethics (Cambridge University Press, 2013).

FREE WILL AND THE REBEL ANGELS IN MEDIEVAL PHILOSOPHY TOBIAS HOFFMANN The Catholic University of America

University Printing House, Cambridge cb2 8bs, United Kingdom One Liberty Plaza, 20th Floor, New York, ny 10006, USA 477 Williamstown Road, Port Melbourne, vic 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 79 Anson Road, #06–04/06, Singapore 079906 Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning, and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781107155381 doi: 10.1017/9781316652886 © Tobias Hoffmann 2021 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2021 A catalogue record for this publication is available from the British Library. isbn 978-1-107-15538-1 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

F R E E W I L L A N D TH E R E BE L A N G E L S IN MEDIEVAL PHILOSOPHY

In this book Tobias Hoffmann studies the medieval free will debate during its liveliest period, from the 1220s to the 1320s, and clarifies its background in Aristotle, Augustine, and earlier medieval thinkers. Among the wide range of authors he examines are not only wellknown thinkers such as Thomas Aquinas, Duns Scotus, and William of Ockham, but also a number of authors who were just as important in their time and deserve to be rediscovered today. To shed further light on their theories of free will, Hoffmann also explores their competing philosophical explanations of the fall of the angels, that is, the hypothesis of an evil choice made by rational beings under optimal psychological conditions. As he shows, this test case imposed limits on tracing free choices to cognition. His book provides a comprehensive account of a debate that was central to medieval philosophy and continues to occupy philosophers today. tobias hoffmann is Professor of Philosophy at the Catholic University of America. He has published widely on medieval thinkers including Albert the Great, Thomas Aquinas, Henry of Ghent, and Duns Scotus, and he is the editor of Weakness of Will from Plato to the Present (2008), A Companion to Angels in Medieval Philosophy (2012), and co-editor (with Jörn Müller and Matthias Perkams) of Aquinas and the Nicomachean Ethics (Cambridge University Press, 2013).

FREE WILL AND THE REBEL ANGELS IN MEDIEVAL PHILOSOPHY TOBIAS HOFFMANN The Catholic University of America

University Printing House, Cambridge cb2 8bs, United Kingdom One Liberty Plaza, 20th Floor, New York, ny 10006, USA 477 Williamstown Road, Port Melbourne, vic 3207, Australia 314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre, New Delhi – 110025, India 79 Anson Road, #06–04/06, Singapore 079906 Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning, and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781107155381 doi: 10.1017/9781316652886 © Tobias Hoffmann 2021 This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2021 A catalogue record for this publication is available from the British Library. isbn 978-1-107-15538-1 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of URLs for external or third-party internet websites referred to in this publication and does not guarantee that any content on such websites is, or will remain, accurate or appropriate.

For Bonnie Kent

Contents

Acknowledgments List of Abbreviations Citation Method

page x xi xiv

Introduction

1

The Scope of This Study Free Will in a Broad and in a Narrow Sense Intellectualism and Voluntarism The Narrative of Angelic Sin

1 4 5 6

part i free will 1 Free Will with and without Aristotle

13

Anselm of Canterbury Bernard of Clairvaux Peter Lombard Aristotle as a Game Changer

13 18 20 22

1.1 1.2 1.3 1.4

2 The Psychological Turn and the Rise of Intellectualism 2.1 2.2 2.3 2.4 2.5

William of Auxerre Philip the Chancellor The Summa Halensis, Albert the Great, and Bonaventure Thomas Aquinas Siger of Brabant

3 Voluntarism and the Condemnation of Intellectualism 3.1 3.2 3.3 3.4

The 1270 and 1277 Censures Henry of Ghent Peter Olivi Other Franciscans and the Correctorium of Brother Thomas

4 Intermediary Theories and Strict Intellectualism 4.1 Giles of Rome

31 32 33 35 40 54

58 59 63 84 91

93 94

vii

Contents

viii 4.2 John of Morrovalle 4.3 Godfrey of Fontaines

5 Refinements and Radicalizations 5.1 5.2 5.3 5.4 5.5

John Duns Scotus John of Pouilly Hervaeus Natalis and Durand of St. Pourçain Peter Auriol William of Ockham

103 105

119 119 133 143 146 151

part ii whence evil? 6 Does Evil Have a Cause?

163

6.1 Augustine 6.2 Pseudo-Dionysius

163 171

7 The Will as the Cause of Evil 7.1 7.2 7.3 7.4 7.5 7.6 7.7 7.8 7.9

Anselm of Canterbury Peter Lombard William of Auxerre Albert the Great and Bonaventure Thomas Aquinas Henry of Ghent Richard of Menneville (Mediavilla) John Duns Scotus John of Pouilly

174 175 175 176 178 182 187 188 190 193

part iii angelic sin 8 Intellectualist Accounts of the Angelic Fall 8.1 8.2 8.3 8.4 8.5

Philip the Chancellor Albert the Great Thomas Aquinas Godfrey of Fontaines John of Pouilly

9 Voluntarist and Intermediary Accounts of the Angelic Fall 9.1 9.2 9.3 9.4 9.5 9.6 9.7

Bonaventure John Pecham Henry of Ghent Peter Olivi John Duns Scotus Giles of Rome Peter Auriol

199 200 201 204 214 216

219 219 221 222 225 230 235 238

Contents 10 Necessary (and Free?) Obstinacy 10.1 10.2 10.3 10.4 10.5 10.6 10.7

Thomas Aquinas Peter Olivi Henry of Ghent John Duns Scotus Durand of St. Pourçain Peter Auriol William of Ockham

ix 243 244 249 250 252 255 257 259

Conclusion

263

Bibliography Index of Manuscripts Index

267 287 288

Acknowledgments

Numerous people helped me in writing this book; while I wish to express my deep gratitude to all of them, I can mention only a few by name. While planning the book, I benefited from conversations with Bonnie Kent during my time as a visiting scholar at the University of California at Irvine in 2012–13. Hilary Gaskin encouraged me to pursue the project with Cambridge University Press and gave me precious advice about the scope of the book. An anonymous reader provided invaluable comments on the whole manuscript, and two readers did so for early drafts of some chapters; the book has gained much from their suggestions. While a visiting scholar at the University of Macerata in 2017–18, I greatly profited from the generous hospitality of the Philosophy Department and especially from conversations with Guido Alliney. I had an analogous experience at the University of Würzburg, where Jörn Müller hosted me during the summer of 2018. Throughout the entire process, Stephen Dumont, Timothy Noone, Riccardo Saccenti, Chris Schabel, and Kevin White provided precious advice and shared reproductions of manuscripts or unpublished papers of theirs. Thomas Jeschke graciously shared a draft of a critical edition. Olivier Boulnois provided the opportunity to test the core of the book by inviting me to teach a course on the same topic at the École Pratique des Hautes Études in Paris in May 2019. I received useful comments on earlier partial or complete drafts of the manuscript from Francis and Sophia Feingold, Peter Furlong, Herb Hartmann, and Logan Weir. It was a pleasure to work with Anjana Karikal Cholan and her team, who very competently produced the typescript. The Alexander von Humboldt Foundation supported the project with an alumni research grant. I wish to dedicate the book to Bonnie Kent, in gratitude for her innovative work in medieval theories of moral psychology and ethics and for the selfless support she has given me since I have come to North America. x

Abbreviations

Book Series and Editions CCCM CCSL CSEL NR OPh OTh PhB PL Viv. WB

Corpus Christianorum Continuatio Mediaevalis Corpus Christianorum Series Latina Corpus Scriptorum Ecclesiasticorum Latinorum John Duns Scotus (2007), Quodlibet q. 16, ed. Noone and Roberts Opera philosophica (of Duns Scotus or William of Ockham) Opera theologica (of William of Ockham) Les Philosophes Belges Patrologia Latina Vivès edition (of Duns Scotus) John Duns Scotus (2004–8), Reportatio I-A, ed. Wolter and Bychkov Aristotle

An. EN GC Int. Met. Phys.

De anima Ethica Nicomachea De generatione et corruptione De interpretatione Metaphysica Physica Augustine

ciu. conf. corrept. c. Iul.

De ciuitate Dei Confessiones De correptione et gratia Contra Iulianum xi

xii c. Iul. imp. ench. en. Ps. Gn. litt. lib. arb. mor. nat. b. nupt. et conc. perf. iust. retr. uera rel.

List of Abbreviations Contra Iulianum opus imperfectum Enchiridion ad Laurentium de fide spe et caritate Enarrationes in Psalmos De Genesi ad litteram De libero arbitrio De moribus ecclesiae catholicae et de moribus Manichaeorum De natura boni De nuptiis et concupiscentia De perfectione iustitiae hominis Retractationes De uera religione Works by Other Authors

DCD DDN DM DNCC DSS DV Expos. . . . In . . . Lect. Ord. Praedest. QD QO QV Q. . . Quodl. Rep. SA SCG Sent. ST I-II II-II

De casu diaboli De divinis nominibus Quaestiones disputatae de malo De necessitate et contingentia causarum De substantiis separatis Quaestiones disputatae de veritate Expositio . . . (e.g., in libros Physicorum Aristotelis) Commentary on . . . (e.g., In DDN, In Sent., In Met.) Lectura Ordinatio Tractatus de praedestinatione et de praescientia divina Quaestio(nes) disputata(e) Quaestiones ordinariae Quaestiones variae Quaestiones in . . . (e.g., QMet, Quaestiones in Metaphysicam Aristotelis) Quodlibet Reportatio Summa aurea Summa contra Gentiles Sententiae Summa theologiae Prima secundae (First part of the Second part) Secunda secundae (Second part of the Second part)

List of Abbreviations Other Abbreviations a. arg. (in opp.) c. concl. intr. q. qc. s.c. sol.

articulus argumentum (in oppositum) corpus articuli (e.g., 6.7c.) conclusio introductio quaestio quaestiuncula sed contra solutio Manuscripts

A Fb N P

Paris, Bibliothèque nationale de France, lat. 14565 Firenze, Biblioteca Nazionale Centrale, Conv. Soppr. A.III.120 Nürnberg, Stadtbibliothek, Cent. III, 75 Paris, Bibliothèque nationale de France, lat. 15372

xiii

Citation Method

Ancient and medieval texts are cited according to their internal divisions (such as Book–Chapter, or Book–Distinction–Question–Article), marking the subdivisions by periods. When useful, references to printed editions (which are listed in the Bibliography) are given by volume numbers (in Roman numerals), page numbers, and line numbers. For example, Bonaventure, In Sent. II.3.2.1.2 ad 2, II: 117a refers to Book II of his Sentences commentary, Distinction 3, Part 2, Article 1, Question 2, response to the second opening argument, Opera omnia, volume II, page 117, left column. When not otherwise specified, Albert the Great is cited according to the Editio Coloniensis (1951–), Thomas Aquinas according to the Editio Leonina (1882–), Henry of Ghent according to the Opera omnia (1979–), and Duns Scotus according to the Editio Vaticana (1950–). Quodlibeta by Henry of Ghent cited without Roman numerals are cited according to the edition by Badius, 1518. When multiple references are indicated, they are listed in chronological order. Because of space constraints, quotations in the footnotes had to be kept to a minimum. All translations are my own.

xiv

Introduction

The reintroduction of Aristotle into the Latin West in the twelfth and thirteenth centuries gave rise to the most innovative theories of free will in the later Middle Ages. The rediscovery of Aristotle’s action theory caused a psychological turn in reflection upon free will: the emphasis shifted from a narrow focus on the relation of free will and sin to the philosophical problem of how free agency is rooted in the powers of the soul. The existence of free will as the basis for the moral responsibility implied in sin was no longer taken for granted on theological grounds, but rather investigated within a philosophical account of action. The reception of Aristotle’s action theory by thinkers of the Latin tradition started in the 1220s, without at first causing any concern. But beginning in the 1260s, some thinkers who followed Aristotle more systematically, especially Thomas Aquinas and Siger of Brabant, prompted strong reactions. According to an Aristotelian action theory, our choices necessarily follow our judgments of what is worth choosing, and bad choices presuppose faulty judgments. But some theologians feared that such a theory threatens free will, for it seems that we do not control what appears to us as worth choosing. Also, moral fault or sin, consisting in a morally reprehensible act, is a voluntary failure; but a miscalculation of what is choiceworthy is an involuntary failure. Medieval thinkers continued to agree about the existence of free will, but began to debate passionately about the relations between cognition and volition, and between faulty cognition and evildoing.

The Scope of This Study The purpose of this book is to discuss free will theories in the first hundred years of the reception of Aristotle’s action theory by thinkers belonging to the Latin tradition. Islamic and Jewish medieval thinkers, and those of the Byzantine tradition, are beyond the scope of this book, so when I write “medieval,” I normally intend “Latin medieval.” For background, the book 1

2

Introduction

studies theoretical presuppositions of the later medieval accounts of free will in the thought of Aristotle, Augustine, Anselm of Canterbury, and others. But the core of the book concerns theories developed between the 1220s and the early 1320s, and thus focuses on some pioneers of the reception of Aristotle’s action theory (such as William of Auxerre, Philip the Chancellor, and Albert the Great), the “big names” in the free will debate (Thomas Aquinas, Henry of Ghent, Giles of Rome, Godfrey of Fontaines, Duns Scotus, and William of Ockham), and some thinkers who are less known today, but important in their time (such as John of Pouilly and Peter Auriol). After the psychological turn, the central issue became the relation between intellect and will. Particular attention was given to the relation between the cognition of a desirable object and its volition and that between cognitive and volitional deficiencies, that is, between error, ignorance, or nonconsideration and an evil will. The concern was to explain control and moral responsibility through a psychologically plausible and metaphysically robust theory. Aristotle became a resource for later medieval theories of free will not only regarding his action theory, but also his psychology, natural philosophy, and metaphysics. The questions Christian theologians had to address were more radical, however, than those tackled by Aristotle. While Aristotle holds that all evildoers are in some sense ignorant of the good, Christian theologians must explain how persons who do not suffer ignorance can do evil. While Aristotle explains that we can act against our better judgment because occurrent passions interfere with practical thinking, Christian theologians must explain that persons possessing unflawed knowledge and lacking passions can do evil. Whereas Aristotle holds that vicious dispositions (habitus), acquired perhaps because of lack of proper moral education, can distort our understanding of what is worth pursuing, Christian thinkers have to explain how persons without bad dispositions can do evil. Furthermore, Aristotle assumes that the universe has no temporal beginning, and so in his view, evil has always existed. Christian theologians, by contrast, assume that the universe has a temporal beginning, that originally all persons were good, but that some sinned while others did not. Finally, Aristotle holds that repeated bad action generates a persistent vice that makes the individual prone to act badly. Christian theologians assumed that a single act can make persons permanently unable to avoid sinning. According to Christian theology, these persons who do evil despite their optimal psychological conditions and who after a single act are beyond the possibility of moral reform are the rebel angels.

Introduction

3

This book aims at presenting the medieval free will debate according to the breadth of the thinkers’ own interests. For the medievals, a theory of free will must not only account for the ability to choose one’s acts, but also explain how an entirely good person can want to commit an evil act for the first time. It must explain this not only for human beings, but also for angels, who act under optimal psychological conditions. Part of the explanation must be an account of the person’s control of the acts – especially in the case of sin, for “sin” means a morally deficient act for which the agent is responsible. For medieval thinkers, sin is not the paradigm case of a free act, since ideally, free will involves the inability to sin. Nevertheless, in their view, the possibility of a good creature sinning for the first time, and the particular problem of angelic sin, are important test cases for the robustness of their theories of free will. The hypothesis of the first choice of the angels allows them to focus exclusively on the interaction of intellect and will, for it results only from thinking and willing some object. Thus this hypothesis prompted thinkers to refine their theories of free will, just as philosophers today employ thought experiments to test commonly accepted opinions and develop innovative ideas.1 Aristotle would have considered the hypothesis of a deficient act of purely intellectual beings absurd.2 Nevertheless, the various philosophical difficulties implied in the Christian understanding of the sin of the angels induced the medievals, with some delay, to treat Aristotle as an important source, leading them to take a more philosophical approach to angelic sin than before. For medieval theologians, angels were no side issue. Particular assumptions about angels were thought to have implications for wider philosophical and theological issues. This explains, for example, why in an important ecclesial condemnation in 1277, more than one-sixth of 219 censured philosophical propositions has to do with angels or “intelligences” (the philosophers’ term). Accordingly, in the last thirty years, scholars have been increasingly interested in philosophical issues raised within medieval discussions of angels, especially in natural philosophy, metaphysics, and theory of knowledge.3 By contrast, little attention has been given to medieval discussions of angelic sin. 1 2

3

Perler 2008 shows that many medieval discussions of angels have the character of thought experiments. According to Aristotle, separate substances are unchangeable (Met. XII.8, 1073a32–4). They are eternal and therefore fully actual and indefectible (Met. IX.9, 1051a19–21). Later medieval authors are aware that Aristotle considers the fall of separate substances impossible; see, e.g., William of Auvergne, De universo II.2.40, I: 884aE; Thomas Aquinas, ST I.63.1 arg. 1; Godfrey of Fontaines, Quodl. VI.7, III: 168. See especially Suarez-Nani 2002a and 2002b, Lenz and Iribarren eds. 2008, and Hoffmann ed. 2012.

4

Introduction

There are three questions in particular that traverse this book. First, do the medieval thinkers’ theoretical presuppositions allow them to explain that persons control their acts? Second, to what extent do they trace a freely made choice to its causes? In other words, do they consider a free choice to be ultimately explainable? Third, do they admit the possibility of “cleareyed evildoing,” that is, of doing evil in full awareness that it is evil, including awareness of the negative impact on oneself? I contend that to account for control, all medieval thinkers studied here hold, or at least concede implicitly, that freely chosen acts cannot be fully traced to cognition and are therefore explainable only up to a certain point. By the same token, an evil choice cannot be sufficiently explained by a cognitive disorder. Nevertheless, most authors think something goes wrong with cognition when someone sins, and so they do not allow for clear-eyed evildoing. To provide background, it is worth offering a few general philosophical and historical considerations concerning free will, followed by an account of the narrative of the fall of the angels that was presupposed in later medieval theories of angelic sin.

Free Will in a Broad and in a Narrow Sense So far, I have employed exclusively the term “free will,” which is also used, albeit somewhat differently, in contemporary philosophy. Medieval thinkers use the corresponding Latin term, libera voluntas, or libertas voluntatis, freedom of the will. But they also use the expression liberum arbitrium – free decision (literally “free adjudication”). Sometimes, they employ these expressions interchangeably, at other times they give them distinct meanings. Two important meanings stand out, depending on which conditions for free will are met. One condition is that an act have its source in the agent’s intellect and will. Borrowing from contemporary parlance, I call this the “sourcehood condition” of free will. Sourcehood condition: An act is free only if the agent is the source of the act through intellect and will.

This formulation leaves open how an act must originate in intellect and will to meet the sourcehood condition. Some medieval thinkers specify that the will must be the source of one’s act without being determined by the intellect. For others, cognitive determination is what alone makes willing possible. So for some, the sourcehood condition involves above all certain cognitive conditions (cognizing an object in a particular way), for others certain volitional conditions (the will moving itself in light of

Introduction

5

a cognized object). They agree, however, that the sourcehood condition involves absence of external coercion. They also agree that acts originating in the sensory appetite, such as unthinkingly scratching one’s beard, are not done by free will, even though they are done voluntarily, without coercion. Another condition is what contemporary philosophers call the “leeway condition.” Leeway condition: An act is free only if the agent has alternative possibilities and is the source of the alternative he or she chooses.

Medieval thinkers generally hold that free will in the broad sense requires only sourcehood and is thus compatible with the inability to do otherwise, while free will in the narrow sense requires, in addition to sourcehood, also the ability to choose among alternatives. They also widely hold that moral responsibility presupposes not only sourcehood, but also leeway.

Intellectualism and Voluntarism Once the medieval theologians’ interest shifts to the psychological foundation of free will, the question becomes for them whether sourcehood and leeway are grounded more in the intellect or in the will. Interpreters often use the labels “intellectualism” and “voluntarism” to distinguish between these approaches to free will. These labels run the risk of oversimplification, and some scholars have argued that they are to be avoided altogether. Yet as we will see, particularly in the debate after Aquinas, the difference of emphasis on either the intellect or the will as the ultimate foundation of free will is so pronounced that these different approaches deserve a name. Nevertheless, medieval theories do not fall into two monolithic camps. There is much variety, especially among voluntarist thinkers, but also among intellectualists. There are also a number of thinkers who develop intermediate positions. It shall suffice to offer broad definitions of intellectualism and voluntarism, which allow for further specification once the details of the corresponding views become clear. Intellectualism: the theory that explains free agency mainly with reference to the intellect. Voluntarism: the theory that explains free agency mainly with reference to the will.

Medieval thinkers consistently held that, properly speaking, it is not the intellect and will that act, but rather the person who acts by intellect and

6

Introduction

will. Nevertheless, once the emphasis is on the relative contributions of intellect and will in free agency, it becomes convenient to speak loosely as if the intellect knows, understands, considers, judges, and deliberates, and the will desires, chooses, and enjoys. The medievals often use this language, and I too will occasionally adopt this way of speaking in expounding their views.

The Narrative of Angelic Sin Sacred Scripture contains quite little about the sin of the angels, yet medieval theologians write rather profusely on it. Allegorical interpretations, validated by authoritative glosses by Church Fathers, expand the number of pertinent scriptural passages. Some difficulties in knowing specific aspects of the fall of the angels are bypassed when the question is not whether and how the fall actually happened, but whether and how it could happen. The fact of angelic sin is a theological question the study of which relies on revelation; its possibility – that is, whether its assumption is noncontradictory and coherent with other assumptions to which medieval thinkers are committed – is a philosophical problem and hence can be investigated apart from revelation. Finally, some theologians even provide empirical evidence for the existence of evil angels: idolatry, magic, demonic possession, and the experience of temptation. On the assumption that God created all things good, the existence of evil angels implies that some sinned.4 As a basis for their reflections about the sin of the angels, later medieval thinkers adopted for the most part the narrative that Peter Lombard formulated in his Sentences (composed in different redactions in the midtwelfth century).5 The Sentences, which mainly consist in a collection of authoritative statements by Church Fathers, were highly influential because by the 1240s they became the obligatory theology textbook. Becoming a master in theology required lecturing on the Sentences, and routinely, theologians published their revised lectures in works that have come to be called Sentences commentaries.6 4

5 6

See William of Auvergne, De universo II.2.39, I: 882bF–883bD; Albert the Great, In Sent. II.5.1 s.c. 1, Borgnet XXVII: 111a; Bonaventure, In Sent. II.6.2.1 arg. 2 in opp. and c., II: 164a–b; Peter Olivi, Summa II.43c., I: 716. For recent research concerning the composition and dating of the Sentences, see Clark 2019, 240–6. An expression not used by the medievals themselves, for their lectures were not commentaries on Lombard’s text, but rather discussions of its themes, often with no connection to Lombard; see Schabel 2020.

Introduction

7

While patristic authors disagreed concerning particular aspects of the angelic fall, Lombard proposed a coherent narrative that selects certain of the Fathers’ specific interpretations of Sacred Scripture, favoring above all Augustine’s. By the second half of the twelfth century, theologians widely accepted Lombard’s narrative,7 and they framed most of the questions they debated about the fall according to the themes covered in the Sentences. Lombard follows Augustine in reading the first verse of the Bible, “in the beginning God created the heavens and the earth,” to mean that the angels (“the heavens”) were created at the same time as the material universe (Sent. II.2.1.3–4, I: 336–7). They were created in a place: the highest heaven of the material universe, the empyreum, that is, the fiery heaven (Sent. II.2.4 nn. 2–3, I: 339–40). Lombard can thus understand the “fall” of the evil angels not only as their act of sin, but also as their consequent fall from heaven to earth, more precisely, “into this misty air” from which they now tempt humans.8 After the Last Judgment, the evil angels will be hurled into hell (Sent. II.6.2–3, I: 355–6). The angels have intellect and will as well as free decision (liberum arbitrium), but they were not all created equal, and hence they do not enjoy these powers to the same degree (Sent. II.3.2–3, I: 342–3). All angels were created good, and, at least for a short interval, all remained good (Sent. II.3.4, I: 343–7). In that interval, they did not enjoy complete beatitude – they did not yet enjoy the beatific vision of seeing God directly – but they did not experience any misery either (Sent. II.4, I: 348–51). With their first act of free decision, some adhered to God out of love and obtained the grace of being “confirmed” in the good, whereas others committed the sins of pride and envy and fell from heaven (Sent. II.3.4 n. 5, I: 344–5; II.5.1, I: 351). Their choice had a permanent effect: thanks to the grace of confirmation, the good angels became unable to sin, whereas the bad angels – who now “were made demons” – became obstinate, that is, permanently unable to have a good will (Sent. II.7.1, I: 359; II.7.3 n. 2, I: 360). Lucifer, the highest angel, is presented as the first angel to sin, while other angels consented to his sin and fell together with him (Sent. II.6.2 n. 1, I: 355). Lombard follows a patristic tradition in interpreting

7 8

See Colish 1995, 91. Lombard inherits the twofold meaning of the fall of the angels from Augustine; for the fall as the sin of the angels, see, e.g., Gn. litt. XI.16, CSEL XXVIII/1: 349 l. 2; ciu. XII.9, CCSL XLVIII: 364 l. 49; for the fall as the fall from heaven, see, e.g., Gn. litt. XI.13 and XI.14, p. 346 ll. 1–2 and l. 17; ciu. XI.14, p. 346 l. 16; ciu. XIV.11, p. 432 l. 56. See also lib. arb. II.20.54.205 and Gn. litt. IV.9, p. 105 l. 20, where “fall” denotes human sin.

8

Introduction

a passage by Isaiah (14: 12–14) to mean that Lucifer sinned by desiring equality with God.

***

Part I provides an account of major theories of free will from the 1220s to the early 1320s and clarifies their historical presuppositions. Chapter 1 sketches the theories of free will by Anselm of Canterbury, Bernard of Clairvaux, and Peter Lombard, which remained points of reference after the psychological turn. It also presents the Aristotelian ideas that impacted later theories of free will. Chapter 2 concerns the psychological turn itself, initiated especially by Philip the Chancellor and brought to maturity by Thomas Aquinas. From the 1220s onward, almost all theologians consider Aristotle a crucial authority regarding free agency, but they differ substantially in their appropriation of key Aristotelian positions. Some thinkers, such as Thomas Aquinas, Siger of Brabant, and Godfrey of Fontaines, follow him more closely and develop intellectualist accounts of free will: they bind volitions tightly to practical knowledge, see freedom rooted primarily in the intellect, and hold that evil willing presupposes deficient cognition (error, ignorance, or inattention). To some theologians, however, a tight connection between intellect and will implies intellectual determinism and hence the loss of free will – at least of free will in the narrow sense. Furthermore, tracing an evil will to faulty cognition would mean reducing moral evil to making a mistake. To avoid such implications, Henry of Ghent, Franciscan thinkers, and many others take a voluntarist approach, seeing freedom as rooted primarily in the will. Thus, decidedly voluntarist theories of free will develop in reaction to Aquinas, Siger, and Godfrey. Ecclesiastical condemnations in 1277 critique intellectualist propositions and thereby deepen the divisions. Some thinkers, such as Giles of Rome and John of Morrovalle, try to mediate between these extremes. Chapter 3 is dedicated to voluntarism and Chapter 4 to intermediary theories and strict intellectualism. Chapter 5 studies the theories developed in the next generation, by Duns Scotus and William of Ockham (who are both voluntarists), John of Pouilly, Hervaeus Natalis, and Durand of St. Pourçain (who are intellectualists), and Peter Auriol (who proposes an intermediary theory). The topic of Part II is the first cause of evil. Chapter 6 studies the accounts of Augustine and Pseudo-Dionysius, which were constant points of reference for later medieval theories. Chapter 7 discusses the theories from Anselm of Canterbury to John of Pouilly. Despite important differences, all thinkers converge on the idea that a perverse choice lacks an ultimate explanation, and among those who discussed the cause of evil

Introduction

9

from the perspective of the relation of intellect and will, there is more or less open agreement that an evil choice cannot be fully traced to an act of the intellect. Part III studies philosophical explanations of angelic sin. Chapters 8 and 9 examine the interplay of intellect and will, and of cognitive and volitional deficiencies, in the angelic fall. Chapter 8 is dedicated to intellectualist accounts, according to which a cognitive deficiency precedes an evil will; Chapter 9 studies voluntarist and intermediary accounts, both of which emphasize that the angels’ will was deficient before their cognition was flawed. Chapter 10 examines theories of the fallen angels’ or demons’ socalled obstinacy, that is, their inability to change their mind and repent. Theologians generally agreed that demonic obstinacy had an external cause: God condemns them to their permanent evil state and refuses to grant the grace of reconciliation. But many thinkers tried to assign, in addition, an internal, psychological cause that explains why the angels cannot change their thinking and willing. Since voluntarists emphasize the will’s causal independence, many of them denied that the demons’ obstinacy had an internal cause, and some went very far in assigning to God an active role in their permanent evil will. Several studies cover certain aspects of later medieval theories of free will, but many are limited to only a few thinkers and to particular texts. Also, studies on free will are usually limited to the thirteenth century, and those concerning the fourteenth century gravitate toward Scotus and Ockham.9 Medieval theories of the first cause of evil have received little scholarly attention, with the notable exception of Aquinas’s theory.10 The problems of angelic sin and demonic obstinacy have been studied more widely, but scholarly interest has focused on Aquinas, Scotus, and Ockham.11 By examining well-known and neglected thinkers concerning a broad spectrum of issues connected to free will, the present study aims to provide a thorough introduction to the later medieval free will debate. 9

10 11

For medieval theories of free will, see Lottin 1957a (from Anselm of Canterbury to the late thirteenth century), Müller 2009, parts IV–VI (from Anselm to Duns Scotus, in connection with the study of incontinence), and Stadter 1971 (on Franciscan thinkers from Bonaventure to Scotus). For the last three decades of the thirteenth century, see San Cristóbal-Sebastián 1958 (which is not always reliable), Kent 1995, ch. 3, and Putallaz 1995. See also Irwin 2007, chs. 18, 25, and 27 (on Aquinas, Scotus, and Ockham) and Perler 2020, chs. 5–6 (on Aquinas, Peter Olivi, Henry of Ghent, and Ockham). For free will in Franciscan thinkers of the early fourteenth century, see Alliney 2015 and his articles indicated there on p. 272, note 3. Studies of Aquinas’s account are indicated in Chapter 7, note 19. An overview of later medieval theories of angelic sin is found in Hoffmann 2012c. See also D’Ercole 2017, who studies the theories of angelic sin by Thomas Aquinas, Peter Olivi, and Richard of Menneville (Mediavilla). References to further literature will be given in Part III.

part i

Free Will

chapter 1

Free Will with and without Aristotle

Before medieval theologians of the Latin tradition had access to Aristotle’s action theory, they had already inherited a rich, albeit unsystematic, theory of free will from Augustine. Yet most of Augustine’s works were not widely available in the Middle Ages, and what largely contributed to the influence of his theory of free will was its indirect transmission. The principal transmitters are Anselm of Canterbury, Bernard of Clairvaux (neither of whom mention Augustine in their writings on free will), and Peter Lombard.1 This chapter gives an overview of the accounts of free will by these three thinkers. Then it sketches the history of the reception of Aristotle’s action theory and offers an account of the principal Aristotelian doctrines that are relevant to later medieval theories of free will.

1.1

Anselm of Canterbury

In the Latin tradition after Augustine, free will was mostly discussed as a side issue of divine predestination, certain understandings of which were thought to threaten free will.2 Anselm of Canterbury makes a new beginning, putting philosophical questions about free will itself at the center of interest, even when treating genuinely theological issues, such as the sin of the devil, that is, of Lucifer. Anselm famously asks: How is freedom of decision (libertas arbitrii) best defined? This question will become standard in medieval discussions. Anselm argues that the “ability to either sin or not” is not a suitable definition, because God and the good angels cannot sin, but must nevertheless be considered free. Hence, he writes, “the ability to sin is neither 1 2

Anselm, Bernard, and Lombard, as well as numerous other theologians of the twelfth and thirteenth centuries, are discussed in Lottin 1957a. See Flasch 1991 and Adamson 2010, 406–9. In the Islamic and Jewish traditions, freedom was primarily discussed in connection to the threat of necessitation by divine primary causality and divine foreknowledge; see Adamson 2010, 400–6.

13

14

Free Will

freedom nor a part of freedom” – a statement that acquired axiomatic status among later medieval thinkers. Furthermore, a will that cannot sin is freer than a will that can sin, for to sin is harmful and unbecoming (De libertate arbitrii 1, I: 207–8). “He who sins is slave to sin,” he writes (De libertate arbitrii 2, I: 209; cf. ch. 11, I: 223), quoting from the Gospel passage where Jesus promises freedom through liberation from sin (John 8: 31–6). Although Anselm does not mention Augustine by name, he follows him closely; Augustine had made the same argument against defining free decision by the ability to sin (c. Iul. imp. VI.10, cf. V.38) and had also argued that freedom increases the more one is unable to sin.3 Augustine held that the blessed in heaven, who are unable to sin, enjoy a much greater freedom than Adam (the first man), who is able to not sin but also able to sin; likewise, the blessed, who are unable to die, enjoy a much greater immortality than Adam, who is able to not die, because by not sinning he can avoid being punished by death (corrept. 12.33; ench. 28.105; ciu. XXII.30). Rather than the ability to sin or not, “freedom of decision is the ability to maintain the rectitude of the will for the sake of rectitude” (De libertate arbitrii 3; 13, I: 212, 225). In short, freedom of decision is the ability to maintain justice, for Anselm defines justice as “the rectitude of the will maintained for its own sake” (De veritate 12, I: 194). Like justice, free decision has not only volitional conditions (willing rectitude and willing it for its own sake) but also cognitive conditions (having the capacity for rational knowledge and thus for the recognition of what justice consists in) (cf. ibid., I: 192–4). The ability to preserve rectitude for its own sake is never lost; by sinning, we forsake rectitude, but not the ability to preserve rectitude if we have it (De libertate arbitrii 3–4; 12, I: 212–14, 224). Thus Anselm proposes a definition of freedom that is valid across the board; it applies to God and creatures, and to the good and the bad. Anselm’s definition of free decision is teleological, insofar as it describes what a free act aims at: to preserve rectitude. His definition says nothing about evildoing being done of one’s own accord and entailing accountability; for although Anselm appropriates this idea of free decision as well, he maintains an asymmetry between free decision’s relation to good and evil. He writes that the evil angel and the first man sinned by their free decision (per liberum arbitrium), but not insofar as it was free (unde liberum erat). Their decision was free because it was done of their own account 3

The only time Anselm names Augustine in any of his treatises is in the prologue to the Monologion, vol. I: 8.

Free Will with and without Aristotle

15

(sponte); but they did not sin by their decision insofar as it was free, because they did not sin by their ability not to sin, that is, by their ability not to be slaves to sin (De libertate arbitrii 2, I: 210). Anselm, then, in addition to free decision in the teleological sense, admits that there is also free decision in a causal sense: to have free decision means to be the source of one’s act, and thereby to control one’s act.4 Anselm is once more in line with Augustine, who uses free decision in fundamentally the same two senses: according to Augustine, by free decision we can do good or evil (that is, we control our acts), but only in doing good are we truly free (corrept. 1.2). Explicit mentions of free decision in the causal sense – doing good or evil by free decision – are rare in Anselm’s writings.5 But the concept is developed in detail in a treatise from which the expression “free decision” is virtually absent: De casu diaboli, an ingenious account of Lucifer’s sin, and more widely of action theory in general. The topic of De casu diaboli is moral responsibility, primarily the responsibility of the evil angels for their decision to abandon justice (that is, the rectitude of the will), and secondarily of the good angels to hold on to it. In order to ascribe moral responsibility to Lucifer, Anselm shows that Lucifer meets all the requirements for being a moral agent, and in doing so he clarifies the necessary conditions for moral responsibility in general. Though the implications hold for all moral agents, the case of the angels is particularly illuminating, because according to Anselm, all of the prelapsarian angels – the angels prior to their good or evil choice – had the same cognitive conditions (DCD 24, I: 271). Although this is not Anselm’s immediate purpose, his theory of Lucifer’s sin must also explain how two individuals under like cognitive conditions can make unlike choices. Anselm establishes that Lucifer sinned by forsaking justice. The cause of his forsaking justice lies in him; the cause lies not in God, who did not bestow on Lucifer the gift of perseverance in the good, which gift God did, however, bestow on the good angels (DCD 3, I: 237–40). Lucifer did not directly desire to abandon justice, but he desired something that he was not supposed to desire at that moment, and this entailed abandoning justice. What was it that he desired? It could not have been equality with God, for Lucifer knew well that nothing can equal God. So he wanted something less than equality to God, and this lesser thing he wanted against God’s will. Thus he wanted to be like God only in a secondary sense: for to be 4 5

See Visser and Williams 2004, 184, who refer to Anselm’s two senses of free decision as “normative definition” and “purely descriptive definition.” In addition to De libertate arbitrii 2, I: 210, see Anselm’s De concordia praescientiae et praedestinationis et gratiae Dei cum libero arbitrio III.1 and III.13, II: 264 and 286.

16

Free Will

nonsubmissive to God, the higher authority, is in a way to be like God, since not submitting to a higher will is a prerogative of God (DCD 4, I: 240–2).6 Anselm confesses that he can only tell us in generic terms what the object of Lucifer’s desire was: he desired his advantage (commodum) rather than justice; he desired something he did not yet possess from creation, but that he was eventually to receive by his merit if he had not forsaken this opportunity through sin (DCD 6, I: 244).7 The most original contribution of Anselm is an investigation into the kind of motivational structure and rationality conditions required for holding Lucifer – or any rational agent for that matter – accountable for his action. Anselm first establishes that in order to move himself to will (or reject) anything at all, an angel’s will must already have a dispositional orientation to an end, a “first will” for some good, by reason of which he can will to move himself to will something in order to attain that end.8 Anselm holds that the angels received this first will from God at creation. There are two kinds of good: the advantageous (commodum), which Anselm identifies with happiness (beatitudo), and justice (DCD 12, I: 252–5). The God-given will, then, must have been either for happiness or for justice – or for both. Anselm now proposes a thought experiment. He first assumes that the first will God gave was only the will for happiness. Could the angel then move himself to have a different will than the one he has received? No, for even in wanting something for the sake of happiness, he would still only will what he has been given to will, namely happiness. Assuming such an angel wanted to be like God, his will would be neither just, for he wills something unfitting for him, nor unjust, for he wills from necessity. His will would be the work of God, just as being alive is the work of God (DCD 13, I: 255–7). Analogous consequences follow from the hypothesis that God gave as first will only the desire for justice. Then the angel could only will what is fitting, and yet his will would not be just because he would will only what he received from God, being unable to will something different from the will he received. Anselm’s point is that if rational agents had a single motivational source, they could not originate a will by themselves other than what they are programmed to will, and so they could neither be called just nor unjust, nor could they be responsible for what they will. Anselm 6 7 8

For a thorough discussion of the sense in which the abandonment of justice was voluntary although not directly willed as such, see Ekenberg 2016. On the object of Lucifer’s desire, see Pini 2013, 64–5. For an argument that the first will is dispositional rather than occurrent willing, see Visser and Williams 2004, 188–9.

Free Will with and without Aristotle

17

argues that in fact God gave the angels not only the will for happiness, but also the will for justice by which he could restrain the will for happiness. The will for justice only restrains, but does not replace, the will for happiness. In fact, the two wills are not symmetrical: while Anselm thinks it is possible to act on the will for happiness and contrary to the will for justice, he does not deem it possible to bypass the will for happiness.9 The pursuit of happiness can conflict with justice, but the pursuit of justice cannot conflict with happiness; rather, the will for justice makes possible genuine happiness, for only a just person can be truly happy (DCD 14, I: 258). According to the account in De casu diaboli, then, moral agency requires that one does not will by necessity – at least not, he seems to think, by natural necessity. Anselm articulates this nonnecessity in terms of sourcehood: the will (that is, the volition) must come from the agent and not from God. Anselm explains sourcehood as not being determined in one’s motivation for an action, and he elucidates this nondetermination as having access to alternative motivations – to act for the sake of happiness or for the sake of justice.10 Another requirement for moral agency is rationality. Anselm tacitly assumes that blameworthiness requires knowledge of what is morally required; thus he argues that Lucifer did not ignore that he was not supposed to will in the way he willed, for otherwise he would not have been unjust (DCD 22, I: 269). Lucifer’s rationality implies that he knew that his sin deserves punishment. But he could not have known that God would actually punish him, for then he could not have willed the sin that would cause his misery, since, as already mentioned, one cannot will against one’s happiness (DCD 23, I: 270). It was fitting that neither the good nor the bad angels had foreknowledge of God’s punishment, for they were supposed to avoid sin from love of justice alone, and not from aversion to punishment (DCD 23–4, I: 271–2). So Anselm’s Lucifer wagered on God’s mercy: he willed the sin and thought he could get away with it.11 The presence of a reason for both the good and the evil choice of the angels is guaranteed by the two-will theory. The good angels acted on the will for justice, and this act was rational. Lucifer acted on the will for 9 10 11

See Normore 1998, 29. On Anselm’s account of moral agency being rooted essentially in sourcehood, see also Adams 1992, 435 and Visser and Williams 2004, 186–90. For a lucid analysis of Anselm’s account of the rationality conditions of Lucifer’s choice, framing it in terms of a rational gamble, see King 2012, 274–80.

18

Free Will

happiness, and this act, too, was rational, albeit evil. The two sources of motivation are incommensurable: if the ultimate motivation of one’s act is justice, then it is not happiness, and vice versa. This implies that the angels’ motivational structure makes it impossible to give a contrastive explanation of why Lucifer acted for the sake of happiness, or why a good angel acted for the sake of justice.12 Though rational, Lucifer’s act is not ultimately explainable; there is no other cause for his act than that he willed it (DCD 27, I: 275). We will return to this point in Section 7.1. Anselm discusses free decision and the sin of Lucifer without entering deeply into moral psychology. He is more interested in the general volitional and cognitive conditions for free agency and moral responsibility than in an analysis of the interaction between intellect and will, a topic that will become important to later medieval thinkers. Anselm’s late treatise De concordia praescientiae et praedestinationis et gratiae Dei cum libero arbitrio offers a few more tidbits on moral psychology. Anselm distinguishes there between three senses in which we speak of will (voluntas): first as instrument (instrumentum), second as proneness (aptitudo) or affection (affectio) – which is how he now classifies the will for advantage or justice – third as exercise, that is as an act, a volition (usus). Voluntarist thinkers, who emphasize the will’s self-motion, will routinely quote Anselm’s statement that the will as instrument “moves itself by means of its affections; hence it can be called a self-moving instrument” (De concordia III.11, II: 279–81 and 284).

1.2 Bernard of Clairvaux Anselm’s definition of free decision was widely known, but it was often considered incomplete because it does not express the relation between freedom and the kind of control that makes one responsible for one’s acts. Bernard of Clairvaux’s response to this incompleteness in Anselm’s theory is to distinguish clearly between freedom as ability to control one’s acts, which he calls freedom from necessity, and two kinds of freedom from evil: freedom from sin and freedom from misery. Thus he makes explicit what is only implicit in Anselm.13 In positive terms, Bernard calls the first type “freedom of decision” (libertas arbitrii), the second “freedom of counsel” (libertas consilii), and the third “freedom of delight” (libertas complaciti) 12 13

See Normore 1998, 28. According to Vanni Rovighi 1964–5, 55–6, Bernard’s threefold distinction is directly influenced by Anselm.

Free Will with and without Aristotle

19

(De gratia et libero arbitrio 3.6–7; 5.15). Bernard’s conception of the two types of freedom from evil is inspired by Augustine, who, as we have seen, compared the smaller freedom from sin and mortality enjoyed by Adam to the greater freedom from sin and mortality enjoyed by the blessed in heaven. While the second and third types of freedom have degrees, freedom of decision does not; it belongs to human nature, is unlosable, and is possessed equally by the good and the evil, in this life and in the next (e.g., 4.9; 7.21; 8.24). Freedom of decision is freedom from external necessity, that is, from coercion; it is what allows one to act of one’s own accord (sponte). Thus Bernard admits implicitly that freedom is after all compatible with necessity, as long as it is not external but only internal necessity. Alluding to the doctrine that after their initial choice, the angels are in a permanent state of either confirmation in the good or of obstinacy in evil, he writes that the good angels remain freely good, while the evil angels persist freely in evil (4.9). Bernard’s freedom from necessity is basically Anselm’s free decision in the causal sense. Bernard does not yet have an elaborate account of the relation between reason (which was at the time the preferred term for the intellect) and will. Reason has an ancillary role in free agency. The will relies on a judgment about what is allowed or not, useful or not, pleasing or not (4.11). Thus the will needs reason so that it can act, and in this sense Bernard calls willing a “rational motion.” But the will need not act according to reason; rather, it can use reason against reason’s advice or judgment (2.3). For Bernard, the will turns to evil or to the good not by a decision of reason (as an Aristotelian account would have it), but by its own decision (arbitrio). Reason does not impose any necessity on the will; if it did, it would not instruct the will but destroy it (2.4). Bernard’s view that free acts originate in the will, while reason only prepares but does not determine the will’s act, is the standard view in his time.14 Bernard’s threefold distinction of freedom was immediately and widely received. Later voluntarist thinkers will adduce Bernard also as an authority for the will’s ability to act against reason. Furthermore, they appreciate the connection Bernard makes between will, absence of necessity, freedom, and moral responsibility (2.5; 3.6; 4.10), and they like to cite his phrase “where there is will, there is freedom [ubi voluntas, ibi libertas]” (1.2).15

14 15

A more explicit discussion of the subordinate role of reason in free decision in this period is found in Robert of Melun, a student of Peter Abelard; see Perkams 2012. For the role Bernard played for later medieval voluntarists, see also Kent 1995, 112–13.

20

Free Will

1.3 Peter Lombard Although less original than Anselm and Bernard, Peter Lombard was more directly influential on later medieval theologians, since his Sentences became the principal reference text in the formation of Catholic theologians from the thirteenth to sixteenth centuries. His account of free decision draws on contemporary theologians and on Augustine. Lombard’s approach to free decision is in part determined by the context within which he discusses it: a long treatise on sin, spanning about half the Second Book of the Sentences. The context of sin implies that an important concern in discussing free decision is moral responsibility, for Christian theologians generally held that, the special case of original sin aside, sin as a moral failure implies moral responsibility. Thirteenth-century theologians placed great importance on the definition of free decision proposed by Lombard, thinking he had taken it from Augustine.16 It contains some of Lombard’s key ideas about free decision: Free decision [liberum arbitrium] is a faculty [facultas] of reason and will by which one chooses good with the help of grace, or evil when grace is lacking. And it is called “free” with respect to the will, which can be turned in either direction, and “decision” [arbitrium] with respect to reason, of which it is the faculty or power and to which it belongs to discern between good and evil. (Sent. II.24.3 n. 1, I: 452–3)

Several points in this definition deserve highlighting, especially in light of a contrast with Aristotle’s approach to free agency, to be discussed in the next section. Lombard attributes free decision to both reason and will, rather than to one or the other faculty. Reason’s role is to discern, while the will’s role is to choose (eligere). Compared to Anselm and Bernard, who had rarely employed the word choice (electio) and its cognates, in Lombard the terminology is consolidated. The discernment and the choice are between good and evil, not between different means to an end. Lombard attributes a role to grace and to its absence in the choice between good and evil, but without working out the relationship between the respective causal roles of grace and the will. Shortly after the quoted definition, Lombard claims that the ability to do good or evil is unequal: the will “chooses what is good only when assisted by grace, whereas it chooses evil

16

See Lottin 1957a, 64, note 3. Instead, Lombard took the definition from the Summa sententiarum (Lottin 1957a, 25 and 28), of anonymous authorship (Otto of Lucca’s authorship, proposed by Lottin, has been shown to be highly unlikely; see Colish 2013).

Free Will with and without Aristotle

21

by itself” (Sent. II.25.3 n. 1, I: 453).17 Evildoing implies a discord between reason and will: the will desires what reason judges should be avoided (Sent. II.25.8 n. 5, I: 467). Lombard also mentions a definition of free decision as “free judgment about a volition [liberum de voluntate iudicium]” (Sent. II.25.1 n. 2, I: 461). It has its origin in Boethius and was reintroduced into the medieval debate by Peter Abelard as the definition of free decision by “the philosophers.”18 It was frequently referred to by twelfth- and thirteenth-century theologians. Lombard takes up Bernard’s distinction between the three types of freedom and makes a few clarifications. Freedom from necessity does not allow for increase or diminishment because the will’s immunity to coercion is permanent. Thus Lombard implicitly identifies freedom from necessity with freedom from coercion, as later thinkers will also do (Sent. II.25.8 n. 2, I: 466). Citing Augustine (ench. 9.30), he calls freedom from sin “true freedom”; analogously, he calls the freedom to do evil “nontrue freedom” (Sent. II.25.8 nn. 4–5, I: 466–7). But the latter remark is not to be taken literally, for he states explicitly that free decision is not only the freedom to do good, but also the freedom to do evil, and so he implicitly rejects Anselm’s asymmetry claim (Sent. II.25.8 n. 8, I: 468). Lombard briefly touches upon the relationship between free decision and time, a point that will be important in Sections 5.1 and 5.5. Free decision relates to future contingents. It does not relate to the present, because the present is determined. In fact, as Lombard remarks, it is not in our power to make something that is presently the case to be presently not the case; we can only make it not be the case in a future moment – provided it is something contingent, that is, something that admits of either being the case or not (Sent. II.25.1 n. 3, I: 461–2).19 The action theory of Peter Lombard has little in common with Aristotle’s. Lombard already had access to John of Damascus’s De fide orthodoxa,20 which, as we will see shortly, contains an account of human action that takes many elements from Aristotle; but Lombard does not integrate any of John of Damascus’s ideas into his theory of free decision. 17

18 19 20

See also Sent. II.27.2 n. 1, p. 481, quoting Augustine, retr. I.9.6, CCSL LVII: 29: “a human being can fall voluntarily [sponte] and by free decision, but cannot rise again.” Cf. Anselm, De libertate arbitrii 11, p. 223; Bernard, De gratia et libero arbitrio 7.23. On the asymmetry between the capacity to do good and evil in the Christian tradition, see Adamson 2010, 406. Boethius, Commentarii in librum Aristotelis Peri Hermeneias III.9, II: 196; Peter Abelard, Theologia Scholarium III.87, CCCM XIII: 536. Lombard here depends on Hugh of St. Victor, De sacramentis I.5, Corpus Victorinum Textus historici I: 125–6. Aristotle holds a similar view; see Int. 9, 19a23–b4. See Sent. Prolegomena, vol. 1: 32*.

Free Will

22

Anselm, Bernard, and Lombard take for granted that we have free decision, that is, control of our acts; if not, we would lack moral responsibility.21 They do not, however, argue that free decision follows upon the nature of intellect and will; for that, we have to wait until the thirteenth century, when the rediscovery of Aristotle’s action theory and psychology helps medieval theologians develop more refined theories of moral psychology.

1.4

Aristotle as a Game Changer

Aristotle’s impact on the later medieval thinkers was enormous, for almost all of them accepted him as the principal authority in philosophy.22 As a rule, they incorporated Aristotelian ideas into their accounts of free will and attempted to show that their accounts are coherent with his teaching. Aristotle has no discussion of what they call free will (libera voluntas) and free decision (liberum arbitrium). Indeed, he has no notion of the will as the medieval thinkers understood it, a fact they were unaware of, because they found the words voluntas (will) and voluntarium (voluntary) in the translations of his works.23 But Aristotle offers a refined theory of action and so provides a solid framework for later conceptions of free actions and hence of free will. His writings also put certain constraints on theories of free will intended to be coherent with Aristotle’s philosophy. Below I outline the stages of the medieval reception of Aristotle’s action theory and the principal Aristotelian teachings that were constant points of reference from the 1220s onward, when Aristotle’s action theory began to be integrated into accounts of free will. What matters for our purposes is primarily how the later medievals understood Aristotle, only secondarily how well they understood him.

21 22

23

Anselm of Canterbury, De libertate arbitrii 5, I: 214–17 and De casu diaboli 13–14, I: 255–9; Bernard of Clairvaux, De gratia et libero arbitrio 2.5; Peter Lombard, Sent. II.25.8 n. 2, I: 466. See Kent 1995, ch. 2, for the attitude toward Aristotle in the 1260s to 1290s. She provides ample evidence that there was “strong opposition to radical Aristotelians and to Aquinas, but very little hostility to Aristotle” (p. 84). An exception to the positive attitude toward Aristotle was that of Peter Olivi (pp. 84–8). William of Auvergne finds it curious, however, that Aristotle and “his Greek and Arab followers” did not seem to care about the will, for while they mentioned it “perhaps” in their ethical works, they wrote almost nothing about it in their works about the soul; see De anima III.7, Opera omnia II, Supplementum, p. 95a. Duns Scotus remarks that Aristotle spoke little about the will; see QMet IX.15 n. 53, OPh IV: 692.

Free Will with and without Aristotle

23

The Stages of the Reception of Aristotle’s Action Theory The reception of Aristotle’s action theory happened in two main stages. For the early stage, the main sources were John of Damascus’s De fide orthodoxa and the first three Books of the Nicomachean Ethics, which circulated separately, as Ethica vetus (Books II–III) and Ethica nova (Book I). These works were translated into Latin by Burgundio of Pisa in the early 1150s, but were widely used by theology and arts masters only from the 1220s onward.24 De fide orthodoxa contains an exposition of human action, compiled from texts by Nemesius of Emesa and Maximus Confessor, which they in turn adopted loosely from Aristotle.25 Without at first recognizing the Aristotelian origin, medieval thinkers found in De fide orthodoxa a description of the psychological stages involved in decision-making, inspired principally by Aristotle’s account in Nicomachean Ethics III.2–4 and fleshed out with Stoic concepts (De fide orthodoxa II.22 [ch. 36] n. 11, p. 137). In De fide orthodoxa they also found the notion of the will as rational appetite in contradistinction to an irrational appetite – a distinction earlier medieval theologians did not make (De fide orthodoxa II.22 [ch. 36] nn. 8–9, 12, pp. 135–8). The notion of rational appetite has its root in Aristotle’s description of decision (προαίρεσις) as intellectual desire or desiderative intellection (EN VI.2, 1139b4–5). A particularly influential passage was John of Damascus’s claim that free decision is implied in the rational nature of human beings. Irrational beings (that is, nonhuman animals) are acted upon (aguntur) by nature rather than acting, whereas the human being, as a rational being, acts (agit) on nature rather than being acted upon. Humans thus control whether they renounce a desire or follow it, and so they are praised or blamed (De fide orthodoxa II.27 [ch. 41] n. 1, pp. 152–3). John of Damascus traces here the ability to do otherwise, and with it moral responsibility, to one’s being the source of one’s action. Until the mid-thirteenth century, De fide orthodoxa was a more important text for accounts of human action than the available portions of the 24

25

Research on the Ethica vetus and Ethica nova is summarized in Gauthier 1970, vol. I: 111–20, Saccenti 2013, 60–2, and Saccenti 2016, 50–2; for Burgundio’s translation of De fide orthodoxa and its influence, see Saccenti 2016, chs. 2–4; for the use of De fide orthodoxa by theologians from the 1220s until Thomas Aquinas, see also Lottin 1957b. For the sources of De fide orthodoxa, see Gauthier 1954. Nemesius of Emesa’s De natura hominis, which the medievals attributed to Gregory of Nyssa, first translated in the eleventh century and then again by Burgundio of Pisa, circulated also independently from the De fide orthodoxa and constituted a further channel through which elements from Aristotle’s action theory became known to the medievals. See Dobler 2000, 11–68.

24

Free Will

Nicomachean Ethics. The Ethica vetus, whose content overlaps in part with De fide orthodoxa, was esteemed because it neatly confirmed and complemented the account of human action contained in De fide orthodoxa, which, written by a saint, enjoyed special authority. What obstructed the correct understanding of Aristotle by the medievals was not only their fragmentary knowledge of the Nicomachean Ethics. Terminological difficulties implied in some translation choices added to the difficulty; for example, in both De fide orthodoxa and the Ethica vetus, Burgundio translates προαίρεσις (decision) with the neologism eligentia, which obscured how Aristotle understood its nature.26 Further confusion was perhaps caused by John of Damascus. In describing the stages of decision-making, John of Damascus clearly separates judgment (iudicium) and decision (electio) – and even adds a stage between them, “verdict” (sententia) (De fide orthodoxa II.22 [ch. 36] n. 11, p. 137). The separation between judgment and decision, which later medieval thinkers understood as acts of intellect and will, respectively, caused them to debate whether a judgment necessarily entails the corresponding decision. For Aristotle, this question would make no sense. In his view, what is chosen coincides with what has been judged after deliberation (EN III.3, 1113a2–5). Decision is “deliberate desire of what is up to us” (EN III.3, 1113a10–11), or, as we have seen, “intellectual desire or desiderative intellection” (EN VI.2, 1139b4–5). What further distorted the reception of Aristotle’s account of deliberation was John of Damascus’s addition of the Stoic notion of consent, which was usually understood as an act of the will. The second phase of the reception of Aristotle’s action theory began with Robert Grosseteste’s translation in 1246/47 of the Nicomachean Ethics, which in addition to providing its complete text also contains explanatory marginal notes by Grosseteste himself. Furthermore, Grosseteste accompanied each of the ten Books of the Nicomachean Ethics with Latin translations of Greek commentaries. Grosseteste also streamlined the use of terminology, for example, translating προαίρεσις with electio (choice). This may not have rendered Aristotle’s intention more transparent, but it facilitated the integration of his thought into the medieval thinkers’ own theories of free will. Deliberation, Decision, and Incontinence In Nicomachean Ethics III.1–5, Aristotle discusses the notions of voluntary/ involuntary (ἑκούσιον/ἀκούσιον), decision (προαίρεσις), deliberation 26

See Wieland 1981, 281–307, Zavattero 2015.

Free Will with and without Aristotle

25

(βουλή, βούλευσις), wish (βούλησις), and control (an act is “up to us” – ἐφ᾽ ἡμῖν; I am “master” – κύριος – of my act). For their theories of free will, the medievals were particularly interested in Aristotle’s account of decision and its relation to deliberation, for they connected decision to their own notion of liberum arbitrium. This connection involves some friction, however, for the traditional account of liberum arbitrium and Aristotle’s account of decision respond to different concerns. What is at issue with the patristic and medieval notion of liberum arbitrium is choice between alternatives, mostly seen as the alternatives of good and evil; what is at issue with Aristotle’s account of decision is finding the best means to an end; so for Aristotle there can be decision without choice, when there is a single means to the end (EN III.2, 1111b26–7).27 (I use “decision” to denote Aristotle’s προαίρεσις and “choice” to denote the medievals’ electio, which they identify with Aristotle’s “decision.”) Medieval authors stressed the appetitive dimension of liberum arbitrium by emphasizing the will’s role in choice-making; for Aristotle, by contrast, decision has an appetitive component only inasmuch as desire follows upon the decision, but what causes a decision is not appetite, but rather deliberation, a thought process. In essence, Aristotle’s account of decision and deliberation is this. The starting points of deliberation are given; “we deliberate not about ends, but about what is related to the ends [τῶν πρὸς τὰ τέλη],” in other words, what promotes the ends (EN III.3, 1112b11–12).28 Aristotle documents this claim through analogies with the crafts; for example, a doctor does not deliberate about whether he or she should heal, only about how to do so most effectively. When the best means has been determined, deliberation has come to its conclusion and the decision is made (1112b12–20). In the parts of the Nicomachean Ethics that became widely available only in the second phase of its reception, Aristotle goes deeper into the conditions for making good decisions, clarifying their relation to practical wisdom or prudence (φρόνησις), to the moral virtues, and to right appetite in general (Book VI). What is particularly relevant to the medieval discussions of free will is Aristotle’s account of incontinence (ἀκρασία), the moral state in which one acts contrary to one’s better judgment. Here Aristotle approaches the relation between cognition and action from the perspective of the connection between morally deficient action and deficient cognition. Aristotle’s target is Socrates, who holds that someone who knows what is 27 28

See Gauthier 1970, vol. 1: 252–5. What is related to the end is either in means-end relation (e.g., bridle-making for the sake of equitation for the sake of military victory, cf. EN I.1) or in part-whole relation (e.g., moral virtue or intellectual virtue as part of what happiness consists in, cf. EN I.7; EN X.7–8).

26

Free Will

good cannot act contrary to this knowledge, thus denying the possibility of incontinence. For Socrates, what explains evildoing is precisely ignorance of the good. Aristotle defends the possibility of incontinence, but he concedes to Socrates that evildoing presupposes some cognitive failure. Aristotle makes two distinctions: first, having knowledge without exercising it as opposed to actually exercising it; second, knowing universal propositions (dry food is good for everyone) as opposed to knowing particular propositions (this is dry food). According to Aristotle, the incontinent act contrary to their better knowledge because, under the influence of passion, they do not exercise their knowledge as it applies to the particular case (EN VII.3, 1146b31–1147b5). So evildoing involves a cognitive deficiency at least in the sense of a failure to actually consider how it is best to act here and now. In Book III, Aristotle had already hinted at the relation between cognitive deficiency and moral deficiency when he wrote that “every evildoer is ignorant” (EN III.1, 1110b28–9) – a statement that acquired axiomatic value in medieval accounts of action. Aristotle’s theory of incontinence clashes with the earlier medieval conception of the relation between reason (or intellect) and will, according to which the will can choose contrary to the particular judgment of reason. For Aristotle, by contrast, it is inconceivable to decide contrary to one’s better judgment, for as mentioned, in his view one’s decision coincides with one’s better judgment. One can only act contrary to one’s better judgment. In fact, Aristotle states that the incontinent act contrary to their decision (EN VII.4, 1148a9–10, VII.8, 1151a6–7). Active and Passive Powers In contrast to Aristotle’s own thinking, his medieval readers saw in deliberation an act of the intellect and in choice an act of the will, so they read his account of deliberation and decision in terms of the relation between intellect and will. They read other texts by Aristotle as shedding further light on this relation. One is Book IX of the Metaphysics, where Aristotle distinguishes between active and passive powers. A thing possessing passive power can undergo a certain type of change; for example, what is oily is inflammable and what can be bent in a certain way is breakable. Active powers, by contrast, belong to the agent that produces a change; for example, something hot causes heat; someone possessing the art of building is able to build (Met. IX.1, 1046a24–8). Aristotle further distinguishes active powers into irrational powers and rational powers, which differ in two respects. First, in determinate circumstances, irrational powers can

Free Will with and without Aristotle

27

produce only one determinate effect; for example, the hot causes only heat, the chilly only cold. By contrast, rational powers can produce contrary effects, for example, medical art can produce health or disease (Met. IX.2, 1046b4–7, 18–19). Second, irrational powers cause their effects necessarily when they approach their corresponding passive power. To provide an example (which is not Aristotle’s): a hot stove must heat the pot of water on it, and the water must undergo the heating. By contrast, rational powers do not cause their effects necessarily; if they did, they would simultaneously cause contrary effects, which is impossible. Therefore something must be in control (τὸ κύριον) of which effect the rational power actually causes (e.g., health versus sickness), and this is desire (ὄρεξις) or decision (προαίρεσις) (Met. IX.5, 1048a5–11). Some voluntarist thinkers quoted this text according to the Arabo-Latin translation transmitted along with Averroes’s Metaphysics commentary, which reads “desire or will ” (appetitus aut voluntas) instead of “desire or decision.”29 So they took this passage to mean that it is the will rather than the intellect that controls one’s acts.30 Desire, Motion, and Self-Motion Another point of reference for medieval theories of the relation between intellect and will was Aristotle’s account of desire or appetite (ὄρεξις, appetitus) as the cause of the motion of animals (human or nonhuman) in Book III of De anima. William of Moerbeke and the Arabo-Latin translation transmitted with Averroes’s Long Commentary on De anima render the word wish (βούλησις), that is, rational desire, with voluntas (will).31 This translational choice facilitated the interpretation of this passage by the medievals as applying to the will as they understood it: while they recognized that in Aristotle, βούλησις / voluntas can refer to the act of willing, they failed to notice that it cannot refer to the will as a power.32 29 30

31

32

Averroes 1969, 43 l. 16. Walter of Bruges, QD 5c., PhB X: 61 (the edition must be corrected to include “appetitus aut voluntas” in Walter’s quotation of Aristotle); Henry of Ghent, Quodl. I.17c., V: 121; Quodl. XI.6 arg. in opp., 452vL; Richard of Menneville (Mediavilla), QD 15c., III: 78. See also Godfrey of Fontaines (referring to the views of others), Quodl. VI.7, PhB III: 163. For the Arabo-Latin translation of De anima, see Averroes Latinus 1953; for the translation of βούλησις as voluntas, see ibid., p. 518 ll. 7 and 9. Moerbeke’s translation of De anima is contained in Aquinas, Sentencia libri De anima; for Moerbeke’s translation of βούλησις as voluntas, see Editio Leonina XLV/1: 244a. The Latin translations of the Nicomachean Ethics and of De fide orthodoxa also translated βούλησις with voluntas. For the resulting semantic shifts in the reception of Aristotle, see Zavattero 2014. See, e.g., Thomas Aquinas, Sententia libri Ethicorum III.10, XLVII/1: 147 ll. 1–19; cf. Sentencia libri De anima III.8, XLV/1: 240–2 ll. 99–162.

28

Free Will

Aristotle describes the cause of desire within the parameters of his general account of motion (κίνησις), for the change from not desiring to desiring is a qualitative change, which in addition to locomotion and diminution and growth falls under Aristotle’s notion of motion (An. I.3, 406a12–13). Motion in general involves three things: an unmoved mover as the ultimate cause of motion; an intermediate mover or a chain of intermediate movers, each of which is a moved mover (moving another in virtue of being moved); and a moved thing, which does not in turn move something else. Aristotle’s example is a human being (as the unmoved mover), who moves his or her hand, which moves a stick (the intermediate movers), which moves a stone (the thing moved but not in turn moving something else) (Phys. VIII.5, 256a6–8). In animal motion, the unmoved mover is the “realizable good” (τὸ πρακτὸν ἀγαθόν) apprehended by the animal, the moved mover is the desiring part (ὀρεκτικόν), and the thing that is moved is the (human or nonhuman) animal (An. III.10, 433b15–18). For “realizable good” the Arabo-Latin translation has bonum intellectum (understood good) while Moerbeke translates more literally with actuale bonum.33 This account raised the worry among some later medieval thinkers that Aristotle professes intellectual determinism: an object understood as good moves the will to will it, and so it seems that the will does not control its own willing. Many medieval thinkers sought to avoid intellectual determinism by ascribing to the will the ability to move itself. In the medieval debate about the will’s self-motion, two issues tend to intersect: one is whether there can be per se self-motion or only self-motion in some incidental sense; the other is whether self-motion extends only to living things or also to inanimate nature. The hypothesis of the will’s per se self-motion would mean that the will moves itself in virtue of itself and not in virtue of being moved by another; in other words, it implies that the will can all by itself reduce itself from potentiality to actuality, similar to firewood setting itself on fire, or to a clock winding itself up. By contrast, incidental self-motion only involves control of the actual mover by removing an impediment (e.g., incidentally moving a sailboat by lifting its anchor, the actual mover being the wind). According to some medieval thinkers, the will moves itself incidentally insofar as it has some control of the presentation of the desirable object by the intellect, the object or intellect being what these thinkers deem the actual mover of the will. 33

For the Arabo-Latin translation of πρακτὸν ἀγαθόν as bonum intellectum, see Averroes Latinus 1953, 523 l. 10; for Moerbeke’s translation as actuale bonum, see Editio Leonina XLV/1: 244b.

Free Will with and without Aristotle

29

The difficulty in conceiving the will as a self-mover is that the will is not composed of parts, whereas according to Aristotle, self-movers must be composed of at least two parts, one moving and the other moved (Phys. VIII.4, 254b14–15, 255a6–17; VIII.5, 257b11–258a2). In his view, there cannot be per se self-movers, for this would mean that the thing as a whole is both moving and being moved; that is, it would be as a whole in potentiality and in actuality (that is, not in motion and in motion; Aristotle’s example is not hot and hot) – which is contradictory (Phys. VIII.5, 257b2–11). Instead, Aristotle holds what is often called the motion principle: “everything that is moved is moved by something” (Phys. VII.1, 241b24, VIII.4, 256a2–3). The medieval thinkers transmitted it in this form: Motion principle: Everything that is moved is moved by another.34

Yet apparently, there are self-movers without being moved by another: animals moving themselves locally and “heavy and light things” moving themselves to their proper places (e.g., stones downward and fire or air upward). Of animals, Aristotle admits that they move themselves (Phys. VIII.4, 254b14–30), but he clarifies that something else moves them from rest to activity (Phys. VIII.6, 259b1–16). So they are not entirely selfmovers. As to heavy and light things, these pose the “greatest difficulty” (Phys. VIII.4, 254b33–255a2). Aristotle denies that they move themselves, since they are not divided into a part that is moving and a distinct part that is moved (255a12–17). Yet it seems that they move themselves, unless something impedes their motion. From where, then, do heavy and light things derive their motion? To resolve the puzzle, Aristotle distinguishes between two senses in which something is in potentiality. He illustrates the two senses with an example of qualitative change: learners of a science know potentially in a different sense than do possessors of the science who do not momentarily use it (Phys. VIII.4, 255a30–4; cf. An. II.5, 417a21–b2). As Averroes puts it, learners are in “essential potentiality” to knowing, while possessors who are not thinking about what they know are in “accidental potentiality” to knowing (In Phys. VIII com. 32, IV: 370A–D). Aristotle remarks that possessors of a science exercise their knowledge whenever there is no impediment (Phys. VIII.4, 255a34–b5, b21–23). Just so, according to Aristotle, heavy and light things, when impeded from being in their natural places, are by nature in (accidental) potentiality toward them. 34

For the medieval interpretation of the motion principle and its application to natural philosophy, see Weisheipl 1965.

30

Free Will

When the impediment is lifted, they move toward them: a stone downward and air upward. A stone on top of a column is impeded by the column from being in its natural place; one who pulls the column away is the accidental mover of the stone’s falling down. (Likewise, one who lifts the lid from a pot of boiling water is the accidental mover of the rising steam.) By contrast, the (essential) mover of heavy and light things is that which generated them, making them heavy or light, that is, making them naturally inclined downward or upward. Thus the motion principle is upheld (255b5–256a3). Aristotle’s explanation, together with Averroes’s commentary on it, played an important role in the later medieval free will debate: some (e.g., Thomas Aquinas) followed Aristotle in denying self-motion of heavy and light things, while others (e.g., Henry of Ghent and Duns Scotus) rearranged the elements of Aristotle’s account to demonstrate that heavy and light things do move themselves, thus obtaining an important analog to self-motion of the will. Medieval thinkers who defend per se self-motion of the will also drew on Averroes’s explanation that what is in accidental potentiality does not need an external essential mover to cause its motion, only the removal of the obstacle (In Phys. VIII com. 32, IV: 370E, 370H–L).35 Aristotle’s chief influences on later medieval theories of free will, then, are his account of decision-making, his theory of the relation between cognitive deficiencies and moral failures, his theory of the causal influence of the desirable object on the will, his account of active and passive powers, and his imposition of limits on the way in which something can be said to move itself. These Aristotelian doctrines themselves, however, do not constitute a theory of free will. Aristotle’s explanation of the causes of action in general and of the causes of morally deficient action in particular is not an argument that actions or moral failures are in the agent’s control. To work out theories of free will that use the Aristotelian insights, or that at least do not contradict them, is the task for the medieval thinkers writing after the rediscovery of Aristotle’s action theory. 35

Solère 2014, 190–4 provides a more detailed account of Aristotle’s and Averroes’s theory of accidental potentiality and its use by medieval thinkers.

chapter 2

The Psychological Turn and the Rise of Intellectualism

Theologians writing in the 1220s still approach free will largely in continuity with earlier medieval thinkers, stressing the relation between free decision, sin, and grace. But they are also pioneers of the reception of John of Damascus and Aristotle and, thanks to these newly available sources, they add a topic to the traditional themes: the psychological foundation of free decision in reason and will. Over the next decades theologians shift the focus toward this new topic, which allows them to explain why rational agents are free, rather than, like earlier theologians, simply assuming the existence of free decision. This psychological turn begins with Philip the Chancellor, is significantly advanced by the Summa Halensis, Albert the Great, and Bonaventure, and matures in the thought of Thomas Aquinas.1 Once the focus shifts toward the psychological root of free action, a key concern becomes what causes the will’s act, and how the act can be caused without threatening free decision. The very nature of causality becomes an issue, particularly whether a cause produces its effect necessarily, as Siger of Brabant thought. This chapter canvasses the reception of Aristotle’s action theory and the psychological turn from the 1220s to 1276, in broad strokes from William of Auxerre to Bonaventure,2 and then in more detail regarding Aquinas and Siger, whose theories are key to the free will debate of the late thirteenth century and beyond. The chapter begins with William of Auxerre’s influential account of free decision, just prior to the psychological turn.

1 2

My understanding of the psychological turn has greatly benefited from Spiering 2010, which focuses on Philip, Albert, and Aquinas. For the theories of free will from William of Auxerre to Albert the Great, see also Zavattero 2017. Some important contributions to the topic of free will during this period fall outside the focus of this chapter. For these, see Lottin 1957a.

31

Free Will

32

2.1

William of Auxerre

William of Auxerre is among the first theologians to take John of Damascus and occasionally Aristotle’s Ethica vetus into account in discussing free decision. His Summa aurea, composed no earlier than 1215 and no later than 1229,3 adopts roughly the structure of Peter Lombard’s Sentences, and thus examines free decision as part of the topic of sin. Auxerre adopts Boethius’s definition of free decision as “free judgment about a volition” and clarifies free decision’s relevance to merit and demerit by distinguishing it from mere judgment about what is right and wrong. The judgment of free decision is a judgment by which one “commands” (imperat) oneself that one thing is to be done and another to be avoided. Only a judgment in this sense, one commanding action, is meritorious. In this context Auxerre mentions in passing Aristotle’s notion of decision as deliberate desire (EN III.3, 1113a10–11), but considers it to be relevant only to the moral philosopher (ethicus), not to the moral theologian (theologus) (SA II.10.3, II: 278–9). Auxerre’s account of free decision is original, but its originality is not owed to Aristotle. One of Auxerre’s contributions is the addition of a new question over and above the question of how to define free decision: he asks what is the freedom (libertas) of free decision, in other words, how to define freedom or what is the ratio libertatis, as later medieval thinkers will say. This question will be routinely asked by Auxerre’s successors. Guided by theological concerns, Auxerre dismisses three candidates as unsuitable: the absence of coercion, the ability to turn to either alternative (that is, to good or evil), and the ability to turn to what one wants. In his own view, freedom is the aptitude to choose according to what synderesis (one’s infallible moral awareness) dictates (SA II.10.4, II: 283–5). In discussing this additional question, Auxerre expresses more clearly a distinction already found in Bernard: freedom from necessity is twofold, freedom from external coercion and freedom from inevitability. The fallen angels, who are irreversibly obstinate, are free in the first sense but not in the second (II: 284). This distinction will be part of the standard toolbox for discussions about free will (in the broad sense) and its compatibility with necessity.

3

The Summa Aurea contains four books, the first two of which exist in a short and a long redaction. My exposition of Auxerre’s thought is based on the latter. For the two redactions and their dating, see the introductions in Magistri Guillelmi Altissiodorensis Summa aurea I: 7–12; V: 16 and 183–237.

The Psychological Turn

2.2

33

Philip the Chancellor

Aristotelian ideas are much more at work in Philip the Chancellor’s Summa de bono of ca. 1225–8.4 Philip’s discussion of free decision has many similarities to Auxerre’s, but Philip makes two important innovations: first, the locus of his discussion is no longer a treatise of sin, but a treatise of the powers of the soul; second, Philip makes use of John of Damascus’s idea of the will as a rational appetite.5 For Aristotle, of course, “rational appetite” denotes not will but decision (EN VI.2, 1139b4–5); but nonetheless the notion of the will as rational appetite reflects Aristotelian thinking inasmuch as it implies a tight connection between willing and practical thinking. With his increased attention to faculty psychology in general and to the will as a rational appetite in particular, Philip pioneers the psychological turn. Philip is original in other ways as well. He makes an influential observation concerning the psychological basis of freedom: according to Aristotle, the possible intellect can understand contrary things because it is separate from the things it understands; Philip seems to refer particularly to the fact that the possible intellect can understand material things because it is itself immaterial. So also, Philip argues, the practical intellect can relate to contrary alternatives because it is not bound up with matter. Freedom thus has its root in the immateriality of reason and will: since they are not bound up with matter, they are open to opposites and hence free. In confirmation, he quotes John of Damascus’s “sourcehood passage”: humans act, nonhuman animals are acted upon (Summa de bono, I: 170 ll. 157–62; I: 97 ll. 134–7; cf. I: 173 ll. 256–9). Philip uses the argument to show that reason and will are both free; for Philip, reason and will are actually one and the same power. Reason and will are only conceptually distinct, insofar as their acts differ, and their acts differ by their objects: the truth and the good (e.g., I: 173 ll. 260–1; I: 174 ll. 296–7; I: 180 ll. 25–6). In line with the conception of the will as an immaterial power and as identical with reason, Philip adopts the notion of the will as distinct from the sensory appetite. Expounding John of Damascus, Philip connects the notion of the will as rational appetite to the process of deliberation, verdict (sententia), and choice of what is related to the end (ea quae sunt ad finem, which translates τὰ πρὸς τὸ τέλος) (I: 160 ll. 31–7). Although Philip does not consider reason and will to be distinct powers, he sometimes speaks as 4 5

For the dating, see Summa de bono, Introduction, I: 63*–66*. On Philip’s theory of free decision see McCluskey 2001b and Saccenti 2013, 115–42; on the originality of his approach to free decision compared to earlier thinkers, see Spiering 2010.

34

Free Will

though they were; but then he intends by reason and will the distinct acts of judging and willing (I: 173 ll. 260–4; cf. I: 180 ll. 27–31). Philip distinguishes the will (that is, willing), which is consequent upon reason, from desire (desiderium), which follows imagination (I: 166–7 ll. 53–8; I: 175 ll. 308–12). He allows for some discrepancy between judgment and volition: when one has freely judged something as good, there remains the freedom to will it or not. This shows for Philip that freedom is primarily in the will (that is, in willing) (I: 173–4 ll. 264–8). Philip furthermore considers the will to be freer than reason, a view that later voluntarist thinkers make their own: the truth coerces (cogit) reason to assent, but the good, however great, does not coerce the will to consent; rather, it is up to the will to will the good or not (I: 177–8 ll. 384–9). Can the will also be inclined contrary to the judgment of reason? Philip at first denies this: irrational appetite can contradict reason, but the will cannot, since the will is “appetite according to reason” (I: 175 ll. 306–8). But shortly thereafter he admits that the will can choose contrary to the judgment of reason because of the habitual inclination to evil (fomes) that is consequent upon original sin and because of the freedom that allows the will to be inclined to what it wants (I: 180 ll. 32–5; cf. I: 96 ll. 93–4). These inconsistencies show that Philip has not yet fully worked out his understanding of the will as rational appetite.6 Philip considers the will as a rational appetite within a discussion of the definition of free decision and the acts of judging and willing. Like William of Auxerre, Philip adds a distinct investigation into the ratio libertatis, that is, the sense in which free decision is free. He searches for a definition that is most universal, applying to different senses of freedom. Among the candidates Philip rejects is a definition of freedom requiring alternative possibilities. The definition he ultimately proposes is “the ability of not being submitted to a defect except by one’s own consent” (I: 188 ll. 162–3). This definition applies by analogy to Bernard’s three senses of freedom: one is not subject to necessity by the ability to resist coercion; one is not subject to sin except of one’s own accord; one is not subject to misery except as a consequence of sins committed of one’s own accord (I: 188 ll. 163–8; I: 188–9 ll. 176–83). Philip’s Summa de bono had an enormous influence on the next generation of theologians, though the importance he gave to the idea of the will as rational appetite was not felt by all his readers. Two prominent early 6

McCluskey 2001b, 197 goes a step further: she sees Philip’s admission that the will can choose contrary to reason as evidence for his voluntarism.

The Psychological Turn

35

readers of the Summa de bono are Alexander of Hales, coauthor of the Summa Halensis, and Hugh of St. Cher, the first Dominican master to lecture on the Sentences, whose influence reached at least as far as Thomas Aquinas.7 In Alexander’s Glosses to Peter Lombard’s Sentences (probably of the late 1220s), the idea of the will as rational appetite remains absent, whereas in Hugh’s Sentences commentary (composed probably between 1230 and 1233), it is integrated into the discussion of free decision, which is highly dependent on Philip’s Summa de bono.8 According to Hugh, the rational appetite is the will taken as an act, not as a power; it is a free appetite of what has been judged or deliberated (In Sent. II.24, ed. Saccenti, 218 ll. 79–88). While these early thirteenth-century theologians begin using Aristotelian ideas in their theories of human free decision, Aristotle is still mostly absent from accounts of angelic sin.9 Only Philip the Chancellor makes reference to him in discussing angelic free decision and peccability (see Section 8.1). Apparently, the early readers of Aristotle did not yet realize the challenge his action theory poses for the Christian belief in the sin of the angels.

2.3

The Summa Halensis, Albert the Great, and Bonaventure

What explains why rational beings have free decision? Beginning in the 1240s, theologians investigated this question by asking whether free decision is found only in rational beings or also in nonhuman animals, thereby clarifying its psychological foundation. These questions receive a similar and complementary treatment in the Second Part of the Summa Halensis, Albert the Great’s De homine (both written in the early 1240s), and Bonaventure’s Sentences commentary from the beginning of the 1250s.10 7 8

9

10

For Hugh of St. Cher’s influence, see Saccenti 2013, 33–4 and the literature he refers to; for his account of free decision and freedom, see ibid., 142–59. The dating of Alexander’s Glosses to the Sentences of Peter Lombard is debated; its editors date it to 1224–5; see Prolegomena, I: 21*; Niklaus Wicki, the editor of Philip the Chancellor’s Summa de bono, argues against this early dating because it would imply the unlikely hypothesis that Alexander’s Glosses influenced Philip’s Summa de bono rather than vice versa; see Introduction, I: 63*–66*. For the dating of Hugh’s Sentences commentary, see Saccenti 2013, 25. Aristotle is entirely absent from the discussion of angelic sin in William of Auxerre’s Summa aurea (II.3, I: 46–84), Alexander of Hales’s Glosses on the Sentences of Peter Lombard (II.4–7, II: 40–72) and Hugh of St. Cher’s Sentences commentary (II.4–7, Assisi, Biblioteca Comunale, Fondo antico, 131, 44rb–46vb; Padova, Biblioteca Universitaria 853, 49rb–52rb). Books I–III of the Summa Halensis were composed in 1236–45 by the Franciscan Alexander of Hales in collaboration with John of la Rochelle and other Franciscans; see Victorin Doucet, Prolegomena, in Summa Halensis (1924–79) vol. IV, CCCLVa–b and CCCLXIXa–b. The Second Part, which is relevant to this chapter, was written in 1240–4; see ibid., CCCXXXIXb. For the dating of the De

36

Free Will

Major innovation is now the work of masters of the new mendicant orders: Alexander of Hales joined the Franciscans in 1236/37; John of la Rochelle, his coauthor of the Summa Halensis, and Bonaventure are Franciscans as well; and Albert is a Dominican. Only in the next generation do Dominicans and Franciscans take antagonistic views, with Aquinas developing a moderately intellectualist account of free will and Bonaventure’s disciple Walter of Bruges a distinctly voluntarist one. The short chapter of the Summa Halensis asking whether free decision is common to human beings and nonhuman animals quotes John of Damascus’s sourcehood passage two times in full: unlike nonhuman animals, rational beings are the source of their actions, and hence they are able to do otherwise and they have moral responsibility. The chapter explains in detail why nonhuman animals’ being bound up with matter results in lack of free decision. The major source of inspiration for linking free decision to immateriality was clearly Philip the Chancellor; in fact, the Summa Halensis mentions his analogy between the immateriality of the intellect as the foundation of intellectual knowledge and the immateriality of reason and will as the foundation of free decision. The key idea of the chapter is that being bound up with matter is precisely what blocks being the source of one’s motion or action, and that blocking sourcehood also blocks the ability to move or act otherwise. The cognitive power of nonhuman animals is material, and thus their appetitive power is necessarily inclined according to their perception: the perception of something agreeable entails pursuit; the perception of something disagreeable entails avoidance. Which things are agreeable and which are disagreeable depends on their natural appetites (Summa Halensis I-II n. 403c., II: 480a–b). Elsewhere the Summa adduces the classic example of sheep that cannot help but flee at the sight of a wolf.11 By contrast, the rational power in human beings is detached from matter, and hence the human appetitive power is not constrained to pursue agreeable or avoid disagreeable things. Unlike nonhuman animals, which are subject to their natural appetites and cannot act contrary to them, human beings can evaluate their natural appetites, approving or disapproving of them; thus humans can act in accord with or contrary to them (n. 403c., II: 480b). So rational judgment allows human beings to rise above the constraints of matter; this allows them to be the source of their actions. Thereby they do not only have

11

homine, see Prolegomena, XXVII/2: XIV–XV. Schlosser 2014, 13 dates Bonaventure’s Sentences commentary to 1250–2, but its dating must be revisited; see Duba and Schabel 2017, 173–4. Summa Halensis I-II n. 502c., II: 718a. The example goes back to Avicenna Latinus, Liber de anima I.5, ed. Van Riet, pp. 86 and 89.

The Psychological Turn

37

alternative possibilities (or the ability to do otherwise), but they also control by the deliberation of reason which alternative they pursue, and so they have free decision. Nonhuman animals, by contrast, have alternative possibilities without control: they can desire something or not, but as soon as they perceive something, their desire is determined to either alternative (n. 403 ad 1, II: 480b). In short, the Summa Halensis fleshes out Philip’s claim that free decision is rooted in immateriality with an argument for why materiality prevents sourcehood and thereby prevents control over which alternative possibility is actualized. It also suggests that sourcehood is grounded by the deliberation and judgment of reason. What it does not yet explain is why deliberation and judgment ground sourcehood. A full argument must wait until Thomas Aquinas, although Albert the Great and Bonaventure develop an important step toward it. They translate the contrast between materiality and immateriality into one between particularity and universality. In De homine, Albert argues as follows: because the cognitive power of nonhuman animals is entangled with matter, their evaluative judgment is constrained to remain within the bounds of particular, sensory goods – the domain of the pleasant. Similarly, their appetitive power is restricted to what is good or evil in the present moment. By contrast, free decision is not restricted to what is pleasant or unpleasant, but rather is concerned with what is good or evil simply speaking (de bono simpliciter et malo simpliciter), that is, with what is noble or base (honestum vel inhonestum). The implication, which Albert spells out only partially, is that understanding the universal notion of good or evil allows human beings to surpass the confines of particular goods, and hence to either pursue a thing or avoid it.12 What Albert adds to the argument of the Summa Halensis is the idea that free decision requires not only that the cognitive power transcend the bounds of the material, particular thing, but also that the appetitive power transcend the domain of the pleasant. Bonaventure expresses the same idea with greater clarity: a power is free only if its proper object is not restricted to any particular genus of desirable things, but rather is capable of pursuing everything desirable and avoiding everything undesirable. The noble good (bonum honestum) fits the description of such a proper object, but the useful good and the pleasant good do not. Irrational creatures can pursue useful and pleasant goods, but only 12

Albert the Great, De homine, “De libero arbitrio” 1 sol., XXVII/2: 508a. For Albert’s theory of free will, see McCluskey 2001a and 2001c, Miteva 2018, and the literature indicated in these studies. For his understanding of the “noble” (honestum), see McCluskey 2001c, 518–25.

38

Free Will

rational creatures (that is, human beings and angels) can pursue noble goods (In Sent. II.25.1.1.1, II: 593a).13 Free decision, then, has cognitive and volitional conditions. The type of cognition and volition required for free decision presupposes reason and will, which transcend the limits of materiality. By contrast, nonhuman animals lack free decision precisely because their purely material nature makes them hardwired to respond to their environment in a determinate way. Accordingly, the source of their action lies not in them but in the environment they are exposed to. For the authors of the Summa Halensis, Albert, and Bonaventure, sourcehood is not only a necessary but also a sufficient condition for free decision, even without alternative possibilities. Taking up the distinction already found in William of Auxerre, they argue that what is incompatible with free decision is not the necessity of immutability, only the necessity of coercion. In fact, God, the holy angels, and the blessed human beings, who cannot will evil but only the good, do so freely.14 Yet in their view, free decision without alternative possibilities is a special case; they take for granted that free decision normally concerns alternatives, and they hold that the ability to do otherwise is the basis for moral responsibility.15 (By contrast, starting with Aquinas, the expression “free decision” tends to be reserved for free will in the strict sense, which presupposes alternative possibilities.) Bonaventure stresses a particular condition for the kind of sourcehood required for free decision that the Summa Halensis and Albert’s De homine do not mention. He writes that “not any kind of motion coming from within makes a power free, but only the kind of motion by which the moving power moves itself” (In Sent. II.25.1.1.1 ad 4, II: 594a). Free decision’s self-motion allows it to control its act and to elicit its act when it wills (II.25.1.1.3 ad 4, II: 599a) and thus to will freely, voluntarily (II.25.2.1.5c., II: 619a). Even necessary acts, such as the love of God by the blessed who cannot fail to love him, result from free decision’s own command and thus from its moving itself, for it wills in such a way that it wills its own willing (II.25.2.1.2c., II: 612b). Bonaventure writes that free 13 14

15

For recent studies of Bonaventure’s theory of free will, see Maglio 2016 and Zavattero 2019. Summa Halensis I-II n. 404c. and ad 1, II: 481b; n. 411 ad 1–2, II: 487b; Albert the Great, De homine, “De libero arbitrio” 3.6 ad 1–2, XXVII/2: 526b; Bonaventure, In Sent. II.25.2.1.3c. and ad 4, II: 612b, 613b. Summa Halensis I-II n. 389c. and ad 2, II: 467b; n. 416, II: 490a–b; Albert the Great, De homine, “De libero arbitrio” 3.2 sol., XXVII/2: 513b–514a; 3.4.1 arg. 4–5 and ad 4–5, XXVII/2: 520a, 521a; Bonaventure, In Sent. II.24.1.1.1c., II: 555a–b; In Sent. II.25.1.1.3c., II: 599a; In Sent. II.25.2.1.3c., II: 612b–613a.

The Psychological Turn

39

decision is in rational creatures “after God the first principle of their acts” (II.25.2.1.1c., II: 611a) and “only subject to God” (II.25.2.1.4c., II: 616b). Bonaventure does not say how precisely he understands free decision’s selfmotion and how it is metaphysically conceivable. His followers will develop the details. But it is clear that in his view free decision is not a passive power determined from without. While the authors of the Summa Halensis, Albert, and Bonaventure build upon Philip the Chancellor, they do not follow him in emphasizing the notion of the will as rational appetite. Mentions of the notion are rare and far from programmatic. For them, this notion does not suggest a tight binding of the will to reason, nor does it imply a strict priority of reason regarding the will. This is clear from two assumptions they make. First, the will has the freedom to follow reason or not, thus the two powers can conflict.16 Second, the act of reason, but not the act of the will, can be coerced; therefore free decision is rooted more in the will than in reason.17 Albert has a unique theory of how possible conflicts between reason and will are resolved: in his view, free decision is a third power distinct from reason and will, allowing one to decide between what has been decreed (decretum) by reason and what has been desired (desideratum) by the will.18 Bonaventure rejects the hypothesis of free decision as a third power and sees a greater unity between reason and will than Albert (In Sent. II.25.1.1.2c., II: 596b–597a). For Bonaventure, the will is an appetite that is “able to reason [ratiocinativus] or conjoined to reason.” This means that the will does not approve or reject an object without reason dictating it as good or evil, as to be pursued or avoided. But, not unlike William of Auxerre, Bonaventure distinguishes two kinds of judgment of reason; one is merely a statement (dictamen), which involves reason alone and is not necessarily followed by the will; another is a “definitive judgment” that something is to be done or not, which is not without the will, for the will pulls this judgment over to the side that the will itself favors (In Sent. II.25.1.1.6c., ad 2, ad 3, II: 605a–606a). Bonaventure adds further down: “however much reason deliberates, unless the will prefers one alternative, it 16

17

18

Summa Halensis n. 403 ad 2, II: 481a; n. 404 ad 2, II: 481b; Albert the Great, De homine, “De voluntate” 2 sol. and ad 10, XXVII/2: 491 ll. 14–15, p. 492 ll. 14–19; “De libero arbitrio” 4.2 ad arg. 2 in opp., XXVII/2: 522 ll. 10–11; Bonaventure, In Sent. II.25.1.1.6 s.c. 2, c., ad 3, ad 4, II: 605a, 605b– 606b. The point is made in tacitly endorsed opening arguments; see Summa Halensis I-II n. 398 arg. 2, II: 476a; Albert the Great, De homine, “De libero arbitrio” 4.2 arg. 2, XXVII/2: 521 ll. 60–3; Bonaventure, In Sent. II.25.1.1.6 s.c. 1, II: 604b. Albert the Great, De homine, “De libero arbitrio” 4.2 sol. and ad 2, XXVII/2: 513b–514a. For a thorough discussion of Albert’s theory of liberum arbitrium as a third power, see McCluskey 2001c.

40

Free Will

never chooses or consents” (In Sent. II.41.2.2 ad 1, II: 952b–953a). Free decision begins in reason and is perfected in the will; hence in free decision, reason and will relate like matter and form (In Sent. II.25.1.1.6c., II: 605b; see also Summa Halensis I-II n. 414, II: 489a). Although reason and will are thus united, Bonaventure consistently ascribes control more to the will than to reason. He claims that free decision resides principally in the will and attributes to the will the mastery over other powers in the soul, for the will “commands to the other powers but is under no other power’s command” (In Sent. II.41.2.2c., II: 952a–953a). By the mid-thirteenth century, theologians converge on a quite advanced theory of free decision. Its existence is no longer simply taken for granted on the basis of theological premises about merit and demerit, but argued for on the assumption of the immaterial nature of reason and will. Sourcehood and the ability to do otherwise are recognized as distinct aspects of free will, and while sourcehood is deemed a sufficient condition for free will, the ability to do otherwise is considered a necessary condition for moral responsibility. There is an incipient realization that the will’s relation to reason needs clarification: Albert posits an arbiter between reason and will, and this arbiter is for him precisely free decision taken as a third power; Bonaventure acknowledges the inseparability between the definitive judgment of reason and the act of the will, albeit by tipping the balance in favor of the will. The notion of the will as rational appetite has become part of the philosophical vocabulary, but it does not yet denote a strict dependence of the will on reason and it does not yet have a foundational role in explaining rational and free action.

2.4 Thomas Aquinas From the beginning of his career, Thomas Aquinas not only builds upon the achievements of his predecessors, but also benefits from an in-depth knowledge of Aristotle’s action theory thanks to attending Albert the Great’s lectures on Grosseteste’s new translation of the entire Nicomachean Ethics in Cologne ca. 1250–2.19 He follows Aristotle in connecting choice strictly to practical knowledge and moral failure to cognitive deficiencies. He conceives of the will purely as rational appetite and grounds the will’s freedom to will otherwise in reason; but he also attributes to the will a certain control of reason and develops a detailed account 19

Regarding the evidence for Aquinas’s presence at the lectures as well as their probable dates, see Wilhelm Kübel, “Prolegomena,” in Albert the Great, Super Ethica, XIV, pp. V–VI.

The Psychological Turn

41

of the interaction between reason and will. While these traits are already present in his Sentences commentary of 1252–6, he refines his theory of free will throughout his career, bringing it to maturity in the Prima secundae of the Summa theologiae and in the De malo, both written in the early 1270s.20 His followers and adversaries alike recognized the significance of Aquinas’s contribution; he revolutionizes the approach to free will and does much to establish the terms of the later debate. Aquinas is also the first to take an Aristotelian approach to free will throughout his discussion of the evil angels.21 Aquinas’s Notions of Freedom Aquinas does not define any overarching concept of freedom, but rather works with different notions.22 In his view, some acts of the will are necessary, such as the love of God by the blessed who behold God in the beatific vision, as we will see below; like his predecessors, he also considers the will’s necessary acts to be free, because they are not coerced but rather have their source in the will itself.23 Yet, some inconsistencies apart, Aquinas fixes the terminology differently than before. For Aquinas, necessary but free acts fall under the general notion of “free will” (libera voluntas; libertas voluntatis). The minimal condition for such acts is merely sourcehood.24 The ability to do otherwise is not essential to the general notion of free will, but to a more narrow notion, for which Aquinas usually reserves the term free decision (liberum arbitrium).25 Moral responsibility, and thus reward and punishment, presupposes not merely free will and hence sourcehood, but also free decision and hence the ability to do otherwise.26 Although it is by free decision that we can do good or evil, Aquinas adopts Anselm of Canterbury’s claim that the ability to sin (posse 20 21 22 23 24 25

26

For the dating of Aquinas’s writings, see Torrell 2015, 432–84. There are numerous studies of Aquinas’s theory of free will. For recent ones, see Perler 2020, chs. 5–6 and the studies cited in note 51 below. See Spiering 2015. See, e.g., De potentia 10.2 ad 5, Contra doctrinam retrahentium a religione 13, XLI: C64 ll. 24–8. For medieval accounts of necessary and yet free acts of the will, including Aquinas’s, see Hoffmann 2019. See, e.g., ST I.60.1 ad 2; ST I.82.1 ad 1. See, e.g., DV 24.1; ST I.83.1; DM 16.5c., XXIII: 304–5 ll. 228–300. By contrast, in DV 22.6c., XXII: 628 ll. 139–46, he speaks of freedom of the will (libertas voluntatis) by which it can will or not, will this or that, and will good or evil. For Aquinas’s use of the expression liberum arbitrium, see also Alliney 2018. See, e.g., DV 24.1c., XXII: 680 ll. 226–30; ST I.83.1c.; DM 6c., XXIII: 148 ll. 248–68. A detailed discussion of Aquinas’s view that the ability to do otherwise is a necessary condition not only for blame, but also for praise, is found in Hoffmann and Michon 2017, 8–15.

42

Free Will

peccare) is not part of free decision – yet, Aquinas adds, in creatures it follows upon free decision (DV 24.7 ad 4). Essentially, free decision is ordered to the good, just as the will is ordered to the good. Accordingly, “freedom simply” (or “true freedom,” or “spiritual freedom”) is freedom from sin, which frees us from what blocks us from the good.27 Rational Appetite Implies Free Appetite A rigorous conception of the will as a rational appetite is key to Aquinas’s entire theory of free will. That the will is appetite implies for Aquinas that it is fundamentally ordered only to the good (ST I-II.8.1). Inversely, every good is loveable (ST I.5.3). That the will is a rational appetite means that it pursues something precisely to the extent it is known as good by reason or intellect – two terms Aquinas and his contemporaries often use interchangeably in discussing free will in human beings.28 The will can pursue something evil only under the guise of the good (sub ratione boni), that is, only if reason cognizes it as good under some aspect (ST I-II.8.1), an axiom Aquinas and most medieval thinkers see confirmed by the authority of Pseudo-Dionysius.29 The will desires by nature, and hence necessarily, the ultimate end, happiness (or beatitude), and it desires everything else in view of happiness (ST I.82.1–2; I-II.1.6). Many of Aquinas’s later voluntarist critics would agree so far.30 But for Aquinas, the will as rational appetite is more deeply connected to the intellect than his later adversaries would admit. In his view, intellect stands to will as not only a necessary but also a sufficient condition: “in whoever possesses intellect, there is will” (ST I.19.1c.). He says the same about free decision: “wherever there is intellect, there is free decision” (ST I.59.3c.). Likewise, the intellect’s judgment that concludes deliberation is not only a necessary but also a sufficient condition for the will’s choice: “We choose that which we judge as to be done after the investigation of deliberation” (ST III.18.4 ad 2). So, assessing what characterizes Aquinas’s conception of

27

28 29 30

In Sent. II.25.1.1 ad 2; II.25.1.5 ad 2; ST II-II.183.4. Freedom from sin is true freedom according to Augustine (ench. 9.30); it is spiritual freedom, since charity, a gift of the Holy Spirit, frees us from sin, according to Aquinas’s reading of 2 Cor 3: 17: “Where the Spirit of the Lord is, there is freedom”; ST II-II.183.4 ad 1. According to Aquinas, reason and intellect refer to the same power, but intellect signifies the power in its nondiscursive function and reason in its discursive function; see ST I.79.8. See e.g., DV 24.7 ad 6; ST I-II.29.4c.; cf. ST I-II.8.1 s.c. Pseudo-Dionysius contends that evil cannot be desired; see Chapter 6 below, p. 173. For example, Henry of Ghent; see Chapter 3, pp. 63–4.

The Psychological Turn

43

the will as a rational appetite requires looking at how he considers the will and its acts to be grounded in the intellect. For Aquinas, the possession of an intellect – and hence the capacity for universal knowledge – entails the possession of a rational appetite inclined to the universal good. Aquinas makes this claim regarding God, angels, and humans (ST I.19.1; I.59.1; I.80.1–2). The most developed version of the argument for the existence of the will as a rational appetite, contained in De veritate (written 1256–9), also demonstrates that the will can will otherwise, that is, in Aquinas’s terminology, that intellectual beings have free decision. The argument uses elements familiar from Aquinas’s predecessors. It starts from two assumptions: first, that intellectual knowledge is found only in immaterial beings, to a degree that is proportional to the degree of their immateriality; second, that everything whatsoever is ordered to something by means of some appetite (a notion that extends also to inclinations of immaterial things) (DV 23.1c., XXII: 652–3 ll. 113–43). Given the assumption that intellectual beings, like all other beings, have appetite, it remains only to be argued that their appetite is free, in the sense that they are able to order themselves to the things to which they are inclined, in contrast to material things, which are not themselves the cause of their ordering, but are ordered from the outside. An example would be a falling stone, which can neither impede its fall nor direct its course. Immaterial, intellectual beings are not fixed in this way; they are ordered to other things by a free ordering of which they are themselves the cause; for this reason, they act voluntarily and of their own accord (voluntarie et sponte) (p. 653 ll. 144–61). So only immaterial beings, not material beings, are the source of their inclinations. Furthermore, only immaterial beings can do otherwise, as Aquinas shows with an example: if the form (that is, the plan) a builder uses to build a house were material and hence determinate, then it would incline only in one determinate way – only to build cookie-cutter houses, we might say. But if the form guiding the builder is the notion of “house” (ratio domus) in the mind, then the builder will be free to build it this way or that, and even either to build it or not, because the notion of the house itself is indifferent to the existence or nonexistence of the house (p. 653 ll. 161–73). From this, Aquinas concludes that intellectual beings have free inclinations, that is, a free appetite: the will (p. 653 ll. 179–87). In De malo, he picks up the example of the builder and clarifies that the indeterminacy of the will regarding various instantiations of a house is rooted in the fact that the notion of “house” is universal, not particularized like material things (DM 6c., XXIII: 148 ll. 269–96).

44

Free Will

Yet in this argument, the crucial step, that intellectual beings are the source of their inclinations, is only affirmed, not demonstrated. Aquinas furnishes this demonstration elsewhere, in arguing that free volitions are caused by free judgments. He has several versions of this argument, which all fit the following pattern: he starts from the premise that the will’s acts follow reason’s judgments about what is to be willed; then he argues that we are the source of our volitions because we are the source of our judgments; he explains that we are the source of our judgments because they are generated from our own reflection rather than resulting from natural instinct. Next, he argues that since we are the source of our judgments, we can judge otherwise, and for this reason we can will otherwise and act otherwise. By contrast, nonhuman animals are not the source of their judgment, and hence they cannot judge otherwise nor act otherwise, as Aquinas illustrates with Avicenna’s familiar example: upon seeing a wolf, the sheep’s judgment that the wolf is a threat is determined by nature and therefore does not admit of being otherwise; thus the sheep cannot help but flee.31 Aquinas thus grounds the sourcehood required for free will principally in reason. Specifically, what makes us the cause of our judgments is our ability to judge our own judgments. We have this capacity for reflexive judgments because we understand the formality (ratio) of an end and the formality of what is related to the end (ea quae sunt ad finem; in most cases, these are means to the end). In other words, we understand the end as end and the means as means; furthermore, we understand how the means are related to the end.32 So the reasons behind our volitions are transparent to us and hence open to reconsideration. Therefore our reasons are the fruit of our own reflection, rather than being determined from the outside, as are the judgments of nonhuman animals. For example, when we are hungry and judge that we should eat now, we can dismiss this judgment out of dietary, social, or even political considerations.33 As we will see, thinkers writing after Aquinas who otherwise disagree strongly among each other generally agree that free will is grounded in self-reflexivity, that is, the ability to bend back on one’s own psychological acts: judging one’s judgments, willing one’s volitions. In these texts, Aquinas moves from sourcehood to the ability to do otherwise and from lack of sourcehood to the inability to do otherwise. He 31 32 33

SCG II.47–8; DV 22.4; DV 24.1–2; ST I.83.1; I-II.6.2; DM 16.5c., XXIII: 304 ll. 215–30. E.g., DV 24.1c. XXII: 681 ll. 288–93; DV 24.2c., p. 685 ll. 92–100; ST I-II.6.2. On Aquinas’s account of reflexive judgments as the cause of free decision, see also Gallagher 1991, 566–74.

The Psychological Turn

45

does so without further argument, presumably because he thinks that by being the source of one’s act, one is also the source of acting in one way or another, provided alternatives are available. Since sourcehood and the ability to do otherwise are normally connected, Aquinas often refers to both indistinctly with the same term, “mastery” (dominium), that is, control.34 Nevertheless, sourcehood and the ability to do otherwise can come apart. In fact, as the Summa Halensis had already remarked, nonhuman animals can do otherwise but are not themselves the source of the alternative they actually pursue: sheep can run or not, dogs can bark or not; but upon seeing a wolf, sheep cannot help but flee, and when excited, dogs cannot help but bark (DV 24.2c., p. 686 ll. 115–33). Furthermore, in cases of necessary willing (to be discussed below), there is sourcehood without the ability to do otherwise. For Aquinas, then, free decision is grounded upon the cognitive abilities of reason. He synthetically describes its root in reason as follows: Freedom’s root, in the sense of subject, is the will; but in the sense of cause, it is reason. For the will can freely be carried toward alternatives because reason can have alternative conceptions of the good. And therefore the philosophers define free decision as the free judgment of reason [liberum de ratione iudicium], as if to say that reason is the cause of freedom. (ST I-II.17.1 ad 2; cf. DV 24.2c., p. 686 ll. 99–100)

Judgment-Volition Conformity and Proportional Deficiency Aquinas, then, traces free decision to free judgment. This move has a presupposition that becomes highly controversial in the later free will debate: for Aquinas, it follows upon the concept of the will as rational appetite that “the appetite follows cognition, because the appetite is only for the good that is proposed to it by the cognitive power. . . . The judgment about this particular possible act right now can never be contrary to appetite” (DV 24.2c., p. 685 ll. 70–2, 79–81). In other words, one wills an object (that is, desires, chooses, or enjoys it) precisely in proportion to reason’s judgment about the object, that is, to what the person subjectively considers best; “the activity [of the will] is proportioned to the activity of the intellect” (DM 16.2 ad 6, XXIII: 290 ll. 420–1).35 I call this assumption “judgment-volition conformity”: 34 35

E.g., in ST I-II.6.2 arg. 2 dominium denotes sourcehood, in ad 2 it denotes the ability to do otherwise. See also In Sent. I.45.1.1c., I: 1033; ST I.64.2c.; I-II.13.3c.; I-II.77.1c.; III.18.4 ad 2; DM 16.2c., XXIII: 288–9 ll. 261–9; 16.5c., XXIII: 305 ll. 317–18. While Aquinas sometimes writes that the choice follows

46

Free Will

Judgment-volition conformity: Willing must conform to what reason judges as to be willed.

Furthermore, according to Aquinas there is proportional deficiency in intellect and will: “in a sin, the deficiency of the intellect or reason and of the will are always proportionally tied in with another” (DM 16.2 ad 4). Deficient willing is in and of itself morally deficient willing, that is, sinful willing (DM 1.3 ad 12), whereas cognitive deficiencies are error (a mistaken judgment), ignorance (failure to know something one should know), or lack of occurrent consideration of what is habitually known, when it ought to be considered. As such, the affirmation that cognitive deficiencies and volitional deficiencies are proportional leaves the causal direction open. Henry of Ghent, for example, insists that sin causes error, but that error does not cause sin. Aquinas does not deny that sin can cause error, but he claims that cognitive deficiencies are necessarily part of the causal history leading to sin. This claim played an important role in the free will debate, especially after Aquinas. It applies to intellect and will an idea usually associated with Socrates, who held that evildoing presupposes ignorance of the good. Aquinas’s claim can be stated as follows: Socratic deficiency thesis: Deficient willing presupposes deficient cognition.

That deficient willing presupposes deficient cognition need not be understood as temporal priority, although for Aquinas, temporal priority seems to be the normal case. According to Aquinas, the deficiency relationship expressed in the Socratic deficiency thesis follows from judgment-volition conformity. As the will wills what is judged as good, it can will something evil only if it is judged as good: Since the object of the will is the good or the apparent good, the will is never moved to evil unless that which is not good appears somehow to reason as good; and therefore, the will only tends to evil with some ignorance or error in reason. (ST I-II.77.2c.)36

Yet the phenomena of acting against one’s better judgment (incontinence) and of acting against one’s conscience seem to be counterexamples to

36

the judgment, at other times he even makes judgment and choice coincide. Choice is a certain judgment: DV 24.1 ad 1; ST I.83.3 ad 2; choice follows the judgment: ST I-II.13.1 ad 2; 13.3c.; choice either is a certain judgment or follows it: DV 24.1 ad 20; choice is simultaneous with the judgment that concludes deliberation: DM 16.4c., XXIII: 299 ll. 279–80; choice is the conclusion of a practical syllogism: ST I.86.1 ad 2; I-II.76.1c. Aquinas makes the same claim in numerous other places throughout his oeuvre. He insists that cognitive deficiency is antecedent to volitional deficiency; see, e.g., DV 24.8, XXII: 700 ll. 92–101.

The Psychological Turn

47

judgment-volition conformity, while deliberate evildoing (peccatum ex certa malitia, also translated as willful wrongdoing) seems to be a counterexample to the Socratic deficiency thesis. But for Aquinas, these phenomena can be explained without denying judgment-volition conformity and rejecting the Socratic deficiency thesis. Loosely following Aristotle (EN VII.3), Aquinas affirms that the incontinent only act contrary to the general knowledge that, for example, fornication should be avoided. In the incontinent, passions impede the consideration of their general knowledge of what is to be avoided; instead, the incontinent act in accord with the judgment that in this particular situation, here and now, it is good to fornicate (DV 24.2c., XXII: 685 ll. 81–4; ST I-II.77.2; DM 3.9).37 By contrast, when acting contrary to one’s conscience, one acts indeed against the particular judgment about what one should do or avoid here and now. But according to Aquinas, the “judgment of conscience” (iudicium conscientiae) is purely cognitive, while the “judgment of free decision” (iudicium liberi arbitrii) implies an affective commitment. The latter is the judgment that concludes deliberation and that defines one’s choice; hence Aquinas also calls it “judgment of choice” (iudicium electionis). Thus when one’s conscience judges that one should avoid fornication here and now because it is immoral, sensual desire can block reason from acting on the prohibitive judgment of conscience, and by considering the pleasurable aspect of fornication, one can act on the permissive judgment of choice, namely, that one should commit the act because it is pleasurable (DV 17.1 ad 4).38 So when Aquinas claims that one wills in proportion to one’s judgment, he means the judgment of choice. The fact that this rather than that cognitive judgment actually becomes a judgment of choice depends on an affective commitment (see also ST I.83.3 ad 2). Unfortunately, Aquinas does not say what causes this affective commitment. Is it itself caused entirely by cognition, or does the will add something that is not fully traceable to cognition?39 If the latter, what allows the will to do so? After Aquinas, theologians debate heatedly whether the will’s choices can be entirely traced to cognitive antecedents or not. Finally, as regards deliberate evildoing, although it is done in full awareness that one’s act is evil, there is nevertheless some ignorance 37 38 39

For more detailed discussions of Aquinas’s account of incontinence, see Kent 1989, Kent 2007, Müller 2009, 512–47, and Pickavé 2013. For a more detailed analysis of Aquinas’s psychological explanation of acting against conscience, see Müller 2009, 535–7. Interpreters disagree whether for Aquinas all acts of the will are fully traceable to the intellect; for references, see Hoffmann and Michon 2017, 22–7.

48

Free Will

implied, namely the so-called “ignorance of choice” (ST I-II.78.4 ad 1). This is the kind of ignorance that corrupts the judgment of prudence concerning what is best to do here and now (In Sent. II.22.2.2c.; De virtutibus in communi 6 ad 3). For example, the intemperate ignore the fact that pleasure is not worth pursuing at the cost of sinning (cf. ST I-II.78.1 ad 1).40 Ignorance of choice (or “error of choice”) is thus implied in every sin, and in Aquinas’s understanding, it is this type of ignorance that Aristotle refers to in saying that “every evildoer is ignorant” (ST I-II.76.4 ad 1; DV 18.6 ad 1; cf. EN III.1, 1110b28–9). For Aquinas, then, judgment-volition conformity implies only that in the moment one considers it best to will something – even if one thinks it is best from selfish considerations, fully aware that it is against right reason – one cannot will contrary to one’s judgment. Nonnecessity and the Will as a Moved Mover Aquinas often uses causal terms to express the idea that volitions conform to judgments: the will is moved to its act by the intellect, or, as he also says, by a cognized good (bonum intellectum) (ST I.82.4c.; I.105.4c.; DM 6c., XXIII: 149 ll. 339–43). In fact, in the Prima pars of the Summa theologiae (completed in 1268), Aquinas adopts the Aristotelian dictum that the will is a moved mover and calls it a passive power (ST I.80.2c.), just as he had called the intellect (ST I.79.2). In 1270 the bishop of Paris Stephen Tempier condemned the statement that free decision is a passive power, presumably out of concern that this would imply determinism,41 and Aquinas in fact subsequently avoids calling the will a passive power (see in particular DM 6 arg. 7 and ad 7). But the idea that the will cannot actualize itself but rather needs actualization from without remains present in his writings (ST I-II.9.1c.; I-II.9.4c.; DM 6c., pp. 148–9 ll. 308–10, 339–422). The basic picture is this: the will’s choice is traced to reason’s judgment, which in turn is grounded in the cognition of some object. How, then, do rational agents control whether something appears choiceworthy to them, that is, which judgment becomes operative in a particular choice they make? Reflexive judgments (judgments of a judgment) are only part of the story, for a question remains open: how do rational agents control the outcome of their reflection and hence their choice – to use the preferred 40 41

Concerning deliberate evildoing, see Kent and Dressel 2016 and McCluskey 2017, ch. 5. Chartularium Universitatis Parisiensis, I: 487; cf. Chapter 3, p. 59.

The Psychological Turn

49

example at the time, whether to fornicate for the sake of pleasure or to abstain for the sake of morality? Aquinas does not provide a full answer to this question in any one place. He provides an important part of the answer, however, by arguing that – special situations aside – the will does not act of necessity, which means for Aquinas that it can will otherwise. In order to explain this ability for alternatives, Aquinas distinguishes in his later writings two respects in which the will is in potentiality to its acts of willing: regarding their specification (which he also calls determination), whether it wills this or that, and regarding their exercise, whether or not it wills at all. In both respects, the will needs to be moved, that is, actualized (ST I-II.9.1; DM 6c., p. 148 ll. 308–19). As to specification, the will is moved by the intellect presenting a cognized good (DM 6c., p. 149 ll. 339–43). In order for a cognized good to move the will to will it, it is necessary and sufficient that it be understood as good and suitable (conveniens), that is, as good for the person willing – not just in general, but “in particular,” by which Aquinas seems to mean: here and now, in light of a certain set of considerations. In other words, anything understood as good and suitable in particular can move the will, and nothing moves the will that is not so understood. An object that is understood as good and suitable from every possible perspective moves the will necessarily, but only so long as one thinks about it. Such an object is happiness, and God, as clearly seen by the blessed. Whatever is not good and suitable from every perspective does not move the will of necessity, but rather admits of being willed or not. For example, a diet seen as beneficial because it is healthy moves the will to will it; but a diet seen as undesirable since it involves renouncing tasty food does not move the will.42 As to the exercise of the act, the powers of the soul are moved by the will; for example, because we want to think about something, we move our intellect to think about it. In fact, to think is itself a certain good, and thus it falls under the proper object of the will, which is the good in general. Regarding the exercise, the will also moves itself (DM 6c., p. 149 ll. 343–54). Aquinas does in fact allow for self-motion of the will – an issue that will be hotly debated after Aquinas. But he denies per se self-motion of the will, that is, for Aquinas the will cannot reduce itself from potentiality to actuality, as Henry of Ghent will maintain. Rather, in virtue of actually willing an end, the will can reduce itself from potentially to actually willing what is related to the end; for example, by willing health, the will moves itself to will what is needed to achieve it. In this, the will is not unique; indeed, the intellect, too, 42

DM 6c., pp. 149–50 ll. 418–49; ST I.82.2; I-II.10.2; I-II.13.6.

50

Free Will

moves itself: in virtue of understanding self-evident principles, it moves itself from potentially to actually understanding a conclusion. When moving itself, the will does not do so independently of the intellect, for it moves itself by means of deliberation: wanting health, the will makes the intellect deliberate about how to achieve it. If it is concluded that health is to be achieved by taking medicine, then one wills to take medicine (DM 6c., p. 149 ll. 360–77 and ad 20; ST I-II.9.3–4).43 According to Aquinas, the fact that the will moves itself by means of deliberation explains why it does not exercise its own act or that of another power of the soul necessarily, for deliberation can conclude either that the will should exercise its act or not. This is the point Aquinas wanted to establish. But there is a further problem: the will does not constantly want the intellect to deliberate, and so willing to deliberate itself requires a cause. The will to deliberate may itself result from a previous deliberation judging that it is worth deliberating. But this cannot proceed ad infinitum. The will to deliberate must begin somewhere; there must be a first will to deliberate. Interestingly, Aquinas does not trace the cause of this first will to deliberate to the intellect. Nor does he trace the first transition from potentially willing to deliberate to actually willing to deliberate to the will moving itself, for this would be per se self-motion, which Aquinas considers impossible. Instead, he traces it to God, who according to Aquinas moves the will in conformity with its nature as contingent cause, and thus without necessitating it.44 Beyond this point, no further explanation is possible. The particular case of God causing the will to deliberate does not seem to be an exception to judgment-volition conformity, as though God caused a volition that corresponds to no judgment. Aquinas seems rather to think that God causes the will to deliberate together with the judgment that it is worth doing so. In any event, Aquinas does not trace the causal history of the will’s acts only to the intellect (and by way of the intellect to the object), but also to God. While Aquinas thinks that God’s causal contribution does not threaten but rather safeguards free decision (e.g., ST I.83.1 ad 3), he says little about how precisely this is the case. Aquinas’s 43

44

Aquinas’s exercise/specification distinction is often referred to as a distinction between “freedom of exercise” and “freedom of specification” (e.g., Jensen 2018b, 245–52). This terminology is misleading; for Aquinas, there is neither freedom of exercise nor of specification – only indeterminacy, that is, potentiality in both these respects. ST I-II.9.1; I-II.9.3–4; I-II.9.6; I-II.10.4; DM 6c., p. 149 ll. 308–417, and ad 17. For a fuller exposition of the interaction of intellect and will in DM 6, see Hoffmann and Furlong 2016, 60–5. Aquinas’s account of the relationship between divine causation and created causation in acts of free will is particularly intricate; on this topic, see Wippel 1984, 358–63, Shanley 1998, Kim 2016, and Costa 2018.

The Psychological Turn

51

exercise/specification distinction was quite successful in the following decades, but his theory of God’s involvement in causing the will to deliberate did not receive much following. Formal, Final, and Efficient Causality Aquinas makes an important clarification: the intellect (or the object it makes present) and the will move in different orders of causality. The object presented by the intellect is the principle regarding specification insofar as it is the formal cause of the will’s act (ST I-II.9.1c.; DM 6c., p. 148 ll. 320–9). The object furthermore moves the will in the order of final causality, because as an end it motivates the will to pursue the object (e.g., ST I.82.4c.).45 The will moves itself and other powers by efficient causality in pursuing the end.46 It is important in Aquinas’s theory that the cognized object, as a formal and final cause of willing, does not produce any willing by efficient causality. In this regard, he differs from the strongly intellectualist theories of Godfrey of Fontaines and his followers, to be discussed in Sections 4.3 and 5.2. For Aquinas, if the will were subject to efficient causality, it would not be moved by its own inclination and hence its willing would not be voluntary. In fact, he argues that the will’s acts cannot be coerced, for one cannot be coerced to will something of one’s own accord, any more than a stone can be made to move upward by its own inclination (ST I.82.1c.; I-II.6.4). Only God can move the will efficiently, as he does, for example, when he moves the will to want to deliberate (DV 22.9; ST I-II.6.4 ad 1). Furthermore, efficient causes produce their effect necessarily, unless they are impeded (cf. De principiis naturae 4, XLIII: 44 ll. 79–86). Formal causes do not produce anything, and final causes do so only remotely and in any event not necessarily. How, then, is the will’s act caused by a formal cause and a final cause? As a formal cause of the will’s act, the object actualizes willing, because there cannot be any willing that is not willing of some object, just as there cannot be any matter that is unformed (cf. De principiis naturae 4, p. 44 ll. 38–40). Reason not only specifies the will’s act (e.g., to willing to take medicine) but also orders the way volitions are related to another, as when 45 46

In clarifying the causal role of the intellect in free decision, Aquinas’s terminology changes over the course of his career, but, as Westberg 1994 has shown, not the substance of his account. E.g., ST I.82.4c.; DV 22.12; DM 6c., pp. 148–9 ll. 329–38. That the will moves by efficient causality is expressed with particular clarity in DV 22.12c. and ST I.82.4c., but it is also clear in DM 6c., p. 148 ll. 322–5.

52

Free Will

one desires health as to be attained by some yet unspecified means (“intention”; ST I-II.12.1 ad 3 and ad 4), or when one wants to attain health by a particular means, such as by taking medicine rather than by exercising (“choice”; ST I.83.3 ad 3), or when one wants to employ one’s limbs to obtain medicine (“use”; ST I-II.16.1 ad 1). According to Aquinas, insofar as reason orders the will’s acts by specifying them and by causing their relational structure (e.g., wanting this in view of that), reason and will correlate like form and matter, and thus like the superior to the inferior (ST I-II.13.1c.; I-II.17.4c.). (Recall that the Summa Halensis and Bonaventure had suggested the reverse relation: the will is superior to reason and it is like the form, while reason is like matter.) Yet insofar as an object is a formal cause, it does not move the will from not willing to willing, for only an efficient cause, not a formal cause, can bring something into existence. The intellect presenting the object can move the will only indirectly, inasmuch as by presenting an object, the intellect presents an end that motivates the will “in the way it is said that the end moves an efficient cause” (ST I.82.4c.; cf. DV 22.9). The end “moves” an efficient cause, because without an end the efficient cause would not act, and so the end causes the efficient cause to be efficient. For example, health, as a final cause, causes someone to walk (De principiis naturae 4, pp. 43–4 ll. 9–36).47 As a final cause, however, the object does not necessitate the will, for final causality implies only conditional necessity. Aquinas writes that if someone wants to cross the ocean, he or she must necessarily take a ship (ST I.82.1c.; III.46.1c.). But taking a ship is not necessary absolutely speaking, because one can renounce the desire to cross the ocean. Control We have now seen how Aquinas explains that the will’s acts are not necessitated. But a crucial question still lacks a satisfying answer: How does one control which alternative one actually chooses? For example, how does someone who cares about pleasure and health end up choosing a diet because it is healthy, despite having entertained the thought of avoiding a diet because it is unpleasant? A theory like Aquinas’s, which assumes that the will’s choice necessarily conforms to the concluding judgment of deliberation, can safeguard control of one’s choice only by admitting 47

For Aquinas’s account of the interplay of different types of causality in acts of the will, see also Osborne 2014, 13–19.

The Psychological Turn

53

that one has some control of deliberation: whether or not one deliberates, how one deliberates, and until what point one deliberates (and hence, with which judgment one concludes one’s deliberation). Aquinas only gives a few hints for a solution. In his view, the will not only moves the intellect to think about something but it controls also whether one actually considers something one knows habitually (DM 16.8c., XXIII: 321 ll. 215–21).48 The will also moves the intellect from one consideration to another (SCG III.10 n. 1950). By controlling one’s considerations, one controls what one judges. During deliberation, one typically makes provisional judgments, for example, “a diet is not choiceworthy because it is unpleasant” and “a diet is choiceworthy since it is healthy.” One controls also which provisional judgment becomes the judgment that concludes deliberation, that is, the judgment of choice. Even when one’s deliberation has already brought to light a sufficient reason for making a particular choice, one can either make the choice on the grounds of that reason, or one can prolong deliberation by reflecting anew upon that reason, and so one can dismiss it in light of new considerations (DM 6 ad 15). What causes the will to act on a particular judgment and thereby to end deliberation, or what causes it instead to continue to deliberate? Aquinas does not tell us.49 It is unlikely that he would think that acting on a particular judgment (e.g., this diet is now to be chosen because it is healthy) requires a metajudgment that one should act on such a judgment, for this would lead to an infinite regress.50 A metajudgment is not needed, since the judgment itself (the diet is now to be chosen) suffices to motivate the will’s adhesion – even though it does not necessitate it, as we have seen. In the absence of metajudgments, there is no contrastive explanation for why one acts on this judgment rather than that. The only explanation lies in the judgment itself, together with the fact that the will adheres to it. If this is indeed Aquinas’s view, then it would follow that an agent can explain the reason for each decision, but ultimately cannot explain why this reason became operative rather than that, other than the fact that he or she wanted to act on this reason. Whether Aquinas indeed ultimately traces the fact that a person makes this rather than that choice to the will, or to the intellect, or jointly to both, is debated among interpreters. There is, 48 49 50

Aquinas’s theory of control of one’s actual consideration will be discussed more fully in Chapter 8 below, pp. 212–13. For a hypothesis based on Aquinas’s theoretical resources, see Jensen 2018b, 272–84. Had Aquinas intended to make such a claim, the natural place would have been the discussion of consent, which he defines as the “application of an appetitive movement to something to be done” (ST I-II.15.2c.). But no such claim is found in his discussion of consent.

Free Will

54

furthermore, disagreement about whether his theory of free decision is ultimately compatibilist (wittingly or unwittingly) or libertarian.51 This is not the place to argue for a particular interpretation. The discussion of Aquinas’s position on the first cause of evil (Section 7.5) and on how angelic sin is possible (Section 8.3) will shed further light on this problem. Aquinas as a Moderate Intellectualist Compared to earlier theories, Aquinas’s free will theory is distinguished by the strict proportion he conceives between intellect and will. This conception of the relation of intellect and will places his theory under the general label of intellectualism. Wherever there is intellect, there is will; to each act of the will corresponds a judgment of intellect; there is no evil willing without a cognitive deficiency. For Aquinas, cognition and volition are distinct principles of an act, and yet they are inseparable, just as form and matter are distinct but inseparable principles of composite substances. The will is rational appetite not merely insofar as willing presupposes cognition, but also insofar as willing conforms to cognition. These intellectualist traits aside, in Aquinas’s theory of free will cognition does not cause a volition efficiently, and the efficient cause of willing is traced not to the intellect but ultimately to God, who causes the will’s act by respecting the will’s nature as a contingent cause. This is a rather voluntarist trait of his theory. On the whole, we may call Aquinas a moderate intellectualist.

2.5

Siger of Brabant

A more rigorously intellectualist, albeit less systematic approach to free will is developed by Aquinas’s contemporary Siger of Brabant, a master at the faculty of art of the University of Paris, who had a great impact on the later free will debate. His influence was both direct, most notably on Godfrey of Fontaines, and indirect, in that he provoked strong reactions from voluntarists and ecclesiastical authorities.

51

For example, Pasnau 2002, 224–33 and more hesitantly Hause 1997 interpret Aquinas as a compatibilist, whereas Gallagher 1994, MacDonald 1998, Stump 2003, 277–306, and Hoffmann and Michon 2017 as a libertarian; Shanley 2007 considers these labels altogether out of place regarding Aquinas’s theory. Furthermore, Hause 1997 and MacDonald 1998 consider his account to be intellectualist, whereas according to Gallagher 1994 and Pasnau 2002, 224–33 there are voluntarist elements in Aquinas’s theory. For a discussion of these interpretations, see Hoffmann and Michon 2017, 22–7.

The Psychological Turn

55

In Siger’s lectures and writings, the topic of freedom and free decision is brought up mostly indirectly, within topics such as punishment, necessity, and causality; yet these different treatments constitute a coherent account. The relevant discussions of which we have written records are dated between 1272 and 1276.52 Siger’s theory of the will and free decision is based upon a metaphysical principle he attributes to Avicenna: Every effect that comes forth [evenit], comes forth from a cause with respect to which its being is necessary, as Avicenna says. (DNCC, p. 14 ll. 9–11)53

Siger argues that if a cause did not produce its effect necessarily, then the cause would still be in potentiality to causing and would require something else that actualizes its causality; once actualized, it cannot but produce its effect. So whenever an effect is actually produced, it comes from a cause that acts necessarily in the moment it acts (DNCC, pp. 14–15 ll. 11–21). Siger puts this idea also as follows: The proposition of Avicenna holds universally true, namely that every cause that is in the condition [in dispositione] in which it is natural for it [nata est] to produce its effect, necessarily produces its effect. This is what Aristotle intends in Book VIII of the Physics, that when the mover is in the condition in which it is natural for it to move, and the movable in the condition in which it is natural for it to be moved, this must [oportet] move and that be moved. (QMet.Ca VI.9, p. 320 ll. 11–16)54

The will is no exception to Avicenna’s principle. The will is always moved to its volition when its cause is in the condition in which it is natural for it to move the will, so that it is impossible for the will not to be moved in that moment (DNCC, p. 34 ll. 45–57). Thus, what causes the will’s volition does so necessarily. Siger connects Avicenna’s principle to Aristotle’s doctrine of the necessity of the present: it is not necessary that a certain event be the case, but when it is the case, it is necessary that it be the case (Int. 9, 19a23– 52 53

54

For the dating of Siger’s work, see the introductions to the critical editions and Duin 1954, ch. 3. See also Siger of Brabant, Impossibilia 5, pp. 86–7; QMet.Ca VI.9, p. 317 ll. 6–7; p. 320 ll. 11–13; p. 323; QMet.Vi VII.1, pp. 377–8 ll. 38–58. Cf. Avicenna, Liber de philosophia prima sive scientia divina I.7, I: 46 ll. 69–71. The Quaestiones in Metaphysicam exist in four student reports (reportationes), which according to Dunphy derive from at least three different lecture cycles; see Siger of Brabant 1981, 21–5. They are commonly referred to by indicating the libraries holding the manuscripts that contain them: Munich, Cambridge, Paris, and Vienna. I refer to them by adding the initial letters, e.g., QMet. Ca for the Cambridge report. The Paris report is most probably by Godfrey of Fontaines; see Siger of Brabant 1983, 10–11. See also QMet.Vi VII.1, p. 378 ll. 53–8. Siger refers to Aristotle, Phys. VIII.1, 251b1–5.

56

Free Will

5). Accordingly, it is not necessary that a certain effect be produced, but if its cause is so disposed to produce its effect, and nothing impedes it, then at that moment it produces the effect necessarily (DNCC, p. 32 ll. 1–14; QMet. Ca VI.9, p. 321 ll. 49–52). So too, when someone has a certain volition (appetitus), it is necessary that he or she have this volition (QMet.Ca V.37, p. 271 ll. 43–5; QMet.Vi V.8, p. 331 l. 38). It would seem, then, that the will wills everything it wills necessarily, and that everything one does, one does necessarily (Impossibilia 5, p. 86 ll. 25–6). Yet according to Siger, the Avicennian principle is no threat to free decision (arbitrii libertas) or the freedom of the will (libertas voluntatis). In fact, there are two kinds of necessity relevant to the present issue: absolute necessity and conditional necessity. Correspondingly, there are two kinds of causes: those that cannot be impeded and those that can be impeded.55 If all effects came from nonimpedible causes, there would be no free decision, “for then every willing of ours would be caused by a cause that the will could not resist” (DNCC, p. 33 ll. 31–9). But as a matter of fact, “such is the nature of the will, that each of the things that are apt to move the will can be impeded from moving [it]” (DNCC, pp. 34–5 ll. 63–4). How does Siger understand that the will can resist a cause? Does the will do so by posing an obstacle to impede what causes its willing?56 Siger nowhere says so.57 It would be wrong to conceive of the will’s resistance to a cause as a veto power,58 as if, once the necessary and sufficient conditions for willing are given, it is up to the will to let itself be moved or not. Siger insists that one should not understand the freedom of the will as though the will were the first cause of its willing and able to move itself to alternatives.59 Rather, the will is moved to its act by cognition, that is, by a judgment of reason. The will wills according to the way it is moved: a thing judged good entails desire, a thing judged bad entails aversion. It is precisely because the will is moved by cognition that its freedom is not threatened; if it willed apart from cognition, then it would not be free.60 Accordingly, Siger explicitly rejects the view of those who hold that, while one has a particular judgment that this is good or evil, one can still choose

55 56 57 59 60

Impossibilia 5, p. 90 ll. 17–27; DNCC, p. 15 ll. 34–7; QMet.Ca VI.9, p. 321 ll. 54–6; QMet.Vi VII.1, pp. 378–9 ll. 66–06; p. 381 ll. 79–81. Lottin 1957a, 264 advances this interpretation. He is followed by Hissette 1977, 234 and Ryan 1983, 168, 171, 174, 185. Thus Siger’s solution would coincide with Aquinas’s in DM 6 ad 15. As was already noted by Putallaz 1995, 25. 58 Pace Ryan 1983, 191. DNCC, p. 34 ll. 50–3, QMet.Ca VI.9, p. 325 ll. 82–90; QMet.Vi VII.1, p. 386 ll. 61–6. Q De causis 25, p. 102 ll. 61–5. DNCC, pp. 34–5 ll. 54–5, 65–6; QMet.Ca V.36, p. 269 ll. 21–5; QMet.Vi VII.1, p. 386 ll. 61–7.

The Psychological Turn

57

either way (QMet.Vi V.8, pp. 330–1 ll. 34–9). In other words, Siger affirms judgment-volition conformity. The important distinction for Siger is not between intellect and will, but rather between sense and intellect. The will necessarily follows the particular judgment of reason, but what is crucial, according to Siger, is that the cognition that moves the will is intellectual, not sensory cognition. While the judgment of the senses is determined by nature to perceive certain things as pleasant and others as unpleasant, intellectual judgment is by nature undetermined; Siger calls it free judgment. Tacitly evoking John of Damascus’s sourcehood passage, Siger says that nonhuman animals are acted upon in pursuing an end, whereas humans act.61 Not surprisingly, then, according to Siger we become good or bad by means of good or bad judgments (QMet.Vi V.8, p. 331 ll. 49–51). Siger need not appeal to an autonomous motion of the will as impeding the necessary action of the judgment of reason on the will. Rather, it is through deliberation that one can impede the cause that, if unimpeded, necessarily causes one’s volition. Just so, taking poison necessarily causes death, but by taking an antidote, one can impede its effect.62 We can put Siger’s point as follows: by means of deliberation, we can reconsider our motivations, that is, we can revise our judgment that, if unrevised, necessarily results in a determinate volition. Aquinas has a similar account, except that Aquinas explicitly assigns to the will the control of whether the intellect deliberates, whereas Siger does not. As we shall see in Section 4.3, Godfrey of Fontaines takes a quite similar approach to free decision as Siger. But first we must turn to initial reactions to intellectualist theories of free will such as Aquinas’s and Siger’s and examine endeavors to propose alternative accounts. 61 62

DNCC, p. 35 ll. 64–73; QMet.Ca V.37, p. 271 ll. 18–41; QMet.Vi V.8, p. 330 ll. 18–33. DNCC, pp. 33–4 ll. 31–45; QMet.Ca VI.9, p. 326 ll. 93–6; QMet.Vi VII.1, p. 380 ll. 43–7, p. 386 ll. 48–50.

chapter 3

Voluntarism and the Condemnation of Intellectualism

On March 7, 1277, three years to the day after Aquinas’s death, the Bishop of Paris, Stephen Tempier, condemned 219 philosophical propositions, some of which could be seen to reflect teachings by Aquinas. Though significantly promoting voluntarism, the condemnations did not put the free will debate to rest; quite the contrary. During this period, the exchange of arguments was largely oral and hence very rapid. The main fora for the debate were “quodlibetal disputations,” public university events held each year in Lent and Advent (the liturgical seasons prior to Easter and Christmas). The theology master presiding over a quodlibetal disputation would receive questions from anyone (a quolibet) – students or colleagues, followers or adversaries – about anything (de quolibet). There are written records of a large number of these oral disputations, many of which are revised versions published by the masters themselves, not long after they delivered the disputations.1 I will present the debate topically rather than strictly chronologically, discussing voluntarism (Chapter 3), intermediary accounts and strict intellectualism (Chapter 4), and the reprise of these approaches a generation later (Chapter 5). Voluntarist tendencies, especially an emphasis on the will’s ability to diverge from reason, are a common phenomenon prior to the reception of Aristotle.2 But voluntarism as a systematic moral psychology presupposes the psychological turn kindled by Aristotelian ideas.3 Full-blown voluntarism, then, develops rather late in the game; nevertheless it is a conservative theory akin to twelfth-century theories of free will. Gerard of Abbeville, for example, who belonged to the conservative faction of the Parisian theology masters, defended some core voluntarist ideas.4 Probably the first to 1 2 4

Hamesse 2006 gives a helpful overview of the nature of quodlibetal disputations and their origin and demise. See, e.g., Perkams 2012. 3 See Kent 1995, 111–12. See Quodl. XVIII.2–3, ed. Pattin, 111–20, denying that the will is a passive power and affirming the will’s per se self-motion.

58

Voluntarism

59

formulate systematically the central claims of a voluntarist conception of free will was Bonaventure’s student and confrère Walter of Bruges in his Disputed Questions of 1267–9.5 At that time, cosmological and psychological determinism suddenly became a topic of general concern; in 1268, Bonaventure mentions determinist views as a threat to the Christian faith, along with the doctrines of the eternity of the world and of the unicity of the intellect for all humankind, while the year before he had mentioned only the latter two doctrines.6 This chapter portrays voluntarism in the late thirteenth century. We will start with Bishop Tempier’s censures of philosophical positions in 1270 and, more importantly, in 1277. Then we will study Henry of Ghent’s theory of free will in some detail, which is greatly influenced by Walter of Bruges. To avoid redundancy, we will not study Walter separately, though points of agreement between Walter and Henry will be mentioned.7 Henry is the greatest protagonist of the Paris free will debate during his long tenure there as theology master, and he also had a massive impact on late thirteenth- and early fourteenth-century theologians. Among the Franciscans active in the late thirteenth century, Peter Olivi deserves attention for his originality and his strong influence within his order. The chapter ends with a brief mention of other prominent Franciscan thinkers, especially William de la Mare, the author of a handbook of corrections to Aquinas’s “errors.”

3.1 The 1270 and 1277 Censures In 1270, Bishop Tempier condemned thirteen philosophical propositions. Two of these directly concern free will: That the human will wills and chooses of necessity. (Article 3) That free decision is a passive power, not an active power, and that it is moved necessarily by the desirable object [ab appetibili]. (Article 9)

Another condemned proposition concerns free will indirectly: That everything that happens here below is subjected to the necessity of the celestial bodies. (Article 4)8

5 6 7 8

For the dating of Walter’s Disputed Questions, see Decorte 1983, 216, note 6. Bonaventure, Collationes de septem donis Spiritus Sancti 8 n. 16, V: 497b; cf. Putallaz 1995, 7 and 129; Müller 2009, 548. For Walter of Bruges’s influence on Henry of Ghent’s theory of free will, see Decorte 1983. On Walter’s theory of free will, see also Kent 1995, 59–68, 117–23, and 175–7, and Kobusch 2006. Chartularium Universitatis Parisiensis, I: 487.

60

Free Will

Tempier did not mention any individual by name, but the intended target seems to be Parisian arts masters; in fact, some of the condemned propositions can be found in pre-1270 writings of Siger of Brabant, although not in writings that bear on free will.9 In January 1277, Pope John XXI, having heard that “errors that damage the Catholic faith are sprouting again,” requested from Bishop Tempier that he investigate by which persons and in which places such errors are maintained, and that he report back to the Pope.10 Bishop Tempier set up a commission of Parisian theology masters to conduct the investigation. The commission, whose members included Henry of Ghent, hastily scrambled together a rather disjointed list of 219 “errors.” Instead of reporting back to the Pope, the Bishop took the matter into his own hands and proceeded to condemn these “errors” on March 7.11 The censured propositions fall into subjects ranging from metaphysics to philosophical anthropology, ethics, moral psychology, and beyond. This time, Tempier wrote an introductory statement that clarifies, without mentioning names, that the condemnation is aimed at arts masters. As a matter of fact, Siger of Brabant and Boethius of Dacia are among its main targets. While it is debated whether this condemnation was also directed against Aquinas, it is clear that some of his views are at least incidentally affected.12 The human will is the direct concern of nineteen of the condemned propositions, several of which have to do particularly with its relation to reason and to the desired object.13 In addition, a few condemned propositions concern the human will indirectly inasmuch as they imply the denial of contingent causes.14 As I mentioned in the introduction, there are also numerous propositions dealing with angels or “intelligences,” but none of these has to do directly with angelic free will. 9 10 11 12

13

14

On the 1270 condemnation and their historical background, see concisely Wippel 2002, 65–7 and in greater detail Hille 2005. Chartularium Universitatis Parisiensis, I: 541. For a thorough study of the doctrinal background of the condemned theses, see Hissette 1977. For Henry of Ghent’s involvement in the drafting of the articles, see Wielockx 2011, 25–6. Hissette (1977 and 1982) argues, in part based on Tempier’s introductory statement, that Aquinas was not targeted by the March 7 condemnation, but Hissette concedes that certain condemned propositions express Aquinas’s views. By contrast, Wippel 1995 argues that Aquinas was directly targeted in the articles expressing his views, backing his claim among other things with the testimony of theologians of the time. Furthermore, Wielockx has repeatedly argued (e.g., 2011, 32–4) that Tempier aimed at Aquinas only in a separate investigation against him that was never brought to conclusion, but according to Wippel 1997 and Thijssen 1998, 52–6, Wielockx’s evidence for such an investigation seems to refer to the process against Giles of Rome, to be discussed below. See also Putallaz 2010, 101–3. Articles 129–36, 158–9, 161–5, 173, 194, 208–9, ed. Piché (Stephen Tempier 1999). Mandonnet 1908– 11, II: 175–91 provides an edition that rearranges the articles thematically; according to his numbering, the articles cited here are 151–69. Piché’s edition adds Mandonnet’s numbers in parentheses. Articles 21, 128, 160, 197.

Voluntarism

61

The condemnation had a massive influence on the subsequent free will debate. Well into the fourteenth century, theologians mention some of its articles, often for polemical purposes. Because of the condemnation’s ongoing effect, in 1325, after Aquinas’s 1323 canonization, the Bishop of Paris considered it necessary to revoke articles “to the extent that they touch on or are said to touch on the doctrine of blessed Thomas.”15 Of the condemned propositions concerning the human will, several affirm that the will is moved by another: by the celestial bodies or spheres16 or by the desirable object.17 Others assert the necessitation of the will by reason or cognition.18 The most discussed articles in the free will debate of the subsequent decades concern the relation between reason and will. The propositions these articles censure express judgment-volition conformity in different ways: As long as passion and particular knowledge remain in act, the will cannot act against it. (Article 129) If reason is right, the will too is right. – Error, because it is against Augustine’s gloss on this Psalm-verse “My soul longed to desire” etc., and because according to this statement, grace would not be necessary for the rectitude of the will, but only knowledge, which was the error of Pelagius. (Article 130) That the will of a human being is necessitated by his or her cognition like the appetite of a nonhuman animal. (Article 159) The will necessarily pursues what is firmly believed by reason, and it cannot abstain from what reason dictates. Yet this necessitation is not coercion, but rather the nature of the will. (Article 163)

Other oft-cited articles concern the relation between the will and its mover: That the will, when it exists in the condition in which it is natural for it to be moved, and when that for which it is natural to move [the will] remains so disposed, it is impossible that the will not will.19 (Article 131) That the appetite, when the impediments are lifted, is necessarily moved by the desirable object [ab appetibili]. – Error concerning the intellective [appetite]. (Article 134)

15

16 19

Chartularium Universitatis Parisiensis, II: 280–1. For the impact of the 1277 condemnation on the philosophical culture of the following decades, see Bianchi 1999, 203–30 and Aertsen, Emery Jr., and Speer 2001. Articles 132, 133, 162; cf. 161. 17 Articles 131, 134, 135, 194, 208. 18 Articles 129, 130, 158, 159, 163. The translation of the articles is based upon Piché’s edition (Stephen Tempier 1999).

62

Free Will That the will is as such undetermined, like matter, concerning opposites; but it is determined by the desirable object, like matter by the agent. (Article 135) When two goods are proposed, the stronger moves [the will] more strongly. (Article 208)

Article 131 clearly evokes Siger of Brabant’s teaching (see Chapter 2, pp. 55–6). Article 208 expresses an issue that Henry of Ghent debated a few months before the condemnations, as we will see shortly. Another censure, later in March 1277, impacted the free will debate at least indirectly: Tempier condemned 51 propositions taken from Book I of the Sentences commentary of Giles of Rome. By that time, Giles had fulfilled the requirements for a master of theology, but this censure prevented his promotion. In preparing the list of articles, Tempier relied again on the Parisian theology masters for the doctrinal investigation. Among the censured propositions, two very similar ones express proportional deficiency in reason and will: There is no evil [malitia] in the will unless there is error in reason. (Article 24) There is never evil [malitia] in the will unless there is error or at least some lack of knowledge [nescientia] in reason.20 (Article 51)

Giles refused to recant, presenting instead a defense of these propositions (Apologia) – to no avail. In 1285, Giles appealed to the newly elected Pope Honorius IV, indicating his willingness to retract the condemned propositions. The Pope ordered Tempier’s successor, bishop Ranulph of Houblonnière, to determine in agreement with the theology masters of the University of Paris which propositions Giles was to retract.21 Giles was promoted to a master in the same year.22 In the course of the reexamination of the articles in 1285, the theology masters conceded the proposition originally condemned in Article 24. Without committing to a particular interpretation – whether deficient reason corrupts the will or vice versa – they were ready to admit its truth at face value. Immediately after its concession, the statement “There is no evil in the will unless there is

20 21 22

Giles of Rome, Apologia, ed. Wielockx, 54, 59. Chartularium Universitatis Parisiensis, I: 633. See Wielockx 1985, 75–113; Putallaz 1995, 209–15; Briggs 2016, 9–10.

Voluntarism

63

error in reason,” referred to by the masters as propositio magistralis, began to play an important role in the disputations about free will.23

3.2 Henry of Ghent Voluntarist conceptions of free will tend to have a remarkable inner coherence. Henry of Ghent, a secular master (a priest, but not belonging to a religious order), develops such a conception in a set of Questions in his first Quodlibet of Advent 1276, where the principal components of his theory of free will are already present.24 In frequent later discussions, Henry will expand on the most controversial positions already taken here and defend them against critics.25 Henry develops some of his signature voluntarist views in contrast with Aquinas, among others; and yet these views are framed by ideas he shares with Aquinas and has in part adopted from him.26 The Shared Framework of Aquinas and Henry Most fundamentally, Henry agrees with Aquinas that the proper object of the will is the good, not evil. He assumes like Aquinas complete symmetry between love (that is, willing: desire, choice, or enjoyment) and goodness: only what is good can be loved, and every good is loveable.27 To will something, it need not be truly good, it only needs to appear to be good. When willing evil, one wills it under the aspect of the good (sub ratione boni), and one can reject evil only under the aspect of evil.28 Henry follows Aquinas almost to the letter in arguing that the will has an ultimate end, which it desires necessarily and for the sake of which it wills 23

24 25 26 27 28

On the propositio magistralis, see Wielockx 1985, 77–81, 91–2, 105–10, and Hödl 1999. On the authority of a statement by John of Pouilly, Wielockx argues that the propositio magistralis was conceded already when the commission of masters met in 1277 to examine Giles’s works and only reconfirmed when they met in 1285 to reexamine his case. But Hödl 1999 argues that the evidence Wielockx adduces does not warrant the early dating of its concession and that the early dating does not explain why there was total silence about it until 1286 and massive discussion of it from 1286 onward. In a forthcoming publication on Giles of Rome, Stephen Dumont adduces compelling evidence against dating the Magisterial Proposition’s concession to 1277. For the dating of Henry’s Quodlibeta, see Porro 2006, 174–8; for the dating of the articles of Henry’s Summa in relation to the Quodlibeta, see Laarmann 1999, 50–2. For recent studies of Henry’s theory of free will, see Putallaz 1995, 177–208, Hoffmann 2008, Teske 1996 and 2011, Pickavé 2012, and Müller 2018. Aquinas’s and Henry’s shared general approach to free will is also highlighted by Alliney 2017. Quodl. IV.11c., VIII: 89 ll. 21–5; cf. p. 42 above. See, e.g., Quodl. I.16c., V: 106–7 ll. 67–87; ibid. V: 110 ll. 46–7; Quodl. XI.7c., 459rN; Quodl. XII.26c., XVI: 153 ll. 19–21; and the extensive discussion throughout Quodl. XIII.9, XVIII: 56–64; cf. p. 42 above.

64

Free Will

everything else that is related to the end. This ultimate end is the absolute good (bonum simpliciter).29 The absolute good understood under a general notion is happiness; as a concrete beatifying thing it is God.30 So everyone loves happiness necessarily. The blessed, who enjoy the direct vision of God, understand him to be the beatifying good, and hence they love God necessarily.31 Just as the will is essentially ordered to the good, so too freedom is essentially related to the good. Like Aquinas, Henry subscribes to Anselm’s view that the ability to sin does not belong to the essence of freedom.32 Henry not only shares these fundamental assumptions with Aquinas, but he follows him also in a terminological choice: freedom of the will (libertas voluntatis) is the kind of freedom that is compatible with necessity of immutability, while free decision (liberum arbitrium) is not; free decision is the more particular kind of freedom that presupposes choice among alternatives.33 Moral responsibility requires not merely free will, but also free decision and hence the ability to do otherwise.34 Following Aquinas closely, Henry sees the indetermination of free decision, that is, its openness to alternative possibilities, as being grounded in the immaterial nature of the intellect. Specifically, he appropriates two of Aquinas’s considerations: the intellect enables one to self-determine the actions done for an end, through knowledge of the end, of the means to the end, and of the relation between the means and the end.35 Furthermore, the intellect grounds free decision by providing universal knowledge. Henry even takes his examples from Aquinas: whereas all swallows build their nests in the same way, an artisan, who understands a chest under a universal notion, is not determined either to build it or not, or to build it in a triangular or round shape.36 Finally, like Aquinas, Henry identifies the will throughout his writings with John of Damascus’s notion of rational appetite. Yet for him, unlike for Aquinas, this notion does not express a strict dependence of the will on 29 30 31 32 33 34 35 36

Quodl. III.17 ad 1, 78vH–79rH; Summa 47.5c., XXX: 24–5 ll. 37–51; Quodl. XIII.9c., XVIII: 61 ll. 5–23; cf. p. 42 above. Quodl. XIII.9c., XVIII: 58–9 ll. 50–74; see also Quodl. IX.5c., XIII: 137 ll. 7–8. Quodl. III.17 ad arg., 79rH; see also Quodl. IV.22c., VIII: 354 ll. 104–10; Quodl. IX.5c., XIII: 135 ll. 38–43; Quodl. XII.5c., XVI: 30–4; cf. p. 49 above. See also Osborne 2013. Quodl. III.17 ad arg., 79rH; Quodl. XIV.5c., 564rY; cf. pp. 41–2 above. See especially Summa 45.4c., XXIX: 124–5 ll. 87–26; see also, e.g., Quodl. X.9 ad argg., XIV: 236 ll. 16–22; cf. p. 41 above. Quodl. I.16c., V: 95 ll. 30–8; Quodl. IX.5c., XIII: 133 ll. 79–81; cf. p. 41 above. Summa 45.3c., XXIX: 112–13 ll. 74–83; cf. p. 44. Summa 45.3c., XXIX: 113–14 ll. 1–31; cf. p. 43 above. For the example of swallows building their nests, see Aquinas, DV 24.1c., XXII: 681 ll. 273–4.

Voluntarism

65

reason. And this makes all the difference: Henry’s principal disagreement with Aquinas’s theory of free will regards the relation between intellect (or reason) and will, as is clear already in a series of four Questions about intellect and will in his first Quodlibet (qq. 14–17).37 A Compact Voluntarist Account of Free Will: the First Quodlibet No voluntarist thinker would deny that reason is crucial for the will. For Henry, the will, as rational appetite, requires rational cognition, not merely sensory cognition. In the insane (in amentibus), who are incapable of rational cognition, there is no will but only sensory appetite (we might say: they act purely instinctively) (Quodl. I.15c., V: 93 ll. 58–65). Yet according to Henry the will is superior to reason, inasmuch as “the will is the universal and first mover in the whole kingdom of the soul” (Quodl. I.14c., V: 85 ll. 45–8). According to Henry, reason directs the will only in the way a servant guides the master by carrying a lantern before him so that he may not trip. In terms of authority, however, the will directs reason, commanding it to consider, to reason, and to deliberate – when the will wills and about what it wills – and likewise to cease these acts. Reason, conversely, does not command the will at all.38 The idea that command is an act of the will, not of reason, and hence that the will is superior to reason, is already present in Bonaventure, as we have seen (Chapter 2, p. 40), and it is a recurrent theme among voluntarist thinkers. For the imagery of the kingdom of the soul and the lantern-carrier, Henry depends on Walter of Bruges, whose influence is palpable throughout these four Questions. The remote origin of the kingdom metaphor lies in William of Auvergne, who about four decades earlier likens the will that commands the acts of the other powers to a king and emperor, the intellect to the king’s counselor, and the sensory appetite and bodily limbs to his ministers.39 Henry’s understanding of free will comes out most clearly in response to a question that seems inspired by Walter of Bruges: When the will is presented with a greater and a smaller good, can it choose the smaller 37 38

39

On Henry’s account of free will in Quodlibet I, see Müller 2007, Hoffmann 2008, and Müller 2009, 581–98. Quodl. I.14c. and ad 5, V: 85–6 ll. 47–57; V: 90 ll. 44–50. On “command” as an act of the will rather than reason, see also Quodl. IX.6, XIII: 139–49, discussed in Macken 1995. On the presuppositions and implications of Henry’s metaphor of the lantern-carrier, see Müller 2018. Walter of Bruges, QD 6c. and ad 1, PhB X: 60–2; William of Auvergne, De virtutibus 9, Opera omnia I: 122aF–G; De anima II.15, Opera omnia II: 85a; De anima III.8–9, II: 95a–96b. Cf. Teske 1994.

66

Free Will

good? (Quodl. I.15–19 intr., V: 91 ll. 9–10).40 The immediate issue here is the dependence of the will’s choice on an object presented by the intellect. But Henry argues that the deeper issue is the foundation of freedom in intellect and will: “The whole force of the question turns on this: does the freedom in question here reside principally in the intellect or in the will?” (Quodl. I.16c., V: 98 ll. 88–9). Henry develops his answer in contrast with a view whose basic ideas are shared by Aquinas, although the wording of Henry’s presentation does not quite fit any of Aquinas’s writings; the hypothesis has been advanced that Henry argues against a text by Siger of Brabant of which we have no record.41 According to this view, the will’s inclinations follow necessarily what has been determined by reason, and humans have free decision precisely because the will’s inclinations result from reason; for deliberation can have different outcomes, whereas the inclinations of nonhuman animals, which are determined by nature, can have only one outcome. Henry comments that, according to this account, free decision would be entirely grounded in reason and not in the will, except insofar as the will depends on reason. Then the will would not be able to choose the smaller good but would necessarily choose the greater good (Quodl. I.16c., V: 98–101 ll. 90–5742). Henry is concerned that this position, which traces freedom of the will to reason, implies cognitive necessitation. Instead, Henry argues that in order to safeguard free will and to account for an evil will one must posit, over and above the freedom to judge in reason, a freedom to choose what is judged in the will. Accordingly, freedom of the will is not grounded in reason alone, but also in the will. Furthermore, Henry argues that the will must be free to act contrary to the judgment of reason. Can the will act against reason’s judgment while reason still holds that judgment, or only after reason has revised it? Intellectualists allow for the second hypothesis, but Henry rejects it, for then the will would be no freer than the appetite of a nonhuman animal, which desires differently only when the senses propose different desirable objects. Instead, for Henry, when reason proposes two objects and judges that one is better, the will can act contrary to that judgment and choose the other. Thus he denies judgment-volition 40

41 42

Cf. Walter of Bruges, QD 6 ad 14, PhB X: 64–5. The same question is contained in an Ethics commentary that one of the manuscripts, Città del Vaticano, Bibliotheca Apostolica Vaticana, Vat. lat. 832, attributes to Egidius. This made some scholars think that it was by Giles of Rome and that it inspired the question put to Henry; see Macken 1977, 129–31. But its author is Radulphus Brito, and its q. 72 repeats that of Henry’s Quodl. I.16; see Iacopo Costa’s introduction to his edition of Brito’s Ethics commentary, pp. 26–7 and 99. Putallaz 1995, 189–91. Cf. Aquinas, ST I.80.1; I.82.2; I.83.1, referred to by Lottin 1957a, 276, note 2. Line numbers in the edition restart at zero after 99, so the second number can be smaller than the first.

Voluntarism

67

conformity. A virtuous choice, however, presupposes that one chooses what reason judges better (Quodl. I.16c., V: 102–4 ll. 85–39; see also Summa 45.4c., XXIX: 123 ll. 49–52). By acting against reason’s judgment, the will nevertheless wills a good known by reason, for a good can either be chosen as it is “simply cognized” apart from deliberation, or as it is judged preferable upon deliberation.43 While the good judged as preferable has a stronger hold on the will than the simply cognized good, it does not determine the will, which remains free to will either good (Quodl. I.17c., V: 125 ll. 16–34). (For example, an attractive proposal may seem less attractive upon further thought; nevertheless, one can accept it.) Why would judgment-volition conformity threaten free decision? For Henry, except insofar as reason is under the command of the will, reason is not free, for it does not control its acts. It cannot avoid understanding simple objects of thought (such as a horse, a stone, a triangle). It also cannot help assenting to self-evident propositions (such as “every whole is greater than its part”). As to nonevident propositions, reason necessarily adjusts its belief or doubt in proportion to the evidence at hand – except if the will, within certain limits, commands otherwise. For Henry, to the extent there is freedom in reason, it is derived from the freedom of the will – not vice versa, as Henry’s anonymous adversary claims.44 Because reason is passive about what it knows and subject to the will’s control, he later calls its acts of knowing and judging servile (Summa 45.4c., XXIX: 124 ll. 70–80). The source of the will’s freedom is not in reason, then, but rather innate to the will.45 When completely spelled out, Henry’s argument is this: reason (prescinding from the will’s control) cannot judge otherwise; but – on the assumption that humans have free decision – the will can choose otherwise; hence the will’s choice cannot be simply derived from the judgment of reason and it must be possible for the will to choose differently from reason’s judgment.

43 44

45

Quodl. I.17c., V: 125 ll. 18–20; see also Summa 45.4c., XXIX: 123 ll. 67–9 and Quodl. X.10c., XIV: 256–7 ll. 32–48. Quodl. I.16c., V: 107–8 ll. 96–05. Later, Henry clarifies that the will cannot make reason dissent from the conclusion of a manifest demonstration (Quodl. IX.6c., XIII: 144 ll. 35–9), but when something merely appears to be true or false, then the will can make reason dissent or assent to it, because of its control (dominium) and freedom (Quodl. XI.7c., 459rP). Quodl. I.16c., V: 108 ll. 5–6. A few years later, attacking Aquinas’s view to the contrary (ST I-II.17.1 ad 2), Henry develops a larger argument that the source or “root” of freedom is in the will and not in reason and that reason is free only by participating in the will’s freedom; see Quodl. IX.6c., XIII: 146–7 ll. 93–18.

68

Free Will

For Henry, judgment-volition conformity would not only destroy free decision, but it would also imply that an evil will is caused by erring reason. Henry addresses this problem extensively in his response to Question 17: Is a disorder in the will caused by an error in reason, or vice versa? (Quodl. I.15–19 intr., V: 91 ll. 11–12). For Henry, a disordered will is a sinful will making perverse use of free decision (Quodl. I.17c., V: 115 ll. 15–16). In his answer, Henry claims that his adversary, who affirms judgmentvolition conformity, would have to say that the order or disorder of the will proceeds from the order or disorder of reason, so that the will can only become disordered if reason has already been disordered (Quodl. I.17c., V: 124 ll. 96–00). The opening argument had made the same claim, appealing to the authority of Aristotle’s account of incontinence (Quodl. I.17 arg., V: 115 ll. 4–10). But for Henry, this cannot be so. He offers two lines of reasoning against tracing a disordered will to a flaw in reason. One concerns nonculpable ignorance, which does not cause sin but rather excuses from sin, as a biblical episode illustrates: Lot’s two daughters, realizing that they could not obtain offspring from other men, inebriated Lot so that he would sleep with them. Lot then had intercourse with them on subsequent nights without realizing it (Genesis 19: 32–5). Henry argues that if Lot had indeed no fault in getting drunk, then his ignorance was nonculpable and the incestuous act committed out of ignorance was no sin (Quodl. I.17 ad arg., V: 139–40 ll. 51–72). Another line of reasoning presupposes an assumption made earlier, that a disordered act is a sin only if it is voluntary (Quodl. I.16c., V: 113 ll. 9–15, quoting Augustine, uera rel. 14.27). Henry here aims to show that sin’s voluntariness implies that the disordered act of the will cannot be determined by reason or the senses, but rather must be self-determined by the will. To make his case, Henry discusses at length the case of incontinence and clarifies the causal relation between the passions inclining to sin, the corruption of reason, and the corruption of the will. He argues that in the incontinent, the passions entice their will to freely consent to evil, and only then does the disordered will make their reason err, rather than the passions first clouding reason and thereby compelling their will to desire in accord with disordered reason (cf. Quodl. I.17c., V: 119 ll. 96–00). In the latter case, the individual’s control would be bypassed, an implication Henry wants to avoid. According to Henry, that sensory passions do not determine the will can be seen in the counterpart of the incontinent, namely the continent. These desire certain pleasures by their sensory appetite, but by their will they do not act upon the sensory inclinations, but rather follow their intellect (ibid., 122 ll. 44–9). Reason’s cognition does not determine the

Voluntarism

69

will either. As mentioned above, cognition can be a “simple apprehension” of reason prior to deliberation, or a judgment or verdict (sententia) of reason after deliberation. That the verdict of reason after deliberation does not determine the will has already been established in Question 16, to which Henry here refers (ibid., 121 ll. 39–43). Simple cognition prior to deliberation does not determine the will, because it is itself indeterminate, as Henry argues with reference to Aristotle’s remark that sciences encompass contraries (Met. IX.2, 1046b10–11). (For simple cognition’s indeterminately encompassing contraries, recall Henry’s example that the idea of a chest does not determine whether one builds it or not.)46 Hence a will that would be bound to follow cognition that is indeterminate about contraries would do contrary acts simultaneously. Instead, what determines the outcome, as Aristotle attests in the translation Henry adduces, is not reason, but desire or the will.47 Neither the passions, then, nor simple apprehension, nor the verdict of reason determine the will. And yet, as Henry reminds us, Aristotle claims that the desirable object is the first mover (cf. An. III.10, 433b15–18 and p. 28 above). How, then, shall we understand the cognized good’s motion of the will? According to Henry, Aristotle left this question open (Quodl. I.17c., V: 122–3 ll. 50–71). Henry’s own answer is this: the cognized good is not a cause of willing in the proper sense, but rather “determines the will to will as that without which it cannot will anything.” Yet the cognized good does not necessitate the will’s act (ibid., 125 ll. 16–32). Instead, the will moves itself to its act. Augustine confirms this: the will becomes slave to lust only because of its own motion, which is voluntary rather than natural. According to Henry, the fact that the motion of the will is voluntary means that it would be against the nature of the will to be determined by reason; rather, it is in the nature of the will to be able to move freely even against reason’s determination.48 The first sin is further confirmation that the will becomes disordered by itself, rather than through a disorder of reason. Adam’s intellect was not subject to error prior to the first sin, since fallibility is a punishment for sin and therefore does not precede the first sin, as was commonly held on the 46 47

48

See note 36 above. Quodl. I.17c., V: 120–1 ll. 23–39, citing Met. IX.5, 1048a10 according to the Arabo-Latin translation (see Averroes 1969, 43 ll. 15–16). Henry also cites Averroes’s commentary on the passage; see Averroes 1969, 45 ll. 69–71. For Aristotle’s account of irrational and rational active powers, see Chapter 1, pp. 26–7. Quodl. I.17c., V: 126–8 ll. 48–91. Henry quotes Augustine, lib. arb. III.1.2.8–9. See also Quodl. I.16c., V: 109 ll. 36–41.

70

Free Will

authority of Augustine.49 Yet Adam sinned; so “it is necessary to admit that this first man sinned without any preceding error in reason, but rather by acting directly against the judgment of right reason” (Quodl. I.17c., V: 128–9 ll. 5–8). The same must be the case for us, since our will is not substantially different from Adam’s, only weaker than his (ibid., 128 ll. 92–5; p. 129 ll. 8–11). Thus, on the authority of Augustine, Henry defends a thesis that is just the opposite of the Socratic deficiency thesis according to which deficient willing presupposes deficient cognition: Augustinian deficiency thesis: Deficient cognition presupposes deficient willing.

Not only does deficient cognition presuppose deficient willing, but for Henry, a disordered will causes reason to err (Quodl. I.17c., V: 129 ll. 12–15). This doctrine plays an important role in his subsequent disputations; later he will formulate it with the Augustinian notion of evil “blinding” the intellect. How does the will corrupt reason? A disordered will can cause reason to be ignorant because it neglects to move reason to study what is necessary for salvation. It can also cause reason to err by failing to prevent its making rash judgments (ibid., 118 ll. 73–80; cf. p. 126 ll. 37–9).50 Recall that the will commands reason’s acts: it can move reason to consider something or to deliberate, and it can stop reason from doing so. A meticulously detailed and rather forced exegesis of Aristotle’s account of incontinence confirms for Henry that passions entice the will to a perverse consent before reason is corrupted (Quodl. I.17 ad arg., V: 130–50). In this context, Henry repeats the claim that the will corrupts reason, rather than vice versa; for example, the will makes reason believe that it is good to commit adultery. Aristotle’s adage that every evildoer is ignorant does not mean that ignorance causes sin, but rather that sin causes ignorance (ibid., 146 ll. 00–6). Henry clarifies that although causally the disorder in the will is prior to the error in reason, chronologically it is simultaneous (ibid., 147 ll. 17–19). Ten Core Theses of Henry’s Voluntarist Approach to Free Will The principal claims of Henry’s voluntarist account of free will are already present in his first Quodlibet. We can single out nine theses that are 49 50

Lib. arb. III.18.52.177–8, quoted in Quodl. I.17c., V: 116. In a later discussion, Henry will claim that every error in reason is caused by the will; see Quodl. XII.25c., XVI: 132–3 ll. 21–36. The similarity with Descartes’s Fourth Meditation is striking.

Voluntarism

71

characteristic of voluntarist approaches to free will; they are found already in Walter of Bruges and are routinely repeated by later Franciscans, and they differ in various degrees from Aquinas’s moderate intellectualist account. Henry adds a tenth thesis that is peculiar to him. (1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

The will is a higher power than reason.51 Command is an act of the will, not of reason.52 The will is an active power, not a passive power.53 By itself, reason is not free.54 The source (principium) of freedom is in the will and not derived from reason.55 The will moves itself.56 The willed object is (only) a necessary condition for the will’s act.57 One can abstain from, or will contrary to, what reason judges best.58 Willing evil presupposes no cognitive deficiency.59 A disordered will necessarily corrupts reason.60

Henry abides by these theses throughout his career and defends them against attacks, especially from Godfrey of Fontaines, who disagrees with the first nine theses (see Chapter 4, p. 111). The corruption of practical knowledge through evil (thesis 10) can be understood in different senses. 51

Quodl. I.14c., V: 89 ll. 16–17. Cf. Walter, QD 6 ad 19, PhB X: 66. By contrast, see Aquinas, ST I.82.3. Quodl. I.14c., V: 85–6 ll. 47–57. Cf. Walter, QD 6c. PhB X: 60–1. By contrast, for Aquinas, command is an act of reason which presupposes an act of the will in virtue of which reason commands (ST I-II.17.1). 53 See Quodl. I.16c., V: 96 ll. 60–2 for the claim that the will is an active power. Against the claim that the will is a passive power, see Summa 45.2c., XXIX: 107 ll. 82–3; Quodl. X.9c., XIV: 229 ll. 50–3. Cf. Walter, QD 4 arg. 3 and ad 3, PhB X: 34, 41. For Aquinas’s contrary view, see p. 48 above. 54 Quodl. I.16c., V: 107–8 ll. 96–7. Later Henry will concede that the intellect is free, albeit less free than the will; see Quodl. XIV.5. Also Walter considers the will to be freer than reason; see QD 5 arg. 15 and c., PhB X: 50, 51. Aquinas, too, at first considered the will freer than the intellect (In Sent. II.7.2.1 ad 2, II: 187–8; In Sent. II.25.1.2c., II: 649), but later effectively describes them as equally free (ST I.82.2c.). 55 Quodl. I.16c., V: 108 ll. 5–6. Cf. Walter, QD 5c. and ad 13 in opp., PhB X: 52, 54. For Aquinas’s contrary view, see pp. 42–5 above. 56 Quodl. I.17c., V: 127 ll. 76–7. Cf. Walter, QD 4c., ad 1, ad 5, PhB X: 40, 41, 42. Aquinas, too, attributes self-motion to the will, but not per se self-motion, as Henry does; see pp. 75–6 below. 57 Quodl. I.17c., V: 125 ll. 16–17. Later, Henry will call the causal role of the known object a causa sine qua non; see p. 80 below. Cf. Walter, QD 5c., PhB X: 52–3; QD 6c., PhB X: 60. For Aquinas’s different view, see pp. 51–2 above. 58 The will is not bound to follow reason but rather can abstain: Quodl. I.16c., V: 102 ll. 85–90; cf. Walter, QD 5c., PhB X: 52, QD 6c., PhB X: 60. The will can act contrary to the judgment of reason: Quodl. I.16c., V: 105–6 ll. 59–62; cf. Walter, QD 5 ad 5 in opp., PhB X: 54, QD 6 ad 10, PhB X: 63. For Aquinas’s contrary view, see pp. 45–8 above. 59 Quodl. I.17c., V: 128–9 ll. 4–11. Cf. Walter, QD 4 ad 11, PhB X: 45; QD 6 ad 14, PhB X: 65. For Aquinas’s contrary view, see pp. 46–8 above. 60 Quodl. I.17c., V: 129 ll. 12–15. 52

72

Free Will

That habitual evil (vice) corrupts practical knowledge is a commonplace accepted by all; that one evil volition immediately corrupts reason, as Henry thinks, is an unpopular view even among voluntarists. What provokes the most discussion is the question of what causes a volition: only the “cognized object” (the desirable object made present to the will by the intellect), or the cognized object and the will, or the will alone? What is at issue here are the self-motion of the will, judgmentvolition conformity, and the Socratic deficiency thesis (cf. theses 6–8). Henry’s principal later developments to his theory regard three themes: the idea of free will as self-determination, the explanation of the metaphysics of self-motion implied in self-determination, and the theory of an evil will causing cognitive deficiencies. We now turn to these developments. Self-Determination and the Essence of Freedom When Aquinas discussed the will’s acts, he focused on the relation between the will and the willed object. This perspective brought to light that while most things are desired contingently, others – God, happiness, and what is understood to be indispensable for happiness – are desired necessarily. For Henry, this is a legitimate perspective on the will, but not the crucial one. According to Henry – whom Scotus will follow on this point – a different perspective on the will’s act is more fundamental: to consider the “mode of eliciting [modus eliciendi] the act of willing.” From this perspective, the issue is what causes the will to elicit its act and whether the will elicits it contingently or necessarily (Quodl. III.17c., 78vG; Summa 47.5c., XXX: 25–6 ll. 59–73). This perspective clarifies the sourcehood conditions of free will, and Henry sees these differently than Aquinas does. Aquinas traced free will to free judgment; accordingly, besides requiring the absence of external coercion, he stressed cognitive requirements for sourcehood: to know the end as end, the means as means, and the relation between the means and the end (Chapter 2, p. 44). In his view, causal determination of the will as such poses no threat to sourcehood, as long as it is not determination by efficient causality (the case of God’s motion of the will aside). In fact, for Aquinas the will is determined by the judgment of reason. Determination need not be necessitating: one can revise the judgment that determines the will. But determination can be necessitating: Aquinas explains the necessary desire for happiness as a determination by nature, insofar as the will has a natural inclination toward it (DV 22.5c. and ad 2), and he explains it also as determination by cognition, since happiness is seen as good from every perspective (DM 6c., XXIII: 150 ll. 429–35).

Voluntarism

73

As we have seen, Henry agrees that freedom requires knowledge of an end, of the means, and of the relation between means and end.61 But he holds that any form of determination is incompatible with the will, and hence with free will. The only determination of the will is self-determination. Against Aquinas,62 he writes that when the will determines itself to its act, it does so not because something impels it violently from without, nor because something inclines it naturally from within or leads it slavishly, but rather because it acts freely and by choice or as if by choice [libere et eligibiliter aut quasi eligibiliter] and wills this by itself, being master [dominus] of its act. (Summa 45.3c., XXIX: 115 ll. 52–4)

Henry does not deny necessary acts of the will: God’s self-love and the love of God by the blessed are necessary, for in both cases, the willed object is the absolute good and it is adequately known (Quodl. III.17 ad 1, 79rH, 79rK; Summa 47.5c., XXX: 25 ll. 43–51). But these acts are not necessary because of determination, only because of selfdetermination. For Henry, this is fundamental for the nature of the will and the essence of freedom. He makes this clear especially in discussing divine self-love. God’s will not only has his essence necessarily as its object, but it also necessarily elicits its act of love. By what sort of necessity does it elicit its act? Henry excludes “necessity of violence or of coercion,” because God cannot be subject to violence, for what is violent has its source outside the agent.63 He quotes Averroes: “the will is delightful, necessity of coercion is distressing [contristabilis].”64 Thus Henry connects the nature of will (and hence its freedom) to delight. Peter Auriol will go a step further: he will identify freedom with delight (see Section 5.4). While in God’s willing there is no necessity of coercion or violence, there is necessity of immutability. But for Henry it is crucial to make a further distinction: necessity of immutability can be conceived as antecedent or as concomitant to the will. If it is antecedent, then the will is subject to this necessity; if it is concomitant, then the necessity is subject to the will. Antecedent necessity of immutability is natural necessity. If God loved himself by this necessity, then his act would be determined like the 61 62 63 64

See note 35 above. Henry of Ghent, Summa 45.3c., XXIX: 115 ll. 32–41, critiquing Aquinas, DV 24.2c., XXII: 685 ll. 70–81. Summa 47.5c., XXX: 26–7 ll. 80–100, where he refers to Summa 30.6c., 184vD, in which he cites Aristotle’s dictum that the violent is that which has its source outside (EN III.1, 1110a1). Averroes, In Met. V com. 6, ed. Ruggero Ponzalli, 100 ll. 49–50 (which comments on Aristotle, Met. V.5, 1015a26–30). See also Quodl. III.17c., 78vG.

74

Free Will

appetite of nonhuman animals; God would not act, but be acted on, and so he would not be free – an allusion to John of Damascus’s sourcehood passage. Concomitant necessity of immutability, by contrast, strengthens (firmat) freedom, for the will eliciting its act by that necessity acts “with delight and as if by choice” (Summa 47.5c., XXX: 27–8 ll. 101–23). The point is that the divine will imposes freely the concomitant necessity of immutability on itself, and it does so because of the strength (firmitas) of its freedom. Henry further articulates this idea in the statement that “it is necessary that God freely wills himself” (Summa 47.5c., XXX: 29–31 ll. 142–90). Thus he can affirm that God necessarily loves himself without undergoing any determination, either by his own nature, or by his own essence as cognized by the intellect: When we say that God necessarily wills himself, this necessity does not originate in the divine nature insofar as it is nature, nor from some determination of the will from the willed object, but only from the mode of freedom in his will. (Summa 47.5c., XXX: 32 ll. 203–6)65

In addition to lack of determination, Henry stresses the relation of freedom to the good: freedom is the ability to adhere to the good that is the ultimate end, and free decision to adhere to goods conducive to the ultimate end; but the ability not to adhere to the good, or to adhere to evil, is a defect of free decision (Quodl. XII.5 ad 2, XVI: 34 ll. 81–91). These considerations are reflected in Henry’s account of the ratio libertatis in a discussion of human freedom: The essential character of freedom [ratio libertatis] consists in the fact that no coercion can prevent the will from inclining itself to the good, if it wants – and not in this alone, but also that human beings do not will what they will by an impetus of nature that precedes the motion of the will and that impels the will to its act, in the way that nonhuman animals desire, but rather insofar as desire and delight [complacentia] of the will are antecedent and the natural impetus concomitant to the will. (Quodl. III.17 ad 1, 79rI; cf. Summa 47.5 ad arg. in opp., XXX: 34 ll. 235–9)66

Clearly, then, Henry has a distinctive understanding of the sourcehood required for free will. Not only must the act originate from within the 65

66

See also Summa 45.3c., XXIX: 119 ll. 46–7: “not by an impetus of nature nor by some determination of the object”; Summa 45.4c., XXIX: 126 ll. 31–5: “not necessarily by a quasi-antecedent necessity, nor by the determination of its nature, nor by the determination of the intellect.” Toward the end of his career, Henry discusses at length the notion of freedom and summarizes the discussion with a definition that is in agreement with these earlier ones; see Quodl. XIV.5c., 565rB and the discussion by Pickavé 2012.

Voluntarism

75

person; it must also originate from within the will, without any determination from nature or cognition. Thus natural necessity – that is, necessity rooted in a natural cause rather than in the will – is incompatible with the will.67 It was common to distinguish nature and will as two fundamental and distinct principles, but Henry goes further: he thinks that what is moved naturally is not free.68 The sharp division between nature and will is going to be at the center of Duns Scotus’s and Ockham’s theories of free will, as we will see in Sections 5.1 and 5.5. The Metaphysics of Self-Motion It is clear, then, that for Henry, unless the will moves itself to its act, there is neither free decision that includes the ability to do otherwise, nor free will as experienced in the will’s necessary acts.69 But the will’s self-motion poses a significant theoretical challenge. In his ninth Quodlibet (of 1286, a few years after Articles 45–7 of the Summa) Henry discusses the will’s selfmotion for the first time in depth, and then returns to it in each of his following Quodlibeta, except for his fifteenth and last one.70 Although he was unable to convince his greatest rival, Godfrey of Fontaines, other critical minds, such as Duns Scotus, adopted the key points of his theory. Henry understands the will’s self-motion as per se self-motion. The will moving itself per se means that it reduces itself from potentially willing to actually willing.71 By contrast, the will as an incidental (per accidens) selfmover would only contribute incidentally to the actualization of its willing, just as sailors move themselves only incidentally when they move a ship that is primarily moved by something else (by the wind, for example) (Quodl. XIII.11 ad 3, XVIII: 126 ll. 86–91). Recall that Aquinas does not allow for per se selfmotion of the will. According to Aquinas, in virtue of actually willing an end – for example, health – the will reduces itself from potentially willing the means to the end to actually willing them (Chapter 2, p. 49). Henry disagrees 67 68 69 70 71

See, e.g., Summa 45.3 ad 2, XXIX: 121 ll. 99–02; Summa 47.5 ad 2, XXX: 33 ll. 225–30; Quodl. XIII.6c., XVIII: 40–1 ll. 79–93. See, e.g., Quodl. XIV.5 ad 2, 566rD: “It would be strange to say that something is a natural active power, and yet free.” Quodl. IX.5c., XIII: 137 ll. 3–8; see also ibid., p. 127 ll. 94–16, pp. 130–1 ll. 3–9, p. 132 ll. 44–54. Quodl. IX.5; X.9; XI.6; XII.26; XIII.11; XIV.2. For Henry’s account of the will’s self-motion, see also Macken 1977 and Teske 1996. Quodl. X.9c., XIV: 225 ll. 26–30; Quodl. XI.6c., 452vM; Quodl. XII.26c., XVI: 154 ll. 46–7; Quodl. XIII.11 ad 3, XVIII: 125 ll. 69–71; Quodl. XIV.2c., 549vB; cf. Quodl. IX.5c., XIII: 131–2 ll. 35–9. That Henry understands the will’s self-motion as per se rather than as incidental self-motion is clear throughout his discussions of the topic.

76

Free Will

with Aquinas on three counts: first, Henry rejects the view that the will can move itself to a volition only in virtue of a prior volition.72 Second, while Aquinas thought that when the will begins to will something, such as when it begins to want the intellect to deliberate, the transition from not willing to willing requires divine intervention (Chapter 2, p. 50), for Henry, such intervention is unnecessary, since the will can move itself from not willing to willing (Quodl. IX.5c., XIII: 121 ll. 18–25). Third, in Aquinas’s view, when the will moves itself from willing an end to willing specific means to the end, it needs to be further actualized by the intellect that specifies which means promotes the end (e.g., wanting to achieve health by taking medicine). By contrast, as we will see below, according to Henry, the cognized object does not cause the will’s act. As we have seen, from an Aristotelian point of view, per se self-motion is highly problematic: there are self-movers in a broad sense – living things, especially animals – but no per se self-movers, for per se selfmotion implies that the same thing is in actuality and in potentiality (and hence not in actuality) simultaneously and in the same respect. There can be self-movers only if the moving parts are distinct from the parts being moved, as is the case in living things (Chapter 1, p. 29). So in order to avoid the implication that the self-moving will is in actuality and potentiality in the same respect, Henry needs to distinguish within the will the mover from what is moved – even though the will is not composed of parts. Furthermore, the mover within the will must be able to move the will to a volition without having to be moved from without to elicit this particular volition. A presupposition of Henry’s in addressing these difficulties is that selfmotion is less problematic in spiritual than in material things, for spiritual things are self-reflexive (super se conversiva), allowing them to act on themselves (Quodl. IX.5c., XIII: 137 ll. 9–17; Quodl. X.9c., XIV: 233 ll. 54– 69). Spiritual things require less of a distinction between mover and moved than material things (Quodl. IX.5 ad 1, XIII: 138 ll. 29–39). Henry claims that when the “philosophers” reject self-motion, they refer to material things, and so the motion principle that “everything that is moved is moved by another” applies above all to these. As we will see momentarily, Henry even allows for self-motion in material things

72

The context is a discussion of Anselm’s view that the will’s moving itself to a new volition requires a previous volition of wanting to move oneself to a new volition; see Quodl. VIII.10c., 320vN–321rO and Quodl. XII.26c., XVI: 155–6 ll. 78–93, discussed in Chapter 9 below, pp. 223–4.

Voluntarism

77

like stones, but he thinks self-motion is present in the will in a more proper sense than in these.73 Henry argues that mover and moved need not be really distinct; a conceptual distinction or “intentional distinction” suffices: the will is in potentiality insofar as it is simply a nature, and in actuality insofar as it is free.74 According to his mature solution, this is understood as follows: the will contains in its nature a formality – the “form of freedom” – in virtue of which it is a mover and in actuality. The form of freedom is analogous to the form of heaviness by which a heavy object is in accidental potentiality to falling down, and will fall down as soon as the impediment is removed.75 The form of freedom enters into composition with the will taken generically as an appetite, and so mover and moved are sufficiently distinct to avoid the will being in actuality and in potentiality in the same respect.76 The will received the form of freedom, and thus its actuality, through its creation. In virtue of its form of freedom, the will can move itself from potentially willing to actually willing an object, without needing to be actualized by the willed object. The only requirement is the removal of the obstacle to the will’s selfmotion. This obstacle is ignorance, that is, the absence of a desirable object. When the object becomes present through the intellect’s cognition, then the will either moves itself necessarily, but freely (vis-à-vis the ultimate end), or it can move itself or not, by free decision (vis-à-vis what is related to the end).77 Put differently, the will is always either actually willing or in accidental potentiality to actually willing. When in accidental potentiality, it becomes actually willing (if it wants, as is implied) when the impediment is removed (Quodl. XIII.11 ad 4, XVIII: 133 ll. 54–64). Henry frequently refers to three analogues to exemplify the will’s selfmotion. One is a thought experiment taken from Walter of Bruges (QD 4 ad 5, PhB X: 42). The sun is “formally” cold, yet it is “virtually” hot, which means that it is cold in itself (as medieval thinkers thought), but causes heat. So in a sense the sun is in actuality regarding heat in that it can cause heat without needing to be further actualized for doing so. If the sun could receive in itself the heat it generates, then it would move itself from being virtually hot but formally cold to being formally hot.78 73 74 75 76 77 78

Quodl. X.9c., XIV: 234 ll. 70–87; Quodl. XIII.11 ad 3, XVIII: 131–2 ll. 28–36; cf. Quodl. IX.5 ad arg. in opp., XIII: 138–9 ll. 40–67. Quodl. IX.5 ad arg. in opp., XIII: 138 ll. 51–3; Quodl. X.9 ad argg., XIV: 231 ll. 5–15. Quodl. XIII.11 ad 3, XVIII: 130–1 ll. 5–19; cf. p. 129 ll. 79–80; Quodl. XIV.2c., 549vB. See also Quodl. X.9c., XIV: 225 ll. 26–35. Quodl. XIII.11 ad 3, XVIII: 126–7 ll. 4–30; pp. 128–9 ll. 51–80; see also p. 133 ll. 48–51. Quodl. X.9c., XIV: 225–6 ll. 26–46; Quodl. XIII.11 ad 3, XVIII: 131 ll. 13–24. Quodl. XIII.11 ad 3, XVIII: 131 ll. 24–8; Quodl. XIV.2c., 549vB.

78

Free Will

Henry sees a second analog to the will’s self-motion in Aristotle’s account of the motion of heavy and light things, interpreted with the help of Averroes (see Chapter 1, pp. 29–30). According to Henry, a stone that is impeded from falling down by a column on which it rests is in a sense in actuality, thanks to its form of heaviness. In Henry’s view, heaviness is an accidental form that enters into composition with the substantial form of the stone; so for Henry, the mover (heaviness) and the moved (the stone) are in a sense distinct.79 According to Henry, the form of heaviness relates to the stone as the form of freedom relates to the will. Once the impediment is removed, the stone falls to its natural place, in virtue of its form of heaviness and without requiring further actualization. The stone has the form of heaviness, and hence its actuality, from what generated it.80 As we have seen, however, Aristotle examined the stone’s downward motion precisely to show that the stone’s free fall is not a case of self-motion; instead, in his view, its essential mover is what generated it and its incidental mover is what removes the obstacle. A third analogue is warm water cooling down. The form of coldness in the water is the mover, the water is what is moved, and the obstacle is the source of heat, the removal of which allows the water to move itself to coldness (Quodl. XI.6c., 453rP, 454rE). In sum, for Henry the form of freedom is the mover of the will, while the will itself is what is moved from potentially willing to actually willing, when the obstacle, ignorance, is removed. So too, the sun’s property of being virtually hot is the mover, while the sun itself is hypothetically what is moved from being formally cold to formally hot. Likewise, the form of heaviness is the mover, the stone itself is what is moved from potentially falling to actually falling down, when the support of the stone is removed. Finally, water cools by itself when heat, the obstacle, is removed. In all of these cases there is self-motion; but only in the case of the will does it occur freely. In this account, the will is never simultaneously in potentiality and in actuality in the same respect. Before moving itself to a volition, the will is simultaneously in potentiality and actuality, but in different respects: in actuality regarding “first actuality,” that is, insofar as it exists and has received from its creator the force (vis) to move itself to be actually willing as soon as the obstacle is removed; at the same time, the will is in 79 80

Quodl. XIII.11 ad 3, XVIII: 126 l. 95; Quodl. XIV.2c., 549vB. Quodl. XIII.11 ad 3, XVIII: 130–1 ll. 5–13. See also Quodl. X.9 ad 5, XIV: 232–3 ll. 32–53; Quodl. XIV.2c., 549vB. Henry’s most detailed account of the movement of heavy and light things is found in Quodl. IX.5c., XIII: 115–19 ll. 66–95.

Voluntarism

79

potentiality regarding actually willing something, and in this sense it is in potentiality to “second actuality.” Expressed differently, the will is made by its creator virtually willing, whereby it is in potentiality to become formally, that is, actually willing (Quodl. XII.26c., XVI: 154–5 ll. 56–65). Henry’s best explanation is in his latest treatment of self-motion: The act by which the will is virtually in this state [of willing] is neither opposed to the [actual] volition, nor to the potentiality for it; but rather, it is simultaneous to both, for it is of a different genus. And the will bringing itself from potentiality to actuality regarding the act of volition is the action of an equivocal agent, just as the sun generates a man or heat. (Quodl. XIV.2c., 549vB)81

If the will were a self-mover as a univocal cause (the effects of which are specifically the same and hence equal in being to the cause, as when a man generates a man), then the will would move itself to its first actuality, that is, it would produce itself from nothing. But the will is an equivocal cause (the effects of which are inferior to the cause, as when the sun generates fire). Therefore the will, which is in actuality in one sense (insofar as it is in first actuality and virtually willing), can move itself from potentiality to actuality in another sense (insofar as it is in second actuality and actually or formally willing).82 The Causality of the Cognized Good As was clear already in Henry’s first Quodlibet, his account of the will’s self-motion has an important implication for the role of the desirable object, that is, the “cognized good,” and of the intellect presenting it. For Aristotle, the cognized good is the first mover that moves the appetite (see Chapter 1, p. 28), that is, the will (as the medievals understood Aristotle). In his later writings, Henry clarifies the causality of the cognized good. For Henry, of course, it does not properly move the will to its volition – the will alone moves itself. The cognized good moves the will only “metaphorically,” in the way that an end, such as health, “moves” someone to pursue it.83 Henry says the same about the intellect or reason 81 82

83

For a brief statement of the difference between virtually and formally willing, see also Quodl. X.9 ad 2, XIV: 230–1 ll. 85–8; ad 3, p. 231 ll. 90–3. Already earlier than in Quodl. XIV.2, Henry briefly appealed to the distinction between univocal and equivocal causation for the explanation of self-motion and explained its relation to the formal/ virtual distinction; see Quodl. XI.6c. and ad 8, 454rL–vL, 458rE. Quodl. XIII.11 ad 3, XVIII: 105–6 ll. 41–62, expounding Aristotle, GC I.7, 324b13–15. See also Quodl. I.17c., V: 127 ll. 61–2; Summa 45.2c., XXIX: 107 ll. 70–5.

80

Free Will

presenting the object.84 He admits, however, that the cognized good causes the volition – not as an acting cause, a causa propter quam sic (a “cause because of which indeed”), only as a causa sine qua non (a “cause without which not”).85 While in a sense the cognized object as causa sine qua non causes the volition, it “effects nothing” (nihil agit) in the will; its only effect is removal of ignorance.86 The notion of “causa sine qua non” was generally understood not as an acting cause, but as the cause that guarantees the conditions for the acting cause to act; for example, by removing an obstacle.87 Henry believes that his theory of the cognized good as sine qua non cause of the will’s act renders Aquinas’s exercise/specification distinction obsolete. Henry argues that if determination (or specification) simply consists in the manifestation (ostensio) of an object, then it comes down to his own idea of causa sine qua non. If determination causes the will to be strongly inclined to an act, then the will retains the freedom to act against it, and the will moves itself to its act. Finally, if determination produces the volition, then the will’s volition is necessitated, and hence it is necessitated in its very exercise – in which case the will is no freer regarding its exercise than regarding its determination (Quodl. IX.5c., XIII: 122–4 ll. 51–08). In fact, if one holds that the will’s acts are dictated by reason, then they are dictated regarding both exercise and determination, and the will is free in neither respect (Quodl. X.9 ad argg., XIV: 239: 98–00). Henry repeats analogous considerations against Giles of Rome, who proposes a distinction that Henry equates with Aquinas’s exercise/specification distinction (see Section 4.1): for Henry, exercise and determination come down to the same thing, and so, unlike in Giles’s theory, Henry does not hold that the will is moved in some respects and self-moving in others; rather, the will moves itself regarding all its acts, including its necessary acts (Quodl. XII.26c., XVI: 150–4 ll. 52–47). The Blinding of the Intellect Revisited To conclude the survey of Henry’s theory of free will, we will now consider an important development in Henry’s account of the blinding of the intellect (deficient willing causing deficient cognition). Henry attaches much importance to the blinding of the intellect and affirms it also in his 84 85 86 87

Summa 4.1 ad 2, XXI: 273 ll. 82–5; Quodl. I.14 ad 2, V: 89 ll. 24–34. Quodl. XII.26c., XVI: 152–3 ll. 6–18. Quodl. X.9 ad argg., XIV: 249–50 ll. 56–61; Quodl. XIII.11c., XVIII: 88 ll. 28–36. For the notion of causa sine qua non and its historical background, see Solère 2014, 195–203.

Voluntarism

81

account of angelic sin (see Section 9.3). In refining his account, Henry makes an important innovation regarding the relation between willing and time that is influential on Duns Scotus’s theory of free will. As already mentioned, in 1285 the Parisian theology masters, including Henry of Ghent, granted the truth of what came to be known as the propositio magistralis: There is no evil in the will unless there is error in reason.88

At face value, this seems to be a restatement of the Socratic deficiency thesis, as, for example, Aquinas affirmed it (see Chapter 2, p. 46): evil in the will presupposes error in reason, and as long as reason is correct, the will cannot become evil. But this claim was condemned in Articles 129 and 130 of the March 7, 1277 condemnation: As long as passion and particular knowledge remain in act, the will cannot act against it. (Article 129) If reason is right, the will too is right. . . . (Article 130)

Henry addresses the compatibility of the propositio magistralis with Article 129 (and Article 130, which he treats as equivalent to Article 129) in Quodlibet X of Advent 1286, Questions 9, 10, and 13. For Henry, resolving the apparent incompatibility involves two issues: causality, that is, whether a disorder in reason causes an evil will or vice versa, and simultaneity, that is, whether reason can be intact while there is evil in the will.89 The gist of Henry’s solution to the causality problem consists in the idea he had already put forth in Quodlibet I, Question 17: evil in the will does not presuppose, but rather causes error in reason. The immediate effect of an evil will, however, is not full-blown error, but rather the darkening (obscuratio) of reason, which means that reason sways from the correct judgment, holds it less firmly, or doubts about it. With reason darkened, the will sins more amply and more freely, leading to a yet greater evil in the will and then to an even greater darkening, until the will is perfectly obstinate and reason fully blinded. The result is the denial of principles of morality and of faith: after thinking that theft is evil and that God exists, one arrives at maintaining that theft is good and that God does not exist (Quodl. X.10c., XIV: 259–63 ll. 93–68). Thus by nature and causally, there is evil in the will before reason errs – contrary to the view of those who 88 89

See note 20 above. For a lucid and detailed analysis of Henry’s reconciliation, see Dumont 1992.

82

Free Will

“follow on this issue the prophecy of Aristotle in Book VII of the Ethics,” where Aristotle holds that incontinence presupposes a momentary cognitive deficiency, “dismissing the doctrine contained in the holy books and expounded by the saints” (ibid., p. 261 ll. 24–8). Nevertheless, temporally, evil in the will and error (that is, darkening) in reason are simultaneous (ibid., ll. 29–44). So we have the intended result: the statements condemned in Articles 129 and 130 are false and the propositio magistralis is true. Upon a closer look, the simultaneity of evil in the will and error in reason poses a major problem. The statement censured in Article 129 essentially affirms that while reason is free from error, the will cannot act against reason. Thus Article 129, that is, the condemnation of this statement, implies that the will can act against flawless reason, in which case the will is evil. The propositio magistralis, in Henry’s understanding, affirms that reason cannot be free from error when the will is evil. In order to reconcile the propositio magistralis with Article 129, one would have to admit that in the moment one begins to have an evil will, one is still without error and yet already in error. Henry bites the bullet and in point of fact allows these contradictory states of affairs to be simultaneously the case. In Quodlibet X, Question 13, which picks up a long-winded and ultimately inconclusive discussion of Question 10, Henry solves the simultaneity problem by means of two devices. First, he now explicitly expands the meaning of the term “error.” Henry writes that, strictly speaking, error is a judgment that is contrary to a principle (e.g., judging that taking a certain thing from another is good when in fact theft is bad). This is an all-or-nothing state of the mind. But Henry’s expanded meaning of error includes the idea of a mere darkening of reason (Quodl. X.13c., XIV: 287–8 ll. 41–5, 60–1). This broad sense of error allows Henry to assume that although there is only a gradual change in reason when the will begins to be evil, reason is immediately in error. Second, and more importantly, he splits the instant of time (instans) in which evil enters the will and in which, as we have seen, in Henry’s account a person is in two contradictory states, into two “marks” (signa). What justifies in Henry’s eyes the distinction of two marks in one instant is the assumption that by nature (although not in time), the will is first evil before reason is in error; in other words, it expresses ontological priority. Thus if at t1 a person’s reason is unflawed – literally, “at the pinnacle of rectitude” (in apice rectitudinis) – and evil has not yet entered his or her will, and if at t2 the person’s will becomes first evil, then at t2 the person’s reason also falls into error. At t2, the individual is in two contradictory mental states: his or her reason is unflawed and flawed – undarkened and darkened. Reason is

Voluntarism

83

still unflawed, because the evil will acts against correct reason (as Article 129 declares possible); reason is already flawed, because there is no evil in the will without error in reason (as the Magisterial Proposition affirms). So the instant of time t2 splits into two marks: t2A in which reason is unflawed, and t2B in which it is darkened (Quodl. X.13c., XIV: 288–9 ll. 55–83).90 In these discussions of Quodlibet X, Henry does not say how his account fits with the principle of noncontraction. The standard understanding of this principle is that contradictory states of affairs cannot be true at the same time. Henry apparently thinks in Quodlibet X that this principle merely requires that contradictories are not both true in the same mark of nature, but it allows that they are both true at the same instant of time (which splits into two marks) – a position no one seems to have held before Henry.91 In his next quodlibetal session, in 1287, the opponent (who is in charge of advancing arguments against the master’s view) attacks Henry’s interpretation of the propositio magistralis because of the implication that reason would be simultaneously “at the pinnacle of rectitude” and darkened, unflawed and flawed (Quodl. XI.6 arg. 9, 452vL). In his response, Henry explains the darkening of reason by an evil will without assuming that reason is in contradictory states in the same instant of time. He claims that as such (per se et simpliciter), the will can act against unflawed reason in any instant of time and even in any mark of nature, since it lies in the nature of the will to be able to do so. Yet incidentally (per accidens) the will cannot effectively (cum effectu) go against reason at the instant of time or the mark of nature in which this would imply contradictory states of affairs, precisely because of this impossible implication (Quodl. XI.6 ad 9, 458vL). Contradictories would be implied in the hypothetical instant in which the will would effectively act against unflawed reason, for to act against unflawed reason is sinful (as is presupposed), and this sin would instantaneously deflect reason (fol. 458rH). Thus at the very instant the will acted against unflawed reason, reason would be unflawed (ex hypothesi) and flawed (as an immediate effect). One wonders, then, whether there ever is an instant in which the will can effectively act against unflawed reason. Henry thinks there is, during the whole period prior to going against unflawed reason: And so I say unqualifiedly that at any given instant, the will can go against reason while reason is at the pinnacle of rectitude, but not with an effect in 90 91

Analogous difficulties are involved in Henry’s account of other cases of instantaneous change; see Knuuttila 2017 and Dumont 1992, 571–7. See Dumont 1992.

Free Will

84

that very instant, but only during the whole time before; and at the instant in which the will wills the opposite, reason is deflected, even though not totally blinded. (Quodl. XI.6 ad 9, 458rH–vH)

By assuming that acting contrary to unflawed reason is possible only prior to actually doing so, Henry is in point of fact talking about a possibility that can never become actual, for it is never available in the present, at the instant of action. But earlier, Henry had assumed that there are cases in which someone’s will effectively acted against unflawed reason: the sins of Adam and Lucifer (see pp. 69–70 above and Section 9.3). Worse, as John of Pouilly remarked twenty years later, Henry’s revised solution infringes anew on the principle of noncontradiction: The fact that the will can will the opposite at the instant in which the rectitude [of reason] is at the pinnacle implies that this can come about [potest poni in esse] and that it is possible; yet the fact that at the instant in which the rectitude is at the pinnacle this cannot be with the effect implies that it is not possible. Thus the same [state of affairs] is possible and not possible; it can and it cannot come about. (Quodl. I.10c., ed. Hödl, 292)92

Ironically, Henry, one of the principal critics of judgment-volition conformity, ends up denying that the will can effectively act against unflawed reason. His account of an evil will instantaneously blinding reason seems contrived, and it was his least successful doctrine concerning free will.

3.3

Peter Olivi

Henry of Ghent’s tenure as master of theology largely coincides with the scholarly activity of Peter Olivi, a highly original Franciscan who was prevented from having a career adequate to his intellectual stature. Olivi received his early formation in the lectorate program of the Parisian Franciscan convent from about 1266 until 1271, and his presence in Paris is still attested in 1273. Thus he was exposed to the thought of Walter of Bruges, John Pecham, and Roger Bacon, among others. For his further studies he was not sent again to Paris, as a bright mind could have expected, nor did he pursue studies to become a master of theology elsewhere. In fact, Olivi was repeatedly the target of censure, the most important of which during his lifetime was by his own order, taking effect in 1283. Olivi was 92

According to John of Pouilly, Henry violated the principle of noncontradiction also in his earlier account of Quodl. X.10 and X.13, where Henry allowed contradictory states to coexist in the same temporal instant; see Pouilly, Quodl. I.10c., ed. Hödl, 289; Quodl. II.11, concl. 6, A 195ra, P 49vb.

Voluntarism

85

rehabilitated in 1285 and later sent to lecture in the Franciscan study houses of Florence, Montpellier, and Narbonne. In 1299, a year after his death, John of Morrovalle (also known as John of Murro), then minister general of the Franciscans, ordered Olivi’s writings to be confiscated and burned. Several popes, too, questioned his orthodoxy; most notably he was condemned by John XXII in 1326. What caused concern were not Olivi’s views regarding free will, but rather regarding poverty, the sacraments, the divine nature, some particular philosophical issues, and posthumously also his apocalyptic speculations.93 His account of free will, by contrast, largely represents the mainstream among Franciscans and is also unaffected by the Parisian 1277 condemnations. As a lecturer in provincial convents rather than a master at Paris, Olivi was not part of the free will debate at the principal center of learning. Though the target of censures, he continued to be read, but caution was in order. For example, Duns Scotus mentions him by name in his early Oxford lectures on the Sentences as the author of a certain view about personhood, but purposely omits the description of this view; in the corresponding passage of his Ordinatio, that is, the version intended for publication, Scotus leaves out even the mention of Olivi.94 Olivi proposes his theory of free will above all in his Summa, also known as Questions on the Sentences (as distinct from his Sentences commentary), in which he collected Disputed Questions held earlier in his career. While most of this work is lost, Book II, which is relevant for our purposes, has survived. (Unless otherwise mentioned, all references to Olivi are to Book II of his Summa.) Olivi shares the fundamental ideas about free will with Henry of Ghent; this can be explained by Walter of Bruges’s influence on both, but also by direct knowledge of Henry’s Quodlibetal Questions. But Olivi and Henry weigh certain topics differently. For example, Olivi treats the self-motion of the will as unproblematic, probably in part because he was not pressed on it in public debates, as Henry was. Likewise, as an outsider to the Parisian faculty of theology, the propositio magistralis is of no concern to Olivi, and Tempier’s March 7, 1277 condemnations are left unmentioned, although he may well allude to them.95 In general, Henry focuses on issues pertaining to moral psychology. In this regard, Olivi also makes important contributions, offering, for example, a profound analysis of how free acts are grounded in the intellect’s and the will’s capacity for self-reflexivity (e.g., 57 ad 19, II: 364; 59c., II: 552–3). But 93 94 95

On Olivi’s life and his condemnations, see Burr 1976, in part updated by Piron 2006, 410–13. Piron 2020a and 2020b offer a chronology of Olivi’s works and provide additional details about his life. See Duns Scotus, Lect. I.26 n. 42, XVII: 327 and note 2 of the editors. Whitehouse 2016 argues at length that Olivi alludes to several of its articles in Summa II.57.

86

Free Will

what he cares most about is freedom as a prerequisite of key aspects of human life and as constitutive of the human person – issues that Olivi treats in greater detail than any of his contemporaries.96 The Significance of Freedom According to Olivi, human experience shows that we have free decision. It is a commonplace among medieval thinkers to argue that without freedom, there is no praise or blame, merit or demerit, counsel, and so forth. Olivi expands considerably the list of human phenomena that show that we are free and discusses them in great detail. For example, by our free decision we have the kind of control without which numerous human affections or attitudes would be fictitious: anger and mercy, glory and shame, friendliness and hostility, gratitude and ingratitude, domination and subjugation, hope and diffidence, foresight and negligence or sluggishness (57c., II: 316–23). Even certain mental acts, such as inquiry and reflection, attest to the existence of free decision, because they presuppose that we freely control them (ibid., 323). Laws and the goods of friendship and association, too, presuppose freedom (ibid., 335–8). We even have direct experience of our freedom: we feel that we control our acts rather than being determined to act one way or the other (ibid., 327). Olivi cites a view (with which he is sympathetic) according to which we feel that our acts proceed from us rather than from things acting on us (58 ad 14, II: 463–4). According to Olivi, freedom is at the core of what we are; in fact, the idea of perpetually renouncing our free decision is as abhorrent to us as being annihilated (57c., II: 335). In conclusion, he makes a famous statement concerning the error of denying free decision: This error destroys every human good, and even every divine good. . . . No wonder, for, if I may say, it takes from us what we truly are, namely our personhood [personalitas], and does not grant us anything more than that we are intellectual beasts, or beasts with intellects. (57c., II: 338)

What is remarkable in this statement is not only the powerful image of “intellectual beasts,” but still more Olivi’s affirmation that freedom is constitutive of personhood. Boethius famously defined “person” in terms of rationality: “the individual substance of a rational nature.”97 Olivi defines it 96

97

For Olivi’s account of free will, see Simoncioli 1956, Stadter 1971, 144–237, Kent 1995, 84–8, 129–37, Putallaz 1995, 127–62, Pasnau 1999, Putallaz 2000, Yrjönsuuri 2002, Normore 2007a, Bobillier 2017, Kent 2017, Szlachta 2018, and Perler 2020, chs. 5–6. Boethius, Contra Eutychen et Nestorium 3, ed. Moreschini, 214 ll. 171–2.

Voluntarism

87

in terms of freedom: personhood (personalitas) is “a ruling [dominativa] and free per se existence” (52c., II: 200; see also 51c., II: 120–1). Voluntarist Themes Beasts with intellects but without freedom – this notion is indicative of a significantly different take on freedom and its psychological root than Aquinas’s. For Thomas, the very fact of having intellect implies free will; in his view, an intellectual or rational appetite is by the same token a free appetite (see Chapter 2, pp. 42–3). Olivi’s notion of intellectual beings without freedom suggests that he thinks differently. In fact, the concept of an intellectual or rational appetite, so central in Aquinas, is virtually absent from Olivi’s works; in Book II of the Summa the expression “intellectual appetite” occurs only twice: once in discussing original sin (110c., III: 263), another time in asking about the difference between the intellectual appetite and the sensory appetite (68c., II: 625). Elsewhere, he uses “superior appetite” to set the will apart from the sensory appetite.98 The nine theses I singled out as characteristic of voluntarist accounts of free will (see p. 71 above) can be found in Olivi’s work as well, but he emphasizes certain things differently. The will is the highest power, because we are free and in control of our acts by our will (De perfectione evangelica 2, pp. 120–1; 57c., II: 330–5). Command – which according to Olivi means to will efficaciously, to move and dominate other powers – belongs to the will; if it belonged to reason, the will would not be free but subject to reason and moved by reason’s command (57c., II: 331–2; 85–6 ad 4, III: 195–6). Freedom is innate to the will, not derived from the intellect’s mode of cognition. The will is essentially free (40c., I: 684), and free decision is essentially the will (In Sent. II.24.2, quoted in Putallaz 2000, 106). Without the will, there would be no freedom (54c., II: 249). The will moves itself freely to its volition (e.g., 59 ad 9, II: 559); only the will, not the intellect, moves itself (72c., III: 26), and self-motion is essential to freedom (57 ad 29, II: 378). While freedom is essential to the will, it is essential to the intellect that it can be freely moved (40c., I: 684). By itself, the intellect relates to its object by natural necessity (54c., II: 246). As a power that is not of itself free, the intellect’s stature (altitudo) consists only in being connected to the will (57c. II: 331).99 In psychological terms, 98 99

See, e.g., 59c., II: 540 and 547; 77 ad 1, III: 155; 84c., III: 183; 111c., III: 275. See also 57 ad 21, II: 365–8, where Olivi presents different views concerning whether freedom is only in the will or in both the will and the intellect, without committing to any view.

88

Free Will

that freedom is innate to the will means that the will is not a passive, but an active power. The will is not only partially, but totally an active power, which means that its acts are not partially produced by other causes such as the willed object, but they are totally produced by the will itself. Moreover, “free decision has the character of a first mover” (58c., II: 409–11). Neither the intellect, nor the object, nor affections move the will. The intellect relates to the will as the counselor to the king (De perfectione evangelica 2 ad 3, p. 127); its function is to represent the object – nothing more (e.g., 57 ad 12, II: 354). Like Henry, Olivi grants the desired object a causal role only in a diminished sense, but his explanation is original: the object is required only as the term of the will’s act; in that function, it does not produce anything in the will. Olivi illustrates this idea with the example of the sun shining on a round vase or a triangular vase; the light in the vase then has a round or triangular shape, without that shape being efficiently produced by the vase. The vase only functions as the term of the sun’s light shining on it. So too, the will can only cause its acts in relation to some willed object, even though the will itself is the total cause of its acts (58 ad 1, ad 2, II: 414–18). That the object does not cause the will’s act is clear from a scenario like Buridan’s ass, where the will serves as a tiebreaker. When choosing among two apples that are alike in every aspect, the choice is made out of freedom alone (57c., II: 326; 85–6 ad 2, III: 194).100 Contrary to a view by John of Morrovalle (to be discussed in Section 4.2), whom Olivi may have in mind but does not mention by name, affections, that is, dispositions (habitus), are not needed for the will’s acts as such. They are needed only for willing in the due manner – for example, charity is needed for meritorious acts – and they render the will’s acts more easy, firm, and efficacious (57 ad 12, II: 376; cf. 58 ad 9, II: 429). Since the will is not a passive power, it need not follow the judgment of the intellect: “However clearly the intellect considers and knows in the universal and in the particular that this good or evil is to be done or not, to be avoided or pursued, the will can still will anything whatsoever” (57 ad 13, II: 356; see also De perfectione evangelica 2c., p. 120). Thus Olivi denies judgment-volition conformity. He also adamantly denies the Socratic deficiency thesis that deficient willing presupposes deficient cognition, for it implies the impossibility of the sin of the angels and of Adam (57 ad 18, II: 360; 85–6c., III: 188–90; cf. Chapter 9, p. 226). Olivi does not profess Henry’s view that an evil will immediately corrupts the judgment of 100

For the significance of Olivi’s account of the will as a tiebreaker, see Kaye 2004 and Normore 2007a.

Voluntarism

89

reason, but he holds that a certain kind of moral knowledge gets corrupted by sin. As we shall see in greater detail in Section 9.4, according to Olivi, one can perceive the truth about virtue and vice in two ways: merely theoretically, or as accompanied by some taste or sensation – an intellectual sweetness in thinking of what is good and a bitter and repulsive insipidity in thinking of vice. Sin corrupts only the second kind of moral knowledge (57 ad 18, II: 361; 85–6c. and ad 1, III: 193–4). An important innovation is the claim that free decision regards present contingents. According to the standard view that goes back to Hugh of St. Victor, free decision concerns only future contingents, for the future is open while the present is fixed.101 This means that one can choose only among options lying in the future. But Olivi objects that one can act only in the present. So unless one can do opposite acts in the very moment of choice, that is, in the present rather than in the future, one cannot at all act otherwise, and so one does not even have the ability to choose between future alternatives (57 ad 10, II: 348–53). Olivi’s innovative thesis was influential on Scotus, as we will see in Section 5.1.102 Freedom and the Good A point where Olivi is rather unconventional is his conception of the relation of free will to the good. This involves two distinct but related issues: the relation of the will to the good and the relation of freedom to the good. Like his contemporaries, Olivi affirms both the Pseudo-Dionysian axiom that one cannot desire or intend evil as such and the related axiom that one can will whatever one wills only under the aspect of the good.103 The scope of the will being anything that can appear as good, the will can desire any end and any means to the end (57c., II: 326). Thus far, he treads common ground. But he furthermore makes the un-Aristotelian claim that the will can set for itself as an end anything that has the aspect of the good, and that it can set for itself an end it did not previously pursue, as when someone begins to love someone for his or her own sake (ibid.). Olivi also speaks of “choosing” an end (83c., III: 179–80). By contrast, Aristotle does not allow for choice of ends, but only of what promotes the end (EN III.2, 1111b26–7), and Aquinas allows that one chooses an end only insofar as one takes it as a means to a further end (ST I-II.13.3). But for Olivi, an end need not be pursued in view of a further end, as in the case of loving someone for 101 103

See Chapter 1, note 19. 102 On Olivi’s theory and its impact on Scotus, see Dumont 1995. 107c., III: 255; see also 57c., II: 326; 77 ad 2, III: 156; 85–6 ad 1, III: 194. Cf. Chapter 6, p. 173.

90

Free Will

his or her own sake. He stresses that the will’s ability to set for itself its ends goes hand in hand with its freedom (57c., ad 14, ad 17, II: 326, 356, 359). Nevertheless, there are ends that the will does not set for itself. We can choose whether or not to love God, who is in fact our ultimate end, but we necessarily love with all our heart happiness under a general notion and cannot will its contrary (57 ad 17, II: 359; 83c., III: 179–80; 85–6 ad 3, III: 195). Elsewhere, in De perfectione evangelica, Olivi makes a clarification that anticipates a famous view of Duns Scotus: objects willed necessarily (e.g., happiness) are willed necessarily insofar as one cannot will their contrary; they are not willed necessarily as though one could not abstain from willing them. The will is able to abstain from willing any object whatsoever because it moves itself to such acts (q. 2c., p. 120). In fact, Olivi insists repeatedly that even when willing necessarily, the will moves itself, and that for this reason these necessary acts are free. This is so, for example, with divine self-love (57 ad 17, II: 359; cf. 59c., II: 547), demonic obstinacy, and the love of God by the holy angels and the blessed (57 arg. 29 and ad 29, II: 316, 378). As we have seen, Henry of Ghent, too, holds that even the will’s necessary acts are self-determined; but unlike Henry, Olivi admits that freedom is compatible with natural necessity (54c., II: 246; 111c., III: 277). Accordingly, what is essential to freedom in Olivi’s theory is only sourcehood, understood as self-motion of the will, not the ability to do otherwise (57 ad 17, ad 29, II: 358–9, 378). For Olivi, the ratio libertatis does not include willing something good. Thus he departs from the standard view, which follows Anselm of Canterbury in claiming that the ability to sin is not included in the essential character of freedom. According to the standard view, even in rational creatures, who are by nature able to sin (see e.g., Aquinas, ST I.63.1c.), the ability to sin is only accidental to freedom (see e.g., Aquinas, DV 24.7 ad 4). By contrast, Olivi considers freedom in God and freedom in creatures to be two distinct notions, and he claims that while the ability to sin is foreign to divine freedom, it is essential to created freedom (40 ad 3, I: 688; 41c., I: 695–6).104 He presents this position not in his own voice but referring to the view of others, possibly for strategic reasons; nevertheless, the favorable manner in which he presents the view indicates that he shares it. Those to whom he refers insist particularly that the ability to sin is essential to personal merit, professing a symmetrical account of moral responsibility: just as there would be no fault if the sinner could not do good instead of evil, so there would be no 104

See the lucid and detailed discussion by Kent 2017.

Voluntarism

91

personal merit if a person could not do evil instead of good (41c., I: 697–8). Duns Scotus and Ockham take up Olivi’s theory of the will setting its own ends, and they also hold that that the ability to sin is essential to created freedom, as we will see in Sections 5.1 and 5.5.

3.4 Other Franciscans and the Correctorium of Brother Thomas The theories of free will by other Franciscans who were Olivi’s contemporaries largely converged with Olivi’s, but they were less original. Here are a few examples of leading Franciscan masters, some of whom we will encounter again in later chapters. In Tractatus de anima, John Pecham writes that the will or free decision is free to withhold its consent to a practical judgment (ch. 11 n. 6, ed. Melani, 41). The intellect does not cause the will’s act but only stimulates the will to act (ch. 12 n. 4, p. 44).105 Matthew of Aquasparta defends the will’s ability to sin without a prior cognitive deficiency in Quodlibet V Question 8 (ed. Simoncioli 1956, 237–45). Richard of Menneville (Mediavilla) argues for the will’s self-motion in his Sentences commentary (II.38.2.2, II: 466b–469a) and Disputed Question 14 (III: 4–44), and he develops detailed arguments for its ability to oppose the intellect’s judgment in the Sentences commentary (II.38.2.4, II: 470b–473b) and in Disputed Question 14 (III: 48–104). Gonsalvus of Spain maintains the will’s self-motion and its ability to oppose the intellect’s judgment in Disputed Question 8 (pp. 113–32) and argues for the will’s innate freedom in Disputed Question 12 (pp. 226–44).106 The remote and proximate sources of inspiration of these Franciscan masters are the same: Bonaventure, Walter of Bruges, and later Henry of Ghent. A general tendency in favor of voluntarism was further bolstered by the Correctorium fratris Thomae, despite its developing weak arguments for the positions taken. The Correctorium is a handbook of corrections of “errors” by Aquinas, written by William de la Mare – incidentally a Franciscan who earlier, in his Sentences commentary, treats Aquinas 105 106

See also Pecham, De beatitudine animae et corporis 12 ad 10, ed. Spettmann, 559. On the free will theories by these Franciscans, see Stadter 1971, 120–43, 253–9, 263–71, 282–4; on Richard, see also Kent 1995, 88–91; on Gonsalvus, see also Martel 1968, 145–77. Melani dates Pecham’s Tractatus de animae between 1270 and 1279 with certain reasons suggesting 1277–9 (Pecham 1948, XLIX–L). Glorieux 1935, 197 dates Matthew’s Quodlibet V to 1282/83. In the introduction to Richard’s Disputed Questions, Boureau dates these between 1292 and 1298 (vol. I: XXI). (For Menneville as Richard’s hometown, see ibid., p. X.) Richard’s Sentences commentary, in its revised and published version, contains references to the Disputed Questions, but not vice versa (Hocedez 1925, 49–55), so at least parts of it must be posterior to the Disputed Questions. Glorieux 1939, 172 dates Gonsalvus’s Disputed Questions to 1302–3.

92

Free Will

with respect and adopts many of his ideas.107 A first redaction of the Correctorium was written after 1277 (for it refers to Tempier’s March 7 censure) but before 1282, when at the meeting of the general chapter of the Franciscans in Strasbourg it was decided that Aquinas’s Summa theologiae was to be made available only to “reasonably intelligent readers” and only as accompanied by the Correctorium.108 The Correctorium did maybe more to define Thomism than to fix Franciscan orthodoxy; in fact, in the following years, several Dominicans wrote replies to William’s handbook, which they called mockingly Correctorium corruptorii, insinuating that William did not correct but corrupted Aquinas’s teachings.109 Regarding the relation between intellect and will, the Correctorium affirms the standard voluntarist theses as professed also by Walter of Bruges, Henry of Ghent, and Olivi.110 Even though Franciscans espousing voluntarist ideas were so numerous, the principal representative of the voluntarist movement and the preferred target of intellectualists from the last decades of the thirteenth century until the second decade of the fourteenth century was not a Franciscan, but Henry of Ghent. 107 108 109 110

See Caldera 2005, 471–84. For the text of the decision, see Fussenegger 1933, 139. William wrote a second, longer redaction, which had limited circulation; see Oliva 2005. For references to the various Correctoria corruptorii, see Oliva 2005, 423, note 3. A printed text of the first redaction is available as part of Richard Klapwell’s Correctorium corruptorii “quare,” ed. Glorieux, where Richard reports William’s text before refuting him. William’s voluntarist views can be associated with Henry of Ghent’s voluntarist theses (see p. 71 above) as follows (page numbers refer to Glorieux’s edition): (1) p. 161, (2) pp. 233–4, (4) cf. p. 241 (reason can be coerced), (5) pp. 237–8, (7) p. 242, (8) p. 232, (9) pp. 332–3. On intellect and will in the Correctorium, see Kent 1995, 81–4 and 182–5, Putallaz 1995, 100–26, and Brungs 2006. The latter two also consider the Dominican responses.

chapter 4

Intermediary Theories and Strict Intellectualism

In an intellectual climate where determinism is a major concern, voluntarists proposed a theory of free will that avoids determinist threats by positing a firewall between genuine causal influences and a person’s will. Thus the will is protected from necessitating influences of the senses, which in turn are subject to external things whose causality is traceable in the last analysis to the necessary motion of the celestial spheres. The will is also protected from cognitive determination by the person’s own rational evaluation of what is worth pursuing or avoiding, all things considered. But this move of voluntarists is also their Achilles’ heel. Postulating a firewall between the will and the rest of reality, including the practical intellect, has at least two problematic implications. One is that one must ascribe to the will per se self-motion, an assumption that – despite efforts by voluntarists to show the contrary – seems to contradict fundamental metaphysical principles. Another implication is a disconnect between the will and its object, that is, between the will and what is grasped as desirable and choiceworthy. This disconnect involves two major problems. First, if the will is moved by itself rather than by the cognized object (that is, by the thought of a desirable thing or activity), what specifies the will’s act to willing this particular thing or action? Furthermore, why would one will something if one does not consider it worth willing, and why would one reject something one considers best, all things considered? These are the main theoretical difficulties of a strictly voluntarist account of free will, such as the one theorized in most detail by Henry of Ghent. In response to these difficulties, many thinkers ascribed a greater causal role to the cognized object than Henry. Godfrey of Fontaines went to the opposite extreme: the will does not move itself (except incidentally); rather, the cognized object moves the will as its efficient cause. This claim alone justifies classifying Godfrey as an intellectualist; but as we will see, he takes intellectualist stances on other issues as well and disagrees with all of Henry’s core voluntarist claims. Other thinkers developed theories that are 93

Free Will

94

located somewhere between Henry and Godfrey. Giles of Rome and John of Morrovalle proposed such intermediary theories in the 1280s – unsuccessfully, however, in the eyes of Henry and Godfrey, who criticized Giles and Morrovalle from opposite perspectives. James of Viterbo, writing in the 1290s, worked out a more refined “middle way,” as he called it; but he too became the target of Godfrey’s extensive critique.1 In this chapter we will study the intermediary accounts of Giles and Morrovalle. Then we will turn to Godfrey, who defends strict intellectualism and whose principal discussions of free will respond not only to Henry of Ghent, his great antagonist, but also to Giles and Morrovalle. Godfrey and Henry were the most prominent theologians at the time, but Giles was also counted among the principal protagonists of the free will debate.2

4.1

Giles of Rome

As mentioned already, Giles of Rome, a student of Thomas Aquinas and a member of the newly founded Augustinian order, began his scholarly career on the wrong foot: in 1277 he was subject to doctrinal condemnation and was refused the license to teach (see Chapter 3, p. 62). After he was rehabilitated and promoted to master of theology in 1285, he was able to hold quodlibetal disputations, allowing him to participate more effectively in the Parisian discussion about free will. Giles had a lasting impact, in part because in 1287 the Augustinian order declared him its official doctor, obliging its students to align with his past and even future writings.3 Two big themes run through Giles’s discussions of free will: the relation between cognitive and volitional deficiencies, and the control of our choices through control of the judgments that cause our choices. The Relation between Cognitive and Volitional Deficiencies Giles’s early view about the relation between cognitive and volitional deficiencies was among the doctrines that came under attack in the process against him. As we have seen, Article 24 condemned the proposition that “there is no evil in the will unless there is error in reason” (and Article 51 censured a very similar formulation) – the very statement that 1 2 3

For a detailed study of James of Viterbo’s theory of free will, see Dumont 2018a. Gonsalvus of Spain mentions Henry, Giles, and Godfrey as representatives of three major theories of free will; see Putallaz 1995, 248–9. For Giles’s life and works, see Del Punta, Donati, and Luna 1993, and Briggs 2016.

Intermediary Theories and Strict Intellectualism

95

after his rehabilitation was called propositio magistralis.4 Giles lectured on the Sentences probably in the academic year 1270–1; the condemned propositions are drawn from Book I of his Ordinatio, the revision of the lectures on the First Book of the Sentences, which he published ca. 1271–3.5 Free will and related issues are mostly discussed in Book II of the Sentences commentaries, and in the passages of Book I from which the censured propositions are taken, Giles discusses the relation between intellect and will only in passing. Article 24 is extracted from a discussion of whether the will needs a disposition (habitus) to elicit an act of love (Ord. I.17.1.1.1, ed. Venice 1521, 89M); Article 51 is taken from a Question that asks whether anything falls outside the divine will (Ord. I.47.2.1, p. 237G). The only time Giles affirms that the error in reason precedes evil in the will, however, is in his actual lectures on the Sentences while discussing angelic sin (Reportatio Monacensis II.29c., III/2: 248). Giles had no reason to think he said anything scandalous: in the second passage of his Ordinatio, he backs the statement by standard authoritative statements from Aristotle and Pseudo-Dionysius (p. 237G). Besides, Aquinas had maintained throughout his writings that there cannot be sin in the will without ignorance in reason, but he was not rebuked. After his condemnation and probably before his rehabilitation, Giles returned to the relation between cognitive deficiencies and an evil will in a Question called De deceptione, which asks whether someone who knows that a deed is evil can do it nonetheless. This Question is programmatic for his future discussions of the relation between cognitive and volitional deficiencies, including his account of angelic sin, where he cites it. His intention in De deceptione is to show that there is no evil in the will without nescience or ignorance in reason – the topic of Articles 24 and 51 – but he explicitly sets aside whether ignorance or nescience precedes evil in the will or only accompanies it (De deceptione, ed. Bruni, 232). Giles shows in some detail that while our choice depends 4 5

On Giles’s early theory of the relation between cognitive and volitional deficiencies and its development after the concession of the Magisterial Proposition, see also Eardley 2006b. For the dating of Book I of the Ordinatio, see Del Punta, Donati, and Luna 1993, 330b. Regarding his actual lectures on the Sentences, Concetta Luna considers 1269–71 or 1270–2 as their dates, but points to two incompatibilities with external evidence; see “Introduzione,” Aegidii Romani Opera Omnia III/2, 16–24. One incompatibility – Giles citing two works by Aquinas dated to the end of 1270 in his Reportatio II, while Aquinas’s Quodl. IV of Lent 1271 knows Giles’s Reportatio III – disappears however, because the Parisian practice was to lecture on the Sentences not in two years, as scholars generally assumed, but in a single year, probably in the sequence I-IV-II-III (see Duba and Schabel 2017). Given the cross-references, for Giles this would have been 1270–1. The other incompatibility rests on the plausible dating to 1272 of Aquinas’s De unione Verbi incarnati, which Giles cites in Rep. III.33, and which therefore should be more plausibly dated to 1271.

96

Free Will

on our judgment, our judgment depends not only “on reason and the light of the agent intellect,” but also on our desire. Inordinate desire (concupiscentia) inclines us in such a way as to make something appear to us as suitable that is not (such as fornication, Giles’s preferred example). Someone who is infected by the desire for pleasure judges falsely concerning pleasures, just as the tongue of a sick person tastes flavors falsely. So desire makes us judge that something is to be done that really must be avoided, or it makes us think that, all things considered, fornication is more good than bad (ibid., 233–7, 244–5). To describe in which sense the evildoer is ignorant about the good, Giles adopts Aristotle’s distinction between acting in ignorance and acting due to ignorance (EN III.1, 1110b18–22). Giles writes that all evildoers act in ignorance, for “every evildoer is ignorant of what ought to be done and what ought to be avoided” (quoting EN III.1, 1110b28–9). But only those who act due to ignorance of what must be done in general and in particular are fully ignorant that what they are doing is evil (De deceptione, 231–2, 242 ll. 34–6). Other evildoers, namely those who act from passion (like the incontinent) or from a vicious disposition (like the intemperate), are not entirely ignorant that what they are doing is evil; rather they simultaneously know and do not know, albeit in different respects. Equally, evildoers simultaneously will and do not will, although in different respects (ibid., 239–40). Regarding the different respects in which someone can be said to know, Giles adopts the distinctions between knowing universally and knowing in particular, and between knowing habitually (in habitu) and knowing occurrently (in actu), which Aristotle introduced to explain incontinence (see Chapter 1, p. 26). Reason can simultaneously judge that fornication is evil and not to be done and that it is good and to be done. The prohibitive judgment (that fornication is evil) is caused by reason and the light of the agent intellect. It concerns the universal, by which reason considers fornication abstractly, “as it is in itself,” namely disordered. The permissive judgment (that it is good) is affected by inordinate desire (concupiscentia). It concerns the particular, by which reason considers fornication “as it is in the thing” that is pleasurable (De deceptione, 233, 235–6). As to different respects in which one can will, Giles adopts Aristotle’s theory of mixed voluntary acts, as exemplified in a captain who during a sea storm in one sense wants to throw cargo overboard and in another sense does not (EN III.1, 1110a4–19). Fornicators do not want fornication as it appears abstractly in the universal and in itself, for one cannot

Intermediary Theories and Strict Intellectualism

97

want evil as such; but while disliking fornication in the universal, they can want it as it is in the particular pleasurable thing, for they want it as pleasurable (De deceptione, 239–40). Evildoers are deceived, and yet their conscience remains intact and accuses them. They would deny that what they are doing ought to be done, but their heart feels differently from what their mouth professes (ibid., 240–1). Giles summarizes the key ideas as follows: Someone who sins from passion or from a disposition knows without qualification [simpliciter] in the universal, but is ignorant in a certain respect [secundum quid], for he is ignorant in the particular. But regarding that person’s willing, we say that he wills without qualification, for he wills in the particular. For someone who sins not due to ignorance but in ignorance has, while he sins, a will that is bad without qualification, but a reason that is erring in a certain respect. (De deceptione, 241)

Thus in De deceptione, Giles shows that evildoing involves ignorance, and that someone who has untainted knowledge does not do evil. He also shows that moral evil does not originate in a cognitive deficiency alone, for the misjudgment that causes evil has itself been caused by base desire. What he has not clarified is what causes such desire and whether one controls its effect on one’s judgment and hence on one’s choice. After his promotion to master, Giles had the opportunity to share his new thinking about the relation between evildoing and ignorance with a larger audience, for in his first quodlibetal disputation of Advent 1285 or Lent 1286, he was asked “whether there can be evil in the will if there is not some error in reason” (q. 19, ed. Leuven 1646, p. 39b). Thus someone in the audience asked him to take a position regarding the propositio magistralis, which had just been conceded by the Parisian theology masters. A similar question was brought to him again two years later in Quodlibet III.6 In the meantime, Henry of Ghent had discussed it in his Quodlibet X, as we have seen. In Quodlibet I Question 19, after mentioning the stance of the Parisian masters who conceded the contemporaneousness of error and evil (p. 40b), Giles repeats the key ideas from De deceptione and formulates his main conclusion as follows: Since no one wants to do something unless he apprehends it as good and judges that it is to be done in this way, it is clear that evil is not in the will

6

On Giles’s discussion in these two Quodlibeta, see also Putallaz 1995, 212–15, Pini 2006, 257–8, 269–70, and Eardley 2006a, 370–5 and 2016, 196–206. For the dating of Giles’s Quodlibeta, see Pini 2006, 240–4; for a more precise dating of Quodl. V, see Osborne 2013, 59 and 75.

98

Free Will unless error is in the intellect, for while the will wants to fornicate, the intellect apprehends fornication as something good and judges it to be done, which cannot happen without some error. (Quodl. I.19, p. 41b)

Here he defends the Socratic deficiency thesis on the ground of judgmentvolition conformity. Again, Giles does not clarify how one controls whether one apprehends something as good and judges that it must be done. How we control our judgments and choices is at the center of Giles’s discussion of free will in Quodlibet III. In Question 16, discussing the propositio magistralis, he uses again the resources from De deceptione to explain how evildoers act in ignorance, though not necessarily due to ignorance. But now he assigns a greater role to the will. Apprehending fornication as good (because it is pleasurable) and evil (because it is inordinate), the intellect can only form the proposition “fornication is inordinate pleasure” if the will makes it connect the terms of the proposition. As we will see in greater detail below, Giles holds that the will controls whether the intellect considers fornication as pleasurable or as inordinate, and thereby the will controls whether one is attracted to or repulsed by the idea of fornication (p. 182a). If attracted by fornication, the will “infects” reason to make it judge perversely that fornication is to be done, just as an infected tongue perverts the judgment about flavors (p. 182b). Giles summarizes this process: In sinners, there is, then, this sequence. First, reason apprehends something under a twofold aspect, and this [happens] without sin, but not without danger. Second, the will determines itself to will to desist [desistere] considering about fornication under this aspect, and to focus [insistere] on the other, and as the will determines itself, so reason carries out as the will wills, and so it considers. Upon this, the will is attracted or repulsed, and if it is attracted, it perverts the judgment or reason. Once it is perverted, we proceed to the pleasing deed. (Quodl. III.16c., p. 182b)

As Giles insists, the defect need not begin in reason, but can begin in the will “that does not want to follow the perfection that the intellect proposed” (ad arg., p. 183b). Giles thus follows Henry in assuming that an evil will blinds the intellect. He now also abandons judgment-volition conformity, with particular clarity in the next Question: What reason judges as simply not to be done, the will can simply will, such that the depraved judgment of reason is not the cause of a depraved will, but rather vice versa. (Quodl. III.17c., p. 186a)

Intermediary Theories and Strict Intellectualism

99

Free Decision and the Causes of the Will’s Act In discussing free decision in his lectures on the Sentences, Giles follows Aquinas closely.7 Later, in his Quodlibeta, he uses two of Aquinas’s resources to develop a theory that goes beyond his teacher.8 One is the idea that in virtue of willing an end, the will can move itself to will what is related to the end; the other is the exercise/specification distinction. With these resources, Giles develops an account according to which the cognized object genuinely causes the will’s act, and yet the will controls which volition the object causes. He returns again to free decision in Book II of his Ordinatio (completed not before 1309), adding a few clarifications to his quodlibetal discussions, partly in response to his critics.9 The central text for Giles’s mature theory of free decision is his Quodlibet III, Question 15 (of 1287 or 1288), where he responds to the question of whether the will moves itself. Giles intends to solve the question within the parameters of Aristotle’s metaphysics. According to Giles, self-motion can be understood either as “self-actuation” or as “selfdetermination.” In Giles’s nomenclature, to “actuate” or to “activate” is the reduction from potentiality to actuality. Throughout his writings, Giles denies the possibility of self-motion as self-actuation, for this would contradict Aristotle’s prohibition of per se self-motion. In fact, Giles repeatedly affirms the Aristotelian motion principle that everything that is moved is moved by another.10 As Giles tacitly assumes, self-determination, in contrast to self-actuation, does not violate this principle. For Giles, it is by means of self-determination that the will controls its own act and that of the other powers. What originally accounts for the reduction of the will from potentiality to actuality, according to Giles, is the (ultimate) end and what is known as good from every perspective. Prior to this “activation,” the will is entirely in potentiality. The end and whatever is cognized as an all-around good necessarily moves the will to pursuit. By contrast, that which is related to the end and that which is not cognized as an all-around good leaves the will indeterminate. Regarding these latter things, there is a twofold dynamic: the will is actuated by the cognized good, and the will determines itself. Giles thus holds, contrary to Henry of Ghent, that the cognized object is 7 8 9 10

Rep. II.65, III/2: 335–46, using Aquinas, DM 6 and ST I.82.2. Eardley 2016, 174–95 provides a detailed study of Giles’s theory of free decision in the various stages of his career. For the dating of Book II of the Ordinatio, see Del Punta, Donati, and Luna 1993, 332a. E.g., Quodl. III.11c., p. 159a; Quodl. III.15c., 177b; Ord. II.25.1.3 dub. lat. 2 ad arg., Venice 1581, II: 301aB.

100

Free Will

the proper cause of a volition (Quodl. III.15c., p. 177b).11 Giles illustrates this twofold dynamic with the classical example of fornication. The thought of it alone does not cause the will to pursue or avoid it. Before the will can want either alternative, the intellect must consider fornication as worthy of pursuit or avoidance. The object causes the volition of wanting the intellect to consider the object, but – the case of complete goods or evils aside – leaves it open whether the will wants to privilege the consideration of one aspect rather than another. The will then determines itself to want to make the intellect stay (sistere) on the thought of fornication as delightful and desist (desistere) from the consideration that it is inordinate, or vice versa.12 Depending on how the will first determines itself to will to determine the intellect’s consideration, the thought of fornication activates the will to want to do or avoid it. For Giles, the will controls directly whether it wants to make the intellect stay on a consideration or desist from it. In support of this, Giles insists that the will, and the will alone, is master of its acts. Granting the will an intrinsic ability of self-determination implies that it is not entirely a passive power (Quodl. III.15c. and ad arg., pp. 177b–180a).13 In the following Quodlibet, Giles clarifies that the will has such control because its freedom is innate rather than derived from the intellect. The intellect grounds the will’s freedom only in the sense that what the intellect proposes – apart from the all-around good – leaves the will undetermined regarding alternatives. But the freedom to determine itself to either alternative is intrinsic to the will itself (Quodl. IV.21c. and ad arg., pp. 257b– 259a). At first sight, Giles’s theory resembles Aquinas’s mature account of free decision. Seemingly, what Aquinas calls specification or determination is the same as Giles’s actuation, and Aquinas’s exercise is Giles’s determination. Henry of Ghent, in fact, considered both theories to be equivalent and attacked both with a single shot (Quodl. XII.26, XVI: 150 ll. 52–63). 11

12

13

Later, citing Averroes, Giles clarifies that the object “in the soul,” that is, as cognized (the thought of a public bath, in Averroes’s example), moves the will by efficient causality, while the object as a real thing outside the soul moves it by final causality (the bath’s benefits motivate a person to go there); see Ord. II.41.2.3c., II: 633aD citing Averroes, In Met. XII.36, Editio Juntina VIII: 318I–K; see also Ord. II.25.1.6 dub. lat. 2, II: 311aA–B. Giles later clarifies that all cognition requires an “intention of the will,” willing the intellect to judge about an impression received in the intellect. Though the intellect’s act precedes the will’s intention to judge “in generation or origin,” both are simultaneous “in time or duration”; if they were not simultaneous, there would be an impression but no cognition. Some of the will’s intentions to judge are natural, beyond control; but after their first occurrence, the will controls whether to stay (persistere) in the intention or not; see Quodl. V.6, pp. 280a–282b and Osborne 2013 for discussion. For the will as a not purely passive power, see also Rep. II.65 ad 4, III/2: 344–6.

Intermediary Theories and Strict Intellectualism

101

But Giles’s understanding of self-determination is profoundly different from what Aquinas understands as exercise, having a clearly voluntarist character that Aquinas’s account lacks. For Giles, self-determination is entirely a matter of the will; the intellect has no say over how the will determines itself to favor one consideration of an object rather than another. By contrast, for Aquinas, the exercise of the will must be traced to an external cause – to the intellect (when one deliberates about the exercise of the will or of another power) and in the last analysis to God (as the ultimate mover of the will’s exercise). For Aquinas, the will’s freedom is derived from the intellect (or God), whereas for Giles, it is innate to the will. Like Aristotle, Aquinas holds that the transition from indetermination to determination is accomplished by deliberation, which results in a determinate choice. For Giles, by contrast, the transition from indetermination (an object cognized prior to the will’s determination) to determination (an object cognized while focusing on certain aspects rather than others) can be made without deliberation; the will’s self-determination suffices to narrow down available options to a determinate choice.14 Originally, Giles had himself used Aquinas’s terminology of specification and exercise (Rep. II.65c., III/2: 337 ll. 48–9). What motivated his change of terminology from specification/exercise to actuation/determination? Possibly he considered “actuation” to be better suited to express the reduction from potentiality to actuality, and “determination” to denote the transition from the indeterminacy of the object insofar as it is cognized prior to the will’s intervention to the determinacy of the object considered from a particular perspective after the will’s involvement. Yet there are several problems in Giles’s theory. First, how do the change from potentiality to actuality and the one from indeterminacy to determinacy differ? It seems that determination, no less than actuation, requires a determinate efficient cause. As we will see, it did not take long for Giles’s critics to formulate objections along these lines. Furthermore, the generation of a volition (e.g., the desire to fornicate) requires a prior volition: wanting to consider the permissive rather than the prohibitive aspects of the object under consideration. How was this prior volition generated? If in just the same way, then it requires itself a prior volition, and so forth ad infinitum. In addition, the ontological status of the will’s selfdetermination is unclear. As mentioned, for Giles, the will determines itself to will to consider fornication as delightful rather than as inordinate (Quodl. III.15c., p. 178b). Is the will’s determining activity an act of the will 14

For the differences between Giles and Aquinas, see also Eardley 2003, 858–61.

102

Free Will

in the full sense?15 Apparently in response to Godfrey of Fontaines’s critique (Quodl. VIII.16c., PhB IV: 151–2), Giles later clarifies that determination is not itself an act; rather, it consists in desisting from one cognized good (e.g., fornication as inordinate), whereby the will is moved by another cognized good (e.g., fornication as pleasurable) (Ord. II.25.1.3 dub. lat. 2, II: 298bB–C). Giles as a Moderate Voluntarist or Intellectualist While at first Giles affirms intellectualism, after his rehabilitation, especially in his third Quodlibet, he shifts toward voluntarism. His mature theory contains elements of both. He constructs his theory of free will upon the intellectualist claim that the will’s act is caused by the cognized object (Quodl. III.15c., pp. 177b, 179b). Hence it is caused from without, not self-caused (as Henry of Ghent thinks), and Giles is in fact intent on respecting Aristotle’s prohibition of per se self-motion (e.g., p. 177b). But Giles also thinks that only the will, not the intellect, is in control of its act (p. 178a), and that for this reason the will is freer than the intellect (Quodl. IV.21 ad arg., pp. 258b–259b). Thus in order to defend free will, he must introduce an apparently minor voluntarist element into his theory: the assumption of the will’s self-determination. This small move considerably attenuates the intellectualist outlook of Giles’s theory, for several voluntarist ideas are implied in this assumption: the will is not an entirely passive power (Quodl. III.15 ad arg., p. 180a); the will is formally free by itself (Quodl. IV.21c., p. 258b); the will can abstain from or even judge contrary to the practical judgment of the intellect (Quodl. III.16 ad arg., p. 183b; Quodl. III.17c., p. 186a); an evil will does not presuppose an error of the intellect but rather entails it (Quodl. I.19c., p. 41b; Quodl. III.16c., pp. 181b– 183a). By proposing an intermediary theory, Giles made himself enemies on both sides. In 1288, shortly after Giles held his third quodlibetal disputation, Henry of Ghent scrutinizes Giles’s theory. He considers the distinction between actuation and determination to be merely verbal, and so he undermines Giles’s attempt at its root (Quodl. XII.26c., XVI: 150–1 ll. 52– 79). Henry basically criticizes every point in which Giles does not agree with his own theory. While Giles argued that the will’s act depends on 15

While Giles does not explicitly call self-determination a positive act of the will, it seems that this is what it amounts to: he speaks of it as “determinately wanting to consider” (e.g., Quodl. III.15c., p. 179a). See also Godfrey of Fontaines’s interpretation, p. 123 below, and Eardley 2016, 191–2.

Intermediary Theories and Strict Intellectualism

103

which consideration the intellect focuses on, Henry retorts that even while commanding the intellect to consider fornication from the viewpoint of pleasure, the will remains free to will to pursue or avoid it (ibid., 152 ll. 93–01). Henry especially rebukes Giles for attributing genuine causality to the object (ibid., 152 ll. 6–11). For Henry, the will need not be actuated from without, but rather can reduce itself from potentiality to actuality (ibid., 154 ll. 46–7). As we will see shortly, Godfrey of Fontaines attacks Giles from the opposite side.

4.2 John of Morrovalle Like Giles, John of Morrovalle, a Franciscan who became a master in early 1289, arrives at a middle ground between intellectualism and voluntarism, coming from the opposite direction. Giles starts from the intellectualist assumption that the cognized object causes the will’s act, but makes a voluntarist concession in assuming that the will controls which aspects of the object end up moving the will. By contrast, Morrovalle starts from the idea that the will moves itself (QD, ed. Longpré, 490–1), but makes the intellectualist concession that the cognized object has a certain influence (influxus) on the will. Regarding Morrovalle’s theory of the cause of the will, only a short Disputed Question has come down to us.16 Henry and Godfrey, who critique his theory, seem to have encountered it in an earlier text or disputation, of which we have no record. Morrovalle seems to have lacked later occasions to defend his view, at least at the University of Paris, where his tenure lasted only one or two years, after which he taught at the Papal court, then served as minister general of the Franciscans, and finally as a cardinal. As minister general, he had more pressing concerns than the moral psychology of free will. The Franciscan Order was deeply divided over the question of evangelical poverty. On this issue, Morrovalle’s position was moderate, his actions severe: he bore down hard on Peter Olivi and his followers, the socalled “spirituals,” who insisted that Franciscans live extreme poverty.17 In his Disputed Question on the cause of the will’s act, Morrovalle wants to avoid two opposite mistakes: the claim that the object does not 16

17

See Alliney 2013, 569–73 for the authenticity of works attributed to John of Morrovalle. Alliney considers the authenticity of the Disputed Question edited by Longpré to be almost certain. As further confirmation, we can add that in a marginal note to Henry of Ghent’s Quodl. XII.26 in Vat. lat. 853, 268va, the scribe refers to Morrovalle and expounds an account of Anselm’s affectiones that closely resembles QD, ed. Longpré, 489. For John of Morrovalle’s biography, see Emili 2010; see also Burr 1976, 74–5.

104

Free Will

cause a volition at all, except in a metaphorical sense (ed. Longpré, 488) and the claim that it (alone) causes the volition (p. 490).18 He seems to have in mind Henry and Godfrey, respectively,19 or else, theologians holding similar views. He observes that if the object has no causality on a volition, then there is no ground for its specification; all volitions would be the same (p. 491). If the object alone causes the volition, then the will is not the source (principium) of its act and is not free (p. 490). So Morrovalle intends to grant the object just enough of a role to allow the specification of volitions, yet without endangering the will’s freedom regarding alternatives. His solution is based upon Anselm of Canterbury’s claim that the will moves itself by means of its affections (p. 489, referring to De concordia III.11, II: 284). Morrovalle construes the affections of the will analogously to the “intelligible species,” that is, the likeness of an essence through which the intellect can cognize universals. In his view, just as the intellect understands something by means of an intelligible species, so too the will moves itself to will something by means of an “influence” (influxus) by the cognized good. He calls this influence “affection” (affectio). Morrovalle’s affections are not quite like Anselm’s, for while Anselm posits only two kinds (for advantage or for justice), Morrovalle holds that “different objects cause in the will different affections” (p. 491), suggesting that there are as many kinds of affections as there are specifically different objects. These affections caused by the objects do not themselves cause the act of willing, for they are only an aptitude (habilitas) for willing. Accordingly, “the object moves [the will] as regards an affection, but not as regards a volition” (p. 492). Although the object does not cause the will’s act, Morrovalle concedes that one can say with Aquinas that the object moves the will as to the determination of the act and the will moves itself as to the exercise of the act (p. 491). When an affection has determined the will to a determinate volition, the will still remains able to determine itself either to will accordingly or not (p. 492). Morrovalle’s theory, probably as formulated on an earlier occasion, provoked enough interest to move certain people in 1289 to present questions about it to Henry of Ghent (Quodl. XIII.11) and Godfrey of Fontaines (Quodl. VI.7). Morrovalle’s middle position did not convince Henry and Godfrey any more than Giles’s earlier attempt. Henry concedes that an affection that the object causes in the will allows one to act more easily and 18 19

For a more comprehensive discussion of Morrovalle’s theory of the cause of the will’s act, see Alliney 2013. See Henry, Quodl. XIII.11 ad 3, XVIII: 105–6 ll. 41–59; Godfrey, Quodl. VI.7c., PhB III: 163.

Intermediary Theories and Strict Intellectualism

105

promptly, as do virtuous dispositions and certain passions (Quodl. XIII.11c., XVIII: 94–5 ll. 91–03). But the will does not need any such affection in order to act; it only needs the presentation of an object (ibid., 89 ll. 49–52). Were the affection indispensable, determinism would follow. Henry argues that since the object causing the affection is a natural cause (in the sense of Met. IX.2, of not being open to contraries), the object makes it so that the affection is itself natural (that is, not open to contraries). Thus the affection can have only one determinate effect, leaving the will without alternative possibilities and hence without freedom (ibid., 92–3 ll. 17–53). For Henry, even if the affection caused the will’s act not as its total cause, but only as a concurrent cause together with the will itself, the act would still be necessitated. A natural cause lacks the flexibility to adjust to the will, hence the will can be a concurrent cause with a natural cause only if it adjusts to the natural cause. The will and a heavy object can cooperate only in downward motion, for example, when someone is falling down and wants to fall down. They cannot cooperate in upward motion, to which the heavy object has no inclination. Thus, the affection, as a natural concurrent cause, would necessitate the will (Quodl. XIII.11 ad 3, XVIII: 111–12 ll. 2–32). We will discuss below Godfrey’s critique of Morrovalle, coming from the opposite direction.

4.3

Godfrey of Fontaines

Until Godfrey of Fontaines entered the scene as master of theology in 1285, Henry of Ghent did not have a rival of his stature. Godfrey, like Henry a secular master, studied theology under Henry,20 but then became his sternest critic. He was also critical of the more moderate theories by Giles of Rome, John of Morrovalle, and James of Viterbo. Godfrey developed Siger of Brabant’s unsystematic ideas concerning free will into a systematic account, thus giving mature expression to intellectualism. Aquinas, for whom Godfrey expressed great esteem (Quodl. XII.5c., PhB V: 102–3), also greatly influenced his theory of free will. Apart from influences from contemporaries, a key factor in Godfrey’s theory of free will is a close reading of Aristotle’s metaphysics and moral psychology.21 Prior to Godfrey, the dominant approach to free will was based upon ethical or psychological assumptions, and metaphysical considerations were a mere afterthought. For Godfrey, metaphysics is the starting point, 20 21

See Wippel 1981, xix. On Godfrey’s theory of free will, see also Lottin 1937a, Langevin 1960, Putallaz 1995, 177–208, 218–52, and Leone 2017b.

106

Free Will

and if metaphysical principles demand it, adjustments are to be made to moral psychology, not vice versa. The Cause of the Will’s Act Our first extensive record of Godfrey’s discussions of free decision and the relation of intellect and will is contained in Quodlibet VI of Advent 1289.22 In Question 7, Godfrey clarifies the theoretical foundations of his account of free will in contradistinction to those of his adversaries. The question brought to him essentially asks whether one who denies Henry of Ghent’s theory of the cause of the will’s act can affirm John of Morrovalle’s: Whether these two affirmations are compatible: that in the presence of an object, the will cannot move itself to an act of willing and that the will can move itself to an act of willing by means of a disposition [per aliquam dispositionem] that exists in the will, even though that disposition and the will are not in distinct subjects. (Quodl. VI.7–8 intr., PhB III: 148)

In his response, Godfrey refutes both Henry’s and Morrovalle’s theories, fundamentally because both affirm per se self-motion of the will, which Godfrey considers impossible. As he points out, this hypothesis implies that the active power and the corresponding passive power are identical: the will as causing a volition and the will as that in which the volition is caused. But such identity contradicts the nature of active and passive powers. The actuality of the active power must correspond to the potentiality of the passive power, hence they are in real relation; but identity allows for no real relation (Quodl. VI.7c., PhB III: 152). Positing that something acts on itself rather than on another implies that the thing is in actuality and potentiality in the same respect. These considerations strike at Morrovalle no less than at Henry, for if the will can act only on something else, then this is so even if – as Morrovalle thought – it becomes active by means of an added disposition (dispositio) (ibid., 153–4).23 As Godfrey says, the nonidentity of the active and the passive 22

23

In Quodl. VI.7, Godfrey refers to his earlier critique of Henry of Ghent’s theory of free will (PhB III: 152). This could have been in Quodl. II.9 of Lent 1286, asking whether intellect or will is the more excellent power, and where an opening argument mentions self-motion, or Quodl. III.15 of Advent 1286, asking whether the will is freer than the intellect. Unfortunately, these Questions contain only the opening arguments and refer for the response to the “liber magistri” (Godfrey’s own copy), which has not come down to us. For the dating of Godfrey’s Quodlibeta, see Wippel 1981, xxvii– xxviii. Godfrey’s axiom that active and passive powers must be distinct and the applications of this axiom are discussed extensively in Wippel 1973.

Intermediary Theories and Strict Intellectualism

107

power grounds his principal objection against both positions (ibid., 152). Concerning Henry, Godfrey also scrutinizes his distinction between virtual and formal actuality and his claim that the cognized object is a mere sine qua non cause. Recall that Henry thought that his hypothesis that the will moves itself from virtually willing to formally willing avoided the implication that the will is simultaneously and in the same respect in actuality and in potentiality (Chapter 3, pp. 78–9). But if the will is virtually willing, Godfrey retorts, it already contains the perfection of willing in a more eminent way; so if it made itself to be formally willing, it would make itself be in a state in which it already is (Quodl. VI.7c., PhB III: 151). In the following Quodlibet, Godfrey offers a more lengthy refutation, centering on the implication that a will moving itself from virtually willing to formally willing would refer to itself by a real relation, which is impossible (Quodl. VII.6c., PhB III: 342–8). (Henry responded by accusing Godfrey of “great crudity of mind” for his inability to grasp the difference between containing a perfection virtually and containing it formally; see Quodl. XIV.2c., 549vC.) Against Henry’s claim that the cognized object is a sine qua non cause for the will’s act, Godfrey objects that a volition is not caused differently from other things. To say that firewood heats up by itself, but needs the presence of fire as a sine qua non cause, is absurd. Just so, it is absurd to hold that the will causes its volition itself but needs the presence of the object as a sine qua non cause (Quodl. VI.7c., PhB III: 158; see also p. 152). Godfrey also finds Henry’s and Morrovalle’s accounts of the self-motion of the will problematic from a theological perspective. If the per se cause of the will’s act is the will itself rather than the cognized object, as Henry holds, or if the affection that disposes the will leaves it the freedom to move itself or not, as Morrovalle thinks, then the intellect of the blessed could see God without their will being moved to love him, contrary to the common view that the blessed love God necessarily. Henry’s view that the will of the blessed moves itself necessarily to love God, imposing the necessity on itself (cf. Chapter 3, p. 73), “is affirmed without reason,” according to Godfrey, for if the object that is the (ultimate) end does not have any greater efficiency on the will than some object that is related to the end, then the will remains no less indifferent to move itself to the end than to what is related to the end (Quodl. VI.1, PhB III: 99–102).24 24

On Godfrey’s critique of Henry’s account of necessary volitions by the blessed, see also Osborne 2013.

108

Free Will

On the presupposition that self-motion is impossible, Godfrey also rejects at a later occasion Giles of Rome’s theory according to which, although the will cannot actualize itself, it can determine itself. According to Godfrey, self-determination is in fact a kind of selfactualization; but “that something that is in potentiality to howsoever small and feeble actuality reduces itself from potentiality to actuality is impossible” (Quodl. VIII.16c., PhB IV: 151). Godfrey takes Giles’s determination to be “a true act of the will,” and hence, just as the will cannot actualize itself to an act of willing, so too it cannot determine itself to an act of willing (ibid., 152; Quodl. X.13c., PhB IV: 374–6). Godfrey’s own view is that, rather than moving itself, the will is moved to its act by the cognized object according to the way it is conceived by the intellect. He argues that something good insofar as it is cognized (the thought of a bath) moves the will as an efficient cause, whereas insofar as it is real (the real bath), it moves the will as a final cause (Quodl. VI.7c., PhB III: 163–4).25 Accordingly, the object’s motion of the will by efficient causality is not to be understood as though a presently existing thing makes the will desire or choose it; in fact, the willed object may have real existence only in the future; for example, health as desired by a sick person (Quodl. XV.3c., PhB XIV: 13). Furthermore, that the cognized object moves the will does not mean that the intellect moves the will. In Godfrey’s view, mover and moved cannot even be in the same subject (Quodl. VI.7c., PhB III: 155). Hence intellect and will, which are in the soul as their subject, cannot move each other. Godfrey occasionally expresses himself loosely and speaks of the intellect moving the will or vice versa, but he holds that strictly speaking the intellect only moves the will in the sense that the object that causes the will’s volition must first cause the intellect’s cognition (first in the order of nature, simultaneously in time). The will moves the intellect to exercise its act only indirectly, by moving the sensory power to form phantasms (sensory likenesses of particulars) – say, to imagine something healthy – that allow the intellect to think about the corresponding thing or activity (ibid., 164; ad arg., 170–1; cf. QO 3, PhB XIV: 125). One may suspect that Godfrey’s account implies determinism and thus threatens free will: what causes our intellection and our volition is the known and willed object. Godfrey insists, however, that problems of moral 25

For the different causality of a good as cognized and as real, Godfrey depends on Averroes. Giles holds the same view (see note 11 above), possibly inspired by Godfrey. For the efficient causality of the object, see also Quodl. VI.12c., PhB III: 228; Quodl. X.14c., PhB IV: 381–2; Quodl. XV.2c., PhB XIV: 7.

Intermediary Theories and Strict Intellectualism

109

psychology must not be solved by giving up the most certain metaphysical principles, such as that one and the same thing cannot be in potentiality and actuality, and that what is in potentiality cannot reduce itself to actuality. Rather, confident about the truth of these principles, we must try to understand how free will is compatible with them (Quodl. VI.7c., PhB III: 170). We will consider below Godfrey’s own attempt to explain free will in agreement with these metaphysical principles. Judgment-Volition Conformity While Godfrey differs from Aquinas in attributing to the cognized object efficient rather than formal causality, his foundational assumptions about the will match with Aquinas’s.26 The primary object of the will is the good insofar as it is cognized, or the “cognized suitable being” (ens conveniens apprehensum) (Quodl. VI.11c., PhB III: 220). The will can only will something under the aspect of the good and suitable (Quodl. X.13c., PhB IV: 368). While Godfrey’s voluntarist contemporaries agreed with him so far, the consensus with voluntarists diminishes regarding Godfrey’s assumption that the will is a rational appetite, and moreover that it is free because it is a rational appetite (Quodl. VIII.16c., PhB IV: 155). More controversial still was a further claim Godfrey shares with Aquinas but emphasizes more strongly than he: the choice of the will must always conform to the judgment of the intellect (e.g., Quodl. VI.11c., PhB III: 220; Quodl. VIII.16c., PhB IV: 161, 164). Thus, when one has concluded one’s deliberation, one cannot fail to will accordingly (Quodl. VIII.16c., PhB IV: 164–6). Godfrey gives several arguments for judgment-volition conformity: if one wanted an end, but not that which one understands to be conducive to the end, one would simultaneously want and not-want the end (Quodl. VI.1c., PhB III: 102). Furthermore, if the will could will differently from the intellect’s judgment, “the will could be in its act without an object” (Quodl. VI.10c., PhB III: 204–5), that is, it could will without willing something. Also, since the will cannot will anything unless it is known, the will cannot want to pursue something unless it is apprehended as to be pursued, and it cannot want to avoid something unless it is apprehended as to be avoided (Quodl. VI.11c., PhB III: 220). As we have seen, voluntarists like Henry and Olivi deny judgmentvolition conformity because they consider it a threat to free will. For Godfrey, this does not follow, for the necessity by which the will must 26

For a more detailed comparison of Godfrey’s and Aquinas’s theories of free will, see Lottin 1937b.

110

Free Will

choose in conformity with the intellect’s judgment is not absolute necessity, but conditional necessity. The choice is conditional upon the intended end; if one has concluded after deliberation that, given one’s end, one must make a certain choice, the choice follows. But one can will otherwise if one gives up the end and thus revises the judgment (Quodl. VIII.16c., PhB IV: 164, 172; see also Quodl. XV.4c., PhB XIV: 29). Someone who wants to cross the ocean must choose an adequate means of transportation – unless he or she renounces the end. Godfrey sees his position confirmed by the propositio magistralis, which, as he writes, was conceded by all (Parisian) masters of theology: if “there is no evil in the will unless there is error or nescience in reason” (as he quotes it), then the choice of the will must necessarily conform to the practical judgment. Godfrey insinuates that some articles of the March 7, 1277 condemnation seem to affirm the contrary of the propositio magistralis, but he argues that since the condemnation took place before the Magisterial Proposition was conceded, the articles must be interpreted, if possible, in conformity with this proposition (Quodl. VIII.16c., PhB IV: 165–6). A few years later, Godfrey goes even further and claims that some of the articles are outright irrational as they stand. Among the statements declared erroneous by these articles, there are some that in one way or other affirm judgment-volition conformity. Godfrey mentions Articles 129, 130, 160, and 163,27 and then adds: But these articles, insofar as the surface of the text claims . . . seem indeed impossible and irrational, so that they cannot be rationally maintained unless they are somehow interpreted differently from what the surface of the text, as it stands, seems to claim. (Quodl. XII.5c., PhB V: 102)

Godfrey says that the statements were rightly condemned at the time, but now that the truth has come to further light, they are to be corrected (Quodl. XII.5 ad arg., PhB V: 104).28 Consistent with his insistence on judgment-volition conformity, Godfrey affirms the Socratic deficiency thesis: an evil volition presupposes a perverse judgment. It is not the case that the incontinent (who sin “from passion”) act against their judgment of what they should do. Rather, they act against their judgment that what is immoral and opposed to the divine 27 28

Cf. Chapter 3, p. 61; Article 160 states “that no agent is open to alternatives; rather, it is determined.” For Godfrey’s critique of the March 7, 1277 condemnation, see also Putallaz 1995, 218–24; for references to it throughout Godfrey’s writings, see Wippel 2007. See also Godfrey’s earlier, more cautious, pronouncement on the condemnation, responding to the question of whether a master in theology must teach what he believes is true when it contradicts a proposition condemned by a bishop; Quodl. VII.18, PhB III: 402–5.

Intermediary Theories and Strict Intellectualism

111

precept is to be avoided, but because of strong passion, they act in conformity with their perverse judgment that now, carnal pleasure is to be pursued. So too, deliberate sinners (who sin ex certa scientia) act, because of their bad disposition, according to their perverse judgment that now the sinful act is to be done rather than avoided. In general, “no sinner sins unless he judges that it is now better to do what he does than to omit it” (Quodl. VIII.16c., PhB IV: 172). Godfrey’s Intellectualism From what we have considered so far, the intellectualist outlook of Godfrey’s moral psychology emerges clearly. His account is at odds with virtually all the core claims about intellect and will defended by voluntarists such as Henry of Ghent. (1) Godfrey argues at length and from different perspectives that the intellect is more eminent than the will (Quodl. VI.10, PhB III, 182–218).29 (2) Though he, like Henry, attributes “command” (imperium) to the will, he makes the will’s command depend on a corresponding judgment (Quodl. X.3c., PhB IV: 373, 375) and clarifies that the will uses the intellect not as the superior uses the inferior, but as the inferior uses the superior (Quodl. VI.10c., PhB III: 209). (3) He treats the will throughout as a passive power and insists that it must be moved by the cognized object. He rebukes those who claim that the will cannot be both passive and free (Quodl. VI.7c., PhB III: 160). (4) He affirms that the intellect is no less free than the will (as we will see shortly), and he rejects Henry’s claim that the intellect acts in a servile way (Quodl. VI.10c., PhB III: 208–9). (5) He denies innate freedom in the will. Rather, what causes freedom of the will is its being “an appetitive power [located] in reason” (ibid., 207). (6) The will does not move itself, except per accidens and indirectly, by moving the intellect to find means to the intended end, which then determine the will to desire them (Quodl. VIII.16c., PhB IV: 156; Quodl. X.13c., PhB IV: 375–6). (7) The cognized object is the per se cause of the will’s act. (8) Godfrey affirms judgment-volition conformity, (9) and he claims that an evil will presupposes a perverse judgment. (10) Unlike Henry, he does not argue that an evil will necessarily corrupts reason. Of these theses, the principal ones are the sixth and the eighth claims, which express the same fact from a metaphysical and a psychological perspective: there is no autonomy of the will.30

29

For discussion, see Leone 2017a, 194–212.

30

Pace Leone 2017a, 224.

112

Free Will Freedom and Its Foundation

In Quodlibet VIII (disputed 1292/93), Question 16 – one of Godfrey’s few Questions directly concerned with free will – Godfrey is asked “whether the appetite of a nonhuman animal is free, such that it can be called ‘will’” (Quodl. VIII.16 intr., PhB IV: 140). This Question offers Godfrey the opportunity to clarify the ratio libertatis and its psychological foundation. Following Aristotle (Met. I.2, 982b25–6), Godfrey understands as free a person who exists for his or her sake (causa sui), and as a slave someone who exists for someone else and who needs to be directed by another. This kind of freedom, as Godfrey argues by tacitly following Aquinas (see Chapter 2, p. 44), has a threefold cognitive condition: Someone is free who can direct himself, knowing the end and the formality [rationem] of the end and of the things related to the end, and the relation of those things to the end. (Quodl. VIII.16c., PhB IV: 145)

Like Aquinas, Godfrey argues that nonhuman animals lack freedom, for although they perceive an end, they do not know it as an end, nor do they know how things relate to the end. Thus they follow their natural instinct and are said to be acted upon rather than to act. By contrast, freedom is rooted in the immaterial nature of a thing. Thanks to their immaterial soul, humans are able to bend back upon themselves. This self-reflexivity allows them not only to know the end, but also the formality of the end and the way certain things are related to the end. Thus human beings not only exist for their own sake (causa sui), but they are also for themselves the cause (sibi causa) of their acts and have control of their acts through intellect and will (Quodl. VIII.16c., PhB IV: 146–7). In short, for Godfrey, freedom requires sourcehood, the threefold cognitive condition of which is made possible by self-reflexivity, which in turn presupposes immateriality. Godfrey also argues that immateriality is a sufficient condition for freedom. Thus all immaterial powers are free, and all their acts are formally (that is, essentially) free. The will is free in willing, choosing, and loving, the intellect in knowing, judging, and deliberating (ibid., 150, 155–6; ad arg., PhB IV: 175). Godfrey insists on this point against Giles, who in Godfrey’s account holds that the intellect has freedom only as the cause of freedom while the will alone has freedom formally, and that it is thanks to the will’s determination that one controls one’s act (ibid., 148–56; cf. p. 102 above).31 Godfrey wonders, in fact, why people always understand freedom 31

See also Quodl. X.13c., PhB IV: 373–5. For Godfrey’s response to Giles, see also Putallaz 1995, 227–45.

Intermediary Theories and Strict Intellectualism

113

as freedom of the will and why they think that humans control their acts because of their will, as if freedom belonged only to the will (ibid., 156). According to Godfrey, the immaterial nature of the soul not only allows for the kind of sourcehood required for freedom, but also for alternative possibilities. He basically presents Aquinas’s argument, in a highly condensed form: when the intellect understands something, it is not determined by its cognized form (that is, by its concept) to one result, but it can understand the form and its opposite; likewise, when acting according to this cognition, one can do one thing or the opposite (Quodl. VIII.16c., PhB IV: 147–8). Recall Aquinas’s example: the concept of a house is not determined to existence or nonexistence or to a particular instantiation; thus one remains free to will it or not, or to will it as a square or a round house (see Chapter 2, p. 43). With explicit reference to Aquinas, Godfrey takes the will’s alternative possibilities in choosing to be grounded in reason’s ability for deliberation (Quodl. VI.10c., PhB III: 207–8). Alternative possibilities are not, however, essential to freedom. Freedom is compatible with necessity of immutability or absolute necessity, by which we necessarily desire our ultimate end, happiness; but it is incompatible with necessity of coercion. Absolute necessity applies only to the desire of the ultimate end; other things are desired by conditional necessity, contingent upon the end pursued and the practical deliberation made. Godfrey suggests that his distinction between what is willed by absolute necessity and what is willed by conditional necessity corresponds to the distinction made by “some” – he apparently thinks of Henry of Ghent – between what is willed freely albeit without free decision and what is willed freely and with free decision (Quodl. VIII.16c., PhB IV: 162–5).32 Godfrey sees his view confirmed by Anselm of Canterbury. Freedom as the ability to preserve rectitude for the sake of rectitude does not presuppose the ability to do otherwise, but rather to act intentionally for an end, for example, doing what is just because it is just. Such freedom involves precisely the threefold cognitive condition of sourcehood: to know the end as end, to know the means as means, and to know how the means relate to the end. Thus, one is oneself the source (sibi causa) of one’s act of desire, as opposed to a heavy object, which desires its end (the lower place) naturally, and to nonhuman animals, who desire their end “quasi-naturally” by the mere perception of the thing that is the end (Quodl. VIII.16c., PhB IV: 32

Godfrey’s own use of “free decision” is not as consistent as Henry’s; in Quodl. VI.7c., III: 160, he presents it as essentially involving alternative possibilities; in Quodl. VIII.16c., PhB IV: 163, he makes the contrary claim.

114

Free Will

162–3). It is clear, then, that Godfrey insists no less on sourcehood than voluntarists, such as Henry of Ghent or Peter Olivi, but Godfrey conceives of it mainly in cognitive terms and does not think it implies per se selfmotion. Moral Responsibility We are free, then, even if we cannot will or act otherwise. Can we also be praised or blamed for things we could not have done otherwise? Godfrey does not answer this question in a consistent manner. In Quodlibet VI, he requires as a necessary condition for praiseworthiness and blameworthiness that the act be in the agent’s power (in potestate ipsius), that is, that the agent have control (dominium) of it. As necessary conditions for such control, in turn, he insists on sourcehood and excludes absolute necessity: But this implies nothing else than that the agent acts of his own accord [sponte] and freely, not from coercion [coacte] nor by absolute necessity (which is not far from coercion), in the way that nonhuman animals act necessarily. (Quodl. VI.12c., PhB III: 233; see also ibid., 237)

Control is here presented as a narrower notion than freedom, for while freedom is compatible with absolute necessity (and hence with the absence of alternative possibilities), control is not. In Quodlibet VIII, by contrast, he insists on sourcehood alone: God and the blessed immutably and freely have a good will and deserve praise; the devil immutably and freely wants evil and, it is implied, deserves blame. What is essential to moral responsibility is to be oneself the cause of one’s volition: “When I will something immutably because it is essentially good or because it is ordered to an essential good, I will in a praiseworthy way despite its immutability . . . because I am by myself [mihi] the cause of this volition” (Quodl. VIII.16c., PhB IV: 163–4). Would Godfrey consider the devil blameworthy even if he could not have helped becoming obstinate? Or does he trace the blameworthiness of the devil’s current state to an earlier state in which he could have done otherwise? Presumably he does. Godfrey quotes approvingly Augustine’s axiom that something counts as sin only if one was able to avoid it (Quodl. X.13c., PhB IV: 377; lib. arb. III.18.50.171). Also, he affirms the common view that one is excused from a mortal sin (that is, a sin that dissolves the friendship with God and neighbor) if one lacked full control – the ability to do an act or not (Quodl. XV.2c., PhB XIV: 8–9). But Godfrey nowhere claims that a praiseworthy act presupposes the ability not to do it, or to do a bad act instead; so on Godfrey’s considered account, moral

Intermediary Theories and Strict Intellectualism

115

responsibility seems to be asymmetrical. What counts, it seems, is the ability to do what is right; one is responsible for a sin because one had the ability to do what is right. For example, one can avoid giving in to unlawful pleasures by repelling the thought of them; if one fails to repel them, one deserves blame (cf. Quodl. XV.4c., PhB XIV: 30–1). Control How can Godfrey explain control – to be the cause of one’s acts – given his conviction that the will is moved by the cognized object? He holds that the reason why nonhuman animals are not praised or blamed for their perception or desire is that they are not for themselves the cause (sibi causa) of perception and desire through any act of theirs, but rather, perception and desire are caused by the object alone (Quodl. VIII.16c., PhB IV: 159). Would not the same apply to human beings? If the object moves our intellect and will, how can we be in control? Godfrey’s solution is that although the object has a determinate causality on the will, the will controls whether an object moves the will. Although every human act is determined by an object, humans determine themselves, for they are determined only to the good simply speaking, not to any particular good (Quodl. V.12c., PhB III: 59). One of his explanations is this: we control the practical judgment that causes our choice. He draws on a statement by Augustine that played an important role in the free will debate: “what one accepts or rejects is in one’s power, but by which sight one is affected is beyond control” (lib. arb. III.25.74.255). Accordingly, writes Godfrey, the first affective or volitional reaction to something that comes to mind is not in one’s control, but then the will can move the intellect to deliberate about whether to pursue or avoid it, and in this way one is for oneself the cause (sibi causa) of the judgment resulting from deliberation, and thus of the choice resulting from the judgment. Therefore, we control what we choose.33 Since it is by deliberation that we control our practical judgment and choice, the big question is how we control our deliberation. Essentially, deliberation consists of an inference about means-end relations or of a series of such practical inferences, each of which takes the form of a practical syllogism (cf. Quodl. V.12c., PhB III: 59). The way we deliberate, then, depends on the premises from which we make our practical inferences. From which premises we reason may depend on our virtuous or vicious dispositions. Godfrey discusses the different deliberations of the 33

Quodl. VIII.16c., PhB IV: 159–61; cf. Quodl. VI.7c., PhB III: 161.

116

Free Will

intemperate and the temperate concerning adultery: the intemperate reason from the premise that carnal pleasure is to be pursued, the temperate from the premise that nothing against right reason is to be done; the intemperate conclude that they should commit adultery, the temperate that they should not (ibid., 50–1). Yet in case of strong dispositions determining what we consider worth doing, there is no open-ended deliberation, and hence no room for control. That we control our deliberation presupposes that we control which things we consider in general worthy of pursuit (say, pleasure or decency) and which things we consider to be suitable as means to the end – in other words, that we control from which premises we make our practical inferences. In Quodlibet XV (of 1303–4), Godfrey presents an elaborate reflection of how we control what we consider. This theory was influential on thinkers of the next generation, and we will return to it when discussing John of Pouilly in Section 5.2. The context is a discussion of the exercise/determination distinction, which Godfrey understands as a distinction of reason, for exercise and determination are inseparable: one cannot bring about an act without bringing about a determinate act. Yet one can distinguish exercise and determination as two different aspects of an act. As Godfrey explains, the same effect can be traced to two causes: to the active principle that causes the effect in the passive principle, and to that which approaches the active and the passive principle. Thus, the warming of a pot of water can be traced to the fire that warms it and to that which applies the fire to the water. That the effect is of a certain kind – warming rather than cooling – must be traced to the fire, the active principle; that the water undergoes the fire’s action in the first place must be traced to that which applies the fire to the water. Just so, the will’s act can be traced to two causes: one is the object presently considered by the intellect, which is the active principle that determines the act of the will; the other is that which makes the object present in the imagination and the intellect in the first place. Though both causes act together to bring about the will’s exercise and determination, its determination can be traced more to the intellect or the cognized object, while its exercise can be traced more to that which moves the intellect to think about the object. As regards control, a volition is not up to the will upon the first encounter with an object, for example, the will’s immediate reaction to the thought of committing adultery or a theft. In such a first volition, the will moves itself neither as to its determination nor as to its exercise, because before the will already wills something, it cannot apply an object to itself. But once the will is made to have an act, it can move itself to exercise other acts, by moving the

Intermediary Theories and Strict Intellectualism

117

cognitive powers to consider something as to be pursued or avoided (Quodl. XV.2c., XV.4 ad 1, PhB XIV: 7–11 and 32). An earlier version of Godfrey’s “application theory” offers further clarification. The will applies the intellect to exercise its act by making it consider certain objects – simple things or propositions – which then cause the intellect’s act. Thus the will controls the intellect by uniting an object (the active principle) to the intellect (the passive principle). But the will does not control whether an object determines the intellect in one way or another – the intellect may even remain indeterminate, such as when the will makes the intellect consider whether the number of stars is even or odd; in that case, the object is not evident enough to determine the intellect to one or the other. Likewise, the will does not determine the intellect to a specific act when it makes the intellect consider what is conducive to an end, for it is not in the will’s control what the intellect judges as expedient and thus as to be chosen. Godfrey assigns to the will only indirect control of what the intellect judges, by generating a condition (dispositio) in the knower that makes him or her apprehend something as suitable. Thus the will controls indirectly what appears as good (Quodl. VI.11c., PhB III: 222–4). As we have seen, to the intemperate, to commit adultery appears as good; to the temperate, it appears as bad (Quodl. V.12c., PhB III: 50). Yet the application theory leaves open the question of what moves the will to apply an object to the intellect in the first place, and to apply one object rather than another. Furthermore, indirect control of what appears as good and hence of what we choose does not explain how we control the acts that originally caused the condition or disposition making certain things appear as good. What is it, then, that moves the will to make the intellect consider a certain object, and how do we control what so moves the will? Giles of Rome assumed that the will controls directly the direction of the intellect’s consideration, for example, whether to stay on the consideration of fornication as pleasant or to desist from it and to consider it instead as indecent. But Godfrey denies that the will controls anything directly. The will cannot move the intellect to its exercise and make it stay on its consideration without the intellect previously judging that such motion of the intellect is to be done, or as Godfrey also puts it, without the cognized object causing the will to want the intellect to stay on this consideration.34 Nevertheless, one can repel the considerations that occur 34

For passages directed against Giles’s account in Quodl. III.15, see Quodl. VIII.16c., PhB IV: 153–5, 159; Quodl. X.13, PhB IV: 374–5. See also Quodl. VI.10c., VI.11c., PhB III: 209, 222; Quodl. XV.3c., XV.4c., PhB XIV: 15, 29.

118

Free Will

during deliberation, thanks to the ability to reflect upon one’s own acts. As Godfrey writes, while a man is thinking about which form of life to adopt, the thought of a woman or of something delightful may come to his mind; he can repel such thoughts, and if he fails to do so, he is responsible for this failure. But no “per se and necessary cause” can be given for the failure to repel the consideration; rather, it can be caused by an infinite number of incidental causes, such as negligence in deliberation, a sensual life (delicata vita), or occupation with worldly things; and these in turn do not have any per se and necessary cause (Quodl. XV.4c., PhB XIV: 30). According to Godfrey, negligence cannot be traced to a determinate cause; nevertheless, a person can avoid being negligent (Quodl. X.13c., PhB IV: 377). This is about as far as Godfrey takes us in explaining how we control what moves our volitions: while deliberating, we have some control of what we think about – we apply certain objects to the intellect and so we consider them, and we do so not blindly but for a reason. Other things may come to our mind without us directly applying our intellect to them, but we can repel such thoughts or fail to do so, and the failure can be explained only up to a certain point. Our moral dispositions are a factor in how we see things but do not explain how we acquired these dispositions in the first place. In the last analysis, the question of how we control the causes of our choices remains unanswered. Voluntarists explicitly assume brute preferences, that is, volitions that are primitive; in fact, they argue that the will can will differently from a practical judgment without requiring a new practical judgment. Aquinas traces at least some of our preferences to God’s causality on our will (see Chapter 2, p. 50), a solution Godfrey is skeptical about.35 Godfrey traces all volitions to our cognition of an object. And yet, it seems that his account cannot do without brute preferences – to dwell on a consideration or to repel it; to be negligent or not – that are in our control and that we are responsible for. We may suspect a hidden residue of voluntarism in his account. Analyzing Godfrey’s account of the sin of the angels (Section 8.4) will further illuminate whether he allows for certain acts of the will that are not fully traceable to the cognized object.

35

Quodl. X.14c., PhB IV: 381; but see also his neutral exposition of this view in Quodl. XV.2c., PhB XIV: 11.

chapter 5

Refinements and Radicalizations

Within fifteen years after Aquinas’s death, by 1289, the principal positions in the free will debate provoked by Aristotelian ideas had already been theorized and publicly defended. In that year Henry of Ghent debated his Quodlibet XIII, Question 11 and Godfrey of Fontaines his Quodlibet VI, Question 7, two landmark treatments of free decision that champion strongly opposed stances. By then, attempts at a compromise, by Giles of Rome and John of Morrovalle, had also been made. Theologians of the following generation, active in the first two decades of the fourteenth century, are intimately familiar with these theories and partly refine or partly radicalize them. In this chapter we will examine John Duns Scotus, who takes voluntarism a step further, loosening the will’s connection to happiness and formulating with greater clarity than before some core voluntarist ideas. Then we will consider John of Pouilly as a strong representative of intellectualism, and Hervaeus Natalis and Durand of St. Pourçain, whose accounts of free will share the same presuppositions as Pouilly’s but provide alternative solutions to a problem that forms a vulnerable point in Pouilly’s theory. Next we will discuss Peter Auriol, who developed a sophisticated intermediary theory. We will close with William of Ockham, who disconnects the will further from the good than anyone had done before.

5.1

John Duns Scotus

With Duns Scotus, a Franciscan whose scholarly activity spans from the 1290s to his death in 1308, the voluntarist movement achieves maturity. Scotus dissolves a tension in the free will theory of Henry of Ghent, who breaks only half-heartedly with Aquinas. Henry calls the will a rational appetite, but does not strictly conceive of it as such, for he thinks that the will can act contrary to reason’s final judgment. Yet he holds that the will has a necessary inclination to what is simply good, that every rational being necessarily desires happiness and everything else for the sake of happiness, 119

120

Free Will

and that the blessed necessarily love God (see Chapter 3, pp. 63–4). By contrast, Scotus abandons the notion of the will as a rational appetite. For Scotus, the will as rational appetite would not be free, but more fundamentally, it is essentially not even an appetite, but rather an active power (or an active principle) – one that is by itself essentially undetermined. This novel notion of the will has far-reaching implications, and Scotus defends them wholeheartedly: rational creatures do not necessarily desire happiness, nor do they want other things necessarily for the sake of happiness. The blessed, who see God’s all-encompassing goodness, could in principle stop loving him. Accordingly, Scotus abandons a eudaimonistic conception of ethics. The Will Voluntarists of the previous generation insisted that the will is an active and not a passive power. Scotus agrees (e.g., Lect. II.25 nn. 28–31, XIX: 236–8), but what matters to him is that it is an active power of a unique kind. He defines the will by means of Aristotle’s notion of rational active power, in contrast to irrational active power, a distinction Scotus maps onto the Augustinian opposition between will and nature (e.g., lib. arb. III.1.2.9). Recall that for Aristotle, rational powers can produce contrary effects, irrational powers only one determinate effect: medical art can produce health or disease, the hot can produce only heat (Chapter 1, pp. 26–7). To describe the nature of rational powers, Scotus adopts Henry of Ghent’s intuition that what characterizes the will primarily is its “mode of eliciting the operation,” and only secondarily its relation to its object, that is, its ability to will this or that, for it is only due to the will’s unique eliciting mode that it has alternative possibilities in a relevant way (QMet. IX.15 n. 21, OPh IV: 680; cf. Chapter 3, p. 72). The different eliciting modes divide active powers into two kinds: There are only two general modes by which a thing can elicit its own operation. For either a power is as such [ex se] determined to act, so that, as far as it is concerned [quantum est ex se], it cannot not act when it is not impeded from without. Or it is not as such determined, but it can do this act or the opposite act, and it can also either act or not act. The first power is commonly called “nature,” the second is called “will.” (QMet. IX.15 n. 22, OPh IV: 680–1)

The will is here defined as a power that is not as such determined to its act, and so it can do otherwise even under determinate circumstances. Natural

Refinements and Radicalizations

121

powers, too, can cause opposite effects: the sun can solidify and liquify. But the eliciting mode of natural powers is determined, and hence, depending on what they act on, they produce determinate effects: the sun hardens mud and melts ice.1 What remained implicit in the above quotation is made explicit further down: the indeterminacy of the will is not such that it needs to be determined from without, as unformed matter must be determined by a form, but rather such that it can determine itself, thanks to what Scotus calls the will’s “superabundant sufficiency.”2 The will’s distinctive eliciting mode marks it off from every other active principle (QMet. IX.15 n. 43, OPh IV: 687), indeed, from everything else in the universe (Lect. II.25 n. 93, XIX: 261; Ord. II.39 n. 25, VIII: 464). For Scotus, the will’s self-determining eliciting mode is precisely what makes it free. Freedom is thus intrinsic to the will, and hence the nature/will dichotomy coincides with the distinction between natural and free powers (e.g., Lect. II.34–7 nn. 107–8, XIX: 353). Aristotle’s rational power, which can have opposite effects, then, is the will, since it can freely determine itself. The intellect, by contrast, is strictly speaking Aristotle’s irrational power, that is, a natural power, for the intellect does not by itself control whether it understands something or not, nor whether it assents to or dissents from a proposition. The intellect needs the will to determine it to one outcome in case it remains indeterminate (as in the case of uncertainty). And it is only insofar as the will controls the intellect that the intellect is a power of opposites.3 Scotus uses the expression “rational appetite” for the will rarely and only when the use of the term is suggested by the context.4 When he discusses the will in its own right, he defines it not as an appetite, but as a power. The specifying difference is the ability for opposites (as in the above block quote) or freedom.5 For Scotus, the will cannot be an intellectual appetite, since the intellect is a natural power; so a will that is bound to follow the intellect would be a natural power as well. Scotus expresses this in terms of Anselm’s two affections of the will: a purely intellectual appetite would be like a will having only the affection for the advantageous, desiring by natural necessity whatever is judged suitable to the intellectual nature. As 1 2 3 4 5

QMet. IX.15 nn. 5, 15, 43, XVII: 676, 678, 687; Ord. I.2.2 nn. 349–50, II: 334. QMet. IX.15 nn. 31–2, 67, OPh IV: 683, 696; see also Ord. I.8.2 n. 298, IV: 324. QMet. IX.15 nn. 35–41, 67, OPh IV: 684–6, 697; Lect. I.2.2 n. 226, XVI: 197–8; Ord. I.2.2 nn. 346–50, II: 333–4; cf. Ord. I.1.1.1 n. 22, II: 16. In the Ordinatio, “rational appetite” is used only four times: Ord. II.30–2 n. 52, VIII: 340; Ord. III.15 n. 68, IX: 509; Ord. III.17 n. 9, IX: 565; Ord. III.33 n. 44, X: 162. On Scotus’s theory of the will as a power rather than an appetite, see also González Ayesta 2014.

122

Free Will

a purely intellectual appetite, it would lack the affection for justice, which according to Scotus grounds freedom and allows one to check the desire for benefit according to the demands of morality. In short, a purely intellectual appetite would not be free.6 Scotus considers deliberation to be crucial for acting virtuously, indeed for acting in a human way, for otherwise one acts without understanding (Ord. III.33 n. 76, X: 175). Yet deliberation does not determine the will: Scotus denies judgment-volition conformity. While the intellect dictates something to be done, the will can abstain from choosing, or choose the opposite.7 While Scotus agrees on that with Henry, he rejects Henry’s theory of the blinding of the intellect. The will cannot make the intellect judge falsely about moral principles or deductions from moral principles, for the principles and syllogistic deductions are self-evident and thus independent from the will (Ord. III.36 n. 65, X: 249).8 Like other voluntarists, Scotus does not think the denial of judgmentvolition conformity implies that the will can will evil under the aspect of evil. Since the will’s primary object is the good, it can will only that in which is found an aspect of good, and it can detest only that in which is found an aspect of evil.9 Contingency For Scotus, the will determines itself, and it does so contingently.10 Contingent self-determination grounds moral responsibility, that is, as 6

7 8 9

10

Lect. II.39 n. 28, XIX: 385; Ord. II.6.2 nn. 49–51, VIII: 48–51; Ord. II.39 nn. 22–3, VIII: 463; Rep. II-A.25 n. 20, Viv. XXIII: 128b. On Scotus’s theory of the two affections, see Williams 1995, King 2010, and Section 9.5. Lect. III.36 n. 69, XXI: 331; Ord. III.36 nn. 65, 72, X: 245, 249; cf. QMet. IX.15 n. 55, OPh VI: 692. See for a more detailed account Hoffmann 2012a. E.g., QMet. IX.15 n. 68, OPh IV: 697; Quodl. q. 16 n. 24, NR, 171–2. See also note 21 below. Nevertheless, in discussing the sin of deliberate evildoing (peccatum ex certa malitia), Scotus mentions an explanation of deliberate evildoing as willing evil under the aspect of evil, which he neither endorses nor refutes; see Ord. II.43 nn. 3–4, VIII: 484–5; cf. Lect. II.43 nn. 4–5, XIX, 402. Also, in Scotus’s Parisian collatio 8 (or 17 in the Wadding/Vivès edition), the will’s ability to desire evil under the aspect of evil is defended with multiple arguments (nn. 10–14, Viv. V: 216b–218b); for discussion, see Alliney 2002. Since Scotus’s collationes are school exercises, they do not necessarily reflect his own view. For the nature of the Parisian collationes, see Alliney 2020, 606–7. Rep. II-A.25, n. 20, Viv. XXIII: 127b–128a: “. . . illud igitur quod contingenter evenit . . . oportet quod in illo instanti, in quo evenit, eveniat a causa indeterminata potente se determinare ⟨contingenter, . . . sed nulla causa creata alia a voluntate, si est indeterminata, potest se determinare⟩.” The passage in angle brackets is missing from the Wadding/Vivès edition, but witnessed in all known manuscripts (apart from an insignificant variant in one manuscript). The manuscripts are listed in Dumont 2018b, 437. Scotus’s Paris lectures on the Sentences have come down to us in student reports. For Book II, there are two versions, marked A and B. Reportatio II-A offers a more

Refinements and Radicalizations

123

Scotus writes, the fact that an act is “imputable”: praiseworthy or blameworthy; deserving reward or punishment (Quodl. q. 18 n. 9, Viv. XXVI: 241b). To connect the ability to will otherwise to the will causing its acts contingently is nothing novel – Aquinas, for example, did the same (e.g., ST I-II.10.4c.). But Scotus interprets this ability to will otherwise differently from the standard view. What is uncontroversial is that there is a diachronic ability to will otherwise, namely the ability to will differently than presently at a future moment: Diachronic ability to will otherwise: the will willing A at t1 can will non-A at t2.

For example, I presently want to eat fish for dinner, but then I can change my mind and prefer meat for dinner. But inspired by Olivi (see Chapter 3, p. 89), Scotus moreover affirms a “less obvious” synchronic ability to will otherwise (or synchronic contingency) in the same instant, that is, in the present moment in which one chooses: Synchronic ability to will otherwise: the will willing A at t1 can will non-A at t1.

For example, I presently want to eat fish for dinner, and while I want this, I can want to eat meat for dinner instead. Scotus writes: “The will willing A in this instant and regarding this instant [in hoc instanti et pro hoc instanti] can will-against [nolle] A in the same [instant] and regarding the same [instant]” (Lect. I.39 n. 50, XVII: 495).11 So if “The will wills A at t1” is true, then “The will can will non-A at t1” is true as well. Scotus understands the synchronic ability to will otherwise not merely in a weak sense, as a logical possibility, as if he were only saying: “it is conceivable that the will, which wills A at t1, will non-A at t1.” So understood, Aquinas would have agreed. But this means only that the will could have willed otherwise, not that it can now will otherwise. Rather, Scotus understands it in a strong sense, as a real ability: the will, which wills A at t1,

11

complete text than II-B, but II-B also has text not contained in II-A. While Reportatio II-B is unedited, Reportatio II-A is printed in Luke Wadding’s edition of Scotus’s works (1639), reprinted by Louis Vivès (1891–5). This edition of the Reportatio II-A has many small divergences from the original text and some omissions; a substantial one is indicated in Chapter 10, note 12. Dumont 2018b clarifies long-standing confusion about the Reportationes for Books I–II and their relation to so-called Additiones. Apart from a few exceptions, Scotus uses the preposition “in” to indicate the moment the psychological act (of cognition or volition) takes place, and the preposition “pro” to indicate the moment the act refers to (e.g., to think that something will happen or to want something to happen tomorrow).

124

Free Will

is able to will non-A at t1. The alternative volition is presently available to the will, not merely logically possible. Thus, under the hypothesis that a will existed in a single instant, it would nevertheless have alternative possibilities.12 Scotus’s point is the same as Olivi’s: one can only act in the present, not in the future; so if the will did not have access to alternative volitions in the present, it would never have access to alternative volitions, and so it would not act contingently but necessarily.13 Scotus addresses some difficulties involved in his theory of the will’s synchronic ability to do otherwise; two difficulties are especially important. The synchronic ability to do otherwise implies that the present is contingent (just as the future), against the common assumption that the present is necessary (like the past). Aristotle seems to affirm the necessity of the present: “whatever is, necessarily is when it is” (Int. 9, 19a23–4). In response to this difficulty, Scotus defends the contingency of the present and interprets Aristotle’s dictum as stating that “everything that is, ‘is necessarily when it is’,” rather than “everything that is when it is, ‘is necessarily’.” The first statement expresses the necessity of the consequence (that is, the necessity of the logical inference), the second expresses the necessity of the consequent (that is, its necessary being). According to another difficulty, if the will that is presently willing A actualizes its power for willing non-A while willing A, then contradictories are simultaneously the case. The contradiction is avoided by restricting the actualization of the power for the opposite to a later moment – but then the will would not have, as assumed, a synchronic ability to do otherwise, only a diachronic one. Scotus solves this difficulty by claiming that the ability for the opposite volition precedes the actualization of this ability in the order of nature, but is temporally simultaneous.14 Like Henry of Ghent in Quodlibet X (see Chapter 3, pp. 82–3), Scotus essentially splits the single temporal moment into two instants of nature and admits that contradictories can be the case at the same instant of time, although not at the same instant of nature.15 12

13 14 15

Lect. I.39 nn. 47–52, XVII: 494–6; Ord. I.38.2–I.39, VI: 417–20; Rep. I-A.39–40 nn. 39–44, WB II: 476–8. The hypothesis of a will existing in a single instant alludes to a thought experiment Robert Grosseteste had employed to explain the possibility of angels sinning in the first instant of their existence; see Grosseteste, De libero arbitrio, recensio prior 9.5, ed. Lewis, 54; recensio posterior 7.14, p. 148. This work was composed before 1235; see Introduction, xxvii–xxx. On the relation between Scotus’s and Grosseteste’s theories of the synchronic ability to will otherwise, see Lewis 1996. QMet IX.15 n. 65, OPh IV: 696; Lect. I.39 n. 60, XVII: 500; Rep. I-A.39–40 n. 42, WB II: 477. Lect. I.39 nn. 55, 57–8 and 60, XVII: 498–500; Ord. I.38.2–I.39, VI: 421–3; Rep. I-A.39–40 nn. 45, 47, 49–50, and 54–5, WB II: 478–80. Rep. I-A.39–40 n. 54, WB II: 480. – For Scotus’s theory of the synchronic ability to will otherwise, see Knuuttila 1981, 217–30; Dumont 1995 (also on the roots of Scotus’s theory in Olivi); Normore 2003, 130–7. For a critique of Scotus’s theory, see MacDonald 1995.

Refinements and Radicalizations

125

Scotus formulates a significant implication of the contingency of the will’s acts: that the will wills something or not is ultimately an irreducible fact, and therefore the will’s act can have at the most a penultimate explanation. As Scotus puts it, “the will wills A” is either itself an “immediate contingent proposition,” or it can be traced to one. His interlocutors were familiar with immediate necessary propositions: self-evident first theoretical principles (such as the principle of noncontradiction), which are immediate because they cannot be explained through other propositions. Immediate contingent propositions are likewise not further explainable, but they are not self-explanatory. Instead, Scotus writes, they are knowable through intuition, that is, a posteriori (Ord. III.7.1 n. 8, IX: 264). To show that a proposition describing an act of will either is itself an immediate contingent proposition or can be traced back to one, Scotus argues that what is contingent does not follow from necessary antecedents. Hence, a contingent proposition expressing a contingent state of affairs must either be itself immediate, or traceable to an immediate contingent proposition. To give an example, that I want to use a ship is a contingent proposition; if it has no further explanation, it is immediate. If it does have an explanation, then no more than a part of that explanation can be a necessary proposition: a ship is necessary to cross the sea. But if necessary propositions fully explained the act, then the act would not be contingent. So at least a partial explanation must lie in another contingent proposition, such as that I want to cross the ocean. If this proposition is not immediate, it must be further traced back – until one stops at an immediate contingent proposition expressing a bare fact lacking further explanation. Put differently, Scotus writes that no other reason can be given for why the will wills A than that it is a will.16 A more radical innovation is Scotus’s claim that rational creatures will contingently in all of their volitions, including the desire for happiness and the love of God by the blessed. Under the knowledge conditions in this life, we grasp the ultimate end in a general notion: happiness. Scotus admits that in a sense the will tends necessarily to happiness, but it does so only by its natural inclination, not by an elicited act.17 Scotus expresses this claim in technical language: the will does not necessarily enjoy (fruitur) the end obscurely known in general.18 Scotus explains that while “enjoyment” can 16 17 18

QMet. IX.15 nn. 26, 28–9, OPh IV: 682; Lect. prol. 3.1 nn. 115–17, XVI: 41–2; Ord. prol. 3 n. 169, I: 112–13; Ord. I.8.2 nn. 299–300, IV: 324–5; Ord. II.1.2 n. 91, VII: 47–8. Ord. II.6.2 n. 56, VIII: 53; Ord. II.39 n. 24, VIII: 464. For Scotus’s theory of natural inclinations of the will, see González Ayesta 2012. Ord. I.1.2.2 nn. 82, 143, II: 62, 96; Rep. I-A.1.2.1 n. 35, WB I: 95.

126

Free Will

mean “delight” (cf. Ord. I.1.2.1 nn. 69–73, II: 52–6), in this context, he understands “to enjoy” as an elicited act of the will: wanting something in itself, whereas “to use” is wanting something for the sake of something else (Ord. I.1.2.1 nn. 64, 68, II: 48, 51). Since the will is free to elicit or not any of its acts, it can will (velle) happiness or not will it (non velle).19 Aquinas too granted that it is possible not to will happiness – but only as long as one does not think about it (DM 6c., XXIII: 150 ll. 435–40). According to Scotus, however, even while thinking of happiness, one can suspend the act of willing it.20 But one cannot will-against (nolle) happiness, for there is no aspect of evil in it, nor can one will misery, for it contains no aspect of good.21 Scotus also holds that it is not necessary that whatever is willed be willed for the sake of happiness. In fact, one can will things without thinking about happiness; one can also will something without ordering it to any further end. Furthermore, one can will something that is incompatible with happiness, as when someone wants to fornicate, fully aware that this act cannot be ordered to the happiness that consists in the enjoyment of God (Rep. IV-A.49.8–9 n. 18, Viv. XXIV: 667a–668a). We might object that even then happiness is still desired, albeit under some other description: sensual enjoyment. Scotus does not address this objection. He certainly considers it possible to set for oneself as one’s ultimate end sensual enjoyment. In fact, he thinks like Olivi that because of the will’s freedom, one can set for oneself anything as one’s ultimate end (Lect. I.1.1.1 n. 20, XVI: 67; Ord. I.1.1.1 n. 16, II: 10). Scotus directly embraces a hypothesis that Godfrey of Fontaines feared was implied in Henry of Ghent’s theory of the will’s self-motion (see Chapter 4, p. 107): the blessed in the beatific vision do not necessarily love God; expressed in technical language: the will that is supernaturally elevated by the disposition (habitus) of charity does not necessarily enjoy the ultimate end clearly seen.22 In the beatific vision, not only the willed object is flawless (as in the case of willing happiness), but so too is its cognition, and yet, according to Scotus, its love does not necessarily follow. The foundation of this claim is his conception of the will as defined by its 19 20 21 22

E.g., Lect. I.1.2.2 n. 118, XVI: 100; Quodl. q. 16 n. 22, NR 170–1. Lect. I.1.2.2 n. 109, XVI: 97; Rep. I-A.1.2.1 n. 35, WB I: 95; cf. Ord. I.1.2.2 n. 77, II: 59. Lect. I.1.2.2 n. 118, XVI: 100; Ord. IV.49.1.6 n. 355, XIV: 379; Rep. IV-A.49.9 n. 11, Viv. XXIV: 664a; Quodl. q. 16 n. 24, NR, 171–2. Ord. I.1.2.2 nn. 82, 145, II: 62, 97; Rep. I-A.1.2.1 n. 48, WB I: 101. For Scotus’s theory of nonnecessary beatific enjoyment and the reception of this theory by Franciscans, see Alliney 2015 and his articles cited there (p. 272, note 3).

Refinements and Radicalizations

127

unique eliciting mode. The eliciting mode belongs to the essence of a power, and hence does not change. Natural powers act by natural necessity, but the will elicits its acts freely – no matter what it wills or under what conditions it wills (e.g., those of this life or the next). The will wills without necessity the things related to the ultimate end; for Scotus, this shows that it wills them freely. Given that the will’s eliciting mode does not change, the fact that the will wills some things freely is proof that it wills everything freely, in whatever situation. Accordingly, Scotus argues, the will also wills the ultimate end freely, that is, contingently (Ord. I.1.2.2 nn. 80, 136, II: 60, 91). Note that in this argument, contingent willing (of what is related to the end) implies free willing (as the will’s characteristic eliciting mode), and free willing implies contingent willing (of the ultimate end). In fact, Scotus often treats “freedom” and “contingency” as equivalent. Sometimes he moves back and forth between “freedom” and “contingency,” as if they were synonyms.23 He repeatedly claims that freedom implies the ability to will otherwise24 and lack of freedom the inability to will otherwise,25 and that necessity or lack of alternative possibilities implies lack of freedom.26 That the blessed do not necessarily enjoy God has a problematic implication. According to Scotus, beatitude consists precisely in the enjoyment of God, that is, in loving him for his own sake.27 So if the blessed suspended their act of enjoyment – their act of loving God – they would no longer be blessed; worse still, they would sin. But according to the Christian faith, the blessed are impeccable and their beatitude is permanent. To avoid the implication that beatitude can be lost, Scotus assumes that the love of God by the blessed is made necessary extrinsically, by God himself, who prevents their will from suspending their act of loving him.28 He takes great pains to show that this divine intervention does not threaten their freedom.29 Necessity It seems, then, as though the will wills everything contingently. In his Lectura, the script for his lectures on the Sentences at Oxford and the earliest 23 24 25 27 28 29

E.g., Lect. I.1.2.2 n. 111, XVI: 97 ll. 18–20. E.g., Lect. I.39 nn. 45–52, XVII: 493–6; Lect. II.25 n. 74, XIX: 255. E.g., Ord. II.39 n. 22, VIII: 463 ll. 127–8. 26 E.g., Lect. III.33 n. 40, XXI: 279 ll. 202–3. Ord. IV.49.1.1–2 nn. 80–98, XIV: 309–14; Ord. IV.49.1.5 nn. 282–5; XIV: 360–1; Rep. IV-A.49.4 nn. 2–5, Viv. XXIV: 636a–638a. Ord. IV.49.1.6 nn. 348–9, 365, XIV: 376–7, 381–2; Rep. I-A.1.2.1 n. 48, WB I: 101. Ord. IV.49.1.6 nn. 357–63, 366–7, XIV: 379–80, 382.

128

Free Will

of his Sentences commentaries, Scotus even says so: “the will wills nothing necessarily” (Lect. I.1.2.2 n. 117, XVI: 99). In considering the divine will, however, Scotus refines his account. God’s will loves necessarily the infinite good (that is, his own essence) by an infinite love; this infinite love proceeds as a distinct subsistence, the Holy Spirit (Ord. I.10 n. 9, IV: 341–2). The production of the Holy Spirit, which the medievals call “spiration,” thus results from a necessary volition. Both divine self-love and the love of God by the blessed concern the same infinitely good and hence lovable object; so why is one necessary, but not the other? Thomas Aquinas and Henry of Ghent took as necessary and jointly sufficient conditions for necessary willing that the object must be the absolute (or infinite) good and that it must be perfectly known. Scotus adds a third condition: the will must be infinite. At first, Scotus argued on moral grounds that under these three conditions the volition is necessary: the infinite good is to be willed and the infinite will cannot fail to be upright.30 But this argument traces spiration to right reason, whereas according to the commonly accepted view, the principle of spiration is the will, not the intellect. So later he prefers an explanation on metaphysical grounds: an infinite will cannot fail to be actually willing (that is, it cannot suspend its willing) and cannot fail to will the object that fully corresponds (adaequat) to its willing capacity. But only the infinite good entails God’s necessary love; a finite good does not, for if it did, then the divine will would depend on this finite good, which implies an imperfection.31 Scotus argues that God’s necessary volition is nevertheless free. One of his arguments for the compatibility of freedom with necessity works with two premises: first, that freedom is intrinsic to the manner in which the will wills; second, that necessity in willing is a perfection. On the assumption that an intrinsic feature of a power is compatible with its perfect mode of operation, Scotus concludes that freedom is compatible with necessity.32 Another argument appeals again to the eliciting-mode thesis: a power’s eliciting mode is the same for all its acts. Since the divine will loves contingently the things that are related to the end (that is, creatures), it loves them freely; and since it loves these freely, it loves everything freely, and so too it loves the ultimate end freely, albeit necessarily. 30 31

32

Lect. I.10 n. 26, XVII: 124; Ord. I.10 nn. 48–9, IV: 359–61; Rep. I-A.10.2 nn. 43–4, WB I: 398–9. Ord. I.10 n. 49, IV: 360, annotation; cf. Rep. I-A.10.2 nn. 41–2, WB I: 398. See also Quodl. q. 16 n. 7, NR 163: the infinite will must “relate in the most perfect way” to the most perfect object. This can be read as an explanation of necessary willing on either moral or metaphysical grounds. Quodl. q. 16 n. 38, NR 179; see also Rep. I-A.10.3 n. 51, WB I: 402; Ord. I.10 nn. 56–7, IV: 363.

Refinements and Radicalizations

129

He briefly justifies the inference from “contingently” to “freely” by arguing that contingent activities must be traced to a free active power, the will (Ord. I.10 n. 44, IV: 358; Quodl. q. 16 n. 36, NR 178–9). Contingent activities must be traced to the will, for only the will, but no natural power, can synchronically produce contrary effects: at a given moment and hence in a given set of circumstances, the will controls whether it acts or not.33 Scotus here uses the eliciting-mode thesis to justify the inference from free willing of what is related to the end to free, albeit necessary willing of the end. By contrast, in discussing the created will’s relation to the ultimate end, Scotus had used the eliciting-mode thesis to justify the inference from free willing of what is related to the end to contingent willing of the end (see pp. 126–7 above), precisely because he had identified contingent willing and free willing.34 This difference reflects a change in his theory of free will, which entails also a new account of the core difference between nature and will. This difference can no longer be described in terms of the necessary versus contingent eliciting mode. Rather, Scotus realizes that it must be described in terms of a natural versus free eliciting mode (Rep. I-A.10.3 n. 53, WB I: 403; Quodl. q. 16 nn. 40, 62, NR 180, 192). What is essential to the will is not contingency, but freedom, and what is incompatible with freedom is not necessity in general, but only natural necessity. Later additions by Scotus to his discussion of enjoyment in the Ordinatio reflect his revised account of the will/nature dichotomy and are in tension with the passages that contain the earlier account.35 Freedom Freedom, then, is compatible with necessity, although not with natural necessity. Scotus clarifies that even when acting necessarily, the will is not determined from without, but determines itself to its act – a view that resembles Henry’s theory that in necessary volitions, the will imposes the necessity on itself. Freedom, then, implies self-determination, although not contingency.36 Contingency is coextensive only with created freedom. 33 34

35 36

Ord. I.2.1.1–2 nn. 79–86, II: 176–9; cf. Rep. I-A.39–40 nn. 26–7, WB I: 472. See Knuuttila 1981, 218–23. Scotus’s use of the eliciting-mode thesis may seem arbitrary. Ockham, in fact, argues that it could just as well be used to infer from the existence of one necessary volition to the fact that all volitions are necessary; see Ord. I.10.2, OTh III: 344 ll. 2–8. In Ord. I.1.2.2 nn. 81, 131, II: 61, 87–8, marked in the Vatican edition as additions by Scotus, he disproves the inference from willing freely to willing contingently he had made in Ord. I.1.2.2 n. 80, II: 60–1. Quodl. q. 16 n. 63, NR 192–3; Rep. I-A.10.3 n. 54, WB I: 403; Ord. I.13 n. 47, V: 90. On Scotus’s account of necessity, freedom, and self-determination, see Normore 2003, 141–5 and González Ayesta 2019.

130

Free Will

Scotus even holds like Olivi and contrary to Anselm that the ability to sin is essential to created freedom; he identifies free decision (liberum arbitrium) in rational creatures with the will, a “free and defectible active power,” where “defectible” means to be able to sin.37 But whereas Olivi considered created and divine freedom to be two distinct notions (Chapter 3, p. 90), Scotus allows, like Anselm, for a univocal notion of freedom (libertas) that applies to God and rational creatures. This univocal notion is freedom as a pure perfection (perfectio simpliciter),38 that is, freedom as it abstracts from the limitations that come with it in creatures (Ord. II.44 nn. 8–9, VIII: 493–4). Since Scotus goes so far as to take freedom to be a univocal notion, we might expect him to propose a complete definition of freedom, but he does not.39 The Causes of the Will’s Act As we have seen, according to Scotus, no ultimate reason can be given why the will wills what it wills. Yet this does not mean that the will acts blindly. At first, in his Oxford Lectura, he argues that Henry’s theory of the cognized object as a mere sine qua non cause of a volition does not attribute a sufficient role to the cognized object. Thus, for example, Henry of Ghent cannot explain what specifies a volition to be the volition of one thing rather than another, and hence what makes loving God more perfect than loving a fly, assuming the love is equally intense (Lect. II.25 nn. 60–3, XIX: 249–50). Scotus also rejects Godfrey of Fontaines’s view that the cognized object is the total cause of the will’s act, for then the active principle is a natural agent and the will a mere passive principle. Thus the will would lose control of its act (ibid., nn. 28–31, 36, XIX: 236–9). No matter how complex Godfrey may conceive the interactions (circulationes) between intellect and will to be, the whole process would happen by natural necessity.40 Scotus thinks that John of Morrovalle’s view, that the object causes in the will an affection, entails that then the affection causes the volition necessarily, and hence he rejects this theory as well (Lect. II.25 nn. 51–2, XIX: 244–5). Instead, Scotus argues that the willed object and the 37 38

39 40

Ord. II.44 nn. 6–7, VIII: 492; cf. Rep. II-A.44 n. 3, Viv. XXIII: 232a. Scotus understands pure perfections, e.g., wisdom and the will, as concepts that apply to God and creatures univocally; see Ord. I.3.1.1–2 nn. 38–40, III: 25–7. He sees a confirmation of the univocity of the concepts of pure perfections in Anselm’s notion of freedom as applying to God and creatures; see Lect. I.3.1.1–2 n. 33, XVI: 237; Ord. I.8.1.3 n. 72, IV: 185–6. But for a reconstruction of how Scotus would define freedom, see González Ayesta 2019, 97. Rep. II-A.25 nn. 6–7, Viv. XXIII: 120a–b; Ord. IV.49.1.4 n. 240, XIV: 351.

Refinements and Radicalizations

131

will itself are partial causes of the will’s act, the will being the superior cause and the object the inferior – in a similar way that father and mother are concurrent causes of offspring (ibid., nn. 69–74, XIX: 253–5). He proposes this theory explicitly as a middle way between the accounts of Henry and Godfrey (ibid., n. 69, XIX: 253). Later, in his Paris lectures, Scotus revises his position, rejecting only Godfrey’s solution and in effect adopting Henry’s, holding that the object is only a sine qua non cause and the will the total cause of its act (Rep. II-A.25 nn. 15, 20, Viv. XXIII: 124b, 127b). The Ordinatio, the version of the Sentences commentary intended for publication, omits this discussion.41 Scotus does not change, however, his more significant view that the will rather than the willed object is the decisive cause of the will’s act. Self-Motion According to Scotus, the will as the principal cause (Lectura) or total cause (Reportatio) of its volition moves itself to its act. His explanation of the will’s self-motion is highly indebted to Henry.42 Like Henry, Scotus rejects the idea that the will must already be actually willing before it moves itself to a new volition. Thus he disagrees with Anselm’s claim that the will cannot have its first volition from itself (Lect. III.18 n. 69, XXI: 22–3) and with Aquinas’s view that the will can move itself to will what is related to the end only in virtue of willing the end (Ord. I.1.2.2 n. 158, II: 107–8). This raises the problem of how the will can cause a new volition, especially as causing a new volition would then seem to imply that the will is in actuality and potentiality simultaneously and in the same respect, that is, that mover and moved are identical. Relying heavily on Henry’s theory of self-motion as equivocal causation (see Chapter 3, p. 79), Scotus solves these difficulties succinctly as follows: 41

42

On Scotus’s change of position, see Dumont 2001. Dumont argues that Scotus’s change of position was occasioned by the fact that in Paris, he read the Sentences under Gonsalvus of Spain, who defended Henry’s theory of free will against Godfrey (pp. 774–7). Alliney 2020 suggests that at the beginning of Scotus’s Parisian sojourn, he still held his original view that the intellect is a partial cause of the will’s act. Scotus’s lectures on the Sentences in Oxford took place in 1298–9 or, less likely, in 1299–1300; see Dumont forthcoming. His lectures in Paris took place in 1302–3; see Duba and Schabel 2017, 165–73. Scotus did not publish any version of his Ordinatio in his lifetime, but rather left it unfinished when he died in 1308; see “De Ordinatione I. Duns Scoti disquisitio historico-critica,” Editio Vaticana I: 173*–175*. For more comprehensive studies of Scotus’s theory of self-motion, see Effler 1962 and King 1994.

132

Free Will But how does the will move itself insofar as it is something other? I say that the will has first actuality and is in first actuality, and it has an equivocal effect, namely the volition. Hence it moves insofar as it is in first actuality, but it is moved insofar as it is in potentiality to second actuality. Thus it is not moved insofar as it is in actuality, for it is not moved according to first actuality, but according to first actuality it moves; whereas it is in potentiality to second actuality, and according to this, it is moved. Hence existing in first actuality, it is moved insofar as it is in potentiality to second actuality. (Lect. II.25 n. 88, XIX: 259)

Mover and moved are distinct, since the mover is the will in its first actuality, that is, insofar as it exists; the moved is the will in second actuality, that is, as having a particular volition. As Scotus explains, univocal agents cannot move themselves, but equivocal agents can. Univocal agents are formally in that actuality to which the thing they are acting on is in potentiality. (A moving billiard ball moves another in virtue of its motion, not by something containing motion virtually.) Univocal self-movers, if such things could exist, would be simultaneously and in the same respect in actuality and potentiality (as if the billiard ball could move on its own). By contrast, equivocal self-movers are virtually (but not formally) in that actuality to which the thing they are acting on – themselves – is in potentiality (Ord. I.3.3.2 n. 513, III: 304). Scotus admits that nothing is simultaneously in formal actuality and potentiality with respect to the same perfection. But Scotus rejects Godfrey’s statement that nothing is in virtual actuality and in potentiality to the perfection that is to be effected thanks to its virtual actuality (see Chapter 4, p. 107). This statement is not a first principle, as Godfrey had claimed; Scotus polemicizes that it is not even a tenth principle, since it has many absurd implications and since knowledge of its terms does not necessarily cause this statement to be grasped.43 Put differently: thanks to its first actuality (its existence), the will is in accidental potentiality to second actuality (to willing), and so it does not need an external mover for the transition from potentially to actually willing (Rep. II-A.25 n. 16, Viv. XXIII: 126a; cf. Chapter 1, pp. 29–30). Scotus explains the self-motion of heavy and light things similarly and argues that since they need an external mover only to be moved to their first actuality, the motion principle does not imply that they are moved by another in each movement (Ord. II.2.2.6 nn. 465–70, VII: 363–7).

43

QMet. IX.14 n. 106, OPh IV: 666–7; see also Ord. I.3.3.2 n. 516, III: 306; Rep. II-A.25 n. 11, Viv. XXIII: 122b.

Refinements and Radicalizations

133

5.2 John of Pouilly Shortly after Duns Scotus developed his advanced voluntarist account of free decision, John of Pouilly elaborated what was probably the most detailed defense to date of an intellectualist approach to free decision.44 Pouilly was a secular master, a student of Godfrey of Fontaines and successor to his chair. He depicts himself as a loyal follower of his teacher: “I want to follow them [i.e., Godfrey] in this and in general in all or in most issues, for compared to others, their teaching seems to me more suitable to be true.”45 Pouilly’s time at Paris overlapped with Scotus’s, but his main target of critique was not Scotus but Henry of Ghent, whom he refutes at length in each of his major treatments of free decision, debated between 1306 and 1309/10: Quodlibet I, Question 10, which argues against Henry’s attempted reconciliation of Tempier’s Article 129 with the propositio magistralis (see Chapter 3, pp. 80–4); Quodlibet II, Questions 11–14, which is about judgment-volition conformity (q. 11), the cause of the will’s act (q. 13), and related issues; and Quodlibet IV, Question 6, which argues against the principal claims of Henry’s Quodlibet I, Question 16.46 As we will see in Section 8.5, Pouilly gives an intellectualist account of angelic sin that is quite solid; here, we will focus on his detailed theory of how one can explain under intellectualist assumptions that free agents control their acts. We begin with his discussion of judgment-volition conformity, where he clarifies the core ideas of his moral psychology, which align with Godfrey’s positions but are elaborated in greater detail. Judgment-Volition Conformity In Quodlibet II, Question 11 (most likely of Advent 1307), Pouilly is asked “whether the practical intellect, by the act of deliberation [consilii] and by 44 45 46

On Pouilly’s theory of free decision, see also Hödl 1999 and 2007, 214–18. Quodl. V.12, P 169vb, N 140rb–va. I am grateful to Chris Schabel for providing this reference. Pouilly debated five Quodlibeta, the first three of which exist in two redactions. For an overview of his Quodlibeta, see Hödl 2007; for their dating and that of his Quaestiones ordinariae de scientia Dei, see Schabel 2014; Schabel further narrowed down the dates in unpublished work. Quodl. I.10 is edited in Hödl 1999; Chris Schabel and I are preparing a critical edition of Quodl. II.11–14 and IV.6. For Quodlibet II, q. 11 (q. 8 in the first redaction), the two redactions contain mostly the same text; for qq. 12–14 (second redaction), Pouilly copies to a large extent text from the first redaction of his Quaestiones ordinariae de scientia Dei, q. 7 (of summer 1307). I am citing Quodl. II.11–14 from Paris, Bibliothèque nationale de France, lat. 14565 (= A), which contains (among other things) the first redaction of Quodl. I–III and the Quaestiones ordinariae de scientia Dei, and from Paris, Bibliothèque nationale de France, lat. 15372 (= P), which contains the second redaction of Quodl. I–V. Both manuscripts are available on https://gallica.bnf.fr. I am citing Quodlibet IV and V according to P and Nürnberg, Stadtbibliothek, Cent. III, 75 (= N).

134

Free Will

its verdict [sententiam], immobilizes and determines the will in such a way that in the moment [pro illo instanti] in which the deliberation or verdict stands, the will cannot will the opposite of the verdict” (A 102rb, P 47va).47 The question was tailored to Pouilly’s doctrinal sensibilities; he took the bait and defended at great length judgment-volition conformity. He divides his response into eight theses (conclusiones) and their demonstrations. Three of these are directly concerned with judgment-volition conformity; the third thesis affirms it regarding the ultimate end: one wills immutably the highest good, or happiness, or the ultimate end – as long as one thinks about it (A 102va, P 47vb); the fifth and sixth affirm it regarding what is judged conducive to the end. The fifth thesis is particularly important to Pouilly: When an end is willed unconditionally [simpliciter], in such a way that one judges and wills to pursue it absolutely and wholly so that no other end is willed more, and if by deliberation and by the judgment of a practical syllogism something is found by which the end can be had but not without it, . . . then [the will] cannot in that moment abstain from willing [non velle] or choosing it. I say, however, judgment and verdict, for while deliberation still wavers and fluctuates, it need not be that way. (Quodl. II.11 concl. 5, A 103vb, P 48va)

The last sentence clarifies that provisional considerations during deliberation do not entail the will’s adhesion, only the concluding judgment and the ensuing verdict. Whereas the fifth thesis excludes abstaining from willing while one judges that something is to be willed, the sixth thesis excludes willing the opposite (A 104vb, P 49va). The time-index is crucial in Pouilly’s fifth and sixth theses: the will is bound to follow the judgment while it lasts. One can indeed choose differently from one’s current all-things-considered judgment – but this requires a corresponding new judgment.48 The first two and the fourth theses provide theoretical foundations for judgment-volition conformity: no intellectual appetite can tend to something unknown (concl. 1, A 102rb, P 47va). Our will does not will anything by necessity of immutability (concl. 2, P 47va, omitted through lineskipping in A) – not even happiness; for although we cannot but desire 47 48

Here and elsewhere, Pouilly uses the proposition “pro” to indicate the moment the psychological act takes place, not the moment to which the act intentionally refers (unlike Scotus; see note 11 above). Quodl. II.11 concl. 8 (A 112ra, P 55ra): “. . . frequenter fertur contra aliquod iudicium rationis, sed tamen per aliud iudicium”; see also Quodl. II.11 concl. 7 (A 107vb, P 51vb): “. . . etiam oppositum possumus, licet non pro nunc.” This position seems to contradict Tempier’s Article 129, as John Baconthorpe pointed out with explicit reference to Pouilly; see Dumont 1992, 592, note 90.

Refinements and Radicalizations

135

happiness while thinking of it, we need not constantly think about it; in fact, we can judge that thinking of it is undesirable, for it keeps us from other activities (concl. 3, A 103va, P 48rb–va). The will – which wills whatever it wills under the aspect of the good – does not will anything else than the ultimate end or happiness or the unqualified good immutably while actually considered, for everything else can be seen under some aspect as lacking and not desirable (concl. 4, A 103va, P 48va). Pouilly marshals numerous arguments in support of the theses that concern judgment-volition conformity, of which we will consider those most characteristic of his way of thinking. In support of the third thesis, that happiness or the ultimate end is willed necessarily while thinking about it, he argues that the will cannot abstain from willing that in which nothing lacking desirability can be found and which is the ground of desirability (ratio appetibilitatis) in every desirable thing (A 102va, P 47vb). Furthermore, mentioning Duns Scotus by name, Pouilly rejects his claim that one need not will happiness even when thinking about it.49 Pouilly argues: if the mover and the moved are in the condition in which it is natural for the one to move and the other to be moved, but if nevertheless at that moment motion does not result, then the mover will never move and the moved never be moved.50 In fact, a reason would have to be given for why at certain times, but not at others, the mover moves and the moved is moved. Pouilly sees his consideration confirmed by Aristotle’s refutation of Empedocles’s view that friendship and discord alternately cause motion while in between there are periods of rest (Phys. VIII.1, 252a5–25). So when happiness is actually considered, it is necessarily willed, for it contains no shortcoming and hence it is to the highest degree in the condition in which it is natural for it to move the will (A 102vb, P 47vb–48ra). Pouilly says that the fifth thesis, according to which one cannot abstain from willing what the intellect judges as necessary to attain one’s end, is supported by some arguments for the third thesis (A 104ra, P 48va). He develops other arguments as well. One calls again for a sufficient reason for the will’s preference: no per se cause could be indicated for why the will does not choose according to the judgment and verdict; if one says that the will itself is the cause, then one must still indicate a cause why the will at one point does not choose according to the verdict when at another point it 49 50

See note 20 above. Scotus’s name is inserted only in the second redaction. Pouilly’s rule resembles a core thesis of Siger of Brabant (see Chapter 2, p. 55), the application of which to the will was condemned by Tempier in Article 131 (quoted in Chapter 3, p. 61). But Pouilly offers an exegesis of Article 131 that dissolves any conflict with his own view; see Quodl. II.11 concl. 8, A 111rb–va, P 54va.

136

Free Will

does (A 104ra, P 48vb). Another argument is borrowed from Godfrey (Quodl. VI.1c., PhB III: 102): when one wills an end unconditionally, then one wills implicitly that by which it is necessarily had and without which it cannot be had, for otherwise, the end would be at once desired and not desired. Thus, when through deliberation it becomes manifest that, all things considered, something is of this sort, it is necessarily desired (A 104rb–va, P 49ra–rb). The sixth thesis, that under the cognitive conditions described in the fifth thesis the will cannot will the opposite of the intellect’s judgment, is also proved by multiple arguments. The first is that this thesis is implied by the impossibility of willing what is unknown (cf. thesis 1); but if, for example, after deliberation fornication is judged to be avoided, then the will cannot desire to fornicate, because it would desire what is not known under the aspect under which it is desired. Furthermore, if the will still desired to fornicate, then it would will something that is not apprehended under the aspect of the good, which is impossible (A 104vb, P 49va). These and a number of other arguments in support of the third, fifth, and sixth theses lead to the conclusion that answers the question Pouilly was asked in the disputation: So I conclude that, when deliberation and judgment are carried through and when the verdict is given that something is to be chosen, the principle of which deliberation being the volition of the end as to be absolutely [omnino] had, the will chooses this necessarily and immutably, and in that moment it cannot abstain from willing [non velle] it. (Quodl. II.11 concl. 6, A 105vb, P 50rb)

Though the answer is now complete, Pouilly adds two more theses, which he discusses at length: the seventh states that judgment-volition conformity does not threaten or diminish free decision or free will (A 105vb, P 50rb); we will return to this below. The eighth thesis states that the opposite view, which denies judgment-volition conformity, is mistaken (A 108rb, P 52ra). In support of his eighth thesis, he submits Henry’s denial of judgmentvolition conformity in his Quodlibet X, Questions 9–10 (A 108rb–vb, P 52ra–va) to a detailed critique, based upon a careful exegesis of Aristotle’s account of incontinence in Nicomachean Ethics VII.3 and confirmed by passages from the “saints”: Augustine, Anselm, and even Bernard (an authority preferred by voluntarists) (A 109ra–110vb, P 52vb–54ra). Thus he intends to show that what Henry called Aristotle’s “prophecy” (see Chapter 3, p. 82) is held also by the saints (A 110ra, P 53va), whereas Henry’s “fiction,” that the will can will something contrary to the

Refinements and Radicalizations

137

judgment of reason, and that the will is evil before reason errs, “is not contained in the holy books, nor expounded by the saints” (A 110va, P 53vb). Pouilly discusses also, more briefly, Gonsalvus of Spain’s denial of judgment-volition conformity in Gonsalvus’s Disputed Question 8 (A 108vb–109ra, P 52va–vb; A 111vb–112va, P 55ra–rb). An argument by Gonsalvus that makes reference to angelic sin provokes a particularly detailed response (A 112va–113ra, P 55va–vb), as we will see in Section 8.5. Pouilly also briefly refutes Scotus’s theory of the synchronic ability to do otherwise on the grounds that it violates the principle of noncontradiction; he argues that when the will wills something, it does not have the power to will otherwise at that moment, but only at a later moment (A 109ra, 112va, P 52vb, 55rb–va). Furthermore, since Henry seems to have Tempier’s condemnation of March 7, 1277 on his side, Pouilly makes the remarkable attempt to show, one by one, that its relevant articles are instead in agreement with his own view (A 108vb, P 52va; A 110vb–111vb, P 54ra–55ra). Control Just as Pouilly affirms judgment-volition conformity, he argues that the will’s act is caused by the cognized object. Pouilly argues that the will’s act must be caused by the cognized object rather than by the will itself, because per se self-motion of the will is impossible. Following Godfrey, Pouilly denies the possibility of per se self-motion of the will because the will is entirely simple (that is, not composed of parts) and thus it cannot contain an active power that produces a perfection and a passive power that lacks the perfection the active power produces. In his view, this impossibility is not avoided by the hypothesis that the will as active power possesses virtually, but not formally, the perfection it causes in itself (Quodl. II.13c., A 209va–210vb, P 56rb–57rb). Many of Pouilly’s contemporaries thought that the cognized object as the cause of the will’s act eliminates free decision and moral responsibility. In 1306, when Pouilly gave his inaugural disputation as newly incepted master of theology in the bishop’s palace (in aula episcopi), as he tried to address just this difficulty, he was impeded by the shouts of his adversaries.51 In a digression during an Ordinary Question, he describes the scene: The cognized object moves in the essential order first the intellect and then the will. I prove the assumption in passing, as I intended to do on another 51

On Pouilly’s inception, see Hödl 1960; for its dating, see Schabel 2014, 257.

138

Free Will occasion, but some – I do not know well by which intention they were moved – impeded me with their clamors, one after the other, to the point that I could not say a word. . . . for since [their opinion] could not be upheld by rational arguments, they wanted to uphold it in whatever way: with words and clamors, or maybe by scaring with threats those holding [the contrary opinion], and inciting people to this [opinion of theirs] with promotions or promises. . . . and since the defenders of this view against those who profess the other view cannot be slayed with swords, they are killed with tongues. (QO de scientia Dei 7c., A 209vb)

In Quodlibet II.13 (second redaction), which is mostly copied from that Ordinary Question, Pouilly skipped this polemical passage; but he included the following further down: And good Lord, not long ago I saw that in Paris only one man dared to maintain the view I hold, and God knows the reason, and so do I. Now several, and the best in Paris, hold this view, and it will be maintained as long as the natural judgment-capacities [naturalia iudicatoria] remain and the nature of things does not change. This is what I wanted to say in the bishop’s palace, but I could not, for I was impeded. But here, by the grace of God, I will not be impeded. (Quodl. II.13c., A 213ra, P 58rb)

To whom and to which circumstances does Pouilly refer? Here are possible answers: Godfrey of Fontaines’s Quodlibet VI maintained the same view as Pouilly’s. The reason no one else dared to profess is apparently the March 7, 1277 condemnation, but perhaps not alone. Those who held this view in Paris at the time (1307) included Hervaeus Natalis, as we will see in Section 5.3. Finally, those who shouted down Pouilly were followers of Henry of Ghent, to whom Pouilly at one point refers as “Gandavistae,” that is, (Henry of) “Ghentists.”52 Pouilly intends to defend free decision and moral responsibility no less than his adversaries. In his view, what is essential to moral responsibility is that an act be in one’s control and that it be caused freely. Control as a condition for moral responsibility implies the ability to do otherwise, as he makes clear by quoting Augustine’s claim that acts we cannot avoid are not sins (lib. arb. III.18.50.171). For moral responsibility, it does not suffice that the act be directly and essentially caused by the will; in fact, if the will caused an act in itself surreptitiously, it would not be imputed to the person (Quodl. II.13c., A 210vb, P 57rb). Alluding to a discussion by Henry of Ghent of so-called “first motions,” that is, unpremeditated emotional responses, Pouilly claims that he and his adversaries (Henry and his 52

For Pouilly’s use of “Gandavistae,” see Hödl 1960, 58–9; Schabel 2014, 262, note 41.

Refinements and Radicalizations

139

followers) in point of fact agree.53 Henry had argued – on the authority of Augustine’s treatment of the admonition in the Sermon on the Mount that looking lustfully at a woman is adultery (Matthew 5:28) – that the first motion of sexual attraction to a woman not one’s own is not yet a mortal sin. According to Henry, it becomes a mortal sin with the addition of deliberate consent, which requires a judgment of reason, or through neglect of deliberation and thus failure to repel the desire (Quodl. VI.32c., X: 269–71 ll. 61–12). On this point at least, Henry and Pouilly agree. Pouilly sums up his own view: “An act is imputable to us, praiseworthy or blameworthy, precisely because [ex hoc per se] it is in our power, from whatever it may be caused and specified” (Quodl. II.13c., A 212ra, P 57rb). Even though our acts are caused by the cognized object, we are responsible for them because we can do otherwise. Pouilly explains from two angles how we control our acts: we control whether an object causes the will’s act, and we control our judgment and verdict regarding what is to be willed. From both angles, he argues that the will’s control is mediated by deliberation, and hence he denies the will direct control of any acts. The first account is developed in Quodlibet II.13 / Ordinary Question 7. To explain our control of the object’s causation of our acts, he adopts Godfrey’s application theory, according to which one controls one’s psychological acts by approaching or removing the objects that are their efficient causes: Even though the will is thus per se and directly reduced to the act by the cognized object, by which, as by an efficient [cause], the act of the will has its form and species, nevertheless, indirectly and by an intermediary [per medium] the will – or the human being by the will and intellect – reduces itself or moves itself to all its acts, apart from the very first ones. Thus in every act or regarding every act, the will and the object have the character of a mover; the object per se and directly, and so it can be said to move as to the determination of the act, but the will secondarily and by an intermediary, as that which applies the active [power] to the passive [power], and it can be said to move as to the exercise of the act. In fact, the will is in its act and it exercises its act when the object is applied, effectively causing the act. And all our acts are imputable to us, except the first ones. (Quodl. II.13c., A 210vb, P 57rb)

Pouilly’s example for the will applying the object to itself is similar to Godfrey’s: fire burns a city only if applied to inflammable material. 53

On medieval theories of first motions, see Knuuttila 2004, 177–95.

140

Free Will

Whether or not fire burns a city is not in the power of the fire, but of the one who applies it. Nevertheless, once applied, the fire causes burning necessarily (A 212ra, P 57rb–va). Just so, we do not at first control what the intellect apprehends under the aspect of the good and to what the will is consequently inclined, but after this first impact, it is in our power to accept or reject it. (The example of spontaneous sexual attraction, to which one can consent or not, could be given here as well.) In fact, we can apply or “disapply” the object to the will by means of intellect and will. Pouilly sees this point confirmed in Augustine’s statement already used by Godfrey that “what one accepts or rejects is in one’s control, but by which sight one is affected [tangitur] is beyond control” (A 212ra, P 57va, quoting lib. arb. III.25.74.255). Whether we accept or reject our first inclination, that is, whether or not we apply an object to the will, depends on our deliberation. During deliberation, intellect and will move the sensory powers to represent suitable and unsuitable aspects of an object, until a phantasm (a representation of an object in the imagination) appears that moves the intellect determinately to its cognition as either suitable or unsuitable, and that accordingly moves the will to pursuit or avoidance. This process may involve, as Pouilly illustrates, remembering having seen something relevant in a book, opening the book, and if that does not settle the issue, going to ask someone for advice. He adds that just as one can apply an object, one can also not apply it if one is negligent, or disapply it by moving the imagination to represent something else (A 212ra–rb, P 57va–vb). In short, for Pouilly, to apply or not an object to our will is to deliberate whether or not an object is worth desiring or choosing. Pouilly concludes that the will is not entirely passive, but also active.54 Above all, he claims that he has proven his point: although the object causes directly the acts of intellect and will, the acts are imputable to us, for we – and not the object – control whether the object is applied so as to cause an assent or not (A 212rb, P 57vb).55 He claims that his theory has several advantages over Henry’s: it does not make any impossible assumptions; it explains moral responsibility better; it safeguards all that is required for free decision; and it accounts for the numeric and specific 54

55

Quodl. II.13c., A 212rb, P 57vb and A 213ra, P 58rb. In Quodl. II.11 concl. 8, A 110vb, P 54ra, Pouilly writes that his theory coheres with the article “that the will is passive and not active, error” (as he quotes it), which is Article 9 of Tempier’s 1270 condemnation but which he mistakes as one of the articles of 1277. Pouilly also makes detailed use of the application theory to explain our control of the intellect’s acts, that is, whether or not it assents to a proposition; see Quodl. IV.7c., P 121rb–vb, N 92rb–va.

Refinements and Radicalizations

141

differences between the will’s acts (that is, the fact that each volition consists in willing something determinate, rather than indistinctly in willing) (A 212va, P 57vb). In the slightly later discussion of Quodlibet II.11, Pouilly gives his second account of how we control our volitions. We do so by controlling the deliberation, judgment, and verdict that cause them. He shows this in proof of his seventh thesis, that the necessitation of the volition by the judgment and verdict of the intellect is compatible with free decision and free will. He supports this thesis through the demonstration of this claim: The necessity or immutability of the will with respect to the things related to the end, which is only from some supposition – a supposition of which the human being is the cause by free decision (which includes intellect and will in such a way that one can will this and the opposite) – in no way undermines free decision or the freedom of the will. (Quodl. II.11 concl. 7, A 105vb, P 50rb)

Most of his proofs for this claim use authoritative statements by Aristotle, John of Damascus, and Nemesius (whom Pouilly cites as Gregory of Nyssa56). But Pouilly’s main proof stands on its own feet. To demonstrate that we control the deliberation, judgment, and verdict that cause our volition, he traces them to their cause: they are the conclusion of a practical syllogism. He argues that we control the practical syllogism because we control both of its premises and the syllogization (pertractatio) of the premises. Generalized, the major premise of a practical syllogism states that “that by which the end can be had in the best way is to be desired and done” (A 106ra, P 50va), and the minor premise states that “this is such a thing by which alone or in the best way the end can be had” (A 106va, P 50vb). So to control the major premise is to control whether we desire an end. When an end is apprehended as the ultimate end, happiness, we do not control whether we desire it so long as we consider it, but we control whether we consider it (cf. theses 3 and 2, respectively, pp. 134–5 above). When an end is not apprehended as the ultimate end, then we do not control whether our will is affected by it, but we control, once affected, whether we continue to consider it or turn to something else, and furthermore, whether we consent to it or not (A 106ra–rb, P 50va). We control the minor premise because it results from the deliberation, judgment, and verdict about what is conducive to the end.

56

See Chapter 1, note 25.

142

Free Will

Deliberation, judgment, and verdict are in our power because the intellect dictates them (A 106va, P 50vb). Pouilly consistently traces control to the alternative of starting or not, to deliberate. We thereby control whether we apply an object to the will (Quodl. II.13c., A 212ra, P 57va; A 212vb, P 58ra). We thereby also control the major premise of the practical syllogism (Quodl. II.11 concl. 7, A 106rb, P 50vb) and the minor premise (ibid, A 106va, P 50vb), and hence the conclusion, which is the judgment that entails our choice. Thus control comes down to the alternative between remaining under the first impression of what appears as good or deliberating whether what appears as good is truly good. If we proceed without deliberation, we act impulsively, like nonhuman animals; if we deliberate, we have alternatives. In short, human freedom is safeguarded by the fact that a human being does not immediately follow that first desire which Augustine calls “being affected” [tactum], but reason holds back this desire and reflects and deliberates whether it is worth desiring or not. (Quodl. II.13c., A 213ra, P 58rb, referring to lib. arb. III.25.74.255)

If we control our acts only by controlling whether we deliberate about how to respond to what affects us, how do we control whether we deliberate? According to Pouilly’s own report, in his inaugural disputation in aula, the master who debated with him – identified in a marginal note with the Augustinian friar Henry of Friemar (A 213vb) – pressed him precisely on this point. Friemar argued that unless the will causes its own act directly (immediate), we control just as little the will to deliberate as we control the first perception (apprehensio) and the ensuing volition of what appears as good. Indeed, if it seems good upon the first perception to deliberate, then the will cannot fail to want to deliberate – unless it causes its own act directly. But then, on Pouilly’s assumption that we control any of our acts only because we control whether we deliberate, we do not control any other acts either (Quodl. II.13c., A 213ra, P 58rb). In other words, the issue is the regress problem that is implied if deliberation is a necessary condition for control, for then the control of whether to deliberate requires itself deliberation. Recall that, in contrast to Pouilly, Aquinas avoided a regress because he attributed our decision to begin deliberation to a previous deliberation or to God acting directly on our will (Chapter 2, p. 50). Pouilly confesses that Friemar’s argument is the most difficult one to refute (A 213ra, P 58rb). His response is disappointing: Although it is not in our control that this first act comes about upon what is seen, its use is nevertheless in our control, because one can stay in it so that

Refinements and Radicalizations

143

one is moved by it, in the manner of a nonhuman animal, or not stay in it, in the way that was said. (Quodl. II.13c., A 213va, P 58va)

“In the way that was said” refers to the explanation we considered above. This explanation, however, relied itself on deliberation; as we have seen, for Pouilly, our consent to an object is the application of the object to our will, which we control by deliberation (see p. 140). So Friemar’s question of how we control the start of deliberation remains unanswered. Shortly after the quoted passage, Pouilly argues that we control whether we deliberate either because something we know occurs to us (aliquo noto nobis occurrente) that makes us deliberate, or something else we know occurs to us that makes us not deliberate (A 213va, P 58va). Earlier he had given examples of what may so occur to us: one might consider that one should not act impulsively, or one might have heard that the thing under consideration is not good (A 212rb, P 57va). Tracing our control of deliberation to what occurs to us or not, however, does not explain how we control what occurs to us, that is, what we think about. To avoid tracing our volitions to things we do not control, Pouilly would have to accept as irreducible at least the control of some of our acts, as Aquinas did regarding the decision to start or not to start deliberating about how to react to a first impression. The issue will resurface when we discuss Pouilly’s theories of the first cause of an evil will in Section 7.9 and of angelic sin in Section 8.5.

5.3

Hervaeus Natalis and Durand of St. Pourçain

About the same time as Pouilly, the Dominicans Hervaeus Natalis and Durand of St. Pourçain offer alternative solutions to the regress problem implied in making deliberation essential to control in free decision.57 Their theoretical framework depends heavily on Godfrey of Fontaines and closely resembles Pouilly’s. Having examined in detail Pouilly’s theory of

57

The theories of free will by Hervaeus and Durand have been investigated only in studies of limited scope: Schöllgen 1927, 56–65 focuses on Hervaeus’s De intellectu et voluntate; Alliney 2005, 360–6 examines the extent to which Hervaeus and Durand engage Scotus in their discussions of the will; Beonio Brocchieri Fumagalli 1969, 113–26 discusses mainly the divine will in the third redaction of Durand’s Sentences commentary; Stella 1962 is a mainly textual study of Durand’s Questions on free decision. In general, late thirteenth- and early fourteenth-century theories of free will by Dominicans are little studied. For free will in Dominican responses in the correctorium controversy, see Putallaz 1995, 112–26 and Brungs 2006; regarding John Quidort (also known as John of Paris), a Dominican involved in the controversy, see also Kent 1995, 106–8. For Thomas of Sutton’s theory of free will, see Schneider 1977 and Szlachta 2020.

144

Free Will

free decision, we need only to consider the role they attribute to deliberation and the way they address the regress problem. In Quodlibet I, Question 1 of Advent 1307 (or less likely, Lent 1308)58, Hervaeus argues that free decision is not founded upon intellect and will as “bare powers” of the soul, but rather upon deliberation, by which alone we control the acts of our will. In fact, nondeliberated acts are not fully in our control (ed. Zimara, Venice 1513, fol. 1vb, 2vb). Nor can control be located after deliberation has been concluded, for the will is bound to follow deliberation’s concluding judgment (fol. 1vb, 3vb). Hervaeus thinks like Pouilly that deliberation explains why we control our acts even though they are caused by the cognized object: when our will is affected by an object, we can hold back our appetitive response until we deliberate about whether we should pursue or avoid the object. Thus we control whether we apply an object to our will (fol. 2rb, 2vb–3ra). The first act that is in our control is either the start of deliberation or its conclusion (fol. 2ra, 2vb). Hervaeus raises the difficulty of the regress problem: if we can control our acts only by deliberation, then we control deliberation itself by means of another deliberation, ad infinitum (fol. 3rb). He responds that, while we do not always control whether we begin to deliberate, we control the reasoning (pertractatio) of deliberation itself, and thus we control whether our deliberation yields a determinate practical conclusion (fol. 3va). Yet how we control this reasoning remains unclear. Hervaeus gives a similar account of free decision in Book II, Distinction 20 of his Sentences commentary. As regards the regress problem in particular, he mentions and endorses Godfrey’s solution, which according to his exposition traces the will to deliberate either to another deliberation or to a “natural impulse” (instinctus naturalis). He explains the natural impulse as a general desire to deliberate, in analogy to the general desire to speak.59 It is striking that Hervaeus, who sees himself as a faithful follower of Aquinas, does not mention Aquinas’s solution of tracing the start of deliberation to another deliberation or a divine impulse, which in Aquinas’s account safeguards free decision. Thus, in point of fact, by tracing the will to deliberate ultimately to a natural impulse, he traces it to something that is beyond

58 59

Dating according to Friedman 2007, 433–6. Hervaeus Natalis, In Sent. II.20.2c., ed. Paris, 267bC (expounding Godfrey, Quodl. XV.2c., PhB XIV: 11); see p. 268bA for the endorsement of the expounded view. Even more than in Hervaeus’s Quodl. I.1, Godfrey’s influence is manifest throughout In Sent. II.20. According to Friedman 2002, 69, Hervaeus read the Sentences probably in 1302–3 and revised his commentary for publication in 1309 or later.

Refinements and Radicalizations

145

our control; the only room for control that remains is the reasoning of deliberation itself, which, however, remains unexplained. Hervaeus not only insists on deliberation as the foundation of free decision; he also denies per se self-motion (Quodl. I.1, 1ra–rb, 2ra; In Sent. II.20.1, p. 262aB–C). Nevertheless, he admits that a choice between alternatives is ultimately unexplainable. In a way that – perhaps unintentionally – echoes the view of Duns Scotus (see p. 125 above), he writes that no cause is to be searched for why free decision is determined to this rather than to that, other than that it is a principle that can determine itself to this or that (Quodl. I.1, 4ra). A few years later, Durand of St. Pourçain, who studied under Hervaeus but was involved in numerous polemics with his former teacher, argues against him that the control of our acts is indeed grounded in the intellect and will as bare powers and need not be mediated by deliberation.60 Only prudent people think that before consenting to a loveable object one must wait and deliberate, while most people do so only rarely, and yet they act by free decision (QD De libero arbitrio q. 1, ed. Stella, p. 474 ll. 36–44). As examples of free acts that are in one’s control without deliberation, Durand mentions acts done from a skill (ars), for example writing a character without deliberating about how to do so, and acts done by the intemperate, who seek pleasure without deliberating whether they should do so (p. 475 ll. 4–11). Deliberation and the conclusion of deliberation are themselves examples of acts that we control “through our bare powers,” that is, directly (p. 475 ll. 34–6).61 Durand, too, considers per se self-motion of the will to be impossible62 and explains free decision by means of the application theory. He insists, with greater emphasis than Pouilly, that the will is active regarding its exercise. If the power of free decision were passive regarding both the determination and exercise, then it would not be free to act or not, or to do one of opposite acts (QD De libero arbitrio q. 2, p. 480 l. 19 – p. 481 l. 13). 60

61

62

Hervaeus and Durand were not merely involved in private disputes. Hervaeus sat on two commissions, in 1314 and 1316/17, that investigated Durand on behalf of the Dominican order because of Durand’s frequent critiques of Aquinas. On their controversy, see Lowe 2003, along with Russell Friedman’s review in Medieval Review 2004 04.02.28 and her response in Medieval Review 2004 04.03.16. Durand has the same, but briefer, discussion in his disputation in vesperis (the last disputation of a bachelor at the eve of the inception in aula as master), ed. Stella 1962, 453. He includes it also in the third redaction (but not in the first two) of his Sentences commentary, In Sent. II-C.24.1 nn. 9–12, ed. Venice 1571, 170ra–rb. For the three redactions (marked by the letters A, B, C), see Koch 1927, 14–85, Fiorella Retucci’s introduction to the critical edition of the first two redactions of In Sent. II.1–5, and briefly Friedman 2002, 71. QD De libero arbitrio q. 2, p. 579 ll. 14–15; In Sent. II-C.24.2 n. 15, fol. 170vb.

146

Free Will

But as a power that is passive regarding its determination but active regarding its exercise, it is free. In fact, we control by intellect and will whether we apply an object or not, or whether we impede its application. Regarding things that are willed contingently, even if we have a deliberation-concluding judgment that something is to be chosen, we can still inhibit the judgment by commanding our lower powers (e.g., imagination or memory) to represent other aspects that make us see the thing at hand differently, that is, by restarting deliberation and applying or not applying different objects that would then move the will to different acts. His example is this: if I need to cross a river and my only options are by bridge or by boat, even if I conclude that I should cross by bridge, I can stop and consider that the bridge might collapse, and then decide to cross by boat (p. 481 l. 14 – p. 482 l. 31, p. 483 ll. 14–44). To avoid a regress, Durand assumes that no further judgment (and hence, it is implied, no further deliberation) is needed either to settle in a judgment or to impede it by commanding our sensory powers to represent other things, but rather, we control this alternative directly (p. 483 ll. 42–4, cf. p. 482 l. 32 – p. 483 l. 3).63 This control, then, of whether we settle or not on a judgment is irreducible in Durand’s account of free action. Pouilly traces control to the alternative between beginning to deliberate or not, Hervaeus to the beginning of deliberation or its conclusion, and Durand to whether we conclude deliberation or continue it. Of all three, Pouilly offers the greatest resistance to admitting that there is something primary, irreducible in acts of free decision. Yet, it seems, there must be something of this sort if free acts are to be imputed to persons rather than to circumstances. In fact, although all three deny per se self-motion of the will, they do allow for self-determination – Hervaeus and Durand openly, and Pouilly implicitly, for his application theory is in fact a theory of selfdetermination.64

5.4 Peter Auriol In light of the theoretical difficulties of strictly voluntarist and of strictly intellectualist accounts of free will, several thinkers had proposed a middle 63

64

See also his disputation in aula, ed. Stella 1962, 454–5, and his resumpta (the reprise of the disputations in vesperis and in aula), ibid., 458–9; In Sent. II-C.24.2 nn. 18–24, fol. 171ra–va; II-C.24.3 n. 15, fol. 172rb–va. The first two redactions of the Sentences commentary lack again this discussion. The three written versions of his lectures on the Sentences date between 1304 and 1327; for details, see Friedman 2002, 71–2. See Hervaeus Natalis, Quodl. I.1, fol. 4ra: free decision is “a principle that can determine itself to this or that”; Durand of St. Pourçain, QD De libero arbitrio q. 2, p. 480 ll. 27–9, p. 484 ll. 18–19.

Refinements and Radicalizations

147

way. We saw the attempts by Giles of Rome and John of Morrovalle in Chapter 4. Duns Scotus, too, in his early career, developed a solution he calls a middle way (see pp. 130–1). A weakness in these approaches is that they devalue the role to deliberation, for even if they make room for deliberation, it is not where choice-making really happens. The Franciscan thinker Peter Auriol, active in the second decade of the fourteenth century, proposes a middle way that locates the control of our acts within the process of deliberation, as did Godfrey and his followers.65 Yet in contrast to them, Auriol allows for a point in this process in which the will takes direct control of the outcome of deliberation, namely of whether a judgment considered during deliberation becomes the conclusive judgment that entails the will’s choice. Independently of this theory, Auriol makes another noteworthy innovation: he defines freedom as complacentia, that is, delight. We will examine first his theory of freedom and then his theory of free decision; in Sections 9.7 and 10.6, we will encounter these theories again in connection to angelic sin. Freedom Associating freedom and delight is no novelty. Aquinas calls freely doing what one wants delightful (In Sent. III.34.2.2 qc. 1c.; ST II-II.162.4 ad 4). Henry of Ghent connects the will’s self-determination to delight (see Chapter 3, pp. 73–4). But Auriol goes a step further – he defines freedom by delight: “the formal account (ratio) of freedom consists in a power from the act of complacentia and delight [delectatio]” (Scriptum I.1.8.3 n. 114, ed. Buytaert, 449). Complacentia – which unlike the English “complacency” does not have any negative connotations in Latin – is a central notion for Auriol: it is a love that underpins all subsequent volitions grounded on this love and to which these volitions tend. Complacentia is caused by the experience of suitableness (convenientia) or conformity (conformitas) with an object (Scriptum I.17.1.2, ed. Vatican 1596, 411bD–E). It follows either upon the cognitive possession or the actual possession of a thing (Scriptum I.1.7.1 n. 54, p. 396). A still absent object, when perceived as suitable, causes love for it; this love is the reason why one desires it; once attained, one delights in it. Auriol at times calls both the first and the third act complacentia, that is, both initial love and the rest in the object. Dislike (odium) 65

Auriol lectured on the Sentences in Paris in 1317–18; see Duba and Schabel 2017, 159–65. Prior to that, he had lectured on the Sentences in Bologna and Toulouse. Auriol reworked Book I of the Sentences into a monumental Scriptum (an ordinatio). For Books II–IV we only have a reportatio; see Friedman 2002, 82–3.

148

Free Will

and fear also have their origin in love, for one dislikes or fears something because it clashes with what one loves (Scriptum I.1.7.1 n. 57, pp. 397–8; I.1.8.3 n. 120, pp. 451–2). Complacentia, delight, gladness (laetitia), joy, and love mean the same thing, with different connotations; for example, “delight” connotes the sweetness and pleasantness of acts of complacentia (Scriptum I.1.7.1 n. 68, pp. 402–3). Auriol insists that an act of complacentia is for its own sake, since it is the origin and ultimate cause of all other acts of the will. This is why freedom is complacentia, for according to Aristotle, to be free is to be for one’s own sake, unlike the slave who is for the master’s sake (Scriptum I.1.8.3 nn. 114 and 120, pp. 449–52; Met. I.3, 982b25–6). In sum, Auriol identifies freedom with complacentia because all acts of the will either are acts of complacentia or have it as their foundation and because all acts of the will are free. Freedom so understood is compatible with necessity. Thus the Father and the Son producing the Holy Spirit, God loving himself, and the blessed loving God do these acts necessarily and freely (Scriptum I.1.8.3 nn. 118, 120, 129, pp. 451–2, 456). Not every free act, then, is also in the control (in potestate) of the agent. Freedom requires only complacentia; control requires contingency – more precisely, the ability to do an act or to let go of it (Scriptum I.46.1c., ed. Vatican 1596, 1088aB–C).66 Control In discussing free decision, Auriol rejects two views he considers to be at opposite extremes. Henry of Ghent’s theory of the will moving itself to its act fails because it postulates per se self-motion (Rep. II.25.2c., Fb 99rb; ed. Rome 1605, 266aD–E).67 Auriol considers per se self-motion impossible, including the self-motion of what is virtually in actuality and formally in potentiality (Rep. II.25.1c., Fb 98rb, 263bA–265bB). Auriol further considers that Godfrey of Fontaines’s view and similar theories, according to which the will is moved by the cognized object, fail because they cannot safeguard the will’s control of its acts. Much like Scotus (see p. 130 in this chapter), Auriol argues that according to such theories, all the active principles involved in producing an act are active by natural necessity and all the passive principles undergo their activity by natural necessity; 66 67

For a more detailed account of Auriol’s theory of freedom, see Hoffmann 2019, 209–12. I established Reportatio II.4–7 and II.25 on the basis of Firenze, Biblioteca Nazionale Centrale, Conv. Soppr. A.III.120 (= Fb) and Padova, Biblioteca Antoniana, 161. I cite the text according to Fb and the Rome 1605 edition. The Rome edition allows for precise references and can be consulted on Google Books, but it does not offer a reliable text. Reportatio II.4.3 is edited in Hoffmann 2012b.

Refinements and Radicalizations

149

thus no matter how many loops may occur in the interaction of active and passive principles, that is, no matter the complexity of the interaction between intellect and will, the result will follow by natural necessity. Against Godfrey and his followers, who think that the necessitating effect of the object is undercut by the will’s ability to command the intellect to deliberate rather than stopping at the first impression, Auriol argues that if the will’s acts were caused by the object, the command to deliberate would itself be caused by the object – and hence it would be caused by natural necessity, and so would the ensuing deliberation. Thus the final judgment (iudicium ultimatum) would result from a series of causal interactions, all of which would happen necessarily. Although in the object so judged there may appear two aspects, one noble and one base, the will would be moved necessarily by the aspect that acts more strongly on it (Rep. II.25.2c., Fb 99rb–va, 266aE–bD). Put differently, Auriol shows that if the will’s act is caused by the object, then it is caused by absolute necessity, not merely by conditional necessity, for whatever control the will may have of the antecedent – the final judgment – this control would itself depend entirely on the necessitating causality of the object.68 Auriol’s own solution can be seen as another version of the application theory, and it may have been influenced by earlier versions of it. But in contrast to earlier versions, Auriol stresses that the will exercises direct and irreducible control of the antecedents to its choice. In his view, the will moves itself to its volition per accidens, by determining its mover. (He considers per accidens self-motion nonproblematic; see Rep. II.25.1c., Fb 98ra–rb, 263aE–bA.) When I want to run, my “moving power” (that is, my legs) elicits my running, not my will; but my will determines the moving power to elicit or not my running. Just so, the will does not elicit its own motion, but it determines that which elicits its motion: it determines the intellect’s final judgment to move the will to its volition (Rep. II.25.2c., Fb 99va–vb, 266bD–267aA). Thus, as Auriol writes, “the will moves the thinking [intellectionem] by way of application [applicative], for it determines it [i.e., the intellect] to act; the thinking so applied moves the willing” (ibid., Fb 99vb, 267aC; cf. Fb 99va, 266bF).

68

Auriol’s critique affects also Pouilly and Hervaeus, for although they attribute to the will the ability to apply to itself a cognized object or not, on their account the application itself would have to be caused by an object. Durand may escape Auriol’s critique because he stressed the will’s active role in the application, but it remains unclear in his account what enables the will to apply an object without having been moved to do so.

150

Free Will

Auriol’s theory of the will determining a final judgment builds upon a distinction inspired by Aquinas (ST I-II.17.1c.): Auriol distinguishes between two judgments, an indicative judgment (actus enunciativus, iudicium enunciativum) and an imperative judgment (actus imperativus, iudicium imperativum). The indicative judgment is expressed in the indicative mood, for example, “One must choose what is fitting according to reason”; the imperative judgment is expressed in the imperative mood (e.g., “Do now what is fitting according to reason!”). The indicative judgment follows “the nature of the thing” and the syllogistic deduction from the first practical principles. The will can will differently from this judgment, even if it concerns a particular act here and now, such as “it is bad to fornicate now with this woman.” Thus Article 129 of the March 7, 1277 condemnation is upheld. The imperative judgment follows upon the determination of free decision (that is, of intellect and will), as Auriol says sometimes, although he mostly mentions only the will as the determinant. The will cannot will differently from the imperative judgment, as long as the will does not suspend its determination of that judgment. The imperative judgment coincides in fact with the “final judgment” that concludes deliberation.69 Thus Auriol affirms judgment-volition conformity only regarding the imperative judgment. He attributes judgmentvolition conformity to Aristotle in Nicomachean Ethics VII, “where he seems to hold that always some verdict [sententia] precedes choice, and that the will cannot at all be moved against the ⟨verdict⟩”70 (Rep. II.4.3c., Fb 30rb–va, 79aA). As we have seen, the notion of verdict is not Aristotelian but was introduced into the discussion by John of Damascus, and the concept of will is foreign to Aristotle as well. But Auriol’s reading reflects Aristotle’s claim that we necessarily act on our own practical inference, unless we are impeded (EN VII.3, 1147a24–b10). Auriol’s theory adopts an intellectualist account of deliberation, according to which various judgments may be considered until one settles on a judgment, thus making it the final judgment that determines the will’s

69 70

Rep. II.4.3c., Fb 30va, 79aB–aF; Rep. II.25.2c., Fb 99vb, 267aC–aF; cf. Scriptum I.1.8.2 n. 90, p. 443. I conjecture sententiam (verdict), although the known manuscripts read scientiam (knowledge) or, in two cases, conscientiam. (The manuscripts of Reportatio II are listed in Schabel 2000, 155.) That the will cannot choose against the verdict that precedes choice makes more sense, for Aristotle argues in EN VII.3 that one can act against knowledge, and in the near context, Auriol acknowledges as much (Rep. II.4.2c., Fb 29va, 75aC; II.4.3c., Fb 30va, 79aE). Aristotle only denies that one can act against knowledge in the unqualified sense (1147b15–17). Scientiam could be a misreading of sententiam in the (lost) original (in the abbreviated script, they can look similar); the mistake could also have been in the original student report.

Refinements and Radicalizations

151

choice. But he makes a voluntarist emendation: to settle conclusively on a judgment is in the will’s direct control. What is important in Auriol’s dual-judgment theory is that the will determines the intellect’s judgment directly (immediate), without in turn being determined to do so. The determination by the will is not itself a volition, for then it would require a prior imperative judgment – which would imply an infinite regress. Auriol describes his notion of determination mostly in negative terms: it is not an elicited act, it does not result in anything produced, it is not something received in the will as its subject. These characteristics do not apply to a volition. As he adds, the will determines the judgment in virtue of the substance of the soul; that the will determines the intellect means that it is in the power of the soul whether the intellect elicits its act and whether the intellect’s act moves the will to its volition.71 For Auriol, then, the locus of control of one’s acts is whether one determines a judgment to be the final judgment that terminates deliberation. Like Durand, he locates this control at the end of deliberation, whereas Pouilly locates it at its beginning and Hervaeus at its beginning or end. More importantly, Auriol assumes that such determination is not mediated by other acts of intellect and will, but rather directly and irreducibly in the person’s power. Auriol affirms, like intellectualists, judgment-volition conformity and thus locates control within the process of deliberation, rather than in the ability to act contrary to the result of deliberation. But like voluntarists, he attributes to the will – or the soul – the ability to take direct control of what results in a person’s choice.

5.5

William of Ockham

Of the thinkers here considered, William of Ockham is the most modern, insofar as his views on the will, its relation to the good, and its freedom are closer to the theories of early modern thinkers than the views of Ockham’s predecessors. Ockham writes the works containing his theory of the will and of freedom between 1317 and 1324, at a time when Scotus’s thought was already widely known among Franciscan theologians in England.72 Ockham 71

72

Rep. II.4.3c., Fb 30vb, 79bD–E; Rep. II.25.2c., Fb 100ra–rb, 267bB, 268aC–bD. For a more detailed study of Auriol’s dual-judgment theory, see Hoffmann 2015; specifically on his notion of determinatio, see ibid., 80–3. On Ockham’s life and works, the intellectual influences he was subject to, and his own influence, see Courtenay 1999. For more details on the dating of his works than those offered by Courtenay, see the introductions of Ockham’s Opera philosophica and Opera theologica.

152

Free Will

also knows Peter Auriol’s Scriptum on the First Book of the Sentences,73 but to my knowledge he does not engage Auriol’s theory of freedom contained therein. Ockham’s moral psychology takes some of its core principles from Scotus, such as the distinction between will and nature and the characterization of intellectual activity as natural. Like Scotus and voluntarists in general, Ockham holds that the will need not follow the judgment of the intellect and can will the contrary. On other points, Ockham goes well beyond Scotus, as we will see.74 Will and Nature According to Ockham, intellect and will are identical – they are the substance of the soul – but the words “intellect” and “will” connote really distinct acts: “intellect” connotes the act of understanding, “will” connotes the act of willing. In other words, the term “intellect” means “the substance of the soul insofar as it is able to understand,” and “will” means “the substance of the soul insofar as it can will.”75 Ockham conceives like Scotus the distinction between nature and will (that is, for Ockham, the soul insofar as it can will) as clear-cut and fundamental: “Every agent other than the will acts naturally” (Quodl. I.17, OTh IX: 90 ll. 12–13). A natural cause can cause contrary effects in different things – the sun can cause hardness and softness – but not in the same thing. But the will can do so, even when the object is considered under the same respect (Expos. Phys. II.5.9, OPh IV: 290 ll. 20–9). Acts of understanding or of knowledge are purely natural, while acts of willing, loving, hating, and desiring are free (Expos. Phys. II.8.1, OPh IV: 323–4 ll. 238–46). Similar to Scotus’s early view (which Ockham identifies tout court as Scotus’s position; Ord. I.1.6, OTh I: 492 ll. 8–9), Ockham holds that volitions have as partial causes the will and the cognition of an object. He adds that God is a partial cause of volitions; in fact, as the primary cause, God is a partial cause and principal cause of all effects produced by secondary causes. Also dispositions (habitus) and the feelings of pleasure and pain can be partial causes of volitions. Yet, no partial cause other than the will controls whether the will actually wills something, abstains from 73 74 75

See the preface to the edition of Ockham’s Ordinatio, OTh I: 36*. For Ockham’s moral psychology, see the concise account in Adams 1999. Rep. II.20, OTh V: 435–9; Ord. I.1.2, OTh I: 402 ll. 12–17. For Books II–IV of Ockham’s Sentences commentary, there is only a reportatio, while Book I exists in an ordinatio. On the real identity of intellect and will and their distinct connotations, see Panaccio 2012, 76–80. Recall that Philip the Chancellor defended a similar thesis about a century earlier.

Refinements and Radicalizations

153

willing it, or wills against it. Dispositions and pleasure and pain only make the will more or less inclined to certain volitions.76 Even God controls our volitions only when he is their total cause, as we shall see. The will is a contingent cause, as we know through experience: we experience in ourselves that we can will, not will, or will against what reason dictates (Quodl. I.16, OTh IX: 88 ll. 25–8). According to Ockham, the will has a real ability to do otherwise only diachronically, not synchronically. He rejects Scotus’s theory of the synchronic ability to do otherwise because it implies that contradictories can be the case in the same instant. Assume that the proposition “The will wills A at t1” is true. Scotus holds that then the proposition “The will can will non-A at t1” is true as well. But Ockham argues that to these propositions about the present correspond propositions regarding the past, which will be necessarily true any moment after t1: “The will willed A at t1” and “The will could will non-A at t1.” Since after t1 the proposition “The will willed A at t1” is necessarily true, the proposition “The will willed non-A at t1” is necessarily false. And so “The will could not will non-A at t1” is necessarily true, which contradicts the proposition that “The will could will non-A at t1.” Thus, pace Scotus, if the will wills A at t1, it cannot will non-A at t1 (Praedest. 3, OPh II: 533–4 ll. 24–55).77 In Ockham’s own view, the will’s control of its acts only requires a diachronic ability to do otherwise, namely the ability to will differently in different instants without any change having to accrue to the will or to a partial cause of its act (p. 536 ll. 92–8). Ockham infers from the will’s causing its acts contingently that per se selfmotion is possible and that Aristotle’s motion principle does not apply to the will, only to natural causes (whether corporeal or spiritual). He claims explicitly that the same thing can be active and passive regarding itself. Furthermore, when all the necessary conditions for an act of willing are present for some time – the cognized object under the aspect of desirability, God as partial cause of the act, and all other prerequisites – the will need not elicit its act. The will can begin willing after some time by itself alone, without any change in the other partial causes. By contrast, a natural cause that is inactive for some time requires something extrinsic to allow it to cause its act.78 76

77 78

For God as principal primary case of all effects by secondary causes, see Rep. II.3–4, OTh V: 62–3. For God, the will, and the cognition of the object as partial causes of volitions, see QV 7.4, OTh VIII: 393 ll. 412–16; for dispositions, see Rep. III.7, OTh VI: 205–6; for pleasure and pain, see QV 8.2, OTh VIII: 447–8. On Ockham’s theory of the causes of the will’s act, see Osborne 2014, 44–56. On Ockham’s intricate argument against Scotus and on his own theory of contingent willing, see in greater detail Normore 2007b, 290–2. Rep. IV.15, OTh VII: 332 l. 7 – 334 l. 5; Quodl. I.16, OTh IX: 89 ll. 39–53. On Ockham’s theory of self-motion in general and of the will in particular, see Normore 1994.

154

Free Will The Scope of the Will

The will also differs from natural causes in that natural causes have natural inclinations toward their perfection, but the will does not. On this, Ockham’s view is radically new. According to Aquinas and Henry, the will is naturally inclined to its ultimate end, happiness, so that one cannot will anything except in view of happiness. Scotus, too, holds that the will has a natural inclination to happiness – the affectio commodi – although he denies that one wills everything in view of happiness. Ockham, by contrast, contends in a discussion of the enjoyment (fruitio) of the ultimate end that the will does not have a natural inclination to the ultimate end, except if “natural inclination” refers to that which happens normally.79 Still, the will cannot act if not for the sake of an end – either an end dictated by reason, or an end that the will sets freely for itself (Ord. prol. q. 10, OTh I: 291 ll. 2–4). The implication is that one can will the opposite of the ultimate end, a possibility Ockham lays out in detail in Ordinatio I.1.6, about whether the will enjoys (that is, loves) the ultimate end contingently, and in an analogous discussion in Reportatio IV.16, about whether the blessed necessarily enjoy God. Ockham makes the general statement that “the will, as a free power, is receptive of willing-against and willing with respect to any object whatever.”80 The will can love happiness or not, and it can even will against happiness. One of Ockham’s several arguments is that one can will against everything one considers unattainable, and one may think that happiness is unattainable. Another argument is this: God wants a damned person not to be blessed, and this person is able to conform his or her will to God’s will so as to want not to be blessed; but a person in this life can want to lack happiness no less than a damned person is able to want that. For Ockham, willing to lack happiness is not sinful, as is implied in a clarification that so willing is compatible with charity. Concerning love of God, the beatifying object, Ockham claims that one can will-against God (nolle Deum), that is, one can will that God does not exist, for God can be perceived as disagreeable insofar as he punishes. That one can willagainst God follows also from the consideration that every will can conform to the divine will, and God can command a person to hate him. To do so is even possible while contemplating the divine essence, for if 79 80

Ord. I.1.6, OTh I: 507 ll. 9–16; see also Rep. III.6, OTh VI: 175 l. 20 – 176 l. 5. On the limited sense in which Ockham allows for natural inclinations, see Adams 1999, 255–6. Rep. IV.16, OTh VII: 350 ll. 18–20; cf. Ord. I.1.6 arg. ad opp., OTh I: 486 ll. 12–13. For a more detailed discussion of Ockham’s view on the desire for happiness, see Osborne 2012, 448–55.

Refinements and Radicalizations

155

commanded by God, hatred of God is an upright act, no less presently (in via) than in heaven (in patria).81 To prevent the will of the blessed from failing to love God and thus from losing beatitude, God must be the total cause of their love of God.82 As we will see in Section 10.7, the will’s freedom is also the reason why the permanence of damnation must be guaranteed by God as total cause of their hatred of God. These are radical ideas, and they did not remain unobserved. Pope John XXII put Ockham to trial in Avignon under charges of heresy and ordered a commission of masters to examine his Sentences commentary. The commission, which included Durand of St. Pourçain (possibly its chair), singled out for condemnation 51 articles, five of which concern Ockham’s views about love of the ultimate end and hatred of God. The commission’s evaluation of these five articles in an earlier report was relatively restrained, but in the final, more severe report, its judgments ranged from “debatable” to “erroneous” and even “heretical.”83 The trial, which lasted from 1324 to 1326, did not lead to a condemnation of Ockham. But the Pope excommunicated him, without connection to the trial, for apostacy after he fled Avignon in 1328 to seek protection under Louis of Bavaria, the Holy Roman Emperor.84 Ockham also takes an extreme view by abandoning the axiom that the will can will any object only under the aspect of good and detest an object only under the aspect of evil. Ockham argues against this axiom as follows. In this present life, meriting concerning an action or an omission presupposes that one can demerit (that is, sin) concerning the same. (Christ and the Blessed Virgin, who in their earthly life merited but did not sin, did not fall under this rule because of divine dispensation.) Ockham assumes that one can merit by willing against worship of foreign gods. Hence, on the assumption of symmetry, one can demerit by willing to worship foreign gods, against one’s judgment that this is evil. Accordingly, one can will something evil that is neither really good (for idolatry is evil) nor apparently good (for one judges it to be evil). Ockham makes an analogous consideration to show that one can will-against a good that is neither really nor apparently evil: if one can merit by willing a good that the intellect presents as good, then one can demerit by willing against it (QV 8.2, OTh VIII: 442–5). This clear commitment to the possibility of willing evil under the aspect of evil is uncommon even among voluntarist thinkers. 81 82 83 84

Ord. I.1.6, OTh I: 503–7; Rep. IV.16, OTh VII: 350–2. Rep. IV.16, OTh VII: 352 ll. 11–21; Rep. II.15, OTh V: 341 ll. 7–19. See Articles 5–6, 9, 35, and 46 of the second report, edited in Koch 1935–6, 88–9, 92, 184, 192. On Ockham’s Avignon trial, see Koch 1935–6 and Brampton 1966.

156

Free Will Judgment-Volition Conformity

As regards the relation between judging and willing, Ockham treads common ground with other voluntarists. The will can will the opposite of what is judged by reason or it can suspend its act (Rep. III.12, OTh VI: 421 ll. 15–18). If the will conformed to the intellect necessarily, then we could not merit or demerit, for no act would be in the will’s control (Rep. IV.16, OTh VII: 353 ll. 10–14). The reason is that only the will controls its acts; the intellect controls its acts only to the extent it is under the will’s command (QV 7.3 concl. 2 proofs 1, 2, and 6, OTh VIII: 363–70). Accordingly, one can act knowingly against one’s better judgment. According to Ockham, when acting incontinently, the incontinent know the prohibitive major premise of the practical syllogism that nothing immoral is to be done, and they know from experience that this act is immoral and hence not to be done. They do not fail to draw the prohibitive conclusion. Nor do they draw a permissive conclusion from a permissive major premise (as Aquinas thought); rather, they act consciously against the prohibitive conclusion (QV 7.3 concl. 2 proof 5, OTh VIII: 367). The Aristotelian adage that “every evildoer is ignorant” (EN III.1, 1110b28–9) receives a new interpretation: not that one could not act consciously against the judgment of reason; rather, an evil person misses out on many occasions to acquire knowledge that a good person can acquire by practicing good actions (Rep. IV.16, OTh VII: 357 l. 17 – 358 l. 2). Ockham admits a case in which a volition inevitably matches a practical judgment – a case of judgment-volition conformity. It is when one wills an end efficaciously and understands which means is necessary to achieve the end. In fact, “whoever wills something efficaciously wills all that without which he thinks he cannot obtain what he wills” (Ord. I.1.6, OTh I: 505 ll. 5–6). Thus, willing health efficaciously, and knowing that it can only be achieved by means of a bitter potion, necessarily entails that one wills to drink it.85 Godfrey of Fontaines and John of Pouilly made a similar consideration: willing an end, one wills implicitly the means to the end; if it has been established by deliberation what the necessary means is, one cannot fail to will it. Pouilly used this consideration as one of his proofs of judgment-volition conformity (see p. 136 in this chapter). Yet Ockham holds that in such cases the act of willing the means is caused not by a new activity of the will, but rather by the previous acts of willing an 85

Ord. I.1.6, OTh I: 493–4, 499–500; cf. QV 7.3, OTh VIII: 353 ll. 283–301. On such necessary volitions, see Panaccio 2012, 86–8.

Refinements and Radicalizations

157

end and of judging which means is indispensable (Ord. I.1.6, OTh I: 500 ll. 1–3). The point is that what makes the volition of the means necessary in the case under consideration is not that it is caused by a practical judgment to which the will necessarily conforms, as Godfrey and Pouilly held, but rather that it is caused by reason and will together. Freedom Although Ockham recognizes the traditional multiple senses of the word freedom, such as freedom from misery, freedom from sin, and freedom from coercion,86 he prefers to reserve the term for freedom from necessity of immutability, which in his understanding is the way the word is used by the philosophers (of whom he cites Aristotle, Averroes, and John of Damascus). Thus “free” means the same as “contingent” and “indifferent.”87 He expresses his understanding of freedom succinctly in this wellknown passage: I call freedom [libertatem] the ability by which I can posit indifferently and contingently different [effects], such that I can cause or not cause the same effect, without there being a difference elsewhere outside that power. (Quodl. I.16, OTh IX: 87)

Freedom so understood is incompatible with necessity, and therefore God does not love himself freely, nor is the Holy Spirit freely produced (Ord. I.10.2, OTh III: 341 ll. 9–17, 343 ll. 21–3). Many, including intellectualists, ground freedom in the will’s indeterminacy. But while Godfrey and his followers hold that the will’s indeterminacy must be determined by deliberation so as to make the will actually will one or the other alternative, Ockham insists that the will does not need any such determinant; rather, it determines itself to one alternative.88 Put differently, the will has immediate control of its act; its control is not mediated by deliberation. In fact, as he argues in detail, if the choice of one alternative over another depended essentially on a cause other than the will, this cause would be a natural cause, the causality of which escapes the will’s control (Expos. Phys. II.8.1, OPh IV: 320 ll. 121–43; De causis 126, OPh VI: 735–6). It is unique to the will that it can act differently under the exact same circumstances. At its root, the alternative up to the will is either willing or 86 87 88

Ord. I.1.6, OTh I: 501–2; Rep. II.15, OTh V: 354–6. Ord. I.10.2, OTh III: 335–8, 341; cf. Ord. I.1.6, OTh I: 501–2. Expos. Phys. II.8.1 OPh IV: 319–20 ll. 105–20; Rep. IV.16, OTh VII: 358 l. 20 – 359 l. 10.

158

Free Will

willing-against. Precisely by means of this alternative, a person controls all other acts, such as acts of understanding or external acts (e.g., walking) (Expos. Periherm. II.7, OPh II: 481 ll. 40–54). Freedom, as a power of opposites, is what allows one to sin. Ockham agrees with Anselm that the ability to sin is not freedom or a part of freedom, for then everyone who is free could sin, which is not the case with God and the blessed. It is rather that everyone who can sin is free (Rep. IV.16, OTh VII: 355). Ockham reads Anselm to hold that sin only decreases the freedom from servitude to sin. But sin does not affect the freedom from coercion that allows one to will or will against something (Ord. I.10.2, OTh III: 342–3). Ockham’s originality does not consist in his notion of freedom of indifference. The idea that the will can choose differently without presupposing a different judgment or a difference regarding other factors affecting the choice is typical to the voluntarist movement as a whole. What is innovative in Ockham is rather the radicalization of a process initiated by Olivi and Scotus, according to whom the will can set an end for itself, rather than receiving an end through rational recognition. For Ockham, the will can set for itself any end; there are no restrictions inherent to the psychological setup of the human person, nor are there naturally given limits within which alone willing an object is possible. Ockham thinks there is a final end for rational beings, but the will is not limited to willing this final end and what is related to it. The disconnect between the will and pre-given ends (that is, goods) is complete.89

***

There is more variety and nuance among the theories of free will considered in the past five chapters than the labels “intellectualism,” “voluntarism,” and “intermediary theories” capture. Yet on a key point, there is a clear divide: regarding the cognitive and volitional process leading up to a free choice, either an account of free will conceives of the person’s control as necessarily mediated by the intellect’s deliberation, or an account allows the person at some point direct, irreducible control of an outcome. While intellectualists (such as Godfrey, Pouilly, and Hervaeus) fall under the first horn of the alternative, voluntarists (such as Henry, Olivi, Scotus, and Ockham) fall under the second horn. Intermediary theories, too, fall under the second horn, since they allow for some direct control. For example, Morrovalle grants that the will controls how to react to an 89

For a more ample discussion of the will’s independence from a natural hierarchy of goods and ends, see Normore 1998, 33–9 and Adams 1999.

Refinements and Radicalizations

159

affection caused by an object; according to Giles, the will controls which aspects of an object to consider; for Durand, intellect and will control upon which practical judgment to settle; and similarly, in Auriol’s view, the will controls what becomes a practical judgment that causes a choice. Aquinas is less clear about direct rather than mediated control, for he traces the will’s acts to deliberation of the intellect – except that in the last analysis he traces them to God, who moves the will according to the will’s nature as a contingent cause. Even for intellectualists, the problem of direct control remains, but is shifted from the will to the intellect: if we control our volitions through our cognitive acts, how do we control our cognitive acts? If the will is granted direct control at some point, a choice of the will is at bottom primitive, unexplainable. If the will does not have direct control at any point, then a choice is traced to the practical consideration that determines it. But even then, the question remains why a person favored this practical consideration rather than that, and if no answer can be given, then a choice is also at bottom unexplainable. Either approach to free will, then, faces the problem of what explains why free agents make the choices they do. The remainder of the book will shed new light on the problem of how to reconcile intelligibility of choices with their control, and more generally, on the relations between cognition and volition, ignorance and evildoing, freedom and necessity, and the scope of the will’s acts.

part ii

Whence Evil?

chapter 6

Does Evil Have a Cause?

For Christian thinkers, the sin of the angels is the first incident of moral evil in a universe created good. That already existing evil produces further evil is not difficult to explain. But it is odd that evil could come about in a flawless universe. The sin of the angels highlights in a special way the problem of the origin of moral evil, but the problem itself applies more broadly to the cause of an evil will in a good person, and thus can be treated separately from the topic of angelic sin, as many medieval thinkers in fact did treat it. The present chapter concerns the theories of the origin of evil by Augustine and Pseudo-Dionysius, two important sources for medieval theories of the nature and origin of evil. The medieval theories themselves are the topic of Chapter 7.

6.1 Augustine Augustine developed his theory of the origin of evil above all in connection with a thorough reflection about the fall of the angels. Failing to identify the cause of this mysterious event, he arrives at the conclusion that it is by its very nature unexplainable.1 “Whence Evil?” While Augustine adhered to Manichaeism, he had a simple solution to the problem of the origin of evil: it has always existed, and every evil in the world can be traced to a first evil principle that is coeternal with the first good principle, God. Having left behind Manichaeism, the problem of the origin of evil becomes pressing for him, for now he believes that God is good, that he made all things good, and that nothing exists that he did not make (e.g., conf. VII.12.18; nat. b. 1–3; ciu. XII.1 and XII.5). He confesses: 1

For Augustine’s account of evil, see Evans 1982; for his account of sin, see Burns 1991.

163

164

Whence Evil? I had this burning question: Whence evil [unde malum]? What torments of my heart in travail, what lament, my God! (conf. VII.7.11)

Pointing to the devil does not answer the question: Whence is it, then, that I want evil and loathe good? . . . If this comes to me from the devil, whence is the devil? For given that he himself, having been a good angel, turned into a devil by his perverse will, whence comes the evil will in him by which he became a devil, since he was made entirely an angel by the wholly good creator? (conf. VII.3.5)

Augustine’s conundrum was not a solely Christian one; his conviction that all things are good by nature also has a Neoplatonic origin. Neoplatonic writings confirmed the Christian doctrine of creation (conf. VII.9.13) and allowed Augustine to understand evil not as a substance but rather as the privation of the good (conf. III.7.12). To be sure, there is evil in the world but evil cannot subsist on its own. Evil is the corruption of the good; specifically, evil is “the corruption of the natural measure, or form, or order [corruptio uel modi uel speciei uel ordinis naturalis]” of a thing (nat. b. 4). Evil is therefore always parasitic upon some good, because corruption presupposes a substance that can be corrupted (mor. II.5.7, conf. VII.12.18, ciu. XII.6, ench. 4.14). Since evil has not always existed, and since it is not caused by God who created all things good, Augustine accepts the idea that the cause of evil is a free act, a “free decision of the will” (liberum voluntatis arbitrium) (conf. VII.3.5). But this does not solve the problem. It is easy to see how someone with an evil will could want to cause evil – but how can someone want to cause evil before he or she has an evil will (cf. ciu. XII.6, CCSL XLVIII: 362 ll. 96–9)? A person may be corrupted by someone else’s evil, all the way back to our first parents, Adam and Eve, and further to the devil tempting them. This leads us back to Augustine’s question of how the devil comes to have an evil will. What Caused the Devil’s Evil Will? What makes the angelic fall particularly challenging is not only that Lucifer sinned without any instigation to sin, but also that some angels sinned and others did not, whereas both Adam and Eve sinned. The good angels’ perseverance in the good can be traced to God (ciu. XII.9), but not the evil will of the evil angels.2 What, then, caused their evil will? 2

For arguments that sin cannot be traced to God, see, e.g., lib. arb. II.20.54 and III.16.46; ciu. XII.5.

Does Evil Have a Cause?

165

Augustine discusses the cause of angelic sin in most detail in Book XII of De civitate Dei.3 God created all angels good. What differentiated the good from the evil angels was not their nature, but their will. The good angels had a good will: they remained steadfast in loving God, the “common good of all.” The evil angels had an evil will: they delighted in their own power and proudly turned from the common good to their own private good. Adhering to God caused the happiness of the good angels, while not adhering to God caused the misery of the evil angels (ciu. XII.1, p. 355 ll. 11–24). For Augustine, sin consists precisely in this: turning away from God, the “unchangeable good,” to some changeable, inferior good (lib. arb. II.19.53.199). What, then, caused them to turn away from God? “What is it that makes the will evil, since it is the will that makes a deed evil?” (ciu. XII.6, p. 360 ll. 15–16). Augustine does not think that the will itself (as a power) caused the evil will (the bad occurrent act). In fact, by “will,” Augustine does not mean a power or faculty of will as distinct from the intellect and as the subject of rational desires and choices, but rather an occurrent or dispositional appetite: the act or disposition of loving or hating.4 If the idea that bad willing is originally caused by the will as a power is unavailable to Augustine, can he say that it was caused by the agent (the angel in the case under consideration)? This does not solve the problem either, as Augustine makes clear through a detailed diaeresis, an argument that starts by considering all logically available alternatives and then eliminates those that are manifestly false, leaving as a result only the correct one. The diaeresis results in the claim that nothing can be the efficient cause of the first evil will. It proceeds as follows: what made the will – the occurrent or dispositional appetite – bad is (1) either a thing (res aliqua) or (2) nothing. If a thing made the will bad, then the cause of an evil will is either (1.1) a thing with a will or (1.2) a thing without a will. If it is a thing with a will, then it is either (1.1.1) a thing with a good will or (1.1.2) a thing with an evil will. Both of the latter hypotheses fail: that a good will might cause an evil will is absurd, while an evil will cannot cause the first evil will. If a thing without a will made the will bad (option 1.2), then it would be 3

4

Augustine’s extensive treatments of the angelic fall are Gn. litt. XI.14–26 and later ciu. XI.11–XII.9. For more detailed studies, see Babcock 1991, MacDonald 1999, King 2012, Wetzel 2012, and Rist 2014, 44–51. See Chappell 1995, 127, Byers 2013, 217–31 (esp. 218–22, concerning ciu. XII and XIV), Rist 2014 (who contrasts Augustine’s notion of will with the will as faculty, which he considers to be an innovation by Anselm of Canterbury), and Couenhoven 2017. For the transformation of the notion of will in medieval philosophy, see Williams 2019.

166

Whence Evil?

a nature that is good insofar as it is a nature; but something good cannot cause evil. Since the will cannot have become bad through the action of a thing (with or without a will), it must be the case that nothing made the will bad. In that case, either (2.1) the will has always been bad, or (2.2) the will became bad without an efficient cause. Augustine rejects the first alternative, from the following consideration: the will must be hosted in a nature, and this nature is necessarily good; for an evil will corrupts a nature, but corruption presupposes a good nature that can be corrupted. Also, this nature must have existed at least at the beginning without a corrupted will, because the good nature must be prior to the will’s corruption. In conclusion, since the will cannot have always been bad (option 2.1), it must have become bad after being good; and since it cannot have been made bad by a thing (option 1), Augustine claims that it must have become bad without an efficient cause (option 2.2) (ciu. XII.6, pp. 360–1 ll. 18–54).5 The core of the diaeresis is that if one assumes that the will causes evil, one is faced with a dilemma: a good will cannot cause moral evil at all (option 1.1.1) and an evil will cannot cause the first moral evil (option 1.1.2). Since Augustine drew attention to it, we may call it “Augustine’s dilemma”: Augustine’s dilemma: A good will cannot cause moral evil, and an evil will cannot cause the first moral evil.

Augustine bypasses the dilemma, for he denies that the will or anything else is the efficient cause of evil. But as we will see, it is a dilemma for medieval thinkers, who assume that the will is the efficient cause of evil (and that Augustine held this view). In order to further illustrate the impossibility of assigning an efficient cause to the evil will, Augustine proposes a thought experiment that enjoyed popularity among his medieval readers: two men, who are equally conditioned as to their psychological and bodily state (animo et corpore), both see the same beautiful woman; one of them harbors lust while the other perseveres in a chaste will. What explains that one sins but not the other? No contrastive explanation can be given, for whatever would be the cause of sin in the lustful man would have had the same effect in the chaste man (ciu. XII.6, p. 361 ll. 58–82). The only explanation for the lustful man’s sin is that his evil will comes not from the fact that he is a nature 5

For a full reconstruction of Augustine’s diaeresis in ciu. XII.6, see Hoffmann forthcoming a, Appendix.

Does Evil Have a Cause?

167

(that is, a being), but rather from the fact that he is made from nothing (p. 361 ll. 89–93). But this is neither a positive cause, nor a contrastive explanation of why one man sins and the other not. Rather, being made from nothing is the reason why a rational creature is not all-perfect like God, and so it explains why the created nature is able to sin and why it becomes unable to sin only through the gift of grace. The evil will has no positive cause at all, as Augustine insists by arguing ad absurdum: “For if a nature is the cause of the evil will, are we not compelled to say that evil comes from something good and that something good is the cause of evil?” (pp. 361–2 ll. 93–5). As he had just observed, something good cannot cause an evil will (p. 360 ll. 21–3, p. 361 ll. 52–4). The Deficient Cause of Evil Augustine concludes his investigation into the cause of evil with a famous statement that concisely summarizes his position: Hence no one should seek the efficient cause of an evil will, for [the cause] is not efficient, but deficient [deficiens], for this [evil will] is not effection [effectio] but defection [defectio]. For to defect [deficere] from that which is highest to that which is lesser is precisely what it is to begin to have an evil will. And wishing to find the causes for these defections, since, as I said, they are not efficient but deficient, is like wishing to see darkness and hear silence. (ciu. XII.7, p. 362 ll. 1–7)

Here he writes that the cause of an evil will is deficient. Shortly thereafter he uses the term “deficient cause”: The things, the greater they are and the more they do good, have efficient causes, for then they do something; but insofar as they defect and thus do evil, they have deficient causes, for what do they do in that case, other than emptiness [vana]? (ciu. XII.8, p. 362 ll. 3–6)

What does Augustine mean by the “deficient cause” of the evil will? He does not give any explicit answer. But he gives some indications that it should not be taken as a cause that acts, albeit not as it should. This would be a deficient efficient cause, hence an efficient cause after all, as, for example, shipwreck is caused by a pilot who does not steer well.6 Nor 6

In a speculative and philosophically ingenious interpretation of Augustine’s theory of the deficient cause of evil, MacDonald 1999 understands it as a deficient efficient cause: the will, understood as a power, is deficient because of carelessness in practical reasoning, and thus it produces a deficient act. His interpretation of Augustine resembles Aquinas’s account of the deficient cause of evil (see Chapter 7).

168

Whence Evil?

does Augustine’s deficient cause of evil correspond to the privation of an efficient cause, in the way that shipwreck is caused or explained by the absence of the pilot who should be present (cf. Aristotle, Phys. II.3, 195a13– 14; Met. V.2, 1013b11–16). Let us first see why Augustine’s deficient cause of evil is not a deficient efficient cause. For Augustine, sin, or an evil will, consists in a “defective motion” away from God and toward creatures (ciu. XII.7, p. 362 ll. 3–4; lib. arb. II.20.54.204). He argues that it has no cause because it is nothing, just as it cannot be known because it is nothing (lib. arb. II.20.54.202–4; cf. ciu. XII.7, p. 362 ll. 10–20). When he writes that defecting (defectio) from God makes the will bad, he does not mean that defecting is the efficient cause of an evil will; rather, defecting is what the evil will consists in. He continues that “the cause of defecting is altogether lacking [deficit]” (ciu. XII.9, p. 363 ll. 4–5). Essentially the same idea is already present in De libero arbitrio. Pressed to explain why some rational creatures (that is, angels and humans) never want to sin, others want to sin permanently, and still others want to sin sometimes but not always, Augustine responds that he can only point out that the root of every evil is avarice, the impious will of wanting more than one needs (lib. arb. III.17.47.161–III.17.48.167). Sinning has the will (i.e., actual or dispositional willing) as its first cause, otherwise one would not be blamed for it. But Augustine cannot indicate a cause for the will that causes sinning; he only says that this will does not come to pass by nature and that it is avoidable (lib. arb. III.17.49.168–III.18.50.171). For the same reason, Augustine refrains from explaining in his thought experiment why one man harbors an unchaste will and the other does not. Neither does Augustine’s deficient cause of evil correspond to the privation of efficient causation, that is, to the inactivity or absence of an efficient cause that should act. The reason is that then the first evil would consist in an omission. But the sin of the angels is not an omission, that is, failure to love; rather, it is a misguided love. As we just saw, Augustine describes this disordered self-love as avarice; he specifies that for the devil, avarice meant loving his own power.7 He also describes it as loving himself “to the point of contempt for God” (ciu. XIV.28) and as pride, that is, putting himself above God and refusing to submit to God.8 Why the devil was proud, that is, why his motive for being proud actually was a motive for him, Augustine does not and cannot say. And

7 8

Gn. litt. XI.15, CSEL XXVIII/1: 347 ll. 4–6, 15–17; Gn. litt. XI.23, p. 355 ll. 15–17. Lib. arb. III.25.76.262–3; Gn. litt. XI.23, p. 355 ll. 15–23; ciu. XII.6, p. 359 ll. 4–6.

Does Evil Have a Cause?

169

this is the point: an evil will ultimately lacks an explanation. Any explanation could only be a penultimate explanation.9 Sourcehood and Alternative Possibilities While an evil will cannot be ultimately explained, it can be traced to the agent. Regarding the evil will, Augustine explicitly stresses sourcehood and the ability to do otherwise: “In him in whom an evil will arises, something arises that would not arise if he did not want it to; and hence the defects, which are not necessary but voluntary, entail just punishment” (ciu. XII.8, p. 362 ll. 6–9). The evil angels “would not have defected from a good will if they had not wanted to” (ciu. XII.9, p. 363 ll. 16–17). Similarly, he says about the deficient motion of turning away from God: “This deficiency [defectus], however, since it is voluntary, has been placed under our control. For if you are afraid of it, you ought to will against it; and if you will against it, it will not happen” (lib. arb. II.20.54.204). He argues that one does not sin in what one cannot guard against (lib. arb. III.18.50). Based on what we have seen so far, the difference between good and bad angels can be explained only up to a certain point: some angels sinned out of pride, which is a disordered love and thus a case of evil willing, and this evil willing cannot be traced to anything anterior and outside the evil angel. Other angels did not sin, and likewise their good will cannot be traced to anything anterior and outside the good angel. But the matter is more complex. Augustine holds that all angels were created with a good will (ciu. XII.9) and that thereafter, the good and evil angels differ only regarding their wills and desires (uoluntatibus et cupiditatibus) (ciu. XII.1).10 But he also thinks that the good angels remained faithful because they received a greater grace than the evil angels (ciu. XII.9). The evil angels would not have sinned if they had wanted to be good, but persistently wanting to be good depended not on them but on God.11 If this means that the angels’ sin or perseverance in the good, and hence their eternal punishment or reward, was not up to the angels but rather depended on God giving them or refusing them the grace that makes them persevere in the good, then 9 10 11

Chappell, 1994 and 1995, ch. 8, makes a good case that, according to Augustine, the first evil will is in principle unexplainable. For the inexplicability of the sin of the angels, see also King 2012, 267–71. In Gn. litt. XI.17 and XI.26, Augustine considers the hypothesis that the good and bad angels already differed prior to their fall, but he thinks this is unlikely. Rist 2014, 45–52 reads Augustine in this way, not without expressing discomfort. See also Babcock 1991, 107–8 who critiques Augustine for making divine grace the efficient cause of the good angels’ will.

170

Whence Evil?

Augustine’s view would have a highly problematic implication concerning God’s justice. The Origin of Evil in Good Things As we have seen, according to Augustine, the reason why evil has no efficient cause is that only something good could serve as an efficient cause, but the good cannot be the cause of evil. Ironically, as we will see in Chapter 7, in medieval discussions of the cause of evil, Augustine often appears as an authority for just the opposite view: that the good is indeed the cause of evil. The confusion was increased by some texts, written shortly after De civitate Dei XII, where Augustine affirms that an evil will presupposes a good nature in which the evil will occurs. He expresses this idea by means of Ambrose’s phrase that “evil originated from good things.”12 Augustine affirms the origin of evil in good things against the Manichaean idea that evil can only originate from evil things, ultimately from a first evil principle. Instead, he argues, evil originated from a nature that is good and created by God, but which can defect from the good because it is from nothing. He insists that “evil things originate [orta sunt] from good things; evil deeds do not originate from good wills, but rather evil wills originate from good natures” (c. Iul. I.8.36–8). By saying that evil originates in something good, Augustine does not claim that something good causes evil. Some medieval readers of Augustine furthermore attributed to him the view that the good that causes evil is the will, taken as a power of the soul. They found this interpretation of Augustine supported by a passage from the Enchiridion: We must have no doubt that nothing but the goodness of God is the cause of the good things that concern us, while the cause of the evil things is a will that defects from the immutable good and seeks a mutable good, first in the angel and then in the human being. (ench. 8.23)

Yet in this passage, “will” is not understood as a power, but rather as occurrent or dispositional love or desire. Furthermore, Augustine does not here address the question of what might have caused such an evil will (or perverse love) in the first place (the problem of ciu. XII.6–8), but only says that an already evil will causes further evil. 12

Cited in c. Iul. I.9.44, ench. 4.14, and nupt. et conc. II.28.48 (all written around 421); see Ambrose, De Isaac vel anima 7.60, CSEL XXXII/1: 685: “ex bonis igitur mala orta sunt.” For the dating of Augustine’s works, see Allan D. Fitzgerald et al., eds., 1999, xliii–il.

Does Evil Have a Cause?

171

Something else that is curious about the medieval reception of Augustine’s theory of the origin of evil is that later medieval thinkers only rarely consider one of his most original treatments of evil, his discussion of the theft of the pears. As a teenager, Augustine went out with some companions late at night to steal pears, not in order to enjoy them, but rather to throw them to the pigs, for the mere pleasure of doing something that was not allowed (conf. II.4.9).13 This seems at least at first sight to be a case of desiring evil for its own sake – a conclusion that Augustine is himself puzzled by, for to desire evil insofar as it is evil would mean to desire nothing, given that evil is a privation (cf. conf. II.8.16).14 In fact, that evil may be desired as such contradicts a fundamental Neo-Platonic tenet. This tenet was transmitted to medieval thinkers by a separate avenue, as we will now see.

6.2

Pseudo-Dionysius

For medieval thinkers, a fundamental text about the nature and origin of evil is Pseudo-Dionysius the Areopagite’s treatise on evil in De divinis nominibus IV.18–35, for which he draws heavily on Proclus’s De malorum subsistentia.15 Pseudo-Dionysius is a late fifth-century theologian who claimed to be the Athenian convert and disciple of Saint Paul (Acts 17: 34). Medieval thinkers did not recognize his dependence on Proclus even after William of Moerbeke translated Proclus’s treatise into Latin in 1280. Thus they continued to believe he was Paul’s disciple and attributed to him quasi-apostolic authority.16 The Nature of Evil Pseudo-Dionysius talks about evil in two senses, without making this distinction explicitly: either as “evil itself” or as the thing that is evil. Evil itself (τὸ αὐτοκακόν, ipsum malum in John Sarracenus’s translation), which Pseudo-Dionysius also calls evil insofar as it is evil, is not something 13 14

15 16

For a very brief medieval discussion of the theft of the pears, see Aquinas, DM 3.12 ad s.c. 1–2. Whether Augustine truly admits that evil can be desired under the aspect of evil is debated. Matthews 2014 argue that he does, while (more convincingly in my view) Chappell 1995, ch. 7 and MacDonald 2014 argue that he does not. Despite its high dependence on Proclus, Pseudo-Dionysius’s account of evil contains some distinctly Christian accents; see Mühlenberg 1969 and Schäfer 2006, 11–21. For a more detailed discussion of Pseudo-Dionysius’s theory of evil, see Schäfer 2002, 440–52 and Schäfer 2006, 137–53. For a thorough study of the Pseudo-Dionysian background of accounts of evil by medieval thinkers, especially Aquinas, see O’Rourke 2016.

172

Whence Evil?

existent (ὄν, existens).17 Everything, insofar as it exists, is good and from the Good (that is, from God); what is entirely deprived of good does not and cannot exist (DDN IV.20, 720B). By contrast, evil is a weakness, a deficiency of good (IV.30, 732B). And yet one cannot deny that there is evil, for one must acknowledge that vice is a different reality from virtue (IV.19, 716D). Accordingly, Pseudo-Dionysius uses “evil” in a second sense to refer to evil things (τὰ κακά, mala; IV.31, 732B–C). His examples for evil in this sense are an intemperate person and anger. An evil thing participates in the good, because evil “exists only in existing things,” that is, it is hosted by things that, as existent, must have some goodness (IV.20, 720B–C). So evil has no existence on its own (ὑπόστασις, subsistentia), only parasitic existence (παρυπόστασις, which Sarracenus renders inaccurately as privatio substantiae) (IV.31, 732C). The Fall of the Angels From the beginning of his treatise on evil, Pseudo-Dionysius frequently refers to the fall of the angels, but he does not propose a narrative or detailed theory of the fall. He begins his treatise with this question: since the “beautiful and good” is desirable and lovable to all things, how is it that the demons do not love it, thus falling away from their angelic identity and becoming the cause of all evil for themselves and for others? (IV.18, 716A). Pseudo-Dionysius denies that they have always been evil, for they are not evil by nature. In fact, nothing can be evil by nature. His argument is essentially the same as Augustine’s: evil is a corruption, and so it presupposes a good nature that can be corrupted (IV.23, 724C–D). Another reason why the demons cannot be evil by nature is that God produces only good things (IV.23, 725A). The demons became evil because of a deficient activity of their own, because of “an irrational rage, a foolish desire, a reckless imagination” (IV.23, 725B–C). But Pseudo-Dionysius insists that the demons’ desire is not entirely corrupted: even they desire the good, because they desire existence, life, and knowledge (IV.34, 733C–D).

17

Pseudo-Dionysius, DDN IV.20, 717B–C, Dionysiaca, I: 243–6. My exposition follows the critical edition by Beate Regina Suchla but refers to the page numbers of Patrologia Graeca, vol. 3. The translation most widely used in the thirteenth century is by John Sarracenus, composed in the midtwelfth century (printed in Dionysiaca, ed. Chevallier). Parenthetical Latin terms are from Sarracenus’s translation.

Does Evil Have a Cause?

173

The Origin of Evil Pseudo-Dionysius’s account of angelic sin, being rather undeveloped, was not very important for medieval theologians. Instead, they relied heavily on his account of the origin of evil. Pseudo-Dionysius holds, like Augustine, that evil does not have an efficient cause: it does not come from anything good, just as coolness does not come from fire (IV.19, 716B). But what is important to the medievals is chiefly Pseudo-Dionysius’s emphasis on the idea that evil cannot be a final cause. Evil itself is never done for its own sake: For no matter what one does, no one does it aiming at evil [τὸ κακὸν ἀποβλέπων, ad malum respiciens]. (IV.31, 732B–C; cf. IV.19, 716C)

The passage just quoted served later medieval theologians as support of their axiom that something evil can only be willed “under the aspect of goodness” (sub ratione boni). Pseudo-Dionysius argues that since evil as such is not the intended outcome of an activity, it does not have its own cause: Being is attributed to evil accidentally [κατὰ συμβεβηκός, secundum accidens] and because of something else, not due to a cause proper to [evil] itself. (IV.32, 732C)

According to Pseudo-Dionysius, evil is thus “outside the aim” (παρὰ τὸν σκοπόν, praeter intentionem; IV.32, 732C) of action; that is, evil is an unintended dimension or side effect of the intended good. Proclus, whom Pseudo-Dionysius follows here very closely, expresses the idea that evil is caused accidentally more clearly than PseudoDionysius.18 Proclus seems to owe it to Aristotle by way of Peripatetic writings with which he was directly acquainted.19 We now turn to medieval theories of the cause of evil. While Augustine and Pseudo-Dionysius are their principal authorities concerning the nature and origin of evil, Aristotle’s notion of accidental causality, which Augustine ignores and Pseudo-Dionysius mentions only in passing, allows them to develop more refined accounts of the origin of evil. 18 19

Proclus, De malorum subsistentia 18.50, ed. Boese, 242–4. For an in-depth discussion of the accidental causation of evil in Proclus, see Opsomer and Steel 1999, 247–55.

chapter 7

The Will as the Cause of Evil

Augustine denies that the first evil will has an efficient cause because a good will cannot cause an evil will and an evil will cannot cause the first evil will. By contrast, most medieval thinkers studied in this chapter, while believing they followed Augustine, held just the opposite view, claiming that something good – the will as a power of the soul – is the cause of moral evil. Their misinterpretation follows partly from a misreading of Augustine by Peter Lombard, partly from the projection of their own faculty psychology onto Augustine. After brief observations about Anselm of Canterbury and Peter Lombard, we will discuss William of Auxerre, Bonaventure, and Albert the Great, whose accounts of the first cause of evil rely heavily on Augustine and Pseudo-Dionysius, while Aristotle remains mostly marginal. Next, we will study Aquinas, who virtually abandons Augustine’s notion of deficient causality and instead integrates Aristotle’s notion of accidental causality more systematically into his account than previous thinkers. We will then consider Henry of Ghent’s rather short but characteristic remarks on the first cause of evil. Next, we will turn to the closely related accounts of Richard of Menneville (Mediavilla) and Duns Scotus, whose solutions depend in part on Aquinas’s, but who do not accept a core aspect of his theory. We will conclude with John of Pouilly, who claims that the cause of evil is free decision, that is, intellect and will, and who traces evil to negligence, which lacks a definite cause. Despite notable differences, at bottom all thinkers here considered converge on the admission that evil has no ultimate explanation – which ironically is the point of Augustine’s “deficient cause” of evil, although they failed to recognize this. Of other thinkers we either have no discussion of the first cause of sin (e.g., Ockham) or only within their account of angelic sin, to be considered later (e.g., Godfrey of Fontaines). 174

The Will as the Cause of Evil

7.1

175

Anselm of Canterbury

In De casu diaboli, Anselm asks about the origin of evil: “Whence came the first evil which is called injustice or sin, in the angel who was created just?” (DCD 27, I: 275). In response, Anselm makes three points. The first echoes one already made by Augustine (see Chapter 6, p. 168): injustice is nothing (for evil is not a thing); hence it does not have a proper cause. Second, properly speaking, the evil angel did not recede from justice, for he abandoned justice only by wanting something he should not want. (We can rephrase this in Aristotelian terminology: the angel wanted something per se, and thereby abandoned justice incidentally; so the abandonment of justice has only a per accidens cause.) Third, and most importantly, Anselm clarifies that the evil angel’s desire for the thing he was not supposed to desire does not have a cause: Why did he will what he ought not? – No cause preceded this will, except that he could will it. . . . – Why did he will it? – Only because he willed it. For this will had no other cause that would have somehow incited or attracted it, but it was for itself its efficient cause [ipsa sibi efficiens causa fuit], if this can be said, and its effect. (DCD 27, I: 275)

While Anselm appropriates the Augustinian theme that evil ultimately lacks a cause, the idea that the will itself causes its first evil willing clearly departs from Augustine. Anselm seems to take the will in two senses here: the will (as faculty) causes its will (i.e., volition). Anselm thus anticipates the view of later medieval thinkers, according to whom the cause of an evil volition is the will as power of the soul, while the fact that the will causes evil lacks a full explanation.

7.2 Peter Lombard In discussing the first cause of evil in the Sentences, Peter Lombard remarks repeatedly that the origin and cause of evil is a good thing; in fact, before the first sin, there was nothing evil from which evil could have originated. He supports this thesis with a quotation from Augustine’s De nuptiis et concupiscentia (Sent. II.34.1–3, I: 525–6), although in this passage Augustine merely argues that evil originates from something good.1 Lombard does not mention De civitate Dei XII, where, as we have seen, Augustine denies that something good can be the cause of evil. In fact, he knows this work only secondhand, rarely cites it in his Sentences, and seems unaware of the treatment of the cause of evil in Book XII.2 He does not go so far as to attribute to Augustine the view that the 1 2

See Chapter 6, note 12. For Peter Lombard’s knowledge and use of Augustine, see Kent 2012, 231–4.

Whence Evil?

176

good that is the cause of evil is a good will. But an evil will can cause further evil, as he writes, expounding Enchiridion 8.23 (quoted in Chapter 6, p. 170). In summary, he writes: The first origin and cause of sin, then, was a good, and the second, an evil that originated from a good. (Sent. II.34.3, I: 526)

7.3

William of Auxerre

In William of Auxerre, we find a remarkably rich discussion of evil, which makes ample use of Augustine and Pseudo-Dionysius. Yet unlike them, Auxerre argues that evil has a cause. Auxerre explains the cause of evil in two closely related texts: his discussions of angelic sin (SA II.3.3) and of evil (SA II.11). According to Auxerre, evil as such is the corruption or privation of good, and in this sense, it is nothing. But evil is spoken of also as an evil act or an evil disposition (habitus), and then it is nothing insofar as it is evil, but something insofar as it is an act or disposition (SA II.11.1.2, II: 314 ll. 22–5). Evil corrupts a good, but it does so not in virtue of its being evil, but rather in virtue of the good with which evil is mixed; thus it acts only incidentally (per accidens) (ibid., pp. 314–15). Auxerre distinguishes between a formal, material, and efficient cause of evil. As to the formal cause, he adopts Augustine’s notion of evil as “privation of the natural form [species], measure [modus], and order [ordo]” (SA II.11.3.2, II: 342–3; cf. nat. b. 4). What is perfect has these three; in what is evil, at least one of them is lacking (SA II.11.2.2.1, II: 326 ll. 50–4). A perfect human body has a nose that is not too big in proportion to the rest of the body (ibid., p. 327 ll. 69–70); love of God is perfect when God is loved for his sake (form), above all else (order), and infinitely above all else (measure) (ibid., p. 329 ll. 123–7). The material cause of evil is the fact that creatures are created from nothing, for this is the reason why by free decision one can sin (SA II.3.3, II: 56–7, SA II.11.3, II: 339). Most important, evil has an efficient cause, which is free decision (liberum arbitrium) (SA II.3.3, II: 56 ll. 82–3; SA II.11.3, II: 339 ll. 17–21). Yet Auxerre makes important qualifications; first because he needs to address Augustine’s dilemma: a good will cannot cause an evil will, nor can an evil will cause the first evil will (SA II.3.3, II: 56 ll. 64–70). Second, he must reconcile conflicting authoritative statements (see especially SA II.11.3.1, II: 340). Through inaccurate or incomplete quotes, he makes Augustine and Pseudo-Dionysius each affirm and deny that evil has an efficient cause. Augustine seems to affirm it by claiming that we sin by free

The Will as the Cause of Evil

177

decision of the will (but Augustine writes that we sin “voluntarily,” without mentioning free decision; lib. arb. II.20.53.200); Pseudo-Dionysius seems to affirm it in writing that the soul is the cause of sin (a truncated sentence that is actually a rhetorical question; DDN IV.30, 732A). PseudoDionysius seems to deny that evil has an efficient cause in writing that evil is not from a principle (ex principio) (DDN IV.32, 732C; see Chapter 6, p. 173); Augustine seems to deny it in De civitate Dei XII.7 (SA II.3.3, II: 55–6). Auxerre addresses Augustine’s dilemma and reconciles the seemingly contradictory authoritative statements by means of a solution that accords with De civitate Dei XII.7: Free decision is the cause of an evil will not as such, but incidentally [secundum accidens], that is, by a deficiency. (SA II.3.3, II: 56 ll. 81–4; cf. p. 57 ll. 101–2) The good is the cause of evil only by way of dis-acting and defecting [deagendi et deficiendi], since, as Augustine says, “the cause of an evil will is not efficient, but deficient, for evil is not effection, but defection.” Although properly speaking, that is, without semantic incorrectness, it is true that something good is the cause of evil; something good is nevertheless not the per se cause of evil. (SA II.11.3.1, II: 342 ll. 67–72)

Auxerre, then, interprets Augustine’s deficient cause of evil in terms of Aristotle’s per accidens cause: it is an efficient cause that causes evil not per se, but incidentally. The notion of accidental cause allows Auxerre to assume that the good that causes evil is free decision, although it causes sin only in combination with a defect. He uses the example of someone limping: the “walking power” is the cause of limping – not per se, only because of a defect in a leg or foot. Just so, free decision is the cause of sin, not as such, only because of a defect (SA II.3.3, II: 57 ll. 97–102; cf. SA II.11.1.2, II: 314–15). The defect because of which a person sins is the failure to consider what should be considered and seen in the proper relation (SA II.11.3.1, II: 340 ll. 12–17). Thus Lucifer sinned because he failed to consider his own excellence and the divine excellence in the proper relation (SA II.3.3, II: 57 ll. 94–8). Lucifer knew the proper relation habitually; he knew that no creature can become equal to God. Without bearing this in mind, he considered his own beauty apart from the proper relation to God, wanting to be superior (presumably to the other angels) without being subordinate to anyone (volens preesse et non subesse; SA II.3.3, II: 50–3). This solution

178

Whence Evil?

raises the question of whether the failure to consider what ought to be considered is not itself the first sin, rather than a precondition for it. Aquinas develops Auxerre’s solution systematically and addresses this difficulty. But first, we will consider Albert and Bonaventure, who build upon Auxerre’s account.

7.4 Albert the Great and Bonaventure The authors of the Summa Halensis, Albert the Great, and Bonaventure take a similar approach to the cause of evil. It will suffice to focus on the more developed accounts of Albert and Bonaventure, whose approaches are, as we will see, aligned enough that we can treat them together.3 On this topic, their main sources are Augustine and Pseudo-Dionysius. They also seem influenced by Auxerre – but not by Philip the Chancellor, who normally is an important resource for their theories of moral psychology and ethics. In fact, in discussing the cause of evil, Philip does little more than refer to Augustine as saying that “evil has no cause but rather has a deficiency [defectum] as its cause” and to Pseudo-Dionysius as saying that “evil is uncaused and ungenerated” (Summa de bono, I: 25 ll. 59–61). Formulating the problem of evil, Albert and Bonaventure write that if all evil had a cause that is itself evil, there would either be an infinite regress of evil causing evil, or there would be an uncreated, highest, first evil, as the Manichaeans held. To address this difficulty, Albert and Bonaventure trace evil to the will as a power of the soul. They state explicitly that, so understood, the cause of evil is something good.4 Augustine’s question of whether the evil will has a deficient or an efficient cause now becomes the question of whether evil comes from a deficient or an efficient will. They argue that it comes from a deficient will, and they take this to be Augustine’s view. In fact, Bonaventure writes: Without a doubt, evil or sin is from the will insofar as it is deficient [ut deficiente], as Augustine says explicitly, and Dionysius agrees. (In Sent. II.34.1.2c., II: 806b)

In a similar vein, Albert refers to Augustine to make the claim that free decision or the will is the cause of evil, albeit insofar as free decision (or the 3 4

For the account of the Summa Halensis, see II-II nn. 3–8, III: 7a–14b. Albert the Great, In Sent. II.34.4 arg. 5, sol., and ad 5, Borgnet XXVII: 553a–b; Bonaventure, In Sent. II.34.1.1 arg. 5 in opp., c., and ad 5, II: 803b–804b; 805b. See also Bonaventure, Breviloquium III.1, V: 231a–b. Albert tells us in In Sent. II.6.9, p. 139a that he is writing in 1246; for the dating of Bonaventure’s Sentences commentary around 1250, see Chapter 2, note 10.

The Will as the Cause of Evil

179

will) is deficient.5 Thus they understand the cause of evil as a deficient efficient cause, rather than denying, like Augustine, that evil has a cause. Bonaventure in particular seems to conflate two issues that Augustine kept distinct: how an evil will is caused in the first place (ciu. XII.6–7), and how further evil is caused by an already evil will, that is, by perverse love (ench. 8.23).6 Echoing Augustine, Albert and Bonaventure take the will’s origin from nothing as an explanation of the will’s ability to sin – its vertibilitas (the ability to turn to good or evil), as Albert calls it, or its defectibility, as Bonaventure names it.7 Bonaventure writes: The evil of sin comes from a will that is deficient by way of a natural defect, which means that this defect precedes sin; for unless the will contained in itself nothingness [vanitatem] and defectibility, it could never be deficient in willing the good. (In Sent. II.34.1.2, II: 807a)

Vertibilitas or defectibility does not, of course, explain the actual occurrence of an evil will. Hence Bonaventure asks whether there is any further ground (ratio) or condition (dispositio) that is the proximate cause of the will’s becoming evil. Albert mentions an analogous problem: the ability to do good or evil (vertibilitas) is permanent, but actual sin is not; hence this ability does not explain a particular sin. In response to the problem of what causes an actual evil will or sin, Albert and Bonaventure make a significant claim: they maintain that the will’s freedom to “turn” one way or another (to good or evil) is sufficient to account for an actual evil will or sin, without any further condition being required for the actual occurrence of sin.8 So for Albert and Bonaventure, the will can have contrary effects without being itself subject to efficient causes that determine it to a good or a bad act. This amounts to Augustine’s position that the evil will has no efficient cause. Yet Albert and Bonaventure do not make the connection 5

6

7 8

In Sent. II.34.3 s.c. and c., p. 550a; see also In Sent. II.5.1 s.c. 3, p. 111a. Nevertheless, Albert also mentions Augustine’s claim that one must not search for efficient causes of evil, only for deficient causes; see In Sent. II.34.2 arg. 1, p. 548b. Bonaventure refers to the Enchiridion in support for the claim that the will causes evil as a deficient cause, but actually quotes the text of ciu. XII.7; see In Sent. II.34.1.2 arg. 1, II: 805a–b. For Albert’s use of ench. 8.23 to argue that the will as a faculty is the cause of sin, see his ST I.6.26.2.3 arg. 1, XXXIV/ 1: 195. Albert, In Sent. II.34.4 sol. and ad 3, p. 553a–b; Bonaventure, In Sent. II.34.1.2, II: 807a–b; see also In Sent. II.34.1.1 ad 4, II: 805b. Albert, In Sent. II.34.4 arg. 4 and ad 4, p. 553b; see also In DDN IV n. 168 sol. and ad 2, XXXVII/1: 253a–b; Bonaventure, In Sent. II.34.1.2 arg. 6 in opp. and c., II: 806b and 807b; see also In Sent. II.25.2.3 ad 4, II: 615b. The Summa Halensis makes the same point; see II-II n. 8, III: 14b.

180

Whence Evil?

with Augustine’s claim in this particular context. Instead, Albert explains the will’s ability to do otherwise by the Aristotelian notion of a rational power, which, unlike natural powers, can have contrary effects in determinate circumstances, and in a similar way Bonaventure contrasts a cause that acts freely (the soul) with a cause that acts naturally (fire).9 They also discuss whether evil can be intended. As was common, they distinguish between two senses of evil. On the one hand, “the subject of evil materially speaking” (Albert) or “evil in the concrete sense” (Bonaventure) is something real: an act, a disposition (habitus), or a thing (e.g., a blind animal). On the other hand, “the form of evil” or “evil insofar as it is evil” (Albert), which Bonaventure calls “evil in the abstract sense,” is nothing: it is merely a privation, like blindness.10 Evil in the first sense can be intended (Bonaventure’s examples are fornication, stealing, and lying), but evil in the second sense cannot be intended, that is, one cannot want the evil of these acts under the aspect of evil.11 Albert uses two examples to explain how the deficient will causes the evil act efficiently and hence intentionally without intending the deficiency of the act. Like Auxerre, he compares the deficient will to a crooked leg that makes a person limp. As he explains, when the person walks, what is intended and efficiently caused is not limping but rather walking. He also compares the deficient will to a defective seed that generates deformed offspring; presumably, Albert’s point is again that what is efficiently caused and intended is the offspring but not its deficiency.12 In 1250, a few years after he wrote the Sentences commentary, Albert’s lectures on Pseudo-Dionysius’s De divinis nominibus address the problem of the cause of evil through a new approach, one which seems to have been influential on Aquinas, thanks to whose report the lectures have come down to us.13 Occasioned by Pseudo-Dionysius’s affirmation that “evil is not from the good” (DDN IV.19, 716B),14 Albert asks “whether evil is from

9 10 11

12

13 14

Albert, In Sent. II.34.4 ad 4, p. 553b; Bonaventure, In Sent. II.34.1.2 ad 1, II: 807b. For Aristotle’s theory of rational vs. natural powers, see Chapter 1, pp. 26–7. Albert, In Sent. II.34.1 sol., p. 547b; Bonaventure, In Sent. II.34.2.3c., II: 815a. Albert, In Sent. II.34.2 sol., p. 549a–b; In Sent. II.34.3 sol., ad 1, and ad 2, pp. 550a–551a; cf. Super Ethica III.6 n. 191, XIV: 170b–171a; Bonaventure, In Sent. II.34.1.3, II: 808a–809b, citing PseudoDionysius, DDN IV.31, 732B–C. Albert, In Sent. II.34.3 sol., p. 550a–b. He attributes the example of the crooked leg to Augustine (see perf. iust. 2.4) and the example of the defective seed to Aristotle (cf. Phys. II.14, 199b3–7), although it is found more clearly in Averroes’s In Phys. II.82, ed. Iuntina IV: 80B. Cf. Wippel 2016, 27. For the dating and Aquinas as reportator, see In DDN, Prolegomena, XXXVII/1: VIa–b. “Malum non est ex bono” (Dionysiaca, I: 234) in Sarracenus’s translation, which Albert takes as his base text; see In DDN I n. 5, XXXVII/1: 3b.

The Will as the Cause of Evil

181

something good as its cause” (In DDN IV n. 153, XXXVII/1: 239a). As he remarks, Augustine and Pseudo-Dionysius seem to give opposite answers: Augustine explicitly says that evil is caused by a good, insofar as free decision of the human being or angel, which is good, is the cause of evil. Hence Dionysius seems to say something false. (In DDN IV n. 153 arg. 2, p. 239a)

Albert gives value to both perspectives. In order to affirm that evil is caused by something good, he makes use of Aristotle’s distinction between per se causes and accidental causes.15 As we have seen, Auxerre had used this distinction already, but Albert’s conception of accidental causes follows Aristotle more closely than Auxerre’s did. According to Aristotle, as expounded loosely by Albert, a per se cause causes through what is essential to itself: thus, heat or something that is hot by nature is the per se cause of something else becoming hot. By contrast, the term “accidental cause” refers to something that is incidental to the causal relation. A doctor can be the cause of a statue, but not insofar as he is a doctor; going to the marketplace can be the cause of finding one’s debtor, but not because one was looking for him.16 Albert explains that evil does not have a per se cause because evil is a privation, and a privation does not result from “the effectiveness [per efficientiam] of a principle” but rather “from the deficiency [ex defectu] of a principle” (which might be an allusion to De civitate Dei XII.7). But evil has an accidental cause, for when doing evil, one aims per se at something good, but one departs incidentally from the due end (debitum finem) (n. 153 sol., p. 240a). (A fornicator departs from the order of morality by aiming at a good – pleasure – not at an evil.) Hence Pseudo-Dionysius is right: the good is not the cause of evil, for it is not the per se cause of evil. Albert’s Augustine, too, is right: the good is the cause of evil, for it is its accidental cause (n. 153 ad 2, p. 240a). Later, in the Summa theologiae, Albert repeats the claim that the good can cause evil only incidentally (accidentali operatione), not per se (essentiali operatione),17 and he clarifies that evil originates in a good not insofar as it is good, but insofar as it is deficient.18 15

16

17 18

Ironically, Albert does not recognize that a little later, Pseudo-Dionysius himself employs the notion of accidental causality (DDN IV.32; see Chapter 6, p. 173). By contrast, Aquinas uses Aristotle’s theory of accidental causality to explain DDN IV.32; see In DDN IV.22 n. 584, ed. Pera, XXXVII/1: 215a. In DDN IV n. 153 sol., XXXVII/1: 239b–240a. While Albert does not mention Aristotle by name, his examples are adaptations of Aristotle’s; see Phys. I.8, 191b4–5 and Phys. II.4, 196a3–5. For Aristotle’s theory of accidental causality, see also Phys. II.3, 195a27–b3. ST I.6.26.2.3 sol., XXXIV/1: 196b. ST II.18.114.1–2 sol., Borgnet XXXIII: 331–2. The Summa theologiae was written after 1268; at least parts of it after 1274 (see Prolegomena, XXXIV/1: XVI–XVII).

Whence Evil?

182

7.5

Thomas Aquinas

The problem of the cause of evil receives its most elaborate and original treatment in Aquinas. He discusses it throughout his career and continually enriches his solution. He approaches the issue usually by asking whether something good is the cause of evil.19 Aquinas does not mention the sin of the angels in his treatments of the cause of evil, but his solution to the cause of evil is highly relevant to his explanations of angelic sin. Aquinas’s treatments of the cause of evil presuppose the distinction between different meanings of evil. His most detailed discussion of these is in his Sentences commentary, where he closely follows Avicenna and adopts his terminology. (1) “Per se evil,” which Aquinas identifies with evil in the abstract sense, is the privation of some perfection (e.g., the lack of a foot or a hand). (2) “Accidental evil,” which he identifies with evil in the concrete sense, is further subdivided. On the one hand, there is (2.1) “the subject of privation” (which he sometimes calls the “subject of evil”): it can be an action (e.g., adultery), a disposition (e.g., intemperance), a passion (e.g., pain or suffering), or a substance (e.g., an evil man or soul). On the other hand, there is (2.2) “the thing that causes the privation in some other thing” (e.g., fire that burns, a sword that cuts, clouds that darken).20 These distinctions make it clear that evil is in one sense nothing (as a privation), but in another sense it is a reality: there are evil things (the subjects of privation), and evil can be harmful (as the cause of a privation in something else).21 But according to Aquinas, we can even say of evil as a privation that it “is,” in a sense – it is not a real being (ens) because it is not a substance, a quality, or another of Aristotle’s ten categories, but it signifies the truth of a statement. We can indeed rightly say that there are privations, for we say of people that they are blind, without implying that blindness is a thing.22 For Aquinas’s account of the cause of evil, I will focus on De malo Question 1, Article 3, which is his most mature discussion of this topic. In this text, he approaches the problem from two perspectives. First, he 19

20

21 22

Aquinas, In Sent. II.34.3; SCG III.10 and III.13; In DDN IV.22; ST I.49.1; ST I-II.75.1; DM 1.3. For discussions of Aquinas’s theory of the first cause of sin, see Wippel 2016, 25–30, the studies referred to in Jensen 2018a, 74, note 3, and those cited in notes 28 and 29 below. Aquinas, In Sent. II.34.2 sol., ed. Mandonnet II: 875–6, summarized in DM 1.1c., XXIII: 5–6 ll. 165–9, 237–42. For Aquinas’s source, see Avicenna, Liber de philosophia prima IX.6, ed. Van Riet, 496–9. For Avicenna’s theory of evil and Aquinas’s use of it, see Steel 2002; for Aquinas’s account of evil in De malo, see Wippel 2016, 16–25. Adultery and intemperance are my own examples; the others, some of which he takes from Avicenna, are given by Aquinas. Brock 2018 provides an ample discussion of what he calls “positive associates of evil” in Aquinas’s thought. In Sent. II.34.1 sol., II: 872; DM 1.1 ad 19. Cf. Bonaventure, In Sent. II.34.2.3 ad 3, II: 815b–816a.

The Will as the Cause of Evil

183

reasons from the effect to the cause, asking whether evil has a cause. He answers that it does and that its cause is something good. His second perspective is from the cause to the effect: here he asks how something good causes evil. To ask about the cause of evil is not the same as to ask how the “subject of evil” is caused. For example, in the case of sin, which is an inordinate act, the subject of evil is the act, and this has a per se cause (ST I-II.75.1c.). This is trivial, and Aquinas’s examples show this clearly: a man is the per se cause of the act of adultery, or, in the natural order, fire is the per se cause of burning. But for Aquinas, the inordinateness (inordinatio) itself (the “per se evil” according to the above distinction) also has a cause. Of what kind? On this point, Aquinas departs more radically and probably more consciously from Augustine than Albert and Bonaventure. For the most part, Aquinas treats Augustine’s insistence that evil does not have an efficient cause with respectful silence. Also, Aquinas rarely makes use of Augustine’s notion of a deficient cause of evil. The few times he does, he understands it as a deficient efficient cause that does evil accidentally (SCG III.10 n. 1940; ST I-II.75.1c.). Augustine’s idea that evil lacks an efficient cause is unacceptable for Aquinas, for the lack of a cause (defectus causae) does not explain evil (ST I-II.75.1c.). Instead, he argues that evil must have a cause, and he clearly understands this as an efficient cause. The reason evil requires a cause is that evil is in a thing not by nature but contrary to nature (preternaturaliter). In fact, a privation means that something is lacking that by nature should be present, as Aquinas illustrates with two examples. To lack wings is not bad for humans, because for them it is not natural to have them; by contrast, to lack sight is bad for humans, although this is not bad for a stone, for which sight is not natural. Evil, then, must have a cause; but it cannot have a per se cause; hence it must have an accidental cause (causa per accidens).23 Aquinas attributes the idea that evil has an accidental cause to Pseudo-Dionysius but, like Albert, he conceives of the notion of accidental cause in distinctly Aristotelian terms.24 Why is it that evil cannot be caused per se? Aquinas gives three reasons. First, a per se cause intends its effect, whereas the effect of an accidental cause is unintended (preter intentionem). His example for the latter is taken from Aristotle: digging a hole for a grave (in Aristotle, for a plant) is the accidental cause of finding a treasure, for this discovery was not intended when digging (cf. Met. V.30, 1025a14–19). Likewise, evil as such is not 23 24

DM 1.3c., p. 15 ll. 175–90. See also SCG III.13; ST I.49.1; ST I-II.75.1. In Sent. II.34.3 s.c. 2 and ad 5, II: 879 and 882; In DDN IV.22 n. 584; DM 1.3 ad 5.

184

Whence Evil?

intended; indeed, when one does something evil, one is pursuing what appears to be something good. Second, every per se effect is in some way similar to the cause. Any cause, insofar as it causes, is in act and hence good, and so its per se effect, as a likeness to the cause, is good. Third, every per se cause has a determinate order to its effect, whereas evil occurs by omitting the order (DM 1.3c., pp. 14–15 ll. 143–74). Having considered evil from the perspective of effect to cause, Aquinas’s second perspective is from cause to effect. The cause of evil is something good, in one of two ways: (1) something good is the accidental cause of evil (to be precise, an accidental efficient cause); (2) something good is the cause of evil insofar as it is deficient (that is, it is a deficient efficient cause). Both ways are found in nature: (1) fire corrupts water, but the directedness (intentio) of fire is not principally to destroy water but rather to induce the form of fire into the matter; (2) a defective seed is the cause of a birth defect. In the realm of nature, the second way of causing evil must be traced to the first: the seed must have been corrupted through accidental causality (DM 1.3c., p. 15 ll. 201–26). Voluntary agents cause evil in two analogous ways. The will causes evil accidentally by wanting a good to which evil is attached. Just as fire accidentally causes the corruption of water, so too, in Aquinas’s example, the pursuit of inordinate sexual pleasure causes evil accidentally, namely the absence of the order of reason and of the divine law (ibid., p. 15 ll. 227–32). Adulterers intend the good of pleasure, without primarily intending the evil implied in the action; yet, rather than renouncing the pleasure, they accept the evil that comes with it.25 The cause of evil, then, is the will, which causes evil accidentally insofar as it is drawn toward what is good in a certain respect (bonum secundum aliquid), but which has something absolutely evil (simpliciter malum) attached to it (DM 1.3c., p. 16 ll. 243–8). In doing so, however, the will is in control. Whereas fire necessarily corrupts water (in specific circumstances), the prospect of pleasure does not necessarily cause the pursuit of pleasure. Aquinas says, “however much an external sensory thing attracts [the will], it is in the power of the will to accept it or not” (ibid., p. 15 ll. 233–9). Explaining sin by accidental causality poses a problem. A treasure is found by chance when digging a hole, but hardly does one cause evil by chance when seeking pleasure in adultery. How, then, can the will do evil, that is, act outside the order of reason and of the divine law, without being already deficient? This difficulty is implied in Augustine’s dilemma: a good 25

DM 1.3 ad 15, p. 18 ll. 428–37; cf. SCG III.6 n. 1907 and DM 3.12c., p. 92 ll. 131–43.

The Will as the Cause of Evil

185

will cannot cause an evil will, and an evil will cannot cause the first evil will. Aquinas cuts Augustine’s Gordian knot: it is indeed the will insofar as it is deficient that causes evil. Yet this deficiency is not evil, nor caused by anything evil (ibid., p. 16 ll. 253–85, and ad 6).26 Jacques Maritain called Aquinas’s solution “one of the most original of his philosophical discoveries.”27 By contrast, Aquinas’s early readers, Richard of Menneville and Duns Scotus, were not convinced by it, as we shall see shortly. In what way, then, is the will deficient before it elicits its first deficient act? According to Aquinas, the will’s initial, blameless deficiency consists in the agent’s failure to consider the rule of reason and of the divine law when making a choice that ought to be governed by them. Aquinas does not ponder the hypothesis of conscious transgression of the rule; he might consider this hypothesis impossible because only an already evil will can act in conscious contempt of the rule of action (cf. ST I-II.78.1), or perhaps because a conscious transgression would imply knowingly rejecting happiness (if the connection between the rule and happiness is obvious). Crucially, he argues further that not attending to the rule is a deficiency (defectus) but not an evil, because its consideration is not required while one is not acting (DM 1.3c., p. 16 ll. 248–52, 263–7, 271–5). It is a pure negation, not a privation (ibid., l. 289 and ad 13). If the nonconsideration of the rule were evil, it would not explain the first evil but would itself require explanation. The will becomes sinful only when, without considering the rule, one makes a choice that was to be measured by the rule. Aquinas compares this to a woodworker who needs to hold a measure in the hand only while cutting wood (DM 1.3c., p. 16 ll. 258–62, 278–81). What caused this deficiency, that is, the failure to attend to the rule of action? In natural things, the deficiency of what causes evil (e.g., a defective seed) has itself a cause; Aquinas traces it to something that caused this deficiency per accidens. Thus the occurrence of physical evil can be explained without gaps, at least in principle. In voluntary agents, Aquinas does the reverse: that the will causes evil per accidens (e.g., seeking pleasure outside the moral rule) is traced to a deficiency in the will, namely to a failure to heed the rule when making a choice. He does not, however, in turn trace this deficiency to any earlier accidental or deficient cause. In fact, he does not trace it to any cause at all:

26

Aquinas first develops this idea in SCG III.10.

27

Maritain 1942, 23.

186

Whence Evil? There is no need to search for any cause of this nonuse of the aforesaid rule, because for this the very freedom of the will [libertas uoluntatis] is sufficient, by which it can act or not act. (DM 1.3c., p. 16 ll. 268–71; see also ad 6, p. 16 ll. 329–30)

If the cause of the nonuse of the rule cannot be traced any further than to the freedom of the will, then neither, it seems, can the will’s choosing-while-not-considering-the-rule be traced any further. Nothing explains why the will acts apart from the rule; the evil will remains at bottom inexplicable. Albert and Bonaventure had made a similar observation, and Augustine likewise insisted that there is no efficient cause of an evil will. Aquinas closes his response, in fact, with a reference to Augustine: And this is why Augustine says in Book XII of the City of God that the will is the cause of sin insofar as it is deficient. But he compares this deficiency to silence and darkness, because this deficiency is only a negation. (DM 1.3c., p. 16 ll. 285–9)

Aquinas is wrong, of course, to make Augustine say in this text that the will is the cause of evil, but he is right to adduce it in support of the idea that the deficiency is a mere negation, which has no efficient cause. Aquinas’s account of the nonconsideration of the rule in the first sin is intricate. What role does the nonconsideration of the rule really have in explaining how a still innocent individual can do evil for the first time? Prior to the moment of choosing, it seems irrelevant whether one considers a moral rule; in the moment of choosing, not considering the rule seems itself a sin of negligent omission. Then this omission would be the first sin, rather than a presupposition for the first sin. Aquinas partly agrees with this reasoning, but he does not draw the conclusion that nonconsideration is the first sin. He distinguishes between blameless nonconsideration – a pure negation – and sinful nonconsideration – a privation; the first occurring prior to choice (at least by a priority of nature), the second while choosing (DM 1.3 ad 13). In Aquinas’s account only blameless nonconsideration explains the possibility of a first sin. Since the cognitive state prior to choosing is indeed irrelevant, we must consider the cognitive state in the moment of choice, in order to see how Aquinas seems to understand the role of the blameless nonconsideration in causing the first sin. In the moment of choice, nonconsideration does not make the will sinful; in fact, it has no causal efficacy on the will. Rather, it is the will’s initiative to choose that renders the nonconsideration sinful, for when choosing, the rule ought to be

The Will as the Cause of Evil

187

considered.28 Nevertheless, the blameless nonconsideration is not irrelevant to the evil choice, for on Aquinas’s presuppositions, without it, the will would not have made that choice. By analogy, faulty cutting of wood is made possible by the nonuse of a measure, but not caused by it; and the nonuse of the measure becomes blameworthy once the woodworker cuts wood that should have specific dimensions (DM 1.3c., p. 16 ll. 271–85).29 As we will see in Section 8.3, nonconsideration of the rule also plays an important role in Aquinas’s account of angelic sin, where he emphasizes its cognitive dimension more than its volitional one.

7.6 Henry of Ghent Medieval theologians who published their Sentences commentary usually offer us an explicit discussion of the first cause of evil in Book II, Distinction 34. Henry of Ghent did not publish his, and we only have a brief discussion in his first Quodlibet, Question 16, where, as we have seen in Section 3.2, his main point is that free decision is principally rooted in the will, not the intellect. In this context, he addresses the problem of the first cause of evil, interpreting Augustine’s thought experiment of the two men in the same psychological conditions who have opposite reactions when seeing the same woman (ciu. XII.6). For Henry, this passage shows that the will is not made evil by a judgment of reason. In fact, both men are ex hypothesi in the same cognitive state, but only one sins. Henry concludes: “Hence besides the freedom to judge in reason it is necessary to posit in the will the freedom to choose what is judged,” allowing it to choose evil apart from reason judging that it should do so – except if reason proposes the (ultimate) end or what is recognized as indispensable to attain it (Quodl. I.16c., V: 102 ll. 83–90; cf. Chapter 3, p. 66). So Augustine’s thought experiment shows that the will can turn to evil entirely by itself. Henry clarifies that it can do so in virtue of its “natural 28

29

Barnwell, 2015 and 2019, argues that Aquinas’s theory implies the following sequence: (1) innocent nonconsideration of the rule when the agent is not yet intending to make a choice; (2) negligent, sinful nonconsideration of the rule when the agent intends to make a choice and should be aware that it needs to be regulated; (3) the act of choice itself. The first sin would then occur at the second instant and consist in the nonconsideration of the rule; thus the theory of nonconsideration does not serve to explain the first sin. But according to Aquinas, (2) and (3) are not and need not be distinct, since a certain nonobservance of a special circumstance is part of every sin. See Chapter 8, note 18 and Jensen 2019, 168–9. My understanding of Aquinas’s position has benefited from Nicolas 1960, 199–204 and from the debate between Barnwell 2015 and 2019 and Jensen 2018a and 2019.

188

Whence Evil?

defectibility,” which derives from its being made from nothing (p. 109 ll. 40–3).30 Henry makes Augustine say, awkwardly, that the cause of the evil will in the one man is his evil will (p. 110 ll. 52–3). In confirmation that evil cannot be traced further than to the will, Henry quotes from De libero arbitrio, where Augustine affirms the impossibility of indicating the cause of an evil will and claims that, whatever its cause may be, one can resist it.31 Henry formulates the central idea in his own solution as follows: The voluntary evil has no other cause than a voluntary defect of the will, and someone who searches for another cause, searches for a knot in a reed. For he searches for a positive cause where there is none, and he searches for the cause of a voluntary defect of the will, of which nothing other is the cause than the will by itself [voluntas sibi]. (Quodl. I.16c., V: 111 ll. 71–5)

To search for a knot in a reed, that is, for something not existing in nature, is a Latin saying that means that one is searching for a problem where there is none.32 Henry does not engage Aquinas’s theory of nonconsideration as a precondition for sin. But he argues that sin cannot be traced in the last analysis to a cognitive deficiency, for it is incumbent on the will to make reason consider what is to be considered to avoid error. Consequently, any error in reason that causes sin originates in a flaw of the will (Quodl. I.16c., V: 112 ll. 89–96). In fact, as we have seen, Henry argues that the first sin in Adam, who was not subject to error, consisted in acting directly against the judgment of right reason (Quodl. I.17c., V: 128–9 ll. 92–08; see Chapter 3, pp. 69–70). Augustine’s dilemma – an evil will cannot come from a good will and the first evil will cannot come from an evil will – is addressed only implicitly in Henry’s solution. Henry assumes that the first evil will is caused by a will that is not entirely unflawed, for it has a natural defectibility, consequent upon its being created from nothing. Later, in discussing angelic sin, he says that the first evil act came from a will that was neither pious nor impious (Quodl. VIII.10c., 322vR).

7.7 Richard of Menneville (Mediavilla) Richard of Menneville addresses the problem of the cause of evil in Question 23 of his Disputed Questions (of the 1290s), which asks 30 31 32

Cf. Bonaventure, In Sent. II.34.1.2, II: 807a; see p. 179 above. Quodl. I.16c., V: 111–12 ll. 77–89, quoting excerpts from lib. arb. III.17.48.164–III.18.50.171. See the entry for scirpus in Harper’s Latin Dictionary.

The Will as the Cause of Evil

189

specifically about the cause of the first sin of the angels, and in Book II, Distinction 34 of his Sentences commentary (published after the Disputed Questions).33 Richard argues that the cause of the sin of the angels (and of rational creatures in general) is the will, understood as a principle that is good but defectible (QD 23c., IV: 26–30; cf. In Sent. II.34.2.3c., II: 424a). In explaining how the will can cause evil, he follows Aquinas’s De malo account partly verbatim, but without acknowledging him. Like Aquinas, he argues that evil as evil – in Richard’s words, the depravity (deformitas) of the sinful act – cannot be caused by the will per se, only accidentally (QD 23c., IV: 30–4; cf. In Sent. II.34.2.3c., II: 424a). He also adopts from Aquinas (SCG III.10 n. 1940, DM 1.3 ad 14) a distinction between two types of accidental causality: what is accidental in a causal relation can be accidental to the effect or to the cause. His example illustrates both types: a doctor builds a house that will cause misfortunes to its inhabitants. It is accidental to the effect of the house that its inhabitants suffer misfortune; it is accidental to its cause, the builder, to be a doctor (QD 23c., IV: 34; cf. In Sent. II.34.2.3c., II: 424b). Evil is caused accidentally with respect to the effect, insofar as an unintended depravity happens to be connected to the intended act (QD 23c., IV: 34; cf. In Sent. II.34.2.3c., II: 424b). Evil is caused accidentally with respect to the cause, because the will, the cause of the act, happens to be defectible (QD 23c., IV: 36). Sinning thus involves both kinds of accidental causation: the first, insofar as some depravity is attached to the intended good; the second, because a good that involves depravity would not be intended “unless there is in the will some defectibility” (QD 23c., IV: 34–6). Richard does not explain in which sense defectibility is accidental to the cause, that is, to the will. He might understand this like Aquinas, who wrote that it is accidental to the agent qua agent to suffer a deficiency of power, for an agent acts insofar as it has power, not insofar as it lacks it (SCG III.10 n. 1940). Or he might think, as Scotus will formulate explicitly, that it is accidental to the notion of will to be defectible, for God’s will is not. Richard addresses directly Augustine’s dilemma, which he formulates in an opening argument: the first sin of an angel did not come from his will insofar as it was well-ordered, for a well-ordered will cannot cause a sin, nor insofar as it is disordered, for a disordered will has already sinned (QD 23 arg. 9, IV: 10). Richard responds: 33

On the dating, see Chapter 3, note 106. For Richard’s account of the cause of evil, see also D’Ercole 2017, 106–23.

190

Whence Evil? Before the first sin, the will of the angel was neither ordered nor disordered, as far as moral order or disorder are concerned, but rather it was ordered according to the order of nature. The moral disorder could happen only on account of the freedom and defectibility [of the will] taken together, and it happened when he sinned. (QD 23 ad 9, IV: 50)

With the will’s defectibility we have the nonculpable deficiency that serves the same function in Richard and in Henry of Ghent as Aquinas’s nonconsideration of the rule. The only difference is that Aquinas insisted on the voluntary nature of this deficiency, while the created will’s defectibility is natural, not voluntary. Yet Richard addresses the concern of voluntariness by clarifying that the will’s defectibility explains the possibility of the first sin only in conjunction with its freedom. By its freedom, the will could either become actually deficient or guard against it (QD 23c., IV: 38; In Sent. II.34.2.4c., II: 425a). Like Henry and in contrast to Aquinas, Richard stresses that no actual defect – no cognitive deficiency – is presupposed in the will causing the first sin. He thinks, however, that some cognitive deficiency did in fact accompany angelic sin: the angel focused on the perfections he had, not considering the perfections he lacked. But Richard insists that he could have sinned also without this cognitive defect (QD 23c., IV: 40).

7.8 John Duns Scotus For reasons that will become apparent shortly, Duns Scotus addresses the problem of the first cause of evil in his Sentences commentaries in close connection with God’s causal contribution to sin. As is typical for Scotus, he builds his solution on the positions he critiques. His foils are Richard of Menneville and a view that resembles Aquinas’s.34 In the Ordinatio, Scotus starts the discussion of the first cause of sin with the claim that “sin, insofar as it can have a cause, is from something good” (Ord. II.34–7 n. 71, VIII: 395) – a formulation found almost verbatim in Aquinas (DM 1.3c., XXIII: 14 ll. 139–41). For Scotus, the good that causes evil is the will. He thinks this is Augustine’s view; in fact, he takes Augustine to assert in his thought experiment of the two men that the fall of the man who lusts for the woman is caused by his will (Ord. II.34–7 n. 75, VIII: 397).

34

For a more detailed discussion of Scotus’s theory of the first cause of sin, see Hoffmann forthcoming a.

The Will as the Cause of Evil

191

To show how the will is the cause of evil, Scotus discusses three approaches that he finds insufficient by themselves, but correct when integrated into a larger theory. The first two are Richard’s ways in which evil is caused accidentally – either insofar as what is accidental regards the cause (the will happens to be defectible) or the effect (depravity happens to be conjoined to the intended effect). According to the third approach, sin does not have an efficient, but a deficient cause. The view Scotus describes in the third approach resembles in part Aquinas’s; Scotus might intend either him or, more likely, someone influenced by him.35 Scotus formulates several objections against these three approaches, principally regarding the danger of tracing sin not only to the will, but further to God. If the will insofar as it is defectible is the cause of sin, then so too is God, who made the defectible will. If the will causes evil insofar as it intends an act to which depravity is attached, then so too does God, for as the primary cause he wills the act in which depravity is implied. If the will as a deficient cause is the cause of sin, then so also is God: Scotus understands the deficient cause as a cause that fails to be efficient. (Scotus refers to Aristotle’s example in Met. V.2, 1013b11–16 of the pilot whose absence causes shipwreck.) Thus God would be the deficient cause of sin insofar as he does not cause (by primary causality) the rectitude of the sinful act.36 The hypothesis of a nonculpable nonconsideration of the rule of action, which Scotus mentions only in the Paris lectures as part of the third view, is briefly dismissed: he objects that since consideration is obligatory while doing an act that needs regulation, nonconsideration would itself be a sin, indeed the first sin.37 Since the three approaches bear upon the relationship between the causality of God and that of sinning creatures, Scotus addresses God’s causality in sin in order to solve the question of the first cause of sin. For Scotus, God as primary cause and the created will cause together the volitions of the created will. The lack of rectitude in acts of sin, however, is not to be attributed to God, but rather to the created will. God causes concurrently the rectitude of our acts if our will causes the rectitude; but if 35

36 37

The resemblance with Aquinas is particularly clear in the Paris lectures, where the nonculpable nonuse of the moral rule is mentioned (Rep. II-A.34, Viv. XXIII, 174a–b). But the language – the will as deficient cause – does not fit with Aquinas; if Scotus intended Aquinas, he formulates his view freely. A better match is John Quidort, In Sent. II q. 87c., II: 211 (dated between 1292 and 1296, see vol. I: XXX). Lect. II.34–7 nn. 73–97, XIX: 343–50, Ord. II.34–7 nn. 76–94, VIII: 398–407; Rep. II-A.34 nn. 9–13, Viv. XXIII: 173a–175a. Rep. II-A.34 nn. 12–13, Viv. XXIII: 174b. His observation thus converges with Barnwell’s; see note 28 above.

192

Whence Evil?

our will does not cause it, then neither does God, for our will poses an impediment to his action.38 Scotus does not tell us what it means for an act to be caused with or without its rectitude. (Fornication, for example, lacks rectitude in his view; but rectitude cannot be simply added to the act to make it right, it would seem.) Scotus then combines the three approaches in a way that avoids tracing sin to God. He starts with the third approach, which is at the heart of his solution: From these three approaches (that is, from the two that posit an accidental cause of evil, and the third, which posits a deficient cause of evil) one integral solution can be gathered, which is this. In a sin, a positive act, as what is material, and the privation of the requisite justice [iustitiae debitae], as what is formal, come together [concurrunt]. With respect to this privation, there is no efficient, but only a deficient cause, in accordance with the third approach, for the will, which is obliged to give rectitude to its act but fails to do so, sins by being deficient. This “being deficient” (that is, not causing or giving the requisite rectitude to its act) comes from a cause that could freely cause in that moment, that is, that could freely give rectitude to its act. Accordingly, formally speaking, this is what it is to sin: for this free cause not to bestow the requisite rectitude which it could bestow in that moment. (Ord. II.34–7 n. 125, VIII: 420–1)

In short, Scotus holds that “the deficient will is the deficient cause of sin, but differently from the divine will, or rather, the deficient will is the cause and the divine will is not the cause” (Ord. II.34–7 n. 129, VIII: 423). Scotus conceives of the will as a deficient efficient cause: it does not cause the requisite effect when it should. Furthermore, as the above quotation indicates, it is crucial that the will causes evil freely, being able to do otherwise.39 Scotus also adopts what he takes to be correct in the first and second approaches. Concerning the first approach, he admits that “defectibility” is in a sense accidental to the will, insofar as it is not included in the will taken as a “pure perfection” (perfectio simpliciter), that is, as a notion that is common to God and creatures.40 The second approach had emphasized 38 39 40

Ord. II.34–7 nn. 142–5, VIII: 428–30; Lect. II.34–7 nn. 150–4, XIX: 365–6. For a detailed discussion of Scotus’s theory of divine primary causality in sin, see Frost 2010, 24–8. See also Lect. II.34–7 n. 72, XIX: 343; Rep. II-A.34 n. 11, Viv. XXIII: 174a; Ord. II.44 nn. 6–7, VIII: 492. Ord. II.34–7 nn. 127–8, VIII: 421–2. The Vatican Edition has misleading punctuation and makes an unneeded conjecture in n. 127; for a better text reconstruction, see Hoffmann forthcoming a, note 67. Cf. Lect. II.34–7 n. 98, XIX: 350–1, Rep. II-A.34 n. 13, pp. 174b–175a. For the notion of pure perfection, see Chapter 5, note 38.

The Will as the Cause of Evil

193

that sin does have an efficient cause, albeit one that intends the evil effect only incidentally. Scotus agrees, but insists that, strictly speaking, the accidental cause causes evil not by efficient but rather by deficient causality. What is caused by efficient causality is a thing that is good (aliquod positivum) to which a deficiency (defectus) is conjoined; this deficiency itself results from deficient, not efficient, causality (Ord. II.34–7 n. 126, VIII: 421). The deficiency of the act is not entirely unintended, however, because by wanting the act to which the depravity (deformitas) is conjoined, one also accidentally wants the depravity of the act (Lect. II.34–7 n. 99, XIX: 351). Scotus does not address the problem expressed in Augustine’s dilemma, of how the will can cause evil without being in some way disordered. In fact, for Scotus, sin does not presuppose any deficiency at all, neither in the will, nor – as he mentions in passing – in the intellect (Ord. II.34–7 n. 142, VIII: 428). Scotus’s view that an entirely unflawed will can cause evil coheres with his stress of the created will’s radical contingency, allowing it in principle to sin even in the beatific vision.

7.9 John of Pouilly Unsurprisingly, Henry of Ghent, Richard of Menneville, and Duns Scotus traced the first evil act to the will’s freedom and no further. Even Aquinas, hardly a voluntarist, argues that the nonuse of the rule of action, presupposed in the will causing evil, need not be traced any further than to the will’s freedom to act or not. How do strict intellectualists solve the problem of the first cause of an evil will? For Godfrey of Fontaines, all we have to go by is his brief discussion of angelic sin, to be considered in Chapter 8. But John of Pouilly addresses this problem briefly in its own right in his Quodlibet IV, Question 6 (of Advent 1309 or Lent 1310). This Question consists largely of a critique of Henry’s Quodlibet I.16, and in this context Pouilly also rejects Henry’s account of the cause of evil, especially the conclusion that the will can turn to evil by itself, and that it does so by acting against reason. Pouilly argues that the simplicity of Henry’s solution is itself a sign that it is wrong: the problem of the cause of an evil will, which Augustine discussed with such difficulty, and the problem of the sins of the first angel and of Adam, which theologians consider to be extremely difficult, would be easy to solve if one could simply say that the will can cause its own evil act by willing against the judgment of reason (P 120rb, N 91rb).

194

Whence Evil?

Pouilly accepts the idea of Augustine’s thought experiment that two men who are presently equally conditioned and placed in the same circumstances can have different futures (P 120va, N 91va). But his story of how their unequal volitions come about differs from Henry’s. The evil will is to be traced to reason and will, not to the will alone, for we control our acts by free decision, which includes both. What causes the two men to act differently is negligence in one and diligence in the other: one considers carefully whether the object is worth desiring and what are the consequences of desiring it; the other does not, fails to draw the right conclusions from moral principles, and falls (P 120ra–rb, N 91ra). Through intellect and will, the objects that cause us to err (about what is worth willing) and to have an evil will can be disapplied (P 120rb, N 91rb) – that is, the attention can be turned away from them toward something else (cf. Chapter 5, p. 140). This is what we should do especially when tempted by inordinate sex (P 120rb, N 91ra–rb). Pouilly concludes: I do not deny that the will is itself the cause of its evil act . . . but not in the way they say, namely that the will causes its evil act by willing contrary to the judgment of reason, but rather because it is in the power of the human being, by intellect and will, to be diligent, by applying good objects and disapplying bad ones and by reasoning diligently toward the good conclusion that is to be pursued and the bad one that is to be avoided . . . or also to be negligent. Through this negligence, first error is caused in reason and then evil in the will, insofar as, because of such negligence, the objects causing this [error and evil] were not disapplied. . . . (Quodl. IV.6c., P 120rb–va, N 91rb–va)

So Pouilly traces our control of acting well or badly to diligence or negligence. He then asks what is the cause of negligence or diligence. He answers that negligence has no per se cause, but there can be infinite things causing it (per accidens, it is implied), such as not wanting to make the effort of thinking carefully, or feeling secure and thinking one is not easily deceived (P 120va, N 91va).41 In Pouilly’s theory of the cause of moral evil, negligence has an antecedent role with respect to an evil will, similar to the role of nonconsideration in Aquinas’s theory. But Pouilly does not clarify its moral status. If he thinks negligence is itself a sin, then the origin of this sin must be explained.

41

For the idea that negligence can have infinite accidental causes, Pouilly depends on Godfrey of Fontaines, Quodl. XV.4c., PhB XIV: 30–1. For the lack of a per se cause of negligence, see also Godfrey, Quodl. V.12c., PhB III: 58; Quodl. X.13c., PhB IV: 377.

The Will as the Cause of Evil

195

Furthermore, the negligence that causes an evil will consists in the failure to deliberate carefully. So we may wonder how we control whether we are negligent or not, in other words, whether we deliberate or not – if not directly by our will. This difficulty in explaining our control of deliberation applies to Pouilly’s theory of free will as a whole; as we have seen, he roots all control of our acts in deliberation, but he cannot tell us how we control whether or not we deliberate, let alone how we deliberate (see Chapter 5, pp. 142–3).

***

Augustine, Bonaventure, Aquinas, Scotus, and Pouilly represent five main approaches to explaining the first cause of sin. According to Augustine, the first evil has no cause, for as first evil, it cannot be caused by an evil cause, nor can any evil be caused by a good cause. According to Bonaventure, who is followed by Henry and Richard, the will causes evil insofar as it is naturally defectible, and hence it is not entirely unflawed. Aquinas develops systematically the idea already found in William of Auxerre and Albert the Great that evil is caused per accidens. He adds that the first cause of moral evil is a will that is deficient without being culpably deficient. Scotus holds that what allows the will to cause evil is simply its freedom for opposites, regardless of other cognitive or volitional failures. Pouilly’s refusal to root the first evil in the will brings him closer to Augustine and distinguishes him from the other thinkers here considered. But Pouilly thinks like Aquinas that evil is in the last analysis caused per accidens; in fact negligence, the cause of sin, does not have a per se cause. As much as these medieval theories of the cause of evil differ, they converge where we might least expect it: on Augustine’s idea that the cause of evil is deficient, that is, that there is no ultimate explanation for why someone does evil. Albert and Bonaventure argue that the will’s free preference for one alternative rather than another is not in turn caused by a previous efficient cause. Aquinas holds that the nonconsideration of the rule has no cause and is sufficiently explained by the will’s freedom. Scotus makes a general argument that the will’s volitions are not fully explainable (see Chapter 5, p. 125). Pouilly considers negligence, the antecedent to sin, to be at bottom unexplainable. So there is a consensus that moral evil cannot be traced entirely to antecedent causes but rather must be traced no further than to the person. Moral evil, then, escapes full intelligibility. As an intentional act, an evil choice is done for a reason, and yet it remains ultimately unexplainable.

196

Whence Evil?

We now turn to angelic sin. The question of how angelic sin is possible is tightly linked with the problem of the first cause of moral evil. But the perspective is shifted toward the question of how a person in optimal cognitive conditions can do evil. The problem of demonic obstinacy will allow us to take a new perspective on free agency, for the issue is what can permanently block a free person from doing good actions and from having alternative possibilities.

part iii

Angelic Sin

chapter 8

Intellectualist Accounts of the Angelic Fall

For the study of the moral psychology of free will, explanations of the fall of the angels are no less important than those of the first cause of evil. When it comes to the angels, the question is not only how a good person can do evil for the first time, but also how someone can do so while acting under ideal psychological conditions. Angels are thought to be supremely intelligent; as pure spirits they cannot experience any conflict between rational desire and sensory desire, nor can sensory desire interfere with their practical cognition; as persons whose crucial decision occurred shortly after coming into existence, their decision does not depend on good or bad character formation. In their accounts of angelic sin, medieval thinkers are pressed in a particular way to clarify what motivates rational agents to make evil choices, whether their sin can be a case of clear-eyed evildoing, and how they control their acts. Certain difficulties in explaining the fall of the angels are particularly challenging for intellectualists, who are considered in this chapter. One difficulty is this: Infallibility problem: As purely intellectual natures, angels seem unable to experience any cognitive deficiency; so if a sinful act presupposes a cognitive deficiency, then angels cannot sin.

In other words, angels cannot sin because they know better. Even if their cognition could be deficient, another difficulty is implied in the Socratic deficiency thesis, according to which a volitional deficiency presupposes a cognitive deficiency. But to cause sin – an evil act for which one is responsible – the cognitive deficiency must itself be culpable, for blameless ignorance excuses from sin rather than causing it. This implies a dilemma that is the cognitive counterpart to Augustine’s dilemma: Blameless ignorance dilemma: Blameless ignorance excuses from sin, and culpable ignorance cannot cause the first sin. 199

Angelic Sin

200

While it is a problem for any person’s first sin, applied to the angels, the problem is this: Blameless ignorance problem: Angels cannot have any culpable ignorance before their first sin; so if a sinful act presupposes a culpable cognitive deficiency, then angels cannot sin.

In short, angels cannot sin, because even if they do what they should not, they cannot be expected to know better. A third difficulty is implied in the assumption that the cognized object causes the will’s act. Some medieval theologians assume that the angels obtained all naturally attainable knowledge along with their creation, apart from knowledge of future contingents, which they obtain when the future becomes present. So the increase of knowledge does not depend on the angels, and hence it is not up to them what they know. If angels do not control their knowledge, then they do not control whether a cognized object causes their act. Assuming that sin presupposes the ability to avoid sin, lack of control entails impeccability: Control problem: Angels cannot know otherwise; so if their actions follow their knowledge, they cannot act otherwise; and since sin presupposes the ability to act otherwise, angels cannot sin.

In short, angels cannot sin, because even if they do what they should not, it was not up to them to know better. We begin this chapter with Philip the Chancellor, probably the first to formulate the infallibility problem on the authority of Aristotle. We then turn to Albert the Great, whose account resembles Philip’s but is more elaborate. Next, we will consider Aquinas, in whose innovative and resourceful treatment the complexity of angelic sin comes into clear light. Last, we will examine Godfrey of Fontaines, who given his theoretical presuppositions struggles to explain angelic sin, and John of Pouilly, who proposes without any hesitation a strictly intellectualist account of angelic sin.

8.1

Philip the Chancellor

Philip the Chancellor introduces into the discussion a popular objection based upon the authority of Aristotle’s dictum that intellectus is always right (cf. An. III.10, 433a26). The word intellectus can mean the power (intellect) or the act (understanding). Philip takes it to mean the power and formulates a version of the infallibility problem: since the angelic appetite is moved only by the intellect (and not, it is implied, by imagination), and

Intellectualist Accounts of the Angelic Fall

201

since the intellect is always right, angels do not have free decision insofar as it involves “flexibility” toward good or evil (Summa de bono, I: 93 ll. 1–6). Philip counters that angels must have free decision not only to do good, but also to do evil, for without it, they could not merit. Hence angels can sin. This argument presupposes a symmetrical account of merit, which he professes explicitly further down (I: 201 ll. 35–7): not only does demerit for a bad act require the ability to do good instead, but merit for a good act implies the ability to do evil instead. He points out that unlike angels, humans have two motivational sources, imagination (phantasia, the sensory power of representing an object even in its absence) and practical intelligence (I: 94 ll. 37–45). So, it is implied, humans can sin by following their imagination rather than the judgment of their practical intelligence. Further below, in discussing human sin, Philip explains that imagination presents what is good in the present moment (bonum ut nunc) and intellect presents what is good as such (bonum simpliciter). In support of this distinction he cites the De anima passage more fully: the intellect is always right, while imagination can be right or not (433a26–7). He explains that the intellect can err only by turning to what is mixed with imagination (I: 161 ll. 60–5). When the human will follows the imagination rather than the intellect, it chooses an apparent good, which is not always the true good (I: 166 ll. 48–51). By contrast, although angels do not possess two motivational sources, they have alternative possibilities because they can turn to the unchangeable good (God) or to the changeable good (creatures), in which latter case they sin (I: 94 ll. 45–9, cf. I: 164 ll. 146–8). To the objection concerning the intellect’s infallibility, Philip responds that the intellect is always right not as such, but only as to its “highest part,” to which belongs synderesis (the infallible moral awareness) (I: 94 ll. 50–4). When Philip is writing, there is not yet a pronounced difference between intellectualist and voluntarist approaches to free will. But Philip’s account of angelic sin has an intellectualist dimension because he recognizes the angels’ infallibility as a threat to their free will and solves the infallibility not by granting that angels act against a correct judgment, but rather by allowing for error in the lower part of the angelic intellect.

8.2 Albert the Great In Albert the Great, who follows the basic lines of Philip’s solution, the problem of how angelic sin is possible takes on greater importance than in Philip. His first of several discussions of how angelic sin is possible is in De IV coaequaevis, within a Question about the cause of the sin of the angels. A few years later, he pioneers what was to become standard practice: he dedicates

202

Angelic Sin

a full Question to the possibility of angelic sin, first in his Sentences commentary and then in his lectures on the Divine Names. In his Summa theologiae, whose treatment of angelic sin follows closely that of De IV coaequaevis, he returns once more to this issue within a Question about the cause of angelic sin.1 For the idea that the angels have a superior and infallible intellect, Albert uses, like his contemporaries, as an important authority Pseudo-Dionysius’s influential notion that the angelic intellect is godlike (deiformis).2 But the formulation of the infallibility problem that he emphasizes is the one based on Aristotle’s authority, which was introduced by Philip. Albert puts it in the Sentences commentary as follows: Aristotle says that the intellect is always right while the imagination is either right or not; in the angels, only the intellect moves the appetite; a sin cannot be elicited by what moves the appetite rightly; hence the angels cannot sin.3 In his responses, Albert follows Philip in denying that the intellect is entirely infallible.4 His line of thought becomes clearer in light of a passage from De homine, where he asks whether the practical intellect is always right.5 In his answer, he clarifies that “practical intellect” can mean three things: first, the rational part of the soul as such; second, that which is in potentiality to all matters of action (just as the theoretical intellect is in potentiality to all things); and third, that which grasps practical principles. In the first sense, as the rational part of the soul as such, the practical intellect is not infallible, as is evinced by the fact that the angelic intellect could turn to evil without being conjoined to imagination. Nor is the practical intellect in the second sense infallible, for matters of action involve mainly particulars, and regarding particulars, one can easily err. Only in the third sense, which according to Albert is the sense intended by Aristotle, is the intellect infallible: it cannot err regarding the universal principles of right (universalia iuris) (De homine, XXVII/2: 483 l. 51 – 484 l. 6). The first practical principles or universal principles of right are, for example, that one must not kill, 1

2 3 4 5

De IV coaequaevis exists in two redactions. The first, which is unedited, was written before 1241; the second, on which my exposition is based, is contemporaneous to In Sent. II (1246/47); see Rigo 2005, 347–58. For the dating of In DDN (1250) and of Summa theologiae (after 1268), see Chapter 7, notes 13 and 18, respectively. De IV coaequaevis IV.63.3 arg. 2, Borgnet XXXIV: 674b; In Sent. II.5.1 arg. 2, Borgnet XXVII: 110b; ST II.5.21.3 arg. 2, Borgnet XXXII: 262b–263a; cf. Pseudo-Dionysius, DDN VII.2, 869C. In Sent. II.5.1 arg. 1, Borgnet XXVII: 110b; see also De IV coaequaevis IV.63.3 arg. 1, Borgnet XXXIV: 674b; In DDN IV n. 149 arg. 1, XXXVII/1: 236a; ST II.5.21.3 arg. 1, Borgnet XXXII: 262b. De IV coaequaevis IV.63.3 ad 1, p. 675b; In Sent. II.5.1 sol., p. 111a; In DDN IV n. 150 ad 1, p. 237a; ST II.5.21.3 ad 1, pp. 263b–264a. De homine was written after De IV coaequaevis (first redaction) and before the Sentences commentary; cf. Chapter 2, note 10.

Intellectualist Accounts of the Angelic Fall

203

one must not commit adultery, one must observe justice, one must venerate God, one must honor one’s parents (In De anima III.4.6, VII/1: 235 ll. 30–5). In his discussions of angelic sin, Albert repeats the idea expressed in the De homine that the intellect is fallible regarding knowledge of particulars. In his Sentences commentary, the explanation of the angelic intellect’s fallibility resembles Philip’s explanation of human sin: the angelic practical intellect is either concerned with the good as such (bonum simpliciter) or with the angels’ own good, which is a “good as to the present moment.” Regarding the good as such, the angelic intellect can only be right, for then it proceeds from the principles of the truth as such and of the good as such (by which Albert seems to intend the first, self-evident, theoretical and practical principles). Regarding the angels’ own good, however, their practical intellect can be deceived (In Sent. II.5.1 sol., p. 111a). In fact, as Albert explains in his De anima commentary, while the “good as such” (bonum simpliciter) is good always and everywhere, this is not the case with the “good as to the present moment.” He cites the famous example that returning borrowed goods to their owner is good, but if the good is a sword and if the owner has gone mad, doing so is bad (In De anima III.4.6, p. 235 ll. 54–75). In the commentary on the Divine Names, Albert explains that deception could happen through mistaken application of universal knowledge to particulars, or by mistakenly relating one particular to another (In DDN IV n. 150 sol. and ad 1, XXXVII/1: 236–7). What here is formulated abstractly corresponds to a hypothesis by Anselm that Albert mentions in De IV coaequaevis: Lucifer knew that his sin deserved punishment, but – applying falsely this knowledge to his personal case – he thought that God would be forbearing (IV.63.3 ad 1 and ad 3, pp. 675b–676b; cf. Anselm, DCD 23). What is important in all these explanations is that Albert takes the infallibility problem very seriously. He adopts the intellectualist view that the will follows cognition, and hence that deficient willing presupposes deficient cognition. His effort therefore consists in showing that the angelic intellect is not entirely infallible. Nevertheless, in his Sentences commentary, he adds to his standard solution an additional explanation: the angels can sin because their will need not follow their cognition (In Sent. II.5.1 s.c. 2 and ad 4, p. 111a–b). But such a voluntarist explanation is absent from his slightly later account in the commentary on the Divine Names, where he argues that since volitions follow from knowledge, the abandonment of the desire of the good follows from mistaken thinking.6 6

In DDN IV n. 149 arg. in opp., p. 236 ll. 68–74, endorsed on p. 237 l. 49.

Angelic Sin

204

Having offered a solution to the infallibility problem, the next task would be to explain how the angels could have avoided the error that led to their sin, and whether failure to avoid it is itself a sin. But Albert does not provide this explanation, and so he leaves the blameless ignorance problem unaddressed.

8.3

Thomas Aquinas

In treating the various problems implied in the narrative of angelic sin, Aquinas raises the discussion to a higher philosophical level than his immediate predecessors. Compared to them, Aquinas’s theory of angelic psychology poses a greater challenge to explaining angelic sin, and he makes greater use of Aristotle’s action theory. By contrast, he pays almost no attention to Anselm’s De casu diaboli, which was, however, an important source for his contemporaries. We begin with Aquinas’s account of angelic psychology, which sets narrow boundaries for his solutions to the philosophical problems implied in angelic sin.7 Angelic Psychology Aquinas brings out in greater detail than his predecessors the optimal psychological makeup of angels. He holds, of course, that no angel is evil by nature.8 In addition, he argues that the angels cannot have any natural inclination to evil. As he explains, every nature is good, and so nothing can have a natural inclination to evil, except incidentally, insofar as it is inclined to a particular good that happens to be contrary to some other good. Only things that are composed of two natures can have such conflicting goods. In human beings, sensory pleasure, a particular good, can be contrary to the good of reason, which is the good simply speaking (bonum simpliciter); thus in humans, there is a natural inclination to the sensory good that can be contrary to the good of reason (or of the intellect). But as purely intellectual beings, angels can only be inclined to the good of the intellect, and thus to the good simply speaking. Hence they cannot be naturally inclined to evil.9 7

8 9

On Aquinas’s account of angelic sin in general, see Maritain 1959, Bonino 2016 ch. 10 and the literature he indicates on p. 197, note 25, and D’Ercole 2017, 75–104; for the control problem in Aquinas’s account of angelic sin, see Hoffmann 2007 and Peck 2014. SCG III.107; In Ioannem 8.6 n. 1246; In DDN IV.19 nn. 529 and 535; ST I.63.4; DM 16.2c., XXIII: 288 ll. 191–9; DSS 17, XL: D70–1; DSS 20, pp. D76–7 ll. 27–64. DM 16.2c., p. 288 ll. 200–57; DSS 20, p. D77 ll. 65–94; see also ST I.63.4.

Intellectualist Accounts of the Angelic Fall

205

Aquinas thinks that angels cannot desire evil under the aspect of the good, for they cannot have erroneous beliefs. Angelic infallibility follows from certain assumptions of Aquinas’s theory of angelic knowledge. He holds that angels, like human beings, know by means of “intelligible species” that inform their intellect, that is, by means of likenesses of essences. Unlike humans, angels do not acquire intelligible species over time, but possess all of them from their creation (ST I.55.1–3). Based in part on the authority of Pseudo-Dionysius (DDN VII.2), Aquinas holds that angels know nondiscursively, that is, they do not gain knowledge of things unknown through things already known, as humans do through syllogistic reasoning (ST I.58.3). Angels cannot err, for they cannot falsely affirm or deny a predicate of a subject, because they know simultaneously and nondiscursively all that can be affirmed or denied of the essences they know through the intelligible species (ST I.58.4–5) – except for future contingents (DM 16.7 ad 6). Thus they have perfect knowledge of everything that falls within the natural scope of their intellect (DM 16.6c., p. 310 ll. 233–44). Nevertheless, the knowledge of higher angels extends further than that of lower angels, because they possess more universal intelligible species (ST I.55.3; DM 16.4c., p. 299 ll. 319–30). Yet no angel is omniscient. Although they know by nature that God exists,10 at first, they did not enjoy the beatific vision (ST I.62.1), and so they did not know any divine mysteries unless they were revealed to them (ST I.57.5). Only if they presumptuously judge about things that fall outside of the scope of their natural knowledge can they err. For example, they can believe that a dead man (that is, Christ) will not rise from the dead, or that Christ is not God. Such presumptuous judgments cannot come about without an evil will (ST I.58.5c.; DM 16.6c., p. 310 ll. 255–91). Despite the angelic intellect’s perfection, Aquinas thinks that it is limited in that it cannot simultaneously consider everything it knows habitually, just as we cannot simultaneously think about all we know. Aquinas professes a theory of undivided thinking: what an angel is actually thinking about at a given moment is restricted to what is knowable through a single intelligible species (In Sent. II.3.3.4; ST I.58.2).11 As he explains, this restriction is due to the fact that the intelligible species is the “form” of the intellect, that is, its cognitive configuration, and nothing can be perfected by several forms, just as a body cannot be configured by different 10 11

ST I.56.3; DM 16.3c., p. 293 ll. 174–80; DM 16.4 ad 14. Dubouclez 2014 gives a lucid explanation of Aquinas’s theory of undivided thinking and documents that it generated controversy through the fourteenth to seventeenth centuries and became widely rejected.

206

Angelic Sin

configurations. Thus, angels can think of different things at a time, but only as many as are knowable through a single species – which, however, is no small number.12 Their habitual knowledge extends to everything that is naturally knowable to them; their occurrent consideration does not (DM 16.6 ad 4). Given the optimal conditions of the angelic intellect and will, it is difficult to see how angels could sin. Philip and Albert had made concessions to angelic infallibility prior to sin; Aquinas does not. He addresses the possibility of angelic sin both in general, arguing that all rational creatures are in principle able to sin, and in particular, showing how the angels could sin despite their infallibility concerning naturally knowable truths. The Peccability of Rational Creatures Augustine traced the rational creatures’ ability to sin to their being created from nothing (ciu. XII.6). Aquinas adds to this explanation that for rational creatures being created implies being subject to God’s rule and measure. Nonhuman animals and God are not under a higher rule, and hence they cannot have an evil appetite. The good of an animal is to follow its sensory appetites, whereas God is his own rule and hence cannot sin, just as a person cutting wood could not fail if his or her own hand were the rule for correct cutting.13 Aquinas consistently applies his theory of the natural peccability of rational creatures to the case of angels.14 Aquinas’s Early Treatment of the Infallibility Problem Aquinas discusses the problem of how angelic sin is possible repeatedly throughout his career.15 By examining his early treatment in the Sentences commentary and his late treatment in the Prima pars of the Summa theologiae and the De malo, we can observe the development of his theory and see how it relates to discussions before and after him. 12 13

14

15

ST I.58.4c. and ad 1; DM 16.4c., p. 299 ll. 335–9. On Aquinas’s theory of angelic knowledge, see for more details Suarez-Nani 2002b, 17–75, and Goris 2012. DM 1.3 ad 9; DM 16.2c., p. 289 ll. 277–87. Aquinas here implicitly rejects Aristotle’s theory that the virtuous person embodies the “rule and measure” of what is truly good (EN III.4, 1113a28–33). For earlier, different explanations, see In Sent. II.23.1.1 and DV 24.7. SCG III.109; ST I.63.1c.; DM 16.2c., p. 289 ll. 300–5. Aquinas’s theory of the peccability of angels generated a lively debate in the mid-twentieth century because of its implications for the relation between nature and grace. The issue is whether angels are capable of sinning only if they are destined by grace toward a supernatural final end. See Marieb 1964 and the literature he cites. In Sent. II.5.1.1; SCG III.108–10; ST I.63.1; DM 16.2; DSS 20 (unfinished).

Intellectualist Accounts of the Angelic Fall

207

The basic narrative Aquinas presents in the Sentences commentary remains substantially the same in later treatments, whereas his thought about how angels could sin undergoes some development. Aquinas holds that all angels are in a state of happiness from their creation, which, however, is not the highest happiness of which they are capable (In Sent. II.4.1.1). What Lucifer, the highest angel (In Sent. II.6.1.1), wrongly desired was his preeminence in happiness and goodness compared to all lower creatures, and he wanted to have his ultimate perfection, complete happiness, by his natural means alone rather than through grace. So he desired in a way to be like God, for to have complete happiness by nature rather than by grace is unique to God (In Sent. II.5.1.2c.).16 Lucifer’s disordered desire for his excellence, triggered by the consideration of his own beauty, is pride (In Sent. II.5.1.3c.). Lucifer also wanted to be the cause of the ultimate perfection of the lower angels, and when he expressed this desire to them, some consented (In Sent. II.6.1.2). Regarding the problem of how angels can sin, Aquinas formulates an opening argument that repeats the objection discussed by Philip and Albert: angels cannot sin because their appetite follows their intellectual cognition, which cannot be false (In Sent. II.5.1.1 arg. 4). In the solution, Aquinas states that “there cannot be sin in the will, unless there is somehow deception in reason, hence ‘every evildoer is somehow ignorant’,” quoting the famous Aristotelian adage (EN III.1, 1110b28–9). His solution builds upon Aristotle’s explanation of how incontinence is possible. Recall that Aristotle distinguished between merely having knowledge and having and exercising knowledge, and furthermore between knowing universal propositions (dry food is good) and knowing particular propositions (this is dry food). According to Aristotle, because of passion, the incontinent do not apply their habitual knowledge to the case at hand (see Chapter 1, pp. 25–6). Aquinas mentions the two distinctions and puts the brunt of the weight on the first one: someone can have habitually (in habitu) the correct assessment, even concerning a particular matter of action, but be impeded from actually considering it, because passion “fetters” the intellect. Angels lack passions, but their intellect can nevertheless be fettered. Based on his theory of undivided thinking, Aquinas argues that by considering one thing, an angel is drawn away from considering other things. The angels can sin by considering an aspect according to which something is choiceworthy, although it is not to be chosen, all things considered. Thus they commit an “error in choosing” (error electionis) (In Sent. II.5.1.1c.). In 16

Regarding Aquinas’s theory of the object of Lucifer’s desire, see Pini 2013, 66–71.

208

Angelic Sin

response to the opening argument, Aquinas clarifies how such nonconsideration can lead to a deficiency in the intellect without causing it to err: a doctor can fail by correctly judging something to be beneficial to a sick person insofar as a certain sickness is concerned, which, however, is not beneficial to the person in view of another sickness the doctor does not consider. Just so, the angelic intellect can be deficient by failing to consider something, although it cannot err by judging something to be the case that is not the case (In Sent. II.5.1.1 ad 4). An “error in choosing” need not be caused by a false intellectual judgment, but rather may be caused simply by the failure to consider properly what was to be chosen (In Sent. II.39.3.2 ad 5). Aquinas’s Mature Treatment of the Infallibility Problem Aquinas’s mature account of the fall, in the Summa theologiae and the De malo, uses again the idea that the angels sinned not because of a mistaken judgment but rather because of some lack of consideration. Aquinas now clarifies what it was that the angels overlooked: the divine rule by which their will was to be regulated. In the Summa theologiae, he gives this explanation in response to an opening argument that formulates a major difficulty concerning angelic sin: the angels could not sin by wanting evil as such, for evil under the aspect of evil is beyond the will’s scope; nor could they sin by wanting evil under the aspect of the good, because mistaking an evil for a good is impossible for their intellect, which – at least before their first sin – could not err (ST I.63.1 arg. 4). So they could only sin by wanting a true good. In the response, the idea of nonconsideration of the rule has a double function: concerning moral theory, it explains why the desire for something truly good can be bad, and concerning moral psychology, it responds to the requirement of some cognitive deficiency for an evil choice. Sin in an act of free decision can happen in two ways. In one way in that something evil is chosen, as when a man sins in choosing adultery, which as such is evil. And such sin proceeds always from some ignorance or error; otherwise that which is evil would not be chosen as a good. . . . In another way it happens that one sins by free decision by choosing something good in itself, but not according to a proper measure or rule. In this case, the defect that induces sin is only on the part of the choice that is not properly regulated, not on the part of the thing chosen; as if one were to pray, but without heeding the order established by the Church. Such a sin does not presuppose ignorance, but merely lack of consideration of the things that ought to be considered. This is the way the angel sinned: by seeking his own

Intellectualist Accounts of the Angelic Fall

209

good, by his own free decision, without ordering [the act] according to the rule of the divine will. (ST I.63.1 ad 4)

As in his earlier discussions, Aquinas holds that the good the sinning angels desired was complete happiness, which they intended to attain by their natural power. According to the divine rule, however, complete happiness was to be obtained as a gift freely bestowed on them by God’s grace.17 Aquinas identifies the nonobservance of the higher rule with the sin of pride (ST I.63.2c.). For our purposes, what is most important is the psychological function of nonconsideration of the rule in angelic sin. In the above quote, Aquinas draws the parallel between an adulterer’s ignorance and the evil angels’ nonconsideration, implying that the Socratic deficiency thesis applies to both cases. He suggests that, just as the adulterer would not have sinned if he or she had known better, so also the angels would not have sinned if they had attended to the divine rule. In the De malo, Aquinas repeats essentially the same solution. The angels failed to cognize the rule and what it ordered, which is a presupposition for sin because “in a sin, the deficiency of the intellect or reason and of the will are always proportionally tied in with another.” Thus sin followed in the angels’ will upon insufficient consideration of the order of the divine governance (DM 16.2 ad 4 and ad 5). Why could the angels not have sinned in full consideration of the rule, contemptuously dismissing it? Aquinas does not directly answer this question, but he provides some hints. By sinning, Lucifer “withdrew from the order of divine justice, and so he abandoned what was better [deseruit meliora], for the rule of divine justice is better than the rule of the angelic will” (DM 16.3 ad 10). Hence, we may assume, if Lucifer had considered the rule, he could have recognized that its ordinance was better than his own plan. Perhaps by considering the rule, he could have concluded – by understanding or by faith – that by observing it, he only had to gain. In fact, according to Aquinas, had Lucifer not sinned, he would have attained precisely what he inordinately desired; Aquinas supports this idea with an oft-cited claim by Anselm of Canterbury: “he desired what he would have obtained if he had persevered [stetisset]” (ST I.63.3c.; cf. Anselm, DCD 6, I: 243–4). If Lucifer had recognized what was in his better interest, he would have acted accordingly. Aquinas thinks that not even the already fallen angels act intentionally against their own interest (perverse as it may be); for example, if they had realized the effects of Christ’s crucifixion for 17

ST I.63.3; DM 16.2 ad 4, p. 289 ll. 358–61; DM 16.3c., p. 294 ll. 219–29 and ad 1.

210

Angelic Sin

human salvation (which they dread), they would never have wanted his crucifixion (ST I.64.1 ad 4). That the angels could only sin while not considering the divine rule does not mean, however, that they sinned inadvertently. In Aquinas’s account, the evil angels were fully aware that they committed a “mortal sin,” that is, one that turns them away from God. Human beings can sin “venially,” that is, without turning away from God, because they do not always realize how a particular act affects their relationship with God. In human beings, such venial sins can happen even without their awareness. But angels, who do not think discursively, cannot know something without knowing all its implications; thus, when they fail to refer a desired good to God, they are aware that this failure implies turning away from God (DM 7.9c. and ad 2). Aquinas’s account of nonconsideration in angelic sin can be compared to people who damage a new technological device because they did not read the instruction manual. In the normal case, this happens not because they read the manual and intentionally disregard its instructions, nor because they are unaware of the existence of the manual, but rather because they feel confident that they can use the device without further instructions. The question, then, is what one should be blamed for: for having disregarded the manual, for the improper use of the device that follows it, or for both. Regarding the angels, the question is whether the nonconsideration of the rule is itself a sin (of negligent omission), or whether the sin is only the unregulated act, or both the act and the nonconsideration that made it possible. The issue here is the blameless ignorance dilemma: blameless nonconsideration of the rule excuses from sin; culpable nonconsideration cannot be a presupposition for the first sin. As already mentioned, this difficulty is analogous to Augustine’s dilemma: a good will cannot cause an evil will, and an evil will cannot cause the first evil will. Aquinas’s solution to Augustine’s dilemma (see Chapter 7, pp. 184–7) applies in a similar way to the blameless ignorance dilemma. Nonconsideration as such is no sin; it becomes sinful only when one does an act that is to be regulated. But even then, it is not a stand-alone sin (a “special sin,” in medieval jargon), but only a dimension of the actual sin. Aquinas does not spell out this solution in his late treatments of angelic sin, but he did formulate it in the Sentences commentary. In an opening argument it is maintained that Lucifer’s first sin consisted in the sin of omission rather than in pride, for he failed to consult a divine ordinance that would have forbidden the desiring of something beyond what was ordered. In response, Aquinas clarifies that omission is in one sense a special sin and in another sense an aspect of every sin. As a special sin, omission concerns something that was specifically

Intellectualist Accounts of the Angelic Fall

211

enjoined; as an aspect of every sin, omission is the failure to observe a relevant circumstance of the act. The latter case applies to Lucifer’s sin.18 The Control Problem The idea of nonconsideration of the rule is so important in Aquinas because he assumes a tight connection between cognition and choice. Nevertheless, the choice of the angels cannot be fully traced to their knowledge. I will first give a thumbnail sketch and then a fuller account of why this is so. Whereas the angels controlled their choice (as Aquinas assumes), they did not control their knowledge. But they controlled the use of their knowledge, and so they controlled how habitual knowledge becomes an occurrent practical judgment that is followed by the corresponding choice. We have already seen an instance in which Aquinas traces the use of knowledge to the will and no further: in De malo 1.3 he writes that the use or nonuse of the moral rule depends on the will’s freedom, and no further cause need be searched (see Chapter 7, p. 186). The case of angelic sin shows even more forcefully that there is no room for further tracing of the first choice beyond the will. In fact, before their first choice, there is no distinction between good and evil angels (In Sent. II.4.1.2c.). According to Aquinas, after creation and before their first choice, there is only one moment in the angelic existence, in which all angels were good. Immediately thereafter, in the second moment of their existence, all angels made either a good or evil choice (ST I.63.6; cf. DM 16.4c., p. 300 ll. 349–74). Thus, Aquinas locates an angel’s control of his choice in the way his will uses his knowledge. Now we must see this in greater detail. Aquinas is committed to the view that the angels could do otherwise. That the angels controlled their choice is implied in the fact that the evil choice was a sin. For Aquinas, what distinguishes moral fault from mere mishap is that the act in question is imputed to the agent. Such imputation presupposes that one control the act; as he phrases it, the act must be in one’s power (in potestate), that is, one must have mastery (dominium) over one’s act.19 Aquinas expresses the angels’ control by claiming that they sinned by their own free decision, by which they could choose to sin or not to sin.20 18 19 20

In Sent. II.5.1.3 arg. 4 and ad 4. Cf. Chapter 7, note 28. ST I.48.5c.; ST I-II.21.2c.; DM 2.2c., p. 33 ll. 136–42. In Sent. II.5.1.1 s.c. 2 and ad 1; In DDN IV.19 n. 541; ST I.63.1 ad 2 and ad 4; ST I.63.7c. and ad 3; ST I.64.2c.; DM 16.5c., pp. 304–5 ll. 215–52. Cf. DM 16.4c., p. 300 ll. 364–5. For the relation between moral responsibility and free decision in Aquinas, see Chapter 2, note 26.

212

Angelic Sin

That the angels did not control their habitual knowledge follows from the fact that they received all intelligible species belonging to their natural knowledge from their first moment of existence. This knowledge is continually updated as future contingents become actual in the present moment (DM 16.7 ad 6). Other than through such updates, angelic knowledge can increase only through divine revelation. Neither change in knowledge depends on them (except for knowledge of their own choices and foreseeable consequences of their choices). Hence, prior to making their first choice, they did not control their habitual knowledge. But ex hypothesi, they controlled this choice. Their situation is roughly analogous to God’s, who cannot know otherwise, but who can nevertheless will otherwise. Aquinas writes: “Whatever God knows, he knows by necessity; however, he does not by necessity will whatever he wills” (ST I.19.3 ad 6). So when the angels make their free choice, there is a transition from necessity (of knowing) to contingency (of willing). Based on Aquinas’s account of angelic sin and on what he says elsewhere about cognition and free decision, this transition happens with the “use” of their knowledge or the “exercise” of their intellect, through which they actually consider something they know habitually. In his mature account, Aquinas explains that the good and evil angels differed in whether they considered the rule governing their choice (e.g., ST I.63.1 ad 4); earlier, he argued that they differed in whether they considered all the conditions relevant to making a good choice (In Sent. II.5.1.1). So depending on how the angels use their knowledge, they judge what is worth choosing and act accordingly.21 Aquinas traces the way in which rational creatures use their knowledge to the will. Knowing something habitually (scire) means possessing its intelligible species without actually considering it.22 But thinking about something (intelligere, cogitare) consists in using an intelligible species, and this use depends on the will.23 How the will uses an intelligible species – whether or not it uses a certain species, and what precisely it considers by means of this species – is not predetermined by that intelligible species or by any other intelligible species. That Aquinas holds this view is implied in a consideration he makes about whether someone who has access to another person’s intelligible species – in the case he considers, an angel knowing a species in a human mind – would know what that person is 21

22 23

In another context, Aquinas describes how the transition from habitual knowledge to a particular choice is mediated by a practical judgment that depends on the use one makes of one’s knowledge; see DM 3.9 ad 7. DV 10.2c., XXII: 301–2 ll. 175–81 and ad 4; ST I.79.6c. and ad 3. ST I.57.4c.; ST I-II.57.1c.; DM 16.8c., p. 321 ll. 193–9, 215–21.

Intellectualist Accounts of the Angelic Fall

213

thinking about. Aquinas denies this, precisely because by knowing someone’s species one does not yet know how that person’s will uses the species (DM 16.8c., p. 321 ll. 200–41).24 Another doctrine of Aquinas makes clear that he does not think that what a person wills follows exclusively from what the person knows. As we have seen, in his explanation of how the will is moved (that is, actualized) by the intellect and how it is moved by itself, Aquinas writes that the will moves itself to the exercise of the act, albeit by means of making the intellect deliberate about whether to exercise its act (Chapter 2, p. 50). To avoid the regress implied if every exercise of a power required a previous deliberation, including the exercise of the activity of deliberation, Aquinas traces the will’s exercise of its act ultimately to God moving it. While tracing it to God bears its own theoretical difficulties, what is important is that Aquinas does not trace the exercise of the will’s act exclusively to cognition (DM 6c.; ST I-II 9.6). It appears, then, that according to Aquinas the angels’ choice at the second moment of their existence was not predetermined by their knowledge in the first moment of their existence. Nor does any other fact in the first moment of their existence determine their choice in the second moment. For this reason, Aquinas argues that prior to their fateful choice, the angels themselves could not know – not even conjecturally – what they would choose; hence they could not foresee whether they would fall or persevere in the good (In Sent. II.4.1.2c.). Aquinas’s account is intellectualist in its general outlook, insofar as he does not admit that angels sin by willing against an all-things-considered judgment of what is worth willing. But his account has a voluntarist dimension in that he traces the angels’ use of their habitual knowledge, especially of the divine rule that was accessible to them but not necessarily considered, to their will and no further. Thus, the control the angels had of their choice and of its antecedents is primitive. Yet the evil angels did not act without a motive. Lucifer sinned because he took pride in his superiority over others (ST I.63.7c., In Sent. II.5.1.2c.). In general, the evil angels were lured to sin by their own beauty (DM 16.2 ad 13, cf. In Sent. II.5.1.3c.). But they were not drawn by these motives necessarily; indeed, the good angels were beautiful, too, yet their beauty did not cause them to sin. While the motives explain the choice, there is no explanation for why an angel acts on this motive rather than on that. Nor is there a contrastive explanation for why Lucifer sinned rather than not, or 24

For an excellent discussion of Aquinas’s theory of mind-reading and its implications for his account of cognition, see Cory 2016. See also ST I.107.1, concerning angels reading other angels’ minds.

214

Angelic Sin

why Lucifer sinned but not Michael. Their choice can be made intelligible only by explaining how they could act in one way or the other, but not why they did act in the way they did.

8.4 Godfrey of Fontaines We will now see whether Godfrey of Fontaines and John of Pouilly manage to explain angelic sin without making any voluntarist concessions. The only discussion of angelic sin by Godfrey of Fontaines that has come down to us is a brief reply to an opening argument in his Quodlibet VI, Question 7. This Question, as we have seen in Section 4.3, contains a lengthy discussion of the impossibility of per se self-motion. Against Godfrey’s idea that self-motion is possible only where the mover and the moved are in distinct subjects, the opponent refers to angelic cognition and angelic sin as counterexamples: an angel who because of an intelligible species understands something in potentiality moves himself to understand the thing in actuality; an evil angel produces in himself an evil volition (Quodl. VI.7 arg. 1, PhB III: 149).25 The issue implied in this argument is the control problem: if the angels are to be blamed for their sin, then they cannot have undergone evil, but it must have been up to them to have an evil will or not. But since Godfrey denies that mover and moved can be in the same subject, he rejects the idea that the will of the evil angels caused their volition. Godfrey starts his reply to the opening argument with the remark that the philosophers denied any transition from potentiality to actuality in separate substances (that is, in immaterial substances: God and angels). Moral deficiency in separate substances, it is implied, is unthinkable to them. But unlike them, Godfrey takes as his guide not natural reason but Sacred Scripture, and hence he accepts the fall of the angels on the authority of faith. He does not deny the angels’ transition from potentiality to actuality; he only denies that they moved themselves from potentiality to actuality (Quodl. VI.7 ad 1, PhB III: 168). His answer builds upon Aquinas’s theory: I say that their fall does not prove that the same subject moves itself, because the devil sinned only by his spontaneous will, desiring something in a manner in which he was not supposed to desire it. Under the 25

Duns Scotus will formulate a similar objection against Godfrey; see Lect. II.25 n. 40, XIX: 241; Rep. II-A.25 n. 9, Viv. XXIII: 121a–b; Ord. I.3.3.2 nn. 430–2, III: 262–3.

Intellectualist Accounts of the Angelic Fall

215

presupposition of the defectibility of the will and nonconsideration and some sort of negligence by reason, the cognized object caused in him the act of willing. By this act he sinned, because he desired this object in a manner he was not supposed to – not because of an error or ignorance that would be temporally antecedent, but rather from a simple unawareness [nescientia] due to a deficiency of considering the manner [of proper desire] he was supposed to consider. But upon this unawareness or lack of consideration there followed an error or ignorance in the moment he consented. Although this error or ignorance was temporally simultaneous with the consent, according to nature it was prior [to consent]. (Ibid.)

Even concerning angelic sin, Godfrey maintains his conviction that a volition is caused by the cognized object – by efficient causation, as is implied and as he affirms explicitly earlier in the same Question, although not in reference to the angels’ volitions. He does not explain, however, how the angels can control what they consider and which object causes their volition; his prohibition of self-motion impedes such an explanation. In his further discussion, he insists that just as the will does not move itself to will something, so too the intellect does not move itself to understand something. Instead, it is the object that moves the intellect to its cognition and the will to its volition. As we have seen in Section 4.3, Godfrey is adamant that the principle that something cannot reduce itself from potentiality to actuality applies to all of being, and thus also to angels and the human soul. He also insists that first and most certain metaphysical principles must not be denied, even if they make it difficult to explain how to safeguard free will (Quodl. VI.7 ad 1, PhB III: 170; cf. Chapter 4, pp. 108–9). In the above block quote Godfrey furthermore affirms that nonconsideration led the angels into error. While Aquinas excluded fallibility from angels prior to sinning, Godfrey does attribute error to them, perhaps because he takes judgment-volition conformity to imply that the devil’s perverse volition presupposes the mistaken practical judgment that it is good to will in this manner. His disciple John of Pouilly will make this point explicitly. Godfrey’s discussion of angelic sin is brief but important, because it brings out the tension between the legitimate reservations against admitting per se self-motion and the attempt to explain a free choice among alternatives. What is significant in Godfrey’s solution is above all what it does not explain: how the angels control which choice they make. His position is vulnerable to the objection that it fails to avoid intellectual determinism.

Angelic Sin

216

8.5

John of Pouilly

Whereas Godfrey discusses angelic sin because he was challenged, John of Pouilly takes it up of his own accord, to reject the claim that angelic sin and the sin of the progenitors Adam and Eve are a counterproof to judgmentvolition conformity. Pouilly develops his position in response to Gonsalvus of Spain. As part of a series of arguments to show that the will can act against the judgment of reason, Gonsalvus maintains that, given the Augustinian deficiency thesis that deficient cognition presupposes deficient willing, the angels and progenitors suffered no ignorance before their sin, and so they could only have sinned by acting against their better knowledge (QD q. 8c., ed. Amorós, p. 128). Gonsalvus’s argument seems inspired by Henry of Ghent (see Chapter 3, pp. 69–70); it was in general popular among voluntarists, as we will see in Chapter 9. Against Gonsalvus’s argument, Pouilly defends the Socratic deficiency thesis in order to safeguard judgment-volition conformity. Pouilly argues that one cannot will what is unknown, and hence one cannot will something under an aspect under which it is not known. In positive terms, what one wills must be cognized and judged as to be willed under the very aspect under which one wills it. The angels and progenitors, then, no matter how they sinned, could not have sinned without making a cognitive error. If they desired an evil as an apparent good, they erred. If they desired a good, but not in the way in which it was meant to be desired, then they erred as well. Assuming that they desired their own perfection as a good in itself rather than as subordinate to a higher good, they erred because they judged that this particular good was to be desired for its own sake, when in fact it was to be desired for the sake of something else. According to Pouilly, Sacred Scripture confirms that the first sin was caused by error. When God asks Eve why she took of the forbidden fruit, she answered: “because the serpent deceived me,” rather than “because I had a bad will” (cf. Genesis 3: 13).26 Pouilly thus sides with Godfrey that sin presupposes an outright error, not merely some nonconsideration. Pouilly knows that in attributing fallibility to the angels and progenitors, he goes against the majority view. He therefore offers his own exegesis of the proof text from Augustine on which some theologians base the Augustinian deficiency thesis. The text is usually quoted as affirming that 26

Quodl. II.11 concl. 8, A 112vb, P 55va. This passage is edited in Hoffmann forthcoming b, Appendix B.

Intellectualist Accounts of the Angelic Fall

217

fallibility in Adam is punishment for his sin; from this it is concluded that before sinning, Adam was infallible; hence error presupposes an evil will (cf. Chapter 3, pp. 69–70). Pouilly observes that so stated, the quotation is incomplete. Augustine does not write that the punishment for sin is to err, simply speaking, but rather to err unwillingly (invitus) (see indeed lib. arb. III.18.52.179). It was possible to err before sin; but the point is that then it was also possible not to err, and error was voluntary. Now, in contrast to Adam, we err willy-nilly (Quodl. II.11 concl. 8, A 112vb, P 55va–vb). Thus, for Pouilly, the infallibility problem is a false problem. One need not explain how the angels could sin despite infallibility, for they were not infallible in the first place. In making these points, Pouilly does not consider other arguments for angelic infallibility, such as those of Aquinas, that are based upon the working of the angelic intellect. If the error that led to the first sin was voluntary, how could the angels and progenitors have guarded against it? Pouilly addresses this question in another context: while discussing the cause of an evil will, he comments that his theory on that subject applies also to the sin of angels and progenitors (see Chapter 7, p. 193). The immediate cause of an evil will is error or deception, the cause of which is negligence. We control whether we are negligent or diligent through intellect and will, by which we can apply good objects and disapply bad objects. By deliberating diligently we can avoid making an error in reason and then committing a subsequent sin in the will (Quodl. IV.6c., P 120rb–va, N 91rb–va).27 Presumably, the case of the angels is analogous for Pouilly: by being diligent, they could apply an object in conjunction with the fitting order in which it was to be desired, or through negligence, they could apply an object that omits the correct order. As we have seen in Section 7.9, Pouilly holds that negligence is not itself to be traced to a per se cause. It seems, then, that for Pouilly, the control of being either negligent or diligent is primitive, and likewise the control of applying or disapplying an object to the will. If this is so, he is not so far from voluntarist thinkers, except that Pouilly traces control equally to intellect and will, whereas voluntarists trace it mainly to the will.

***

Of the three problems mentioned at the beginning of the chapter, Philip, Albert, Aquinas, and Pouilly address above all the infallibility problem. Philip and Albert conceded that angels can err about particulars and hence about the 27

The cited passage is edited in Hoffmann forthcoming b, Appendix C.

218

Angelic Sin

cognitive antecedents of a choice. Aquinas did not allow for angelic fallibility, only for nonconsideration of habitual knowledge. Pouilly’s concern is less to explain angelic sin than to show that it does not demonstrate the will’s ability to act against the judgment of reason, and he is willing to sacrifice the angels’ infallibility for that. Aquinas does not explicitly address the control problem, but his moral psychology makes room for angels controlling their choices by controlling what they actually consider. The control problem poses a serious difficulty for Godfrey, who all but admits that his premises leave him unable to explain how an angel can sin of his own free will. Pouilly feels more confident than his master about the problem of the angels’ (and progenitors’) sin. But he can safeguard their free decision only by covertly making the power to apply or disapply an object to the will primitive. Control of choices among alternatives comes with a price: although a freely made choice is made for a reason, it remains at bottom unexplained why the agent acted on this rather than on that reason. Intellectualists are in general more reluctant to accept that free choices escape full intelligibility. In the case of the angels, however, who under the same initial conditions made opposite choices, this acknowledgement is inevitable. Voluntarist thinkers are generally more forthright in acknowledging the ultimate inexplicability of acts of the will in general and of evil acts in particular. We now turn to their accounts of how a sinful choice can originate in the angelic mind.

chapter 9

Voluntarist and Intermediary Accounts of the Angelic Fall

Angelic sin is a challenge above all for intellectualists. But not only for them. Voluntarists and thinkers defending an intermediary theory of free will can use the fall of the angels as evidence against intellectualists that evil willing does not presuppose a cognitive deficiency, and some theologians limit their discussions to this purpose. The topic also gives them the opportunity to buttress their theories of free will. But nonintellectualists need to give a plausible account of how the evil angels’ choice could be intentional, and hence not an irrational act. They must also explain how their will could be defective while their cognition was unflawed. We begin with theories developed from a voluntarist perspective, from Bonaventure to Duns Scotus. We omit William of Ockham, for apart from passing remarks, he left us no discussion of how angels can sin.1 As representatives of intermediary theories of free decision, we will examine the accounts of Giles of Rome and Peter Auriol.

9.1 Bonaventure The principal discussion of the angelic fall in Bonaventure’s Sentences commentary follows closely the themes treated by Lombard himself. Bonaventure intends above all to clarify the Christian doctrine of the fall through an attentive reading of “the saints” (especially Augustine). In discussing the topic, he displays great sensitivity to the moral psychology of angelic affections (for example, of self-love, pride, and envy), but rarely to the relation of intellect and will. The Aristotelian dictum that Philip the Chancellor introduced into the discussion of angelic sin, that the intellect is always right, surfaces in a context that is not about moral psychology, but

1

See Quodl. II.6c., OTh IX: 137 ll. 38–9, arguing that the angels could sin, even in the first instant, since they could freely conform or not conform to right reason; and QV 7.3c., OTh VIII: 366 ll. 580–95, affirming the fallen angels’ ability to act against right reason and insinuating that they had this ability also prior to the fall.

219

220

Angelic Sin

rather about nature and grace (In Sent. II.5.3.1 arg. 2 and ad 2, II: 154a, 155b). Bonaventure maintains the traditional view that the first sin of Lucifer was pride. He explains that his pride began as presumption, occasioned by the sight of his own beauty, and was consummated in aspiration (ambitio), for by being presumptuous about himself, he aspired to things above his means (In Sent. II.5.1.1c., II: 146b), or as he says later, he aspired to an excellence beyond the excellence he had received (Breviloquium II.7, V: 223a).2 To see how Bonaventure understands the relation between intellect and will in causing Lucifer’s pride, we must turn to a different context. In a Question that asks whether the will is the subject of every sin, an opening argument makes the case that the subject of sin is more properly reason than the will. The argument proceeds as follows: the first inordinate judgment (of Adam or Lucifer, it is implied) was prior to inordinate love; this inordinate judgment was sinful; hence the subject of sin is reason (which made the judgment). That there must be disorder in the judgment before there is disorder in love is shown by the assumption that no one loves something unless he or she judges it as to be loved. But the inordinate judgment must be either sinful or a punishment for sin; and, given Augustine’s axiom that punishment does not come before fault, it appears that the inordinate judgment must itself be the fault (In Sent. II.41.2.2 arg. 6, II: 951b). Bonaventure’s reply reads like a manifesto for the later voluntarist movement: To the objection that a bad judgment precedes every bad will, it must be said that this is true regarding the act of the will that is full consent. Nevertheless, in a rational creature as it was instituted [from creation], an erroneous judgment would never intervene unless the will mixed itself in with this judgment. In fact, Adam would never have been presumptuous, or he would never have judged presumptuously about himself, unless he had adhered too much to himself by the will’s love. Thus love astonishingly perverts a judgment. Hence, although the first angel, or Adam, when he sinned was first presumptuous before he aspired [improperly], the perversity of presumption nevertheless had its origin in the will, and the wickedness of aspiration was consummated in the will. Thus sin regards the will above all, for it begins in the will and is consummated in the will. And so the will is rightly called the first subject of sin, for it is first as to the origin and as to the completion or consummation. (Ibid., ad 6, II: 953b) 2

Faes de Mottoni 1991 offers a greater variety of themes in Bonaventure’s theory of the fall, which, however, do not include the relation of intellect and will.

Voluntarist and Intermediary Accounts of the Fall

221

That the will is the subject of sin was uncontroversial in the later debate. But the ideas that sin has its origin in the will, and that a corrupt will blinds the intellect, became contentious issues, as was seen in Chapters 3–5 and as we will see below. Bonaventure adds that Lucifer’s sin was unique in that there was nothing that inclined his will either from without or from within. Bonaventure therefore ranks it as involving the highest degree of voluntariness compared to other kinds of sin (In Sent. II.41 dub. 3, II: 956a). He does not engage Aquinas’s position because he writes before him. For almost everyone else considered in this chapter, Aquinas’s view is the preferred target of critique.

9.2 John Pecham Bonaventure’s confrère John Pecham follows him regarding the order of deficiencies in angelic sin. In Question 4 of his first Quodlibet (of Lent 1270), he is asked whether the evil angel was first disordered in his cognition or volition.3 He answers this Question by engaging Aquinas’s early account (In Sent. II.5.1 a. 1 and a. 3), which he summarizes as follows: In the angels, the intellect was disordered before the will was disordered. By being intent on one thing, the angelic intellect is withdrawn from another; by being too intent on his own beauty, the evil angel did not consider the rectitude of virtue, and so he erred by not considering what he was supposed to (Quodl. I.4 n. 4, ed. Etzkorn, 12). Pecham formulates three arguments against Aquinas. First, the angel would not be intent on himself without willing to do so, and so he erred voluntarily; thus there was already an error in the will before the disorder in the intellect (n. 5, p. 12).4 Second, before his sin, everything was well-ordered in the angel; hence the angel could not be more intent on something than he was supposed to be (ibid.). The antecedent is implied in the Augustinian deficiency thesis, which is based upon Augustine’s claim that cognitive error was only a punishment for Adam’s sin and hence could not precede sin; Pecham extends the scope of the thesis from Adam to the angels (n. 3, p. 11). Finally, the assumption that the angel was immoderately intent on himself implies that his first sin

3 4

For the dating, see the introduction to the critical edition, p. 23*. “Contra, ⟨non⟩ [ms.: si] intendit sibi erronee si [= ms.; Etzkorn conjectures “sed”] non intendit sibi volens, ergo voluntarie erravit; iam ergo erat error in voluntate.” I conjecture “non,” while the sole manuscript, Firenze, Biblioteca Nazionale Centrale, Conv. Soppr. J.I.3, fol. 40ra, reads “si.” It is not unlikely that a scribe would confuse non (abbreviated n¯) and si (written ſı).

Angelic Sin

222

is an omission (the failure to restrain the focus on himself, as is implied), contrary to the common assumption that the first sin was pride (n. 5, p. 12). In his own answer, Pecham formulates some typical voluntarist ideas. He insists like Bonaventure that the first disorder of the evil angels was in their will. Lucifer delighted in himself at first legitimately, but then immoderately, at which point his intellect was darkened; finally, he fell into the perverse desire to “rob the divinity,” that is, to be equal to God5 (n. 6, p. 12). Pecham adds that to know is delightful, but this delight must be checked; immoderate desire is blinding and leads to an erroneous choice. But an evil choice does not presuppose a cognitive disorder, for cognition is not the total cause of an act of free decision, only its precondition (dispositio). Thus after a thing’s cognition, free decision remains free to want it or not. Nevertheless, a consummated evil choice is not without error, as is attested by Aristotle’s dictum that “every evildoer is ignorant” (EN III.1, 1110b28–9). But the order of priority is important: a cognitive error does not cause an evil choice, but rather an evil will causes the intellect to err (n. 7, p. 13). The first disorder, then, was the transition from legitimate to immoderate self-delight, caused by the will alone, not by a cognitive deficiency. No further explanation is available for why Lucifer delighted immoderately in himself.

9.3

Henry of Ghent

Some of Bonaventure’s and Pecham’s ideas are taken up and developed by Henry of Ghent. In his eighth Quodlibet of Advent 1284, Henry was asked whether the first act of the human being or angel in the state of innocence could be evil (q. 10, 320rL). Henry restricts his response to the sin of angels. Whether the angels actually sinned in their first act was no longer up for debate, for in 1241, William of Auvergne, bishop of Paris, and Odo of Châteauroux, the chancellor of the University of Paris, in accord with the University’s theology masters, condemned the statement “that the evil angel was evil in the beginning of his creation, and that he has always been evil.”6 Nevertheless, there is still room for philosophical investigations about the possibility of the angels’ first act being evil. Henry uses this topic to clarify the necessary preconditions for the sin of the angels, and in 5 6

Cf. Glossa ordinaria in Ps. 68.5, citing Augustine, en. Ps. 68.1.9. Chartularium Universitatis Parisiensis, I: 170–1. On the historical circumstances of the condemnation, see Dondaine 1955, 214–17.

Voluntarist and Intermediary Accounts of the Fall

223

general, for producing a free act. In particular, he clarifies that the will can move itself to a new act from scratch, so to speak, without already actually willing something. In his response to the question, Henry discusses a view according to which the angels’ first volition proper, or “deliberate volition” (actus voluntatis deliberativus), cannot be evil because it follows directly upon the impetus of a natural volition (actus voluntatis naturalis) conferred on the angels by God. Accordingly, a first evil volition would have its remote origin in God, which is impossible because God does no evil (Quodl. VIII.10c., 320vM). Since the reported view claims support in Anselm’s theory that the first volition of the angels is given to them by God, Henry uses this opportunity to scrutinize Anselm’s theory itself. Anselm argued that the angels could not have moved themselves from not willing to willing without first having received, through their creation, a volition in virtue of which they willingly moved themselves to their first new volition (DCD 12; see Chapter 1, p. 16). For Henry, by contrast, all that is needed for a new volition is a lovable good, made present through cognition. In his view, the voluntary character of the new volition need not be guaranteed by a preceding, second-order volition by which one wants to have that new volition, for every volition is as such essentially willed. According to Henry, to say “I want to will” is redundant, unless a volition is not in one’s power – as for example in the case of someone who lacks the virtue of charity and wants to love God by charity (fol. 320vN–321rO). Henry agrees with Anselm’s theory only if the volition presupposed in the will’s moving itself to have a new volition is understood as a dispositional, natural volition to pursue what is beneficial (commodum) and avoid what is detrimental (incommodum) (fol. 321rP). He accordingly solves the question he was asked: the angels’ first volition cannot be evil because it is a natural volition, placed in the angel by God. This natural volition does not, however, produce through its impetus the first volition proper of the angels, as the view under discussion claimed; it merely allows the will to produce by itself its first volition proper (fol. 321rQ). The difference between Henry’s own position and the anonymous view he critiques is subtle and seemingly trifling. But to Henry, it is highly significant that the will can produce in itself its own act of willing, without being moved to its act by a previous volition that was received from God. The will’s self-motion to a new volition presupposes only the cognition of an object and a dispositional, natural volition to seek good and avoid evil. The importance of this claim emerges more clearly in Quodlibet XII, Question 26, where Henry refers back to the discussion

224

Angelic Sin

of Quodlibet VIII within a critique of rival accounts of the will’s selfmotion: Henry rejects the theory of Thomas Aquinas and Giles of Rome that the will can move itself to will the means to an end only in virtue of already actually willing an end, and he also rejects the theory of John of Morrovalle that the will can move itself to will something only in virtue of being actualized by an “affection.” Against these views, Henry contends that the will has its volition – of an end or a means – from itself alone, without any prior motion (such as the will for an end) or antecedent condition (dispositio, such as an affection). For Henry, the fact that a volition is not obtained from another – not even from God – applies also to the will’s first volition. The first volition can only be said to be from God, Henry now clarifies, insofar as the will has from God the natural power (naturalem vim) by which it moves itself (Quodl. XII.26c., XVI: 155–6 ll. 78–93). Another issue Henry discusses in Quodlibet VIII.10 concerns the famous topic of the relation between a cognitive deficiency and an evil will in the sin of Lucifer. According to Henry, Lucifer pondered his prerogative over the other angels and considered it good to rule over them; and he took delight (complacentia) in this thought. This was no sin; indeed, other, good angels were superior to inferior angels as well, and they too were aware of their prerogative and delighted in it, but with moderation (fol. 322vR). Lucifer, however, surpassing innocent delight, burst into pride (elatio) in his unique status, taking inordinate glory in it (fol. 322vR–323rR). Henry points out that this sin did not proceed from a cognitive disorder or error, for his intellect proposed to the will only what he truly was. His intellect could in fact not err prior to his sin of pride (fol. 323rR). Instead, it is pride that corrupts his thinking: Hence reason erred in him only after he prided himself. For after his first act of pride [elationem], his reason darkened in him regarding the desirable objects it proposed, and [he considered] his leadership over the others as choiceworthy and good for him, without reflecting – because the will did not command the intellect to reflect – that it was not due to him on purely natural grounds [ex puris naturalibus]. Therefore, without awaiting reflection, he immediately desired that leadership . . . not attending to the mode and order of attaining it. . . . He presumptuously thought that God’s common and natural influence sufficed for him, and that he should obtain on purely natural grounds that he rule over others, although he should have expected to receive this [leadership] only as a special gift of God, if indeed God so wanted. Hence he should have suspended the judgment of reason and the desire of the will regarding what he wished for, until he received a divine precept, as the good angels did. (Fol. 323rR)

Voluntarist and Intermediary Accounts of the Fall

225

Henry’s view that a new volition does not presuppose any altered cognition allows him to affirm, like Bonaventure and Pecham, that Lucifer’s will became deficient without any prior cognitive deficiency. Henry also agrees with them that Lucifer’s disordered will caused the intellect to err; the deficient will allowed for the presumptuous belief that he could obtain leadership over others apart from a special grace. But Henry takes their voluntarist theories of error formation a step further: Lucifer’s cognitive error must be traced to the failure of the will to command the intellect to reflect upon the limits of his natural powers. Henry’s discussion in Quodlibet VIII implicitly honors a principle he had formulated earlier: sin cannot be caused by an error, unless the error is itself caused by a corrupt will that distorts cognition or that fails to direct reason to consider what it should (Quodl. I.16c., V: 112 ll. 89–96; cf. Chapter 7, p. 188). As we have seen, in Quodlibet X, Henry develops systematically the theory that an evil will does not presuppose, but rather causes a cognitive deficiency (Chapter 3, pp. 80–4). Henry does not consider whether Lucifer’s choice was, or could have been, clear-eyed in the sense that he was fully aware of its negative implications for himself. But he reiterates in his discussion of Lucifer’s sin that the will need not follow the judgment of reason (Quodl. VIII.10 ad arg. in opp., 323rT), and earlier he had said of Adam that he acted contrary to his clear judgment of what he should do (Quodl. I.17c., V: 128–9 ll. 95–08). As we will now see, Peter Olivi has a more nuanced position.

9.4

Peter Olivi

That the angelic fall is an intriguing problem even for voluntarists comes to light especially in Olivi, who takes great pains to make the angels’ decision to sin intelligible.7 In his principal treatment of the possibility of angelic sin in Question 43 of Book II of the Summa, Olivi reports theoretical difficulties concerning angelic sin by certain people whom he calls pejoratively “philosophantes” (as opposed to philosophi, his word for genuine philosophers). He seems to intend Christian theologians with Aristotelian inclinations, such as Aquinas.8 For example, it seems implausible (incredibile) to the philosophantes, writes Olivi, that the first angel (Lucifer) preferred what he plainly 7 8

For Olivi’s theory of angelic sin, see also D’Ercole 2017, 132–47. The term “philosophantes” with negative connotations is found especially among thirteenth-century Franciscans. Regarding Roger Bacon, Matthew of Aquasparta, and Roger Marston, see Gilson 1952.

226

Angelic Sin

saw to be of no value compared to the divine good, which was offered to him clearly and without any strain or hurdle. It seems even more implausible that Lucifer chose to glory in himself and to rule over others apart from any connection to God’s glory and rule, for that connection would have made his self-glory and rule better and more useful, delightful, and glorious. It is also completely implausible that Lucifer and the other evil angels, who knew that their fall was irreparable and entailed perpetual damnation, chose nevertheless their irreversible fall and damnation (43 argg. 1–2, 6, I: 714–15; all references are to Summa II). These arguments make strong cases against the hypothesis of clear-eyed angelic sin. Action against one’s better knowledge without the admission of clear-eyed evildoing is precisely what Aristotle’s account of incontinence is intended to explain. His tool, as we have seen, is a distinction between what we may call weak and strong practical knowledge: incontinence is possible because one can act against one’s habitual and general knowledge, but clear-eyed incontinence is impossible because one cannot act against one’s actual and particular knowledge (see Chapter 1, pp. 25–6). In Questions 85–6, while discussing Aristotle’s account of incontinence, Olivi mentions “contemporary theologians” – that is, Aquinas – who apply Aristotle’s account to the fall of Lucifer and Adam, claiming that they could not have consented to evil if they had considered that the act is evil, and thus they could not have sinned unless they did not know or failed to consider that their sin should be avoided. But Aquinas’s account is unacceptable to Olivi, who marshals seven arguments against it. According to one argument, this position implies intellectual determinism. As Augustine points out, we do not control by which sight we are affected (lib. arb. III.25.74.255); Olivi adds that at least prior to an act of the will we do not control what affects us. As Olivi understands Aquinas’s position, it implies transfer of necessity from cognition to judgment to consent: cognition that is beyond control necessarily entails an actual and particular judgment of what is good, which in turn necessarily entails the will’s consent (85–6c., III: 188–9; cf. 57 ad 29, II: 382). Another argument shows that this position implies the blameless ignorance dilemma (cf. Chapter 8, p. 199): the nonconsideration leading to first sin cannot be culpable, for then there would be sin prior to the first sin; it cannot be nonculpable, for then the angels who sin because of nonconsideration would fall into sin without their fault (85–6c., III: 189; see also 57 ad 18, II: 360).9 9

The chronological order of the questions most relevant to angelic sin seems to be this: 57, 85–6, 43. Questions 85–6 and 43 contain cross-references to 57. Piron, 2020a and 2020b, dates Question 57 to ca. 1277–8 and Question 43 to ca. 1290–1. Piron does not discuss the dating of

Voluntarist and Intermediary Accounts of the Fall

227

What is apparent throughout Olivi’s own explanation of the angelic fall is a certain anthropomorphism in his conception of angels. This seems intentional; Olivi maintains the unique view that the angelic mind is essentially the same as the human mind, only more powerful (56, II: 298–304). Olivi intends to avoid intellectual determinism without attributing to the evil angels the kind of irrational behavior implied in clear-eyed evildoing. His explanation of the fall of the angels has two principal components. One is a distinction between weak and strong practical cognition that replaces Aristotle’s distinction between general, habitual cognition and particular, actual cognition. Through this distinction Olivi can argue that the angels’ cognitive conditions were not as ideal as the arguments he attributes to the philosophantes suggest. Another element is Anselm’s theory of the two affections of the will, which clarifies that the angels are not set up from creation as maximizers of self-interest. Olivi changes the nomenclature: he calls one affection natural (affectio naturalis), the other – which he explicitly equates with Anselm’s affectio iustitiae – virtuous (affectio virtualis). By their natural affection the angels are attracted to themselves and to their visible companions in proportion to the natural goodness found in them; their virtuous affection is principally rooted in the love of God above all things. The natural desire for happiness flowing from the natural affection needs to be checked by the virtuous affection, so as to avoid that natural, inferior desires pursue their objects infinitely and without measure. The order of natural love was more manifest to the angels than the proper order of virtuous love for God, the knowledge of which requires an intimate, loving relationship with God (43c., I: 717–18). The interdependency between cognition and love is also the idea behind Olivi’s own distinction between weak and strong practical knowledge.10 For Olivi, doing the right thing requires a commitment that does not simply flow from cognition – however actual and particular the cognition may be. With varying terminology, this idea spreads through the Summa. Strong practical knowledge involves “tasting” the object: one can perceive the truth of goodness and badness on the one hand through discursive reasoning or faith, and on the other hand through taste and sensation (per modum gustus et sensus), by which one tastes an intellectual sweetness in the good and a bitter and repulsive insipidity in vice (57 ad 18, II: 361). Again,

10

Questions 85–6, but Question 43 seems later than 85–6 because 85–6 lacks a cross-reference to 43, although one would expect it, since 43 has a much richer discussion than 85–6. Prior to Olivi, a similar distinction had been made by William of Auxerre, Bonaventure, and Thomas Aquinas; see Chapter 2, pp. 32, 39, and 47.

228

Angelic Sin

we can know, think, opine, or believe on the one hand by simple knowledge or theorizing (per speculationem), and on the other hand with an experienced taste and an affective approval or disapproval (85–6c., III: 193). The contrast between merely knowing and tasting is a Franciscan theme; Bonaventure reports in his Lectures on the Hexaëmeron, which appear to have had a great impact on Olivi, that St. Francis wanted his friars to study, as long as they first practice in their own lives what they teach; “for what good is it to know, but not to taste [gustare]?”11 What can render practical knowledge strong is also experience or sight. The first angel knew theoretically (speculative) that he could not sin without being maximally disordered, but he did not know this by experience, for no one had ever experienced sin (43c., I: 724–5). Similarly, the angel knew with certainty (certitudinaliter) that God, the uncreated good, is infinitely better than a created good, but the created good appeared to him as more visible and therefore more accessible (43 ad 1, I: 726).12 Whether practical knowledge is weak or strong can also depend on the object: it is one thing to know that something is a noble good (bonum honestum), and quite another to know that it is a delightful or useful good (85–6c., III: 194). In light of these psychological presuppositions, Olivi describes in remarkable detail how Lucifer and the other angels may have sinned. His aim is to describe no more than a plausible scenario, for he remarks that apart from what Scripture tells us, we can know about the nature of their fall only through similitudes based upon lower things familiar to us (43, I: 734). According to Olivi, the root of Lucifer’s sin was self-referential selflove, that is, loving himself for his own sake rather than for the sake of God. Quoting Augustine, he phrases this as self-love to the point of contempt of God rather than love of God to the point of contempt of self (ciu. XIV.28). Thus his sin consisted not in a choice of unfitting things related to a fitting end, but rather in pursuing an unfitting end (43c., I: 718–19; cf. 57 ad 17, II: 359). The inordinate self-love entailed presumption, for he did not want to be under anyone else. Since he did not submit to any rule, his desire for honor and power was without limit (43c., I: 719). Olivi explains what made it attractive for Lucifer to rule over the other angels and for them to be ruled by him. By being united to the rest of the 11 12

Collationes in Hexaëmeron 22.21, V: 440b; cf. Schlosser 2014, 27. For echoes of the Collationes in Olivi’s writings, see Piron 2003. On Olivi’s two types of practical knowledge, see also the lucid analysis by Müller 2009, 625–8. While Müller argues that Olivi defends clear-eyed evildoing, I find this is not the case in his account of angelic sin.

Voluntarist and Intermediary Accounts of the Fall

229

angels as their ruler, Lucifer felt as it were duplicated in each of them, and therefore magnified. The other, inferior angels found the companionship of the first angel delightful, much as Adam delighted in Eve’s companionship. The companionship among the angels was more visible and hence more enjoyable to them than God’s, for although the angels knew God, they did not see him as they saw each other, and so “their intellect related to God as absent.” Since Lucifer was more visible to the other angels than God, they could easily be drawn to him in a relation that was distinct from the relation to God. They also felt greater in union with him than alone; separation from him was like being cut off from the benefits of a king and his kingdom. Accordingly it was easier for them to be disordered in their love of Lucifer (43c., I: 717, 719–21). Besides, when trying to draw the other angels to consent, Lucifer displayed himself in all the decency and affability he was capable of (43 ad 5, I: 728–9). So the sin of Lucifer and the other angels had a certain logic, albeit a perverse logic. The evil angels did not reject what they understood to be in their own best interest. The preference of self and one’s companions to God, and thus of an infinitely lower good, was not clear-eyed. Following Anselm, Olivi furthermore holds that they did not know that God would damn them irrevocably. Olivi also considers it unlikely that from the first beginning of their sin, they deliberately wanted to separate themselves eternally from justice and virtue (43 ad 6, I: 729), or that their sin was from the start an open rejection of virtue, “just as if a pious virgin publicly prostituted herself to every passerby.” More probably, they sinned at first hypocritically, upholding an appearance of virtue (43, I: 731). Their sin was less intense and less ramified in the first moment than thereafter (43c., I: 723). If the angels sinned under less than optimal knowledge conditions, unaware of the consequences of their act, it may seem that they should be pitied rather than blamed. But for Olivi, this would miss the point. From the outset, the angels were so perfectly connected to the order of justice that they could not excusably draw away from it; yet they were not so deeply rooted in it that they could not by their free decision move in the opposite direction (43c., I: 726). What caused the evil angels’ sin was not imperfect knowledge of God and ignorance of the consequences of their act; in fact, the good angels shared these knowledge conditions. Rather, what caused their sin was their perverse will, setting as the final end their own good rather than the divine good. If they had pursued a good end, albeit with unfitting means, then they would indeed have simply made an involuntary and inculpable miscalculation (103 ad 8, III: 242). Setting one’s

Angelic Sin

230

end in one thing rather than another, then, seems to be primitive in Olivi’s view, rather than the result of a judgment about which end is worth pursuing in view of a higher end. So for Olivi, the evil angels fell because of their evil will, not because of their faulty knowledge. They did not simply make a miscalculation, but rather their love was disordered. Nonetheless, they could intentionally make their evil choice only because they had less than ideal knowledge conditions, for example, ignoring the punishment that awaited them. Nor did the good angels act out of pure calculation: if they had refrained from sinning only because they feared punishment, they would have acted from self-interest, not from love (43c., I: 725).

9.5

John Duns Scotus

Olivi had a strong impact on Duns Scotus’s account of angelic sin, although Scotus leaves him unacknowledged, apparently to stay out of trouble (cf. Chapter 3, p. 85). Scotus’s own answer to the question of how angelic sin is possible is mostly a by-product of his discussions of other issues concerning the fall. He brings up the topic directly in critiquing a thesis maintained by Aquinas, which essentially states that if the intellect cannot think otherwise, the will cannot will otherwise.13 Aquinas uses the thesis to explain the permanence of the angels’ fallen state, for he holds that the angelic intellect can think otherwise before the angels sin but not thereafter (ST I.64.2; see Section 10.1). But Scotus argues that if Aquinas’s maxim is true, then the angels could not have sinned at all: Since the intellect of the angels was correct in apprehending (for punishment does not precede fault), it moves the will to desire something correctly. Nor could the intellect move differently, for it moves by way of nature; hence it moves the will to will correctly. Therefore the will could never sin. (Ord. II.7 n. 20, VIII: 83)

That the intellect moves “by way of nature” means that it does not control its act; recall that for Scotus the intellect is a power of opposites only insofar as it is under the will’s control (Chapter 5, p. 121).14

13 14

The issue is the angels’ occurrent thinking, not their habitual knowledge, most of which they do not control, according to Aquinas; cf. Chapter 8, pp. 211–13. For Scotus’s use of this idea in discussing angelic sin, see also Lect. II.6.2 n. 40, XVIII: 382; Ord. II.6.2 n. 58, VIII: 54.

Voluntarist and Intermediary Accounts of the Fall

231

We will first consider Scotus’s explanation of how the angels’ will could defect from a moral rule, and then we will turn to his account of how their will can be flawed while their intellect does not err. Scotus offers the first explanation while arguing that Lucifer’s sin did not consist in pride. Like Olivi, he thinks that Lucifer’s sin developed gradually; according to Scotus, it started with inordinate self-love, followed by immoderate desire and envy, and ended in hatred of God – although he expresses the doubt that hatred of God, understood as the wish for God’s nonexistence, is psychologically possible, since in God there is no aspect of evil.15 Scotus’s most important contribution to the question of how angelic sin is possible is found within his discussion of Lucifer’s immoderate desire. In order to clarify the progression of Lucifer’s sin, Scotus analyzes the structure of his evil volitions. This analysis depends on a similar one by Olivi, which, however, is less central in Olivi’s own explanation of the fall.16 Most volitions are complex volitions that have more than a single object. As regards acts of love, most are composed of what is loved and for whom this is loved. Scotus expresses this twofold object through the traditional vocabulary of desire-love (amor concupiscentiae) and friendshiplove (amor amicitiae) – or “desire-willing” and “friendship-willing,” as he also calls them. Love of something for someone is composed of “desirelove” for some good and “friendship-love” for the beneficiary of the desired good. The desired good can be a thing or a person, the beneficiary can be someone else or oneself. Scotus argues in some detail that most plausibly, Lucifer’s first disordered friendship-love was love of self, while his first disordered desire-love was wanting happiness for himself.17 That self-love can be disordered is clear. But that love of happiness can be disordered is not obvious. Scotus formulates an objection against his position: everyone desires happiness; what everyone desires is desired naturally; the natural desire is God-given and thus always correct; hence no one sins by desiring happiness.18 Scotus replies that the natural inclination to happiness cannot be excessive, but the elicited act – the deliberate volition – of willing happiness can indeed be disordered.19 According to 15

16 17 18 19

Lect. II.6.2 n. 44, XVIII: 383–4; Ord. II.6.2 nn. 68–72, 78; VIII: 60–2, 65–6. Regarding hatred of God, see Ord. II.6.2 n. 78, inserted text, VIII: 66; Rep. II-A.6.1 n. 9, Viv. XXII: 617a–b; Rep. II-A.43 n. 5, Viv. XXIII: 229b. Summa II.43c. and ad 3, I: 720, 727–8; II.103, III: 232–44. Lect. II.6.2 nn. 24–32, XVIII: 376–80; Ord. II.6.2 nn. 34–45, VIII: 39–47; Rep. II-A.6.2 nn. 4–7, Viv. XXII: 618b–620b. Lect. II.6.2 n. 33, XVIII: 380–1; Ord. II.6.2 n. 46, VIII: 47; Rep. II-A.6.2 n. 8, Viv. XXII: 620b–621a. Lect. II.6.2 n. 39, XVIII: 382; Ord. II.6.2 nn. 55–6, VIII: 53; Rep. II-A.6.2 n. 9, Viv. XXII: 621b–622a.

232

Angelic Sin

Scotus, one must check the desire for happiness.20 Lucifer’s desire for happiness could be inordinate because he desired happiness more for himself than for God, or more as a good for himself than as a good in itself, or because he wanted it at once rather than at the appointed time, or because he wanted it by purely natural means rather than as a gift of grace.21 Apart from the case of wanting happiness through natural means alone, Scotus conceives of Lucifer’s sin, like Olivi, as setting an unfitting object as his end, rather than as choosing inadequate means to a fitting end. Scotus formulates the idea that one is able to and supposed to check the desire for happiness by means of Anselm’s two-will theory (or twoaffection theory). Following De casu diaboli 13, he writes in the Ordinatio that a fictitious angel who has the affection for advantage but not the affection for justice could not help but will maximally his own advantage; but this would not count as sin because he could not do otherwise. The affection for justice allows the angels to moderate their desire for happiness (Ord. II.6.2 n. 49, VIII: 48–9). Scotus’s story of how the two affections provide the mechanism that allows one to adjust one’s desire to the demands of justice changed from the earlier to the later Sentences commentaries. At first, in the Lectura, he understood both affections as sources of motivation, holding like Anselm that both incline to their respective goods (advantage or justice) and that a person can elicit an act from either affection.22 Later, in the Ordinatio and the Paris lectures, detaching himself from Anselm’s account, Scotus thinks only of the affection for advantage as a source of motivation and identifies it with the will insofar as it is an intellectual appetite. The affection for justice is no longer a source of motivation, but becomes the will’s freedom, by which the will can moderate its volitions according to the demands of the rule of justice, which comes from the will of God.23 In the Paris 20

21

22 23

The idea that one must check one’s desire for happiness amounts to a rejection of eudaimonism, the theory according to which morality coincides with the pursuit of authentic happiness. Concerning Scotus’s rejection of eudaimonism, see Boler 1993 and Williams 1995. Ord. II.6.2 nn. 52–4, VIII: 51–2; cf. Lect. II.6.2 nn. 37–8, XVIII: 381–2; Rep. II-A.6.2 n. 11, Viv. XXII: 622b–623a. By comparison, for Aquinas, there are two ways in which one can sin regarding happiness: either by desiring inauthentic happiness, for example, as sought in carnal pleasures (cf. DV 22.7c.), or by desiring supernatural happiness apart from grace (see Chapter 8, pp. 207 and 209). Lect. II.6.2 nn. 29, 36, 38, and 42, XVIII: 378, 381–3; cf. Anselm, DCD 4, I: 241 and De concordia III.11, II: 281. Ord. II.6.2 nn. 50–1, VIII: 49–51; Rep. II-A.6.2 nn. 9–10, Viv. XXII: 621a–622a. See also Chapter 5, pp. 121–2. The corresponding passage in the Reportatio II-B contains the same doctrine, but lacks the expression affectio iustitiae, using instead regula iustitiae and libertas arbitrii. Scotus’s equating the affectio iustitiae with the will’s freedom may have been influenced by Olivi; see Summa II.57

Voluntarist and Intermediary Accounts of the Fall

233

lectures, he goes a step further than the Ordinatio: he clarifies that the affection for justice is the ultimate specifying difference of the genus “intellectual appetite,” constituting the will as “free intellectual appetite” (Rep. II-A.6.2 n. 9, p. 621a–b).24 Thus the affection for justice has become inseparable from the will. Since it is the will’s freedom and since all acts of the will are free, Scotus now holds that the will does not elicit its act by the affection for advantage without the affection for justice.25 What Scotus expresses with the Anselmian language of the two affections is at bottom a simple point: if the will were purely an intellectual appetite, one could not help but will one’s maximal happiness, and so one could not sin. What matters is that one can check one’s affection for happiness – whether one does so by an affection that is an alternative source of motivation, as in the Lectura account, or by means of the will’s innate freedom, as in the later accounts, does not matter much.26 Furthermore, if the will were an intellectual appetite only, then Aquinas would be right that if the intellect cannot think otherwise, the will cannot will otherwise (see p. 230 above). It remains to be seen how according to Scotus the angels’ will can be flawed although their intellect is correct. He explains this with an ingenious theory according to which not only the intellect but also the will is a “vis collativa,” a power capable of causing mental relations (relationes rationis), that is, of mentally relating things to each other. Scotus defines the mental relation as something the intellect causes in an object under consideration without any causal influence from the object itself (Ord. I.2.2 n. 395, II: 353). In other words, it is a pure mental product. As a vis collativa, the intellect can combine things even if they are not combined in reality, such as “gold” and “mountain,” forming the notion of “golden mountain” (QMet. V.11 n. 39, OPh III: 579). When Scotus claims that the will, too, is a vis collativa, he attributes to it analogous capacities.

24 25

26

ad 12, II: 376. The Ordinatio still contains a trace of the earlier account; in one passage, Scotus presents the two affections as alternative motivational sources of the will’s “elicited acts” (Ord. II.6.2 n. 40, VIII: 43). For a thorough study of Scotus’s transformation of Anselm’s two-affections theory, see King 2010. Scotus only says that it could elicit its act as if it had been elicited by the affection for advantage alone; see Rep. II-A.6.2, n. 9, Viv. XXII: 621b, corrected according to the text witnessed by all known manuscripts: “Nec potest voluntas * elicere actum per affectionem commodi tantum, sine affectione justi, etsi possit elicere actum conformiter, ac si esset solum per affectionem commodi.” At the place indicated by the asterisk, the Wadding/Vivès edition adds “libere et recte operans,” which completely distorts the meaning. Cf. King 2010, 375.

234

Angelic Sin

As mentioned earlier, most volitions are complex, involving more than one term, for example, willing something for someone, willing something as a means to an end, or hoping to receive something from someone. Are the compositions in these volitions caused by the intellect or the will? Aquinas addresses this question regarding the act of choice and argues that the relation between means and end involved in an act of choice derives from the act of the intellect judging that this means is to be willed for that end. For Aquinas, the collatio, that is, the mental arrangement of means and end, is the work of the intellect. Scotus, by contrast, thinks that the will itself makes the collatio, that is, that it is the will that connects the means and the end.27 This explains why the will can will differently from what the intellect judges, and why the will can err when the intellect is correct. Scotus uses the theory of the will as vis collativa to explain how the angels could have a flawed volition without flawed cognition, and also more particularly, how an angel could desire equality with God. (Although the desire to be equal to God was thought to be Lucifer’s sin in particular, Scotus does not single him out by name.) That an angel might desire equality with God seems problematic, for an angel being equal to God implies a contradiction (the angel would be angel and not-angel; notGod and God). Thus the angel’s intellect could not present equality with God to his will as a desirable object. The hypothesis that this contradiction could be the mistake of an erring intellect must be excluded, since the intellect could not err before the first sin (Ord. II.6.1 nn. 6–7, VIII: 26–7; cf. Rep. II-A.6.1 n. 2, p. 614a–b). But Scotus argues that the will can desire equality with God without the intellect presenting it as an object. The intellect need only understand the terms, and the will can connect them – just as when the intellect presents “whiteness” and “crow,” the will can then want the crow to be white. The angel’s intellect can understand “equality with God” without error, for in the Trinity, the Son is equal to the Father. His will can then desire equality with God for himself. This sin involves an erroneous composition, yet it is not the angel’s intellect that commits the error, but his will (Ord. II.6.1 nn. 18–19, n. 23, VIII: 32–5).28 The same is the case when the angels sin by immoderately desiring happiness for themselves. Simply willing happiness by itself is never sinful; the sin lies in willing it to the highest degree for someone to whom this is 27

28

See Aquinas, ST I.83.3 ad 3 and ST I-II.13.1 ad 1; Scotus, Ord. II.38 n. 5 and n. 15, VIII: 450, 453; Ord. III.26 n. 100, X: 30. For a detailed discussion of Scotus’s theory of the will as vis collativa, see Hoffmann 2013. For an excellent discussion of Scotus’s theory of the evil angel’s desire for equality with God, see Pini 2013, 71–81. See also Hoffmann 2013, 1080–3.

Voluntarist and Intermediary Accounts of the Fall

235

not due (Rep. II-A.6.2 n. 11, p. 622b). To cause this volition, the intellect need not erroneously compose “highest happiness” with “self,” that is, the intellect need not propose “highest happiness for me” as an object to be willed. Rather, the will can cause this erroneous composition by itself (Rep. II-A.6.1 n. 6, p. 616a). Thus, no cognitive error or nonconsideration is presupposed in angelic sin. This theory leaves other questions unanswered, especially whether the angels could sin even in full awareness of the severity of the punishment. Scotus does not address this question directly, but he might have denied it, for elsewhere he denies that one can desire one’s misery.29

9.6 Giles of Rome We now turn to Giles of Rome and in the next section to Peter Auriol, who both proposed intermediary theories of free will, according to which the will’s choice is caused by a cognized object or a practical judgment, although the will has some direct, underived control of the intellect’s activity leading to the choice (see Sections 4.1 and 5.4). We will see to what extent this idea plays a role in Giles’s account of angelic sin. In discussing whether angels could sin, Giles addresses both the question of what the deficiency of their act could consist in and how their intellect and will could produce such an act. His answer to the first question is the same in his lectures on the Sentences (of probably 1270–1) and his Ordinatio of Book II of the Sentences (completed 1309 or later).30 It is essentially Aquinas’s position, expressed with Augustine’s notion of sin as privation of form (species), measure (modus), and order (ordo) (nat. b. 4). Giles argues that what was flawed in the angels’ sin was not the form of their act, for they did not desire anything evil; nor the measure, for they did not desire anything that, though good in itself, was unsuitable to them; rather, they sinned by desiring something good and suitable, namely happiness, but outside the correct order, for they desired to achieve it by natural means apart from grace.31 29 30 31

See Chapter 5, note 21. For the dating of the lectures, see Chapter 4, note 5; for that of Ordinatio II, see Chapter 4, note 9. Rep. II.29c., p. 249; Ord. II.5.1.1c., I: 279bA–280aC. The same language – deficiency of order, but not of form and measure – is used in Aquinas’s DM 16.2 ad 1 and ad 4, but nowhere else in Aquinas’s accounts of angelic sin. Who follows whom is an open question; it is possible that Giles functioned as a respondent in Aquinas’s disputed question and thus could have formulated these replies. DM 16 was published in 1272, which does not mean that it was not debated earlier; see Editio Leonina XXIII: 4*–5*; Torrell 2015, 260–4. In Rep. II.33, Giles critiques the explanation of demonic obstinacy of DM 16.

236

Angelic Sin

In explaining how the angels’ sin came about, Giles begins the response in his lectures on the Sentences by following Aquinas’s Sentences commentary – the same text Pecham critiqued in his first Quodlibet. Aquinas started his solution with the intellectualist premise that there cannot be sin in the will unless there is deception in reason, and he backed this claim with considerations from Aristotle. Aquinas does not say in this text that deception precedes sin, let alone causes it. But Giles argues that “in a human being there is never evil in the will unless error precedes in reason” (Rep. II.29c., p. 248). With this affirmation – a first version of Giles’s statement that came to be known as propositio magistralis – he takes a clearly intellectualist stance. But in explaining angelic sin in the same Question, he inverts the priority of the cognitive and volitional disorder, professing a voluntarist account: Hence [the angel] desired inordinately that which he wanted to achieve on natural grounds. But this inordinateness [inordinatio] need not have been first in the intellect, for there was no inordinateness of the intellect before the inordinateness of the will, making him think he could achieve it on natural grounds. For then punishment would precede sin, since sin is in the will. Rather, the will sinned from an omission of thinking. And all this must be attributed particularly to the will, for it wanted the intellect to think only of these things, namely of the excellence of natural [qualities] as such and of beatitude, and to omit thinking that on natural grounds he could not have these. (Rep. II.29c., pp. 249–50)

The idea that no inordinateness of the intellect could precede any inordinateness of the will, since punishment cannot precede sin, is precisely the point Pecham insisted upon in his first Quodlibet about a year earlier, and Giles may well have been impressed by Pecham’s critique of Aquinas. In his Ordinatio, Giles explains in greater detail how the evil angels’ sin could unfold in their mind, in close analogy to his explanation of the origin of human sin in Quodlibet III.16 (see Chapter 4, p. 98). In the first moment of the angels’ existence, they fully understood their own being, and they experienced delight (complacentia) in it. These acts were caused not by them but by God creating them in the state of understanding and willing. Since they did not move themselves to their acts of understanding and delight, these acts could not be sinful. Nevertheless, their delight in this first moment was dangerous, for from it could later have arisen a flawed desire. In fact, after the first moment, they did control the focus of their attention and so they controlled their delight. Seeing their own beauty and splendor, they could move themselves to want to attain happiness by nature rather than by grace (Ord. II.3.2.1.2c., I: 219aB–bC). Thus

Voluntarist and Intermediary Accounts of the Fall

237

they could have sinned by commission (that is, through a flawed act). Giles also considers the hypothesis that they sinned by omission. After creation, they remained fully intent on themselves; the self-cognition faded only if they intentionally turned to something else. They could thus sin through omission by failing to turn elsewhere, remaining too absorbed by themselves rather than referring their own good to the praise of their Creator (Ord. II.3.2.1.3c., I: 221aB–222aB). What is important to Giles is that the angels’ sin involved ignorance, but was not caused by ignorance. To explain this, he reminds the reader of how ignorance is involved in human sin. One can fornicate because one ignores the inordinateness of fornication, but one can also do so in awareness of its inordinateness, but – because of one’s attraction to it – wanting to omit thinking about the disorder of fornication and thinking instead of its delightfulness. Thus what is first infected is one’s will, and the will then infects the judgment of reason to make it judge that fornication is to be done, although really it is not to be done. Passion corrupts the fornicators’ appetite and judgment, and so they sin in ignorance, although they do not sin due to ignorance (Ord. II.5.1.1c., I: 279bA–B). Something similar happened in angelic sin, although the angelic will and judgment were corrupted not by passion, but by delight in their own beauty. This delight made the evil angels fail to consider what they needed to consider regarding how to obtain happiness (p. 281aA), and so they sinned by thinking falsely that they could obtain it by their natural power (p. 281aC–D). They should have thought that they are not competent to judge how happiness is to be obtained and that they needed a divine ordinance – but they failed to think about this, and so they sinned in ignorance (p. 281bB). Giles makes it very clear that although the evil angels sinned in ignorance, they were fully aware that they sinned. He makes an explicit reference to his Question De deceptione, recalling that, apart from those who sin due to ignorance of what is right, sinners apprehend their act as good in one respect and bad in another; but because of the corrupting effect of desire, they judge that, all things considered, the act is to be done (Ord. II.6.1.2c., I: 298aA–C; see Chapter 4, pp. 96–7). Giles speaks particularly about Lucifer: he did not want to wait for an ordinance about how God wanted him to attain happiness – and yet he knew that he had to await such an ordinance. (We might say, he was under a metarule to await a rule.) Although he knew this, he did not take care to act accordingly; the delight in his own excellence blinded him to think that he could pursue happiness without divine guidance. Since he must have been aware that he did not wait for divine guidance, he sinned in ignorance, rather than due to

Angelic Sin

238

ignorance (p. 298aC–bB). Lucifer must also have known that his sin deserved punishment; but following Anselm, Giles holds that Lucifer could think that God would forgive him (pp. 298bB–299aA). It is central in Giles’ account of angelic sin that there was first a disorder in the will and then in the intellect. The angel’s first sin presupposes no occurrent cognitive deficiency, such as the nonconsideration of the moral rule. Nor, it seems, does it presuppose any nonculpable deficiency in the will, as Aquinas assumed; Giles nowhere mentions such a hypothesis in discussing angelic sin in Ordinatio II.3–7 or in examining the cause of evil in Ordinatio II.34. The excessive delight the angel’s will took in his own good caused an error in his intellect, making him judge that something is to be done, although in fact the moral rule forbids it (Ord. II.5.1.1c., I: 279aB). Analogously, according to De deceptione and the Quodlibeta, inordinate desire (concupiscentia) causes an error in reason, making something appear suitable that is not. Giles’s theory contains the intellectualist idea that a disordered desire for happiness as attainable by purely natural means is accompanied by the corresponding judgment that happiness is to be pursued in this way, and it contains the voluntarist idea that the will can move itself to prefer that the intellect focus on one thing rather than another, without presupposing for this preference a judgment that this focus is to be preferred.

9.7

Peter Auriol

Like Giles, Peter Auriol explains the sin of the angels by attributing to the will direct control of cognition. Giles emphasizes the will’s control of the intellect’s actual consideration and thereby of the all-things-considered judgment of what is to be done. Auriol focuses instead on the practical judgment itself. As we have seen in Section 5.4, Auriol holds that the will can determine the practical judgment’s motivating power. To explain angelic sin, he adds that the will’s determination can cause an erroneous practical judgment. Auriol begins his discussion of whether the angels can sin with a critique of Aquinas’s explanation in the Summa theologiae, which Auriol expounds as follows: sin presupposes a defect in the intellect; in the case of the angels, this defect could not have been ignorance of what one is held to know, nor an error, but only nonconsideration, namely of the necessity of grace to attain happiness (Rep. II.4.3c., Fb 30ra, 78aC–E).32 Auriol’s objections, 32

Regarding the textual situation of Auriol’s Reportatio II, see Chapter 5, note 67.

Voluntarist and Intermediary Accounts of the Fall

239

among others, are these: Lucifer’s nonconsideration could not have happened by chance, for sin occurs only in deliberate acts; nor could it have happened because of Lucifer’s natural constitution, for he is by nature more splendid than some other angels who were good, and thus his actual consideration would have been more intense; hence Lucifer’s nonconsideration must have happened because Lucifer did not want to consider grace, while the good angels did. Thus sin does not begin with nonconsideration, a failure of the intellect, but rather with not wanting to consider, a failure of the will (Fb 30rb, 78aF–bA).33 Similarly, if adequate consideration was required in order to avoid sin, then the nonconsideration is itself a sin, in addition to the sin that followed upon it (Fb 30rb, 78bE–79aA). Auriol also thinks that Aquinas’s account contradicts Article 129 of the March 7, 1277 condemnation, which by censuring the contrary view establishes that “When the intellect is at the pinnacle of its act, the will can go against it” (Fb 30rb, 78bE).34 Auriol agrees with Aquinas at least to the extent of holding that in the moment in which an evil angel sinned, there was some ignorance in him. He backs this thesis with Aristotle’s claim that every evildoer is ignorant and does evil because of ignorance (EN III.1, 1110b28–30). The specific ignorance Auriol has in mind is an erroneous verdict (sententia), that is, a false practical judgment. In fact, as we have seen in Section 5.4, Auriol holds that choice follows a verdict, and, it is implied, a bad choice follows an erroneous verdict. Crucially, however, Auriol claims the verdict’s falsity is not beyond control, for the ignorance that corrupts the verdict followed upon free decision, rather than “upon the nature of the thing” and syllogistic deductions from first practical principles (Rep. II.4.3c., Fb 30rb–va, 79aA–B). He formulates this idea by means of his dualjudgment theory. The “indicative judgment” (“one must do this”) is uninfluenced by free decision and does not move the will to a corresponding choice, and so Auriol can defend the idea that the will can act against a judgment, even if it regards the particular. (So Auriol agrees with the position enforced by Article 129, which censures the 33

34

The argument ends with the statement, witnessed in most manuscripts: “habeo propositum quod prima radix peccati non fuit in angelo ex parte voluntatis sed ex parte intellectus [I have shown my point that the first root of sin in the angel was not on the side of the will but on the side of the intellect].” But the thrust of the argument and of the engagement with Aquinas goes in the opposite direction: to show that the root of the sin is in the will, not the intellect. I suppose that locating the root of sin in the intellect rather than in the will is a mistake in the student report. Two manuscripts correct the mistake, one by leaving out “non,” the other by swapping “voluntatis” and “intellectus.” For the manuscripts of Reportatio II, see Schabel 2000, 155. For this formulation of Article 129, see Henry of Ghent, Quodl. X.10c., XIV: 271 ll. 41–2.

240

Angelic Sin

opposite idea.) By contrast, the “imperative judgment” (“Do this now!”) is based upon a corresponding indicative judgment, but has undergone the determination of free decision and consequently moves the will without its being able to act against it (Fb 30va, 30vb, 79aB–bA, 80aB). On this basis, Auriol explains the evil angels’ error as follows: In the intellect of the sinning angel . . . no error of an indicative statement preceded . . . but speaking of error as pertaining to the verdict of the intellect which is . . . the imperative statement by which the intellect commands definitively [ultimate], saying “do this!”, I say that there was in the intellect of the angel some error preceding the will, namely preceding the act of the will which is a volition, but not preceding the determination by the will; rather, this [judgment] became such by the determination of the will. (Rep. II.4.3, Fb 30vb, 80aC–bA)

In short, the angelic intellect was initially unflawed and had a correct indicative judgment of what must be willed, after which the will’s determination led to an erroneous imperative judgment, which then caused a flawed act of the will. Auriol adds that, in the last analysis, the reason why the angels could sin is because they controlled the judgments that caused their sins (Fb 30vb–31ra, 80bB–C). Auriol is silent about the content of the evil angels’ indicative and imperative judgments. Perhaps he was unable to make his dual-judgment theory fit with his theory of the nature of the angels’ sins, especially Lucifer’s, which he describes in some detail. Auriol reminds his audience that “affections of the appetite” correspond to judgments of the intellect and then describes Lucifer’s sins as matched with “inordinate,” that is, erroneous judgments. Auriol portrays these as theoretical rather than practical judgments: the first inordinate judgment was to consider his own properties to be the greatest; the second inordinate judgment was to consider the excellences of the others as smaller than they actually were. Taking himself to be the greatest resulted in vainglory, which Auriol describes as inordinate complacentia and delight in one’s own excellence, and which can be called pride, if “pride” is understood as a common trait of all seven capital vices. Taking others to be smaller than they were resulted in haughtiness (erectio appetitus), which is pride in the proper sense (as a vice on its own, distinct from the other capital vices). Haughtiness concerning someone superior (that is, God) is irreverence; haughtiness regarding those who are equal is contempt (Rep. II.6.1c., Fb 32va, 85aF–bE). How do erroneous practical judgments fit into this account of Lucifer’s sins? According to Auriol’s own assumptions, the acts of vainglory and

Voluntarist and Intermediary Accounts of the Fall

241

pride (and perhaps the theoretical judgments about Lucifer’s excellence) presuppose erroneous imperative judgments commanding these acts, and these imperative judgments presuppose correct indicative judgments. It is not clear what the correct indicative judgments would have consisted in, and how the imperative judgments could have been rendered false by a determination of the will if the corresponding indicative judgments were correct. Perhaps a correct indicative judgment, such as “excellence is worth rejoicing in,” is rendered erroneous as an imperative “rejoice in excellence!” because then it implicitly becomes an all-things-considered judgment that “it is good for me, now, to rejoice in my excellence.” In Lucifer’s case, this judgment was false, for he overestimated his excellence (he is not God). How an imperative commanding haughtiness might be matched by a correct indicative judgment is more difficult to see. In any event, the key idea in Auriol’s theory of angelic sin is that what corrupts the angelic mind is the will’s determination of the judgments that underlie their sins. In his account, determination is primitive, and thus it cannot be further explained why Lucifer had a disproportionate perception of his excellence and why he inordinately delighted and prided in himself. But this sin is not contrary to Lucifer’s own judgment; rather, it is part and parcel of his perverse judgment.

***

The approaches to angelic sin considered in this chapter are remarkably diverse, but some core ideas are common. All agree that the deficiency that caused the evil angels’ sin began in the will, not in the intellect. Thus they reject the Socratic deficiency thesis. There is furthermore widespread disagreement with Aquinas’s account, which was thought to give an inadequate explanation of how angels made an evil choice or how they made it freely. Despite the emphasis on the angels’ sin originating in the will without an antecedent cognitive deficiency, even some strict voluntarists doubt that the angels could have sinned under optimal cognitive conditions, fully aware that their act brings them no benefit but rather entails neverending misery. In fact, all the thinkers we have considered understand angelic sin as pursuit of or delight in some good. This sin may have been committed in defiance of the moral law, but not in defiance of the angels’ own expected benefit. As a pursuit of a good, the sin had a certain rationale. Whatever the rationale, voluntarists and thinkers holding an intermediate position do not think that the act is sufficiently explained by its

242

Angelic Sin

reasons, and as far as the explanation goes, it terminates in the will, not the intellect. While granting that evildoing involves error, they locate the origin of the error in the will, with Scotus holding the unique view that the error itself is not cognitive, but rather an inappropriate volitional composition.

chapter 10

Necessary (and Free?) Obstinacy

The Christian narrative of the fall of the angels pressed medieval theologians to explain two contrary phenomena: the contingency of the angels’ first choice and the necessity of their continuing commitment to this choice. In the beginning, all angels could either sin or not, but then, the good angels were “confirmed in the good” so that they could no longer sin, and the evil angels became “obstinate in evil,” which was understood as their inability to do significant good, to avoid sinning, or to escape their sinful condition. Theologians at first explained the fallen angels’ obstinacy without entering into its psychological foundation. A prominent explanation is that God decided, from considerations of justice, to end the period in which the angels could receive grace – their so-called “state of meriting” – right after the angels’ first choice, just as for humans the state of meriting ends with death. So the fallen angels or demons are obstinate because God does not bestow on them the grace of repentance and the grace of justification. This explanation is given, for example, in the Summa Halensis, by Albert the Great, and by Bonaventure, and it is typically backed with John of Damascus’s statement that “What death is to human beings, the fall is to angels.”1 Among the explanations of why God finds angelic sin inexcusable, but human sin forgivable, three stand out. One concerns the incentive for sin: as Gregory the Great remarked, the angels fell through their own wickedness, while human beings fell through that of another (an allusion to the temptation of Adam and Eve by the devil). Two further explanations are based on the superior angelic nature: the angels’ possession of godlike intellect (according to Pseudo-Dionysius) means that they sinned even though they could understand immediately what they ought to do, whereas humans can understand what they ought to do only through discursive reasoning; furthermore, according to Gregory the Great, the angels, 1

Summa Halensis II-II n. 75 sol., III: 93b–94a; Albert the Great, In Sent. II.5.6 sol., Borgnet XXVII: 120a–b; ST II.5.25.1.4 sol., Borgnet XXXII: 276a, Bonaventure, In Sent. II.7.1.1.1c., II: 175a. The quotation is from De fide orthodoxa II.4 (ch. 18) n. 15, p. 77. For an earlier discussion by Alexander of Hales, see Wierzbicki 2016; for Bonaventure’s account, see Faes de Mottoni 1991, 108–13.

243

244

Angelic Sin

as pure spirits, could have avoided sin because they are stronger than humans, who are weakened by bodily fragility.2 Thomas Aquinas is the first to seek a psychological cause of demonic obstinacy. But he and later thinkers have generally greater difficulties in giving a psychological account of demonic obstinacy than of the good angels’ confirmation. As we have seen in Part I, the latter can be explained by the will’s relation to its object. The good angels are thought to enter the beatific vision immediately after their first choice. Seeing clearly God’s goodness and each thing’s connection to God, they cannot help but love God, nor can they be attracted to anything that distances them from God. No analogous explanation is available for the obstinacy of the evil angels. With the increased focus on the moral psychology of demonic obstinacy, the fallen angels’ freedom also became a topic of discussion. Many theologians argued that their free will is preserved, even though they are not only unable to do otherwise, but also separated from the good to which their will is naturally inclined. This chapter discusses the most original contributions to the moral psychology of the fallen angels’ obstinacy, from Thomas Aquinas to William of Ockham. The issue was whether their obstinacy has an internal cause, and if so, whether it is the intellect or will. Theologians generally took God to be the external cause of demonic obstinacy, but debated whether he causes it merely by denying grace, or – as several Franciscan authors hold – by acting on the demons’ will, even causing in them an evil act or disposition. The philosophical issues involved in theories of obstinacy are the necessitation of the will by something other than the perfectly known all-encompassing good, the compatibility of free will with such necessitation, God’s causal role in evil, and even the conception of ethics.

10.1 Thomas Aquinas The difficulty of Aquinas’s novel enterprise, to explain demonic obstinacy as a psychological block, is apparent from the fact that he changes his account three times.3 2

3

These three explanations are given, for example, in Summa Halensis II-II n. 75 ad 2–ad 4, III: 94a–b, citing, respectively, Gregory the Great, Moralia in Iob IV.3.8, CCSL CXLIII: 169; Pseudo-Dionysius, DDN IV.20 in the translation of John Scotus Eriugena, Dionysiaca, I: 309; Moralia in Iob IX.50.76, p. 510. William of Auxerre spells out the Dionysius-based argument in greater detail than the Summa Halensis; see SA II.3.6, II: 66–7. For Aquinas’s accounts of the cause of demonic obstinacy, see also Dowd 2015 and D’Ercole 2017, 192–7.

Necessary (and Free?) Obstinacy

245

In the Sentences commentary, Aquinas grounds the explanation of demonic obstinacy in the notion of disposition (habitus), which fixes the fallen angels’ will on an evil end and thus prevents them from doing good acts. They can do acts that are “good from their genus” (bona ex genere), but they cannot do them with a good will. Thus whatever act they do is corrupted by an evil will, just as someone’s act of almsgiving (which is good from the genus), if done out of the desire to earn praise, is corrupted by the bad intention. Aquinas’s theory is modeled after Aristotle’s account of the intemperate. Aquinas refers to two passages; in one, Aristotle writes that the intemperate are unable to regret their intemperate action and are incurable, whereas the incontinent regret their action and are curable (EN VII.8, 1150b29–32). In another passage, Aristotle writes that vice corrupts the first principle, that is, the end (ibid., 1151a14–17). Tacitly alluding to another statement of Aristotle (EN III.5, 1114a32–b1), Aquinas clarifies that dispositions influence what people perceive as their end; he gives the example of the lustful who take sexual pleasure as their end. To explain why a corrupt end makes a person incurable, he writes that just as those who are mistaken about first theoretical principles cannot be corrected, for there is nothing more evident available for their correction, so too those who are fixed on an evil end cannot be corrected (In Sent. II.7.1.2c.). Elsewhere he spells out the analogy more fully: those who take a sin as their end put in it their happiness, so that there is nothing else they love more, for the sake of which they would abandon that sin (In Sent. II.43.1.4c.). Aquinas considers the hypothesis that the fallen angels’ disposition could be removed – whether by the fallen angels’ own efforts or by healing grace, he does not say. But he rejects this hypothesis, since after the angels made their good or evil choice, the status viae and the “state of mutability” have ended (In Sent. II.7.1.2c.). By the status viae the medievals intend the state in which human beings and angels are “wayfarers” (viatores) to their final destiny, their “homeland” (patria). In other words, the status viae is the state of meriting. By saying that the angels’ status viae has ended, Aquinas makes clear that they can no longer receive grace. But by saying that their state of mutability has ended, he presupposes what was to be explained. In later discussions, Aquinas abandons the disposition-based account of demonic obstinacy. Its biggest problem is that it contradicts his own view about the possibility of moral reform for those who are inclined to an evil end by a vicious disposition. On other occasions, when commenting on Aristotle’s claim that the intemperate are incurable – including in the Sentences commentary – Aquinas attenuates Aristotle’s claims about the

246

Angelic Sin

intemperate, writing that the intemperate are less curable than the incontinent, that they do not easily abandon their belief that pleasure is absolutely worth pursuing, and that they do not easily repent.4 Behind these discreet corrections of Aristotle is a different conception of disposition. According to Aristotle, dispositions can become so ingrained in a person’s character that the individual is beyond reform; by contrast, Aquinas holds that the use of one’s disposition is always in one’s control.5 A few years later, in De veritate, Aquinas explains angelic obstinacy by an external cause and by two internal causes. God, for reasons of justice, is its external cause, “not as effecting or conserving wickedness, but as not giving grace” (DV 24.10c., XXII: 706 ll. 224–5). The internal causes for the fallen angels’ obstinacy are their will and intellect. In the angels, there is a single appetite, the will, whereas in humans, there is also the sensory appetite. Thus the angels have an undivided appetite, and so whatever their will is inclined to, it is inclined to entirely (ibid., p. 707 ll. 303–8). The angelic intellect causes obstinacy because of its mode of cognition. Aquinas writes: Since they do not have reason but intellect, whatever they judge [aestimant], they grasp in an intellectual manner. But what is grasped intellectually, is grasped irreversibly, as when someone grasps that every whole is greater than its part. Hence the angels cannot give up the judgment that they made once and for all, whether it is true or whether it is false. (Ibid., p. 707 ll. 308–15)

This account, too, is problematic. It is hard to see how false intellectual knowledge – if there can be such a thing – is beyond correction; it would seem that the reason why intellectual knowledge is fixed is precisely because it is evidently true. Furthermore, this account implies that the evil angels sinned by an erroneous judgment. But as Aquinas clarifies later, the angels’ nondiscursive knowledge makes them exempt from error, unless they judge by a perverse will about things that are beyond the scope of their natural knowledge (ST I.58.1). Since such erroneous judgments presuppose an evil will, they cannot be what caused the angels’ first sin. For this reason, Aquinas argues later that the angels sinned because of nonconsideration rather than because of an error (ST I.63.1 ad 4; DM 16.2 ad 4). In his next attempt, in the Summa theologiae, Aquinas mentions only briefly the external cause of demonic obstinacy: after the angels’ fall, their sins are irremissible because their status has ended (ST 4 5

In Sent. II.43.1.4c.; Sententia libri Ethicorum VII.8, XLVII/2: 414 ll. 28–31; p. 415 ll. 116–19; DM 3.13c., XXIII: 94–5 ll. 73–90. For Aquinas’s account of the freedom to use one’s disposition, see Kent 2013, 106–9, and Löwe 2019; for his more optimistic view of moral reform compared to Aristotle, see Hoffmann 2018, 138–9.

Necessary (and Free?) Obstinacy

247

I.64.2c.); God does not extend his mercy to those who are incapable of penitence (ibid., ad 2). The main focus of his account in the Summa theologiae is on the internal cause of obstinacy, that is, on what makes the fallen angels incapable of penitence. Aquinas further develops the idea that this internal cause is angelic intellectual cognition. In contrast to his De veritate account, the argument here does not posit the angelic undivided desire as a distinct cause, for the appetitive power is not independent from the intellect. In fact, Aquinas builds his new explanation on an intellectualist premise: The appetitive power is completely proportioned to the cognitive power by which it is moved, like a movable to its mover. (ST I.64.2c.)

In his further explanation, Aquinas again points out that the angels’ intellectual mode of cognition causes their cognitive fixity, and since the will is bound to follow the intellect’s cognition, the angelic will is fixed as well. The gist of his explanation is this: The angel understands immovably by the intellect, just as we understand immovably the first principles. . . . The will of the angel adheres in a fixed and immovable way. Hence, as regards [the will] before the adhesion, it can freely adhere to this or that (apart from what it wills naturally); but after it has already adhered, it adheres immovably. (Ibid.)

The greatest challenge is to explain how the angels are at first free to make either choice, but not thereafter. In the Summa, he only assumes this idea. Aquinas’s most mature account, in the De malo, furnishes the missing explanation. This account adds to the previous ones the distinction between the natural and the supernatural order. As we have seen, Aquinas holds that the angels possess from their first moment all knowledge they can naturally have, but they do not know supernatural truths unless they are revealed to them (Section 8.3). On this premise, he explains the fallen angels’ obstinacy as follows: Their intellect relates immovably to everything they know naturally. And since the will is proportioned to the intellect, it follows that also their will is naturally immobile regarding the things that belong to the order of nature. They are, however, in potentiality regarding the motion toward supernatural matters [supernaturalia], by turning either toward or away from them. But since everything that accrues to a thing accrues to it according to the mode of the thing’s nature, it follows that the angels persevere immovably either in the aversion from the supernatural good or in the conversion toward it. (DM 16.5c., XXIII: 305 ll. 315–30)

248

Angelic Sin

The angels are able to choose at the beginning to turn toward or away from supernatural matters because they are in potentiality toward them. They are unable to change that choice afterwards, because they are no longer in potentiality toward them: the good angels, entering the beatific vision, behold God in actuality and love him; the evil angels, having sinned, remain confined to their natural knowledge, because they no longer deserve access to supernatural knowledge. Aquinas clarifies that God could still, by his absolute power, turn the will of the fallen angels to the good, but this would not be suitable to their nature (DM 16.5 ad 13). The nonsuitability of God overriding the fallen angels’ own decision is expressed by the thesis in the above quotation that what accrues to a thing accrues to it according to the mode of its nature. Aquinas also expresses this idea with the traditional notion that the “state of wayfarer” (another word for status viae) has ended for the angels after their first choice; hence it would go against divine wisdom to infuse in the demons the grace that would revoke their evil choice (DM 16.5c., pp. 305–6 ll. 331–7). The angels’ free decision between good and evil, that is, their being “flexible” (uertibile), is limited to the supernatural order and ceases after their first choice (DM 16.5 ad 10). Though unable to do anything other than evil, the demons persevere voluntarily in evil (DM 16.5 ad 8). (Aquinas avoids saying that they persevere freely in evil, probably to honor the Anselmian axiom that the ability to sin is not part of freedom.)6 They know that they sinned and they regret having sinned (DM 16.6 ad 10), but they regret it for the wrong reason; in fact, they do not perceive their act as evil because their mindset has not changed (DM 16.5 ad 5; 16.6 ad 8). Their attitude resembles that of the damned human beings, who according to Aquinas regret their sin not because they dislike it, but because it brought them misery rather than the happiness they expected (Compendium theologiae I.175, XLII: 150 ll. 20–32). The demons can still make other choices than those between good and evil, and so they maintain their free decision, but they sin in all their subsequent choices, for each of them is tainted by their initial choice (DM 16.5c., p. 306 ll. 337–40; 16.5 ad 17). The good angels, too, maintain their free decision. For example, they can decide with whom they communicate (ST I.107.1). They also make decisions concerning their role as guardian angels; these decisions are aided by the so-called gift of counsel (ST II-II.52.3 ad 1). 6

For Aquinas’s discussions of whether the fallen angels maintain their free decision in the Anselmian sense, according to which the ability to sin is not a part of freedom, see Alliney 2018.

Necessary (and Free?) Obstinacy

249

In Aquinas’s De malo, then, the explanation of demonic obstinacy rests on the distinction between the natural and the supernatural order. But while he claims in Article 5 that before their choice, the angels are in potentiality only regarding the supernatural order, in Article 4 he admits that they have some potentiality in the natural order. Based on his theory of undivided thinking, he stresses that the angels cannot consider their habitual knowledge all at once, but only successively, since only one intelligible species can actualize their intellect at a time (DM 16.4c., p. 299 ll. 335–45). So we may wonder why, on Aquinas’s own premises, the fallen angels cannot change their volition by changing the focus of their attention. Giles of Rome, in his lectures on the Sentences, makes precisely this objection against Aquinas’s De malo account. In his own solution, Giles returns to the idea of the Summa Halensis, Albert the Great, and Bonaventure, according to which demonic obstinacy is grounded not in angelic psychology, but solely in the divine decision to end the angels’ status viae after their first choice.7

10.2

Peter Olivi

Once Aquinas has put the problem of the psychological foundation of the fallen angels’ obstinacy on the table, even his critics will have to take it into account. Peter Olivi introduces a long discussion of obstinacy with a detailed critique of Aquinas’s explanation of the cause of obstinacy in the damned human beings and demons. Olivi summarizes his critique by reproaching Aquinas for “detracting very much from the freedom of the will,” for he holds that the will follows in effect necessarily the laws of the intellect and the sensory powers (Summa II.57 ad 29, II: 387). In Olivi’s own view, “the immobility of the will in hell” has three causes: a negative cause, which is the lack of a new divine influx (of grace) that would bring the damned to the good; a predisposing cause, which is the sin that entailed damnation; and a consummating cause, which is the stable and unmovable fall (praecipitatio) into the final misery and punishment. Olivi argues on biblical grounds that this fall involves two things, which he attributes to God’s action: God ties the damned persons and their psychological powers to a mode of existence that is vehemently repugnant to them, and God reveals to them that their damnation is irrevocable. The consequence is infinite despair, extreme wrath and hatred of the divine judge and of all who belong to him, and a perturbance of their appetite that 7

Giles of Rome, Rep. II.33c., III/2: 256–8; cf. Ord. II.7.1.2c., p. 322bA–B.

Angelic Sin

250

makes everything appear dreadful to them. Despite this divine influence (impressio) that impedes the damned from turning their psychological powers to something else, they maintain their freedom, as Olivi claims (ibid., 387–90; see also p. 378 and 57 ad 17, II: 358–9). The reason why the will of the damned remains free although it cannot will otherwise is that they move themselves to their evil consent. They do so necessarily, however, because the perturbance of their appetitive power prevents any good from appearing suitable to them; so there is no good to which they could be drawn under the aspect of good. Furthermore, God and any good of virtue appear to them as completely inaccessible (57 ad 29, II: 391–2). In binding the will of the damned human beings and angels to evil, it seems that God does something incompatible with his goodness. But Olivi clarifies that while God cannot directly cause an evil act, disposition, affection, or inclination in the damned, he can punish them in such a way that they cannot move to the good nor withdraw from evil (p. 391). Later, in discussing specifically the sin of the angels, Olivi maintains the irrevocable damnation of the fallen angels, but he doubts that it can be proved that they do not keep any ability to cooperate (with God) in doing penance (43 ad 6, I: 729). Olivi’s idea that God acts on the psychological powers of the damned so as to block them from good acts would be popular among Franciscans, as we will see.

10.3

Henry of Ghent

According to Henry of Ghent, whose discussions of the topic postdate Olivi’s Summa II.57, the evil angels’ freedom is closely connected to their obstinacy. Thus the compatibility of freedom with obstinacy is not an issue. In his view, the angels are freer than human beings precisely because after their first choice they cannot change their inclination to the ultimate end, while humans can repeatedly incline toward or away from it. The fixity of the angels’ choice is consequent upon their superior nature, by which their freedom is stronger and firmer than that of humans, for it is a mark of weakness of freedom (libertatis infirmitas) not to adhere once and for all to that which one has decided to adhere to (Summa 45.3c., XXIX: 116–18 ll. 61–72 and ll. 97–23). As is implied in this theory, the cause of obstinacy is not the fixity of the mind, as in Aquinas’s account, but rather the forcefulness of the will by which the evil angels made their initial choice. Henry elaborates on this idea a few years later in Quodlibet VIII, where he is asked whether the

Necessary (and Free?) Obstinacy

251

damned (humans and angels) can return to the good. Henry discusses distinctly the ability of the damned to attain the natural good, the moral good, the good of grace, and the good of glory (that is, the good enjoyed in the beatific vision). Their natural good is no issue, for their sin entails no harm to their nature (Quodl. VIII.11c., 323rV). Their attainment of the goods of grace and of glory is unproblematic; a decree of God’s justice determined that they will not receive them (fol. 323vY). The difficult case is the moral good. Henry narrows down the problem to the question of what keeps the damned from willing the moral good for the sake of the moral good, rather than for the sake of avoiding punishment. As he remarks, the inability to have such volitions cannot be explained by appealing to the end of the status viae, for the nature of the will does not change after it has ended. He builds his solution upon the view of unnamed “others” who hold that the more the will is free and perfect, the more perfectly it immerses itself in the willed object, and so an immaterial intellectual nature (that is, an angel or the separated soul after death) immerses itself so perfectly into the willed object that it cannot pull back (fol. 323vX).8 Henry fleshes out this theory as follows: In keeping with this statement in Proverbs 18, “When a sinner reaches the depth of sins, he is contemptuous,” it must be said that when the will is in its perfect freedom, after completion of deliberation, it hastens into the willed object of sin with great efficacy and puts in it its end. Thus when it arrives at the obstacle of the infallible moral awareness [synderesis], it does not stop there, but rather contemptuously thrusts into it, which makes [the will] blunt so that it cannot pull itself back, nor does it want to nor can it want to pull itself back – just as a sharp iron object that penetrates the flesh becomes blunt in the inner bones, and so it cannot be pulled out. But as wayfarers, human beings cannot thrust with such an efficacy [into the willed object]; hence they can always return to some morally good act. (Ibid.)

Thus Henry explains the impossibility of the damned angels or human beings attaining the good of grace or glory through an external cause (God not giving them grace and glory) and their inability of attaining the moral good with an internal cause. Whereas Aquinas locates the internal cause of obstinacy in the intellect that determines the will, Henry locates it directly in the will, namely in its forcefulness and freedom.

8

Bonaventure expresses a similar view in a brief passage, see In Sent. II.6.1.1 ad 3, II: 162b, but Henry may well have in mind someone else who develops the view more than Bonaventure.

252

Angelic Sin

10.4 John Duns Scotus As is often the case, Scotus adopts the way Henry sets up the problem but rejects his solution. In examining demonic obstinacy, Scotus, too, considers it from the perspective of different kinds of good volitions the fallen angels can still have. For Scotus, the compatibility of freedom with obstinacy is not an issue, for in his view, the demons do not lose the fundamental freedom to will otherwise. Scotus’s most elaborate treatment of demonic obstinacy is in his Ordinatio, which expands substantially on his early theory in the Lectura. His account in the Paris lectures is largely in agreement with the other two, except for God’s role in securing obstinacy.9 Scotus introduces his discussions in his Sentences commentaries with critiques of Aquinas’s account in the Summa theologiae and of Henry’s account in Quodlibet VIII. Scotus’s principal concern with Aquinas’s theory regards the assumption that the appetitive power is fully proportional to the cognitive power by which it is moved (ST I.64.2c.), which Scotus presents as the claim that the intellect is the total mover (sufficiens motor) of the will’s acts. He rejects this claim because it contradicts his own view that the will is the principal cause of its volitions (see Chapter 5, pp. 130–1). As we have seen in Chapter 9 (p. 230), Scotus also argues that under Aquinas’s premises, the angels cannot sin at all, for their intellect could only move their will rightly. Scotus moreover finds Aquinas’s view inconsistent, for the intellect cognizes equally unchangeably before and after the angels’ choice, and so the will’s act, as moved by the intellect, would be unchangeable not only after sinning, but also before (Ord. II.7 nn. 19–20, VIII: 83). (Scotus does not consider Aquinas’s De malo account, in which he solves this problem.) Against Henry, Scotus argues that the angels do not will whatever they will with maximal force. Since the will is perfectly free, it controls the force with which it wills an object; hence it can will an object more or less intensely, or not at all. Scotus also points out against Henry that a will that is perfectly free is not only free in moving toward an object, but also in resting in it. Since the angels adhere contingently to the object of their choice, they can withdraw from it (nn. 23–4, pp. 85–6).10 In his own solution, Scotus distinguishes, like Henry, different types of goodness in order to clarify which good volitions the demons can still have. A volition has natural goodness to the degree it is a being, hence the 9 10

For a detailed analysis of Scotus’s theory of demonic obstinacy, see Alliney forthcoming. For Scotus’s critiques of Aquinas and Henry, see also Lect. II.7 nn. 11–12, 19, and 21, XIX: 3 and 6; Rep. II-A.7 nn. 9 and 11, Viv. XXIII: 628a–629a.

Necessary (and Free?) Obstinacy

253

demons can have naturally good volitions simply because they can have volitions. Besides natural goodness, there are three ascending levels of moral goodness: “goodness from the genus” (bonitas ex genere), virtuous goodness, and meritorious goodness (Ord. II.7 n. 28, p. 88). Goodness from the genus does not require any suitable circumstances, such as willing or acting for a good reason, but only that the object – what is willed or done – accord with right reason. The example is almsgiving (n. 29 and n. 32, pp. 89–90). By contrast, virtuous goodness, which Scotus also calls goodness from the circumstance and goodness in the species, involves a generically good volition that additionally contains all the circumstances required according to right reason. He gives the example of almsgiving, from one’s own means, to a poor person in need, in the most suitable place, and out of love of God (nn. 30 and 32, pp. 89–90). Finally, meritorious goodness requires the previous two kinds of goodness and additionally that the volition be elicited from charity or grace, not merely from a natural inclination (nn. 31–2, pp. 89–90). Scotus admits that the fallen angels can have many volitions that are good from the genus, such as loving themselves and hating punishment (Ord. II.7 n. 41, p. 94). (Aquinas had said the same in his Sentences commentary.) Meritorious goodness receives a longer examination. He clarifies that the issue is not whether an angel can simultaneously be evil and have a meritorious volition, for when an angel has a meritorious volition, he is not evil (n. 42, p. 94). The gist of Scotus’s solution, to which he sees Holy Scripture attests, is that the only reason why the demons cannot have meritoriously good volitions is God’s decision not to give them the grace required for such volitions (for a summary statement, see n. 60, pp. 104–5). God cannot cause evil as such, and so he does not cause the demons’ obstinacy; their obstinacy can be said to be from God only as “abandoning them and not granting grace” (nn. 57–8, pp. 103–4). By God’s absolute power, which extends to everything that is not contradictory, he could provide this grace, for the demons’ will (as a power) is still receptive to it; but by his ordained power, which extends to what is feasible within the rules of God’s wisdom and will, he cannot bestow grace on them (nn. 43 and 52–4, pp. 95 and 100–1).11 Thus, concerning meritorious goodness Scotus maintains the widespread view that the demons’ obstinacy is definitive because the status viae has ended.12 11 12

That God could, by his absolute power, restore the fallen angels’ good will is affirmed also by Bonaventure, In Sent. II.7.1.1.1c. and ad 1, II: 175a and 177a, and Aquinas, DM 16.5 ad 13. For the demons’ access to meritorious volitions, see also Lect. II.7 nn. 25–32, XIX: 7–11 and Rep. II-A.7 nn. 17–20, Viv. XXII: 632a–633b. Note that after n. 19, the Wadding/Vivès edition omits ca. 460 words.

254

Angelic Sin

Scotus’s discussion of virtuous goodness – his third perspective on obstinacy – is the most interesting. He examines two perspectives: whether the fallen angels’ volitions are never virtuously good and whether they are always evil. In both regards, he is more optimistic than his contemporaries. Henry of Ghent had already remarked that the nature of the angelic will does not change after their fall; Scotus, too, claims that the natural capacity of the will remains unchanged. He cites the Pseudo-Dionysian thesis that sin does not corrupt the evil angels’ nature and argues that, since their natural inclination to the good remains intact, they can have volitions that correspond to this natural inclination, and hence that are not evil (Ord. II.7 n. 67, p. 107, citing DDN 4.23, 725C). “It would be strange,” he writes, “to deny the natural power in such an excellent nature, when nothing appears because of which it is to be denied” (n. 75, p. 111). For this reason, he rejects the view that any act of the demons is necessarily corrupted by inordinate self-love (nn. 63 and 67–9, pp. 106–7). He does not exclude that in principle the demons could have a volition that is moral according to all required circumstances; at least he considers it likely that they can have acts that are good from the genus without being corrupted by unsuitable circumstances. So Scotus admits that in principle, by acting according to their natural capacities, the fallen angels can have virtuous acts; nevertheless, he considers it more likely that in fact they act according to their “vehement viciousness [malitia]” (n. 75, p. 111).13 Scotus makes his most original contribution in discussing a view that explains demonic obstinacy by means of an evil disposition (habitus). This view claims that an evil disposition inclines the demons perfectly to evil, and hence they continually will evil (Ord. II.7 nn. 64–5, pp. 106–7). Scotus objects that the fallen angels’ obstinacy cannot be secured by a disposition. Just as the will controls the acts of other powers, such as those of the intellect, so too it controls its own acts. It follows that the will could be without any act. Since the will can suspend its acts, it does not necessarily will evil; so if the demons suspended their volitions, they would not be obstinate (n. 70, p. 108; cf. Ord. I.1.2.2 nn. 91–2, II: 66–7). (Shortly after this affirmation, Scotus expresses doubt about their ability to suspend all volitions, but admits that they that can suspend certain of them; Ord. II.7 nn. 71, 74–5, pp. 108–10). Also, a disposition does not incline a power contrarily to its characteristic “mode of action” (that is, its eliciting mode). Hence just as a nonhabituated will is a power for contraries, so too is 13

For the fallen angels’ ability to have virtuously good volitions, see also Lect. II.7 nn. 33–7, XIX: 11–12 and Rep. II-A.7 nn. 28–33, Viv. XXII: 636b–639b.

Necessary (and Free?) Obstinacy

255

a habituated will; so in neither case does the will necessarily will evil. Thus Aristotle’s dictum that the intemperate person (the evildoer, as Scotus writes) is impenitent (EN VII.8, 1150a21–2; 1150b29–30) must be interpreted as “penitent with difficulty” (n. 72, p. 109). In the Paris lectures, Scotus makes a similar observation but takes it a step further. In the Lectura and Ordinatio, he merely supposed God to have abandoned the fallen angels to their own evil. In the Paris lectures he makes the remarkable claim that, since a disposition or any other “intrinsic cause” cannot necessitate the will to have an evil volition, it is God himself who causes the fallen angels’ evil volition of loving themselves immoderately or wanting beatitude immoderately. In the Ordinatio, Scotus writes that God cannot cause evil as such. His view in the Paris lectures does not contradict this position: God only causes the “substance” of the act, not its inordinateness (inordinatio).14 God continues the angels’ good or bad volitions as he finds them at the end of their status viae. God furthermore guarantees that the fallen angels do not suspend their act of abhorring (nolle) hellfire, for if they stopped abhorring it, it would no longer be distressing (contristans) to them. (The medievals held the peculiar view that hellfire, though corporeal, torments the fallen angels and the damned souls before the resurrection of the body.) Therefore Scotus assumes that God either causes their distress (tristitia) directly, or that he conserves their act of willing against (nolitio) the hellfire as to the substance of the act. Scotus raises the objection that God would be a cause of sin, because he necessitates the demons regarding the substance of their inordinate acts. He responds that he will address the objection “at its proper place” (Rep. II-A.7 nn. 29–31, XXII: 637b–638b). But he seems not to have gotten around to doing so.

10.5

Durand of St. Pourçain

In discussing demonic obstinacy, Durand of St. Pourçain is mainly in critical conversation with Aquinas, but his treatment is also connected to Scotus, although it is not certain that he writes about obstinacy with Scotus in mind.15 Durand’s discussion is found in the first redaction of his Sentences commentary; he repeats this discussion verbatim in the third

14 15

This claim accords with Scotus’s general theory of God’s causality in acts of sin; see Chapter 7, pp. 191–2. There is evidence elsewhere in the first redaction of Durand’s Sentences commentary that he knows Scotus’s Ordinatio; see, e.g., In Sent. II-A.3.3 n. 2, vol. II/1: 129, expounding Scotus, Ord. II.3.1.7 n. 229, VII: 501.

256

Angelic Sin

redaction, but omits it from the second redaction, in which he generally avoids criticism of Aquinas.16 The scenario is the same as in Henry and Scotus: Durand asks whether the fallen angels can have acts that are good by nature, by morality, and by grace. According to Durand, all acts of the fallen angels have natural goodness, because this goodness coincides with the being of these acts. Their acts lack the goodness of grace, but this lack does not by itself make their acts sinful. The real question for Durand regards the fallen angels’ access to moral goodness, namely whether they can do acts that are good from the genus and the circumstances (In Sent. II-C.7.2 n. 5, ed. Venice 1571, 145b).17 Concerning the demons’ access to moral goodness, Durand critiques two of Aquinas’s solutions. As we have seen, in the Sentences commentary, Aquinas holds that demonic obstinacy consists in being fixed on an evil end. But Durand argues that this is inconsistent with Aquinas’s theory of how the angels sinned (with which Durand agrees; see In Sent. II.5.1 in all three redactions). In fact, according to Aquinas, the evil angels sinned not by pursuing an evil end, for they did not want something evil, but rather something good, without heeding the proper way in which it was to be obtained. Durand remarks that even if the angels were fixed on an evil end, this would not guarantee that all their acts are evil, just as those who make adultery their end can still give alms without doing so for the sake of committing adultery (In Sent. II-C.7.2 n. 7, 146a). Durand rejects also the idea in the Summa theologiae that demonic obstinacy can be grounded in immobile adherence to an object. Besides finding this theory implausible, he also argues that this hypothesis does not imply that the demons could not have any good volitions at all; it implies only that they cannot have those good volitions that conflict with the object of their choice (n. 11, p. 146b). Durand writes that neither a rational consideration nor an authoritative statement shows convincingly the impossibility of the demons willing something that is good from the genus and the circumstances. Up to this point, Durand’s theory matches Scotus’s. But he adds, contrary to Scotus’s Parisian account, that “it is entirely unsuitable to attribute this [impossibility of the demons having good volitions] to God, who cannot be the cause of someone willing badly or persisting in willing evil” (n. 12, p. 146b). Durand presents no argument for the demons’ inability to attain the moral good. He only explains their inability to attain the good of grace: by 16 17

On Durand’s different attitude toward Aquinas in the three redactions, see Koch 1927, 74. Since Durand’s In Sent. II-A.7 is still unpublished, I cite the text of q. 2 according to In Sent II-C.7. I am grateful to Thomas Jeschke for having shared his edition of In Sent. II-A.7.

Necessary (and Free?) Obstinacy

257

a divine ordinance, they cannot merit the remission of their fault, for the state of meriting ends with their fall (ibid.). In light of the difficulties of explaining the fallen angels’ obstinacy by a psychological cause, Durand thus returns to the standard account prior to Aquinas, which has no reference to their psychological conditions.

10.6 Peter Auriol Whereas Durand rejects the hypothesis that God secures the fallen angels’ obstinacy, Peter Auriol picks it up again, but handles it differently than Scotus. In his Paris lectures, Scotus said that God maintains directly a volition in the demons that is essential to their obstinacy; Auriol holds instead that God infuses in them a disposition that renders them obstinate. Although Auriol does not refer to Scotus in his treatment of obstinacy, he is omnipresent in it. Part of what motivates Auriol’s view of rooting demonic obstinacy in an external cause – God’s infusing a disposition – is the failure of Aquinas’s and Henry’s explanations by intrinsic causes, namely the fixity of intellectual cognition or the will’s thrust toward the object. Against Aquinas, Auriol argues that the demons’ intellect could be moved to a different conclusion by a more plausible reason. Auriol thinks that their obstinacy consists in the displeasure (displicentia) about God. Since such displeasure is what is most detestable, it should be easy for their intellect to judge that it is better to let go of it (Rep. II.7.2c., Fb 34rb, 90bF–91aB). As for Henry, Auriol argues that his metaphors, when analyzed, turn out to be empty (ibid., Fb 34rb–va, 91aE–F).18 Auriol’s own explanation is this: The only reason the obstinate cannot let go [of their act] is because by a just judgment of God, with the mortal sin, an extremely strong disposition of wickedness [habitus malitiae intensissimus] is infused in the will, against which the will cannot at all resist, and this is their obstinacy and continuous misery. (Rep. II.7.1c., Fb 33va, 89aC)

The disposition of wickedness causes the act in which the demons’ misery essentially consists and which is the opposite of the beatitude of the blessed: an intense displeasure about God. Paradoxically, they have complacentia about their act of displeasure, for the disposition makes it connatural to them to feel displeasure (ibid., Fb 33va, 89aD–F).

18

The word metaphysica in the Rome edition, fol. 91aE, must be read as metaphorica.

258

Angelic Sin

Auriol seems aware of Scotus’s view that a disposition cannot immobilize the will, and apparently in opposition to Scotus, he argues that a disposition can become so strong that the will cannot resist it (ibid., Fb 33rb–va, 88aE–89aA). He finds the explanation of obstinacy by means of a disposition more plausible than the idea that God acts directly on the will of the fallen angels (as Scotus argued in his Paris lectures), for it is more natural for the will to be immobilized by a disposition than by an act, and it is more typical for God to act on creatures through secondary causes than directly (Rep. II.7.2c., Fb 34va, 91bC–E). Scotus had argued that a disposition cannot necessitate the will, for it cannot change the will’s contingent eliciting mode.19 Auriol addresses an analogous problem: it seems that the nature of the will is freedom, and it also seems that freedom implies control regarding alternative possibilities. So a disposition that immobilizes the will would remove its freedom, and thus it would change the nature of the will. To address this difficulty, Auriol reminds us that freedom implies only complacentia, but complacentia does not imply control; hence freedom is compatible with necessity. The disposition that causes obstinacy removes control and contingency, but not freedom, since it causes complacentia regarding the act in which the demon is damned (Rep. II.7.1c., Fb 33vb, 89bA–F; cf. Chapter 5, pp. 147–8). So Auriol considers the demons to be free, despite their misery. Another problem Auriol mentions is that God seems to be the cause of evil, for the disposition he infuses is formally evil. Auriol solves this problem by distinguishing between three aspects of evil: what is positive in something evil (which seems to be what Scotus calls the substance of the act), evil insofar as it is culpable and imputable, and evil insofar as it is depraved (deformis) but not culpable and imputable. God can cause whatever is positive, that is, absolute in evil; thus God can even cause in a person the hatred of God – but in that case, hatred of God would not have the character of imputability and fault. The disposition that God infuses into the will of the demons lacks the character of fault; it only has the character of punishment (ibid., Fb 33vb, 89bF–90aB). Peter Olivi, Duns Scotus, and Peter Auriol argue in different ways that God directly contributes to the demons’ obstinacy: by binding their wills to evil (Olivi), by causing their immoderate self-love as to their substance (without causing their inordinateness) and by causing their abhorrence of hellfire (Scotus in his Paris lectures), and by causing their habit of wickedness (Auriol). William of Ockham follows this trend and takes it further. 19

Ord. II.7 n. 72, VIII: 109; cf. Rep. II.7 nn. 23 and 29, Viv. XXII: 634b and 637b.

Necessary (and Free?) Obstinacy

259

10.7 William of Ockham William of Ockham develops his highly original account of demonic obstinacy in response to Scotus, especially to his idea that the demons could suspend all their volitions, in which case they would not necessarily will evil (p. 254 in this chapter). In the Ordinatio, Scotus did not see any problematic implications in the hypothesis that the angels suspend their volitions; he only raised a doubt about its psychological possibility. Ockham, by contrast, thinks this hypothesis is psychologically possible, but finds it highly problematic because of two assumptions he makes: the demons’ punishment necessarily involves distress (tristitia) – taken here as a passion of the will (cf. Quodl. II.17, OTh IX: 187) – and their distress presupposes an act of the will, namely that of willing against (nolle) their punishment. So by suspending the act of willing against their punishment, the fallen angels could evade their punishment (Rep. II.15, OTh V: 338–40). We may wonder why, for Ockham, distress cannot be caused without the act of willing against the state of affairs that is experienced as distressing. Ockham expands on this in a polemic against Scotus, who holds that distress can be caused either by an act of willing against or without such an act, namely either when something is experienced as disagreeable precisely because one wills against it, or when something is naturally disagreeable or disagreeable to the senses (Ord. III.15 nn. 47–60, IX: 498–505). By contrast, Ockham argues that experience shows that distress always involves the act of willing against a state of affairs opposed to one’s wishes.20 In light of these presuppositions, Ockham formulates in three theses – which he presents cautiously “without assertion” – how he sees the demons’ distress, and hence their punishment, secured; the same three theses apply not only to the volition that causes their distress, but also to the one that causes their obstinacy. First, a damned angel necessarily has a certain act of the will; second, there is an act in a damned angel that is not in his power; third, God totally and immediately causes some act in that angel’s will. The demons’ distress is secured, because they necessarily have the act of willing against punishment; it is not in their power to suspend this volition; they cannot suspend it, because the act is caused by God alone. Possibly as an implicit rebuke of Auriol’s theory, Ockham adds that a disposition (habitus) could not necessitate the will – only God can (Rep. II.15, OTh V: 339–41). 20

QV 6.9, OTh VIII: 257 ll. 126–8; pp. 265–6 ll. 297–327; see also pp. 259–60 ll. 182–4: “Et ex sola nolitione potest causari tristitia in voluntate, puta si nolo aliquid et illud nolitum eveniat.” The editors read “. . . si nolo aliquid et illud nolitum non eveniat” with one of the four collated witnesses. My reconstruction of this clause follows another of the four witnesses.

260

Angelic Sin

(Ockham knows Book II of Auriol’s Sentences commentary21 but does not refer directly to Auriol in discussing demonic obstinacy.) Ockham’s theory of the cause of the demons’ obstinacy is analogous to his theory of what causes their distress. He conceives of the evil angels’ obstinacy in strict symmetry with the holy angels’ confirmation. According to Ockham, in the holy angels, God is the total cause of their acts of seeing and of loving God, without their intellect and will being partial causes of these acts. Their act of loving God is incompatible with any sin; hence so long as God conserves this act, they cannot sin. Analogously, in the evil angels, God causes as a total cause the act of hating God. This act is incompatible with love of God; hence so long as God conserves this act, they cannot will the opposite, and so they are obstinate. Ockham clarifies that God causes only that which is “absolute” in the act of hating God, without causing its depravity (deformitas) and evil (malitia), and so God does not cause what includes an “evil of fault” (malum culpae) (ibid., pp. 341–2). So the hypothesis that Auriol mentioned in passing, that God could cause in a person the hatred of God without that hatred having the character of fault, becomes central in Ockham’s account of obstinacy. As an alternative account, Ockham argues that God causes concurrently in the evil angel’s will the act of hating God, but not any (morally) good act, such as love; thus the evil angel is obstinate because he cannot elicit any good act. Yet by failing to cause concurrently the good act, God does not sin, for God is under no obligation; he cannot fail to do what he must do. In this account, God is the partial cause of the act that makes the fallen angels obstinate, but he is nevertheless the total cause of the acts that belong to their punishment, such as the act of willing against their punishment. Thus they can cease hating God, but they cannot stop their punishment, nor can they do any good acts (pp. 343–4). Since for Ockham the ability to do otherwise is essential to freedom in the proper sense of the word, he holds that the confirmed angels and the demons are not free regarding loving or hating God, although they remain free concerning other acts: the good angels in their office as guardian angels, the bad angels in tempting human beings (pp. 344–5). Nevertheless, while the will of the confirmed and of the fallen angels necessarily either loves or hates God, these acts are not coerced. A heavy object that is moved upward is coerced, for it is moved against its natural inclination and contrary to what it can cause naturally. But 21

See the preface to the edition of Book II of Ockham’s Reportatio, OTh V: 26*. Cf. Chapter 5, note 73.

Necessary (and Free?) Obstinacy

261

the will that is necessitated to love or hate God is not coerced, for it can cause these acts naturally, by itself, and it is not naturally inclined to their opposite (pp. 351–2, 355). The assumption that God causes in the fallen angels the hatred of himself is of course the most problematic part of Ockham’s theory. Ockham himself mentions three difficulties. According to Aristotle, certain acts have evil and depravity (malitia et deformitas) inherently connected to them; Aristotle mentions adultery, theft, and murder (EN II.6, 1107a8–17). But inherently evil acts cannot be done by God. Hatred of God is one such act; hence it cannot be caused by God. Furthermore, if hatred of God can be done by God without sin, God could command a creature to hate him, and so one could merit by hating God, which seems absurd (magnum inconveniens). Finally, hatred of God is evil, for it is done for an improper end (propter indebitum finem), and so it cannot be caused by God (pp. 347–8). Ockham’s response to the first difficulty is that hatred, theft, and adultery are bad because one is obliged by a divine precept to avoid them; yet whatever is absolute in these acts can be done by God, and if they fell under a divine precept, they could be done meritoriously by a “wayfarer” (that is, by one who is in the state of meriting). Yet if they are done meritoriously, they would have different names, for the names of theft, adultery, and hatred have the connotation of falling under a negative divine precept (p. 352). Thus Ockham has already in part addressed the second difficulty; he reiterates that as long as there is a divine precept to love God, a created will cannot hate God in a morally good way (bene odire Deum), but rather necessarily does a morally evil act. He further clarifies that while a created will is obliged by a divine precept, God is under no obligation. Thus God can cause any “absolute act” without moral evil, or cause its opposite. God can cause an act of hatred of God without moral evil, because moral evil connotes that the act violates an obligation, whereas God falls under no obligation (p. 353). In response to the third difficulty, Ockham says that when God causes the act of hating God, he does so for a good end; what is evil is to hate God for an improper end (pp. 353–4).22 When Ockham was put to trial in Avignon, what raised the censors’ concern was not his view that God causes the demons’ hatred of God, but rather his theory that hatred of God could be commanded by God and 22

These passages are often associated with the view that Ockham is a divine command ethicist. Regarding this point, see the different perspectives of Adams 1999 and Osborne 2005.

262

Angelic Sin

hence be an upright act. This claim is found in the list that John Lutterell, the former chancellor of Oxford, compiled to implicate Ockham’s teachings (Article 23); Lutterell remarks that hatred of God is against right reason and hence cannot be commanded by God.23 The first report of the Avignon process, which is based on Lutterell’s list, contains this claim as well (Article 20), and the second report repeats it (Article 5). The second report remarks that hating God is evil by itself, so if God commands it, he would be an evildoer (actor mali).24 Recall that Durand of St. Pourçain was a member of the commission, possibly its chair. As we have seen, Durand saw difficulties in explaining demonic obstinacy by intrinsic psychological causes, but he did not admit that God is the cause of the demons’ persistence in evil.

***

The hypothesis of demonic obstinacy highlights completely different difficulties about free will. Apart from theories that explain obstinacy by a stable disposition, such as Aquinas’s first account and Auriol’s account, it has no analog in Aristotle’s action theory. Indeed, the problem itself is specific to Christian theology, though it is not without philosophical relevance: once the issue is approached from the psychological rather than from the merely theological side, theologians must address psychological difficulties, even if in the end they favor an explanation that guarantees demonic obstinacy externally, by divine intervention or nonintervention. As we have seen, this topic led medieval thinkers to take new perspectives on the connection between intellect and will, the causes of the will’s act, the conditions under which the will acts necessarily, and the compatibility of freedom with necessity. The cause of obstinacy poses a challenge to intellectualists and voluntarists alike; we have seen that Aquinas, who tries to give an intellectualist explanation, struggles with this topic. But for voluntarists, there is an added layer of difficulty: according to Scotus and Ockham, the will is so free that it could even, if left to itself, suspend its act of distress at having lost eternal beatitude. To prevent this last trick of the devil, God must intervene – which poses a whole new set of difficulties. 23 24

Edited in Fritz Hoffmann 1959, 58–60. Edited in Koch 1935–6, 88–9. The reports of the Avignon process quote the claim according to Rep. IV.16, OTh VII: 352.

Conclusion

As the foregoing study has shown, beginning in the 1220s, the reception of Aristotle’s action theory by Christian thinkers enabled the development of a psychological approach to free will, which raised philosophical reflection on free will to an unprecedented level. Aristotle’s thought enabled later medieval thinkers to articulate more clearly the rational process that leads to a choice, and it provided the tools for them to develop more refined theories of the way in which decision-making is rooted in the powers of the soul. According to the close reading of Aristotle proposed by Thomas Aquinas and others, there cannot be any discrepancy between a choice and the practical judgment that concludes deliberation. But this hypothesis raised concerns about the relation between cognition and volition: Is a choice (or any other kind of volition, such as desire or enjoyment) an inevitable response to a judgment about what is worth choosing? That is, is what we choose caused by what we judge? Assuming that we judge according to our best knowledge, tracing our choices to our judgments implies that the moral quality of our choices is proportional to our knowledge – which comes down to Socrates’s intellectualism. A further, more consequential implication concerns our control of our volitions, which is essential to free decision and which was assumed to be a presupposition for moral responsibility. If our volitions are caused by our practical judgments, then we control our volitions only if we control our judgments. Our practical judgment is the conclusion of our practical deliberation, so we control our practical judgment by means of our deliberation. But how do we control our deliberation? The difficulty concerning how we control our judgments leads later medieval thinkers to take different stances on the psychological foundation of free decision. The principal question is whether the control of our choices is indeed grounded in deliberation alone, or whether the will makes a contribution to our choice-making that is not itself the result of deliberation. Strict intellectualists argue that our volitions are caused exclusively by our practical judgment, that they are in our control only 263

264

Conclusion

because of practical deliberation, and that the will does not make any contribution to our deliberation that is not itself the result of deliberation. We have seen that Godfrey of Fontaines and his students John of Pouilly and Hervaeus Natalis take this view. What they do not adequately explain is how we control whether or not to start deliberation, when to end deliberation (and hence on which judgment to settle), and how to evaluate different hypotheses considered during deliberation. At the other extreme, strict voluntarists argue that our volitions have the will as their only cause (e.g., Henry of Ghent) or at least as their principal cause (e.g., Duns Scotus’s early account) and that our volitions are in our control precisely because they are caused by the will. In their view, one deliberates about which choice to make, but the choice does not depend on the deliberation, for the will remains free to act contrary to the practical judgment that results from deliberation. What they fail to explain is why we would choose contrary to what we consider worth choosing, all things considered. Thinkers taking intermediary views deny the will the ability to choose contrary to our practical judgment, but grant the will immediate control of what causes the practical judgment. Such positions can explain the will’s choice by the judgment that causes it, but they cannot explain why earlier in the process the will privileged one judgment over another. Peter Auriol, for example, develops this position systematically. Thomas Aquinas, too, grants that deliberation does not mediate every instance of the will’s contribution to our choice-making: deliberation does not mediate every volition to begin deliberation. He traces the will to begin deliberation to a divine intervention that respects the will’s nature as a contingent cause, but he does not seem to think that the will controls the way in which God determines it to will to begin deliberation. None of the authors considered in this book followed Aquinas on this point. All these attempts have difficulty reconciling the control of one’s acts, and hence the ability to do otherwise, with the rationality of one’s acts. Rationality requires that we act for a reason; control requires that the reason be up to us. If we are in control of our reasons, or if we control whether or not we act on our reasons, there is some arbitrariness about our choices done for these reasons. A perfectly rational agent, it seems, would only act on the best reasons, and thus would not truly have alternative possibilities. Nevertheless, Christian thinkers could not admit the view that perfectly rational agents necessarily make the best choices. Nor could they in the last analysis admit that volitions are entirely accounted for by cognition. When discussing the problem of the first cause of evil and the related problem of

Conclusion

265

the sin of the angels, they must explain how something can occur in a person that cannot be traced to previous states of affairs. The first sin is a novel event: previously, there was no evil in the world. The first sin of the angels highlights particularly that what is novel in the first sin is not cognition, but volition, for it is assumed that, from the outset, the cognitive state of the angels who would sin and of those who would not did not differ. As if the problems implied in free decision – that is, in the ability to make rational choices between alternative possibilities – were not difficult enough, the medieval thinkers were confronted with an additional challenge: they had to explain why the blessed humans and holy angels in heaven necessarily love God, and the more difficult problem of why the damned in hell necessarily persist in evil. The challenge was that the same assumptions about moral psychology that are meant to explain free decision, and hence the ability to do otherwise, were supposed to explain the limited ability or the inability to do otherwise of the evil angels after they sinned. Though in discussing this topic the later medievals did not develop new insights about free decision, they brought out some of the implications of their theories of free decision and they showed inconsistencies in the theories of their adversaries. In addition, this topic gave them the opportunity to discuss voluntariness and freedom in a state of misery and permanent inability to do good. All Christian thinkers admit that rational creatures – angels and human beings – are in control of their choices. Voluntarists claim unambiguously that the control rational agents have of the psychological process that results in their choices is primitive, in other words, that their control is not fully traceable to antecedent causes. I have argued that thinkers proposing intermediary theories and even the strict intellectualists examined in this book concede this point as well – at least, implicitly – in their discussions of angelic sin. They trace choices not simply to cognition, but to the use rational agents make of their cognition; and this use is not itself fully explained by cognition. Freely made choices have an explanation – there is a story behind them – but at some point, the explanation stops at a bare fact, at a person willing something rather than not willing it, a bare fact expressed in what Scotus calls an immediate contingent proposition. But the bare fact is not the whole story. Anselm of Canterbury, and later Peter Olivi (who generally takes strong voluntarist stances), work out a possible story behind Lucifer’s fall. What is important is not its plausibility as such, but rather the fact that they describe Lucifer’s act as a rational act, even though it remains in the last analysis unexplained. In fact, there is

266

Conclusion

ultimately no contrastive explanation of why a person freely makes one choice rather than another, or why two persons who are in the same conditions make different choices (e.g., Lucifer and Michael). Nevertheless, the medieval thinkers did not think that bad choices happen because of bad luck. They happen because the agent wants them to happen, and for this reason the agent is responsible for them. As Bonaventure writes about the evil angels: “they could remain standing and not turn away, if they wanted to; and for this reason, the turning away is imputed to them as fault” (In Sent. II.5.3.1 ad 6).

Bibliography

Primary Literature Albert the Great (1890–99). B. Alberti Magni Ratisbonensis episcopi, Ordinis Praedicatorum, Opera omnia, ed. Auguste Borgnet. 38 vols. Paris: Vivès. (1951–). Alberti Magni Opera omnia, ed. Bernhard Geyer et al. Münster: Aschendorff. (Editio Coloniensis.) Alexander of Hales (1951–57). Glossa in quatuor libros Sententiarum Petri Lombardi, ed. PP. Collegii S. Bonaventurae. 4 vols. Quaracchi (Florence): Collegium S. Bonaventurae. Ambrose (1896). De Isaac uel anima, ed. C. Schenkl, 639–700. CSEL 32/1. Vienna: Tempsky; Leipzig: Freytag. Anselm of Canterbury (1938–61). S. Anselmi Cantuariensis Archiepiscopi Opera omnia, ed. Franciscus Salesius Schmitt. 6 vols. Edinburgh: Thomas Nelson & Sons. Aristotle (1831–70). Opera. Edited by the Academia Regia Borussica. 5 vols. Berlin: Reimer. Aristoteles Latinus (1972–74). Ethica Nicomachea, ed. René Antoine Gauthier. 5 vols. Aristoteles Latinus 26/1–3. Leiden: Brill; Brussels: Desclée de Brouwer. Augustine (1865). Contra Iulianum, ed. J.-P. Migne, 641–874. PL 44. (1891). De duabus animabus, ed. J. Zycha, 49–80. CSEL 25/1. Vienna: Tempsky; Leipzig: Freytag. (1892). De natura boni liber, ed. J. Zycha, 853–89. CSEL 25/2. Vienna: Tempsky; Leipzig: Freytag. (1894). De Genesi ad litteram libri duodecim, ed. J. Zycha. CSEL 28/1. Vienna: Tempsky; Leipzig: Freytag. Repr. New York: Johnson, 1970. (1902). De nuptiis et concupiscentia ad Valerium comitem libri duo, ed. K. F. Urba and J. Zycha, 207–319. CSEL 42. Vienna: Tempsky; Leipzig: Freytag. (1902). De perfectione iustitiae hominis, ed. K. F. Urba and J. Zycha, 1–48. CSEL 42. Vienna: Tempsky; Leipzig: Freytag. (1955). De ciuitate Dei libri XXII, ed. B. Dombart and A. Kalb. 2 vols. CCSL 47–48. Turnhout: Brepols. (1962). De uera religione liber unus, ed. K.-D. Daur, 169–260. CCSL 32. Turnhout: Brepols. 267

268

Bibliography

(1969). Enchiridion ad Laurentium de fide et spe et caritate, ed. E. Evans, 21–114. CCSL 46. Turnhout: Brepols. (1970). De libero arbitrio libri tres, ed. W. M. Green, 205–321. CCSL 29. Turnhout: Brepols. (1974 and 2004). Contra Iulianum opus imperfectum. 2 vols. Books I–III, ed. E. Kalinka and M. Zelzer; Books IV–VI, ed. M. Zelzer. CSEL 85/1–2. Vol. 1: Vienna: Hoelder-Pichler-Tempsky; vol. 2: Vienna: Verlag der Österreichischen Akademie der Wissenschaften. (1981). Confessionum libri XIII, ed. L. Verheijen. CCSL 27. Turnhout: Brepols. (1984). Retractationum libri duo, ed. Almut Mutzenbecher, 5–143. CCSL 57. Turnhout: Brepols. (1992). De moribus ecclesiae catholicae et de moribus Manicheorum libri duo, ed. J. B. Bauer. CSEL 90. Vienna: Hoelder-Pichler-Tempsky. (2000). De correptione et gratia, ed. Georges Folliet, 127–284. CSEL 92. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. (2001–). Enarrationes in Psalmos, ed. Franco Gori et al., CSEL 93/1A–95/5. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Averroes latinus (1562–74). Aristotelis opera cum Averrois commentariis. 12 vols. Venice: Apud Iunctas. (Editio Juntina.) Repr. Frankfurt am Main: Minerva, 1962. (1953). Averrois Cordubensis Commentarium Magnum in Aristotelis De Anima libros, ed. F. Stuart Crawford. Cambridge, MA: The Medieval Academy of America. (1969). Das neunte Buch (Θ) des lateinischen Großen Metaphysik-Kommentars von Averroes, ed. Bernhard Bürke. Bern: Francke Verlag. (1971). Averrois in librum V (Δ) Metaphysicorum Aristotelis commentarius, ed. Ruggero Ponzalli. Bern: Franke Verlag. Avicenna latinus (1968–72). Liber de anima seu sextus de naturalibus, ed. Simone Van Riet. 2 vols. Avicenna Latinus 1–2. Leuven: Peeters; Leiden: Brill. (1977–80). Liber de philosophia prima sive scientia divina, ed. Simone Van Riet. 2 vols. Avicenna Latinus 3–4. Leuven: Peeters; Leiden: Brill. Bernard of Clairvaux (1957–77). Sancti Bernardi opera, ed. J. Leclercq, C. H. Talbot, H. M. Rochais. 9 vols. Rome: Editiones Cistercienses. Boethius, Anicius Manlius Severinus (1877–80). Commentarii in librum Aristotelis Peri Hermeneias, ed. Karl Meiser, 2 vols. Leipzig: Teubner. (2005). Contra Eutychen et Nestorium. In De consolatione philosophiae, Opuscula theologica, ed. Claudio Moreschini, 206–41. 2nd edn. Munich: K. G. Saur. Bonaventure (1882–1902). Opera omnia. 10 vols. Quaracchi and Grottaferrata: Editiones Collegii S. Bonaventurae ad Claras Aquas. Chartularium Universitatis Parisiensis (1889–97). Edited by Heinrich Suso Denifle and Émile Chatelain. 4 vols. Paris: Delalain. Durand of St. Pourçain (1571). D. Durandi a Sancto Porciano, Ord. Praed. et Meldensis Episcopi, In Petri Lombardi Sententias Theologicas Commentariorum libri IIII. Venice: Guerraea. Repr. Ridgewood, NJ: Gregg Press, 1964.

Bibliography

269

(1962). Quaestiones disputatae de libero arbitrio, ed. Stella 1962, 471–99. (2012–). Durandi de Sancto Porciano Scriptum super IV libros Sententiarum, ed. Fiorella Retucci, Thomas Jeschke, et al. Leuven: Peeters. Gerard of Abbeville (1993). Quodlibeta, ed. Adriaan Pattin, L’anthropologie de Gérard d’Abbeville, 3–122. Leuven University Press. Giles of Rome (1521). In primum librum Sententiarum. Venice. Repr. Frankfurt am Main: Minerva, 1968. (1581). In secundum librum Sententiarum. 2 vols. Venice. Repr. Frankfurt am Main: Minerva, 1968. (1646). Quodlibeta. Leuven. Repr. Frankfurt am Main: Minerva, 1966. (1939–40). Aegidii Romani Quaestio de deceptione, ed. Gerardo Bruni, Analecta Augustiniana 17: 229–45. (1985–). Aegidii Romani Opera omnia, ed. Francesco Del Punta, Gianfranco Fioravanti, et al. Florence: Leo S. Olschi; Sismel – Edizioni del Galluzzo. Godfrey of Fontaines (1904). Les quatre premiers Quodlibets de Godefroid de Fontaines, ed. Maurice De Wulf and Auguste Pelzer. Les Philosophes Belges 2. Leuven: Institut Supérieur de Philosophie. (1914). Les Quodlibet cinq, six et sept de Godefroid de Fontaines, ed. Maurice De Wulf and Jean Hoffmans. Les Philosophes Belges 3. Leuven: Institut Supérieur de Philosophie. (1924–31). Le huitième, neuvième et dixième Quodlibets de Godefroid de Fontaines, ed. Jean Hoffmans. Les Philosophes Belges 4. Leuven: Institut Supérieur de Philosophie. (1932). Les Quodlibets onze–quatorze de Godefroid de Fontaines, ed. Jean Hoffmans. Les Philosophes Belges 5. Leuven: Éditions de l’Institut Supérieur de Philosophie. (1937). Le Quodlibet XV et trois Questions ordinaires de Godefroid de Fontaines, ed. Odon Lottin. Les Philosophes Belges 14. Leuven: Éditions de l’Institut Supérieur de Philosophie. Gonsalvus of Spain (1935). Quaestiones disputatae et de Quodlibet, ed. León Amorós. Quaracchi: Editiones Collegii S. Bonaventurae Ad Claras Aquas. Gregory the Great (1979–85). Moralia in Iob, ed. M. Adriaen. CCSL 143, 143A, 143B. Turnhout: Brepols. Henry of Ghent (1518). Quodlibeta. Paris: Jodocus Badius. Repr. Leuven: Bibliothèque S.J., 1961. (1520). Summa quaestionum ordinariarum. Paris: Jodocus Badius. Repr. St. Bonaventure, NY: The Franciscan Institute, 1953. (1979–). Henrici de Gandavo Opera omnia, ed. Raymond Macken et al. Leuven University Press; Leiden: Brill. Hervaeus Natalis (1513). Quodlibeta, ed. M. A. Zimara, Venice. Repr. Ridgewood, NJ, The Gregg Press, 1966. (1647). In quatuor libros Sententiarum commentaria. Paris. Repr. Ridgewood, NJ, The Gregg Press, 1966.

270

Bibliography

Hugh of St. Cher (2013). In II Sententiarum, distinctiones 24–5, ed. Saccenti 2013, 215–32. Hugh of St. Victor (2008–). Corpus Victorinum Textus historici, ed. Rainer Berndt. Münster: Aschendorff. John Duns Scotus (1891–95). Joannis Duns Scoti doctoris subtilis, ordinis minorum opera omnia. 26 vols. Paris: Louis Vivès. Repr. Westmead: Gregg International Publishers, 1969. (1950–). Doctoris subtilis et mariani Ioannis Duns Scoti Ordinis Fratrum Minorum Opera omnia, ed. C. Balić et al. Rome: Typis Vaticanis. (Editio Vaticana.) (1997–2006). Opera Philosophica, ed. Girard J. Etzkorn, Timothy B. Noone, et al. 5 vols. St. Bonaventure, NY: Franciscan Institute; Washington, DC: The Catholic University of America Press. (2004–2008). The Examined Report of the Paris Lecture: Reportatio IA. Latin Text and English Translation, ed. Allan B. Wolter and Oleg V. Bychkov. 2 vols. St. Bonaventure, NY: The Franciscan Institute. (2007). “John Duns Scotus’ Quodlibet: A Brief Study of the Manuscripts and an Edition of Question 16,” ed. Timothy B. Noone and H. Francie Roberts. In Schabel ed. 2006–2007, vol. 2: 131–98. John of Damascus (1955). De fide orthodoxa. Version of Burgundio of Pisa and Cerbanus, ed. Eligius M. Buytaert. St. Bonaventure, NY: The Franciscan Institute. John of Morrovalle (1947). Quaestio disputata “utrum obiectum voluntatis moveat ipsam ad actum volendi finem,” ed. Longpré 1947, 488–92. John of Paris (1961–64). Commentaire sur les Sentences, Reportation, ed. Jean-Pierre Muller. 2 vols. Rome: Herder. John of Pouilly (1999). Quodlibet I, q. 10, ed. Hödl 1999, 264–97. (Forthcoming). Quodlibet II, q. 11 and Quodlibet IV, q. 6 (selections), ed. Hoffmann forthcoming b, Appendix B and C. John Pecham (1948). Tractatus de anima Ioannis Pecham, ed. Gaudenzio Melani. Florence: Edizioni “Studi Francescani.” (1989). Quodlibeta Quatuor. Quodlibeta I–III ed. Girard J. Etzkorn; Quodlibet IV (Romanum) ed. Ferdinand Delorme, O.F.M., revised by Girard J. Etzkorn. Grottaferrata: Editiones Collegii S. Bonaventurae ad Claras Aquas. (2002). Quaestiones disputatae, ed. Girard J. Etzkorn, Hieronymus Spettmann, and Livarius Oliger. Grottaferrata: Editiones Collegii S. Bonaventurae ad Claras Aquas. Matthew of Aquasparta (1956). Quodlibet V q. 8, ed. Simoncioli 1956, 237–245. Peter Abelard (1987). Theologia Scholarium, ed. Eligius M. Buytaert and Constant J. Mews, 313–549. CCCM 13. Turnhout: Brepols. Peter Auriol (1596). Commentariorum in primum librum Sententiarum. Rome: Ex typographia Vaticana. books.google.com/books?id=gEPqoVz0z9MC. (1605). Commentarium in secundum librum Sententiarum. Rome. books.google.com/books?id=XvO1qQE8yU8C.

Bibliography

271

(1952). Scriptum super Primum Sententiarum. Prologue and distinctions 1–8, ed. Eligius M. Buytaert. 2 vols. St. Bonaventure, NY: The Franciscan Institute. (2012). Reportatio in Secundum librum Sententiarum d. 4, q. 3, ed. Hoffmann 2012b, 65–75. Peter Lombard (1971–81). Sententiae in IV libris distinctae. 2 vols. 3rd edn. Grottaferrata: Editiones Collegii S. Bonaventurae ad Claras Aquas. Peter Olivi (1922–26). Quaestiones in secundum librum Sententiarum, ed. Bernhard Jansen. 3 vols. Quaracchi: Editiones Collegii S. Bonaventurae ad Claras Aquas. (1964). De perfectione evangelica. Quaestio secunda, ed. Aquilino Emmen and Feliciano Simoncioli, Studi Francescani 61: 113–67. Philip the Chancellor (1985). Philippi Cancellarii Parisiensis Summa de bono, ed. Niklaus Wicki. 2 vols. Bern: Francke. Proclus (1960). Procli Diadochi Tria Opuscula (De providentia, libertate, malo), ed. Helmut Boese. Berlin: De Gruyter. Pseudo-Dionysius (1937). Dionysiaca: Recueil donnant l’ensemble des traductions latines des ouvrages attribuées au Denys de l’Aréopage, ed. Philippe Chevallier. 2 vols. Paris: Desclée de Brouwer. (1990). De divinis nominibus, ed. Beate Regina Suchla, 93–231. Berlin: De Gruyter. Radulphus Brito (2008). Le questiones di Radulfo Brito sull’‘Etica Nicomachea’, ed. Iacopo Costa. Turnhout: Brepols. Richard of Menneville (Mediavilla) (1591). Super quatuor libros Sententiarum Petri Lombardi quaestiones subtilissimae. 4 vols. Brixen. Repr. Frankfurt am Main: Minerva, 1963. (2011–14) Questions disputées, ed. Alain Boureau, 6 vols. Paris: Les Belles Lettres. Robert Grosseteste (2017). The Two Recensions of On Free Decision, ed. Neil Lewis. Oxford University Press. Siger of Brabant (1954). De necessitate et contingentia causarum, ed. Duin 1954, 14–44. (1972). Les quaestiones super librum de causis de Siger de Brabant, ed. Antonio Marlasca. Leuven: Publications universitaires; Paris: Béatrice-Nauwelaerts. (1974). Impossibilia. In Écrits de logique, de morale et de physique, ed. Bernardo Bazán, 67–97. Leuven: Publications universitaires; Paris: BéatriceNauwelaerts. (1981). Quaestiones in Metaphysicam. Édition revue de la reportation de Munich; texte inédit de la reportation de Vienne, ed. William Dunphy. Louvain-laNeuve: Éditions de l’Institut Supérieur de Philosophie. (1983). Quaestiones in Metaphysicam. Texte inédit de la reportation de Cambridge; Édition revue de la reportation de Paris, ed. Armand Maurer. Louvain-laNeuve: Éditions de l’Institut Supérieur de Philosophie. Stephen Tempier (1999). Articuli condemnati anno 1277, ed. David Piché, La Condamnation Parisienne de 1277: texte latin, traduction, introduction et commentaire. Paris: Vrin.

272

Bibliography

Summa Halensis (1924–79). 6 vols. Quaracchi: Editiones Collegii S. Bonaventurae ad Claras Aquas. Thomas Aquinas (1852–73). Sancti Thomae Aquinatis, Doctoris Angelici, Ordinis Praedicatorum Opera omnia, ad fidem optimarum editionum accurate recognita. 25 vols. Parma: Typis Petri Fiaccadori. Repr. New York: Musurgia, 1948–50. (1882–). Opera omnia, iussu Leonis XIII edita cura et studio Fratrum Praedicatorum. Rome. (Editio Leonina.) (1929–47). Scriptum super libros Sententiarum magistri Petri Lombardi Episcopi Parisiensis, ed. Pierre F. Mandonnet (vols. 1–2) and Marie Fabien Moos (vols. 3–4). Paris: Lethielleux. (1950). In librum Beati Dionysii De divinis nominibus expositio, ed. C. Pera, P. Caramello, and C. Mazzantini. Turin: Marietti. (1965). Quaestiones disputatae, ed. P. Bazzi, M. Calcaterra, T. S. Centi, E. Odetto, and P. M. Pession. 2 vols. 10th edn. Turin: Marietti. (1961–67). Summa contra Gentiles, ed. P. Marc. 3 vols. Turin: Marietti. (1972). Super Evangelium S. Ioannis lectura, ed. R. Cai, edn. Turin: Marietti. Thomas of Sutton (1977). Quaestiones ordinariae, ed. Johannes Schneider. Munich: Verlag der Bayerischen Akademie der Wissenschaften. Walter of Bruges (1928). Quaestiones disputatae du B. Gauthier de Bruges: texte inédit, ed. Ephrem Longpré. Les Philosophes Belges 10. Leuven: Institut Supérieur de Philosophie de l’Université. William de la Mare (1927). Correctorium fratris Thomae. In Les premières polémiques thomistes. I.–Le Correctorium Corruptorii “Quare,” by Palémon Glorieux. Le Saulchoir, Kain: Revue des sciences philosophiques et théologiques. William of Auvergne (1674). Guilielmi Alverni opera omnia. 2 vols. Paris, 1674. Repr. Frankfurt: Minerva, 1963. William of Auxerre (1980–87). Magistri Guillelmi Altissiodorensis Summa aurea, ed. Jean Ribailler. 5 vols. Paris: Éditions du CNRS; Grottaferata (Rome): Editiones Collegii S. Bonaventurae ad Claras Aquas. William of Ockham (1967–89). Opera philosophica et theologica. St. Bonaventure, NY: The Franciscan Institute.

Secondary Literature Adams, Marilyn McCord (1992). “St. Anselm on Evil: De casu diaboli.” Documenti e studi sulla tradizione filosofica medievale 3: 423–51. (1999). “Ockham on Will, Nature, and Morality.” In Spade ed. 1999, 245–72. Adamson, Peter (2010). “Freedom and Determinism.” In The Cambridge History of Medieval Philosophy, edited by Robert Pasnau, 399–13. Cambridge University Press. Aertsen, Jan A., Kent Emery, Jr., and Andreas Speer eds. (2001). Nach der Verurteilung von 1277: Philosophie und Theologie an der Universität von Paris im letzten Viertel des 13. Jahrhunderts. Studien und Texte. Berlin: De Gruyter.

Bibliography

273

Alliney, Guido (2002). “Velle malum ex pura libertate: Duns Scoto e la banalità del male.” Etica e Politica 4/2. www2.units.it/etica/2002_2/indexalliney .html (2005). “La ricezione della teoria scotiana della volontà nell’ambiente teologico parigino (1307–1316).” Documenti e studi sulla tradizione filosofica medievale 16: 339–404. (2013). “Giovanni di Morrovalle e le affectiones anselmiane.” Archivum Franciscanum Historicum 106: 569–84. (2015). “Landolfo Caracciolo, Peter Auriol, and John Duns Scotus on Freedom and Contingency.” Recherches de Théologie et Philosophie médiévales 82: 271–300. (2017). “Fra Aristotele e Agostino: Per un paradigma della libertà nel XIII secolo.” In Leone and Valente eds. 2017, 171–96. (2018). “Demoni e dannati: Libero arbitrio e ostinazione al male nel pensiero di Tommaso d’Aquino.” Medioevo 43: 259–83. (2020). “Quando Duns Scoto ha cambiato idea sulla volontà? La causa del volere secondo la quaestio 6 delle Collationes parisienses.” Quaderni di Noctua 7: 605–41. (Forthcoming). “Angeli mali: Ostinazione al male e libertà del bene secondo Duns Scoto.” In Will and Cognition: John Duns Scotus Reconsidered, edited by Pasquale Porro and Leonardo Sileo. Turnhout: Brepols; Bari: Edizioni di Pagina. Babcock, William S. (1991). “Augustine on Sin and Moral Agency.” In The Ethics of St. Augustine, edited by William S. Babcock, 87–113. Atlanta, GA: Scholars Press. Barnwell, Michael (2015). “The Problem with Aquinas’s Original Discovery.” American Catholic Philosophical Quarterly 89: 277–91. (2019). “The Root of Sin is Still Undiscovered: A Counter-reply to Jensen.” American Catholic Philosophical Quarterly 93: 147–60. Beonio Brocchieri Fumagalli, Mariateresa (1969). Durando di S. Porziano: Elementi filosofici della terza redazione del “Commento alle Sentenze.” Florence: La Nuova Italia. Bianchi, Luca (1999). Censure et liberté intellectuelle à l’Université de Paris (XIIIe– XIVe siècles). Paris: Les Belles Lettres. Bobillier Stève. (2017). “Éthique et personne: la volonté et le choix du mal dans la pensée de Pierre de Jean Olivi (1248–1298).” Ph.D. dissertation, Paris: École des Hautes Études en Sciences Sociales. Boler, John F. (1993). “Transcending the Natural: Duns Scotus on the Two Affections of the Will.” American Catholic Philosophical Quarterly 67: 109–26. Bonino, Serge-Thomas (2016). Angels and Demons: A Catholic Introduction, translated by Michael J. Miller. Washington, DC: The Catholic University of America Press. Brampton, C. K. (1966). “Personalities at the Process against Ockham at Avignon, 1324–26.” Franciscan Studies 26: 4–25.

274

Bibliography

Briggs, Charles F. (2016). “Life, Works, and Legacy.” In A Companion to Giles of Rome, edited by Charles F. Briggs and Peter S. Eardley, 6–33. Leiden: Brill. Brock, Stephen L. (2018). “Dead Ends, Bad Form: the Positivity of Evil in the Summa Theologiae.” In Aquinas’s Summa theologiae: A Critical Guide, edited by Jeffrey Hause, 29–46. Cambridge University Press. Brungs, Alexander (2006). “Intellekt, Wille und Willensschwäche im Korrektorienstreit.” In Hoffmann, Müller, and Perkams eds. 2006, 265–83. Burns, J. Patout (1991). “Augustine on the Origin and Progress of Evil.” In The Ethics of St. Augustine, edited by William S. Babcock, 67–85. Atlanta, GA: Scholars Press. Burr, David (1976). “The Persecution of Peter Olivi.” Transactions of the American Philosophical Society 66/5: 3–98. Byers, Sarah Catherine (2013). Perception, Sensibility, and Moral Motivation in Augustine: A Stoic-Platonic Synthesis. Cambridge University Press. Caldera, Federica (2005). “Guglielmo de la Mare tra Bonaventura, Tommaso d’Aquino e Pietro di Tarantasia: Dipendenze testuali e originalità del Commento alle Sentenze.” Archivum Franciscanum Historicum 98: 465–508. Chappell, T. D. J. (1994). “Explaining the Inexplicable: Augustine on the Fall.” Journal of the American Academy of Religion 62: 869–84. (1995). Aristotle and Augustine on Freedom: Two Theories of Freedom, Voluntary Action, and Akrasia. New York: St. Martin’s Press. Clark, Mark J. (2019) “An Early Version of Peter Lombard’s Lectures on the Sentences.” Traditio 74: 223–47. Colish, Marcia L. (1995). “Early Scholastic Angelology.” Recherches de Théologie ancienne et médiévale 62: 80–109. (2013). “Otto of Lucca, Author of the Summa sententiarum?” In Discovery and Distinction in the Early Middle Ages: Studies in Honor of John J. Contreni, edited by Cullen J. Chandler and Steven A. Stofferahn, 57–72. Kalamazoo: Medieval Institute Publications, Western Michigan University. Cory, Therese Scarpelli (2016). “Attention, Intentionality, and Mind-Reading in Aquinas’s De malo, q. 16, a. 8.” In Dougherty ed. 2016, 164–91. Costa, Iacopo (2018). “Dieu et la volonté humaine: sur les motions divines chez Thomas d’Aquin.” In Actualité de Thomas d’Aquin, edited by Olivier Boulnois, 117–34. Paris: Parole et silence. Couenhoven, Jesse (2017). “Augustine’s Moral Psychology.” Augustinian Studies 48: 23–44. Courtenay, William J. (1999). “The Academic and Intellectual Worlds of Ockham.” In Spade ed. 1999, 17–30. Decorte, Jos (1983). “Der Einfluß der Willenspsychologie des Walter von Brügge OFM auf die Willenspsychologie und Freiheitslehre des Heinrich von Gent.” Franziskanische Studien 65: 215–40. Del Punta, Francesco, Silvia Donati, and Concetta Luna (1993). “Egidio Romano.” In Dizionario biografico degli italiani vol. 42: 319–41. Rome: Istituto della Enciclopedia Italiana.

Bibliography

275

D’Ercole, Angela Federica (2017). “Il Peccato dell’Angelo: I dibattiti scolastici tra la fine del XIII e gli inizi del XIV secolo.” Ph.D. dissertation, Università degli studi di Salerno. Dobler, Emil (2000). Zwei syrische Quellen der theologischen Summa des Thomas von Aquin: Nemesios von Emesa und Johannes von Damaskus. Ihr Einfluss auf die anthropologischen Grundlagen der Moraltheologie (S. Th. I–II, qq. 6–17; 22–48). Fribourg: Universitätsverlag. Dondaine, Hyacinthe-François (1955). “Le premier instant de l’ange d’après Thomas d’Aquin.” Revue des sciences philosophiques et théologiques 39: 213–27. Dougherty, Michael V., ed. (2016). Aquinas’s Disputed Questions on Evil: A Critical Guide, Cambridge University Press. Dowd, Joseph Suk-Hwan (2015). “Aquinas on Demonic Obstinacy.” American Catholic Philosophical Quarterly 89: 699–718. Duba, William, and Chris Schabel (2017). “Remigio, Auriol, Scotus, and the Myth of the Two-Year Sentences Lecture at Paris.” Recherches de Théologie et Philosophie médiévales 84: 143–79. Dubouclez, Olivier (2014). “Plura simul intelligere: Éléments pour une histoire du débat médiéval et renaissant sur la simultanéité des actes de l’intellect.” Recherches de Théologie et Philosophie médiévales 81: 331–67. Duin, J. J. (1954). La doctrine de la providence dans les écrits de Siger de Brabant. Leuven: Éditions de l’Institut Supérieur de Philosophie. Dumont, Stephen D. (1992). “Time, Contradiction and Freedom of the Will in the Late Thirteenth Century.” Documenti e studi sulla tradizione filosofica medievale 3: 561–97. (1995). “The Origin of Scotus’s Theory of Synchronic Contingency.” The Modern Schoolman 72: 149–67. (2001). “Did Duns Scotus Change His Mind on the Will?” In Aertsen, Emery, and Speer eds. 2001, 719–94. Berlin: De Gruyter. (2018a). “James of Viterbo on the Will.” In A Companion to James of Viterbo, edited by Antoine Côté and Martin Pickavé, 249–305. Leiden: Brill. (2018b). “John Duns Scotus’s Reportatio Parisiensis Examinata: A Mystery Solved.” Recherches de Théologie et Philosophie médiévales 85: 377–438. (Forthcoming). “John Duns Scotus’s Life in Context.” In Interpreting Duns Scotus, edited by Giorgio Pini. Cambridge University Press. Eardley, Peter S. (2003). “Thomas Aquinas and Giles of Rome on the Will.” Review of Metaphysics 56: 835–62. (2006a). “The Foundations of Freedom in Later Medieval Philosophy: Giles of Rome and His Contemporaries.” Journal of the History of Philosophy 44: 353–76. (2006b). “The Problem of Moral Weakness, the Propositio Magistralis, and the Condemnation of 1277.” Mediaeval Studies 68: 161–203. (2016). “Ethics and Moral Psychology.” In A Companion to Giles of Rome, edited by Charles F. Briggs and Peter S. Eardley, 173–211. Leiden: Brill. Effler, Roy (1962). John Duns Scotus and the Principle “Omne quod movetur ab alio movetur.” St. Bonaventure, NY: The Franciscan Institute.

276

Bibliography

Ekenberg, Tomas (2016). “Voluntary Action and Rational Sin in Anselm of Canterbury.” British Journal for the History of Philosophy 24: 215–30. Emili, A. (2010). “Minio, Giovanni (Giovanni da Morrovalle).” In Dizionario biografico degli italiani vol. 74: 650–4. Rome: Istituto della Enciclopedia Italiana. Evans, G. R. (1982). Augustine on Evil. Cambridge University Press. Faes de Mottoni, Barbara (1991). “Bonaventura e la caduta degli angeli.” Doctor Seraphicus 38: 97–113. Fitzgerald, Allan D. et al., eds. (1999). Augustine through the Ages: An Encyclopedia. Grand Rapids, MI: Eerdmans. Flasch, Kurt (1991). “Freiheit des Willens: 850–1150.” In Die abendländische Freiheit vom 10. zum 14. Jahrhundert: Der Wirkungszusammenhang von Idee und Wirklichkeit im europäischen Vergleich, edited by Johannes Fried, 17–47. Sigmaringen: Jan Thorbecke Verlag. Friedman, Russell L. (2002). “The Sentences Commentary, 1250–1320: General Trends, the Impact of the Religious Orders, and the Test Case of Predestination.” In Mediaeval Commentaries on the Sentences of Peter Lombard: Current Research, vol. 1, edited by G. R. Evans, 41–128. Leiden: Brill. (2007). “Dominican Quodlibetal Literature, ca. 1260–1330.” In Schabel ed. 2006–2007, vol. 2: 401–91. Frost, Gloria (2010). “John Duns Scotus on God’s Knowledge of Sins: A Test-Case for God’s Knowledge of Contingents.” Journal of the History of Philosophy 48: 15–34. Fussenegger, Gerold (1933). “Definitiones Capituli Generalis Argentinae celebrati anno 1282.” Archivum Franciscanum Historicum 26: 127–40. Gallagher, David M. (1991). “Thomas Aquinas on the Will as a Rational Appetite.” Journal of the History of Philosophy 29: 559–84. (1994). “Free Choice and Free Judgment in Thomas Aquinas.” Archiv für Geschichte der Philosophie 76: 247–77. Gauthier, René-Antoine (1954). “Saint Maxime le Confesseur et la psychologie de l’acte humain.” Recherches de Théologie ancienne et médiévale 21: 51–100. (1970). L’Éthique à Nicomaque. 2nd edn., vol. 1: Introduction. Leuven: Publications Universitaires; Paris: Béatrice-Nauwelaerts. Gilson, Étienne (1952). “Les ‘philosophantes’.” Archives d’histoire doctrinale et littéraire du Moyen Âge 19: 135–40. Glorieux, Palémon (1935). La littérature quodlibétique. Vol. 2. Paris: Vrin. (1939). “Notations brèves sur Godefroid de Fontaines.” Recherches de Théologie ancienne et médiévale 11: 168–73. González Ayesta, Cruz (2012). “Duns Scotus on the Natural Will.” Vivarium 50: 33–52. (2014). “¿Es la voluntad un apetito o un ‘poder’? La perspectiva de Duns Escoto.” Anuario filosófico 47: 77–102. (2019). “Duns Scotus on Freedom as a Pure Perfection: Necessity and Contingency.” In Philosophy of Mind in the Early and High Middle Ages, edited by Margaret Cameron, 83–102. London: Routledge.

Bibliography

277

Goris, Harm (2012). “Angelic Knowledge in Aquinas and Bonaventure.” In Hoffmann ed. 2012, 149–85. Hamesse, Jacqueline (2006). “Theological Quaestiones Quodlibetales.” In Schabel ed. 2006–2007, vol. 1: 17–48. Hause, Jeffrey (1997). “Thomas Aquinas and the Voluntarists.” Medieval Philosophy and Theology 6: 167–82. Hille, Markus (2005). Die Pariser Verurteilung vom 10. Dezember 1270 im Spannungsfeld von Universitätsgeschichte, Theologie und Philosophie. Leipzig: St.-Benno-Verlag. Hissette, Roland (1977). Enquête sur les 219 articles condamnés à Paris le 7 mars 1277. Leuven: Publications Universitaires; Paris: Vander-Oyez. (1982). “Albert le Grand et Thomas d’Aquin dans la censure parisienne du 7 mars 1277.” In Studien zur mittelalterlichen Geistesgeschichte und ihren Quellen, edited by Albert Zimmermann, 226–46. Berlin: De Gruyter. Hocedez, Edgar (1925). Richard de Middleton: sa vie, ses œuvres, sa doctrine. Leuven: Spicilegium sacrum Lovaniense; Paris: Librairie Ancienne Honoré Champion, Édouard Champion. Hödl, Ludwig (1960). “Die Aulien des Magister Johannes von Polliaco und der scholastische Streit über die Begründung der menschlichen Willensfreiheit.” Scholastik 35: 57–75. (1999). “Die Diskussion des Johannes de Polliaco über die Lehrentscheidung der Pariser Theologen von 1285/86 ‘Non est malitia in voluntate . . . ’.” Archives d’histoire doctrinale et littéraire du Moyen Âge 66: 245–97. (2007). “The Quodlibeta of John of Pouilly († ca. 1328) and the Philosophical and Theological Debates at Paris 1307–1312.” In Schabel ed. 2006–2007, vol. 2: 199–229. Hoffmann, Fritz (1959). Die Schriften des Oxforder Kanzlers Johannes Lutterell: Texte zur Theologie des vierzehnten Jahrhunderts. Leipzig: St. BennoVerlag. Hoffmann, Tobias (2006). “Voluntariness, Choice, and Will in the Ethics Commentaries of Albert the Great and Thomas Aquinas.” Documenti e studi sulla tradizione filosofica medievale 17: 71–92. (2007). “Aquinas and Intellectual Determinism: The Test Case of Angelic Sin.” Archiv für Geschichte der Philosophie 89: 122–56. (2008). “Henry of Ghent’s Voluntarist Account of Weakness of Will.” In Weakness of Will from Plato to the Present, edited by Tobias Hoffmann, 115–37. Washington, DC: The Catholic University of America Press. (2012a). “Duns Scotus’s Intellectualist Account of Practical Knowledge.” In John Duns Scotus 1308–2008: The Opera Theologica of Scotus. Proceedings of “The Quadruple Congress” on John Duns Scotus, part 2, edited by Richard Cross, 35–52. Münster: Aschendorff. (2012b). “Peter Auriol on Practical Judgment and Angelic Sin.” In Contingenza e libertà: Teorie francescane del primo Trecento, edited by Guido Alliney, Marina Fedeli, and Alessandro Pertosa, 45–75. Macerata: Edizioni Università di Macerata.

278

Bibliography

(2012c). “Theories of Angelic Sin from Aquinas to Ockham.” In Hoffmann ed. 2012, 283–316. (2013). “Freedom Beyond Practical Reason: Duns Scotus on Will-Dependent Relations.” British Journal for the History of Philosophy 21: 1071–90. (2015). “Peter Auriol on Free Choice and Free Judgment.” Vivarium 53: 65–89. (2018). “Aquinas on Moral Progress.” In Aquinas’s Summa theologiae: A Critical Guide, edited by Jeffrey Hause, 131–49. Cambridge University Press. (2019). “Freedom without Choice: Medieval Theories of the Essence of Freedom.” In Williams ed. 2019, 194–216. (Forthcoming a). “Augustine, Thomas Aquinas, and Duns Scotus on the First Cause of Moral Evil.” In Will and Cognition: John Duns Scotus Reconsidered, edited by Pasquale Porro and Leonardo Sileo. Turnhout: Brepols; Bari: Edizioni di Pagina. (Forthcoming b). “The First Sin in the Early Fourteenth-Century Free Will Debate: Alexander of Alessandria and John of Pouilly.” In Libertés médiévales, edited by Kristell Trego. Paris: Vrin. Hoffmann, Tobias, and Peter Furlong (2016). “Free Choice.” In Dougherty ed. 2016, 56–74. Hoffmann, Tobias, and Cyrille Michon (2017). “Aquinas on Free Will and Intellectual Determinism.” Philosophers’ Imprint 17/10: 1–36. http://hdl.handle.net/2027/spo.3521354.0017.010. Hoffmann, Tobias, ed. (2012). A Companion to Angels in Medieval Philosophy. Leiden: Brill. Hoffmann, Tobias, Jörn Müller, and Matthias Perkams, eds. (2006). Das Problem der Willensschwäche in der mittelalterlichen Philosophie / The Problem of Weakness of Will in Medieval Philosophy. Leuven: Peeters. (2013). Aquinas and the Nicomachean Ethics. Cambridge University Press. Irwin, Terence (2007). The Development of Ethics. Vol. 1: From Socrates to the Reformation. Oxford University Press. Jensen, Steven (2018a). “Aquinas’s Original Discovery: A Reply to Barnwell.” American Catholic Philosophical Quarterly 92: 73–95. (2018b). Sin: A Thomistic Psychology. Washington, DC: The Catholic University of America Press. (2019). “Proto-Sin: A Case Study.” American Catholic Philosophical Quarterly 93: 161–71. Kaye, Sharon M. (2004). “Why the Liberty of Indifference Is Worth Wanting: Buridan’s Ass, Friendship, and Peter John Olivi.” History of Philosophy Quarterly 21: 21–42. Kent, Bonnie (1989). “Transitory Vice: Thomas Aquinas on Incontinence.” Journal of the History of Philosophy 27: 199–223. (1995). Virtues of the Will: The Transformation of Ethics in the Late Thirteenth Century. Washington, DC: The Catholic University of America Press. (2007). “Aquinas and Weakness of Will.” Philosophy and Phenomenological Research 75: 70–91.

Bibliography

279

(2012). “Reinventing Augustine’s Ethics: the Afterlife of City of God.” In Augustine’s City of God: A Critical Guide, edited by James Wetzel, 225–44. Cambridge University Press. (2013). “Losable Virtue: Aquinas on Character and Will.” In Hoffmann, Müller, and Perkams eds. 2013, 91–109. (2017). “Our inalienable ability to sin: Peter Olivi’s rejection of asymmetrical freedom.” British Journal for the History of Philosophy 25: 1073–92. Kent, Bonnie, and Ashley Dressel (2016). “Weakness and Willful Wrongdoing in Aquinas’s De malo.” In Dougherty ed. 2016, 34–55. Kim, Yul (2016). “Thomas Aquinas on the Divine First Motion of the Human Will.” Freiburger Zeitschrift für Philosophie und Theologie 63: 483–98. King, Peter (1994). “Duns Scotus on the Reality of Self-Change.” In Self-Motion from Aristotle to Newton, edited by Mary Louise Gill and James G. Lennox, 227–90. Princeton University Press. (2010). “Scotus’s Rejection of Anselm: The Two-Wills Theory.” In Johannes Duns Scotus 1308–2008: Die philosophischen Perspektiven seines Werkes / Investigations into His Philosophy. Proceedings of “The Quadruple Congress” on John Duns Scotus, part 3, edited by Ludger Honnefelder et al., 359–78. St. Bonaventure, NY: Franciscan Institute Publications; Münster: Aschendorff. (2012). “Augustine and Anselm on Angelic Sin.” In Hoffmann ed. 2012, 261–81. Knuuttila, Simo (1981). “Time and Modality in Scholasticism.” In Reforging the Great Chain of Being: Studies of the History of Modal Theories, edited by Simo Knuuttila, 163–257. Dordrecht: D. Reidel Publishing Company. (2004). Emotions in Ancient and Medieval Philosophy. Oxford: Clarendon Press. (2017). “Change and Contradiction in Henry of Ghent.” Vivarium 55: 22–35. Kobusch, Theo (2006). “Willensschwäche und Selbstbestimmung des Willens: Zur Kritik am abendländischen Intellektualismus bei Heinrich von Gent und in der franziskanischen Philosophie.” In Hoffmann, Müller, and Perkams eds. 2006, 249–63. Koch, Joseph (1927). Durandus de S. Porciano O.P.: Forschungen zum Streit um Thomas von Aquin zu Beginn des 14. Jahrhunderts. Münster: Aschendorff. (1935–36). “Neue Aktenstücke zu dem gegen Wilhelm Ockham in Avignon geführten Prozess.” Recherches de Théologie ancienne et médiévale 7: 353–80; 8: 79–93; 168–97. Laarmann, Matthias (1999). Deus primum cognitum: Die Lehre von Gott als Ersterkanntem des menschlichen Intellekts bei Heinrich von Gent († 1293). Münster: Aschendorff. Langevin, Paul-Émile (1960). “Nécessité ou liberté, chez Godefroid de Fontaines.” Sciences Ecclésiastiques 12: 175–203. Leone, Marialucrezia (2017a). “‘Non quaecumque necessitas excludit libertatem.’ Goffredo di Fontaines e la libertà della volontà e dell’intelletto.” Medioevo 42: 187–225. (2017b). “Utrum voluntas moveat se ipsam: Goffredo di Fontaines e l’automovimento della volontà.” In Leone and Valente eds. 2017, 259–307.

280

Bibliography

Leone, Marialucrezia, and Luisa Valente, eds. (2017). Libertà e determinismo: Riflessioni medievali. Canterano: Aracne editrice. Lewis, Neil (1996). “Power and Contingency in Robert Grosseteste and Duns Scotus.” In John Duns Scotus: Metaphysics and Ethics, edited by Ludger Honnefelder, Rega Wood, and Mechthild Dreyer, 205–25. Leiden: Brill. Longpré, Éphrem (1947). “L’œuvre scolastique du cardinal Jean de Murro, O.F.M. († 1312).” In Mélanges Auguste Pelzer: Études d’histoire littéraire et doctrinale de la Scolastique médiévale offertes à Auguste Pelzer à l’occasion de son soixante-dixième anniversaire, 467–92. Leuven: Bibliothèque de l’Université; Éditions de l’Institut Supérieur de Philosophie. Lottin, Odon (1937a). “Le libre arbitre chez Godefroid de Fontaines.” Revue néoscolastique de philosophie 40: 213–41. (1937b). “Le thomisme de Godefroid de Fontaines en matière de libre arbitre.” Revue néoscolastique de philosophie 40: 554–73. (1957a). “Libre arbitre et liberté depuis saint Anselme jusqu’à la fin du XIIIe siècle.” In Psychologie et morale aux XIIe et XIIIe siècles. 2nd edn., vol. 1: 11–389. Gembloux: Duculot. (1957b). “La psychologie de l’acte humain chez saint Jean Damascène et les théologiens du XIIIe siècle occidental.” In Psychologie et morale aux XII e et XIII e siècles, 2nd edn., vol. 1: 393–424. Gembloux: Duculot. Löwe, Can Laurens (2019). “Thomas Aquinas on Our Freedom to Use Our Habitus.” In The Ontology, Psychology and Axiology of Habits (Habitus) in Medieval Philosophy, edited by Nicolas Faucher and Magali Roques, 167–84. Cham: Springer. Lowe, Elizabeth (2003). The Contested Theological Authority of Thomas Aquinas: The Controversies between Hervaeus Natalis and Durandus of St. Pourçain. New York and London: Routledge. MacDonald, Scott (1995). “Synchronic Contingency, Instants of Nature, and Libertarian Freedom: Comments on ‘The Background to Scotus’s Theory of Will.’” The Modern Schoolman 72: 169–74. (1998). “Aquinas’s Libertarian Account of Free Choice.” Revue internationale de philosophie 52: 309–28. (1999). “Primal Sin.” In The Augustinian Tradition, edited by Gareth B. Matthews, 110–39. Berkeley: University of California Press. (2014). “Petit Larceny, the Beginning of All Sin: Augustine’s Theft of the Pears.” In Debates in Medieval Philosophy: Essential Readings and Contemporary Responses, edited by Jeffrey Hause, 5–27. London: Routledge. Macken, Raymond (1977). “Heinrich von Gent im Gespräch mit seinen Zeitgenossen über die menschliche Freiheit.” Franziskanische Studien 59: 125–82. (1995). “The Human Will as Decision-Maker in the Human Person according to the Philosophy of Henry of Ghent.” Mediaevalia, Textos e Estudos 7–8: 427–39. Maglio, Gianfranco (2016). Libero arbitrio e libertà in San Bonaventura. Padua: Wolters Kluwer.

Bibliography

281

Mandonnet, Pierre (1908–11). Siger de Brabant et l’averroïsme latin au XIIIme siècle. 2nd edn. 2 vols. Leuven: Institut Supérieur de Philosophie. Marieb, Raymond E. (1964). “The Impeccability of the Angels Regarding Their Natural End.” The Thomist 28: 409–74. Maritain, Jacques (1942). Saint Thomas and the Problem of Evil, translated by Gordon Andison. Milwaukee: Marquette University Press. (1959). The Sin of the Angel: An Essay on a Re-interpretation of Some Thomistic Positions, translated by William L. Rossner. Westminster, MD: Newman Press. Martel, Benoît (1968). La psychologie de Gonsalve d’Espagne. Montréal: Institut d’études médiévales; Paris: Vrin. Matthews, Gareth (2014). “Wanting Bad Things.” In Debates in Medieval Philosophy: Essential Readings and Contemporary Responses, edited by Jeffrey Hause, 28–37. London: Routledge. McCluskey, Colleen (2001a). “Albertus Magnus and Thomas Aquinas on the Freedom of Human Action.” In Albertus Magnus: Zum Gedenken nach 800 Jahren. Neue Zugänge, Aspekte und Perspektiven, edited by Walter Senner, 243–54. Berlin: Akademie Verlag. (2001b). “The Roots of Ethical Voluntarism.” Vivarium 39: 185–208. (2001c). “Worthy Constraints in Albertus Magnus’s Theory of Action.” Journal of the History of Philosophy 39: 491–533. (2017). Thomas Aquinas on Moral Wrongdoing. Cambridge University Press. Miteva, Evelina (2018). “I Want to Break Free: Albert the Great’s Naturalistic Account of Freedom of Choice and Its Limitations.” Bochumer Philosophisches Jahrbuch für Antike und Mittelalter 21: 11–28. Mühlenberg, Ekkehard (1969). “Das Verständnis des Bösen in neuplatonischer und frühchristlicher Sicht.” Kerygma und Dogma 15: 226–38. Müller, Jörn (2007). “Willensschwäche im Voluntarismus? Das Beispiel Heinrichs von Gent.” Archiv für Geschichte der Philosophie 89: 1–29. (2009). Willensschwäche im Denken der Antike und des Mittelalters: Eine Problemgeschichte von Sokrates bis Johannes Duns Scotus. Leuven University Press. (2018). “Der Herr und sein Diener mit der Lampe: Heinrich von Gent über Wille und Intellekt.” In Freiheit und Geschichte: Festschrift für Theo Kobusch zum 70. Geburtstag, edited by Jörn Müller and Christian Rode, 95–120. Münster: Aschendorff. Nicolas, J.-H. (1960). “La permission du péché.” Revue thomiste 60: 5–37, 185–206, 509–46. Normore, Calvin (1994). “Ockham, Self-Motion, and the Will.” In Self-Motion from Aristotle to Newton, edited by Mary Louise Gill and James G. Lennox, 291–303. Princeton University Press. (1998). “Picking and Choosing: Anselm and Ockham on Choice.” Vivarium 36: 23–39. (2003). “Duns Scotus’s Modal Theory.” In The Cambridge Companion to Duns Scotus, edited by Thomas Williams, 129–60. Cambridge University Press.

282

Bibliography

(2007a). “Freedom, Contingency, and Rational Power.” Proceedings and Addresses of the American Philosophical Association 81/2: 49–64 (2007b). “What Was Contingency.” In Laudemus viros gloriosos: Essays in Honor of Armand Maurer, edited by R. E. Houser, 283–96. South Bend, IN: University of Notre Dame Press. Oliva, Adriano (2005). “La deuxième rédaction du Correctorium de Guillaume de la Mare: les questions concernant la I Pars.” Archivum Franciscanum Historicum 98: 423–64. Opsomer, Jan, and Carlos Steel (1999). “Evil without a Cause: Proclus’ Doctrine on the Origin of Evil, and Its Antecedents in Hellenistic Philosophy.” In Zur Rezeption der hellenistischen Philosophie in der Spätantike. Akten der 1. Tagung der Karl-und-Gertrud-Abel-Stiftung vom 22.–25. September 1997 in Trier, edited by Therese Fuhrer and Michael Erler, 229–60. Stuttgart: Franz Steiner Verlag. O’Rourke, Fran (2016). “Evil as Privation: the Neoplatonic Background to Aquinas’s De malo, I.” In Dougherty ed. 2016, 192–221. Osborne Jr., Thomas M. (2005). “Ockham as a Divine-Command Theorist.” Religious Studies 41: 1–22. (2012). “William of Ockham on the Freedom of the Will and Happiness.” American Catholic Philosophical Quarterly 86: 435–56. (2013). “Giles of Rome, Henry of Ghent, and Godfrey of Fontaines on Whether to See God Is to Love Him.” Recherches de Théologie et Philosophie médiévales 80: 57–76. (2014). Human Action in Thomas Aquinas, John Duns Scotus, and William of Ockham. Washington, DC: The Catholic University of America Press. Panaccio, Claude (2012). “Intellections and Volitions in Ockham’s Nominalism.” In Emotion and Cognitive Life in Medieval and Early Modern Philosophy, edited by Martin Pickavé and Lisa Shapiro, 75–93. Oxford University Press. Pasnau, Robert (1999). “Olivi on Human Freedom.” In Pierre de Jean Olivi (1248–1298): Pensée scolastique, dissidence spirituelle et société, edited by Alain Boureau and Silvain Piron, 15–25. Paris: Vrin. (2002). Thomas Aquinas on Human Nature: A Philosophical Study of Summa theologiae 1ª 75–89. Cambridge University Press. Peck, John W. (2014). “Angelic Primal Sin: A Test Case for Aquinas’ Intellectualism.” Gregorianum 95: 105–26. Perkams, Matthias (2012). “Bernhard von Clairvaux, Robert von Melun und die Anfänge des mittelalterlichen Voluntarismus.” Vivarium 50: 1–32. Perler, Dominik (2008). “Thought Experiments: The Methodological Function of Angels in Late Medieval Epistemology.” In Iribarren and Lenz eds. 2008, 143–53. (2020). Eine Person sein: Philosophische Debatten im Spätmittelalter. Frankfurt am Main: Klostermann. Pickavé, Martin (2012). “Que signifie ‘être libre’? Le cas Henri de Gand.” Médiévales 63: 91–104.

Bibliography

283

(2013). “Aquinas on Incontinence and Psychological Weakness.” In Hoffmann, Müller, and Perkams eds. 2013, 184–202. Pini, Giorgio (2006). “Giles of Rome.” In Schabel ed. 2006–2007, vol. 1: 233–86. (2013). “What Lucifer Wanted: Anselm, Aquinas, and Scotus on the Object of the First Evil Choice.” Oxford Studies in Medieval Philosophy 1: 61–82. Piron, Sylvain (2003). “The Formation of Olivi’s Intellectual Project: ‘Petrus Ioannis Olivi and the Philosophers’ Thirty Years Later.” Oliviana 1. http:// journals.openedition.org/oliviana/8. (2006). “Franciscan Quodlibeta in Southern Studia and at Paris, 1280–1300.” In Schabel ed. 2006–2007, vol. 1: 403–38. (2020a). “Chronologie des écrits de Pierre de Jean Olivi. Première partie: avant 1279.” Oliviana 6. http://journals.openedition.org/oliviana/1035. (2020b). “Chronologie des écrits de Pierre de Jean Olivi. Deuxième partie: après l’été 1279.” Oliviana 6. http://journals.openedition.org/oliviana/1050. Porro, Pasquale (2006). “Doing Theology (and Philosophy) in the First Person: Henry of Ghent’s Quodlibeta.” In Schabel ed. 2006–2007, vol. 1: 171–231. Putallaz, François-Xavier (1995). Insolente liberté: Controverses et condamnations au XIIIe siècle. Paris: Cerf. (2000). “Entre grâce et liberté: Pierre de Jean Olivi.” In Geistesleben im 13. Jahrhundert, edited by Jan A. Aertsen and Andreas Speer, 104–15. Berlin: De Gruyter. (2010). “Censorship.” In The Cambridge History of Medieval Philosophy, edited by Robert Pasnau, 99–113. Cambridge University Press. Rigo, Caterina (2005). “Zur Redaktionsfrage der Frü hschriften des Albertus Magnus.” In Albertus Magnus und die Anfänge der Aristoteles-Rezeption im lateinischen Mittelalter: Von Richardus Rufus bis zu Franciscus de Mayronis, edited by Ludger Honnefelder et al., 325–74. Münster: Aschendorff. Rist, John M. (2014). Augustine Deformed: Love, Sin and Freedom in the Western Moral Tradition. Cambridge University Press. Ryan, Christopher J. (1983). “Man’s Free Will in the Works of Siger of Brabant.” Mediaeval Studies 45: 155–99. Saccenti, Riccardo (2013). Conservare la retta volontà: L’atto morale nelle dottrine di Filippo il Cancelliere e Ugo di Saint-Cher (1225–1235). Bologna: Società editrice il Mulino. (2016). Un nuovo lessico morale medievale: Il contributo di Burgundio da Pisa. Canterano: Aracne editrice. San Cristóbal-Sebastián, Antonio (1958). Controversias acerca de la voluntad desde 1270 a 1300. Madrid: Editorial y librería CO. CUL. S.A. Schabel, Christopher (2000). “Place, Space, and the Physics of Grace in Auriol’s Sentences Commentary.” Vivarium 38: 117–61. (2014). “John of Pouilly’s Quaestiones ordinariae de scientia Dei.” Recherches de Théologie et Philosophie médiévales 81: 237–72. (2020). “Were There Sentences Commentaries?” In Commenter au Moyen Âge, edited by Pascale Bermon and Isabelle Moulin, 243–66. Paris: Vrin.

284

Bibliography

Schabel, Christopher, ed. (2006–2007). Theological Quodlibeta in the Middle Ages. 2 vols. Leiden: Brill. Schäfer, Christian (2002). Unde malum: Die Frage nach dem Woher des Bösen bei Plotin, Augustinus, und Dionysius. Würzburg: Königshausen & Neumann. (2006). The Philosophy of Dionysius the Areopagite: An Introduction to the Structure and the Content of the Treatise on the Divine Names. Leiden: Brill. Schlosser, Marianne (2014). “Bonaventure: Life and Works.” In A Companion to Bonaventure, edited by Jay M. Hammond, J. A. Wayne Hellmann, and Jared Goff, 9–59. Leiden: Brill. Schneider, Johannes (1977). “Der Wille und seine Freiheit.” In Thomas of Sutton (1977), 165*–197*. Schöllgen, Werner (1927). Das Problem der Willensfreiheit bei Heinrich von Gent und Hervaeus Natalis. Düsseldorf: Schwann. Shanley, Brian J. (1998). “Divine Causation and Human Freedom in Aquinas.” American Catholic Philosophical Quarterly 72: 99–122. (2007). “Beyond Libertarianism and Compatibilism: Thomas Aquinas on Created Freedom.” In Freedom and the Human Person, edited by Richard Velkley, 70–89. Washington DC: The Catholic University of America Press. Simoncioli, Feliciano (1956). Il problema della libertà umana in Pietro Giovanni Olivi e Pietro de Trabibus. Milan: Vita e pensiero. Solère, Jean-Luc (2014). “Sine qua non Causality and the Context of Durand’s Early Theory of Cognition.” In Durand of Saint-Pourçain and His Sentences Commentary, edited by Andreas Speer et al., 185–227. Leuven: Peeters. Spade, Paul Vincent, ed. (1999). The Cambridge Companion to Ockham. Cambridge University Press. Spiering, Jamie Anne (2010). “An Innovative Approach to Liberum Arbitrium in the Thirteenth Century.” Ph.D. dissertation. Washington DC: The Catholic University of America. (2015). “‘What Is Freedom?’: An Instance of the Silence of St. Thomas.” American Catholic Philosophical Quarterly 89: 27–46. Stadter, Ernst (1971). Psychologie und Metaphysik der menschlichen Freiheit: Die ideengeschichtliche Entwicklung zwischen Bonaventura und Duns Scotus. Munich: Schöningh. Steel, Carlos (2002). “Avicenna and Thomas Aquinas on Evil.” In Avicenna and His Heritage: Acts of the International Colloquium, Leuven – Louvain-laNeuve, September 8 – September 11, 1999, edited by Jules Janssens and Daniel De Smet, 171–96. Leuven University Press. Stella, Propero T. (1962). “Le ‘Quaestiones de libero arbitrio’ di Durando da S. Porciano.” Salesianum 24: 450–524. Stump, Eleonore (2003). Aquinas. London: Routledge. Suarez-Nani, Tiziana (2002a). Les anges et la philosophie: subjectivité et fonction cosmologique des substances séparées à la fin du XIIIe siècle. Paris: Vrin. (2002b). Connaissance et langage des anges selon Thomas d’Aquin et Gilles de Rome. Paris: Vrin.

Bibliography

285

Szlachta, Michael (2018). “Peter John Olivi, Free Will, and the Threefold aspectus.” Recherches de Théologie et Philosophie médiévales 85: 285–307. (2020). “Thomas of Sutton’s Intellectualist Doctrine of the Will’s Self-Motion.” Oxford Studies in Medieval Philosophy 8. Teske, Roland J. (1994). “The Will as King over the Powers of the Soul: Uses and Sources of an Image in the Thirteenth Century.” Vivarium 32: 62–71. (1996). “Henry of Ghent’s Rejection of the Principle: ‘omne quod movetur ab alio movetur.’” In Henry of Ghent: Proceedings of the International Colloquium on the Occasion of the 700th Anniversary of His Death (1293), edited by Willy Vanhamel, 279–308. Leuven University Press. (2011). “Henry of Ghent on Freedom of the Human Will.” In A Companion to Henry of Ghent, edited by Gordon A. Wilson, 315–35. Leiden: Brill. Thijssen, J. M. M. H. (1998). Censure and Heresy at the University of Paris 1200–1400. Philadelphia: University of Pennsylvania Press. Torrell, Jean-Pierre (2015). Initiation à Saint Thomas d’Aquin: Sa personne et son œuvre. Paris: Cerf. Vanni Rovighi, Sofia (1964–65). “Notes sur l’influence de saint Anselme au XIIe siècle.” Cahiers de civilisation médiévale 7/28: 423–37; 8/29: 43–58. Visser, Sandra, and Thomas Williams (2004). “Anselm’s account of freedom.” In The Cambridge Companion to Anselm, edited by Brian Davies and Brian Leftow. Cambridge University Press. Weisheipl, James A. (1965). “The Principle Omne quod movetur ab alio movetur in Medieval Physics.” Isis 56: 26–45. Westberg, Daniel (1994). “Did Aquinas Change His Mind about the Will?” The Thomist 58: 41–60. Wetzel, James (2012). “Augustine on the Origin of Evil: Myth and Metaphysics.” In Augustine’s City of God: A Critical Guide, edited by James Wetzel, 167–85. Cambridge University Press. Whitehouse, Dominic (2016). “Peter of John Olivi’s Allusions to the Condemnation of 7 March 1277 in Question 57 of His Quaestiones in secundum librum Sententiarum.” Archivum Franciscanum Historicum 109: 47–98. Wieland, Georg (1981). Ethica – scientia practica: Die Anfänge der philosophischen Ethik im 13. Jahrhundert. Münster: Aschendorff. Wielockx, Robert (1985). “Commentaire.” In Giles of Rome 1985–, III/1: 67–225. (2011). “Henry of Ghent and the Events of 1277.” In A Companion to Henry of Ghent, edited by Gordon A. Wilson, 25–61. Leiden: Brill. Wierzbicki, Mateusc J. (2016). “L’irremissibilità del peccato degli angeli secondo Alessandro di Hales.” Archivum Franciscanum Historicum 109: 585–96. Williams, Thomas (1995). “How Scotus Separates Morality from Happiness.” American Catholic Philosophical Quarterly 69: 425–45. (2019). “Will and Intellect.” In Williams ed. 2019, 238–47. Williams, Thomas ed. (2019). The Cambridge Companion to Medieval Ethics. Cambridge University Press.

286

Bibliography

Wippel, John F. (1973). “Godfrey of Fontaines and the Act-Potency Axiom.” Journal of the History of Philosophy 11: 299–317. (1981). The Metaphysical Thought of Godfrey of Fontaines: A Study in Late Thirteenth-Century Philosophy. Washington, DC: The Catholic University of America Press. (1984). “Divine Knowledge, Divine Power, and Human Freedom in Thomas Aquinas and Henry of Ghent.” In Metaphysical Themes in Thomas Aquinas, 243–70. Washington, DC: The Catholic University of America Press. (1995). “Thomas Aquinas and the Condemnation of 1277.” The Modern Schoolman 72: 233–72. (1997). “Bishop Stephen Tempier and Thomas Aquinas: A Separate Process against Aquinas?” Freiburger Zeitschrift für Philosophie und Theologie 44: 117–36. (2002). “The Parisian Condemnations of 1270 and 1277.” In A Companion to Philosophy in the Middle Ages, edited by Jorge J. E. Gracia and Timothy B. Noone, 65–73. Oxford: Blackwell. (2007). “Godfrey of Fontaines and the Condemnation of March 7, 1277.” In Laudemus viros gloriosos: Essays in Honor of Armand Maurer, edited by R. E. Houser, 213–37. South Bend, IN: University of Notre Dame Press. (2016). “Metaphysical Themes in De malo, I.” In Dougherty ed. 2016, 12–33. Yrjönsuuri, Mikko (2002). “Free Will and Self-Control in Peter Olivi.” In Emotions and Choice from Boethius to Descartes, edited by Henrik Lagerlund and Mikko Yrjönsuuri, 99–128. Dordrecht: Kluwer Academic Publishers. Zavattero, Irene (2014). “La βούλησις nella psicologia dell’agire morale della prima metà del XIII secolo.” In Il desiderio nel medioevo, edited by Alessandro Palazzo, 133–50. Rome: Edizioni di storia e letteratura. (2015). “Voluntas est duplex: La dottrina della volontà dell’anonimo Commento di Parigi sull’Ethica nova e vetus (1235–1240).” Medioevo 40: 65–96. (2017). “Liberum arbitrium est facultas voluntatis et rationis: Sulla definizione di libero arbitrio all’inizio del Duecento.” In Leone and Valente eds. 2017, 143–69. (2019). “Il libero arbitrio imago della libertà divina secondo Bonaventura.” In L’uomo nel pensiero di Bonaventura da Bagnoregio, edited by Irene Zavattero. Canterano: Aracne editrice.

Index of Manuscripts

Assisi, Biblioteca Comunale, Fondo antico, 131, 35 Città del Vaticano, B.A.V., Vat. lat. 832, 66 Città del Vaticano, B.A.V., Vat. lat. 853, 103 Firenze, B.N.C., Conv. Soppr. A.III.120, 148–51, 238–40, 257–8 Firenze, B.N.C., Conv. Soppr. J.I.3, 221

Nürnberg, Stadtbibliothek, Cent. III, 75, 133, 140, 193–4, 217 Padova, Biblioteca Antoniana, 161, 148 Padova, Biblioteca Universitaria, 853, 35 Paris, BnF, lat. 14565, 84, 133–43, 216–17 Paris, BnF, lat. 15372, 84, 133–43, 216–17

287

Index

ability to do otherwise. See also control; free decision and moral responsibility. See moral responsibility compatible with conditional necessity, 110 despite inability to think otherwise, 67, 212, 230, 233 essential to free decision, 5, 41, 75, 264–5 essential to freedom, 34, 127, 260 not essential to freedom, 38, 90, 113, 258 rooted in deliberation, 37, 113, 142 rooted in sourcehood, 23, 36–7, 43–5 rooted in the will, 100, 120 synchronic vs diachronic, 89, 122–4, 129, 137, 153 Adam, 14, 19, 69–70, 84, 88, 164, 188, 193, 217, 220, 221, 225, 226, 229, 243 affection affectio commodi vs iustitiae, 16–17, 18, 121–2, 154, 227, 232–3 as aptitude for willing, 88, 104–5, 107, 130, 159, 224 akrasia. See incontinence Albert the Great, 6, 31, 35–6, 37–40, 178–81, 183, 186, 195, 201–4, 206, 207, 217, 243, 249 Alexander of Hales, 34–5, 36, 243 alternative possibilities. See ability to do otherwise Ambrose, 170 angelic sin avarice, 168 desire of equality with God, 8, 15, 177, 222, 234 envy, 7, 219, 231 hatred of God, 231, 249, 260–1 pride, 7, 165, 168–9, 207, 209, 210, 213, 219–20, 222, 224, 231, 240–1 self-love, 168, 228, 231–2 animals, nonhuman cannot sin, 206 have alternative possibilities without control, 37, 45, 66

intellectual beasts, 86–7 lack sourcehood, 23, 33, 36, 38, 57, 74, 112 lack the ability to do otherwise, 44 Anselm of Canterbury, 13–22, 41, 64, 76, 90, 103, 104, 113, 130, 131, 136, 158, 165, 175, 203, 204, 209, 223, 229, 232–3, 238, 248, 265 appetite, rational. See will Aristotle, 1–3, 13, 20, 21–30, 31, 32, 33, 35, 40, 47, 48, 55, 68, 69, 70, 73, 78, 79, 82, 89, 95, 96, 99, 105, 112, 120, 124, 135, 136, 141, 148, 150, 157, 168, 173, 180, 181, 183, 191, 200, 206, 207, 222, 236, 239, 245–6, 255, 261, 262, 263 Augustine, 7, 13, 14, 15, 19, 20, 21, 42, 61, 68, 69–70, 115, 136, 138, 139, 140, 142, 163–71, 172, 173, 175, 181, 183, 186, 187–8, 190, 193–4, 195, 206, 217, 219, 221, 222, 226, 228, 235 Augustinian deficiency thesis, 70, 216, 221 Averroes, 29–30, 69, 73, 78, 100, 108, 157, 180 Avicenna, 36, 55, 182 Bernard of Clairvaux, 13, 18–22, 32, 34, 136 Boethius, 21, 86 Boethius of Dacia, 60 Bonaventure, 6, 31, 35–6, 37–40, 52, 59, 65, 91, 178–80, 182, 183, 186, 188, 195, 219–21, 222, 225, 227, 228, 243, 249, 251, 253, 266 causa sine qua non. See object, cognized censure of 1241, 222 of 1270, 48, 59–60, 140 of 1277 (March 7), 3, 58, 60–2, 81, 85, 110, 134, 137, 138, 150, 239–40 of Giles of Rome, 62, 94–5 of Peter Olivi, 84–5 process against Durand of St. Pourçain, 145 process against Ockham, 155, 261–2 clear-eyed evildoing, 4, 225, 226–7, 229, 235, 237, 241 cognition. See also object, cognized

288

Index actual vs habitual, 26, 29, 46, 53, 96, 177, 205–6, 207, 211–13, 218, 226–7, 249, 265 angelic, 7, 17, 203, 205–6, 207–8, 212–13, 215, 216–17, 227, 228, 230, 236–8, 240, 246–7 deficient. See Augustinan deficiency thesis; Socratic deficiency thesis intellectual vs discursive, 205, 210, 243, 246–7, 257 reflexive, 44, 48, 76, 85, 112, 118 supernatural vs natural, 247–8 universal vs particular, 26, 37, 43, 47, 57, 61, 64, 81, 96–7, 203, 207, 226–7 compatibilism. See freedom condemnation. See censure conscience, 46–7, 97 consent, 24, 34, 40, 68, 70, 91, 139, 141, 143, 215, 220, 226, 250 contingency future contingents. See free decision immediate contingent propositions, 125, 265 synchronic. See ability to do otherwise control. See also ability to do otherwise; free decision essential to free decision, 15, 263 essential to moral responsibility. See moral responsibility of actual consideration, 116–18, 215, 218, 236 of deliberation, 52–3, 115–16, 142–3, 144–5, 195 of the practical judgment, 48, 98, 100, 115, 240, 263 rooted in deliberation, 37, 57, 139–42, 144, 158, 194–5, 263–4 rooted in intellect and will, 112, 194, 217 rooted in sourcehood, 15 rooted in the use of knowledge, 211–13 rooted in the will, 40, 87, 129, 152, 156 underived, 100, 117, 139, 142, 145–6, 149, 151, 157, 158–9, 213, 217, 218, 235, 238, 264, 265 control problem, 200, 211–13, 214–15, 218 Correctorium fratris Thomae, 91–2 deliberation. See also control basic account of, 25 regress problem, 50, 142, 143–4, 146, 213 determinism cosmological, 59 intellectual, 8, 28, 48, 59, 105, 108, 215, 226–7 Durand of St. Pourçain, 143, 145–6, 149, 151, 155, 159, 255–7, 262 Empedocles, 135 envy. See angelic sin error. See Augustinan deficiency thesis; Socratic deficiency thesis; propositio magistralis eudaimonism, 120

289

Eve, 164, 216, 229, 243 evil. See also clear-eyed evildoing as privation of some good, 164, 171, 176, 180, 181, 182, 183, 192, 235 as such, 171, 176, 180, 182, 183, 189, 191, 193, 260 Augustine’s dilemma, 166, 176–7, 184, 188, 189–90, 193, 199, 210 can be intended as such, 155 cannot be intended as such, 42, 63, 89, 122, 126, 171, 173, 180, 183, 208 deficient cause of, 167–9, 174, 177, 178–9, 183, 185, 191, 192–3, 195 deficient efficient cause of, 167, 179, 183, 184, 192 efficient cause of, 176–7, 183, 184 inexplicability of, 166–7, 168–9, 186, 195, 218 lacks an efficient cause, 165–7, 173, 193 per accidens cause of, 173, 175, 176, 177, 181, 183–4, 185, 189, 191, 192–3, 194, 195 subject of, 170, 172, 176, 180, 182, 183, 258 free decision (liberum arbitrium). See also ability to do otherwise; control as a power distinct from intellect and will, 39–40 definition, 13–14, 18, 20–1, 32, 34, 41 free will in the broad vs narrow sense, 4–5, 32, 41, 64, 113 in fallen angels, 248, 252, 260 regards future contingents, 21, 89 regards present contingents. See ability to do otherwise, synchronic vs diachronic rooted in immateriality, 33, 36–7, 40, 43, 64, 112–13 ultimate inexplicability of free acts, 4, 18, 53, 125, 145, 159, 166–7, 186, 195, 213–14, 218, 241, 265–6 free judgment, 21, 44–5, 57, 141–2 free will. See free decision; freedom freedom and ability to sin, 3, 13–14, 41, 64, 90–1, 130, 158, 248 and contingency, 127, 128–9, 148, 157, 258 and the good, 42, 64, 74, 89–90, 147–8 compatible with necessity, 5, 19, 32, 38, 64, 73–5, 90, 113, 128–9, 148, 244, 250, 258, 262 definition (ratio libertatis), 32, 34, 74, 90, 112, 130, 147 existence for one’s own sake (causa sui), 112, 148 from coercion, 18–19, 21, 32, 157, 158 from inevitability, 32, 157 from misery, 18–19, 157 from necessity. See freedom from coercion from sin, 18–19, 42, 157, 158 in the beatific vision, 14, 19, 38, 41, 90, 114, 125, 126–7, 148, 154, 158

290

Index

freedom (cont.) in the state of obstinacy, 32, 90, 114, 248, 250, 258 innate to the will, 67, 71, 87–8, 91, 100–1, 111, 121, 128, 233 true freedom, 21, 42 ultimate inexplicability of free acts. See free decision Gerard of Abbeville, 58 Giles of Rome, 60, 62–3, 66, 80, 94–103, 104, 105, 108, 112, 117, 159, 224, 235–8, 249 God divine justice, 170 divine self-love, 73–4, 90, 128 hatred of, 155, 231, 249, 258, 260–2 love of God by the blessed, 38, 41, 73, 90, 125, 127, 128, 154, 155 Godfrey of Fontaines, 3, 27, 51, 54, 55, 57, 71, 75, 93–4, 102, 103–4, 105–18, 126, 130–1, 132, 133, 136, 137, 138, 139, 143, 144, 147, 148–9, 156–7, 158, 174, 193, 194, 214–16, 218, 264 Gonsalvus of Spain, 91, 94, 131, 137, 216 Gregory the Great, 243–4 happiness desired necessarily, 17, 42, 64, 72, 90, 113, 134, 135, 141 desired not necessarily, 125–6, 134, 154 desired sinfully, 18, 209, 231–2, 234–5, 236–8, 255 Henry of Friemar, 142–3 Henry of Ghent, 27, 30, 42, 46, 49, 59, 60, 62, 63–84, 85, 88, 90, 91–2, 93–4, 97, 98, 99, 100, 102–3, 104–6, 109, 111, 113–14, 119, 120, 122, 124, 126, 128, 129, 130–1, 133, 136–7, 138–9, 140, 147, 148, 154, 158, 187–8, 190, 193–4, 195, 216, 222–5, 239, 250–1, 252, 254, 256, 257, 264 Hervaeus Natalis, 138, 143–5, 146, 149, 151, 158, 264 Hugh of St. Cher, 35 Hugh of St. Victor, 21, 89 ignorance. See also Socratic deficiency thesis; propositio magistralis acting in ignorance vs due to ignorance, 96–8, 237–8 blameless ignorance dilemma, 199, 210 blameless ignorance problem, 200, 204, 226 cause of sin, 26, 46, 208 every evildoer is ignorant, 26, 48, 70, 96, 156, 207, 222, 239 nonculpable, 68, 199–200 of choice, 48

incontinence, 25–6, 46–7, 68, 70, 82, 96, 110, 136, 156, 207, 226 clear-eyed. See clear-eyed evildoing infallibility problem, 199, 200–1, 202–3, 208–11, 217 intellect enjoys freedom, 111, 112 entails will, 42–3 identical with the will, 33, 152 infallibility of, 200–1, 202–3, 205, 219 lacks control, 67, 87, 102, 230, 233 lacks freedom, 71, 121, 152 self-motion of, 50 intellectualism (definition), 5 James of Viterbo, 94, 105 John Baconthorpe, 134 John Duns Scotus, 22, 30, 72, 75, 81, 85, 89, 90, 91, 119–33, 135, 137, 145, 148, 151–2, 153, 154, 158, 185, 189, 190–3, 195, 214, 230–5, 242, 252–6, 257–8, 259, 262, 264, 265 John Lutterell, 262 John of Damascus, 21, 23–4, 31, 32, 33, 36, 57, 64, 74, 141, 150, 157, 243 John of la Rochelle, 35, 36 John of Morrovalle, 85, 88, 94, 103–5, 106, 107, 130, 158, 224 John of Pouilly, 63, 84, 144, 146, 149, 151, 156–7, 158, 193–5, 215–18, 264 John Pecham, 84, 91, 221–2, 225, 236 John Quidort, 143, 191 John XXII, 85, 155 judgment-volition conformity, 45–8, 50, 57, 61–2, 66–8, 72, 84, 88, 98, 109–11, 122, 133–7, 150–1, 156–7, 215, 216 knowledge. See cognition Lucifer, 7–8, 13, 15–16, 17–18, 84, 164, 177, 203, 207, 209, 210–11, 213–14, 220, 221, 222, 224–6, 228–9, 231–2, 234, 237–8, 239, 240–1, 265–6 Manichaeism, 163, 170, 178 Matthew of Aquasparta, 91 Maximus Confessor, 23 moral responsibility asymmetrical account, 114–15 requires access to alternative motivations, 15–17 requires control, 3, 18, 22, 114, 122–3, 138–9, 140, 156, 211, 263 requires rationality, 17 requires sourcehood, 5, 15–17, 23, 41, 114 requires the ability to do otherwise, 5, 36, 38, 40, 41, 64, 114, 138–9, 169–70, 200, 211, 232

Index symmetrical account, 90–1, 155, 201 motion principle, 29–30, 99, 132 does not apply to spiritual things, 76 does not apply to the will, 153 necessity absolute vs conditional, 52, 56, 110, 113, 114, 141, 149 natural, 73, 75, 87, 90, 121, 127, 129, 130, 148–9 of coercion, 32, 38, 73, 113 of immutability, 32, 38, 64, 73–4, 113, 134, 141, 157 of the present, 55–6, 89, 124 negligence, 86, 118, 140, 186–7, 194–5, 210–11, 215, 217 Nemesius of Emesa, 23, 141 nonconsideration, 26, 46, 177–8, 185–7, 188, 190, 194, 195, 207–11, 215, 216, 218, 221, 235, 238–9, 246 object, cognized causa sine qua non, 69, 71, 80, 107, 130–1 efficient cause, 99–100, 108, 109, 111, 137, 139, 144, 148, 215 final cause, 51–2 formal cause, 51–2 partial cause, 130–1, 152, 153 terminating cause, 88 Peter Abelard, 21 Peter Auriol, 73, 146–52, 159, 238–41, 257–8, 259–60, 262, 264 Peter Lombard, 6–8, 13, 20–2, 175–6 Peter Olivi, 6, 22, 84–91, 92, 103, 109, 114, 123, 124, 126, 130, 158, 225–30, 231, 232, 249–50, 258, 265 Philip the Chancellor, 31, 33–5, 36–7, 39, 152, 178, 200–2, 203, 206, 207, 217, 219 power active vs passive, 26–7, see also will free vs natural, 120–1, 152 rational vs irrational, 26–7, 120–1, 180 pride. See angelic sin Proclus, 171, 173 propositio magistralis, 62–3, 81–3, 85, 95, 97–8, 110, 133, 236 Pseudo-Dionysius the Areopagite, 42, 95, 171–3, 176–7, 178, 180–1, 183, 202, 205, 243–4, 254 psychological turn, 1, 31, 33, 58 Radulphus Brito, 66 reason. See intellect Richard Klapwell, 92 Richard of Menneville (Mediavilla), 27, 91, 185, 188–91, 193, 195

291

Robert Grosseteste, 24, 124 Roger Bacon, 84 self-determination, 73, 98, 99–101, 102, 108, 115, 122, 129, 146, 157 self-motion. See also motion principle as equivocal causation, 79, 131–2 by means of affections, 18, 104 of animals, 29, 76 of heavy and light things, 29–30, 78, 132 of the will, 18, 28–9, 30, 38–9, 49–50, 51, 69, 71, 72, 75, 76, 77–80, 85, 87, 90, 91, 99, 103, 104, 107–8, 114, 131–2, 137, 149, 213, 214–15, 223–4 per accidens, 28, 75, 111, 149 per se, 28–9, 50, 75–6, 99, 102, 106–8, 137, 145, 146, 148, 153, 214–15 virtual/formal distinction, 77–9, 107, 132, 137, 148 self-reflexivity of the intellect. See cognition of the will, 38, 44, 76, 85, 223 Siger of Brabant, 1, 31, 54–7, 60, 62, 66, 105, 135 sin. See also angelic sin; freedom implies moral responsibility, 1, 3, 20 of Adam and Eve, 70, 84, 88, 164, 188, 193, 216–17, 220, 221, 225, 226 of omission, 168, 186–7, 210–11, 221, 236, 237, see also negligence original sin, 20, 34 Socrates, 25–6, 46, 263 Socratic deficiency thesis, 46–7, 70, 72, 81, 88, 98, 110, 199, 209, 216, 241 sourcehood. See also ability to do otherwise; control; moral responsibility necessary condition for freedom, 15, 104, 112, 169 rooted in cognition, 36–7, 43–4, 57, 72, 112, 113–14 rooted in self-motion, 38, 90 rooted in the will, 73–5 sufficient condition for freedom, 38, 40, 41, 90, 114 Stephen Tempier, 48, 58, 59–60, 62 Summa Halensis, 31, 35–40, 45, 52, 178, 179, 243, 244, 249 Thomas Aquinas, 1, 3, 5, 22, 27, 30, 31, 35, 36, 37, 38, 40–54, 56, 57, 58, 60, 61, 63–5, 66, 67, 71, 72–3, 75–6, 80, 81, 87, 89, 90, 91–2, 94, 95, 99, 100–1, 104, 105, 109, 112, 113, 118, 119, 123, 126, 128, 131, 142, 143, 144, 145, 147, 150, 154, 156, 159, 171, 178, 180, 181, 182–7, 188, 189, 190, 191, 193, 194, 195, 204–14, 215, 217–18, 221, 224, 225, 226, 227, 230, 232, 233, 234,

292

Index

235–6, 238–9, 241, 244–9, 250, 251, 252, 253, 255–6, 257, 262, 263, 264 Thomas of Sutton, 143 voluntarism (definition), 5 Walter of Bruges, 27, 36, 59, 65, 71, 77, 84, 85, 91–2 will. See also judgment-volition conformity; propositio magistralis; self-motion act or disposition, 165 act, disposition, or faculty, 18, 175 active power, 71, 88, 120–1, 129, 130, 140 actuation vs determination, 99–103, 108 affectio commodi vs iustitiae. See affection cannot be coerced, 21, 51 cause of volitions. See object, cognized; selfmotion desire-love vs friendship-love, 231 divine motion of, 50–1, 76, 213 eliciting mode of, 72, 120–1, 127–9, 254, 258 exercise vs specification, 49–51, 80, 100, 104, 116–17, 139, 145–6, 213

first mover, 65, 88 identical with the intellect, 33, 152 inferior to the intellect, 111 nature vs will, 75, 120–1, 129, 152 necessary acts of, 38, 41, 42, 45, 49, 64, 73–4, 90, 107, 128, 129, 134, 135, 247–8, 249–50, 251, 257–8, 259–60 ordered to the good, 42, 46, 63, 89, 109, 122, 135, 208 passive power, 48, 59, 100, 111 rational appetite, 23, 33–5, 39, 40, 45, 64–5, 87, 109, 119–20, 121–2, 134, 232–3 subject of sin, 220–1 superior to the intellect, 65, 71, 87 vis collativa, 233–4 William de la Mare, 91–2 William of Auvergne, 3, 6, 22, 65, 222 William of Auxerre, 31–3, 34, 35, 38, 39, 176–8, 180, 181, 195, 227, 244 William of Ockham, 75, 91, 129, 151–8, 174, 219, 258–62