Oxford Studies in Medieval Philosophy, Volume 4 [1 ed.] 9780198786368, 9780198786375, 0198786360

Citation preview

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

OXFORD STUDIES IN MEDIEVAL PHILOSOPHY VOLUME 4

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

ADVISORY BOARD Marilyn McCord Adams, Rutgers University & Australian Catholic University Peter Adamson, Ludwig-Maximilians University, Munich Peter King, University of Toronto Henrik Lagerlund, University of Western Ontario John Marenbon, Trinity College, Cambridge Calvin Normore, University of California, Los Angeles Dominik Perler, Humboldt University, Berlin Eleonore Stump, St. Louis University Editorial Assistant Tyler Huismann, University of Colorado

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Oxford Studies in Medieval Philosophy Volume 4 Edited by

ROBERT PASNAU

1

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

3

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © the several contributors 2016 The moral rights of the authors have been asserted First Edition published in 2016 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available ISBN 978–0–19–878636–8 (hbk) ISBN 978–0–19–878637–5 (pbk) Printed in Great Britain by Clays Ltd, St Ives plc Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Contents Articles Modern Toleration through a Medieval Lens: A “Judgmental” View Cary J. Nederman On a Possible Argument for Averroes’s Single Separate Intellect Stephen R. Ogden Scotus on Universals John Hawthorne Duns Scotus, the Natural Law, and the Irrelevance of Aesthetic Explanation Jeff Steele Al-Taftāzānī on the Liar Paradox David Sanson and Ahmed Alwishah Suárez’s Non-Reductive Theory of Efficient Causation Jacob Tuttle How Not to Be a Truthmaker Maximalist: Francisco Peinado on Truthmakers for Negative Truths Brian Embry

1

27 64

78 100 125

159

Critical Notice Reconstructing Aquinas’s World: Themes from Brower Thomas M. Ward

184

Discussion Don’t Mind the Gap: A Reply to Adam Wood Turner C. Nevitt

198

Briefly Noted

215

Notes for Contributors Index

217 219

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Modern Toleration through a Medieval Lens A “Judgmental” View Cary J. Nederman

During the past twenty years or so, a remarkable development has occurred in the study of medieval intellectual history, namely, the realization that numerous writings from the twelfth century onward yielded an array of defenses of conceptually coherent and philosophically well-grounded principles of religious, social, and cultural toleration.1 Included in this panoply are works by some of the best known figures in the theology and philosophy 1 Among the contributors to this development are, in chronological order: chapters by Cary J. Nederman, Stephen Lahey, Kate Langdon Forhan, and Paul J. Cornish contained in Cary J. Nederman and John Christian Laursen (eds), Difference and Dissent: Theories of Toleration in Medieval and Early Modern Europe (Lanham, MD: Rowman & Littlefield, 1996); István Bejczy, “Tolerantia: A Medieval Concept,” Journal of the History of Ideas 58 (1997), 365–84; chapters by Georg Wieland, Rainer C. Shwinges, Klaus Schreiner, and Alexander Patschovsky in A. Patschovasky and H. Zimmermann (eds), Toleranz im Mittelalter (Sigmaringen: Thorbecke, 1998); chapters by Constant J. Mews, Cary J. Nederman, and Gary Remer in J. C. Laursen and C. J. Nederman (eds), Beyond the Persecuting Society: Religious Toleration Before the Enlightenment (Philadelphia: University of Pennsylvania Press, 1998); Cary J. Nederman, Worlds of Difference: European Discourses of Toleration, c.1100–c.1550 (University Park, PA: Pennsylvania State University Press, 2000); John Christian Laursen and Cary J. Nederman, “Problems of Periodization in the History of Toleration,” Storia della Storiografia 37 (2000), 55–65; the contributors to M. Lutz-Bachmann and A. Fidora (eds), Juden, Christen und Muslime: Religionsdialoge in Mittelalter (Darmstadt: Wissenschaftliche Buchgesellschaft, 2004); Manfred Svensson, “A Defensible Conception of Toleration in Aquinas?,” The Thomist 75 (2011), 291–308; Cary J. Nederman, “Toleration in a New Key: Historical and Global Perspectives,” Critical Review of International Social and Political Philosophy 14 (2011), 349–61; Enzo Solari, “Contornos de la Tolerancia Medieval,” Ideas y Valores 62 (2013), 73–97; Cary J. Nederman, “Toleration in Medieval Europe: Theoretical Principles and Historical Lessons,” in J. Muldoon (ed.), Bridging the Medieval/Modern Divide: Medieval Themes in the World of the Reformation (Farnham, UK: Ashgate, 2013), 45–64.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

2

Cary J. Nederman

of the Western Middle Ages, ranging from Peter Abelard to John of Salisbury to Ramon Llull to Marsiglio of Padua to Nicholas of Cusa. However, scholars have seldom or never asserted that, as a matter of historical “anticipation,” any of these thinkers constituted a “precursor” to typically modern, liberal approaches to tolerance, such as those found in the writings of John Locke, John Stuart Mill, or (more recently) John Rawls. The latter theories supposedly share several key properties: 1) they are secular, in the sense of removing conflicts concerning the diversity of religious beliefs and comprehensive doctrines from the realm of public reason; 2) they rest on some doctrine of the individual as a sovereign, autonomous, and inviolable being, possessing a fundamental “liberty of conscience” or “freedom of expression”; and 3) they regard as arbitrary many of the “limits” to tolerance typical of the Middle Ages, such as the denigration of alternate belief systems or the denial of certain foundational Christian dogmas about the nature of God.2 I wholeheartedly agree that none of these features are to be found anywhere in medieval writings, so far as I know. Perhaps unsurprisingly, scholars who have demonstrated cognizance of the line of interpretation that posits forms of toleration during the Middle Ages have been, with rare exceptions,3 respectful of it, yet somewhat dismissive of its relevance to latter-day debates. The reasoning behind this treatment requires some extended exposition. Kristen Deede Johnson remarks: Some recent scholarship attempts to expand current conceptions of toleration, in which tolerance is almost exclusively linked to liberalism, by finding examples of tolerant political arrangements and principled defenses of toleration that pre-date the 2 For a useful summary statement of how these three claims are conceptually and historically interrelated, see David A. J. Richards, “Toleration and the Struggle against Prejudice,” in D. Heyd (ed.), Toleration: An Elusive Virtue (Princeton: Princeton University Press, 1996), 126–35. 3 In How the Idea of Toleration Came to the West (Princeton: Princeton University Press, 2003), for example, the late Perez Zagorin drew upon R. I. Moore’s thesis in Beyond the Persecuting Society to insist that the Middle Ages could not have permitted room for truly tolerant ideas (1–45 and esp. 313–14). Zagorin apparently did not realize that Moore’s perspective is one of the social, not the intellectual, historian, the former of which stands quite at odds with the argument of the latter that ideas of toleration were a significant driving force behind the emergence of tolerant practices in early modern Europe. See Karen Bollermann and Cary J. Nederman, “Standing in Abelard’s Shadow: Gilbert of Poitiers, the 1148 Council of Rheims, and the Politics of Ideas,” in K. Bollermann, T. M. Izbicki, and C. J. Nederman (eds), Religion, Power and Resistance from the Eleventh to the Sixteenth Centuries: Playing the Heresy Card (New York: Palgrave/ Macmillan, 2014), 13–36. Preston King’s review of Nederman, Worlds of Difference (American Political Science Review 95 [December 2001], 981–2) adopts a highly belligerent tone consonant with a willful misreading, and thus misunderstanding, of the central claims about the various authors analyzed in that book.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Modern Toleration through a Medieval Lens

3

rise of liberalism . . . Although other forms of tolerance have existed and continue to exist, it nevertheless seems safe to say that the tolerance that predominates in contemporary Western society has its roots in liberalism . . .4

In similar fashion, Nadia Urbanati comments: The aim of toleration before liberal toleration was dual: the attainment of social peace by imposing state authority over religious communities and the practice of benevolence toward, although not acceptance of, religious difference. This made it the opposite of a secularist project, as for instance, toleration during the Enlightenment, but also not naturally disposed toward liberal toleration as an art of separation of spheres because it was inspired by an ideal of community rather than individual freedom of conscience. Toleration before liberal toleration was internal to religious discourse and was conceived not as a process that would help the differentiation of the spheres of life, but as one that was meant to promote an articulation of those spheres within a unitary horizon of meaning and value that was profoundly religious.5

Likewise, Rainer Forst asserts: The first conclusion should be to contradict the widespread notion that toleration is a child of modernity, whereas antiquity and the Middle Ages represent nothing more than a pre-history of minor importance. The idea of toleration was reinterpreted and radicalized during the modern period, especially following the Reformation, and the spectrum of toleration justifications was greatly extended; nevertheless a large proportion of the new justifications proposed took their inspiration from existing arguments which, as we have seen, were always developed in conflict, that is, in the context of particular social and philosophical controversies.6

Finally, Gerson Moreno-Riaño observes: Tolerance as a general political principle was first advanced during the era commonly known as the Enlightenment. . . . This is not to suggest that previous historical eras, communities, or thinkers did not advance tolerance related arguments. But it is in the Enlightenment that one can find tolerance being advanced as a general and universal principle of politics and the modern state.7

4 Kristen Deede Johnson, Theology, Political Theory, and Pluralism: Between Tolerance and Difference (Cambridge: Cambridge University Press, 2007), 8–9. 5 Nadia Urbanati, “Half-Toleration: Concordia and the Limits of Dialogue,” in A. Stepan and C. Taylor (eds), Boundaries of Toleration (New York: Columbia University Press, 2014), 135–6; also 146, 147–8. 6 Rainer Forst, Toleration in Conflict: Past and Present, tr. C. Cronin (Cambridge: Cambridge University Press, 2013), 93. 7 G. Moreno-Riaño, “Tolerance, Identity, and the Problem of Citizenship,” in Gerson Moreno-Riaño (ed.), Tolerance in the Twenty-First Century (Lanham, MD: Lexington Books, 2006), 81, 91.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

4

Cary J. Nederman

The key theme emphasized by these scholars (and others who share their views8) should be evident: While they acknowledge that there may have been conceptions of toleration prior to the dawn of the Enlightenment, such theories are entirely or largely irrelevant to modern approaches to tolerance, whether rooted in Lockean natural rights or Millian intellectual liberty. In other words, there is a fundamental discontinuity, an insurmountable break, between medieval and later, essentially liberal ideas of toleration toward dissonant religious, social, and cultural ways of life. On this view, the identification of tolerant principles in a variety of writings dated to the Middle Ages, even if accurate, is utterly irrelevant to current discussions of tolerance. Insofar as the liberal principle of tolerance goes, this thesis seems to me to be correct. The observations of Brian Tierney, perhaps the leading exponent of the view that there is no significant divide between the political ideas of the twelfth century and those of the eighteenth century,9 confirm this conclusion. According to Tierney, given aspects of medieval thought such as “an insistence on the primacy of the individual conscience, a rejection of forced coercion, a nascent theory of natural rights—one might reasonably hope that at least some medieval thinkers would have moved on from these premises to assert a full fledged doctrine of religious liberty. Nothing of the sort happened . . . real religious rights emerged in the early modern world.”10 This reflects what I have already identified as a dominant assumption concerning toleration, namely, that any modern conception of toleration must be derived from or dependent upon some principles unique to liberalism. But is this equivalence of liberal doctrines with toleration entirely accurate? Do all ideas of tolerance in play today require a pre-existing commitment to liberal doctrines? If not, might there be some contemporary approach to toleration consonant with and capable of being informed by elements of medieval thought? Although, in order to avoid the charge of “ancestor hunting,” previous scholarship failed to examine very directly any potential linkages between

8 See also Andrew Murphy, Conscience and Community: Revisiting Toleration and Religious Dissent in Early Modern England and America (University Park, PA: Pennsylvania State University Press, 2001), xv–xvii; Hans Oberdiek, Tolerance: Between Forbearance and Acceptance (Lanham, MD: Rowman & Littlefield, 2001), 72; and Ingrid Creppell, Toleration and Identity: Foundations in Early Modern Thought (New York: Routledge, 2003), 34. 9 See Brian Tierney, Religion, Law and the Growth of Constitutional Thought 1150–1650 (Cambridge: Cambridge University Press, 1982), 103–8. 10 Brian Tierney, “Religious Rights: A Historical Perspective,” in N. B. Reynolds and W. C. Durham, Jr (eds), Religious Liberty in Western Thought (Grand Rapids, MI: Eerdmans, 1996), 41–2.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Modern Toleration through a Medieval Lens

5

medieval and modern approaches to toleration, this does not mean that no such continuity whatsoever may be found. The present paper, in particular, proposes that some of the key dilemmas faced by contemporary theories and practices of toleration, built around the three essentially liberal principles stated earlier, would in fact dissipate if one looks back to certain medieval frameworks. In order to support this case, I build upon the idea of “judgmental toleration” that the eminent political philosopher Michael Sandel once proposed in a brief essay of the same title.11 By this phrase, Sandel means that tolerance may be afforded to otherwise unacceptable ideas or practices for the sake of achieving another, greater good or of avoiding some greater evil. Nor is Sandel entirely alone among recent philosophers in endorsing the significance of the basic insight on which “judgmental toleration” is founded. Hence, I hope to demonstrate the relevance of some medieval defenses of tolerance both as illustrations and as extensions of the position articulated by Sandel and other philosophers. Finally, the paper considers how “judgmental toleration” may prove relevant and useful as a lens through which to view certain issues arising in modern political philosophy, by turning to the influential conception of liberty proposed by John Stuart Mill and to debate among contemporary proponents of the theory of deliberative democracy. In sum, I contend that, in the case of an explicit and demonstrably medieval defense of tolerance, the lessons of the past are salient in the present, so that the early history of toleration provides vital and fruitful resources for understanding the limitations of liberalism and the potential for alternative non-liberal routes to the realization of tolerant social and political attitudes and practices.

1. JUDGMENTAL TOLERATION Sandel’s paper draws an analytical distinction between a liberal version of toleration, which he labels “non-judgemental,” and a “judgemental” alternative. Taking aim at the Rawls of Political Liberalism, as well as at Richard Rorty, Sandel asserts that liberal toleration has as an essential feature a “bracketing” effect, “in the sense that it seeks to bracket substantive moral and religious controversies; it seeks to avoid passing moral judgement on the

11 Michael Sandel, “Judgemental Toleration,” in R. P. George (ed.), Natural Law, Liberalism, and Morality (Oxford: Clarendon Press, 1996), 107–12. Note that the spelling of the word “judgemental” was Anglicized in its published form; I will employ this spelling whenever quoting from or directly referring to the text. Otherwise, I shall adopt the American spelling, per the style of this journal.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Cary J. Nederman

6

practices that it permits.”12 In other words, liberal toleration does not wish to get involved in the messy business of adjudicating the worthiness of the doctrines and practices to be tolerated. All positions are deemed to be tolerable unless they require, as a condition of their existence, the violation of the justice or liberty necessary for the maintenance of liberal society or mandate that argumentation from a comprehensive doctrine constitutes the only valid form of human reasoning (such as fundamentalists and scriptural literalists). To the extent that those viewpoints or ways of life of which I do not approve do not interfere with my own viewpoint or chosen way of life, liberal toleration requires that I refrain from entering into any direct evaluation, let alone condemnation, of them, at least in the public sphere. (Of course, “non-judgmental toleration” still permits me as a private person to converse with neighbors, friends, or indeed complete strangers about their comprehensive doctrines, should they be willing to engage in such dialogue. And presumably, I could express in that venue my revulsion or other moral disapprobation of their views. But such private dialogue is not mandatory and interlocutors always retain the freedom to walk away, or ignore me, or even announce their judgment about my own evil beliefs and behaviors to my face.) In this sense, the dominant form of liberal toleration assumes a strictly “political” orientation, setting aside theological, philosophical, or metaphysical considerations entirely.13 In the argument that I am summarizing, Sandel develops an internal critique of this position that shows “non-judgemental toleration” to involve a self-contradiction concerning its estimation of the truth-values of the doctrines tolerated. I am not, however, so interested in analyzing Sandel’s critical argument on this count as I am in exploring the “judgmental” stance that he articulates as an alternative. “Judgemental toleration,” he says, “does not bracket. It assesses the moral worth or permissibility of the practice at issue, and permits or restricts it according to the weight of these moral considerations in relation to competing moral and practical considerations.”14 On the one hand, it seems that such judgments might at times lead to intolerance of practices deemed morally intolerable—a concession that Sandel freely admits, indeed, seems to regard as inescapable. But Sandel presumes that we can honestly and fairly evaluate the relative merits of competing moral claims and restrict or defend their practice as a result of deep reasoning about the consequences thereof. No determination of whether or not to tolerate any given practice can “be detached from the question of whether the practice itself is morally permissible.”15 This implies, it seems to me, that there are three possible categories of judgments 12 Ibid., 107.

13 Ibid., 108.

14 Ibid., 107.

15 Ibid., 108.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Modern Toleration through a Medieval Lens

7

available: 1) an act is permissible because there is no clear or compelling moral reason to restrict it; 2) an act is permitted even though there are moral reasons to restrict it, because such restriction would itself produce a greater evil or preclude a greater good; and 3) an act is restricted because the moral reasons for such restriction are clear and compelling on the basis of considered deliberation. The second of these three possibilities is really the interesting and important one. On the one hand, it allows individuals and institutions to espouse moral judgments, yet, on the other, it constrains their ability to act in restrictive fashion on the basis of such judgments.16 In a book that slightly predates the publication of Sandel’s essay, the analytical philosophers Nick Fotion and Gerard Elfstrom provide an account of toleration that parallels and, I think, elaborates on the basic idea of “judgmental toleration.” They consider the distinction between skepticism and dogmatism, where the former is ordinarily associated with toleration and the latter with intolerance. A dogmatic perspective, unlike a skeptical one, would seem to involve commitment to a proposition of the following sort: “We should not tolerate this behavior because we all know right from wrong, and this behavior is wrong.”17 Fotion and Elfstrom argue, however, that this claim by no means logically entails intolerance in all instances. Rather, they observe, “The behavior that concerns the dogmatist may be wrong in and of itself; but other (moral) considerations may override its presumed wrongness, so that it is still logically open for the speaker to practice or preach toleration. This kind of reason giving, in which a lesser evil is chosen in order to avoid a greater evil, is one of the most important

16 It may be reasonably asked, as did a colleague of mine, Hilaire Kallendorf, whether Sandel’s position reflects simply some updated version of casuistry. I have given this question serious consideration and conclude that it does not. An important study of the subject by Albert R. Johnsen and Stephen Toulmin, The Abuse of Casuistry: A History of Moral Reasoning (Berkeley: University of California Press, 1988), points out: “There lies a deeper intellectual conflict between two very different accounts of ethics and morality: one that seeks eternal, invariable principles, the practical implications of which can be free of exceptions or qualifications, and another, which pays closest attention to the specific details of particular moral cases and circumstances” (2). The latter, of course, is casuistry. In my view, the “judgmental” approach favored by Sandel, as well as his medieval predecessors, stands between these two supposedly antithetical accounts, in the sense that principles still matter, but not in the invariant way mentioned. “Judgmental” toleration requires comparison of competing principles, not the rejection of principles per se in favor of a narrow concentration on specifics. In other words, casuistry is wholly contextual in a way that “judgmental” toleration is not. See also Hilaire Kallendorf, Conscience on Stage: The Comedia as Casuistry in Early Modern Spain (Toronto: University of Toronto Press, 2007). 17 Nick Fotion and Gerard Elfstrom, Toleration (Tuscaloosa, AL: University of Alabama Press, 1992), 33.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

8

Cary J. Nederman

when applying the concept of toleration.”18 As with Sandel, Fotion and Elfstrom emphasize the importance of invoking “rational methods to justify our saying that we have been tolerant in a specific situation. . . . dogmatism is clearly no more tied to intoleration than skepticism is to toleration.”19 Indeed, they assert that such “normative reason giving,” in which different types of values are compared in such fashion, entails that “goods are . . . chosen over other goods . . . or one evil over another.”20 Normative reason giving demands that values be rationally weighed and balanced, they say, to the extent that “all other tolerating examples involve comparisons.”21 In other words, for Fotion and Elfstrom, true expressions of tolerance cannot be disentangled from the rational evaluation of competing conceptions of what is good and/or evil, a view that converges with and reinforces Sandel’s “judgmental toleration.” Sandel recognizes that his concept of “judgmental toleration” is rooted in a longstanding intellectual tradition that can be traced back to the Middle Ages, in particular, to St. Thomas Aquinas’s attempt to grapple with the problem of tolerating various forms of faithlessness.22 Sandel acknowledges that St. Thomas certainly upheld that the “sinfulness” of unbelief constituted sufficient warrant for its punishment. Thus, we might expect him to advocate the penalizing of Jews and other infidels. Yet Aquinas demurs, drawing an analogy between God’s tolerance for evils in the created world that might be avoided “lest without them greater goods might be forfeited or greater evils ensue” and the situation of “those who are in [earthly] authority rightly tolerat[ing] certain evils, lest certain goods be lost or certain greater evils incurred.”23 On this point, Aquinas cites a well-known dictum of St. Augustine to the effect that “if you do away with harlots, the world will be convulsed with lust.”24 Thus, for Aquinas, the conferral of toleration depends upon a calculation of the relative moral merits of the beliefs to be judged. For example, from the Jews, who “of old foreshadowed the truth of the faith which we hold, there follows this good— . . . that our faith is represented in a figure, so to speak. For this reason, they are tolerated in 18 Ibid., 34. 19 Ibid., 34. 20 Ibid., 40. 21 Ibid., 40. 22 Sandel, “Judgemental Toleration,” 107–8. For further discussion of Aquinas’s views concerning these matters, see Oberdiek, Tolerance, 69–72; Shadia B. Drury, Aquinas and Modernity: The Lost Promise of Natural Law (Lanham, MD: Rowman & Littlefield, 2008), 45–74; and Svensson, “A Defensible Conception of Toleration in Aquinas?,” 291–308. 23 St. Thomas Aquinas, Summa Theologiae II-II, Q10, A11; translation from W. P. Baumgarth and R. J. Regan (eds), St. Thomas Aquinas on Law, Morality, and Politics (Indianapolis: Hackett, 1988), 254–5. 24 Ibid., 255; cf. Augustine, De Ordine II.4, in L. Schopp, D. J. Kavanagh, R. P. Russell, and T. F. Gilligan (eds), Divine Providence and the Problem of Evil (New York: Cosmopolitan Science & Art, 1942), 94–5.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Modern Toleration through a Medieval Lens

9

the observance of their rites.”25 Likewise, “the rites of other unbelievers” will have to be judged according to their moral bases and consequences. In some cases, “they are by no means to be tolerated”; in other instances, they may be tolerable “in order to avoid an evil, e.g., the scandal or disturbance that might ensue or some hindrance to the salvation of those who, if they were unmolested, might gradually be converted to the faith.”26 In sum, we don’t have to “bracket” the beliefs of infidels (not to mention our own beliefs) in order to accept good reasons to tolerate them, according to Aquinas. It must be noted that Aquinas was not simply a forerunner of an essentially concordant or modus vivendi idea of toleration. Nor may his position be assimilated to the so-called “permission” conception of tolerance, according to which an asymmetry of power between institutions (such as the Church) and individuals whose conduct or beliefs violate institutional norms (morality, orthodoxy) means that any tolerance afforded “could be revoked at any time” and was thus “a fragile, precarious condition.”27 Rather, St. Thomas insists that the decision about what to tolerate and what not to tolerate depends upon doing the hard work of moral differentiation between what can be tolerated and what is entirely beyond the pale in the name of some greater good or lesser evil. In this way, Aquinas exemplifies precisely the core principles that Sandel ascribes to “judgmental toleration” (and indeed, that Fotion and Elfstrom attribute to “normative reason giving”).

2. MEDIEVAL JUDGMENTALISM Sandel does not realize—and perhaps should not be expected to know— that the position he attributes to St. Thomas was hardly unique or sui generis during the Middle Ages. Inasmuch as toleration, conceived in judgmental terms, occupied an influential position in a wide range of medieval legal and political theories, it is possible to extend its application to moral problems far beyond those envisioned by Sandel. In an important article, István Bejczy located one source of such a medieval version of “judgmental toleration” in the work of canon lawyers. Citing Raymond de Peñafort and a host of other canonists, Bejczy argues that the word tolerantia was “applied in order to prevent a greater evil than the tolerated one”: Minus malum toleratur ut maius toleratur, exactly the position advocated by St. Thomas.28 Raymond and his 25 Aquinas, Summa Theologiae II-II, Q10, A11; 255. 26 Ibid., II-II, Q10, A11; 255. 27 Forst, Toleration in Conflict, 61; also see Moshe Halbertal, “Autonomy, Toleration, and Group Rights,” in Heyd, Toleration, 114–26. 28 Bejczy, “Tolerantia: A Medieval Concept,” 370.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

10

Cary J. Nederman

fellow canon lawyers distinguished three different forms of “permission.” None of these principles of “permission” accords with the modern liberal understanding of that term, but they have parallels to elements of Sandel’s “judgmental toleration” and of Fotion and Elfstrom’s “normative reason giving” previously discussed. To quote Raymond: Permission is taken in three different ways. First, when something is allowed that is not forbidden by law. . . . Second, when something is indulged that runs counter to human rules. . . . This is properly called the true and absolute permission, and it excuses from sin. The third type of permission occurs when lesser evils are permitted so as to prevent greater ones. This is called permissio comparativa, and it does not excuse from sin. It should, however, be called tolerantia rather than permission.29

Whereas liberal critics of “judgmental toleration” would point to permission as a matter of temporary institutional forbearance, medieval canonists saw at work precisely the comparative aspect of tolerance in the sense that was later adopted by Aquinas. Thus, the modern critique of the “permission model” misses the point: permission was understood in medieval legal thought to be a form of rational normative judgment which assured unqualified protection even to those who violate established religious or moral standards, whether they be infidels or sinners, in a manner not subject to the insecurity posited by its critics.30 The lines of medieval legal thought (followed by Aquinas) about the character of tolerantia (presaging Sandel’s “judgmental toleration”) were embraced by a host of later thinkers and adapted to suit a number of circumstances in which seemingly sinful actions could nonetheless be countenanced for the sake of a greater good or lesser evil. One example may be found in the treatise De moneta (On Money) by the fourteenth-century French schoolman Nicole Oresme, who focused on the problem of the morality of market transactions. Oresme held that economic activities are generally subject to regulation according to moral and religious standards. In this connection, he invoked what appears to be a conventional Christian/ Aristotelian prohibition on activities such as usury and money-changing (campsoria).31 Yet his analysis of these supposedly “unnatural” economic transactions is qualified. Having soundly condemned both usury and money-changing on the predictable grounds that they are incompatible 29 Quoted from ibid., 369–70. 30 A more extensive discussion of permission in general, and permissio comparativa more particularly, may be found in Brian Tierney’s recent book, Liberty and Law: The Idea of Permissive Natural Law, 1100–1800 (Washington, DC: Catholic University of America Press, 2014). 31 Nicole Oresme, De moneta, in C. Johnson (ed.), The De Moneta of Nicholas Oresme and English Mint Documents (London: Thomas Nelson and Sons, 1957), 27–8.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Modern Toleration through a Medieval Lens

11

with the true purpose of money—namely, to facilitate the exchange of produced goods—Oresme nonetheless refuses to prohibit them unconditionally: “Yet sometimes from necessity or advantage some vile business, such as the art of money-changing, or some evil one, such as usury, is permitted.”32 These enterprises (which he compares with the maintenance of “public houses of prostitution,” an evident allusion to Augustine’s De ordine) may be sanctioned in order to avoid a greater evil or scandal within the community. “ . . . Inasmuch as the usurer’s interest is not so excessive, nor so generally prejudicial to the multitude,”33 Oresme reasons, the usurer-borrower relationship remains tolerable. Here, features of “judgmental toleration” are put to work to overturn a universal prohibition on certain sorts of voluntary interpersonal relations in the marketplace. Note that Oresme does not take issue with the traditional moral condemnation of usury; indeed, he invokes and endorses it. Rather, he merely asserts that the suppression of such an evil may not be strictly entailed by the plain fact that it is evil, since it at least sometimes contributes to a greater good. Thus, there are countervailing moral reasons why usury may be permitted under certain circumstances. It is necessary for us to do the hard job of moral reasoning in order to determine when circumstances forbid and when they allow the practice. In similar fashion, the French courtier-author Christine de Pizan, writing c.1400, focused several of her works on advising members of the royal family and their associates (including, and especially, women) about the appropriate uses of power. Among her recommendations, she advocated the legitimate employment of deception in order to serve the ends of maintaining political peace (a key theme in her writings). Thus, for instance, she suggests that the princess should dissemble with her enemies, even when she has definite knowledge of their plots and machinations. Christine regards the employment of deceptive conduct to be a practical guide to maintaining public order as well as personal reputation (so long as her deceit is not discovered). As she remarks, “The wise lady will use this prudent device of discreet dissimulation, which should not be considered vicious but rather a great virtue when employed for the common good, to maintain peace, or to avoid detriment or greater harm. Not only will she escape trouble but she will also achieve great benefit if she pretends not to notice conspiracy against her.”34 Mendacity also is proposed in the case of charitable works and benefactions. Christine counsels that the public endowment of gifts to the poor and religious, irrespective of its intrinsic moral worth, aids the princess 32 Ibid., 29. 33 Ibid., 28. 34 Christine de Pizan, A Medieval Woman’s Mirror of Honor: The Treasury of the City of Ladies, ed. C. C. Willard, (New York: Persea, 1989), 106–7.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

12

Cary J. Nederman

and her family in the maintenance of their standing and reputation in the community, no matter how hypocritical the act itself: “Justifiable hypocrisy is necessary for princes and princesses who must rule over others and thus be accorded more respect than others. Moreover, expedient hypocrisy is not unworthy for others desiring honor, as long as they practice it for worthy ends.”35 Christine’s point is not the Machiavellian one that the ends justify the means. Rather, she considers the maintenance of a decorous reputation to matter a great deal for the well-being not merely of the governors but of the governed. The public image of the royal family (and the order it engenders), even if it is achieved by means of deception, is a good sufficiently worth upholding that otherwise reprehensible conduct may be employed to attain it. In this sense, an action otherwise acknowledged to be evil should be tolerated on account of the effect it exercises upon the more notable goal of engendering public peace. Another intriguing illustration of medieval “judgmental toleration” may be found in the De legibus of Francisco Suárez, a leading figure of the so-called Spanish “Second Scholastic” of the sixteenth and seventeenth centuries. Although chronologically Suárez may seem to have flourished in the early modern rather than the medieval period, his primary work, as a legal theorist, is thoroughly grounded in the intellectual framework of the Middle Ages. More specifically, he sought to unpack some of the complexities of preceding, essentially Thomistic, doctrines, for example, concerning the relationship between the many different types and layers of law that Aquinas had posited in the Summa. In this connection, one of Suárez’s significant contributions was to articulate the distinction between natural law and so-called ius gentium (the law of nations or peoples).36 The Romans who invented and deployed such terminology were often notoriously equivocal about whether ius gentium was a form or branch of natural law, or an independent type of law intermediary between natural and positive human law. Aquinas seems to have been equally uncertain about the character of ius gentium.37 By contrast, Suárez treats ius gentium as a product of longstanding custom and usage pretty much everywhere in the world, regardless of location or cultural or religious circumstance.38 Thus, while not strictly natural—in the sense that its tenets are not identical with the law of nature—ius gentium constitutes something like a second nature.39 Intriguingly, this implies that ius gentium may in some cases override natural law, even though natural law in principle is supposed to forbid all evil acts in such a way as to be strictly intolerant. 35 36 37 38

Ibid., 109. Francisco Suárez, De legibus, (Madrid: Instituto Francisco de Vitoria, 1973) I.17–20. See St. Thomas Aquinas, Summa Theologiae, I-II Q95 A4. Suárez, De legibus II.19.10. 39 Ibid., II.20.1.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Modern Toleration through a Medieval Lens

13

But Suárez claims that certain practices contrary to natural law may nonetheless “be tolerated by ius gentium; since this very toleration may be so necessary, in view of the frailty and general character of mankind or business affairs that almost all nations agree in manifesting it.”40 In this regard, he cites prostitution and commercial caveat emptor as cases in point; prostitution is forbidden by natural law and complete honesty is required. Yet “the toleration of prostitutes” is permissible under ius gentium because a greater evil (presumably, universal licentiousness, as Augustine said) is avoided by condoning prostitution; moreover, “the toleration of deception that is not excessive in the matters of contracts” is licit because no seller can be reasonably expected to report all the defects of his wares to a buyer, so long as there is no malicious attempt to hide those defects.41 Law, therefore, may legitimately permit what a narrow standard of morality would forbid. Suárez admits that prostitution and sharp business practices are incompatible with the dictates of natural law, but also insists that it is widely recognized that there can exist countervailing moral reasons why human beings may still permit such conduct without falling into utter sinfulness. Suárez thus endorses a position that converges with key features of the judgmental form of toleration. Another figure who, writing about a half-century before Suárez, remained grounded in the thoroughly medieval worldview I have associated with “judgmental toleration” was Bartolomé de Las Casas. A Dominican friar, Las Casas’s work focused on the status of the inhabitants of the newly discovered Americas and, in particular, of the forms of “heathen” worship practiced by them. Although educated in and wholly conversant with the philosophical, legal, and political doctrines of Spanish late scholasticism, he observed the practices of indigenous civilizations as well as the results of European conquest first-hand, initially as a member of the Spanish colonial elite, and later as Bishop of Chiapas. During decades of concerted attack on Spanish mistreatment of the indigenous population of the Americas, Las Casas unleashed a broad range of arguments against the policies and conduct of his fellow Europeans. Over the course of his lengthy career, Las Casas produced a vast body of works in defense of the Indians, including polemical tracts and historical and anthropological treatises. In addition, he produced a lengthy refutation of Juan Ginés de Sepúlveda’s scholastic defense of the Spanish conquest of the “New World,” and the subsequent enslavement, slaughter, or forced conversion of its inhabitants. The latter, entitled simply Apologia (translated as In Defense of the Indians), was composed in the aftermath of a public disputation with Sepúlveda held at Valladolid in 1550–1551. Written in typical scholastic format, the Apologia represents 40 Ibid., II.20.3.

41 Ibid., II.20.3.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

14

Cary J. Nederman

Las Casas’s most extensive reflections on the philosophical and theological bases for respecting the Amerindian ways of life, including their religious beliefs and cult. Given the sheer length of the Apologia, I am compelled to concentrate on a single argument proposed by Las Casas that reflects the judgmental framework that I have been highlighting. One of the central arguments deployed by Sepúlveda in favor of Spanish suppression of the way of life of the Indians arises from their supposed practices of human sacrifice and even cannibalism as part of their religion. Sepúlveda charged that such forms of worship were so morally repugnant that the Spanish conquerors had a duty to God to intervene by force in order to protect the innocent. In Chapter 28 of the Apologia, Las Casas responds to this claim by invoking the principle of permissio comparativa: “In trying to prevent the death of a few innocent persons we should not move against an immense multitude of persons, including the innocent, and destroy whole kingdoms. . . . Instead, war must be avoided and that evil tolerated for a while—and, in some cases, permanently . . . ”42 Beyond his observation, based on his own experience, that the slaying of innocents is uncommon among native peoples, Las Casas appeals to the normative weighing of reasons from multiple sources. First, he says, Aristotle countenances as a “rule of right reason [that] when we are confronted by two choices that are evil both as to moral guilt and punishment and we cannot avoid both of them, we ought to choose the lesser evil. For in comparison with the greater evil, the choice of a lesser evil has the quality of a good.”43 This position, according to Las Casas, is reinforced by natural law, which also teaches the immorality of preferring the greater evil to the lesser.44 He calls such a violation of natural law “a sin, which, although not mortal, is very serious indeed,” invoking the authority of St. Thomas.45 Finally, he appeals to the doctrine that “the doctor jurists give concerning the well-known permission, which happens when evils and even serious sins are permitted so that more serious evils may be avoided or so that the good by which the condition of the state is strengthened should not be obstructed,” which he supports by pointing to familiar examples of “permission for prostitutes and for Jewish rites.”46 It should be noted that, throughout the chapter, Las Casas overtly employs the language of tolerantia, in the same manner as both Aquinas and the canonists. In other words, he equates the comparing of evils (and even the equation of the lesser evil with a good) with a tolerant outlook, conceived in judgmental terms. One may be repulsed by the idea of allowing the approved death of 42 Bartolomé de Las Casas, In Defense of the Indians, ed. S. Poole (DeKalb: Northern Illinois University Press, 1974), 190. 43 Ibid., 191. 44 Ibid., 191–2. 45 Ibid., 192. 46 Ibid., 192.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Modern Toleration through a Medieval Lens

15

innocents—as Las Casas surely was—but far greater injury and immorality arises from the destruction of a civilization and its people in order to eliminate that practice. This is not an easy conclusion to accept; yet Las Casas, at any rate, sees no other reasonable alternative, given that the saving of small numbers of Amerindians from sacrifice and the protection of all inhabitants from death are not both possible.

3 . S O D O M Y A N D H E RE S Y One may well wonder how far medieval thinkers might be prepared to extend tolerance. After all, the examples thus far examined seem to apply to a range of moral practices (prostitution, mendacity, commerce, religious rite) that fall within the category of “venial” (as opposed to “mortal”) sins that do not necessarily endanger salvation. As we have just seen, Las Casas admits that even widespread slaughter in the name of preventing the deaths of a few innocents does not count as a mortal sin. But was it possible for at least some thinkers during the Middle Ages to extend toleration to the really hard cases of far more serious (mortal) sins on judgmental grounds? Bejczy states that there were universally acknowledged limits to the tolerable. “Heretics and homosexuals were not even in theory allowed to dwell in the margins of society, like Jews and prostitutes,” he remarks, “Heresy and homosexuality were not seen as minor evils that society could afford.”47 In the instance of homosexuality, a number of recent studies have emphasized that early Church strictures concerning its practice came to be reinforced by the reintroduction into the West of Aristotle’s scientific, ethical, and political writings during the course of the thirteenth century.48 Writing c.1300, the Parisian schoolman Ptolemy of Lucca borrows upon both Augustine and Aristotle in order to revile homosexuals. Abstaining from intercourse with women, he proposes, would lead men “to approach males sexually . . . and for this reason it is less evil to have carnal relations with women than to fall into vile and shameful acts.”49 Ptolemy then adapts De ordine to suit the purposes of his own agenda: “Augustine says . . . ‘Take away whores from the 47 Bejczy, “Tolerantia,” 375. 48 John Boswell, Christianity, Social Tolerance, and Homosexuality: Gay People in Western Europe from the Beginning of the Christian Era to the Fourteenth Century (Chicago: University of Chicago Press, 1980), 303–32; Mark D. Jordan, The Invention of Sodomy in Christian Theology (Chicago: University of Chicago Press, 1997), 114–58; and Joan Cadden, Nothing Natural is Shameful: Sodomy and Science in Late Medieval Europe (Philadelphia: University of Pennsylvania Press, 2013). 49 Ptolemy of Lucca, On the Government of Rulers, ed. J. M. Blythe (Philadelphia: University of Pennsylvania Press, 1997), 4.14.6.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

16

Cary J. Nederman

world, and you will fill it with sodomy.’ For this reason Augustine says that the earthly city made the use of harlots a licit foulness,”50 a claim for which, once again, there is no direct corollary in the writings of the saint. For good measure, Ptolemy cities the discussion of “continence” and “incontinence” in Aristotle’s Nicomachean Ethics 1148b28–30 to the effect that “the vice of sodomy exists because of depraved nature and perverse usage . . . ”51 Arguably, Ptolemy’s position nicely encapsulates the attitude toward homosexuality endorsed by his rough contemporaries. But the combination of clerical and Aristotelian teachings about homosexuality did not always produce wholesale condemnation of it. Brunetto Latini, a mid-thirteenth century Florentine civil servant and teacher of Dante, adopted a rather different point of view. Dante, in the Inferno, placed his master in that part of Hell reserved for sodomites, the reasons for which have been widely debated.52 Perhaps one reason for this stems from Latini’s remarks in his Li livre de tresor, a vast (and influential) compendium of knowledge concerning theology, nature, ethics, and politics composed while its author was exiled in France. Among the significant features of the Tresor was a detailed summary, in French, of the entirety of the Nicomachean Ethics, which had been translated into Latin c. 1245. Significantly, Latini barely touches upon the seventh book of the Ethics, in which Aristotle mentions and attempts to account for the supposed depravity and “brutishness” of “sexual relations between males” previously mentioned.53 Instead, his treatment of homosexuality is subsumed under his discussions of the virtue of temperance, in which he employs a judgmental logic. Broadly speaking, he says that temperance is concerned with “the pleasures of the touch,” which he divides into two types: “one which is prompted by lust, and the other which comes from the other bodily members.”54 Among the latter he counts “fine clothes, baths, armor, dice playing, or similar things which corrupt man’s life if they are done with a lack of moderation.”55 Yet such pleasures are not inherently evil; if engaged in temperately, Latini remarks, a man who indulges in them “should be overlooked, if it does not harm the person or his honor or his possessions.”56 In other words, in determining whether to tolerate activities that are deemed morally dubious, one must judge according to the manner in which they are

50 Ibid., 4.14.6. 51 Ibid., 4.14.6. 52 See Richard Kay, Dante’s Swift and Strong: Essays on Inferno XV (Lawrence: The Regents Press of Kansas, 1978), 1–23; Giovanni della’Orto, “L’omosessualità nella poesia volgare italiana fino al tempo di Dante,” Sodoma 3 (1986), 13–35. 53 Brunetto Latini, The Book of the Treasure, ed. P. Barraette and S. Baldwin (New York: Garland, 1993), II.41–2. 54 Ibid., II.76. 55 Ibid., II.76. 56 Ibid., II.76.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Modern Toleration through a Medieval Lens

17

conducted. In a different section of the Tresor that addresses temperance, Latini expands on this position. Here, he adopts the position that a distinction must be made between those who act moderately “in drinking and eating, in clothing and in all other bodily pleasures” and those who are blameworthy on account of their intemperance.57 He praises the virtue of the former, who take into consideration the context of their conduct and restrain themselves accordingly. By contrast, he asserts that “the delight of the world which is against nature”—namely, gross and unconstrained intemperance—“is demonstrably blameworthy, more so than sodomy (avoutire), by which is meant lying with a male.”58 We do not find here an unqualified moral endorsement of homosexual behavior, but instead a pattern of thought typical of the medieval version of “judgmental toleration.” Engaging in homosexual acts may constitute an evil, but it is a lesser evil than that of an irredeemably intemperate man. Presumably the reason for Latini’s view is connected with the full-fledged immorality of the man who lacks any self-control whatsoever in comparison with the man who engages only in a single evil activity. Is Latini lauding homosexuality as a good? Assuredly not. But nor is he treating it as the worst of vices, as an absolute and implacable mortal sin. Even more unforgivably sinful than sodomy is heresy. As Brian Tierney has declared, “Every writer . . . saw heresy as a sin and a crime that was properly judged by the church and properly punished by the secular power. . . . They thought that heresy must somehow stem from malice, from a perverted will that deliberately chose evil over good.”59 Hence, the heretic had no place whatsoever in Christian society and, as a threat to the unity of the faith, was to be exterminated in the name of defending dogma. It is well nigh impossible to find a scholar who would dispute this conclusion. But at least some evidence to the contrary, based on a fundamentally judgmental standard of reasoning, may be located in the writings of Marsiglio (Marsilius) of Padua. According to Marsiglio’s best-known work, the Defensor pacis (Defender of Peace), completed in 1324, the material welfare and harmony of citizens were the chief goals of the political community. These aims were best achieved when all of the goods and services necessary for bodily human existence were exchanged in orderly fashion. Thus, those who adopted supposedly heretical views in matters of faith, while they could be detached from the spiritual community of the church, ought not to be excluded thereby from interactions with fellow citizens involving purely earthly needs. It remained beyond the power of clerics, in Marsiglio’s view, to

57 Ibid., II.33. 58 Ibid., II.33. 59 Tierney, “Religious Rights,” 43–4, 45.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

18

Cary J. Nederman

regulate the social and economic exchanges that were required or useful for maintaining the corporeal existence of the orthodox and the heterodox alike. Rather, only the community itself enjoyed competence to determine how temporal human conduct should be governed. Religious dissent per se constituted an invalid basis for determining who deserved to be excluded from or punished by secular society.60 In a later work, the Defensor minor (written around 1340), Marsiglio developed his claim that the priesthood lacks any rightful power whatsoever to declare a member of the church anathema: “Inasmuch as this authority must involve coercion over goods or persons or both, and thus must be applied (however moderately) in this world by the civil power, such authority never pertains to priests.” Excommunication involved a coercive dimension precisely because it imposes upon the sinner “the punishment of exile” and removes “the wealth and income” by which one sustains oneself and one’s family.61 In the conventional view, the excommunicant stands outside of the temporal community as well as the spiritual one. But the priesthood lacks standing to impose such a punishment. By contrast, Marsiglio holds that, like all other applications of coercive authority, the decision to remove a person from civil association through excommunication can only be the determination of the whole body of citizens or its appointed executives. Indeed, Marsiglio poses the further query of “whether it is expedient to separate heretics from, or deprive them of, the fellowship of believers.”62 In other words, he wonders about the validity of excommunication at all, even in a purely spiritual sense. Is it ever appropriate that those who decline to accept the established tenets of religious dogma should be excluded from worship and the sacraments? To answer this question, Marsiglio examines the purpose of excommunication. It cannot be the case that individuals are excommunicated for the sake of compelling them to accept dogma. Marsiglio denies that non-believers can ever truly or effectively be coerced into accepting the faith: It is not a tenet of Christianity that “any individual ought to be compelled to profess the Christian faith.” This view he employs explicitly against the practice of crusading: “If a foreign journey is made or will be made in order to subdue or restrain infidels for the sake of the Christian faith, then such a foreign journey would in no way seem to be meritorious.”63 Thus, if the aim of excommunication is the coercion of heretics to return to the faith, 60 An especially cogent discussion of the issues at stake here is offered by Bettina Koch, “Marsilius of Padua on Church and State,” in G. Moreno-Riaño and C. J. Nederman (eds), A Companion to Marsilius of Padua (Leiden: Brill, 2012), 139–79. 61 Marsiglio of Padua, Writings on the Empire: Defensor Minor and De Translatione Imperii, ed. C. J. Nederman (Cambridge: Cambridge University Press, 1994), 10.3. 62 Ibid., 10.5. 63 Ibid., 7.3.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Modern Toleration through a Medieval Lens

19

this enterprise is likewise unjustified.64 Compulsion is not to be employed as a means to return the heterodox to the faith. Consequently, the only valid reason for excommunication must be the protection of the eternal souls of orthodox believers. Marsiglio admits that “heretics and other infidels . . . are to be shunned, especially in connection with domestic or social relations, or cohabitation or conversation, concerning those matters which pertain to the preservation of the rituals of the faith, . . . lest they taint the remaining believers.”65 He quotes several biblical passages in support of the practice of excommunication on these grounds. But he immediately emphasizes that the separation of believer from heretic may at most occur in connection with spiritual matters only: “This is to be understood in regard to belief and the observance of the rituals of the faith, rather than in regard to other domestic or civil intercourse.”66 There is neither a scriptural nor a rational basis for extending spiritual prohibition to the temporal domain of communal life. Marsiglio’s defense of this claim rests on what he regards to be the final cause of human law and its coercive force, namely, “the tranquility and finite happiness of this world,” with the result that such precepts and their enforcement regulate “human beings insofar as they are disposed and affected towards tranquility and power and many other [earthly] things.”67 If excommunication imposed by the priestly caste were to encompass temporal as well as spiritual affairs, then the legitimate concerns of public order and welfare would be disturbed. Indeed, the very purpose of excommunication, the well-being of the faithful, would be undermined. By denying intercommunication of a purely secular nature between the orthodox and non-believers of all sorts, it is the faithful themselves who will be harmed: They will not be able to take advantage of those material benefits that constitute the very foundation of communal life. It is a violation of the social bond, the very law of nature that joins human beings together, to deprive anyone (let alone orthodox Christians) of civil comforts and associations, such as by purchasing bread, wine, meat, fish, pots or clothes from them [persons guilty of heresy], if they abound in such items and others of the faithful lack them. This is likewise also true of the rest of the functions and comforts to which they might possibly be judged susceptible in connection with their positions or civil duties and services. For otherwise this punishment would redound in like fashion or for the most part to innocent believers.68

The extension of excommunication to civil association constitutes for Marsiglio an attack on the most basic principle of community rooted in 64 Ibid., 15.7. 67 Ibid., 15.5.

65 Ibid., 10.5. 68 Ibid., 15.6.

66 Ibid., 10.5.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

20

Cary J. Nederman

the interchange of functions. Neither orthodox Christians nor religious dissenters would be permitted to perform those tasks which form the basis of their citizenship. Inevitably, the breakdown of the community would occur and intranquility would ensue. Here again is a form of “judgmental toleration” at work in support of a principled policy of not merely affording, but requiring, protection to heretics and other religious dissidents in all temporal dimensions. Persecution destroys the natural and necessary communal order.

4 . M OD E R N A P P LI C A T I O N S My rationale for this Cook’s Tour through the writings of Aquinas, Oresme, Christine, Suárez, Las Casas, Latini, and Marsiglio may be condemned (rightly) as an exercise in the “ancestor seeking” that previous scholarship on medieval tolerance consciously sought to avoid. However, I am not attempting to lend support to the lessons derived from Sandel’s version of tolerance by multiplying early authorities who have advocated a similar conception of toleration. Rather, I hope to have shown that a judgmental perspective has been applied and adapted over time to advocate tolerance of a wide range of beliefs and behaviors that might otherwise be found immoral, unnatural, or impious. On the one hand, all of the medieval thinkers whom I have examined make and express judgments about matters of moral and religious right or wrong. There is no need to bracket one’s moral convictions, whether about religious beliefs or social or economic practices. On the other hand, these authors offered reasons and made compelling arguments about why their judgments are so overriding that they should be translated into formal restrictions subjected neither to political institutions nor intolerant individuals. As Suárez teaches us, even direct appeal to the prohibition of some activity by natural law does not end the debate; human beings may licitly override natural law for cause. In turn, the general features of “judgmental toleration” are by no means foreign or irrelevant to more recent debates about and defenses of tolerance. John Stuart Mill’s On Liberty, for instance, builds much of its case for a maximal liberty of thought and expression on a presumption that moral argumentation and pursuit of the truth are central features of human progress. In Mill’s view, extensive liberty should be accorded to individuals in order that small truths may be pried out of great falsehoods, as well as because of the perpetual need for well-established truths to retain their vitality and not descend into unthinking dogma.69 These arguments may 69 John Stuart Mill, On Liberty, ed. E. Rapaport (Indianapolis: Hackett, 1978), 36–41.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Modern Toleration through a Medieval Lens

21

be traced back to Mill’s broadly conceived utilitarianism: liberty of inquiry is useful to the pursuit of truth and the improvement of mankind, which together constitute the moral imperative of civilization. Mill thus sometimes stands accused of promoting a consequentialist defense of toleration, according to which individual liberty is valued only for the results it yields.70 The pitfalls of such a position have been exposed.71 But perhaps Mill can be rescued from such charges if we read him through the medieval framework previously articulated. Let us reframe Mill’s defense of individual liberty in essentially moral terms. On this account, human improvement is understood as a moral rather than a pragmatic goal. If this is so, then it constitutes a value that may be compared to others, that is, potentially competing expectations and/or requirements for human conduct demanded by institutionalized religions or the “forces of moral order.” In order to justify the liberty upon which human improvement depends, then, it is necessary to give reasons why those few individuals who “experiment” with their lives and intellects should be afforded tolerance. To the extent that the giving of normative reasons for human improvement, in contrast to salvation or moral conformity, is convincing, liberty of thought and opinion ought to be accorded to individuals who choose to use it. Note that this claim does not entail the demonstration of the absolute and unlimited superiority of the exercise of liberty. Indeed, nothing in Mill’s argument, construed in moral terms, necessitates that the state or any group of individuals must surrender their commitment to their own fundamental values. That is to say, those who adopt ends incompatible with human improvement are by no means required to bracket judgment when they tolerate beliefs and ways of life deemed otherwise abhorrent. Rather, insofar as the free exploration of “deviant” views is granted forbearance, it is because a greater injury would ensue from the prohibition and suppression of such activity. To the extent that the worthiness of human improvement may be rationally demonstrated, disapproval may not be translated into official criminalization: we have to tolerate beliefs and behaviors that we may consider to be profoundly untrue, immoral, evil, and so on just because a greater good is served or a greater harm is avoided by doing so. This is precisely the logic of “judgmental toleration”: so long as good moral reasons can be adduced for tolerance of a lesser evil, we are obliged to permit it, no matter how offended or disgusted we may be. Of course, “judgmental toleration” does not preclude open debate and dispute about the conflicting 70 Ibid., 10–11. 71 As by Richard H. Dees, “The Justification for Toleration,” in M. A. Razavi and D. Ambuel (eds), Philosophy, Religion, and the Question of Intolerance (Albany: State University of New York Press, 1997), 139–42.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

22

Cary J. Nederman

moral conceptions that justify the extension of toleration to individual variations of thought and action; indeed, it requires such on-going engagement concerning moral disagreements, as Mill acknowledges in On Liberty.72 Moreover, even the limits that Mill himself places on the uses of individual liberty, based on the harm principle, are not self-evident. Just as with any other moral norm, it is necessary to make a reasoned case for why direct harm is impermissible and thus intolerable. Mill’s own position would be self-contradictory if it failed to admit this. A similar point pertains to the nature of toleration as conceived by recent proponents of deliberative democracy, such as Amy Gutmann and Dennis Thompson. In Democracy and Disagreement, Gutmann and Thompson maintain that the very point of democracy is not to hold fundamental divisions among citizens about moral questions at bay, but instead to create a public realm capable of sustaining and nourishing deep and difficult debate about such critical issues. As a result, they reject the standard (“non-judgmental” liberal) principle of toleration on the grounds that it “would not go far enough for the purposes of deliberative democracy. It provides no basis on which citizens can expect to resolve their moral disagreements . . . . [M]ere toleration also locks into place the moral divisions in society and makes collective moral progress far more difficult.”73 In place of the liberal conception of tolerance, they propose a rather complex conception of mutual respect and reciprocity in order to achieve their purposes. Had they employed the elements of “judgmental toleration,” by contrast, they would have found a theory of tolerance far more appropriate to the needs of the deliberative democracy that they defend. Incorporating judgment as a feature of public discourse means that the mechanisms for the expression of disagreement and for mutual normative reason-giving are built into the processes of democratic deliberation. While liberal tolerance forecloses any opportunity for breaking down barriers to the sort of rational engagement capable of resolving moral disputes, even pro tempore, “judgmental toleration” provides a framework within which open reason-giving might fruitfully proceed. Such openness comports well with the uncertainty of democratic deliberation as conceived by Gutmann and Thompson. “Judgmental toleration” compels decision-making in the face of moral disagreement out into the public realm, which is precisely the vision of the democratic process promoted by them. Although the medieval thinkers surveyed above were scarcely advocates of democracy, they nonetheless offer

72 Mill, On Liberty, 16–21. 73 Amy Gutmann and Dennis Thompson, Democracy and Disagreement (Cambridge, MA: Harvard University Press, 1996), 62.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Modern Toleration through a Medieval Lens

23

illustrations of the range of moral issues (including those that involve a society’s core values) to which “judgmental toleration” may be applied.

5 . C O N C L U D I N G R E M AR K To continue to cling to the hegemonic secular, individualistic principle of “non-judgmental toleration” associated with mainstream liberalism seems folly in light of the real-world conditions of religious, ethnic, and cultural conflict that one encounters almost daily. In Europe and North America, no less than in the Middle East or Central and South Asia, fundamental disputes about the “right” way of life—what has sometimes, and problematically, been termed “the clash of civilizations”—have produced violence, hatred, and disrespect among peoples. It seems to me that liberalism’s insistence upon the absolute singularity of its particular account of tolerance only exacerbates many conflicts by removing from consideration the moral sources whence they emerge and are sustained. By contrast, we ought to keep in mind Michael Walzer’s guiding precept that tolerant individuals and communities “make room for men and women whose beliefs they don’t adopt, whose practices they decline to imitate; they coexist with an otherness that, however much [or little] they approve of its presence in the world, is still something different from what they know, something alien and strange.”74 Telling people that they must keep these differences to themselves—whether for the sake of human rights or modus vivendi or some other standard external to their concrete experiences of otherness— has proven futile and is likely to continue to do so. But, as I have tried to illustrate, the past, when liberalism was not ensconced as an article of faith in the Western world, offers us clues as to how we might move successfully beyond the present impasse.75 Texas A&M University 74 Michael Walzer, On Toleration (New Haven: Yale University Press, 1997), 11. 75 The original version of this paper was composed as a talk to the Department of Political Studies at the University of Otago, Dunedin, New Zealand in 2006. Later drafts were presented at Center for Thomistic Studies, University of St. Thomas, Houston in 2008 and the annual meeting of the Lone Star Chapter of the Conference for the Study of Political Thought, University of Houston in 2015, as well as to my colleagues in the Medieval Studies Working Group at Texas A&M University. The paper also formed the basis for my Medieval Academy of America Plenary Lecture at the 50th International Congress on Medieval Studies, May 2015, Western Michigan University. Thanks are due to Karen Bollermann for her expert intellectual and editorial advice, as well as to the anonymous reader for this journal for extensive comments and suggestions. Thiago Talzzia provided valuable assistance in preparing the final manuscript for publication.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

24

Cary J. Nederman BIBLIOGRAPHY

Bartolomé de Las Casas. In Defense of the Indians, ed. S. Poole (DeKalb, IL: Northern Illinois University Press, 1974). Bejczy, István. “Tolerantia: A Medieval Concept,” Journal of the History of Ideas 58 (1997), 365–84. Bollermann, Karen and Cary J. Nederman. “Standing in Abelard’s Shadow: Gilbert of Poitiers, the 1148 Council of Rheims, and the Politics of Ideas,” in K. Bollermann, T. M. Izbicki, and C. J. Nederman (eds), Religion, Power, and Resistance from the Eleventh to the Sixteenth Centuries: Playing the Heresy Card (New York: Palgrave/Macmillan, 2014), 13–36. Boswell, John. Christianity, Social Tolerance, and Homosexuality: Gay People in Western Europe from the Beginning of the Christian Era to the Fourteenth Century (Chicago: University of Chicago Press, 1980). Brunetto Latini. The Book of the Treasure, ed. P. Barrette and S. Baldwin (New York: Garland, 1993). Cadden, Joan. Nothing Natural is Shameful: Sodomy and Science in Late Medieval Europe (Philadelphia: University of Pennsylvania Press, 2013). Christine de Pizan. A Medieval Woman’s Mirror of Honor: The Treasury of the City of Ladies, ed. C. C. Willard (New York: Persea, 1989). Creppell, Ingrid. Toleration and Identity: Foundations in Early Modern Thought (London: Routledge, 2003). Dees, Richard H. “The Justification of Toleration,” in M. A. Razavi and D. Ambuel (eds), Philosophy, Religion, and the Question of Intolerance (Albany: State University of New York Press, 1997). della’Orto, Giovanni. “L’omosessualità nella poesia volgare italiana fino al tempo di Dante,” Sodoma 3 (1986), 13–35. Drury, Shadia B. Aquinas and Modernity: The Lost Promise of Natural Law (Lanham, MD: Rowman and Littlefield, 2008). Forst, Rainer. Toleration in Conflict: Past and Present, tr. C. Cronin (Cambridge: Cambridge University Press, 2013). Fotion, Nick and Gerard Elfstrom. Toleration (Tuscaloosa, AL: University of Alabama Press, 1992). Gutmann, Amy and Dennis Thompson. Democracy and Disagreement (Cambridge, MA: Harvard University Press, 1996). Halbertal, Moshe. “Autonomy, Toleration, and Group Rights,” in D. Heyd (ed.), Toleration: An Elusive Virtue (Princeton: Princeton University Press, 1996), 114–26. Johnsen, Albert R. and Stephen Toulmin. The Abuse of Casuistry: A History of Moral Reasoning (Berkeley: University of California Press, 1988). Johnson, Kristen Deede. Theology, Political Theory, and Pluralism: Between Tolerance and Difference (Cambridge: Cambridge University Press, 2007). Jordan, Mark D. The Invention of Sodomy in Christian Theology (Chicago: University of Chicago Press, 1997).

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Modern Toleration through a Medieval Lens

25

Kallendorf, Hilaire. Conscience on Stage: The Comedia as Casuistry in Early Modern Spain (Toronto: University of Toronto Press, 2007). Kay, Richard. Dante’s Swift and Strong: Essays on Inferno XV (Lawrence: The Regents Press of Kansas, 1978). King, Preston. “Review of Cary J. Nederman, Worlds of Difference,” American Political Science Review 95 (2001), 981–2. Koch, Bettina. “Marsilius of Padua on Church and State,” in G. Moreno-Riaño and C. J. Nederman (eds), A Companion to Marsilius of Padua (Leiden: Brill, 2012), 139–79. Laursen, John Christian and Cary J. Nederman (eds). Beyond the Persecuting Society: Religious Toleration before the Enlightenment (Philadelphia: University of Pennsylvania Press, 1998). Laursen, John Christian and Cary J. Nederman. “Problems of Periodization in the History of Toleration,” Storia della storiografia 37 (2000), 55–65. Lutz-Bachmann, Matthias and Alexander Fidora (eds). Juden, Christen und Muslime: Religionsdialoge im Mittelalter (Darmstadt: Wissenschaftliche Buchgesellschaft, 2004). Marsiglio of Padua. Writings on the Empire: Defensor Minor and De Translatione Imperii, ed. C. J. Nederman (Cambridge: Cambridge University Press, 1994). Mill, John Stuart. On Liberty, ed. E. Rapaport (Indianapolis: Hackett, 1978). Moreno-Riaño, Gerson. “Tolerance, Identity, and the Problem of Citizenship,” in G. Moreno-Riaño (ed.), Tolerance in the Twenty-First Century (Lanham, MD: Lexington Books, 2006), 81–96. Murphy, Andrew R. Conscience and Community: Revisiting Toleration and Religious Dissent in Early Modern England and America (University Park, PA: Pennsylvania State Press, 2001). Nederman, Cary J. Worlds of Difference: European Discourses of Toleration, c. 1100–c. 1550 (University Park, PA: Pennsylvania State Press, 2000). Nederman, Cary J. “Toleration in a New Key: Historical and Global Perspectives,” Critical Review of International Social and Political Philosophy 14 (2011), 349–61. Nederman, Cary J. “Toleration in Medieval Europe: Theoretical Principles and Historical Lessons,” in J. Muldoon (ed.), Bridging the Medieval-Modern Divide: Medieval Themes in the World of the Reformation (Farnham, UK: Ashgate, 2013). Nederman, Cary J and John Christian Laursen (eds). Difference and Dissent: Theories of Toleration in Medieval and Early Modern Europe (Lanham, MD: Rowman and Littlefield, 1996). Nicole Oresme. De Moneta, in C. Johnson (ed.), The De Moneta of Nicholas Oresme and English Mint Documents (London: Thomas Nelson and Sons, 1957). Oberdiek, Hans. Tolerance: Between Forbearance and Acceptance (Lanham, MD: Rowman and Littlefield, 2001). Patschovsky, Alexander and Harald Zimmermann (eds). Toleranz im Mittelalter (Sigmaringen: Thoerbecke, 1998). Ptolemy of Lucca. On the Government of Rulers, ed. J. M. Blythe (Philadelphia: University of Pennsylvania Press, 1997).

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

26

Cary J. Nederman

Richards, David A. J. “Toleration and the Struggle against Prejudice,” in D. Heyd (ed.), Toleration: An Elusive Virtue (Princeton: Princeton University Press, 1996), 126–35. Sandel, Michael. “Judgemental Toleration,” in R. P. George (ed.), Natural Law, Liberalism, and Morality (Oxford: Clarendon Press, 1996), 107–12. Schopp, Ludwig, Denis J. Kavanagh, Robert P. Russell, and Thomas F. Gilligan (eds), Divine Providence and the Problem of Evil (New York: Cosmopolitan Science & Art, 1942). Solari, Enzo. “Contornos de la Tolerancia Medieval,” Ideas y Valores 62 (2013), 73–97. Suárez, Francisco. De legibus (Madrid: Consejo Superior de Investigaciones Cientificas, Instituto Francisco de Vitoria, 1973). Svensson, Manfred. “A Defensible Conception of Tolerance in Aquinas?,” The Thomist 75 (2011), 291–308. Thomas Aquinas. On Law, Morality, and Politics, ed. W. P. Baumgarth and R. J. Regan (Indianapolis: Hackett, 1988). Tierney, Brian. Religion, Law, and the Growth of Constitutional Thought, 1150–1650 (Cambridge: Cambridge University Press, 1982). Tierney, Brian. “Religious Rights: A Historical Perspective,” in N. B. Reynolds and W. C. Durham Jr (eds), Religious Liberty in Western Thought (Grand Rapids, MI: Eerdmans Publishing, 1996). Tierney, Brian. Liberty and Law: The Idea of Permissive Natural Law, 1100–1800 (Washington, DC: Catholic University of America Press, 2014). Urbanati, Nadia. “Half-Toleration: Concordia and the Limits of Dialogue,” in A. Stepan and C. Taylor (eds), Boundaries of Toleration (New York: Columbia University Press, 2014). Walzer, Michael. On Toleration (New Haven: Yale University Press, 1997). Zagorin, Perez. How the Idea of Toleration Came to the West (Princeton: Princeton University Press, 2003).

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect Stephen R. Ogden

Despite some confusion in the primary and secondary literature, I think most scholars now agree that Averroes was, indeed, an “Averroist.”1 That is to say, in his most mature philosophical Long Commentary on Aristotle’s De Anima, Averroes (Ibn Rushd) argues not only (along with much of the Graeco-Arabic tradition) for a single, separate “active” or “agent” intellect, but also for the quite novel concept of a single, separate “material” or “possible” intellect (hereafter MPI) for all humankind.2 The philosophical significance of this view is not limited to its originality as an interpretation of Aristotle. It implies nothing less than evacuating individual human beings of intellect (properly speaking). Rather than seeing either the MPI or the active intellect as some intellectual faculty of our own, Averroes sends the two intellects into orbit, like satellites among the heavenly spheres which we “log 1 The primary text issue is that Averroes wrote his customary three commentaries (Short, Middle, Long) on Aristotle’s De Anima, and his view developed. His most mature and controversial thesis on the intellect appears only in the Long Comm. on De Anima (LCDA). On the debate in the secondary literature concerning whether Averroes was himself an “Averroist” (the term came to be used for a number of controversial doctrines associated with him among the Latin philosophers), see Gómez Nogales, “Saint Thomas, Averroès et l’Averroïsme,” in G. Verbeke and D. Verhelst (eds), Aquinas and Problems of His Time (Louvain: Leuven University Press, 1976); R. A. Gauthier, “Notes sur les début (1225–1240) du premier ‘Averroïsme,’ ” Revue des sciences philosophiques et théologiques 66 (1982), 321–74; Bernardo Carlos Bazán, “Le commentaire de S. Thomas d’Aquin sur le Traité de l’âme,” Revue des sciences philosophiques et théologiques 69 (1985), 521–47; and the conclusive summary in Richard Taylor’s introduction to his translation of Averroes’s Long Commentary on the De Anima of Aristotle (New Haven: Yale University Press, 2009), c–cii. 2 The active, agent, or maker intellect is found in Aristotle’s De Anima (DA) III.5 and is called [nous] poiētikos. The MPI is generally regarded as the intellect described in III.4 as being capable of receiving all intelligible forms, described as dunatos (possible). It is also compared to matter (hulē) at the beginning of III.5. Furthermore, many interpreters identify the MPI with the nous pathētikos at the end of III.5, but Averroes does not.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

28

Stephen R. Ogden

on to” in order to play a role in individuated acts of understanding. Ever since, this idea has been the source of much philosophical debate, some admiration, and plenty of scorn. After a brief introduction to Averroes’s view, someone once asked me, “Why would anyone want to think that?” Certainly the defender of Averroes may feel like David Lewis presenting his view of possible worlds, sharing as perhaps their most commonly faced objection the infamous “incredulous stare.”3 Nonetheless, it stands to reason that Averroes (like Lewis well after him) may provide interesting arguments in favor of his position—not only strong textual arguments from the De Anima, but also independent philosophical arguments. We have been treated in recent years to enlightening explanations of Averroes’s overall position by scholars such as Herbert Davidson, Deborah Black, and Richard Taylor, but the possible philosophical arguments for the view as presented by Averroes have been either entirely overlooked or discussed only briefly—and, even in the latter case, conflated and misunderstood.4 So here I aim to provide a new, thorough analysis of one prominent text which some scholars (such as Taylor, Black, and a younger

3 Lewis, Counterfactuals (Cambridge, MA: Harvard University Press, 1973), 86. 4 For good accounts of Averroes’s overall view, see Herbert Davidson, Alfarabi, Avicenna, and Averroes on Intellect (Oxford: Oxford University Press, 1992), esp. ch. 7; Deborah Black, “Consciousness and Self-Knowledge in Aquinas’s Critique of Averroes’s Psychology,” Journal of the History of Philosophy 31 (1993), 349–85, and “Models of Mind: Metaphysical Presuppositions of the Averroist and Thomistic Accounts of Intellection,” Documenti e studi sulla tradizione filosofica medievale 15 (2004), 319–52; many articles by Taylor, esp. “Averroes’ Epistemology and Its Critique by Aquinas,” in R. E. Houser (ed.), Medieval Masters: Essays in Memory of Msgr. E. A. Synan (Houston: Center for Thomistic Studies, 1999), “Intellect as Intrinsic Formal Cause in the Soul According to Aquinas and Averroes,” in M. Elkaisy-Friemuth and J. Dillon (eds), The Afterlife of the Platonic Soul: Reflections of Platonic Psychology in the Monotheistic Religions (Leiden: Brill, 2009), and the introduction to Averroes’s Long Commentary (n. 1); and also Bazán, “Intellectum Speculativum: Averroes, Thomas Aquinas, and Siger of Brabant on the Intelligible Object,” Journal of the History of Philosophy 19 (1981), esp. 435–6. Davidson sets out to give a comprehensive account of Averroes’s position, yet surprisingly nowhere discusses any philosophical arguments for his view, only how he criticizes and adapts previous commentary on Aristotle. Black and Bazán also do not discuss Averroes’s positive arguments, only his defense against objections. Taylor conflates what I consider two separate arguments in “Averroes’ Epistemology,” 162 and in “Intelligibles in Act in Averroes,” in J. B. Brenet (ed.), Averroès et les averroïsmes juif et latin (Turnhout: Brepols, 2007), 130. Although he rightly and clearly distinguishes the arguments in “Aquinas and ‘the Arabs’: Aquinas’s First Critical Encounter with the Doctrine of Averroes on the Intellect, In 2 Sent. d. 17, q. 2, a. 1,” in L. X. López-Farjeat and J. Tellkamp (eds), Philosophical Psychology in Arabic Thought and the Latin Aristotelianism of the 13th Century (Paris: Vrin, 2013), 153–4—and less clearly in “Averroes’ Philosophical Conception of Separate Intellect and God,” in A. Hasnawi (ed.), La lumière de l’intellect (Leuven: Peeters, 2011), 395–6—they are not discussed in any detail.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

29

version of myself )5 have taken to constitute an important independent philosophical argument for Averroes’s single separate MPI—I call it the Determinate Particular Argument because it argues that the MPI cannot be a determinate particular if it is to be the ontological receptacle of universal intelligible forms (or concepts). I will reveal several aspects of the argument heretofore unobserved and overturn one mistaken reading.6 I ultimately argue, contra the aforementioned scholars and some misleading indications in the text itself, that the Determinate Particular Argument should not be taken as a sufficient and independent argument for Averroes’s full thesis on the MPI; only a critical part of it. I begin by reconstructing the Determinate Particular Argument, and I demonstrate that there is much more to it than even defenders have previously seen. After offering support for one of its most crucial premises, however, I proceed to highlight a way in which ‘determinate particular’ must be qualified to avoid an inconsistency with Averroes’s substance metaphysics. My argument there will require a brief excursus into his thoughts on certain complex problems in Aristotle’s Metaphysics regarding substance and particularity, but the results of that inquiry will confirm the need for a revision. The objection from Averroes’s metaphysics and a subsequent consideration of his views of species membership for separate substances like the MPI will lead us to the solution, namely a qualified understanding of ‘determinate particular’ as it features in the Determinate Particular Argument. My final assessment, however, is that the qualified understanding would considerably weaken the argument in the face of competing views on the MPI, like those of Avicenna (Ibn Sīnā) and Aquinas, at least if taken on its own as the sole support for Averroes’s complete position. Therefore, I conclude by looking at what precisely the argument does prove for Averroes and where it fits into the larger philosophical case for his view. A final introductory note. Critics of Averroes, like Aquinas, often develop objections from what we might call “the human side” of the debate, arguing that Averroes’s position cannot make sense of human knowledge, human nature, and so forth. These issues are worth debating, which I do elsewhere, and Averroes does have responses to them. Yet, Black has rightly stressed

5 See Taylor, “Averroes’ Epistemology,” 159–61; “Intelligibles in Act,” 130; Long Commentary, lviii; and “Aquinas and ‘the Arabs,’ ” 153. Also Black, “Mental Existence in Thomas Aquinas and Avicenna,” Mediaeval Studies 61 (1999), 75, fn. 102—she calls it “Averroes’s principal argument.” 6 More specifically, I will argue later that at least one aspect of Taylor’s treatment, regarding species membership, should be revised. I also show how Averroes’s metaphysics are crucial to interpreting this argument in the psychology of the LCDA.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

30

Stephen R. Ogden

that to evaluate Averroes fairly we must also turn to the objects of knowledge rather than the supposed knowing subject(s).7 Thus, I concentrate precisely on those objects, the universal intelligibles, since that is where we shall find Averroes’s primary philosophical motivation. Generally, then, I hope this paper will contribute towards a better grasp of the strengths and weaknesses of Averroes’s reasoning than previous defenders, detractors, and (certainly) incredulous stare-ers.

1 . O U T L I N I N G T H E D ET E R M I N A T E P A R T I C U L AR A R G U M E N T It is ironic that Averroes (along with others in the Peripatetic tradition) calls the intellect capable of receiving intelligible forms in Aristotle’s DA III.4 the “material intellect,” even though he takes the arguments in that chapter to prove that the MPI is robustly immaterial (being separate, unmixed, impassible, without a bodily organ, etc.).8 It not only leads to a paradoxical name; it highlights a tension in the original text, namely that in Aristotle’s system a being is material just in case it has some potentiality.9 This gave rise to what Averroes calls “the question of Theophrastus,”10 which is really a bundle of

7 Black, “Models of Mind.” 8 Of course, the source of the name is Aristotle’s own comparison of the MPI’s “becoming all things” to matter in DA III.5, 430a10. One might argue that incorporeal is the more strictly accurate terminology for Aristotle instead of immaterial. However, Aristotle does also say that the MPI is “without matter” (aneu hulēs) (430a2–3), and Averroes describes it as “altogether unmixed with matter” and not even a power or form in the body (LCDA, 386 [Eng., 302]; cf. Arabic fragments in Taylor, Long Commentary, 302–3, fns. 28 and 29). Thus, I will continue to speak of the MPI as immaterial. 9 He implies this, for example, in Phys. II.2, 193b6-8. Cf. Met. VIII.6. It is true that the distinction between incorporeal and immaterial mentioned earlier might help Aristotle in this regard, as he seems to make room for non-bodily albeit “intelligible matter” (hulē noētē)—see Ross in Aristotle’s Metaphysics (Oxford: Oxford University Press, 1958), vol. 1, cxl. In fact, Averroes seems to have his own notion of intelligible matter which will be mentioned later (see n. 100). However, the problem is still pointed because apparently even so close a contemporary as Theophrastus (see later) was puzzled about incorporeal beings which are capable of being affected in some way. 10 Averroes, LCDA, bk. III, comm. 5, Latin 399 [Eng., 315]. N.B. For primary sources in this paper, I often list the page number of the edition of the original (or extant ancient) language, followed by the page number of a standard English translation and any other translation (where available) in square brackets. In many of these cases I have made my own modifications to the translation. Otherwise, all translations are my own, i.e., when lacking a bracketed page number. See the bibliography for the relevant editions and translations. The LCDA survives in toto only in Latin, along with a few fragments in Arabic.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

31

questions, including, most importantly for our purposes, how the MPI can be both incorporeal and yet somehow in potency.11 The task, then, for Averroes is to show how the MPI is both receptive and immaterial: “Now we must consider . . . these two claims about the intellect, namely, that it is in the genus of passive powers and that it is unchangeable because it is neither a body nor a power in a body. For these two claims are the principles of all the things which are said about the intellect.”12 He, like Aristotle, starts with the MPI’s receptive potency for intelligible forms and compares it to the sense faculty’s reception of sensible forms. This potency for receiving all the intelligible forms constitutes the MPI’s definitive nature.13 However, given this definition of the MPI in terms of pluri-potency (like a kind of intellectual stem cell), Averroes must be clear to distinguish it (at least) from prime matter,14 and it is in this context that he gives what I call the Determinate Particular Argument: Since that is the definition of the material intellect, it is evident that according to him it differs from prime matter in this respect: it is in potency all the intentions15 of the universal material forms, while prime matter is in potency all those sensible forms [and is not] something which knows or apprehends. The reason why that nature [the MPI] is something which discerns and knows while prime matter neither knows nor discerns, is because prime matter receives diverse forms, namely, individual and particular forms (individuales et istas), while this [the MPI] receives universal forms 11 Theophrastus in Themistius, De Anima Paraphrasis, 108.1–3 [Eng., 133; Ar., 196]: “How can [the potential intellect] become the objects of thought, and what is [the activity] of being affected them? . . . But what affection [is produced] on an incorporeal [object] by an incorporeal [object]?” For a good discussion of Averroes on the question(s) of Theophrastus, see Taylor, Long Commentary, lxxx–lxxxi. 12 Averroes, LCDA, III, c. 4, 384 [301]. Something that should make Averroes’s question easier, here, is that Aristotle at least does not seem to think certain kinds of actualization of potentiality are strictly speaking a “change” (kinēsis). See Aristotle, DA II.5 and Burnyeat, “De Anima II 5,” Phronesis 47 (2002), 28–90. 13 “[T]he definition of the material intellect, therefore, is that which is in potency all the intentions of universal material forms and is not any of the beings in act before it understands any of them”—LCDA, III, c. 5, 387 [304]. Averroes also sometimes calls an important subset of the intelligible forms the “intelligible material forms” or “universal material forms” since they are abstracted from the encounter with the forms of composite material substances. 14 This also is one of the questions Theophrastus mentions as cited earlier. 15 An “intention” (maʿnā or intentio), although it has a notoriously wide range of meanings in Arabic, is in this context “a form or essence insofar as it is apprehended by any cognitive faculty and serves as an object for that faculty”—Black, “Psychology: Soul and Intellect,” in P. Adamson and R. Taylor (eds), The Cambridge Companion to Arabic Philosophy (Cambridge: Cambridge University Press, 2005), 311. It is what we might call the content or information borne by a certain particular or universal form. I discuss this further in a later note.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

32

Stephen R. Ogden

( formas universales). From this it is apparent that this nature is not a determinate particular (aliquid hoc) nor a body nor a power in a body. For if it were so, then it would receive forms inasmuch as they are diverse and particular (istas); and if it were so, then the forms existing in it would be intelligible [only] in potency; and thus it would not discern the nature of the forms inasmuch as they are forms, just as is the disposition in the case of individual forms, be they spiritual or corporeal.16 For this reason, if that nature which is called intellect receives forms, it must receive forms by a mode of reception other than that by which those matters receive the forms whose contraction by matter is the determination of prime matter in them. For this reason it is not necessary that it be of the genus of those matters in which the form is included, nor that it be prime matter itself. Since if this were so, then the reception in these would be of the same genus; for the diversity of the received natures causes the diversity of the nature of the recipient. This, therefore, moved Aristotle to set forth this nature, which is other than the nature of matter, other than the nature of form, and other than the nature of the composite.17

By the end, it is evident that Averroes is doing something quite novel with his Aristotelian sources. The important part of the passage for my purposes is the argument that the MPI cannot be a determinate particular or aliquid hoc—literally “some this,” normally used to pick out some individual, particular substance. While there is much to this passage, the crux of the argument is rather simple: how could the ontological receiver of universal forms itself be particular, since universals and particulars are thought to be contradictory? Here is a neater reconstruction of the argument as I see it in the passage: (1) The MPI receives only universal, intelligible forms.18 [def. of MPI] (2) If the MPI were a determinate particular (aliquid hoc), it would receive only particular forms. (3) It doesn’t receive particular forms. 16 I.e., both the forms received in sensation (spiritual forms) and the forms that naturally inhere in subjects and prime matter (corporeal forms). It is clear here that Averroes is extending his argument beyond the original aim of simply distinguishing the MPI from prime matter. 17 LCDA, III, c. 5, 387–8 [304–5]. 18 I am using Averroes’s language here, but it is ambiguous. What I think he means and what the argument should rely on is only that the received forms (universal or particular) are representing forms which have certain universal or particular content. That content or information is what he describes at the beginning of the passage as “the intentions of the universal material forms.” If Averroes were assuming that the received forms have their universality or particularity as entities per se, it would beg the question against other positions like those of Avicenna and Aquinas which assume, given individual intellects, that discrete acts of understanding are ontologically particular accidents (albeit with universal content) inhering in distinct human minds. I thank Gyula Klima for reminding me of this point.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

33

i. It receives only universal forms [(1)]. Universal forms are not particular forms. Therefore, (3). ii. If it received particular forms, those forms would be intelligible only in potency. But the forms are actually (not just potentially) intelligible [(1), and the non-equivocal categorization of the MPI as an intellect19]. Therefore, (3) (4) Therefore, the MPI is not a determinate particular (aliquid hoc). [(2), (3)] So far we have an argument that the MPI is not a determinate particular. However, the way the text proceeds shows immediately why it seems like an argument for the Averroistic doctrine par excellence—i.e., that the MPI is a separate substance, another incorporeal celestial intellect. He continues in the next sentences: This same [argument] brought Theophrastus, Themistius, and several commentators to hold the opinion that the material intellect is a substance which is neither generable nor corruptible. For every generable and corruptible thing is a determinate particular (hoc). But it was just demonstrated that this [the MPI] is not a determinate particular (hoc), nor a body, nor a form in a body.20

For continuity’s sake, then: (5) If a substance is generable and corruptible, it is a determinate particular. [accepted Aristotelian doctrine] (6) The MPI is not a determinate particular. [(4)] (7) Therefore, the MPI is a substance21 neither generable nor corruptible. [(5), (6)] (8) Therefore, the MPI is an immaterial, separate celestial substance. [(7), accepted Aristotelian doctrine] Thus we see how the universality of intelligible forms, featuring in (1), leads Averroes to the conclusion of a substantially separate MPI in (8). 19 Averroes suggests this when he says in the passage, “For this reason, if that nature which is called intellect receives forms, it must receive forms by a mode of reception other than that by which those matters receive the forms . . . ” (emphasis mine). It is clear enough from the context in the DA that Aristotle is explaining a different activity from sensation, which consists in the reception of particular, sensible forms. 20 LCDA, III, c. 5, 389. In the Arabic fragments, in Taylor, Long Commentary, 305, fn. 39, corresponding to this passage, the only authority listed here is Aristotle himself. The others are discussed in detail by Averroes after this point in Book III, and he mentions Themistius and other “ancient commentators” in the Short Comm. DA (SCDA), 83, on a similar point regarding the MPI’s being neither generable nor corruptible. 21 See my discussion later regarding a shortcut here; it should strictly substitute “a thing” for “a substance” here in premise (7) and premise (5).

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

34

Stephen R. Ogden

I will go on to explain premise (2) more in what follows, and I think much of Averroes’s (rather fluid) terminology and overall position will be known to many readers. However, (5)–(8) significantly shape the way in which the argument can be interpreted as proving Averroes’s full thesis on the MPI, so it is worth briefly clarifying that thesis and these steps towards it. What Averroes ultimately must demonstrate for his view of the MPI is that it is a single, immaterial, separate, celestial substance. It is single or unique insofar as there is only one which is shared by all human beings.22 I have touched on immateriality earlier in my notes, but we may add here that this means in the case of the MPI that it is fully incorporeal, lacking both sensible matter and topical matter (the latter of which Aristotle assigns to the eternal celestial bodies with respect to their change in place).23 The closely related term “separate” for Averroes seems to mean eternally separate from bodies and all material forms,24 including separate from celestial bodies and motion.25 Of course, the only beings that meet these requirements in the Aristotelian-Arabic system are eternal unmoved movers, or celestial intellects. Averroes in the latter part of the argument takes a few shortcuts, but largely obtains most of these requirements for the MPI. First, on strict logical grounds and confined only to the passage at hand, we should substitute “a thing” for “a substance” in premise (7), and perhaps also in premise (5). Averroes in the passage switches from “substance” to “thing,” but his identification of the MPI as a separate substance, as we have seen, is most pertinent. We could easily add a premise, “The MPI is a substance,” 22 Despite the naturalness of thinking of the MPI’s singularity and uniqueness as universal in the context of this argument (i.e., as contrasted with particulars), Averroes never to my knowledge speaks of the MPI or its uniqueness using the term “universal.” That term is limited to the intelligible forms and their content, as already mentioned. I will discuss this point further later, especially in section 3. 23 See, among other places, Aristotle, Met. XII.2, 1069b24–6. 24 Averroes, LCDA, III, c. 4, 385–6 [302–3]. This is highlighted by Averroes in the earlier quote, that the MPI is not “a body, nor a form in a body.” 25 Ibid., c. 5, 403 [318–19]; Long Comm. Met. (LCM), vol. IV, Λ, c. 5—on Aristotle Met. XII.1, 1069a30–34, where Aristotle says there are three kinds of substance: perishable sensible substances, eternal sensible substances (celestial bodies), and immovable substances—esp. 1427/r: “He established that the substances are three and that one of them is the separate (al-mufāriq) [kind].” It is clear that “separate” here corresponds to Aristotle’s category of the immovable substances. Cf. Met. XII.1, 1422/f, ln. 4. Note here that Averroes’s understanding of the MPI’s separation is stronger than other medieval philosophers who choose to emphasize the legitimate modal reading of Aristotle’s Greek chōristos as “separable”—for instance, Thomas Aquinas, who thinks the rational human soul is separable from the body (capable of separate existence), but not like an eternal and ungenerated separate substance. Though that modal reading is possible for the Arabic as well (and this is occasionally brought out in the Latin translation of the LCDA), I choose “separate” here to underscore this particularly strong aspect of Averroes’s own thinking.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

35

since, in addition to his argument from authorities in this passage, he also previously argued for the MPI’s standing as a substance (from Aristotle’s argument that the MPI is unmixed in DA 429a18-20).26 Second, the move from (7) to (8) is secured because only celestial substances are non-generable and incorruptible on standard Aristotelian doctrine.27 Since the MPI is by definition an intellect, as in (1), it cannot be a celestial body; hence, it will also be an immaterial and separate substance in the robust way outlined above—i.e., an unmoved, eternal celestial intellect. However, it is crucial from my perspective that Averroes, thus far, does not seem to explicitly argue that the MPI is single or unique. For all that is said here, there might, in principle, be many of these separate MPIs. Perhaps its singularity could be implied, but that would largely depend on how Averroes understands the denial that the MPI is a determinate particular and the nature of celestial intellects, topics to which I will return in earnest later.

2 . D E F E N D I N G P R E M I S E (2 ) Is this a good argument? As we’ve already witnessed, certainly not all of these premises are uncontroversial, especially premises (5)–(8), but the only premise for which I refrained from giving at least some support was the (rather important) conditional (2). Why does Averroes think that any aliquid hoc must receive only determinate and particular forms? I think there are at least two possible principles in the background to which he may be appealing. The first is a principle that Averroes would have known through the Kalām fī Mah.d. al-Khayr (Discourse on the Pure Good ), called in the Latin tradition the Liber de Causis. It was originally falsely attributed to Aristotle but actually largely based on Proclus’s Elements of Theology. I will call the principle in question “the Receiver Principle,” and it is stated in the Liber de Causis (LDC ) as: “And similarly everything only receives (yunālu) what is higher in the mode in which it is able to receive it, not in the mode of the thing received.”28 Or, to use Aquinas’s catchier version: “whatever is 26 LCDA, III, c. 4, esp. 385–6 [302–3]. 27 Again, see Met. XII.1, 1069a30–34, mentioned in n. 25, and De Caelo I.10–12 and II.1. 28 Liber de Causis (LDC ), prop. 9, Arabic, 81 [Lat., 174]. My translation from the Arabic. For an English translation of the Latin, see the translation by Guagliardo, et al., of Aquinas’s Commentary on the Book of Causes (Washington, DC: Catholic University of America Press, 1996), prop. 10, 75. This entire proposition is based on Proclus, Elements of Theology, prop. 177, though, as Guagliardo, et al. point out on 74, fn. 1, the Receiver Principle is more explicitly stated in Proclus, ET, prop. 173, 150 [151]: “For each principle participates its superiors in the measure of its natural capacity (hōs pephuken), and not in the measure of their being . . . participation varies with the distinctive character

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

36

Stephen R. Ogden

received is received according to the mode of the receiver.”29 Clearly, this Receiver Principle, if true, would justify (2), as (2) would just be a specific instance of the principle, stating that if something is a determinate particular, then it will receive determinate and particular forms (i.e., in the mode of the receiver).30 In at least one place (Sent. II.17), Aquinas considers virtually this exact Determinate Particular Argument from Averroes’s LCDA, but states it as explicitly relying upon the Receiver Principle: “Moreover, everything which is received in something is received in it through the mode of the recipient, not through its own mode, as it is held by Dionysius and the Book of Causes.”31 To my knowledge, no one other than Aquinas and myself have discussed the possibility that Averroes might hold this NeoPlatonic principle for purposes of his argument. But perhaps Aquinas’s reading is mistaken. Deborah Black, in expounding a different part of the LCDA, argues that Averroes does not adopt this principle, even though she says (rightly, I think) that the Receiver Principle would be his most economic option for explaining certain features of his cognitive psychology.32 She quotes a passage on sensation in the LCDA, Bk. II, in which Averroes resists using the Receiver Principle in order to solve an aporia regarding the intentionality of sensation:

and capacity (kata tēn idiotēta . . . kai dunamin) of the participants.” Versions of the principle appear in other places in the LDC, including propositions 11 [Eng., prop. 12], 19 [20], and 23 [24]. 29 See Guagliardo, et al., Book of Causes, xxvi, fn. 55 and Wippel, “Thomas Aquinas and the Axiom ‘What Is Received Is Received According to the Mode of the Receiver,’ ” in R. Link-Sallinger (ed.), Studies in Medieval Philosophy and Culture (Washington, DC: Catholic University of America Press, 1988). 30 See again n. 18. Again, the argument relies on these particular forms being particular at least in content. While the Receiver Principle may also apply to entities per se, the important point here is that it can be and often is applied to the content of the form received in different types of cognition (and the same goes, mutatis mutandis, for the other principle, the Identity Principle, discussed further later in the paper). This will be made clear in Averroes’s case, since he at least considers the Receiver Principle as a means of explaining the intentions or intentional being of sensible forms, which plainly deals with content, as also noted in n. 15. Although, as we shall see, he seems to reject the Receiver Principle there, it is for a different reason. One might worry that the principle is thereby committing a version of what has been called the “content fallacy,” much discussed in particular connection with Aquinas—see esp. Pasnau, “Aquinas and the Content Fallacy,” The Modern Schoolman 75 (1998), 293–314. I cannot go into the debate here or exactly how Averroes’s premise (2) differs from similar moves featured in Aquinas. However, I will briefly give some further philosophical defense for the content reading of premise (2) at the end of this section. 31 Aquinas, Sent. II, d. 17, q. 2, a. 1, obj. 3. 32 Black, “Averroes on the Spirituality and Intentionality of Sensation,” in P. Adamson (ed.), In the Age of Averroes: Arabic Philosophy in the Sixth/Twelfth Century (London: Warburg Institute, 2011), 172–3.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

37

One can say that sensibles do not move the senses in accord with the way they exist outside the soul, for they move the senses insofar as they are intentions, since in matter they are not intentions in act, but in potency. And one cannot say that this difference occurs by virtue of the difference of subject such that the intentions come to be on account of a spiritual matter which is the sense, not on account of an external mover. For it is better to think that the reason for the difference of matter is the difference of forms, rather than that the difference of matter is the reason for the difference of forms.33

Without getting into the details, the important point for our purposes is that, as Black says, Averroes denies the idea: . . . that we can explain intentional being simply by appealing to the special nature of the patient receiving the intention. . . . Averroes, then, is effectively rejecting the maxim that ‘whatever is received is received according to the mode of the receiver.’ [That would be] . . . to reverse the order of priority between form and matter. Matter is for the sake of form, and form has explanatory and causal priority within an Aristotelian framework.34

There is also evidence in our main text, cited earlier, that he does not hold the Receiver Principle, and for the same reason: “ . . . for the diversity of the received natures causes the diversity of the nature of the recipient.”35 This is, of course, just the converse of the Receiver Principle. So, although the Receiver Principle would clearly support (2), there seems to be good evidence, even in the very same argument, that Averroes would not hold (2) on these grounds. The other principle to which Averroes might be appealing is a prominent one within Aristotle’s De Anima itself, namely that of the identity of the knower and known, which I shall call the “Identity Principle.” As we will see, though the route is a bit more circuitous, the Identity Principle may also support (2), and, in fact, could be part of the historical and logical origins of the Receiver Principle itself. Aristotle lays out the Identity Principle in several places, but most succinctly at DA III.5, 430a20 (“Actual knowledge is the same as its object”).36 I cannot hope to wholly clarify this Aristotelian principle here, but III.4, 430a3–5 sheds a bit more light: “For in the case of things without matter, what understands and what is understood are the same. For theoretical knowledge and what is thus known are the same.” Recall in DA III.4 that the nature of the MPI is said to be a kind of pure potentiality for receiving intelligible forms. For the MPI to receive a 33 LCDA, II, c. 60, p. 221 [172], emphasis mine. 34 Black, “Averroes on the Spirituality,” 173. 35 “ . . . diversitas enim nature recepti facit diversitatem nature recipientis.” LCDA, III, c. 5, 388 [305]; cf. Black, “Averroes on the Spirituality,” 173, fn. 44. 36 See also DA III.7, 431a1–2 and 431b17, and Met. XII.7, 1072b19–21.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

38

Stephen R. Ogden

form is for it to be actualized by that form, and, indeed (having no other nature or matter to it), it becomes that form; it becomes the object of understanding.37 The Identity Principle, then, is stronger than the Receiver Principle, and the latter may rest on the former. For, at least insofar as understanding is concerned, if the knower and known are identical then that means the form received will be received in the mode of the receiver; indeed, that mode is somehow identical. But in the same way, the Identity Principle will also imply the Receiver Principle’s converse (what is received, is received according to the mode of the received) for the received and the receiver are again identical! Therefore, it is more likely that Averroes is drawing on the Identity Principle since he states the converse of the Receiver Principle in the main passage.38 Indeed, if the Identity Principle entails both the Receiver Principle and its converse, then Averroes’s apparent rejection of the Receiver Principle above must be qualified in some way.39 Although the Identity Principle implies a similar mode of existence for intellect and its universal objects, it doesn’t necessarily apply to determinate particulars as featured in (2). However, this extension is justified when also considering Aristotle’s remarks about the close analogy between understanding and sensation and the definition of their objects: “Just as sensation is related to the sensible, so also intellect is related to the intelligible” (429a17–18). In fact, in DA III.8, Aristotle provides more clarification on

37 For more analysis on this principle in DA and its application in the Met., see Defilippo, “Aristotle’s Identification of the Prime Mover as God,” The Classical Quarterly 44 (1994), 405ff. 38 Taylor, Long Commentary, xxi, fn. 14, suggests a related though slightly different point about applications of the Identity Principle and its corollaries in a brief discussion of the LDC. He explains that the LDC argues in prop. 6 [Eng., prop. 7] that because intellects are indivisible and immaterial, they will not be able to know extended, material particular substances, perhaps because of a lack of proper identity in the two modes of existence. Taylor conjectures: “This seems to indicate an awareness on the part of the author . . . of De Anima 3.7, 431b17-19,” i.e., the Identity Principle. Furthermore, as Taylor notes, the converse of the LDC’s conclusion about intellect not being able to know extended particulars would imply that “insofar as human beings are divisible particular entities in bodies, they cannot know separate entities with an identity of knower and known while in the body.” This is clearly where Averroes ends up and is further suggestion that something like the Identity Principle may be behind his thinking (even if he was also influenced by the LDC ). 39 My suggestion would be that while his acceptance of the Identity Principle will commit him to the truth of both the Receiver Principle and its converse, he might be opposed at times to the Receiver Principle as the best explanation of truths like premise (2). In support of this conclusion, although Averroes claims in the passage at n. 33 that one “cannot say” the Receiver Principle, his further argumentation regards only what it is “better to think.”

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

39

the Identity Principle and seems to envisage its holding somehow for both knowledge and sensation: In the soul the powers of sensation and knowledge are potentially these [objects], the knowable object and the sensible object. But they are necessarily either things themselves or the forms. But they are clearly not the things themselves, for the stone is not in the soul, but rather the form.40

This analogy, then, seems to suggest that there must be a kind of similar formal identity between, on the one hand, intellect and intelligibles, and, on the other, sensation and sensibles.41 This formal identity between the two faculties and their objects is in neither case numerical or material identity. Aristotle states that sense is “what has the power of receiving into itself the sensible forms of things without the matter. . . . ”42 And we have already seen that the identity between knower and known is also without matter. More specifically, what this formal identity should entail is the reception in the relevant faculty of a numerically distinct form of the same essence or species as the corresponding form instantiated in the external object in the world.43 40 Aristotle, DA III.8, 431b26–432a1. 41 Black, “Conjunction and the Identity of Knower and Known in Averroes,” American Catholic Philosophical Quarterly 73 (1999), 167, shows that Averroes applies the same relation of “conjunction” (ittis.āl / continuatio) not only to intellect, but also to sensation. We see this sort of extension of the Identity Principle to both sensation and understanding in other, earlier Arabic philosophers, such as al-Kindī—On the Intellect, par. 4, 355 [96–7]. Avicenna also seems to hold something close to premise (2)—e.g., in Najāt (The Salvation), II.6.9, 217 [50]. 42 Aristotle, DA II.12, 424a17–19. Smith’s translation in The Complete Works of Aristotle. 43 Though passages from Aristotle (such as the last one quoted) and Averroes could be interpreted differently—i.e., to read that the form received is numerically identical—the above interpretation is a common one, and I think it is the more plausible and implied by other texts, at least in Averroes. As will be mentioned again later, Averroes thinks the principle of individuation for forms of the same essence or species is matter, which is why I equate numerical and material identity. So the particular form of green in the crown of broccoli in front of me, on this principle, will be numerically distinct from the sensible form of that green in the different subject of my sense faculty (which obviously does have a corporeal and material organ), and from the intelligible universal form of greenness in the intellect. This is already suggested by the distinction between the natural inherence of the (same species) form in the world and the intentional reception of it in a cognitive faculty. As Averroes says, “in this way each of the senses is affected by the things by which they are naturally constituted to be affected, either by color or by sound, but it is not affected by the thing according to which there is color or sound. For, if it were so, it would happen that when it would receive it, it would be a color or a sound, not an intention. . . . For the intention of color is other than color”—LCDA, II, c. 121, 317 [243] (on Aristotle, DA 424a17–24); cf. LCDA, II, c. 62, 223 [174] (on Aristotle, DA 417b32-418a6) and Black, “Averroes on the Spirituality,” 172. Numerical distinctness is also strongly implied in the passages at n. 33 where Averroes seemed to resist the Receiver Principle, since he speaks there of a difference of forms (plural) as the more proper cause for the difference of the same (in species)

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

40

Stephen R. Ogden

The formal identity, then, is an identity in this wider sense of sameness in form, i.e., of the same essence or species form. The Identity Principle, if extended to sensation as the DA III.8 quote directs, tells us that there must be some identity between the receiver and received in both understanding and sensation. In both, the identity can be described as formal identity (identity in essence or species) since it is not numerical/material in either case. But there must be some difference between sense and intellect, and the primary way Aristotle draws the distinction is by defining them exclusively in terms of the particularity or universality of their received objects: “actual sensation is of the particular (kath hekaston), but knowledge is of the universal (katholou).”44 That distinction also means the formal identity between a sense faculty and the sensible form cannot be exactly the same formal identity as in understanding, which is somehow more immaterial.45 That is why many medieval philosophers speak of a third kind of existence between the material and the fully immaterial or separate (like the MPI and the intelligible form)—it is often called “spiritual” being.46 Averroes upholds this ordering of types of being for each faculty: sensible form’s existence in potency (in the external object) and, on the other hand, in act (in the sense faculty). I do think there is another prima facie tension here, especially considering matter as the principle of individuation, with Averroes’s apparent rejection of the Receiver Principle above, but this can be explained by his thinking that the latter is not the best explanation of certain cognitive facts, even if it is true. See n. 39. 44 DA II.5, 417b22–23. Averroes’s Latin text in the LCDA has particularia and universalia, and the Arabic of the Middle Comm. DA (MCDA), 63, has juz ʾīya and kullīya. Cf. Aristotle, DA III.4, 429a16–17. 45 Note that the use of “immaterial” here is tracking Aristotle’s usage of receiving the form “without the matter” and a lack of what I have called material or numerical identity. It is a weaker and different sense than the use of “immaterial” used elsewhere in the paper to describe the MPI—which I shall distinguish immediately below as fully immaterial or separate. Indeed, although the sense faculty receives the form without the matter, it still has a corporeal and material organ, which on my view, does undergo some material changes during sensation. 46 I cannot here discuss the large debate in the contemporary interpretation of Aristotle on sensation between so-called “literalists”—e.g., Sorabji, “Body and Soul in Aristotle,” Philosophy 49 (1974), 63–89—and “spiritualists”—e.g., Burnyeat, “Is an Aristotelian Philosophy of Mind Still Credible? A Draft,” in M. Nussbaum and A. Rorty (eds), Essays on Aristotle’s De Anima (Oxford: Oxford University Press, 1992), 15–26, and Burnyeat, “De Anima II 5.” I will simply claim here that it is far from obvious that medieval philosophers like Averroes who posit “spiritual being” for the received sensible form are “spiritualists” in the contemporary understanding; indeed, I think Averroes is not. For an overview of the debate and some mention of the medieval position, see Victor Caston, “The Spirit and the Letter: Aristotle on Perception,” in R. Salles (ed.), Metaphysics, Soul and Ethics in Ancient Thought (Oxford: Oxford University Press, 2005). For another argument that at least one medieval philosopher (Aquinas) was not a spiritualist, see Nussbaum and Putnam, “Changing Aristotle’s Mind,” in M. Nussbaum

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

41

For the sense [faculty] perceives forms only insofar as they are individual (shakhs.) and, generally, insofar as they are in matter and determinate particulars (fī hayūla wa-mushār ilayhā). But it does not receive them materially (hayūlānīya), according to the way they exist outside the soul, but rather in a more spiritual way (rūh.ānīya). . . . But as for the intellect, it alone can extract the form from the material determinate particular and conceptualize it detached (mufrada) according to its being. . . . [i.e., in] the quiddities of things.47

Importantly, the spiritual being of the sensible form is still particular, which corresponds to Aristotle’s definition of the sense faculty according to its distinctly particular objects; so the formal identity of the sense faculty and the sensible form must be in terms of their mutual particularity.48 Thus, there will be a similar mode of particular existence between the sensing material determinate particular (i.e., an individual creature’s particular sense faculty) and the received determinate and particular sensible forms.49 This last step of applying the Identity Principle to particulars would seem to support (2); if the MPI were a determinate particular, then it would only receive determinate particulars. In both sensation and understanding, there is an analogous relation between receiver and received, albeit the marked difference between sensation and understanding means that the relations

and A. Rorty (eds), Essays on Aristotle’s De Anima. My own view is that Aristotle was neither a literalist nor a spiritualist, and good alternative interpretations may be found in Caston, “The Spirit and the Letter;” Hendrik Lorenz, “The Assimilation of Sense to Sense-Object in Aristotle,” Oxford Studies in Ancient Philosophy 33 (2007), 179–220; and David Charles, “Aristotle’s Psychological Theory,” Proceedings of the Boston Area Colloquium in Ancient Philosophy 24 (2008), 1–29. 47 Short Comm. Met. (SCM), ch. 2, 61 [77]. Cf. SCDA, 64 and 30. Again, my interpretation here is that Averroes can speak, in a way, of “the [same] form” as present in the external object, the sense faculty, and the intellect because of identity in species form, even though these are numerically distinct instances of that species form. As I discuss in section 3c, I think this will be a natural reading considering that Averroes often identifies substantial form with species, yet, nevertheless, thinks substantial forms are numerically distinct in individuals of the same species. This also means that though sensation and understanding are both importantly connected to external objects in the world according to formal identity, there is nonetheless a real distinction between the proper objects of each. 48 The identity between knower and known (the abstracted intelligible form), on the other hand, must be a formal identity at least in terms of the full immateriality or separateness of both; it remains to be seen whether the MPI and the intelligibles are also identical in universality (as the sense power and sensible form are in particularity). 49 Presumably, a fortiori, a determinate particular receiving a form naturally in the way first described in the main passage (i.e., prime matter receiving some accidental or substantial form) would be even more obviously particular, since there (unlike in sensation) the identity between the recipient and the form in question is material or numerical identity.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

42

Stephen R. Ogden

occur in two different modes—a particular mode and an immaterial/separate mode, respectively. As I said before, Averroes’s intuition is that the proper subject of universal intelligibles cannot be particular itself, otherwise the intelligibles would become particularized (and, therefore, not universal). The thought here is similar to those in many contemporary arguments in philosophy of mind against reductive physicalism, especially ones that argue from the non-identity of physical and mental properties to two different subjects or substances as the bearers of those properties.50 Of course, we should remember that the contemporary mind-body problem is orthogonal to the Aristotelian nous-body relation.51 However, (2) is not just motivated by such intuitions. Now that we have investigated the principles behind (2), I want to put forward quite briefly a more simplified statement of the philosophical reasoning behind this ancient and medieval background. Something like the following is true, I think, for Aristotle and Averroes. Given a fairly Aristotelian account of what sensation is, it becomes clear why the objects of sensation must be particular. That is, given that our bodies and individual sense organs are individuated and particularized by occupying certain and distinct segments of space-time, they will be limited to receiving only forms within a certain range determined by those segments and the underlying matter. Aristotle thinks that each of the five senses is rendered determinate by its own particular objects. Not only will a certain kind of matter and material change be necessary to receive those certain objects (transparency in the eyes, air in the ears, etc.), but also the forms they receive are necessarily particular and individuated by their own original instantiation in the things of the world which are the common objects of sense.52 I can only see the particular patch of green before me at this time and only hear the particular D major chord played before me at this time. If that is a necessary truth about sensation,53 it 50 For an example of an argument from (the more popular) property dualism to (the less popular) substance dualism, see Dean Zimmerman, “From Property Dualism to Substance Dualism,” Proceedings of the Aristotelian Society Suppl. 84 (2010), 119–50. 51 We can thank Deborah Modrak for the title in her “The Nous-Body Problem in Aristotle,” Review of Metaphysics 44 (1991), 755–74. Cf. Burnyeat, Aristotle’s Divine Intellect (Milwaukee, WI: Marquette University Press, 2008), 10–15. Indeed, many scholars are interested in Aristotelian hylomorphism because it seems to promise a potential avenue around some of the more intractable issues in the contemporary mindbody debate. 52 Sensation of common sensibles, objects cognized by more than one external sense power, will also be necessarily individual and particularized in virtue of the component proper sensibles. 53 Aristotle seems to think it is a necessary truth—see An. Post. I.31, 87b28–30 (aisthanesthai ge anagkaion tode ti kai pou kai nun), and 37–9 (aisthanesthai men gar anagkē kath hekaston).

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

43

suggests a more general truth that looks very much like (2): If something is a (sensing) determinate particular (e.g., a sense faculty), then it must receive determinate and particular forms.54

3 . A N O B J E C T I O N F R O M AV E R R O E S O N A RIS T OT LE ’ S M ET A P H Y S I C S The Determinate Particular Argument is valid, and, if we accept the above justification for (2) through a combination of the Identity Principle and the analogy between sensation and understanding, the premises all seem fairly supported. But even if I am right up to this point, both (2) and the conclusion of Averroes’s argument seem plagued by a much deeper difficulty. How can Averroes possibly hold that the MPI is not a determinate particular? All of the previous discussion would seem to suggest that there must be some sort of strong identity between the mode of existence of the MPI and its universal intelligible objects. But if that is so, and if the MPI is not a determinate particular, then shouldn’t we think of the MPI as universal like its objects? That proposal, I submit, would clash with the lights of Aristotle and Averroes himself, who interestingly never calls the MPI universal. In fact, rather than being some multiply instantiable property or content, on Averroes’s view the MPI is one, singular substance, and it must be distinct from the other celestial intellects—all of which would seem to make it a kind of particular. In this section, I will develop this objection, namely that according to Averroes’s own appropriation of Aristotelian metaphysics, he must think the MPI is in fact some sort of determinate particular. In the following section 4, I will then explore Averroes’s views concerning the species membership of separate substances like the MPI, which will then solidify how he is only ruling out a qualified, lower kind of determinate particular in the conclusion of the Determinate Particular Argument.

54 Cohoe, “Why the Intellect Cannot Have a Bodily Organ: De Anima 3.4,” Phronesis 58 (2013), 347–77, argues that this basic understanding of sensation and intentionality in Aristotle is what drives Aristotle’s own arguments for the intellect’s lack of a bodily organ in DA III.4. Cohoe defends those arguments in a similar way as my comments above, though in more detail. Also, to generalize Averroes’s point here, any substantial or accidental forms received naturally by prime matter will also be rendered particular by the matter, since matter is the principle of individuation. Cf. n. 49.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Stephen R. Ogden

44

a. Substance, But Not al-Mushār Ilayhi / Aliquid Hoc? There are strong linguistic and conceptual reasons in Aristotle and Averroes to think that the MPI should be some kind of determinate particular simply in virtue of its being a substance. As Taylor tells us, the term “ ‘[d]eterminate particular’ reflects the Latin aliquid hoc, which corresponds to Aristotle’s τόδε τι at De Anima 402a24, and ‫ ﺍﳌﺸﺎﺭ ﺇﻟﻴﻪ‬al-mushār ilayhi, meaning ‘designated’ or ‘determinate’. . . . ”55 As that DA passage and a number of other texts (in the DA and Met.) reveal, there is a very close link if not equivalence in Aristotle’s usage of tode ti (‘this something’) and ousia (‘substance’). The following table may help the reader keep track of the relevant terminology in the different languages: Greek Arabic Latin English

τόδε τι (tode ti) ‫( ﺍﳌﺸﺎﺭ ﺇﻟﻴﻪ‬al-mushār ilayhi) aliquid hoc or hoc determinate particular

οὐσία (ousia) ‫( ﺍﻟﺠﻮﻫﺮ‬al-jawhar) substantia substance

The DA text says, “First, clearly it is necessary to determine in which of the genera [soul is] and what it is, I mean whether [it is] a tode ti and an ousia, or a quality, or a quantity, or also some other of the already established categories. . . . ”56 We see that tode ti and ousia are taken together, referring to the one category of substance, the other options referring to accidental categories.57 Averroes’s text also preserves this close relationship between the terms: “ . . . utrum sit hoc et substantia. . . . ”58 In a similar reference in DA, 410a14, tode ti (in Aristotle) and hoc (in Averroes) are also conceptually aligned with the category of substance. So much so that the Arabic translation of the text reads “the being of a thing and its substance” (āniya al-shayʾ wa-jawhar) and Averroes’s Middle Comm. on DA simply says “substance” (al-jawhar).59 These co-references are unsurprising since, in Met. VII.3, Aristotle says, “For it seems that separateness and being a tode ti belong especially to substance (tēi ousiai).”60 Averroes himself uses the Arabic terms 55 Taylor, Long Commentary, lvi, fn. 104. Cf. Crawford (ed.), Averrois Cordubensis Commentarium Magnum in Aristotelis De Anima Libros (Cambridge, MA: Mediaeval Academy of America, 1953), 560 and 563. As Crawford records, sometimes the correspondence is between tode ti and hoc or, alternatively, tode and aliquid hoc. 56 Aristotle, DA I.1, 402a23–25. 57 Cf. Aristotle, Cat. 4–5. 58 LCDA, I, c. 6, 9. See Taylor, Long Commentary, 7, fn. 25—he translates both technical terms hoc and aliquid hoc as “determinate particular,” remarking that the latter only occurs in Book III. 59 Aristotle, DA I.5, Arabic, 24. Averroes, MCDA, 36. Cf. Taylor, Long Commentary, 89, fn. 275. He notes that Smith’s English translation of the DA in The Complete Works of Aristotle also adds “substance.” 60 Met. VII.3, 1029a26–27. In the context, these are listed as reasons why matter should not be considered substance.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

45

(al-shakhs.) al-mushār ilayhi and jawhar virtually interchangeably in some of his comments on Met. VII.61 According to these passages, then, if Averroes holds that the MPI is a substance, he should also affirm that it is somehow a mushār ilayhi / aliquid hoc.

b. Not a Universal I also mentioned earlier that, if the MPI is a substance, there seems to be no way in which either Aristotle or Averroes could think of it as a universal. Metaphysics VII.13 argues in several ways against the idea that universals could be substances, again, in part because of an affinity between substance and tode ti (e.g., 1038b34–1039a3 and 1039a14–19), and Averroes certainly agrees with these anti-Platonic arguments in his own commentary. Although much of this evidence seems to point toward Aristotle’s concern with material particular substances as tode ti, we should note that in Met. XII.5, a section which leads directly into the famous chapters on separate unmoved movers, Aristotle argues again that the first or proximate principles (prōtai archai) of all things are individual (todi) and particular (to kath hekaston), and not universals (1071a18–21).62 Averroes agrees: “Since it is known that ‘principles’ is said in two ways [universal and particular], know that the true way between them is the particular (juz ʾī). . . . The true principles are those the substance of which is a thing existing in actuality outside the soul and a determinate particular (mushār ilayhi).”63 Here (and in the rest of the comment) Averroes uses the term mushār ilayhi (“determinate particular”) and juz ʾī (“particular”) in a context where it is evident that the very first principles are going to be immaterial separate substances. Similarly, in Met. VII.15, both Aristotle and Averroes seem to countenance the sun, a unique eternal celestial body, as among particular things (tōn kath hekasta and min al-ashyā’ al-juz ʾīya, respectively), just like Cleon and Socrates.64 Averroes, in his discussion, also employs the term mushār ilayhi.65 On the 61 E.g., LCM, vol. II, bk. Z, c. 48, 968/a. 62 Aristotle emphasizes in XII.6 (1071b12–22) that the eternal, unmoved substances (specifically) should not be identified with Platonic universals or forms (eidē), for these substances, although without matter (aneu hulēs) must (like those in the passage from XII.5) be actual (energeia). 63 LCM, vol. III, Λ, c. 27, 1544/c. 64 Aristotle, Met. VII.15, 1040b1–2. Averroes, LCM, vol. II, Z, c. 55, 993/k. 65 LCM, c. 55, 996/j. He uses the term in an argument which states that if any other thing (shayʾ ākhar mushār ilayhi) were to have the attributes of the sun, then that thing would be a sun. We should remember that, for Averroes, the celestial bodies are only equivocally material, thus linking the bodies more closely to the immaterial celestial intellects than perhaps Aristotle does. Furthermore, we will see later that Averroes holds

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

46

Stephen R. Ogden

principle alluded to earlier from XII.5 (viz., that the principles of particulars are particulars),66 it stands to reason that the celestial intellects (as movers to particular bodies like the sun) should also be particular. Thus, not only can the MPI as a separate substance not be thought of as a universal, there are texts that seem to push Averroes to say that it must be, indeed, a “determinate particular,” mushār ilayhi / aliquid hoc, i.e., some individual thing, and not further specifiable.

c. Substance as Form and Particular My initial linguistic analysis in 3a and 3b, which argues from the MPI’s being a substance to its somehow being a determinate particular and not a universal, is rendered more complicated by several serious difficulties in interpreting Aristotle’s metaphysics. The overarching problem for my considered objection here is that some commentators do not interpret tode ti as implying particularity, in turn envisioning substance as a kind of universal.67 If that interpretation were correct and adopted by Averroes, it would suggest an immediate solution to the apparent contradiction between his denial that the MPI is a determinate particular and my objection that the metaphysics necessitate that it is one. I will break down the specific issues around this central problem as follows: (i) the relationship between the Categories and the Metaphysics, (ii) substance as tode ti, (iii) further questions of particular forms in Met. VII, and (iv) particular-universal language in Met. XII. I cannot here fully discuss these issues, but I will briefly explain what Averroes thinks and how his interpretations are not implausible from the standpoint of contemporary Aristotelian scholarship.68 Despite the for a close affinity between the celestial bodies and the intellects on the question of their species membership. 66 Aristotle, Met. XII.5, 1071a20–21. Averroes, LCM, vol. III, Λ, c. 27, 1541/f [128]; cf. ibid., c. 27, 1543–5, esp. 1543/a [128–9]. 67 For a good overview of the problematic triad of Met. VII (substance is form, form is universal, no universal is a substance), see S. Marc Cohen, “Substances,” in G. Anagnostopoulos (ed.), A Companion to Aristotle (Malden, MA: Wiley-Blackwell, 2013), 207ff, and “Aristotle’s Metaphysics,” in E. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Fall 2015 Edition) , }10. 68 I argue these points at length and in much more detail elsewhere. For other accounts of Averroes’s metaphysical views discussed here, and especially their relation to contemporary Aristotle scholarship, see Matteo Di Giovanni, “Individuation by Matter in Averroes’ ‘Metaphysics,’ ” Documenti e studi sulla tradizione filosofica medievale 18 (2007), 187–210; Gabriele Galluzzo, “Averroes and Aquinas on Aristotle’s Criterion of Substantiality,” Arabic Sciences and Philosophy 19 (2009), 157–87, and The Medieval Reception of Book Zeta of Aristotle’s Metaphysics, vol. 1 (Leiden: Brill, 2013).

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

47

complication, the investigation here will actually strengthen my case in sections 3a and 3b. For although Averroes does envision a move towards an understanding of primary substance as species-form (as opposed to particular mundane composites), separate substances (like the MPI) are simply separate forms, and even they should not be thought of as universals. (i) Like many contemporary scholars, Averroes sees a decisive shift in the identification of primary substance (prōtē ousia) from composite particulars in the Categories to substantial form in the Metaphysics.69 Averroes explains this turn in terms of the form as the cause of (i.e., “the substance of ”) the composite particular.70 But this shift is felicitous for contemplating what must be true of the MPI’s substantiality, since Averroes considers separate substances (like the MPI) simply as separate or pure substantial forms (al-s.uwar al-mufāriqa).71 (ii) Averroes interprets the beginning of Met. VII in line with this shift, arguing that form, as primary substance, should be somehow both essence (to ti esti) AND tode ti,72 even if the form’s being tode ti is derivative upon the composite particular of which it is the cause.73 In fact, in the contemporary field regardless of where an interpreter lies on 69 Scholars as diverse with respect to issue (iii) as Michael Frede and Günther Patzig, on the one hand, and Michael Loux, on the other, agree on some basic understanding of this Cat.-Met. shift. See Frede and Patzig, Aristoteles, “Metaphysik Z.” Text, Übersetzung und Kommentar, vols. 1–2 (München: Verlag C.H. Beck, 1988), and Loux, Primary Ousia (Ithaca, NY: Cornell University Press, 1991). 70 See Averroes, SCM, ch. 1, 14–16 [30–1] and ch. 2, 43 [58]; LCM, vol. II, Z, c. 5, 761/i–762/l. For “substance of each thing” (ousia . . . hekastou), see Aristotle, Met. VII.3, 1028b35. Burnyeat, A Map of Metaphysics Zeta (Pittsburgh: Mathesis Publications, 2001), 14, reads it as a causal analysis, like Averroes. 71 E.g., LCM, vol. III, Λ, c. 5, 1421/d [73]. Though I cannot fully substantiate this point here, despite Averroes’s insistence that the MPI is a “fourth kind of being” (e.g., LCDA, III, c. 4, 386 [303]), it seems that the MPI is closest (if not identical) to a separate form in that he always gives it the same analysis as the other celestial intellects, which he considers separate forms. 72 Averroes, LCM, vol. II, Z, cc.1–2, 746–8. Cf. Laurence Bauloye, La question de l’essence. Averroès et Thomas d’Aquin, commentateurs d’Aristote, Metaphysics Z1 (Louvainla-Neuve: Editions Peeters, 1997), esp. 75. For contemporary arguments that VII.1 requires that primary substance be both to ti esti and tode ti, see G. E. L. Owen, “Particular and General,” Proceedings of the Aristotelian Society 79 (1978), 2; Alan Code, “Aristotle: Essence and Accident,” in R. Grandy and R. Warner (eds), Philosophical Grounds of Rationality (Oxford: Oxford University Press, 1986), 435; Jonathan Lear, Aristotle: The Desire to Understand (Cambridge: Cambridge University Press, 1988), 273; and Cohen, “Substances,” 209. 73 LCM, vol. II, Z, c. 3, p. 751/d–f. He claims that when Aristotle says, “the [type] of substances which belong to each one of the particulars (li-kull wāh.id al-juz ʾīyāt),” he means to indicate “the essences (al-māhīyāt).” Cf. Aristotle, Met. VII.1, 1028a27, and Loux, Primary Ousia, described below.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

48

Stephen R. Ogden

the spectrum of views concerning particular forms, virtually everyone agrees with my basic argument in 3a.—viz., that if something (like form or the MPI) is substance, it must also be tode ti.74 Also, because both species (nawʿ) and form (s.ūra) make known the essence of a thing,75 Averroes identifies form and species. One could criticize this identification, but such a move is not without contemporary defenders.76 (iii) There are three basic positions on particular or individual forms in Met. VII. First, there are what I call strict particularists, who hold that substantial forms are somehow intrinsically particular.77 Second, there are strict non-particularists who not only argue that matter is the principle of individuation for forms, but that there is really only numerically one substantial form per species which is shared by the composites of that kind.78 Finally, there is a range of intermediate positions. One typical moderate approach is to maintain that there are particular forms for each composite particular, but that they are

74 The most prominent deniers of particular forms nonetheless endorse substantial form as tode ti (in the sense of being individual and indivisible). See Code, “Essence and Accident,” esp. 435; Cohen, “Substances,” 202, 209; and Michael Woods, “Problems in Metaphysics Ζ, Chapter 13,” in J. Moravcsik (ed.), Aristotle: A Collection of Critical Essays (Notre Dame, IN: University of Notre Dame Press, 1967), 226–27. Even Loux, Primary Ousia, 144–6, is at least willing to say that form is tode ti “derivatively” from the particular composite. Of course, those who espouse particular forms rest some of their case on the apparent tode ti requirement. See Frede and Patzig, Metaphysik Z, vol. 1, 39, 52, 56; T. H. Irwin, Aristotle’s First Principles (Oxford: Oxford University Press, 1988), 255–9; and Charlotte Witt, Substance and Essence in Aristotle (Ithaca, NY: Cornell University Press, 1989), 161–2. For a very similar move to my argument in 3a, see Jennifer Whiting, “Metasubstance: Critical Notice of Frede-Patzig and Furth,” The Philosophical Review 100 (1991), 613: “ . . . a necessarily unique substance which is not a specimen of any kind (like, e.g., the Prime Mover) should be tode ti if Aristotle takes being tode ti as a requirement for being a substance.” 75 For form, see Averroes, SCM, ch. 1, 15–16 [31]. For species, see SCM, ch. 1, 14–15 [30]. 76 Esp. Woods, “Problems in Metaphysics Z,” who influentially argued for species-form as part of a solution to the problem of particular forms in Met. VII. Cf. Owen, “The Platonism of Aristotle,” Proceedings of the British Academy 50 (1965), 125–50, and Woods, “Substance and Essence in Aristotle,” Proceedings of the Aristotelian Society 75 (1975), 167–80. The Greek term eidos can mean either “form” or “species.” For critique, see Driscoll, “ΕΙΔΗ in Aristotle’s Earlier and Later Theories of Substance,” in D. O’Meara (ed.), Studies in Aristotle (Washington, DC: Catholic University of America Press, 1981), and Loux, Primary Ousia, esp. 197–8, fn. 1. 77 For example, Frede and Patzig, Metaphysik Z, who seem to deny any type of existence to universals in Aristotle. For different, but still intrinsic particularist views, see Irwin, Aristotle’s First Principles, and Witt, Substance and Essence. 78 See, for example, Loux, Primary Ousia, and Montgomery Furth, Substance, Form and Psyche (Cambridge: Cambridge University Press, 1988). Woods, “Substance and Essence in Aristotle,” 178, and “Particular Forms Revisited,” Phronesis 26 (1991), 82 and 85, also seems to fall into this camp.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

49

only extrinsically particular since matter is the principle of individuation.79 Averroes is a moderate particularist. Despite how his view of species-forms resembles some non-particularists, his commitment to particularism is manifest in several contexts.80 One example seems decisive: “All of this is necessary in order to clarify that the forms of individual substances are substances and that nothing is in the individual substance (al-shakhs. jawhar) except the particular matter and form (al-mādda wa-l-s.ūra al-juz ʾīya) of which it is composed.”81 However, he consistently adopts matter as the principle of individuation (so forms cannot be individuated intrinsically).82 When it comes then to separate forms, like the MPI, since matter will not work as the principle of individuation, it would be easy to assume that the separate forms are simply per se individual or particular. Yet, we will see in section 4 that even at this level Averroes develops a hierarchy of individuation among the separate forms, placing them analogically within the same species. This should suggest a further layer of (and textual support for) Averroes’s moderate particularism. (iv) As the last point and his categorization of separate substances as forms in (i) illustrate, Averroes is happy to extend the logic of substance and form from Met. VII into his commentary on Met. XII. Some of the strongest evidence already mentioned in favor of the MPI’s being somehow a determinate particular comes from XII, especially the individual causal principles in XII.5.83 Furthermore, it would also be impossible for Averroes to conceive of the celestial intellects in XII as universals.84 He sees in Aristotle generally, and especially in VII.13, an overarching critique of all self-subsisting Platonic universals. As he summarizes in 79 On my liberal understanding, this is a quite wide group of scholars, normally categorized as particularists or non-particularists. For a few examples, see Rogers Albritton, “Forms of Particular Substances in Aristotle’s Metaphysics,” Journal of Philosophy 54 (1957), 699–707; Driscoll, “ΕΙΔΗ;” Whiting, “Metasubstance;” Mary Louise Gill, Aristotle on Substance (Princeton: Princeton University Press, 1989); and Stephen Menn, “On Myles Burnyeat’s Map of Metaphysics Zeta,” Ancient Philosophy 31 (2011), 161–202. 80 E.g., Averroes, MCDA, 113; SCM, throughout ch. 2, e.g., 56 [72]. 81 Averroes, LCM, vol. II, Z, c. 44, 960/a. Also decisive is LCDA, II, c. 20, 158 [127]—speaking of plants and animals that can be divided and continue living as two distinct plants or animals, he says the single original soul can be “divided into souls (animas) which are the same in form as the soul existing in it.” 82 LCM, vol. II, Z, c. 28, 866–70. Cf. SCM, ch. 2, 49 [64]. 83 Various particularists also use Met. XII for some of their best arguments. For a good example, see Menn, “On Myles Burnyeat’s Map,” 172–3, on Aristotle, Met. XII.5, 1071a27–29. 84 I have not found any strict non-particularist arguing that the unmoved movers are separate universal forms.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

50

Stephen R. Ogden the SCM: “it has become clear from the preceding that [genus and differentia],85 insofar as they are universals, have no existence outside the mind (dhihn) and are in no way causes of the things defined [by them in the study of logic].”86 Hence, although Met. XII concerns separate forms, Averroes insists that they are not the causally useless Platonic universal forms.87 In contrast, it is particulars that have causal efficacy.

All of these points confirm my original argument in 3a and 3b—the MPI, in virtue of being a primary substance and separate form, must be somehow a determinate particular (tode ti/mushār ilayhi/aliquid hoc). Though he never explicitly uses such a term to describe the MPI, it seems that his metaphysics is committed to the implication.

4. U N I Q U E I N S P E C I E S ? If the argument of the previous section is correct that the MPI must be some sort of determinate particular, what might Averroes actually mean when he denies this in the Determinate Particular Argument? Richard Taylor explains in several publications that Averroes’s conclusion that the MPI is not a determinate particular means that the MPI is, like the other celestial intellects, unique in its own species.88 His clearest explication of this interpretation is in a 1999 paper, where he elaborates: The sense of non-particularity relevant here is that of the Material Intellect’s uniqueness in its species, which is responsible for its nature as what we might call an individual, not some other characteristic which might distinguish it such as matter. It is not ‘a this’ in the sense that it is a particular individual distinguished from other members of its species.89

So, the MPI is one “individual,” but it is not, as we might say, “one of a kind”—it is not a single member of a species with multiple members. In contrast, according to Taylor, “What is a ‘this,’ ‫[ ﺍﻟﺸﺨﺺ ﺍﳌﺸﺎﺭ ﺇﻟﻴﻪ‬al-shakhs. al-mushār ilayhi], aliquid hoc, is a particular which is a member of a species containing more than one member and which derives its particularity from 85 jins wa-fas.l in the preceding sentence. 86 SCM, ch. 2, 64. Cf. LCM, vol. II, Z, c. 45, 962/b. 87 SCM, ch. 2, 53 [68]. Cf. SCM, ch. 2, 56–7 [72–3], and LCM, vol. III, Λ, 1403 [63–4]—a brief critique of al-s.uwar al-āflāt. ūnīya, “ . . . I mean the separate forms posited by Plato.” 88 Taylor, “Averroes on Psychology and the Principles of Metaphysics,” Journal of the History of Philosophy 36 (1998), 516, fn. 27; “Averroes’ Epistemology,” 160; “Intelligibles in Act,” 130; Long Commentary, lviii, lxi, and 299, fn. 17. 89 Taylor, “Averroes’ Epistemology,” 160.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

51

the contraction of the form to matter in a composite.”90 This is certainly one way to go in resolving the complex issue of the individuation of the various immaterial celestial intellects (or unmoved movers) in Aristotle’s system.91 However, there is strong evidence in Averroes’s corpus that he does not envision each celestial intellect as unique in its species; rather, they are distinct members of one and the same shared species. In the Short Comm. or Epitome on Metaphysics, he says in several places that the separate celestial intellects are one in species, yet distinguished according to an order of priority and posteriority: “[the separate immaterial principles] differ . . . by excellence in nobility within one species and not by any difference in species (yughāyiru . . . bi-l-tafād.ul fī l-sharaf fī l-nawʿ al-wāh.id lā bi-ikhtilāf al-naw ʿiyya).”92 He explains further in another passage: We say that if you consider the case concerning them carefully, one will find that some of them are prior in nobility to others, for evidently the moving cause of the diurnal motion is nobler than all others. . . . And if you consider the case of the other [movers] carefully, one will find that they are disparate in this respect. Now, if things are disparate in excellence without being disparate in species (i.e., without constituting a multiplicity of species . . . ), the difference found in them [emerges] only from [the fact] that some of them are prior to others with respect to one thing shared by them.93

He repeats this view, albeit extended to celestial bodies as well, in the Long Comm. on De Caelo with a scold of Ibn Sīnā: “the celestial bodies (stelle) are the same in species, not in genus as Avicenna thinks.”94 He goes on to affirm that the celestial bodies agree in species according to priority and posteriority, and that the movers, i.e., the intellects, will agree in species (and differ from one another) in the same way, i.e., secundum prius et posterius.95 On much of this same evidence as well as material in the Long Comm. on Met., Matteo Di Giovanni has argued for the shared species view for both celestial intellects and bodies, his case for the bodies resting on the more explicit

90 Taylor, Long Commentary, 299, fn. 17. 91 This is, for example, Aquinas’s view, which he applies to angelic beings. Cf. Harry A. Wolfson, “The Plurality of Immovable Movers in Aristotle and Averroes,” Harvard Studies in Classical Philology 58 (1958), 233–53. 92 SCM, ch. 4, 149. 93 Ibid., 142 [153–4], translation modified and emphasis mine; cf. 144 [155], 150 [160], 153 [163]. 94 Long Comm. De Caelo (LCDC), II.8, c. 49, 369. 95 Ibid., 369–70: “ . . . convenientia in specie convenientia secundum prius et posterius, non secundum univocationem; et ideo natura cuiuslibet individui eorum est alia a natura alterius aliquo modo, et nature que movent ea sunt convenientes huiusmodi convenientia, et diverse huiusmodi diversitate.”

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

52

Stephen R. Ogden

evidence in favor of the intellects.96 He helpfully explains the difference in priority and posteriority (Arabic: bi-taqdīm wa ta ʾkhīr) as one of analogy, specifically in the case of the intellects what he calls a more Platonic notion of “synonymic analogy.”97 According to Averroes, Aristotle’s notion of pros hen predication is one type of analogy by priority and posteriority, which implies difference in both being and definition. But Averroes thinks the intellects partake in a second kind of analogy by priority and posteriority, differing in their degree of being, but not in definition—they all really share the immaterial intellectual nature of the first intellect (and are not merely effects of the first).98 Each intellect differs in degree of excellence (or simplicity) based on how (simply) it thinks of the first unmoved intellect, i.e., God, all while sharing the same definition and essence.99 This is Averroes’s own notion of purely “intelligible matter,” individuating separate intellects according to their potential for contemplating other higher intellects.100

96 Di Giovanni, “Averroes on the Species of Celestial Bodies,” in A. Speer (ed.), Miscellanea Mediaevalia, bd. 33 (Berlin: Walter de Gruyter, 2006), 440. 97 In SCM, ch. 4, 144 [155], Averroes also calls the relation between the movers one “by order and analogy” (bi-tartīb wa tanāsub). It is neither univocal (for that kind of diversity is only based on matter), nor equivocal (“as it has become clear that they are of the same kind”). Cf., LCDC in the quote footnoted above. For a full discussion of the notion of synonymic analogy here, see Di Giovanni, “Species,” 449–55. 98 See Di Giovanni, “Species,” 450–1. He points out that Averroes makes this distinction between two types of analogy by priority and posteriority in the Long and Short Comm. Met. See especially LCM, vol. I, Γ, c. 2, 302/b–303/c. Rüdiger Arnzen, Averroes on Aristotle’s “Metaphysics”: An Annotated Translation of the So-called “Epitome” (Berlin: Walter de Gruyter, 2010), 190, n. 27 and 199, n. 61, observes that the terminology sometimes used in the SCM—“by order and analogy” (bi-tartīb wa tanāsub)—is carefully distinguished in the LCM from pros hen predication (Arabic: asmāʿ mushakkaka). As Arnzen points out, the three terms (including bi-taqdīm wa ta ʾkhīr), are used somewhat interchangeably in the SCM (especially ch. 4, 144), but, if Di Giovanni is correct, then that may be because pros hen is a sub-species of analogy by priority and posteriority. 99 Di Giovanni, “Species,” 459ff. 100 LCDA, III, c. 5, 409–10 [326]: “For just as sensible being is divided into form and matter, so too intelligible being must be divided into . . . something similar to form and into something similar to matter. This [latter thing] is necessarily in every separate intelligence which understands something else. And if not, then there would be no multiplicity in separate forms. It was already explained in First Philosophy that there is no form free of potency without qualification except the First Form, which understands nothing outside itself.” On this point about the individuation of the intellects, I think there is probably an agreement between Di Giovanni and Taylor—in “Averroes’ Philosophical Conception of Separate Intellect,” esp. 401, and “Averroes on Psychology.” It is Taylor, however, who emphasizes this latter LCDA passage wherein Averroes presents a novel way of putting the individuating ordering in terms of potentiality, even in unmoved, immaterial intellects.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

53

Thus, the separate intellects, while individuated, still share in one species by analogy.101 One might object to this evidence that the proper movers of celestial bodies are actually celestial souls, rather than intellects.102 Without fully adjudicating this debate here, I will point out that at least in the SCM, where there are the most indications of a one species view, Averroes explicitly argues that the movers are intellects: If there are these forms whose existence is according to their not being in matter (laysat fī hayūla), then these are necessarily separate intellects (ʿaqūl mufāriqa) since there isn’t for forms qua forms a third [type of ] existence. Since it is clear that the existence of these movers (al-h.arakāt) is only insofar as they are intellects (ʿaqūl), we should consider in what way they move the celestial bodies.103

He explains in this same passage that the study of the soul (in the DA) proves that the nature of separate forms must be intellectual, a point which Taylor has also emphasized.104 Averroes also straightforwardly holds the principle in the LCM, the latest of his philosophical long commentaries: “it has been explained in the eighth book of the Physics that the mover of these celestial bodies is without matter and a separate form, and in the de Anima that the separate forms are intellect. It follows that this mover is an intellect. . . . ”105 In the places where Taylor puts forward his thesis of each celestial intellect constituting a unique species, he normally does not cite textual evidence other than that which simply supports some kind of diversity among the intellects. In fact, he references some of the same passages we have considered earlier (about an ordering of excellence based on different degrees of knowing in each intellect).106 On just this point, then, Taylor most likely 101 Arnzen, Averroes on Aristotle’s “Metaphysics,” 317–18, n. 606, commenting on the SCM, shares a very similar view to Di Giovanni’s and confirms the shared species interpretation: “Unlike Ibn Sīnā, Ibn Rushd conceives the celestial movers as pertaining to one species.” 102 E.g., David Twetten, “Averroes on the Prime Mover Proved in the Physics,” Viator 26 (1995), esp., 132, 134, would distinguish the movers of the spheres as celestial souls, distinct from celestial intellects. I thank Jon McGinnis for raising this objection. 103 SCM, ch. 4, 139. 104 Cf. LCDA, III, c. 5, 410 [327]. Taylor, in “Averroes on Psychology” and “Averroes’ Philosophical Conception of Separate Intellect,” esp. 392ff., puts forward evidence that the celestial intellects are the movers in a number of works by Averroes, and Di Giovanni develops his own argument for this point—Di Giovanni, “Species,” 458–9. Cf. Davidson, Alfarabi, Avicenna, and Averroes, 230 and 291. 105 LCM, vol. III, Λ, c. 36, 1594/d [149]. 106 E.g., Taylor, “Averroes on Psychology” and “Averroes’ Philosophical Conception of Separate Intellect,” esp. 401–2. I want to briefly note two textual exceptions—the Tahāfut (Incoherence of the Incoherence), d. 3, 260 [154–5] and the LCDA, III, c. 5, 403

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

54

Stephen R. Ogden

means to say that such diversity arises without having univocally the same species, which would have to involve sensible matter.107 Our discussion has now led us to a greater understanding of what Averroes could mean in claiming that the MPI is not a determinate particular, even though his metaphysics seems to imply the opposite. I argued in the last section that he must think the MPI is some sort of determinate particular. Furthermore, we now see that he cannot mean that the MPI is not a distinct individual in the one species of separate intellects, simpliciter, since he says explicitly in a number of texts that the separate intellects are in the same species, albeit analogically. Rather, in agreement here with one of Taylor’s most detailed characterizations, to say that the MPI is not a determinate particular means more specifically that it is not “a member of a species containing more than one member and which derives its particularity from the contraction of the form to matter in a composite.”108 The MPI is not one (determinate particular) of a kind only if we mean a univocal species of material composites.

5. W H A T D O E S T H E A R G U M E N T P R O V E ? What, then, shall we finally make of Averroes’s Determinate Particular Argument? I have surveyed evidence from his thinking on the Met. that Averroes could use the term for determinate particulars (mushār ilayhi / aliquid hoc) quite generically, even extending it to immaterial separate substances such as the MPI. But, if we also take into consideration our investigation into species membership, the most plausible explanation of the Determinate Particular Argument is to assume that Averroes means that the MPI cannot be a material determinate particular or a member of a univocal species. That seems to be his main emphasis in the context of the DA.109 In [318–19]. Taylor quotes these (though not as direct support for the unique species view) in “Averroes on Psychology,” 522, fn. 48, and 512, fn. 13. We must be circumspect about this evidence because the Tahāfut is a dialectical work where Averroes says in the very same passage that he is not giving his decided philosophical views. Also, the quote in the LCDA comes in the midst of a considered objection to Averroes’s own interpretation of the MPI. 107 However, to my knowledge Taylor never indicates the possibility of their being analogically the same in species, as discussed earlier. 108 Taylor, Long Commentary, 299, fn. 17, emphasis mine. 109 As Bazán, “The Human Soul: Form and Substance? Thomas Aquinas’ Critique of Eclectic Aristotelianism,” Archives d’histoire doctrinale et littéraire du Moyen Age 64 (1997), 105–6, says speaking only of the LCDA, even though the MPI is a pre-eminent substance, Averroes does not call the MPI a “hoc aliquid, because for him this notion applies only to material substances.”

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

55

fact, perhaps Averroes carefully avoids using the term in the DA commentary, since (as I noted earlier) the references in the DA all seem to suggest the more mundane, Categories analysis of substance and tode ti as applied only to composite particulars with sensible matter. It is only when we investigate the Metaphysics that we find a more philosophically accurate account of substantial form as primary substance. Although the MPI fits that Met. framework for being a determinate particular of a higher kind, Averroes adheres to the texts in DA and denies that the MPI is a determinate particular—at least not one of this lower, composite type.110 But if that is the case, premise (2) of the Determinate Particular Argument is not true generally, since the MPI most likely is a determinate particular of a higher type which, nevertheless, does not receive determinate particular forms. The argument and its conclusion should be re-written with the qualified “material determinate particular.” Once we make this qualification, I think we can see why this is not a sufficient argument for Averroes’s complete position on the intellect. If the Determinate Particular Argument were supposed to serve as an independent argument for his entire view, it would be a weak one. It would not be effective against other important contending interpretations of the MPI, namely those of Avicenna and Aquinas, which hold that the MPI, albeit multiple, is also an immaterial determinate particular.111 Avicenna and Aquinas have different views of the MPI, both of which have their fair share of difficulties, and obviously Averroes was ignorant of Aquinas’s position. Still, Averroes does have Avicenna in view as a major rival interpretation, and the MPI is equally immaterial on both of their accounts.112 Of course, Avicenna thinks there is more than one MPI, but Averroes cannot simply beg that question against him.

110 Though I cannot develop the case here, certain passages in Averroes and Aristotle point to a flexibility of meaning in tode ti/al-mushār ilayhi / aliquid hoc. Just as Averroes seems to have both a single universal species-form and individual species-forms in the particulars, so too will these terms have at least two different connotations respectively: tode ti insofar as it is not further specifiable; and tode ti in virtue of being numerically distinct in the composite particular. Cf. Gill, Aristotle on Substance, 31–8, esp. 38. 111 However, the MPI on their views is not robustly separate in Averroes’s sense—it is only separable (capable of independent existence). Although it is incorruptible, it is, in fact, generable on their views, coming to exist only in conjunction with a body. So the MPI for Avicenna and Aquinas is not an eternal, celestial intellect. 112 Of course, Averroes dismisses Avicenna as having led many to confusion regarding the interpretation of the DA (LCDA, III, c. 30, 470 [374–5]). Even so, the Determinate Particular Argument is an independent philosophical argument, and one might have expected something that would be potent against Avicenna in that regard. He criticizes Avicenna’s views on the MPI and active intellect for being incoherent in SCDA, 83–4.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

56

Stephen R. Ogden

If Taylor’s unique species view of celestial intellects were the correct conception, then the Determinate Particular Argument might have proven the entire Averroistic doctrine of the MPI on its own.113 Not only would the MPI not be a material determinate particular, there would also only be one (since, on that view, among separate substances there is only one member per species). However, having undermined that interpretation of species membership for separate intellects in Averroes, we can conclude that the argument only demonstrates that the MPI is not a material determinate particular and, thus far, it will not have motivated his truly distinct position. Therefore, I think we should surmise that Averroes never meant for this argument to be sufficient for his complete position on the MPI. To be sure, there are indications in the text that might lead one to see it as an independent and sufficient argument. In the main text of the argument he does show a concern about not only the matter or material particularity of forms, but also their diversity. Hence one can easily get the impression that this argument is simultaneously aiming to secure the unity of the MPI. Averroes also gives the follow-up argument in premises (5)–(8), which conclude that the MPI is a separate celestial substance. But again, while this secures many of Averroes’s desiderata, this does not necessarily prove uniqueness, and certainly not without more work. Moreover, Avicenna and Aquinas would definitely contend with those further premises, particularly (5) and Averroes’s robust interpretation of “separate” featuring in (8), even if the argument were revised using “material determinate particular.’ Maybe, then, it is best to think of the Determinate Particular Argument as an opening salvo at the beginning of comment 5, demonstrating the immaterial nature of the MPI. Then, after answering several objections, it is significant that Averroes virtually ends the comment with another possible philosophical argument for his view, what I call the Unity Argument. This argument is beyond the scope of the current paper, and I will analyze it elsewhere. However, while I can only briefly mention it here, I shall quote the core of it: “And if we assume it [the understood intelligible] to be many, 113 This would explain the repeated conjunction of Taylor’s unique species view of the MPI and his apparent characterization of the Determinate Particular Argument as independently sufficient—see again Taylor, “Averroes’ Epistemology,” 159–61; “Intelligibles in Act,” 130; Long Commentary, lviii. In the last reference, he does say the Determinate Particular Argument “provides part of the argument” for the full view, but the other “part” identified is primarily Averroes’s defense against objections. In fact, Taylor states the conclusion of the Determinate Particular Argument as decisive: “ . . . [the MPI] is unique in its species and not a determinate particular belonging individually to any determinate particular human being.” That is precisely what I have argued the correctly understood argument fails to establish.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

57

then it would happen that the thing understood in me and in you would be one in species and two in individual [number]. And in this way the thing understood will have a thing understood, and it thus proceeds into infinity.”114 This is very compact, but the idea seems to be a kind of “Third Man Argument,” as Taylor helpfully classifies it.115 By one-over-many considerations, if two things are alike in species but differ numerically, including two individuated thoughts of the same universal F, there should be some universal intelligible over them. But, in the case of two understandings of an already supposedly universal F, this would require a further universal intelligible F*, and so on ad infinitum. Averroes’s argument proceeds from the datum of unified knowledge to establish that this regress is not actual and that there must be some one, authoritative stopping point for the universal intelligibles in act. The interesting point about this Unity Argument for our purposes is that its conclusion is that the MPI must be one. So if the Determinate Particular Argument can establish for Averroes at least that the MPI is not a material determinate particular, the Unity Argument might get him the rest of the way to his signature doctrine of the MPI—viz., that the MPI is unique and one for all humankind.

6 . C O N C LU S I O N While the Determinate Particular Argument can be shown to rely on plausible interpretations of Aristotelian principles, I have argued that its key term, ‘determinate particular,’ must be qualified in light of Averroes’s metaphysical commitments. It seems that Averroes must think of the MPI (a separate form and substance) as somehow a determinate particular (tode ti/mushār ilayhi / aliquid hoc), albeit on a different order from material determinate particulars. This revision would render the argument otiose when compared to rival interpretations of the MPI—viz., those of Avicenna and Aquinas, which equally deny that the MPI is a material determinate particular, even if there are multiple MPIs. Therefore, the Determinate Particular Argument should no longer be considered as a sufficient, independent argument for Averroes’s doctrine of the single separate MPI. In my 114 LCDA, III, c. 5, 411. See the whole argument in c. 5, 411–12, ll. 713–28 [328–9]. Averroes considers similar arguments in the SCDA, 82–3, and the SCM, ch. 2, 57–9 [73–5]. The same argument appears in Averroes’s predecessor Ibn Bājja—Risālat Ittis.āl al-ʿAql bi-l-Insān (Treatise on the Conjunction of the Intellect with Man), 163–4. 115 Taylor, Long Commentary, lxi. Taylor’s comments on this argument are brief (lx–lxi, and 328–9, fn. 114). He considers it again as a distinct point in “Aquinas and ‘the Arabs,’ ” 154.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

58

Stephen R. Ogden

view, we must look elsewhere, to the Unity Argument at the end of comment 5, to see the remainder of his reasoning in favor of the view. Nevertheless, I hope this close analysis of the Determinate Particular Argument has shed much greater light on Averroes’s philosophical case and, to that extent, fended off at least some incredulous staring. Johns Hopkins University

BIBLIOGRAPHY

Primary Sources Aristotle. Analytica Posteriora [An. Post.], in Analytica Priora et Posteriora, ed. W. D. Ross (Oxford: Oxford University Press, 1964). Aristotle. Categories [Cat.], in Categoriae et Liber de Interpretatione, ed. L. MinioPaluello (Oxford: Oxford University Press, 1936). Aristotle. The Complete Works of Aristotle, The Revised Oxford Translation, vols. 1–2, ed. J. Barnes (Princeton: Princeton University Press, 1984). Aristotle. De Anima [DA]. Greek. Aristotelis De Anima: Recognovit brevique adnotatione instruxit, ed. W. D. Ross (Oxford: Oxford University Press, 1956). Greek. Aristotle, De Anima, ed. W. D. Ross (Oxford: Oxford University Press, 1961). Arabic. Arist. ūt. ālīs fī-l-Nafs, ed. A. Badawi (Cairo: Imprimerie Misr S.A.E., 1954; repr. Kuwait: Dār al-Qalam, 1980). English. In The Complete Works of Aristotle, tr. J. A. Smith. Aristotle. De Caelo, in Aristote, Du ciel, ed. P. Moraux (Paris: Les Belles Lettres, 1965). Aristotle. Metaphysica [Met.], in Aristotle’s Metaphysics, vols. 1–2, ed. W. D. Ross (Oxford: Oxford University Press, 1958). Aristotle. Physica [Phys.], in Aristotle’s Physics, ed. W. D. Ross (Oxford: Oxford University Press, 1936). Averroes (Ibn Rushd). Long Commentary on De Anima [LCDA]. Latin. Averrois Cordubensis Commentarium Magnum in Aristotelis De Anima Libros, ed. F. S. Crawford (Cambridge, MA: Mediaeval Academy of America, 1953). English. Long Commentary on the De Anima of Aristotle, tr. R. Taylor, sub-ed. T.-A. Druart (New Haven: Yale University Press, 2009).

For helpful comments on different parts and versions of this paper, I’d like to thank the colloquium audiences at University of Toronto and Cornell University, Verity Harte, John Hare, Deborah Black, Gyula Klima, Frank Griffel, and especially an anonymous reviewer who provided very useful feedback.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

59

Averroes (Ibn Rushd). Long Commentary on De Caelo [LCDC], in Averrois Cordubensis Commentum Magnum Super Libro De Celo Et Mundo Aristotelis, vols. 1–2, ed. F. J. Carmody and R. Arnzen (Louvain: Peeters, 2003). Averroes (Ibn Rushd). Long Commentary on Metaphysics [LCM]. Arabic. Tafsīr mā ba ʿd at. -T.abī ʿat, vols. 1–4, ed. M. Bouyges (Beirut: Imprimerie Catholique, 1938–52). English. Averroes’s Metaphysics: A Translation with Introduction of Ibn Rushd’s Commentary on Aristotle’s Metaphysics, Book Lām, tr. C. Genequand (Leiden: Brill, 1984). Averroes (Ibn Rushd). Middle Commentary on De Anima [MCDA], in Middle Commentary on Aristotle’s De Anima: A Critical Edition of the Arabic Text with English Translation, Notes and Introduction, ed. and tr. A. Ivry (Provo, UT: Brigham Young University Press, 2002). Averroes (Ibn Rushd). Short Commentary on De Anima [SCDA], in Talkhīs. Kitāb al-Nafs, ed. A. F. El-Ahwani (Cairo: Imprimerie Misr, 1950). Averroes (Ibn Rushd). Short Commentary on Metaphysics [SCM]. Arabic. In Averroes: Compendio de Metafísica, ed. C. Q. Rodríguez (Madrid: Imprenta de Estanislao Maestre, 1919). English. Averroes on Aristotle’s “Metaphysics”: An Annotated Translation of the So-Called “Epitome,” ed. and tr. R. Arnzen (Berlin: de Gruyter, 2010). Averroes (Ibn Rushd). Tahāfut al-Tahāfut (The Incoherence of the Incoherence). Arabic. Averroès Tahafot at-Tahafot (L’Incohérence de l’incohérence), 2nd edn, ed. M. Bouyges (Beirut: Dar el-Mashreq, 1987). English. The Incoherence of the Incoherence, vols. I–II, ed. and tr. S. Van Den Bergh (Cambridge: Gibb Memorial Trust, 1954). Avicenna (Ibn Sīnā). Najāt (The Salvation). Arabic. Kitāb al-Najāt, ed. M. Fakhry (Beirut: Dār al-Āfāq al-Jadīda, 1985). English. Avicenna’s Psychology, tr. F. Rahman (London: Oxford University Press, 1952). Ibn Bājja. Risālat Ittis.āl al-ʿAql bi-l-Insān (Treatise on the Conjunction of the Intellect with Man), in Rasāʾil Ibn Bājja al-Ilāhīya, ed. M. Fakhry (Beirut: Dār al-Nahār li-l-Nashr, 1968). al-Kindī. On the Intellect. Arabic. Rasā ʾil al-Kindī al-Falsafīya, ed. M. Abū Rīda (Cairo: Dār al-Maʿārif, 1950–53). English. In The Philosophical Works of al-Kindī, tr. P. Adamson and P. Pormann (Oxford: Oxford University Press, 2012). Liber de Causis/Kalām fī Mah.d. al-Khair [LDC]. Arabic and Latin. Die pseudo-aristotelische Schrift Ueber das reine Gute bekannt unter dem Namen Liber de causis, ed. O. Bardenhewer (Freiburg: Herdersche Verlagshandlung, 1882). Proclus. The Elements of Theology, Greek and English in The Elements of Theology, 2nd edn, ed. and tr. E. R. Dodds (Oxford: Oxford University Press, 1963). Themistius. De Anima Paraphrasis. Greek. In Libros Aristotelis De Anima Paraphrasis, ed. R. Heinze (Berlin: G. Reimeri, 1899). CAG V.3.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

60

Stephen R. Ogden

Arabic. An Arabic Translation of Themistius’ Commentary on Aristotle’s De Anima, ed. M. C. Lyons (Columbia, SC: B. Cassirer, 1973). English. Themistius. On Aristotle’s On the Soul, tr. R. Todd (Ithaca, NY: Cornell University Press, 1996). Thomas Aquinas. Scriptum Super Sententiis [Sent.], in Scriptum super libros Sententiarum magistri Petri Lombardi episcopi Parisiensis, vols. 1–2, ed. P. Mandonnet (Paris: P. Lethielleux, 1929). Thomas Aquinas. Super Librum de Causis. Latin. Super librum De Causis exposition, ed. H. D. Saffrey (Fribourg-Louvain: Société Philosophique-Nauwelaerts, 1954). English. Commentary on the Book of Causes, tr. V. Guagliardo, C. Hess, and R. Taylor (Washington DC: Catholic University of America Press, 1996).

Secondary and Contemporary Sources Albritton, Rogers. “Forms of Particular Substances in Aristotle’s Metaphysics,” Journal of Philosophy 54 (1957), 699–707. Arnzen, Rüdiger (ed. and tr.). Averroes on Aristotle’s “Metaphysics”: An Annotated Translation of the So-called “Epitome” (Berlin: de Gruyter, 2010). Bauloye, Laurence. La question de l’essence. Averroès et Thomas d’Aquin, commentateurs d’Aristote, Metaphysics Z1 (Louvain-la-Neuve: Editions Peeters, 1997). Bazán, Bernardo Carlos. “Intellectum Speculativum: Averroes, Thomas Aquinas, and Siger of Brabant on the Intelligible Object,” Journal of the History of Philosophy 19 (1981), 425–46. Bazán, Bernardo Carlos. “Le commentaire de S. Thomas d’Aquin sur le Traité de l ”âme,” Revue des sciences philosophiques et théologiques 69 (1985), 521–47. Bazán, Bernardo Carlos. “The Human Soul: Form and Substance? Thomas Aquinas’ Critique of Eclectic Aristotelianism,” Archives d’histoire doctrinale et littéraire du Moyen Age 64 (1997), 95–126. Black, Deborah. “Consciousness and Self-Knowledge in Aquinas’s Critique of Averroes’s Psychology,” Journal of the History of Philosophy 31 (1993), 349–85. Black, Deborah. “Conjunction and the Identity of Knower and Known in Averroes,” American Catholic Philosophical Quarterly 73 (1999), 159–84. Black, Deborah. “Mental Existence in Thomas Aquinas and Avicenna,” Mediaeval Studies 61 (1999), 45–79. Black, Deborah. “Models of Mind: Metaphysical Presuppositions of the Averroist and Thomistic Accounts of Intellection,” Documenti e studi sulla tradizione filosofica medievale 15 (2004), 319–52. Black, Deborah. “Psychology: Soul and Intellect,” in P. Adamson and R. Taylor (eds), The Cambridge Companion to Arabic Philosophy (Cambridge: Cambridge University Press, 2005), 308–26. Black, Deborah. “Averroes on the Spirituality and Intentionality of Sensation,” in P. Adamson (ed.), In the Age of Averroes: Arabic Philosophy in the Sixth/Twelfth Century (London: Warburg Institute, 2011), 159–74.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

61

Burnyeat, Myles. “Is an Aristotelian Philosophy of Mind Still Credible? A Draft,” in M. Nussbaum and A. Rorty (eds), Essays on Aristotle’s De Anima (Oxford: Oxford University Press, 1992), 15–26. Burnyeat, Myles. A Map of Metaphysics Zeta (Pittsburgh: Mathesis Publications, 2001). Burnyeat, Myles. “De Anima II 5,” Phronesis 47 (2002), 28–90. Burnyeat, Myles. Aristotle’s Divine Intellect (Milwaukee, WI: Marquette University Press, 2008). Caston (2005). “The Spirit and the Letter: Aristotle on Perception,” in R. Salles (ed.), Metaphysics, Soul and Ethics in Ancient Thought (Oxford: Oxford University Press, 2005), 245–320. Charles, David. “Aristotle’s Psychological Theory,” Proceedings of the Boston Area Colloquium in Ancient Philosophy 24 (2008), 1–29. Code, Alan. “Aristotle: Essence and Accident,” in R. Grandy and R. Warner (eds), Philosophical Grounds of Rationality (Oxford: Oxford University Press, 1986), 411–39. Cohen, S. Marc. “Substances,” in G. Anagnostopoulos (ed.), A Companion to Aristotle (Malden, MA: Wiley-Blackwell, 2013), 197–212. Cohen, S. Marc. “Aristotle’s Metaphysics,” in E. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Fall 2015 Edition) . Cohoe, Caleb. “Why the Intellect Cannot Have a Bodily Organ: De Anima 3.4,” Phronesis 58 (2013), 347–77. Crawford, F. Stuart (ed.). Averrois Cordubensis Commentarium Magnum in Aristotelis De Anima Libros (Cambridge, MA: Mediaeval Academy of America, 1953). Davidson, Herbert. Alfarabi, Avicenna, and Averroes on Intellect (Oxford: Oxford University Press, 1992). Defilippo, Joseph. “Aristotle’s Identification of the Prime Mover as God,” The Classical Quarterly 44 (1994), 393–409. Di Giovanni, Matteo. “Averroes on the Species of Celestial Bodies,” in A. Speer (ed.), Miscellanea Mediaevalia, bd. 33 (Berlin: Walter de Gruyter, 2006), 438–64. Di Giovanni, Matteo. “Individuation by Matter in Averroes’ ‘Metaphysics,’ ” Documenti e studi sulla tradizione filosofica medievale 18 (2007), 187–210. Driscoll, John. “ΕΙΔΗ in Aristotle’s Earlier and Later Theories of Substance,” in D. O’Meara (ed.), Studies in Aristotle (Washington, DC: Catholic University of America Press, 1981), 129–59. Frede, Michael and Günther Patzig. Aristoteles, “Metaphysik Z.” Text, Übersetzung und Kommentar, vols. 1–2 (München: Verlag C.H. Beck, 1988). Furth, Montgomery. Substance, Form and Psyche: An Aristotelean Metaphysics (Cambridge: Cambridge University Press, 1988). Galluzzo, Gabriele. “Averroes and Aquinas on Aristotle’s Criterion of Substantiality,” Arabic Sciences and Philosophy 19 (2009), 157–87. Galluzzo, Gabriele. The Medieval Reception of Book Zeta of Aristotle’s Metaphysics, vol. 1 (Leiden: Brill, 2013). Gauthier, R. A. “Notes sur les début (1225–1240) du premier ‘Averroïsme,’ ” Revue des sciences philosophiques et théologiques 66 (1982), 321–74.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

62

Stephen R. Ogden

Gill, Mary Louise. Aristotle on Substance (Princeton: Princeton University Press, 1989). Gómez Nogales, Salvador. “Saint Thomas, Averroès et l’Averroïsme” in G. Verbeke and D. Verhelst (eds), Aquinas and Problems of His Time (Louvain: Leuven University Press, 1976), 161–77. Irwin, T. H. Aristotle’s First Principles (Oxford: Oxford University Press, 1988). Lear, Jonathan. Aristotle: The Desire to Understand (Cambridge: Cambridge University Press, 1988). Lewis, David. Counterfactuals (Cambridge, MA: Harvard University Press, 1973). Lorenz, Hendrik. “The Assimilation of Sense to Sense-Object in Aristotle,” Oxford Studies in Ancient Philosophy 23 (2007), 179–220. Loux, Michael. Primary Ousia: An Essay on Aristotle’s Metaphysics Z and H (Ithaca, NY: Cornell University Press, 1991). Menn, Stephen. “On Myles Burnyeat’s Map of Metaphysics Zeta,” Ancient Philosophy 31 (2011), 161–202. Modrak, Deborah. “The Nous-Body Problem in Aristotle,” Review of Metaphysics 44 (1991), 755–74. Nussbaum, Martha and Hilary Putnam. “Changing Aristotle’s Mind,” in M. Nussbaum and A. Rorty (eds), Essays on Aristotle’s De Anima (Oxford: Oxford University Press, 1992), 27–56. Owen, G. E. L. “The Platonism of Aristotle,” Proceedings of the British Academy 50 (1965), 125–50. Owen, G. E. L. “Particular and General,” Proceedings of the Aristotelian Society 79 (1978), 1–21. Pasnau, Robert. “Aquinas and the Content Fallacy,” The Modern Schoolman 75 (1998), 293–314. Sorabji, Richard. “Body and Soul in Aristotle,” Philosophy 49 (1974), 63–89. Taylor, Richard. “Averroes on Psychology and the Principles of Metaphysics,” Journal of the History of Philosophy 36 (1998), 507–23. Taylor, Richard. “Averroes’ Epistemology and Its Critique by Aquinas,” in R. E. Houser (ed.), Medieval Masters: Essays in Memory of Msgr. E. A. Synan (Houston: Center for Thomistic Studies, 1999), 147–77. Taylor, Richard. “Intelligibles in Act in Averroes,” in J. B. Brenet (ed.), Averroès et les averroïsmes juif et latin. Actes du colloque tenu à Paris, 16–18 juin 2005 (Turnhout: Brepols, 2007), 111–40. Taylor, Richard. “Intellect as Intrinsic Formal Cause in the Soul According to Aquinas and Averroes,” in M. Elkaisy-Friemuth and J. Dillon (eds), The Afterlife of the Platonic Soul: Reflections of Platonic Psychology in the Monotheistic Religions (Leiden: Brill, 2009), 187–220. Taylor, Richard. “Averroes’ Philosophical Conception of Separate Intellect and God,” in A. Hasnawi (ed.), La lumière de l’intellect. La pensée scientifique et philosophique d’Averroès dans son temps (Leuven: Peeters, 2011), 391–404. Taylor, Richard. “Aquinas and ‘the Arabs’: Aquinas’s First Critical Encounter with the Doctrine of Averroes on the Intellect, In 2 Sent. d. 17, q. 2, a. 1,” in L. X. López-Farjeat and J. Tellkamp (eds), Philosophical Psychology in Arabic

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

On a Possible Argument for Averroes’s Single Separate Intellect

63

Thought and the Latin Aristotelianism of the 13th Century (Paris: Vrin, 2013), 142–83. Twetten, David. “Averroes on the Prime Mover Proved in the Physics,” Viator 26 (1995), 107–34. Whiting, Jennifer. “Metasubstance: Critical Notice of Frede-Patzig and Furth,” The Philosophical Review 100 (1991), 607–39. Wippel, John. “Thomas Aquinas and the Axiom ‘What Is Received Is Received According to the Mode of the Receiver,’ ” in R. Link-Sallinger (ed.), Studies in Medieval Philosophy and Culture (Washington, DC: Catholic University of America Press, 1988), 279–89. Witt, Charlotte. Substance and Essence in Aristotle: An Interpretation of Metaphysics VII–IX (Ithaca, NY: Cornell University Press, 1989). Wolfson, Harry A. “The Plurality of Immovable Movers in Aristotle and Averroes,” Harvard Studies in Classical Philology 58 (1958), 233–53. Woods, Michael. “Problems in Metaphysics Ζ, Chapter 13,” in J. Moravcsik (ed.), Aristotle: A Collection of Critical Essays (Notre Dame, IN: University of Notre Dame Press, 1967), 215–38. Woods, Michael. “Substance and Essence in Aristotle,” Proceedings of the Aristotelian Society 75 (1975), 167–80. Woods, Michael. “Particular Forms Revisited,” Phronesis 26 (1991), 75–87. Zimmerman, Dean. “From Property Dualism to Substance Dualism,” Proceedings of the Aristotelian Society Suppl. 84 (2010), 119–50.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Scotus on Universals John Hawthorne

Suppose one believes that the humanity of Socrates exists. One will then usually go on to affirm one of the following two theses. First, one might claim that the humanity of Socrates is distinct from the humanity of Plato. Following contemporary nomenclature, we can call the natural generalization of this thesis the trope theory for humanity: for every human there exists the humanity of that human which is not the humanity of any other human. Alternatively, one might claim that the humanity of Socrates is identical to the humanity of Plato. This is the universals theory of humanity: there is one thing—humanity—that is possessed both by Plato and Socrates and indeed every other human. If one opts for a trope theory, one has the burden of explaining what the humanities have in common; if one opts for a universals theory, one has the burden of explaining away the conceptual oddity of a single thing being fully present in many places. So each view has its puzzles. But what seems pretty clear is that if the humanity of Socrates exists, then either tropes or universals exist (and perhaps both). Intriguingly, Duns Scotus appears upon first reading to offer a different view than either of these concerning the metaphysical status of the humanity of Socrates. The trope and universals views both claim that the humanity of Socrates stands in relations of numerical identity and difference to other natures. On the trope view, the humanity of Socrates is numerically distinct both from the humanity of Plato and from the angelhood of Gabriel. On the universals view, the humanity of Socrates is numerically identical to the humanity of Plato but numerically distinct from the angelhood of Gabriel. According to Scotus, however, the humanity of Socrates has “less than a numerical unity”—which would appear to imply that both the universals and trope theories are off the mark. But what can it mean to say that a nature has “less than a numerical unity?” And how does such a claim relate to claims of identity and difference between natures? Is there even an intelligible position in the vicinity here? This paper attempts to map out Scotus’s view, display some arguments that Scotus offers for the view, and address Ockham’s critique of its coherence.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Scotus on Universals

65

1 . BA C K G R O U N D According to Scotus, a material substance can be metaphysically factorized into two components: a nature that determines what the individual substance is, and an individual difference (or “haecceity”) that determines which substance it is. The singularity-generating individual difference “contracts” with the nature to yield an individual substance. The individual difference, lacking any quiddity at all, has no capacity for commonness. By contrast, the nature has an intrinsic capacity for a commonness of sorts. It does not have a capacity to be numerically one in different substances. But it has a capacity to be quiddatively like the natures of certain other substances. We must distinguish then two notions of sameness: numerical identity and quiddative kinship. (I will introduce another in due course.) I shall use the relational predicate “quiddatively like” to express the latter. What of questions of numerical sameness? I shall codify Scotus’s views on numerical sameness in regard to natures by four theses about Socrates’s nature (that can obviously be generalized to all natures), which I shall label P1–4. P1: Socrates’s nature is not intrinsically essentially a universal. P2: Socrates’s nature is not intrinsically essentially a trope. P3: Socrates’s nature is extrinsically essentially unique to Socrates. P4: Socrates’s nature is essentially unique to Socrates.

In the section that follows, I expand on these theses along with Scotus’s reasons for embracing them.

2. A R G U M E N T S F O R P 1 –P 4

a. Argument for P1 In the Categories Aristotle had written that the definition of the secondary substance man—that is, the species man—is such that its definition distributes to individual men.1 Anything true by definition of the species man will be true of the individual man.2 Avicenna spotted that as a consequence one 1 Thus in Categories 5 (2a19ff.), we find “ . . . if something is said of a subject, then both its name and definition are necessarily predicated of the subject.” 2 Of course there are issues here that Aristotle and the medievals were aware of concerning whether one and the same characteristic—animality—can really be predicated of both a species and certain concrete individuals. (For example, don’t we risk giving the wrong answer to the question “How many animals are there?”?) I shall not engage with these issues here.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

66

John Hawthorne

cannot say that the species man is in and of itself a universal, for that would mean that it was, by definition, a universal.3 Applying the rule of the Categories, one would then be led to the absurd conclusion that each man is a universal. Call this the distribution argument. Scotus quotes with approval Avicenna’s view that horseness is not of itself a universal. The modifier “of itself,” deployed in both Scotus’s and Avicenna’s arguments, requires careful treatment. Indeed, I think it is key to understanding their views. Contemporary metaphysicians recognize a distinction between contingent and essential properties on the one hand and between intrinsic and extrinsic properties on the other.4 But it is less common for the intrinsic/extrinsic distinction to be applied to abstracta such as natures, and also uncommon for it to be applied to properties of being essentially this way or that. But in this context we need to avail ourselves of both these resources.5 In particular, we shall need to apply the fine-grained distinction between essential properties of natures that are intrinsically determined and essential properties that are extrinsically determined. (More carefully, it is the distinction between having the property of being essentially 3 “For if [animal] were universal in itself, in such a way that animality from the fact that it is animality is universal, no animal could possibly be singular; rather every animal would be universal,” Logica, as translated in Richard Bosley and Martin Tweedale, Basic Issues in Medieval Philosophy (Peterborough, ON: Broadview 1997), 402. 4 Regarding intrinsicality, there is a subtle distinction between an intrinsic property and a property that is not intrinsic but which is had intrinsically by some thing. Suppose we follow David Lewis in using a duplication heuristic for testing for intrinsicality (see Lewis, “Extrinsic Properties,” Philosophical Studies 44 (1983), 197–200). An intrinsic property is one that can never divide duplicates (intra- or transworld). Meanwhile, some x has a property p intrinsically just in case no possible duplicate of x lacks that property. Thus a square thing might have the property of being either a square thing or near a zebra intrinsically while a circular thing near a zebra will have that property extrinsically. And the second case indicates that the relevant disjunctive property is extrinsic. The question relevant in what follows is whether certain natures have various essential properties intrinsically. I certainly do not want to suggest that the medievals would think that the duplication idea is definitionally adequate. But there is certainly continuity between medieval discussions and Lewis’s intuitive gloss on an intrinsic property as one which “things have in virtue of the way they themselves are” (Lewis, On The Plurality of Worlds (Oxford: Blackwell, 1986), 61). 5 Another closely related way to reconstruct a view in the vicinity of Scotus’s would be to make use of Kit Fine’s distinction between having a property necessarily and the feature of having a property essentially where the latter entails the former but not vice versa (see, for example, Fine, “Essence and Modality,” Philosophical Perspectives 8 (1994), 1–16). One of Fine’s motivating examples is the relation between Socrates and singleton Socrates. He suggests that while singleton Socrates essentially and necessarily contains Socrates, it is necessary but not essential to Socrates that he belong to singleton Socrates. A reconstruction of some of Scotus’s thinking could be carried out along these lines, where the basic idea would be that a given nature is (i) neither essentially a trope nor essentially a universal but is (ii) necessarily a trope.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Scotus on Universals

67

F intrinsically and having the property of being essentially F extrinsically.)6 That fine-grained distinction is, I shall suggest, key to Scotus’s view. (Intrinsically essential properties, we may note, come in two varieties: primary ones come from—are grounded in—the whole of the thing itself and not from any essential proper part of the thing. Non-primary but per se essential properties come from some essential part of the thing, and not from the whole of the thing itself.)7 Let me introduce two complex predicates: “intrinsically essentially F” and “extrinsically essentially F.” Both of these entail “essentially F.” Of course “not intrinsically essentially F” does not entail “possibly not F” since the former (“not intrinsically essentially F”) but not the latter (“possibly not F”) is compatible with “extrinsically essentially F.”8 Using this terminology, we can reconstruct the relevant distribution argument as relying on the principle that if x’s nature is intrinsically essentially F, then x is F (which is naturally combined with the idea that the quiddity of a nature is given by what is intrinsically essential to it). This principle, combined with the premise that Socrates is not a universal, yields the conclusion: P1: Socrates’s nature is not intrinsically essentially a universal.

b. Argument for P2 With Avicenna, Scotus also believes that Socrates’s nature is not intrinsically essentially a trope—i.e., unique to just one substance. We may note that if Socrates’s nature were intrinsically singular, then there would be no explanatory 6 In various places that follow, for relevant F’s, there is logical space for a view according to which some nature is essentially F, is not intrinsically essentially F, but is intrinsically F. But these positions in logical space would not have seemed particularly appealing to the relevant metaphysicians. At least when it comes to natures, it is arguable that any property that is had intrinsically is also such that it is intrinsic that it have it essentially. This is arguably a difference between individuals and abstracta. It is easy to provide examples of properties that Socrates has intrinsically but not essentially. But it is much harder to provide compelling examples of this sort when it comes to numbers or Socrates’s nature. 7 See Ordinatio [= Ord.] II, d. iii, pt. 1, q. 6. My discussion will be exclusively focused on Ord. II, d. iii, pt. 1, qq. 1–6 (Opera VII: 391–494). Scotus’s discussion of these topics is not confined to that text. However, by all accounts, that text contains the most worked out version of his thoughts on individuation and natures. 8 I shall not have use for “x is intrinsically possibly F,” but were I to introduce it there would be a definitional choice point. It could mean “x is not intrinsically necessarily not F.” Or it could mean “x has a potentiality of being F that is grounded in the intrinsic profile of x.” On the former gloss, “x is intrinsically possibly F” does not entail “possibly F,” but on the latter it does.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

68

John Hawthorne

need for an individual difference. We could perhaps say that Socrates just is Socrates’s nature. So why not say this and throw out the individual difference? Scotus offers various arguments for the thesis that Socrates’s nature cannot be intrinsically essentially a trope (or more generally, cannot be intrinsically essentially inseparable from Socrates). Let me mention one such argument, also derived from Avicenna,9 which is implicit in the following passage: . . . I say that a material substance is not on account of its own nature this of itself, because, if it were, the intellect would not be able to ideate it under an aspect opposed [to this] without ideating its object under an aspect of ideation that conflicts with the character of such an object.10

Let me spell out what I take to be the line of thought here: (a) When we, by abstraction, grasp Socrates’s nature, we go on to use the abstracted object of thought as a universal that is truly predicated of things that are not inseparable from (i.e. really distinct from) Socrates. (We do this by forming a second intention of universality directed at the abstracted nature.)11 (b) An object of the intellect that is intrinsically F cannot truly be predicated of things that are not F. (c) So if the nature we abstract from Socrates were intrinsically inseparable from Socrates then it could not be truly predicated of things that are not inseparable from Socrates. (d) So the nature we abstract from Socrates is not intrinsically inseparable from Socrates. (Note this argument does not show that the nature is not intrinsically numerically identical to itself. But it does show that it is not intrinsically identical to, or inseparable from, Socrates.)12

9 “But if animal from the fact that it is animal were singular, it would be impossible for there to be more than one singular, viz. the very singular to which animality is bound, and it would be impossible for another singular to be an animal” (Logica, tr. Bosley and Tweedale, Basic Issues, 402). 10 Scotus, Ord. II, d. iii, pt. 1 q. 1 n. 29; tr. Bosley and Tweedale, Basic Issues, 407. (Cf. Paul V. Spade, Five Texts on the Mediaeval Problem of Universals [Indianapolis, IN: Hackett, 1994], 61.) The Ordinatio translations that follow are from Bosley and Tweedale. 11 See Ord. II, d. iii, pt. 1, q. 1, where universality is explained as a mode of ideating an abstracted nature (and thus a second intention) rather than as part of the abstracted concept itself. 12 It might be suggested that Scotus has neglected the possibility that Socrates’s nature is intrinsically such that it is inseparable from some individual even though it is not intrinsically inseparable Socrates. This issue will be addressed later.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Scotus on Universals

69

We have, then, the following conclusion: P2: Socrates’s nature is not intrinsically essentially a trope.

c. Argument for P3 and P4 So now it seems that neither tropehood nor universalhood is intrinsically essential to Socrates’s nature. Is either extrinsically essential? Especially pertinent here is whether it is extrinsically essential to Socrates’s nature that it belong to Socrates. What would it be for this to be extrinsically essential? Well, it would be for it (a) to be essential so that it is really impossible for Socrates’s nature to belong to a different thing but (b) for this essential feature to be explained by something other than its own intrinsic nature. The obvious candidate explanans is the contracting individual difference. Scotus writes: Once the contracting factor is given it [the nature] cannot belong to something else, although it does not of itself reject belonging to something else.13

What the passage seems to be saying is that while Socrates’s nature is not intrinsically essentially unique to Socrates, Socrates’s nature is extrinsically essentially unique to Socrates (and thus essential to Socrates). In his paper “Duns Scotus on the Common Nature and the Individual Differentia,” Peter King contends that Scotus is not committed to its being essential to Socrates’s nature that it belongs to Socrates. He writes: The actualization of a common nature by an individual differentia does not ‘use up’ the real potencies belonging to the nature, which are retained even while contracted. . . . A nature that is contracted to a given individual cannot metaphysically be contracted to a different individual, but it retains a real power to have been and to be so contracted.14

The idea is that while Socrates’s nature cannot simultaneously be Socrates’s nature and the nature of another, it is such that it could have been the nature of another. King goes on to say this: Socrates’ human nature, as contracted, is not and cannot be the contracted nature of another: as contracted, it is always Socrates’ humanity and wherever it exists there Socrates is. Nevertheless, Socrates’ contracted human nature retains a real potency to be the contracted nature of another.15

13 Ord. II, d. iii, pt 1, q. 1, n. 41. 15 Ibid.

14 Philosophical Topics 20 (1992), 67.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

70

John Hawthorne

How is the modifier “as contracted” working here? Suppose we think that a lump can be a statue so long as it is statue-shaped and freestanding and can be a doorknob if it is doorknob-shaped and stuck to a door in the right way. We can say then: as a doorknob, the lump cannot be a statue. What we mean is that in no possible world is the lump both a statue and a doorknob at the same time. So King’s view is that Socrates’s nature could have been Plato’s though it could not have been both Socrates’s and Plato’s at the same time. There seem to be two problems with King’s account. First, it allows that a nature in one substance at one time could be numerically identical to a nature in another substance at another time. This would be inter-substantial commonality in reality by the greater unity—numerical unity—which Scotus explicitly rejects. Second, it renders strange Scotus’s opinion that, while formally distinct (more on this ideology later), Socrates’s nature and the Socrates’s individual difference are really the same. Their being really the same, as Scotus understands things, entails that they are inseparable. But on King’s account, Socrates’s nature and the Socrates’s individual difference appear to be separable in that the former can exist without the latter. (One further observation: King cites Scotus as claiming that Socrates’s nature could belong to something else. But all Scotus says is that the nature in itself is compatible with its belonging to something else. This does not entail that the nature could belong to something else, given that some essential properties are externally determined.) Here’s the alternative picture I am proposing. What the nature is like quiddatively is intrinsically determined but the features that determine it as a trope are extrinsically determined. Socrates’s nature is singular, meaning that it is essentially unique to Socrates, though it is not singular in itself (i.e. intrinsically essentially unique to Socrates). Rather, it is singular thanks to the individual difference that it contracts with. And so we have: P3: Socrates’s nature is extrinsically essentially unique to Socrates.

From which follows: P4: Socrates’s nature is essentially unique to Socrates.

P4 is just the trope view. So vis-à-vis contemporary positions, Scotus’s view is not distinctive after all: Scotus believes that the natures that exist in reality are tropes. What he adds to contemporary trope theory is a refinement along a dimension that contemporary theorists do not typically discuss, namely: granted that Socrates’s nature is a trope, is it a trope intrinsically (per se or primarily) or extrinsically (denominatively)? Neither Socrates nor the Socratizer nor the Socratic nature is bi-locatable so as to be fully present in more than one individual. But in the first case (Socrates) this is a per se

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Scotus on Universals

71

feature, in the second (the individual difference) it is a primary feature (each of this pair being, in a different way, intrinsically essentially a trope), and in the third (Socrates’s nature) it is a denominative feature.

3. SCOTUS ON NUMERICAL UNITY I mentioned at the start that Scotus speaks about natures as having less than “numerical unity.” What does this amount to? Here is a relevant passage: . . . what has a real unity peculiar to it and sufficient for it, but less than a numerical unity, is not of itself one by a numerical unity (i.e. is not of itself this). But the nature existing in this stone has a real and sufficient unity peculiar to it, and one less than numerical unity. Therefore.16

Consider the nature of Socrates. That nature enjoys the sort of unity peculiar to natures, namely that of being quiddatively like something. Let us allow that quiddative unity is less than numerical unity in the sense that numerical unity with a thing is more intimate than quiddative unity. (We cannot quite say that whenever there is numerical unity between A and B, there is quiddative unity, since an individual difference has numerical unity but has no quiddative unity with anything.) Now insofar as we read “numerical unity” as “numerically identical to something” (as that expression would be understood by modern philosophers), it is hard to fashion a cogent view out of the above passage and those similar to it. Consider the inference scheme: ∃x (a R x) Therefore ∃y (a is identical to y)

This looks like a perfectly acceptable inference scheme.17 But if it is, then what is sufficient for being a nature is sufficient for numerical unity too. (Take the instance of the schema that uses quiddative kinship for R.) Of course having inferred that x has numerical identity to something from the fact that it is has quiddative unity with something, we do not have to require that the things it is has quiddative unity with are the very things that it is numerically identical to. In sum, granted that quiddative unity with a thing 16 Ord. II d. iii pt. 1 q. 1 n. 8. 17 I leave it to others to explore the radical option of representing Scotus’s views within a positive free logic that denies the validity of that inference. (With that kind of logic in place, the truth of both “I hate the Grinch” and “There is nothing identical to the Grinch” can be taken at face value.)

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

72

John Hawthorne

does not entail numerical identity with that thing, it may still entail numerical identity with something or other. In that sense quiddative unity does entail numerical identity. Were we to construe matters in this way and still deny that a nature has numerical unity intrinsically while affirming it has quiddative unity intrinsically, we would somehow have to deny that the logical consequences of that which is had intrinsically are also had intrinsically. But making such a position cogent is an uphill task. That said, it is not plausible to interpret “numerical unity” as “identical to something.” Famously, Scotus made use of the ideology of both real distinctions and formal distinctions. Real distinctness is standardly understood as involving a kind of separability.18 Meanwhile, things can be formally distinct even if they are not relevantly separable. And, not surprisingly, formally distinct things can have different features, since inseparability does not entail duplication of characteristics. Thus it is formal distinctness rather than real distinctness that is the correlate of the numerical distinctness of modern philosophy, since being really the same (the opposite of real distinctness) does not satisfy Leibniz’s Law. (It is arguable, though, that the relation of being really the same is reflexive, symmetrical, and transitive and thus is an equivalence relation. It is thus analogous in that respect to numerical identity—and one can use it as the basis for a kind of counting— though of course there are plenty of equivalence relations other than numerical identity.) “Numerical unity,” in the current context, roughly amounts to being really the same as some individual substance, where being really the same as something crucially does not entail being numerically identical to it, and where having the property of being numerically identical to something certainly does not entail the property of being really the same as some individual substance. Once we are reminded of this, and take advantage of the distinction between intrinsic and extrinsic determination, it is easy enough to see what is going on. While Socrates’s nature is really the same as Socrates’s individual difference, that feature of Socrates’s nature is not intrinsically determined (since, as we have seen, Scotus denies that it is intrinsic to Socrates’s nature that it is inseparable from, and so really the same as, Socrates’s individual difference). Indeed, it is not even intrinsic to Socrates’s nature that it belongs to just one individual substance. Hence it is not intrinsic to Socrates’s nature that it has numerical unity.

18 Things may not be so straightforward however. See Joshua Blander, Dependence, Separability and Theories of Identity in Late Medieval Philosophy: Case Studies from Scotus and Ockham (Ph.D. Dissertation, UCLA, 2014). What is important for my purposes is that having numerical unity in Scotus’s sense involves something more than being numerically identical to something or other.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Scotus on Universals

73

(We can also get clear in what sense the lesser unity is sufficient for, say, Socrates’s nature. To have the lesser unity is to have quiddative kinship with something, and the particular lesser unity a nature has is given by its quiddative kinship profile. The lesser unity is obviously not sufficient for being numerically identical to Socrates’s nature nor being really the same as Socrates’s nature, since other natures will have the very same quiddative kinship profile. But the lesser unity is sufficient for being the same as Socrates’s nature is in itself. That is, the lesser unity is sufficient for being intrinsically just like Socrates’s nature.) With this in place, we can make good sense of another argument that Scotus presents for P2. Here is how Scotus goes on to justify his rejection of the view that a thing’s nature could “from its own nature alone” enjoy both the greater unity of numerical unity and the lesser unity of quiddative unity: But these when taken as about the same item and in respect of the same item are opposed, because a multiplicity opposed to the greater unity can co-exist without contradiction with the lesser unity, but this multiplicity cannot co-exist with the greater unity because it rejects it: therefore, etc.

The lesser unity is qualitative kinship. Let me try to unpack various aspects of Scotus’s thinking here. One underlying thought is that qualitative kinship can be shared by a multiplicity that are really distinct. But if each nature was intrinsically determined to be really the same as some particular individual, this would preclude qualitative kinship with the natures of other things. Consider, by analogy, that the nature of a particular dog is not quiddatively like that of the nature of a particular cat. And that is because the natures of that cat and dog intrinsically involve not just the genus of animality but also the differentia that separate cats from dog. But if the natures of Socrates and Plato intrinsically involve such features as “belonging to Plato” and “belonging to Socrates” then, by analogy, their natures would not even be quiddatively alike. Thus in order that the nature of Socrates be capable of enjoying qualitative kinship with the natures of other individuals, it is important that this nature not be intrinsically determined to belong to Socrates. The above reflections are faithful to Scotus’s thinking. But the passage brings in a further thought. We have just seen that for Plato’s and Socrates’s natures to enjoy qualitative kinship, it is important that it not be that one is intrinsically tied to Plato and the other to Socrates. But it does not immediately follow that both cannot intrinsically have the characteristic of being inseparable from some individual or other. Allowing this would not induce an intrinsic dissimilarity between the two natures. Thus the requirement that a multitude be able to have intrinsically matching natures will only get one so far. In the above passage, Scotus actually suggests something stronger.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

74

John Hawthorne

He suggests that it is not incompatible with what is intrinsic to the nature that there be a multitude of really distinct things that share the nature. This does not merely preclude its being intrinsic to Socrates’s nature that it be really the same as Socrates’s individual difference. It also precludes its being intrinsic to Socrates’s nature that it belong exclusively to some individual substance or other.

4 . O C K H AM ’ S CR IT IQ U E Let us now turn to Ockham’s critique of Scotus’s position in Ordinatio Book I, d. 2 q. 6, one that Marilyn Adam thinks is decisive.19 I lack the space to engage with all aspects of Ockham’s discussion, but let me touch on four themes. My overarching conclusion will be that the critique is less than decisive and that Ockham does not properly grasp the contours of Scotus’s position. One noticeable feature of Ockham’s discussion is that he deploys a conception of sameness that is very much of a piece with modern conceptions of numerical identity. For example, he notes that “contradiction is the strongest way to prove a distinction of things.”20 On this basis he argues that if Socrates’s nature is a thing in the world it is a thing that is different from Socrates’s contracting difference (his haecceity), since they would have different features. Given this, he sees trouble in Scotus’s view that the nature is a thing that is really the same as the contracting difference. But there is not really a problem here. The relation of being really the same, as deployed by Scotus, does not entail numerical identity. And, in particular, while x and y’s having contradictory properties entails they are numerically distinct, that does not entail they are really distinct. Again, supposing we run with the toy theory that x and y are really the same if and only if they are inseparable from the same individual substance, then it is clear enough that x and y’s having contradictory properties will not show that they are not really the same. A second argumentative theme is that any given nature is not really common to many and so is not a universal.21 There are terminological issues here that are relatively unimportant. If we make it part of the definition of “universal” that it be separable from, not really the same as, 19 See Adams, William Ockham, vol. 1 (South Bend, IN: University of Notre Dame Press, 1987), 53. 20 Ockham, Ordinatio Bk. I, d. 2, q. 6 par. 24 (Opera theol. II:174). Here and elsewhere I am using John Kilcullen’s online translation . 21 See for example, ibid., par. 40 (Opera theol. II:179).

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Scotus on Universals

75

an individual substance, then we can agree that Scotus does not admit universals. Yet that does not yet settle whether Socrates’s nature is truly predicable of many things. And if being truly predicable of many things is taken to be the defining feature associated with “universal” then Socrates’s nature may well be universal in that sense. The most interesting issue is how two notions of commonality line up. One way that a nature can be common to many things is to be really present as a constituent of each of those things. Another way that a nature can be common is by being predicable of many things. On a simple, arguably naïve, conception of predication, a thing can be truly predicated of a substance only if it is a constituent of that substance. It is clear that Scotus must reject this. Scotus’s basic thought is that a nature is unfit to be predicated of an object only if the features intrinsically associated with the nature do not distribute to the object. What one needs to show is that this aspect of his thinking is somehow incoherent or misguided. Third, Ockham suggests a very interesting constraint on grounding. His basic thought is that facts of numerical identity and distinctness are basic and not grounded in further facts. On this view, facts of identity and difference are so bound up with the being of the objects that it is a mistake to look for something else in the world to ground them. Thus he tells us: “to be the same and diverse follow upon being immediately.”22 And so we have: Ockham’s Grounding Constraint: Facts of identity and distinctness are not grounded in further facts.

I suspect that this grounding constraint has quite a bit going for it. But does it tell against Scotus’s view? Ockham clearly thinks it does, since he accuses Scotus of violating that constraint by supposing that the non-identity of Socrates’s nature and Plato’s nature is grounded in some further facts, namely, the relation each has to a certain contracting difference. Again, it is not clear to me that this worry is decisive. Suppose we adhere to the grounding constraint and admit that it is a basic fact that Socrates’s nature is numerically distinct from Plato’s nature. It is perfectly consistent with this that there be some fact that grounds the inseparability of Plato’s nature with Plato and hence that grounds Plato’s nature being really the same as Plato. And relatedly, it is perfectly compatible with Ockham’s grounding constraint that there be something that grounds Plato’s nature having numerical unity in the sense outlined earlier. The basic point is that Ockham’s grounding constraint is quite consistent with the hypothesis that certain cases of being really the same are grounded. (Of course it is a further issue whether Scotus’s particular views here are well motivated or compelling.) 22 Ibid., par. 55 (Opera theol. II:184).

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

76

John Hawthorne

Finally, I’d like to point out that what initially looks like a successful master argument against Scotus’s view is ultimately a failure on account of its neglect of the role of intrinsic determination. The line of argument goes like this. According to Scotus, Socrates’s nature is really the same as Socrates’s individual difference, and as a consequence Socrates’s nature cannot exist in a multiplicity of beings. But then Scotus has no right to say that things with a lesser unity (quiddative kinship) can exist in a multitude of beings, since those things are really the same as individual differences and so cannot exist in a multitude of beings after all: Moreover his (i.e. Scotus’) statement that “the multitude opposed to the greater unity can stand without contraction with the lesser unity” seems repugnant with another of his statements, that nature and individual difference do not differ really; because when some two are the same really, whatever either can be really by divine power, the other can be. But the individual difference cannot be several in number really distinct; therefore the nature, which is really the same with that contracting difference, cannot be many really; neither, consequently, can anything be except the contracting difference; and thus the nature is not compatible without contradiction with numerical multitude.23

This argument is telling only if we can saddle Scotus with the view that Socrates’s nature can exist in a multitude (thanks to its lesser unity). But Scotus does not hold this. He does not hold that thanks to how Socrates’s nature is in itself, it is possible for Socrates’s nature to exist in a multitude. Ockham is right to think that this would be a disaster for Scotus. What Scotus holds is that thanks to how Socrates’s nature is in itself, it is not intrinsically necessary that it contracts with Socrates’s individual difference and not intrinsically necessary that it exists in only one individual substance. But a claim to the effect that it is not intrinsically necessary that P does not entail that it is possible that not P, and indeed is quite compatible with the claim that it is necessary that P. Scotus holds that it is necessary that Socrates’s nature exist in Socrates but that it is not intrinsically necessary that Socrates’s nature does so. There is no failure of coherence here. Ockham’s critique once again misses the mark. University of Southern California

23 Ibid., par. 39 (Opera theol. II:179). Many thanks to Tamar Gendler for comments on a draft of this paper and to members of the audience at the Medieval Philosophy Colloquium, Cornell University, June 5, 1999, where a version of this paper was presented. I had not looked at the paper for many years until Robert Pasnau recently expressed some interest in publishing it. I am grateful to him and an anonymous referee for providing help and encouragement.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Scotus on Universals

77

BIBLIOGRAPHY Adams, Marilyn. William Ockham, vol. 1 (South Bend, IN: University of Notre Dame Press, 1987). Aristotle. Complete Works, ed. J. Barnes (Princeton, NJ: Princeton University Press, 1984). Blander, Joshua. Dependence, Separability and Theories of Identity in Late Medieval Philosophy: Case Studies from Ockham and Scotus (Ph.D. Dissertation, UCLA, 2014). Bosley, Richard and Martin Tweedale (eds). Basic Issues in Medieval Philosophy (Peterborough, ON: Broadview, 1997). Fine, Kit. “Essence and Modality,” Philosophical Perspectives 8 (1994), 1–16. John Duns Scotus. Opera omnia, ed. C. Balić et al., 21 vols. (Vatican City: Typis Polyglottis Vaticanis, 1950–2013). King, Peter. “Duns Scotus on the Common Nature and the Individual Differentia,” Philosophical Topics 20 (1992), 51–76. Lewis, David. “Extrinsic Properties,” Philosophical Studies 44 (1983), 197–200. Lewis, David. On The Plurality of Worlds (Oxford: Blackwell, 1986). Spade, Paul V. Five Texts on the Mediaeval Problem of Universals (Indianapolis, IN: Hackett, 1994). William Ockham. Opera philosophica et theologica (St. Bonaventure, NY: Franciscan Institute, 1967–89). William Ockham. Ordinatio, I, dist. 2, q. 6, tr. J. Kilcullen .

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Duns Scotus, the Natural Law, and the Irrelevance of Aesthetic Explanation Jeff Steele 1. INTRODUCTION At the heart of Duns Scotus’s ethics lies the distinction between natural law in the strict sense and the broad sense. Natural law in the strict sense contains only those moral propositions whose truth value is known from their terms alone (veritas nota ex terminis), or conclusions that necessarily follow from them.1 Roughly, the natural law in the strict sense includes the First Table of the Decalogue (i.e., our obligations to God himself), which are necessary truths that not even God can alter. In contrast, the natural law in the broad sense includes the precepts of the Second Table of the Decalogue, namely, these precepts involving our relationship with one another. Scotus claims that God can make dispensations from these Second Table precepts, since these precepts are not logical deductions following necessarily from the First Table. Nevertheless, they are “highly consonant” with it.2 However, Scotus does not offer an exact explanation of what he means by saying the Second Table precepts are “highly consonant” with the natural law in the strict sense. Consonance must be weaker than entailment, since these precepts do not necessarily follow from our obligations to God; Scotus is also clear that consonance does not require God to prescribe the exact laws he did in fact prescribe. But these claims only tell us what consonance is not. What is the positive content of the claim that the precepts of the Second Table are “highly consonant” with the law of nature in the strict sense? Let us call this the “Consonance Problem.” 1 See Ordinatio III, d. 37, q. un., n. 16; Ordinatio IV, d. 17, q. un, n. 19; Ordinatio IV, d. 33, q. 1, n. 22. 2 Ordinatio III, d. 37, q. un., nn. 25–26.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Duns Scotus, the Natural Law

79

Until recently, two basic strategies for solving the Consonance Problem were locked in a stalemate: the teleologist solution and the reductionist solution. While the teleologist approach attempts to explain consonance in terms of features that fit our natures and our appropriate ends,3 the reductionist approach reduces consonance to the role the Second Table precepts play in fulfilling our obligation to the First Table (loving God).4 Recently, a fresh and interesting proposal interprets consonance in terms of aesthetic considerations: while God does not have moral reasons for choosing specific Second Table precepts, he chooses particular contingent precepts on the basic of aesthetics. I shall argue, however, that the aesthetic solution to the Consonance Problem falters in establishing that consonans ought to be interpreted aesthetically, and furthermore, the interpretation contradicts explicit statements Scotus makes about divine freedom and what God can do: namely, it places constraints upon God’s will beyond mere logical possibility. 3 Some have desired to soften Scotus’s claims about morality and God’s dispensations by interpreting the Second Table’s precepts as being “highly consonant” with human nature as the reason God chose them. On this view, Scotus is not really a voluntarist about Second Table precepts, but a crypto-Aristotelian teleologist, even though God has sometimes made a dispensation from such laws. For example, Allan Wolter writes: Scotus maintains . . . that while the second table represents what is “valde consonans” with natural law, certain aspects of the second table of the decalogue can be dispensed with according to right reason, when their observation would do more harm than good. But God could obviously not dispense from all its precepts at once, for this would be equivalent to creating man in one way and obligating him in an entirely different fashion, something contrary to what he “owes to human nature in virtue of his generosity” (Duns Scotus on the Will and Morality (Washington, DC: Catholic University of America Press, 1986), 24). So on this interpretation, the Second Table precepts are intimately linked with human nature and are only dispensed with in special circumstances. But in general, Second Table precepts align with the teleological direction human nature provides. Mary Beth Ingham also endorses something like this view. See Ingham, “Duns Scotus, Morality, and Happiness: A Reply to Williams,” American Catholic Philosophical Quarterly 74 (2000), 173–95, and “Letting Scotus Speak for Himself,” Medieval Philosophy and Theology 10 (2001), 172–216. 4 In other words, consonance should be reduced to conformity with the First Table. Thus no positive content exists beyond being serviceable to express love of God. Thomas Williams has endorsed something like this view in various places. See “How Scotus Separates Morality from Happiness,” American Catholic Philosophical Quarterly 69 (1995), 425–45; “Reason, Morality, and Voluntarism in Duns Scotus: A PseudoProblem Dissolved,” The Modern Schoolman 74 (1997), 73–94; “The Libertarian Foundations of Scotus’s Moral Philosophy,” The Thomist 62 (1998), 193–215; and “A Most Methodical Lover? On Scotus’s Arbitrary Creator,” Journal of the History of Philosophy 38 (2000), 169–202. John E. Hare also seems to endorse this view. See Hare, God and Morality: A Philosophical History (Malden: Blackwell Publishing Ltd, 2009), 97–105.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

80

Jeff Steele 2 . N AT U R A L L A W

According to Scotus, natural law in the strict sense contains those [moral] propositions whose truth is known from their terms alone (veritas nota ex terminis), or conclusions that necessarily follow from them.5 Scotus identifies these necessary, a priori moral truths with the following principle: “If God exists, then he alone ought to be loved as God.”6 For Scotus, the content of the First Table of the Decalogue—namely, those commandments and prohibitions pertaining to our relationship with God—follows directly from this principle.7 According to Scotus, the commandment and prohibitions contained in the First Table have their respective truth values prior to any decision on the divine will, and thus are independent of God’s will or commands.8 These natural law principles are “necessarily true,” and “not even God could make them false.”9 Nor can God make a dispensation from these precepts, in such a way that someone could rightly will the opposite of what these precepts command or prohibit.10 So these precepts are always obligatory and no exception can be made—not even by God himself. In contrast, Scotus argues that the Second Table of the Decalogue is a part of natural law only in a broad sense. These commandments and prohibitions (e.g., do not murder, do not steal, etc.) are not necessary, a priori truths, like the First Table, nor do they follow necessarily from such truths.11 Scotus argues that God could make a dispensation from them. For example, in arguing against the position that all the commandments in the Decalogue belong to natural law in the strict sense (Aquinas’s view12), Scotus makes the following claim: I ask therefore, whether leaving all the circumstances in the act of ‘killing a human’ the same, except the one circumstance of whether the act is prohibited or not 5 See Ordinatio III, d. 37, q. un., n. 16; Ordinatio IV, d. 17, q. un, n. 19; Ordinatio IV, d. 33, q. 1, n. 22. 6 Ordinatio III, d. 37, q. un., n. 20. The first two commandments, in the form of prohibitions (“you shall have no other gods,” and “you shall not show irreverence for God”) follow from this principle. 7 Ordinatio III, d. 37, q. un., n. 19. 8 See Ordinatio III, d. 37, q. un., nn. 14, 20. 9 Ordinatio III, d. 37, q. un, n. 3. “Those precepts which belong to the law of nature are either practical principles known from their terms, or are conclusions that follow necessarily from such principles; in either case, they have necessary truth. Therefore, God cannot make it the case that they are false.” While these statements appear in the initial pro and con arguments, it is clear from the context (see n. 40) that Scotus accepts this opinion, and the arguments made there support his overall viewpoint. 10 Ordinatio III, d. 37, q. un., n. 20. 11 Ordinatio III, d. 37, q. un., n. 18, 25. 12 For Aquinas’s view, see Summa Theologiae I–II, q. 100, a. 1 and a. 8.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Duns Scotus, the Natural Law

81

prohibited, can God bring it about that that very act, which with those same circumstances is prohibited and illicit at one time, would be not prohibited and licit at another time? If so, then God can make a dispensation without qualification, in the same way he changed the Old Law—with respect to the ceremonial laws— when he gave the New Law. Indeed, he did not bring it about that the precepts of the ceremonial law remained binding but were not to be observed, but he made it the case that the act remained the same, but people were no longer obligated to it as they had previously been. And so this is how any legislator makes an unqualified dispensation: when he revokes a precept of his own positive law, namely, by making it the case that the prohibited or illicit act remains the same in its nature, but the feature of its being prohibited or illicit is removed, and the act is made licit. Yet if God cannot bring it about that this act [of killing], which, with a particular set of circumstances was prohibited, is no longer prohibited with the same circumstances as when it was prohibited, then God cannot make it the case that killing is not prohibited—But it is obvious that God did make it that case that killing was not prohibited with Abraham and others.13

Since natural laws (strictly considered) are necessary moral truths that cannot be changed or dispensed with, and since God has made a dispensation with respect to the various precepts of the Second Table in the Ten Commandments, then contra Aquinas, not every one of the Ten Commandments belongs necessarily to the natural law. And more importantly, God has the ability to make the same act—under the same circumstances— either obligatory or prohibited, simply by divine fiat. Thus precepts of the Second Table depend upon God’s will for their moral worth, since the very same act could be prohibited or obligatory simply by God’s choice. However, Scotus notes that while these precepts in the Second Table of the Decalogue are not logical deductions following necessarily from the First Table, they are nevertheless “highly consonant” with it: There is another way that some precepts can be said to belong to the law of nature, because they are highly consonant with that law, although they do not follow necessarily from the first practical principles, which are known from their terms and are necessarily known by every intellect. And in this [second] way, it is certain that all the precepts—even the second table—belong to the law of nature, because their rectitude is highly consonant with the first practical principles that are known necessarily.14

This Consonance Passage has created serious disparity among Scotus scholars, and no one entirely knows what he means by this, since Scotus nowhere gives a precise account of what he means by saying that the 13 Ordinatio III, d. 37, q. un., n. 13. 14 Ordinatio III, d. 37, q. un., nn. 25–26.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

82

Jeff Steele

precepts of the Second Table are “highly consonant” with the natural law in the strict sense. He is clear that consonance is something weaker than entailment, since such precepts do not necessarily follow from our obligations to God; he is also clear that consonance does not require God to prescribe the exact laws he did in fact prescribe. But these claims only tell us what consonance is not. What is the positive content of the claim that the precepts of the Second Table are “highly consonant” with the law of nature in the strict sense?

3 . T HE A E S T H E T I C S O L U T I O N While the standoff between teleologist and reductionist continues, a third solution to the Consonance Problem has recently been developed by Richard Cross15 and Oleg Bychkov.16 This view denies that God has any moral obligation to prescribe certain laws to fit our nature. It also denies that God has moral reasons for preferring one ordering of laws to another. But to say that God has no moral reasons to prescribe certain laws is not to say that he has no reasons at all. According to this view, he has non-moral reasons: aesthetic reasons. So when Scotus says that the contingent, Second Table precepts are “highly consonant” (valde consonans) with the First Table precepts, he means that God chooses these contingent precepts because they beautifully fit with the First Table precepts concerning our obligation to God. Unfortunately, the aesthetic solution to the Consonance Problem fails to establish that consonans and its various synonyms should be taken aesthetically, while simultaneously creating constraints upon Scotus’s God that Scotus adamantly rejects.

15 Richard Cross, “Natural Law, Moral Constructivism, and Duns Scotus’s Metaethics: The Centrality of Aesthetic Explanation,” in J. A. Jacobs (ed.), Reason, Religion, and Natural Law: From Plato to Spinoza (Oxford: Oxford University Press, 2012), 175–97. 16 Oleg Bychkov, “ ‘In Harmony with Reason’: John Duns Scotus’s Theo-aesth/ ethics,” Open Theology 1 (2014), 45–55. Bychkov has been working on Scotus’s aesthetics and its relationship to ethics for a number of years. For example, see Aesthetic Revelation: Reading Ancient and Medieval Texts after Hans Urs von Balthasar (Washington DC: Catholic University of America Press, 2010), and “ ‘Aesthetic’ Epistemology: Parallels between the Perception of Musical Harmony and the Cognition of Truth in Duns Scotus,” in L. Honneflder, et al. (eds), John Duns Scotus 1308–2008: Investigations into his Philosophy (Münster: Aschendorff, 2010), 345–56.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Duns Scotus, the Natural Law

83

a. Interpretational Problems i. Richard Cross. According to Richard Cross, Scotus holds a constructivist view of the Second Table prescripts: while God has reasons for choosing the specific contingent laws he chose, these reasons are not moral, but aesthetic. Here, Cross appeals to various passages from Scotus purporting to show various aspects of the moral realm are laced with aesthetics; more specifically, Scotus’s ethical writings are filled with terms that have aesthetic connotations, such as “harmony” (consonantia) and “fittingness” (convenientia), which Cross thinks Scotus uses equivalently.17 In order to show that Scotus believes the universe to be governed by aesthetic considerations, Cross appeals to what he sees as aesthetically minded passages. For example, he appeals to the following two passages on the cognitive powers of the intellect perceiving God’s infinite being, stating that Scotus appeals to “aesthetic considerations” here: The sense powers, which are less cognitive than the intellect, immediately perceive inappropriateness (disconvenientiam) in their object: this is clear in the case of hearing relative to an inappropriate object. Therefore if infinite were incompatible with being, the intellect would immediately perceive this inappropriateness and incompatibility, and then it could not have infinite being as its object.18

And, The intellect, whose object is being, finds no incompatibility understanding something infinite: rather, it seems perfectly intelligible. But it is remarkable if a contradiction in relation to the first object is made evident to no intellect, when discord (discordia) in sound so easily offends the hearing.19

Cross’s argument seems to be that Scotus appeals to aesthetics at various points in his philosophy and this thereby lends support to an appeal to aesthetics in solving the Consonance Problem. However, aesthetic considerations do not seem to play a significant role in the above passages at all. The disconvenientia of “infinite” and “being” (if they were in fact disconveniens to each other) would not be aesthetic at all, but logical/conceptual. It is not that infinite and being would be like checks and plaids or whatever— jarring and hideous in combination—but that they would be logically incompatible with each other. The point of Scotus’s analogy is not that 17 Cross, “Natural Law, Moral Constructivism, and Duns Scotus’s Metaethics,” 189. 18 Scotus, Reportatio I-A, d. 2, part 1, qq. 1–3, n. 72, tr. Wolter and Bychkov, John Duns Scotus: The Examined Report of the Paris Lecture: Reportatio I-A. Cf. Cross, “Natural Law, Moral Constructivism, and Duns Scotus’s Metaethics,” 190. 19 Ordinatio I, d. 2, part 1, qq. 1–2, n. 136. Translation from Cross, “Natural Law, Moral Constructivism, and Duns Scotus’s Metaethics,” 190.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

84

Jeff Steele

the disconvenientia of “infinite” and “being” is the same thing as the disconvenientia of hearing and an ugly sound; it is that both are known in the same way: by immediate, non-inferential perception in a sense of “perception” that is not limited to the senses. In other words, this passage is not about aesthetics at all; it is a defense of the idea that there is such a thing as logical intuition. This point is made clearer if the last quote above by Cross were carried a little further: The intellect, whose object is being, finds no incompatibility understanding something infinite: rather, it seems perfectly intelligible. But it is remarkable if a contradiction in relation to the first object is made evident to no intellect, when discord (discordia) in sound so easily offends the hearing. . . . For if the unfittingness immediately offends when it is perceived, why does no intellect naturally flee the understood infinite as if from something incompatible, and thus destructive of its first object?20

Scotus’s point is not that perception of the infinite being is some kind of aesthetic taste we have for confluence of certain metaphysical concepts, but that both a discord in harmony and a logical/conceptual contradiction are known immediately and intuitively. The harmony of a discordant sound merely serves as an example. Furthermore, even if Scotus appeals to aesthetic considerations at various places in his thought, it does not mean that such an appeal is relevant to the current topic. Suppose Scotus’s use of disconvenientia above happens to be an aesthetic notion. What follows from it about the nature of the Second Table precepts? Nothing: there is no reason consonans or convenientia has to be an aesthetic notion when talking about the moral law even if it is an aesthetic notion here, unless we can demonstrate that is how Scotus always uses these terms. And as the passages assembled earlier in the paper suggest, Cross seems to take it for granted that convenientia and disconvenientia are always aesthetic terms.21 While convenientia and disconvenientia do apply to beauty,22 the notions in themselves are much broader. Scotus uses convenientia throughout his works not simply in aesthetic contexts, but in order to convey a 20 Ordinatio I, d. 2, part 1, qq. 1–2, n. 136. The first half of the translation is from Cross; the added, italicized translation is mine. 21 Cross, “Natural Law, Moral Constructivism, and Duns Scotus’s Metaethics,” 188–90. Cross states: “On this view, God has reasons, but not moral ones, for endorsing certain moral claims. It might be thought that these aesthetic considerations are too weak to do the relevant grounding. But Scotus believes that they are much more robust than we might imagine” (190). He then goes on to cite the passages concerning perceiving God’s infinite being as an example of this (i.e., Ordinatio I, d. 2, part 1, qq. 1–2, n. 136), and suggests that Scotus’s use of disconvenientia is an aesthetic notion. 22 See Ordinatio I, d. 17, part 1, q. 2, n. 62 and Ordinatio II, d. 40, q. un, n. 7.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Duns Scotus, the Natural Law

85

proper relationship among entities that ought to stand in some relationship to each other.23 For example, in his various questions on the works of Aristotle, Scotus often asks whether Aristotle appropriately or correctly applies or defines something, or whether some definition suitably captures the things being defined. And in each of these cases, he uses convenientia.24 Furthermore, in Questions on Porphyry’s Isagoge, Scotus uses “convenientia” to refer to the fittingness or properness of the definition or concept, given the nature of the terms involved. A few cases will suffice to make the point: Q.15: “Whether the definition of ‘genus’ is fittingly (convenienter) assigned.” Q.17: “Whether ‘differing in species’ is fittingly (convenienter) posited in the definition of ‘genus.’ ” Q.19: “Whether ‘in quid’ is appropriately (convenienter) posited in the definition of ‘genus.’ ”25

Now we would be foolish to say that Scotus’s use of convenientia in his writings on Aristotle and Porphyry conveys some deep aesthetic truth about reality! On the contrary, Scotus uses convenientia to signify a proper relationship between two entities that ought to stand in a particular relationship with one another. And one type of convenientia happens to be beauty. But simply pointing out that Scotus uses the word convenientia to refer to beauty does not necessarily mean that in every usage of the term Scotus has some aesthetic consideration in mind. In most cases he does not. The same concern applies to the term “consonance” itself: purveyors of the aesthetic interpretation simply assume that Scotus’s use of the term has aesthetic undertones, and read that back into the passage about the Second Table, thereby falsely assuming Scotus’s God chooses these precepts on the basis of their beautiful harmony with natural law in the strict sense. It is true that consonans has a semantic range that includes the harmony (of sounds). But I cannot find a single passage in Scotus’s writings that uses consonance in this way. Rather, Scotus uses the term to signify some type of “agreement” between two things with absolutely no aesthetic connotations whatsoever. Consider some examples from Scotus’s Questions on the Metaphysics and the Ordinatio, where Scotus uses consonance or its various cognates for a 23 Quodlibet 18, a.1, n. 9. See also n. 39. 24 For example, see Scotus, Quaestiones super Praedicamenta Aristotelis qq. 30, 35, 38, 43, and 44; Quaestiones super Libros Metaphysicorum Aristotelis V, q. 4 and q. 12; and Quaestiones in Libros Perihermenias Aristotelis q. 4. 25 Scotus, Quaestiones in Librum Porphyrii Isagoge. See also questions 21, 23, 25, and 35. And lest one object that these usages of convenientia are only from his early writings, he seems to speak the same way in later, more mature writings, such as Quaestiones super Libros Metaphysicorum Aristotelis, X.15.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

86

Jeff Steele

general “agreement” between two things, without any aesthetic undertones. In the Questions on the Metaphysics VIII, Scotus says: If therefore it is asked why there is a per se unity from matter and form, I respond according to the Philosopher in the text: that because ‘one is in actuality, the other in potentiality, then [the question] is without difficulty.’ This is explained as follows: because that same thing was first in potentiality on account of its matter, and now it is in actuality through its form. ‘For by the extraction of the potency in the act,’ according to the Commentator here, ‘it is not a growth in multitude but perfection.’ And it seems that the exposition of the commentary agrees (consonare) with this text.26

In this passage, the importation of aesthetics is nowhere implied. The term simply refers to an agreement between two ideas. The same point could be made from many passages, but consider two from the Prologue to the Ordinatio: One ought to deny the assumption that being is naturally known to be the first object of our intellect, and this according to the total indifference [of being] to the sensible and insensible, and this is what Avicenna says is naturally known. For he has mixed his own sect—which was the sect of Mohamed—with philosophical matters, and certain things he says are proven by philosophy and reason, and other things consonant (consona) with his own sect.27 The reply to the second argument, concerning the agreement (concordia) of the Scriptures, is evident as follows: in those things not evident from their terms, nor having principles thus evident from their terms, many diversely disposed people do not firmly and infallibly agree (consonant), unless they are inclined to assent from a cause superior to their intellect. But the writers of the sacred Canon, variously disposed and existing in diverse times, were in agreement (consonabant) entirely in such inevident things. Augustine treats this topic at great length in City of God, XVIII ch. 42: “Our authors should have been few, lest on account of their multitude they become worthless, and yet they are not so few that their agreement (consensio) is not to be wondered at: for neither in the multitude of philosophers will one easily find agreement (conveniant) between all the things they thought.” And this Augustine proves there in the examples.28

These passages show that Scotus often uses the term consonans and its various cognates to refer to an agreement between two things, without any aesthetic implications whatsoever. These examples could be repeated indefinitely.29 26 Quaestiones super Libros Metaphysicorum Aristotelis VIII, q. 4, n. 7: See also Quaestiones super Libros Metaphysicorum Aristotelis VII, q. 20, n. 36 and VIII, qq. 2–3, n. 49. 27 Ordinatio, Prol., part 1, q. un., n. 33. 28 Ordinatio, Prol., part 2, q. un., n. 102. 29 For example, there are plenty of other passages in the Ordinatio where consonans lacks any aesthetic undertones. For a sampling of passages from just the second half of Book IV of the Ordinatio, see Ordinatio IV, d. 43, q. 2, n. 104; n. 141, Ordinatio IV, d. 44, part 1, q. un., n. 50; Ordinatio IV, d. 45, q. 1, nn. 17–18; Ordinatio IV, d. 46, q. 1, n.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Duns Scotus, the Natural Law

87

In the last passage cited, interestingly enough, Scotus seems to take consono as synonymous with concordia, consensio, and convenio—all of which mean “agreement” or “harmony,” but not in any way that necessarily implies any aesthetic connotations. One final and rather amusing passage concerns the Devil’s first two sins, the first of which Scotus associates with a kind of pride, and the second a disordered desire for happiness. About the latter, Scotus claims: “And so if it ought to be reduced to something, it seems to fit better (magis consonare) with the sin of lust: for although lust properly considered is in a carnal action, nevertheless, any immoderate delight in pleasure can be called lust . . . .”30 In all of these passages, aesthetics connotations remain entirely absent from Scotus’s usage of consonans, unless of course, we are willing to conclude that Scotus thought Avicenna’s harmony of philosophy with Islamic thought is a beautiful thing, or the Devil’s inordinate desire for friendship is an aesthetically pleasing companion to lust, and so on. Furthermore, in examples where Scotus appeals to the concept of harmony in an aesthetically relevant way, he does not use the term consonans, but harmonicus.31 If Scotus had any aesthetic understanding of consonans, he would presumably use the term in those cases where it would appropriately describe some beautiful situation. But he does not. So in aesthetically appropriate settings, Scotus fails to use consonans, and in a plurality of different contexts where aesthetics plays absolutely no role, Scotus uses consonans. The only appropriate conclusion to draw from this data is that Scotus does not take consonans to be an aesthetically loaded term, but mere agreement between two things. ii. Oleg Bychkov. Oleg Bychkov has also recently argued for God’s choosing the Second Table precepts on the basis of aesthetic considerations. After a tour de force in contemporary neurobiology and aesthetics, Bychkov appeals to various passages in Scotus that purport to show that God has aesthetic tastes—just like humans.32 His argument seems to be that if God has aesthetic tastes in certain aspects of the moral realm, then he probably relies on such tastes in establishing contingent moral norms. Let us consider these passages in turn.

34; Ordinatio IV, d. 47, q. 1, n. 33; Ordinatio IV, d. 49, part 1, q. 6, n. 319; Ordinatio IV, d. 49, part 2, q. un., n. 463. 30 Ordinatio II, d. 6, q. 2, n. 71. 31 See Lectura I, d. 17, part 1, q. un, n. 95 and Ordinatio I, d. 17, part 1, qq. 1–2, n. 152. 32 Oleg Bychkov, “In Harmony with Reason,” 53–4. Cf. Lectura I, d. 17, part 1, q. un, n. 57; Ordinatio 1, d. 17, n. 62.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

88

Jeff Steele

Bychkov claims that “Scotus compares moral goodness as a type of spiritual harmony to musical harmony.”33 In support, he cites the following passage: A sound is more from the percussion of the sounding body than from the order of the percussion; but the sound’s being agreeable to the sense of hearing is more from the order of the percussion than from the efficacy of the power that strikes [the object]. Indeed, the power that strikes [the object] could be more efficacious and yet [the sound it produces could be] less agreeable, indeed not agreeable at all to the sense of hearing, because the sound isn’t harmonious.34

In order to see the significance of this passage, its context must be kept in mind, namely, the requirements for a meritoriously good act. Already, one should see a problem: Bychkov began by claiming that Scotus compares moral goodness as a type of spiritual harmony with musical harmony. However, the passage does not concern moral goodness at all, but meritorious goodness. At any rate, according to Scotus, in order for an act to be meritoriously good, the act must also be generically and specifically/morally good. That is to say, the kind of act must be appropriate to the agent, his nature, his causal powers, and be performed with an appropriate object in mind, for one or many good ends, in the right manner, time, and place.35 But the act must also be elicited in conformity with the source of merit (conformiter principio merendi), which Scotus identifies as the theological virtue of charity.36 The debated question between Scotus and his contemporaries concerns how charity and the will together produce an act acceptable to God, without denigrating the will’s intrinsic freedom on the one hand, or making the virtue of charity superfluous to the meritorious act, on the other.37 Scotus responds to such worries by arguing that the will and the virtue of charity concur as two partial agents in eliciting an act of love, resulting in its meritoriousness. According to Scotus, each cause is perfect in itself as a cause, and so each cause is independent of the other. However, one of the “agents” in the act is more perfect and the principal agent, while the other is less perfect and less principal.38 Scotus argues that the principal mover or agent is the will, while the disposition to love is the 33 Bychkov, “In Harmony with Reason,” 54. 34 Ordinatio I, d. 17, part 1, qq. 1–2, n. 152. 35 Ordinatio II, d. 7, n. 31. 36 Ibid. 37 Scotus’s primary theological targets on this issue are Henry of Ghent and Godfrey of Fontaines. See Henrici De Gandavo, Quodlibet IV, q. 10, nn. 15–290; Godfrey of Fontaines, Quodlibet XI, q. 4. 38 Lectura I, d. 17, part 1, q. un, n. 71.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Duns Scotus, the Natural Law

89

second and less principal cause in the act, but nevertheless makes the act more perfect than it would have been with merely the will acting.39 So while the act’s cause is primarily due to the will and the agent’s freedom, the act is better (more perfect/complete) if it is also accompanied by the disposition of charity. But what exactly makes the act meritorious? Here, Scotus claims that while the act’s cause derives more from the will, the act’s being meritorious and accepted by God results more from the virtue of charity than the will.40 In other words, the causes of the agent’s act are reversed in importance: the act’s meritoriousness stems more from the act’s being accompanied by the virtue of charity and secondarily because of the act’s free elicitation.41 And this is where the example of sounds comes in: the comparison is meant to illustrate how, when two causes concur in an effect, one cause can be more important in one respect (e.g., striking a sounding body or a will) and the other cause (e.g., the ordered arrangement of the striking or charity) can be more important in another respect. This has nothing to do with aesthetics, and certainly nothing to do with comparing moral or spiritual harmony with musical harmony—it merely serves as an example about how various causes can work together at eliciting various effects, while differing in their importance. Furthermore, what is relevant to aesthetics is that, in the musical example, a relationship exists between the object (the sound) and the perceiver (the one who hears it). And the analogues to this would be the acceptable act and God, who accepts it. But unlike the auditory case, where the nature of the perceiver and of the sound determines whether the sound will be pleasant, in the moral case, it is God’s will that determines whether or not the act will be acceptable—and according to Scotus, God does so on the basis of the concurrent causes of free will and charity, of which the latter is more important in terms of divine acceptation. So God ordains that a reward is attached to a charitable act.42 Consequently, Scotus seems more concerned about God’s freedom than his aesthetic sensibilities. As such, this passage provides no support for God’s having aesthetic tastes, and even less support for God’s choosing Second Table precepts on the basis of them.

39 Lectura I, d. 17, part 1, q. un, nn. 76–80; 87. 40 Lectura I, d. 17, part 1, q. un, n. 94; Ordinatio I, d. 17, part 1, qq. 1–2, nn. 130, 146–147, 151–152. 41 See Lectura I, d. 17, part 1, q. un, n. 94; Ordinatio I, d. 17, part 1, qq. 1–2, nn. 151–152. 42 Ordinatio I, d. 17, part 1, qq. 1–2, n. 153.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

90

Jeff Steele

In order to show that God has aesthetic tastes, and makes decisions based upon them, Bychkov also cites the following passage from the Lectura, also about the virtue of charity: The habit [of charity] . . . is the decoration of the soul and makes the soul acceptable to God (n. 58); just as the wooing [of the beloved by her lover] . . . is accepted on account of the beauty of the lover, in the same way . . . since charity is the decoration of the soul which makes it beautiful, a [charitable] act is perceived and accepted on this account by God who loves the soul on account of the beauty of virtue (n. 59).43

This passage fails to accomplish the task Bychkov supposes it does. First, contrary to what Bychkov implies about Lectura I, d. 17, it is not at all clear that Scotus “suggests that the moral beauty of the human soul is the cause of its acceptance by God.” The idea that God loves the soul because of its beauty is part of an argument that Scotus clearly rejects.44 It is true that Scotus rejects the argument for reasons that have nothing to do with God’s loving the soul because of its beauty or lack thereof,45 but talk of beauty does not reappear at all in his discussion about divine acceptation of acts that result from charity. What is more, Scotus ends this discussion by noting that God’s acceptance of acts elicited from the virtue of charity is part of God’s ordained power (de potentia ordinata) and not his absolute power (de potentia absoluta). In terms of God’s absolute power, God could accept acts not elicited from charity—and thus, even assuming Bychkov’s point about charity being beautiful, God could accept acts without regard for their aesthetic value.46 Second, charity’s being beautiful really does not help the case for the aesthetic interpretation of consonans, because charity is about love of God, which is necessarily good; beauty/consonantia was supposed to ground the contingent commandments. So even if one were to grant that God had aesthetic tastes, and accepted a charitable act on the basis of such beauty grounded in the virtue of charity, this says nothing about the establishment of Second Table precepts. Furthermore, virtues are different sorts of entities from moral norms, and pointing out that God sees them as beautiful differs entirely from God’s establishment of moral norms on the basis of some unmentioned aesthetic consideration. It’s a stretch at best: suppose something’s being beautiful is a reason for God to love it. Unless we have some reason to think that the commandments of the Second Table are beautiful, 43 Lectura I, d. 17, part 1, q. un, nn. 57–59. Bychkov, “In Harmony with Reason,” 54. Quoted from Mary Beth Ingham, The Harmony of Goodness (Saint Bonaventure: Franciscan Institute Publications, 2012), 106–7. 44 Lectura I, d. 17, part 1, q. un, n. 60. 45 Lectura I, d. 17, part 1, q. un, n. 60. See nn. 46–55 for the context. 46 Lectura I, d. 17, part 1, q. un, nn. 102–103.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Duns Scotus, the Natural Law

91

that does not give us any basis for thinking that God has a reason (an aesthetic reason) to love them and thus will them. Where is the support for thinking that the commandments of the Second Table are beautiful? It cannot be found here, because charity is a virtue distinct from moral precepts, and directs us to God as our necessary good, and so is more associated to the First Table of the Decalogue than the Second. And as I have already shown, it cannot be found in the terms consonans or conveniens themselves, as these terms are used extensively by Scotus for mere agreement between two things—without any aesthetic connotation. Finally, Bychkov appeals to yet another passage in the Scotist corpus on the topic of charity: “For justice is a great beauty of the soul, namely which distinguishes the just from the sinner. . . . in this way, justice is a certain beauty and godlikeness. All likeness and beauty, however, is a reason for loving any thing.”47 Absent in the quoted material, however, is that Scotus was stating the opinion of Augustine in De Trinitate, and he does not confirm whether he endorses any aspect of this view.48 And later in the same question, Scotus explicitly states that beauty is irrelevant to the case at hand because it is not the reason for divine acceptance (non est ratio acceptandi).49 So even if God appreciates the beauty of charity in some aesthetic way (which Scotus nowhere suggests), that appreciation plays no role in divine acceptation and is therefore irrelevant. While Bychkov recognizes this, it is highly misleading when he then concludes that “there is clear textual evidence” that God has aesthetic tastes. Consequently, there is no textual evidence to treat God’s establishment of these moral norms on the basis of aesthetic considerations, and, as I will show a little later, plenty of evidence that doing so would contradict Scotus’s notion of divine freedom. Bychkov’s conclusion is rather telling, as it all but admits that no textual evidence exists to support his view that the Second Table precepts are governed by aesthetic considerations: Whatever the exact scenario may be, there is clear textual evidence that according to Scotus, God has aesthetic taste, just like us. And if he has aesthetic taste in perceiving moral beauty, there is a good chance that his legislative activity in the area of moral law will be similarly based, in order that his laws might allow for beautiful harmonious patterns in the soul to play out.50

Simply because God has aesthetic tastes in perceiving moral beauty—which I have shown is a tenuous claim to make—is no reason to suspect that such 47 48 49 50

Reportatio I-A, d. 17, part 1, q. 2, n. 67. Bychkov, “In Harmony with Reason,” 54. Augustine, De Trinitate VIII.10. Cf. Reportatio IA, d. 17, part 1, q. 2, n. 67. Reportatio I-A, d. 17, part 1, q. 2, n. 69. Bychkov, “In Harmony with Reason,” 54. Emphasis mine.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

92

Jeff Steele

perceptions of beauty govern or regulate the moral norms God wishes to establish. There may be a chance, as Bychkov says, but it is an unlikely one, since he has failed to provide any clear evidence for taking the commandments of the Second Table as beautiful, and he has failed to show God chooses them because of that beauty. But establishing all this is not the only problem with the aesthetic solution to the Consonance Problem; indeed, taking consonans as indicating aesthetic constraints on God’s creative legislation threatens Scotus’s radical conception of divine freedom.

b. The Conceptual Problem: Divine Freedom Given the radical notion of freedom Scotus wishes to ascribe to God, and the true lack of limitations he wishes to place on God’s absolute power (God can justly will anything that does not involve a contradiction), this view seems to limit God’s freedom of choice to only prescriptive options with some intrinsic aesthetic appeal. On this view, God could only prescribe laws for humans that have some aesthetically appealing feature to them. But does this not limit God’s sovereignty in precisely the way Scotus wishes to avoid? Does this not limit God’s overall options as far as what he can choose? In a footnote, Cross responds to similar worries raised by Terence Irwin in another context.51 Cross claims that “God has options in cases that each alternative choice can be perceived to be aesthetically pleasing; in other cases, he has no choice.” Thus, so long as God has multiple aesthetically pleasing options, he could retain the veneer of libertarian freedom. In cases where he does not, God must choose the aesthetically pleasing option. Yet this seems to limit Scotus’s God in precisely the way in which Scotus wishes not to limit God: namely, constraining God beyond mere logical possibility.52 Cross seems to realize this, but nevertheless concludes that “Scotus is more concerned with providing a non-empty account of the grounding of the secondary principles than he is with preserving radical voluntarism.”53 But is he? This is a particularly odd claim to make, considering that we have no evidence for taking consonans in the debated passage aesthetically, and plenty of evidence for Scotus’s desire to preserve his radical voluntarism: at every possible place in his discussions of divine freedom, Scotus constantly

51 See Irwin, The Development of Ethics: A Historical and Critical Study, Vol. I: From Socrates to the Reformation (New York: Oxford University Press, 2007), 694–5. 52 Ordinatio I, d. 44, q. un., nn. 7–8; Reportatio I-A, d. 44, q. 1, nn. 7–10. 53 Cross, “Natural Law, Moral Constructivism, and Duns Scotus’s Metaethics,” 192, n. 57.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Duns Scotus, the Natural Law

93

emphasizes the freedom of the divine will over everything else.54 In these texts and many others, Scotus desires to save God’s will not only from the Islamic necessitarianism found in Avicenna, but also from the restrictions that might be placed upon God’s nature by anything beyond mere logical possibility. For example, in Reportatio I-A, d. 43, Scotus says: And so I say that God can bring about not only those things which he has not brought about, but even the opposite of those things he has brought about, because it is necessary that each opposite is possible for God. . . . For nothing is impossible for God, unless it includes a contradiction.55

In light of this, I pose the following question to the aesthetic solution to the Consonance Problem: can God choose a non-aesthetically-pleasing option? If yes, then it is hard to see how adding aesthetics does anything to supply content to the Second Table precepts, since nothing determines or inclines God to act in such a manner. But if God cannot choose a non-aestheticallyappealing option, as Cross surmises, then aesthetic reasons end up collapsing into logical reasons, since logical reasons are the only restrictions Scotus wishes to place on the divine will. In other words, if God cannot bring about a non-aesthetically pleasing option, doing so would have to involve a contradiction—per Scotus’s own words concerning divine power and what God can do. But Scotus explicitly claims that the Second Table precepts are not logically necessary in the way the First Table precepts are, and their denials do not result in contradictions.56 Furthermore, Scotus explicitly argues that the divine will is not determined to anything but God’s own goodness—and I take it this would apply to aesthetic considerations too. When we consider Scotus’s claims about divine justice, he explicitly states that God is not bound to anything but his own goodness and those necessary moral truths, and that with respect to everything else, God could have rightly and justly willed otherwise than he did. For example, after defining justice in one sense along Anselmian lines as “rectitude of the will preserved for its own sake,”57 Scotus remarks that God not only has this rectitude of will, but that his will cannot even be slanted (inobliquabilem). Why? Because God’s will is the chief standard or norm itself.58 In order for there to be a deviation from some norm, there must first 54 See Ordinatio I, d. 8, part 2, q. un., n. 299; Ordinatio I, d. 44, q. un., nn. 7–8; Ordinatio IV, d. 46, q. 1, n. 25; Reportatio I-A, dd. 39–40, qq. 1–3, a. 3, nn. 38–44; Reportatio 1-A, d. 44, q. 1, nn. 7–10; Also see Ordinatio I, d. 39, qq. 1–5, n. 21 [Appendix A of Vatican Edition vol. VI, 425–6]. For an explanation of why this question appears in the appendix of the Ordinatio, see Steven Dumont, “The Origin of Scotus’s Theory of Synchronic Contingency,” The Modern Schoolman 72 (1995), 150, n. 7. 55 Reportatio I-A, d. 43, q. 2, n. 32. 56 Ordinatio III, d. 37, q. un., nn. 16–18. 57 Anselm, De Veritate 12. 58 Ordinatio IV, d. 46, q. 1, n. 10.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

94

Jeff Steele

be a distinction between that norm and the will which deviates from it. But this is impossible in the case of God, because his will provides the basis for moral norms themselves: what God wills is right, and hence no deviation by the divine will is possible.59 But if we consider justice in the Aristotelian sense as a virtue or habit which naturally inclines the will to rectitude, then the divine will does not have any rectitude that inclines it determinately to anything other than its own goodness. The divine will, Scotus claims, is related to every object other than itself only contingently, and so could be equally inclined to it and its opposite. Scotus concludes that God has only one justice—the justice that inclines him to render what is due to his own goodness.60 Scotus further elaborates that the divine will is not inclined determinately to anything in such a manner that it cannot be justly inclined to the opposite of that thing. In fact, God could justly will the opposite.61 In Ordinatio I, d. 44, Scotus applies his distinction between God’s absolute power and his ordained power to the moral realm. According to Scotus, by God’s absolute power, God can act in any manner not involving a contradiction.62 He goes on to claim: I say that he can do many other things ordinately, and that many things different than those that are done in conformity with those laws can be done ordinately, and there is no contradiction involved when the rightness of the law—according to which someone is said to act rightly and ordinately—is in the power of the agent himself. Therefore, just as he can act otherwise, so also he can establish a different right law, which, if established by God, would be right, because no law is right except insofar as it is established by the divine will’s acceptance.63

Here, Scotus unapologetically and emphatically suggests that not only could God establish a different set of laws than he actually did, but that if he did so, those laws would be right or correct. Divine acceptation establishes a law as right. But if Cross is correct, then Scotus’s God would be bound to something beyond his own goodness, namely, he would be required to prescribe laws that are aesthetically pleasing. Yet this would contradict what Scotus specifically says. So, God cannot be bound to choosing some aesthetically pleasing option. As I see it, the only way out of this conclusion would be to place beauty under the conception of infinite goodness, such that, in making judgments about beauty, God would not need to appeal to something independent of his goodness: beauty would have to be an absolute quality, 59 See also Ordinatio IV, d. 46, q. 1, n. 25. 60 Ordinatio IV, d. 46, q. 1, n. 29. 61 Ordinatio IV, d. 46, q. 1, n. 32. 62 Ordinatio I, d. 44, q. un., n. 7. 63 Ordinatio I, d. 44, q. un., n. 8.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Duns Scotus, the Natural Law

95

convertible with God’s infinite being and goodness. In other words, beauty would have to be transcendental, only formally distinct from being and goodness. Things get complicated rather quickly here, as Scotus does claim that beauty and (one type of) goodness are similar; however, Scotus distinguishes two types of goodness—primary goodness and secondary goodness—only one of which is convertible with being and transcendental. According to Scotus, primary or essential goodness is an intrinsic perfection or completeness that a thing has “in itself and in reference to itself;”64 that is, as an absolute quality of a thing, there is no need to reference something outside the thing’s being to explain its goodness. Since this type of goodness is the perfection of being itself,65 everything that has being will have this corresponding goodness. In other words, primary goodness is transcendental goodness, coextensive with being. Scotus holds that being is predicated univocally of everything of which it is predicated in quid—both God and creatures. Before being is divided into the ten categories, it is “quantified” under two modes, infinite and finite. Finite being is then divided into Aristotle’s ten categories which represent finite modes of being.66 Since primary goodness converts with being, Scotus conceives of two types of transcendental goodness: infinite goodness characterizing God himself, and finite goodness, which runs throughout Aristotle’s categories. Scotus’s terminology for this type of goodness varies throughout his writings: in different places he calls the goodness that is convertible with being “primary goodness,”67 “natural goodness,”68 and “essential goodness.”69 Scotus often contrasts primary goodness with “secondary goodness” or “accidental goodness,”70 which corresponds to the second and extrinsic sense of perfection. As a secondary perfection, secondary goodness accidentally perfects its subject—the entity could continue in its existence without being good in the secondary sense. Secondary goodness, as John Hare characterizes it, is in substances, whereas primary goodness characterizes them as substances.71 The virtue of courage, for example, would be a

64 Reportatio II, d. 34, n. 3. 65 See Quodlibet 18.1. For the Quodlibetal Questions, I have used the Spanish-Latin semi-critical edition by Felix Alluntis, Obras del Doctor Sutil Juan Duns Escoto: Cuestiones Cuodlibetales, Edicion Bilingue (Madrid: Biblioteca De Autores Christianos, 1968). 66 Ordinatio I, d. 8, part 1, q. 3, n. 113. See also Lectura I, d. 8, part 1, q. 3, n. 107. 67 Quodlibet 18, a. 1, n. 9 and Ordinatio IV, d. 49, q. 2, n. 24. 68 Ordinatio II, d. 7, n. 28 and Ordinatio II, d. 40, n. 7. 69 Quodlibet 18, a. 1, n. 9. 70 See Ordinatio I, d. 17, nn. 62–64, Quodlibet 18, a. 1, n. 9, and Ordinatio IV, d. 49, q. 2, n. 24. 71 John E. Hare, “The Supervenience of Goodness on Being,” in K. Timpe (ed.), Metaphysics and God: Essays in Honor of Eleonore Stump (New York: Routledge, 2009), 152.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

96

Jeff Steele

secondary perfection, since a human could exist—as a human—without having the virtue. Thus a secondary perfection qualifies a subject in an accidental way: it is a real perfection, but the kind of perfection extrinsic to the nature of the thing, whereby the subject could exist without the secondary perfection. So if the aesthetic interpretation wishes to avoid the problem of binding God to something beyond his infinite goodness, beauty would have to be a non-absolute quality which converts with being, and thus, in its infinite mode, is a formal property of God’s infinite being. Only then could one claim that God would not violate the restriction that he is only bound to his own goodness when making an appeal to aesthetics for choosing moral norms. Unfortunately, Scotus draws a parallel between beauty and secondary goodness, not primary goodness: both secondary goodness and beauty are non-absolute, secondary perfections that require a harmony of a number of external factors in order to obtain.72 Thus, as a non-absolute quality beauty is not convertible with being, and thus not convertible with goodness— infinite or otherwise. And since God is only bound to his own infinite goodness, his will cannot be determined by aesthetic considerations. Finally, recall that the aesthetic reading of consonans claimed that the contingent, Second Table laws were consonans with what belongs to the natural law in the strict sense—the First Table of the Decalogue. So their consonantia (whatever that turns out to be) is a relation to the commandments of the First Table. Beauty, however, does not seem to be a relationship one set of norms has in relation to another. Rather, it is a relationship to the one perceiving it, namely, the person with aesthetic sensibilities. So even if the commandments of the Second Table are beautiful (which no one has established), and even if God chooses them because of that beauty (which no one has established), that beauty cannot be what Scotus means when he talks about their being valde consonans with the natural law in the strict sense. And given the further restrictions this would place on God’s creative activity, the aesthetic solution to the Consonance Problem remains highly improbable.

4 . C O N C LU S I O N Consequently, the aesthetic solution to the Consonance Problem reduces to a matter of fetishizing a particular group of words. But as I have shown, consonans and conveniens are not even exclusively or primarily about 72 Ordinatio I, d. 17, part 1, q. 2, n. 62.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Duns Scotus, the Natural Law

97

aesthetics at all, and so their appeal as an explanation for God’s choosing certain ethical norms on their supposed aesthetic basis remains quite tenuous. Scotus uses consonans as mere “agreement” between two things without any aesthetic connotations. It is still an open question, however, as to how and why these precepts agree with the First Table, but do not follow from it necessarily. What does Scotus mean by claiming that the contingent aspects of the moral law can be changed by God but somehow agree with the necessary moral truth that God must be loved? I have no clue. But in light of the dearth of evidence that consonance should be interpreted aesthetically, and in light of the abundance of passages defending Scotus’s radical view of divine freedom, whatever Scotus means by “valde consonans,” it has nothing to do with aesthetics. Perhaps the solution lies within Scotus’s radical view of the divine will. For as Scotus says: And if you ask why the divine will is more determined to one pair of contradictories than to another, I respond: “the untutored seek causes and demonstration of everything,” according to the Philosopher in Metaphysics IV, “for the principle of demonstration is not demonstrable” . . . and thus there is no cause why the will willed, except because the will is a will, just as there is no cause why heat heats, except that heat heats, because there is no prior cause.73

University of South Florida

BIBLIOGRAPHY Anselm of Canterbury. De Veritate. Sancti Anselmi Cantuariensis Archiepiscopi Opera Omnia, ed. F. S. Schmitt (Stuttgart: F. Frommann, 1984). Augustine. De Trinitate. CCSL 50, eds W. J. Mountain and F. Glorie (Turnhout: Brepols, 1968). Bychkov, Oleg. “ ‘Aesthetic’ Epistemology: Parallels between the Perception of Musical Harmony and the Cognition of Truth in Duns Scotus,” in L. Honneflder, et al. (eds), John Duns Scotus 1308–2008: Investigations into his Philosophy. Archi Verbi, Subsidia 5 (Münster: Aschendorff, 2010), 345–56. Bychkov, Oleg. Aesthetic Revelation: Reading Ancient and Medieval Texts after Hans Urs von Balthasar (Washington DC: Catholic University of America Press, 2010). Bychkov, Oleg. “ ‘In Harmony with Reason’: John Duns Scotus’s Theo-aesth/ ethics,” Open Theology 1 (2014), 45–55.

73 Ordinatio I, d. 8, part 2, q. un., n. 299. All translations are mine unless noted otherwise. A special thanks to Thomas Williams, Matt Dee, Kimberly Heil, and Robbie Hirsch for helpful comments on an earlier draft of this article.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

98

Jeff Steele

Cross, Richard. “Natural Law, Moral Constructivism, and Duns Scotus’s Metaethics: The Centrality of Aesthetic Explanation,” in J. A. Jacobs (ed.), Reason, Religion, and Natural Law: From Plato to Spinoza (Oxford: Oxford University Press, 2012), 175–97. Dumont, Steven. “The Origin of Scotus’s Theory of Synchronic Contingency,” The Modern Schoolman 72 (1995), 149–67. Godfrey of Fontaines. Les Quodlibets onze-quatorze de Godefroid De Fontaines. Les Philosophes Belges, Tome V, Fascicules I–II, ed. J. Hoffmans (Louvain: Editiones de L’Institut Superieur de Philosophie, 1932). Hare, John E. God and Morality: A Philosophical History (Malden: Blackwell, 2009). Hare, John E. “The Supervenience of Goodness on Being,” in K. Timpe (ed.), Metaphysics and God: Essays in Honor of Eleonore Stump (New York: Routledge, 2009), 143–56. Henry of Ghent. Henrici De Gandavo Opera Omnia VIII, eds G. A. Wilson and G. J. Etzkorn (Leuven: Leuven University Press, 2011). Ingham, Mary Beth. “Duns Scotus, Morality, and Happiness: A Reply to Williams,” American Catholic Philosophical Quarterly 74 (2000), 173–95. Ingham, Mary Beth. “Letting Scotus Speak for Himself,” Medieval Philosophy and Theology 10 (2001), 172–216. Ingham, Mary Beth. The Harmony of Goodness (Saint Bonaventure: Franciscan Institute Publications, 2012). Irwin, Terence. The Development of Ethics: A Historical and Critical Study, vol. I: From Socrates to the Reformation (New York: Oxford University Press, 2007). John Duns Scotus. Reportatio. Joannis Duns Scoti Opera Omnia (Paris: Vivès, 1891–95). John Duns Scotus. Lectura. Iohannis Duns Scoti Doctoris Subtilis et Mariani Opera Omnia, eds C. Balic, et al. (Rome: Typis Polyglottis Vaticanae, 1950–2004). John Duns Scotus. Ordinatio. Iohannis Duns Scoti Doctoris Subtilis et Mariani Opera Omnia, eds C. Balic, et al. (Rome: Typis Polyglottis Vaticanae, 1950–2013). John Duns Scotus. Obras del Doctor Sutil Juan Duns Escoto: Cuestiones Cuodlibetales. Biblioteca de Autores Cristianos 277, ed. F. Alluntis (Madrid: La Editorial Católica, 1968). John Duns Scotus. Quaestiones super Libros Metaphysicorum Aristotelis, Opera Philosophica 3–4, eds G. Etzkorn, et al. (St. Bonaventure, NY: The Franciscan Institute, 1997). John Duns Scotus. Quaestiones in Librum Porphyrii Isagoge, Opera Philosophica 1, eds R. Andrews, et al. (St. Bonaventure, NY: The Franciscan Institute, 1999). John Duns Scotus. Quaestiones super Praedicamenta Aristotelis, Opera Philosophica 1, eds R. Andrews, et al. (St. Bonaventure, NY: The Franciscan Institute, 1999). John Duns Scotus. The Examined Report of the Paris Lecture: Reportatio I-A, vol. 1, tr. and ed. A. B. Wolter and O. V. Bychkov (St. Bonaventure, NY: Franciscan Institute, 2004). John Duns Scotus. Quaestiones in Libros Perihermenias Aristotelis, Opera Philosophica 2, eds R. Andrews, et al. (St. Bonaventure, NY: Franciscan Institute, 2004). Thomas Aquinas. Summa Theologiae (Turin: Marietti, 1950).

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Duns Scotus, the Natural Law

99

Williams, Thomas. “How Scotus Separates Morality from Happiness,” American Catholic Philosophical Quarterly 69 (1995), 425–45. Williams, Thomas. “Reason, Morality, and Voluntarism in Duns Scotus: A Pseudo-Problem Dissolved,” The Modern Schoolman 74 (1997), 73–94. Williams, Thomas. “The Libertarian Foundations of Scotus’s Moral Philosophy,” The Thomist 62 (1998), 193–215. Williams, Thomas. “A Most Methodical Lover? On Scotus’s Arbitrary Creator,” Journal of the History of Philosophy 38 (2000), 169–202. Wolter, Allan B. Duns Scotus on the Will and Morality (Washington, DC: The Catholic University of America Press, 1986).

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Al-Taftāzānī on the Liar Paradox David Sanson and Ahmed Alwishah 1. INTRODUCTION Saʿad al-Dīn Masʿūd ibn ʿUmar ibn ʿAbd Allāh al-Taftāzānī (1322–1390) was a Persian polymath who wrote on a wide range of subjects— grammar, logic, theology (kalām), jurisprudence (fiqh), and literature. Here we translate and discuss his remarks on the Liar Paradox, in which he (a) presents the first example of a Liar Cycle or Deferred Liar in the tradition, (b) gives the paradox a puzzling name—the fallacy of the “irrational root” (al-jadhr al-as.amm)—which became the standard name for the paradox in the tradition, and (c) suggests a connection between the paradox and what it tells us about the nature of truth and falsehood, and related puzzles concerning reason and the nature of goodness and badness. Al-Taftāzānī’s discussion occupies a unique position in the tradition. The century prior had yielded a flurry of attempts to solve the paradox, centered around the Maragha School, with the work of Athīr al-Dīn al-Abharī (1200–1265), Nas.īr al-Dīn al-T.ūsī (1201–1274), Kātibī al-Quzwīnī (d. 1276), Saʿad Ibn Mans.u.̄r Ibn Kammūna (d. 1284), and Shams al-Dīn Muḥammad al-Samarqandī (1240–1304).1 The following century would yield the first treatises dedicated to the problem, centered around the Shīrāz

1 For a brief overview of the place of the Maragha School in the history of Arabic logic, see Tony Street, “Arabic and Islamic Philosophy of Language and Logic,” in E. N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Winter 2013 Edition) . For extended discussion of the solutions offered by al-Abharī and al-T.ūsī, and references to al-Quzwīnī, Ibn Kammūna, and al-Samarqandī, see Ahmed Alwishah and David Sanson, “The Early Arabic Liar: The Liar Paradox in the Islamic World from the Mid-Ninth to the MidThirteenth Centuries CE,” Vivarium, 47 (2009), 97–127. Also see Reza Pourjavady and

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Al-Taftāzānī on the Liar Paradox

101

school, with the work of S.adr al-Dīn al-Dashtakī (d. 1497) and Jalāl al-Dīn al-Dawwānī (1426–1502).2 These later authors seem to give al-Taftāzānī pride of place. They begin with him, and follow him in calling it the fallacy of the irrational root. But they also seem dismissive. Al-Dawwānī says: Since he pointed to its weakness, we leave [his proposed solution] aside to discuss the others, and we should not explore what is in it.3

And al-Dashtakī says much the same: Since the weakness in [his] solution is apparent to ‘the people of understanding’4 and [he] admits that he has departed from the right path, for the sake of brevity, we leave aside the details.5

Even so, it seems that his introduction of Liar Cycles was influential, as they play an important role in shaping al-Dawwānī’s own views about how to solve the paradox.

2 . T H E CO N T E X T The earliest discussions of the Liar in the tradition situate the problem as a putative counterexample to the definition of “declarative sentence” (khabar). So, for example, al-Baghdādī argues that: There is no declarative sentence that is both true and false together, except one: namely, the declaration by he who has not lied at all, about himself, that he is a liar, and this declarative sentence, from him, is false. And a liar who declares that he is a liar says the truth. And therefore this one declarative sentence is true and false, and it has one subject.6

Sabine Schmidtke, A Jewish Philosopher of Baghdad: ʿIzz al-Dawla Ibn Kammūna (d. 683/ 1284) and His Writings (Leiden: Brill, 2006), 41–4. 2 We are indebted to Ahad Qarāmalekī for editing and collecting most of the texts in the tradition that discuss the Liar, most of which have been collected together in his 12 Treatises on Liar Paradox in Shirāz School (Tehran: Iranian Institute of Philosophy, 2007). 3 Al-Dawwānī, al-Muntakhab min h.āshiyatihi ʿalā al-sharh. al-jadīd lil-tajrīd, in Qarāmalekī, 12 Treatises, 67. 4 “The people of understanding” (li-ūlā al-albāb) is a quote from Qurʾān verse 2, 179. 5 Al-Dashtakī, al-Muntakhab min h.āshiyatihi ʿalā al-sharh. al-jadīd lil-tajrīd, in Qarāmalekī, 12 Treatises, 5. 6 Al-Baghdādī, Us.ūl al-dīn 13, 217. For further discussion of this passage from al-Baghdādī, see Alwishah and Sanson, “Early Arabic Liar,” 101.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

102

David Sanson and Ahmed Alwishah

Within the Maragha School, and later, the Shīrāz School, the Liar is instead seen as a “difficult fallacy” (mughālat. a .saʿba) which raises questions about the nature of truth and falsehood, and the relationship between a sentence and what it is about. But al-Taftāzānī’s discussion of the Liar occurs in a very different context. It appears as an aside in the midst of a broader anti-rationalist polemic against the Muʿtazila and in favor of a broadly Ashāʿira theology.7 The goal of the polemic is to establish that goodness and badness must be shar ʿiyān— that is, grounded in divine command and so knowable only by revelation. The discussion occurs in the chapter, “No Judgment for the Intellect On Goodness and Badness,” in al-Taftāzānī’s Sharh. al-maqās.id (Commentary on the Aims of [Kalām]). That chapter begins by setting up the terms of this broader dispute: Goodness and badness are in our view shar iʿ yān, but for the Muʿtazila are rational ( aʿ qlī).8

To say that goodness and badness are shar iʿ yān is to say that they are grounded in divine command, and so only knowable by divine revelation. To say that goodness and badness are “rational” is to say that they are knowable by reason, because they are grounded in the natures of the acts, or the acts together with their consequences, or, as al-Taftāzānī often says, because they are “intrinsic” (dātī) to the acts and their consequences.9 The aim of the chapter, then, is to establish that goodness and badness are not intrinsic in this sense. To this end, al-Taftāzānī presents several arguments, each meant to show that we fall into incoherence if we try to use reason to judge whether or not something is good or bad. The Sharh. al-maqās.id is a commentary on Fakhr al-Dīn al-Rāzī’s al-Mat. ālib al-ʿāliyya min al- ʿilm al-ilāhī (The Higher Inquiry of the Science of Metaphysics), and many of al-Taftāzānī’s arguments derive from al-Rāzī.10 For example, they both argue that, as judged by reason alone, no human act is good or bad, since all human acts are determined by prior causes. But, they say, this is an absurd consequence, and it follows from the assumption that we can use reason alone to judge the goodness and badness of human acts, and so that assumption should be rejected.11 7 For an overview of the debate between Muʿtazila and Ashāʿira on the problem of good and evil see George Hourani, Reason and Tradition in Islamic Ethics (Cambridge: Cambridge University Press, 1985). Qarāmalekī briefly discusses the connection between these debates and al-Taftāzānī’s treatment of the Liar in his introduction to 12 Treatises. 8 Al-Taftāzānī, Sharh. al-maqās.id, 282. 9 See also al-Taftāzānī, Tahdhīb al-mant. q wa al-kalām, 92. 10 Al-Rāzī, al-Mat. ālib, 317–58. 11 Al-Taftāzānī, Sharh. al-maqās.id, 288. See al-Rāzī, al-Mat. ālib, 323–4, 328–9.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Al-Taftāzānī on the Liar Paradox

103

Here is another of al-Taftāzānī’s arguments concerning the goodness and badness of lying: If the badness of lying is intrinsic, then by necessity there is no case in which this would be undermined. But this is not true, for what if you have to lie to save the prophet from death? Then you must, and thus the lie becomes good.12

The same argument was pressed by al-Rāzī, and, before him, by al-Ghazālī.13 The position that is being attacked—that lying is intrinsically and absolutely bad, and so wrong in all cases—was defended by one of the most prominent Muʿtazila, ʿAbd al-Jabbār al-Qād.ī (d. 1024).14 Lying to save the prophet is meant to be a counterexample to this claim. One might

12 Al-Taftāzānī, Sharh. al-maqās.id, 285. 13 See al-Rāzī, al-Mat. ālib, 336–7. Al-Ghazālī discusses lying to save a prophet or saint in at least two places. The first is at Mi ʿyār al- ʿilm: Another example is when one judges truthfulness to be good because it is found to agree with one’s ends in view, desirable in most cases, but forgets that it is bad on the part of someone who is asked to reveal the place of a prophet or a holy man sought by a questioner in order to kill him. One (on the other hand) may (in such an instance) believe the falsehood involved in concealing the place of the prophet to be a bad thing because one has found badness and falsehood to be associated in most cases. (Translation taken from Michael Marmura, “Ghazālī on Ethical Premises,” Philosophical Forum 1 (1969), 401.) He later discusses lying to save a saint, at al-Qust. ās al-mustaqīm, 60: For someone says: “Every lie is bad in itself.” Then we say: “If one saw one of the saints who has hidden himself from an aggressor, then the aggressor asks him about place of the saint, and he conceals the truth [by saying I do not know], then is what he said a lie?” He says: “Yes.” We say: “Is it, then, bad?” He says “No, rather what is bad is telling the truth which would result in the death of [the saint].” We say to him: “Look to the balance.” And then we say: “His statement—in which he concealed the place of [the saint]—is false, and this is a known principle, and this statement is not bad, and this is the second known principle. Therefore not every lie is bad.” Now reflect: “Is doubt about this conclusion conceivable, after the admission of the two principles?” (This is our translation, but we benefited from Richard McCarthy’s translation, in his al-Ghazālī, Just Balance (Boston: Twayne Publishers, 1980), 301.) For discussion of the broader context of these passages, see Hourani, Reason and Tradition, 155. For an overview of al-Rāzī and his relation to al-Ghazālī, see Ayman Shihadeh, “From al-Ghazālī to al-Rāzī: 6th/12th Century Developments in Muslim Philosophical Theology,” Arabic Sciences and Philosophy 15 (2005), 141–79. 14 Al-Jabbār al-Qād.ī devotes volume IV of al-Mughnī to debates on the problem of goodness and badness. For the claim that al-Jabbār’s view is that lying is absolutely wrong and bad in all cases, see Hourani, Reason and Tradition, 112–13.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

104

David Sanson and Ahmed Alwishah

take the counterexample to show that reason does not judge lying to be absolutely bad, but al-Taftāzānī takes it to show that reason falls into error when it attempts to judge goodness and badness on its own. This suggests that he thinks that reason alone does judge that lying is absolutely bad, but also judges that lying to save the prophet is good, and so leads us to contradiction. Al-Taftāzānī considers a possible response to the counterexample: One might object that in this case lying is still bad, but not saving the prophet is worse than lying, and thus one must do the lesser evil. Thus the requisite good is the saving of the prophet, not the lying.15

Here the idea seems to be that lying is always bad, but not always wrong, since the badness of lying can be outweighed by other goods. His response is brief: Given that the lying is the cause and the reason for the requisite saving, it is also required, and thus it becomes good.

Here the claim seems to be that, even if we distinguish these two acts—lying and saving the prophet—each is required in this case, and so both are good: the saving is required, and the lying is the “cause and reason”—i.e., the necessary means—for the saving, and so it is also required. So once again al-Taftāzānī draws the moral that attempting to judge goodness and badness by reason alone leads to incoherence.

3. THE TEXT We turn now to a third argument, again about the goodness or badness of lying. This is the argument that leads to a discussion of the Liar Paradox. This argument does not seem to be derived from al-Rāzī, and seems to be original to al-Taftāzānī. Since it has not been translated before, and we will be discussing it throughout the rest of the paper, we begin with a translation of the passage in full.16 [1.] If goodness (h.asan) and badness (qubh.) are intrinsic, then this entails the conjunction of two contradictories, as in the case of the one who says, “I will lie tomorrow.” For either he is telling the truth (s.idq), which must be both a good in virtue of his truthfulness and a bad in virtue of entailing lying tomorrow; or he is 15 Al-Taftāzānī, Sharh. al-maqās.id, 285. 16 Al-Taftāzānī, Sharh. al-maqās.id, 286–7. Note that the paragraph numbers are our own, and that al-Dawwānī, al-Dashtakī, and al-Khafrī only quote paragraphs [4]–[9] and do not discuss paragraphs [1]–[3].

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Al-Taftāzānī on the Liar Paradox

105

lying, which must be both a bad in virtue of the lying and a good in virtue of entailing refraining from lying tomorrow. And thus, it is possible to establish that there is a conjunction of two contradictories in the statement, “I will lie tomorrow.” [2.] For, in virtue of the lying it [i.e., lying tomorrow] is bad and in virtue of entailing the truth of the original sentence it is good. Or, it is [only] good, but then badness cannot be intrinsic to lying. Or, it is [only] bad, but then refraining from lying becomes good, while also entailing the falsity of the original sentence, and so is [also] bad. And the ground of this entailment is the restriction of the “tomorrow” saying in this case. And thus it can be either true or false, and either way, this entails the conjunction of goodness and badness in it. And the ground for all this is that goodness must be good, and that badness must be bad, and that each good or bad is intrinsic. [3.] And this fallacy can be set up in such a way that truth and falsehood are gathered in one sentence (kalām), and thus goodness and badness are gathered. For if we consider a proposition (qad.iyya) whose purport is to declare itself not to be true, then truth and falsehood are entailed in it, as when you say that the sentence (kalām) I now speak is not true, for its truth entails the non-being of its truth and vice versa. [4.] And this can be expressed in the form of “tomorrow” and “yesterday” sentences (kalām). For if one says that the sentence (kalām) I speak tomorrow is not true, or that nothing I speak tomorrow is extrinsically true, and then tomorrow his only statement is, “the sentence (kalām) I spoke yesterday is true,” then the truth of either the “tomorrow” sentence or the “yesterday” sentence entails the non-being of the truth in both cases, and vice versa. [5.] This is a fallacy that has perplexed the minds of the most intelligent and smartest people. For this reason I call it the fallacy of the irrational root (jadhr al-as.amm). I have reviewed many claims, and found nothing that quenched my thirst, and I have contemplated it many times, but nothing came to me except the littlest bit of a little bit, namely: [6.] In the same way that truth and falsehood are conditions for judgment (h.āllan li-l-h.ukm)—i.e., affirmative and negative properties, required of all propositions— they can be a judgment (h.ukman)—i.e., what is judged [in the sense that it] is predicated of something indirectly, as when we say, this is true and that is false.17 And they are not contradictory to each other unless we consider them as two conditions for one judgment, or two judgments upon one subject, as opposed to if we consider one of them as a condition for judgment and the other as a judgment. For what differentiates the subject (al-marja ʿ) is either some explicit difference, as in our saying, “ ‘the sky is beneath us’ is true (or false),” or something hidden, as in the specific proposition that belongs to this fallacy. [7.] So, if we suppose it [i.e., the proposition that belongs to this fallacy, “the sentence I now speak is not true”] to be false, then nothing beyond the truth of its contradiction is entailed, that is, our saying of this sentence that it is true, so then

17 “Judgment” (h.ukman) is the nominalization of the verb “to judge” (h.ukm), and it is ambiguous. It can refer to a judgment as a whole (e.g., “the sky is beneath us”) or it can refer to the predicate judged (e.g., “beneath us”).

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

106

David Sanson and Ahmed Alwishah

truth occurs as a judgment on this specific [proposition], not as a condition for its judgment; rather the condition for its judgment is falsehood, as we are supposing, and truth is the affirmative property which is a condition for the judgment of the opposite [proposition], and a judgment for the original specific [proposition], and thus it is impossible to gather two conditions for one judgment or two judgments upon one subject. [8.] And likewise, if we assume it [i.e., the proposition that belongs to this fallacy, “the sentence I now speak is not true”] to be true, then one may deny the contradiction of the truth and falsity which necessitate each other by referring one of them to the judgment of the specified [proposition] and the other to the judgment of its subject. [9.] However, the correct judgment regarding this proposition is to give up on a solution and admit the incapability of [solving] this paradox.

4 . “ T H E C O N J U N C T I O N O F G OO D N E S S A N D B AD N E S S ” We begin by reconstructing the argument, in paragraphs [1] and [2], for the conclusion that goodness and badness are not intrinsic. The central claim of the argument is that if I were to say today, “I will lie tomorrow,” then my doing so would be both good and bad, which is impossible; and so once again the attempt to judge goodness and badness by reason alone leads to incoherence. For the sake of concreteness, suppose today is t1 and tomorrow is t2, and that what I say tomorrow is “I am not a believer:” t1: I say, “I will lie tomorrow.” t2: I say, “I am not a believer.”

At t1, either I told the truth or I lied. (Following al-Taftāzānī’s lead, we here ignore any distinction between lying and speaking falsely.) So suppose I told the truth at t1, and so lied at t2. Reason tells us that, T. Truth-telling is always good,

and F. Lying is always bad.

So, at t1, I did something good, and at t2, I did something bad. It remains to be shown that what I did at t1 was bad. For this, we need a bridge principle, to get us from the badness of what I did at t2 to the badness of what I did at t1. Al-Taftāzānī says that what I did at t1 is “bad in virtue of entailing lying tomorrow.” This suggests the principle EBB. If doing A entails doing B and doing B is bad, then doing A is bad.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Al-Taftāzānī on the Liar Paradox

107

By applying (EBB), we can infer the badness of what I did at t1 from the badness of what I did at t2. So what I did at t1 was both good and bad. If we suppose instead that I lied at t1, then we can immediately infer from (F) that what I did at t1 is bad and, from (T), that what I did at t2 is good. Again, we need a bridge principle, this time to get us from the goodness of what I did at t2 to the goodness of what I did at t1. Again, al-Taftāzānī appeals to entailment, claiming that what I did at t1 “must be both a bad in virtue of the lying and a good in virtue of entailing refraining from lying tomorrow.” This suggests the principle EGG. If doing A entails doing B and doing B is good, then doing A is good.

Using this principle, we can again infer that what I did at t1 is both good and bad. So when I said “I will lie tomorrow,” I did something both good and bad, whether I was speaking truthfully or lying. (He also argues that what I do at t2 is both good and bad, for parallel reasons, but we will not discuss those arguments separately.) Finally, al-Taftāzānī assumes that C. Goodness and badness are contradictories,

and so concludes that our attempt to apply reason to matters of good and bad has led to incoherence. Al-Taftāzānī does not say anything in defense of (T), (F), (EBB), (EGG), or (C), and we will not make an attempt to defend these principles on his behalf, or assess the fairness of attributing such principles to the Muʿtazila. But we will note that the argument, as we have reconstructed it, fails, because neither (EBB) nor (EGG) applies. What I do at t1—saying “I will lie tomorrow”—does not entail that I lie tomorrow or that I do not lie tomorrow. For those entailments to hold, we either need to assume both that I say it and that it is true, or that I say it and that it is false. But the truth or falsehood of what I say at t1 is not part of what I do at t1. Perhaps al-Taftāzānī has slightly different entailment principles in mind, namely, EGBB. If the goodness of doing A entails doing B and doing B is bad, then doing A is bad. EBGG. If the badness of doing A entails doing B and doing B is good, then doing A is good.

These principles seem about as plausible as (EBB) and (EGG), but their primary virtue is that they actually apply to the cases. Although what I do at t1 does not entail that I lie at t2, the goodness of what I do at t1 does. When I said, “I will lie tomorrow,” what I did was good because what I said was

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

108

David Sanson and Ahmed Alwishah

true, so the goodness of what I did entails that I do something bad—tell a lie—at t2.

5. THE FALLACY IN ONE SENTENCE These entailment principles are messy, and it is not obvious that they are true. Would it not be better to find an example that combined the goodness and badness in a single act, at a single time? This is precisely what al-Taftāzānī proceeds to do. The example he gives is the Liar Paradox: If we consider a proposition (qad.iyya) whose purport is to declare itself not to be true, then truth and falsehood are entailed in it, as when you say that the sentence (kalām) I now speak is not true, for its truth entails the non-being of its truth and vice versa [3].

His presentation suggests the influence of al-T.ūsī, who was the first in the tradition to describe the paradox abstractly in this way, as a problem generated by self-reference: Thus, the following paradox can be generated: The first declarative sentence, which is a declaration (khabar) about itself, namely that it is false, is either false or true. If it is true, then it must be false, because it declares itself to be false. If it is false, then it must be true, because if it is said falsely, then it will become true, which is absurd.18

But al-Taftāzānī’s immediate goals are quite different from al-T.ūsī’s. At first, al-Taftāzānī does not seem especially interested in the Liar Paradox as a paradox about truth and falsehood, but in what follows from it about goodness and badness. Given (T) and (F), since what I have said is both true and false, the act of saying it was both good and bad. The point of introducing the Liar Paradox, for al-Taftāzānī, was to simplify the earlier argument, involving two separate sentences and dubious goodness and badness entailment principles. And note that, if al-Taftāzānī is serious about this, he needs to accept the consequence that the sentence is both true and false, and so cannot dismiss it as a fallacy, as al-T.ūsī does.

18 For discussion and translation of T.ūsī on the Liar, including translation of this passage, along with references to the original sources, see Alwishah and Sanson, “Early Arabic Liar,” 125–7.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Al-Taftāzānī on the Liar Paradox

109

6 . A T E M P O R A L LI AR C Y C L E Before he stops to consider the Liar Paradox more carefully, al-Taftāzānī first extends the analogy between this case and the earlier case, in which I say, “I will lie tomorrow,” by describing a Liar Cycle, that is, a set of two or more sentences that, taken separately, generate no paradox, but taken together generate a paradox: For if one says that the sentence (kalām) I speak tomorrow is not true, or that nothing I speak tomorrow is extrinsically true, and then tomorrow his only statement is, ‘the sentence (kalām) I spoke yesterday is true,’ then the truth of either the ‘tomorrow’ sentence or the ‘yesterday’ sentence entails the non-being of the truth in both cases, and vice versa [4].

Let us name the two sentences, s1: “What I say tomorrow is not true.” s2: “What I said yesterday is true.”

Suppose s1 is true. Then it follows that s2 is not true, since s2 is what I say tomorrow. But if s2 is not true, then s1 is not true, since s1 is what I said yesterday. Likewise, suppose s1 is not true. Then it follows that s2 is true, from which it follows that s1 is true. So the two sentences together constitute what is known as a Liar Cycle or Deferred Liar. Liar Cycles are interesting because they suggest that we cannot solve the Liar Paradox by a simple ban on direct self-reference. Al-T.ūsī’s solution involves just such a ban: If a declarative sentence is the same as that-about-which-it-declares, then it cannot be conceived to be true and false. For agreement cannot be conceived except as between two things, and we cannot conceive them as opposed. For if one thing is affirmed, then nothing will be denied, and if one thing is denied, then nothing can be conceived to be affirmed.19

Al-T.ūsī’s claim is that, when a sentence is its own subject—when it makes a claim about itself—it is not the sort of thing that can be true or false, because truth requires an agreement between two distinct things and falsehood requires disagreement between two distinct things. And so, he says: It is clear that this fallacy arises as the result of a judgment that applies truth and falsity to something to which they in no way apply, and to apply them in any way is the misuse of a predicate.20

19 Alwishah and Sanson, “Early Arabic Liar,” 127.

20 Ibid.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

David Sanson and Ahmed Alwishah

110

But neither sentence s1 nor sentence s2 declares something about itself. Each is distinct from the other, so they are two distinct things, and so, for all that al-T.ūsī has said, they are the sorts of things that can be true or false. Al-Taftāzānī does not mention al-T.ūsī, and he does not present his Liar Cycle as a counterexample to any proposed solution to the Liar, so we do not know whether this was something he saw but did not mention, or something he did not see. But Liar Cycles would prove important to the later tradition, as al-Dawwānī uses them both to argue against al-T.ūsī and to develop his own explanation of why such sentences are neither true nor false.21

7. “THE LITTLEST BIT OF A LITTLE BIT” Al-Taftāzānī now turns, in paragraph [5], to consider the Liar Paradox itself. He tells us that he has “found nothing that quenched [his] thirst,” and says that nothing has come to him, by way of a solution, but “the littlest bit of a little bit.” The idea he presents mirrors an idea from al-T.ūsī. Al-T.ūsī distinguishes two roles that a sentence can play. A sentence, G, can either be used as a sentence, as when I say, G. Grass is purple.

Or it can be the subject of a sentence, as when I say, A. “Grass is purple” is false.

Since (G) is false, there is a sense in which, in the sentence (A), truth and falsehood are gathered, but this does not involve any contradiction: the subject of (A) is false, but (A) itself is true. We only get a contradiction when truth and falsehood are gathered together in a single sentence in the same role. In similar fashion, al-Taftāzānī distinguishes two roles that truth and falsehood can play: In the same way that truth and falsehood are conditions for judgment (h.āllan li-l-h.ukm)—i.e., affirmative and negative properties, required of all propositions— they can be a judgment (h.ukman)—i.e., what is judged [in the sense that it] is predicated of something indirectly, as when we say, this is true and that is false [6].

21 Al-Dawwānī, Nihāyat al-kalām fī h.al shubhat kull kalāmī kādhib, 129.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Al-Taftāzānī on the Liar Paradox

111

So, consider his example, B. The sky is beneath us.

(B) is false. This is a property (B) has, but it is not part of the content of (B). Supposing (B) expresses my judgment, my judgment predicates a certain location to the sky; it does not predicate truth or falsehood to anything. So, in this case, falsehood is a condition for judgment, but it is not the judgment— that is, it is not what is judged of the sky. By contrast, in (C), falsehood is what is judged, C. “The sky is beneath us” is false.

But (C) is true. So, in this case, the condition for judgment is truth, but what is judged is falsehood. Note that, just as we saw with (A), there is a sense in which truth and falsehood are conjoined in (C), but without contradiction. Al-T.ūsī describes the case as one in which the sentence is true, but its subject is false. Al-Taftāzānī describes it as one in which what is judged is falsehood, but the condition for that judgment is truth. And, like al-T.ūsī, he points out that this is not enough for a contradiction: And they are not contradictory to each other unless we consider them as two conditions for one judgment, or two judgments upon one subject, as opposed to if we consider one of them as a condition for judgment and the other as a judgment [6].

How might one attempt to use this insight to solve the paradox? To see this, it helps to first sketch out the argument that generates the paradox. Suppose our sentence is, L. L is false.

(L) is either true or false. We generate the paradox by first supposing that it is true, and arguing from its truth to its falsehood, and then supposing that it is false, and arguing from its falsehood to its truth. There are various ways to present each argument, but here is a simple version of the argument from false to true: 1. 2. 3. 4.

“L is false” is false. For any S and P, if “S is P” is false, then S is not P. Therefore, L is not false. Therefore, L is true. With this sort of argument in mind, consider what al-Taftāzānī says:

If we suppose it [i.e., the proposition that belongs to this fallacy, “the sentence I now speak is not true”] to be false, then nothing beyond the truth of its contradiction is entailed, that is, our saying of this sentence that it is true, so then truth occurs as a

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

David Sanson and Ahmed Alwishah

112

judgment on this specific [proposition], not as a condition for its judgment; rather the condition for its judgment is falsehood, as we are supposing, and truth is the affirmative property which is a condition for the judgment of the opposite [proposition], and a judgment for the original specific [proposition], and thus it is impossible to gather two conditions for one judgment or two judgments upon one subject [7].

It is not clear how to interpret the text here, but perhaps al-Taftāzānī is telling us that in (1)—our assumption that L is false—falsehood is the condition for judgment but not the judgment, while in (4)—the conclusion that L is true—truth is the judgment but not the condition for judgment. And so we do not have a contradiction, because we have not managed to gather truth and falsehood together in the same sentence in the same role. If this were right, then, by carefully policing the distinction between truth and falsehood as conditions for judgment and truth and falsehood as judgments, one might hope to block both the argument from false to true and the argument from true to false, and so avoid the paradox.

8 . “ T HE W E AK N E S S IN H IS S OL U T IO N IS APPARENT” We have already seen that al-Dawwānī and al-Dashtakī think little of al-Taftāzānī’s proposal. Al-Taftāzānī’s seems to share their dim assessment, as he immediately goes on to say: However, the correct judgment regarding this proposition is to give up on a solution and admit the incapability of [solving] this paradox [9].

Only Shams al-Dīn Muh.ammad b. Ah.mad al-Khafrī (d. 1550), writing half a century after al-Dawwānī and al-Dashtakī, attempts to explain what is wrong with the solution: Since the weakness in [his] solution is apparent—it is necessary that the one and only subject of this proposition is itself, which entails the conjunction of truth and falsity in one proposition, which is impossible—al-Taftāzānī admits that he has departed from the right path.22

It is not immediately obvious how this objection applies, as al-Taftāzānī does not say that the Liar Sentence has multiple subjects. Consider again (1) and (4) from the argument from false to true: 1. “L is false” is false. 4. L is true. 22 Al-Khafrī, H.ayrat al-fud.alāʾ, in Qarāmalekī, 12 Treatises, 276.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Al-Taftāzānī on the Liar Paradox

113

Al-Taftāzānī insists that (1) expresses the fact that falsehood is a condition for judgment, while (4) judges L to be true. One might hope to read this distinction off of the structure of each sentence. A sentence like (1) is of the form, “A is B” is C,

and so ascribes a property to a judgment, and so seems apt for expressing a condition for judgment. But (4) is of the form, A is B,

which ascribes a property to an object, and so seems apt for expressing a judgment. But this distinction collapses once we remember that L is an abbreviation for “L is false,” so that (1) can just as well be written as, 1*. L is false,

and (4) can just as well be written as, 4*. “L is false” is true.

So it is hard to see how we are supposed to tell whether a given sentence expresses a condition for judgment or a judgment. But perhaps we can find the materials for resisting this collapse in al-Taftāzānī’s puzzling remarks about explicit and hidden differences: For what differentiates the subject (al-marja ʿ) is either some explicit difference, as in our saying, “ ‘the sky is beneath us’ is true (or false),” or something hidden, as in the specific proposition that belongs to this fallacy [6].

The passage is difficult, but perhaps al-Taftāzānī means to distinguish metalinguistic subject terms that make the subject “explicit” from metalinguistic subject terms that “hide” the subject. So, in the sentence, “ ‘the sky is beneath us’ is true,” the sentence that is the subject, “the sky is beneath us,” is explicitly stated. But in the sentence “what I said yesterday is true,” the sentence that is the subject—whatever sentence it was that I said yesterday—is not explicitly stated, and so remains “hidden.”23 This distinction nicely captures an important difference between (1) and (1*), and (4) and (4*): 1. “L is false” is false. 1*. L is false. 4. L is true. 4*. “L is false” is true. 23 Thanks to an anonymous referee for suggesting this reading of the passage.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

114

David Sanson and Ahmed Alwishah

In (1) and (4*), the subject is “explicit”; in (1*) and (4), it is “hidden.” Armed with this distinction, one might insist on a syntactic test for the distinction between judgments and conditions for judgment: a sentence with an explicit sentential subject term expresses a condition for judgment; a sentence with a subject term that hides the sentential subject expresses a judgment. But it is hard to see what to make of this claim when the underlying subject, whether made explicit or left hidden, is in fact the same sentence. When a sentence has an explicit sentential subject term, as in “ ‘the sky is beneath us’ is true,” there will often also be an explicit difference between the inner predicate—“is beneath us”—and the outer predicate, “is true.” But this is not always true. When we express “the specific proposition that belongs to this fallacy” with an explicit sentential subject term, the inner predicate and the outer predicate appear to be the same: 1. “L is false” is false.

What al-Taftāzānī needs, if he is going to block the argument that generates the paradox, is some way of positing a “hidden” difference between these two predicates. We might try to make such a difference explicit by marking the distinction between judgments and conditions for judgment with subscripts. Recall that the argument from false to true begins with (1) and ends with (4), which together appear to show that the same sentence, L, is both false and true: 1. “L is false” is false. 4. L is true.

And recall that al-Taftāzānī wants to insist that this is not a contradiction, because in (1), falsehood is a condition for judgment, while in (4), truth is what is judged. So, using subscripts to make his claim explicit, we can rewrite (1) and (4) as, 1**. “L is falsejudgment” is falsecondition for judgment. 4**. L is truejudgment.

The key premise of the argument from true to false, is premise (2): 2. For any S and P, if “S is P” is false, then S is not P.

If al-Taftāzānī is right, and the argument begins with (1**) and ends with (4**), then that premise should be disambiguated as, 2**. For any S and P, if “S is Pjudgment” is falsecondition Pjudgment.

for judgment,

then S is not

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Al-Taftāzānī on the Liar Paradox

115

From (1**) and (2**), we can infer, 3**. L is not falsejudgment,

and from (3**), given the assumption that truthjudgment and falsehoodjudgment are contradictories, we can infer (4**). But since truthjudgment and falsehoodcondition for judgment are not contradictories, this result is not paradoxical. The problem with this attempt to disarm the paradox is that (2**) is not the only plausible disambiguation of (2). In particular, (2***) looks equally plausible: 2***. For any S and P, if “S is Pjudgment” is falsecondition Pcondition for judgment.

for judgment,

then S is not

(2***) says that, if, as a matter of external condition-for-judgment fact, it is false to judge that S is P, then, as a matter of external condition-forjudgment fact, S does not have P. And this seems correct: if, for example, as a matter of external fact, it is false to judge that the sky is beneath us, then, as a matter of external fact, the sky is not beneath us. But once (2***) is granted, we have what we need to argue from falsehoodcondition for judgment to truthcondition for judgment. From (1**) and (2***), it follows that, 3***. L is not falsecondition for judgment.

and so, assuming that falsehoodcondition for judgment are contradictories, it follows that,

judgment

and truthcondition

for

4***. L is truecondition for judgment.

And so we have the same sentence, L, that is both truecondition for judgment and falsecondition for judgment, and al-Taftāzānī’s gambit fails. So if al-Taftāzānī’s’ gambit is to succeed, he needs to give us some reason to accept principles like (2**), but reject principles like (2***). But this is not something he does, and it is not clear what such a reason could be. 9. A R E T R U T H A N D F A LS E HO O D S H A R ʿI YĀ N ? Al-Taftāzānī’s goal in this chapter is to show that goodness and badness are not intrinsic, by showing that attempts to use reason alone to judge goodness and badness lead to fallacies and contradictions. That, in turn, is meant to support the view that goodness and badness are shar ʿiyān—that is, grounded in divine command and knowable only by revelation. So, is the Liar Paradox merely a digression, with no real relevance to these broader goals, or is it meant to be of a piece with them?

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

116

David Sanson and Ahmed Alwishah

On the one hand, al-Taftāzānī might think that truth and falsehood are knowable by reason, but goodness and badness are not. If so, the Liar is a digression, introduced as yet another way to show that reason leads us to incoherence by dictating that truth-telling is always good and lying is always bad. And if so, it is a problem for al-Taftāzānī that the Liar Paradox remains unsolved. Reason alone ought to be able to judge truth and falsehood without falling into incoherence, even if it cannot judge goodness and badness. On the other hand, al-Taftāzānī might think that the Liar Paradox shows that truth and falsehood, like goodness and badness, are also not knowable by reason alone. On this reading, the Liar shows that the problems raised for goodness and badness are also problems for truth and falsehood. On this reading, it is not a problem for al-Taftāzānī that the Liar Paradox remains unsolved, because that is the point: reason alone cannot judge truth and falsehood without falling into incoherence: “the correct judgment regarding this proposition is to give up on a solution and admit [our] incapability” of solving it. Note that this reading of al-Taftāzānī’s final remarks on the incapability of solving the paradox crucially depends on the broader context of the argument that occurs in paragraphs [1] through [3]. So here it might matter that al-Dawwānī, al-Dashtakī, and al-Khafrī focused only on paragraphs [4] through [9], and so read al-Taftāzānī’s remarks instead as an admission of failure. To put the question bluntly: we know that al-Taftāzānī is a divine voluntarist, and that he thinks that goodness and badness are shar ʿiyān. Might he go a step further, and embrace what we might call alethic voluntarism, the view that truth and falsehood are grounded in divine command, and so likewise shar ʿiyān? This seems like a radical view indeed, and we hesitate to attribute it to al-Taftāzānī. But recall that al-Taftāzānī’s Sharh. al-maqās.id is a commentary on al-Rāzī’s al-Mat. ālib al-ʿāliyya. As we have seen, al-Taftāzānī uses the Liar Paradox to show that the same speech act can be both good and bad, and so seems willing to argue that truth and falsehood are problematic, in order to show that goodness and badness are not intrinsic. And in a similar spirit, al-Rāzī attacks truth and falsehood—arguing that no sentence is ever true or false—in order to show that truth and falsehood cannot ground the goodness and badness of telling the truth or lying. Al-Rāzī begins with some definitions: Truth is taken to be the agreement between a declarative sentence (khabar) and that about which it declares, and falsity, the non-agreement between a declarative sentence and that about which it declares. And it is well known that truth and falsity

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Al-Taftāzānī on the Liar Paradox

117

are two species under the genus, declarative sentence (khabar). And declarative sentence [is a species] under the genus of sentence (kalām), and a sentence is taken to be a vocalization (lafz.) composed of a successive sequence of syllables.24

A vocalization (lafz.) is a produced sound, but it need not have any syntax or semantics. The screech of a monkey, for example, is a lafz.. A kalām is a lafz. composed of successive syllables. Here we translate “kalām” as “sentence,” but it could also be translated as “speech.” Note that by al-Rāzī’s definition, examples will include sub-sentential bits of speech, like an utterance of a kalima (word), like “lollipop,” or a phrase, like “the lollipop in the bag,” or a bit of nonsense, like “do doop dee do.” A khabar (declarative sentence) is a kalām (sentence) that is true or false: so, not the utterance of a word or a phrase, or a string of nonsense, or a command or question, but the utterance of a complete declarative sentence, like “I want the lollipop in the bag.” Al-Rāzī proceeds to argue that each declarative sentence exists successively, one syllable after the next, so there is never a time at which a complete sentence exists: And what exists of [a declarative sentence] is always nothing but a single syllable, and when [that syllable] is completed, the second syllable occurs, and so on, in this arrangement, until the last syllable of the word (kalima) occurs, and at this point the word (kalima) is complete. Based on this, the word (kalima) does not exist at all at any time or circumstance, and all that exists of it [at any time or circumstance] is a single syllable. And a single syllable is not a single sentence (kalām).25

Finally, he argues that, since complete sentences never exist, and truth and falsehood are properties of sentences, nothing is ever true or false. Furthermore, since sentences—that is, spoken sentences—are good and bad only if they are true or false, no sentence is ever good or bad: If this is established, then we say: a single syllable is not a declarative sentence (khabar), and it is neither true nor false, and it is impossible for it to be what necessitates goodness or badness. With respect to the totality of a word (kalima), it does not exist at all, and that which does not exist at all cannot be the cause of something’s being good or bad. By this proof, we establish that it is impossible for the sentence (kalām) to be good or bad, for it cannot be true or false.26

This is a bold and clever argument, to say the least. Whatever you think about its merits as an argument, note al-Rāzī’s target. He wants to convince you that no sentence is ever true or false because he wants to convince you that no sentence is ever intrinsically good or bad. So, as with al-Taftāzānī,

24 Al-Rāzī, al-Mat. ālib, part 3, 335.

25 Ibid.

26 Ibid., 335–6.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

118

David Sanson and Ahmed Alwishah

the detour into semantics is in service to the broader goal of showing that goodness and badness are not intrinsic. He says: The difference is that you [Muʿtazila] say that being true necessitates the attribute of goodness, and being false necessitates the attribute of badness. We say that the truth is the totality, and this totality does not exist at all, and it is impossible for what does not exist to be a necessitator of an actual fixed attribute, and that claim [i.e., the claim that truth is a necessitator of an actual fixed attribute] is in contrast to what you say.27

But we do still use the words “true” and “false.” What are we getting at when we do so, if not an intrinsic attribute of sentences? Instead, we say when we hear this sequence of syllables—which they agree make us aware of some meanings—without doubt we understand some of these meanings, and at this point we form a belief about something. If the sentence (kalām) is false, then it becomes clear to us that our act is wrong (bāt. il ). And there is no other meaning of what is to be false except this. And thus the difference [between us and the Muʿtazila] becomes evident.28

So it appears that al-Rāzī wants to reverse the order of explanation: to say that a sentence is false is to say that it produces a belief in us that is wrong; to say it is true is to say that it produces a belief in us that is right. But in what sense could a belief be right or wrong, if not in the sense that the belief is true or false? Remember that for al-Rāzī goodness and badness, and right and wrong, are shar ʿiyān—grounded in divine command, and knowable not by reason, but only by revelation. So, putting two and two together, this suggests that for al-Rāzī, what it is good or bad to believe is likewise shar ʿiyān, and that a belief, therefore, is true or false only insofar as it agrees or disagrees with divine command. The text suggests, but does not force this reading. Perhaps al-Rāzī thinks that a right belief is one that corresponds to the facts, and a wrong belief is one that does not. On this reading, his primary concern is to reject truth and falsehood as properties of sentences—bits of language—and, instead treat them as properties of beliefs. But note that he does not say that a false sentence is one that produces a false belief. What he says is that a false sentence is one that causes one to form a belief, where the act of forming that belief is wrong. And if he were to go on to say that the rightness and wrongness of belief is a matter of correspondence to the facts, then he would seem to be admitting that at least some rightness and wrongness is intrinsic rather than shar ʿiyān, and we doubt this is something he would want to admit.

27 Ibid., 336.

28 Ibid., 336, our emphasis.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Al-Taftāzānī on the Liar Paradox

119

10 . “ T H E F A L L AC Y O F T H E ‘ I R R A T I ON AL R O O T ’ ” We have nothing more to say about the substance al-Taftāzānī’s remarks on the Liar. In this last section of the paper, we turn to consider the puzzling name that he gives to the paradox. As we’ve mentioned, the later tradition follows al-Taftāzānī in calling the Liar Paradox the fallacy of the “irrational root.” It is not clear to us what lies behind this choice of name, or why the later tradition would have found it apt. We have translated “jadhr al-as.amm” as “irrational root.” It is a standard piece of mathematical terminology for a magnitude, like the square root of two, or the ratio of circumference to diameter, which cannot be expressed as a ratio of two numbers. But what does the Liar have in common with such magnitudes? For the Greeks, the only numbers are the whole numbers. What we now think of as rational numbers they think of as magnitudes expressible as ratios of numbers. What we now think of as irrational numbers, they think of as magnitudes not expressible as ratios of numbers, and they think of those magnitudes primarily in geometric terms. So, for example, the square root of 2 is the ratio of the length of the side of a unit square to the length of its diagonal, and π is the ratio of the circumference of a circle to its diameter. For them, the problem posed by irrational magnitudes in general, and irrational roots in particular, was this: some quantities, essential to geometry, could not be expressed as ratios of numbers.29 Perhaps al-Taftāzānī’s thought is this: when you begin doing geometry, and develop the concept of unit and then the concept of a ratio, it seems natural to suppose that all magnitudes resolve into a ratio of two numbers. But then you discover irrational roots, and realize that some magnitudes cannot be resolved in this way. So too, when you begin doing logic, and develop the concept of a declarative sentence, and the concept of truth and falsehood, it seems natural to suppose that all declarative sentences can be resolved into the true or the false. But then you discover the Liar, and realize that some sentences cannot be so resolved. Al-Taftāzānī’s own remarks provide little by way of explanation of what sort of analogy he had in mind: This is a fallacy that has perplexed the minds of the most intelligent and smartest people. For this reason I call it the fallacy of the irrational root ( jadhr al-as.amm) [5].

29 For an overview of the Greek concept of number in relation to irrational ratios, see Morris Kline, Mathematical Thought from Ancient to Modern Times (Oxford: Oxford University Press, 1990), 32ff.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

120

David Sanson and Ahmed Alwishah

ʿIs.mat Allāh b. Kamāl al-Dīn Mah.mmūd al-Bukhārī (16th c.), writing more than a century after al-Taftāzānī’s death, was equally puzzled. In his al-Muntakhab min sharh. risālat al-mughālat. āt (a Selection from Commentary on the Treatises of Fallacies), he offers up three possible explanations for the label, which trade on three different meanings of the word “as.amm.” The first meaning is the one from mathematics, and so his first explanation attempts to draw an analogy between the Liar Paradox and a corresponding problem about irrational roots in mathematics: This is based on the inability of the intelligent people to solve it, just as they are unable to know the irrational root. And the root is the origin of a thing, and as he said in the Muhadhdhab,30 “the root is the origin of arithmetic, and the irrational [root] of a number has no fraction, from the half to the tenth, as 11 and 13.” And it is reported in some of the books of arithmetic that no one knows the irrational root except God, and it is reported from some of the sages that, “[God] is to be praised, by saying, ‘Praise He who knows the irrational root and no one else.’ ”31

So the idea is that the solution to the Liar Paradox is beyond the grasp of human knowledge, just as irrational roots are beyond the grasp of human knowledge. The report from the sages appears to be a reference to the famous mathematician, Muh.ammad ibn Mūsā al-Khawārizmī (780–850), who says that we cannot apprehend the true nature of an irrational root, but God can: An irrational root is that for which there is no way of knowing its truth (h.aqīqatahu) through number, such as the root of two, root of three, or even root of ten, and it can be taken by an approximation, and its truth (h.aqīqatahu) cannot be apprehended. And it is mentioned that one of the Brāhima’s praises in India was, “Praise the one who knows the roots.”32

Al-Bukhārī does not attempt to push the mathematical analogies any further than this epistemic idea, that the Liar and irrational roots are both unknowable by human reason. Or, perhaps, that the truth (s.idq) of the Liar is 30 Qarāmalekī (12 Treatises, 315 n. 272) suggests that this is a reference to Sajzī Mah.mūd bin ʿUmar’s Muhadhdhab al-asmāʾ fi murattab al-ashyāʾ. 31 Al-Bukhārī, al-Muntakhab min sharh. risālat al-mughālat. āt, in Qarāmalekī, 12 Treatises, 315–16. 32 Al-Khawārizmī, Mafātīh. al- ʿulūm, 221. See also the marginal note in the Oxford manuscript of al-Khawārizmī’s Algebra, as reported and translated by Frederic Rosen: Nobody can ascertain the exact truth of this, and find the real circumference, except the Omniscient: for the line is not straight so that its exact length might be found. This is called an approximation, in the same manner as it is said of the square-roots of irrational numbers that they are an approximation, and not the exact truth: for God alone knows what the exact root is (The Algebra of Mohammed Ben Musa [London: The Oriental Translation Fund, 1831], 200).

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Al-Taftāzānī on the Liar Paradox

121

unknowable by human reason, just as the truth (h.aqīqa)—i.e., true essence or nature—of irrational roots is unknowable by human reason. Are there any deeper analogies here, which might help explain al-Taftāzānī’s choice of label? Perhaps, but here we can only speculate. One might try to press a structural analogy: just as rational roots involve a ratio between two numbers, you might think that truth and falsehood involve an agreement or disagreement between the judgment and the conditions for the judgment. And so just as irrational roots appear to be quantities that cannot be expressed as ratios between two numbers, the Liar Paradox might appear to give us a sentence that cannot stand in any such relation of agreement or disagreement. If al-Taftāzānī had this sort of analogy in mind, it might help explain his own dissatisfaction with his proposed solution. You cannot “solve” the problem posed by π by insisting that every time we attempt to construct a ratio of circumference to diameter, we end up constructing some other rational ratio instead. So too you cannot “solve” the Liar by insisting that every time we attempt to compare the judgment with the conditions for judgment, we instead end up making some other non-paradoxical comparison.33 Arabic mathematicians focused a great deal of energy on developing algorithms for calculating ever more precise rational approximations of irrational roots. Each approximation might seem to get us closer to the true nature of the irrational root, but, as a matter of fact, we never get any closer, since the true nature always recedes further away. Perhaps alTaftāzānī saw an analogy here: it is not just that both are unknowable, but both are elusive. With the Liar, just when you think you’ve got it pinned down, it slips between your fingers and remains unresolved.34 So much, then, for speculative mathematical analogies. In addition to this explanation, al-Bukhārī offers two alternative non-mathematical suggestions for why al-Taftāzānī might have chosen to call this the fallacy of the “jadhr al-as.amm.” First, as al-Bukhārī points out, the word “as.amm” can also mean “solid” or “concrete,” or even “backbone,” understood as the principle that holds a solid thing together. So “jadhr al-as.amm” could mean something like “solid root”: It is possible that the naming of it is in virtue of ‘root’ (‘jadhr’) being found in the language, in the meaning of ‘az bikh barkandan’ (‘to pull up from the roots’). And, as

33 Thanks to an anonymous referee for suggesting that we make explicit the possible relevance of this analogy to al-Taftāzānī’s solution, and his attitude toward the solution. 34 Thanks to Roshdi Rashed, who suggested a possibility along these lines in private correspondence.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

122

David Sanson and Ahmed Alwishah

was declared in al-S.arāh., ‘as.amm’ means backbone (al-s.ulb).35 Thus, given that this fallacy invalidates the [principle] of the impossibility of conjunction of two contradictories, it is as if it cuts the origin of the thing, i.e., the backbone (al-s.ulb), which is the impossibility of the conjunction of two contradictories.36

So here the idea seems to be that the Liar Paradox threatens the principle of non-contradiction, and so threatens the “solid root” or “backbone” of logic itself. Whatever one might think about this speculation as an explanation of al-Taftāzānī’s use of the label, “jadhr al-as.amm,” it suggests that by al-Bukhārī’s time there was a clear sense in the tradition that the Liar Paradox posed a serious threat to the foundations of logic. Finally, “as.amm” can mean “mute” or “deaf ” (which is perhaps why the mathematicians chose it to translate the Greek word “alogos,” which can mean “speechless”). Al-Bukhārī’s most entertaining speculation trades on this meaning of the word: perhaps the fallacy is so-called because the person who came up with it (and so the person who is its “root” or origin) was deaf: It is possible that ‘root’ (‘jadhr’) is a name attributed to the author of this fallacy, and he was deaf (as.amm), unable to hear. But this cannot be the case provided what we have reported of the scholar al-Taftāzānī, and [given] that it is reported that the author of this fallacy is Ibn Kammūna al-Baghdādī, who was one of the heretics (malāh.ida).37

1 1. C O N C L U S I O N Al-Taftāzānī’s discussion of the Liar is both intrinsically interesting and historically important. He was the first to introduce Liar Cycles into the 35 Qarāmalekī (12 Treatises, p. 316 n. 273) suggests that this is a reference to Qurashī Jamāl al-Dīn’s al-S.arāh. min al-s.ih.āh., a Persian-Arabic dictionary that would have been available to al-Bukhārī. 36 Al-Bukhārī, al-Muntakhab, in Qarāmalekī, 12 Treatises, 316. 37 Al-Bukhārī, al-Muntakhab, in Qarāmalekī, 12 Treatises, 316. According to Pourjavady and Schmidtke, al-Bukhārī is among the earliest scholars to misattribute the paradox to Ibn Kammūna: As a result of the popularity of the writings of Dawānī and Khafrī’s supercommentary on T.ūsī’s Tajrīd among the philosophers of subsequent generation, Ibn Kammūna now became primarily known as a formulator of philosophical and logical sophistries (shubahat). [ . . . ] One of them was the liar paradox, known among Muslim philosophers as shubhat jadhr al-as.amm or shubhat kull kalāmī kādhib. ʿIs.mat Allāh b. Kamāl al-Dīn al Bukhārī (10th/16th c.) was one of the earliest scholars to claim that this paradox originated with Ibn Kammūna. He even states that it is called as.amm (“deaf ”) because, so he maintained, Ibn Kammūna was deaf (A Jewish Philosopher of Baghdad, 44). Note, however, that al-Bukhārī does not in fact endorse the claim that Ibn Kammūna was deaf, nor does he endorse this explanation of the meaning of “jadhr al-as.amm.”

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Al-Taftāzānī on the Liar Paradox

123

tradition, and the only figure in the tradition who seems to have connected the paradox with the broader debate between the Muʿtazila and Ashāʿira over the nature of goodness and badness. His own “littlest bit of a little bit” of a solution is suggestive even if it fails. And his decision to call the paradox the fallacy of the irrational root remains, for us, as for al-Bukhārī, an enigma. Illinois State University and Pitzer College

BIBLIOGRAPHY Alwishah, Ahmed and David Sanson. “The Early Arabic Liar: The Liar Paradox in the Islamic World from the Mid-Ninth to the Mid-Thirteenth Centuries CE,” Vivarium 47 (2009), 97–127. Al-Baghdādī. Us.ūl al-dīn (Dār al-Kutub al-ʿIlmiyya: Beirut, 1981). Al-Bukhārī. al-Muntakhab min sharh. risālat al-mughālat. āt, in A. Qarāmalekī, 12 Treatises, 311–28. Al-Dashtakī. al-Muntakhab min h.āshiyatihi ʿalā al-sharh. al-jadīd lil-tajrīd, in A. Qarāmalekī, 12 Treatises, 3–14. Al-Dawwānī. Nihāyat al-kalām fī h.al shubhat kull kalāmī kādhib (the Ultimate Saying in Solving the Paradox of “All what I say is False”), in A. Qarāmalekī (ed.), Nāma Mofid 5 (Qom: University Mofid, 1996). Al-Dawwānī. al-Muntakhab min h.āshiyatihi ʿalā al-sharh. al-jadīd lil-tajrīd, in A. Qarāmalekī, 12 Treatises, 65–92. Al-Ghazāl. al-Qust. ās al-mustaqīm, ed. V. Shalh.at (al-Mat. baʿa al-Kāthūlīkiyya, 1959). Al-Ghazālī. Mi ʿyār al- ʿilm, ed. S. Dunyya (Cairo: Dār al-Maʿārif, 1961). Al-Ghazālī. Just Balance, tr. Richard McCarthy (Boston: Twayne Publishers, 1980). Al-Jabbār al-Qād.ī. al-Mughnī, VI, ed. A. al-Ahwānī (Cairo: al-Sharika al-ʿArabiyya li-l-T.ibāʿa wa al-Nashr, 1962). Al-Khafrī. H.ayrat al-fud.alāʾ, in A. Qarāmalekī, 12 Treatises, 271–309. Al-Khawārizmī. Mafātīh. al- ʿulūm, ed. I. al-ʾAbyyārī (Beirut: Dār al-Kitāb al-ʿArabī, 1989). Al-Rāzī. al-Mat. ālib al- ʿāliyya min al- ʿilm al-ilāhī, III, ed. A. al-Ssaqā (Beirut: Dār al-Kitāb al-ʿArabī, 1987). Al-Taftāzānī. Tahdhīb al-mant. q wa al-kalām (Cairo: Mat. baʿat al-Saʿāda, 1912). Both authors would like to thank Calvin Normore, Terry Parsons, Roshdi Rashed, and an anonymous referee. The first author would also like to thank Andy Arlig, Jeff Behrends, Daniel Breyer, Sunil Chebolu, Harry Deutsch, Keota Fields, and Todd Stewart for helpful discussion, comments, and feedback. Material from this paper was presented at the 2015 NEH Summer Institute, “Between Medieval and Modern: Philosophy from 1300–1700,” directed by Robert Pasnau, at the University of Colorado Boulder. Any views, findings, conclusions, or recommendations expressed in this paper do not necessarily represent those of the National Endowment for the Humanities.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

124

David Sanson and Ahmed Alwishah

Al-Taftāzānī. Sharh. al-Maqās.id, part IV, ed. ʿUmara ‘Abd al-Rah.mān, 2nd edn. (Beirut: ʿĀlam al-Kitāb, 1998). Hourani, George. Reason and Tradition in Islamic Ethics (Cambridge: Cambridge University Press, 1985). Kline, Morris. Mathematical Thought from Ancient to Modern Times (Oxford: Oxford University Press, 1990). Marmura, Michael. “Ghazālī on Ethical Premises,” Philosophical Forum 1 (1969), 393–403. Pourjavady, Reza and Sabine Schmidtke. A Jewish Philosopher of Baghdad: “ ʿIzz al-Dawla Ibn Kammūna (d. 683/1284) and His Writings (Leiden: Brill, 2006). Qarāmalekī, Ahad. 12 Treatises on Liar Paradox in Shirāz School, with the collaboration of T.aiiebe ʿĀref Nīyā (Tehran: Iranian Institute of Philosophy, 2007). Rosen, Frederic. The Algebra of Mohammed Ben Musa (London: The Oriental Translation Fund, 1831). Shihadeh, Ayman. “From al-Ghazālī to al-Rāzī: 6th/12th Century Developments in Muslim Philosophical Theology,” Arabic Sciences and Philosophy 15 (2005), 141–79. Street, Tony. “Arabic and Islamic Philosophy of Language and Logic,” in E. N. Zalta, (ed.), The Stanford Encyclopedia of Philosophy (Winter 2013 Edition) .

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation Jacob Tuttle

In the last twenty-five years or so, scholars have increasingly come to recognize the importance of the late scholastic philosopher Francisco Suárez (1548–1617), both for the intrinsic interest of his philosophy, and for his influence on subsequent early modern thinkers.1 However, in spite of this recognition, Suárez’s views on a wide variety of topics remain poorly understood. One such topic is efficient causation. Although his treatment of this topic is often cited in the literature on figures such as Descartes and Malebranche, relatively little attention has been paid to Suárez’s theory of efficient causation in its own right.2 My aim in this paper is to fill part of this 1 Two recent collections of essays on Suárez’s philosophy are Benjamin Hill and Henrik Lagerlund, (eds), The Philosophy of Francisco Suárez (New York: Oxford University Press, 2012); and Daniel Schwartz, (ed.), Interpreting Suárez (New York: Cambridge University Press, 2012). Both volumes include introductions with details about Suárez’s life and work, as well as about the current state of Suárez scholarship. The first of these volumes also includes two invaluable bibliographies: one of Suárez’s own published works in their various editions (including translations), and another of the existing secondary literature on Suárez. Notably, many of the existing English-language translations have been completed within the last twenty-five years. For further discussion of Suárez’s life and the significance of his work, see the preface of Sydney Penner, “Francisco Suárez on Acting for the Sake of an Ultimate End” (Ph.D. Dissertation, Cornell University, 2011), xii–xxxix; John P. Doyle, “Suárez, Francisco,” in E. Craig (ed.), Routledge Encyclopedia of Philosophy (London: Routledge, 1998) ; and Jorge J. E. Gracia, “Francisco Suárez: The Man in History,” American Catholic Philosophical Quarterly 65 (1991), 259–66. Doyle’s assessment of Suárez’s significance and influence is particularly glowing; for a more sober assessment, see Robert Pasnau, Metaphysical Themes 1274–1671 (Oxford: Clarendon Press, 2011), 253. For references to Suárez in the secondary literature on early modern philosophy, see subsequent notes. 2 For two excellent discussions of Suárez’s influence on Descartes’s theory of causation, see Tad Schmaltz, Descartes on Causation (New York: Oxford University Press, 2008), 24–48; and Helen Hattab, “Conflicting Causalities: The Jesuits, their Opponents, and Descartes on the Causality of the Efficient Cause,” in D. Garber and S. Nadler (eds),

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

126

Jacob Tuttle

gap in the scholarship. I do this by examining an important but neglected issue in Suárez’s account of efficient causation—namely, the precise nature of the relationship between an efficient cause and its effect. The central component of Suárez’s account of this relationship is his claim that efficient causation is to be identified with action (actio), one of Aristotle’s ten highest genera (summa genera) or categories (praedicamenta).3 Thus, according to Suárez, an efficient cause brings about its effect in virtue of performing an action. Or, to put the point in another idiom, he thinks that action plays the role of a connection (connexio) or link (vinculum) between an efficient cause and its effect.4 This aspect of Suárez’s theory of efficient causation has been noted several times in the literature.5 However, its significance has not been fully appreciated. In what follows, I show how Suárez’s identification of efficient causation with action helps to shed light on his views about the precise nature of efficient causation, and its role in his ontology. More specifically, I show that Suárez understands efficient causation to be a distinctive or sui generis type of entity, and that he thinks we must adopt this view in order to account for the facts of efficient causation. A few remarks about my strategy are in order. Because, I shall argue, Suárez’s views about action constitute the core of his theory of efficient causation, the bulk of this paper is devoted to explaining and motivating his account of the category of action. In keeping with this focus, the paper is divided into two parts. In the first part (sections 1–2), I sketch the theoretical background against which Suárez presents his theory of efficient causation. In section 1, I introduce Suárez’s concept of efficient causation, with special attention to his understanding of the relationship between efficient causation and action. The most important conclusion of this Oxford Studies in Early Modern Philosophy, vol. 1 (Oxford: Clarendon Press, 2003), 1–22. For a wide-ranging treatment of scholastic influences on Descartes’s philosophy of nature, see Dennis Des Chene, Physiologia: Natural Philosophy in Late Aristotelian and Cartesian Thought (Ithaca, NY: Cornell University Press, 1998). For an illuminating discussion of scholastic views of natural necessity, focusing on Suárez, see Walter Ott, Causation and Laws of Nature in Early Modern Philosophy (New York: Oxford University Press, 2009), ch. 3. For a book-length treatment of Suárez’s theory of efficient causation, with a strong emphasis on theological contexts, see Alfred J. Freddoso, “Introduction,” in Creation, Conservation, and Concurrence. For a more general treatment of Suárez’s theory of efficient causation, see Jacob Tuttle, “Suárez's Metaphysics of Efficient Causation” (Ph.D. Dissertation, Purdue University, 2013). For a discussion of issues connected with the “propinquity,” or nearness, of efficient causes to their effects, see Dennis Des Chene, “Suárez on Propinquity and the Efficient Cause,” in The Philosophy of Francisco Suárez, 89–100. 3 See DM 18.10.5. 4 DM 18.10.8. 5 For discussion, see Schmaltz, Descartes on Causation, 30–3; Hattab, “Conflicting Causalities,” 1–6; and Freddoso, “Introduction,” xliii–l. An older and less helpful treatment is Patout Burns, “Action in Suarez,” The New Scholasticism 38 (1964), 453–72.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

127

section, which guides the remainder of the paper, is that Suárez identifies efficient causation with action. Next, in section 2, I identify some of the most important philosophical and theological commitments that inform Suárez’s account of the nature of efficient causation or action. In the second, longer part of the paper (sections 3–6), I argue for a detailed interpretation of Suárez’s views about the category of action. As it turns out, he develops the bulk of his account in the context of a scholastic dispute about whether action is reducible to any of the other Aristotelian categories. For this reason, my interpretation emphasizes Suárez’s own position in this dispute. In section 3, I offer a precise characterization of what Suárez regards as the two main dialectical options in this controversy— what I call “reductive” and “non-reductive realism about action”—and I show that he endorses the latter of these options. In section 4, I present and explain Suárez’s arguments for his own non-reductive theory of action. In section 5, I present two objections to Suárez’s non-reductive theory, and I show how his responses to these objections shed light on his further characterization of efficient causation or action. Finally, in section 6, I make some concluding remarks about the philosophical significance of Suárez’s theory of efficient causation. Although Suárez discusses efficient causation in several places in his corpus, his most thorough and systematic treatment is found in the Metaphysical Disputations.6 Accordingly, in this paper I focus primarily on material from this work. I draw most heavily on DM 12, On the Causes of Being in General; DM 18, On the Proximate Efficient Cause, and its Causality, and Everything that it Requires in order to Cause; DM 20, On the

6 Suárez provides a brief overview of the topics covered in the Metaphysical Disputations, as well as his rationale for writing this work, in a note to the reader, “Ratio et discursus totius operis: ad lectorem.” This note precedes the “Index locupletissimus in Metaphysicam Aristotelis,” which begins on p. i of the Vivès edition. These items have been included together in translation by John Doyle. See Francisco Suárez, A Commentary on Aristotle’s Metaphysics: Index locupletissimus in Metaphysicam Aristotelis, trans. John Doyle (Milwaukee, WI: Marquette University Press, 2004). In preparing this paper, I have consulted reprints of volumes 25 and 26 of the Vivès edition of Suárez’s Opera omnia. References to this work are indicated by “DM,” followed by disputation number, chapter number, and paragraph number. For example, “DM 17.1.2” indicates Disputation 17, Chapter 1, paragraph 2. All translations are my own. However, I have consulted existing English translations where available, and have in some cases adopted their wording without significant changes. See especially Francisco Suárez, On Efficient Causality: Metaphysical Disputations 17, 18, and 19, tr. A. J. Freddoso (New Haven, CT: Yale University Press, 1994); and Francisco Suárez, Creation, Conservation, and Concurrence: Metaphysical Disputations 20–22, tr. A. J. Freddoso (South Bend, IN: St. Augustine’s Press, 2002). For Suárez’s other works, I have consulted the Vivès edition of his Opera omnia.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

128

Jacob Tuttle

First Efficient Cause, and his First Action, which is Creation; and DM 48, On Action.7

1. E F F I C I E N T C A US A T I O N A N D A C T I O N Suárez presents his theory of efficient causation within the context of a much broader account of causation and causal explanation. Accordingly, in order to understand his views about efficient causation, it will be helpful to have a rough outline of this broader account in mind. Like other Aristotelians, Suárez acknowledges four different types of causes—namely, formal causes, material causes, efficient causes, and final causes.8 Against some unnamed opponents, he argues that the four Aristotelian causes do in fact share a common notion (conceptus) or account (ratio), and he spends the bulk of DM 12 attempting to clarify exactly what this account is.9 This discussion culminates in his definition of a cause as a principle (principium) or source “instilling (influens) being in another.”10 Accordingly, Suárez’s view is that each of the Aristotelian causes somehow “instills” (influit) being in its effect. So, how are we to understand what it means for one thing to instill being in another? The Latin verb “influere” is most literally translated as “to flow in” or “to pour in.”11 However, it is important to emphasize that in his definition of a cause, Suárez is using the term in a figurative or metaphorical sense. In fact, elsewhere in the Metaphysical Disputations, he explicitly denies 7 DM 18, De causa proxima efficiente, ejusque causalitate, et omnibus quae ad causandum requirit; DM 20, De prima causa efficiente, primaque ejus actione, quae est creatio; DM 48, De actione. 8 DM 12.3. 9 See especially DM 12.2. 10 DM 12.2.4. 11 None of the options for translating Suárez’s use of “influere” is completely satisfying. “To flow being” is an ungrammatical expression in contemporary English, given that “to flow” is an intransitive verb. Moreover, as I shall explain below, both “to flow” and “to pour” involve a misleading transference metaphor. Two alternatives would be “to influence” and “to have influence,” which are Freddoso’s standard translations. But these translations suggest that the being in question is preexistent and is somehow changed by the cause, and so is also misleading. In adopting “to instill,” I am following Gracia. See the glossary to Suárez on Individuation: Metaphysical Disputation V: Individual Unity and its Principle, tr. J. J. E. Gracia (Milwaukee: Marquette University Press, 1982), 186. For Fredosso’s use of “to influence,” see Suárez, On Efficient Causality, and Suárez, On Creation, Conservation, and Concurrence. For the literal translation given above, see Cassell’s Latin Dictionary, ed. D. P. Simpson (New York: MacMillan Publishing Company, 1968), 305. For a treatment of the various senses of “influxus” in late medieval and early modern philosophy, see Eileen O’Neill, “Influxus Physicus,” in S. Nadler (ed.), Causation in Early Modern Philosophy: Cartesianism, Occasionalism, and Preestablished Harmony (University Park, PA: Pennsylvania State University Press, 1993), 27–56.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

129

that causes literally transfer anything to their effects.12 Moreover, his subsequent treatment of each of the four causes provides further confirmation that he intends “influere” to be taken figuratively. This is perhaps easiest to appreciate for the final and efficient causes, which Suárez calls “extrinsic causes” (causae extrinsicae) because they do not enter into the composition of their effects. Characterizing these causes, he writes: [An extrinsic cause] does not communicate to its effect its own proper and (as it were) individual being; rather, [it communicates] another [being], really flowing forth (profluens) and emanating from such a cause [ . . . ].13

In contrast, Suárez understands the so-called “instrinsic causes” (causae intrinsicae)—namely, formal and material causes—as constituents of hylomorphic compounds. For example, a material substance is composed of both a formal and a material cause. Accordingly, to the extent that such causes do enter into the composition of their effects, he claims that they contribute their own being to those effects. Nevertheless, Suárez plainly does not think that in making this contribution, formal and material causes lose any of their own being, as a literal reading of “influere” would suggest. It is not as if being is a liquid that can be moved around from one vessel to another. But if this is right, then what does Suárez mean when he says that causes instill being in their effects? He presents his clearest answer to this question in DM 12, several paragraphs after his generic definition of a cause. There, he argues that what is common to each of the four causes can also be captured by appealing to the notion of existential dependence. On this view, to say that a cause instills being in its effect “means the very same thing” as to say that an effect depends for its being on its cause.14 This equivalence helps to explain why Suárez also routinely characterizes causes in terms of the existential dependence of their effects. And in fact, although he says that for expository reasons he prefers to define a cause in terms of its instilling being in another, he also acknowledges what he takes to be a coextensive definition, according to which a cause is anything on which another depends for its being.15 Although this definition is perhaps not completely transparent, it does have the advantage of avoiding the misleading transference metaphor that appears in Suárez’s preferred definition. For this reason, in the remainder of this paper I shall generally frame Suárez’s talk of causes in terms of existential dependence, rather than instilling being. Before turning to the details of Suárez’s views about efficient causes, it is worth pausing to make one final observation about causes in general. So far 12 For one very clear denial, see DM 12.2.7. 13 DM 17.1.6. 14 DM 12.2.7. 15 DM 12.2.4.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

130

Jacob Tuttle

I have been emphasizing that Suárez thinks the generic account of a cause can be captured by appealing to the notion of existential dependence: C is a cause of E if and only if E depends on C for its being. However, it bears emphasizing that according to Suárez, it is precisely in virtue of this relationship of existential dependence that C qualifies as an actual cause of E. He sometimes expresses this point by noting that existential dependence is to be identified with causation or “causality” (causalitas) itself. For example, he writes: [T]he causality of each cause is that by virtue of which [ . . . ] it achieves its effect, and conversely that by virtue of which the effect depends on such a cause.16

Now, as it turns out, Suárez thinks that each of the four Aristotelian causes involves its own distinctive type of causality or existential dependence. In fact, the existence of four different kinds of causality is one of the chief reasons he offers for thinking that there must be four different kinds of causes. Accordingly, Suárez’s characterization of efficient causes emphasizes how their causality differs from that of formal, material, and final causes. He writes: [M]atter and form [ . . . ] do not cause by means of action, but through formal and intrinsic union. And an end [i.e., a final cause] causes only through metaphorical motion, insofar as it is an end. On the other hand, an efficient [cause] causes through a proper action flowing from it.17

As this passage makes clear, Suárez thinks that, unlike these other causes, an efficient cause brings about its effect by performing an action. We have seen already that Suárez classifies formal and material causes as intrinsic causes, because he takes them to be constituents of hylomorphic compounds. For this reason, it should not be surprising that he understands the causality of formal and material causes to be their union with one another, which he elsewhere describes as the “inherence” (inhaerentia) of the formal cause in the material cause. Moreover, although final and efficient causes are both extrinsic causes, Suárez insists that final causes do not perform actions. Instead, they are the ends for the sake of which efficient causes act, and so “move” efficient causes only metaphorically. Of the four causes, then, only efficient causes bring about their effects via action. Accordingly, action is the causality of an efficient cause.18 That is to say, action is the peculiar sort of existential dependence that an effect has on an efficient cause. Or, to put the point in a more modern idiom, efficient causation just is action.

16 DM 18.10.6. 17 DM 17.1.6. 18 Suárez makes this claim routinely, but he defends it most explicitly in DM 18.10.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

131

In order to appreciate the significance of this result, it will be helpful to consider one of Suárez’s favorite examples of efficient causation, in which a fire heats some water. Imagine that a pot of cold water is placed over a fire, where it remains until the water is boiling. In this case, Suárez will say that the agent or efficient cause is the fire, and its effect is an accidental form— namely, a quality of heat—that comes to exist in the water. And, most importantly for our purposes here, the causal connection between the fire and the quality of heat just is the fire’s action of heating. That is to say, it is precisely in virtue of heating the water that the fire causes or instills being in the quality of heat. Or, described alternatively, it is precisely in virtue of the fire’s heating the water that the quality of heat depends on the fire for its existence. With this example in mind, I want to address two potential sources of misunderstanding about how Suárez conceives situations involving efficient causation. The first of these concerns the extension of the term “action.” Philosophers nowadays generally reserve “action” for the purposive or intentional behavior of conscious agents. A paradigmatic example of an action in this sense would be a person moving her hand with the aim or intention of grasping an object. However, it turns out that Suárez understands action in a much broader sense. To be sure, he acknowledges that conscious, intentional behavior qualifies as action. Nevertheless, as his example of fire heating water makes clear, he thinks that action extends well beyond what is done by conscious agents. Suárez’s reasons for holding this view can be best appreciated by observing that he understands action in terms of making or producing something. In fact, he often points out that performing an action is nothing over and above making or producing an effect. For example, he writes: [A]ction, as action, if it is true and proper, is nothing besides production (productio) or making (effectio) [ . . . ].19 [W]hat is it to make (efficere) except to act (agere)?20

Accordingly, to the extent that an unconscious agent (such as fire) is able to produce an effect (such as heat), Suárez regards it as being capable of acting. Another potential source of misunderstanding concerns the nature of causes and effects. In the contemporary philosophical literature, causes and effects are routinely taken to be events. This can be illustrated by a common example of causation, made famous by David Hume, in which the event of one billiard ball striking another causes the event of the second billiard ball rolling.21 19 DM 48.2.16. 20 DM 18.10.3. 21 See David Hume, A Treatise of Human Nature, ed. D. F. Norton and M. J. Norton (Oxford: Clarendon Press, 2007), 111.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

132

Jacob Tuttle

However, in Suárez’s example of efficient causation outlined above, the efficient cause is a substance (a fire), and its effect is an accidental form (a quality of heat). And as it turns out, he thinks that this case is representative of efficient causation in the natural world. In such instances, the agent or efficient cause is always a substance, and its effect is always a form (either an accidental or a substantial form). This view fits well with Suárez’s understanding of action as the making or production of an effect. For while it makes good sense to say that substantial and accidental forms are produced by their causes, it is not immediately clear what it would mean to say that events are made or produced. This completes our introduction to Suárez’s concept of efficient causation. To sum up the results, Suárez understands the four Aristotelian causes in terms of four distinctive types of existential dependence or causality. Unlike the other causes, efficient causes bring about their effects by acting, and for this reason Suárez identifies action with the special sort of existential dependence or causality associated with efficient causes. For the same reason, on Suárez’s view, action just is the connection between an efficient cause and its effect. However, it is important to recognize that, by itself, this characterization of efficient causation is fairly neutral about its nature or ontological status. After all, it will be illuminating only to the extent that we already have a clear account of what action is in itself. Indeed, Suárez emphasizes this very point near the beginning of his treatment of efficient causation in DM 17, where he defines an efficient cause as a “per se principle from which an action first exists.”22 Immediately after proposing this definition, he acknowledges that action’s precise nature is itself “somewhat obscure,” and thus in need of further clarification. As it turns out, Suárez provides this clarification in the context of a dispute about whether the Aristotelian category of action is reducible to any of the other categories. Accordingly, the remainder of this paper focuses on clarifying and motivating Suárez’s position in this dispute. But in order to follow the details of the dispute itself, it will be helpful to have in mind a few of the philosophical and theological commitments that inform his views about the nature of efficient causation or action. I turn to these now.

2 . P H I L O S O P HI C A L A N D T HE O LO G I C A L COMMITMENTS Suárez’s views about the nature of efficient causation or action were strongly influenced by two of the main intellectual currents of the medieval era, 22 DM 17.1.5.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

133

Aristotelianism and Christianity. Fundamental to all medieval theorizing about the nature of action is the Aristotelian doctrine that action is one of the nine categories of accidents.23 Aristotle and his medieval followers divide real beings into accidents, which they construe as existing in a subject, and substances, which they construe as not existing in a subject. It is important to notice that nearly all medievals take this conception of accidents to rule out the possibility of a single accident spanning two or more subjects. That is to say, medievals typically understand their substance–accident ontology to preclude the existence of what we would nowadays call polyadic (or many-place) properties.24 Even accidents such as action, which clearly play a relational role, must somehow be accommodated within a framework of monadic properties. This conception of action naturally raises the question of which subject it exists in. Aristotle addresses this question in Book 3 of the Physics, where he articulates a general model for understanding situations involving action. According to this model, an action is always performed on some patient, where a patient is understood as something that undergoes a change. This can be illustrated by our earlier example of efficient causation, in which some fire (an agent) heats some water (a patient), which changes from being cold to being hot. Not surprisingly, Aristotle thinks that something qualifies as an actual agent in virtue of an action, and that something qualifies as an actual patient in virtue of a passion (being-acted-on). However, he argues that action and passion both exist in the patient, and moreover that they are one and the same change or motion. In our example, the fire’s action of heating, as well as the water’s passion of being-heated, both exist in the water, and they are the same change or motion undergone by the water— namely, its becoming hot.25 As already noted, another important influence on Suárez’s views about the nature of action is Christianity. According to the Christian theological tradition, God causally interacts with the world in a variety of ways. Perhaps 23 The Aristotelian categories are substance, quantity, quality, relation, action, passion, where, when, situation, and habit. The most important source for Aristotle’s category theory is his short treatise Categories, 1a1–15b32. 24 See Jeffrey Brower, “Medieval Theories of Relations,” in E. N. Zalta (ed.), The Stanford Encyclopedia of Philosophy, (Winter 2013 Edition) ; and Mark Henninger, Relations: Medieval Theories 1250–1325 (New York: Oxford University Press, 1989), 4–6. For an apparent exception to this nearly unanimous medieval view, see Heine Hansen, “Strange Finds, or Nicholas of Paris on Relations,” in J. L. Fink, H. Hansen, and A. M. Mora-Marquez (eds), Logic and Language in the Middle Ages: A Volume in Honor of Sten Ebbesen (Leiden: Brill, 2012), 139–54. 25 See Physics 3.3, 202a15–202b20. For a recent treatment of this passage, see Anna Marmodoro, “The Union of Cause and Effect in Aristotle: Physics III 3,” Oxford Studies in Ancient Philosophy 32 (2007), 205–32.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

134

Jacob Tuttle

most strikingly, this tradition claims that God creates the world ex nihilo, or from nothing. Like other medievals, Suárez understands God’s creation of the world as an instance of efficient causation, and accordingly, he takes creation to be a type of action. However, he understands this type of action to depart from the Aristotelian paradigm in at least one important respect— namely, that it precludes the existence of a patient on which the agent acts. When God creates, he does not change any preexisting stuff, but rather brings into being his effect and whatever is presupposed by it, all at once. Suárez thus recognizes two fundamentally different types of action. On the one hand, there is motion, which conforms to the model proposed by Aristotle, and consequently presupposes the existence of a patient on which the agent acts. And on the other hand, there is creation, which does not conform to this model, insofar as it precludes the existence of a patient.26 Suárez’s theological commitments have an important implication for his views about the subject in which action exists. Like others in the Aristotelian tradition, he conceives action as a kind of accidental being—that is to say, a being that exists in another. However, because he acknowledges the possibility of creation ex nihilo, he must abandon the Aristotelian view that every action exists in a patient. Instead, Suárez claims that every action exists in an effect. Although he does not offer a sustained argument for this thesis, he seems to have been led to it by his view that action just is the peculiar sort of existential dependence that an effect has on an efficient cause. But once this dependence is construed as an accidental being, it is difficult to see what it could exist in, besides a dependent thing. Indeed, according to the standard medieval analysis of accidental property-possession, some subject qualifies as an actual F precisely in virtue of having some accidental being—an F-ness— existing in it. All this helps to explain why Suárez endorses the general principle that “a dependence must necessarily exist in a dependent thing.”27 Now, in light of what we have seen so far, one might naturally assume that Suárez takes efficient causation to involve only the initial production of an effect, whether this production is an instance of motion or creation ex nihilo. However, according to the Christian theological tradition, God not only creates the world in the first instance, but also preserves or sustains it over time. In keeping with this tradition, Suárez refers to God’s continual 26 For Suárez’s definition of creation, see DM 20.1.1. For discussion of Suárez’s views about creation and its distinction from motion, see Tuttle, “Suárez's Metaphysics of Efficient Causation,” ch. 3; and Freddoso, “Introduction,” xxxvii–xxxviii and lxxiii– lxxxvii. For a more general treatment of issues connected with creation and efficient causation in medieval philosophy, see Taneli Kukkonen, “Creation and Causation,” in R. Pasnau and C. Van Dyke (eds), The Cambridge History of Medieval Philosophy, 2 vols. (Cambridge: Cambridge University Press, 2010), 232–46. 27 DM 20.4.6.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

135

preservation of the world as “divine conservation” (conservatio divina), and he argues that “all entities apart from God depend on divine conservation for their being.”28 Accordingly, he thinks that if God should ever stop efficiently causing some particular creature, that creature would simply go out of existence.29 Because Suárez’s defense of his non-reductive theory of efficient causation presupposes two additional points about how he understands conservation, I will briefly note them here. First, although Suárez believes that creatures depend on God for their continuing existence in a special way, he is also willing to grant that in some cases creatures are conserved by other creatures. For example, he thinks that light produced by the sun depends on the sun not only for its initial coming-to-be, but also for its continued existence thereafter.30 Suárez thus understands the term “conservation” broadly enough to include any action by which something is sustained in existence over time, whether this action is performed by God or by a creature.31 The second point concerns the relationship between instances of production and conservation. Imagine a case in which some light is initially produced by the sun at dawn, and is thereafter conserved by the sun for the remainder of the day. In such a case, will the sun’s action of conservation be really distinct from its action of production? In answer to this question, Suárez acknowledges that there is a conceptual distinction between the sun’s production of the light and its subsequent conservation of it. It is included in the concept of production that the relevant effect has not yet existed, whereas it is included in the concept of conservation that this effect has already existed. Nevertheless, Suárez thinks that in the case described above, there is only a single, continuous action performed by the sun throughout 28 DM 21.1.4. 29 See DM 21.1.14. For discussion of Suárez’s theory of divine conservation, see Freddoso, “Introduction,” lxxxviii–xcv. 30 DM 21.3.10. 31 As it turns out, Suárez thinks that creatures never perform actions on their own, without God’s participation or concurrence (concursus). This is another important way in which his account of efficient causation is influenced by the Christian theological tradition. However, because Suárez’s views about divine concurrence are only tangentially related to the topic of this paper, I do not discuss them in detail here. For his lengthy treatment of the topic, see DM 22. For discussion, see Freddoso, “Introduction,” xcv–cxxi; Freddoso, “God’s General Concurrence with Secondary Causes: Pitfalls and Prospects,” American Catholic Philosophical Quarterly 67 (1994), 131–56; and Freddoso, “God’s General Concurrence with Secondary Causes: Why Conservation is not Enough,” Philosophical Perspectives 5 (1991), 553–85. For treatments of Descartes’s theory of causation that rely on Suárez’s account of concurrence, see Geoffrey Gorham, “Cartesian Causation: Continuous, Instantaneous, Overdetermined,” Journal of the History of Philosophy 42 (2004), 389–423; and Andrew Pessin, “Descartes’s Nomic Concurrentism: Finite Causation and Divine Concurrence,” Journal of the History of Philosophy 41 (2003), 25–49.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

136

Jacob Tuttle

the day. Accordingly, even though production and conservation can be distinguished in thought, they are not necessarily distinct in reality. Not surprisingly, Suárez thinks this result holds equally well for God’s creation and conservation of creatures.32 With this background in mind, let us turn now to Suárez’s views about the nature of efficient causation or action.

3 . T H E D I A L E C T I C AL O P T I O N S : R E D U C T I V E A N D N O N - RE D U CT I V E R EA L I S M Throughout the Metaphysical Disputations, Suárez takes for granted what I shall call realism about action—namely, the view that action exists in extramental reality. One statement of this view can be found in the Prologue to DM 48, where he notes that action is “something in the nature of things (in rerum natura)” and is “contained within the latitude of being (entis).” He does not dwell on this thesis, and as far as I can tell, he does not explicitly argue for it anywhere. However, his contemporaries probably would have regarded it as uncontroversial. One reason for thinking that this is correct has to do with medieval accounts of the categories more generally. According to the dominant medieval view, Aristotle’s categories constitute a scheme for classifying extramental reality, so that the categories include only extramental beings.33 Suárez follows the tradition on this point, noting approvingly that it is “for all of the philosophers the common way of conceiving and speaking” of the categories.34 Thus, his realism about action is just part and parcel of a more thoroughgoing realist program for the categories as a whole. Consequently, Suárez’s discussion of the nature of action focuses on specifying what sort of extramental being action is. As we shall see in what follows, much of this discussion aims to clarify action’s precise relationship to the other categories. However, before proceeding to the details of this discussion, it will be helpful to make a few additional remarks about how he understands the categories themselves. Because he thinks that the categories include only extramental beings, it is perhaps natural to assume that Suárez also takes them to classify such beings exhaustively and exclusively, so that each is included in at least one category, and none is included in more than 32 See DM 21.2.2–7 and DM 49.2.11. For a brief discussion, see Freddoso, “Introduction,” xci–xcii. 33 See Pasnau, Metaphysical Themes, 229–30; and Brower, “Medieval Theories of Relations.” 34 DM 32.2.18.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

137

one. After all, philosophers normally think of a realistic category theory as one that attempts to “carve up reality at the joints,” and two desiderata of such a theory are that it should neither omit any part of reality, nor count any part of reality twice. Nevertheless, Suárez denies that the categories are exclusive. That is to say, he thinks that at least some of the categories can overlap, in the sense that the same extramental beings may fall under more than one category. He expresses this most clearly in the context of his discussion of the so-called “sufficiency” (sufficientia) of the categories, where he rejects the view that entities falling under different categories must be distinct “according to being” (sub ente), or in extramental reality. He writes: But this opinion cannot be defended in general. For in the first place, concerning all categorical relations, it [seems] most likely to me that they are not actually distinguished on the side of reality (a parte rei) from their foundations [ . . . ], but only according to reason or the connotation of the other extreme. Moreover, concerning action and passion it is very nearly the common opinion that they are not actually distinguished in extramental reality (ex natura rei), yet nevertheless that they constitute diverse categories.35

In this passage, Suárez’s main point merely echoes what is, if not the unanimous opinion of medievals, at any rate the opinion of the overwhelming majority. Although medievals construe the categories as a system for classifying extramental beings, most nevertheless understand this system to classify such beings via our ways of conceiving or speaking of them.36 However, because it is not obvious that our ways of conceiving or speaking are exactly isomorphic with reality, this way of understanding the categories 35 The quotation is taken from DM 39.2.21. For Suárez’s characterization of the view he is attacking, see DM 39.2.19–20. In most contexts, Suárez understands a distinction ex natura rei as a specific type of real or extramental distinction—namely, the distinction between a thing (res) and its mode (modus). Suárez is emphatic that such a distinction obtains in extramental reality. He makes this point most explicitly in DM 7.1.16, where he notes that a distinction ex natura rei “can be called ‘real’ in the general sense, because it is truly on the side of reality (ex parte rei), and is not through an extrinsic denomination from the intellect.” However, he sometimes also appears to use “ex natura rei” to signify any extramental distinction—even a distinction between thing and thing—and it is not always clear which sense he intends. I introduce Suárez’s notion of a mode later in this section. For Suárez’s theory of distinctions, see DM 7, De variis distinctionum generibus. For an English translation, see Francisco Suárez, On the Various Kinds of Distinctions, tr. C. Vollert (Milwaukee: Marquette University Press, 1947). For discussion, see Sydney Penner, “Suárez on the Reduction of Categorical Relations,” Philosophers’ Imprint 13 (2013), section 2.2. 36 See Pasnau, Metaphysical Themes, 229–30. This view of the categories can be traced back to Categories commentaries from late antiquity. See Giorgio Pini, Categories and Logic in Duns Scotus: An Interpretation of Aristotle’s Categories in the Late Thirteenth Century (Leiden: Brill, 2002), 9–10.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

138

Jacob Tuttle

leaves open the possibility that a single entity might fall under more than one category. And as it turns out, nearly all medievals think that at least some of the categories overlap in this way. Suárez’s example of the category of relation is a controversial one. However, as he notes, most medievals think that the categories of action and passion include the same entities. This is at least partly because of Aristotle’s opinion in the Physics that action and passion are the same motion. Although the precise relationship between action and passion is an important topic in Suárez’s treatment of the nature of efficient causation, here it will suffice to make a few brief remarks. As the quoted passage makes clear, Suárez endorses the standard medieval opinion about the categories of action and passion—namely, that they are constituted by the same reality, or as he notes elsewhere, by “the same dependence” of an effect on an efficient cause.37 As a result, we can think of the categories of action and passion as together performing a distinctive, efficiently-causal role in Suárez’s ontology. Because it should be clear from the context that I am focused exclusively on efficient causation, I shall sometimes refer to the categories of action and passion simply as the causal categories, and the remaining categories as the non-causal categories. Nevertheless, it is important to emphasize that for Suárez and other Aristotelians, action and passion have different notions or accounts (rationes). One way of understanding this difference is by recalling the metaphysical or functional roles of action and passion, as articulated in the first two sections of this paper. As we have seen, Suárez thinks something qualifies as an agent or efficient cause in virtue of an action, whereas something qualifies as a patient in virtue of a passion. Making the same point, he sometimes characterizes action and passion as the actuality of an agent and a patient, respectively.38 Suárez also emphasizes that these distinctive metaphysical or functional roles are associated with the distinctive relations in which a given instance of existential dependence stands. On his view, the same instance of existential dependence both proceeds from the agent or efficient cause, and exists in the patient in which the effect is produced. Thus, when we think of this dependence as proceeding from an agent or efficient cause, we are conceiving of it as an action. On the other hand, when we think of it as existing in a patient, we are conceiving of it as a passion. This difference helps to explain why Suárez treats the categories of action and passion separately (in DM 48 and 49, respectively), as well as why he raises questions about action that he does not about passion, and vice versa. Following Suárez, in the

37 DM 49.1.8.

38 Ibid.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

139

remainder of this paper I shall focus on questions that are relevant to his characterization of action as such.39 For the purposes of this paper, the most important of these questions concerns the relationship between action and the non-causal categories. Exploiting the conception of the categories already outlined, some medievals endorse a kind of reductionism about action. According to this view, which I shall call reductive realism about action (or simply “reductive realism,” where the context is clear), action is to be identified with the effect that serves as its end-point (terminus). As noted already, Suárez insists that, whatever action is in itself, it must be something that exists in an effect. The idea behind reductive realism is that actions may in a certain sense exist “in” effects without being extramentally distinct from them. For example, on this view, when a fire heats an object, the action of heating is not anything over and above the heat produced in the object. Of course, we might be able to conceive this action of heating as if it were a distinct property existing in the heat produced. Nevertheless, on this reductive view, in extramental reality the action of heating just is the heat produced. Consequently, the action of heating also falls under the category of the heat produced. More generally, things in the category of action also fall under the categories of the effects that serve as their end-points. If the effect in question is a substance, then the action falls under the category of substance; if the effect is a quality, then the action falls under the category of quality, and so on. I am calling this view “reductive” because (at least if we take for granted Suárez’s view that action just is efficient causation) it tries to account for efficient causation not by appealing to a distinctive or sui generis sort of causal entity, but rather by appealing to ordinary non-causal entities. Thus it is reductive in much the same way as theories that account for modality by appealing to the actual, or theories that account for the mental by appealing to the physical. Suárez attributes reductive realism to the nominalists William Ockham and Gregory of Rimini, as well as to the Thomist thinkers Hervaeus Natalis (d. 1323) and Chrysostomus Javellus (1470–1538).40 However, Suárez himself emphatically rejects reductive realism about action, instead 39 Suárez’s most detailed treatment of the differences between action and passion is found in DM 49.1. Valuable insights are also found throughout DM 48, especially in }4, where he examines whether action as such includes a subject of inherence. For discussion of the relationship between action and passion in Aquinas, as well as relevant citations from Aquinas’s work, see Michael Rota, “Causation in Contemporary Metaphysics and in the Thought of Thomas Aquinas” (Ph.D. Dissertation, St. Louis University, 2006), 126–37. 40 For Suárez’s attributions and his explanation of these views, see DM 18.10.2–3, DM 20.4.6, and DM 48.1.9–12.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

140

Jacob Tuttle

endorsing what I shall call non-reductive realism about action (or “nonreductive realism,” for short).41 According to him, actions are distinct in extramental reality from the effects in which they exist. More generally, efficient causation or action is not to be explained by appeal to ordinary non-causal entities. Here are three examples of what is a common refrain in Suárez’s work on action: [A]n action is not [ . . . ] the very effect produced by an agent [ . . . ].42 For in all of these [examples] the action [ . . . ] is distinguished in extramental reality from its end-point [ . . . ].43 [A]ction must be some medium between an agent and an effect, and distinct in extramental reality from both of them.44

Now, in light of the fact that Suárez regards efficient causation or action as a distinctive, irreducible type of entity, it is natural to wonder exactly how we are to understand this entity. That is to say, how does Suárez characterize efficient causation’s sui generis nature? This is a question that shall occupy much of the remainder of this paper. For the moment, however, I want to focus on two specific ways in which Suárez thinks non-reductive realism might be developed. As noted already, he understands efficient causation or action as an accidental being that exists in an effect, and thereby accounts for that effect’s existential dependence on its efficient cause. However, it turns out that Suárez acknowledges two very different types of accidental beings. On his view, Aristotle’s nine categories of accidents can themselves be divided into what he calls “things” (res) and “modes” (modi). (In keeping with the convention in the secondary literature, I shall refer to the former as “real accidents.”) Accordingly, in order to best appreciate Suárez’s views about action’s sui generis nature, it will be helpful to briefly clarify how he understands this division.45 41 I take the terms “reductive realism” and “non-reductive realism” from Brower, “Medieval Theories of Relations.” 42 DM 18.10.8. 43 DM 20.4.12. 44 DM 48.1.10. 45 In order to avoid unnecessary complications, my discussion here is limited to things and modes that fall in the accidental categories. But it is worth noting parenthetically that Suárez regards substances as things, and he also locates some modes in the category of substance. Suárez’s most detailed treatment of modes is found in DM 7. Suárez also makes important remarks about the distinction between things and modes, as well where they fall in the Aristotelian categories, in DM 32.1 and DM 37.2. For a treatment of Suárez’s theory of modes, see Stephen Menn, “Suárez, Nominalism, and Modes,” in K. White (ed.), Hispanic Philosophy in the Age of Discovery (Washington, DC: Catholic University of America Press, 1997), 226–56. For a broader discussion of scholastic theories of modes, including Suárez, see Pasnau, Metaphysical Themes, 244–75. For an older discussion of Suárez on modes that includes citations from throughout the Metaphysical Disputations, see Peter Nolan, “The Suárezian Modes,” Proceedings of the Annual Convention of the Jesuit Educational Association, Midwest Division (1931), 184–200.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

141

Before turning to the details of the division, however, I need to address a complication in the way Suárez understands accidents. As we have seen already, Aristotelians construe accidents as existing in a subject, and for this reason one might naturally expect them to think that accidents cannot exist apart from a subject. Indeed, Aristotle appears to draw just this conclusion in the Metaphysics, where he mentions several examples of accidents and insists that none of them is “self-subsistent or capable of being separated from substance.”46 Thus, for example, Socrates’s whiteness is his whiteness, and it is difficult to see how it could exist without actually inhering in Socrates. Suárez appreciates the force of this intuition, and he is happy to grant that accidents cannot naturally exist without inhering in a subject. Nevertheless, he accepts the standard scholastic view that certain types of accidents can be miraculously conserved apart from any subject. Suárez portrays this view as a corollary of the Catholic doctrine of transsubstantiation in the Eucharist. According to this doctrine, when a priest consecrates the host, the bread and wine are converted into the body and blood of Christ. In the course of this conversion the substance of the bread and wine is destroyed, but the sensible accidents of the bread and wine—their quantity, color, odor, and so on— continue to exist. Because he takes it to be obvious that these sensible accidents do not inhere in Christ’s glorified body and blood, Suárez believes that they must instead be miraculously conserved by God apart from any subject.47 Accordingly, he insists that some kinds of accidents can be separated from their subjects, even if only by divine power. This aspect of Suárez’s theory of accidents forms the basis for what is, for the purposes of this paper, the most important difference between real accidents and modes. He characterizes the former as accidents for which actual inherence in a subject is not part of their essence. Instead, he argues that what is essential to real accidents is their aptitude to inhere.48 As examples of real accidents, he cites quantity and some kinds of quality, such as color.49 In contrast, Suárez characterizes modes as accidents that

46 Metaphysics 7.1, 1028a20–25. 47 This is actually a simplification of Suárez’s view. Following Aquinas, Suárez thinks that in transsubstantiation, the quantity of the bread and the wine exists as a subsistent entity, without any subject, whereas the sensible qualities inhere in this quantity. See Suárez’s theological work De Sacramentis in genera, Baptismo, Confirmatione, Eucharistia, Missae Sacrificio, Question 77, Articles 1–2. This work is a commentary on the same topics in Part 3 of Aquinas’s Summa Theologiae, and includes the text of the same question and articles. For an easily accessible collection of Aquinas’s works, see the Corpus Thomisticum site at . 48 DM 37.2.9. 49 See DM 37.2.3–4 and the passages cited earlier from De Sacramentis.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

142

Jacob Tuttle

cannot be conserved apart from their subjects, even by divine power. Thus, in his presentation of his theory of modes in DM 7, he writes: [A] mode so intimately includes a union (conjunctionem) with the thing of which it is a mode, that it cannot exist by any power without that [union].50

Modes appear throughout Suárez’s metaphysics, but some of his most prominent examples are the union or inherence (inhaerentia) of form in matter; shape (figura), which he takes to be a mode of quantity; and the Aristotelian categories of where (ubi) and position (situs).51 Because this last example is perhaps the easiest for us to appreciate, I want to briefly explain it here. As an example of something in the category of position, consider Socrates’s sitting (sessio). If Socrates’s sitting were a real accident, then on Suárez’s view it should be possible for God to conserve Socrates’s sitting apart from Socrates himself. However, Suárez takes it as obvious that under no circumstances can there be a sitting position apart from any seated thing. Accordingly, he classifies instances of position as modes of positioned things. As it turns out, Suárez also classifies efficient causation or action as a mode, rather than a real accident. More specifically, he takes it to be a mode of the effect that serves as its end-point. His clearest argument for this thesis focuses on action’s relational character. As I have emphasized so far, Suárez posits action in order to account for the causal connection between efficient causes and their effects. Accordingly, he describes action variously as the causality of an efficient cause, as the existential dependence of an effect on its efficient cause, and as the production of an effect. However, each of these ways of understanding action evidently presupposes the existence of some effect that is characterized by the action. For this reason, Suárez takes it to be inconceivable that an action should exist apart from the effect that it characterizes. Thus, he writes: [I]t is impossible even to conceive by the mind a true production through which there is not something produced, or an actual causality without something caused. Therefore it is also impossible to understand an action without an end-point.52

Because an action cannot be separated from its end-point, even in the understanding, Suárez concludes that it must be a mode of that end-point. Of course, it bears emphasizing that in this argument, Suárez is taking for granted that efficient causation or action must be some sort of accidental being that is extramentally distinct from its end-point. The purpose of the 50 DM 7.1.20. See also DM 37.2.10. 51 See DM 7.1.18–19, DM 7.2.10, and DM 7.2.18. 52 DM 48.2.16. For similar reasoning expressed in terms of existential dependence, see DM 7.1.18.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

143

argument, then, is to motivate the specific version of non-reductive realism that Suárez endorses. I shall have more to say about how he develops this view in section 5. However, before turning to this, I want to examine Suárez’s reasons for endorsing non-reductive realism as such. These reasons are best appreciated in the context of two methodological arguments that he presents on behalf of his reductive-realist opponents. Let us begin, therefore, by considering these.

4 . T HE F I R S T R O U N D O F A R G U M E N T S

a. Two Methodological Arguments for Reductive Realism i. The Argument from Parsimony. Suárez reports that some proponents of reductive realism argue for their view by appealing to considerations of parsimony. Although he mentions this argument in several contexts, his clearest statement of it is found in DM 48, where he attributes it to Natalis and Javellus. He writes: [They] say that the action of heating, for example, is nothing other than the heat produced by the fire, from which the fire is denominated an agent [ . . . ]. The foundation of this opinion is that this is sufficient for the agent to be constituted an actual agent. Therefore it is also sufficient for the nature of action. Therefore whatever else is posited (fingitur) is superfluous and scarcely intelligible, and therefore should not be granted.53

Suárez’s opponents are here arguing that actions, construed as extramentally distinct properties of effects, are theoretically dispensable. We are able to account for the world’s causal connections without positing any sui generis causal entities, over and above agents and effects. For example, in order to account for the fact that some specific fire produces some specific quality of heat, it is not necessary to posit a sui generis action of heating. Instead, we need only posit the fire and the quality of heat. One way of understanding this view is in terms of the contemporary distinction between “internal” and “external” relations. According to the standard way of drawing this distinction, an internal relation is one that obtains just in virtue of the relata and their intrinsic (or non-relational) properties. One common example is the relation of similarity. Thus, if Simmias and Socrates each have the intrinsic property of whiteness, they will evidently be similar to one another just in virtue of themselves and their respective instances of whiteness. We might say that, in this case, the 53 DM 48.1.9.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

144

Jacob Tuttle

similarity of Simmias and Socrates will obtain automatically, without any addition of being. In contrast, for external relations this is not the case. For example, to account for the fact that Simmias is standing to the left of Socrates, it seems that we cannot simply appeal to Simmias, Socrates, and their intrinsic properties. In this case, it appears that we must posit something further in order to account for the relational fact in question.54 On the account that Suárez attributes to his opponents, efficient causation is something akin to an internal relation. Of course, the opponents’ point is not that efficient causation belongs in the category of relation. Rather, the point is that the existence of an agent and an effect necessitates the causal connection between them, without any addition of being. But if this is right, then non-reductive realism runs afoul of the methodological principle known as “Ockham’s Razor,” which admonishes us not to posit any more entities (or types of entities) than are theoretically required.55 Considerations of parsimony thus favor reductive realism over non-reductive realism. ii. The Argument from Essential Dependence. Another argument alleges that non-reductive realists fail to account for the modal character of creatures’ existential dependence on God. This argument focuses on the common medieval view that creatures are essentially dependent on God for their existence. That is to say, it is an essential attribute of a creature that it be conserved by God at every moment at which it exists. But if this is right, then it is difficult to make sense of the view that God’s actions of conservation are extramentally distinct accidents of creatures. After all, one of the primary reasons for positing accidents would seem to be that they enable us to account for the contingent or non-essential features of their subjects. For example, part of what motivates the view that Socrates’s whiteness is extramentally distinct from him is the fact that he can change from being white to being some other color (if he gets a tan, or becomes jaundiced). Accordingly, if Socrates’s existential dependence on God were an extramentally distinct accident like his whiteness, then in a similar way Socrates could change from depending on God to not depending on him, and his dependence on God would not be essential after all.

54 For a helpful explanation of the distinction between internal and external relations, see John Heil, “Relations,” in R. Le Poidevin (ed.), The Routledge Companion to Metaphysics (New York: Routledge, 2009), 310–21. For discussion of how the contemporary notions of internal and external relations shed light on Suárez’s theory of the category of relation, see Penner, “Suárez on the Reduction of Categorical Relations,” especially section 5.2. 55 For Ockham’s own use of this principle in his ontology, see Marilyn McCord Adams, William Ockham (Notre Dame, IN: University of Notre Dame Press, 1987), 156–61.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

145

It thus looks as if reductive realism is not only the most parsimonious explanation of the data of efficient causation, but also the only explanation of some of that data—namely, the facts about creatures’ essential dependence on God. Consequently, even if non-reductive realism is not rejected outright, it must at least be softened to exclude cases of divine conservation. Moreover, once it is granted that agents and effects are by themselves sufficient to account for some instances of efficient causation, it is difficult to see why this should not be granted for all instances of efficient causation.56

b. Suárez’s Arguments for Non-Reductive Realism i. The Separability Argument. In order to see how Suárez responds to his opponents’ methodological arguments, it will be helpful to begin by considering his own arguments in favor of non-reductive realism. The most prominent of these arguments, and the one that he develops in the most detail, proceeds from the insight that action and effect are separable, in the sense that an effect can exist without the action whereby it is produced. However, Suárez thinks that if action and effect are separable in this way, then action must be an extramentally distinct property of an effect. He presents this argument in its most general form in DM 18, where he writes: [A]n action is not [ . . . ] the very effect produced by an agent, but rather is a mode of such an effect, extramentally distinct from it. An adequate argument for this [thesis] is that an effect can persist without such a mode [ . . . ].57

The argument can be represented formally as follows: Separability Argument (1) If A can exist without B, then A and B must be extramentally distinct. (2) But an effect can exist without the action whereby it is denominated an effect. (3) Therefore, effects and the actions whereby they are denominated effects must be extramentally distinct.

Suárez regards premise (1), which is implicit in the quoted passage, as the best available rule (regula) or method (modus) for showing that things signified by two distinct concepts are themselves distinct in extramental reality.58 Indeed, whenever he approaches the issue of whether entities in two given categories are extramentally distinct, one of the first questions he considers is whether either can exist without the other. This rule would 56 See DM 20.4.8.

57 DM 18.10.8.

58 DM 7.2.2.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

146

Jacob Tuttle

appear to be a truth of reason, for, as he observes, to deny premise (1) apparently commits one to saying that the very same thing can both exist and not exist at the same time. Moreover, he notes that it is overwhelmingly endorsed by various authorities. It is not only taught by Aristotle, but is also the “common opinion of the doctors,” and is widely employed by other philosophers and theologians.59 In defense of premise (2), Suárez adduces two kinds of cases in which he thinks it is clear that an effect can exist apart from any particular action. The first sort of case involves what we might call diachronic separability, whereas the second involves what we might call synchronic separability. Let us consider each type of case in turn. Characterizing the first sort of case, he writes: [A]n effect can now depend on this cause, and later on another; and thus in respect of one and the same effect the dependence in it can be varied (variatur), and so nothing else is varied besides the causality of the agent.60

In order to appreciate the point that Suárez is making here, it is important to recall that he, like other medievals, believes that contingent beings depend on their efficient causes not only in being-made (fieri), or at the first moment of their existence, but also in being (esse), or at every subsequent moment at which they exist. To put the point somewhat differently, Suárez thinks that contingent beings require not only producing causes, but also conserving causes. Now, if this is right, then it is sensible to ask whether a contingent being must be conserved by numerically the same agent throughout its entire career, or whether it may instead depend on different agents at different times. And as it turns out, Suárez thinks that in a wide variety of cases, a single effect need not be conserved by the same agent at every time at which it exists. Suárez’s clearest illustrations of this point are taken from the natural world. One of his favorite examples is the production of light by a lantern. Medieval Aristotelians understand light as a quality that exists in the air. Suárez thinks it is evident that the same quality of light can be conserved in the air by different lanterns at different times. Likewise, he would say that the quality of heat in a pot of boiling water could be sustained by a different fire than the one that first brought it to a boil. As contingent beings, both of these qualities require that some agent or other conserve them, lest they go out of existence. Nevertheless, this plainly need not be the agent that first brought them into existence, and this is sufficient to show that these effects may depend on different efficient causes at different times. From here it is a short argument to premise (2). Suárez thinks that an effect’s depending on different efficient causes at different times can only be

59 DM 7.2.2–3.

60 DM 18.10.8.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

147

explained by its possessing numerically distinct dependencies at those times. But given that these dependencies just are actions, it follows that the same effect can be conserved via numerically distinct actions at different times. Thus, if the quality of heat in some water successively depends on two different fires, this is because numerically distinct actions of heating successively exist in that quality of heat. It is worth emphasizing that Suárez thinks premise (2) holds not only for actions performed by natural agents, but also for the actions by which God creates and conserves creatures. Responding to the Argument from Essential Dependence—which, recall, claims that a creature’s essential dependence on God can only be explained if the creature and God’s action of conservation are identical—he distinguishes two propositions: Essential Dependence 1: Every creature is essentially such that it depends on God via some action or other. Essential Dependence 2: Every creature is essentially such that it depends on God via this particular action.

While Suárez thinks the Catholic faith commits us to Essential Dependence 1, which is thus “absolutely true,” it does not commit us to Essential Dependence 2, which is what is needed to show that the action of divine conservation is identical to its created end-point.61 Moreover, Suárez thinks he can show that creatures and the actions by which God actually creates and conserves them are separable, in much the way that he has shown that effects and the natural actions by which they are conserved are separable. In his discussion of the nature of creation ex nihilo in DM 20.4, he runs the Separability Argument again, this time specifically for God’s creation and conservation of creatures. Here he also defends premise (2) by devising cases involving diachronic separability. In the simplest of these cases, Suárez imagines that God first annihilates and later re-creates the same angel. He argues that although God could certainly ensure that the angel is produced by means of the very same action in both instances, there is no reason to suppose that God must perform the same action. On the contrary, if we accept, as Suárez has already argued, that two created agents can successively cause the same effect via distinct actions, then we should also accept that God could do this on his own. After all, it is a scholastic dictum that God can accomplish by his absolute power whatever any combination of created agents can accomplish by their finite power. So much for cases of diachronic separability. Let us now turn to cases of what we called “synchronic separability.” Suárez imagines such cases in DM 20.4, where he claims that, although God actually does create the world 61 DM 20.4.32.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

148

Jacob Tuttle

without the participation of any other agent, he nevertheless might have created it by using a creature such as an angel as an instrument. But since actions involving instruments are specifically different from those not involving instruments, they must also be numerically distinct, and thus the action by which God actually does create the world is numerically distinct from some action by which he might have created it.62 Finally, Suárez baldly asserts that, in his opinion, it is probable that God could simply decide, by fiat, to create the numerically same effect by performing an action that is specifically the same, but numerically distinct, from the one that he actually does decide to perform.63 It should also be clear at this point how Suárez responds to the Argument from Parsimony. Recall that according to this argument, we can account for the world’s causal connections without positing any actions over and above agents and effects, and hence such actions are dispensable. Suárez denies the opponents’ claim that we can account for the world’s causal connections without positing actions over and above effects. On the contrary, his defense of premise (2) of the Separability Argument makes clear that he thinks it is precisely in order to account for these connections that we must suppose that action and effect are separable in the sense specified. For, if actions and the effects that are their end-points were not separable in this way, then we could not account for the fact that the same effect depends for its being on different agents at different times, nor for the fact that it could have depended on a different agent at the same time. ii. The Argument from Action’s Relational Essence. Although Suárez’s case for non-reductive realism focuses primarily on his Separability Argument, he does sketch another argument in DM 20.4. This argument appeals to the idea that efficient causation or action has a fundamentally different type of nature or essence than do items in the absolute (or non-relational) categories of substance, quantity, and quality. According to Suárez, action differs from the absolute categories to the extent that action stands in an essential relation to an agent and its active power, whereas items in these other categories do not. But if this is correct, then surely action cannot be identified with the entities in these other categories, as Suárez’s opponents insist that it should be.64

62 DM 20.4.13. 63 Ibid. 64 See DM 20.4.14. For a brief discussion of the argument, see Freddoso, Creation, Conservation, and Concurrence, 74–5, note 24. This sort of argument appears to have been prominent in medieval debates about the nature of categorical relations, and so probably would have been well-known to Suárez’s audience. See Brower, “Medieval Theories of Relations.”

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

149

Although Suárez does not spell out his support for this difference in detail, his remarks suggest two reasons for accepting it. First, as he emphasizes repeatedly, action cannot even be conceived apart from an agent and its active power.65 However, the concepts of substance, quantity, and quality evidently do not include any relation to an agent or its power. This is perhaps easiest to appreciate in the case of substance, which is supposed to be unique among the Aristotelian categories because it exists per se, or in itself, whereas accidents exist per aliud, or in another. However, even if we cannot understand the categories of quantity and quality apart from any subject, we still do seem to be able to understand them apart from any agent or active power. Accordingly, Suárez’s point appears to be that this difference between our conception of action and the absolute categories is best explained by a fundamental ontological difference between action and these categories. Suárez’s remarks also suggest that he thinks the difference between action and the absolute categories can be appreciated by reflecting on the metaphysical or functional role of action. As we have seen, he identifies action with the existential dependence of an effect on its efficient cause. But again, whatever this dependence is in itself, it is plainly some sort of relational entity. After all, scholastics posited this sort of dependence precisely in order to account for the connection or link between an efficient cause and its effect. However, Suárez thinks that if we reflect on the account (ratio) of substance, quantity, or quality, we will see that it does not include this sort of relational aspect.66 Now that we have seen Suárez’s case for non-reductive realism, it is worth pausing to reiterate an important aspect of the dialectical situation. In section 3, I argued that Suárez and his opponents each endorse some version of realism about efficient causation, to the extent that they each construe efficient causation as a mind-independent entity. What Suárez and his opponents disagree about is not whether efficient causation exists in extramental reality, but rather what sort of extramental being it is. (This is, after all, why I have referred to their theories “reductive” and “non-reductive realism.”) Not surprisingly, this agreement on the truth of realism about efficient causation is reflected in the arguments we have seen so far. In claiming that their respective theories best account for the facts of efficient causation, Suárez and his opponents are of course taking for granted that there are some such facts. Accordingly, none of the arguments we have seen so far is directed against anti-realism about efficient causation, where this is understood as the view that efficient causation is a fiction or a being of reason. It is not immediately clear how Suárez would respond to an 65 See especially DM 48.1.17.

66 DM 20.4.14.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

150

Jacob Tuttle

opponent who rejects realism about efficient causation outright, but at any rate, that is a topic for a different paper.

5. OBJECTIONS TO SUÁREZ’S N ON -R ED U C T IV E R EA L IS M In the previous section, I outlined what Suárez takes to be the main methodological challenges to his non-reductive theory of efficient causation, and I showed how he attempts to overcome these challenges by appealing, first, to considerations about separability, and second, to considerations about action’s relational nature or essence. However, Suárez also responds to two reductio ad absurdum arguments on behalf of his opponents. Because his responses to these arguments yield important insights into how he understands action’s sui generis nature, I turn to them now.

a. The Vicious Regress Objection One such argument alleges that if we admit into our ontology actions, construed as extramentally distinct properties of effects, we thereby generate a vicious regress. Setting up the alleged regress in the context of creation, Suárez writes: [I]f creation were something in a creature distinct from it, and as it were intermediate between it and God, it would be something created and produced. For it would be distinct from God, and flowing forth ( profluens) from him. Therefore one would have to distinguish in it its creation from itself, and so it would proceed to infinity [ . . . ].67

Although Suárez reports this argument in the context of his discussion of creation, there is nothing about the argument that confines it to this domain. Indeed, he notes that as far as this argument goes, “the same thing can be inferred for any action.”68 The regress requires three assumptions to get going: (1) Every action is extramentally distinct from its end-point. (2) Every action is produced or efficiently caused by the agent that performs it. (3) Whatever is produced or efficiently caused, is produced or efficiently caused in virtue of an action. 67 DM 20.4.7.

68 DM 20.4.31.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

151

Now, consider a given action, A (for example, the action of heating by which a fire produces heat in some water). By (1), A must be extramentally distinct from its end-point, the accident of heat in the water. Moreover, by (2), A is efficiently caused by the fire that performs it. But then, by (3), A must be efficiently caused in virtue of an action, A’. By (1), A’ must be extramentally distinct from its end-point, A. And by (2), A’ is efficiently caused by the agent that performs it, the fire. And by (3), A’ must be efficiently caused in virtue of an action, A’’, and so on in infinitum. Because scholastics generally believe that the simultaneous existence of an actual infinity of real beings is impossible, both Suárez and his opponents would regard this regress as vicious. What the regress evidently shows, then, is that any theory that includes the conjunction of (1)–(3) is untenable. Suárez’s reductive-realist opponents will, of course, advertise the fact that they can escape the regress by denying (1). According to them, actions are all identical to their end-points, so there is no room for the regress to get started. For Suárez, however, escape is more difficult. What makes this argument so powerful is that the regress is generated from claims that Suárez himself explicitly endorses, and in some cases emphatically so. As we have seen, he thinks the facts of efficient causation require the existence of actions that are extramentally distinct from their end-points. Thus, (1) is a nonnegotiable commitment. Likewise, (3) is simply a restatement of Suárez’s thesis that action just is efficient causation, or the existential dependence of an effect on its efficient cause. Because this thesis is one of the cornerstones of Suárez’s theory of efficient causation, it too is non-negotiable. Finally, although there might appear to be a bit more room for him to reject (2), in certain passages Suárez explicitly endorses it. For example, in DM 48, he writes: [An action’s] efficient cause is that principle from which it flows.69

It is clear from the context that the “principle” he has in mind is the agent that performs the action. Although he is not completely forthcoming about why he endorses this claim, two considerations seem relevant. First, it would seem to be part of the concept of an action that its occurrence is to be explained by reference to the agent that performs it. For example, if Polycletus is singing, it certainly looks as if this singing is to be explained by reference to Polycletus. It is precisely because Polycletus performs this action that it occurs. This is why Suárez frequently describes action as flowing from, arising from, or depending on, an agent. However, this explanatory connection is, at least prima facie, very similar to causation. 69 DM 48.5.2.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

152

Jacob Tuttle

Another reason for endorsing (2) has to do with medieval views about contingent beings. As noted already, medievals generally agree that contingent beings require some conserving cause as long as they exist. Since actions are contingent beings, they too require efficient causes throughout their careers. But, granted that actions must have efficient causes, it is difficult to see what these causes could be, if not the agents that perform these actions. In his response to the Regress Objection, Suárez highlights an ambiguity in assumption (2). In some passages, such as the one quoted earlier, he makes it sound as if he accepts without qualification that actions are efficiently caused by the agents that perform them. However, his considered opinion turns out to be more nuanced than such passages suggest. In fact, when he is being most careful, he always qualifies the claim that actions are products or effects of the agents that perform them. For example, defending himself against the Regress Argument in DM 18, he writes: [I]f by the term “effect” (effectus) we understand not only the thing produced, but whatever flows from the power of an agent, then we concede that action is in some way an effect of an agent, since it is dependent [on an agent]—or better, just is the very dependence [of an effect on an agent].70

Suárez is willing to grant that an action is an effect of an agent, if the term “effect” is taken in what he later calls the “broad way” (lato modo), so that anything whatever that flows from or depends on an agent counts as its effect. However, he urges that it does not follow from this that actions flow from or depend on their agents via other actions. On the contrary, he thinks that an action’s role as the efficient-causal connection between agent and effect requires that it not be related to an agent in the same way as this effect is. After noting that every type of causality is, in the broad sense, an effect of the cause it denominates, Suárez writes: And neither is it necessary for that reason that a causality be caused through another causality, or that an action be performed (fiat) through another action, because it is of the nature (ratione) of a causality that it have an immediate relation (habitudinem) to a cause, and that it be as it were a medium or link (vinculum) between a cause and an effect.71

Suárez’s point here is that insofar as action is what connects efficient causes and their effects, it must stand in what he calls an “immediate” relation to a cause—where what he means by an “immediate” relation is one that a thing stands in just in virtue of itself. The consequences of denying this are, he thinks, dramatically brought home by the regress outlined earlier. Thus, Suárez thinks that in advancing the Regress Objection, his reductive-realist 70 DM 18.10.8. Italics are mine.

71 DM 18.10.8.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

153

opponents are suffering from confusion about the theoretical role that a causal connection is posited to perform. That is to say, he is accusing his opponents of what we would nowadays call a “category mistake.” To sum up his response, Suárez thinks the regress can be avoided by rejecting the second assumption that gives rise to it. On the analysis he has given, (2) is true only if “produces” and “efficiently causes” are understood so broadly that entities that depend on agents immediately, without any intervening actions, count as effects of those agents. Although Suárez is himself willing to talk this way in some contexts, he also thinks that this way of understanding efficient causation can be misleading, because it obscures the fundamental ontological differences between actions and the non-causal entities that serve as their end-points. Alternatively, if an opponent insists on understanding efficient causation in this overly broad sense, then Suárez will say that assumption (3) is false. On his view, only what is produced or efficiently caused in the strict sense requires the mediation of an action.

b. The Vicious Circle Objection Another argument exploits an apparent tension between Suárez’s view that action is an extramentally distinct, accidental feature of its end-point, and his view that action just is efficient causation. Although his presentation of this argument is compressed, the thought appears to be that, taken together, these two claims generate a vicious circle of existential priority.72 This can be appreciated by considering the relationships of dependence between a subject and its accident, on the one hand, and efficient causation and its end-point, on the other. As we have seen, medieval Aristotelians take the characteristic feature of accidents to be their existence in a subject. Indeed, this is precisely the feature by which accidents are supposed to be distinguished from substances. Moreover, and most importantly for our purposes here, medievals typically understand accidents to be existentially dependent on the subjects in which they exist. This is, after all, the reason why they often refer to an accident’s subject as its material cause. Thus, for example, Socrates’s accident of whiteness depends on Socrates for its existence, and accordingly Socrates is said to be its material cause.73 However, it also appears that an effect depends for its existence on the action whereby it is produced, since it is precisely in virtue of this action that the effect exists. As Suárez sometimes puts the point, an action is a 72 DM 20.4.4. Suárez also clarifies the argument in the course of responding to it; see DM 20.4.25–26. 73 For Suárez’s arguments for the thesis that accidents are existentially dependent on their subjects or material causes, see DM 14.1.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

154

Jacob Tuttle

path (via) to its end-point, and accordingly this end-point “has no real being that has not come to it through that path.”74 It thus looks as if, on Suárez’s account, action will be existentially prior to effect (insofar as it is that by virtue of which the effect is produced), and effect will be existentially prior to action (insofar as it is the subject or material cause of the action). And this sort of circularity certainly appears vicious, for if the existence of each is a necessary condition for the existence of the other, neither could ever begin to exist. Suárez’s way out of the circle is to deny that actions are, strictly speaking, accidents of the effects that serve as their end-points. He writes: [N]o dependence or being made, as such, has the character of an accident in respect of the end-point at which it aims (tendunt) [ . . . ].75

He sometimes puts the same point in a different idiom, noting that action does not inhere in its end-point, or that it does not exist in its end-point “as in a subject.” Although this response does appear to avoid the circle, it suggests a further question about exactly how action is supposed to be related to its end-point. After all, as we have seen, Suárez insists that action exists in an effect, and that it is a certain sort of property of that effect— namely, a mode of it. But once he has denied that this relationship can be explained in terms of the inherence of an accident in a subject, it is not immediately clear how we are to understand it. In response to this question, Suárez insists that action is related to its endpoint via a distinctive or sui generis type of property-possession. He claims that action exists in its end-point “not as in a subject, but in keeping with the special (specialis) and distinctive (propria) relation by which a path is related to its end-point.”76 He elsewhere describes this relation as “a peculiar (peculiaris) way of establishing the end-point in reality, [as] depending on and flowing from its principle.”77 In order to appreciate what Suárez thinks is distinctive about this relation, it will be helpful to briefly compare it to an accident’s inherence in its subject. As it turns out, inherence plays two different roles in Suárez’s property theory. As we have seen, one of these roles is to account for an accident’s existential dependence on its subject. To reiterate the example used earlier, it is in virtue of inhering in Socrates that Socrates’s whiteness depends on him for its existence. And perhaps not surprisingly, the other role of inherence is to account for an accident’s intrinsic characterization of its subject. Thus, according to Suárez, it is also in virtue of inhering in

74 DM 20.4.25. 76 DM 20.4.27.

75 DM 32.1.17. 77 DM 48.4.13.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

155

Socrates that Socrates’s whiteness intrinsically affects (afficit) or denominates him as being white.78 Now, it should be clear from what we have seen so far that Suárez also takes the relation between an action and its end-point to involve intrinsic characterization. After all, it is precisely in order to account for intrinsic differences in an effect’s dependence on an agent that Suárez thinks we must posit a sui generis type of causal entity. In this way, the relation between action and end-point is similar to an accident’s inherence in its subject. However, Suárez insists that, unlike the relation of inherence, action’s relation to its end-point does not render action existentially dependent on that end-point. He writes: Regarding this [relation], there is not required a priority of nature on the part of the end-point, because it does not require true [material] causality, but only termination (terminationem).79

That is to say, the relation between action and end-point is completely onesided: it involves characterization without material causation. Suárez’s description of the relation between an action and its end-point appears to be consistent. After all, there is no obvious reason to think that a property’s intrinsic characterization of a subject entails its existential dependence on that subject. Even so, the fact that Suárez’s theory of efficient causation requires him to appeal to a sui generis type of property-possession should be regarded as a significant cost of the theory.

6 . C O N C LU S I O N My primary aim in this paper has been to clarify the details of Suárez’s account of the nature of efficient causation or action. As mentioned in the introduction, this project has significant historical value, to the extent that Suárez’s views on this topic have not been adequately explored by scholars. However, I also hope that my interpretation has gone some way toward persuading readers of the intrinsic philosophical interest of Suárez’s theory of efficient causation. Since the time of Hume, philosophers in the Anglophone world have tended to assume that any adequate theory of causation must be reductive, in the sense that it must account for facts about causation by appealing to ordinary, non-causal facts. In my view, this assumption is far from obvious. Suárez’s theory of efficient causation may or may not 78 For a useful overview of relations of union or inherence in scholastic philosophy, see Pasnau, Metaphysical Themes, ch. 11. 79 DM 20.4.27.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

156

Jacob Tuttle

ultimately succeed; but in any case, it is one example of a sophisticated, nonreductive account of causation, and it would be a mistake to discard it in the dustbin of history without careful scrutiny. Loyola Marymount University

BIBLIOGRAPHY Adams, Marilyn McCord. William Ockham, 2 vols. (Notre Dame, IN: University of Notre Dame Press, 1987). Aristotle. The Complete Works of Aristotle: The Revised Oxford Translation, 2 vols., ed. J. Barnes (Princeton: Princeton University Press, 1984). Brower, Jeffrey. “Medieval Theories of Relations,” in E. N. Zalta (ed.), The Stanford Encyclopedia of Philosophy, (Winter 2013 Edition) . Burns, Patout. “Action in Suarez,” The New Scholasticism 38 (1964), 453–72. David Hume. A Treatise of Human Nature, ed. D. F. Norton and M. J. Norton (Oxford: Clarendon Press, 2007). Des Chene, Dennis. Physiologia: Natural Philosophy in Late Aristotelian and Cartesian Thought (Ithaca, NY: Cornell University Press, 1998). Des Chene, Dennis. “Suárez on Propinquity and the Efficient Cause,” in B. Hill and H. Lagerlund (eds), The Philosophy of Francisco Suárez (New York: Oxford University Press, 2012), 89–100. Doyle, John. “Suárez, Francisco,” in E. Craig (ed.), Routledge Encyclopedia of Philosophy (London: Routledge, 1998) . Francisco Suárez. Opera omnia, 28 vols., ed. C. Berton (Paris: Vivès, 1856–61). Francisco Suárez. On the Various Kinds of Distinctions, tr. C. Vollert (Milwaukee: Marquette University Press, 1947). Francisco Suárez. Disputationes metaphysicae, 2 vols. (Hildesheim: Georg Olms, 1965). Francisco Suárez. Suárez on Individuation: Metaphysical Disputation V: Individual Unity and its Principle, tr. J. J. E. Gracia (Milwaukee: Marquette University Press, 1982). Francisco Suárez. On Efficient Causality: Metaphysical Disputations 17, 18, and 19, tr. A. J. Freddoso (New Haven, CT: Yale University Press, 1994). Francisco Suárez. Creation, Conservation, and Concurrence: Metaphysical Disputations 20–22, tr. A. J. Freddoso (South Bend, IN: St. Augustine’s Press, 2002).

I presented early versions of this paper at Purdue University, Franciscan University of Steubenville, and Loyola University Chicago. I would like to thank the audiences at these venues for their helpful questions and suggestions. I would also like to thank Michael Bergmann, Susan Brower-Toland, Jan Cover, Peter Hartman, John Kronen, Robert Pasnau, an anonymous referee for Oxford Studies, and especially Jeffrey Brower, for their comments on various drafts of the paper. Finally, I am grateful to the Purdue Research Foundation, which provided two years of financial support for my dissertation, from which this paper is partly taken.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Suárez’s Non-Reductive Theory of Efficient Causation

157

Francisco Suárez. A Commentary on Aristotle’s Metaphysics: Index locupletissimus in Metaphysicam Aristotelis, tr. J. Doyle (Milwaukee, WI: Marquette University Press, 2004). Freddoso, Alfred J. “God’s General Concurrence with Secondary Causes: Why Conservation Is Not Enough,” Philosophical Perspectives 5 (1991), 553–85. Freddoso, Alfred J. “God’s General Concurrence with Secondary Causes: Pitfalls and Prospects,” American Catholic Philosophical Quarterly 67 (1994), 131–56. Gorham, Geoffrey. “Cartesian Causation: Continuous, Instantaneous, Overdetermined,” Journal of the History of Philosophy 42 (2004), 389–423. Gracia, Jorge J. E. “Francisco Suárez: The Man in History,” American Catholic Philosophical Quarterly 65 (1991), 259–66. Hansen, Heine. “Strange Finds, or Nicholas of Paris on Relations,” in J. L. Fink, H. Hansen, and A. M. Mora-Marquez (eds), Logic and Language in the Middle Ages: A Volume in Honor of Sten Ebbesen (Leiden: Brill, 2012), 139–54. Hattab, Helen. “Conflicting Causalities: The Jesuits, their Opponents, and Descartes on the Causality of the Efficient Cause,” in D. Garber and S. Nadler (eds), Oxford Studies in Early Modern Philosophy, vol. 1 (Oxford: Clarendon Press, 2003), 1–22. Heil, John. “Relations,” in R. Le Poidevin (ed.), The Routledge Companion to Metaphysics (New York: Routledge, 2009), 310–21. Henninger, Mark. Relations: Medieval Theories 1250–1325 (New York: Oxford University Press, 1989). Hill, Benjamin and Henrik Lagerlund (eds). The Philosophy of Francisco Suárez (New York: Oxford University Press, 2012). Kukkonen, Taneli. “Creation and Causation,” in R. Pasnau and C. Van Dyke (eds), The Cambridge History of Medieval Philosophy, 2 vols. (Cambridge: Cambridge University Press, 2010), 232–46. Marmodoro, Anna. “The Union of Cause and Effect in Aristotle: Physics III 3,” Oxford Studies in Ancient Philosophy 32 (2007), 205–32. Menn, Stephen. “Suárez, Nominalism, and Modes,” in K. White (ed.), Hispanic Philosophy in the Age of Discovery (Washington, DC: Catholic University of America Press, 1997), 226–56. Nolan, Peter. “The Suárezian Modes,” Proceedings of the Annual Convention of the Jesuit Educational Association, Midwest Division (1931), 184–200. O’Neill, Eileen. “Influxus Physicus,” in S. Nadler (ed.), Causation in Early Modern Philosophy: Cartesianism, Occasionalism, and Preestablished Harmony (University Park, PA: Pennsylvania State University Press, 1993), 27–56. Ott, Walter. Causation and Laws of Nature in Early Modern Philosophy (New York: Oxford University Press, 2009). Pasnau, Robert. Metaphysical Themes 1274–1671 (Oxford: Clarendon Press, 2011). Penner, Sydney. “Francisco Suárez on Acting for the Sake of an Ultimate End” (Ph.D. Dissertation, Cornell University, 2011). Penner, Sydney. “Suárez on the Reduction of Categorical Relations,” Philosophers’ Imprint 13 (2013), 1–24. Pessin, Andrew. “Descartes’s Nomic Concurrentism: Finite Causation and Divine Concurrence,” Journal of the History of Philosophy 41 (2003), 25–49.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

158

Jacob Tuttle

Pini, Giorgio. Categories and Logic in Duns Scotus: An Interpretation of Aristotle’s Categories in the Late Thirteenth Century (Leiden: Brill, 2002). Rota, Michael. “Causation in Contemporary Metaphysics and in the Thought of Thomas Aquinas” (Ph.D. Dissertation, St. Louis University, 2006). Schmaltz, Tad. Descartes on Causation (New York: Oxford University Press, 2008). Schwartz, Daniel (ed.). Interpreting Suárez (New York: Cambridge University Press, 2012). Simpson, D. P. (ed.). Cassell’s Latin Dictionary (New York: MacMillan Publishing Company, 1968). Thomas Aquinas. Opera omnia S. Thomae, Corpus Thomisticum . Tuttle, Jacob. “Suárez’s Metaphysics of Efficient Causation” (Ph.D. Dissertation, Purdue University, 2013).

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

How Not to Be a Truthmaker Maximalist Francisco Peinado on Truthmakers for Negative Truths Brian Embry

Much of recent metaphysical debate hinges on the notion of a truthmaker. According to David Armstrong’s characterization, a truthmaker for a true proposition is “some existent, some portion of reality, in virtue of which that truth is true.”1 For example, Lucy is the truthmaker for the proposition that Lucy exists. And the fact that Lucy is running (if there is such a thing) might be the truthmaker for the proposition that Lucy is running. The idea of a truthmaker is often used to “regiment metaphysical inquiry.”2 The thought is that asking what makes true propositions true is a good guide to ontological commitment. Perhaps the most trenchant problem in contemporary truthmaker theory is the problem of negative truths: it is hard to see what could make negative truths true, and intuitively, such truths do not even need truthmakers. For example, it is hard to see what portion of reality could make it true that Pegasus does not exist, and intuitively, “Pegasus does not exist” is true because Pegasus does not exist, not because something else does exist. Very roughly, there are three approaches to the problem of negative truths in the contemporary literature. Some philosophers posit exotic entities like totality facts and negative facts as truthmakers for negative truths;3 others 1 D. M. Armstrong, Truth and Truthmakers (Cambridge: Cambridge University Press, 2004), 5. 2 Armstrong, Truth and Truthmakers, 4. 3 Armstrong posits totality facts in Truth and Truthmakers; Stephen Barker and Mark Jago posit negative facts in “Being Positive about Negative Facts,” Philosophy and Phenomenological Research 85 (2012), 117–38.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

160

Brian Embry

argue that more mundane entities, like the world, are truthmakers for negative truths;4 others reject the assumption that generates the problem in the first place—the assumption that negative truths have truthmakers.5 Iberian and Italian scholastics of the seventeenth century also had the notion of a truthmaker (“verificativum” ), and they engaged in sophisticated debates about truthmakers for various problematic truths, including modal truths, tensed truths, and negative truths.6 In response to the problem of negative truths, seventeenth-century scholastics carved out roughly the same conceptual space that contemporary philosophers have carved out, typically endorsing one of three positions: (i) negative truths are made true by negative entities called “lacks”; (ii) negative truths are made true by positive entities, including divine decrees; (iii) negative truths do not have truthmakers.7 While the conceptual space was broadly the same then as it is now, the details were developed quite differently. In this paper I explain how the third solution to the problem of negative truths was developed by Francisco Peinado (1633–1696). In the contemporary literature the fully general truthmaker principle— “all truths have truthmakers”—goes by the name “Truthmaker Maximalism.” I will argue that Peinado effectively restricts the truthmaker principle to positive truths, thereby rejecting Truthmaker Maximalism. But the position that I attribute to Peinado is widely rejected in the contemporary literature. George Molnar calls restriction of the truthmaker principle “the

4 Ross Cameron, “How To Be a Truthmaker Maximalist,” Noûs 42 (2008), 410–21; Jonathan Schaffer, “The Least Discerning and Most Promiscuous Truthmaker,” The Philosophical Quarterly 60 (2010), 307–24. 5 Kevin Mulligan, Peter Simons, and Barry Smith, “Truth-Makers,” Philosophy and Phenomenological Research 44 (1984), 289, 314ff. 6 On the advice of Sven Knebel, John Doyle was the first to translate “verificativum” as “truthmaker.” See Doyle, “Another God, Chimerae, Goat-Stags, and Man-Lions,” Review of Metaphysics 48 (1995), 773 n. 12. For more detailed discussions of truthmaking in seventeenth-century scholasticism, see Jacob Schmutz “Verificativum,” in D. Calma, et al. (eds), Mots médiévaux offerts à Ruedi Imbach (Turnhout: Brepols, 2011), 739–48; Brian Embry, “Truth and Truthmakers in Seventeenth-Century Scholasticism,” Journal of the American Philosophical Association 1 (2015), 196–216. 7 For discussion of this debate, see Tilman Ramelow, Gott, Freiheit, Weltenwahl: Die Metaphysik der Willensfreiheit zwischen Antonio Pérez S.J. (1599–1649) und G.W. Leibniz (1649–1716) (Leiden: Brill, 1997), 230–50; Sven Knebel, Wille, Würfel und Wahrscheinlichkeit: Das System der moralischen Notwendigkeit in der Jesuitenscholastik, 1550–1700 (Hamburg: Felix Meiner Verlag, 2000), 175–9; Jacob Schmutz, “Réalistes, nihilistes et incompatibilistes: le débat sur les negative truthmakers dans la scolastique Jésuite Espagnole,” Cahier de philosophie de l’Université de Caen [Special issue titled “Dire le Néant”] (2007), 131–78. For a detailed exposition of the first position, see Brian Embry, “An Early Modern Scholastic Theory of Negative Entities: Thomas Compton Carleton on Lacks, Negations, and Privations,” British Journal for the History of Philosophy 23 (2015), 22–45.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

How Not to Be a Truthmaker Maximalist

161

way of ontological frivolousness” and “truly desperate.”8 David Armstrong asks incredulously, “But why should one who accepts truthmakers baulk at truthmakers for negative truths? Are we not giving up when the ontological going gets a bit tough?”9 Trenton Merricks puts a finer point on the objection. Writing “Truthmaker” for a fully general truthmaker principle, he says: The only reason to scale back Truthmaker to exempt negative existentials is that there do not seem to be truthmakers for negative existentials. Truthmaker theorists who proceed in this way have no principled objection to the cheater who, when confronted with her own apparently truthmakerless truths, scales back Truthmaker accordingly. For this cheater is simply adopting the strategy of the timid truthmaker theorist, concluding that since there do not seem to be any truthmakers for a certain kind of truth, none are required.10

Ross Cameron writes in a similar vein: Either there’s something wrong with accepting truths that don’t have an ontological grounding or there isn’t: if there is, then every truth requires a grounding; if there isn’t, then no truth requires a grounding. Truthmaker theory is a theory about what it is for a proposition to be true; it’s not the kind of theory that can apply only in a restricted domain. What possible reason could one have for thinking of some propositions that they need to be grounded in what there is that doesn’t apply to all propositions? Why should it be okay for negative truths to go ungrounded and not okay for positive truths to go ungrounded?11

The objection in these passages seems to be that restricting the truthmaker principle is ad hoc. What is needed is an independent reason to think that positive truths are grounded in reality, while negative truths are not. My aim in this paper is to show (i) how Peinado restricts the truthmaker principle to positive truths, and (ii) how Peinado would answer the objection from Merricks and Cameron. I will argue that Peinado provides independent reasons to think not only that negative truths do not have truthmakers, but that negative truths cannot have truthmakers. The paper proceeds as follows. In section 2 I sketch a prevalent seventeenth-century theory of truth and truthmaking that is in the background of Peinado’s discussion. This stage-setting will help to make sense of Peinado’s positive view as well as the objections he considers likely to arise against his view. In sections 3–4 I explain Peinado’s view of negative truths and the reasons he 8 George Molnar, “Truthmakers for Negative Truths,” Australasian Journal of Philosophy 78 (2000), 85. 9 Armstrong, Truth and Truthmakers, 55. 10 Trenton Merricks, Truth and Ontology (Oxford: Clarendon Press, 2007), 40. 11 Cameron, “How Not To Be a Truthmaker Maximalist,” 411–12.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

162

Brian Embry

gives for restricting the truthmaker principle to positive truths. In section 5 I explain how he responds (or could respond) to select objections. As we will see, the claim that negative truths do not have truthmakers does not perfectly capture Peinado’s position, but it is a good first pass, to be modified later on in light of some problematic texts. For simplicity I will proceed through the next two sections as if Peinado’s position is simply that negative truths do not have truthmakers. I will then introduce and accommodate the textual difficulty in section 4. It is also important to note that here I focus on negations of atomic propositions, entirely setting aside the problem of finding truthmakers for universally quantified truths. The reason for this choice is that scholastics did not treat quantifiers the way we typically do, and their treatment deserves a discussion of its own. However, the scholastics did treat existence as a predicate, so a negative existential like ‘Peter does not exist’ counts as a negation of an atomic proposition and will therefore be of interest here.

1 . BA C K G R O U N D That early modern scholastics had a notion of truthmaking at least approximating the contemporary notion is clear from their characterizations of truthmaking.12 Here are three examples (with “truthmaker” translating “verificativum” ): The immediate and formal truthmaker of a proposition is that by which a proposition is immediately and formally rendered true. (Francisco Peinado, 1633–1696)13 You have to say what it is for a proposition to be true and what is required in reality on the part of the object for the proposition to be true. The latter I call a truthmaker. (Giovannbattista Giattini, 1651)14 In all propositions some existence, or the ultimate actuality that is the truthmaker of the proposition, is affirmed. (Silvestro Mauro, 1619–1687)15

These passages suggest that for early modern scholastics a truthmaker is a portion of reality in virtue of which a truth-bearer is true. 12 The material in this section follows the more thorough discussion in Brian Embry, “Truth and Truthmakers in Early Modern Scholasticism,” Journal of the American Philosophical Association 1 (2015), 196–216. 13 Peinado, Disputationes in octos libros physicorum, lib. 1, d. 7, s. 2, nn. 12, p. 177. All translations are my own. Seventeenth-century works are cited according to their original divisions. 14 Giattini, Logica, q. 6, a. 5, p. 292. 15 Mauro, Quaestionum philosophicarum liber secundus, 170 (ad 5).

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

How Not to Be a Truthmaker Maximalist

163

The theory of truthmaking in the background of Peinado’s discussion arises from a prevalent answer to the seventeenth-century scholastic disputation, “What is truth?” Before proceeding it will be useful to clarify what exactly this question about truth amounts to. “What is truth?” is naturally understood in one of two ways. First, it might be asking for an analysis of what it is to be true. We could call this the “analytic question.” An answer to the analytic question would fill in the right-hand side of the schema: x is true =df –––. As a point of comparison, an analytic question is what philosophers typically have in mind when they ask, “What is knowledge?,” and it is what Socrates has in mind when he asks Euthyphro, “What is piety?” When the early modern scholastics ask “What is truth?,” however, they are not primarily interested in the analytic question. Rather, they are interested in what I call the “metaphysical question.” The metaphysical question asks what sort of thing truth is. Another way to put this question is to ask, in what ontological category does truth belong? Is truth a relation, a property, a being of reason, or something else altogether? Because early modern scholastics were primarily concerned with the metaphysical question, I will begin with the metaphysical question, and then I will show how a prevalent answer to the metaphysical question entails an answer to the analytic question. Early modern scholastics standardly conceive of truth-bearers as token mental acts.16 These mental acts go by various names, including “proposition,” “judgment,” “affirmation,” “cognition,” “mental act,” and sometimes just “act.” (Some of these terms can be seen in quotations throughout the paper.) Going forward I will arbitrarily adopt the term “proposition” for the relevant truth-bearing mental acts. But propositions in this sense are not abstract entities that provide semantic values for sentences. In the present context it is best to think of propositions as mental sentences. I will accordingly set off propositions with single quotation marks. The “content” of a proposition is its intentional object. Many seventeenth-century scholastics use the language of parts and wholes to describe truth. The common view seems to have been that the truth of a proposition p is the mereological sum of p and its intentional object. Giuseppe Polizzi describes this sort of view in the following passage. 16 Gabriel Vázquez held the minority view that truth-bearers are something like the contents of mental acts. Suárez rejects Vázquez’s view, and subsequently Suárez’s view becomes the mainstream Jesuit position. See Vázquez, Commentariorum ac disputationum in primam partem Sancti Thomae tomus primus d. 76, ch. 1, pp. 452ff. Suárez rejects Vázquez’s view in Disputationes metaphysicae (Hereafter DM) 8.1 (XXV, 275–7). The parenthetical reference indicates the volume and page number of the Vivès edition of Suárez’s Opera.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

164

Brian Embry

(Polizzi is an opponent of the view, but his description is useful on account of being especially straightforward.) The first conclusion is that the object is not a part intrinsically composing formal truth and falsity. This is against Arriaga . . . , Oviedo . . . and other contemporaries who think that truth is composed of two intrinsic parts, one of which is intrinsic to the proposition—indeed, it is the proposition itself—insofar as it affirms, for example, Peter’s running, and the other part is extrinsic to the proposition but intrinsic to the truth, and it is the object existing in the way in which it is affirmed by the proposition.17

According to this passage, many seventeenth-century scholastics thought that truth is a composite entity. The parts of an instance of truth are (i) a true proposition, and (ii) that proposition’s intentional object. As far as I have seen, seventeenth-century scholastics do not provide semantics for determining the intentional objects of propositions. Most often they give examples. So for example, they say that the object of the proposition that Peter is running is Peter’s running. Generally, the object of a proposition of the form “a is F” is a’s Fness. So suppose that “Peter is running” is about Peter’s running. On the view described by Polizzi, if Peter’s running exists, then the mereological sum of “Peter is running” and Peter’s running exists. Hence, it is true that Peter is running. It should be noted that early modern scholastics tend to assume that we can think about things that do not exist. I can think about Peter’s running even if Peter’s running does not exist. In that case, the proposition that Peter is running would be false. The assumption that we can think about non-existent objects is controversial in the contemporary literature and may strike some as needing explication. Briefly, early modern scholastics standardly say that intentional objects that do not exist nonetheless have “objective being in the intellect.”18 For simplicity we can think of “objective being in the intellect” as delimiting a Meinongian third realm. But it should 17 Polizzi, Philosophicarum disputationum tomus tertius, d. 25, s. 4, n. 33, p. 366. 18 This is a standard line that can be found in many disputations on beings of reason. It can also be seen in this colorful passage by Arriaga: “First, it cannot be denied that some acts are false and others are true. Second, it cannot be denied that false acts are false because their objects do not exist in reality as they are affirmed by the acts, whereas true acts are true because their objects are as they are affirmed to be by the acts. These two claims are really most certain and cannot be denied by anyone. From these claims I infer, therefore, that there are some acts whose objects do not exist in reality. I will give you an example: ‘A horse is rational.’ The object of this act is a horse identified with rational, or a rational horse. But this horse does not exist in reality. Who would deny that? But it has being [est] through the intellect—that is, it is cognized by the intellect. . . . Therefore the rational horse does not have being in reality, but only objective being in the intellect. (I really do not think this is denied by anyone, nor can it be)” (Cursus philosophicus, Metaph., d. 6, s. 2, n. 8, p. 783).

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

How Not to Be a Truthmaker Maximalist

165

be noted that there were further debates about the ontological status of the third realm. Very roughly, some thought that objective being is reducible to some existing mental item, such as a token thought, and others denied that objective being is so reducible.19 I do not want to enter that debate here, but I hope these remarks will help contemporary readers entertain the conceptual framework that supports the early modern scholastic debates about truthmaking. I’ll call the view described above by Polizzi “the composite theory of truth.” As I understand it, the composite theory can be formulated as follows: The Composite Theory of Truth: The truth of a proposition p = the mereological sum of p and the intentional object of p.

Given the above remarks about objective being, a worry now arises:20 if merely objective beings can enter into composition relations, then every proposition will be trivially true just in virtue of having an intentional object. As far as I have seen, early modern scholastics did not explicitly address this worry. But one way to block the worry is to restrict composition to things that exist. This move is in keeping with the way the composite theory is often formulated in terms of propositions and their existing objects. (See the earlier Polizzi quotation and the first Arriaga quotation below.) At any rate, the composite theory of truth appears to have been standard among seventeenth-century Jesuits, as suggested by the following remark from Luis de Losada: The common opinion in our School is that truth consists in a complex of an act and an object.21

On the plausible assumption that the parts of truth are essential parts,22 the composite view of truth entails what I call the standard analysis of truth: Standard Analysis of Truth: p is true =df p exists and the intentional object of p exists.

19 I take it this is what Suárez is concerned with at DM 54.2.3–4 (XXVI, 1019). 20 Thanks to a reviewer for pressing this worry. 21 Losada, Cursus philosophici prima pars, t. 5, d. 2, ch. 1, n. 2 p. 231 (erroneously marked as p. 131). See also Mangold, Philosophia recentior, vol. 1, Ontology, diss. 4, a. 1, s. 4, n. 98, p. 96. 22 This idea is implicit in the quotation from Giattini cited earlier. In one place Polizzi claims that the parts of truth, on the composite theory, are like the body and soul of a human being, precisely because body and soul are not only parts but essential parts of a human being (Polizzi, Philosophicarum disputationum tomus tertius, d. 25, s. 5, n. 49, p. 369).

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

166

Brian Embry

The standard analysis is often endorsed explicitly. To give just one passage, Roderigo de Arriaga writes: For a proposition to be true is nothing other than for it to affirm [for example] Peter’s running, and for this to exist in reality. But this implies two things: the proposition and the existence of the object [my emphasis].23

It is not hard to see why early modern scholastics would have been attracted to the standard analysis of truth. Most early modern scholastics pay tribute to tradition by endorsing a conformity theory of truth. However, there was disagreement about how to understand the conformity theory. As Sebastián Izquierdo (1601–1681) tells us, “Both ancient and modern philosophers think that truth must be conformity between a cognition and the thing cognized, even if they are not in conformity in explaining what such a conformity amounts to.”24 The disagreement arises because the claim that truth is conformity is not very illuminating. The standard analysis dispenses with the mysterious notion of conformity and analyzes truth in terms of (i) intentionality and (ii) existence. Existence is arguably primitive, admitting of no further analysis.25 And while it may be difficult to explain how intentionality works, most philosophers will accept the occurrence of intentionality, and it is therefore available in most philosophical toolkits. At any rate, if intentionality is a mystery, it is everyone’s mystery, not only the conformity theorist’s. The standard analysis of truth therefore improves on the conformity theory by reducing conformity to existence and intentionality. The composite theory and the standard analysis of truth drive the account of truthmaking that is in the background of Peinado’s discussion of truthmakers for negative truths. Many early modern scholastics speak of truthmakers both as parts of truth and as intentional objects of propositions. So the idea seems to be that a proposition’s intentional object is that proposition’s truthmaker because the proposition is true just in case that object exists. The standard analysis of truth is fully general, so it entails that every truth has a truthmaker. Several seventeenth-century scholastics explicitly commit themselves to a fully general truthmaker principle. To reiterate the earlier quotation from Mauro, “in all propositions some existence, or the ultimate actuality that is the truthmaker of the proposition, is affirmed”

23 Arriaga, Cursus philosophicus, Logic, d. 14, s. 1, subs. 5, n. 27, p. 170. 24 Izquierdo, Pharus scientiarum, t. 2, d. 3, q. 1, n. 3, p. 110. 25 There was actually a debate about whether being is primitive. For the affirmative, see Giattini, Logica, q. 6, a. 2, pp. 259ff.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

How Not to Be a Truthmaker Maximalist

167

(my emphasis).26 And the author of an anonymous disputation on truthmaking tells us, “every act that is strictly true in the present ought to have a truthmaker strictly existing in the present.”27 The fully general truthmaker principle was often used to motivate realism about negative entities called “negations” and “lacks.” For example, Giovannbattista Giattini constructs the following truthmaker argument for the existence of negations: It seems like you have to say that there are negations in reality. . . . This conclusion is proven because there must be a formal truthmaker for a negative proposition—for example, for the proposition by which I say “Light is not in the air.” But this truthmaker is not something positive. Therefore, it is a negation, distinct from everything positive.28

Although he does not use the term “truthmaker” in the following passage, Thomas Compton Carleton’s realism about negative entities is clearly motivated by the idea that truths have truthmakers. He writes: That the man is dead, that the fire is extinguished, that Peter is not sitting and is not reading, that Paul is blind, and six hundred other such propositions, are no less true, really and mind-independently, than that the man lives, that Peter sees, &c. Therefore there must be something in reality from which the denomination [‘true’] comes to these propositions. But that could not be anything except negations.29

As we will see, Peinado wants to excise negations from his ontology, and he wants to resist these truthmaker arguments for the existence of negations.

2. PEINADO ON NEGATIVE TRUTHS Peinado’s view of negative truths is largely driven by a distinctive view of truthbearers, which emerges in a debate about whether contradictory propositions p and ¬p have the same intentional object. Peinado answers this question in the affirmative, his opponents in the negative. Peinado’s opponents typically claim that positive propositions are about positive entities, and negative propositions are about negative entities. More specifically, they claim that 26 Mauro, Quaestionum philosophicarum liber secundus, 170 (ad 5). This claim does not entail that every proposition is true but that every proposition affirms the existence of a truthmaker. If I affirm the existence of Pegasus, my affirmation is true if and only if Pegasus exists. Similarly, if p affirms the existence of a truthmaker T, p is true if and only if T exists. Hence, every true proposition has a truthmaker. 27 Anon., Disputatio de obiecto et verificativo propositionum, f. 220v. 28 Giattini, Logica, q. 6, a. 2, p. 265. 29 Carleton, Philosophia universa, Logic, d. 18, s. 2, n. 6, p. 82.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

168

Brian Embry

if p is about some positive entity x, then ¬p is about the lack of x, where the lack of x is a sui generis negative entity that is incompatible with the existence of x.30 I will call this the “object view” of affirmation and negation, since it asserts that contradictory propositions differ according to their intentional objects. When paired with the composite view of truth, the object view entails that negative truths are made true by negative entities; indeed, it is noteworthy that Polizzi, an advocate of the object view, describes those who are opposed to the object view as motivated by the desire to reject negative entities.31 Note that on the object view, the term “not” plays a role in determining the intentional object of a proposition: if “Peter is running” is about Peter’s running, then “Peter is not running” is about the lack of Peter’s running. Others reject the assumption that “not” plays a role in determining the intentional object of a proposition. This can be seen in the following passage by Luis de Losada: In conditional acts, copulative acts, and disjunctive or vague acts, the particles ‘if,’ ‘and,’ and ‘or’ do not represent some conditionality, copulation, or disjunction or vagueness on the part of the object; they merely indicate a certain way of representing. The same thing must be said about the particle ‘not’ in negative acts.32

According to Losada’s view, the disjunction “Peter is running or Paul is running” is about Peter’s running on the one hand and Paul’s running on the other, and it represents those objects in a disjunctive way. Losada applies this view to negation. The idea is that if “Peter is running” is about Peter’s running, then “Peter is not running” is also about Peter’s running. Hence, contradictory propositions p and ¬p differ not according to their intentional objects but according to a “way of representing” intentional objects. The relevant “way of representing” is often called a “modus tendendi,” which I translate henceforth as “intentional mode.” Peinado illustrates distinct intentional modes with the case of love and hate. Love and hate have intentional objects, and they are different not because they have different intentional objects but because they are different ways of being cognitively related to their objects. Similarly, affirmation and negation are different ways of being cognitively related

30 See ibid., s. 6, nn. 2–4, p. 84. For a detailed exposition of a seventeenth-century theory of lacks, see Embry, “An Early Modern Scholastic Theory of Negative Entities.” 31 Polizzi, Philosophicarum disputationum tomus tertius, d. 50, s. 9, n. 160, p. 668. 32 Losada, Cursus philosophici tertia pars, Metaphysics, d. 4, ch. 3, n. 29, p. 157. See also Izquierdo, Pharus scientiarum, d. 2, q. 4, n. 171, p. 83; d. 2, q. 4, prop. 4, pp. 89ff; Lugo, Disputationes scholasticae et morales, d. 1, s. 3, nn. 13–14, pp. 3–4; nn. 17–25, pp. 4–5. Lugo’s discussion seems to have been influential on the intentional mode camp.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

How Not to Be a Truthmaker Maximalist

169

to an intentional object. I call this the “intentional mode view” of affirmation and negation: The intentional mode view: p and ¬p have the same object and different intentional modes.

The fact that Peinado endorses the intentional mode view is clear from the following passage: The affirmative and negative judgments mentioned above [‘Peter exists’ and ‘Peter does not exist’] do not differ according to object, but both have for their object the existence of Peter. But they differ according to their intentional modes.33

In this passage, Peinado says that the propositions “Peter exists” and “Peter does not exist” are both about the existence of Peter, and they differ only according to their intentional modes. Peinado has several arguments for the intentional mode view. Here I will paraphrase two of them. Peinado thinks that we can deny content as well as affirm it. Using contemporary terminology, we could say that Peinado distinguishes between the force and the content of a proposition, and he further distinguishes between two kinds of force: positive and negative. The intentional mode view can account for these distinctions. The “content” of a proposition is its intentional object; the “force” of a proposition is its intentional mode. Thus, a positive proposition p affirms its object, and ¬p denies its object. With this in mind, Peinado raises a dilemma for the object view. Consider the proposition that Peter does not exist. Does “Peter does not exist” have an affirmative or a negative intentional mode? The advocate of the object view cannot say that “Peter does not exist” has a negative intentional mode for the following reason: on the object view, “Peter does not exist” is about the lack of Peter. So if “Peter does not exist” had a negative intentional mode, it would deny the lack of Peter, and it would therefore be true when Peter exists, which is absurd. So the object view is committed to the claim that “Peter does not exist” has an affirmative intentional mode. But in that case, the object view cannot account for the distinction between positive and negative force. On the object view, then, we can only ever affirm content; we cannot deny it.34 This consequence of the object view is sometimes explicitly embraced. In a quotation we have already seen, Mauro claims that “in all propositions some existence, or the ultimate actuality that is the truthmaker of the proposition, is affirmed” (my emphasis).35 Here Mauro explicitly states that all propositions are affirmative. But since we apparently 33 Peinado, De anima, lib. 2, d. 3, s. 2.3, n. 25, p. 262. 34 Ibid., s. 2.4, n. 26, p. 262. 35 Mauro, Quaestionum philosophicarum liber secundus, 170 (ad 5).

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

170

Brian Embry

can deny as well as affirm things, Peinado thinks we ought to endorse the intentional mode view of affirmation and negation.36 Peinado also thinks that intentional objects are not fine-grained enough to account for the difference between positive and negative propositions. This is because propositions with the same object might nonetheless differ with respect to polarity—one might be positive and the other negative. To see this, consider an example. According to the object view of negation, “Peter does not exist” is about the lack of Peter. “The lack of Peter exists” is also about the lack of Peter, yet “The lack of Peter exists” is not syntactically negative. So here we have a positive and a negative proposition about the same intentional object. The intentional object alone therefore cannot discriminate between the positive and negative propositions. On the intentional mode view, however, “Peter does not exist” is negative because it has a negative intentional mode, and “The lack of Peter exists” is positive because it has a positive intentional mode. Peinado’s view of negative truths is driven by his intentional mode view of negation, together with what I call “the aboutness constraint on truthmaking.” Aboutness Constraint: If T makes p true, then p is about T.

Recall that the standard analysis of truth entails that truths are about their truthmakers, so the aboutness constraint was widely accepted by Peinado’s contemporaries. Peinado also accepts the aboutness constraint, as is clear from the following passage: It seems obvious that no internal or external speech and no act of the intellect is rendered true by something that it does not say, by an object that it does not represent. For who would say that this act, ‘Peter runs,’ is formally made true by the existence of God or by any other entity distinct from the running of Peter, which the act represents?37

Peinado thinks that the aboutness constraint is “obvious,” and he brings the obviousness out with an example. Imagine someone saying that God is the

36 Of course Peinado has not eliminated all the options with this argument. One might follow Ockham and say that propositions—rather than the objects of propositions—are the objects of affirmation and denial. On this view affirmation and denial are not built into propositions at all. This view does not appear to have been on Peinado’s radar, but it is not without difficulties. As Walter Chatton pointed out, the phenomenology of affirmation and denial tells against Ockham’s view, since we typically are not thinking about a mental act when we affirm or deny something. For more objections to Ockham’s view, see Susan Brower-Toland, “Facts vs. Things: Adam Wodeham and the Later Medieval Debate About Objects of Judgment,” The Review of Metaphysics 60 (2007), 597–642. 37 Peinado, De anima, lib. 2, d. 3, s. 2.4, n. 32, p. 265.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

How Not to Be a Truthmaker Maximalist

171

truthmaker for “Peter runs.”38 Peinado finds such a view obviously absurd. Peinado’s diagnosis of the situation is that “Peter runs” is about Peter’s running, not God. More generally, true propositions are about the things that make them true. It is worth noting that many contemporary truthmaker theorists subject truthmaking to a relevance constraint due to worries about trivial truthmakers for necessary truths.39 One way to cash out such a relevance constraint is to insist that truths be about their truthmakers. Peinado’s restriction of the truthmaker principle is driven by the intentional mode view of affirmation and negation and the aboutness constraint on truthmaking. Peinado nowhere gives an argument for the claim that negative truths do not generally have truthmakers. Instead, he proceeds piecemeal. His target in the following quotation is a particular version of the view that negative truths are made true by lacks.40 The details of this theory are unimportant, for as we shall see, Peinado’s argument generalizes. Peinado argues as follows: A negative proposition (for example, ‘Peter does not exist’) does not have for its object the lack of Peter, but that existence which it negates. . . . But no proposition is made true by something that is not its object. . . . Therefore, a lack identified with something positive is not necessary for making negative propositions true.41

The argument here is simple: “Peter does not exist” is not about the lack of Peter; so, by the aboutness constraint, “Peter does not exist” is not made true by the lack of Peter. Peinado can run this kind of argument on any proposed truthmaker for “Peter does not exist.” For example, suppose one of Peinado’s opponents were to claim that “Peter does not exist” is made true by God’s decree that Peter not exist.42 Peinado will insist that “Peter does not exist” is 38 Some of Peinado’s contemporaries thought that God’s decrees are truthmakers for negative truths and tensed truths, and that God himself is the truthmaker for all necessary truths. God’s decrees are handy truthmakers because it is necessarily the case that if God decrees that p, then p is true. Against this, Peinado claims that propositions about ordinary creaturely objects ought to be made true by ordinary creaturely objects, not by God’s decrees. See Mauro, Quaestionum philosophicarum liber secundus, qq. 47, 49, 51. 39 The worry is that, absent a relevance constraint, anything can make true any necessary truth, since anything necessitates any necessary truth. But this trivialized truthmaking for necessary truths. See Greg Restall, “Truthmakers, Entailment, and Necessity,” Australasian Journal of Philosophy 74 (1996), 333–4; Armstrong, Truth and Truthmakers, 10–12; Merricks, Truth and Ontology, 22–34. 40 Some of Peinado’s contemporaries claimed that negative truths are made true by lacks, but then they identified these lacks with positive entities in order to make them seem more respectable. That is the view in Peinado’s sights here. 41 Peinado, De generatione et corruptione, lib. 1, t. 2, d. 4, s. 2, n. 23, p. 211. 42 In fact, one of Peinado’s opponents makes precisely that claim: Mauro, Quaestionum philosophicarum liber secundus, q. 49, pp. 197ff. For arguments against Mauro’s position, see Giattini, Logica, q. 6, a. 2, pp. 260–5.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

172

Brian Embry

about Peter, not a divine decree. So “Peter does not exist” is not made true by a divine decree. More generally, for any x (x ¼ 6 Peter), “Peter does not exist” is not about x, and “Peter does not exist” is therefore not made true by x. Because “Peter does not exist” is about Peter, Peter is the only thing that could make it true that Peter does not exist. But since Peter makes it true that Peter exists, Peter cannot be the truthmaker for “Peter does not exist,” on pain of contradiction. It follows that “Peter does not exist” does not have a truthmaker. This argument generalizes. Hence, negative truths do not have truthmakers. In order to make Peinado’s commitments explicit, it is worth noting that the above argument hinges on the assumption that “Peter exists” is true if its intentional object—Peter—exists (assuming the proposition “Peter exists” exists). Peinado needs a general version of this claim in order for the argument to generalize. I will call the general version “Positive Truth”: Positive Truth: A positive proposition p is true if p exists and the intentional object of p exists.

Peinado’s opponents would have accepted Positive Truth because it follows from the standard analysis of truth. And although Peinado rejects the standard analysis of truth as conflicting with his view of negative truths, he accepts the standard analysis as restricted to positive truths. Hence, Peinado too accepted Positive Truth. The intentional mode view, the aboutness constraint on truthmaking, and Positive Truth together entail that negative truths do not have truthmakers. 3 . A T E X T UA L D I F F I C U LT Y Unfortunately, Peinado does not explicitly say that negative truths do not have truthmakers. In fact, he sometimes states that negative truths do have truthmakers. For example, he says: The truthmaker for the negative act, ‘Peter does not exist,’ is Peter negatively represented by that act. Say the same thing about any other negative act.43 The formal truthmaker of an affirmative proposition is the existence affirmed. Therefore, the formal truthmaker of a negative proposition is nothing other than the existence denied.44

Far from endorsing the position I have attributed to Peinado, in these passages Peinado explicitly states that negative truths have truthmakers. Moreover, his claims about truthmakers for negative truths in these passages 43 Peinado, De anima, lib. 2, d. 3, s. 2.4, n. 32, p. 266. 44 Ibid., n. 34, p. 266.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

How Not to Be a Truthmaker Maximalist

173

are deeply puzzling. In the first passage Peinado says that “Peter negatively represented” makes it true that Peter does not exist. Is Peter negatively represented the same thing as Peter? If so, how can Peter make it true that Peter does not exist? If not, what sort of thing is Peter negatively represented? The sense of the second passage seems to be that one existence makes contradictory propositions true. But if that is the case, then that one existence would make a contradiction true. So what is going on in these passages? Peinado is aware of the fact that his view is non-standard and potentially confusing. In response to the worries I just raised about his position, he offers some clarification: When we say, “A non-existent object makes a negative proposition true,” by this act we take the object negatively, since it is the same as to say, “The object does not exist, and the proposition says that and nothing else.”45

This passage makes it clear that Peinado thinks that non-existent objects make negative propositions true. As he explains, when he says that a non-existent object makes a negative proposition true, he means that the proposition says that its object does not exist, and its object does not exist. So the idea seems to be that Peter, by failing to exist, makes it true that Peter does not exist. It therefore seems likely that when Peinado speaks of “Peter negatively represented,” he means something like “Peter, considered as non-existent.” Peinado claims in several passage that non-existent objects are truthmakers for negative truths. In one such passage, he claims that the Antichrist makes it true that the Antichrist does not exist. He then considers the following objection: When this act [‘The Antichrist does not exist’] is true, its truthmaker exists, for it is denominated ‘true’ from its truthmaker. But when this true act exists, the Antichrist does not exist. Therefore something else exists, which is its truthmaker.46

Peinado responds to this objection by rejecting the major premise. He writes: I deny the major premise. . . . For if the truthmaker of a negative act—namely, its object—existed when the act exists, the act would be false; hence, for the act to be true, it is required that, when the true act exists, its truthmaker does not exist. On the contrary, the act is true precisely because its truthmaker does not exist.47

In the face of the objection Peinado maintains quite explicitly that negative truths have truthmakers that do not exist. The Antichrist, by failing to exist, makes it true that the Antichrist does not exist. 45 Ibid., n. 37, p. 268. 46 Peinado, De generatione et corruptione, lib. 1, t. 2, d. 4, s. 5, n. 55, p. 224. 47 Ibid.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

174

Brian Embry

Put this way, I think Peinado’s view is still a bit confusing to contemporary ears. How can a proposition be made true by something that does not exist? Peinado’s position can be illuminated with some terminology that was introduced by his colleague at the University of Alcalá, Juan de Ulloa Madritano (1639–1723). Madritano endorses Peinado’s position, and to help explain the position he introduces a distinction between truthmakers in the positive sense and truthmakers in the negative sense. He describes truthmakers in the positive sense as follows: The truthmaker of any act in the positive sense is that which the act is about [attingitur per actum], and additionally is such that when it exists, in proportion to the intentional mode, the act is true, and when it is absent, the act is false.48

Based on this passage, truthmakers in the positive sense are what early modern scholastics standardly think of as truthmakers. For the sake of clarity, we can define a truthmaker in the positive sense as follows: T is a truthmaker in the positive sense for p =df p is about T, and p is true if and only if T exists.

Madritano then describes truthmakers in the negative sense as follows: A truthmaker in the negative sense I call that which the act is about and is such that the act is true if [and only if] it does not exist. Notice the difference here with a truthmaker in the positive sense.49

Based on this passage and others like it, we can define truthmakers in the negative sense as follows: T is a truthmaker in the negative sense for p =df p is about T, and p is true if and only if T does not exist.

And now we can express Peinado’s view using the new terminology. Peinado’s truthmaker principle states: Peinado’s truthmaker principle: Positive truths have truthmakers in the positive sense but not in the negative sense. (First order) negative truths have truthmakers in the negative sense but not in the positive sense.

Madritano nicely summarizes the Peinadian view of negative truths as follows: You will say, so what is required for a negative proposition to be true? I respond that nothing is required in the positive sense, since negative propositions are not of the

48 Madritano, Prodromus, d. 7, ch. 7, n. 53, p. 776. 49 Ibid., n. 54, p. 777. I add the material in brackets in light of other remarks made by Madritano, but nothing in this paper hangs on the addition.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

How Not to Be a Truthmaker Maximalist

175

sort that strictly and properly require anything in order to be true; rather, they are of the sort whose truth is precisely prevented by something—namely, by the existence of their object.50

So when Peinado says that ‘Peter does not exist’ is made true by Peter negatively represented, he means that Peter is the truthmaker in the negative sense for ‘Peter does not exist’. And this amounts to the claim that ‘Peter does not exist’ is true because it is about Peter, and Peter does not exist. This clears up the paradoxical air of the claim that negative truths have nonexistent truthmakers. When Peinado says that, all he means is that negative truths are true because their objects do not exist, and there is nothing obviously paradoxical about that. It is also clear that Peinado’s considered view is a terminological variant of the claim that negative truths do not have truthmakers. To say that negative truths have truthmakers in the negative sense but not in the positive sense is to say that there is no portion of reality responsible for the truth of negative propositions. Any contemporary truthmaker theorist would put that point by saying that negative truths do not have truthmakers.

4. OBJECTIONS

a. First Objection: Alethic Pluralism As we have seen, Peinado’s truthmaker principle is largely motivated by his intentional mode view of affirmation and negation. Notice that the standard analysis of truth is not available to someone who endorses the intentional mode view. This can be seen by way of example. According to Peinado, “Peter is running” is about Peter’s running. From the intentional mode view, it follows that “Peter is not running” is also about Peter’s running. Hence, if the standard analysis of truth is correct, “Peter is not running” is true if and only if Peter’s running exists, which is obviously incorrect. So Peinado is not entitled to the standard analysis of truth. As mentioned, Peinado accepts the standard analysis of truth with respect to positive truths, but in order to accommodate his view of negation, he must modify the analysis with respect to negative truths, and this is precisely what he does: For a negative act to be true now is for the act to exist now and for its object not to exist, since the act says that its object does not exist.51 50 Ibid., n. 59, p. 782. 51 Peinado, De generatione et corruptione, lib. 1, t. 2, d. 4, s. 5, n. 55, p. 224.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

176

Brian Embry

So it seems that Peinado has distinct analyses of truth for positive and negative propositions. His analysis of truth for positive propositions is given by the standard analysis. And Peinado analyzes truth for negative truths as follows: Negative Truth: A negative proposition ¬p is true =df ¬p exists and the object of ¬p does not exist.52

So Peinado appears to be a sort of alethic pluralist. Yet Peinado’s opponents commonly reject alethic pluralism. As Francisco Suárez writes, “truth ought to have the same account [ratio] in every case.”53 It is not clear exactly why Peinado’s opponents reject alethic pluralism. Perhaps they were concerned with theoretical parsimony. But it is worth noting that alethic pluralism has some problems well known from the contemporary literature. To give one example, we say that an argument is valid if and only if it is truth-preserving. If there is a unified analysis of truth, the concept of truth can help us explain validity. We can say that when an argument is truth-preserving, there is some one condition or property such that, if the premises satisfy it, then the conclusion satisfies it. But if there are different kinds of truth for different proposition types, it is no longer clear what truth-preservation amounts to. Further, if truth is different for positive and negative truths, it is unclear what we should say about the truth of a conjunction of a positive and a negative truth, or about iterated negations. Still further, Peinado thinks that every logical connective is an intentional mode,54 and he presumably would want to make the same sort of moves with respect to other logically complex truths as he does with respect to negative truths. But then his analyses of truth must proliferate according to the infinitely many kinds of logically complex truths. Unfortunately, Peinado does not respond to the charge of alethic pluralism. I will briefly deviate from the project of explaining Peinado’s account of truthmakers for negative truths in order to explain how someone attracted to Peinado’s account could provide an analysis of truth that preserves Peinado’s other commitments and dodges the charge of alethic pluralism.

52 Ibid. 53 DM 8.2.2 (XXV, 277). Some of Peinado’s opponents argue to the effect that someone who endorses the intentional mode view cannot account for truth. See Giattini, Logica, q. 6, a. 5, pp. 292–3; Polizzi, Philosophicarum disputationum tomus tertius, d. 50, s. 9, nn. 161–166, pp. 668–9. 54 It is not clear what Peinado would say about the intentional modes of logically complex propositions with multiple logical connectives. Does such a proposition feature nested intentional modes? Or do the various logical connectives somehow work together to form a new intentional mode? It is not immediately clear how Peinado would answer these questions, or even what hangs on it.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

How Not to Be a Truthmaker Maximalist

177

The idea is to use a recursive analysis inspired by Alfred Tarski, with a correspondence theoretic base clause that helps preserve Peinado’s commitments about truthmaking.55 For simplicity I want to start by assuming that there is only one logical connective, negation (¬), and I want to set aside tense, modality, the quantifiers, and other complicating syntactical features. I will understand atomic propositions to be propositions free of logical connectives—hence, only positive propositions are atomic. Focusing thus on a part of a language, we can define truth as follows: The Recursive Analysis of Truth* 1. Where p is any atomic proposition, p is true if and only if the intentional object of p exists. 2. Where p is any proposition a. ¬p is true if and only if p is not true.

Clause 1 expresses Peinado’s analysis of truth for positive truths. It therefore entails the first part of Peinado’s truthmaker principle, that every atomic proposition has a truthmaker in the positive sense. The recursive analysis, together with the intentional mode view and the aboutness constraint, entails that negative truths have truthmakers in the negative sense but not in the positive sense. This can be seen as follows. Consider the negation ¬p of the atomic proposition p. According to clause 2, ¬p is true if and only if p is not true. According to clause 1, p is not true if and only if its object does not exist. According to the intentional mode view of negation, the object of ¬p is the object of p. So, ¬p is true if and only if its object does not exist. Given the definition of “truthmaker in the negative sense,” it follows that ¬p has a truthmaker in the negative sense (its object) but not in the positive sense. So the recursive analysis preserves Peinado’s truthmaker principle. A further benefit of the recursive analysis is that it defines truth for iterated negation.56 And crucially, it is not a form of alethic pluralism, precisely because it is recursive. It should be noted, however, that the recursive analysis does not fit well with the idea that primary truth-bearers are 55 This idea is suggested in Mulligan, Simons, and Smith, “Truth-Makers.” For brief discussion of the recursive correspondence theory of truth, see Marian David, Correspondence and Disquotation: An Essay on the Nature of Truth (New York: Oxford University Press, 1994), 119–24. David notes that someone adopting the recursive analysis cannot define the logical connectives in terms of truth, on pain of circularity. Special thanks to Adam Murray and Jessica Wilson for discussion here. 56 According to clause 2, ¬¬p is true if and only if ¬p is not true; applying clause 2 again, ¬p is not true if and only if p is true. And now according to clause 1, p is true if and only if its object exists. Putting all this together, we get ¬¬p is true if and only if the object of p exists. But according to the intentional mode view, the object of ¬¬p is the object of p. So ¬¬p is true if and only if its object exists. It follows that ¬¬p has a truthmaker in the positive sense.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

178

Brian Embry

token mental sentences. If primary truth-bearers are token mental sentences, then the recursive analysis defines the truth of one token in terms of the truth of another token, and there is no guarantee that the second token exists. One way to avoid this problem is to pair the recursive analysis with the idea that primary truth-bearers are mental sentence types that exist even when their tokens do not.57 As we have seen, Peinado and many of his contemporaries endorse an intentional mode view not only of negation but of all logical connectives. For reasons similar to those explained in section 3, it follows that the standard analysis of truth is not adequate with respect to molecular propositions. As far as I have seen, Peinado does not discuss molecular propositions involving conjunction and disjunction, but we can account for them in the recursive analysis by adding a clause for conjunction: The Recursive Analysis of Truth 1. Where p is any atomic proposition, p is true if and only if the intentional object of p exists. 2. Where p and q are any propositions a. ¬p is true if and only if p is not true; b. p ∧ q is true if and only if p is true and q is true.

The remaining logical connectives can be defined in terms of conjunction and negation. Clauses 1–2b will tell us what is required on the part of the object(s) of a molecular proposition in order for that proposition to be true.

b. Second Objection: Privative Propositions One might wonder what Peinado would say about truths like “Peter is blind.” Because “Peter is blind” is syntactically positive, Peinado must assign it a truthmaker (if it is true), by his own lights. But it seems that “Peter is blind” is in some sense negative. So it seems that Peinado must assign truthmakers to some negative truths.58 I do not think that the above consideration presents any special problems for Peinado. His account of negative truths is neutral with respect to metaphysical theories of blindness, for example. Peinado insists that “a is F” is true just in case a’s Fness exists. Hence, Peinado can say that “Peter is blind” is true just in case Peter’s blindness exists. Nor does it follow that 57 For a contemporary defense of the view that propositions are mental act types, see Scott Soames, What Is Meaning? (Princeton: Princeton University Press, 2010). 58 Peinado considers this objection in De anima, lib. 2, d. 3, s. 2, subs. 4, n. 39, p. 268. For a contemporary version of the objection, see Molnar, “Truthmakers for Negative Truths.” Interestingly, Molnar’s objection implicitly assumes the object view of affirmation and negation, which Peinado rejects.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

How Not to Be a Truthmaker Maximalist

179

“Peter is blind” is negative in Peinado’s sense. Peinado’s account of negative truths is an account of syntactically negative truths. Which truths are syntactically negative is built into the metaphysical structure of the truths themselves: the negative truths are all and only the ones with negative intentional modes.59 Because “Peter is blind” is affirmative, it is not negative in Peinado’s sense. As it turns out, Peinado rejects the existence of things like blindness and lacks more generally. For our purposes it does not matter why Peinado rejects the existence of lacks. The point I want to focus on is that because Peinado rejects the existence of lacks, Peinado claims that propositions such as “Peter is blind” are literally false: These propositions seem to be affirmative: ‘Peter fails to be [deficit],’ ‘Peter is excluded from things,’ ‘There is a lack of Peter.’ In these and similar propositions the verbs are alienated from their proper to their improper signification. . . . For those verbs, properly taken, properly signify that the defect, exclusion, and the lack of Peter exist in reality. . . . Thus all the above propositions, if they are not taken in a sense equivalent to negative propositions, are false.60

I take it Peinado is making two moves in this passage. In one move, he is admitting that propositions like “There is a lack of Peter” are literally false. In light of his analysis of truth for positive propositions, it is easy to see why Peinado would say that such propositions are literally false. “There is a lack of Peter” is about the lack of Peter, so it is true just in case the lack of Peter exists. Peinado thinks that lacks do not exist; a fortiori, the lack of Peter does not exist. Hence, “There is a lack of Peter” is literally false, even if Peter does not exist. Peinado would say the same sort of thing about “Peter is blind.” (False because Peter’s blindness does not exist, even if Peter cannot see.) However, in the second move, Peinado claims that propositions like “Peter is blind” can be taken as somehow equivalent to negative propositions, in which case they come out true. There is a question here how exactly syntactically positive propositions can be “taken in a sense equivalent to negative propositions.” This has to do with Peinado’s doctrine of alienation, which occurs when a verb receives non-standard signification in virtue of other terms in a proposition.61 A detailed discussion of the semantics of 59 For simplicity I ignore iterated negation here, which these claims can easily be adjusted to accommodate. 60 Peinado, De generatione et corruptione, lib. 1, t. 2, d. 4, s. 4, n. 39, p. 217. See also Peinado, De anima, lib. 2, d. 3, s. 2.3, n. 26, p. 263. 61 Peinado gives an explanation of alienation in his discussion of tense. See De anima, lib. 2, d. 6, s. 4, n. 66, pp. 400–1: “The verb in a proposition is often so affected by the character [ratione] of the particle placed on the part of the predicate that, although the

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

180

Brian Embry

alienation would take us too far afield, and we need not understand the details of alienation in order to appreciate the kind of move, broadly speaking, that Peinado is making here. The claim is that a certain natural language sentence that we routinely treat as true turns out false upon closer metaphysical or scientific inspection. (“The sun is rising,” for example.) However, there is a nearby sentence that more closely tracks the purported metaphysics and accordingly turns out true, and we treat the latter (true) sentence as somehow equivalent to the former (false) sentence. It is a difficult and controversial matter what could be the relationship between the true and the false sentence: alienation, regimentation, paraphrase, synonymy, analysis, translation, meaning, etc. At any rate, the sort of move Peinado is making in the passage cited earlier should be familiar to contemporary philosophers in the Quinean tradition, even if the details are not.

5 . C O N C LU S I O N Peinado thinks that negative truths do not have truthmakers in the positive sense. In contemporary terminology, Peinado’s view effectively restricts the truthmaker principle to positive truths. But contemporary philosophers often reject such a restriction of the truthmaker principle as ad hoc. It is now clear how Peinado would answer the charge of adhocitas.62 Peinado verb itself properly signifies the present, nevertheless when it is conjoined with that particle it does not signify anything about the present, but about the past or future. Thus the verb ‘is,’ taken in itself, signifies an existence of the present and nevertheless when it is conjoined with the particle ‘dead’ in this proposition ‘Peter is dead,’ it does not signify that Peter presently exists, but that he existed in the past, since it is equivalent to this proposition, ‘Peter existed, and does not now exist.’ ” In Phys. (lib. 1, d. 7, s. 2, n. 25, p. 183) Peinado attributes the alienation doctrine to Francisco Alonso (Alphonsus), who describes alienation as follows: “In propositions de tertio adiacente the verb ‘to be’ is sometimes taken from its proper signification to another signification. It happens this way in the following propositions: ‘Peter is dead,’ and ‘The Antichrist is future.’ If the particle ‘is’ were placed without another predicate, it would signify that Peter and the Antichrist now exist. But because other predicates are added to it, in the first place it signifies that Peter existed and now does not, and in the second place it signifies that the Antichrist does not exist now but will at some later time. The same goes for these propositions, ‘Peter is possible’ or ‘Peter is not contradictory.’ In these propositions the particle ‘is,’ by reason of the predicate adjoined to it, is taken to signify that if Peter were to exist, a contradiction would not follow. Grammarians call this ‘alienation,’ since, the copula is transferred from its proper to an alien signification because of an adjoined term” (translated from Jacob Schmutz, La querelle des possibles: Recherches philosophiques et textuelles sur la métaphysique Jésuite Espagnole, 1540–1767 (Ph.D. Dissertation, Université Libre de Bruxelles, 2003), vol. 2, 1091, n. 33). 62 I have long felt that philosophers need an abstract form of ad hoc, and I prefer the scholastic barbarism “adhocitas” to the linguistic mongrel, “ad-hocness.”

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

How Not to Be a Truthmaker Maximalist

181

would answer Cameron by saying that it is “ok for negative truths to go ungrounded” because the intentional mode view of affirmation and negation, the aboutness constraint on truthmaking, and the analysis of truth for positive truths jointly entail that negative truths do not have truthmakers in the positive sense. And Peinado would simply deny Merricks’s claim that “the only reason to scale back Truthmaker to exempt negative existentials is that there do not seem to be truthmakers for negative existentials.” Although he agrees that there do not seem to be truthmakers for negative existentials, Peinado restricts the truthmaker principle because he independently endorses the intentional mode view of affirmation and negation, a certain analysis of truth for positive truths, and the aboutness constraint on truthmaking. Of the three doctrines, only the intentional mode view was not mainstream among his Jesuit contemporaries. And as we have seen, Peinado endorses the intentional mode view because he thinks it alone can account for the distinction between positive and negative force, and it alone can account for the difference between positive and negative propositions. University of Toronto

BIBLIOGRAPHY Anonymous. Disputatio de obiecto et verificativo propositionum (Madrid: Biblioteca de la Real Academia de la Historia), Ms. 9/3081, ff. 213r–23r. Armstrong, David. Truth and Truthmakers (Cambridge: Cambridge University Press, 2004). Augustine de Herrera. Tractatus de voluntate Dei (Complutum [Madrid], 1673). Baker, Stephen and Mark Jago. “Being Positive About Negative Facts,” Philosophy and Phenomenological Research 85 (2012), 117–38. Brower-Toland, Susan. “Facts vs. Things: Adam Wodeham and the Later Medieval Debate About Objects of Judgment,” Review of Metaphysics 60 (2007), 597–642. Cameron, Ross. “How to Be a Truthmaker Maximalist,” Noûs 42 (2008), 410–21. David, Marian. Correspondence and Disquotation: An Essay on the Nature of Truth (New York: Oxford University Press, 1994).

This paper descended from papers read at the 2013 Princeton Penn Columbia Graduate Conference, the 2015 Truth and Grounds conference in Ascona, Switzerland, and the 2015 NEH Institute, “Between Medieval and Modern,” in Boulder, CO. Thanks to the respective audience members for helpful input. This paper has also benefited from the comments of several readers and interlocutors, including (but probably not limited to): Megan Embry, Peter King, Adam Murray, Martin Pickavé, Gonzalo Rodriguez-Pereyra, Marleen Rozemond, David Sanson, Jessica Wilson, and audience members at Columbia University. Finally, a reviewer for this journal provided insightful comments that helped me clarify the paper considerably.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

182

Brian Embry

Doyle, John. “Another God, Chimerae, Goat-Stags, and Man-Lions,” Review of Metaphysics 48 (1995), 771–808. Embry, Brian. “An Early Modern Scholastic Theory of Negative Entities: Thomas Compton Carleton on Lacks, Negations, and Privations,” British Journal for the History of Philosophy 23 (2015), 22–45. Embry, Brian. “Truth and Truthmakers in Early Modern Scholasticism,” Journal of the American Philosophical Association 1 (2015), 196–216. Fine, Kit. “Guide to Ground,” in F. Correia, and B. Schnieder (eds), Metaphysical Grounding (Cambridge: Cambridge University Press, 2012), 37–80. Francisco Peinado. Disputationes in octos libros physicorum Aristotelis (Alcalá, 1680). Francisco Peinado. Disputationes in tres libros Aristotelis de anima, opus posthumum (Alcalá, 1762). Francisco Peinado. Disputationes in duos libros de generatione et corruptione, opus posthumum (Alcalá, 1765). Francisco Suárez. Opera omnia (Paris: Luis Vivès, 1856). Gabriel Vázquez. Commentariorum ac disuptationum in primam partem Sancti Thomae tomus primus (Lyon, 1609). Giovannbattista Giattini Sr. Logica (Rome, 1651). Giuseppe Polizzi. Philosophicarum disputationum tomus tertius (Palermo, 1675). Juan de Lugo. Disputationes scholasticae et morales de virtute et sacramento poenitentiae item de suffragiis et indulgentiis (Lyon, 1666). Juan de Ulloa Madritano. Prodromus seu prolegomena ad scholasticas disciplinas (Rome, 1711). Knebel, Sven. Wille, Würfel und Wahrscheinlichkeit: Das System der moralischen Notwendigkeit in der Jesuitenscholastik, 1550–1700 (Hamburg: Felix Meiner Verlag, 2000). Luis de Losada. Cursus philosophici tertia pars (Salamanca, 1735). Luis de Losada. Cursus philosophici prima pars (Salamanca, 1747). Martin, C. B. “How It Is: Entities, Absences, and Voids,” Australasian Journal of Philosophy 74 (1996), 57–65. Maximus Mangold. Philosophia recentior (Ingolstadt, 1765). Merricks, Trenton. Truth and Ontology (Oxford: Clarendon Press, 2007). Molnar, George. “Truthmakers for Negative Truths,” Australasian Journal of Philosophy 78 (2000), 72–86. Mulligan, Kevin, Peter Simons, and Barry Smith. “Truth-Makers,” Philosophy and Phenomenological Research 44 (1984), 287–321. Ramelow, Tilman. Gott, Freiheit, Weltenwahl: Die Metaphysik der Willensfreiheit zwischen Antonio Pérez S.J. (1599–1649) und G.W. Leibniz (1649–1716) (Leiden: Brill, 1997). Restall, Greg. “Truthmakers, Entailment, and Necessity,” Australasian Journal of Philosophy 74 (1996), 331–40. Rodrigo de Arriaga Cursus philosophicus (Paris, 1639). Rodriguez-Pereyra, Gonzalo. “Truthmaker Maximalism Defended,” Analysis 66 (2006), 260–4. Schaffer, Jonathan. “The Least Discerning and Most Promiscuous Truthmaker,” The Philosophical Quarterly 60 (2010), 307–24.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

How Not to Be a Truthmaker Maximalist

183

Schmutz, Jacob. La querelle des possibles: Recherches philosophiques et textuelles sur la métaphysique Jésuite Espagnole, 1540–1767 (Ph.D. Dissertation, Université Libre de Bruxelles, 2003). Schmutz, Jacob. “Réalistes, nihilistes et incompatibilistes: le débat sur les negative truthmakers dans la scolastique Jésuite Espagnole,” Cahier de philosophie de l’Université de Caen [Special issue titled “Dire le Néant”] (2007), 131–78. Schmutz, Jacob. “Verificativum,” in D. Calma, et al. (eds), Mots médiévaux offerts à Ruedi Imbach (Turnhout: Brepols, 2011), 739–48. Sebastián Izquierdo. Pharus scientiarum (Lyon, 1659). Silvestro Mauro. Quaestionum philosophicarum liber secundus (Rome, 1670). Soames, Scott. What Is Meaning? (Princeton: Princeton University Press, 2010). Tarski, Alfred. “The Semantic Conception of Truth,” Philosophy and Phenomenological Research 4 (1944), 341–76. Thomas Compton Carleton. Philosophia universa (Cursus philosophicus universus) (Antwerp, 1698).

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Reconstructing Aquinas’s World Themes from Brower Thomas M. Ward

Few, probably, would deny that scholarship in medieval philosophy ought to aim at explanation, where the data to be explained are old philosophical texts. Explanation of this kind of data takes the old words (the texts) and puts them into new words (the scholarship). The scholar might have a certain audience in mind when choosing the new words. An audience of other medievalists will expect the new words to be similar to the old words. Students or a non-academic audience will expect the new words to be a simplification of the old words. And so on. Jeffrey E. Brower’s new words,1 predominantly, belong to the discourse of contemporary analytic metaphysics; I surmise, then, that the audience for whom Brower’s book can be explanatory comprises analytic metaphysicians and historians of philosophy who are well versed in analytic metaphysics. Brower says that one of his central goals in the book is “to offer a precise reconstruction of the essential elements of Aquinas’s ontology of the material world” (vii). Reconstruction is not an unusual metaphor for a certain kind of scholarship in the history of philosophy, but it is a controversial one and one worth thinking about. We might think of the old words as the materials for the old construction (the timbers of the old house) and the new words, the words of contemporary analytic metaphysics, as the materials for the new construction. Let’s extend the metaphor: Brower the scholar is Brower the builder and he is reconstructing Aquinas’s ontology of the material world for his audience, those who speak the language of contemporary analytic metaphysics. But what sort of reconstruction? St. Martin’s Cathedral, Ypres, was reconstructed after World War I according to its original design. By contrast, 1 Jeffrey E. Brower, Aquinas’s Ontology of the Material World: Change, Hylomorphism, and Material Objects (Oxford: Oxford University Press, 2014).

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Reconstructing Aquinas’s World

185

Christopher Wren’s reconstruction of St. Paul’s Cathedral, London, bears little outward resemblance to its eponymous medieval predecessor destroyed in the fire of 1666. There is also an extended sense in which Thomas Jefferson’s original Rotunda at the University of Virginia was a reconstruction of the (still standing) Pantheon in Rome. So what sort of work is Brower up to?2 My overall impression from the book is that Brower conceives of his own work more along the lines of Jefferson’s Rotunda than either the new St. Paul’s or the new St. Martin’s. Far from thinking there is anything seriously the matter with the old construction, he deeply admires it, as Jefferson admired the Pantheon. He is committed to the preservation of medieval buildings and is happy to see Aquinas’s works endure in Paris, Cologne, and Naples in their gothic splendor. This admiration inspires Brower to share these works with contemporary metaphysicians, as Jefferson shared the Pantheon with the people of Virginia. But precisely in order that it be appreciated by his intended audience, Brower of West Lafayette3 cannot just copy the old works—the contemporary metaphysicians are not typically in a position to appreciate achievements of the old world; he must reimagine them, expressing them in a manner which gives his contemporaries the opportunity to value them as much as he does. The metaphor of reconstruction in scholarship in the history of philosophy is controversial because its most ordinary sense implies that there is something wrong with the old words. I take it that this is not the sense Brower intends. Nor, even, do I take it that Brower thinks the new words, the language of contemporary analytic metaphysics, are better than the old, the language of scholastic philosophy. It seems to me simply that Brower has, for whatever reason, found it helpful for his own understanding of Aquinas’s ontology of material objects to think about it in relation to the contemporary analytic literature, and (perhaps) therefore has guessed that other contemporary analytic folks might find it easier to see what is valuable about Aquinas’s ontology of the material world if it were expressed, and its merits judged, in the language of contemporary analytic metaphysics. However, some will find fault with Brower’s new words even with this clarification of the sense of his reconstruction. The new and the old are different enough from each other that one might reasonably appreciate one

2 Jefferson’s Rotunda was destroyed in the late nineteenth century and rebuilt to a somewhat modified design. The building has since undergone several renovations designed in part to restore it to Jefferson’s plans and to make repairs. Ignore this later history of the building. I am only interested in the relation between Jefferson’s original building and the Roman Pantheon. 3 Brower is Professor of Philosophy at Purdue University, in West Lafayette, Indiana.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

186

Thomas M. Ward

and not the other, just as an admirer of the Pantheon might be unmoved by Jefferson’s original Rotunda (and vice versa). Those who do not already possess some fluency in and appreciation of the discourse of contemporary analytic metaphysics are unlikely to find Brower’s book very useful. This is because the vocabulary and style of contemporary analytic metaphysics are no easier to understand, or even any likelier to be on the track of truth, than the (reliably translated, if need be) vocabulary and style of medieval scholastic metaphysics. So for those not already fluent in analytic metaphysics, much of Brower’s book just will not work for them as an explanation of Aquinas’s views. But the new stuff is more widely taught and read than the old, which makes it the de facto canonical language of metaphysics in the English-speaking philosophical world. This means that many philosophers need a Brower-style reconstruction of an old philosophy in order to take it seriously as philosophy. For this reason, among others, Brower has done a great service to the Anglophone philosophical world. My guess is that some medievalists will sniff at Brower’s book for being insufficiently historical and insufficiently wedded to Aquinas’s own way of expressing himself. For many others, however, these qualities will be among its virtues. We medievalists need a book like this, not to prove that what we do is worth doing but to help outsiders see this. The thirteen chapters of the book are broken up into five parts: Part I. Introduction, which gives an opinionated account of Aquinas’s fundamental ontology; Part II. Change; Part III. Hylomorphism; Part IV. Material Objects; and Part V. Complications, which considers several natural and supernatural phenomena that might be seen to challenge Aquinas’s canonical formulations of the issues discussed in the first four parts. As a way of offering readers a sense of both the scope of the book as a whole along with its precision and argumentative rigor, I will focus on four somewhat narrow but important topics. These are: 1. Brower’s discussion of the relation between Aristotle’s Ten Categories and the not-obviously-connected four-fold division of being into substance, form, prime matter, and accidental unity. 2. Brower’s interpretation of prime matter as “non-individual stuff.” 3. Brower’s account of numerical sameness without identity along with some discussion of how this is supposed to be useful for solving a particular puzzle about accidental change. 4. I will close with some reflections on Brower’s somewhat novel solution to a longstanding problem with Aquinas’s metaphysics of the afterlife. There is far more worth talking about here than I will get to. The book is deep, complex, and much of it, in particular Parts II and III, is among the most exciting contemporary work I have read either in metaphysics or in medieval philosophy. I will quibble over a few things later, so let me make it clear now: this is an excellent book, which I will be consulting for a long time.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Reconstructing Aquinas’s World

187

1. THE DIVISION OF BEING Early on, Brower makes the claim that according to Aquinas, “the world or ‘all that there is’ can be exhaustively described in terms of two general ontological types—namely God and creature” (3). Shortly thereafter Brower defines a type of being in the following way: “F is a type of being if and only if some Fs (one or more) can or do exist” (4). Given this, Brower thinks that “being turns out to be the most general possible ontological type, with all other ontological types serving as specifications or subtypes of it” (4). These claims, as claims about what Aquinas thinks, are quite frankly astonishing. The most strident critic of Duns Scotus on univocity could not have written a better summary of what is supposed to be problematic with Scotus’s theory. Aquinas, the defender of analogy, is supposed to be the champion who keeps God separate from the world, transcending any concept through which we humans conceive anything we can naturally conceive. A few pages later Brower takes up Aquinas’s venerable distinction between essence and existence in creatures. Given Aquinas’s insistence that this distinction is real, does it not imply that among the things there are we must include not just substances but “individual acts of being or existence” (17)? No, says Brower. “As I see it, the whole point of Aquinas’s distinction between essence and existence in creatures is to emphasize their contingency or lack of aseity. Likewise, I think his talk of creatures being ‘composed’ of essence and existence is a figurative way of expressing the same point, and hence not to be taken literally” (17). Here is one who speaks not as the Thomists speak. Brower’s method here is bound to seem cavalier to some. He does, later on, give an account of Aquinas on analogy, but these early violations of Thomistic shibboleths set a sort of tone for the book: Brower is not primarily writing for those who already know a lot about Aquinas. He is framing Aquinas’s views in his own way, on his terms, in order to make them suitable for engaging contemporary analytic debates and to persuade outsiders that they are cogent. That said, I think what Brower says about Aquinas is on the whole accurate, once his interpretive moves are properly seen within the context of his project as a whole. Brower’s Aquinas thinks that being divides into four irreducible types: prime matter, form, substance, and accidental unity. Brower distinguishes irreducibility from fundamentality or basicality in order to allow for extra divisions among these four irreducible types; substance, for example, is divided into God and created substance, but both alike are substances and hence the irreducible type, substance, is not reducible to types other than

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

188

Thomas M. Ward

substance. Also, form divides into substantial form and accidental form, but not into any type other than form. Prime matter’s main characteristics are being non-individual and being the ultimate substratum of material substances and accidental unities (more on this in the following section). The special mark of form is to be an inherent, rather than a subsistent being: substantial forms inhere in prime matter and accidental forms inhere in substances. Substances themselves (e.g., Socrates and God) are specially characterized as basic particulars, whereas accidental unities (e.g., seated Socrates) are non-basic particulars. A particular in this context means “a being that is both subsistent and individual.” A basic particular in this context means “a particular that is not itself composed of any particulars,” and a non-basic particular means “a particular that is composed of at least one particular” (24). Having laid out this four-fold division of being, Brower then introduces the perhaps better-known division of being into the Ten Categories. Aquinas clearly thought that the four- and ten-fold divisions were compatible with each other, but it is not clear exactly how he envisioned this compatibility. Of the four, only substance has a clear match in the ten. How should we make sense out of the other nine in relation to the original four? Brower here divides form into substantial and accidental, and claims that all the remaining nine Categories find their proper place as determinations of accidental form. But there is a wrinkle here. Aquinas himself says that the Ten Categories apply to created modes of being rather than to created beings (45). An F is the sort of thing that can be a type, whereas an F-ness is the sort of thing that can be a mode. Thus, “F is a type of being if and only if some Fs (one of more) can or do exist” (4), and “F-ness is a mode of being if and only if something does (or can) exist as an F ” (45). Brower thinks that an accurate mapping of the Ten onto the Four has somehow got to preserve this obscure insistence on modes over types. This switch from types to modes complicates Brower’s presentation of Aquinas’s fundamental ontology. We now have to explain both the compatibility of the four-fold and ten-fold division of being, along with the switch from talk of categories or types of being to modes of being. Here I wish Brower were a little clearer, though I admit I find the material intrinsically very difficult so failure to understand may be due more to the reader than the author. Despite the difficulty, here is Brower’s picture as I see it. Belonging to a type of being can itself be explained, namely, by existing in a certain fundamental way. As Brower puts it: [God and created substances] belong to fundamentally different types—God and created substance, respectively. But as we have seen, they appear to belong to such types, for Aquinas, precisely because they exist in fundamentally different ways. Thus,

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Reconstructing Aquinas’s World

189

it is because God exists in the mode of absolute independence, whereas created substances do not, that the general type substance must itself be further subdivided into the more specific, also more fundamental, subtypes God and created substance (46).

So types of being are explained by modes of being, and therefore modes are in some sense the more fundamental target of the task of dividing being. Since types are posterior to modes, we can look for correlates of the four-fold division of types of being among the modes of being. And this is what we find: prime matter exists in the mode of non-individuality, form in the mode of inherence, substance in the mode of basic particularity, and accidental unities in the mode of non-basic particularity. As we have just seen, that substance divides into two types, created substance and God, is to be explained by the modes of absolute and non-absolute independence. That form divides into two types is to be explained by two fundamentally different modes of inherence: substantial forms exist in the mode of inherence in prime matter, whereas accidental forms exist in the mode of inherence in substances. And that accidental form divides into nine types is to be explained by nine fundamentally different modes of inherence in substance. Given this picture, it looks like we get a one-to-one correlation between types and modes. But things evidently are more complicated even than this, in a way that raises a worry about the cogency of Brower’s mapping of the ten onto the four. The problem is that, as Brower is well aware, Aquinas denies that each accidental category has as its correlate, in the world of concrete substances, its own distinct type of accidental form. There are no relation forms, for example; the mode of being toward another is ultimately explained by a substance’s possessing some non-relative form, such as an accidental form of quality. So it is not clear to me, either in Aquinas or in Brower, how the architecture of types and modes ultimately maps onto concrete things in the world. Later on in the book, Brower explains that Aquinas insists that the division of the accidental categories “must be understood not in terms of distinct classes of beings but rather in terms of distinct modes of being” (206). So far, so good: we are supposed to let the modes of being, rather than concrete individual forms or substances or accidental unities, shape our understanding of the categories (types) of being. But Brower continues, “What is more [ . . . ] Aquinas insists that one and the same form or property can be associated with distinct accidental modes of being, and hence fall under more than one of the accidental categories” (206). Brower focuses on action and passion, showing how Aquinas thinks that these two modes of being, associated with two accidental categories, are fully accounted for, ontologically speaking, by one thing, a motion. But then it is not at all clear to me how the division of the nine accidental categories can be mapped onto the type, form. If you can get an accidental mode and/or category of being

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

190

Thomas M. Ward

without a distinct type of form, what ultimately explains the decision to call these nine accidental modes/categories, determinations or specifications of the type, form? Brower may have a totally reasonable answer to this question, but I was not able to figure it out from the book. I will close this section with a quibble. As I mentioned earlier, Brower does address Aquinas’s doctrine of the analogy of being. Ultimately, for Brower, the analogy of being is a way of marking the fact that being comes in several different, fundamental modes. To be a form is a fundamentally different mode of being (inherence) than the mode of being a substance (basic particularity). That God and created substances are only analogously spoken of as substance is, for Brower, precisely (and completely?) because each exists in a fundamentally different mode of being. I myself am inclined to think that this is not all there is to Aquinas’s doctrine of analogy as it applies to the difference between God and creatures (it seems dandy as an account of analogy as applied to the differences between different modes of creaturely being). First, the account here still conceives of being as the most general type, with God as a type of being, albeit a fundamental type. It looks like we are still in the land of the Scotus-caricatures. Second, Brower’s account seems to be too metaphysical and insufficiently epistemological. Part of what drives Aquinas to say that the words we use to talk about God only apply to God analogously is that God’s essence is beyond our ken. But on Brower’s account, we are supposed to be able to say what the distinctive mode of being God is—the mode of being absolutely independent. What is there not to understand about this? If absolute independence is that whereby God instances the type, God, then it seems like we actually do have pretty good access to the divine essence—something Aquinas vehemently denies. What role can be left for human cognitive limitation to play in forcing an analogical construal of our God-talk? Aquinas obviously wants our cognitive limitation to play a big role, but Brower’s account of Aquinas seems to disregard its role altogether.

2 . P R I M E M A T T E R AS N O N - I N D I V I D U A L S T UF F Brower says that prime matter is in some ways “the most distinctive type of being there is for Aquinas” (18). In the previous section we saw that Brower thinks its characteristic mode of being is non-individuality. A thing is individual in the relevant sense if and only if it has actuality through itself and belongs to some natural kind or species. Prime matter does not have actuality through itself; famously, Aquinas says that prime matter is “pure potentiality.” Nor does prime matter belong to any natural kind or species.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Reconstructing Aquinas’s World

191

Now I confess to not really understanding the use of the term “nonindividual” to describe Aquinas’s view of prime matter; that is, I do not know why “individuality” and “non-individuality” are terms used in this context. But it is pretty clear from Brower what “non-individual” is supposed to denote. There are not kinds of prime matter, such as gold, wood, water, etc.—prime matter is uncharacterized. And it does not exist through itself or on its own; it only exists and only can exist united with form. There can be distinct portions of prime matter, and Aquinas does indeed assign to prime matter the role of individuating substances, but prime matter plays this role without itself being individual in the relevant sense. Aquinas’s account of prime matter as pure potentiality strikes some as incoherent. This apparent incoherence is well summarized by the following argument, which Brower labels The Simple Argument: (1) If prime matter is a being in pure potentiality, it has no actuality. (2) But prime matter exists, and whatever exists has actuality. Therefore (3) Prime matter cannot be a being in pure potentiality, but must rather be a being in actuality (31). Brower thinks Aquinas would reject (1). There are at least two senses of actuality, and the claim that prime matter is pure potentiality is incompatible with just one of these senses of actuality. In the broad sense of actuality, anything that exists in any way whatsoever is actual. In the narrow sense, only what has actuality through itself is actual. Prime matter is actual in the broad sense, not the narrow. So (1) is false. That prime matter lacks narrowsense actuality, that it can only exist insofar as it composes something with form, is for Aquinas a theoretical primitive (33). Brower thinks of Aquinas’s prime matter as atomless gunk, which is a type of being every part of which is composed of parts and is therefore infinitely divisible (119). But what is really distinctive of Aquinas’s conception of prime matter, as opposed to other medievals’ conceptions, is that prime matter is not merely gunky but that it is non-individual and hence, as Brower argues, closely resembles a contemporary conception of stuff. (Non-individuality, again, means belonging to no natural kind and lacking actuality through itself.) Ned Markosian thinks that stuff is distinct from things, that it is what ordinary objects are made of, that it cannot exist without things, can enter into part-whole relations, and that there are distinct portions of stuff (34, 126). One important difference between Aquinas’s and Markosian’s conceptions of stuff is that for Markosian stuff is not gunky, but Aquinas thinks that prime matter is what Brower calls gunky stuff. Additionally, for Aquinas there is just one type of prime matter—it does not come in the kinds, such as water, gold, wood, which are often recognized by gunk theorists as kinds of stuff. Brower’s interpretation of prime matter entails that it is not a basic particular, and this fact about prime matter comes to play an important

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

192

Thomas M. Ward

role in the hylomorphic substratum theory Brower develops on Aquinas’s behalf. I will say more about this in the following section, but for now it is worth saying that on Brower’s substratum theory some portion of prime matter serves as the substratum of a substantial form, where the substratum and substantial form together constitute a material substance. Prime matter is just the sort of thing in which substantial forms (or properties of a special, essence-constituting kind) can inhere but which substantial forms cannot characterize. Prime matter does not become a horse or have a horsey soul when the soul of horse begins to inform it. Instead, when the soul of a horse inheres in prime matter a new substance is generated, a horse, and this horse has a horsey soul. And the reason prime matter cannot be so characterized by substantial form is precisely because it is not a particular (basic or otherwise). The horse, not its prime matter, is for Aquinas and Brower the basic particular and the bearer of properties.

3. NUMERICAL SAMENESS WITHOUT IDENTITY Substances and accidental unities are peculiarly related. Socrates is a substance; he is composed of prime matter (his non-individual substratum) and a substantial form (his soul). But Socrates himself is the substratum of his various accidental properties, such as his height, weight, and skin color. With each and every one of his accidental properties, Socrates composes a distinct accidental unity. Thus there is the accidental unity composed of Socrates and his paleness, another composed of Socrates and his being six feet tall, and so on. That’s a lot of things! How are they related? Brower develops the idea, on Aquinas’s behalf, that a substance and an accidental unity are numerically the same but not identical. Brower deploys the notion of numerical sameness without identity in a very specific way. He tells us: it just means that the things so related (numerically the same but not identical) completely overlap with respect to their matter (94). More officially, numerical sameness without identity says that for any hylomorphic compounds x and y, where x does not equal y, and any time t, x is numerically the same material object as y at t if and only if x and y share all their prime matter in common at t (94). So a substance and an accidental unity which has that substance as a part are numerically the same material objects but are not identical to each other. The reason they are not identical to each other is, of course, that the accidental unity is a composite of a substance plus an accidental form. Brower puts this distinction to many magnificent uses. Here I will focus on just one, which concerns a solution to a puzzle about accidental change.

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Reconstructing Aquinas’s World

193

Distinctive of accidental change is that “it involves a subject (more specifically, a substance) that not only possesses distinct forms or properties over time, but is successively characterized by these same properties.” That is, it is not just that some substance has one property, F-ness, at one time and another property, G-ness, at another time, but rather that the substance is F at one time and is G at another time. Now, on what Brower calls an Aristotelian conception of forms, a conception that Aquinas undoubtedly endorses, a hylomorphic compound is characterized by some form or property by having that form or property as a constituent. So consider the substance which is F. In order to be F, it must have F-ness as a constituent. So the substance plus F-ness constitute one hylomorphic compound. But inasmuch as hylomorphic compounds are what they are in virtue of their formal and material constituents, it would seem that for that compound to lose F-ness just would be for it to cease to exist. It would seem then that hylomorphic compounds non-contingently have the constituents they have, and hence that accidental change is impossible. Numerical sameness without identity lets us say that a substance and some other hylomorphic compound, an accidental unity, can share numerically the same prime matter but fail to be identical. So the hylomorphic compound consisting of the substance and F-ness is an accidental unity which has the substance as one of its parts. When the substance ceases to be F the accidental unity is destroyed, but the substance is not destroyed. When it begins to be G a new accidental unity begins to exist, an accidental unity which has the substance and G-ness as constituents. In every case of accidental change, a new accidental unity comes into existence; but because the enduring substance is the substance-constituent of each successive accidental unity, it is derivatively characterized by the property constituents of each accidental unity. So, for example, the accidental unity white-Socrates is white primarily, but the material substance, Socrates, is white derivatively, because Socrates is a part of white-Socrates.

4 . M E T A P H Y S I C S O F T H E A F T E R LI F E A current hot topic in Aquinas scholarship concerns Aquinas’s metaphysics of (human) afterlife. Brower takes up this topic in the concluding chapter of the book and offers an original solution to what he calls The Problem of the Afterlife. For philosophical and theological reasons, Aquinas thinks that the human soul is not only the substantial form of a human being, but also that it is immortal. So when a human being dies, its soul lives on. But it is not obvious,

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

194

Thomas M. Ward

taking all that Aquinas has to say about the afterlife into account, that the survival of a human soul is sufficient to guarantee the survival of the human being whose soul it was. The problem is that Aquinas seems to accept the following three claims but they also seem to be inconsistent: (1) human beings are essentially human; (2) human beings cease to be human at death; and (3) human beings do not cease to exist at death, but rather survive. (1) is completely uncontestable; no one doubts Aquinas thought this. (2) and (3) are well-supported by textual evidence, both explicit and implicit. In their attempts to present an Aquinas with a consistent view about the metaphysics of the afterlife, most commentators fall into one of two camps. The cessationists hold that a human being such as Socrates ceases to exist at death, but that he will exist again when his soul begins again to inform prime matter. The cessationists reason that from (1) and (2) we can infer that at death human beings cease to be. The inference is not unreasonable given a modal conception of essence according to which for all x, if x is essentially F then x cannot exist and not be F. Cessationists then have the interpretive burden of suppressing or explaining away texts which support (3). The other camp, the survivalists, hold that even if Socrates ceases to be human at death, he does not cease to be; the continued disembodied existence of Socrates’s soul is sufficient to guarantee the continued disembodied existence of Socrates. Moreover, since (1) is uncontested in this debate, the survivalists hold that since Socrates is essentially human and since on their view Socrates continues to exist, it follows that Socrates continues to be human. So the survivalists assume the interpretive burden of suppressing or explaining away texts which support (2). The easy route, and the one I once felt forced to take, is simply to judge that Aquinas was inconsistent on this particular issue. Brower takes a harder route. He claims to have found a way to show that properly understood (1)–(3) are consistent. The most important move is to deny the modal conception of essence. It does not follow, according to Brower (and Brower’s Aquinas) that if x is essentially F then x cannot exist without being F (or, in possible worlds talk, it does not follow that if x is essentially F then there is no possible world in which x exists and is not F ). Instead, Brower attributes to Aquinas a dispositional conception of essence according to which for all x, if x is essentially F, then x cannot exist and not be disposed to be F. Brower’s Aquinas evidently thinks it follows from being by nature disposed to be F that x cannot permanently be not-F without ceasing to exist. I do not think this follows and will come back to this later. For now, the important point is that the dispositional account of essence allows for the possibility that x is essentially F and that it is not-F. Applied to the human case, we can say that possibly, Socrates is both essentially a human being and at death ceases to be a human being without ceasing to exist. Taking all the textual evidence into

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Reconstructing Aquinas’s World

195

consideration, we can infer that Aquinas thinks that actually, Socrates is essentially a human being, that Socrates ceased to be a human being at death, and that he continued to exist after his death. A consistent triad. Brower distinguishes between human persons and human beings as a way to establish some tidy language with which to discuss his new interpretation. The idea is that Socrates is always (necessarily?) a human person, but that he is a human being only when he has both prime matter and his soul as his metaphysical constituents. When his soul is separated from prime matter at death, Socrates ceases to be a human being but goes on being a human person. Brower labels his new view, somewhat misleadingly, I think, as non-human survivalism, which he summarizes as “the view according to which all human beings survive their death as human persons but not as human beings” (300). Brower motivates non-human survivalism first by considering it as the best way to understand the metaphysics of the Triduum. During Christ’s three days of death, Aquinas tells us, the second person of the Trinity remained hypostatically united with Christ’s prime matter (existing in the tomb under the form of a corpse) and Christ’s soul. Since soul and body were separated, Christ was really dead and he really ceased to be a human being. But because Christ is also a person suppositing divine nature, Christ cannot cease to exist and therefore does not cease to exist at death. Here we have an example of a human death which results in both the ceasing to exist of a human being and the continued existence of the person who was a human being. This example points the way toward Brower’s generalization of non-human survivalism to cover all cases of human death. As Brower himself says, there are important differences between Christ’s personhood and merely human personhood. For the most part Brower does a fine job making all the relevant qualifications needed to keep his view cogent. But properly qualified, there is one big metaphysical cost to buying Brower’s non-human survivalism, which for me at this point is insurmountable. Here is the cost. Brower must hold that a human person in the interim state after death and before the resurrection is a substance which has just one part. His view requires a rejection of the mereological axiom of weak supplementation, a highly intuitive axiom. Weak supplementation says that if x is really and truly a part (a proper part in the language of mereology) of some whole, y, then y has some proper part, z, other than or disjoint from x. In other words, anything composed of parts has at least two of them. To talk about a thing having just one part is a forced way of talking, and mereology highlights this awkwardness with the label “improper part” for any part, x, of some whole, y, such that y has no other part besides x. To be an improper part is not really to be a part, say I. I think it follows from weak supplementation that the only possible relation between a whole and its one fake part is identity. Thus, I think that Brower’s claim that Socrates

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

196

Thomas M. Ward

continues to be a human person while he has soul as his one fake part commits him to the view that Socrates, the human person, is identical with his soul. But then Socrates is always identical with his soul, even when he is united with prime matter. But this is just Platonism or Cartesianism about human persons—a bad result for Thomists. Brower and Aquinas deny that Socrates is identical with his soul, of course. Brower wants there to be a genuine part-whole relation between Socrates’s soul and Socrates. To get this result he needs to deny supplementation, as he does (309). Doing so lets him continue to talk of Socrates’s soul as a part of Socrates without reducing the relation between Socrates and his soul to identity. At this point I am prepared to claim, by way of objection, only that I just cannot make sense of the idea that a whole can be composed of just one part, and I especially cannot make sense of the idea that a person can be composed of, not identical with, its disembodied soul, if this soul is its one improper part. I wish Brower had spent more time reflecting on the aftermath of his rejection of weak supplementation. As if this were not bad enough, Brower closes with a suggestion that strikes me not just as unintuitive and difficult to understand but just plain wrong. Brower reflects that, on his non-human survivalist view, “Socrates clearly continues to be a composite or compound of some type after death, insofar as he remains composed of a proper part.” But, Brower adds, the single part which composes Socrates is a “hylomorphic part,” a substantial form. Hence, Brower concludes, “the relation that he bears to it presumably remains a type of hylomorphic composition” (310). Here is the problem. Either Brower has in mind that Socrates is related to his one part in some way analogous to the way matter and form are related to each other in a hylomorphic compound, or Brower has in mind that Socrates is related to his one part in some way analogous to the way in which a whole hylomorphic compound is related to its matter and form. Clearly it cannot be the former, since Socrates is supposed to be the whole composed of his one part. But almost as clearly it cannot be the latter, because ex hypothesi there is simply nothing there with which Socrates’s soul can make a composition. So, ultimately, until I can find a way to make sense of the idea of one part composing a whole with which it is not identical, I cannot accept Brower’s version of non-human survivalism. But the real achievement here is that Brower really has, I think, shown that Aquinas’s (1)–(3) are consistent. The original formulation of The Problem of the Afterlife (these three seem to be inconsistent) is therefore finally solved. But now we have a New Problem of the Afterlife: can Thomists make any sense of the idea of a person being related to its soul and only its soul as whole to part? I have elsewhere (and independent of Brower’s book) made exactly the distinction between human beings and human persons which Brower here

OUP CORRECTED PROOF – FINAL, 8/9/2016, SPi

Reconstructing Aquinas’s World

197

makes, and for pretty much exactly the same reasons.4 But, digging in my heels when it comes to supplementation, I argue that a human soul must receive, by a special act of God, some supplemental part so that the soul can continue to compose, rather than be identical with, a human person. I draw on the mystical tradition of deification to suggest that this part is a personconstituting divine grace or maybe even God himself. I write this description of my own paper as a way of helping Brower out. If this is what it takes to hold on to both Brower’s way of making the triad consistent and supplementation, then, you may well think, so much the worse for supplementation. One final criticism, which is a bit of a nitpick. As I mentioned earlier, Brower’s Aquinas thinks it follows from being by nature disposed to be F that x cannot permanently be not-F without ceasing to exist. Brower characterizes the “Thomistic Conception of Natures” in the following way: If x is essentially F, and F-ness is x’s primary nature, then x is non-contingently disposed to be F (and hence such that x cannot permanently cease to be F without ceasing to exist)” (299, italics mine, for emphasis). I do not think this follows. Here is why: suppose Socrates spends 10,000 years waiting for his reunion with matter. What is to prevent him from going on disembodied for 10,001, or 100,000, or 10 billion years? There is no lifespan to a disembodied soul. Surely it can indefinitely subsist without union with matter. But if indefinitely, why not permanently? I know Brower is following Aquinas on this issue but I have never been able to take Aquinas seriously on this point. I understand that it is fitting that a disembodied soul, which is after all a substantial form, should be eventually reunited with matter. But that it must be (after some amount of time?) as a condition for its existence, strikes me as totally unmotivated. Again, a nitpick probably, but as Brower is normally an extremely careful and clear thinker and writer, I wish he had clarified or defended this strange teaching of Aquinas. Loyola Marymount University

BIBLIOGRAPHY Brower, Jeffrey E. Aquinas’s Ontology of the Material World: Change, Hylomorphism, and Material Objects (Oxford: Oxford University Press, 2014). Ward, Thomas M. “Transhumanization, Personal Identity, and the Afterlife: Thomistic Reflections on a Dantean Theme,” New Blackfriars 96 (2015), 564–75.

4 Thomas M. Ward, “Transhumanization, Personal Identity, and the Afterlife: Thomistic Reflections on a Dantean Theme,” New Blackfriars 96 (2015), 564–75.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Don’t Mind the Gap A Reply to Adam Wood Turner C. Nevitt

In the previous volume of Oxford Studies in Medieval Philosophy Adam Wood takes up a challenge to the contemporary consensus on Aquinas’s view of the possibility of intermittent or “gappy” existence, especially as it relates to human death and resurrection.1 According to the contemporary consensus, Aquinas thinks the separated soul’s continued existence immediately after death is necessary for the general resurrection: if my soul perished at death along with my body, then not even God could bring me back to life. Yet, as Wood notes, this consensus rests on shaky foundations. For on the basis of a widely neglected and quite late quodlibet question both John Capreolus and Francis Sylvester Ferrara, two of Aquinas’s best Renaissance interpreters, recognized that Aquinas reserves to God the absolute power to annihilate material substances, which include human beings, and then re-create them numerically the same.2 That means that Aquinas thinks the continued existence of nothing other than God’s own power is necessary for God to bring me back to life. If Aquinas’s Renaissance interpreters are right, then the contemporary consensus is in bad need of revision. But Wood argues that his Renaissance interpreters are not right. First Wood defends an alternative reading of Aquinas’s late quodlibet question on annihilation and re-creation. Then he presents evidence of Aquinas’s endorsement of what Christina Van Dyke

1 Adam Wood, “Mind the Gap? The Principle of Non-Repeatability and Aquinas’s Account of the Resurrection,” Oxford Studies in Medieval Philosophy 3 (2015), 99–127. 2 John Capreolus, Defensiones theologiae divi Thomae Aquinatis, In quarto Sententiarum, ed. C. Paban and T. Pegues, vol. 7 (Tours: Cattier, 1908), 1–29; Francis Sylvester Ferrara, Commentaria in libros quatuor contra gentiles S. Thomae de Aquino, in Sancti Thomae Aquinatis opera omnia, vol. 15 (Rome: Leonine Commission, 1950), 254–9.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Don’t Mind the Gap

199

calls “the principle of non-repeatability,”3 which basically says that what ceases to exist cannot return numerically the same. Aquinas’s Renaissance interpreters were aware of many of the texts in which he appears to endorse this principle, and they developed strategies for reading them in a way compatible with their own understanding of his late quodlibet question. Wood mentions a few of these interpretative strategies and briefly explains why he finds them implausible. Wood then argues that on at least one interpretation of his theory of matter as the principle of individuation Aquinas had a strong philosophical reason for endorsing the principle of non-repeatability, at least where sublunary material substances such as human beings are concerned. Wood then concludes with a corroborating text from another of Aquinas’s best Renaissance interpreters, Cardinal Cajetan, who appears to share his interpretation. I think Wood’s attempt to vindicate the contemporary consensus fails. First of all, his alternative reading of Aquinas’s late quodlibet question is impossible to square with the text itself. Given that, I think Wood’s evidence for Aquinas’s endorsement of the principle of non-repeatability is worth reconsidering. Although I agree that not all of them are plausible, a close reading of the relevant texts shows that at least some of the strategies of Aquinas’s Renaissance interpreters are a good deal more plausible than Wood allows, which means that the textual case in favor of Wood’s interpretation is a good deal weaker than he realizes. And the further philosophical reason Wood presents for Aquinas’s endorsement of the principle of non-repeatability seems to me to beg the question against the possibility of annihilation and re-creation. Finally, his commentary on the Summa makes it clear that Cajetan in fact agreed with Capreolus and Ferrara about Aquinas’s view of the possibility of intermittent existence. All in all, then, I think the consensus of Aquinas’s Renaissance interpreters continues to give us good reason to doubt the contemporary consensus. Let me begin with Aquinas’s late quodlibet question on annihilation and re-creation, which he composed in Lent of 1271.4 In Quodlibet IV, q. 3, a. 1 Aquinas asks whether God can annihilate anything. In reply he says that as far as God’s absolute power is concerned he can annihilate the whole of creation. In a. 2 Aquinas asks whether God can restore something

3 Cf. Christina Van Dyke, “Human Identity, Immanent Causal Relations, and the Principle of Non-Repeatability: Thomas Aquinas on the Bodily Resurrection,” Religious Studies 43 (2007), 373–94. 4 For the dating of Aquinas’s quodlibet questions, see Jean-Pierre Torrell, Saint Thomas Aquinas: The Person and His Work, rev. ed. (Washington, DC: Catholic University of America Press, 2005), 207–12.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

200

Turner C. Nevitt

numerically the same if it has been annihilated. His reply is worth quoting in full: Response. It should be said that there is a certain difference to be noticed among those things that can be reduced to nothing. For there are some things whose unity has continuity of duration in its very notion, as is clear in the case of motion and time, and for this reason the interruption of such things is indirectly contrary to their unity according to number. But things that imply a contradiction are not contained under the number of things possible to God, since they fall short of the notion of being. And therefore if things of this kind are reduced to nothing, God cannot restore them the same in number: for this would be to make contradictories true at the same time, namely if there were one interrupted motion. But there are other things whose unity does not have continuity of duration in its very notion, as the unity of permanent things, except per accidens, insofar as their existence is subjected to motion; for as such, things of this kind are both measured by time and their existence is one and continuous according to the unity and continuity of time. And since a natural agent cannot produce these things without motion, hence it is that a natural agent cannot restore things of this kind the same in number, if they have been reduced to nothing or if they have been corrupted according to substance. But God can restore things of this kind both without motion, since it is in his power to produce effects without intermediate causes; and therefore he can also restore them the same in number, even if they have fallen into nothing.5

Wood notes a few immediately obvious features of this passage.6 Aquinas clearly thinks that there are some things that God cannot bring back from nothing and other things that God can. The relevant distinction is between successive things, whose unity involves continuity of duration in its very notion, and permanent things, whose unity does not. The annihilation and re-creation of the former successive things is impossible even to God, Aquinas explains, because their annihilation would involve an interruption in the continuity of duration that makes them one and the same thing. But the annihilation and re-creation of the latter permanent things would not involve such a contradiction, since the notion of their unity does not involve continuity of duration. And therefore God can annihilate the latter permanent things and re-create them numerically the same. But only God can do this; natural agents cannot. So Aquinas clearly thinks that there are some things God can annihilate and re-create numerically the same. To that extent he recognizes the metaphysical possibility of intermittent or “gappy” existence. But, according

5 Quaestiones disputatae de quolibet (QQ) IV, q. 3, a. 2. References to the works of Aquinas are to the Busa-Alarcon edition available online at . All translations are my own. 6 Wood, “Mind the Gap?” 103–5.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Don’t Mind the Gap

201

to Wood, the important question is still which things Aquinas thinks God can annihilate and re-create.7 As Wood notes, Capreolus and Ferrara both answer, “all permanent things.” But Wood suggests that they have failed to notice another crucial distinction in the above reply.8 After noting that there are some things whose unity does not involve continuity of duration in its very notion, as the unity of permanent things, Aquinas adds, “except per accidens, insofar as their existence is subjected to motion; for as such, things of this kind are both measured by time and their existence is one and continuous according to the unity and continuity of time.” I agree with Wood’s suggestion that Aquinas is distinguishing here between permanent things such as immaterial substances and the heavenly bodies on the one hand and permanent things such as sublunary material substances on the other.9 The latter permanent things have a unity involving continuity of duration per accidens because their existence is subjected to motion and is therefore one and continuous according to the unity and continuity of time. But I disagree with Wood’s further suggestion that Aquinas introduces this distinction here in order to set aside the latter permanent things from those that God can annihilate and re-create numerically the same.10 For it is precisely the latter permanent things that Aquinas goes on to explain cannot be restored numerically the same by any natural agents, but can be restored numerically the same by God. A few further features of the above reply make this clear. From the point at which he first distinguishes the permanent things whose unity only involves continuity of duration per accidens until the very end of the reply, Aquinas repeats the phrase “things of this kind,” which I put in bold earlier, in order to continually refer back to just these permanent things whose unity only involves continuity of duration per accidens. Aquinas does not set these permanent things aside after distinguishing them; his whole subsequent discussion is about just these permanent things. And there is no room to doubt that Aquinas means to include sublunary material substances among the permanent things that he concludes can be annihilated and restored numerically the same by God, because Aquinas explicitly mentions such substances when drawing his conclusion. He says, “a natural agent cannot restore things of this kind the same in number, if they have been

7 Ibid., 105. 8 Ibid., 106. 9 Ibid., 107. 10 Ibid., 108–9. The same two suggestions are made by Gyula Klima and followed by Patrick Toner. See Klima, “The Problem of ‘Gappy Existence’ in Aquinas’s Metaphysics and Theology,” in S. Ogden, A. Hall, and G. Klima (eds), The Metaphysics of Personal Identity: Proceedings of the Society for Medieval Logic and Metaphysics, vol. 13 (Newcastle Upon Tyne: Cambridge Scholars Publishing, 2016), 119–133 and Patrick Toner, “St. Thomas Aquinas on Gappy Existence,” Analytic Philosophy 56 (2015), 94–110.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

202

Turner C. Nevitt

reduced to nothing or if they have been corrupted according to substance. But God can restore things of this kind . . . the same in number, even if they have fallen into nothing.” The fact that Aquinas mentions things that have been corrupted according to substance among the permanent things that cannot be restored by any natural agent, but can be restored by God, makes it clear that he means to include sublunary material substances among such permanent things, since sublunary material substances are the only corruptible ones; angels and the heavenly bodies are incorruptible. So it is not that Capreolus and Ferrara have failed to notice the distinction Aquinas draws between permanent things such as angels and the heavenly bodies and permanent things such as sublunary material substances. It is rather that Aquinas draws that distinction precisely in order to go on to explain that the latter permanent things can still be annihilated and re-created numerically the same, but only by God. The annihilation and re-creation of permanent things such as sublunary material substances would not involve a contradiction because the notion of their unity only involves continuity of duration per accidens insofar as their existence is subjected to motion. Elsewhere I have tried to explain what Aquinas means by saying that such permanent things have an existence subjected to motion, which is therefore measured by time and is one and continuous according to the unity and continuity of time.11 Wood comes very close to suggesting that Aquinas means that their existence is itself composed of parts ordered one after another in time.12 If their existence were so composed, then an interruption in their existence would involve the same contradiction Aquinas says would be involved in the interruption of motion itself. But if Aquinas thought this about the existence of such permanent things, then he would be contradicting himself right in the reply I cited earlier, since it is precisely these permanent things that he concludes God can annihilate and re-create numerically the same. Fortunately there is a good case to be made that Aquinas does not in fact think that the existence of sublunary material substances is itself composed of parts ordered one after another in time. Since I have tried to make this case elsewhere, I shall not say more about it here. I admit that much more work needs to be done on Aquinas’s view of permanent and successive entities and especially on the relation of their 11 Turner C. Nevitt, “Annihilation, Re-creation, and Intermittent Existence in Aquinas,” in S. Ogden, A. Hall, and G. Klima (eds), The Metaphysics of Personal Identity: Proceedings of the Society for Medieval Logic and Metaphysics, vol. 13 (Newcastle Upon Tyne: Cambridge Scholars Publishing, 2016), 101–17. 12 Wood, “Mind the Gap?” 107–8. Klima makes the same suggestion. See his “The Problem of ‘Gappy Existence’ in Aquinas’s Metaphysics and Theology.”

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Don’t Mind the Gap

203

existence to motion and time. Perhaps such work will make what he says in the reply cited earlier even clearer, but it is not needed in order to see that Aquinas clearly concludes there that God alone can annihilate permanent things such as sublunary material substances—again the only corruptible ones—and then re-create them numerically the same. Given the view Aquinas clearly expresses in this late quodlibet question on annihilation and re-creation, I think the evidence Wood presents for Aquinas’s endorsement of the principle of non-repeatability bears reconsidering. Some of Wood’s evidence comes from Aquinas’s replies to a number of objections to the possibility of the resurrection in the Commentary on the Sentences, the Summa Contra Gentiles, the Compendium of Theology, and the Questions on the Soul.13 The rest of his evidence comes from Aquinas’s discussions of a problem about the Eucharist in the Commentary on the Sentences and in the Summa Theologiae.14 The principle of non-repeatability actually appears in a number of other texts as well.15 I presume that Wood focuses on just a few of these texts because they make the best case for Aquinas’s endorsement of the principle of non-repeatability. And I agree: if Aquinas’s endorsement of the principle cannot be established from these texts, then it is very unlikely that it can be established at all, at least not without some new evidence. But I think Aquinas’s endorsement of the principle cannot be established from these texts. The texts containing objections to the resurrection all have a similar form: what corrupts or falls into nothing cannot return numerically the same; something essential to a human being, such as the sense powers or the form of corporeity, corrupts or falls into nothing at death; therefore, the numerically same human being cannot return at the resurrection. Aquinas’s general strategy in reply to these objections is to deny that anything essential to a human being ceases to exist at death. But he has two different versions of this strategy. Sometimes Aquinas points out that, although essential, the thing in question does not in fact corrupt or fall into nothing at death, since it survives in the separated soul. When Aquinas adopts this version of the strategy he does not challenge the principle of non-repeatability stated in the objection. In fact that is true of his replies to a host of similar objections. His 13 Scripta super libros sententiarum (In Sent.) IV, d. 44, q. 1, a. 1, q.c. 2, obj. 3& ad 3; Summa contra gentiles (SCG) IV, c. 80, n. 3 & c. 81, n. 6; Compendium theologiae (CT) I, c. 154; Quaestiones disputatae de anima (QDA) q. 19, ad 13. 14 In Sent. IV, d. 12, q. 1, a. 2, qc. 4; Summa theologiae (ST) III, q. 77, a. 5. 15 Cf. In Sent. IV, d. 22, q. 1, a. 1; In Sent. IV, d. 44, q. 1, a. 1, qc. 1, obj. 4 & ad 4; In Sent. IV, d. 44, q. 1, art. 1, qc. 2, obj. 1–2 & ad 1–2; In Sent. IV, d. 44, q. 3, a. 3, qc. 1, obj. 6 & ad 6; SCG IV, c. 80, nn. 2 & 4 & c. 81, nn. 5 & 7–11; QDA q. 19, obj. 5 & ad 5; Commentaria in libros Physicorum V, l. 6, n. 6 & n. 8.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

204

Turner C. Nevitt

silence in these replies could be read as an implicit endorsement of the principle of non-repeatability. Elsewhere I have tried to give what I think is a better explanation of Aquinas’s silence here.16 But sometimes in reply to these objections Aquinas points out instead that the thing in question, such as the sense powers or the form of corporeity, is merely accidental, so that even if it corrupts or falls into nothing at death, its loss is irrelevant to the resurrection of the numerically same human being. And, as Wood points out, by adopting this version of the strategy Aquinas does in fact seem to be granting the principle of non-repeatability.17 Why else would Aquinas admit that we will not get back the numerically same sense powers at the resurrection, even though he says we will have the numerically same sensitive soul and the numerically same sense organs? Capreolus and Ferrara suggest that perhaps Aquinas is merely granting the principle of non-repeatability for the sake of argument in order to show that the resurrection of the numerically same human being is still possible. But Wood finds this suggestion implausible. “Why,” he asks, “would Aquinas concede strong non-repeatability for the sake of argument in a discussion of the resurrection? Moreover, why would he do so without explaining that this is what he is up to?”18 Well, Capreolus and Ferrara also suggest that perhaps Aquinas concedes the principle of nonrepeatability because that was the general opinion of the learned at the time and particularly of the philosophers to whom Aquinas is replying.19 That does not seem implausible to me, especially considering that the medieval debates about intermittent existence did not heat up until after Aquinas’s death. And in some of his replies Aquinas does in fact indicate that he is merely granting the principle of non-repeatability for the sake of argument. In reply to the objection about the sense powers from the Questions on the Soul, for example, Aquinas concludes, “Hence it is not necessary for the resurrected eye to be different, granted that (licet) the sense power is different.”20 He says the same thing in reply to the objection about the form of corporeity in the Compendium of Theology. “Hence granted that (licet) it does not return numerically the same, there is no impediment to the unity of the subject . . . ”21 Now we usually translate “licet” as “although,” but “although” is stronger and more assertive, whereas “licet” is weaker and more concessive. When used as a verb, “licet” means “it is allowed” or “it is 16 Turner C. Nevitt, “Survivalism, Corruptionism, and Intermittent Existence in Aquinas,” History of Philosophy Quarterly 31 (2014), 1–19. 17 Wood, “Mind the Gap?” 110–12. 18 Ibid., 113. 19 Capreolus, Defensiones, 23; Ferrara, Commentaria, 256. 20 QDA q. 19, ad 13 (emphasis mine). 21 CT I, c. 154 (emphasis mine).

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Don’t Mind the Gap

205

permitted,” which shades its meaning when used as a conjunction. Thus some dictionaries offer “granting that” or “granted that” as a possible translation of the conjunction “licet.” And I think “licet” likely has this more reserved sense in these replies. For Aquinas shows a similar reserve in reply to the same kind of objection in the Summa Contra Gentiles. “Even if (etsi) this corporeity fell into nothing when the human body corrupts,” he says, “nevertheless it cannot impede the numerically same [human being] rising.”22 Aquinas’s use of “licet” and “etsi” in these replies makes it sound as if he is merely granting for the sake of argument that the sense powers and the form of corporeity will not be restored numerically the same at the resurrection. I should admit that in a passage from the Sentences Commentary that Wood does not mention Aquinas raises the same objection about the corruption of the sense powers and there his reply seems less reserved. At the resurrection, he says, “the organs will be numerically the same, although (quamvis) the powers are not numerically the same.”23 But if you translate this “quamvis” more literally as “however much you like” or “as much as you will,” then it will again sound as if Aquinas is merely granting the principle for the sake of argument. And in the reply that Wood himself quotes from the Sentences Commentary Aquinas only says, “according to some the sense powers do not remain.”24 Given his reserve in so many of these replies, it does not seem implausible to me to suggest, as Capreolus and Ferrara do, that Aquinas is merely granting the principle of non-repeatability for the sake of argument in these contexts. Wood’s other evidence for Aquinas’s endorsement of the principle of non-repeatability is from Aquinas’s discussions of a problem about the Eucharist. The question under discussion is whether anything can be generated from the Eucharistic species. Aquinas thinks it is evident that something can be generated from them, since, for example, consecrated hosts turn to ash when they are burned and turn to worms when they are eaten by them. But Aquinas still considers a number of rival theories about how this could be possible. One theory says that the substance of bread and wine is annihilated at the consecration and then returns when the Eucharistic species are corrupted. But Aquinas says that this is impossible. As he explains in the Summa: If the substance of bread and wine is annihilated, it cannot return again, since what has fallen into nothing does not return numerically the same, unless perhaps the

22 SCG IV, c. 81, n. 7 (emphasis mine). 23 In Sent. IV, d. 44, q. 3, a. 3, qc. 1, ad 6 (emphasis mine). 24 In Sent. IV, d. 44, q. 1, a. 1, qc. 2, ad 3 (emphasis mine). Cf. Wood, “Mind the Gap?” 110, n. 30.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Turner C. Nevitt

206

aforementioned substance is said to return because God creates a new substance de novo in place of the first.25

In the Sentences commentary Aquinas offers the same criticism.26 And in both places he seems to be endorsing the principle of non-repeatability, as Wood notes.27 But Capreolus suggests that when Aquinas says that what is annihilated does not return, “he means that this cannot happen without a new miracle, as is clear from his words at the end of the article.”28 Now Capreolus is right that at the end of this article of the Summa Aquinas criticizes and rejects another rival theory because it would involve a new miracle other than the consecration itself. But Aquinas does not reject the possibility of the return of the numerically same substance here because it would involve a new miracle; he rejects it because it would violate the principle of non-repeatability, which he explicitly cites. In fact Aquinas grants the possibility of a new miracle other than the consecration here when he says that God could create de novo a new substance in place of the first. So it seems very unlikely that he only means “without a new miracle” when he says that what has fallen into nothing does not return numerically the same. Ferrara suggests two other ways of taking Aquinas’s use of the principle of non-repeatability here. First he suggests that Aquinas is only talking about the limits of the power of natural agents, rather than the limits of divine power.29 Capreolus makes the same suggestion.30 Again Wood finds this implausible. And it is implausible at first glance, since Aquinas is discussing the Eucharist, which is itself a miracle, as Wood notes.31 But Ferrara points out that Aquinas uses the principle of non-repeatability here to criticize a rival theory that itself involves the power of natural agents.32 Remember the question under discussion is how anything can be generated from the Eucharistic species when they are corrupted by being burned by fire or eaten by worms. And the theory Aquinas is criticizing claims that “the substance of bread and wine returns in the very corruption of the species.”33 Talk of generation and corruption clearly signals the involvement of natural agents. So it does not seem implausible to me to suggest that when Aquinas says here that the substance of bread and wine does not return numerically the same once it has fallen into nothing, he just means that it does not return through the power of natural agents such as fire or worms. This suggestion

25 27 29 31 33

ST III, q. 77, a. 5. 26 In Sent. IV, d. 12, q. 1, a. 2, qc. 4. Wood, “Mind the Gap?” 112. 28 Capreolus, Defensiones, 23. Ferrara, Commentaria, 256. 30 Capreolus, Defensiones, 23. Wood, “Mind the Gap?” 113. 32 Ferrara, Commentaria, 256. ST III, q. 77, a. 5.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Don’t Mind the Gap

207

gets further support from Aquinas’s own interpretation of the principle of non-repeatability from Aristotle’s De generatione et corruptione. Wood laments the fact that Aquinas did not finish his commentary on the De generatione, since he thinks Aquinas’s views might have been clearer if we had his own reading of Aristotle’s principle of non-repeatability.34 But Capreolus points out that Aquinas did in fact give his own reading of the principle in reply to one of the objections I discussed earlier, which cites the principle against the possibility of the resurrection.35 The objection quotes Aristotle in book two of the De generatione saying, “what has a moved corruptible substance does not return numerically the same.” And in reply Aquinas says: In response to the first objection it should be said that the philosopher is speaking about the iteration that is made through motion or natural mutation. For he shows the difference between the circulation that occurs in generation and corruption and the circulation that occurs in the motion of heaven: for by local motion heaven returns numerically the same to the starting point of its motion, since it has an incorruptible moved substance; but by generation generable and corruptible things return specifically the same, but not numerically the same; since from a human being semen is generated, from which blood [is generated]; and so successively it gets to a human being, not numerically the same, but specifically; and from fire air is generated, from which water [is generated]; from water earth [is generated], and from earth fire [is generated], not numerically the same, but specifically. Hence it is clear that the reason brought forward, according to the intention of the philosopher, is not relevant to the proposed.36

In reply to a similar objection in the Summa Contra Gentiles Aquinas explains why the principle of non-repeatability is not relevant to the proposed question about the possibility of the resurrection. “The power of nature falls short of the divine power,” he says. “Hence by an operation of nature, what is corrupted cannot be restored numerically the same. But . . . divine power . . . can restore corrupted things to wholeness.”37 On the basis of his reply Capreolus concludes that Aquinas understands the principle of non-repeatability as a comment about the limits of natural mutation rather than the limits of supernatural action.38 And of course this is exactly the position Aquinas takes in the reply of his late quodlibet question on annihilation and re-creation. In fact Aquinas cites Aristotle’s principle of non-repeatability in the first objection to the quodlibet question, as Wood also notes.39 And after giving his reply Aquinas says, “Hence

34 Wood, “Mind the Gap?” 113. 35 Capreolus, Defensiones, 22–3. 36 In Sent. IV, d. 44, q. 1, a. 1, qc. 2, ad 1. 37 SCG IV, c. 81, n. 5. 38 Capreolus, Defensiones, 23. 39 Wood, “Mind the Gap?” 102.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

208

Turner C. Nevitt

the response to the first [objection] is clear.”40 Given Aquinas’s own interpretation of Aristotle’s principle of non-repeatability as a comment on the limits of natural power, it does not seem implausible to me to suggest, as Capreolus and Ferrrara do, that Aquinas is only using the principle as a comment on the limits of natural power when he cites it against a theory involving natural agents in his discussions of the Eucharist. Aquinas gives the same merely natural reading of the principle in reply to some of the objections citing it against the resurrection, as we just saw. But then why does Aquinas go on to say that the same substance could return in the sense that God could create de novo a new substance in place of the first? If Aquinas thought that God could restore the numerically same substance, as he says in his quodlibet question cited earlier, then why wouldn’t he just say that here instead? I think Ferrara’s second suggestion is helpful at this point. He suggests that perhaps by “possible” and “impossible” here Aquinas just means “fitting” and “unfitting,” since it would be unfitting for God to superfluously annihilate something if he intended to re-create it.41 If he is right that Aquinas thinks the annihilation and re-creation of the numerically same substance would be unfitting, then it would make sense for Aquinas not to grant that possibility to the theory he is criticizing. And Aquinas does in fact think that annihilation would be unfitting, as Wood himself notes.42 Aquinas says in a number of places that he thinks God would never annihilate anything.43 But if Aquinas thinks God would never annihilate anything, then he also thinks God would never annihilate and re-create anything either. So perhaps that is why Aquinas does not grant such an absolute possibility to the theory he is criticizing in his discussion of the Eucharist. Another possible explanation is that the proponents of the theory Aquinas is criticizing did not themselves believe in the possibility of annihilation and re-creation. If they did not believe in such a possibility, then it would make sense for Aquinas not to grant them that possible way out, but to suggest something else instead. So in light of a close reading of the relevant texts I think that at least some of the strategies Capreolus and Ferrara developed for dealing with Aquinas’s use of the principle of non-repeatability appear far more plausible than Wood allows, which means that his textual case for Aquinas’s own endorsement of the principle is far weaker than he realizes. In this connection I think it is also worth repeating a point made by Capreolus.44 He points out 40 QQ IV, q. 3, a. 2, ad 1. 41 Ferrara, Commentaria, 256. 42 Wood, “Mind the Gap?” 113. 43 E.g. Quaestiones disputatae de potentia q. 5, a. 4; QQ IV, q. 3, a. 1; ST I, q. 65, a. 1, ad 1; ST I, q. 104, a. 4. 44 Capreolus, Defensiones, 22.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Don’t Mind the Gap

209

that in all the texts in which Aquinas appears to grant or endorse the principle of non-repeatability he is discussing the views of others on another topic and in the course of doing so he mentions the principle incidentally. So it is possible that Aquinas’s use of the principle in these contexts says more about the mind of the people whose views he is discussing than it does about his own mind, as Ferrara also suggests.45 The only place where Aquinas directly and thoroughly answers the question whether an annihilated material thing can be restored numerically the same is in his late quodlibet question on the topic. And there he clearly says that it can, but only by God. Of all the texts, then, his late quodlibet question should be taken the most seriously, since it presumably represents his most considered personal view on the question. In addition to the textual case, however, Wood also makes a philosophical case for Aquinas’s endorsement of the principle of non-repeatability.46 Since his philosophical case is based upon a disputed interpretation of Aquinas’s theory of matter as the principle of individuation, it would be possible to avoid Wood’s conclusion simply by rejecting his interpretation of Aquinas’s theory of individuation. But I think one can accept his interpretation of the theory and still avoid his conclusion, because the conclusion does not in fact follow from his interpretation of the theory. According to Wood’s interpretation, a material substance is individuated by its form, which is itself individuated by the particular spatial and temporal dimensions of the matter in which it starts to exist. This interpretation, he argues, has the consequence that once a material thing’s substantial form is destroyed, not even God can bring the material thing back numerically the same. As he explains: So if a given substantial form informs matter under some particular spatio-temporal dimensions it is necessarily numerically distinct from any substantial form coming into matter at a different place or time. Hence the substances resulting from uniting form and matter at two different times or places are also necessarily numerically distinct. No single substance can have two distinct sets of spatial and temporal starting points.47

Wood goes on to note the further consequence that the continued existence of the separated soul, a human being’s substantial form, between death and resurrection is necessary for the numerically same human being to be restored to life at the resurrection. If my soul perished at death along with my body, then not even God could bring me back to life, since that would give my substantial form, and therefore me, two different spatio-temporal starting points. 45 Ferrara, Commentaria, 256. 47 Ibid., 117.

46 Wood, “Mind the Gap? ” 114–18.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

210

Turner C. Nevitt

Wood’s argument here seems to me to beg the question against the possibility of annihilation and re-creation by assuming that every coming to exist is a start rather than a re-start. Given Wood’s interpretation of Aquinas’s theory of individuation, it is true that one and the same thing cannot have two different spatio-temporal starting points.48 But a re-created thing does not have two different spatio-temporal starting points; it has the one and the same spatio-temporal starting point that it always had. A thing’s spatio-temporal starting point is the point in space and time at which it came to exist first after never having existed. But re-creation is not a different start; re-creation is a re-start, that is, a second coming to exist after having once existed and then ceased to exist. Given Wood’s interpretation of Aquinas’s theory of individuation, a contradiction would indeed follow from supposing that one and the same material thing has two different spatio-temporal starts—two different first comings to exist—but no contradiction would follow from supposing that one and the same thing has a start and then a re-start—a first coming to exist and then a second one after the first.49 That is not to say that no contradiction whatsoever would follow from supposing that one and the same material thing is annihilated and then re-created. Perhaps a contradiction can be generated in some other way. But it cannot be generated by assuming that every coming to exist is a start, since that assumption begs the question against the very possibility of re-starts, and so also against the possibility of re-creation. Perhaps I am wrong about this. Perhaps Wood’s interpretation of Aquinas’s theory of individuation does indeed imply that the annihilation and recreation of the numerically same sublunary material substance is impossible. If that is the case, though, I think it counts against Wood’s interpretation of Aquinas’s theory of individuation. For the view Aquinas expresses on annihilation and re-creation in his late quodlibet question is much clearer than his view of matter as the principle of individuation. I presume that is why his Renaissance interpreters all agreed about Aquinas’s view on the possibility of annihilation and re-creation.

48 In fact on one reading this seems trivially true, as Dean Zimmerman says, since nothing can have two earliest moments of existence. See Dean Zimmerman, “The Compatibility of Materialism and Survival: The ‘Falling Elevator’ Model,” Faith and Philosophy 16 (1999), 194. 49 I presume this is why the standard thought experiments in favor of origins essentialism involve conceiving of a possible world in which a thing has a different origin from the one it has in the actual world, rather than conceiving of a possible world in which a thing has its actual origin and then a re-origin after having ceased to exist. The necessity of origin has no direct bearing on the possibility of intermittent existence precisely because coming back into existence after having ceased to exist is not a matter of a different origin, but of a re-origin.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Don’t Mind the Gap

211

Wood argues that Cajetan in fact disagreed with Capreolus and Ferrara about this. But I think Wood is mistaken, which can be seen by examining Cajetan’s commentary on Aquinas’s discussion of the Eucharist in the Summa.50 In the course of that discussion, as we saw earlier, Aquinas considers a few theories about how something can be generated from the Eucharistic species when they corrupt. According to one theory, the substance of bread and wine is annihilated at the consecration and then returns in the corruption of the Eucharistic species. But Aquinas says that this is impossible, since what has fallen into nothing does not return numerically the same. In his commentary on this passage Cajetan says: Do not be mistaken about [the author’s] refutation of the other opinions, when in this article you read many things asserted [as] impossible, which, according to the teaching of the author, can be done by divine omnipotence. It is necessary to remember the author’s custom. For on each kind [of topic] he is accustomed to speak and judge and assert [things] according to the order of that kind [of topic]. And, therefore, when he treats of the sacraments here, he means and asserts “possible” or “impossible” according to the sacramental order, as we also said above. It is according to that order that what is annihilated is said not to be able to return. . . . So do not be worried about responding to [those who] object to the author [on the grounds] that he contradicts himself.51

Now I think Cajetan was wrong to suggest that Aquinas’s use of the principle of non-repeatability here is restricted merely to what is possible or impossible according to the sacramental order that God has established. I gave my reasons for thinking so earlier when I discussed a similar suggestion by Capreolus. But Cajetan was clearly impelled to make such a suggestion, however bad, because he recognized that elsewhere Aquinas says that by his omnipotence God can restore what has been annihilated numerically the same. Cajetan says as much himself. And there can be no doubt that he has sublunary material substances in mind, since Aquinas is discussing the annihilation and return of the substance of bread and wine. So what should we make of the passage that Wood cites from Cajetan’s commentary on the De ente et essentia?52 Well, Cajetan does not say there that the annihilation and re-creation of the numerically same human being is impossible. All he says is that the soul “should not be annihilated if the 50 ST III, q. 77, a. 5. 51 Thomas de Vio (Cajetan), Commentaria in summa theologiae S. Thomae de Aquino, in Sancti Thomae Aquinatis opera omnia, vol. 12 (Rome: Leonine Commission, 1906), 200–1. 52 Cajetan, Commentary on Being and Essence (In De Ente et Essentia d. Thomas Aquinatis), tr. L. Kendzierski and F. Wade (Milwaukee: Marquette University Press, 1964), 201. Cf. Wood, “Mind the Gap?” 121.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

212

Turner C. Nevitt

notion of the resurrection is to be saved.”53 Given his sensitivity to what God has established in this or that order, Cajetan probably just means that the theological notion of the resurrection requires the survival of the soul after death. Annihilation and re-creation is not an acceptable model of the resurrection for Cajetan, not because he thinks it is absolutely impossible, but because he thinks it is not what God has in fact established. Other Catholic doctrines such as purgatory and the intercession of the saints make the survival of the separated soul a theological constraint for Cajetan. Any model of the resurrection that violates that constraint, as annihilation and re-creation does, would not be theologically acceptable to him. Since he thinks God has established that souls survive death, Cajetan must also think that the theological notion of the resurrection—what God has in fact promised to do in promising to bring us back to life—involves the survival of the soul as well. The same is true for Aquinas. So I do not think the view Aquinas clearly expresses in his quodlibet question on annihilation and re-creation requires any revision in our understanding of what he thinks is required for the resurrection of the dead. Aquinas does think the survival of the soul is required for the resurrection. But his reasons for thinking so are not what the majority of his contemporary interpreters have supposed. His contemporary interpreters have wrongly supposed that Aquinas includes the soul’s survival in his account of the resurrection because he thinks the uninterrupted existence of a material substance’s form is metaphysically necessary for the identity of such a substance over time. Aquinas’s Renaissance interpreters rightly recognized that Aquinas reserves to God the absolute power to annihilate sublunary material substances and then re-create them numerically the same. That means that Aquinas thinks the uninterrupted existence of nothing other than God’s power is metaphysically necessary for the identity of such substances over time. But the soul’s uninterrupted existence is still required for the resurrection, because the resurrection is a theological notion with its own theological constraints. So all his Renaissance interpreters challenge us to do is to revise our understanding of what Aquinas thinks is metaphysically necessary for the identity of material substances over time. And I hope to have shown that, in spite of Wood’s attempt to prove otherwise, this Renaissance challenge still retains much of its force. University of San Diego

53 Cajetan, Commentary, 201 (emphasis mine).

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Don’t Mind the Gap

213

BIBLIOGRAPHY Francis Sylvester Ferrara. Commentaria in libros quatuor contra gentiles S. Thomae de Aquino, in Sancti Thomae Aquinatis opera omnia, vol. 15 (Rome: Leonine Commission, 1950). Klima, Gyula. “The Problem of ‘Gappy Existence’ in Aquinas’s Metaphysics and Theology,” in S. Ogden, A. Hall, and G. Klima (eds), The Metaphysics of Personal Identity: Proceedings of the Society for Medieval Logic and Metaphysics, vol. 13 (Newcastle Upon Tyne: Cambridge Scholars Publishing, 2016), 119–33. Nevitt, Turner C. “Survivalism, Corruptionism, and Intermittent Existence in Aquinas,” History of Philosophy Quarterly 31 (2014), 1–19. Nevitt, Turner C. “Annihilation, Re-creation, and Intermittent Existence in Aquinas,” in S. Ogden, A. Hall, and G. Klima (eds), The Metaphysics of Personal Identity: Proceedings of the Society for Medieval Logic and Metaphysics, vol. 13 (Newcastle Upon Tyne: Cambridge Scholars Publishing, 2016), 101–17. Thomas Aquinas. Sancti Thomae de Aquino opera omnia, ed. R. Busa and E. Alarcon (2000–) . Thomas de Vio (Cajetan). Commentaria in summa theologiae S. Thomae de Aquino, in Sancti Thomae Aquinatis opera omnia, vol. 12 (Rome: Leonine Commission, 1906). Thomas de Vio (Cajetan). Commentary on Being and Essence (In de ente et essentia d. Thomae Aquinatis), tr. L. Kendzierski and F. Wade (Milwaukee: Marquette University Press, 1964). Toner, Patrick. “St. Thomas Aquinas on Gappy Existence,” Analytic Philosophy 56 (2015), 94–110. Torrell, Jean-Pierre. Saint Thomas Aquinas: The Person and His Work, rev. ed. (Washington, DC: Catholic University of America Press, 2005). Van Dyke, Christina. “Human Identity, Immanent Causal Relations, and the Principle of Non-Repeatability: Thomas Aquinas on the Bodily Resurrection,” Religious Studies 43 (2007), 373–94. Wood, Adam. “Mind the Gap? The Principle of Non-Repeatability and Aquinas’s Account of the Resurrection,” Oxford Studies in Medieval Philosophy 3 (2015), 99–127. Zimmerman, Dean. “The Compatibility of Materialism and Survival: The ‘Falling Elevator’ Model,” Faith and Philosophy 16 (1999), 194–212.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Briefly Noted Richard Cross. Duns Scotus’s Theory of Cognition (Oxford University Press, 2014), 224 pp. As the incredibly messy remains of John Duns Scotus’s literary output come increasingly into critical focus, our philosophical understanding of his work progresses apace. In this volume, Richard Cross attempts a detailed survey of Scotus’s main ideas about cognitive theory. The volume puts on display an impressive grasp of the far-flung textual evidence, taking advantage of all the latest critical editions, and frequently attending to manuscripts that remain unedited. Scotus’s ideas in this area have been somewhat less studied than in some other domains, and understandably so, because he has not in general seemed to be the innovator here that he was in metaphysics, ethics, and will. Cross’s study bears this conclusion out in some respects. But in other respects Cross shows Scotus to have wonderfully inventive and (of course) subtle things to say: particularly on the distinction he founded between intuitive and abstractive cognition, and on the notion of esse intelligibile, where Cross argues, controversially but compellingly, that Scotus’s account should be understood reductively: “esse intelligibile is not a case of esse at all” (192). Peter of Spain. Summaries of Logic: Test, Translation, Introduction, and Notes. Translated by Brian P. Copenhaver with Calvin Normore and Terence Parsons (Oxford University Press, 2014), 547 pp. Although there have been several previous English translations of this famous treatise, this one will surely become canonical for the foreseeable future. The text’s various intricacies are mapped onto English with an abundance of care—to give just one example, where one Latin word is translated by several in English, the English words are hyphenated. Detailed footnotes stud the text throughout. The Latin of De Rijk’s edition is supplied on facing pages. A long introduction supplies extensive information of both an historical and a philosophical kind. Why lavish so much care? The answer from the Preface is worth quoting at length: “In the case of Peter’s Summaries of Logic, because his book was the basis for teaching logic for nearly four centuries when that subject was the heart of what we now call the ‘undergraduate curriculum.’ Since no book about logic was read by more people until the twentieth century, the Summaries has extensively and profoundly influenced the distinctly Western way of speaking formally and writing formal prose by constructing well-formed sentences, putting them together in valid arguments, and then defending those arguments in debate against refutation. Few works, apart from the Authorized Version of the English Bible and the collected plays of Shakespeare, have been more influential than Peter’s Summaries” (ix). Jacopo Zabarella, On Methods and On Regressus, tr. John P. McCaskey (Harvard, 2014), vol. 1: xxvi + 323 pp.; vol. 2: 470 pp. These beautifully produced volumes provide, for the first time, a large chunk of Zabarella’s philosophy in English— together with a carefully revised edition of the Latin on facing pages, a useful introduction, and punctilious scholarly endnotes. The works translated are ones

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

216

Briefly Noted

for which very large claims of influence on modernity have sometimes been made. McCaskey’s introduction, however, as well as the works themselves, show that in fact these are characteristic products of the scholastic era, and if anything rather conservative examples of the genre. Zabarella’s concern here is to promote the orthodox methods of the Posterior Analytics. In doing so, he introduces various interesting innovations of his own, particularly concerning the famous regressus that takes us from quia to propter quid demonstrations. Even so, there is little to forewarn of the deluge that was to come in the following century. Christopher Hughes, Aquinas on Being, Goodness, and God (Routledge, 2015), 370 pp. This is a contribution to the genre of history for the philosopher, the sort of book that would rather arrive at philosophical truth than historical accuracy. The topics are just as the title indicates, which is to say they range widely over central themes in Aquinas, but by no means attempt to cover the full range of his philosophy. The discussion is pitched at a sophisticated philosophical level—this is not a book for beginners, nor for anyone who does not share the interests in analytic philosophy that preoccupy Chris Hughes. But for the right sort of reader, this is quite a stimulating book, which engages with the ideas in a way that is thoroughly informed about both Aquinas and recent analytic philosophy. Hughes brings to his topic a creative and original mind, and a flair for winning examples, as when he contemplates what happens when you mix espresso, grappa, and cocaine—a caffè supercorretto. It’s an example for something, though of course the example lingers longer than does the elaborate philosophical point. Jari Kaukua, Self-Awareness in Islamic Philosophy: Avicenna and Beyond (Cambridge UP, 2015), 257 pp. In contrast to the usual philosophical history of the self that runs through Augustine and Descartes, Jari Kaukua argues here for the importance of the Islamic tradition, and of Avicenna in particular. Avicenna is argued to have developed a wholly new philosophical conception of self-awareness, one that it is primitive and immediate, on which I grasp my self in a way that is wholly distinct from my grasp of the objects that cognition makes available to me—“an I abstracted from all other considerations” (228). This in turn grounds my knowledge of my existence and identity over time, and underlies key arguments for the soul’s immateriality. Building on a comprehensive reading of Avicenna’s discussions of self-awareness (including, of course, the famous Flying Man Argument), Kaukua proceeds to show how these ideas make their way from Avicenna into later Islamic thought, particularly that of Suhrawardī and Mullaˉ S.adraˉ , who reconsiders Avicenna’s conception of self in important ways. Ruth Glasner, Gersonides: A Portrait of a Fourteenth-Century Philosopher-Scientist (Oxford UP, 2015), 139 pp. This brief but informative book works through the central aspects of Gersonides’s scientific thought: his conceptions of motion, celestial mechanics, empiricism, and cosmology. The philosophical aspects of his views—as we now would think of it—are not given much space, and indeed Glasner documents how Gersonides came to have increasing doubts about the value of metaphysical speculation. It is as a scientist that Gersonides has his principal interest here. On Glasner’s telling his views begin with Aristotle but often end up diametrically opposed, making him an important forerunner of the mechanical science of the seventeenth century.

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

Notes for Contributors OSMP welcomes submissions in all areas of medieval philosophy. Papers received will be evaluated in a timely manner, and an effort to provide significant feedback will be made in every case. To the fullest extent possible, all papers will be refereed according to a triple-blind process, so that neither editor nor referee will know the identity of the author, nor will the author know the identity of the referee. In addition to articles, we welcome editions of texts and brief critical discussions of recently published articles (both in OSMP and elsewhere). Book reviews, however, will be published only when solicited. Submissions should be in English, without author’s name or any other information that would impede blind refereeing. Papers may be of any length, and in particular we welcome the submission of longer works that fall outside the parameters of most journals. Contributors should bear in mind, however, that the lengthier the work, the higher the standard for acceptance. Papers should be submitted as either .pdf or Word-compatible files. The formatting of the initial submission is immaterial, but accepted papers will ultimately need to adhere to OSMP style, as on display in this present volume. All submissions, as well as queries, should be addressed to [email protected].

OUP CORRECTED PROOF – FINAL, 30/8/2016, SPi

OUP CORRECTED PROOF – FINAL, 20/9/2016, SPi

Index Aristotle 45, 133 Averroes 27–58 Bartolomé de Las Casas 13–15 Brower, Jeffrey 184–97 Brunetto Latini 16–17 Bychkov, Oleg 87–92 Cajetan (Thomas de Vio) 199, 211–12 Christine de Pizan 11–12 Cross, Richard 83–4 Elfstrom, Gerard 7–8 Fotion, Nick 7–8 Francisco Peinado 160–3, 167–81 Francisco Suárez 12–13, 125–56 Giuseppe Polizzi 164–5 Gutmann, Amy 22–3 John Capreolus 206, 208–9 John Duns Scotus 64–76, 78–97 Juan de Ulloa Madritano 174 King, Peter 69–70

Luis de Losada 168 Marsiglio of Padua 17–20 Mill, John Stuart 20–2 Nicole Oresme 10–11 Ptolemy of Lucca 15–16 al-Rāzī 116–18 Sandel, Michael 5–8 al-Taftāzānī 100–23 Taylor, Richard 53–4 Thomas Aquinas 8–9, 184–97, 198–212 Thompson, Dennis 22–3 al-T. ūsī 109–11 William of Ockham 74–6 Wood, Adam 198–212