Oxford Studies in Medieval Philosophy Volume 7 9780198845515, 9780198845522, 0198845510

Table of contents :
Cover
Oxford Studies in Medieval Philosophy Volume 7
Copyright
Contents
“Not So Ridiculous”: Avicenna on the Existence of Nature (ṭabī a͑) contra Aristotle and the Ash a͑rites
1. Introduction
2. Nature as a Power
3. Historical-Philosophical Background
4. The Avicennian Argument
5. A Question and a Suggestion
6. Conclusion
Bibliography
Only God Can Make a Tree: Abaelard on Wholes and Parts and Some Evidence of His Later Thinking about Them
1. Introduction
2. Abaelard on Parts, Wholes, and the Nature of Increase
a. The Loci from a Whole and from a Part
b. The Division of Integral Wholes
c. The Growing Square
d. The Identity of the Amputee
3. The Testimony of the Introductiones Montanae Maiores
Bibliography
Accidental Forms as Metaphysical Parts of Material Substances in Aquinas’s Ontology
1. Introduction
2. Two Competing Models
3. Setting the Standard I: Substantial and Accidental Change
4. Setting the Standard II: Unum per Accidens, Unum Simpliciter
5. The Case for Expansion I: Divine Simplicity
6. The Case for Expansion II: Essence and Suppositum
7. Further Expansion I: “Substance” Is Said in Many Ways
8. Further Expansion II: Making Room for Accidental Unities
9. Further Expansion III: Unity, Identity, and Persistence
10. Conclusion
Bibliography
Aquinas on the Intension and Remission of Accidental Forms
1. The Problem of Intension and Remission and Aquinas’s Critique of the Prominent Medieval Solutions
2. Aquinas’s Participation Solution
a. Aquinas on Accidental Forms and Accidental Qualities
b. How Substances Participate in their Accidental Forms
3. Causes of Changes in Degree of Participation in a Form
4. Conclusion
Bibliography
Primary Sources
Secondary Sources
Scotus versus Aquinas on Instrumental Causality
1. Aquinas’s Argument in the Sum of Theology
2. Scotus’s Criticism
3. Aquinas’s Theory of Instrumental Causality
a. Double Action
b. Dispositive and Perfective Causes
4. Scotus’s Theory of Instrumental Causality
a. Secondary Causality
b. Part–Whole Causality
c. Dispositive Causality
d. Borrowed Instrumentality
e. Artificial Instrumentality
4. Conclusions
Bibliography
Primary Sources
Secondary Sources
The Relation-Theory of Mental Acts: Durand of St.-Pourçain on the Ontological Status of Mental Acts
1. The Historical Background
2. Durand’s Motivations
3. John Duns Scotus and the Change Objection
4. The Objection from Agency: Adam Wodeham and an Anonymous Thomist
a. Durand’s Reply to the Objection from Agency
5. Conclusions
Bibliography
Walter Chatton’s Rejection of Final Causality
1. Introduction
a. Plan of the Paper
b. The Problem of Final Causes and Causality: First Approximation
2. Chatton’s Position
a. Final Causes
b. Rejection of Final Causality and Real Final Causes
c. Final Causes as Causes Metaphorically Speaking
d. Why Keep Final Causes?
3. Ockham’s Reaction to Chatton
4. Conclusions
Bibliography
Aquinas on the Sources of Wrongdoing: Themes from McCluskey
Bibliography
Briefly Noted
Notes for Contributors
Index of Names

Citation preview

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

Oxford Studies in Medieval Philosophy Volume 7

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

ADVISORY BOARD Peter Adamson, Ludwig-Maximilians University, Munich Deborah Black, University of Toronto Peter King, University of Toronto Henrik Lagerlund, Stockholm University John Marenbon, Trinity College, Cambridge Calvin Normore, University of California, Los Angeles Dominik Perler, Humboldt University, Berlin Eleonore Stump, St. Louis University Editorial Assistant Mark Boespflug, University of Colorado

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

Oxford Studies in Medieval Philosophy Volume 7 Edited by ROBERT PASNAU

1

OUP CORRECTED PROOF – FINAL, 1/8/2019, 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 2019 The moral rights of the authors have been asserted First Edition published in 2019 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2019947026 ISBN 978–0–19–884551–5 (hbk.) ISBN 978–0–19–884552–2 (pbk.) Printed and bound in Great Britain by Clays Ltd, Elcograf S.p.A. Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

Contents Articles “Not So Ridiculous”: Avicenna on the Existence of Nature (ṭabī a) ͑ contra Aristotle and the Ash ͑arites Davlat Dadikhuda

1

Only God Can Make a Tree: Abaelard on Wholes and Parts and Some Evidence of His Later Thinking about Them Christopher J. Martin

44

Accidental Forms as Metaphysical Parts of Material Substances in Aquinas’s Ontology Jeremy W. Skrzypek

67

Aquinas on the Intension and Remission of Accidental Forms Gloria Frost Scotus versus Aquinas on Instrumental Causality Jean-Luc Solère The Relation-Theory of Mental Acts: Durand of St.-Pourçain on the Ontological Status of Mental Acts Peter John Hartman Walter Chatton’s Rejection of Final Causality Kamil Majcherek

115 147

186 212

Critical Notice Aquinas on the Sources of Wrongdoing: Themes from McCluskey Thomas Williams

243

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

vi



Briefly Noted Adriaenssen—Scotus—Gorman—Wyclif

257

Notes for Contributors Index of Names

259 261

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“Not So Ridiculous” Avicenna on the Existence of Nature (ṭabī a)͑ contra Aristotle and the Ash ͑arites Davlat Dadikhuda

We will indulge in whatever the truth itself reveals of its form, testifying against [its] opponent through [what he] accepts and rejects.¹

1. Introduction In this paper, I set out to explicate what I take to be a distinctive argument that Avicenna offers for the existence of nature (ṭabī a)͑ as a causal power (quwwa) in bodies (ajsām). In doing this, I first clarify the philosophical and historical context of the argument, showing that its two main targets were the Aristotelian tradition on the one hand and the Ash arite ͑ theological (kalām) tradition on the other. With regards to the Aristotelian tradition, which took the existence of nature as a given, I show that the shaykh departs from it in this regard for at least two reasons. The first has to do with a certain feature of how Avicenna conceptualized, consistent with the Aristotelian tradition, the hierarchical relation between the various sciences. That feature is: principles of a lower science must sometimes be proven by a higher one on the hierarchy. The second reason has to do with the influence of Avicenna’s theological contemporaries; for although such thinkers held, with the

¹ Avicenna, The Physics of The Healing, tr. J. McGinnis, 2, modified. (Henceforth as The Healing; Physics, followed by book, chapter, paragraph, and page number(s).)

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

2

 

Aristotelians, that bodies display various kinds of activity or motion, they were anti-realists about any sort of powers and, hence, of nature understood as a power. That is, they denied that bodies behaved in their characteristic ways in virtue of some internal power identifiable with “nature” in the technical Aristo-Avicennian sense. Instead, bodies, they argued, do what they appear to be doing in virtue of a single, powerful, and transcendent being, i.e., God. Avicenna had to meet this challenge, and I show that he meets it in a unique manner—namely, by allowing, with the Ash aris, ͑ the causal involvement of a transcendent being in the production of some effect e from some body x, and yet still showing, against them, that e must occur in virtue of some property F in x, where F makes a real causal contribution to e’s occurrence. In this way, Avicenna attempts to establish the existence of nature qua power and thereby refute occasionalism.

2. Nature as a Power At the beginning of The Healing; Physics I.5, the shaykh states: [T1] We set it down as a posit, which the natural philosopher grants and the metaphysician demonstrates, that the bodies undergoing these motions are moved only as a result of powers (quwā) in them which are principles of their motions and actions.²

The question of causal powers is then later taken up at Metaphysics IV.2 of The Healing.³ In that context, we are presented, by way of an exhaustive division, with an intricate argument for the existence of causal powers in bodies, the conclusion of which is that bodies must have a nature (ṭabī a), ͑ understood in the technical sense as the intrinsic ² The Healing; Physics I.5.3, 39 (tr. J. McGinnis, modified). ³ Avicenna, The Metaphysics of The Healing IV.2.20–3, 137–9, tr. M. Marmura. (Henceforth as The Healing; Metaphysics, followed by book, chapter, paragraph, and page number(s).) Both Bertolacci and McGinnis hence err when they suggest the argument for nature’s existence comes at The Healing; Metaphysics IX.5 or IX.2 respectively. For the former, see The Reception of Aristotle’s Metaphysics in Avicenna’s Kitāb al-Shifā :͗ A Milestone of Western Metaphysical Thought (Leiden: Brill, 2006), 289; for the latter, see his translation of The Healing; Physics, 39 footnote 2. The same argument, with slight differences in nuance and much more condensed, appears in other Avicennian works as well. See for instance Kitāb al-Najāt; Ilāhiyyāt III.1.13, 526–9, as well as Uyūn ͑ al-ḥikma III.3, 49–50.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

3

source or principle of a body’s activities and rests. In employing this strategy, Avicenna is running the following two claims together: 1. Bodies have causal powers, and 2. Bodies have natures The reason for this is that, on the Avicennian view, nature just is a sort of power; or, equivalently, a causal power just is nature when considered in a certain respect. As he says at The Healing; Physics I.1 (emphasis mine): [T2] And natural things [ . . . ] are called natural in relation to the power (quwwa) which is called nature (ṭabī a).⁴ ͑

This identification between being a nature and being a power is made even more explicit in ͑Uyūn al-ḥikma where, immediately after having offered a condensed version of the argument for causal powers we consider below, the shaykh writes: [T3] And because every body is characterized, as we said, by a place, a quality, and the remaining [categories] – in sum, by motion and rest – these, therefore, belongs to them on account of a power (quwwa) that is a principle of the movement to those states. And this is [the meaning of] the term ‘nature’ (al-ṭabī a).⁵ ͑

It is clear from T2–3, then, that, according to the shaykh, the issue about bodies having causal powers and the issue about them having natures amount to the same issue considered in different ways. Hence, to establish the one is to establish the other. Having said all that, let me now, before going through the argument at The Healing; Metaphysics IV.2 in detail, first contextualize it by saying something about its philosophical and historical background.

3. Historical-Philosophical Background There are two background motivations for the Avicennian argument, each of which is related to the two main traditions he was influenced by

⁴ The Healing; Physics I.1.1, 3 (tr. McGinnis, modified). ⁵ Uyūn ͑ al-ḥikma III.3, 49.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

4

 

and directly engaged with in his context. I say these two traditions are the “motivations” for the argument in the sense that it is meant to address them specifically. First, there is the Aristotelian philosophical tradition which Avicenna inherited; and second, there is the kalām theological tradition, which he was in one way or another influenced by.⁶ Let me take up each in turn. With respect to the Aristotelian tradition, the Avicennian argument fills an important gap left by Aristotle—namely, a justification of the existence of nature (ṭabī a)͑ as a causal power which serves as a principle of the science of physics.⁷ For Aristotle taught that nature’s existence is something in no need of argument because it was obvious and “known through itself.” At Physics II.1, he states: [T4] That nature exists, it would ridiculous to try to prove; for it is obvious that there are many things of this kind, and to prove what is obvious by what is not is the mark of a man who is unable to distinguish what is self-evident from what is not.⁸

If we then take T4 in light of Aristotle’s definition of nature earlier at Physics II.1, i.e., that Nature df = “a principle or cause of motion and rest in that to which it belongs primarily, in virtue of itself and not accidentally”⁹ ⁶ For the influence of these two traditions on Avicenna’s metaphysics more generally, see Robert Wisnovksy, Avicenna’s Metaphysics in Context (London: Duckworth, 2003). ⁷ The same refusal to establish the existence of nature seems to be true of the Aristotelian tradition in general, as far as I can tell. For example, two of the most famous commentators on Aristotle in late antiquity, Philoponus (d. circa 570) and Simplicius (d. circa 560), in their respective commentaries on Physics II.2, both endorse Aristotle’s claim that nature’s existence is self-evident. For the former, see Philoponus: On Aristotle Physics 2, tr. A. R Lacey (London: Duckworth, 1997), 19–21; for the latter, see Simplicius: On Aristotle Physics 2, tr. B. Fleet (London: Duckworth, 1997), 24–6. Avicenna for his part was certainly familiar with Philoponus’ commentary, but it is less clear that he knew of Simplicius’, because it is not clear whether that work was translated into Arabic. For a general discussion of Avicenna’s critical engagement with the Aristotelian tradition in terms of resolving major points of conflict that he saw within it, see Dimitri Gutas, Avicenna and the Aristotelian Tradition (Leiden: Brill, 2014), ch. 6; for specific cases of such critical engagement, see Andreas Lammer, The Elements of Avicenna’s Physics: Greek Sources and Arabic Innovations (Berlin: De Gruyter, 2018). ⁸ Aristotle, Physics II.1.193a4–7, tr. R. P. Hardie and R. K. Gaye, in The Complete Works of Aristotle, ed. J. Barnes (Princeton, NJ: Princeton University Press, 1984). ⁹ See Physics II.1.192b21–3. I deliberately leave Aristotle’s definition ambiguous between an active sense and passive sense of being a principle of motion, to reflect the apparent difference, noted by both the ancients and moderns, between the account of nature he offers at Physics II.1 and Physics VIII.4. For two recent studies of this, see Richard Sorabji, Matter, Space, and Motion: Theories in Antiquity and Their Sequel (Ithaca, NY: Cornell University Press, 1988),

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

5

we realize that, in the passage at T4, when it is said that nature’s existence is evident and known through itself, Aristotle must mean that the claim bodies have an internal principle of their motions and rests, in a primary way and not accidentally

is a self-evident claim. And hence, according to him, the existence of “nature” in the technical sense above is not in need of argument. Avicenna knew the passage at T4 and found it perplexing if taken in just this way. In explicitly addressing it, he even first charitably distinguishes a sense in which it could turn out true: [T5] What is puzzling is the statement that “the one inquiring into establishing [nature’s existence] ought to be ridiculed.” I suppose what is meant by that is the one who inquires into establishing [its existence], while engaged in the science of physics, must be ridiculed – since he wants to demonstrate, in the discipline [of physics] itself, its principles.¹⁰

The claim in T5 accords well with general Peripatetic scientific methodology, which states, roughly speaking, that a given science, as a systematic body of knowledge, has (i) a subject matter, i.e., what the science is about; (ii) principles, i.e., the propositions that are used to construct demonstrative arguments in the science; and (iii) conclusions (maṭlūbāt), which are the theses the demonstrations prove in that science. In any given science, the existence of its subject matter and its principles are taken for granted.¹¹ With regards to (ii), the principles of a science, the shaykh more fully expresses the point in his work on the philosophy of science: [T6] ‘Principle of demonstration’ is said in two ways. So [in one way] ‘principle of demonstration’ is said with regard to science absolutely and [in another way] ‘principle of demonstration’ is said with regard to a certain science. The principle of demonstration with regard to science absolutely is a premise that has no middle [term] at all, i.e., it is not ch. 12, 219–26, and Aristotle, Physics: Book VIII, tr. D. Graham (Oxford: Clarendon Press, 2007), 74–89. ¹⁰ The Healing; Physics I.5.4, 40 (tr. J. McGinnis, modified). ¹¹ The Healing; Metaphysics I.1.8, 2–3, where Avicenna briefly alludes to this point.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

6

  such that the proof of the relation of its predicate and its subject – be it an affirmation or a negation – depends on a middle term, so that another premise would be prior to it and before it. The principle of demonstration with regard to a certain science may have a middle term in itself, yet it is laid down in that science as a posit and, in that stage (martabatihi) in that science, does not have a middle [term]. Rather, either its middle is in a science before it or on a par with it or its middle is in that [same] science after that stage [ . . . ].¹²

The passage above makes at least three claims of interest for what follows. First, a principle of demonstration may be a principle for science as such; second, it may be a principle for some science and not another; and third, if the latter, then this type of principle may be demonstrated in another science. If we then suppose p is some proposition that one employs in the constructions of demonstrative arguments in a given science, we can say that the truth of p itself is either not demonstrable at all by a proposition more known than it—in which case p would not be demonstrable by something more evident than itself in any science; or, the truth of p is demonstrable, but not by anything within the science that employs it as a principle. The first type of principle we can call the absolutely self-evident and the second the relatively self-evident. The latter type is not demonstrable within the science that employs it; rather, the science simply assumes its truth, i.e., as evident within it. But, it must be established in another science. In light of the above, when Aristotle says at T4 that nature’s existence is “evident” and “known through itself,” on the basis of what is said at T6 the shaykh understands him to be saying either that nature’s existence is evident, but relative to the science of physics or natural philosophy, or he intends that it is unqualifiedly evident. If the former, there is no disagreement; for, as we have seen in T5, the claim that (i)

Nature (as an internal principle of motion) exists

is a premise the truth of which physics assumes when it employs it in proving claims about what it studies, i.e., mobile being.¹³ But if so, then

¹² Avicenna, The Healing; Demonstration I.12, 110. ¹³ And again, Aristotle himself suggests as much when at Physics VIII.3.253b5–6 he states (emphasis mine) “it is a hypothesis that nature is a principle of motion.”

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

7

(i) needs to be justified in another, higher science. If the latter—i.e., if Aristotle takes (i) to be evident absolutely speaking, something in no need of justification in any science—then Avicenna thinks he is mistaken: [T7] If [the former] or some other related interpretation is not what is meant, but rather what is meant is that this power’s existence is selfevident (bayyin bi-nafsihi), then it is not something that I am willing to listen to and endorse. How could it be, when we frequently find ourselves forced to undertake a great deal of work to prove that every [body] undergoing motion has a mover? And so how can he be ridiculed who sees a motion and seeks an argument proving that it has a mover – let alone one who grants [that there is] a mover but makes it external?¹⁴

That is, if (i) were self-evident unqualifiedly, then it would also be equally self-evident, e.g., that (ii)

Every body in motion has a mover.

Proposition (ii) would not be in need of argument just as (i) is not in need of one on this interpretation. But the consequent is clearly false; for it is not obvious that every body in motion has a mover. Perhaps some bodies undergo motion in a brute-fact sort of way, i.e., simply on their own in a manner such that the motion isn’t analyzable in terms of some division within the body. Or, even granting (ii), it is not obvious that that mover is something intrinsic to the body, which nature is supposed to be on the Aristotelian account. That is, perhaps the mover of every body undergoing some sort of motion is an entity extrinsic to them.¹⁵ And Aristotle himself recognized the non-evidentness of the claim that everything in motion has a mover, insofar as he explicitly attempts to justify it in places like Physics VIII.4.

¹⁴ The Healing; Physics I.5.4, 40 (tr. J. McGinnis, modified). ¹⁵ As we will see in a moment, in the Islamic context the view that the motions of bodies are due to an agent external to them is precisely the option the Ash arite ͑ theologians, as well as independent thinkers like Abū Bakr al-Rāzī, took. So insofar as Avicenna knew the thought of both, it is not unlikely that a view like theirs may have played some part in his disagreement with the Peripatetics over the self-evidentness of the existence of nature. For the Ash arite ͑ kalām tradition, see footnote 22. For Rāzī, see Maqāla fī mā ba d͑ al-ṭabī a,͑ 116–17, 120. For a study of these passages in Rāzī, see Peter Adamson, “Against Nature: Two Critics of Naturalism in the Islamic World” (forthcoming).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

8

 

In sum, that bodies undergo motion is uncontroversial—it is a datum of experience that we must reason from. But that they have a nature in the Aristotelian sense of an internal principle of that motion is disputable, and hence that claim is one we must reason to. This is the import of the shaykh’s claim at the end of T7 that in many cases justificatory work must be done to show that bodies do in fact have something in them which is the source of their motions. So the basic point is that either nature’s existence is evident in the sense delineated at T5–6, with the consequence that Aristotle owes us, according to an agreed upon scientific methodology, an argument for nature’s existence. Or, it is not evident in that sense but rather evident in a non-relative way, which Avicenna thinks is false for the reasons mentioned above. Now one might think that Aristotle does indeed offer, or can be plausibly interpreted as offering, in at least two places, some sort of justification for (i), and from that infer that he must have then taken the claim about its existence to be self-evident but not in the absolute way. The first place one might appeal to is the Physics itself, specifically I.2, where Aristotle is responding to the Parmenidean threat to the very possibility of studying the natural world (or doing physics).¹⁶ The Parmenidean there claims that (iii)

What exists is one and unchangeable

and Aristotle criticizes (iii) in at least two ways. In the first, he notes that (iii) would entail that there are no principles (presumably of the objects that physics studies, i.e., natural bodies); and the reason for this is because “a principle must be the principle of some thing or things,”¹⁷ which implies multiplicity. But, the Parmenidean thesis (iii) forbids multiplicity, and hence principles too will be done away with. So insofar as “nature,” per Physics II.1, is for Aristotle a principle of sorts, Physics I.2 can then be seen as a defense in a way of nature’s existence qua principle. The second way Aristotle criticizes (iii) at Physics I.2 is that he argues it would equally do away with the existence of motion as such, which Aristotle rightly thinks is an undeniable feature of at least some bodies.¹⁸ Now insofar as nature, for Aristotle, is a principle of motion, he can be understood as saying that a denial of motion would entail a denial of ¹⁶ See Physics I.2.185a1–186a1–3.

¹⁷ Ibid., 185a2–4.

¹⁸ Ibid., 185a12–15.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

9

nature as a principle (of motion). And to that extent, he can be seen again at Physics I.2 as offering some kind of defense of (i), i.e., the claim that nature exists. The second place one might appeal to as evidence for the view that Aristotle does offer, or can be interpreted as offering, a justification of nature’s existence is Metaphysics IX.3. There, Aristotle is responding to the Megarian thesis that (where F is some activity) (iv) Nothing has the capacity to F before it actually Fs. That is, there are no inactive powers in things.¹⁹ Now to the extent that nature for Aristotle is a causal power of sorts, and a power is a principle, to that extent, one might reason, his defense of the existence of powers at IX.3 can be viewed as a defense, again, of the existence of nature qua principle. For nature, according to him, is the source of a thing’s activities or motions, and so, one might think, is the capacity or power for those activities or motions. However, even granting the above interpretations of Physics I.2 and Metaphysics IX.3 as legitimate, I think they fail to show that, when Aristotle claimed at T4 that nature’s existence is “known through itself” or self-evident, he meant that its self-evidentness is relative in the sense specified above. And this for the simple reason that what Aristotle offers in both contexts is a dialectical justification, against the Parmenidean and Megarian theses respectively, of (i). This is why his arguments at Physics I.2 and Metaphysics IX.3 all have the form of a reductio ad absurdum: if principles or powers don’t exist, then various absurd consequences follow. And the method of reductio style argumentation is usually applied when defending propositions one takes to be self-evident or “known through themselves” in an absolute way—the reason for this being that self-evident first principles do not themselves have further, better known principles by which they can be demonstrated.²⁰

¹⁹ For recent studies of Metaphysics IX.3 that see it as having to do with the existence of powers, see Aristotle, Metaphysics Book Θ, tr. S. Makin (Oxford: Clarendon Press: 2006), 60–8; Jonathan Beere, Doing and Being: An Interpretation of Aristotle’s Metaphysics Theta (Oxford: Oxford University Press, 2009), ch. 3; and Rebekah Johnson, “The Existence of Powers,” Apeiron 41 (2008), 171–90. ²⁰ Avicenna himself employs the dialectical reductio ad absurdum method of argumentation at The Healing; Metaphysics I.6–7 when defending the first proper principles, in the domain of assent (taṣdīq), of the discipline of metaphysics. For discussion of these chapters, see Daniel De

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

10   The shaykh, by contrast, is explicit about (i)’s qualified or relative self-evident character. At Physics I.5, he writes: [T8] The truth, however, is that the claim that “nature exists” is a principle of the science of physics; it is not on the natural philosopher to address anyone who denies it. Establishing [its existence] belongs only to the metaphysician, whereas it belongs to the natural philosopher to study its quiddity.²¹

The above, then, is more or less the general Aristotelian background to which, in my view, the Avicennian argument is intended, partly, as a response. But as I said earlier, I think the argument is also meant to address Avicenna’s theological contemporaries, and so I now want to turn to characterizing the relevant kalām background. In the Islamic tradition, there are basically two competing models of the natural world: atomism and hylomorphism. Generally speaking, thinkers who usually held the former were theologians (mutakallimūn), belonging either to the Ash arī ͑ or Mu tazilī ͑ schools, while those who held the latter were philosophers (falāsifa/ḥukamā )͗ usually belonging to the Avicennian camp. As far as the ontology of the former is concerned, the most basic division for them is between beings that are eternal (qadīm) and those that come to be or occur in time (muḥdath). To the former category belong only God and His attributes; to the latter, everything other than God (mā siwā Allāh), which is the world as a whole (al- ālam) ͑ understood as, to use Sabra’s phrase, “all that occurs” (kullu mā yuḥdath).²² That which occurs refers to a spatio-temporally conditioned event (ḥadith) which, as such, is either something that occupies space (mutaḥayyiz) or something that inheres in something that occupies space. The former is called “substance” (jawhar), and comes in two types: either as simple (basīt), which is a fundamental—i.e., an indivisible, Haan “Where Does Avicenna Demonstrate The Existence Of God?,” Arabic Sciences and Philosophy 26 (2016), 109–21. ²¹ The Healing; Physics I.5.4, 40 (tr. J. McGinnis, modified). For the historical and philosophical background to Avicenna’s account of the quiddity or essence of nature at The Healing; Physics I.5, see Andreas Lammer, “Defining Nature: From Aristotle to Philoponus to Avicenna,” in A. Alwishah and J. Hayes (eds.), Aristotle and the Arabic Tradition (Cambridge: Cambridge University Press, 2015), 121–42. ²² A. I. Sabra, “The Simple Ontology of Kalām Atomism: An Outline,” Early Science and Medicine 14 (2009), 68–78.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

11

particle or atom; or as composite (murakkab), which is an arrangement of two or more simple substances and as such is called “body” (jism). That which exists in what occupies space is called “accident” ( araḍ ͑ ), understood as a phenomenal property. According to the mutakallimūn, there are four types of accident which necessarily characterize every substance, whether simple or composite, called “modes of being” or “happenings” (al-akwān, sing. kawn). These are: (1) motion, (2) rest, (3) separation, and (4) contact/aggregation.²³ Motivated by scriptural concerns, the Ash arite ͑ thinkers combined this atomic model of the world with an occasionalist model of God’s causal relation to that world—a doctrine that was properly founded by the eponymous Abū’l-Ḥasan al-Ash arī ͑ (d. 936).²⁴ According to Ash arī ͑ and his followers, God is the true and exclusive causal agent of events—a doctrine which was endorsed in direct conscious opposition to a view of the world on which there are natures that are sources of effects. Ibn Fūrak (d. 1015), a later member of the school, reports the following: [T9] [Ash arī] ͑ denied the doctrine of natural disposition and nature (al-ṭab ͑wa’l-ṭabī a), ͑ and said that all events are actions of God, by His choice, will, governance, and determination – without any of these necessitating another event, or a nature that engenders it – but, rather, all are His invention by His choice and in the manner He chooses and knows.²⁵

²³ Ibid., 77–8. ²⁴ Abū l-Ḥasan al-Ash arī, ͑ Al-Ibānah, 9. For an important later Ash arite ͑ theologian, see Abū Bakr al-Bāqillānī, Al-Inṣāf, 31. For a general history of occasionalism in the Islamic world, see Ulrich Rudolph, “Occasionalism,” in S. Schmidtke (ed.), The Oxford Handbook of Islamic Theology (Oxford: Oxford University Press, 2016), 347–63. Rudolph argues that, strictly speaking, early Mu tazilite ͑ theologians e.g., Abū l-Hudhayl (d. 842) and his successors, whom Ash ͑arī critically engaged with, did not endorse full-fledged occasionalism. Rather, they only laid the foundations for it, which Ash arī ͑ later built on to finally articulate an occasionalist doctrine proper. And in fact Asha arī ͑ himself reports that, in general, the Mu ͑tazila, with the exception of Jubbā ͗ī (d. circa 915), believed in causal relations involving “natures.” See his Maqālāt al-islamiyyīn wa-l-Ikhtilāf al-muṣallīn, 314 and 412. See also Ibn Fūrak, Mujarrad maqālāt al-Asharī, ͑ 275–6. Whatever the case may be, to the extent that some of these early Mu ͑tazilī theologians’ denial of powers was local, i.e., confined to non-human entities, to that extent they also serve as a target for the Avicennian argument we shall consider below. ²⁵ See A. I. Sabra, “Kalām Atomism as an Alternative Philosophy to Hellenizing Falsafa,” in J. E. Montgomery (ed.), Arabic Theology, Arabic Philosophy. From the Many to the One: Essays in Celebration of Richard M. Frank (Leuven: Peeters, 2006), 244 (tr. modified). See also ibid., 245–6. For Ibn Fūrak, see Mujarrad 76.8–11, as well as 271, where it is reported that Ash arī ͑ in a work called al- Idrāk aimed to show that fire does not cause burning in cotton upon contact but rather God does so (i.e., occasionalism).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

12   The key consequence of this combination of atomism and occasionalism for our purposes here is that all bodies were deemed as being devoid of any sort of nature understood in the technical Aristo-Avicennian sense, i.e., as a causal power for various kinds of activity. Let me now offer a concise characterization of Ash arite ͑ occasionalism by drawing on a later member of that school, Abū Ḥāmid al-Ghazālī (d. 1111).²⁶ My focus will be not so much on the reasons why the Ash arites ͑ endorsed occasionalism, but rather on its central thesis. I’ll give a general description of this thesis, with a view to setting up a target for the Avicennian argument we will consider next.²⁷ At the end of the chapter on God’s attribute of power (qudra) in his Al-Iqtiṣād fī’l- I tiqād, ͑ Ghazālī faithfully reproduces the above Ash arite ͑ view of God’s causal relation to the world when he writes: [T10] [A]ll events (ḥawādith), their substances and accidents, and those that occur in the animate or inanimate beings – come into existence through the power of God (Glorious and Exalted is He), who is their exclusive originator. It is not true that some created things come into existence through other created things, but all of them come into existence through divine power.²⁸

Now there are different versions of occasionalism and its proponents differ over details of the doctrine. But, its common core is the thesis that God is the only efficient cause of beings in the world, such that the claim:

²⁶ I choose Ghazālī deliberately, for reasons having to do with the famous (in modern scholarship) 17th discussion of his Tahāfut al-falāsifa. For some studies, see Michael Marmura, “Al-Ghazâlî’s Second Causal Theory in the 17th Discussion of His Tahâfut,” in P. Morewedge (ed.), Islamic Philosophy and Mysticism (Delmar, NY: Caravan Books, 1981), 85–112; Richard M. Frank, Creation and the Cosmic System: Al-Ghazâlî & Avicenna (Heidelberg: Carl Winter, 1992), and Al-Ghazali and the Ash arite ͑ School (Durham, NC: Duke University Press, 1994); Michael Marmura, “Ghazâlian Causes and Intermediaries,” Journal of the American Oriental Society 115 (1995), 89–100; Stephen Riker, “Al-Ghazali on Necessary Causality in ‘The Incoherence of the Philosophers,’ ” Monist 79 (1996), 315–24; Blake D. Dutton, “Al-Ghazâlî on Possibility and the Critique of Causality,” Medieval Philosophy and Theology 10 (2001), 23–46; E. O. Moad, “Al-Ghazali’s Occasionalism and the Natures of Creatures,” International Journal for Philosophy of Religion 58 (2005), 95–101; Frank Griffel, Al-Ghazâlî’s Philosophical Theology (Oxford: Oxford University Press, 2009). See also footnote 63. ²⁷ For arguments justifying occasionalism, see for instance Al-Ash arī, ͑ Al-Luma , ͑ in R. J. McCarthy (ed.), The Theology of al-Asharī ͑ (Beirut: Catholic Press, 1953), 33–44; Abū Bakr al-Bāqillānī, Kitāb al-tamhīd, 286–7; Al-Inṣāf, 41, 44–5; and Al-Juwaynī, Kitāb al-irshād, 215–25. ²⁸ Al-Ghazālī, Moderation in Belief II.1, 103, tr. A. M. Yaqub, modified. (Henceforth as Al-Iqtiṣād, followed by treatise, part, and page number(s). All translations from this work are Yaqub’s unless otherwise noted.)

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

13

For any event that comes to be at t, God is the “exclusive originator” of that event at t turns out true. We can capture this occasionalist thesis in the claim: (OC) Only God possess efficacious (ta ͗thīr) causal power where A causal power, F, is “efficacious” just in case F makes a real causal contribution to some effect, e. Let us take the standard example: the occurrence of burning upon fire touching a piece of cotton. When this happens, the common sense view—that it is the fire that acts on the cotton, causing it to burn—is false according to proponents of OC. Rather, in their view, what happens, metaphysically speaking, is that God directly produces or originates, on the “occasion” of fire’s coming into contact or conjunction with the cotton, the effect of being burnt in the cotton.²⁹ The fire has no efficacy at all. But, this description of the scenario should not mislead one into thinking that on this account God merely produces an effect (i.e., burning of cotton) at some moment t when two otherwise independent things (the fire and the cotton) come into contact at t. The occasionalists state something much stronger, namely, that God directly creates the atoms arranged “fire-wise” as contacting—i.e., with the accident of “contact”—and directly produces the atoms arranged cotton-wise as being in contact (with the fire at t) and with the accident of being burnt, i.e., the accident of separation, at t. That is, God creates, at t, the whole event fire’s burning—cotton’s being burnt. On this latter formulation, God is the direct and only efficient cause of everything about the fire, cotton, and the burning at t—“all ḥawādith, their substances and accidents”—and not just the effect of burning. All power is located in God—or more precisely, God’s will. The above is the gist of the Ash arite ͑ atomist-occasionalist model of the natural world. The Avicennian view, on the other hand, states that bodies are composites of a material and formal principle. It challenges the occasionalist tenet of the Ash arite ͑ system insofar as the hylomorphic ²⁹ Al-Ghazālī, Incoherence of the Philosophers VII.2, 166–7, tr. M. Marmura. (Henceforth as Tahāfut, followed by chapter, paragraph, and page number(s).)

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

14   principles of a body endow that body with a set of natural capacities or powers that explain its characteristic behavior. As such, hylomorphism entails that bodies are equipped with capacities that, contra the proponents of occasionalism, make genuine causal contributions to the world. Importantly, however, Avicenna does not try to reject a claim like OC by appealing to hylomorphic considerations; for of course doing so would amount to appealing to premises which the Ash arites, ͑ from the start, simply would not grant. Were the shaykh to proceed in that way, he would be refuting them from the presuppositions of his own system, and so either beg the question against his opponents or simply talk past them. Instead, the argument the shaykh presents at The Healing; Metaphysics IV.2 is supposed to appeal only to premises which the opponents already concede. I now clarify what these premises are. Principally, they are two in number. The first is empirical in nature. Generally stated, it says something like: (EP) Events in the world display regularity. A particular example of which would then be, say, that Burning regularly proceeds from fire when it touches cotton. According to the shaykh, EP as instanced in the fire–cotton case is a premise that falls under those he terms “the experientials” (al-majarrabāt). These constitute one class of proposition the assent to which is immediate, i.e., does not involve any intermediary.³⁰ The second premise is a version of the principle of sufficient reason (PSR). The PSR, as regards existence and non-existence, states that For anything that exists (or does not exist), there is a sufficient reason/ cause ( illa) ͑ why it exists (or does not exist). The shaykh endorses the PSR in this form.³¹ It is not clear, however, that his theological opponents do. But whatever the case may be, what is

³⁰ For the account of this type of immediate premise, see Kitāb al-najāt; manṭiq II.43, 113–14. ³¹ For Avicenna’s commitment to the PSR understood generally, see The Healing; Metaphysics I.6.1, 4–6. For an analysis of these passages which sees them as employing it, see Kara Richardson “Avicenna and the Principle of Sufficient Reason,” Review of Metaphysics 67 (2014), 743–68.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

15

important in the context at hand—i.e., as far as the argument for nature’s existence is concerned—is that he employs a particular version of the PSR that his theological opponents are in fact committed to. This particular version of the PSR is known, in the parlance of the postAvicennian tradition, as “the rule of determination or preponderance” (qā ida ͑ al-tarjīḥ). This rule, which we can call QT, can be defined as follows, where x is some entity (dhāt) and F some attribute (ṣifa): QT: for any x that is contingently F (or not-F, for any not-F), there is a sufficient reason/cause ( illa) ͑ why x is F (or not-F). Or, more simply put, there is some explanatory factor that determines—in the sense of necessitates—why something contingent is the case or not the case.³² On the Avicennian view, QT, as a specific version of the PSR, would be a proposition that falls under those called “the primaries” (al-awwaliyyāt). As with the experientials, the assent to the primaries is immediate.³³ With this Aristotelian and kalām background in place, I now turn to the argument in Healing; Metaphysics IV.2 for the existence of nature as a power intrinsic to bodies.

4. The Avicennian Argument The shaykh’s argument, as I said, presupposes EP and QT—two selfevident propositions that he and his opponents share. The syllogistic order (naẓm al-qiyās) that the argument as a whole takes is a compound repetitive disjunctive conditional (sharṭiyya munfaṣila istithnā ͗ī murakkaba), a highly abridged version of which can be stated like this: If any effect, e, proceeds from a body neither accidentally or by force, then e proceeds from it either (1) because of its being a body, or (2) ³² Why do I think the PSR qua QT is implicit in the Avicennian argument for nature’s existence? Two reasons: first, the various disjuncts of that argument (as well as the reasoning offered for or against them, as we will see below) make no sense unless something like QT is presupposed. Second, because for Avicenna the PSR (as is evident from The Healing; Metaphysics I.6) is a first principle, in the domain of assent, of the science of metaphysics; as such, it makes perfect sense that he appeals to a particular version of it in the construction of demonstrations in that science. ³³ For an account of this type of immediate premise, see Kitāb al-najāt; manṭiq II.50, 121–3.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

16   because of some feature F inhering in it, or (3) because of some entity E really distinct from it, which can be either (3.1) another body, or (3.2) something incorporeal. But not (1) and not (3.1–2). Therefore, (2). This F is what is called a power. Therefore, powers exist.

Let us now turn to the detailed version of the argument. The shaykh begins with a statement of the conclusion to be established (maṭlūb): [T11] Every body from which an act proceeds (ṣadara) neither accidentally (bi’l-araḍ) nor by compulsion (bi’l-qasr) from another body acts through some power (quwwa) in it.³⁴

The conclusion is formulated in non-causal language (“proceeds from” as opposed to “produced or caused by”) in order to avoid any begging of the question against the opponent. This is meant as a precise demarcation of the conclusion to be proved, namely, that for any body from which some activity proceeds in (i) a non-accidental and (ii) a non-forced way, the act proceeds from the body in virtue of a power in it, i.e., a power for the act.³⁵ Conditions (i)–(ii) are important for two reasons, one

³⁴ The Healing; Metaphysics IV.2.20, 137. (All translations from this work from here on are my own.) ³⁵ An accidental act or event, on Avicenna’s view, is the opposite of the essential act or event. As such an opposite, the accidental is said in many ways. At The Healing; Physics I.13.2, 74, he explains the difference like this: The essential efficient cause is like [ . . . ] the fire when it heats; it is a principle of that very act itself and is taken insofar as it is its principle. The accidental efficient cause is anything apart from that [ . . . ]. Avicenna then goes on to specify the latter notion. In one sense, the accidental is what happens by chance or coincidentally, where this is opposed what occurs either always or for the most part, i.e., regularly (The Healing; Physics I.13.10, 87). In another sense, an act or event is bi’l-araḍ ͑ (accidental) just in case it proceeds from something in virtue of something else—an example is of a man at rest on a boat which is in motion (The Healing; Physics I.5.6, 42). Though at rest, he is also moving insofar as the boat he is on is moving. But he is not in motion in the way the man on the street walking is in motion. There is clearly a difference between the two, even though motion “proceeds” (ṣadara) from both just in the sense that both are moving. In the former case, the act of moving proceeds from the man because of something else, i.e., the boat, and so his being in motion is said to be in an accidental way. (For a detailed discussion of acts or motions bi’l- araḍ ͑ , see The Healing; Physics IV.13.) An act that proceeds from a body by force (bi’l-qasr) is also accidental in a way. But this case differs from the previous two cases in that the body in question is truly undergoing the act, but it proceeds from the body because of something external to it. The example given is of a stone that is being dragged along the ground (The Healing; Physics IV.14.1); the stone itself really is moving, unlike the man at rest on the

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

17

related to the kalām background, the other to the Aristotelian. As regards the former, conditions (i)–(ii) limit the target of Avicenna’s argument to cases of regularity, where regularity is opposed to what happens randomly or by chance. For, on Avicenna’s view, that which happens randomly, by chance, or coincidentally is a subset of what happens by accident, and hence to rule out the latter kind of act/event (i.e., the accidental) entails ruling out the former type (i.e., the random/chance/ coincidental event).³⁶ So events or acts that proceed from a body in the non-accidental way that is specifically of relevance here are those that are said to occur in a regular and constant way. It is more evident that this is the crucial feature of the act or event in the abridged Uyūn ͑ alḥikma version of the argument, which begins by precisely noting such regularity: [T12] Every body from which, in what is habitually sensed (fī’l- āda ͑ al-maḥsūsa), an act always proceeds – either that act proceeds from it due to its corporeality, or due to a power (qūwwa) in it, or because of an external cause (bi-sabab khārij).³⁷

The expression fī’l- āda ͑ al-maḥsūsa—“in what is habitually sensed” or, even more broadly, “in sense experience”—is supposed to designate what happens in a consistent or predictable way. In fact, the Ash arite ͑ theologians themselves explicitly use it to refer to the “habitual course of things,” which they chalk up to God’s constantly creating them in that way (more on this below). So insofar as that expression describes the character of the events that occur in the world, namely, that they happen in a regular way, the claim at T11 then basically is: An act which proceeds regularly, i.e., non-accidentally, from a body (or an event which regularly happens), proceeds due to a power in it (or, in the event case, due to powers in the bodies involved in the event).

ship, but it moves only because something else is forcing that act or motion on it, and not of itself. The upshot of all this is that, on the shaykh’s view, any act that is not accidental—in the sense in which accidental is opposed to the essential at The Healing; Physics I.13.2, 74—will proceed from a body in virtue of a power in that body. ³⁶ For Avicenna’s own extensive, positive account of chance or the coincidental, see The Healing; Physics I.13. ³⁷ Uyūn ͑ al-ḥikma III.3, 49.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

18   The regularity condition is important for the argument in that it is a datum the Ash arite ͑ theologians concede and for which they too attempt to offer an explanation. The second reason, related to the Aristotelian background, is that conditions (i)–(ii) allow Avicenna to say that, although bodies have powers, such powers are not attributable to them in an indiscriminate way; that is, it is not the case that a body possesses a power for whatever act happens to proceed from it. Rather, only if an act is one that is essential and not forced is there a power in the body for the act. So, if the argument turns out sound, (i)–(ii) will be criteria for determining whether a body has a genuine causal capacity for an act we observe to proceed from it. The shaykh continues: [T13] In the case of that [act] which is by volition (irāda) and choice (ikhtiyār), it is obvious (ẓāhir).³⁸

The Ashā ira ͑ maintain that there is, at least in the case of human beings, a difference—one that is known of necessity (ḍarūrī)—between a voluntary act like raising one’s hand and an involuntary one like a tremor. The difference is that the former, unlike the latter, involves a power or capacity (istiṭā a)͑ for the act in the human being.³⁹ As T13 indicates, this is a claim Avicenna agrees with. In fact, as we will see below, he would be willing to go as far as to agree, for the sake of argument, with their further claim that the capacity for the act is a generated one (hādith), i.e., one that God creates in the human being and does so at the same moment as creating the act of lifting his arm. For granting all of this so far is consistent with anti-occasionalism. What Avicenna will deny is their insistence that the created capacity in the creature for the act is not in any sense causally responsible for the act, i.e., that it actually does not produce the act in any way.⁴⁰ The shaykh can certainly reject this last claim on independent grounds but, as far as his present argument is concerned, we will see that a consequence of it will be that, in the case of creatures that have volition and choice, even if God endows them, as the Ashā ira ͑ say, with a capacity to act but brings about the act Himself, still the capacity He endows them with must in some sense ³⁸ The Healing; Metaphysics IV.2.20, 137. ⁴⁰ Cf., ibid., II.1, 94–5.

³⁹ Cf., al-Iqtiṣād II.1, 92–3.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

19

causally contribute to the act. And in the case of those creatures which lack volition and choice, Avicenna will argue that even if an external entity is somehow involved in the production of their acts, the explanation for this too must ultimately be grounded in factors (= powers) internal to them. The argument focuses exclusively on this latter, i.e., non-volitional, category, since it is the non-evident case. How then does the shaykh establish that the acts of non-volitional bodies must proceed from them in virtue of powers in them? He begins with an initial disjunction: [T14] As for [acts] which are not by volition and choice, the act proceeds either [1] from [the body’s] essence (dhātihi) or [2] from a distinct thing that is corporeal or [3] from a distinct thing that is incorporeal (ghayr jismānī).⁴¹

Here we have a first use of the QT principle in combination with the EP premise: if a given act proceeds regularly from some non-volitional body, there must be some reason or cause why. That reason is either because of some factor intrinsic to the body or because of one outside it. Let us assume that e is some effect (or type of effect), e.g., burning, that occurs in a regular manner on the occasion of body x (or bodies of type x), e.g., fire, conjoining or contacting some body y (or bodies of type y), e.g., cotton. Then, for any connection between x, y, and occurrence of e, e is either due to (1) x qua body, or (2) some material entity, c, different from x, or (3) some immaterial entity, G, different from x. Option (3) will be further distinguished, as we will see below at T21; but as stated, it does not yet amount to the occasionalist thesis of the Ashā ira, ͑ although it includes that thesis. For at this stage of the argument it is an open question whether or not that immaterial entity G is a divine agent, and whether or not it is a volitional agent.⁴² The line of argument that ensues will run as follows. With respect to option (1), the shaykh will argue that, on analysis, it turns out to be the ⁴¹ The Healing; Metaphysics IV.2.20, 137. ⁴² The occasion to deal with the Ash arite ͑ view proper will come only at T23 below.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

20   conclusion sought at T11. As for option (2), he will reject it because it contradicts an assumption laid down earlier. Finally, with respect to option (3), he will show that, however we construe the relation of that separate, incorporeal agent G to the body x from which, on this assumption, G produces the effect e, that agent G can produce e only in virtue of some property F in x which makes a causal contribution to e. So option (3), too, turns out to be the conclusion sought at T11. Let us now take up each disjunct at T14. With respect to (1), it returns to the desired conclusion. The shaykh argues as follows: [T15] If [the act] proceeds from [1] the [body’s] essence, and its essence shares corporeality with other bodies but differs from them in [being] that from which that act proceeds, then in its essence there is a factor (ma nā) ͑ additional to corporeality, which is a principle for the procession (ṣudūr) of this act from it – and this is what is called power (quwwa).⁴³

The thought here is: bodies have the fact that they are bodies in common but obviously differ in the effects that arise from, or issue out of, them. If then an effect e issues from some body merely because of its being a body, then e would equally arise or issue from all bodies; for they are all equivalent in that regard, i.e., qua bodies. But the effect clearly does not arise from every body equally, as sense experience bears out. For fire burns cotton in a regular way, and not gold; and heat arises out of fire in a regular way, and not out of snow. If so, that must be because the body in question has some attribute not identical to the attribute of merely being a body, and is peculiar to it in that it explains why it specifically, and not some other kind of body, produces the specific effect in question in a regular way. Whatever that attribute is, we call it a power. Therefore, bodies have powers; and therefore, powers exist. So, if option (1) is true, then nature understood as a power exists and occasionalism is refuted. The argument against option (2) runs as follows: [T16] If that [act proceeds] from [2] another body, then the act would [proceed] from that [initial] body by compulsion or accidentally.

⁴³ The Healing; Metaphysics, IV.2.20, 137–8.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

21

But it had been supposed [to proceed] neither by compulsion from another body nor accidentally.⁴⁴

According to this option, two different bodies are involved in the causal scenario, such that from one of them some act or effect proceeds in virtue of the other body—in the way, e.g., heat proceeds from a piece of metal because it had previously been heated by fire. The argument against option (2) is fairly clear: it contradicts the initial assumption that the act or effect proceeds from the first body non-accidentally and in a nonforced way. For if e proceeds from body x because of body c, then e would proceed from x either accidentally or by force—in just the same way that the man at rest on a boat moves because the boat is moving, or that the metal heats a body of water as a result of being made hot by fire. For “an act to be due to another” is just a part of what it means for it to be accidental and forced. The shaykh then takes up option (3) and sub-divides it: [T17] If [the act is] from [3] something separate (mufāriq) [i.e., from matter], then either this body’s being distinguished by this mediating [of the act] from that separate [entity] is [3.1] because of that by which it is a body, or [3.2] due to a power in it, or [3.3] due to a power in that separate [entity].⁴⁵

In what follows, the shaykh is going to argue that options (3.1) and (3.2) land us back to where we started; for both options entail the conclusion to be established. The thought in T17 involves another appeal to the QT principle; for if the external, non-material entity G produces e from body x in a regular way, there must be a reason for that. That reason either has to do with the body in question—that is, with the fact that the body is fire or cotton—or with the external immaterial entity G outside that body. On the one hand, if the former, then the reason that G produces the act exclusively from x is, again, either because of the mere fact that x is a body or because of some special property that x has—that is, it either has to do with the mere fact that fire or cotton are both bodies, or with the fact that they possess certain special, say, chemical power-properties F and F* (however

⁴⁴ Ibid., IV.2.20, 138.

⁴⁵ Ibid.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

22   F and F* are empirically specified). On the other hand, if the reason has to do with G itself, it must be grounded in some power-property possessed by G, which it brings to bear on body x and thereby produces effect e from it. With this last disjunct, i.e., (3.3), we do not yet actually have the occasionalist view, although (3.3) indistinctly contains it. (We will see the reason for this below, at T21.) How are options (3.1) and (3.2) handled? Against (3.1), which states that G regularly produces e from body x because of the fact that x is a body, the shaykh urges the following: [T18] If [3.1] it is because of that by which it is a body, then every body would share in [mediating the act]; but they do not share in [mediating the act].⁴⁶

What is meant by “body x mediating the act,” i.e., from G? I take it in the following way: the act’s always occurring by means of x (as opposed to some other body y). So, at (3.1), what is essentially being asked is: why does G regularly produce burning, say, from fire—where “burning being regularly produced from fire” just means “fire’s regularly mediating that act or effect (i.e., of burning)”—and not from some other body, e.g., snow? In other words, fire mediates the act or effect in that it serves as an intermediary between the external entity G and the effect or act produced by G in some other body, e.g., cotton. The move whereby (3.1) at T18 is eliminated is familiar from T15. If body x mediated the effect or act e that comes about through G because of the mere fact that x is a body, then, because every body other than x is equally a body, these other bodies would equally be suitable for mediating, and so should mediate, e from G. But, as a matter of fact, bodies other than x clearly do not; for we observe that e proceeds out of x in a regular way and not out of any other kind of body. But given the QT premise, that must be due to some causal factor, and that factor cannot be located, as (3.1) would have it, in the fact that x is a body. In that case, and this implicit in the reasoning at T18, it must be due to a factor in x ⁴⁶ Ibid.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

23

over and above its being a body. And insofar as this factor causally explains e’s occurrence, it is a power. Hence, powers exist in bodies. As for T17’s option (3.2), it too, like option (1) at T11, and option (3.1) at T17, turns out on analysis to be the desired conclusion (maṭlūb). The shaykh explains: [T19] If [3.2] it is due to a power in [the body], then this power is a principle for the procession of the act from it and, further, would be so even if [the act] emanated from and with the aid of the separate [entity] or due to its being the first principle of [the act].⁴⁷

That is to say, the view that x mediates the effect or act e from G in virtue of some power-property F in x entails that F in a real sense is a principle for the production of e. If fire burns cotton because it has property F, then that property actually contributes to cotton’s being burned. And insofar as F causally contributes to e’s occurrence, it would then in a real sense be, in Avicennian terms, a mu ͗aththir, i.e., a causally efficacious agent with respect to the effect e. That is why F is there in x in the first place; otherwise, F’s existence would be otiose, and we’d have to fall back on one of the earlier disjuncts in order to account for e’s regular occurrence from x. But as things stand, given option (3.2), F’s being causally efficacious with regards to e would hold true, the shaykh insists, even if G’s causal activity was a necessary condition for e’s occurrence. That is, he is claiming that x would still be a genuine cause of e, even if the separate entity G somehow aided it in producing e or, equivalently, even if that separate entity was the primary or first principle (and so where x would be a non-primary principle) of the effect that issues from x.⁴⁸ This last point is crucial—as we will see, it will be reiterated below at T34 when the final conclusion of the argument is drawn. In making that point, Avicenna is attempting to accommodate to the extent that he can—as I noted above—a key feature of the Ash arite ͑ account of God’s causal relation to events in the world. It all depends on what the effect in question is, as can be gleaned from the way natural efficient causes are divided up at The Physics I.10:

⁴⁷ Ibid. ⁴⁸ My thanks to an anonymous OSMP referee for forcing me to clarify this point.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

24   [T20] The principle of motion is either what prepares or what completes. What prepares is that which makes the matter suitable, like what moves semen during the preparatory states. What completes is that which gives the form. And it seems that the giver of the form by which the natural species subsist is outside of the natural order.⁴⁹

From T19 in combination with the passage above, we see that what the shaykh intends with the general claim: x mediates e from G in virtue of F is that F causes e’s occurrence insofar as F prepares for it. And the exact preparatory contribution of x’s F to e depends on whether e is some accidental modification of another thing (i.e., some type of motion) or it is a substantial form. If the former, F’s being a preparatory principle of e is just for it to cause e’s very occurrence; if the latter, F’s being a principle of e is for F to actually modify something else, y, such that G, the separate entity, produces e in y. To this extent, the shaykh is accommodating the Ash arite ͑ insistence on the involvement of a separate immaterial cause in the events of the world; for in certain events, i.e., substantial changes, the causal activity of an extrinsic and non-bodily cause, the Giver of forms, is required.⁵⁰ In both accidental and substantial changes, however, x’s power-property F makes a genuine causal contribution—qua specifying or preparing—to the occurrence of e even if, as the end of T19 states, in the latter case it does so “with the aid of [the separate entity].” For now, let us set aside this distinction between the cause that prepares and the one that completes and attend to the next step in the argument, which concerns T17’s disjunct (3.3). It is further divisible:

⁴⁹ The Healing; Physics I.10.3, 65 (tr. J. McGinnis, modified). For an account of Avicenna’s views on substantial generation, see Kara Richardson, “Avicenna and Aquinas on Form and Generation,” in D. Hasse and A. Bertolacci (eds.), The Arabic, Hebrew, and Latin Reception of Avicenna’s Metaphysics (Berlin: De Gruyter, 2011), 251–74; and Jon McGinnis, “On the Moment of Substantial Change: A Vexed Question in the History of Ideas,” in J. McGinnis (ed.), Interpreting Avicenna: Science and Philosophy in Medieval Islam, Proceedings of the Second Annual Symposium of the Avicenna Study Group (Leiden: Brill, 2004), 42–61. ⁵⁰ But note: the Giver of forms, according to Avicenna, is not a divine agent, though it is outside the natural order. This is another difference between his view and that of the Ash aris. ͑

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

25

[T21] If [the act] is [3.3] due to a power in that separate [entity], then either [3.3.1] that very power itself necessitates that [act] or [3.3.2] it is specified [by] a will (irāda).⁵¹

At this stage of the argument, we get a clear demarcation of an occasionalist position, namely option (3.3.2), which has it that the separate entity G, i.e., a divine agent according to the theologians, the only entity with power, produces e by an act of volition. Option (3.3.1), on the other hand, seems to have a Greek background—specifically certain Neoplatonist philosophers whom Avicenna was familiar with. In this regard, John Philoponus (d. circa 574) stands out; he taught that nature is an active, non-volitional principle that, though existing outside bodies, nevertheless somehow permeates them and thereby manages or regulates them (tadbīr).⁵² A number of Arabic philosophers also maintained this Neoplatonic view of nature.⁵³ Whoever the proponents of (3.3.1) were, the shaykh rejects the view by noting that it reduces to one of the divisions already dealt with: [T22] If [3.3.1] the very power necessitates that [act], then necessitating that [act] from this body itself must be due to one of the reasons mentioned [above] – and so the argument is brought back to where it was before.⁵⁴

The reasoning seems to be: insofar as e is due to G’s power in a necessary way, that causal power, considered as such, is indifferently related to all bodies. Why then, one might wonder, does e regularly occur when, say, x and y are related or when x and z are related, and not when say z and y— or some other connection not involving x—are related? It cannot be because of some fact about G’s power itself; for, again, that power is equally related to all bodies. The following example may illustrate the force of the argument at T22: consider a source of light, e.g., the sun. Although it equally or indifferently shines on rocks x and y, yet we observe only rock x reflecting its light. Why is this the case? Nothing ⁵¹ The Healing; Metaphysics IV.2.21, 138. ⁵² See Philoponus, On Aristotle’s Physics 2, 197.34–198.1. ⁵³ For the details of this Neoplatonist view of nature, as well as its reception among Arabic philosophers, and Avicenna’s critical reaction to it in the context of Aristotle’s definition of nature, see Lammer, The Elements of Avicenna’s Physics, ch. 5. ⁵⁴ The Healing; Metaphysics IV.2.21, 138.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

26   about the sun’s rays can explain the difference, for they fall on both rocks equally. The explanation must then be sought on the side of x, namely, that it has some property that rock y lacks—say, the fact that it is a diamond and so is of a particular material constitution which reflects, and so has the capacity to reflect, light. The point then is that if e’s occurrence is due to something on the side of body x, then, as T22 indicates, we just run again those steps of the argument that we did when dealing with options (3.1) and (3.2) earlier. The upshot of the argument contra (3.3.1), then, is that in order to explain the characteristic behavior of a body, there is no need to posit nature as an entity extrinsic to that body, which produces acts from it by somehow being diffused in it or governing it. Rather, we must inevitably end up appealing to some factor immanent in the body in question to account for the production of acts or effects from it. Hence, the explanation will necessarily be in terms of properties the body possesses, and not some transcendent universal nature. And so, even if nature as the Neoplatonists conceive it exists, it cannot explain the phenomenon they posit it to explain. In order to avoid these consequences, the anti-realist about powers as intrinsic features of bodies must then hold that effect e is due to some additional factor on the side not of x, but of G, which he then identifies with a specifying act of volition (irāda). If so, that move amounts to (3.3.2) of T21, which is the one I noted the Ash aris ͑ ultimately opt for. In addressing (3.3.2), the shaykh again appeals to the QT principle: [T23] As for [3.3.2] its being by way of volition, that volition distinguished this body from all other bodies either [3.3.2.1] by a property peculiar to [the body], or [3.3.2.2] randomly (juzāfan) and in however a chance way.⁵⁵

That is, if per (3.3.2) e issues from x in a regular manner because G wills its occurrence on every occasion, then either there must be some principled reason why G’s volition constantly specifies x (as opposed to any other body x*) as that from which to produce e specifically (as opposed to any other effect e*), or there is no such reason, in which case it just so happens that G does so, in the sense that the specification by G is ⁵⁵ Ibid.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

27

completely arbitrary.⁵⁶ The latter alternative implies that it is possible for G to will e from y just as much as it in fact does from x, or to will e* from x just as much as it in fact wills e from x (e.g., heat from snow just as much as from fire). Disjunct (3.3.2.2) is then ruled out on the grounds that it fails to account for the consistent way in which e in fact proceeds from x: [T24] If [3.3.2.2] randomly and in however a chance way, [the event] would not continue [happening] in this perpetual or for the most part order (niẓām al-abadī wa’l-aktharī); for chance affairs are those that are neither always nor for the most part. But natural things are either always or for the most part. Therefore, they are not by chance.⁵⁷

That is, random or chance events, by definition, are those that do not occur either always or for the most part. But the putative causal connections in the world happen either always or for the most part, as we can empirically observe. Hence, the divine will qua operating in an arbitrary way cannot be the explanation. The Ash arite ͑ will surely have to concede this point, insofar as he denies that God does things in a random way in the sense characteristic of someone who lacks knowledge or is ignorant and whose behavior is hence irrational. That much of a concession is enough for the Avicennian to infer that there must be a reason why the divine will regularly specifies the events that it does. However, the issue bears further consideration (fīhi naẓar). For although the Ash arite ͑ may grant this conclusion, namely, that the divine will operates in a non-arbitrary manner, he will still have to deny, contra what the shaykh wants at T23, that the explanatory factor must be located on the side of the body that the divine will specifies for the production of the act or effect, e. Instead, the explanatory factor appealed to, the Ash arite ͑ may urge, remains in God Himself. And this factor is what they call the divine custom or habit (al-āda).⁵⁸ ͑ So, on their view, God’s habit or custom explains the regular connections between events in that those events display order and regularity due to God’s ⁵⁶ In other words, the proponent of (3.3.2) may hold that although there is a reason why the effect occurs (as opposed to not occurring), there is no reason why that effect occurs, and no reason why it proceeds from the specific body it in fact consistently proceeds from. ⁵⁷ The Healing; Metaphysics IV.2.21, 138. ⁵⁸ Ibn Fūrak, Mujarrad 131–2, 134.5–8, and 176. Bāqillānī, Kitāb al-bayān, 50–5.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

28   constantly creating them in that way on each particular occasion. He constantly chooses to will e from x as opposed to y and that is the sufficient reason for e regularly following x as opposed to y. Nothing in the argument at T24, one might think, tells against this. But one would be wrong to think that. For what is at issue in T24 is not the cause of x’s regularly mediating e’s occurrence. At this point, we already know the identity of that cause: it is the divine will. Rather, what is at issue is the reason for the divine volition’s consistently choosing x (as opposed to some non-x) to mediate the production of e. In other words, the divine habit explains e’s regular procession from x, not why the volition itself regularly chooses x over its alternatives. At T24, the shaykh is asking about the reason for the divine regularity (= habit), not the regularity on the side of x, y, and e. The Ash arite ͑ at this point might simply insist that the regularity of the divine will has no ground in anything over and above mere acts of volition—i.e., the choosing of one possibility over another. This response, however, is no good, for the Avicennian will point out that he is committed to the following claim: The divine will in itself is indifferently related to every possibility. The possibilities in this case are effects and the occasions to produce them. As one later member of the school, Juwaynī (d. 1085), put it in his al-Irshād: [T25] For every event, God wills its coming into being. And it is not the case that God’s volition is specifically related to one class of events to the exclusion of another class.⁵⁹

If that is the case, then a difficulty arises for the Ash arite—namely, ͑ that the QT principle, a premise both parties share, is violated. For the divine will’s relation, per T25, to both x and ~x (for some ~x, e.g., y) being one of equality means that its specification or attachment (taalluq) ͑ to either alternative is contingent in the sense that there is no necessity (ḍarūra) that it should (consistently) choose one alternative over the other. But, if so, the contingency involved in the divine will’s choice on any given occasion demands, given QT, a reason/cause (illa) ͑ why it opts for one ⁵⁹ Juwaynī, Kitāb al-irshād, 238.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

29

alternative over the other on that occasion. Otherwise, the Ash arite ͑ is then effectively saying: (a) x is contingently F, but (b) there is no reason or cause why x is F, in which case we have a determination of some contingent without a determining factor—a violation of the QT principle. At this point, the Ash arite ͑ may bite the bullet and reject QT by insisting that no additional explanatory factor is needed despite the contingency of the divine will’s choice (of x over ~x, for any ~x). That is, he may insist that the divine will just brutely and contingently specifies x over ~x in a regular way, given that [T26] [ . . . ] the will is nothing but an attribute whose function is to distinguish one thing from among its counterparts.⁶⁰

The trouble with this view is that it opens the door to Humeanism; for, as far as contingency goes, the divine will and any other event are equivalent. Hence, if no determining factor is required in the case of the divine volition’s (contingent) specification of x over ~x, then it is also not required in the case of the (contingent) event of x occurring at the moment it does, as opposed to ~x’s occurring. And if there is no difference at all between the two with respect to contingency holding of both, then not only is the regularity of events that we started with rendered unintelligible, but any way to establish a divine cause of such events is blocked. A second difficulty also arises at this point—namely, the principle of non-contradiction (PNC) will be violated. For the divine will cannot both regularly choose x specifically as opposed to ~x, and be equally related to the two. In other words, according to the Avicennian, the following two commitments of the Ash arite: ͑ (i) The divine will is equally related to x and ~x and (ii) The divine will regularly chooses x over ~x

⁶⁰ Ghazālī, Al-Iqtiṣād II.1.4, 109.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

30   are inconsistent with each other. For “equality,” i.e., of the divine will in relation to x and ~x, and “preponderance” or determination, i.e., of that will for x over ~x, are contradictory properties. Hence, it would be a contradiction to hold that both are true of the divine will in one and the same respect. And so, on analysis, it turns out that claims (i) and (ii) are denials of each other, so that if the divine will regularly chooses x over ~x (for some ~x), this means that it is not equally related to ~x; and vice versa. The upshot of this argument is that there is pressure on the occasionalist opponent to give up either claim (i) or claim (ii). If he gives up (ii), he gives up a self-evident experiential premise. If he gives up (i), then the Avicennian can conclude that the divine will is not equally related to x and ~x, and hence that, as T24 argued, the explanation for the regularity of events is not the divine will qua operating arbitrarily or randomly but qua operating on the basis of some principled reason. Now, with T23’s option (3.3.2.1), we have seen that the shaykh wants to locate this reason in some factor belonging to the body from which God, according to the occasionalists, produces an effect. But suppose the Ash arite ͑ again grants this much, but again denies, against (3.3.2.1), that the explanatory reason must be located in the body from which the effect is willed. Instead, he might claim, given that God’s activities are wise activities, the explanatory reason, whatever it may actually be, falls under His attribute of knowledge. In my view, this move is also no good on Ash arite ͑ principles. The reason is that they hold the view that knowledge as such [T27] [ . . . ] depends on what is known, attaches to it as is, and does not affect it or change it.⁶¹

In other words, knowledge on the Ash arite ͑ view is conditioned by its object, which in this case would be x, i.e., the body from which God regularly produces some effect e. So on this account, God’s knowledge of x, i.e., that it mediates e, must be determined by some relation intrinsic to x and e themselves. And this in turn explains why He regularly chooses to bring e into existence from x as opposed to any ~x. But if so, when we inquire into the nature of this intrinsic relation between x and e, we ⁶¹ Ibid., II.1.4, 107.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

31

realize it must hold either because of something about x qua body or because of some factor additional to that. And hence, God must regularly choose x to mediate e either because of the mere fact that x is a body or because of some additional feature that it has. With this, we see that disjuncts (3.1) and (3.2) of the Avicennian argument resurface once again, in which case we run the same steps as we did before when we first encountered them. Having eliminated (3.3.2.2), the shaykh concludes to (3.3.2.1): [T28] It [3.3.2.1] remains, therefore, that [the will specifies the body] due to a property by which [the body] is distinguished from the rest of the bodies, and from this property is willed the procession of that act.⁶²

That is, in terms of bringing about e’s existence, the relation of the divine will to x must be grounded in some special property of x. At this stage of the argument, it might seem that the shaykh has his conclusion—namely, there exists some power-property F in x in virtue of which G wills e from x. But that doesn’t seem right. For occasionalism still stands, the Ash arite ͑ might urge, insofar as the conclusion at T28 is compatible with an account of power-properties on which they are purely passive, in the sense of equally related to the effects that we observe occurring. And if so, then such power-properties would not be making a real causal contribution to the observed effects. In Avicennian terms, they would not be active necessitating principles of those effects—which is what they need to be for occasionalism to be false. Moreover, an account on which such power-properties are indifferently passive is one on which they are immediately dependent on some external agent that activates them and thereby produces e upon their activation. So on this view, although the occasionalist gives up his anti-realism about powers by granting the Avicennian that the divine agent regularly chooses to produce e from x because of a power-property F in x, he can still hold that F is indifferently related to effects e and e*. That is, the occasionalist might still hold that F is causally inert, such that the only determining reason why e regularly proceeds from F is the specification of the divine will. If so, then F in fact

⁶² The Healing; Metaphysics IV.2.22, 138.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

32   is not really a principle (mabda ͗) of e’s occurrence, and so does not make any real causal contribution to that occurrence.⁶³ It seems to me that the shaykh explicitly recognizes something like this possibility with option (3.3.2.1.3) at T29 below, which is why he pushes the analysis further instead of terminating the argument at T28. We’ll deal with (3.3.2.1.3) later, at T33. For now, what is important is that the claim at T28 is not the desired conclusion. This is also why the next move the shaykh makes is to zero in on the relation obtaining between property F, which his opponent just granted exists in body x, and effect e (that God, on the occasionalist picture, wills from F) and inquire into why the external cause G regularly brings it specifically about: ⁶³ Some interpreters seem to attribute this passive view of powers, or at least one like it—which is a sort of intermediate position between traditional Ash arite ͑ occasionalism and the Avicennian theory of efficient causation—to Ghazālī himself in the 17th discussion of the Tahāfut, as well as in certain other works like the Iḥyā ͗ Ulūm ͑ al-Dīn, al-Maqṣad al-Asnā, and the Al-Iqtiṣād itself. See e.g., Jon McGinnis, “Occasionalism, Natural Causation and Science in Al-Ghazālī,” in Montgomery (ed.), Arabic Theology, Arabic Philosophy, 441–63. Other interpreters agree with this sort of view insofar as they too think that Ghazālī did not endorse traditional Ash arite ͑ occasionalism— see e.g., Frank, Creation and the Cosmic System, 12–21, and Griffel, Al-Ghazālī’s Philosophical Theology, 204–5. For my part, basing myself strictly on Al-Iqtiṣād, I think that Ghazālī endorses classic occasionalism, and so I disagree specifically with McGinnis’s reading of the Ghazālian texts, especially in the Tahāfut, given that in that work, Ghazālī explicitly says that he is not putting forth any positive doctrines of his own. But that is an interpretative issue the details of which I am not interested in pursuing here. My basic point is, even assuming that McGinnis is right that Ghazālī believed in creaturely passive powers, he did not think that those powers were efficacious (mu aththir) ͗ in any sense. As the al-Iqtiṣād puts it, although the creaturely power “attaches” to the object of power (maqdūr), it does not “create” or “produce it.” The only reason why the Ash arites, ͑ Ghazālī included, posit a creaturely power in the first place is to preserve the distinction, known of necessity, but which they claim the Mujbirites deny, between a voluntary movement and a tremor. It is not to show that (human) creatures make a real causal contribution to an event’s occurrence—see Al-Iqtiṣād II.1.1, 94. But if so, this is just classic Ash arite ͑ occasionalism—i.e., the view that all efficacious power belongs to God, not that all power belongs to God. Relatedly, I would also disagree with McGinnis’s analysis (at “Occasionalism,” 443–4) of what Avicenna is up to in Kitāb al-najāt; ilāhiyyāt III.2.1, 546–7. In that context, Avicenna is thought to be offering an argument for necessary causal relations, an argument which McGinnis says Ghazālī then addresses at Tahāfut VII. This, however, does not seem right. For clearly that section of the Najāt, as its very title indicates, is about what the notions of necessary and possible existence mean, on which the shaykh will draw in the subsequent chapters to prove certain theses. One of these is that the relation between a cause and its effect is one of necessity, and the argument proper for that only comes two chapters later at Najāt III.2.3, 548–9, where it is established that nothing exists unless it is necessitated to exist, i.e., by whatever causes it. And this argument at III.2.3 548–9 is just a more condensed version of the one offered for the same conclusion at The Healing; Metaphysics I.6, which McGinnis, correctly I think, believes Ghazālī accepts because it is “innocuous for [his] own theory of efficient causation” (“Occasionalism,” 448) in so far as Ghazālī goes on make the necessitating factor an act of (divine) volition. But if so, then the argument at Najāt III.2.3, 548–9 is equally harmless for Ghazālī’s theory, and so he need not be interpreted, à la McGinnis, as seriously responding to it at Tahāfut VII. As far as the Avicennian argument proper for causal powers is concerned, Ghazālī nowhere addresses it in the 17th discussion of the Tahāfut.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

33

[T29] It must then be that either that [act] is willed because [3.3.2.1.1] that property necessitates that act or [3.3.2.1.2] it comes to be from [the property] for the most part or [3.3.2.1.3] it neither necessitates [the act] nor does it come to be from [the property] for the most part.⁶⁴

If, as we have seen, the explanation for e’s regular occurrence can neither be grounded in G’s divine will itself, nor in the fact that F’s possessor, x, is a body, then it must be grounded in the power F itself. If so, then F itself will either be a power essentially directed to the production of e specifically as opposed anything else e*, or F will be a power indifferently related to whatever effect G happens to produce from it at any given moment. This latter alternative is the same as the occasionalist option I just delineated above and identified with option (3.3.2.1.3) of T29, on which F is a causally inert passive power, i.e., indifferently related to effects e and e*. Now if the former is the case, i.e., if F is essentially directed to e, then, as T29 has it, either F is such that it essentially necessitates e, which is option (3.3.2.1.1) of T29, or F essentially renders e to occur for the most part, which is option (3.3.2.1.2) of T29. Obviously, no bona fide occasionalist, Ash arite ͑ or otherwise, can opt for (3.3.2.1.1); otherwise, it is straightforwardly clear that occasionalism is false: [T30] If [3.3.2.1.1] [the property] necessitates [the act], then it is a principle (mabda ͗) of that [act].⁶⁵

That is just what it means for something to be a principle, given QT. Nor will disjunct (3.3.2.1.2) be any good for the occasionalist; for the problem with that option is that it reduces, on analysis, to the first one, i.e., (3.3.2.1.1). The shaykh explains: [T31] If [3.3.2.1.2] for the most part, that which is for the most part – as you learned in The Physics – is identical (bi- aynihi) ͑ to that which necessitates, but it has an impediment. For its being characterized as having the thing come to be for the most part is through an inclination (mayl) from its nature in the direction of what is from it. Hence, if [the thing] does not exist, this is due to an impediment.

⁶⁴ The Healing; Metaphysics IV.2.22, 138.

⁶⁵ Ibid.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

34   Therefore, what is for the most part is also in itself necessitated, if there is no impediment [ . . . ].⁶⁶

Recall what option (3.3.2.1.2) states: e is regularly brought about by G from F because F is of itself such that it leads to the production of e, not necessarily or always, but only for the most part. Avicenna responds to this by noting that the difference between a case where F leads to e for the most part and a case where it leads to e necessarily or always is that, in the former case, there is some obstacle present which bars F from producing its characteristic effect, e. Hence, were the obstacle not there, F would always in fact produce e.⁶⁷ The upshot of the argument at T31, then, is that the external entity G regularly produces e from F because F is essentially a principle of e. For F’s rendering e’s occurrence for the most part is because of some obstacle which, were it absent, F would always bring e about. And that which is always is, in a sense, that which is necessary. Therefore, F is related to e as the principle which necessitates it, which is just option (3.3.2.1.1) of T29. With (3.3.2.1.1) and (3.3.2.1.2) eliminated, option (3.3.2.1.3) remains to be treated. According to this account, which I noted is the one the Ash arite ͑ must ultimately choose, F is equally related to e and ~e (for some ~e). More precisely, the claim is that F, despite e’s regularly proceeding from it through G, neither necessitates e nor renders it to occur for the most part. This move consists in the Ash arite ͑ introducing an indifference, similar to the one he affirms in the divine will (as stated in T25), into the power F, which power he concedes exists in x. The shaykh rejects this view by arguing that it would again entail that e’s occurrence is a random event: [T32] If [3.3.2.1.3] this property [in x] neither necessitates [the act] nor does it [occur] from it for the most part, then [the act’s] being from it and from something other than it would be one and the same; and hence, [the act’s] being specifically by [that property] would be random. But it has been stated that it is not random.⁶⁸

⁶⁶ Ibid., IV.2.22, 138–9. ⁶⁷ For the full argument for the reduction, see The Healing; Physics I.13.6, 83–4. ⁶⁸ The Healing; Metaphysics IV.2.22, 139.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

35

In other words, if, per (3.3.2.1.3), F is completely neutral with respect to the occurrence of e specifically, then e’s proceeding from F and from anything other than F would be identical. And hence, e’s occurrence from F would be a purely chance affair. But we decidedly know that its being by chance is false; since, again, we observe that e regularly proceeds from F. Hence, F cannot be an indifferent power. Regularity is opposed to randomness; so this last option too fails to account for the fact that e constantly occurs from F. The same argument can be put another way: if the fact that effect e is regularly brought about by G from power F is in need of an explanation, then we are in need all the more of an explanation for why this happens not only regularly but with the additional condition, per (3.3.2.1.3), that F is itself a purely passive or indifferent power. For if an indifferently related divine will cannot explain (as we have seen) the regular occurrence of e from x, then a fortiori an indifferently related divine will plus an indifferently related power F in x cannot explain e’s regular occurrence from x. For, again, a thing’s being indifferent to e and its regularly bringing about e are incompatible properties. Thus, either F is not a purely indifferent power or, if it is, e’s regular occurrence is inexplicable. The occasionalist is in a bind. Before drawing the final conclusion, the shaykh considers one last possibility—which does not concern the occasionalist per se—that someone may opt for in order to avoid the conclusion that F itself is what necessitates e. The opponent suggests that e’s production is more properly attributable not to F but to that which possesses F as a property, namely, body x, so that x has a better claim than F to occupying the role of the principle of e. [T33] Likewise, if it is said: “[the act’s] being [from] the possessor (ṣāḥib) of that property is more appropriate (awlā)”, then its meaning is that [the act’s] procession from it is all the more fitting (awfaq). And so [the property], then, [either] necessitates [the act] or facilitates (muyassir) its necessitation. That which facilitates is either a cause essentially ( illa ͑ bi’l-dhāt) or accidentally (bi’l- araḍ ͑ ). And so if there is no essential cause other than [the property], then it is not an accidental [cause]; for that which is an accidental [cause] is in accordance with one of the two mentioned ways.⁶⁹

⁶⁹ Ibid., IV.2.23, 139.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

36   The response to the objection states that it makes no real difference whether the production of e is more properly attributable to F or to what possesses F as a property, i.e., to x. That is, e’s occurrence will still be accounted for in terms of F—whether alone or conjunctively with its bearer, x. This is because if we assume that it is xF jointly, as opposed to just F singly, that is responsible for e’s occurrence, F’s causal role in that, as the shaykh states, would either have to be (i) one of necessitation or (ii) one in which F merely aids its bearer x in necessitating e. If (i), then obviously F necessitates e, in which case F is properly a principle of e. But this is just the conclusion sought (maṭlūb). As for (ii), as already indicated in T20, in the context of The Physics Avicenna calls the “facilitating” or “aiding cause” a “preparing” (muhayyi ͗) principle, and defines it as “that which makes the matter suitable,” i.e., for the action of something else so that some effect may be produced.⁷⁰ If we then assume, per option (ii), that F merely aids e’s being necessitated to exist, then F will then, as the middle of T33 states, either be (ii.1) a per se (bi’l-dhāt) or (ii.2) a per accidens (bi’l-araḍ) facilitating cause of e.⁷¹ For F to be a per se or essential cause of e is for it to be a “principle of that very [effect] itself and taken insofar as it a principle of it,” whereas for F to be an accidental cause of e is for it to not be an essential cause in the sense just defined.⁷² The accidental cause comes in various types, all of which presuppose the essential.⁷³ So if, then, F alone (i.e., F without any other causal factors involved) is (ii.1) a per se facilitating cause of e, such that it brings e about regularly, then F cannot be an accidental cause of e in any sense, especially not in the two senses we have been considering so far, namely, as a chance or a forcing cause. But as a per se cause of e, F is a necessitating principle of e—which, again, is just the conclusion sought. But should one opt to hold that F is (ii.2) a per accidens cause of e, there must then exist some cause F* that is a per se cause and which F is parasitic upon. And the argument will then be about F*. The upshot: either we end in a per se cause of e or an infinite regress threatens, in which case e’s existence will never be necessitated, with the result, given the QT principle, that e

⁷⁰ The Healing; Physics I.10.3, 65 (tr. J. McGinnis). ⁷¹ For Avicenna’s explanation of this distinction, see ibid., I.12.2, 74–5. ⁷² Ibid., I.12.2, 74. ⁷³ See footnote 35.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

37

will not in fact exist. But this contradicts the datum that it in fact exists, and does so regularly. Therefore, the property F must be a per se, necessitating cause of e. The shaykh then concludes: [T34] Therefore, it remains that that property in itself a necessitating [principle]. And so the necessitating property is called power (quwwa). And this power is [one] from which bodily activities proceed, even if by the aid of a more remote (ab ad) ͑ principle.⁷⁴

The argument for the existence of causal powers in bodies ends here. As I noted at the beginning of this inquiry, T34’s conclusion is also meant to be taken as showing that nature (ṭabī a)͑ in the technical Aristotelian sense exists; for on Avicenna’s view, nature just is a power, and a power, considered in a certain respect, just is nature (in that technical sense). If so, the argument fulfills both its aims; on the one hand, it establishes the existence of nature qua power, and on the other, it refutes occasionalism, which is probably the greatest challenge to Aristo-Avicennian natures. For it shows that some body produces an effect e in virtue of F where F is a power-property of that body—and this even on the assumption, per the end of T34, that the body requires the causal cooperation of an immaterial entity to do so. In sum, if it is true that Some body possesses efficacious causal power then occasionalism, understood as a commitment to OC above, i.e., that Only God possesses efficacious causal power is not true. In light of the above analysis, we can summarize the shaykh’s argument in a fifteen-premise argument: 1. Upon contact between body x and y, some effect e regularly occurs (empirical premise). 2. e is due to either (1) x qua body, (2) a power F in x, or (3) some incorporeal entity G.

⁷⁴ The Healing; Metaphysics IV.2.23, 139.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

38   3. ~ (1); otherwise, every body would produce e. But not every body does. 4. If (2), that is the desired conclusion. 5. If (3), G regularly produces e from x either due to (3.1) x qua body, (3.2) a power F in x, or (3.3) a power F* in G. 6. ~ (3.1); otherwise, e would regularly occur from any body. But it does not. 7. If (3.2), that is the desired conclusion. 8. If (3.3), then F* either (3.3.1) necessitates e or (3.3.2) specifies e because of a will, w (= OC). 9. If (3.3.1), that is either because of (3.1) x qua body or (3.2) a power F in x. Whatever the case, the desired conclusion is established. 10. If (3.3.2), then w regularly specifies x (over some non-x) for the production of e either (3.3.2.1) because of something, F**, peculiar to only x or (3.3.2.2) w does so randomly. 11. ~ (3.3.2.2); otherwise, there would be no regularity. This contradicts 1. 12. If (3.3.2.1), then w wills e from x’s F** either because F** (3.3.2.1.1) necessitates e, or (3.3.2.1.2) renders e to occur for the most part or (3.3.2.1.3) neither necessitates nor renders e to occur for the most part. 13. ~ (3.3.2.1.3); otherwise, e would be random. This contradicts 1. 14. If (3.3.2.1.2), this reduces to option (3.3.2.1.1). 15. Therefore, (3.3.2.1.1). And this is the conclusion sought (maṭlūb). I leave the argument at that.

5. A Question and a Suggestion Assuming he is serious, Avicenna’s willingness to admit the causal relevance of a separate entity in the events of nature raises the question of what sort of causal model of the Creator–creature relation he endorses. In the contemporary literature, there are three main accounts: conservationism,

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

39

concurrentism, and occasionalism.⁷⁵ Roughly, on conservationism, the separate entity only conserves a body and its powers in being, but makes no causal contribution to the effects of that body’s powers at all. That is, the body itself is the only cause of those effects. On the concurrentist view, the separate entity not only conserves the body and its powers in being, it also, together with the body, contributes to the body’s effects. As for occasionalism, as should be clear by now, it maintains that the separate entity is the only cause, both of the body, its powers, and the effects that are (mistakenly) attributed to those powers. Where does Avicenna stand on this scheme? The answer, it seems, depends on what the creaturely-effect is. It is likely that he is a conservationist if the effect in question is motion of any sort (or an accidental form more generally). But if the effect is a substantial form, then he can be understood as a modified occasionalist—I say “modified” because, on this view, the separate causal agent, as already noted, is not God à la the Ashā ira ͑ but an immaterial entity lower on the chain of being, i.e., the Giver of forms. The explication and justification of these claims, though, I leave for another occasion.

6. Conclusion In relation to the two traditions it is meant to address, the Avicennian argument we have been considering in detail is interesting and unique for at least two reasons. First, unlike Aristotle, Avicenna does not think that the existence of nature as a power is something unqualifiedly selfevident. That is, it is not self-evident in such a way that it can be justified only indirectly or dialectically, i.e., by the absurdities that would follow were it to be denied. This, as we have seen, is Aristotle’s preferred method, at least on a certain interpretation, in the Physics and Metaphysics. There were at least two reasons why the shaykh denied that nature’s existence is self-evident in this sense. The first has to do with a reason internal to the Aristo-Avicennian system: namely, with the way ⁷⁵ For a useful description of these three models of God’s causal relation to the world, see Alfred Freddoso, “God’s General Concurrence with Secondary Causes: Pitfalls and Prospects,” American Catholic Philosophical Quarterly 67 (1994), 131–56. See also Thomas V. Morris (ed.), Divine and Human Action: Essays in the Metaphysics of Theism (Ithaca, NY: Cornell University Press, 1988).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

40   the hierarchical relation among the various sciences is conceived, such that the principles of a lower science must be proved in a higher one. The second reason plausibly has to do with pressures from his Ash arite ͑ contemporaries who, as we have seen, accepted the fact that bodies are involved in events of various sorts but denied that those events occurred because of some intrinsic power the bodies involved in them possessed. In connection with this last point, the second reason the Avicennian argument is interesting is that, when addressing occasionalism, it concedes, as is evident from the end of T34, a fundamental tenet of that doctrine while still showing that there must be some factor F in a body in virtue of which some putative effect e is produced. So it does not deny outright an important facet of Ash arism, ͑ namely, the causal role of an external (in their view divine) agent in the events of the world. Rather, it accommodates that tenet in some way, and says: even so, there must be some feature in a body in virtue of which the phenomenal effect is either always or for the most part, i.e., regularly, produced with the assistance of that external cause. In my view, that is a feature of the argument that gives it powerful appeal.⁷⁶ Ludwig Maximilian University of Munich University of Jyväskylä

Bibliography Adamson, Peter. “Against Nature: Two Critics of Naturalism in the Islamic World,” in P. Adamson and C. Rapp (eds.), State and Nature: Essays on Ancient Political Philosophy (Berlin: De Gruyter, forthcoming). Al-Ash arī, ͑ Abū l-Ḥasan. Al-Ibānah an ͑ Uṣūl ͑ al-Diyānah (Hyderabad, 1903). Al-Ash arī, ͑ Abū l-Ḥasan. Al-Luma ͑, in R. J. McCarthy (ed.), The Theology of al-Asharī ͑ (Beirut: Catholic Press, 1953). Al-Ash arī, ͑ Abū l-Ḥasan. Maqālāt al-Islamiyyīn wa-l-Ikhtilāf al-Muṣallīn, ed. H. Ritter (Wiesbaden: Franz Steiner Verlag, 1963). Al-Bāqillānī, Abū Bakr. Kitab al-Inṣāf, ed. M. Z. Al-Kawtharī (Cairo, 1950). ⁷⁶ My thanks to Peter Adamson, Andreas Lammer, and an OSMP referee for helpful comments and criticisms.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

41

Al-Bāqillānī, Abū Bakr. Kitāb al-Tamhīd, ed. R. J. McCarthy (Beirut: Librairie Orientale, 1957). Al-Bāqillānī, Abū Bakr. Kitāb al-Bayān, ed. R. J. McCarthy (Beirut: al-Maktab al-Sharkiyya, 1958). Al-Bāqillānī, Abū Bakr. Al-Insāf, ed. M. Z. Al-Kawtharī (Cairo: Al-Maktabah al-Azhariyya li’l-Turāth, 2000). Al-Ghazālī, Abū Ḥāmid. Incoherence of the Philosophers, tr. M. Marmura (Provo, UT: Brigham Young University Press, 2000). Al-Ghazālī, Abū Ḥāmid. Moderation in Belief, tr. A. M. Yaqub (Chicago: University of Chicago Press, 2013). Al-Juwaynī, Abd ͑ al-Malik Ibn Yūsuf. Kitāb al-Irshād ilā qawāṭi ͑al-adilla fī ed. M. Y. Mūsā and A. Abd ͑ al-Ḥamīd (Baghdad: al-uṣūl al-I tiqād, ͑ Maktabah al-Khānjī, 1950). Aristotle. The Complete Works of Aristotle, ed. J. Barnes (Princeton, NJ: Princeton University Press, 1984). Aristotle. Metaphysics Book Θ, tr. S. Makin (Oxford: Clarendon Press, 2006). Aristotle. Physics: Book VIII, tr. D. Graham (Oxford: Clarendon Press, 2007). Avicenna. The Demonstration of the Healing, ed. Abū al- Alā ͑ Afīfī ͑ (Cairo: Al-Matba ah ͑ al-Amiriyya, 1954). Avicenna. Uyūn ͑ al-ḥikma, ed. A. R. Badawī (Beirut, 1980). Avicenna. Kitāb al-Najāt, ed. M. T. Danishpazhuh (Tehrān: Danishgah-i Tihran, 1985). Avicenna. The Metaphysics of The Healing, tr. M. Marmura (Provo, UT: Brigham Young University Press, 2005). Avicenna. The Physics of The Healing, tr. J. McGinnis (Provo, UT: Brigham Young University Press, 2009). Beere, Jonathan. Doing and Being: An Interpretation of Aristotle’s Metaphysics Theta (Oxford: Oxford University Press, 2009). Bertolacci, Amos. The Reception of Aristotle’s Metaphysics in Avicenna’s Kitāb al-Shifā ͗: A Milestone of Western Metaphysical Thought (Leiden: Brill, 2006). De Haan, Daniel. “Where Does Avicenna Demonstrate the Existence of God?,” Arabic Sciences and Philosophy 26 (2016), 97–128. Dutton, Blake D. “Al-Ghazālī on Possibility and the Critique of Causality,” Medieval Philosophy and Theology 10 (2001), 23–46. Frank, Richard M. Creation and the Cosmic System: Al-Ghazālī and Avicenna (Heidelberg: Carl Winter, 1992).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

42   Frank, Richard M. Al-Ghazali and the Ash arite ͑ School (Durham, NC: Duke University Press, 1994). Freddoso, Alfred. “God’s General Concurrence with Secondary Causes: Pitfalls and Prospects,” American Catholic Philosophical Quarterly 67 (1994), 131–56. Griffel, Frank. Al-Ghazālī’s Philosophical Theology (Oxford: Oxford University Press, 2009). Gutas, Dimitri. Avicenna and the Aristotelian Tradition (Leiden: Brill, 2014). Ibn Fūrak. Mujarrad Maqālāt al-Ash arī, ͑ ed. D. Gimaret (Beirut: Dar El-Machreq, 1987). Johnson, Rebekah. “The Existence of Powers,” Apeiron 41 (2008), 171–92. Lammer, Andreas. “Defining Nature: From Aristotle to Philoponus to Avicenna,” in A. Alwishah and J. Hayes (eds.), Aristotle and the Arabic Tradition (Cambridge: Cambridge University Press, 2015), 121–42. Lammer, Andreas. The Elements of Avicenna’s Physics: Greek Sources and Arabic Innovations (Berlin: De Gruyter, 2018). McGinnis, Jon. “On the Moment of Substantial Change: A Vexed Question in the History of Ideas,” in J. McGinnis (ed.), Interpreting Avicenna: Science and Philosophy in Medieval Islam, Proceedings of the Second Annual Symposium of the Avicenna Study Group (Leiden: Brill, 2004), 42–61. McGinnis, Jon. “Occasionalism, Natural Causation and Science in Al-Ghazālī,” in J. E. Montgomery (ed.), Arabic Theology, Arabic Philosophy. From the Many to the One: Essays in Celebration of Richard M. Frank (Leuven: Peeters, 2006), 441–63. Marmura, Michael. “Al-Ghazâlî’s Second Causal Theory in the 17th Discussion of His Tahâfut,” in P. Morewedge (ed.), Islamic Philosophy and Mysticism (Delmar, NY: Caravan Books, 1981), 85–112. Marmura, Michael. “Ghazâlian Causes and Intermediaries,” Journal of the American Oriental Society 115 (1995), 89–100. Moad, E. O. “Al-Ghazali’s Occasionalism and the Natures of Creatures,” International Journal for Philosophy of Religion 58 (2005), 95–101. Morris, Thomas V. (ed.). Divine and Human Action: Essays in the Metaphysics of Theism (Ithaca, NY: Cornell University Press, 1988). Philoponus. On Aristotle Physics 2, tr. A. R Lacey (London: Duckworth, 1997). Rāzī, Abū Bakr. Maqāla fī mā ba d͑ al-ṭabī a,͑ in P. Krauss (ed.), Rasā ͗il falsafiyya (Cairo, 1939).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

“  ”

43

Richardson, Kara. “Avicenna and Aquinas on Form and Generation,” in D. Hasse and A. Bertolacci (eds.), The Arabic, Hebrew, and Latin Reception of Avicenna’s Metaphysics (Berlin: De Gruyter, 2011), 251–74. Richardson, Kara. “Avicenna and the Principle of Sufficient Reason,” Review of Metaphysics 67 (2014), 743–68. Riker, Stephen. “Al-Ghazali on Necessary Causality in ‘The Incoherence of the Philosophers,’ ” Monist 79 (1996), 315–24. Sabra, A. I. “Kalām Atomism as an Alternative Philosophy to Hellenizing Falsafa,” in J. E. Montgomery (ed.), Arabic Theology, Arabic Philosophy. From the Many to the One: Essays in Celebration of Richard M. Frank (Leuven: Peeters, 2006), 199–272. Sabra, A. I. “The Simple Ontology of Kalām Atomism: An Outline,” Early Science and Medicine 14 (2009), 68–78. Simplicius. On Aristotle Physics 2, tr. B. Fleet (London: Duckworth, 1997). Sorabji, Richard. Matter, Space, and Motion: Theories in Antiquity and Their Sequel (Ithaca, NY: Cornell University Press, 1988). Ulrich, Rudolph. “Occasionalism,” in S. Schmidtke (ed.), The Oxford Handbook of Islamic Theology (Oxford: Oxford University Press, 2016), 347–63. Wisnovsky, Robert. Avicenna’s Metaphysics in Context (London: Duckworth, 2003).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

Only God Can Make a Tree Abaelard on Wholes and Parts and Some Evidence of His Later Thinking about Them Christopher J. Martin

1. Introduction Questions about the relations between parts and wholes played an important role in the development of logic and metaphysics in the twelfth century and drove the development of new theories of universals, topical inference, and the semantics of conditional propositions. Central to these discussions was the work of Peter Abaelard, which I began to examine a number of years ago in an investigation of the origins of the thesis that nothing grows associated with his followers the Nominales.¹ Since I wrote on the subject, Andrew Arlig² has explored Abaelard’s mereology further, arguing that it grounds a theory of objects, both natural and artificial, quite at odds with our folk intuitions about them, and that Abaelard thus regards such quotidian beliefs as false and only to be retained as useful fictions. I take the opposite view and hold that while Abaelard does indeed present an account of wholes and parts which supports the thesis that nothing is increased, or grows, this is a philosophical truth about quantitative wholes which he seeks to reconcile with our ordinary intuitions about living creatures, and perhaps about artifacts too. In the end, I think, Abaelard believes that he has saved these intuitions with a theory of the part–whole relationship and of the nature ¹ C. J. Martin, “The Logic of Growth: Twelfth-Century Nominalists and the Development of Theories of the Incarnation,” Medieval Philosophy and Theology 7 (1998), 1–15. ² Arlig, “Abelard’s Assault on Everyday Objects,” American Catholic Philosophical Quarterly 81 (2007), 209–27; Arlig, “Some Twelfth-Century Reflections on Mereological Essentialism,” Oxford Studies in Medieval Philosophy 1 (2013), 83–112.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

45

of increase which acknowledges the truths of mereology but insists that there is more to the world than fusions of parts.³ In the following paper, I return to Abaelard’s discussion of parts and wholes of different kinds to try to clarify further his views on identity over time, first by looking again in some detail at his own treatment of the problem, and then briefly considering some information about his views on artifacts that we may extract from the discussion of wholes and parts in the mid-twelfthcentury Introductiones Montanae Maiores.⁴ The Montani⁵ were the followers of Alberic of Paris the master who, according to John of Salisbury, was the most bitter opponent of the Nominales,⁶ i.e. Abaelard and his followers in the middle of the twelfth century. Its recent publication makes available fascinating material on the debates which took place between Alberic and Abaelard and the death blow which Alberic was able to deliver to Master Peter’s logical ambitions.

2. Abaelard on Parts, Wholes, and the Nature of Increase Mereology is the study of what medieval logicians called integral, or constitutive, wholes and parts. For them, and for the purposes of this paper, a part is always a proper part, included in, but not equal to, the whole. In its modern form mereology has been of interest to metaphysicians because of its purported minimum of ontological commitment.⁷

³ Abaelard seems in fact not to consider fusion in general but only the more restricted operation of composition-fusion in which no two items overlap. See Ned Markosian, “Restricted Composition,” in T. Sider, J. Hawthorne, and D. Zimmerman (eds.), Contemporary Debates in Metaphysics (Oxford: Blackwell, 2008), 341–63. ⁴ Egbert P. Bos and Joke Spruyt (eds.), Anonymi Introductiones Montanae Maiores (Louvain la Neuve: Peeters, 2017). ⁵ So called after Mont Sainte Geneviève where Alberic taught. See Yukio Iwakuma, “Alberic of Paris on Mont Ste Geneviève against Peter Abelard,” in J. L Fink, H. Hanesen, and A. M. Mora-Márqez (eds.), Language and Logic in the Middle Ages (Leiden: Brill, 2013), 27–47. ⁶ “ . . . nominalis sectae acerrimus impugnator,” Iohannes Sarisberiensis, Metalogicon lib. 2, ch. 10, p. 70. On Abaelard’s logic and Alberic’s attack on it see Christopher J. Martin, “Logic,” in J. Brower and K. Guilfoy (eds.), The Cambridge Companion to Abelard (Cambridge: Cambridge University Press, 2004), 158–99. ⁷ For a general survey of contemporary mereology see Achille C. Varzi, “Mereology,” . For a rather idiosyncratic survey of medieval mereology including a discussion of Abaelard’s contribution see D. P. Henry, Medieval Mereology (Amsterdam: John Benjamins, 1991).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

46  .  Wholes, or composites, are, to use the modern jargon, the sum, or fusion, of their parts and, according to at least some contemporary friends of mereology, fusion is ontologically cost free⁸—a mereological whole, or composite, is simply identical with its parts. Abaelard is among those who accept this principle: Every composite is the same as all of its parts taken together, whence whatever is attributed to the whole is attributed to all of its parts taken together, and conversely, and to whatever the whole applies, its parts taken all together apply.⁹

The scope of fusion is a considerably more controversial question and while one of Abaelard’s remarks, which I will mention below, does accept the fusion of items from different categories, it seems doubtful to me that he would accept what contemporary mereology calls unrestricted composition,¹⁰ allowing that any completely arbitrary fusion produces an object mereologically entirely on a par with all others.¹¹ This, however, is perhaps simply because there is no occasion for him to discuss the legitimacy of arbitrary fusions when his interest in general in composites is in much more than their merely mereological character. Abaelard’s ancient sources provided him with four occasions to explore the nature of the part–whole relationship and the related questions of identity and persistence over time. Two of these are traditional and so a commentary is to be expected, but two are curious and surprising. The traditional sources are, first, the account of the topical loci from a whole and from a part described by Boethius in De differentiis topicis and, second, the account of the division of an integral whole which Boethius gives in De divisione.¹² The surprising sources are what I will call the problem of the growing square and the problem of the identity of the amputee. The first two are discussed by Abaelard only in his Dialectica, the third in the Dialectica and Logica “Ingredientibus,” and the last in the “Ingredientibus” and the “Nostrorum.” The Dialectica dates

⁸ D. M. Armstrong, A World of States of Affairs (Cambridge: Cambridge University Press, 1997), 12–13. ⁹ Petrus Abaelardus, Dialectica, ed. L. M. De Rijk (Van Gorcum: Assen, 1970), p. 554:25–8. ¹⁰ See Varzi, “Mereology,” sec. 4.4. ¹¹ See Varzi, “Mereology,” sec. 4.4. ¹² On topical, or local, inference and Abaelard’s development of the ideas he found in Boethius see Martin, “Logic.”

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

47

from around 1110, perhaps even earlier, and reflects the intense debates over the correct account of inference which took place in Paris in the first decade of the twelfth century; the “Ingredientibus” and “Nostrorum” are later: the “Ingredientibus” probably from around 1115 and the “Nostrorum” from around 1120 or later. Both are commentaries on the logica vetus but also contain extended digressions on issues which had captured Abaelard’s interest.

a. The Loci from a Whole and from a Part Abaelard’s central concern in discussing topical differences and their associated maximal propositions in his Dialectica is to establish which of them will ground the truth of conditional propositions. For this he requires that there be a hyperintensional connection between antecedent and consequent. The antecedent must be such that it cannot be true without the consequent’s being true, and this must be so because in its meaning there is included the meaning of the consequent. In his rather brief discussion of them, Abaelard argues that the loci from a whole and from a part fail to provide such a warrant. Properly speaking, he notes, only one maximal proposition belongs to the locus from a whole and only the maximal proposition that is in effect its contrapositive to the locus from a part. For the locus from a whole we have the maximal proposition “if a whole exists, it is necessary that its part exists,” and from a part “if a part is removed, its whole is removed.” These fail to warrant true conditionals because a whole and its part taken singly are ontologically distinct and no true conditional can bridge such a gap. Thus, according to Abaelard, the conditional “if there is a house, then there is a wall” is false even though it is not possible for a house to exist without a wall.¹³ What follows from there being a house is rather that a foundation, a wall, and a roof, or rather, presumably, whatever it is that is really required in the proper constitution of a house, all exist together as a house. This consequence and others which are true of houses as integral wholes are warranted, Abaelard holds, not “ . . . by the from a whole, but rather may be said to hold by the locus from a defined term; ¹³ Abaelard, Dialectica, p. 346:21–8.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

48  .  for by a house nothing else is to be understood other than those three having the composition of a house.”¹⁴ The existence of a house thus requires more than the mere existence of its parts; it requires that they exist combined in the appropriate way. Houses are more interesting than mere house-parts. As is usual in his treatment of those loci that he holds are not powerful enough to ground a hyperintensional connection between antecedent and consequent, Abaelard appeals to the rejection rules provided by his underlying commitment to connexive logic to show that conditionals putatively warranted by the locus from a whole must be false. If, for example, we were to accept as true “if a house exists, a wall exists” then we would also have to accept “if only a house exists, a wall exists,” whereas what is in fact true is “if only a house exists, a wall does not exist” which cannot, Abaelard insists, be true together.¹⁵

b. The Division of Integral Wholes In his discussion of division in the Dialectica Abaelard considers three kinds of wholes: those with respect to substance, those with respect to form, and those with respect to both substance, or matter, and form.¹⁶ The first of these is again divided into wholes with respect to extension of quantity (comprehensio quantitatis), that is integral wholes, and wholes described strikingly by the twelfth century’s most famous anti-realist as with respect to “the diffusion of a common essence.” Integral wholes are further classified according to the relation of their parts to one another into those, such as a line, that are continuous, and those, such as a people, that have their parts separated (disgregatae) from one another.¹⁷ ¹⁴ Abaelard, Dialectica, p. 344:19–36. ¹⁵ While claiming to show that embarrassment would follow if the conditional “if there is a house, then there is a wall” were true, Abaelard in fact gives the corresponding argument for relative opposites to show that the conditional “if paternity exists, then filiation exists” is false. See Abaelard, Dialectica, p. 345:34–46:3. ¹⁶ Abaelard, Dialectica, pp. 546–62. Abaelard initially characterizes the third kind of whole as with respect to both substance and form and in his discussion of it as the division into matter and form gives the example of human beings as composites of the “substance of animal” and the form of rationalness or mortalness (Dialectica, pp. 560–2). Abaelard does not define wholes with respect to form and considers only the example the division of the soul into powers (Dialectica, pp. 555–60). ¹⁷ Abaelard, Dialectica, p. 547.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

49

Under continuous wholes are included both highly structured wholes such as the human body, stuffs such as water, and mixtures, such as watered down wine, whose actual parts are smaller volumes of watered down wine, but which reason divides into parts consisting of wine alone and water alone.¹⁸ About loosely integrated wholes in which each part remains spatially separate from the others, Abaelard remarks: In those wholes in which there is a separation of the parts there can be no unity of nature. For this people is not by nature one thing but naturally many. The plurality of things alone constitutes this totality, not some composition of things with respect to the same substance.¹⁹

Abaelard seems to be on his way to advocating unrestricted mereological fusion when he remarks shortly after this that it is not necessary that the parts of such wholes are subject to the same general genus. “This whiteness and this finger,” he says, “agree in one whole,” where one is a substance and the other a quality.²⁰ The context of the remark indicates that he is talking about one finger with its whiteness, but it is possible that he would have held that the same is true of a finger and, for example, the whiteness of the wall that it is pointing at. Perhaps in favor of this and perhaps, indeed, of his being an advocate of mereological universalism is a comment that he makes later in the Dialectica on the necessity for some form of composition, or organization, in an integral whole. “Perhaps,” he says, . . . composition is not necessary in all cases, as in those which are in no way one whole either, that is, by nature or by some operation, but are said to be whole only in virtue of embracing a plurality, this people, for example, or this heap of stones. But even here a certain property of congregation seems to be necessary. For neither this people nor this heap is said to exist in dispersed substances located far apart from one another. Granted, however, that what are separated by the

¹⁸ Abaelard, Dialectica, pp. 574–5. Also classified by Abaelard as integral wholes are wholes of time, place, and quantity, where he characterizes the latter as any collection of things quantified over with “omne” (Dialectica, p. 561). ¹⁹ Abaelard, Dialectica, p. 548:11–15. ²⁰ Abaelard, Dialectica, p. 548:18–21.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

50  .  greatest distance are not said to be a people or a heap, there is still here a whole with respect only to the multiplication of parts. So that the oneness of this man remaining here, and of that one living in Rome, produce this twoness, whence those who have these onenesses are said to be two.²¹

Does Abaelard’s acknowledgement of the fact that one plus one equals two indicate that he accepts unrestricted composition? It seems to me that it does not but rather the reverse; that our collection of objects to form new objects must be guided by some principle for integration, at a minimum that we wish to count them, in this case to calculate the number of men. The main problem for Abaelard in the discussion of integral wholes in the Dialectica is that of properly characterizing the relationship of the parts to the whole. He accepts, as every mereologist does, the principle that the parts of the parts of a whole are themselves parts of the whole but holds that one division of the whole into parts is privileged, that into the principal parts into which it is proximately divided. In arguing for his own account of principal parts he first sets out and claims to refute two alternatives.²² The first alternative is constructive. Its proponents hold the principal parts of a whole are just those parts of it which are not parts of parts. Abaelard objects that gerrymandering will always defeat such a division, since for any proposed partitioning of a whole into principal parts in this way we may divide differently so as to include each of those parts in a larger part of the whole. In the case of discrete parts we may simply construct the non-singleton subsets of the set of putative principal parts to produce something of which each is a proper part. For continuous wholes we have to carve differently, taking, for example, the foot of a man as one part, and the rest of his body plus his soul as the other. This is what it takes, Abaelard suggests, to defeat the claim of his soul to be a principal part and elevate his foot to that status.²³ Whether this argument can be avoided depends on whether there can be a privileged division into parts which resists this kind of restructuring, and so again depends on whether there is a limitation of some sort to composition. For now, ²¹ Abaelard, Dialectica, pp. 575:37–576:7. ²³ Abaelard, Dialectica, p. 549:26–32.

²² Abaelard, Dialectica, p. 549:4–12.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

51

however, Abelard presents it as a refutation of the constructive account of principal parts. The second account of principal parts which Abaelard rejects is destructive. Its proponents hold that the principal parts of a whole are only those parts which are such that if they are destroyed, the substance of the whole is destroyed. Abaelard notes that it is crucial that the criterion be formulated in terms of the substance rather than the quantity of the whole since the quantity is simply the whole considered as a mereological fusion of its parts, and so the destruction of any one of them destroys it. The quantity of Socrates identified as the sum of his parts, he remarks, does not survive the cutting of his nails, but the substance of Socrates does.²⁴ Against this theory Abaelard argues that the destruction of the smallest quantitative part of a house destroys not only the quantity but also the substance of the whole. His argument, unlike that which saves Socrates from the manicurist, thus relies on identifying the “substance” of a house, i.e. the house itself, with its quantitative parts. The destruction of the smallest stone in this wall, destroys this house in substance. For if this stone does not exist, this house which depends upon it does not remain in its substance, indeed it does not exist . . . For this house is taken to be nothing other than all its parts taken together. And since this smallest stone is included in the parts it is necessary that this house be this stone and all other parts together.²⁵

Abaelard notes that the argument does not show that with the removal of the smallest stone there ceases to be a house. A house remains but it is not quantitatively and so, he argues here, not substantially the same house as the one that existed before the stone was removed. That which we now refer to as “this house” is a part of what we referred to as “this house” before the removal of the stone. To the objection that since the stone which has been removed continues to exist the original substance of the house remains but a little more spread out than it was before, Abaelard replies that the house is to be identified not simply with its

²⁴ Abaelard, Dialectica, p. 549:12–19.

²⁵ Abaelard, Dialectica, pp. 549:38–550:9.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

52  .  parts but with these parts organized to form a house. Take a pebble out of the wall and that structure is destroyed.²⁶ According to this destructive criterion, Abaelard concludes, every part of a whole is a principal part. The argument relies, of course, upon the identification of substance and quantity. Applying it to Socrates, he argues, would require us to accept that he too ceases to exist as soon as one of his nails is trimmed, something which is quite contrary to what we observe: But what will we say if, from the destruction of any part, there follows the destruction of the whole substance, when we see that Socrates remains and the entire substance of a man persists in him when he has lost a nail or even a hand or foot?²⁷

Abaelard thus reinvents the ancient Growing Argument and anticipates the conclusions of contemporary mereological essentialism when he concludes that the identification of Socrates’ substance with his quantitative parts would entail that the name “Socrates,” and referring expressions like “this human being,” are equivocal, changing their reference every time the smallest amount of matter is lost or gained.²⁸ Developing the point dramatically he argues that such an identification of substance and quantity would entail, according to the destructive account of principal parts, that the amputation of a man’s hand would be murder. Against this unacceptable consequence of the destruction of the smallest part of the body, Abaelard appeals to the division of a human being into both body and soul. This division is special in that not any destruction of

²⁶ Abaelard, Dialectica, pp. 550:32–551:2. A house for Abaelard is thus not merely a “scattered object” whose parts may be located more or less close to one another. For the connection between the principle of universal fusion and the existence of scattered objects see Richard Cartwright, “Scattered Objects,” in K. Lehrer (ed.), Analysis and Metaphysics (Dordrecht: Reidel, 1975), 153–71. ²⁷ Abaelard, Dialectica, p. 552:15–21. ²⁸ Abaelard, Dialectica, p. 552:18–27. The Academic Skeptics’ Growing Argument purports to show that the identification of a substance with its corporeal parts entails that growth and diminution are in fact the corruption of an old substance and the generation a new one. Mereological essentialism is its contemporary equivalent, the theory that a material object is necessarily identical with its material parts. For the former see A. A. Long and D. N. Sedley, The Hellenistic Philosophers (Cambridge: Cambridge University Press, 1987), vol. 1, n. 28. For the latter and some of the difficulties with it see Peter Simons, Parts: A Study in Ontology (Oxford: Clarendon Press, 1987), pp. 1–3, 272–80.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

53

the quantity of a man is murder but only that which also removes his soul. Manicure is not a crime! Abaelard is here relying on his own account of principal parts, which he next states after noting that his arguments against both the constructive and destructive theories have shown them to be “completely empty” (cassata). His own account requires rather that there be a principle for privileging certain parts of a whole which is immune to arbitrary dissection and for which principal parthood, unlike parthood, is not transitive. Those parts . . . appear to us to be called principal on the conjunction of which to one another the completion of the whole immediately follows, for example, if the roof, floor, and wall are conjoined, the house is immediately completed.²⁹

Abaelard provides no argument in any of his surviving works to show how this definition avoids the destruction of the house with the removal of the smallest pebble from the wall. We will see shortly, however, that he does seem to have eventually developed an account of principal parts which provides one. Before that, however, let me now turn to the two less obvious places where Abaelard discusses the part–whole relationship and identity over time.

c. The Growing Square In chapter 14 of the Categories, Aristotle provides the only account of his views on change available in the first half of the twelfth century. He classifies it into six kinds: generation, corruption, enlargement or growth (crementum), diminution, alteration, and translation in place; each of these, he argues, may occur without the other. In particular, increase occurs without alteration when to a square there is adjoined an appropriately sized gnomon.³⁰ Then, according to Aristotle, the square, by which he means the original square, has increased, or grown, without any alteration, by which, as Boethius helpfully explains, he means alteration

²⁹ Abaelard, Dialectica, pp. 552:38–553:7. ³⁰ Aristotle, Categories, 15a29–31. See Boethius, In Categorias Aristotelis, PL 64, 296D.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

54  .  in shape. As I said above, Abaelard discusses this claim in both the Dialectica³¹ and the Logica “Ingredientibus.” The significant difference between the two treatments is that in the “Ingredientibus” he apparently believes that he has a simple solution to the problem of growth, one that he rather labors with in the Dialectica. In both works, in discussing the first two kinds of change, generation and corruption, Abaelard makes a crucial distinction and proposes a theory of creation and of the ontological distinction between natural kinds and artifacts. God creates the world as we know it ex nihilo but, according to Abaelard, he does so in stages. In his Hexameron, he attempts to provide an account of this process with a literal reading of Genesis, but for the purposes of the Dialectica and the “Ingredientibus” only the most fundamental distinction is important. In what Abaelard calls first creation, God brings into being both matter and form; since no matter preexists this combination, first creation is not generation.³² The informed matter produced first is a confused mixture of the four elements. In second creation God creates all of the various kinds of creatures by introducing their substantial forms into the preexisting primordial matter in the first act of generation. Every generation succeeding the first must likewise, according to Abaelard, be attributed to God alone and not in any way to the creatures produced and reproduced. Humans can organize bodies into artifacts but that is the limit of their efficient power since only God can generate a plant, a human being, or an individual of any other natural kind.³³ God operates to create, form, and complete the natural unities constituted as such by the substantial forms which he provides.³⁴ Thus, according to Abaelard, “a thing is generated when it takes on substantial being, as when some body is vivified it assumes the substance of an animate body, ³¹ The treatment of change in the Dialectica is located in Book II of the Topica where Abaelard discusses Themistian loci not already dealt with as of interest to logicians. He prefaces his account of the topical inferences associated with these loci with a general discussion of the notions involved and it is here that we find the four Aristotelian causes with generation discussed under efficient cause followed by a discussion of the six forms of change. ³² E.g. Abaelard, Dialectica, p. 415:18–24. Abaelard’s discussions of creation and generation and corruption are to be found in Dialectica, tract. 3, lib. 2, pp. 418–21, De substantia; Petrus Abaelardus, Logica “Ingredientibus,” Glossae super Praedicamenta, ed. B. Geyer, Peter Abaelards Philosophische Schriften, BGPM, Bd. 21, Hft. 2, 1921, pp. 297–9; Petrus Abaelardus, Expositio in Hexameron, ed. M. Romig (CCCM XV) (Turnhout: Brepols, 2004) (dated to early to mid-1130s by the editor). ³³ Abaelard, Dialectica, pp. 416:32–417:9. ³⁴ Abaelard, Dialectica, p. 419:13–15.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

55

either of an animal or of a human being.”³⁵ Generation, then, is a change in preexisting matter and that matter is ultimately corporeal, the product of first creation. The substantial forms introduced into this matter by God are introduced into preexisting bodies and so it is bodies alone, according to Abaelard, which are subject to generation. In a striking pair of parallel passages in the Dialectica and “Ingredientibus” he contrasts them in this respect with accidents and the souls of beasts. In the “Ingredientibus” he notes that unlike the change that occurs in the generation of a body there is nothing preexistent which changes when a new soul or a new individual accident comes into existence: . . . this soul was never a before it was a soul or this soul nor was this whiteness a quality or something else before it was whiteness or this whiteness, rather as soon as it existed somewhere it had all of its nature. Bodies are not like this. Certainly this body was a body before it was animate, or an animal, or a man. And note that there are many things which suffer corruption but which have no generation, such as whiteness itself or the souls of beasts which die with the flesh.³⁶

The ontological status of the souls of animals thus corresponds to that of accidents; they are incorporeal and cannot exist apart from the subjects with which they are associated. Abaelard goes on to observe again that only God is able to generate by informing corporeal substance with substantial form. So the production of a child is not an effect of the activity of the father, but rather of “nature, that is, God” who acts on the seed implanted in the mother “forming and vivifying it into a human being.”³⁷ ³⁵ Abaelard, Dialectica, p. 418:7–9. ³⁶ Abaelard, “Ingredientibus,” p. 298:12–27. See Dialectica, p. 419, where, however, Abaelard claims that in the case of accidents and the souls of animals no process of corruption is to be found. He makes the same point about incorporeal things in the Hexameron, p. 10:187–94: “Which does not happen in the case of incorporeal things. For the being of no quality passes into different species, so, that is, that what is now whiteness, ceases to be this, and becomes blackness or something else, or has parts which through some flow into or leave it; nor did what is whiteness ever exist in any way as something before it was whiteness, and when whiteness ceases to exist, it will not retain the status of any being.” ³⁷ Abaelard, Dialectica, p. 419:23–7: “Those creations which never lack prejacent matter are subject to generation and corruption, as when the supreme artificer vivifies the mud of the earth and from the inanimate forms it into a man and the substance of the earth already created turns into a human being by the adjunction of forms”; see “Ingredientibus,” p. 298:28–36. The account in the Hexameron is very different (p. 85, #378–9) “FROM THE MUD OF THE EARTH, that is from earth which is wet and as it were compressed, not dissolved, and thus infuses a soul

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

56  .  Moving on to change in quantity, in the Dialectica Abaelard characterizes it as a change in the quantity of the being, or substance, of a thing by increase or decrease. The problem is to locate just what it is that properly is said to be changed in this way, since it seems that there is no increase in anything by addition and so, as he says, “it seems that nothing grows.”³⁸ This follows immediately from the definition of an integral whole as identical with its constitutive parts. Neither that whole nor the object adjoined in forming a new whole changes in the addition, and the whole formed did not exist before the addition, so it too does not change. Abaelard’s solution to the puzzle in the Dialectica is to insist that augmentation must be characterized as a comparative relation between two distinct objects. Not between an object and itself since, characterized as an integral whole, its only quantitative relationship to itself is that of part-by-part identity. Augmentation thus relates a whole to a part and, as Abaelard defines it in the Dialectica, it is independent of time. One object is augmented with respect to another if it includes that other as a part, and so it is augmented at a given time with respect to all of its parts at that time. There is augmentation over time if an object which is part of a whole at a given time exists at an earlier time but not as part of that whole. Thus, according to Abaelard, the square which is augmented is the whole which results from the adjunction of the gnomon with respect to the square to which it is adjoined. Contrary to Aristotle, though his name is not mentioned here, the square which we start with is not augmented.³⁹ There is no difficulty in locating the relata in the augmentation which results from adjoining a gnomon to a square since the quadrilateral part which is spatially connected in the augmented whole is spatially into a body already created. From which clearly indicates that the human soul in virtue of its manner of creation is different from other souls. For in the creation of other living things the earth or water is said to have produced these with body and the soul together. By which it is indicated that their souls also come from the elements, as, as it were, a certain rareness or subtlety of them, on account of which subtlety those souls are also called spirits, as the wind is sometimes called a spirit, in comparison with the earth and water which are more gross and corpulent.” See also Hexameron, p. 86, # 382. ³⁸ Abaelard, Dialectica, p. 421:34–5: “ . . . no increase seems to come about in something by the adjunction of something, indeed nothing seems to be increased”; “Ingredientibus,” p. 299:19–20: “If someone were to say that to grow is to be made greater than before by the adjunction of something, nothing would seem to grow.” ³⁹ Abaelard, Dialectica, pp. 422:37–423:4.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

57

disconnected from the other parts prior to the augmentation. Abaelard points out, however, that mere relocation is not augmentation. There is no augmentation in a collection of two men if one of them moves to join the other in his house rather than remaining outside. At the end of his discussion of augmentation in the Dialectica, however, Abaelard does note that location may be relevant to increase: If . . . you add one to three, their number does not seem to increase except perhaps with respect to place, that is, in that there are more substances in this place than before when one was not yet added to the already here.⁴⁰

Giving its place at a time is one way of locating a whole without specifying its parts individually. It is this possibility which Abaelard appeals to in the “Ingredientibus” as an easy solution to the problem of growth. There is no longer any mention of the comparative criterion for increase in the “Ingredientibus” and the gnomon appears only in a brief commentary reporting that the figure, a square, is increased without alteration by the addition of a gnomon.⁴¹ Abaelard asks once more just what it is that grows but now says that he has an easy answer: something grows if by the addition of something else it “passes into” a composite which retains its nature and property. The rest of the discussion provides us with examples: water is increased by the addition of water, a heap of stones by the addition of a stone, and a heap by the addition of wood to the stones. What is important is to characterize the whole independently of the quantity. Abaelard concludes by claiming that “it is clear from what has been said that growth should be understood rather in the fact that something passes into a composite con-similar with it than in that there is a multiplication of parts.”⁴² Con-similarity will certainly guarantee sameness of kind but will it guarantee that one and the same individual persists through the change and so preserve our ordinary way of speaking about persistence? Abaelard’s solution to the problem of the identity of the amputee offers to help us here.

⁴⁰ Abaelard, Dialectica, p. 423:34–7. ⁴¹ Abaelard, “Ingredientibus,” p. 304:21–4. ⁴² “Ingredientibus,” p. 300. “Consimilitudo” is sameness of some sort, not similarity, see e.g. Dialectica, pp. 383:35–384:3.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

58  . 

d. The Identity of the Amputee The principal problem with an amputee for Abaelard is to explain how he retains the bipedality that is a differentia of human beings.⁴³ A second problem, however, discussed in both the “Ingredientibus” and the “Nostrorum” is one a version of which had already been formulated in antiquity by Chrysippus concerning Dion and Theon.⁴⁴ What exists after the amputation of a foot is a man, but did this man first come into being as a result of the amputation, or did he exist before as part of the earlier whole man? Neither alternative is palatable. Against the first, human operations have no power to generate substantial forms; against the second, there would then be two men, and many more, in the whole body before the amputation, which no one will accept.⁴⁵ Abaelard once again argues that a manicurist should not be charged with murder. Murder requires the removal of the soul, just as the generation of a human being requires the production of a soul,⁴⁶ but A body vegetated by the same soul is always said to be the same man, granted that the quantity of the being is less in respect of parts when he is a child, and greater when he is a youth, and granted that the body that existed before and that which exists now is not the same with respect to the quantity of being. They are, nevertheless, judged to be the same on account of the same effect of the soul. Otherwise when his nails are cut we would have to say that there is the substance of another man. Granted, however, that this is true with respect to the quantity of

⁴³ See Martin, “Abaelard on Modality, Some Possibilities and Some Puzzles,” in T. Bucheim, C. H. Kneepkens, and K. Lorenz (eds.), Potentialität und Possibilität: Modalaussagen in der Geschichte der Metaphysik (Stuttgart: Frommann-Holzboog, 2001), 97–124. ⁴⁴ Chrysippus defines Theon as the part of the man Dion that coincides with him apart from one of his feet. He then asks whether it is Dion or Theon who exists after that foot is amputated and answers that it is Dion. Abaelard knew nothing of Chrysippus’ version of the problem, which is reported only by Philo of Alexandria writing in Greek in his De aeternitate mundi, first published in a Latin translation in 1554. For the argument see Long and Sedley, The Hellenistic Philosophers, n. 28P, and David Sedley, “The Stoic Criterion of Identity,” Phronesis 27 (1983), 255–70. Dion and Theon appear in many discussions of the versions of the same question raised in contemporary treatments of mereology. See e.g., most recently, Alex Moran, “The Paradox of Decrease and Dependent Parts,” Ratio 31 (2018), 273–84. ⁴⁵ Abaelard, “Ingredientibus,” p. 105:14–21. ⁴⁶ Abaelard, “Ingredientibus,” p. 105:25–7.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

59

being, with respect to the effect of the soul, there is not said to be another person.⁴⁷

So identity is guaranteed by continued vegetation, but vegetating, according to Abaelard, is what every soul does, plant, animal, and human.⁴⁸ The substantial forms of living things include the powers of their souls and it is a fundamental principle of Abaelard’s anti-realist ontology that substantial forms are individuals. It is substantial forms that individuate the composites of which they are the substantial forms. To be Socrates, or Brownie the donkey, is thus simply to be this very human or donkey, quite independently of whatever accidental features Socrates or Brownie might possess.⁴⁹ The identity and persistence of every animate being is guaranteed by its soul and it exists just as long as it remains vegetated by that soul, where sameness of soul is to be judged by sameness of effect. Arlig has pointed out that, in the Hexameron, Abaelard claims that the human soul is different from the souls of all other living creatures in that it is not formed from primordial matter and so is distinct from the body in a way that animal souls are not.⁵⁰ Such a difference is not noticed in his philosophical works, however, and as we have just seen he insists rather that the souls of brute animals are incorporeal. Indeed, in the Hexameron, Abaelard maintains that the production of the vegetative and sensitive souls from the mixing of elements produces something more subtle than their gross and corpulent matter, a spirit, and so perhaps the emergence of the incorporeal from the corporeal. On the other hand, making no distinction between human and non-human souls, Abaelard also argues in the Dialectica that the soul is essentially

⁴⁷ Abaelard, “Ingredientibus,” p. 105:27–30. Petrus Abaelardus, Logica “Nostrorum Petitioni Sociorum,” Glossae super Praedicamenta, ed. B. Geyer, Peter Abaelards Philosophische Schriften (BGPM Bd. 21, Hft. 4, 1933), p. 579:22–30: “We will not, therefore, say that the same whole remains but rather the same man, if with ‘the same’ there is a reference to the soul. For we say that a man is part of a man and Socrates of Socrates; what is above his knees, for example, is said to be a man and Socrates. Nor should someone who cuts off say that he makes a man, which is the work of god alone and not of man. And although Socrates is part of Socrates, there are nevertheless not many men in Socrates, since in him there are not found vegetated by many souls. Nor is ‘Socrates’ predicated of many, since are not numerically distinct.” ⁴⁸ Abaelard, Dialectica, p. 555:20–5. ⁴⁹ E.g., “Ingredientibus,” p. 64:20–4. See Martin, “Abaelard on Grounding in Ontology and Logic” (forthcoming). ⁵⁰ See footnote 38.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

60  .  localized at some place in a living body but exercises its vegetative and vivificatory powers everywhere in that body.⁵¹ This last point agrees nicely with his claim here; amputated feet are no longer vegetated but the body that remains vegetated is the body of the same man. Nothing that he says indicates that Abaelard does not suppose that this not also true of animals and plants. Rather, and independently of the nature of the soul, mimicking Abaelard’s argument, if we humans had the power, for example, to destroy a tree by trimming its branches, we would have the power to generate a new tree. But only God can make a tree! So long, then, as a principal part of a human, an animal, or a plant is not destroyed, one and the same human, animal, and plant remains. The only way to preserve something of the same kind is by preserving the same individual.⁵² Abaelard draws the conclusion that it seems that nothing grows⁵³ from the assumption that a substance is to be identified with its quantitative parts and shows that it is false by explaining the persistence of individual substances over time. Quantity is an accident and the suggestion that the cutting of Socrates’ nails would produce the substance of another human would thus require that individuation be by accidents.⁵⁴ But, as a corollary of the principle that it is substantial forms and only substantial forms that individuate substances, Abaelard holds that accidents are individuated and depend entirely for their existence upon the substances that support them. He thus does not accept and could not accept that merely quantitative changes are in fact generations and corruptions.⁵⁵ ⁵¹ Abaelard, Dialectica, p. 70:20–7. ⁵² With respect to living things Abaelard thus anticipates Peter Van Inwagen’s proposal that it is the life of a living thing which unifies the parts into a whole and preserves the identity of the whole over time. Van Inwagen notes that “Locke was probably using the word in this way when he wrote that the identity of a man consists ‘in nothing but a participation of the same continued life by constantly fleeting particles of matter.’ The Greeks sometimes meant by psuche what I mean by ‘life,’ though I do not suppose, as some Greeks did, that a psuche can detach itself from the material whose activity had constituted it, after that material has become quiescent” (Material Beings [Ithaca, NY: Cornell University Press, 1990], p. 83). ⁵³ See footnote 39. ⁵⁴ Unless, that is, we suppose that God produces many new substantial forms every time that Socrates visits his manicurist or Brownie’s coat is brushed. ⁵⁵ Abaelard’s reply to his version of the growing argument thus corresponds to that given by Chrysippus to the argument formulated by the Academic Skeptics in antiquity. He requires that a distinction be made between the corporeal quantity of a substance and the substance as individuated by its substantial form just as Chrysippus made a distinction between substrate and proper quality. It seems to me that the same strategy is to be found in the distinction between essential and personal predication, which I believe can be invoked to explain the acceptance by

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

61

Abaelard allows, as we have seen, that although humans cannot generate living things, they can fabricate new artifacts. So there is no parallel argument to be had against us accidentally destroying our houses by knocking the smallest stones from their walls. It seems, however, that Abaelard did perhaps also develop an account of principal parts to protect us from this loss of real estate. It is reported in the anonymous Introductiones Montanae Maiores, to which I now turn.

3. The Testimony of the Introductiones Montanae Maiores The Introductiones contains a very extensive discussion of the loci from whole and part, stating a number of rules and considering many objections. Principal parts are first mentioned in discussing the locus from a universal whole between subjects where we are offered the maximal proposition “whatever is predicated of a universal whole in some way is predicated of a part of that whole.” A sophism very familiar from later medieval logic provides an objection. The rule apparently allows us to infer from the truth of “man is more worthy than all other creatures” that Socrates is more worthy than all other creatures.⁵⁶ The solution proposed by the Introductiones has two parts. First, we are told that what appears to be a complex predicate, the phrase “more worthy than all other creatures,” is not in fact a single predicate at all but rather a linguistic device which, taken with the subject, abbreviates a conjunction of two distinct propositions: “man is a worthy creature” and “no creature other than man is as worthy as man.” Our rule allows the substitution of the name of an individual man into the first conjunct but not into the second and so is not invalidated by the counterexample. According to the Introductiones, the proposition “a wall is a principal part of a house” must be handled in the same way. The rule seems to fail here because it allows us to infer that “this wall is a principal part of a

the Nominales of the principle that nothing grows. For Chrysippus’ solution see Sedley, “The Stoic Criterion of Identity”; for the Nominales, see Martin, “The Logic of Growth.” ⁵⁶ “Introductiones,” p. 120:2–11.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

62  .  house,” which according to the Introductiones is false of any given wall. Exposition reveals multiplicity and once more two distinct claims are being made: “a wall is part of a house” and “there can be no house without such a wall.” Both propositions are true of walls and houses in general where there is a “relatio simplex” between the two occurrences of “wall,” that is to say the second refers to the kind of thing picked out by the first, walls, rather than to an individual.⁵⁷ If we substitute an expression denoting a particular wall into both of them, however, the first is true and the second false. The first is an instance of the application of the rule but the second is not.⁵⁸ The problem of a house and its parts is most fully dealt with by the Introductiones in its consideration of the rule which underpinned the destructive argument rejected by Abaelard: “if a part is destroyed, then the whole is destroyed.” We are first told that the rule must be restricted only to principal parts, where a wall is a principal part of a house but no particular wall is. The exposition of the claim that “a wall is a principal part of a house” just mentioned is repeated with the conclusion that It does not follow that this wall or that wall is a principal part of a house. For then the sense would be that this wall constitutes the house and without this wall that house cannot exist . . . . And this theory was discovered by Master P., and may be conveniently maintained, granted that many objections may be made to it.⁵⁹

Master P. is certainly Abaelard, to whom the many references in the Introductiones are almost always in a past tense, usually the imperfect. Unfortunately, it is not at all clear how much of the discussion of parts and wholes here the author wishes to credit to Abaelard, but that he should be credited with anything at all is rather remarkable in a text whose main interest in him seems to be as an opponent thoroughly bested by Alberic. It may be that only the grammatical point is due to Abaelard and the theory of principal parts that follows in the Introductiones is due to the Montani. ⁵⁷ On relatio simplex see C. H. Kneepkens, “ ‘Mulier Quae Damnavit, Salvavit,’ A Note on the Early Development of the Relatio simplex,” Vivarium 14 (1976), 1–22. Kneepkens locates the theory of relatio simplex, called by Abaelard relatio indifferens, in his discussion of the sentence “Mulier qui damnavit salvavit,” and does not cite any earlier discussion. ⁵⁸ Introductiones, p. 120:11–19. ⁵⁹ Introductiones, pp. 132:25–133:5.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

63

It would, however, sit nicely with the position that I have attributed to Abaelard on the identity of living things if we could also associate with him at least the general solution to the destructive argument given in the Introductiones. According to this solution to be a principal part is to play a functional and structural role in the constitution of a house. So long as that role is played by something, the house continues to exist. The Introductiones allows, indeed, that we may accept the claim that “this wall is a principal part of a house,” but only in the sense that a principal part is such that if neither it nor any other in place of it played that role in the constitution of the house, the house would not exist at all: Whence although this wall which is removed ceases to exist and is destroyed entirely, nevertheless the principal part is not destroyed, although that which was the principal part is annihilated. To destroy a principal part is to destroy the leading structural role which that part plays in the composition of the house.⁶⁰

As I said, it would be nice if this were Abaelard’s theory of principal parts and it seems compatible with the constructive account which he proposed in the Dialectica. To be a principal part would thus be to play a role in the constitution of a whole which can be played by different objects but which are such that as soon as all of the relevant roles are filled the whole exists, and it continues to exist despite changes in the objects filling those roles.⁶¹ The Introductiones goes on to offer a direct refutation of the destructive argument that Abaelard employed as a reductio of the destructive account of principal parts.⁶² The argument and the language in which it is formulated is interestingly close to the destructive argument as reported in what Arlig calls the Anonymous Fragment, from Orléans 266,⁶³ published as the first section of De generibus et speciebus.⁶⁴ The author of the Introductiones maintains that the destructive argument fails as an ⁶⁰ Introductiones, p. 133:28–32. ⁶¹ For a contemporary outline of a mereology which makes just this distinction between roles and their occupants, see Karen Bennett, “Having a Part Twice Over,” Australasian Journal of Philosophy 91 (2013), 83–103. ⁶² Introductiones, pp. 134:19–135:4. ⁶³ Arlig, “Some Twelfth Century Reflections,” p. 97. ⁶⁴ Peter King, “Pseudo-Joscelin: Treatise on Genera and Species,” Oxford Studies in Medieval Philosophy 2 (2014), 104–210.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

64  .  application of transitivity because the annihilation of a part does not imply the annihilation of the whole but rather only its destruction in the sense that it no longer exists as a whole with respect to those parts. We thus cannot infer from the annihilation of the smallest pebble in a wall the annihilation of the house, but rather only that it no longer exists as a whole with respect to that pebble. This, for the Introductiones, means that the whole remains the same (idem), although it is altered (alteratum).⁶⁵ This is perfectly sensible, and surely what we would ordinarily say about the effect of even relatively large losses and gains to the structure of our house. In Abaelard’s works, then, and in the Introductiones Montanae Maiores, we find, I believe, sustained and sophisticated attempts to preserve our ordinary intuitions about living things and artifacts from philosophical arguments which might threaten our belief in their identity over time. Whether they are entirely successful is, of course, another matter.⁶⁶ University of Auckland

Bibliography Arlig, Andrew. “Abelard’s Assault on Everyday Objects,” American Catholic Philosophical Quarterly 81 (2007), 209–27. Arlig, Andrew. “Some Twelfth-Century Reflections on Mereological Essentialism,” Oxford Studies in Medieval Philosophy 1 (2013), 83–112. Armstrong, David M. A World of States of Affairs (Cambridge: Cambridge University Press, 1997). Bennett, Karen. “Having a Part Twice Over,” Australasian Journal of Philosophy 91 (2013), 83–103. Boethius. In Categorias Aristotelis libri quatuor, in J. P. Migne (ed.), Patrologiae Cursus Completus, Series Latina (Paris, 1847), vol. 64. Bos, Egbert P. and Joke Spruyt (eds.). Anonymi Introductiones Montanae Maiores (Louvain la Neuve: Peeters, 2017).

⁶⁵ Introductiones, p. 135:5–16. ⁶⁶ Work on this study was supported by an award from the Marsden Fund Council administered by the Royal Society of New Zealand.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

65

Cartwright, Richard. “Scattered Objects,” in K. Lehrer (ed.), Analysis and Metaphysics (Dordrecht: Reidel, 1975), 153–71. Henry, D. P. Medieval Mereology (Amsterdam: John Benjamins, 1991). Iohannes Sarisberiensis. Metalogicon, ed. J. B. Hall and K. S. B. Keats-Rohan (CCCM 98) (Turnhout: Brepols, 1991). Iwakuma, Yukio. “Alberic of Paris on Mont Ste Geneviève against Peter Abelard,” in J. L Fink, H. Hanesen, and A. M. Mora-Márqez (eds.), Language and Logic in the Middle Ages (Leiden: Brill, 2013), 27–47. King, Peter. “Pseudo-Joscelin: Treatise on Genera and Species,” Oxford Studies in Medieval Philosophy 2 (2014), 104–210. Kneepkens, C. H. “ ‘Mulier Quae Damnavit, Salvavit,’ A Note on the Early Development of the Relatio simplex,” Vivarium 14 (1976), 1–22. Long, A. A. and D. N. Sedley. The Hellenistic Philosophers (Cambridge: Cambridge University Press, 1987). Markosian, Ned. “Restricted Composition,” in T. Sider, J. Hawthorne, and D. Zimmerman (eds.), Contemporary Debates in Metaphysics (Oxford: Blackwell, 2008), 341–63. Martin, Christopher J. “The Logic of Growth: Twelfth-Century Nominalists and the Development of Theories of the Incarnation,” Medieval Philosophy and Theology 7 (1998), 1–15. Martin, Christopher J. “Abaelard on Modality, Some Possibilities and Some Puzzles,” in T. Bucheim, C. H. Kneepkens, and K. Lorenz (eds.), Potentialität und Possibilität: Modalaussagen in der Geschichte der Metaphysik (Stuttgart: Frommann-Holzboog, 2001), 97–124. Martin, Christopher J. “Logic,” in J. Brower and K. Guilfoy (eds.), The Cambridge Companion to Abelard (Cambridge: Cambridge University Press, 2004), 158–99. Martin, Christopher J. “Abaelard on Grounding in Ontology and Logic” (forthcoming). Moran, Alex. “The Paradox of Decrease and Dependent Parts,” Ratio 31 (2018), 273–84. Petrus Abaelardus. Logica “Ingredientibus,” in B. Geyer (ed.), Peter Abaelards Philosophische Schriften (BGPM, Bd. 21, Hft. 2) (Münster: Aschendorff, 1921). Petrus Abaelardus. Dialectica, ed. L. M. De Rijk (Van Gorcum: Assen, 1970). Petrus Abaelardus. Expositio in Hexameron, ed. M. Romig (CCCM XV) (Turnhout: Brepols, 2004).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

66  .  Sedley, David N. “The Stoic Criterion of Identity,” Phronesis 27 (1983), 255–70. Simons, Peter. Parts: A Study in Ontology (Oxford: Clarendon Press, 1987). Van Inwagen, Peter. Material Beings (Ithaca, NY: Cornell University Press, 1990). Varzi, Achille C. “Mereology,” in E. Zalta (ed.), Stanford Encyclopedia of Philosophy .

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

Accidental Forms as Metaphysical Parts of Material Substances in Aquinas’s Ontology Jeremy W. Skrzypek

1. Introduction Following in the hylomorphic tradition of Aristotle, Thomas Aquinas holds that all material substances are composed of matter and form.¹ Like Aristotle, Aquinas also recognizes two different types of forms that material substances can be said to possess: substantial forms and accidental forms. Of which form or forms, then, are material substances composed? It seems clear that Aquinas means to include at least the substantial form of a material substance within its hylomorphic constitution, but what about the accidental forms, the various non-essential attributes of a material substance? Are they also meant to be included in the composite whole that is a material substance? In this paper, I explore two competing models of Aquinas’s ontology of material substances, which diverge on precisely this issue. According to what I will refer to as the “Standard Model,” Aquinas’s view is that a material substance is composed of prime matter and substantial form.

¹ That Aquinas holds that material substances are composed of matter and form, and that matter and form are thereby parts of material substances in some sense is evidenced in several passages, among which are the following: “Matter and form are said to be intrinsic to a thing, in that they are parts constituting that thing” (Aquinas, DPN, Ch. 3); “matter and form are said to be related to one another . . . they are also said to be related to the composite as parts to a whole and as that which is simple to that which is composite” (DPN, Ch. 4). All references to the works of Aquinas are to the Latin versions of the texts available at . All English translations are my own, though I have made use of the standard English translations of these texts in preparing them.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

68  .  On the Standard Model, the accidental forms of a material substance, its various non-essential attributes, do not enter into its composition; they are in no way parts of it. According to what I will refer to as the “Expanded Model,” Aquinas’s view is that a material substance is composed of prime matter, substantial form, and all of its accidental forms. On the Expanded Model, the accidental forms of a material substance do enter into its composition; each of its accidents is included among its metaphysical parts.² The Standard Model of Aquinas’s ontology of material substances is the Standard Model for a reason. It is not difficult to find passages that seem to support such a reading. In Chapter Two of his De ente et essentia, for example, Aquinas explains that “a human being is said to be composed of soul and body, as a certain third thing constituted out of two things, which is neither of them. For a human being is neither soul nor body.”³ And in Book I of his Compendium theologiae, Aquinas reiterates a similar point: “in other human beings, the union of soul and body constitutes an hypostasis and a suppositum, since there is nothing else over and above these two components.”⁴ Elsewhere, Aquinas also seems to deny that an accidental form could ever be a part of a material substance,⁵ which accords well with what Aristotle himself says in his Categories.⁶ Beyond these (and various other) passages, which seem to straightforwardly support the Standard Model, there are also at least two big-picture reasons for interpreting Aquinas in this way. As will be

² The term “metaphysical part” is meant to distinguish form and matter from any of the “physical parts” that a human person can be said to have, such as its functional or integral parts (hands, eyes, blood, bones) and its simple or elemental parts (water, fire, earth). For more on the various sorts of parts that a human person can be said to have in Aquinas’s ontology, see, for example: Eleonore Stump, Aquinas (New York: Routledge, 2003), 35, 42; Christopher M. Brown, Aquinas and the Ship of Theseus: Solving Puzzles about Material Objects (New York: Continuum, 2005), 53, 92–8; Robert Pasnau, Metaphysical Themes 1274–1671 (Oxford: Clarendon Press, 2011), 7–11. By referring to them as “parts” here, I do not mean to imply that matter and form are themselves physical objects of some sort. If the reader objects to my use of the term “parts” to describe matter and form, then she may substitute the terms “component” or “constituent” throughout. I do not think that this has any effect on any of my arguments either way. ³ Aquinas, DEE, Ch. 2, emphasis added. ⁴ Aquinas, CT, B. 1, N. 211, emphasis added. ⁵ See, for example, Aquinas, In Met., B. 7, Ch. 13, L. 13, N. 1579–80. ⁶ Aristotle, Categories, tr. J. L. Ackrill, in J. Barnes (ed.), The Complete Works of Aristotle, Vol. 1 (Princeton, NJ: Princeton University Press, 1984), 3 (Ch. 2, 1a20–5).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

69

explored below, a realist interpretation of the general account of accidental and substantial change that Aquinas inherits from Aristotle, to which Aquinas commits himself in several of his works, seems to entail the Standard Model. Moreover, it seems that the Standard Model is the only model that preserves the sort of robust, substantial unity that Aquinas attributes to material substances. Despite all of this evidence that can be given in favor of the Standard Model, I think that there are several other passages in Aquinas’s texts that show that the Expanded Model more accurately reflects his view of material substances. I also think that there are at least a few other big-picture reasons for interpreting Aquinas in this way. The purpose of this paper, then, is to offer a case for the Expanded Model based on these observations. In the next section, I first outline the main claims of each of the two competing models for Aquinas’s ontology of material substances, making note of their most significant points of disagreement. In section 3, I present what I take to be the main argument in favor of the Standard Model: that it is required by Aquinas’s Aristotelian account of substantial and accidental change. In section 4, I consider a second major argument in support of the Standard Model: that including accidental forms within the hylomorphic constitution of material substances would undermine their substantial unity. In sections 5 and 6, I offer two arguments in favor of the Expanded Model. I argue that, given the way in which he argues for God’s simplicity in question three of the Prima pars, and the way in which he consistently describes the difference between an essence and a suppositum, or individual substance, throughout his works, there is good reason to believe that Aquinas thinks that the accidental forms of a material substance are included among its metaphysical parts. In sections 7, 8, and 9, I attempt to motivate the Expanded Model further by offering three sets of replies to arguments for the Standard Model. In section 7, I offer a principled way of explaining away many of the passages that seem to support the Standard Model. In section 8, I reply to the main argument in favor of the Standard Model, arguing that a proponent of the Expanded Model can accommodate Aquinas’s account of substantial and accidental change without having to surrender any of the main commitments of the view. Finally, in section 9, I explain how material substances can include their accidental forms

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

70  .  within their hylomorphic constitution while still maintaining their substantial unity and without collapsing into accidental unities.

2. Two Competing Models The Standard Model of Aquinas’s ontology of material substances is built on a distinction between two sorts of composite wholes: material substances and “accidental unities” (sometimes referred to as “accidental beings” or “accidental compounds”). The difference between these two sorts of composite wholes is as follows. Material substances are composed of prime matter and substantial form. Material substances include things such as elements, minerals, plants, non-human animals, and human beings. You and I are material substances. And so each of us is composed of prime matter and some particular substantial form (in our case, a rational soul). Accidental unities, on the other hand, are what we might call “second-order” wholes, composed of material substances (which, as we have just seen, are themselves composed of prime matter and substantial form) and accidental forms. When an accidental form comes to “inhere” in a material substance, that is, when a material substance comes to possess a certain non-essential attribute, this gives rise to an accidental unity. And for every accidental form possessed by a material substance, there exists an accidental unity that is composed of that accidental form and the material substance in which it inheres.⁷ Accidental unities, then, include so-called “kooky objects,”⁸ such as ⁷ That Aquinas is committed to the existence of such entities is supported by several texts, among which is the following from Chapter 6 of his DEE, “just as a substantial being results from form and matter when they are composed, so too an accidental being results from accident and subject when the accident comes to the subject.” For a complete list of references in Aquinas’s texts to “accidental beings” or “accidental unities,” see Brown, Aquinas and the Ship of Theseus, 64, fn27. In section 3 below I explore in more detail how Aquinas’s account of substantial and accidental change might be seen as committing him to the existence of accidental unities. ⁸ The phrase “kooky objects” comes from Gareth Matthews, “Accidental Unities,” in M. Schofield and M. Nussbaum (eds.), Language and Logos: Studies in Ancient Greek Philosophy (Cambridge: Cambridge University Press, 1982), 223–40, and it refers to those accidental unities in Aristotle’s ontology that are composed of material substances and accidental forms. For similar accounts of “kooky objects” in Aristotle’s ontology, see, for example: Frank Lewis, “Accidental Sameness in Aristotle,” Philosophical Studies 42 (1982), 1–36; S. Marc Cohen, “Kooky Objects Revisited: Aristotle’s Ontology,” Metaphilosophy 39 (2008), 3–19; S. Marc Cohen, “Accidental Beings in Aristotle’s Ontology,” in G. Anagnostopoulos and

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

71

white-Socrates (the accidental unity composed of Socrates and his pallor) and seated-Socrates (the accidental unity composed of Socrates and his seatedness), as well as single-substance artifacts, such as bronze statues (accidental unities composed of bronze and some particular shape) and thresholds (accidental unities composed of wood and some particular location). Now, you and I are material substances. And so, once again, according to the Standard Model, each of us is composed of prime matter and substantial form. But each of us also has several non-essential attributes, such as our particular height, weight, and various qualities that we possess, as well as any and all of our particular thoughts and actions. And so each of us is also a part of several different accidental unities, one for every non-essential attribute that we possess. The key point to emphasize here is that, on the Standard Model, a material substance does not possess its non-essential attributes as parts. It possesses each of its non-essential attributes via inherence, and it composes, together with those attributes, various accidental unities. One recent proponent of the Standard Model is Jeffrey Brower. In his 2014 book, Aquinas’s Ontology of the Material World, Brower characterizes the difference between a material substance and an accidental unity in Aquinas’s ontology in the following way: Aquinas thinks that the corporeal world is completely analyzable in terms of two different types of hylomorphic compound—what he calls material substances and accidental unities, respectively. These two types of compound are distinguished both by their matter and by their form—that is to say, both by the type of being that serves as their substratum and by the type that inheres in their matter . . . Aquinas thinks of all material substances as composed of prime matter

F. D. Miller, Jr. (eds.), Reason and Analysis in Ancient Greek Philosophy (Dordrecht: Springer, 2013), 231–42; Michael J. Loux, “Aristotle’s Constituent Ontology,” in D. W. Zimmerman and K. Bennett (eds.), Oxford Studies in Metaphysics 2 (Oxford: Oxford University Press, 2006), 207–50; Michael J. Loux, “Aristotle’s Hylomorphism,” in L. Novak and D. D. Novotny (eds.), NeoAristotelian Perspectives in Metaphysics (New York: Routledge, 2014), 138–63. For interpretations of Aristotle that reject the existence of kooky objects, see: Theodore Scaltsas, Substances and Universals in Aristotle’s Metaphysics (Ithaca, NY: Cornell University Press, 1994), 97–113, 150–4; Christopher Shields, Order in Multiplicity: Homonymy in the Philosophy of Aristotle (Oxford: Clarendon Press, 1999), 155–75.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

72  .  and substantial form, whereas he thinks of all accidental unities as composed of substances and accidental forms.⁹

Other proponents of the Standard Model include Christopher Brown,¹⁰ David Oderberg,¹¹ Robert Pasnau,¹² and Ross Inman.¹³ The Expanded Model of Aquinas’s ontology of material substances is built on a distinction between two sorts of metaphysical parts that material substances can be said to possess: essential parts and accidental parts. The difference between these two sorts of parts is as follows. Essential parts are those metaphysical parts of a material substance that comprise its essence or nature. The essential parts of a material substance include its matter (further specified in some way) and its substantial form.¹⁴ Typically, when a substance loses one or more of its essential parts that substance ceases to exist.¹⁵ Accidental parts, on the other hand, are those metaphysical parts of a material substance that lie outside of its essence. The accidental parts of a substance include all of its accidental forms. Material substances can, and frequently do, lose and gain such parts over time.¹⁶ Now, you and I are material substances. And so, according to the Expanded Model, each of us has both essential parts (matter and substantial form), and accidental parts (an accidental form

⁹ Jeffrey E. Brower, Aquinas’s Ontology of the Material World: Change, Hylomorphism, and Material Objects (Oxford: Oxford University Press, 2014), 9. See also, Jeffrey E. Brower, “Matter, Form, and Individuation,” in B. Davies and E. Stump (eds.), The Oxford Handbook of Aquinas (Oxford: Oxford University Press, 2012), 85–103; Jeffrey E. Brower, “Aquinas on the Individuation of Substances,” in R. Pasnau (ed.), Oxford Studies in Medieval Philosophy 5 (Oxford: Oxford University Press, 2017), 127–8. ¹⁰ Brown, Aquinas and the Ship of Theseus, 53, 64. ¹¹ David S. Oderberg, Real Essentialism (New York: Routledge, 2007), 167–70. ¹² Robert Pasnau, “Form and Matter,” in R. Pasnau (ed.), The Cambridge History of Medieval Philosophy (Cambridge: Cambridge University Press, 2010), 642; Pasnau, Metaphysical Themes, 101–2; Robert Pasnau, “Mind and Hylomorphism,” in J. Marenbon (ed.), The Oxford Handbook of Medieval Philosophy (Oxford: Oxford University Press, 2012), 501. ¹³ Ross Inman, “Neo-Aristotelian Plenitude,” Philosophical Studies 168 (2014), 588–96. ¹⁴ There are complications here concerning which sort of matter belongs to the essence of a material substance and how this matter relates to prime matter, but I will set aside those complications for the moment (I revisit them in section 6 below). Let the term “matter” here stand in for whatever sort of matter it is that belongs to the essence of a material substance. ¹⁵ Though on some interpretations of Aquinas’s account of the afterlife, human persons can and do survive the loss of their matter (see, Brower, Aquinas’s Ontology, Chapter 13 for an excellent overview of the relevant debate). ¹⁶ “Proper accidents,” those accidents that “flow” necessarily from a substance’s substantial form would, however, be an exception to this generalization. For Aquinas’s account of proper accidents, see, for example: Aquinas, DEE, Ch. 6; Aquinas DPN, Ch. 2; Aquinas, ST, I, Q. 3, A. 6, Co.; Aquinas, ST, I, Q. 77, A. 1, Ad 5; Aquinas, QDA, Q. 1, A. 12, Ad 7.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

73

corresponding to each of our non-essential attributes). The key point to emphasize here is that, on the Expanded Model, a material substance’s prime matter and substantial form do not exhaust its metaphysical parts. An individual material substance, or suppositum,¹⁷ includes among its metaphysical parts its substantial form, its prime matter and all of its accidental forms. One recent proponent of the Expanded Model is Eleonore Stump. In her 2003 book, Aquinas, she characterizes individual material substances, or “supposits” as follows: any thing which has a substantial form necessarily also has accidents, even though it is not necessary that it have one accident rather than another. So a substantial form is not the only metaphysical constituent of a thing; any thing will also have accidental forms as metaphysical constituents. In addition, for material substances, the matter that makes the substantial form of a material supposit a particular is also a constituent of the supposit. So any supposit has more metaphysical constituents than just a substantial form. Insofar as all these constituents compose the supposit, the supposit is not identical to any subset of them.¹⁸

Other proponents of the Expanded Model include Christopher Hughes,¹⁹ Richard Cross,²⁰ J. L. A. West,²¹ Michael Gorman,²² and John Wippel.²³ ¹⁷ The term “suppositum” or “supposit” is one that Aquinas himself uses to refer to individual substances. The reader may notice that in some of the passages featured below Aquinas also frequently uses the term “hypostasis” to refer to such entities. While the terms “individual substance,” “suppositum” and “hypostasis” have slightly different meanings for Aquinas (see Aquinas, ST, I, Q. 29, A. 2, Co. for more on this), these differences are not relevant to the present discussion, and so for our present purposes, these three terms can be seen as interchangeable. ¹⁸ Stump, Aquinas, 50; see also Stump, Aquinas, 44–5, 56, and 112–13. ¹⁹ Christopher Hughes, Aquinas on Being, Goodness, and God (New York: Routledge, 2015), 68. ²⁰ Richard Cross, “Aquinas on Nature, Hypostasis, and the Metaphysics of the Incarnation,” The Thomist 60 (1996), 175–6. ²¹ J. L. A. West, “The Real Distinction Between Supposit and Nature,” in P. Kwasniewski (ed.), Wisdom’s Apprentice: Thomistic Essays in Honor of Lawrence Dewan, O.P. (Washington, DC: Catholic University of America Press, 2007), 92, 97. ²² Michael Gorman, “Uses of the Person–Nature Distinction in Aquinas’s Christology,” Recherches de théologie et philosophie médiévales 67 (2000), 59, 66; Michael Gorman, Aquinas on the Metaphysics of the Hypostatic Union (Cambridge: Cambridge University Press, 2017), 31–3. ²³ John F. Wippel, The Metaphysical Thought of Thomas Aquinas (Washington, DC: Catholic University of America Press, 2000), 241, 243, 244, 245–6, 247, 248. At pages 101–2 of his Metaphysical Themes, Pasnau also briefly traces a distinction in the works of some

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

74  . 

3. Setting the Standard I: Substantial and Accidental Change One of the main reasons for thinking that the Standard Model more accurately reflects Aquinas’s ontology of material substances is that the account of substantial and accidental change that he inherits from Aristotle seems to require it. Aquinas most clearly articulates his account of substantial and accidental change in two places: Chapter One of his De principiis naturae and in Lecture 12 of his commentary on Chapter 7 of Aristotle’s Physics.²⁴ In both of these texts, Aquinas explains, following Aristotle, that there are at least three things required for any real change. First, there must be something that underlies or undergoes the change, something that persists through the change. This is the subject of the change. Second, there must be some characteristic or attribute that the subject possesses prior to the change, which does not survive the change. Sometimes this is simply the lack of the characteristic or attribute to be gained; other times it is a characteristic or attribute that is contrary to (and thus incompatible with) the characteristic or attribute to be gained. Third, there must be some characteristic or attribute that the subject possesses after the change, the characteristic or attribute that it has gained in the change. These are the three key aspects or elements of any real change, according to both Aristotle and Aquinas. But there appear also to be two further aspects or elements that Aristotle and Aquinas wish to include. According to Aristotle and Aquinas, in any real change, there must be something that is corrupted (something that ceases to be) and something that is generated (something that comes to be).²⁵ And in any real change, that which is corrupted and that which medieval philosophers between “thin metaphysical substances” (composites of prime matter and substantial form) and “thick concrete substances” (composites of prime matter, substantial form, and various accidental forms), which seems to correspond to my distinction here between essences and individual substances. ²⁴ Aquinas, DPN, Ch. 1; Aquinas, In Phys., B. I, Ch. 7, L. 12. ²⁵ Commenting on Chapter 7 of Aristotle’s Physics, Aquinas remarks that “He [Aristotle] says, therefore, first that, with these suppositions having been put forward, if one wishes to consider [coming to be] in all the things which come to be naturally, he will accept this: that there must always be some subject to which the coming to be is attributed, and that that subject, although one in number and subject, is nevertheless not the same in species or account. For when it is attributed to a human being that he becomes musical, the human being is indeed one in subject, but two in account. For human being and the non-musical are not the same

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

75

is generated must be understood as composite entities. That which is corrupted in any real change is the composite whole that includes the subject and the characteristic or attribute that is lost in the change. That which is generated in any real change is the composite whole that includes the subject and the characteristic or attribute that is gained in the change. Now, there are two kinds of change according to Aristotle and Aquinas: substantial and accidental. In a substantial change, the subject is prime matter, matter with no characteristics or attributes of its own.²⁶ The characteristic or attribute that inheres in prime matter prior to substantial change, which does not survive the change, is some particular substantial form. The characteristic or attribute that inheres in prime matter after a substantial change, which it has gained in the change, is some other substantial form. That which is corrupted in a substantial change is the composite whole that includes prime matter and the first substantial form. That which is generated in a substantial change is the composite whole that includes prime matter and the second substantial form. The example that Aquinas gives in the De principiis of this sort of change is the generation of a human being.²⁷ He explains that in the generation of a human being sperm and menstrual blood are the composites that are corrupted, a human being is the composite that is generated, and prime matter is the subject that remains throughout. In an accidental change, the subject is itself composite. The subject of an accidental change is a composite whole that includes prime matter and some particular substantial form. Both of the constituents of this composite remain throughout the change. The characteristic or attribute possessed by this composite prior to accidental change, which does not survive the change, is some particular accidental form. The characteristic or attribute possessed by the composite after an accidental change, which according to account. Aristotle does not, however, mention here the third point, namely, that in every generation there must be something generated, since this is obvious” (Aquinas, In Phy., B. I, Ch. 7, L. 12, N. 104, emphasis added). ²⁶ See, for example: Aquinas, DPN, Ch. 2. For an excellent analysis of various passages in Aquinas’s texts on prime matter, as well as a survey of the scholarly literature on this topic, see Wippel, Metaphysical Thought, 312–27. ²⁷ Aquinas, DPN, Ch. 1.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

76  .  it has gained in the change, is some other accidental form. This much of the account parallels the substantial account rather closely. But is there also some composite entity that is corrupted in an accidental change and some composite entity that is generated? Both Aristotle and Aquinas appear to say so. What is corrupted in an accidental change is a composite whole that includes the composite of prime matter and substantial form and the first accidental form. And what is generated in an accidental change is a composite whole that includes the composite of prime matter and substantial form and the second accidental form. The example that Aristotle gives us of this sort of change in the Physics is one in which a human being comes to be characterized as musical.²⁸ He explains that, in such a change, there are three composite entities involved: (1) an “unmusical man,” a composite of prime matter, substantial form, and non-musicality, which is corrupted in the change, (2) a “musical man,” a composite of prime matter, substantial form, and musicality, which is generated in the change, and (3) a human being, a composite of prime matter and substantial form, which remains throughout the change. As Aristotle himself explains in the relevant passage, “One part survives, the other does not: what is not an opposite survives (for the man survives), but ‘not-musical’ or ‘unmusical’ does not survive, nor does the compound of the two, namely the unmusical man.”²⁹ Importantly, Aquinas appears to follow Aristotle in this understanding of accidental change. In his commentary on this passage, he affirms that in an accidental change there is a composite entity, the unmusical man that ceases to exist: when someone has already been made musical, the human being remains, but the opposite does not, were it to be a negative opposite, such as the non-musical, or a privation or contrary, such as the unmusical. Nor does the composite of subject and the opposite remain, for the non-musical man does not remain after the man has been made musical. And so coming to be is attributed to these three things: for it was said that the man becomes musical, and the non-musical becomes musical, and the non-musical man becomes musical. Of these three,

²⁸ Aristotle, Physics, tr. R. P. Hardie and R. K. Gaye, in Barnes (ed.), The Complete Works of Aristotle, Vol. 1 (Princeton, NJ: Princeton University Press, 1984), Ch. 7. ²⁹ Aristotle, Physics, Bk. I, Ch. 7, 190a18–21.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

77

only the first remains at the completion of the production, the other two do not.³⁰

And, on the other side of the change, in the De principiis, he describes the accidental change in which a human being becomes white as generating “a white thing,” that is, a white human being: when a substantial form is introduced, something is said to come to be absolutely, just as we say that a human being comes to be or is generated. When an accidental form is introduced, however, something is not said to come to be absolutely, but rather it is said to come to be in this way, just as when a human being comes to be white, we do not say that a human being comes to be absolutely or that a human being is generated; rather, we say that a white thing comes to be or is generated.³¹

Aquinas’s account of substantial and accidental change, then, an account which he inherits from Aristotle, and to which he commits himself in several of his works, can be seen to support the Standard Model in two ways. First, in the context of these discussions both he and Aristotle refer to the composite of prime matter and substantial form by the name of a material substance. That which is composed of prime matter and a rational soul, for example, that which undergoes the accidental change from unmusical to musical, is referred to as a human being. Second, if we take seriously Aristotle’s and Aquinas’s remarks that in every change there must be something corrupted and something generated, and their remarks that in an accidental change it is a composite of prime matter, substantial form, and an accidental form that is corrupted, and a composite of prime matter, substantial form, and an accidental form that is generated, then this appears to commit them both to an understanding of accidental unities as complexes of precisely this sort: each accidental unity is a composite of a single material substance composed of prime matter and substantial form, and a single accidental form. Aquinas’s preferred accounts of substantial and accidental change, then, appear to require the Standard Model. As Brower puts it, summarizing his case for the Standard Model: just as Aquinas’s realism about matter (or enduring substrata) is an immediate consequence of his general account of change, so too his ³⁰ Aquinas, In Phy., B. I, Ch. 7, N. 102.

³¹ Aquinas, DPN, Ch. 1.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

78  .  realism about prime matter (or enduring substrata for substantial change) is an immediate consequence of his specific account of change. Again, just as his realism about hylomorphic compounds (including his commitment to postulating distinct compounds to serve as the termini for every change) is an immediate consequence of his general account of change, so too his realism about accidental unities (including his commitment to postulating distinct accidental unities to serve as the termini for every accidental change) is an immediate consequence of his specific account of change.³²

4. Setting the Standard II: Unum per Accidens, Unum Simpliciter One reason to think that accidents or accidental forms cannot be included among the metaphysical parts of material substances in Aquinas’s ontology is that including accidental forms within the hylomorphic constitution of material substances would seem to diminish the robust, substantial unity that Aquinas says material substances possess. The worry here is that by including accidental forms within the hylomorphic constitution of material substances the Expanded Model so diminishes the internal unity of material substances that it makes material substances themselves into mere accidental unities. The Standard Model, on the other hand, preserves the sort of robust, substantial unity that Aquinas attributes to material substances, and properly distinguishes the kind of unity exhibited by material substances from the kind of unity exhibited by the accidental unities of which they are parts. For that reason, the Standard Model seems to more accurately reflect Aquinas’s own ontology of material substances. To see the real force of this argument, we need to first explore the various kinds of unity found in Aquinas’s ontology. For Aquinas, unity or oneness comes in various kinds and in various degrees. Following Aristotle, Aquinas recognizes several ways in which something can be

³² Brower, Aquinas’s Ontology, 82.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

79

unified, several ways in which something can be considered one thing.³³ And these various ways of being unified, ways of being one, can be placed along a spectrum, from least unified to most unified. The primary distinction that Aquinas makes between kinds of unity in composite entities is between unities that are “accidentally one” (unum per accidens) and those that are “simply” or “absolutely one” (unum simpliciter).³⁴ Those entities that are accidentally one are less unified, are one in a lesser sense, than those that are simply or absolutely one. Throughout his works, Aquinas gives three sorts of examples of composite entities that are merely accidentally one (unum per accidens) and not simply or absolutely one (unum simpliciter). The first of these are what we might call multi-substance artifacts, such as houses or heaps of stones. As Aquinas explains, a multi-substance artifact is an entity composed of two or more material substances (and often two or more material substances of different types) joined by some accidental form, such as spatial proximity, some more specific spatial arrangement, or some coordinated function or action.³⁵ The second example of composite entities that are merely accidentally one are the accidental unities discussed in section 2 above: “kooky objects” such as white-Socrates and musical-Socrates, and singlesubstance artifacts such as bronze statues and thresholds.³⁶ As we have seen, an accidental unity is an entity composed of a single material substance or subject and a single accidental form. White-Socrates, for example, is composed of a rational soul and prime matter, and an

³³ See, for example, Aristotle, Metaphysics, tr. W. D. Ross, in J. Barnes (ed.), The Complete Works of Aristotle, Vol. 2 (Princeton, NJ: Princeton University Press, 1984), 1603–6 (B. 5, Ch. 7) and Aquinas’s In Met., B. 5, Ch. 7, L. 7–8. ³⁴ See, for example, Aquinas, SCG, I, Ch. 18, N. 2; Aquinas, SCG, II, Ch. 56, N. 10–18; Aquinas, SCG, II, Ch. 57, N. 2–3; Aquinas, SCG, II, Ch. 58, N. 108; Aquinas, ST, I, Q. 11, A. 1, Ad. 2; Aquinas, ST, I, Q. 76, A. 1, Co.; Aquinas, ST, I, Q. 76, A. 3, Co.; Aquinas, ST, I-II, Q. 17, A. 4, Co.; Aquinas, ST, III, Q. 2, A. 1, Co.; Aquinas, QDA, A. 11, Co.; Aquinas, QDSC, A. 3, Co.; Aquinas, QDUVI, A. 3, co.; Aquinas, QDUVI, A. 3, Ad. 5; Aquinas, QDUVI, A. 4, Co.; Aquinas, QQ, I, Q. 4, A. 1, Co.; Aquinas, CT, B. I, Ch. 90; Aquinas, In PH., B. I, L. 8, N. 11; Aquinas, In Phy., B. 1, Ch. 2, L. 3, N. 21–2; Aquinas, In Met. B. 5, Ch. 6, L. 7, N. 843–7; Aquinas, In Met., B. 7, Ch. 17 L. 17, N. 1672–4; Aquinas, In Sent., B. I, D. 24, Q. 1, A. 1, Co. ³⁵ See, for example, Aquinas, SCG, II, Ch. 56, N. 10; Aquinas, SCG, II, Ch. 57, N. 3; Aquinas, SCG, II, Ch. 58, N. 5; Aquinas, ST, I-II, Q. 17, A. 4, Co; Aquinas, ST, III, Q. 2, A. 1, Co.; Aquinas, QDA, Q. 1, A. 11. Co.; Aquinas, In Met., B. 7, Ch. 17, L. 17, N. 1672–4. ³⁶ See, for example, Aquinas, SCG, II, 56, Ch. 12; Aquinas, ST, I, Q. 76, A. 3, Co.; Aquinas, QDSC, A. 3, Co.; Aquinas, QQ, I, Q. 4, A. 1, Co.; Aquinas, In Met., B. 5, Ch. 6, L. 7, N. 843–5; Aquinas, In Sent. B. I, D. 24, Q. 1, A. 1, Co.; Aquinas, In Sent., L. 3, D. 7, Q. 1, A. 1, Ad. 5.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

80  .  accidental form of whiteness. And musical-Socrates is composed of the same rational soul, the same prime matter, and an accidental form of musicality. In some of these “unity passages,” Aquinas also refers to what we might call multi-accident accidental unities, such as white-musicalSocrates, or just-musical-Socrates.³⁷ According to Aquinas, in some contexts, we seem to speak of accidental unities that include more than one accidental form. For example, we may want to concentrate on two of Socrates’ features: his pallor and his musicality. The object of our consideration, white-musical-Socrates, would, then, be understood as composed of a rational soul, prime matter, an accidental form of whiteness, and an accidental form of musicality. Similarly, just-musical-Socrates would be understood as composed of the same rational soul, the same prime matter, the same accidental form of musicality, and an accidental form of justice. Now, it is not clear from these passages whether Aquinas means to recognize the extra-mental existence of such multi-accident accidental unities. He speaks of them in the same sorts of contexts in which he speaks of single-accident accidental unities, and so if we are willing to recognize the existence of the former, then perhaps we ought, for the same reasons, to recognize the existence of the latter. However, unlike single-accident accidental unities, multi-accident accidental unities do not seem necessary to account for any kind of real changes in the world. It would seem that any kind of accidental change involving two or more accidents could be accounted for just as well by means of the corruption of two single-accident accidental unities. It is also not clear whether Aquinas means for multi-accident accidental unities to extend further so as to include triple or quadruple-accident accidental unities. To my knowledge, the only multi-accident accidental unities that Aquinas mentions in any of his works are those that involve only two accidents. The main point here, however, is that Aquinas seems to say that any entity that includes among its metaphysical parts one or more accidental forms is only accidentally one (unum per accidens) and not simply or absolutely one (unum simpliciter).

³⁷ Aquinas, QDA, Q. 1, A. 11. Co.; Aquinas, QDUVI, A. 3, Co.; Aquinas, CT, B. I, Ch. 90; Aquinas, In PH., B. I, L. 8, N. 11; Aquinas, In Met., B. 5, Ch. 6, L. 7, N. 843.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

81

In contrast to these three types of composite entities that possess mere accidental unity, there are at least three sorts of examples that Aquinas gives of composite entities that are indeed simply or absolutely one and not merely accidentally one. The first of these are compounds of matter and form, in particular of prime matter and substantial form. These sorts of unions produce composite substances. The union of soul and body in a human being, for example, is an absolute, substantial union, and this union produces a single, composite substance.³⁸ The second example that Aquinas gives of a simple, absolute unity is the unity exhibited by the various material parts of a material substance. According to Aquinas, every material substance possesses at least three kinds of material parts: quantitative parts, such as a right half and a left half, functional parts, such as a human being’s head, hands, flesh, bones, and blood, and elemental parts, such as the various atoms and molecules of which the material substance is composed. Importantly, for Aquinas, none of these material parts are substances in themselves.³⁹ What this means is that none of these parts have their own independent identity apart from the whole. Each one depends for both its existence and its identity on the whole of which it is a part. When a water molecule becomes a part of the human body, for example, the water molecule loses its substantial form and comes to be informed by the substantial form of the whole of which it is now a part. And this applies to all other material parts. As Aquinas explains, the simple, absolute unity that is a material substance is composed not of smaller material substances with their own substantial forms, but of various virtual or potential parts, all enformed by the single substantial form of the whole.⁴⁰ ³⁸ See, for example, Aquinas, SCG, II, Ch. 56, N. 18; Aquinas, SCG, II, Ch. 57, N. 3; Aquinas, SCG, II, Ch. 58, N. 8; Aquinas, QQ, XII, Q. 7, A. 1, Co.; Aquinas, In Sent., L. 3, D. 27, Q. 1, A. 1, Ad 5; Aquinas, In Sent., L. 4, D. 49, Q. 2, A. 1, Co. ³⁹ See, for example, Aquinas, In Met., B. 5, Ch. 26, L. 21, N. 1102; Aquinas, In Met., B. 7, Ch. 13, L. 13, N. 1588, 1591; Aquinas, In Met., B. 7, Ch. 16, L. 16, N. 1631–2; Aquinas, QQ, IX, Q. 2, A. 1, Co. It should be noted that Aquinas’s claim that each individual substance has one and only one substantial form (and thus has no substances as parts) sets him against most other philosophers in the medieval period. For discussion, see Pasnau, Metaphysical Themes, 574–8; Thomas M. Ward, John Duns Scotus on Parts, Wholes, and Hylomorphism (Leiden: Brill, 2014), 76–109. ⁴⁰ See, for example, Aquinas, ST, I, Q. 76, A. 4, Ad. 4; Aquinas, In BT, Q. 4, A. 3, Ad. 6; Aquinas, DME; Aquinas, QDA, A. 9, Ad. 10; Aquinas, QQ, I, Q. 4, A. 1, Ad. 3; Aquinas, SCG, II, Ch. 56, N. 4. For more on Aquinas’s theory of virtual presence and how exactly we should understand this notion, see, for example: Joseph Bobik, Aquinas on Matter and Form and the Elements (Notre Dame, IN: University of Notre Dame Press, 1998); Christopher Decaen,

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

82  .  The third example that Aquinas gives of a composite entity that is absolutely or simply one is the incarnate Christ. According to Aquinas, the divine and human natures of Christ are not merely accidentally united to another. They are made absolutely or simply one thing by virtue of the absolute, simple unity of the divine essence.⁴¹ Setting aside the special case of the incarnation (for now), then, Aquinas recognizes three sorts of accidental unities and two sorts of simple or absolute unities among composite entities. These two sorts of simple or absolute unities are the union of substantial form and prime matter in a material substance and the unity exhibited by the various material parts of a material substance. Anything less than either of these robust, substantial unities turns out to be a mere accidental unity of one of the three types outlined above. But what is it that makes the union of substantial form and prime matter in a material substance, or the unity exhibited by the various material parts of a material substance, an absolute or simple unity? And what is it that makes the union of one or more material substances and an accidental form, or a single subject and one or more accidental forms, a mere accidental unity? In the context of these sorts of discussions, Aquinas often suggests that the general metaphysical principle at work in discerning whether something possesses accidental or absolute unity is that no union between two actual things can ever produce an absolute or simple unity.⁴² Multi-substance artifacts, for example, fail to possess absolute or simple unity because they include among their constituents two or more actual material substances. Each individual material substance, on the other hand, does possess absolute or simple unity because it does not include among its constituents any actual material substances as parts, only virtual or potential substances. Similarly, single and double-accident accidental unities fail to possess absolute or simple “Elemental Virtual Presence in St. Thomas,” The Thomist 64 (2000), 271–300; Michael Hector Storck, “Parts, Wholes, and Presence by Power: A Response to Gordon P. Barnes,” The Review of Metaphysics 62 (2008), 45–59. ⁴¹ See, for example, Aquinas, ST, III, Q. 2, A. 1, Co.; Aquinas, QDUVI, A. 3, Co., Ad. 5; Aquinas, QDUVI, A. 4, Co. ⁴² See, for example, Aquinas, SCG, I, Ch. 18, N. 2; Aquinas, SCG, II, Ch. 56, N. 12, 14; Aquinas, SCG, II, Ch. 58, N. 8; Aquinas, ST, III, Q. 2, A. 1, Co.; Aquinas, QDA, Q. 1, A. 11. Co.; Aquinas, QDSC, A. 3, Co.; Aquinas, QQ, I, Q. 4, A. 1, Co.; Aquinas, In Sent., L. 3, D. 2, Q. 1, A. 1, QC. 3, Ad. 1.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

83

unity because they include among their constituents two or more actual forms: at least one substantial form and at least one accidental form. Each compound of substantial form and prime matter, on the other hand, does possess absolute or simple unity because it does not include among its constituents any actual forms other than its substantial form. The prime matter in each of these compounds is, as we have seen, pure potentiality, possessing no actuality of its own. Putting all of these pieces together, we can begin to see why Aquinas’s account of accidental and substantial unity would seem to pose problems for the Expanded Model. Aquinas is clear that he thinks that material substances are simply or absolutely one (unum simpliciter) and not merely accidentally one (unum per accidens). But, setting aside (for the moment) the case of the incarnation, Aquinas only recognizes two instances of simple or absolute unity among composite entities: the union of substantial form and prime matter in a material substance, and the unity exhibited by the various material parts of a material substance. All other unions turn out to be mere accidental unities of one of the three types outlined above. Now, based on my descriptions of the various kinds of unities found in Aquinas’s ontology above, and my earlier description of the Expanded Model in section 2, it seems that the Expanded Model’s conception of material substances places material substances under the category of multi-accident accidental unities. For, like multi-accident accidental unities, material substances on the Expanded Model include among their metaphysical parts substantial form, prime matter, and various accidental forms. As I said earlier, it is not clear whether Aquinas means to recognize the existence of accidental unities that include more than two accidental forms, but he does provide, in the context of these discussions, the general framework for modeling such entities. Such entities would include among their metaphysical parts the substantial form and the prime matter of the subject, as well as all of the material substance’s accidental forms. And because such an entity would include among its parts several actual forms, by Aquinas’s own criteria it seems that the greatest unity that it could possess would be mere accidental unity or oneness. Understanding material substances as possessing these sorts of metaphysical parts, then, which the Expanded Model does, would seem to significantly diminish the robust, substantial unity that Aquinas himself clearly and consistently attributes to material

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

84  .  substances throughout his works, making them into nothing more than complex accidental unities.

5. The Case for Expansion I: Divine Simplicity The main disagreement between the Standard Model and the Expanded Model pertains to accidental forms. On the Standard Model, accidental forms are “outside” of material substances, as it were; they are not included among their metaphysical parts. On the Expanded Model, accidental forms are, in a sense, “inside” of material substances; they are included among their metaphysical parts. If it can be shown, then, that Aquinas does in fact consider accidental forms to be included among a material substance’s metaphysical parts, then this would go a long way toward showing that the Expanded Model is a more accurate representation of Aquinas’s ontology. And so, in making my case for the Expanded Model, I would like to begin by focusing on those texts that seem to support such a conclusion. One reason to think that, on Aquinas’s view, accidents or accidental forms are indeed parts of material substances is the way in which he argues for God’s simplicity in question three of the Prima pars.⁴³ Here Aquinas argues for God’s simplicity negatively, that is, by giving an exhaustive list of all of the ways in which something can be composite and all of the different sorts of parts a composite thing can have, and then systematically arguing that the God whose existence he has demonstrated in the previous question cannot be composite in any of those ways or have any of those sorts of parts. The structure of question three is as follows: In article one, Aquinas considers whether God is a body, that is, whether God has any physical or quantitative parts. In article two, Aquinas considers whether God is composed of matter and form, that is, whether God has any essential parts. In article three, Aquinas considers whether God is the same as His essence or nature, that is, whether God has any non-essential parts. In article four, Aquinas considers whether essence and existence are the same in God, that is, whether God’s existence (His “esse”) is some part of Him distinct from His ⁴³ Aquinas, ST, I, Q. 3.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

85

essence. In article five, Aquinas considers whether God is contained in a genus, that is, whether God has any “definitional parts.” In article six, Aquinas considers whether there are any accidents in God, and here I want to interpret this article as asking whether God has any accidental parts (I will return to this point shortly). In article seven, Aquinas considers whether God is altogether simple, that is, whether God has any other sorts of parts not mentioned in the previous articles. And, in article eight, Aquinas considers whether God enters into the composition of other things, that is, whether God Himself is a part of anything else. Now, based on this brief summary of question three, it is clear that there is a very important shift that occurs between articles seven and eight. In article seven, Aquinas considers the various ways in which something can be composite and the various sorts of parts a composite thing can have. In article eight, however, Aquinas considers the ways in which some entity can itself be a part of some further composite. The shift here is between an analysis of the possibility of a decomposition of God into parts and the composition of some further composite that has God as one of its parts. Now, one natural way of reading this question is to see the shift between article seven and article eight as signaling a shift between not only articles seven and eight, but between article eight and all of the other articles that came before. In this way, we might read article seven as a sort of summary of the preceding articles, summarizing all of the ways discussed earlier in which something can be composite and all of the parts that a composite thing can be said to have. We might then read it as inquiring further whether these ways of being composite and these various sorts of parts form a comprehensive list, that is, whether there are any other ways God could be composite, or any other sorts of parts that God might be said to have. Once Aquinas has established the comprehensiveness of articles one through six, he can then move on to ask a different sort of question in article eight, one that is only indirectly about what sorts of parts God might be said to have. In support of this interpretation of the structure of his argument, consider the beginning of his reply in article seven: That God is entirely simple can be shown in several ways. First, from what has been said above. For, in God there is no composition of quantitative parts, since He is not a body, nor of form and matter.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

86  .  Nor is there any difference in God between His nature and His suppositum, nor any difference between His essence and His existence. Nor is there any composition of genus and difference in Him, or of subject and accident. It is clear, then, that God is in no way composite, but is entirely simple.⁴⁴

Here we see that Aquinas himself views each of the preceding articles as detailing a particular way in which something can be composite. He also views the articles collectively as offering a comprehensive list of all of the possible ways in which something can be composite. Article eight, then, cannot be asking the very same question as the other articles (whether God is in any way composite), since Aquinas already takes himself to have decisively proven that point. Article eight asks a very different, though importantly related, question about God’s nature—whether He can be said to be a part of any composite.⁴⁵ There are two important conclusions to be drawn from my interpretation of Aquinas’s argument in question three. First, if I am right that articles one through six all detail various ways in which something can be composite, then article six, which considers whether there are any accidents in God, should be read as asking whether God has any accidental parts. Now, if Aquinas held that accidents were no part of any substance, then it would not make any sense to ask whether they are parts of God. All of the other ways of being composite mentioned in the other articles are in fact true descriptions of other substances, according to Aquinas. And so the fact that, on my interpretation, Aquinas is asking whether God has any accidents as parts is an indication that he holds that in the case of other individual substances, they do have accidents as parts. On my reading, that God is simple in that he does not have any accidents as parts is one more way in which God differs from other existing things.⁴⁶ ⁴⁴ Aquinas, ST, I, Q. 3, A. 7, Co. ⁴⁵ For a similar interpretation of Aquinas’s mode of argumentation in his treatment of Divine simplicity in the Summa, see Peter Weigel, Aquinas on Simplicity (Bern: Peter Lang, 2008). ⁴⁶ Given that article four asks whether God’s existence is distinct from his essence, my argument here might also lend support to the claim that, for Aquinas, the “esse” or “act of existence” of every other thing should be included among its metaphysical parts. Since the focus of this paper is on the formal constituent(s) of material substances, I will have to sidestep issues pertaining to esse or acts of existence. I do not mean to reject the claim that material substances

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

87

Secondly, my interpretation of question three also counts against the Standard Model. Recall that, according to the Standard Model, for Aquinas, accidental forms are not parts of material substances. Rather, accidental forms are parts of accidental unities, the other parts of which are material substances. According to the Standard Model, then, all other substances besides God are parts of larger composites, namely, accidental unities. We should, expect, then, that there would be some mention of this in article eight, since it is here that Aquinas considers whether God is a part of any composite. At the very least, we should expect that Aquinas would explain to the reader that, since it has already been shown that there are no accidents in God, we can conclude that God is not part of any accidental unity. In many of the other articles, Aquinas freely refers back to previous articles when the conclusions there are relevant for the argument at hand.⁴⁷ But we do not see any reference to article six in article eight. Instead, the only place in which God’s relation to accidents is discussed is in article six itself (as well as the summary in article seven), which, according to my interpretation, is on the decomposition side of question three. If the Standard Model were the correct reading of Aquinas, then Aquinas would not have to ask both whether there are any accidents in God and whether God is a part of any composite. The former would be a species of the latter. Now, none of this is a knockdown argument against the Standard Model. Aquinas could be moving from discussing a way in which God might be said to be a part of some further composite in article six, to discussing the ways in which God might be said to have parts in article seven, and then back to discussing other ways in which God might be said to be a part of some further composite in article eight. But a virtue of the Expanded Model is that it supports a much more natural reading of Aquinas’s argument in question three, according to which articles one through seven are all about the sorts of parts that God might be said to

include among their metaphysical parts their esse or act of existence. I only mean to bracket these issues for the time being. ⁴⁷ In the Sed contra of Article 2, he refers back to Article 1; in the Corpus of Article 4, he refers back to Articles 1 and 3; in Article 6, he refers back to Articles 2 and 3; and in Article 7, he refers to “the previous articles.”

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

88  .  have, and only article eight shifts to a discussion of the things of which God might be said to be a part.⁴⁸

6. The Case for Expansion II: Essence and Suppositum A second reason to think that, for Aquinas, accidents or accidental forms are to be included among the metaphysical parts of material substances is the way in which he distinguishes between the nature or essence of a material substance and the material substance itself, the suppositum, throughout his corpus. According to Aquinas, the essence or nature of a thing is that which makes the thing the kind of thing that it is. For any two members of the same kind or species, then, there is a sense in which those two members can be said to possess the same essence or nature.⁴⁹ In the case of material substances, essences or natures are composite: the essence or nature of a material substance includes both its form and its matter. The form that is included in the essence of a material substance is the substantial form that is characteristic of its kind. In the case of human beings, for example, the essence includes the substantial form that makes something a human being, which is a rational soul. Aquinas calls the

⁴⁸ In addition to his treatment of divine simplicity in the Prima pars, Aquinas also argues for the absolute simplicity of God in three other places: Aquinas, SCG, I, Chs. 18–27, Aquinas, QDPD, Q. 7, A. 1–4, and Aquinas, CT, B. I, Chs. 9–24. In each of these texts, the order in which Aquinas argues for his conclusions is slightly different. In the SCG, QDPD, and CT, for instance, Aquinas argues first that there is no composition of any sort in God, and then he goes on to argue that God must therefore be said to lack all of the various sorts of parts that other things have. And the order in which Aquinas considers the different sorts of parts that God might be said to have varies. In all four of these texts Aquinas concludes that there is no composition of subject and accident in God, but only in the ST (A. 8) and the SCG (Chs. 26–7) does Aquinas also conclude that God is no part of any other thing. Nevertheless, in the SCG, as in the ST, that God is no part of any other thing is a conclusion reached after he has finished his discussion of the sorts of parts that God must be said to lack. In the Sed contra of article four of question seven of the QDPD (a Sed contra to which Aquinas does not offer a reply), there is also an interesting passage that reads: “Every accident depends on something else. There can be no such thing in God, because that which depends on something else is caused to exist. And God is the first cause and is in no way caused. Therefore, in God there can be no accident.” Note here that the concern is that, since every accident is dependent on something else, placing an accident in God would make God dependent on something else. Why would placing an accident in God make Him dependent on something else? I would like to suggest that what Aquinas has in mind here is that placing an accident in God would make God dependent on something else because God would thereby depend on something else for the existence of one of His parts. I think that this passage should be read, then, as supporting the Expanded Model. ⁴⁹ See, for example, Aquinas, DEE, Ch. 1; Aquinas, ST, I, Q. 3, A. 3, Co.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

89

matter that is included in the essence of a material substance its “common matter.”⁵⁰ Common matter is the kind of matter that something must possess in order to be the kind of thing that it is. According to Aquinas, the common matter of human beings is flesh and bones; every human being is composed of this same kind of stuff. Because the essence or nature of a thing is that which makes the thing the kind of thing that it is, there is a sense in which every material substance of a particular kind or species has the same substantial form and the same common matter. It is, however, important to understand what this sameness amounts to. Two human beings do not possess numerically the same substantial form and numerically the same common matter. Rather, they possess numerically distinct substantial forms and numerically distinct common matters, which are nevertheless qualitatively identical. The sense in which both the common matter and the substantial form of which the essence of a material substance is composed are “the same,” then, is that they are intrinsically indistinguishable. They are numerically distinct from one another only by virtue of one or more extrinsic individuating principles.⁵¹ What, then, is the relationship between an individual material substance, or suppositum, and its essence or nature? Here I think that a proponent of the Standard Model has two options. First, he or she could identify an individual material substance with its individuated essence or nature. In such case, “common” matter would be understood as nothing more than prime matter specified to a certain extent by the substantial form that inheres in it, and an individual material substance would be nothing more than a particular instance of the common matter and substantial form that place it in its kind. Now, as was mentioned above, an individual essence or nature is an individual as a result of certain extrinsic individuating principles. And so according to this first way of understanding the relationship between individual material substances and their natures, individual material substances would not

⁵⁰ See, for example, Aquinas, DEE, Ch. 2 and Ch. 6; Aquinas, ST, I, Q. 29, A. 2, Ad. 3; Aquinas, QDPD, Q. 9, A. 1, Ad. 6. ⁵¹ My interpretation of Aquinas on this point is indebted to Jeffrey Brower’s analysis of common natures in his “Aquinas on the Problem of Universals,” Philosophy and Phenomenological Research 90 (2015), 1–21. See also, Brower, “Matter, Form, and Individuation,” 94–100.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

90  .  include among their metaphysical parts those principles that individuate them from others in their kind. The difference between an essence or nature and an individual material substance would not be a mereological one; an individual material substance would not have any additional metaphysical parts outside of its extrinsically individuated essence.⁵² Alternatively, a proponent of the Standard Model could maintain a distinction between an individual material substance and its nature or essence by including the individuating principles that make that nature or essence an individual nature or essence among the metaphysical parts of the individual substance. In such a case, common “matter” would be understood as nothing more than certain general specifications for an individual material substance’s matter that are built into its substantial form, and the prime matter of an individual substance would serve as the individuating principle that individuates the individual substance’s nature or essence from others of its kind.⁵³ The key thing to note here is that, on either option, an individual material substance would still be nothing more than its prime matter and its substantial form. Importantly, both of these options for understanding the relationship between an individual material substance and its essence or nature on the Standard Model appear to be in conflict with what Aquinas himself says on this issue. When Aquinas himself describes the difference between the nature or essence of an individual material substance and the individual material substance itself he clearly and consistently describes that difference in mereological terms. Early on in his De ente et essentia, for example, Aquinas makes a distinction that will carry on through the rest of his works between two sorts of matter found in a material substance: undesignated and designated.⁵⁴ Undesignated or common matter is, as we have seen, included in the essence of an individual material substance. Common matter is the matter which, together with a thing’s substantial form, places that thing in its kind. This matter is common in that it does not include any of the designations that make the matter of one member of a substantial kind distinct from the matter of other members of that kind. Designated matter, on the other hand, falls outside ⁵² This seems to be the strategy pursued by Brower in his Aquinas’s Ontology, 112. ⁵³ This seems to be the strategy pursued by Brower in his “Matter, Form, and Individuation,” 96 and in his “Aquinas on the Problem of Universals,” 16. ⁵⁴ Aquinas, DEE, Ch. 2.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

91

of a thing’s essence and so does not place a thing in its kind. Designated matter does, however, include the designations that make a thing the particular thing that it is, distinct from other members of its kind. What is most pertinent about this distinction for the present discussion is the way in which Aquinas characterizes the relationship between individual material substances, their common matter, and their designated matter. According to Aquinas, common matter, together with substantial form, comprises the essence of a thing. And the relationship between this composite essence and any individual substance that possesses it is one of part to whole. An individual material substance includes among its parts a composite essence and designated matter.⁵⁵ Now, what is it, precisely, that designates matter? Aquinas is clear that prime matter includes no designations whatsoever, since it is nothing else but pure potency.⁵⁶ Elsewhere we learn that prime matter becomes designated by certain dimensions, which as Aquinas explains, are accidents falling under the category of quantity.⁵⁷ It seems, then, that for Aquinas, an individual material substance does include among its metaphysical parts at least some of its accidents. It must include at least the quantities that designate its matter. Neither the first nor the second option for understanding the relationship between individual substances and nature or essences outlined above reflect this feature of Aquinas’s account. That individual material substances or supposita include among their metaphysical parts the accidental forms or quantities that designate their matter and thus individuate them from other members of the same kind is a claim found in several of Aquinas’s works. Importantly, in many of these works, Aquinas also expands the list of accidental forms to be numbered among the metaphysical parts of individual material substances to include more than just their particular spatial dimensions. For instance, when comparing the complexity of human beings to the

⁵⁵ See Aquinas, DEE, Ch. 2, as well as the passages cited in my discussion of this point below. ⁵⁶ For descriptions of prime matter as “pure potency,” see, for example: Aquinas, SCG, I, Ch. 17, N. 7; Aquinas, ST, I, Q. 115, A. 1, Ad. 2. ⁵⁷ The clearest explication of this point is in Aquinas’s commentary on Boethius’s De Trinitate (Aquinas, In BT, Q. 4, A. 2 and A. 4), but he also makes note of this fact in his DEE, Ch. 2.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

92  .  simplicity of God in article three of question three of the Prima pars (considered above), Aquinas explains: in things composed of matter and form, the suppositum and its nature or essence must be distinguished, since the essence or nature includes in itself only that which falls under the definition of the species, just as humanity includes in itself that which falls under the definition of human being. For this human being is a human being, and this refers to his humanity, namely, that by which the human being is a human being. But the individual matter, with all of the individuating accidents, does not fall under the definition of the species, for this flesh and these bones, or this whiteness, or this blackness, or other things of this sort do not fall under the definition of a human being. Hence this flesh and these bones and the accidents individuating this matter are not included in humanity. Nevertheless, they are included in that which is a human being. Hence, that which is a human being possesses in itself something that humanity does not. And for this reason a human being and his humanity are not totally the same, but humanity refers to a formal part of the human being, since the defining principles are possessed formally with respect to the individuating matter.⁵⁸

In this passage, Aquinas tells us that the relationship between the essence of a thing and the thing itself, the suppositum, is a mereological one: the essence of a thing is one of its parts. What other parts does an individual material substance have beyond its essence? Aquinas says here that the other parts of an individual substance include its “individuating accidents,” which accords well with what he says in the De ente about designated matter. But he also says here that the other parts of an individual substance beyond its essence include various other nonessential attributes, such as its color and “other things of this sort.” This passage, then, would seem to support the claim, espoused by proponents of the Expanded Model, that an individual substance includes among its metaphysical parts not only the accidents that help to individuate from others of the same kind, but also its various qualities and “other things of this sort,” that is, its other accidental forms. (This passage also

⁵⁸ Aquinas, ST, I, Q. 3, A. 3, Co.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

93

lends further support to my interpretation of question three of the Prima pars outlined above.) And this is not an isolated remark. In his Commentary on Aristotle’s Metaphysics, for example, Aquinas makes the very same point: Humanity is not altogether the same as a human being, since humanity signifies only the essential principles of a human being and excludes all of the accidents. For humanity is that by which a human being is a human being, and so includes none of the things that are accidental to a human being. Hence, all of the accidents possessed by a human being are excluded from the definition of humanity. Now, it is this particular human being himself that is a human being, that possesses the essential principles, and is that to which accidents can inhere. And so, even though the definition of a human being does not include any of the accidents that a human being possesses, nevertheless ‘human being’ does not refer to something separate from those accidents. And therefore ‘human being’ refers to the whole, and ‘humanity’ refers to a part.⁵⁹

Here again we have the claim that a human being, an individual material substance, includes among its metaphysical parts more than its essence. Here Aquinas explains that “humanity” is only one part of the larger whole that is an individual human being. What else does this larger whole include? While Aquinas does not explicitly say that the individual material substance that is a human being includes the accidents that it possesses among its parts, he does say that a human being’s humanity (her essence or nature) is only one part of her, and that the term “human being” does not refer to something separate from those accidents. It is not very much of a leap, then, to suggest that in this passage Aquinas means for these accidents to be included among the additional parts to which he is referring. What is most interesting about this particular passage is that it describes accidents as both inhering in a material substance and as being included in the whole that is that material substance. This would seem to count against the Standard Model’s claim that accidents must inhere in material substances “from the outside,” as it were. ⁵⁹ Aquinas, In Met., B. 7, L. 5, 1379.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

94  .  Finally, that Aquinas holds that there is a mereological difference between the nature or essence of a thing and the suppositum is perhaps most clearly illustrated in his account of the incarnation. So, for instance, in his treatment of the incarnation in the Tertia pars, he states: ‘Nature’ refers to the essence of the species, which the definition signifies. And if nothing else were found joined to that which pertains to the notion of the species, then it would not be necessary to distinguish the nature from the suppositum of the nature, which is the individual subsisting in that nature, since, in this case, every individual subsisting in some nature would be altogether identical to its nature. Now, in certain subsisting things something can in fact be found which does not pertain to the notion of the species, namely accidents and the individuating principles, as is most clear in those things that are composed of matter and form. And therefore, in such cases the nature and the suppositum are different things, not as if they were altogether separate from one another, but since the suppositum includes the nature of the species and certain other things over and above the notion of the species, ‘suppositum’ refers to the whole, which possesses the nature as a formal part and is perfective of it.⁶⁰

Once again, in this passage, Aquinas explains that the essence or nature of a material substance does not exhaust its metaphysical composition. He explains here that any individual material substance will also include among its metaphysical parts its “individuating principles,” which, as we have seen, are accidental forms in the category of quantity, and its accidents. As in the two previous passages, Aquinas places no restriction here on the sorts of accidents that he wishes to include. The most straightforward way of reading these sorts of passages, then, I submit, is to see them as outlining a view according to which all accidental forms ⁶⁰ Aquinas, ST, III, Q. 2, A. 2, Co. For other passages in Aquinas’s corpus that make similar points, see, for example: Aquinas, QDPD, Q. 9, A. 1, Co.; Aquinas, QDPD, Q. 7, A. 4, Co.; Aquinas, QDUVI, A. 1, Co.; Aquinas, QDUVI, A. 3, Co.; Aquinas, QDUVI, A. 3, Ad. 14; Aquinas, QQ, II, Q. 2, A. 2, Co.; Aquinas, QQ, II, Q. 2, A. 2, Ad. S.C.; Aquinas, CT, B. 1, Ch. 10; Aquinas, CT, B. 1, Ch. 154; Aquinas, In DA, B. 3, L. 8, 706; Aquinas, DEE, Ch. 2; Aquinas, In Met., B. 7, L. 11, 1521–2; Aquinas, In Met., B. 7, L. 12, 1535–6; Aquinas, In Met., B. 8, L. 3, 1710; Aquinas, QDA, Q. 1, A. 17, Ad. 10; Aquinas, SCG, I, Ch. 21, N. 2; Aquinas, SCG, IV, Ch. 81, N. 10; Aquinas, In Sent., B. 1, D. 23, Q. 1, A. 1. For more on Aquinas’s distinction between individual substances or supposita and essences or natures, see Wippel, Metaphysical Thought, 238–53.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

95

are metaphysical parts of material substances. These sorts of passages can be found in several of Aquinas’s works, spanning several stages of his career. This fact is, I think, rather strong evidence that, for Aquinas, a suppositum, an individual material substance, is not exhaustively composed of its matter (either prime or common) and its substantial form, as it is on the Standard Model. Rather, for Aquinas, a suppositum includes its accidental forms as further metaphysical parts.⁶¹

7. Further Expansion I: “Substance” Is Said in Many Ways As I said at the start of this essay, the Standard Model of Aquinas’s ontology of material substances is the Standard Model for a reason. It is not difficult to find passages that seem to support such a reading. This prompts what is perhaps the most obvious, most immediate objection to the Expanded Model: if the Expanded Model is indeed the correct interpretation of Aquinas, and accidental forms are indeed parts of individual material substances in Aquinas’s ontology, then why does Aquinas almost never include the accidental forms of a material substance in the list of its metaphysical parts? And why does he regularly refer to the composite of substantial form and prime matter as substance, or the composite of a particular kind of substantial form and prime matter by the name of a particular substance? I think that this is a legitimate concern for the Expanded Model. But I also think that there are at least two strategies available to a proponent of this model for contending with these sorts of passages. In general, I think that a proponent of the Expanded Model has to interpret those passages in which Aquinas refers to the composite of a particular kind of substantial form and prime matter by the name of a particular substance as a sort of shorthand. When Aquinas speaks of a human being, for ⁶¹ While I have focused in this paper on Aquinas’s account of material substances, there are similar passages in Aquinas’s texts that suggest that angels, too, have their accidental forms as further metaphysical parts, and for similar reasons (see, for example: Aquinas, ST, I, Q. 77, A. 1, Ad. 5; Aquinas, QQ, II, Q. 2, A. 2, Co.). For discussion of this point, see Christopher Hughes, “Aquinas on Continuity and Identity,” Medieval Philosophy and Theology 6 (1997), 96–7; Hughes, Aquinas on Being, Goodness, and God, 68–74, 138–41; and Wippel, Metaphysical Thought, 244–5, 246, and 247.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

96  .  example, as composed of substantial form and prime matter, a proponent of the Expanded Model has to interpret Aquinas as referring only to the essential parts of a human being, setting aside the various accidental parts that she also possesses. Indeed, if what I have argued in the previous section is correct, we must read him this way. For Aquinas clearly thinks that the matter of which a human being is composed is designated matter, matter with some particular dimensions, and these dimensions, as we have seen, are accidents in the category of quantity. It cannot be the case, then, that a human being, or any individual material substance, is composed of only its prime matter and substantial form. There must be more to it than that. Concerning those passages in which Aquinas refers to the composite of substantial form and prime matter as substance, a proponent of the Expanded Model has a second sort of response that he or she can offer. Following Aristotle’s remarks in Book V, Chapter 10 of his Metaphysics, Aquinas often distinguishes at least three ways of speaking of substances.⁶² First, we can speak of substantial forms as substances inasmuch as substantial forms are the “cause of being” for the others. Let us refer to this category of substance as substance₁. Second, we can speak of essences, natures, or quiddities as substances, inasmuch as they are signified by the definitions of things. Let us refer to this category of substance as substance₂. Finally, we can speak of individual substances (elements, minerals, plants, animals, human beings, angels, etc.) as substances, inasmuch as they are the ultimate subjects of predication. Let us refer to this category of substance as substance₃. This threefold distinction between ways of speaking of substance runs throughout Aquinas’s works, though occasionally Aquinas mentions only the second and third. For instance, in his discussion of the Trinity in the Prima pars, Aquinas writes: Following what the Philosopher says in Book V of the Metaphysics, ‘substance’ is spoken of in two ways. In one way substance is spoken of as the quiddity of a thing, which the definition signifies. In this way we say that the definition signifies the substance of a thing, which the Greeks called ousia, and which we may speak of as the essence of a ⁶² Aquinas, In Met., B. 5, L. 10, 898–905. For more on this threefold distinction in Aquinas, see Wippel, Metaphysical Thought, 198–208.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

97

thing. In another way, substance is spoken of as the subject or suppositum, that which subsists in the genus of substance.⁶³

This general distinction is also reiterated in his treatment of the incarnation in the Tertia pars: “Substance, as is clear from Book V of the Metaphysics, is spoken of in two ways: in one way as the essence or nature of a thing; in another way as the suppositum or hypostasis.”⁶⁴ The significance of this distinction for my purposes becomes clear when we consider once again the fact that, according to Aquinas, the essence or nature of a material substance includes both the matter and the substantial form of that substance. As Aquinas explains in his De ente et essentia: From the conjunction of both [substantial form and matter] there results that being in which a thing subsists in itself, and from them something essentially one is produced. And, therefore, from their conjunction a certain essence results. Hence, the form, though in itself it is not considered to have the complete notion of an essence, nevertheless, is part of a complete essence . . . [On the other hand,] from accident and subject something essentially one is not produced, but something accidentally one. And therefore from their conjunction a certain essence does not result, as it does from the conjunction of form and matter. As a result, an accident has neither the notion of a complete essence, nor is part of a complete essence.⁶⁵

And so, keeping this understanding of natures or essences in mind, when Aquinas refers to the composite of matter and substantial form as substance, we need not infer from these remarks that he holds that a material substance has substantial form and prime matter as its only metaphysical parts. That would be to collapse the distinction between substance₃ and substance₂. What he means when he refers to the composite of matter and substantial form as substance is that matter and substantial form comprise the essence, nature or quiddity of an individual material substance, which we can call a substance, as long as we keep in mind that by doing so we are calling it a substance₂, not a substance₃, which is the individual material substance or suppositum of which it is a part. ⁶³ Aquinas, ST, I, Q. 29, A. 2, Co. ⁶⁵ Aquinas, DEE, Ch. 5.

⁶⁴ Aquinas, ST, III, Q. 2, A. 6, Ad. 3

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

98  .  Admittedly, these two strategies for contending with those passages in Aquinas’s texts that are difficult to reconcile with the Expanded Model do not decisively resolve this issue. Nor do these two strategies address all of the passages that can be seen to support the Standard Model.⁶⁶ I think that these strategies do, however, take a bit of the sting out of the more commonly cited passages. They also provide proponents of the Expanded Model with a way of trying to reconcile these passages with what Aquinas himself says about the composition of material substances in the context of his discussions of divine simplicity and the real distinction between the essence or nature of individual material substance and the individual material substance itself.

8. Further Expansion II: Making Room for Accidental Unities Proponents of the Expanded Model typically do not speak of accidental unities when describing Aquinas’s ontology of material substances. Perhaps this is because they do not think that Aquinas actually does recognize the existence of such things, or at least that he does not take them to be additional entities in his ontology beyond the material substances with which they are associated. Perhaps proponents of the Expanded Model would rather think of accidental unities as “beings of reason,” found only in the mind. Perhaps “white-Socrates” is just a helpful abstraction, something that we consider when we abstract away several of Socrates’ metaphysical parts in order to concentrate on some important subset of those parts for purposes of explanation in a particular context. And there may be good reasons, interpretive or otherwise, for espousing such a deflationary reading of Aquinas’s references to accidental unities.⁶⁷ It should be

⁶⁶ The two passages cited at the start of this essay in support of the Standard Model are particularly difficult to deal with, since here Aquinas not only lists matter and form as components of a material substance, he also seems to explicitly deny that there are any other further components. ⁶⁷ Some of these are discussed in Lindsay Cleveland, “Groundwork for a Thomistic Account of Contemporary Property Roles” (Doctoral Dissertation, Baylor University 2018), Ch. 6. Cleveland, a proponent of the Expanded Model, favors this approach to accidental unities.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

99

noted, however, that a proponent of the Expanded Model could recognize the real, extra-mental existence of accidental unities without compromising any of the main commitments of his or her view. And since, as was shown in sections 3 and 4 above, there are several key passages in which both Aristotle and Aquinas speak of accidental unities, it might be beneficial for a proponent of the Expanded Model to do so. On the Expanded Model, material substances include among their metaphysical parts their prime matter, their substantial form, and all of their accidental forms. If we hold fixed the claim made by proponents of the Standard Model above that accidental unities include among their metaphysical parts prime matter, substantial form, and a single accidental form, then a proponent of the Expanded Model can recognize the existence of accidental unities. Importantly, however, these entities would be found within the larger composite that is the individual substance. For example, on the Expanded Model, white-Socrates may be understood as a hylomorphic compound that includes Socrates’ prime matter, Socrates’ substantial form, and Socrates’ pallor, but which excludes all of Socrates’ other accidents. And seated-Socrates may be understood as a hylomorphic compound that includes Socrates’ prime matter, Socrates’ substantial form, and Socrates’ seatedness, but which excludes all of Socrates’ other accidents. In this way, on the Expanded Model, accidental unities would turn out to be parts of material substances, not the other way around. Each accidental unity within a particular material substance would include some particular subset of that substance’s metaphysical parts, and each would also overlap with every other inasmuch as each would include among its own parts the same prime matter and substantial form. If proponents of the Expanded Model were to understand accidental unities in this way, then they would also be able to capture everything that proponents of the Standard Model want to say about accidental change. Recall that on the Standard Model’s understanding of accidental change, when a material substance undergoes accidental change, say, when it gains an accidental form, there are no new substances that are generated, but there is something that is generated. On the Standard Model, there is an accidental unity that is generated when a material substance gains an accidental form. And, conversely, when the material substance loses that accidental form, that same accidental unity

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

100  .  is corrupted. The Standard Model’s account of accidental change preserves the idea that in every change, some hylomorphic compound is generated or corrupted. In substantial change, material substances are generated or corrupted; in accidental change, accidental unities are generated or corrupted. Now, with the account of accidental unities introduced earlier in this section, the Expanded Model can also accommodate the claim that there is something generated or corrupted in every real change. And, in particular, it can also accommodate the claim that material substances are generated or corrupted in substantial change and accidental unities are generated or corrupted in accidental change. Recall that, on the Expanded Model of accidental unities outlined above, accidental unities are understood as parts of material substances and not the other way around. An accidental unity is a hylomorphic compound comprised of a material substance’s prime matter, its substantial form, and one of its accidental forms. On an Expanded Model of accidental change, then, when a material substance gains an accidental form, there is something generated: the accidental unity consisting of the three parts just mentioned. And when the material substance loses an accidental form, the accidental unity that possesses that accident as its principal metaphysical part is corrupted. Naturally, the material substance itself continues to exist in such changes, since the material substance does not lose its prime matter, its substantial form, or its other accidental forms. It simply loses a non-essential part. Like a cat losing a tuft of fur (or worse, a limb), the material substance carries on through the loss of one of its nonessential parts by virtue of its continued possession of what is essential to it.⁶⁸ Nevertheless, a proponent of the Expanded Model can admit that some hylomorphic compound (found “within” the material substance) has ceased to exist. And in this way, the Expanded Model can preserve the Standard Model’s basic account of accidental change without having

⁶⁸ As Aquinas himself explains at ST, III, Q. 2, A. 3, Ad 1: “A difference in accidents makes a thing ‘other’; a difference of essence makes ‘another thing.’ Now, it is clear that, in created things, the otherness that results from a difference in accidents can pertain to the same hypostasis or suppositum. In such cases, numerically the same thing can underlie different accidents. However, in created things, it is not the case that numerically the same thing can subsist in different essences or natures. Hence, that, in creatures, one thing is said to be ‘other’ does not signify a difference in suppositum, but only a difference in accidental forms.”

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

101

to surrender its claim that accidental forms are to be understood as metaphysical parts of individual material substances in Aquinas’s ontology. Now, whether proponents of the Expanded Model need to recognize the existence of accidental unities in the first place will depend on how seriously we want to take Aquinas’s references to such entities. But, as I have shown, a proponent of the Expanded Model is free to espouse even a strongly realist interpretation.

9. Further Expansion III: Unity, Identity, and Persistence In section 4, I considered an argument in favor of the Standard Model, according to which including accidental forms within the hylomorphic constitution of material substances would so diminish their internal unity that they themselves would become no more unified than mere accidental unities. As we have seen, according to Aquinas, no union of two actual forms produces a simple or absolute unity (unum simpliciter). The greatest degree of unity that any entity possessing two actual forms can achieve, then, appears to be mere accidental unity (unum per accidens). Earlier it was also discovered that, on the Expanded Model, individual material substances appear to have the very same mereological structure as one of the three major kinds of accidental unity recognized by Aquinas, which we called multi-accident accidental unities. The Expanded Model, then, would seem to significantly diminish the robust, substantial unity that Aquinas himself clearly and consistently attributes to material substances throughout his works, with the result that, on this model, material substances turn out to be nothing more than complex accidental unities. In the previous section, I offered one way in which a proponent of the Expanded Model can distinguish between individual material substances and accidental unities, but several key questions remain: what sort of unity does an individual material substance possess on the Expanded Model, and how is an individual material substance to be distinguished from the multi-accident accidental unity that shares its parts? In this final section of the paper I respond on behalf of proponents of the Expanded Model to both of these concerns.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

102  .  The first claim that I would like to defend is that, on the Expanded Model, individual material substances remain simply or absolutely one, and do not become merely accidentally one. The key to understanding how a composite entity can be considered a simple or absolute unity while still possessing two or more actual things among its constituents is to understand that something can be both simply or absolutely one and also accidentally many. Let us begin with the following passage from Aquinas’s Summa theologiae: nothing prevents something from being in one way divided and in another way undivided, just as that which is divided by number is undivided according to species, and thus it may turn out that something is one in one way and many in another. But, nevertheless, if something is undivided simply, either because it is undivided according to that which pertains to its essence (though it may be divided with respect to those things that are outside of its essence, like something which is one in subject and many according to its accidents) or because it is undivided in act and divided in potency (like something which is one as a whole and many according to its parts), in such case it will be simply or absolutely one (unum simpliciter) and accidentally many (multa secundum quid). On the other hand, if something is undivided accidentally and divided simply, inasmuch as it is divided according to its essence, and undivided accidentally, or according to its principle or cause, it will be simply or absolutely many and accidentally one, like those things that are many in number and one in species or in principle. Hence, being is divided with respect to one and many in this way, as it were, by means of simply or absolutely one and accidentally many.⁶⁹

In this passage, Aquinas explains that something can be one or undivided in one way and many or divided in another way. Here Aquinas describes two ways in which this can occur. First, something can be simply or absolutely one while also having many parts. One example he gives of this is the unity exhibited by the various material parts of a material substance. As we have seen, for Aquinas, a material substance is simply or absolutely one in that it possesses no substantial parts, but it is also ⁶⁹ Aquinas, ST, I, Q. 11, A. 1, Ad. 2.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

103

virtually or potentially many, in that many of its material parts are such that they could, if separated from the whole, become substances in their own right. What is especially pertinent for my purposes is the other example that Aquinas gives of something that is simply or absolutely one while also having many parts. Here Aquinas says that something counts as simply or absolutely one as long as there is a simple or absolute unity within its essence, that is, as long as what makes that thing what it is essentially is a simple or absolute unity. Importantly, this is so even if that thing possesses further accidental parts outside of its essence. In such case, the individual in question will be simply or absolutely one by virtue of the union of its essential parts and accidentally many by virtue of the plurality of its various accidental parts. The second way described in this passage in which something can be one or undivided in one way and many or divided in another is that something can be simply or absolutely many while also exhibiting sufficient unity to count as one thing in a certain respect. Aquinas does not give us a clear example of this sort of case in this particular passage, but he does in similar passages elsewhere. In question seventeen, article four of the Prima-secundae, for example, Aquinas says: nothing prevents certain things from being many in one respect and one in another. Indeed every plurality is one in some respect . . . Nevertheless, a difference is to be noted here: that certain things are simply many and one in a certain respect and certain other things are the reverse. Now, ‘one’ is predicated in the same way as ‘being’. And a substance is a being absolutely, but an accident or a being of reason is a being only in a certain respect. And therefore those things that are one in substance are simply or absolutely one and many in a certain respect, just as a whole in the genus of substance composed of its integral or essential parts is simply or absolutely one (unum simpliciter), for the whole is a being and a substance simply whereas the parts are beings and substances in the whole. On the other hand, those things that are distinct with respect to their substance and one by means of an accident are simply or absolutely many and one accidentally, or one in a certain respect (unum secundum quid), just as many human beings are one people and many stones are one heap, which is a unity of composition or order. Similarly, many individuals, which are one in genus or species are simply or absolutely many and one in a certain

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

104  .  respect, since to be one in genus or species is to be one according to nature (ratio).⁷⁰

Here again we see Aquinas’s distinction between being simply or absolutely one and accidentally many (one in a certain respect) and being simply or absolutely many and accidentally one. And once again one of the examples that Aquinas gives of something that is simply or absolutely one and accidentally many is an individual material substance. An individual material substance is simply or absolutely one in that it is one substance, but it is also many in that it possesses various parts (integral, essential, and accidental). In this passage, Aquinas also gives us a helpful example of something that is simply or absolutely many and accidentally one. Here Aquinas says that a single group of people is unified in this way, as are many stones in one heap: each is simply or absolutely many in that it includes among its constituents two or more material substances, and each is accidentally one in that those material substances are joined or united by a single accidental form. What we are given, in this passage, then, is a distinction between the kind of unity exhibited by individual material substances and the kind of unity exhibited by accidental unities, one that does not rely on a distinction between possessing accidental forms as parts and not possessing accidental forms as parts. What makes the unity of an individual material substance different from the unity of an accidental unity is not that an accidental unity includes among its metaphysical parts an accidental form and an individual material substance does not. Rather, it is that an accidental unity has as the source of its being, and as the source of its unity, the union of some accidental form with one or more subjects, whereas an individual material substance has as the source of its being, and as the source of its unity, the union of its essential parts, its substantial form and prime matter. That Aquinas recognizes the possibility of something being simply or absolutely one despite including among its metaphysical parts several actual forms is perhaps most clearly illustrated in his account of the

⁷⁰ Aquinas, ST, I-II, Q. 17, A. 4, Co. A similar passage can also be found in Aquinas, In DN., Ch. 11, L. 2, in which Aquinas says that a human being, qua material substance, is in itself one (in seipso unum), but is within itself many (intra seipsum non unum).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

105

incarnation. In article three of his Disputed Questions on the Union of the Word Incarnate, for example, Aquinas explains: Although both ‘one’ and ‘many’ can refer to the suppositum and to the nature, it is clear that if some single suppositum were to have many substantial natures it would be simply or absolutely one (unum simpliciter) and many in a certain respect (multa secundum quid). A sign of this is that those things which are distinct with respect to suppositum and are one with respect to that which pertains to the nature in itself are simply or absolutely many but one in genus or species. And, therefore, conversely, if one suppositum were to have many natures it would be simply or absolutely one and many in a certain respect. Since, therefore, Christ is one suppositum having two natures, it follows that He is simply or absolutely one and two in a certain respect.⁷¹

What we see here is that the incarnate Christ is a case in which Aquinas clearly thought that something can include among its metaphysical parts two actual forms or things and still be considered an absolute or simple unity. Here the incarnate Christ includes within his suppositum two actual natures, one human and one divine. Christ is many in that he possesses these two actual natures, and yet He is also simply or absolutely one by virtue of those two natures being found within the same suppositum and by virtue of the absolute simplicity of the divine essence. And this is not meant to be an isolated case. Aquinas himself anticipates the implications of this account of the incarnation for understanding individual material substances. Earlier in the same article, for example, Aquinas says: that which has unity is said denominatively to be one, just as that which has whiteness, or that which is the subject of whiteness is said to be white. And for the same reason, that which has plurality is said denominatively to be many and that which has duality is said denominatively to be two. Now, since one is convertible with being, just as there is accidental being and substantial being, something is said to be either one or many according to an accidental form or according to a substantial form. Indeed, with respect to accidental forms something is said to be many which is the subject of different forms, either ⁷¹ Aquinas, QDUVI, A. 3, Co.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

106  .  successively or simultaneously. Successively, as, for example, Socrates sitting is different from Socrates standing; for that reason Socrates, inasmuch as he is first standing and then sitting is many successively. Simultaneously, as, for example, Socrates inasmuch as he is white and musical is many.⁷²

In the same article, Aquinas goes on to explain: Socrates, insofar as he is white and musical, is many, not simply or absolutely, but in a certain respect. For, with respect to an accident something is said to be, not simply, but in a certain respect. But with respect to substance something is said to be, as a being, simply or absolutely one or simply or absolutely many.⁷³

Altogether what I think that these texts reveal is that, for Aquinas, whether something possesses simple or absolute unity or accidental unity is not simply a matter of whether something includes among its metaphysical parts one or more actual forms. The real issue is whether something has as the source of its unity, the source of its identity, the union of two or more actual forms or whether it has as the source of its unity and identity something that is itself simply or absolutely one. An accidental unity is accidentally one because it has as the source of its unity, the source of its identity, the union of one or more accidental forms with one or more subjects. The incarnate Christ is simply or absolutely one despite the fact that he includes among his metaphysical parts two actual natures because he has as the source of his unity and identity the absolutely undivided divine essence. And an individual material substance is simply or absolutely one despite the fact that it includes among its metaphysical parts various accidental forms because it has as the source of its unity and identity the union of its substantial form and prime matter. Including accidental forms within the hylomorphic constitution of material substances, then, does not prevent them from possessing the robust, substantial unity that Aquinas attributes to them. Based on my interpretation of the passages above, then, I think that a proponent of the Expanded Model should give the following account of

⁷² Ibid.

⁷³ Ibid.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

107

the distinction between accidental unities and individual material substances. An accidental unity is produced, and is made one thing, by the union of an accidental form and its subject. For material substances, that subject includes prime matter and substantial form. And so any particular accidental unity is produced, and is made one thing, by the union of two actual things: a subject that includes prime matter, substantial form and an accidental form. White-Socrates, for example is produced, and is made one thing, by Socrates’ prime matter and substantial form and the accidental form of whiteness that inheres in him. As a result, in order to describe what white-Socrates essentially is we need to make reference to two actual things, and for white-Socrates to continue to exist, he must maintain possession of both of these actual parts. If white-Socrates were to lose the accidental form of whiteness that inheres in Socrates, that is, if Socrates were to undergo an accidental change pertaining to the color of his skin, then white-Socrates would cease to exist, necessarily. It is the coming together of two actual things, then, that makes white-Socrates what it essentially is and the continued union of these two actual things that makes it the case that it continues to be. In this way, the unity found in an accidental unity is, on the Expanded Model, both rigid and fragile. It is rigid in that the parts of an accidental unity are fixed. White-Socrates always and everywhere possesses the very same metaphysical parts. It is fragile in that any change in any of those parts would bring about the destruction of that unity. As Socrates undergoes various accidental changes over the course of his life, accidental unities are coming in and going out of existence at a remarkably rapid rate. There may be some accidental unities that remain for much of Socrates’ life, and there may even be some that remain for all of Socrates’ life,⁷⁴ but there are many more that are only present for a fraction of Socrates’ life. Accidental unity is a unity that is easily and frequently broken. A material substance, on the other hand, is produced, and is made one thing, by the union of substantial form and prime matter. As we have seen, for Aquinas, prime matter is pure potentiality, completely lacking in any actuality of its own. And so a material substance is produced, and is made one thing, by the union of one actual thing and one potential ⁷⁴ Presumably, those accidental unities that are composed of necessary accidents would be such.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

108  .  thing. Socrates, for example, is produced, and is made one thing, by his rational soul and his prime matter. As a result, in order to describe what Socrates essentially is we need to make reference to both his form and his matter, but we do not need to make reference to any of his accidents. Any accidents that Socrates possesses, even if they are included among his metaphysical parts, are accidental, not essential, to what he is. They are in no way constitutive of his identity. Consequently, for Socrates to continue to exist, he must maintain possession of both of his essential parts, his form and his matter,⁷⁵ but he need not maintain possession of any particular accidental part. Socrates can and does gain and lose such parts over time without ceasing to exist and without changing what he essentially is. It is not the coming together of two actual things, then, that makes Socrates what he essentially is; it is still just the coming together of his rational soul and his prime matter. And it is not the continued union of any two actual things that makes it the case that he continues to be; he can survive the dissolution of any accidental unities found within him.⁷⁶ In this way, the unity found in a material substance is, on the Expanded Model, more flexible than the unity present in an accidental unity, but it is also more robust, capable of surviving the ebb and flow of accidental change. Substantial unity is a unity far less easy to break. But what about multi-accident accidental unities? How is an individual material substance to be distinguished from the multi-accident accidental unity that shares all of its parts? Let us consider, then, the possibility that Aquinas does mean to recognize the extra-mental existence of multi-accident accidental unities, and the possibility that in Aquinas’s ontology there is indeed some multi-accident accidental unity that includes among its metaphysical parts the very same metaphysical parts that a material substance is said by a proponent of the Expanded Model to possess at any particular time. For example, let us say that there is some multi-accident accidental unity associated with Socrates that includes his prime matter, his rational soul, and all of the accidental forms that he possesses at that moment. Let us call such an accidental unity a comprehensive accidental unity. I would like to argue

⁷⁵ Setting aside the possibility of Socrates surviving into the afterlife without any kind of matter. ⁷⁶ Setting aside any accidental unities composed of necessary accidents.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

109

that even if there is such a comprehensive accidental unity in Aquinas’s ontology, this accidental unity would still be importantly different from the sort of composite whole that the Expanded Model understands Socrates to be in at least two ways. First, unlike Socrates, this comprehensive accidental unity would depend for its identity on the accidental forms of which it is composed. The accidental unity would be what it essentially is because of the accidental forms that it possesses as metaphysical parts. Second, and as a result of the first, unlike Socrates, this comprehensive accidental unity would depend for its existence on the accidental forms of which it is composed. The existence of such an accidental unity would be so fragile that once it lost any of its accidental forms, or gained any new ones, that is, once Socrates changed any one of his accidental features, that accidental unity would cease to exist and would be replaced by some numerically distinct comprehensive accidental unity with slightly different parts. Socrates himself, on the other hand, the mereologically flexible and robustly unified composite whole that includes Socrates’ prime matter, Socrates’ rational soul, and all of the accidental forms that Socrates happens to possess at that time, does not depend for his identity on any of the accidental forms of which he is composed; they are not what makes him what he essentially is, and so he is able to survive the addition and subtraction of various accidental parts over the course of his existence. The sort of composite whole with which the Expanded Model identifies a material substance, then, is a very different kind of composite whole from an accidental unity. It is in at least two ways distinguishable even from the sort of comprehensive accidental unity that includes all of the material substance’s accidental forms. It has its identity, its unity, and its existence from a different source. While the comprehensive accidental unity gets its identity, its unity, and its existence at least in part from its constituent accidents, Socrates gets his identity, his unity, and his existence from his essence, which includes his substantial form and prime matter but none of his accidents. And so, for this reason, I think that it is false to say that on the Expanded Model material substances turn out to be mere accidental unities. Material substances have unique identity and persistence conditions that set them apart from any accidental unities to which they may be related and with which they may share their parts.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

110  .  In summary, then, on the Expanded Model, individual material substances are both simply or absolutely one and also accidentally many. An individual material substance is simply or absolutely one by virtue of the substantial form–prime matter compound that serves as the source of its unity, identity, and existence. And an individual material substance is accidentally many by virtue of the various accidental forms and accidental unities that are included within its hylomorphic constitution. As I have argued here, I think that, like the Standard Model, the Expanded Model preserves the robust, substantial unity that Aquinas attributes to material substances, and I think that it also properly distinguishes the kind of unity exhibited by material substances from the kind of unity exhibited by the accidental unities with which they are associated. And for that reason, I do not think that Aquinas’s account of accidental and absolute unity gives us any reason to think that the Standard Model more accurately reflects Aquinas’s ontology of material substances.

10. Conclusion There are, then, two main models for understanding Aquinas’s ontology of material substances: the Standard Model, according to which accidental forms are not included among their metaphysical parts, and the Expanded Model, according to which they are included among their metaphysical parts. In this paper, I have attempted to illustrate the major differences between these two models and to outline the main arguments that can be offered in support of each. I have argued that despite a number of passages in Aquinas’s texts that seem to support the Standard Model, and two big-picture reasons for attributing to him this sort of view, there are several other passages in Aquinas’s texts that support the Expanded Model, and at least two other big-picture reasons for attributing to him this sort of view. I have also argued that a proponent of the Expanded Model can accommodate Aquinas’s account of substantial and accidental change without having to surrender any of the main commitments of the view, and preserve the substantial unity of material substances without having them collapse into accidental unities. Based

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

111

on these considerations, I think that there are good reasons to think that the Expanded Model better reflects Aquinas’s views on material substances, and good reasons to think that, for Aquinas, accidental forms are indeed metaphysical parts of material substances.⁷⁷ University of Mary

Bibliography Aristotle. Categories, tr. J. L. Ackrill, in J. Barnes (ed.), The Complete Works of Aristotle, Vol. 1 (Princeton, NJ: Princeton University Press, 1984), 3–24. Aristotle. Metaphysics, tr. W. D. Ross, in J. Barnes (ed.), The Complete Works of Aristotle, Vol. 2 (Princeton, NJ: Princeton University Press, 1984), 1552–78. Aristotle. Physics, tr. R. P. Hardie and R. K. Gaye, in J. Barnes (ed.), The Complete Works of Aristotle, Vol. 1 (Princeton, NJ: Princeton University Press, 1984), 315–446. Bobik, Joseph. Aquinas on Matter and Form and the Elements (Notre Dame, IN: University of Notre Dame Press, 1998). Brower, Jeffrey E. “Matter, Form, and Individuation,” in E. Stump and B. Davies (eds.), The Oxford Handbook of Aquinas (Oxford: Oxford University Press, 2012), 94–100. Brower, Jeffrey E. Aquinas’s Ontology of the Material World: Change, Hylomorphism, and Material Objects (Oxford: Oxford University Press, 2014). Brower, Jeffrey E. “Aquinas on the Problem of Universals,” Philosophy and Phenomenological Research 90 (2015), 1–21. Brower, Jeffrey E. “Aquinas on the Individuation of Substances,” in R. Pasnau (ed.), Oxford Studies in Medieval Philosophy 5 (Oxford: Oxford University Press, 2017), 122–50. Brown, Christopher M. Aquinas and the Ship of Theseus: Solving Puzzles about Material Objects (New York: Continuum, 2005).

⁷⁷ I would like to thank Eleonore Stump, Susan Brower-Toland, Jeff Brower, Tim Pawl, and audiences at the 2017 meeting of the American Philosophical Association (Central Division) in Kansas City, the 2016 Symposium on Medieval and Renaissance Studies at Saint Louis University, and the 2016 Meeting of the Kentucky Philosophical Association at Bellarmine University for helpful comments on earlier drafts of this paper.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

112  .  Cleveland, Lindsay. “Groundwork for a Thomistic Account of Contemporary Property Roles” (Doctoral Dissertation, Baylor University, 2018). Cohen, S. Marc. “Kooky Objects Revisited: Aristotle’s Ontology,” Metaphilosophy 39 (2008), 3–19. Cohen, S. Marc. “Accidental Beings in Aristotle’s Ontology,” in G. Anagnostopoulos and F. D. Miller, Jr. (eds.), Reason and Analysis in Ancient Greek Philosophy (Dordrecht: Springer, 2013), 231–42. Cross, Richard. “Aquinas on Nature, Hypostasis, and the Metaphysics of the Incarnation,” The Thomist 60 (1996), 171–202. Decaen, Christopher, “Elemental Virtual Presence in St. Thomas,” The Thomist 64 (2000), 271–300. Gorman, Michael. “Uses of the Person–Nature Distinction in Aquinas’s Christology,” Recherches de théologie et philosophie médiévales 67 (2000), 58–79. Gorman, Michael. Aquinas on the Metaphysics of the Hypostatic Union (Cambridge: Cambridge University Press, 2017). Hughes, Christopher. “Aquinas on Continuity and Identity,” Medieval Philosophy and Theology 6 (1997), 93–108. Hughes, Christopher. Aquinas on Being, Goodness, and God (New York: Routledge, 2015). Inman, Ross. “Neo-Aristotelian Plenitude,” Philosophical Studies 168 (2014), 583–96. Lewis, Frank. “Accidental Sameness in Aristotle,” Philosophical Studies 42 (1982), 1–36. Loux, Michael J. “Aristotle’s Constituent Ontology,” in D. W. Zimmerman and K. Bennett (eds.), Oxford Studies in Metaphysics 2 (Oxford: Oxford University Press, 2006), 207–50. Loux, Michael J. “Aristotle’s Hylomorphism,” in L. Novak and D. D. Novotny (eds.), Neo-Aristotelian Perspectives in Metaphysics (New York: Routledge, 2014), 138–63. Matthews, Gareth. “Accidental Unities,” in M. Schofield and M. Nussbaum (eds.), Language and Logos: Studies in Ancient Greek Philosophy (Cambridge: Cambridge University Press, 1982), 223–40. Oderberg, David S. Real Essentialism (New York: Routledge, 2007).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

   ’ 

113

Pasnau, Robert. “Form and Matter,” in R. Pasnau (ed.), The Cambridge History of Medieval Philosophy (Cambridge: Cambridge University Press, 2010), 635–46. Pasnau, Robert. Metaphysical Themes 1274–1671 (Oxford: Clarendon Press, 2011). Pasnau, Robert. “Mind and Hylomorphism,” in J. Marenbon (ed.), The Oxford Handbook of Medieval Philosophy (Oxford: Oxford University Press, 2012), 486–504. Scaltsas, Theodore. Substances and Universals in Aristotle’s Metaphysics (Ithaca, NY: Cornell University Press, 1994). Shields, Christopher. Order in Multiplicity: Homonymy in the Philosophy of Aristotle (Oxford: Clarendon Press, 1999). Storck, Michael Hector, “Parts, Wholes, and Presence by Power: A Response to Gordon P. Barnes,” The Review of Metaphysics 62 (2008), 45–59. Stump, Eleonore. Aquinas (New York: Routledge, 2003). Thomas Aquinas. Opera omnia, ed. E. Alarcón, . The following abbreviations are used:

Compendium theologiae (CT). De ente et essentia (DEE). Expositio super Dionysium De divinis nominibus (In DN). Expositio super librum Boethii De Trinitate (In BT). In duodecim libros Metaphysicorum Aristotelis expositio (In Met.). In libros De anima expositio (In DA). In libros De generatione et corruptione expositio (In DGC). In libros Peri Hermeneias expositio (In PH). In libros posteriorum Analyticorum expositio (In PA). In octo libros Physicorum expositio (In Phy.). De principiis naturae (DPN). Quaestiones disputatae de anima (QDA). Quaestiones disputatae de potentia Dei (QDPD). Quaestiones disputatae de spiritualibus creaturis (QDSC). Quaestiones disputatae de unione Verbi incarnati (QDUVI). Quaestiones quodlibetales (QQ). Scriptum super libros Sententiarum (In Sent.). Summa contra Gentiles (SCG). Summa theologiae (ST).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

114  .  Ward, Thomas M. John Duns Scotus on Parts, Wholes, and Hylomorphism (Leiden: Brill, 2014). Weigel, Peter. Aquinas on Simplicity (Bern: Peter Lang, 2008). West, J. L. A. “The Real Distinction between Supposit and Nature,” in P. Kwasniewski (ed.), Wisdom’s Apprentice: Thomistic Essays in Honor of Lawrence Dewan, O.P. (Washington, DC: Catholic University of America Press, 2007), 85–106. Wippel, John F. The Metaphysical Thought of Thomas Aquinas (Washington, DC: Catholic University of America Press, 2000).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

Aquinas on the Intension and Remission of Accidental Forms Gloria Frost

Everyday experience reveals that qualities increase and decrease in their degrees of intensity.¹ There are many familiar examples of this: after several washings, the blue color of a T-shirt is less vibrant. Upon adding more sugar, tea which is already sweet becomes sweeter. Through repeated actions on the battle field, the already courageous soldier becomes even more courageous. This familiar phenomenon of qualitative increase and decrease raises a host of metaphysical puzzles. For instance, when the blue color of the T-shirt fades, is the shirt’s former property of blueness replaced by a new slightly varying property of blue—or, does the shirt have the same property of blueness throughout ¹ I quote Aquinas’s texts from Sancti Thomae Aquinatis, Doctoris Angelici, opera omnia, iussu impensaque Leonis XIII P.M. edita, 50 vols. (Rome: S.C. de Propaganda Fide, 1882–). I use the following abbreviations: De anima = Quaestiones disputatae de anima [vol. 24/1, ed. B. C. Bazán, 1996]; De ente = De ente et essentia [vol. 43, 1976]; De malo = Quaestiones disputatae de malo [vol. 23, 1982]; De oper. occul. = De operationibus occultis [vol. 43, 1976]; De prin. = De principiis naturae [vol. 43, 1976]; De ver. = Quaestiones disputatae de veritate [vols. 22/1–3, ed. A. Dondaine, 1970–6]; De sub. sep. = De substantiis separatis [vol. 40D, 1968]; In Ethic. = Sententia libri Ethicorum [vol. 47/1–2, ed. R.-A. Gauthier, 1969]; In de heb. = Expositio De ebdomadibus [vol. 50, 1992]; In Physic. = In VIII libros Physicorum [vol. 2, 1884]; Quodl. = Quaestiones de quolibet [vol. 25/1–2, ed. R.-A. Gauthier, 1996]; ScG = Summa contra gentiles [vols. 13–15, 1918–30]; ST = Summa theologiae [vols. 4–12, 1886–1906]. For texts not included in the Leonine edition, I quote from the following editions, using the following abbreviations: De virt. = Quaestiones disputatae, t. 2: Quaestiones disputatae de virtutibus [5th ed.: Rome: Marietti, 1922]; In de div. nom. = In librum Beati Dionysii De divinis nominibus expositio, ed. C. Pera, P. Caramello, C. Mazzantini (Rome: Marietti, 1950); In lib. De causis = Super librum De causis expositio, ed. H. D. Saffrey (Fribourg: Société Philosophique, 1954); Sent = Scriptum super libros Sententiarum magistri Petri Lombardi, ed. R. P. Mandonnet and R. P. Maria Fabianus Moos (Paris: Lethielleux, 1929–47). I cite the divisions within the text according to the following convention: 1.1 = question 1, article 1. Roman numerals are used to indicate book numbers of works with multiple books, e.g. I Sent. = first book of the Sentences. Citations refer to the “corpus” of the text unless otherwise noted. All English translations are my own.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

116   the fading process? If so, are qualitative properties themselves able to undergo change? The topic of the intensification and weakening of qualities was vigorously debated by medieval philosophers.² Their interest was occasioned by the theological belief that the goal of human life on earth was to grow in charity. Charity was seen as a quality (i.e. a habit) which perfected the human person, enabling her to perform the act of loving God with ease and pleasure.³ Medieval Christians believed that in the afterlife human beings in heaven would differ in the degree to which they received the beatific vision based on the degree of their charity. Given these theological beliefs, it was crucial for medieval thinkers to develop a metaphysics of qualities which allowed for the same subject’s quality to increase in its perfection over time and for two different subjects’ qualities to be such that one is more perfect than the other’s. Thirteenth- and fourteenth-century thinkers developed and analyzed a wide array of theories about the metaphysics of qualitative increase and decrease. The broad contours of the medieval debate—referred to by specialists as the debate about the “intension and remission of forms”— have been outlined in a series of important studies.⁴ Historians have credited some medieval views on qualitative increase with paving the way

² Interest in the question of intensification died out in the early modern period when reductive approaches to sensible qualities became prominent. On a reductive approach, sensible qualities, such as color and heat, are reduced to quantifiable features of matter, such as mass, extension, and position. Qualitative intensification has a straightforward explanation on the reductive account of qualities: Every change we experience in the intensity of a perceived quality is reducible to a quantitative change in the location or motion of particles. The perceived increase in the water’s heat is nothing more than an increase in the speed of particulate motion, while the “fading of the shirt’s blueness” is nothing beyond changes in the lengths of the light waves it emits. The major drawback of reductive approaches is that they undercut the reliability of our perceptual faculties. For example, when we sense color or heat we believe that we are perceiving qualities which really belong to material objects. But this is false on the reductive account. For a defense of Aquinas’s color realism in light of contemporary science see Christopher Decaen, “The Viability of Aristotelian–Thomistic Color Realism,” The Thomist 65 (2001), 179–222. ³ For Aquinas’s discussion of charity see for instance ST II-II.23.1 & 2. For secondary literature see Eberhard Schockenhoff, “The Theological Virtue of Charity,” in S. J. Pope (ed.), The Ethics of Aquinas (Washington, DC: Georgetown University Press, 2002), 244–58; and Michael Sherwin OP, By Knowledge and By Love: Charity and Knowledge in the Moral Theology of St. Thomas Aquinas (Washington, DC: Catholic University of America Press, 2005). ⁴ See, for example, Annaliese Maier, Das Problem der intensiven Grösse in der Scholastik (De Intensione et remissione Formarum) (Rome: Verlag Heinrich Keller, 1939) and Jean-Luc Solère, “Plus ou moins: le vocabulaire de la latitude des forms,” in J. Hamesse and C. Steel (eds.), L’Élaboration du vocabulaire philosophique au Moyen Âge (Turnhout: Brepols, 2000), 437–88.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

       

117

for modern (i.e. mathematical) approaches to physics in so far as they employed quantitative models to analyze qualitative changes.⁵ Along with the general literature on the medieval debate, there have been several recent articles on the views of particular medieval thinkers who held historically or conceptually significant positions.⁶ There has been very little work, however, on the intensification of qualities in the thought of Thomas Aquinas.⁷ Aquinas was not a cutting-edge thinker when it came to medieval physics. In fact, he countered the more novel positions which analyzed qualitative increase in terms of quantificational increase. Aquinas’s position on intension and remission is nevertheless important because it provides a unique window into crucial, but neglected, aspects of his views on the metaphysics of accidental qualities and accidental change.⁸ To explain qualitative increase, Aquinas claims that a subject can exist more or less perfectly through one and the same invariable accidental form. Aquinas describes the relationship between a substance and its ⁵ On the significance of these views in the history of science see for instance Edith Sylla, “Medieval Concepts of the Latitude of Forms: The Oxford Calculators,” Archives d’histoire doctrinale et littéraire du Moyen Âge 40 (1974), 223–83 and The Oxford Calculators and the Mathematics of Motion, 1320–1350: Physics and Measurement by Latitudes (New York: Garland Publishing, 1991). ⁶ See, for example, Can Laurens Löwe, “Gregory of Rimini on the Intension and Remission of Corporeal Forms,” Recherches de théologie et de philosophie médiévales 81 (2014), 273–330; Brian Francis Conolly, “Dietrich of Freiburg on the Succession of Forms in the Intensification of Qualities,” Recherches de theólogie et philosophie médiévales 81 (2014), 1–35; Magali Roques, “Quantification and Measurement of Qualities at the Beginning of the Fourteenth Century: The Case of William of Ockham,” Documenti e studi sulla tradizione filosofica medievale 27 (2016), 347–80. ⁷ Aquinas’s positions are discussed alongside other figures in the studies cited in notes 4 and 5. There is only one study exclusively devoted to Aquinas’s views of intension and remission: Jean-Luc Solère, “Thomas d’Aquin et les variations qualitatives,” in C. Erismann and A. Schniewind (eds.), Compléments de substance: Études sur les propriétés accidentelles offertes à Alain de Libera (Paris: Vrin, 2008), 147–65. Solère also discusses Aquinas’s position at length in his article “Les variations qualitatives dans les théories post-thomistes,” Revue Thomiste 1 (2012), 157–204 and there is some discussion of Aquinas’s views on qualitative intensification in Barry Brown, Accidental Esse (New York: University Press of America, 1985), 168–84. Major studies of Aquinas’s metaphysics and natural philosophy have left out his views on this topic. For example, the topic is not covered in John Wippel’s magisterial work The Metaphysical Thought of Thomas Aquinas (Washington, DC: Catholic University of America Press, 2000) and Jeffrey Brower’s more recent Aquinas’s Ontology of the Material World: Change, Hylomorphism, and Material Objects (Oxford: Oxford University Press, 2014). ⁸ Both of these topics have been the subject of recent scholarly interest. See, for instance, Brower’s major study of Aquinas’s hylomorphism and views on change cited in note 7 and the alternative analysis of the ontological status of form in Robert Pasnau’s recent chapter “On What There Is in Aquinas,” in J. Hause (ed.), Aquinas’s Summa Theologiae: A Critical Guide (Cambridge: Cambridge University Press, 2018), 10–28.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

118   accidental forms as a “participation” relationship which is susceptible to degree.⁹ In changes of intensification, Aquinas thinks that the substance comes to exist more perfectly through its form while the form in itself remains invariable. Unlike many of his contemporaries, Aquinas rejected that change as such requires that a subject gain or lose a form. For him, some changes merely involve a subject participating more or less perfectly in an already inherent form. For Aquinas, not only do subjects have potentialities for acquiring forms, but they also have potentialities for participating more or less perfectly in forms.¹⁰ The goal of this article is to explore these intriguing aspects of Aquinas’s views on qualitative increase. Aquinas’s approach to this topic has farreaching implications for understanding his broader thought as he draws on his account of intension to address many other topics. For example, when explaining what it is for an object to be in motion, he says that a moving object “gradually participates more and more” in form.¹¹ He likewise appeals to the claim that subjects can be actualized more or less perfectly by a form in order to account for differences in the strength of numerically distinct powers of the same specific kind.¹² In his epistemology, he appeals to differences in the degree to which a form actualizes a subject to explain how different subjects might know an object more or less perfectly in virtue of the same intelligible species.¹³ My discussion of Aquinas’s positions will rely on texts from a variety of different contexts written over the course of his career; and my approach to reconstructing his views will be systematic.¹⁴ Section 1 gives a brief overview of the problem which qualitative increase posed for Aristotelian hylomorphists and it discusses Aquinas’s critique of the ⁹ While there is much literature on Aquinas’s use of the notion of participation in his discussions of the relationship between creatures and God, there has not been much study of Aquinas’s views on how substances participate in their accidental forms. ¹⁰ These aspects of Aquinas’s views on change have not been recognized in recent scholarship. Brower, for instance, claims that for Aquinas change necessarily involves two distinct contrary forms. See, for instance, Brower, Aquinas’s Ontology, p. 72 and Brower, “Matter, Form, and Individuation,” in E. Stump and B. Davies (eds.), The Oxford Handbook of Aquinas (Oxford: Oxford University Press, 2012), 85–103. ¹¹ In III Physic. lec. 2, n. 3 (Leonine 2, p. 105). ¹² See, for instance, De virt. 1.11 (Marietti, p. 524). ¹³ See, for instance, In IV Sent. 49.2.4 where he applies his theory of intension and remission to explain how the blessed might know God with varying degrees of perfection. ¹⁴ Aquinas’s major discussions of the intension and remission of forms occur in the following texts which treat of questions surrounding the increase of habits and virtues: I Sent. 17.2.1–5; Quodl. IX.6.un.; ST I-II.52.1–3; ST I-II.66.1–2; ST II-II.24.4–10; De virt. 1.11; De virt. 5.3.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

       

119

prominent medieval solutions. Section 2 outlines the details of Aquinas’s own preferred solution: the participation model of intensification. Section 3 discusses Aquinas’s views on the causes of changes in degree of participation.¹⁵

1. The Problem of Intension and Remission and Aquinas’s Critique of the Prominent Medieval Solutions The shared background assumption for the debate about the intension and remission of forms in the medieval period was a commitment to Aristotelian hylomorphism and its model for analyzing change. According to Aristotelian hylomorphism, each material substance is composed of matter and a substantial form. Substantial forms account for a substance’s unity and species (e.g. human, feline). Accidental forms inhere in substances to account for their sensible qualities like heat and blue, as well as character traits like courage and intellectual dispositions, such as the ability to think about geometry. The substance as a whole (i.e. the matter–substantial form composite) functions as the matter which is actualized by accidental forms. Aristotle’s motivation for his theory of hylomorphism was accounting for change. When one thing is corrupted and another is generated from it or when a substance loses a quality and gains a new one, it is clear that there is some aspect which stays the same and another aspect which differs as a result of the change. In Aristotle’s view, the best way to explain the novelty and persistence associated with change is by postulating that material substances are composed of actuality (form) and potentiality (matter). According to the Aristotelian paradigm for explaining change, intrinsic changes involve (1) some matter, prime matter in a substantial change or a material substance which functions as matter in accidental change, and (2) contrary forms which successively actualize

¹⁵ There are other important aspects of Aquinas’s views on intension which I cannot address here due to space constraints. For example, in his latest works Aquinas claimed that some composite and dispositional forms could vary in themselves. I plan to discuss this topic and others elsewhere.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

120   a potentiality of the matter.¹⁶ For example, when Socrates changes from being pale to tan, Socrates functions as the matter which is successively actualized by the form of paleness and then the form of tan-ness. The form of tan-ness replaces the form of paleness in Socrates, yet Socrates remains the same material substance throughout the change. The intensification and remission of qualities posed a challenging problem because such changes do not straightforwardly fit into the Aristotelian paradigm for explaining change. When a subject becomes hotter or more charitable, it does not lose a quality it once had in order to gain a new one. It has a quality of the same nature throughout the change. What varies is the intensity of the quality. Given that the change is not a change in the nature of the subject’s quality, it is not obvious that this change can be analyzed at the metaphysical level in terms of a subject losing one form and gaining another contrary form in its place. In response to this problem, medieval thinkers proposed a number of different hylomorphic models for analyzing the metaphysics which underlies increase and decrease in qualitative intensity. As already mentioned, it was not until after Aquinas’s death that some of the most historically significant aspects of the debate took place. Yet, the main solutions, which would later be refined, were already known to Aquinas. In what follows, I will briefly outline the three main solutions and then I will discuss Aquinas’s critique of them.¹⁷ The first theory is the “replacement” or “succession of forms” theory. This view comes the closest to upholding the Aristotelian paradigm of change: according to the replacement or succession model, changes of intension and remission do indeed involve a subject losing one form and gaining another. Both forms involved in the change are of the same species, yet they differ in the degree of intensity which they confer on the subject. According to this view, it is not the case that there are as many forms as there are species or natures of qualities. Rather, there are as many forms as there are determinate degrees of intensity within the species of quality that admit of intension and remission. For example, there is not merely a single form of whiteness which a substance can gain

¹⁶ Aristotle, Physics I.7.190a13–191a22. ¹⁷ In addition to discussing the prominent medieval solutions, Aquinas also considers various ancient theories of intensification in ST I-II.52.1.1.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

       

121

or lose. Rather, there is a series of forms of whiteness, each of which differs from the other in virtue of conferring a different degree of intensity on its subject. A substance will successively instantiate several of these forms in a change of whitening. The “succession” model preserves the Aristotelian paradigm of change in so far as it maintains that changes of intension and remission involve a subject losing one form and gaining another. Yet, it departs from the Aristotelian paradigm in so far as the forms which are gained and lost are not of contrary species. It appears that Aquinas’s teacher, Albert the Great, held the “succession” theory.¹⁸ Yet, Aquinas himself rejects the theory because he thinks that it cannot provide an explanation for why in a change of intensification the previous form is destroyed when the greater form of the same nature arrives.¹⁹ For example, the theory cannot explain why the form of whiteness-to-degree n0 is destroyed when the form of whiteness-to-degree n0 +¹ is introduced into the subject. Aquinas presupposes that forms can only be destroyed by introducing a contrary form into the subject or by destroying the subject itself. Consider, for example, a curved iron rod. To destroy the form of curvedness, an agent must either destroy the rod itself or straighten the rod which just is to cause a form contrary to curvedness in the rod. On the replacement model, the two forms involved in the change of intensification are of the same nature. They are not forms of contrary natures. So, Aquinas supposes that the introduction of the new form would not entail the destruction of the old form. Thus, in his estimation this model fails. The next model, the “addition-to-subject” model, departs even farther from the Aristotelian paradigm of change. On the “addition-to-subject” model, substances do indeed gain a form in accidental changes of intensification. However, in contrast with the Aristotelian paradigm, subjects do not lose a form when they undergo such a change. On this

¹⁸ See Albertus Magnus, Liber de sex principiis, tr. 1 c. 3, p. 310. This position has been attributed to Aquinas’s later follower Godfrey of Fontaines. See Stephen Dumont, “Godfrey of Fontaines and the Succession Theory of Forms at Paris in the Early Fourteenth Century,” in S. Brown, T. Dewender, and T. Kobusch (eds.), Philosophical Debates at the University of Paris in the Early Fourteenth Century (Leiden: Brill, 2009), 39–125. For a different interpretation of Godfrey’s view see John Wippel, “Godfrey of Fontaines on Intension and Remission of Accidental Forms,” Franciscan Studies 39 (1979), 316–55. ¹⁹ I Sent. 17.2.1 (Mandonnet-Moos 1, p. 411). Aquinas never discusses the “replacement” view in his later texts.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

122   view, intensification happens by adding another accidental form of the same nature to the subject.²⁰ The form which is added is of the same nature as the form which the subject already has, yet it is numerically distinct from the one that already exists in the subject. According to this model, it is possible for a subject to have more than one, and in fact a great number, of accidental forms all of which are of the same nature. Intensification involves a quantitative increase in subject’s number of forms. The more accidental forms of F-ness which a subject S has, the more intensely S exists as F. For instance, a hotter pot of water has more forms of heat than a less hot pot of water. Aquinas rejects the “addition-to-subject” model because it presupposes that forms of the same nature can be numerically individuated of themselves. For example, the theory maintains that there can be a numerically distinct form of heat which is such that it (i) is not the heat of any other substance and (ii) is able to be added to an already hot pot of water to make it hotter. In Aquinas’s view, this supposition is incoherent. Aquinas writes: “A duality of forms of the same species is not possible without there being two subjects. For forms of the same species are not diversified in number except through a subject.”²¹ Forms are natures of substances and what it is for a form to exist is for a substance to exist through that nature. The only way to get two numerically distinct forms of the same kind is for two different substances to exist through the same form or nature. The final theory is the “addition to the form” theory. This theory perhaps departs from the Aristotelian paradigm of change even more so than the previous two. On this theory, in changes of intensification, an accidental form is neither gained nor lost. Rather, this theory postulates that the accidental forms which admit of intensification are quantitatively divisible into separable parts. The theory maintains that a subject can be informed by an accidental form without being actualized by all of its parts. In a change of intensification, a substance gains a part of a form which it previously lacked. Accordingly, a change of intensification involves a quantitative increase in a subject’s number of form-parts.

²⁰ According to Dumont, “Godfrey of Fontaines,” p. 40, addition theories were favored among thirteenth-century Franciscans, including Bonaventure. ²¹ De virt. 1.11 (Marietti, p. 523). See also ST I-II.52.2.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

       

123

What is especially significant about this theory is that it implies that the form in the subject varies in itself. The form acquires a new part which it previously lacked through the process of intensification. For instance, in the case of two pots of water, one which is hotter than the other, there is a difference between their forms of heat. The form of heat in the hotter pot has more parts than the form of heat in the less hot pot. The less hot pot’s form of heat is lacking parts which are present in the form of the hotter pot and if it were to be heated more it would gain some of these parts in the process. Aquinas thinks that the “addition-in-form” theory goes wrong in so far as it requires that forms themselves vary.²² Aquinas thinks that any variation in a form would amount to a change of its species. Aquinas claims, for instance, that adding to or subtracting from a substance’s form of paleness will not give rise to a more or less intense paleness; rather it will give rise to black or white, which is an altogether different species of color.²³ In his Disputed Questions on the Virtues he writes: It is necessary to understand the specific principle of something as indivisible. For differences of this kind of principle vary the species. If this principle were added to or subtracted from, the species would necessarily be varied. Accordingly the Philosopher says in Metaphysics 8 that the species of things are like numbers in so far as an added or subtracted unit varies the species.²⁴

In Aquinas’s view, a form is that through which a substance has a quality of a certain nature. A difference of form entails a difference in the specific kind of quality which the subject has. Now that we have seen Aquinas’s reasons for rejecting three important medieval analyses of qualitative increase, we can begin to see how Aquinas himself approaches the problem. In order to appreciate the philosophical significance of Aquinas’s approach to intension, it is helpful to begin by noting a fundamental similarity between the three theories already discussed: all three theories

²² ST I-II.52.1. ²³ ST I-II.52.2 (Leonine 6, p. 334). ²⁴ De virt. 5.3 (Marietti, p. 618). See also ScG III.97 and ScG IV.14. In the latter text Aquinas attributes the view that “forms are like numbers” to Plato as well. On Aquinas’s claim that forms are like numbers, see Rudi te Velde, Participation and Substantiality in Thomas Aquinas (Leiden: Brill, 1995), 227–30.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

124   assume that a change in the degree of a quality’s intensity entails a difference at the level of a substance’s forms. The theories disagree about how to ontologically analyze the formal difference: on the succession view, intensification happens through a new form replacing an old one. On the “addition-to-subject theory,” a form is added to the subject without it losing a form. On the “addition-to-form theory,” a form-part is added to a form in the subject. Thus, on all three theories, a change of intensification entails a modification at the level of form. All three of the theories assume that forms or form-parts of themselves are determined to confer a definite degree of intensity on their subject. Thus, the manner in which a subject’s quality increases is by the subject gaining a new form or form-part. Aquinas, on the other hand, rejects the assumption that formal elements of themselves determine the degree of intensity of the qualities which they give rise to when they inhere in a subject. In his view, accidental forms are not of themselves determined to confer a definite degree of qualitative intensity. Rather, the subject’s determinate degree of qualitative intensity at any given moment is determined by a variable relationship between the subject and an inherent accidental form. Aquinas refers to this relationship as “participation.”

2. Aquinas’s Participation Solution In Aquinas’s mature works, the notion of “participation” is prominent in his account of qualitative increase.²⁵ Consider these textual examples: De virt. 1.11: “For a quality to be increased is nothing other than for the subject to participate more in the quality; for a quality has no existence except for in a subject.”²⁶

²⁵ In his earliest discussion of intension, Aquinas does not appeal to “participation” as he does in later works. Rather, he claims that intensification happens in virtue of a form reducing its subject more in act. Cf. I Sent. 17.2.2 co. (Mandonnet-Moos 1, p. 416). In his later De virt. 1.11, Aquinas identifies these two views. See Marietti, p. 524. Between his earliest discussions of intension in the Sentences and his later treatments of the topic in the Summa Theologiae and the Disputed Questions on the Virtues, Aquinas read the ancient philosopher Simplicius’s commentary on Aristotle’s Categories and this text greatly influenced him in how he framed the questions associated with qualitative increase. For the relevant passage in Simplicius see his Commentaire sur les Catégories ďAristóte, vol. 2, 392. ²⁶ De virt. 1.11 (Marietti, pp. 523–4).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

       

125

ST I-II.52.2: “The increase of habits and other forms of this kind does not happen through an addition of form to form, rather it happens through the subject participating more or less perfectly in one and the same form . . . ”²⁷ ST I-II.52.2 ad 2: “[T]he cause which increases a habit always effects something new in a subject, however not a new form. Rather it makes the subject participate more perfectly in a pre-existing form . . . ”²⁸ Aquinas similarly appeals to participation in other contexts as well. In his Physics commentary, which is also a mature work, he explains motion in the following way: When a subject participates in something of heat but imperfectly, then it begins to be moved with respect to heat, for that which is being heated, gradually participates more and more in heat.²⁹

Though Aquinas repeatedly refers to a subject’s “participation” in its accidental forms, I suspect that many of his recent readers suppose that “participation” is merely a figure of speech.³⁰ On this interpretation “participation” is just a metaphorical way of speaking about which forms or form-parts inhere in a subject and differences in how a subject “participates in a form” reduce to differences in which forms or form-parts inhere in the subject. On this interpretation, Aquinas would thus need to appeal to one of the theories he rejected in order to interpret the literal meaning of “participating more and less.” Accordingly appeal to “participation” would be merely a verbal way of putting off the real ontological complexities of intension.³¹ It is worth noting that others employed the notion of “participation” to explain intension long before Aquinas. Boethius, for example, similarly accounted for qualitative increase and decrease in terms of a variable ²⁷ ST I-II.52.2 (Leonine 6, p. 334). ²⁸ ST I-II.52.2 ad 2 (Leonine 6, p. 334). ²⁹ Ibid. ³⁰ This seems to be how Pierre Conway, OP interprets “participation” in his translation of Aquinas’s Commentary on the Physics which is available here: He translates the phrase “participat calorem magis ac magis” as “gradually acquires heat step by step.” ³¹ Along these lines, Richard Cross writes: “Perhaps in any case we should not regard participation as anything more than a metaphor. And in this case, any explanatory value it might have would have to be spelled out in terms of an appeal to those real features of a thing which admit of non-metaphorical description” (The Physics of Duns Scotus [Oxford: Clarendon Press, 1998], p. 183).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

126   participation between a subject and a form.³² Boethius, like other Platonists, accepted a transcendent realm of forms which existed separate from material particulars.³³ Material particulars possessed their qualities in virtue of participating in these transcendent forms. The particular qualities which inhered in the individuals were more or less intense in virtue of participating more or less perfectly in the corresponding transcendent form. Aquinas, on the other hand, following Aristotle, denied that forms existed separate from material particulars. In his view forms are particulars which inhere in material substances.³⁴ Aquinas’s ontology of form makes it difficult to see how a “participation” theory could do ontological work for him. If there are no transcendent forms, it seems that the source of perfection which is participated in collapses into the received perfection which is the individual’s quality. The goal of this section is to unpack some of the relevant details of Aquinas’s ontology of accidental qualities in order to illuminate his understanding of how substances participate in their inherent accidental forms. First, I will discuss some of Aquinas’s basic commitments regarding accidental forms and qualities. Next, I will show how these commitments enabled him to interpret the relationship between a substance and its inherent accidental forms as “participation.”

a. Aquinas on Accidental Forms and Accidental Qualities There are three elements of Aquinas’s views on the metaphysics of qualities which are crucial for understanding his participation account of intensification. In this section I will show that Aquinas held that for every quality such as whiteness and courage, there is: (1) a nature or form (2) a receiver of the form (i.e. a substance) (3) a way in which the receiving substance exists through the form ³² Boethius, In Categorias Aristotelis, lib. III, p. 257. Others have noted that Aquinas is following Boethius in adopting the “participation” solution to qualitative increase. See, for example, Jean-Luc Solère, “Les variations qualitatives,” at p. 161. ³³ See Richard Cross, “Form and Universal in Boethius,” British Journal for the History of Philosophy 20 (2012), 439–58. ³⁴ For Aquinas’s rejection of Platonic forms see for instance ST I.6.4 and ST I.84.5.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

       

127

I will begin by summarizing Aquinas’s views on forms and then I will segue to discussing these other two aspects. Forms are principles through which a substance has a certain sort of intrinsic character. For example, it is through the inherence of the form of whiteness that a material substance is white, rather than green or orange. In Aquinas’s view, forms are particulars. Each white object has its own inherent numerically distinct form of whiteness.³⁵ Though forms are particulars, Aquinas denies that forms are numerically individuated of themselves. We saw above that Aquinas rejected the “addition-to-subject” view because it supposed that they were. In Aquinas’s view, forms receive their numerical individuation through inhering in numerically distinct substances.³⁶ Each numerically distinct form of the same species is an exact duplicate of every other. Put otherwise, each accidental form of the same kind is internally the same as every other.³⁷ Aquinas explicitly claims that forms of the same species do not vary even according to greater or lesser perfection. He writes: “[I]t is impossible . . . for a form to be found greater than another. It is not that one whiteness considered in itself is a greater whiteness than another.”³⁸ Forms of the same species must be exactly alike because, as we saw above, Aquinas thinks that any difference in a form changes its species.³⁹ According to Aquinas, the perfect resemblance which obtains between forms of the same species is what allows the mind to form a concept which is common to all individuals of the same kind on the basis of considering the form of just one individual.⁴⁰ For instance, by considering the form of one white object, the intellect can have a concept which applies to all white objects.⁴¹ Though each accidental form within a species is internally the same as every other, Aquinas maintains that there can be differences in how

³⁵ See, for instance, In II Sent. 17.1.1 co. and In I Sent. 19.5.1 co. ³⁶ See, for instance, In II de Anim. 12.5 and note 21 above. ³⁷ “Internal sameness” is a term coined by Jeffrey Brower in his paper “Aquinas on the Problem of Universals” (Philosophy and Phenomenological Research 92 [2016], 715–35) to describe the sameness that obtains between numerically distinct individuals which are individuated derivatively rather than of themselves. ³⁸ De virt. 5.3 (Marietti, p. 618). ³⁹ See note 24. ⁴⁰ See for instance In II Sent. 17.1.1 co. ⁴¹ On Aquinas’s views on abstraction see Therese Scarpelli Cory, “Rethinking Abstractionism: Aquinas’s Intellectual Light and Some Arabic Sources,” Journal of the History of Philosophy 53 (2015), 607–46.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

128   substances exist through such forms. For example, though all numerically distinct forms of whiteness are internally the same, one white substance might be more perfectly white through its form of whiteness than another. Aquinas writes: “ . . . snow is said to be whiter than a wall because whiteness is more perfectly in snow than in a wall, but nevertheless it is of the same nature (ratio).”⁴² The forms of whiteness in the wall and in snow must be in themselves exactly alike because if there were any difference in their forms, they would not be of the same nature (ratio). Nevertheless one substance is whiter than the other. To make sense of Aquinas’s views that (a) all forms of the same species F are exactly alike and (2) substances can be more or less perfectly F through their inherent forms of F-ness, there must be some real difference between a substance’s accidental form and the way in which the substance is through that form. A “way of existence” is simply how the substance is. It is not an entity over and above the substance itself. The relevant accidental form, however, cannot be reduced to a way in which the substance is.⁴³ A form is a principle of essence or nature which accounts for why a substance exists in a specific way. Accidental forms, in particular, are the grounds for why a substance exists in certain qualitative ways which are not included in its substantial kind essence. This interpretation of Aquinas’s views follows from his claims about how the wall and the snow are white. If the way in which the wall is white differs from the way in which snow is white, and yet each substance’s form of whiteness is exactly alike, then each substance’s accidental form cannot be reduced to the way in which each is white. Furthermore, there is ample textual evidence to support that Aquinas saw accidental forms not as ways of existing, but rather as causes or grounds for why a substance has a certain way of existing. For instance, he writes: “In this way whiteness is said to exist not because it subsists in itself, but because something has existence-as-white by it (habet esse album ea).”⁴⁴ Elsewhere he writes: “ . . . every existence is from some inhering form, just as existing as white is from whiteness, and existing as a substance is from a substantial form. It is unintelligible that a wall be white without [the form ⁴² De pot. 7.7 ad 3. ⁴³ For a “reductive” reading of Aquinas’s views on accidental forms, see Pasnau, “On What There Is in Aquinas,” pp. 20–4. ⁴⁴ Quodl. IX.2.2 (Leonine 25/1, p. 94, ll. 56–8).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

       

129

of] whiteness inhering . . . .”⁴⁵ These passages claim that forms are that “by which” or “from which” a substance has a certain way of existing. At times, Aquinas goes so far as to explicitly claim that existence is an “effect” of form.⁴⁶ These passages suggest that forms are not reducible to ways in which a substance is. Rather forms are principles of essence or nature which explain why a substance exists in a determinate way. For example, the wall exists as a white thing (or to use Aquinas’s language it has esse album) rather than as a green or blue thing because the form whiteness is in it. In other contexts, Aquinas claims that for each specific way that a substance is, there are aspects of how the substance is which are not determined by the content of the relevant form. In the case of the snow and the wall, it is clear that the degree to which each exists as white is not determined by the content of their respective forms since their forms are internally the same and yet they are each white to a different degree of intensity. Aquinas claims in several texts that for every way in which a substance is there is both a “species” and a “mode.” (He also mentions order, but this is not relevant for our purposes.) He writes in the Summa theologiae: “[E]very existence (esse) is according to some form, consequently with respect to every existence (esse) of a thing, there follows a mode, species and order . . . in so far as a man is white, he similarly has mode, species and order.”⁴⁷ At another point in the same Summa theologiae article, Aquinas explains that the species of how a substance exists just is the form. He writes: “The very form is signified by species because each one is constituted in its species through a form.”⁴⁸ For example, consider how the snow exists as a white thing. The species of this way in which the snow is just is the form of whiteness through which it exists as a white thing.

⁴⁵ I Sent. 17.1.1 (Mandonnet-Moos 1, p. 393). ⁴⁶ See for instance ST Ia.42.1 ad 1 (Leonine 4, p. 436). For discussion of Aquinas’s view that “forms give esse,” see John Wippel, “Aquinas on Creatures as Causes of Esse,” Metaphysical Themes in Thomas Aquinas II (Washington, DC: Catholic University of America Press, 2007), 172–93, at pp. 175–9. ⁴⁷ ST I.5.5 ad 3 (Leonine 4, p. 63). In the sed contra of this article, Aquinas cites Augustine as the source for this threefold consideration of each existence according to mode, species, and order. For literature on Augustine, see Christian Schäffer, “Augustine on Mode, Species and Natural Order,” Augustinian Studies 31 (2000), 59–77. ⁴⁸ ST I.5.5 (Leonine 4, p. 63).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

130   In a De veritate passage about species, mode, and order, Aquinas explains that the “mode” of a substance’s existence is determined by the substance in which the form inheres. He writes: “For species pertains to the very nature (ratio) of the species, which since it has being in some subject, is received according to some determinate mode, since everything which exists in a subject is in it through the mode of the receiver . . . .”⁴⁹ He further explains “mode” in a reply to an objection in the same De veritate article: “[W]henever there is something received, it is necessary that there is mode, since what it received is limited according to the mode of the receiver; therefore since both the accidental and essential being of a creature is received, mode is found not only in accidental being but also in substantial.”⁵⁰ In these passages, Aquinas is employing a more general Neo-Platonic axiom which he uses throughout his works: whatever is received is received according to the mode of the receiver.⁵¹ According to this axiom, whenever a perfection is received from a source, the perfection is not in the receiver in the same exact manner as it is in the source. Rather, there are limiting and individuating conditions which arise from the receiver. As we have seen, substances exist in determinate qualitative ways through inherent accidental forms. The wall, for example, is not white through its very self. Rather, it exists as a white thing in virtue of the accidental form of whiteness inhering in it. Since its existence as white is received from the form, the wall’s existence-as-white has some aspects which are from the form of whiteness and others which arise from the wall’s particular features as a receiver of the form. These latter aspects lie outside of the form of whiteness considered in itself and Aquinas refers to these as “modes” of existence. For Aquinas, a difference in mode is a difference which pertains to existence as opposed to form or essence.⁵² The difference between a brown thing and a green thing is a difference of essence or form. The differences between various white things are differences of “mode of existence.”

⁴⁹ De ver. 21.6 (Leonine 22 3/1, p. 609). ⁵⁰ See also De ver. 21.6 ad 5 (Leonine 22, 3/1, p. 610). ⁵¹ For a study on Aquinas’s use of this axiom see Wippel, “Thomas Aquinas and the Axiom ‘What Is Received Is Received According to the Mode of the Receiver,’ ” in Metaphysical Themes in Thomas Aquinas II, 113–22. ⁵² See for instance De ver. 21.1 (Leonine 22.3, pp. 135–6) and De ver. 1.1 (Leonine 22.1, p. 5).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

       

131

In Aquinas’s view, substances can exist in different ways through one and the same form. As Aquinas noted above, it is the receiver which determines how the substance exists through the form. Material vs. immaterial existence is one division among modes of existence.⁵³ The very nature or ratio of whiteness, for example, does not determine whether it actualizes a substance to be corporeally white or whether it actualizes an immaterial thought of whiteness in an intellectual substance. On Aquinas’s theory, the form in the white wall is the same (according to ratio) as the form by which an intellect thinks about whiteness.⁵⁴ What determines whether the form of whiteness is that through which a physical subject exists as corporeally white or that through which an intellect exists as thinking about whiteness is the sort of receiver in which the form of whiteness inheres. If whiteness inheres in a corporeal substance that substance will exist through the form of whiteness as a corporeal white body. Conversely, if the form inheres in an immaterial intellect, the intellectual substance will exist through it as actually thinking about whiteness. The substances themselves determine the mode of existence and they are the subjects which have differing modes of existence through their forms. Aquinas denies that forms are subjects which have a mode of existence.⁵⁵ For example, it is not that the form of whiteness is such that it exists as a material or immaterial form. Rather, it is the substance in which the form inheres that exists as materially white or immaterially white. Aquinas maintains that degree of qualitative intensity is likewise a mode of existence. He writes in his Summa contra gentiles: “[S]omething is more or less white according to the mode by which its whiteness is completed in it.”⁵⁶ He explains that in the case of two white substances which are such that one is whiter than the other, “they share in forms of the same nature (ratio), but not according to the same mode.”⁵⁷

⁵³ On Aquinas’s understanding of “mode” see John Tomarchio’s papers “Aquinas’s Division of Being According to Modes of Existing,” Review of Metaphysics 54 (2001), 585–613; “Four Indices for the Thomistic Principle Quod recipitur in aliquo est in eo per modum recipientis,” Mediaeval Studies 60 (1998), 315–67; “Thomistic Axiomatics in an Age of Computers,” History of Philosophy Quarterly 16 (1999), 249–75. ⁵⁴ See ST I.75.5 for an example of Aquinas’s claims that physical objects are known in virtue of their forms being in the knower. ⁵⁵ In I Sent. 3.2 (Mandonnet-Moos I, p. 104): “ . . . accidents are more modes . . . of substances, rather than that which has a mode . . . .” ⁵⁶ ScG I.43 (Leonine 13, p. 124). ⁵⁷ ST I.4.3 (Leonine 4, p. 53).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

132   He claims that the most perfect similarity obtains between substances which are the same both according to species of form and according to mode of existence.⁵⁸ In Aquinas’s view, accidental forms abstract from a determinate degree of intensive perfection. The degree of perfection with which a substance exists according to a form is not established by the form, but rather by the receiving substance in which the form inheres. In Aquinas’s view, forms only establish the species of how a substance exists. There are other aspects of how a substance is which are determined by features of the substance which are extrinsic to the form. By way of analogy, consider a computer monitor which is displaying a webpage w. The precise way in which the monitor appears at that moment arises from two distinct sources. The intelligible content w explains why the monitor is displaying w rather than w0 . This content is the source of commonality between computer c and all other electronic devices which are displaying w. Yet, c’s particular hue of colors and level of brightness as it displays w arise from its monitor’s own intrinsic settings, numbers of pixels, etc. Thus, the precise way in which c looks is due both to the intelligible content of w and to factors intrinsic to c. Aquinas similarly believes that the precise qualitative way in which a substance exists arises both from the nature or form inhering in it and from factors of the subject in which the form inheres. Degree of intensive perfection is determinated by the substance, rather than the form.

b. How Substances Participate in their Accidental Forms With these elements of Aquinas’s views on the table, we are now in a position to understand why he thought that there was a “participation” relationship between substances and their inherent accidental forms. I will first discuss Aquinas’s understanding of what participation is and then I will show how the various elements of Aquinas’s views on accidental qualities fit within the participation schema. It is well known that Aquinas invokes the relation of “participation” in a variety of contexts. The most scholarly interest has been given to his claim that

⁵⁸ Ibid.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

       

133

creatures participate in God’s being.⁵⁹ Close attention to the texts in which Aquinas discusses “participation” reveal that it is a technical term which designates a specific kind of metaphysical relationship which obtains between a receiving subject, a received perfection and a source from which the perfection is received. Aquinas describes “participation” in the following passages: De subs. separ. ch. 3: “Every one which participates in something receives that which it participates from the one in which it participates: and on account of this, that in which it participates is the cause of it; just as air has participated light from the sun, which is the cause of its illumination.”⁶⁰ In lib. De Causis, lec. 4: “that which is participated is not received in the one participating according to its full infinity, but rather particularly.”⁶¹ ScG I.32: “for everything which is possessed by participation is determined according to the mode of the participator, and so it is had partially and not according to the full mode of perfection.”⁶² In light of these passages, we can see that Aquinas accepts that there are real, metaphysical “participations” and they involve these four conditions:⁶³ 1. A source S which (i) can confer a perfection F on another to a full or unrestricted degree of perfection; and (ii) can do so through itself (i.e. it does not depend on a higher cause for the capacity to confer F on another); 2. A participant P which possesses F in a limited or restricted manner; 3. P’s possession of F in some way depends on S; 4. The restricted or limited manner of P’s possession of F depends on P.

⁵⁹ For an overview of the literature on this topic, see Wippel, Metaphysical Thought, pp. 94–131. ⁶⁰ De sub. sep. 3 (Leonine 40D, p. 46). ⁶¹ In lib. De Causis 4 (Saffrey, p. 30). ⁶² ScG I.32 (Leonine 13, p. 97). ⁶³ I have based the first three criteria with some modifications on W. Norris Clarke’s discussion in Explorations in Metaphysics (South Bend, IN: University of Notre Dame Press, 1994), p. 93. Wippel explains Aquinas’s understanding of participation similarly in his Metaphysical Thought, pp. 96–7.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

134   Aquinas allows for a variety of kinds of entities to fulfill the source and participant roles. In his commentary on Boethius’s De Hebdomadibus, he explicitly claims that the participation relationship obtains between substances and their accidental forms.⁶⁴ In light of what was covered in the last sub-section, we can understand how Aquinas saw the four participation criteria above satisfied in the case of substances and their inherent accidental forms. The substance itself is the participant and accidental form is the source. The determinate way in which the substance exists is the received perfection which the substance has from the accidental form. For instance, existing as a white thing is what the wall receives through the form of whiteness inherent in it. The wall depends on the form of whiteness to be the way it is with respect to color. As we have seen, the form of whiteness abstracts from a determinate degree of intensive perfection. The substance itself determines the particular degree of perfection to which it exists according to the form; and it will always be a limited or imperfect degree. Some further remarks are in order about this last point. In a number of passages, Aquinas makes clear that forms have no limit in the degree of intensive perfection which they can confer on their subjects. The form of whiteness, for instance, could be the essential principle through which a white object exists as white to the highest possible degree. Every white object, however, is less than perfectly white on account of its own reception of the form—not on account of the form. Aquinas writes in his Summa contra Gentiles: Every act which inheres in another is limited by the one in which it exists. This is because whatever exists in another exists in it according to the mode of the receiver. Therefore, an act which exists in nothing has no limitation. If, for example, there were a whiteness which did not exist in any subject, the perfection of whiteness would not be limited in it, rather it would have whatever it is possible to have concerning the perfection of whiteness.⁶⁵

⁶⁴ In de heb. 2.24 (Leonine 50, p. 271): “Similarly also a subject participates in an accident and matter in form because substantial and accidental form, which is of its nature common, is determinated to this or that subject.” On a substance’s participation in accidents, see Wippel’s Metaphysical Thought, pp. 86, 98. ⁶⁵ ScG I.43 (Leonine 13, p. 124). See also In de div. nom. (p. 234, n. 629).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

       

135

Aquinas states that if there were a whiteness which was not the whiteness of any subject it would be as perfect as possible. The reason why no white substance is perfectly white is not on account of the essential principle or nature through which it is white, but rather on account of the substance which exists as white through the form of whiteness. Elsewhere in the Summa contra Gentiles Aquinas explains that “defects” in the receiver prevent subjects from existing with full perfection. He writes: Just as if there were some separate whiteness, nothing concerning the excellence of whiteness would be lacking from it. For any white thing lacks something of the excellence of whiteness because of a defect in the receiver of whiteness, which receives it according to its mode, and perhaps not according to the whole power (secundum totum posse) of whiteness.⁶⁶

This passage makes clear that the form in itself can actualize a maximal degree of intensive perfection in its subject. To be clear, the form which is the source of participation inherent in an individual material substance is not infinite in its degree of perfection. Rather, its content abstracts from a degree of perfection. No form of whiteness, for instance, includes within it a determinate degree of perfection. This is why all white objects, regardless of how perfectly white each is, can have forms of whiteness that are exactly alike. The sense in which the form of whiteness is unlimited in its perfection and entirely perfect is that of itself it can be that in virtue of which an object has existence-as-white with the highest possible degree of perfection. The form does not cause the existence which is had through it to be limited and imperfect. The reason why every subject which exists through the form exists imperfectly is because each subject has features which prevent it from existing through the form as perfectly as possible. In this regard, receiving subjects are the cause of imperfection. Aquinas’s view that the receiving subject determines or limits the extent to which the form perfects it represents a departure from Aristotle. Aristotle, like his Greek predecessors, conceived of determinacy and limitation as a mark of perfection and indeterminacy and lack of ⁶⁶ ScG I.28 (Leonine 13, p. 87). See also Quodl. III 1.1 (Leonine 25/2, p. 241, ll. 24).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

136   limitation as imperfections.⁶⁷ Finitude and limitation were linked with perfection because to be perfect was seen as being complete and completion required a definite limit. That which was infinite or unlimited was seen as imperfect because it was unending, indefinite, and unknowable.⁶⁸ Since, according to hylomorphism, form is that which perfects matter by actualizing it, Aristotle conceived of form as a limiting principle which restricts or contains the amorphous, indefinite infinity of matter. In his Physics, he writes: “For the matter and the infinite are contained inside what contains them, while it is the form which contains.”⁶⁹ In Aristotle’s view, matter does not limit or contract the perfection which the subject receives from the form. Rather, form limits and contracts matter. Aquinas, by contrast, assigns the receiving subject (which is the matter for accidental forms) the function of limiting and determining the degree of perfection to which the form actualizes or perfects the subject. Aquinas did not see an accidental form’s indeterminacy or lack of limit with regard to a degree of intensive perfection as an imperfection. Through his reading of Boethius’s De Hebdomadibus, works of PseudoDionysius such as De Divinis Nominibus, and the influential Liber de Causis, Aquinas encountered Neo-Platonism’s reversal of the earlier Greek correlations between limitation and perfection, on the one hand, and lack of limitation and imperfection on the other hand.⁷⁰ In the Neo-Platonic system which traces back to Plotinus, limitation is seen as degradation and the unmeasurableness of infinity is viewed as perfection.⁷¹ Aquinas transposes this reversal onto the Aristotelian notions of matter and form: Form as perfect is linked with indeterminacy and lack of limitation, while matter as imperfect is conceived of as a finite principle of limitation which determines the extent to which the subject ⁶⁷ On the concept of infinity in ancient and medieval thought see Norman Kretzmann (ed.), Infinity and Continuity in Ancient and Medieval Thought (Ithaca, NY: Cornell University Press, 1998); Adam Drozdek, In the Beginning Was the Apeiron: Infinity in Greek Philosophy (Ann Arbor, MI: University of Michigan Press, 2008); Leo Sweeney, Divine Infinity in Greek and Medieval Thought (New York: P. Lang, 1998). ⁶⁸ Aristotle, De Generatione Animalium 1.1.715b14: “Nature flees from the infinite, for the infinite is unending or imperfect, and nature ever seeks an end” (Basic Works of Aristotle, ed. R. McKeon [New York: Random House, 1941], p. 666). ⁶⁹ Aristotle, Physics 3.6.207a30–7 (Basic Works, p. 267). ⁷⁰ For a recent summary of how and what Aquinas knew of Platonism see Wayne Hankey, “Aquinas, Plato, and Neo-Platonism,” in Stump and Davies (eds.), The Oxford Handbook of Aquinas, 55–64. ⁷¹ See, for instance, Plotinus, Enneads IV.3, 8; V.5, 4–6 and 9–11.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

       

137

is perfected by the form.⁷² He writes in the Summa theologiae: “ . . . the fullness of form is contracted by matter. Hence the infinity which belongs to a form not determined by matter has the nature of perfection.”⁷³ It is on account of matter’s limiting function that no substance exists through a form with an infinite degree of perfection. Form of itself, however, is not determined to confer existence according to a limited degree of perfection.⁷⁴ The three positions on intensification which Aquinas rejects differ from him on this fundamental point. Those views, as we have seen, presuppose that forms (and form-parts) are of themselves determined to confer a limited, determinate degree of intensive perfection on their subject. The receiving subject plays no role in determining the degree of intensive perfection which the subject has through the form. Forms are already of themselves determined to confer a particular degree of perfection. To increase in intensive perfection, the subject must acquire something new at the formal level (either a form or form-part). In Aquinas’s view, by contrast, forms themselves are undetermined in how much intensive perfection they confer on their subjects. A single form of itself can confer a maximal degree of perfection. In the next section, we will see the factors or causes that determine the degree of perfection to which a given subject exists through a form.

3. Causes of Changes in Degree of Participation in a Form The goal of this section is to examine Aquinas’s thinking about the causes which account for the degree to which a substance participates ⁷² Fran O’Rourke claims that Pseudo-Dionysius was an influential source in Aquinas’s understanding of intensive perfection. See his Pseudo-Dionysius and the Metaphysics of Aquinas (Leiden: Brill, 1992), 156–87. Cornelio Fabro had previously identified PseudoDionysius as the source for Aquinas’s understanding of the intensive perfection of esse. See his Participation et causalité selon s. Thomas d’Aquin (Louvain: Publications universitaires de Louvain, 1961), p. 229. For an example of a passage in which Aquinas discusses intensive perfection see De ver. 2.9 (Leonine 22 1/2, p. 73). ⁷³ ST Ia.7.1 (Leonine 4, p. 72). On Aquinas’s notion of infinity see John Tomarchio, “Aquinas’s Concept of Infinity,” Journal of the History of Philosophy 40 (2002), 163–87. ⁷⁴ While Aquinas goes beyond Aristotle by positing that matter limits the perfection of a form, he still retains the Aristotelian idea that matter is contracted by form to a determinate species. See for instance ST Ia.44.2 (Leonine 4, p. 458).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

138   in a form and the causes of changes in degree of participation. What explains, for example, why the wall participates in whiteness less perfectly than snow and why Plato participates more perfectly in his whiteness at time t2 as opposed to at t1? Later figures, such as William of Ockham and John Duns Scotus, objected to the participation theory on the grounds that it implied that a change could occur without the patient acquiring any new form or form-part. Scotus for example writes: “In qualitative increase there is some change . . . therefore it is necessary to posit in the term of change some new reality which was not there before.”⁷⁵ Aquinas, by contrast, thought that some changes consist of a subject’s coming to exist more or less perfectly in an already inherent form. First, we will see how Aquinas thought such changes happen; and then we will be in position to draw some conclusions about Aquinas’s views on the metaphysics of change. In the following passage, Aquinas describes how changes of intensification happen. He writes: Primary and simple qualities are intensified by their causes, i.e. by the agent and the recipient. The agent intends to reduce the patient from potency to the act which is its likeness in so far as possible. Just as that which is not hot is in potency to heat; so too is that which is less hot in potency to becoming hotter. Just as through the potency of heat that which is not hot is made hot . . . so too is something made hotter through the action of heat . . . And this happens according to the potency of the subject of the act which, insofar as it is in itself, is open to being more or less actualized (terminatur) by its act; either because the power of the agent is augmented, just as when light is intensified by the action of many luminaries; or on account of the matter of the patient in so far as it is made more susceptible to its act, just as when air, in so far as it more rarified, is made more susceptible to light.⁷⁶ ⁷⁵ Scotus, I Lect. 17.2.3 (Vat. XVII, p. 248, n. 208). See also Ockham, I Sent. 17.6 (Opera theol. III, p. 506, ll. 8–16). ⁷⁶ I Sent 17.2.2 (Mandonnet-Moos 1, pp. 415–16). Note that in this passage Aquinas is talking about how primary and simple qualities are intensified. This passage occurs just after a claim that secondary and composite qualities are intensified in virtue of the intensification of primary and simple qualities. On medieval theories of primary and secondary qualities see Anneliese Maier, “The Theory of the Elements and the Problem of Their Presence in Compounds,” tr. S. Sargent in On the Threshold of Exact Science (Philadelphia: University of

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

       

139

There are a lot of details in this passage which need to be unpacked. On the Aristotelian model of change, whenever a substance goes from not having a form to having it, two explanatory factors are involved: (1) a prior potential in the patient for receiving the form and (2) an agent with an active power to actualize the form. Aquinas appeals to these two factors to explain changes in degree of participation. He states in the passage above that there are two ways in which a change of intensification can be caused: First, a quality can be increased in a patient in virtue of an increase in the agent cause’s power. Second, a quality can be increased in the patient in virtue of an increase in the patient’s potential or capacity to be actualized by the quality. As this implies, Aquinas thought that both an agent’s active power and a patient’s potentiality admitted of degrees. Regarding an agent’s power and how it admits of degrees, Aquinas thought that a substance’s active power to cause F-ness is identical with the accidental form by which the substance itself exists as F.⁷⁷ A fire’s power to heat, for instance, is identical with the form of heat through which it is hot.⁷⁸ The degree of a substance’s power corresponds to the degree to which the substance participates in the relevant form. Aquinas writes: “A subject acts more intensely because it participates more in a quality; because each agent acts in so far as it is in act. Thus, the more perfectly a subject is reduced to act, the more perfectly it acts.”⁷⁹ For instance, if substance A is hotter than substance B (i.e. A participates more perfectly in heat than B does), then A will have a greater active power to heat than B does.

Pennsylvania Press, 1982), 124–42; and Robert Pasnau, “Scholastic Qualities, Primary and Secondary,” in L. Nolan (ed.), Primary and Secondary Qualities: The Historical and Ongoing Debate (New York: Oxford University Press), 41–61. ⁷⁷ In the case of God, Aquinas believes that power is identical with God’s substance. In all created agents, however, powers are accidents. See for instance ST Ia.77.1 co. Unlike many later medieval philosophers, Aquinas maintained that a substance’s substantial form was distinct from its powers. ⁷⁸ See, for instance, De ver. 27.3 ad 25 (Leonine 22 3/1, p. 801, ll. 522–3). ⁷⁹ De virt 1.11 (Marietti, p. 524). See also ST Ia.25.1 co. and ScG II.8 n.1. Aquinas’s principle that “everything acts in so far as it is in act” has Neo-Platonic origins. Aquinas supports the principle with the premises that “act is perfection” and the Neo-Platonic idea that “perfection is by nature diffusive” (or communicative). For relevant literature see Julien Péghaire, “L’axiome ‘bonum est diffusivum sui’ dans le néoplatonisme et le thomisme,” Revue de l’Université d’Ottawa 1 (1932), 5–30.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

140   Regarding a patient’s potential, Aquinas posits that substances not only have potential for acquiring determinate forms, but also for determinate degrees of participation in forms. He writes: [T]hat which is not already white is in potency to the form of whiteness, since it does not yet have that form: thus the agent [of the change] causes a new form in the subject. But that which is less hot or less white is not in potency to a form since it already has a form in act, but it is in potency to a more perfect mode of participation; and this it receives through the agent’s action.⁸⁰

For example, a piece of wood has not only the potential to be heated, but also the potential to be hot to a particular degree. Aquinas thought that even after a subject acquired a form, it could retain potential to exist through that form more or less perfectly. Substances which are F in actuality could vary in their level of potential for participating more or less in F-ness. For instance, two hot substances could be such that one has more potential than the other to be hotter. Furthermore, one and the same substance can vary over time in its degree of potential for participating more in a form. With this background in place we can more fully understand Aquinas’s thinking about how changes of intension happen. The first way an agent can cause an increase in a subject’s participation in a form is by applying a power to the subject which is stronger than the power of the agent which originally caused the form in it. Aquinas thought that agents cause likenesses of their qualities in their patients to the greatest extent possible.⁸¹ If nothing interferes, an agent will cause its patient to have the same quality as it has and to the same degree. If the degree of the agent’s active power to cause a form is not as great as the degree of the patient’s potential to receive the form, then even after the agent acts on it, the patient will still have unactualized potential for participating more in the form.⁸² In such a case, the patient’s quality

⁸⁰ ST I-II.52.2 ad 3 (Leonine 6, p. 334). ⁸¹ See, for instance, I Sent 17.2.2 quoted above. ⁸² This aspect of Aquinas’s view seems to go against one of the criteria of “participation” relationships, as well as the Neo-Platonic axiom everything is received according to the mode of the receiver. As seen above, an essential aspect of a participation relationship is that the limited mode of the participant’s perfection arises because of a limitation in the participant. In cases in which the limitation in the patient’s perfection occurs on account of a deficiency in the agent cause’s power, the limitation is not coming from the participant. Furthermore,

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

       

141

can be intensified in virtue of the action of an agent (or group of agents) with a stronger power. The second way in which a change of intensification can happen is in virtue of an increase to the patient’s potential for participating in a form.⁸³ Aquinas thought that the prior ways in which a subject is actualized make it more or less susceptible to becoming actual in other ways. If a piece of wood is cold or wet, for example, it has a lesser degree of potentiality for the form of heat than another dry, lukewarm piece of wood. According to Aquinas’s Aristotelian view of sensible qualities, there are two types of relationships which can obtain between different kinds of qualitative forms. Forms which have a relationship of contrariety are such that the degree to which a subject exists according to one form limits the degree to which the subject can exist according to the other. Forms which have a dispositional relationship to one another are such that one form disposes or enables the subject to receive the other more perfectly. Thus the more that the subject is actualized by one quality, the more perfectly it can be actualized by the other.⁸⁴ Because of the relationships of disposition and contrariety between accidental qualities, there are two ways an agent can increase a patient’s potential for a form F-ness: (1) an agent can act on a patient to remove forms contrary to F-ness or (2) the agent can actualize forms in the patient which dispose the subject for receiving F-ness. To illustrate this way of causing a change of intensification, Aquinas provides the example of an agent which causes air to participate more perfectly in light by purifying it of vapors which impede its reception of light.⁸⁵ Similarly, the form of dryness can be intensified in wood by heating it.⁸⁶ the intrinsic “mode” of the subject is not what determines the “mode” according to which it has the received form. ⁸³ Aquinas mentions this way of causing a change of intension or remission again in ST Ia.48.4 (Leonine 4, p. 494). ⁸⁴ Aquinas explains in the following text that a subject’s reception of an accident is mediated by other accidents: De virt. 1.3 (Marietti, pp. 494–5). Aquinas likewise thought that there were relationships of disposition and contrariety between the forms that perfected the immaterial powers of intellect and will. He claims that a subject’s natural dispositions, other habits, and gifts of grace all affect the degree to which he is able to receive the form of any particular habit. See ST I-II 66.1. ⁸⁵ I Sent.17.2.4 (Mandonnet-Moos 1, p. 423). ⁸⁶ Later figures criticized the participation theory because they thought it explained a subject’s degree of participation in a form solely in terms of how the subject participates in

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

142   From this section, we can conclude about Aquinas’s views on change in general that for him what is essential to change is only the actualization of potential. He explicitly maintains that changes can happen without a subject’s gaining a new form.⁸⁷ Aquinas is able to separate the actualization of potential from the acquisition of a new form because, in his view, substances not only have potential for acquiring determinate forms, but they also have potential for varying degrees of participating in forms more or less perfectly. One and the same form can actualize a subject’s potential for participating in it (i.e. existing according to it) to varying degrees. Thus, a new form is not required for there to be a difference in how a substance’s potential is actualized.

4. Conclusion The goal of this study has been to bring greater clarity to Aquinas’s views on the intension and remission of forms. As explained in the introduction, Aquinas’s views on this topic have far-reaching applications for understanding his approach to other philosophical and theological problems. Moreover, his approach to this topic reveals crucial nuances in his views about accidental qualities and accidental change. In Aquinas’s view, qualities involve both an accidental form and a way in which the substance is through that form. The accidental form establishes the species or nature of a how a substance is. Yet, for every way in which a substance exists there are also modes of existence determined by the subject’s reception of the form rather than the form considered in itself. Aquinas claims that degree of qualitative intensity is a mode of existence. Regarding Aquinas’s conception of change, his account of intension and remission reveals that he thought that it was possible for a subject to undergo a change without gaining or losing any form. For

other forms. They worried that this would lead to infinite regress or else offer no explanation of the degree to which the subject participated in the most basic forms. See Scotus, I Ord. 17.2.2 (Vat. V, pp. 256–8, n. 45); Godfrey, Quodl. 2.10, pp. 142–3. Aquinas’s theory does not run into this problem since he also appeals to the degree of the agent cause’s power to explain the degree to which substances participate in their inherent forms. ⁸⁷ ST I-II.52.2 (Leonine 6, p. 334).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

       

143

him, some changes consist merely in a subject existing more or less perfectly through an already inherent form. For Aquinas what is essential to change is only the actualization of potentiality and one and the same form can actualize its subject’s potential to varying degrees.⁸⁸ University of St. Thomas (St. Paul, MN)

Bibliography Primary Sources Albertus Magnus. Liber de sex principiis, ed. A. Borgnet (Paris: Vivès, 1890). Aristotle. Basic Works of Aristotle, ed. Richard McKeon (New York: Random House, 1941). Boethius. In Categorias Aristotelis, ed. J. P. Migne, Patrologiae Latina 64 (Paris, 1847). Godfrey of Fontaines. Les quatre premiers Quodlibets de Godefroid de Fontaines, ed. de Wulf and Pelzer (Louvain: Institut supérieur de Philosophie de l’Université, 1904). John Duns Scotus. Opera omnia, ed. C. Balić et al., 21 vols. (Vatican City: Typis Polyglottis Vaticanis, 1950–2013). Plotinus. Ennéades, ed. E. Bréhier (Paris: Les Belles lettres, 1924–38). Simplicius. Commentaire sur les Catégories ď Aristóte. Traduction de Guillaume de Moerbeke, ed. A. Pattin, vol. 2 (Leiden: Brill, 1975). Thomas Aquinas. Sancti Thomae Aquinatis, Doctoris Angelici, opera omnia, iussu impensaque Leonis XIII P.M. edita, (Rome: S.C. de Propaganda Fide, 1882–). Thomas Aquinas. Quaestiones disputatae de virtutibus, 5th ed. (Rome: Marietti, 1922). Thomas Aquinas. Scriptum super libros sententiarum magistri Petri Lombardi, ed. R. P. Mandonnet and R. P. Maria Fabianus Moos, 4 vols. (Paris: Lethielleux, 1929–47). ⁸⁸ Earlier ancestors of this paper were presented at the University of St. Thomas philosophy colloquium and the Society for Thomistic Personalism session at the 2014 ACPA meeting. I am grateful to those audiences and several other people for helpful comments and discussions of ideas in this paper. I would particularly like to thank anonymous referees, Jeffrey Brower, Brandom Dahm, Stephen Dumont, Can Löwe, and Bob Pasnau.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

144   Thomas Aquinas, In librum Beati Dionysii De divinis nominibus expositio, ed. C. Pera, P. Caramello, and C. Mazzantini (Rome: Marietti, 1950). Thomas Aquinas. Super librum De Causis expositio, ed. H. D. Saffrey (Fribourg: Société Philosophique, 1954). William of Ockham. Opera Theologica (St. Bonaventure, NY: Franciscan Institute, 1967–86). Secondary Sources Brower, Jeffrey. “Matter, Form, and Individuation,” in E. Stump and B. Davies (eds.), The Oxford Handbook of Aquinas (Oxford: Oxford University Press, 2012), 85–103. Brower, Jeffrey. Ontology of the Material World: Change, Hylomorphism, and Material Objects (Oxford: Oxford University Press, 2014). Brower, Jeffrey. “Aquinas on the Problem of Universals,” Philosophy and Phenomenological Research 92 (2016), 715–35. Brown, Barry. Accidental Esse (New York: University Press of America, 1985). Clarke, W. Norris. Explorations in Metaphysics (South Bend, IN: University of Notre Dame Press, 1994). Conolly, Brian Francis. “Dietrich of Freiburg on the Succession of Forms in the Intensification of Qualities,” Recherches de theólogie et philosophie médiévales 81 (2014), 1–35. Cory, Therese Scarpelli. “Rethinking Abstractionism: Aquinas’s Intellectual Light and Some Arabic Sources,” Journal of the History of Philosophy 53 (2015), 607–46. Cross, Richard. The Physics of Duns Scotus (Oxford: Clarendon Press, 1998). Cross, Richard. “Form and Universal in Boethius,” British Journal for the History of Philosophy 20 (2012), 439–58. Decean, Christopher. “The Viability of Aristotelian–Thomistic Color Realism,” The Thomist 65 (2001), 179–222. Drozdek, Adam. In the Beginning Was the Apeiron: Infinity in Greek Philosophy (Ann Arbor, MI: University of Michigan Press, 2008). Dumont, Stephen. “Godfrey of Fontaines and the Succession Theory of Forms at Paris in the Early Fourteenth Century,” in S. Brown, T. Dewender, and T. Kobusch (eds.), Philosophical Debates at the University of Paris in the Early Fourteenth Century (Leiden: Brill, 2009), 39–125.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

       

145

Fabro, Cornelio. Participation et causalité selon s. Thomas d’Aquin (Louvain: Publications universitaires de Louvain, 1961). Hankey, Wayne. “Aquinas, Plato, and Neo-Platonism,” in E. Stump and B. Davies (eds.), The Oxford Handbook of Aquinas (Oxford: Oxford University Press, 2012), 55–64. Kretzmann, Norman (ed.). Infinity and Continuity in Ancient and Medieval Thought (Ithaca, NY: Cornell University Press, 1998). Löwe, Can Laurens. “Gregory of Rimini on the Intension and Remission of Corporeal Forms,” Recherches de théologie et de philosophie médiévales 81 (2014), 273–330. Maier, Annaliese. Das Problem der intensiven Grösse in der Scholastik (De Intensione et remissione Formarum) (Rome: Verlag Heinrich Keller, 1939). Maier, Annaliese. “The Theory of the Elements and the Problem of their Presence in Compounds,” tr. S. Sargent, On the Threshold of Exact Science (Philadelphia: University of Pennsylvania Press, 1982). O’Rourke, Fran. Pseudo-Dionysius and the Metaphysics of Aquinas (Leiden: Brill, 1992). Pasnau, Robert. “Scholastic Qualities, Primary and Secondary,” in L. Nolan (ed.), Primary and Secondary Qualities: The Historical and Ongoing Debate (New York: Oxford University Press, 2011), 41–61. Pasnau, Robert. “On What There Is in Aquinas,” in J. Hause (ed.), Aquinas’s Summa Theologiae: A Critical Guide (Cambridge: Cambridge University Press, 2018), 10–28. Péghaire, Julien. “L’axiome ‘bonum est diffusivum sui’ dans le néoplatonisme et le thomisme,” Revue de l’Université d’Ottawa 1 (1932), 5–30. Roques, Magali. “Quantification and Measurement of Qualities at the Beginning of the Fourteenth Century: The Case of William of Ockham,” Documenti e studi sulla tradizione filosofica medievale 27 (2016), 347–80. Schäffer, Christian. “Augustine on Mode, Species and Natural Order,” Augustinian Studies 31 (2000), 59–77. Schockenhoff, Eberhard. “The Theological Virtue of Charity,” in S. J. Pope (ed.), The Ethics of Aquinas (Washington, DC: Georgetown University Press, 2002), 244–58. Sherwin, Michael. By Knowledge and By Love: Charity and Knowledge in the Moral Theology of St. Thomas Aquinas (Washington, DC: Catholic University of America Press, 2005).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

146   Solère, Jean-Luc. “Plus ou moins: le vocabulaire de la latitude des forms,” in J. Hamesse and C. Steel (eds.), L’Élaboration du vocabulaire philosophique au Moyen Âge (Turnhout: Brepols, 2000), 437–88. Solère, Jean-Luc. “Thomas d’Aquin et les variations qualitatives,” in C. Erismann and A. Schniewind (eds.), Compléments de substance: Études sur les propriétés accidentelles offertes à Alain de Libera (Paris: Vrin, 2008), 147–65. Solère, Jean-Luc. “Les variations qualitatives dans les théories post-thomistes,” Revue Thomiste 1 (2012), 157–204. Sweeney, Leo. Divine Infinity in Greek and Medieval Thought (New York: P. Lang, 1998). Sylla, Edith. “Medieval Concepts of the Latitude of Forms: The Oxford Calculators,” Archives d’histoire doctrinale et littéraire du Moyen Âge 40 (1974), 223–83. Sylla, Edith. The Oxford Calculators and the Mathematics of Motion, 1320–1350: Physics and Measurement by Latitudes (New York: Garland Publishing, 1991). te Velde, Rudi. Participation and Substantiality in Thomas Aquinas (Leiden: Brill, 1995). Tomarchio, John. “Four Indices for the Thomistic Principle Quod recipitur in aliquo est in eo per modum recipientis,” Mediaeval Studies 60 (1998), 315–67. Tomarchio, John. “Thomistic Axiomatics in an Age of Computers,” History of Philosophy Quarterly 16 (1999), 249–75. Tomarchio, John. “Aquinas’s Division of Being According to Modes of Existing,” Review of Metaphysics 54 (2001) 585–615. Tomarchio, John. “Aquinas’s Concept of Infinity,” Journal of the History of Philosophy 40 (2002), 163–87. Wippel, John F. “Godfrey of Fontaines on Intension and Remission of Accidental Forms,” Franciscan Studies 39 (1979), 316–55. Wippel, John F. The Metaphysical Thought of Thomas Aquinas (Washington, DC: Catholic University of America Press, 2000). Wippel, John F. “Aquinas on Creatures as Causes of Esse,” in Metaphysical Themes in Thomas Aquinas II (Washington, DC: Catholic University of America Press, 2007), 172–93. Wippel, John F. “Thomas Aquinas and the Axiom ‘What Is Received Is Received According to the Mode of the Receiver,’ ” in Metaphysical Themes in Thomas Aquinas II (Washington, DC: Catholic University of America Press, 2007), 113–22.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

Scotus versus Aquinas on Instrumental Causality Jean-Luc Solère

The medieval notion of instrumental cause is not limited to what we call today “instruments” or “tools.” It includes, for instance, the accidents by which a substance acts on another substance, sensible species acting on a visual faculty, bodily organs, sacraments, and sometimes creatures with respect to God’s action. In all these cases, instrumental causes are causes that are subordinated to a principal cause and contribute to its action and effects; however, they do so in a specific manner that makes them different from regular secondary causes. This manner has to be spelled out, which turns out to be not so easy. How exactly instrumental causes act is the object of a vigorous rebuttal of Thomas Aquinas by John Duns Scotus. Aquinas’s conception of instrumental causality does not strongly distinguish between artifacts and natural agents. Moreover, it postulates a complex superposition of layers of causation (the action a thing has by its own nature, and the action it has inasmuch as it is used as an instrument). In reaction, Scotus offers a novel view that clearly separates artificial instrumentality and natural instrumentality, and in both cases explains causation with great economy (one action per thing). More specifically, Scotus’s attack on Aquinas’s conception hinges on the question whether all instrumental causes have a preparatory role for the action of the main agent. I will first present the argument in which Aquinas introduces this thesis, and then Scotus’s criticism of this argument, which consists in pointing out that, as Aquinas himself seems to agree, there are instrumental causes that act on the very effect of the main agent, rather than carrying out a mere preparation for this effect. Next, I will systematically analyze Aquinas’s statements on instrumental causality in order to show that, for him, instrumental causes that act on the effect of the

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

148 - `  main agent also do preparatory work. Making this point salvages the consistency of his account, but at the same time positing that a same instrumental cause may have that double function complexifies his theory even more. Finally, I will systematically analyze Scotus’s account of instrumental causality and show that he offers a completely different classification of instrumental causes, with the result that he is able to reject Aquinas’s claim and to establish an exclusive disjunction between preparatory instrumental causes and perfective instrumental causes. By the same token, Scotus proposes a conception that is much more simple and convincing. His input greatly clarifies the theory of instrumental causality and might constitute a turning point for it.

1. Aquinas’s Argument in the Sum of Theology Medieval commentaries of Peter Lombard’s Sentences generally broach the topic of instrumental causality at the beginning of book IV, which addresses questions on the sacraments, seen as instrumental in God’s dispensation of grace. However, Scotus, before turning to the sacraments, devotes a single but very long question to the metaphysical problem of whether a creature can create in its turn. Although it seems out of place, this question does relate to the general topic of instrumentality, since in the background of the inquiry stands the Avicennian thesis—devised to explain how multiplicity can derive from the One—that higher, immaterial substances may transmit God’s creative action, that is to say, may be endowed by God with the power of creating substances below them. As Scotus notes, theologians generally answer the question in the negative. Aquinas does so. Nevertheless, while agreeing with his conclusion, Scotus is dissatisfied with Aquinas’s arguments and, in the typical medieval fashion, devotes a good number of pages to showing why they are not compelling. The discussion quickly expands beyond the particular issue of creation (let alone that of the sacraments), and involves Aquinas’s and Scotus’s metaphysical views about causation and existence. I will focus here on the part of the debate that provides valuable insights on their respective theories of instrumental causality.¹ ¹ For an overview of the whole discussion, see Marilyn McCord Adams, “Can Creatures Create?,” Philosophia 34 (2006), 101–28.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

149

Overall, Scotus discusses four proofs by which Aquinas tries to demonstrate that a creature cannot create. I will concern myself with the first of these proofs, which is found in the Sum of Theology, I.45.5 and contains a crucial statement on how instrumental causes in general work. Scotus rightly isolates three units in this text: (A) a direct argument that shows that creation is God’s exclusive privilege because giving existence (esse) is the proper effect of God’s action alone, (B) a confirmation by a passage of the Liber de Causis, (C) an additional argument that eliminates the possibility that creatures are used by God as instruments in an act of creation.² Here is the full passage under discussion: . . . creating cannot be the proper action of any being other than God. (A) For effects that are more universal must be traced back to causes that are prior and more universal. But the most universal of all effects is existence (esse) itself. Hence, existence must be the proper effect of the first and most universal cause, that is, God. (B) This is why the Liber de Causis says that neither an intelligence nor a noble soul gives existence except insofar as it acts by God’s action. But to give existence absolutely speaking—and not just insofar as it is this-existence or suchexistence—pertains to the nature of creation. Thus, it is clear that creation is the proper action of God himself. (C) However, it is possible for one thing to participate in another’s proper action, not by its own proper power but, rather, instrumentally, insofar as it acts by that other’s power. For instance, air gives warmth and ignites through the power of fire. Accordingly, some have suggested that even though creation is the proper action of the universal cause, certain lower causes can nonetheless create insofar as they act by the power of the first cause. Thus, Avicenna claimed that the first separated substance, having been created by God, creates another separated substance after itself, along with the substance of the heavens and its soul, and that the substance of the heavens creates the matter of lower bodies. And in the same manner, in Sentences IV, dist. 5, the Master [Peter Lombard] says that God can communicate to a creature the power to create, in such a way that it creates as God’s minister, not on its own. But this is impossible. (C.1) For a secondary instrumental cause

² Ord. IV.1.1.1, nn. 17–18, XI:9–11; Rep. IV-A.1.1, nn. 15–18, vol. I, 4–5.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

150 - `  participates in the action of a higher cause only insofar as, (C.1.a) through something proper to itself, (C.1.b) it contributes dispositively to the principal agent’s effect. If in a given case it did nothing through what is proper to itself, then it would be pointless to put it to work; nor would there have to be determinate instruments for determinate actions. For instance, we see that, by cutting wood, a saw, which has this power as something proper to its own form, produces the form of a bench, which is the principal agent’s proper effect. (C.2) But the proper effect of God as creator is that which is presupposed for every other effect, namely, existence absolutely speaking. (C.3) Hence, nothing can contribute dispositively and instrumentally to the effect of creation, since creation is not from any presupposed thing that could be disposed by an instrumental agent’s action. So, then, it is impossible that it should belong to any creature to create, either by its own power, or instrumentally, i.e., ministerially.³

Unit (C) is the one of interest for our purpose here. A possible objection to the conclusion of Aquinas’s argument in unit (A) is the following. Granted, creatures cannot create by their own power; nonetheless, they might be instrumental causes; that is to say, they might merely relay the power of the First Cause, without adding anything to it. They would be “ministers” rather than “authors” of the action, as Peter Lombard puts it,⁴ in the same way as warm air warms up something by passively transmitting the heat that comes from a fire, as Aquinas explains. In other words, they could be simply an intermediary through which God exercises his creating power. After all, according to the sacramental theory developed by Aquinas himself, the water poured during baptism, for instance, contributes by no power of its own to the cleansing of the soul, but is simply the instrument by which God’s grace is infused in the soul.⁵ In response to this objection, Aquinas adds to his demonstration a corollary, (C), in order to establish that the act of creation cannot involve ³ Translation A. Freddoso, modified. (NB: for the Summa Theologiae, I will use either Freddoso’s translation or the English Dominicans’. I have often heavily modified these translations as well as the translations of other works referenced in the bibliography. All other translations are mine.) ⁴ Lombard, Sent. IV.5.3, n. 3, II:267. ⁵ It is a cause, though, which means that it is more than a mere sign. See Lombard, Sent. IV.1.4, n. 2, II:233, and Aquinas, Sent. IV.1.1.4; STh III.66.1.c., III.66.3.c.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

151

creatures at all, even as instrumental causes.⁶ Aquinas first specifies what the characteristics of an instrumental cause are: (C.1.a) Contrary to what the objection implies, an instrumental cause does contribute something of its own to the action of the main cause.⁷ For instance, a saw contributes by its shape (its serrated blade) to the making of a bench. Otherwise, it would be in vain that it has been designed as an instrumental cause, and the principal agent could indifferently operate through any intermediary whatsoever. (C.1.b) The principal agent uses an instrumental cause to prepare the patient in order to bring about its own effect. In other words, an instrument acts on a preexisting subject, and disposes it so that it can receive the form actualized by the main cause. In accordance with the previous example, the saw, by removing parts of a piece of wood, readies it for the actualization of the form that the carpenter has in mind. Next, (C.2) Aquinas states that the action that brings a patient into existence is a necessary condition for any other type of causality having any other effect on that patient. Obviously, one cannot act on what does not exist. Therefore, an instrumental cause has to act on something that already is. But per (A), God is the sole source of existence. This means that neither the instrumental cause under consideration nor any other creature can independently bring about the patient. Thus, an instrumental cause cannot carry out its own action (the necessity of which was underlined in C.1) if God has not already produced his proper effect, i.e. given existence to the patient. However, (C.3) in a situation of creation, nothing preexists by definition, since creation properly speaking is ex nihilo. Therefore, there is nothing on which an instrumental cause could act.⁸ But, by (C.1.a), an instrumental cause has to contribute something of its own, and, ⁶ See also De Pot. 3.4.c. ⁷ Cf. De Pot. 3.4.c.: “Even when a thing is used as an instrument, its action must come from the potentiality of that thing.” ⁸ What is said here applies of course to the creation at the beginning of the world (provided that one creature has already been created, about which one wonders whether God could use it as an instrument to create other creatures), but one may also imagine a situation in which God, right now, would decide to add to the world an extra cow by making it from nothing, rather than waiting for one existing cow to give birth. Could God use an instrumental cause to create that extra cow from no preexisting material (as opposed to changing a stone into a cow)? Let us note, however, that what Aquinas says here does not apply to natural processes of generation of substances by other created substances, like the normal making of a new cow by a cow. Aquinas believes that, in these processes, creatures are like instrumental causes by which God

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

152 - `  by (C.1.b), it does this as a preparation to the action of the main cause. As a consequence, an instrumental cause cannot contribute to the action of creating. Using an instrumental cause to create is logically impossible, since God would have first to create what the instrument would be supposed to act on. Before God’s action, there is nothing to be acted upon. After God’s action, it is too late to contribute to creation—so to speak, since the order between the actions is not necessarily an order of temporal succession, but an order of logical priority. In short: the nature of an instrumental cause is to prepare; but there is nothing to prepare in view of creation; therefore, an instrumental cause cannot be used to create.

2. Scotus’s Criticism Scotus’s criticism of Aquinas’s corollary (C) targets clause (C.1.b), namely, Aquinas’s claim that the function of an instrumental cause is to prepare a preexisting subject to receive the specific action of the main agent.⁹ This is not always the case, Scotus objects first.¹⁰ It is only when several successive actions have to occur in a certain order that an instrument acts to prepare the patient. The preparation of the patient by the instrument is one of these actions, and is followed by the action of the main agent. But it also happens that the action of the instrument directly has for its terminus the final effect of the main agent. For instance (Scotus’s example), when a worker mints coins, the action of the die (the instrument that the worker uses to imprint an image and words on a face of a coin) contributes directly to the effect that the main agent (the worker) intended, which is just to mint a coin. There is nothing beyond and above this result, which is obtained by a single action.¹¹ Likewise, if an artisan uses several tools to make a bench, admittedly the first tool (say, a plane) might only prepare the wood. But the last tool communicates existence. This is possible because there is some preexisting matter for the creatures to work on (see section 3). ⁹ Ord. IV.1.1.1, nn. 44–9, XI:19–21; Rep. IV-A.1.1, nn. 26–34, 8–13. ¹⁰ Ord. IV.1.1.1, n. 45, XI:20. ¹¹ See also Rep. IV-A.6.5–6, n. 47,160–1.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

153

(say, a chisel) introduces the form in the artifact, and thus its action extends to the effect and end point of the action of the principal agent (the artisan) without having acted in a preparatory manner. Now, let us imagine that the artisan can use a single tool that can perform all the functions of other tools. It would be non-essential for this single tool to act first in a preparatory way, before attaining the final effect and end goal of the main agent.¹² At the moment of the final shaping, it co-acts with the hand of the craftsman. As a consequence, acting in a preparatory manner is not part of the definition of an instrument. The reason why the saw, in Aquinas’s example in Sum of Theology I.45.5, acts dispositively is that multiple actions succeed each other in the making of a bench. The action of an axe is required to cut the wood, before, for instance, a plane is applied to the wood, and then a saw and a chisel. But this is only accidental. If, by its action, a saw could directly contribute to the finishing of a product, the saw would be an instrument that is not confined to preparatory work, but acts at the very moment the form intended by the artisan is finally actualized.¹³ Aquinas’s premise, therefore, is true only when it receives the following restriction: the action of an instrumental cause prepares the patient when that cause cannot contribute directly to the effect of the main agent.¹⁴ Then, for Aquinas’s reasoning to be valid, it must be granted that, regarding creation, the action of a created agent, even under God’s impulse, can never directly contribute to the end point of the process, namely, the giving of existence to a creature. For if it could, it would not be impossible that God uses a creature as an instrumental cause to create, given that the alleged impossibility only results from the absurdity of an instrumental cause acting in a preparatory way on what does not yet exist. If instead of preparing, an instrumental cause could bring about the final effect, as in the case of the die, then that contradiction would disappear. However, Aquinas has not proved that the action of a creature cannot end in the production of existence otherwise than by the argument in ¹² Rep. IV-A.1.1, n. 30, 11. ¹⁴ Ord. IV.1.1.1, n. 46, XI:20.

¹³ Rep. IV-A.1.1, n. 31, 11.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

154 - `  (A), which restricts to God the communication of existence. But Scotus thinks this argument is flawed too.¹⁵ Aquinas, therefore, has not demonstrated that God cannot create by the intermediary of creatures. It might be that God creates through a creature, using it as an instrument, the action of which extends up to the final stage of the whole action, i.e., the giving of existence from nothing. Scotus does not believe this can be the case; he simply challenges Aquinas’s way of barring that possibility. In the course of this refutation, Scotus additionally points out that Aquinas contradicts himself since he sometimes acknowledges that an instrument sometimes contributes to producing the proper effect of the main agent, instead of only preparing this production.¹⁶ There are in effect some texts in which Aquinas says that (1) in natural processes, the action of accidental elementary qualities (such as cold, heat, etc.) on a patient induces a new substantial form in that patient; and that (2) in the production of artifacts, the action of the instrument extends to the introduction, in matter, of the form that the designer intended to implement.¹⁷ Admittedly, none of these instrumental causes achieves this end without the action of the main agent: in the case of natural processes, the elementary qualities act as instruments of the substantial form of the agent; and, obviously, in the production of artifacts, the action of the tool is conditioned by the action of the main agent. Still, this seems to fly in the face of clause (C.1.b), that is, the characterization of the work done by an instrumental cause as a preparatory one.

¹⁵ Ord. IV.1.1.1, nn. 33–7, XI:15–16. See a detailed commentary of Scotus’s refutation of this argument in McCord Adams, “Can Creatures Create?,” pp. 113–15. In brief: either there is an equivocation on “more universal” (which can be taken either intensionally or extensionally), or the minor premise is false. ¹⁶ Rep. IV-A.1.1, n. 45, 19. ¹⁷ For instance, Sent. IV.12.1.2.2.ad2: “ . . . [an instrumental cause], by its own power, does not act beyond its species; but by the power of another, of which it is the instrument, it can act beyond its species, as for instance a saw acts towards the form of a bench”; De Pot. 3.7.c.: “An instrument is somehow a cause of the effect of the main cause, not by its own form and virtue, but inasmuch as it participates in the power of the main cause by its motion . . . the inferior power is not conjoined to the effect but by the superior power”; Q. D. de An., a. 12, c.: “In the elements, there are no principles of actions but the active and passive qualities, which however act by the power of the substantial forms. Hence, their action not only ends at accidental dispositions, but at [the actualization of] substantial forms.” In this light, even the passage of Sum Theol. I.45.5, (C.1.b), discussed by Scotus, can be interpreted in this sense: “by cutting wood, a saw, which has this power as something proper to its own form, produces the form of a bench, which is the principal agent’s proper effect.”

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

155

3. Aquinas’s Theory of Instrumental Causality Pace Scotus, an obvious way to reconcile Aquinas’s apparently contradictory statements is to assume that, for him, even instrumental causes that contribute to the ultimate stage of an action first carry out some kind of preparation. If so, all instrumental causes do preparatory work, both those that stop at that stage, and those that in addition carry out their action up to the final stage. This would not only clear Aquinas from the charge of self-contradiction, but would also defend against Scotus the validity of his argument (C) that, in Sum of Theology I.45.5, bars the possibility of using instrumental causes in creation ex nihilo. If all instrumental causes have to do preparatory work on some preexisting substrate, no instrumental cause can contribute to creation, since there is yet no substrate on which they could apply their dispositive causality. Even if the argument in (A) is inconclusive and does not show that the action of a created agent can never reach the giving of existence to another creature, it remains that since dispositive action is impossible in the case of creation, an instrumental cause cannot contribute to creating. Admittedly, Scotus may disagree with the claim that all instrumental causes begin to act by doing preparatory work, and maintain that some instrumental causes, like the die, get right at the final effect without any preliminaries, as we shall see. But at least it will be clear that the disagreement hinges on this. What exactly an instrumental cause does proves to be the crucial point in this debate. So let us review Aquinas’s theory of instrumental causality, starting with this passage from his Sentences commentary: (i) An instrument has two actions: one due to its own nature, the other due to its being moved by the primary agent. For instance, fire, which is an instrument for the nutritive power, as it is said in De anima II, by virtue of its own nature is able to dissolve and consume, and like effects: but inasmuch as it is an instrument of the vegetative soul, it generates flesh. (ii) However, one should know that the action of an instrument sometimes extends to the final perfection that the principal agent brings in, and sometimes not. (iii) But the action of an instrument always extends to a stage that is beyond that which it can achieve by its nature alone, whether this stage be the final form or a

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

156 - `  disposition: otherwise, it would not be acting as an instrument. (iv) For instance, the active and passive qualities of the elements instrumentally extend to the [actualization of] material forms to be educed from matter—not, however, to the production of the human soul, which comes from outside [matter].¹⁸

a. Double Action Aquinas begins the text just quoted with the claim that (i) an entity x can have simultaneously two different actions: an action φ that results simply from its nature, and another type of action, call it χ, when it is used by a principal agent, in such a way that: if x is an instrumental cause, φ(x) Λ χ(x). For example, fire (or more precisely, the heat that belongs to fire) has by nature the power of breaking down materials, but inasmuch as this power is employed by a vegetative soul in the process of digestion, it contributes to the production of the animal’s matter from the nutrients.¹⁹ This claim of a double action seems counter-intuitive. Is it not rather the case that an agent always simply performs the only action it is capable of—its action φ—while the effects of this action can be used for a larger purpose when this agent is enlisted as an instrumental cause? Heat only dissolves and consumes, but these actions are embedded in a longer process, managed by the main cause, in which the breaking down of food enables the nutritive and growing action of the vegetative function of a soul to use the nutriments. Likewise, a hatchet does not seem to have an action φ distinct from χ, because if it is not handled by someone, it simply does not do anything, and when it is used by someone, it simply does what it is designed to do, namely, split wood.

¹⁸ IV.1.1.4.1.c. ¹⁹ See Aristotle De anima II.4, 416b 28–9. Aquinas seems to have wavered about this. He writes in ScG II.89.15 that the action of the natural heat in the body does not extend to the end of the operation that is carried out under the direction of the soul, namely, the production of flesh; the breaking down of nutrients by the heat only is a preparation. However, he later returns to his initial view. See below the quote from Q. D. de An., and De Pot. 3.8.ad13: “An accidental form acts by virtue of the substantial form, as if it were its instrument; in this way, Aristotle says in book II of On the Soul that the heat of fire is the instrument of the nutritive power. Therefore, there is nothing impossible in the action of an accidental form terminating in a substantial form.”

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

157

However, Aquinas contends in clause (iii) of the text quoted above that all instrumental causes are such that their subordination to a principal agent, or directing cause, makes them do something that is otherwise beyond their reach. For him, that is their distinctive characteristic. Defining an instrumental cause as an agent that is empowered or moved by a higher cause is not accurate enough, because that description fits all secondary causes, not just instrumental causes. In the first place, no creature acts without being moved by God, the non-moved mover and first cause, who is the cause not only of their existence but also of their actions, and “applies” their powers to their effect.²⁰ The chain of cosmological intermediaries, such as celestial movements, sublunar movements, and so on, also actualizes and governs inferior causes.²¹ Thus, a human would not engender another human without the influence of these superior causes. An individual only transmits individual characteristics, while human nature as such has to be actualized by a more general cause, namely, the celestial bodies.²² Therefore, an individual human being is causally subordinated to celestial bodies and acts as a secondary cause. But here the higher cause does not make the lower agent produce something that it is beyond its nature to produce: the effect is just what is to be expected from a human being. In contrast, we saw that, according to Aquinas, no creature has a nature such that it can give existence (esse) to something through its own power. This does not

²⁰ De Pot. 3.7.c.: “And since an agent of a lower nature does not act except when it is moved, because lower bodies of that sort both cause and are subject to alteration, whereas the heavens cause alteration but are not subject to it, and yet do not cause movement unless being themselves moved, this [regress] does not cease until we arrive at God, so that we must eventually trace its movement to God, it follows of necessity that God causes the action of all natural things by moving them and applying their power to action.” Cf. ScG III.67.4: “Moreover, any agent that applies its active power to some action is said to be the cause of that action. Thus, an artisan who applies the power of a natural thing to some action is said to be the cause of that action; for instance, a cook is the cause of the cooking which is done by means of fire. But every application of power to operation is originally and primarily made by God. For operative powers are applied to their proper operations by some movement of body or of soul. Now, the first principle of both types of movement is God. Indeed, He is the first mover and is absolutely not moved, as we have shown above.” ²¹ De Malo 16.9.c.: “Indeed, since the highest beings have the most universal powers, the lowest, passive beings are no match for receiving immediately their universal effect, and must receive it by the mediation of powers that are more particular and limited, as can be seen even in the order that exists between bodily things.” ²² STh I q. 115, a. 3, ad 2m. De Malo 16.9.c.: “ . . . celestial bodies are principles for the generation of humans and other sophisticated animals, by the intermediary of a particular power which is in their semen.” On this power, see ScG II.89.8.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

158 - `  entail that creatures do not contribute at all to the communication of existence in natural processes of generation (by contrast with cases of creation from nothing).²³ On the contrary, even in the ordinary course of things (as opposed to miracles), they are like instrumental causes for God’s action of giving existence.²⁴ Concretely, this means that for Aquinas, when a cow engenders a cow, God acts through it and makes its natural power contribute to an effect that is above what it can achieve by itself or with the celestial causes: giving existence to another creature.²⁵ By its own action, it particularizes the universal effect of the universal cause into the being of this individual cow.²⁶ Hence, Aquinas makes a distinction between, on the one hand, causality by creation, that is, communication of existence, which comes from God, and, on the other hand, causality by information, which is the contribution of a creature inasmuch as it adds determining limits to pure existence.²⁷ In this ²³ This point, in opposition to what is commonly alleged about Aquinas, has been highlighted by John Wippel, “Thomas Aquinas on Creatures as Causes of Esse,” International Philosophical Quarterly 158 (2000), 197–213. ²⁴ De Pot. 3.4.c.: “ . . . whatever other cause gives being does this insofar as it is the recipient of the power and operation of the First Cause, and not by its own power. Thus, an instrument performs an instrumental operation not by the power of its own nature, but by the power of the agent who handles it . . . ”; De Pot. 3.7.c.: “Thus, God is the cause of all actions, inasmuch as any agent is an instrument of the divine power’s operating.” ²⁵ ScG III.66.6: “Therefore, being is the proper effect of the primary agent, and all other things produce being because they act through the power of the primary agent”; De Pot. 3.1.c: “ . . . no thing gives being except in so far as it partakes of the divine power”; De Pot. 3.7.c.: “Accordingly, this or that individual thing cannot by its action produce another individual of the same species except as the instrument of that cause that includes in its scope the whole species and, besides, the whole being of the inferior creature. Wherefore no action in these lower bodies attains to the production of a species except through the power of the heavenly body, nor does anything produce being except by the power of God. For being is the most common first effect and more intimate than all other effects. Wherefore it is an effect which it belongs to God alone to produce by his own power; and for this reason (De Causis, prop. ix) an intelligence does not give being, except inasmuch as the divine power is in it”; STh I.45.5.c: “This is why the Liber de Causis says that neither an intelligence nor a noble soul gives esse except insofar as it acts by God’s action.” ²⁶ ScG II.21.4: “All other agents are not the cause of the act of being as such, but of being this—of being a man or being white, for example”; ScG III.66.6: “But the first among all effects is the act of being, since all other things are certain determinations of it. Therefore, being is the proper effect of the primary agent, and all other things produce being because they act through the power of the primary agent. Now, secondary agents, which are like particularizers and determinants of the primary agent’s action, produce as their proper effects other perfections which determine being”; STh I.45.5.ad1: “ . . . no created entity can produce any entity absolutely speaking except insofar it causes esse in this [subject] . . . .” ²⁷ De Pot. 3.1.c.: “ . . . a natural agent produces a being not simply, but determines a preexistent being to this or that species . . . Wherefore it is stated in De Causis (prop. xviii) that being is by creation, whereas life and the like are by information: for all causation of absolute being is traced to the first universal cause, while the causation of all that is in addition to being, or

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

159

respect, creatures are not just secondary causes for God’s productive action, but are like instrumental causes. As noted above,²⁸ it is only in the case of creation from absolutely nothing that creatures cannot be instrumental causes; in natural processes of generation, there obviously is some preexisting patient on which creatures can act to prepare the action of the main agent; thus, they contribute to giving existence (and this is not by delegation of God’s power, but because of God’s acting in them). According to Aquinas, then, in a general way the very fact that an agent is used as an instrumental cause makes it perform an action that is different from the action it can perform by itself, inasmuch as its instrumental action is enhanced, compared to the natural one. Conversely, if an agent does not do something that is beyond its normal reach, then it is not being used as an instrumental cause. This is why we have to distinguish between actions φ and χ. To answer the hatchet counterexample taken above, suppose that I accidentally drop the hatchet and that it gashes my foot: that’s an action φ, which is due only to the sharpness of the blade. When, on the contrary, the hatchet is skillfully used to carve wood, its action still is to split—it still is the same action φ—but it also has a χ-type action, because its operation is, so to speak, enhanced by the main agent’s action and results in a design that a hatchet falling randomly would not achieve. It is not just that its cutting ability is actualized by a carpenter “applying” the power of the tool to a piece of wood, as opposed to simply falling.²⁹ It is directed by a superior causality, that of the mind guiding the hand, which does more than merely move the hatchet: it implements a design through this instrument, which is an effect that the instrument cannot bring about by its own

specific of being, belongs to second causes which act by information, on the presupposition as it were of the effect of the first cause.” ²⁸ Note 8. ²⁹ De Pot. 3.7.c.: “ . . . thirdly, a thing is said to cause another’s action by moving it to act: whereby we do not mean that it causes or preserves the active power, but that it applies the power to action, even as a man causes the knife’s cutting by the very fact that he applies the sharpness of the knife to cutting by moving it to cut”; ScG II.21.4: “The nature of an instrument is to be a cause of motion that itself is moved”; STh III.64.5: “ . . . an instrument acts not by reason of its own form, but by the power of the one who moves it”; STh III.72.3.ad2: “An instrument acquires instrumental power in two ways: when it receives its form of instrument, and when it is moved by a principal agent to produce an effect.” See also STh III.63.5.ad2.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

160 - `  power only.³⁰ This is the action χ of the hatchet. The arrangement of the different cuts is not explainable solely by the properties of the hatchet (its shape, hardness, and sharpness). This action goes far beyond the mere cutting, inasmuch as it carries or participates in, as Thomas says, the causality of the main agent. Likewise, in the digestion process, beyond its breaking-down power, heat comes to generate flesh inasmuch as a power of the soul directs it.³¹ Making flesh is not reducible to consumption and dissolution. But because heat is harnessed by the vegetative faculty of a soul, its dissolving and consuming activity results in the assimilation of nutriments, which means that heat has an action χ that goes beyond what its action φ does. Its natural action is enhanced on account of being embedded in, and directed by, the action of the principal cause. Hence Aquinas speaks of the instrumental cause’s double action. One is its own action, which it performs by virtue of its own form and power; the other is its instrumental action properly speaking, which it performs by virtue of the power of the main agent, and which is superior to the first one.³² Importantly, according to Aquinas, action χ does not replace action φ: these two actions are performed simultaneously, as strange as it may seem. The subordinated cause executes its instrumental function (χ) and takes part in the operation of the main cause by carrying out its own operation (φ). It is by dissolving and consuming that heat contributes to the production of living matter. It is by cutting the wood that the

³⁰ De Pot. 3.7.c.: “Furthermore we find that the order of effects follows the order of causes, and this must needs be so on account of the likeness of the effect to its cause. Nor can a secondary cause by its own power have any influence on the effect of the first cause, although it is the instrument of the first cause in regard to that effect. For an instrument is in a manner the cause of the principal cause’s effect, not by its own form or power, but insofar as it participates somewhat in the power of the principal cause through being moved thereby. Thus the axe is the cause of the craftsman’s handiwork not by its own form or power, but by the power of the craftsman who moves it so that it participates in his power. Hence, fourthly, one thing causes the action of another, as a principal agent causes the action of its instrument . . . .” ³¹ De Pot. 3.4.c.: “ . . . the natural heat engenders living flesh by the power of the soul, while by the power of its nature it merely causes heat and dissolution.” ³² STh I-II.122.1.ad1: “An instrument does not execute the action of the principal agent by its own power, but by the power of the principal agent”; STh III.62.1.ad2: “An instrument has a twofold action; one is instrumental, in respect of which it works not by its own power but by the power of the principal agent: the other is its proper action, which belongs to it in respect of its proper form. Thus, it belongs to an axe to cut asunder by reason of its sharpness, but to make a bed, insofar as it is the instrument of an art. But it does not accomplish the instrumental action save by exercising its proper action: for it is by cutting that it makes a bed.” See also De Ver. 27.4.ad2.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

161

hatchet contributes to the making of the bench.³³ The cutting is not explainable solely by the properties of the tool, but it is not reducible to the sole action of the main agent either. The hatchet cuts, not the mind or the hand. Importantly too, for Aquinas an instrumental cause does not produce a part of this effect while the other main agent produces another part. Rather, both the main agent and its instrument wholly produce a single effect, that is, each are “total causes” of the effect.³⁴ There is not one detail of the woodwork that did not result from both the action of the tool and the carpenter’s action; one does not do only one part of the bench and the other another part, like partial causes would do. For the same reason, creatures really are efficient causes in the production of other substances, even though they do not communicate existence because of their own power, but because God acts through them, as we saw earlier. Their specific action (acting on a patient, modifying its accidents, eventually actualizing a new substantial form) is precisely that through which God gives existence to a new substance. Divine causality combines with their causal power in a single action of generation. In a general way, action χ is in fact an action that is common to the instrument and the main agent. Their two causalities are interlaced into one action, which produces a single effect.³⁵ In this unique action, the causality of the main agent is, of ³³ Sent. IV.1.1.4.1.c.: “ . . . every instrument reaches its effect as instrument by performing the action that is its own as a thing of a certain nature; thus, a hatchet instrumentally reaches the form of bed by splitting thanks to its sharpness”; ScG II.21.7: “ . . . it is by an action proper and connatural to itself that every instrumental agent carries out the action of the principal agent; thus, by processes of dissolving and dividing, natural heat generates flesh, and a saw, by cutting, plays its part in completing the work of making a stool.” ³⁴ ScG III.70.8: “ . . . the same effect is not attributed to a natural cause and to divine power in such a way that it is partly done by God and partly by the natural agent; rather, it is wholly done by both, each in a different way, just as the same effect is wholly attributed to the instrument and also wholly to the principal agent.” ³⁵ STh I.105.5.ad2: “ . . . one action does not proceed from two agents of the same order. But nothing prevents the same action from proceeding from a primary and a secondary agent”; STh I-II.14.3.ad4: “the principal agent and the instrument are, in a way, one cause, since one acts through the other”; STh III.19.1.ad2: “ . . . the action of the instrument as instrument is not distinct from the action of the principal agent; yet it may have another operation, inasmuch as it is a thing.” However, in Q. D. de An., 2, Aquinas seems to maintain that the action of the main agent and of the instrument are two numerically different actions: “Even though, of two different substances, one can be the cause of the operation of the other, as the principal agent is for an instrument, nevertheless the action and the principal agent is not numerically identical with the action of the instrument, since the action of the principal agent is to move the instrument, while the action of the instrument is to be moved by the principal agent and to move something else.” A way of resolving this contradiction is to interpret this passage as

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

162 - `  course, preeminent.³⁶ Accordingly, Aquinas often insists that the power of the main agent penetrates or permeates, and therefore “acts in” the power of the instrumental cause.³⁷ For instance, the instrument receives or participates in the “virtue” of the artist or of the artisan, he says. Or as he likes to quote from the Liber de causis, the first cause has more influence than the secondary causes on their effect.³⁸ This combination of causal powers in one action is possible only if the causes are hierarchically ordered, one being superior and the other inferior—which is precisely the case of a main agent and the instrument it uses.³⁹ When, on the contrary, causes are of the same nature, there are as many actions as there are causes.⁴⁰ However, contrary to subordinated causes that really and fully receive a form from the principal agent, as for instance the moon receiving light from the sun, instrumental causes receive the power of the main agent on an “intentional” mode only (as opposed to a real, complete mode), in the same way as colors of things are present in the air.⁴¹ Whereas a relevant to instrumental causality in a very broad sense (the third kind of subordination in the passage from De Pot. quoted in note 29), that is, when, for instance, one moves a billiard ball in order to move another ball, as opposed to working with an instrument, as in the example of the hatched and the bench. ³⁶ ScG II.89.14: “ . . . since the action of the primary efficient cause upon the effect of the secondary efficient cause is more powerful even than that which is exercised by the latter, hence the observation that an effect produced by a principal agent through an instrument is more properly attributed to the principal agent than to the instrument.” ³⁷ Sent. IV.1.1.4.3.c.: “An instrument, in the sense defined earlier, does not receive any power except inasmuch it is in continuity with the principal agent, so that the power of the latter is so to speak transfused into the instrument”; Sent. IV.1.1.1.4.ad1: “ . . . the power of the principal agent works in a hidden way in the instrument, as for instance the power of the art or of the artisan in a saw.” ³⁸ De Pot. 3.7.c.: “For the higher the cause, the greater its scope and efficacy; and the more efficacious the cause, the more deeply does it penetrate into its effect, and the more remote the potentiality from which it brings that effect into act.” ³⁹ ScG II.89.14: “In the third objection, the remark that the actions of diverse agents do not terminate in the production of one thing must be understood to refer to diverse agents that are not ordered to one another. For, if they are so ordered, they must have one effect . . . .” ⁴⁰ STh I.52.3.c.: “ . . . it is impossible for there to be two complete and immediate causes of one and the same thing.” See also ScG II.57.6. ⁴¹ Sent. IV.1.1.4.2.c.: “The principal agent acts according to what its form requires, and so the active power in it is a form or a quality that has its complete, real existence. By contrast, an instrument acts to the extent it is moved by something else, and so it receives a power that is in accordance with motion. But motion is not a complete being; it is a process in a being, a sort of intermediary between pure potentiality and pure actuality, as is said in Physics III. Therefore, the power of an instrument as instrument, according to which it acts on an effect that is beyond its natural reach, is not a completed being that has a stable existence in reality, but a sort of incomplete being, such as is, in the air, the power to affect the sight, inasmuch as it is an

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

163

non-instrumental and univocal agent truly and really contains the form that will be actualized in an effect, the form that an instrumental cause carries does not have a full ontological status. That is, it does not have a complete existence, and is in flux, like the form of a thing that is undergoing change and is on the way from potentiality to actuality.⁴² The instrumental cause only transmits a form that it does not genuinely possess. This explains the fact that the effect does not bear a resemblance to its instrumental cause, and can be more perfect than it, contrary to the two laws of normal causality: an agent produces something like itself, and a cause is more noble than its effect.⁴³ This is also why the power received lasts only as long as the main agent applies its operation.⁴⁴ Finally, let us note that Aquinas’s claims (i) and (iii) include no restriction or condition whatsoever. Thus, it can be said of every instrumental cause, in the strict sense, that it has two actions, and this in such a way that the effect of χ is always superior to the effect of φ.

b. Dispositive and Perfective Causes Claim (ii) in the passage of Sent. IV.1.1.4.1 quoted above posits, on the contrary, a distinction between instrumental causes—which we have instrument put in motion by an external, visible thing. These kinds of being are usually named ‘intentions’ . . . ”; ibid., ad4: “ . . . a spiritual power cannot be in a corporeal thing with its complete existence; but it can be in it as an intention, in the way the power of the art is in the instruments put in motion by the artisan . . . .” ⁴² See Sent. IV.1.1.4.4.c., Sent. IV.5.2.2.2.c. ⁴³ See STh III.62.1c., STh III.79.2.ad3. ⁴⁴ De Pot. 3.7.ad7: “ . . . that which God does in a natural thing to make it operate actually is a mere intention, incomplete in being, as colours in the air and the power of the craftsman in his instrument. Hence even as art can give the axe its sharpness as a permanent form, but not the power of the art as a permanent form, unless it were endowed with intelligence, so it is possible for a natural thing to be given its own proper power as a permanent form within it, but not the power to act so as to cause being as the instrument of the first cause . . . ”; De operationibus occultis naturae: “ . . . a lower agent acts, or is moved by the power of a higher agent, in two manners. One, inasmuch as the action proceeds from the form or power that the higher agent has impressed on it, as when the moon is illuminated by the light received from the sun. In the other manner, the lower agent acts only by the power of the higher agent, without receiving any form for acting, but acting merely by the motion by which it is moved by the higher agent, as when a carpenter uses a saw to cut. The cutting primarily is the action of the artisan, and secondarily is the action of the saw inasmuch as it is moved by the artisan, and this action does not proceed from a form or power that would remain in the saw after the motion given by the artisan.” See also De Pot. 3.11.ad14, Sent. IV.1.1.4.4.c.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

164 - `  already seen pointed out by Scotus.⁴⁵ Consequently, we need to subdivide action χ into two types.⁴⁶ In some cases, the activity of the instrument only results in the preparation of a patient for the reception of the main agent’s own effect—even though, as was just established, this preparation is more than what the instrumental cause can do by its action φ. Let us call this action χ₁, and if α is the effect of the instrument’s action φ, let us call α0 the effect of χ₁, since the latter, compared to the former, is enhanced by the power of the main agent. These are the action and effect of instrumental causes that we can label “dispositive.” In other cases, the instrumental activity results directly in the actualization of the main agent’s form in the patient. Let us call this action χ₂, and let us say that it ends in effect ω. These are the action and effect of instrumental causes that we can name “perfective,” since, under the action of the main agent, they co-bestow a final form on the patient. The examples Aquinas gives in (iv), the last sentence of the text from Sent. IV.1.1.4.1, cover both cases and highlight the three following points: 1. Elemental qualities—Hot and Cold, Dry and Wet—are instrumental causes in natural processes of the generation of substances. It is through them that a substance acts on another and eventually actualizes a new material substantial form. They are instrumental because, while Hot heats, Cold cools down, etc. (which are their actions φ, when they act alone), the result is something they would not cause just by themselves, namely, the actualization of a substantial form in the patient, which occurs only under the action of the substance to which they belong. Therefore, in all cases of the generation of a substance, their action under this main agent is an action χ, which is distinct from their natural action φ. 2. In the case of the generation of a human being, however, the action of the elemental qualities does not reach as far as the stage of the actualization of the substantial form, namely, a human soul. Even under the direction of a principal cause, their action is of the χ₁ ⁴⁵ See section 2. ⁴⁶ This is a division from the point of view of the effects, which intersects with the division of efficient causality from the point of view of the causes themselves, between principal and instrumental causes (Sent. IV.1.1.4.1.c.).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

165

type, limited to the preparation of matter for the reception of a human soul. Their instrumental causality is merely dispositive. The same observation applies to the causality of semen: in the process of the generation of a human being, its causality, even though instrumental, stops at an effect α0 , the body of the new human being, short of the ultimate phase of the process, the creation of a human soul by God directly. It only disposes matter for the reception of the rational soul given by God.⁴⁷ God alone can bring about the effect ω. 3. On the contrary, in the case of material forms, the action of the elemental qualities as instruments of a substance is a χ₂-type action, which goes beyond the modification of accidental dispositions and extends to ω, the actualization of a substantial form. Thus, the distinction between the two kinds of instrumental causality: dispositive and perfective, is clear. Now, if there were an exclusive disjunction between χ₁ and χ₂, that is, if instrumental causes did either preliminary work only or perfective work only, Scotus’s objections would be justified, as we saw. However, there is an indication that Aquinas does not regard the distinction as an exclusive disjunction. In the following passage of the Quaestiones disputatae de anima, Aquinas explains that the action of the elemental qualities, although it reaches the actualization of the substantial form, is also dispositive: . . . But the accidental form has to act by the power of the substantial form, as the instrument of the latter; otherwise, by its action, it would not introduce a substantial form [in the patient]. And because of this, in the elements there are no principles of actions except the active and

⁴⁷ ScG II.89.15: “Therefore, since every active power of nature is compared to God as an instrument to the primary and principal agent, nothing prevents the action of nature, in that self-same generated subject which is man from terminating in a part of man, and not in the whole, the production of which is due to the action of God. The human body, therefore, is formed at the same time both by the power of God, as principal and first agent, and by the power of the semen, as secondary agent; but it is God’s action that produces the human soul, which the seminal power cannot produce, but to which it disposes”; ibid., 16: “ . . . a man begets that which is like himself in species, so far as his seminal power acts in a dispositive manner toward the ultimate form from which he derives his specific nature.” See also De Pot. 3.9.ad21. This does not contradict the rule that both the main agent and the instrumental cause are total causes of the effect. As far as concerns effect α0 , the human body, the semen contributes to the making of its totality.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

166 - `  passive qualities, which nevertheless act by the power of the substantial forms. Hence, their action ends not only at accidental dispositions, but also at [the actualization of ] substantial forms.⁴⁸

In this light, let us read again the example given by Aquinas in Sum of Theology I.45.5, clause (C.1.b): “by cutting wood, a saw, which has this power as something proper to its own form, produces the form of a bench, which is the principal agent’s proper effect.”⁴⁹ The causality of a tool is yet another case in which the activity of an instrumental cause can both prepare the patient and directly contribute to the effect of the main agent. The action of the saw would not take place if it was not for the form that is in the carpenter’s mind, whose design is to make a bench; but the saw cuts the wood by virtue of its own form, and that’s its action φ. Action φ is preparatory for the actualization of the form of the bench, even though the saw has an action χ₂ that goes all the way to the final effect. This is precisely the thrust of Aquinas’s argument. What is in discussion is whether instrumental causes could contribute to creation from nothing by co-giving existence. Denying that instrumental causes can act all the way up to the terminus of the operation would be begging the question. Instead, Aquinas points out that even when they are perfective, instrumental causes are also dispositive, and it is for this reason that they cannot contribute to making something from nothing. Therefore, we can posit that, for Aquinas, perfecting instrumental causes, which have a χ₂ action that leads to an effect ω, also do preparatory work. Other instrumental causes, which are limited to having a χ₁-type action and an α0 effect, are merely dispositive causes. This entails that Aquinas holds as a general truth that all instrumental causes have a dispositive role (whereas not all are perfective). As a result, not only does Aquinas not contradict himself, contrary to Scotus’s charge, but if Aquinas is right, this makes premise (C.1.b) in STh I.45.5 unchallengeable, and the whole reasoning unassailable: if it is true that every ⁴⁸ Q. D. de An. 12.c. ⁴⁹ See above p. 151. See also ScG II.21.7: “Every instrumental agent executes the action of the principal agent by an action that is proper and connatural to itself; thus, natural heat generates flesh by dissolving and digesting, and a saw works towards the perfection of a bench by cutting”; STh III.62.1.ad2: “[an axe] does not complete its instrumental action except by exercising its own action: by splitting, it produces the bed.”

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

167

instrumental cause, even those the action of which extends up to the last effect of the main agent, do preparatory work, then it is impossible to use an instrumental cause where there is nothing to prepare, as is the case in creation.

4. Scotus’s Theory of Instrumental Causality Given what we have just established, in order to validate his criticism of Aquinas’s demonstration Scotus must prove that at least some instrumental causes do not do preparatory work, preferably with a more sophisticated argument than the quick counter-example of the die he gives in Ord. IV.1.1.1. Naturally, Scotus does not fail to elaborate on the topic, and in Ord. IV.6.3.1, nn. 115–26, he provides a systematic analysis of auxiliary causality. In the classification he offers, the different types of subservient causes, as Richard Cross remarks, “are arranged in decreasing order of independent causal activity”:⁵⁰ • Secondary causality • Dispositive causality • Borrowed instrumentality • Part–whole instrumentality • Artificial instrumentality As Scotus notes in distinction 1, if “acting as principal agent” means “acting independently of any superior cause,” then only God is a principal agent; by contrast with divine causality, all other causes can be viewed as instrumental causes, because they all are essentially subordinated to the First Cause, which is the source of all creaturely activity.⁵¹ However, if we leave aside this universal subordination to God, there are major differences between secondary causes and instrumental causes, as well as among instrumental causes.⁵²

⁵⁰ Richard Cross, “Some Varieties of Semantic Externalism in Duns Scotus’s Cognitive Psychology,” Vivarium 46 (2008), 275–301, at p. 287. ⁵¹ Ord. IV.1.1.1, n. 119–20, XI:43:842–51. See also Ord. IV.1.3.1–2, n. 117, XI:113:453–4. ⁵² Cf. Ord. IV.1.3.1–2, n. 317, XI:113.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

168 - ` 

a. Secondary Causality Four features characterize secondary causes:⁵³ (i) They possess the form that is the principle of their action; (ii) The main cause does not bestow this form on them at the time of the action (that is to say, a secondary cause has that form before the action takes place, “in esse quieto”); (iii) They act by their own form and action on the same effect as the first cause; (iv) They do so only by virtue of the main cause, to which they are subordinated. The last criterion means that, instead of a circumstantial or accidental combination of two causal powers, there is an essential order, or hierarchy of essence, between powers,⁵⁴ such that the secondary cause cannot act without a cause that is of a different, superior nature. As an example of that type of causality, Scotus offers the cooperation of sun and man in the production of man.⁵⁵ The sun indispensably contributes to the processes of generation and corruption by its heat and its movement on the ecliptic circle,⁵⁶ and it is a more perfect cause because it is a more universal cause (its action is narrowed down to a particular generation process by a specific inferior cause).⁵⁷ Both the sun and an individual ⁵³ Ord. IV.6.3.1, n. 116, XI:330: “A secondary cause has its own form in the state of rest, a form which is the principle of its action at its own level, and this with respect to the same effect on which the primary cause acts at its own level too, so that, without exception, a secondary cause in the strict sense extends its action to the effect of the main cause thanks to its own form; this, however, it does by virtue of the primary cause, because it is subordinated to that cause.” See also Rep. IV-A.6.5–6, n. 50, 161–2 (modified): “ . . . when an action of some agent concurs with the action of the principal agent to [produce] one and the same effect, and this agent does not act, strictly speaking, insofar as it is moved, then this agent is not an instrumental agent but a secondary agent, acting as a secondary to [produce] the same effect as the principal agent intends to . . . and in such an agent, a proper form is required, by means of which it immediately achieves the effect . . . .” ⁵⁴ Ord. IV.6.3.1, n. 122, XI:332. ⁵⁵ Ord. IV.6.3.1, n. 123, XI:332. Cf. Aristotle, Physics II.2, 194b13: “Man is begotten by man and by the sun as well.” ⁵⁶ Generation of Animals II.3, 737a3–6: “The heat of the sun does effect generation, and so does the heat of animals.” See also On Generation and Corruption 336a31, b6, b17. ⁵⁷ Ord. I.3.3.2, n. 545, III:326:8–10: “ . . . the action of a higher cause is limited [by a lower cause], like the sun to producing a human by the concurring action of a human being, and to producing an ox by the concurring action of an ox . . . .” See also Lect. I.3.3.2–3, n. 379, XI:372:7–10, and n. 415, XVI:391:18–22.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

169

human are partial causes of the effect; acting together, they constitute one total cause (one not by formal unity, but by a unity due to their hierarchical relation).⁵⁸ Their connection is not accidental as it is the case between concurring causes of the same nature. Among the latter, a cause may lack power and need the cooperation of another cause, as when we request help to carry a load. However, as Scotus explains, this conjunction is only accidental because, if all the concurring forces that pull a load could be transferred and added up in one individual, this individual would pull the load all by himself.⁵⁹ On the contrary, the dependency of a secondary cause toward a cause of a different, superior nature is such that no increase in the amount of its power would make it able to act alone. It needs a complement of a different order. Thus, in conformity with criteria (i) to (iv), a secondary cause, properly speaking, is a fully-fledged being, endowed with its own power of action, which in its own right co-acts on the same effect as another agent, but it so happens that in the natural order of things this secondary cause cannot bring about its effect without this other

⁵⁸ Ord. I.3.3.2, n. 503, III:298:10–14, and n. 509, III:301:15–17; Lect. I.Prol.1.1, n. 35, XVI:15:14–16. See also Lect. II.34–7.5, n. 125, XIX:358: “There is another order between causes, when one is higher and the other lower, the higher one moves the lower one, and none is the total cause of the effect, but they combine in one total cause, like the sun and the procreator with respect to the procreated”; Ord. II.34–7.5, n. 117, VIII:417:725–9: “ . . . sometimes the order [between causes] is such that the principal cause moves the less principal cause, either to its second actuality or to its first actuality, in such a way, however, that they are both partial causes, and combine into one total cause per se, like the hand and the stick with respect to the movement of the ball, or the sun and the father with respect to the son.” However, given that in Quodl. XV, 142b, Scotus writes: “ . . . sometimes the posterior [cause] gets its power from the primary cause. Example: the sun and the father in the engendering of a human being” (which is more than just moving the lower cause), one may wonder if the sun-and-father example is the best one for secondary causes as defined here, since by criterion (ii) they shouldn’t receive their power from their agent. Perhaps a better example would be that of the co-action of the male and the female in procreation. See Ord. I.3.3.2, n. 496, III:294:3–8: “ . . . if one posits that the mother has an active power in the engendering of a child, her power and the father’s power concur as two partial powers, between which there nonetheless is an order, because one is more perfect that the other. However, the less perfect one does not receive its causality from the more perfect one, nor is its causality contained eminently in the more perfect cause, but the less perfect cause adds something [to the process], insofar as the effect is more perfect for being the effect of both the more perfect and the less perfect causes, than from the more perfect cause alone.” And Quodl. XV, 142b: “ . . . the mother is somehow active, which seems reasonable, since from a form of the same species results a power of the same nature [as the father’s] . . . . But the mother does not get her active power from the father, and does not depend on him while having it; she depends on him for acting by this power, and she acts less primarily than him.” ⁵⁹ Lect. I.3.3.2–3, n. 366, XVII:367:20–2.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

170 - `  agent.⁶⁰ This other agent is not of the same nature, but is a superior cause, whose own, independent action conditions the effectiveness of the operation of the secondary cause. In contrast to secondary causes in this specific sense, any other essentially subordinated cause is an instrumental cause inasmuch as it fails to satisfy one or several of conditions (i) to (iv). Instrumental causality is in its turn subdivided into different types, depending on which of these four conditions is not met.

b. Part–Whole Causality Let us modify the order of Scotus’s list of instrumental causes to go quickly over a very particular kind of instrumental causality: the instrumentality of a part within a whole, of which bodily organs are the paradigmatic case. In this particular configuration, a whole acts by one of its parts. This part is the principle of action in the first place; the whole can be said to carry out this action too, but only by virtue of the part.⁶¹ This is possible for the reason that, according to Scotus, each organ is a substance in its own right and has therefore a substantial form.⁶² Moreover, for Scotus, forms themselves, not the composite of which they are forms, cause directly.⁶³ All this entails that each organ is an agent by itself and carries out its own operation. However, the forms of all the organs are unified under a form of the whole,⁶⁴ with the result that the form of an organ also belongs to the whole,⁶⁵ and the operations of an organ are also operations of the whole.⁶⁶ Since the part is part of a whole that

⁶⁰ See Ord. I.36.1, n. 65, VI:297:16–298:1: “ . . . I concede that when there is an order between two causes, both cause the effect – but in a different way, for the higher cause causes more.” ⁶¹ Ord. IV.6.3.1, n.120, XI:331:150–1. See also Ord. IV.1.3.1–2, n. 317, XI:113:456–9. ⁶² See Richard Cross, The Physics of Duns Scotus: The Scientific Context of a Theological Vision (Oxford: Clarendon Press, 1998), 68–71. ⁶³ See Cross, “Varieties of Semantic Externalism,” 277–8, 282. ⁶⁴ Ord. IV.11.1.2.1, nn. 237–8, XII:250, and n. 254, XII:256. See Thomas Ward, John Duns Scotus on Parts, Wholes, and Hylomorphism (Leiden: Brill, 2014), 76–109. ⁶⁵ The text of the edition of the Ordinatio being apparently faulty (Ord. IV.6.3.1, n. 121, XI:332:155–8), I am here following the text given in the Vivès edition, XVI:565b: “The active form of the part belongs both to the part and to the whole; but it belongs primarily to the part, and belongs to the whole of itself (per se) but not primarily . . . .” ⁶⁶ Cross, “Varieties of Semantic Externalism,” 282.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

171

has substantial unity, the part is an instrument of the whole, for example: “the organ of the nutritive power and the whole animal with respect to nutrition.”⁶⁷ The whole animal is said to eat, even though the organ accomplishes this in fact. Nonetheless, the form is primarily the form of the part, and secondarily the form of the whole. Here, then, the essential order is between a lesser or greater participation of the same power. This is what is unique with this type of instrumental causality: the power of the part and the power of the whole are in fact one and the same power, and this power is owned “primarily” by the instrument (the part) and “secondarily” by the main agent (the whole).⁶⁸ Thus, like secondary causes, organs satisfy criteria (i), (ii), and (iii): they have their own principle of action, which they do not receive from the main agent, and they act on the same effect as the main agent. But, contrary to secondary causes, they do not satisfy criterion (iv): they do not act because of the main agent; rather, the main agent acts because of them.

c. Dispositive Causality A dispositive cause also satisfies criteria (i) and (ii), but does not satisfy criterion (iii): it does not directly contribute to the effect of the principal agent, and only causes in the patient a disposition that is a condition for the patient to receive the effect of the main agent.⁶⁹ It does not satisfy criterion (iv) either. On the face of it, this form of causality might look identical to the χ₁-type causality we have seen in Aquinas. Scotus gives the same example of heat as an altering quality that, subordinated to the action of the substance fire, prepares the introduction of the substantial form of fire in a patient.⁷⁰ Yet, for Scotus, there is here no essential order between these causes. Hence, as he puts it, in dispositive causality the order of subordination is between effects, not

⁶⁷ ⁶⁸ ⁶⁹ ⁷⁰

Ord. IV.6.3.1, n. 123, XI:332:171–2. Ord. IV.6.3.1, n. 122, XI:332. Ord. IV.6.3.1, n. 117, XI:330:126–331:128. Ord. IV.1.3.1–2, n. 317, XI:113:459–60. Ord. IV.6.3.1, n. 123, XI:332:166–8.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

172 - `  between the causes themselves.⁷¹ This means that a dispositive cause can act without a main agent. It could be an autonomous cause, because it is “a perfect principle” of action.⁷² What happens is simply that the effect of the dispositive cause obtains first, and is followed by the effect of the principal cause, as for instance the cutting of planks is generally followed by the assembling of a bench. The rift with Aquinas broadens even more when Scotus specifies that, in fact, a dispositive cause hardly qualifies as an instrumental cause.⁷³ In the very action of alteration, heat is not an instrument, since it carries out this action just by itself, in virtue of its own nature, and would do so even in the absence of a main agent. Heat would alter the patient, Scotus claims, even if, per impossibile, it existed by itself as an accidental form separate from any substance, that is, even if a substance, as the main agent, did not act through it.⁷⁴ Even when such extraordinary circumstances are not hypothesized, Scotus goes as far as maintaining that an accident has a complete entity and does not in essence act on behalf of another, but acts in its own right as efficient cause of its own effect (another accident).⁷⁵ An efficient cause need not be a subsistent being, but only a being in actuality, which an accident is.⁷⁶ As we saw with part/ whole instrumentality (organs), for Scotus forms themselves, not the composite of which they are forms, cause directly; this applies to accidental forms too.⁷⁷ In other words, Scotus denies that such causes have any χ action at all. They only have a φ action. Accordingly, Scotus attacks Aquinas’s view that substances act by the intermediary of passive and active qualities as instrumental causes, and that the action of the latter terminates at the introduction, not only of accidental forms, but also of most substantial forms—which is possible, for Aquinas, only because the substance acts through them as its main agent.⁷⁸ It is true, Scotus replies, that one may call these qualities ⁷¹ Ord. IV.6.3.1, n. 122, XI:332:161: “an order between effect and effect.” See also Ord. IV.12.2.1, n. 196, XII:355:408. ⁷² Ord. IV.6.3.1, n. 117, XI:331:130. ⁷³ Ord. IV.6.3.1, n. 117, XI:331:131–5. ⁷⁴ See also Ord. I.7.1, nn. 11–12, IV:110. ⁷⁵ Theoremata, 4, n. 11, 627:5–8. ⁷⁶ Theoremata, 4, n. 13, 627:20–2. ⁷⁷ See Cross, “Varieties of Semantic Externalism,” 277–8, 282. ⁷⁸ Aquinas, Sent. IV.12.1.2.2.ad2, quoted by Scotus, Ord. 4.12.2.1, n. 187, XII:352: “ . . . by its own power, [an accident] does not do anything that is beyond its species, but by the power of another [agent], of which it is the instrument, it can act beyond its species, just as a saw acts towards the form of a bench.”

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

173

instrumental causes inasmuch as they pave the way for the introduction of a substantial form.⁷⁹ However, they carry out this preparation by their own power. Given that “nothing is an instrumental cause with respect to that which its active power is sufficient to produce,”⁸⁰ they are therefore per se agents, whereas in all the other types of subordinated causes, the main agent must concur with them, and when it does not, they do not cause anything, at any level.⁸¹ The opposition, Scotus says, is here not so much instrumental versus principal cause, but rather dispositive versus perfective cause.⁸² There are two principles of action and two actions, with no intertwining in a common action; each has a different effect, at a different level. According to Aquinas’s view,⁸³ not only are the accidental forms instruments of the substantial forms, and not only do they act by the latter’s power, but this very fact entails that substantial forms are not immediate and proximate principles of these changes. On the contrary, for Scotus the active principle, in the case of digestion and growth, is just heat itself with respect to the preparing alteration, and the soul with respect to the introduction of the substantial form.⁸⁴ The terminus of alteration is not generation; the action of accidents never attains the introduction of a substantial form. The substance alone produces a substance.⁸⁵ If accidents were to contribute directly to the actualization of substantial forms, they would have to be principles immediately acting on the matter that receives the substantial form and to actualize their likeness in it, which is impossible, first because an accident must be received in a substance already constituted, and second because an accident must act on a substance already constituted and not on matter alone (or on the form alone).⁸⁶

⁷⁹ Ord. IV.1.1.1, n. 147, XI:53. ⁸⁰ Ord. IV.12.2.1, n. 192, XII:353:256–7. ⁸¹ Ord. 4.12.2.1, nn. 194–5, XII:354:387–355:393; Rep. IV-A.6.5–6, n. 49, 161. ⁸² Ord. IV.6.3.1, n. 117, XI:331:134–5; Ord. IV.12.2.1, n. 196, XII:355:408–9. See also Rep. IV-A.6.5–6, nn. 48–9, 161. ⁸³ Sent. IV.12.1.2.2.c. ⁸⁴ Ord. IV.12.2.1, n. 196, XII:355:404–6. See also Ord. IV.1.1.1, n. 88, XI:33: “Thus, the response taken from the example of heat in De anima II (4, 416b27–29) is clear: heat is said to be the instrument of the soul for the generation of living flesh insofar as it is the principle of an alteration that is preliminary to generation, and not because, in the instant of generation, its action would extend to the form of flesh as terminus of the action; it does not even extend to the matter of the substance flesh as patient.” ⁸⁵ Ord. IV.12.2.1, n. 198, XII:355:414–16; n. 202, XII:357:458–62. ⁸⁶ Ord. IV.1.1.1, n. 87, XI:33.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

174 - `  In summary, dispositive causes, for Scotus, have a simple, φ-type action, and produce only α-type effects, not ω-type effects. Contrary to Aquinas’s view, it is not the case that the action of the main agent permeates and directs their action so that they cause a higher effect than what they can do by nature. They are—improperly⁸⁷—said to be instruments only because it just so happens that their action produces a state of affairs that is favorable for the action of the main agent.

d. Borrowed Instrumentality Certain subservient causes do not possess by nature the form that is the principle of their action. They receive it at the time of the action and only as long as the action lasts.⁸⁸ Moreover, they receive it only from the main agent in this action, not from another cause, as might be the case for the other kinds of subservient causes (for instance, a tool made by a blacksmith and used by a carpenter). We can call this type of causality borrowed instrumentality, and these causes temporary intermediaries. They satisfy condition (i), that is to say, they have in themselves the principle of their action—even if temporarily only— and therefore perform an action of their own. Contrary to dispositive causes, they satisfy condition (iii) too, since they directly contribute to the effect of the main agent.⁸⁹ But, also contrary to dispositive causes, they do not meet criterion (ii), since they do not possess that form by nature. For instance, Scotus explains, air, as medium of vision, does not possess by nature the sensible species of color which affects the visual faculty.⁹⁰ The species is actualized in it by the illuminated color in the thing, and disappears as soon as this source of actualization ceases its action, with the result that the medium also stops acting on the visual faculty. Here, the order consists in the subordination of a power acquired during the action (in fieri, for the power of the medium to

⁸⁷ Rep. IV-A.6.5–6, n. 48, p. 161 (as opposed to what McCord-Adams says, “Can Creatures Create?,” 116). ⁸⁸ Ord. IV.6.3.1, n. 118, XI:331:137–40. ⁸⁹ Ord. IV.6.3.1, n. 118, XI:331:136–7. ⁹⁰ Ord. IV.6.3.1, n. 123, XI:332:169–70.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

175

affect the sight) to a power present before the action (in esse quieto, for the colored thing).⁹¹ Hence, again contrary to dispositive causes, criterion (iv) is satisfied: the action of the intermediary takes place only because of the action of the main agent. Furthermore, temporary intermediaries are even more dependent upon the main agent than secondary causes are.⁹² Granted, without the activity of the main cause, a secondary cause cannot have any actual action on their conjoined effect either.⁹³ A temporary intermediary, however, which receives the active form during the action, is “entirely moved” by the superior cause.⁹⁴ A secondary cause, on the contrary, permanently has a power of acting in its own order, and at the moment of the action does not receive any supplementary influence from the main cause; the superior cause simply activates it.⁹⁵ The example of the colors in the medium might induce us to think that, in the terms we used for Aquinas, we have here at last a χ-type instrumental causality (the main cause literally acts through the instrumental cause), and more precisely a χ₂ action, since the intermediary contributes to an ω-type effect (the perception of the colors in the thing by the visual faculty). However, whereas it is true that the action of this temporary intermediary terminates in the effect of the main cause, it

⁹¹ Ord. IV.6.3.1, n. 122, XI:332:161–2. ⁹² Ord. IV.1.1.1, n. 168, XI:59:211–14: “An instrument is more properly than a secondary cause said to act by the power of another thing, because a cause that does not have an active form included in its being, and only receives it at the moment it is actually moved, is more dependent towards a higher agent than a cause that has at rest an active form.” ⁹³ Ord. IV.1.1.1, n. 168, XI:59:209–10. ⁹⁴ Ord. IV.1.1.1, n. 120, XI:44:860–1. ⁹⁵ Cf. Ord. IV.1.1.1, n. 169, XI:60:215–18: “The fact that a secondary cause—that is, a cause that at rest has the active form—acts by virtue of another cause does not mean that it receives something from this other cause, but only that in the order of things it is subordinated to this other agent . . . .” Ibid., n. 170, XI:60:221–5: “ . . . When a primary cause and a secondary cause in the exact sense of the term act together, there is no new influence [of the primary cause] on the secondary cause, which would be the creation of some entity inhering in the secondary cause; rather, the influence consists in the determined order between these causes in their action on their shared effect.” Ord. IV.13.1, n. 140, XII:478:932–40: “ . . . I do not think that a secondary cause, which is sometimes called ‘instrument,’ always receive a special motion from the primary cause; rather, all there is is the subordination of its active form to the form of another thing, in some sort of order. The primary cause existing in actuality and causing at its own level, the secondary cause is bound to spring forth in act at its own level of causality. Thus, the secondary cause is called ‘moving-moved’ not because it receives motion from the primary cause by which it is moved, but because for its motion it depends on this other that moves first.”

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

176 - `  does so by an action that is no other than the one it is capable of at an ordinary level.⁹⁶ In other words, like dispositive causes, temporary intermediaries have a φ-type action, which is not boosted, so to speak, by the main agent and does not become a χ action. It only contributes to the effect of the main agent by being subordinated to it and receiving its form from it. According to the standard medieval Aristotelian vision theory, the “diaphanous element” present in the air is actualized only as long as colors in objects shine through it, and does nothing but transmit the chain of actualized “species” that reach the eyes and the visual faculty, that is, nothing beyond and above what this element is supposed to do by itself. Moreover, this type of instrumental cause, according to Scotus, does not carry out any preparatory work. One might perhaps be tempted to think that, due to its role as intermediary, the medium prepares the act of seeing. However, this is not the case: it only transmits the species and does not act on the sight before the action of the main agent attains, through it, its end, namely, the actualization of a species in the visual organ. In fact, when the species in the medium acts on the organ, this is supposed to mean that the thing itself acts on the organ. The temporary intermediary is much like the die of the money minter (except that the die is passive, as we will see): with the main agent, it acts straight on the final effect. In short, a temporary intermediary has only a φ-type action and an ω-type effect.

e. Artificial Instrumentality Like secondary causes, instrumental causes in each of the types presented previously have in themselves, temporarily or not, the form that is the principle of their action at their own level, and are therefore active.⁹⁷ In contrast, artifacts—that is to say tools, such as saws, axes, etc., and

⁹⁶ Ord. IV.1.1.1, n. 170, XI:60:225–7: “When an instrument acts, it receives an influence specifically from the thing of which it is the instrument, because [it receives] motion in actuality, and in the motion the form by which it acts at its own level.” ⁹⁷ Ord. IV.6.3.1, n.119, XI:331:145–7.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

177

presumably all other sorts of instruments⁹⁸—do not satisfy criterion (i), according to Scotus. This means that, surprisingly, Scotus contends that they do not have any intrinsic principle of motion, any active power of their own, either before the action or during the action, and therefore do not have any action of their own on an effect.⁹⁹ They are not per se active, but merely receive a local motion imparted by the main cause.¹⁰⁰ Scotus perhaps took this idea from a rather cryptic passage of Aristotle’s De generatione and corruptione (II.9, 336a 1–12), which involves a saw as an example and seems to deny that the tool is a cause of the things that come-to-be. Whatever the case may be, Scotus offers the following argument.¹⁰¹ A saw in action only has the properties of quantity, shape, and local motion. None of these properties are active forms: according to the Aristotelian doctrine of the categories of being, shape is a quality of the third species, and size or quantity are based on matter, a passive principle. Therefore, neither of them can explain an action. Thus, the saw is not properly speaking active; paradoxically, “cutting apart insofar as it is an action does not come from the saw.”¹⁰² One might object that the hardness of its blade is an active quality.¹⁰³ But first, as Richard Cross remarks, in Scotus “active” has a technical and restrictive sense: only the forms that produce something—namely, other forms of the same kind—are active. Thus, hardness is not active because it does not cause anything to become hard, whereas heat is active because it causes other things to become hot. In fact, only substantial forms and some forms in the category of quality are active.¹⁰⁴

⁹⁸ Ord. IV.1.3.1–2, n. 317; Ord. IV.6.3.1, n.124, XI:332:173–6 (“Instruments of workers,” “instruments of a profession”). ⁹⁹ Ord. IV.6.3.1, n. 126, XI:333. ¹⁰⁰ Ord. IV.1.1.1, n. 120, XI:44:853–7. ¹⁰¹ Ord. IV.1.3.1–2, n. 318, XI:113. ¹⁰² Rep. IV-A, 2. Cf. Ord. I.17.2, n. 152, V:212. ¹⁰³ Ord. IV.1.3.1–2, n. 319, XI:113–14. ¹⁰⁴ In Metaph. IX.6, n. 11, 579–80: “Not only is a relation of active power not founded in every being, but also [it is not founded] in every form. For, setting aside various less perfect things, it is clear that what is extended, as extended, does not make another nearby extended thing, previously unequal to it, to be made like it or equal to itself in quantity, just as what is hard does not make something hard that was previously not hard. But what is active, contrariwise, immediately changes what is nearby it, making it like itself, provided [every] impediment is removed, as is clear in the case of what is hot. Thus it can be argued that there is no active form other than some substantial form perhaps, and some in the genus of quality” (quoted and translated by Cross, “Varieties of Semantic Externalism,” 281).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

178 - `  Second, Scotus argues that, by his absolute power (de potentia absoluta), God could decide that a “soft hatchet” (say, a hatchet with a rubber blade) will keep the same quantity and shape while being moved to a place where a piece of wood is, and as a result will split wood. Therefore, hardness in the hatchet is not the formal principle of its action.¹⁰⁵ What happens is that when the blade is being placed where the wood is, it expels from the wood parts in proportion to its quantity and shape, because the parts of the wood cannot stay in the same place where the blade is. Ordinarily, it keeps the same quantity and shape because of its hardness, but God could obtain the same effect with a soft blade by miraculously conserving in it the same quantity and shape. Moreover, the dislodging of parts of wood is not due to any causal efficiency of the tool. Rather, it is based on the formal impossibility (impossibility due to forms) that two bodies occupy the same place (“incompossibilitas respectu ubi”). Whenever there is formal incompatibility between two properties, one of them expels the other, not in the way of an efficient cause, but formally.¹⁰⁶ Thus, the cutting of the wood consists in the expulsion of pieces of wood due to the formal incompatibility regarding place between these parts and the blade.¹⁰⁷ By now, then, it should be clear why the action of cutting apart is not an action of the tool. To act is to act by one’s own power; that which does not have an intrinsic form as principle of action, not even a form received only for the time of the action, does not act at all.¹⁰⁸ The only efficient cause of the splitting of the wood is the carpenter who moves the hatchet. Because a piece of the wood yields to the blade, one might believe that the blade moves the wood. Nevertheless, this is not true; rather, the carpenter’s arm moves both the hatchet and the wood shavings, even though it moves the wood by the intermediary of the hatchet.

¹⁰⁵ Ord. IV.1.3.1–2, n. 319, XI:113:472. ¹⁰⁶ Ord. IV.1.3.1–2, n. 320, XI:114:479–81. ¹⁰⁷ Ord. IV.1.3.1–2, n. 320, XI:114:486–9. In Ord. II.3.2.1, n. 281, VII:531, Scotus applies exactly the same analysis to a stick striking a ball: the hand gives local motion to the bat, which propels the ball because of the incompossibility of two hard bodies in the same place. This shows that artificial instrumental causes correspond to the category of subordinated causes that are moved by a higher cause but receives from it only motion, not a form, in the classification of Ord. I.3.3.2, n. 495–6, III:293–4, for which Scotus provides the example of this proto-bat-andball game or club-and-ball game. ¹⁰⁸ Ord. I.1.1.1, n. 167, II:59:201–3.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

179

As with dispositive causes, we again have here simply an order between effects.¹⁰⁹ Consequently, Scotus characterizes an instrumental artifact as something “moved, receptive of a first effect directed to a subsequent effect.”¹¹⁰ The motion imparted by the arm has two subsequent and coordinated effects: (1) the blade is moved to a certain place, which is the first effect; (2) because two bodies cannot be in the same place, the wood fiber yields, which is the second effect. The hatchet is not an efficient cause, but serves only passively as a body which is put in the place of another. The worker—the principal agent—is the efficient cause of both effects. Thus, properly speaking, artifacts have no operation of their own.¹¹¹ Naturally, they eminently satisfy criterion (iv), since they cannot cause without the main agent. But they do not satisfy criterion (i), since they do not have in themselves the principle of action (which makes criterion (ii)—receiving the action principle from the main agent at the time of the action, or not—irrelevant to this case). As for criterion (iii)—acting on the effect of the main agent or not—it is only in their case that Scotus admits what is for Aquinas a characteristic of all other instrumental causes too, namely, the dual possibilities of merely doing preparatory work or contributing directly to an effect ω. They can simply act dispositively, like the hatchet used to rough-hew a beam, or actualize the form intended by the main agent, together with the main agent, like the minter’s die. Dispositive causes never extend their action to the main effect, and temporary instrumental causes, which do extend their action to the main effect, do not introduce a preparing disposition. Tools, and tools only, can do both.¹¹² However, as we just saw, Scotist tools are nothing like Thomist tools, since for Scotus tools do not carry out any action as agents. Not only do they properly speaking have no action χ, but they do not even have an action φ.

¹⁰⁹ Ord. IV.6.3.1, n. 125, XI:333. ¹¹⁰ Ord. IV.6.3.1, n. 125, XI:333:185–6. See also Ord. IV.1.3.1–2, n. 321, XI:114:490–2: “ . . . artificial instruments are not formally active, but only receive an effect of the primary cause [motion] directed at a final effect . . . .” ¹¹¹ Theoremata, 6, nn. 1–2, 653. Ord. IV.1.1.1, n. 120, XI:44:861–2: “ . . . if a thing does not have a form in either of the two ways [at rest and in the action], it does not act at all properly speaking.” ¹¹² Ord. IV.6.3.1, n. 124, XI:333:179–83. See also Rep. IV-A.6.5–6, n. 47, 160.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

180 - ` 

4. Conclusions

Instrumental causes

Scotus’s views on instrumental causality can be summarized in the following table: (i) Contain the formal principle of their action

(ii) Do not receive it from the main agent

(iii) Act on same effect as the main agent

(iv) Act only because of the main agent

Secondary causes

Y

Y

Y

Y

Dispositive causes

Y

Y

N

N

Temporary intermediaries

Y

N

Y

Y

Parts

Y

Y

Y

N

Artifacts

N

–

Y/N

Y

Active

Passive

Among all these features of Scotus’s doctrine, I will close by highlighting the following ones. Artificial causes are in a separate category inasmuch as they do not contain at all the principle of their action and are purely passive. Scotus can therefore dismiss all the tool-based examples that Aquinas uses to explain the action of natural instrumental causes, since artificial instrumentality is completely different from natural instrumentality. Temporary intermediaries are, among subordinated causes that contain their formal principle of action (that is to say, the active subordinated causes), the only ones that do not possess it by nature and have to receive it from the main agent at the moment of the action. To the opposite, dispositive causes, are, apart from cases of part–whole causality, the only subordinated causes that act without the main agent moving them (which is why they are not instrumental causes properly speaking). Leaving likewise part–whole instrumentality aside, temporary intermediaries are the only active instrumental causes that are perfective (that is, act directly on the effect of the main agent). Moreover, this is their sole action: they do not carry out preparatory work. Conversely, dispositive causes are the only active instrumental causes that do preliminary work.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

181

Moreover, this is their sole action: as opposed to secondary causes, temporary intermediaries, and some artifacts, they do not act directly on the effect of the main agent. In other words, Scotus breaks apart the combination of dispositive and perfective activity on which Aquinas’s solution relies. Regarding the problem that was the starting point of the discussion, namely, whether creatures can create, Scotus can block the defense of Aquinas’s demonstration in Sum of Theology I.45.5 we tried to mount. Contrary to Aquinas’s claim (C.1.b), not all instrumental causes carry out preparatory work: temporary intermediaries do not. Therefore, the argument in (C.3), namely, that in a situation of creation from nothing there is nothing to be prepared, is not sufficient to bar the possibility of a creature contributing instrumentally to creation. In his own response to the question, Scotus does employ the same argument that Aquinas made, but this argument becomes one of two alternatives in a disjunction, the other member of which precludes the possibility left open by Aquinas.¹¹³ An instrumental cause, Scotus says, either produces a disposition to the effect of the main agent, or acts on the same effect as the main agent. But, if one is talking about creation, nothing exists that could be disposed for a subsequent effect, so dispositive causes cannot be employed. And if it is not a dispositive cause, but, rather, a perfective one, an instrumental cause should contain, at least temporarily (as an active intermediary, for instance, leaving aside tools and organs, which are obviously irrelevant here), the appropriate intrinsic form that enables it to contribute to creation. But such an intrinsic form cannot belong to a creature, even temporarily, as Scotus says he has shown earlier.¹¹⁴ Beyond this answer to the theological question, Scotus’s in-depth discussion of instrumental causality has farther-reaching implications. It radically transforms the understanding of instrumental causality in general. First, it may be a good thing to separate clearly, as Scotus does, artificial instrumentality from natural instrumentality. His view that instruments and tools are completely passive may be disconcerting at first sight;¹¹⁵ yet it offers some theoretical advantages. There is some ¹¹³ Ord. IV.1.1.1, n. 152, XI:54:95–55:101. ¹¹⁴ Ord. IV.1.1.1, n. 81 sqq. ¹¹⁵ Cf. Cross, “Varieties of Semantic Externalism,” 291.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

182 - `  appeal in the idea of explaining the action of tools purely in terms of local motion and parts pushing other parts. That is something that seventeenth-century mechanistic philosophies will explore farther. There is no “opening power” of any sort in a key opening a lock, nor, in a clock, is there any “power” of showing the time; there are only pieces moved by others due to the property of impenetrability. I am in no way contending that Scotus anticipates modern mechanicism, as he remains a committed Aristotelian; I am only saying that there is something promising in Scotus’s view, which could be defended. Second, Scotus radically simplifies and clarifies the theory of instrumental causality. Wisely, he avoids ascribing to instrumental causes the multi-tasking Aquinas wants them to perform, namely, having both a φ-type action and a χ-type action at the same time. For Scotus, each cause has only one type of action, the one it is capable of by its own nature. Admittedly, instrumental causes can do more, by being subordinated to a main agent, than what they can do by their own power alone, because they are integrated in a larger causal scheme. Without the main agent, they would not produce the common effect (like secondary causes in general). However, even with the main agent, their action does not extend beyond the domain of what they are by nature able to cause at their own level.¹¹⁶ By the same token, Scotus does not appeal to the mysterious enhancing of φ-type actions by a permeating influence of the main agent, which one might view as a case of magical thinking. The resulting difference with Aquinas in the conception of artificial instrumentality is striking. Whereas the skill and design of the craftsman is supposed by Aquinas to be present “intentionally” in the chisel,¹¹⁷ for Scotus the chisel remains just a chisel: a piece of metal that takes the place of parts of wood. Alternatively, think of musical instruments: Aquinas might be tempted to say that because a pianist expresses her art through a piano, she communicates something to the piano; but, for Scotus, and more aptly, the piano would be just a mechanism that is activated to produce certain sounds in an orderly way. As Bach is said to have wittily explained, there is nothing mysterious in it, you just have to hit the

¹¹⁶ Ord. IV.1.1.1, n. 171, XI:60.

¹¹⁷ See above p. 162–3.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

183

right key at the right moment.¹¹⁸ To paraphrase this: you do not have to pour your soul into the piano. Or, if you were told you should, this would be a metaphorical way of speaking, not a theory of instrumental causality. True, within the framework of borrowed instrumentality, temporary intermediaries receive a form from the principal agent. But without it, they simply do not act, and so it is not the case that they have a natural action of their own plus a higher action when they receive a form from the main agent.¹¹⁹ As for dispositive causes, we saw that they do what they do without the help of the main cause; their action is only coordinated, in subordination, with the action of the main cause. Likewise, secondary causes in general do not receive any form or power from the higher causes, but are simply conjoined and subordinated to them.¹²⁰ Accordingly, Scotus drastically limits the number of cases in which the main cause and the instrumental cause are both total causes of the same effect—which is the norm for Aquinas, even for tools, as we saw. In artificial instrumentality, like a hand striking a ball with a bat, as in natural subordinated causes, like the sun and the father, the main agent and the subservient cause are each only partial causes—even if, taken together, they form one total efficient cause with respect to their common effect. It is only when the subordinated cause holds its causal power from the main agent (as when it receives from it its essence and existence), that both are total causes.¹²¹ Finally, Scotus avoids the superposition of preparatory work and contribution to the final effect that Aquinas ascribes to certain instrumental ¹¹⁸ It is noteworthy that Scotus uses the model of artificial instruments to explain the causality of sacraments, Ord. IV.1.3.1–2, n. 321, XI:114. There is no need to suppose a power as a real form in the instrument, he adds, ibid. n. 323, XI:115:503–5. See also Ord. IV.13.1, n. 150, XII:481:12–15, where he uses the same model for the minister: “ . . . his action is an instrumental action with respect to the action of the principal agent, somewhat in the same way as cutting is with respect to the form of a bench, because this action is constantly followed by [the actualization of] the form, as per the order [between actions] established by the principal agent.” The minister’s words or doings are dispositive for the reception of grace (they certainly do not extend to this final effect of the main agent), and are, by divine institution, nomically followed by the bestowing of grace. ¹¹⁹ Ord. IV.1.1.1, n. 170, XI:60:225–7 (see note 96). ¹²⁰ Ord. IV.1.1.1, n. 170, XI:60:221–5. See also Ord. Ord. II.3.2.1, n. 281, VII:531:6–15, and Ord. IV.12.2.1, n. 190, XII:353:344–6: “ . . . the influence that the lower agent receives from the higher agent is not some form caused [in the lower agent], but only is the determined order between these causes, which act or cause simultaneously.” ¹²¹ Lect. II.34–7.5, nn. 124–7, XIX:357–9; Ord. II.34.5, n.117, VIII.416–17.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

184 - `  causes (χ₂-type causality, yielding both α0 and ω effects). The clear distinction between dispositive causes and temporary intermediaries is particularly important in this respect. Either an instrumental cause does preparatory work, but in that case it does not directly contribute to the effect of the main agent (dispositive causes); or it directly contributes to the effect of the main agent, but does not do preparatory work (temporary intermediaries).¹²² Only artifacts can do one or the other; but then again, they do not do it properly speaking: rather, it is the main agent who does both. In sum, contrary to his reputation, Scotus de-clutters rather than clutters the theory of instrumental causality. Beyond the theological question that was the occasion of his reflection, the minuteness of his analysis brings important and novel insights on how accidents, substances, and artifacts function. Boston College, Philosophy Department CNRS, PSL, LEM (UMR 8584)

Bibliography Primary Sources Aristotle. The Complete Works of Aristotle: The Revised Oxford Translation, ed. J. Barnes, 2 vols. (Princeton, NJ: Princeton University Press, 1984). John Duns Scotus. Questiones in Quartum Librum Sententiarum, dist. 1–7 [Opus Oxoniense], in Opera Omnia, vol. XVI (Paris: Vivès, 1894). John Duns Scotus. Quaestiones quodlibetales XIV–XXI [= Quodl.], in Opera Omnia, vol. XXVI (Paris: Vivès, 1895). John Duns Scotus. Ordinatio [= Ord.], in Opera Omnia, vols. I–XIV, ed. Commissio Scotistica (Vatican: Typis Vaticanis, 1950–2013). John Duns Scotus. Lectura in libros Sententiarum [= Lect.], in Opera Omnia, vols. XVI–XXI, ed. Commissio Scotistica (Vatican: Typis Vaticani, 1960–2004). John Duns Scotus. Quaestiones super libros Metaphysicorum Aristotelis [= In Metaph.], in Opera Philosophica, vol. IV, ed. R. Andrew et al. (St. Bonaventure, NY: Franciscan Institute Publications, 1997). ¹²² Ord. IV.6.3.1, n. 117–18, XI:330–1.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

185

John Duns Scotus. Theoremata, in Opera Philosophica, vol. II, ed. M. Dreyer et al. (St. Bonaventure, NY: Franciscan Institute Publications, 2004). John Duns Scotus. The Report of the Paris Lecture. Reportatio IV-A [= Rep.], ed. and trans. O. V. Bychkov and R. T. Pomplun (St. Bonaventure, NY: Franciscan Institute Publications, 2016). Peter Lombard. Sententiae in IV Libris Distinctae, 3rd ed., 2 vols. (Grottaferrata: Collegium S. Bonaventurae, 1971–81). Thomas Aquinas. Opera Omnia, [Quaestiones disputatae de anima = Q. D. de An., Quaestiones disputatae de malo = De Malo, Quaestiones disputatae de potentia Dei = De Pot., Quaestiones disputatae de veritate = De Ver., Scriptum super Sententiis = Sent., Summa contra gentiles = ScG, Summa Theologiae = STh]. Thomas Aquinas. The Summa theologica, trans. Fathers of the English Dominican Province, ed. S. K. Perry (New York: Benziger Bros., 1947), . Thomas Aquinas. Quaestiones disputatae de potentia Dei. On the Power of God, trans. The English Dominican Fathers, ed. J. Kenny (Westminster, MD: The Newman Press, 1952), . Thomas Aquinas. Contra gentiles. On the Truth of the Catholic Faith, ed. and trans. J. Kenny (New York: Hanover House, 1955–7), . Thomas Aquinas. Summa theologiae, trans. A. J. Freddoso . Secondary Sources Adams, Marilyn McCord. “Can Creatures Create?,” Philosophia 34 (2006), 101–28. Cross, Richard. The Physics of Duns Scotus: The Scientific Context of a Theological Vision (Oxford: Clarendon Press, 1998). Cross, Richard. “Some Varieties of Semantic Externalism in Duns Scotus’s Cognitive Psychology,” Vivarium 46 (2008), 275–301. Ward, Thomas. John Duns Scotus on Parts, Wholes, and Hylomorphism (Leiden: Brill, 2014). Wippel, John F. “Thomas Aquinas on Creatures as Causes of Esse,” International Philosophical Quarterly 158 (2000), 197–213.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

The Relation-Theory of Mental Acts Durand of St.-Pourçain on the Ontological Status of Mental Acts Peter John Hartman

Sometimes I change my mind. A moment ago I was thinking about my macchiato—indeed, I was looking at its color and had its taste on the tip of my tongue. But now I am thinking about this piece of paper before me. Last night, in a deep and dreamless sleep, I was thinking about nothing at all, perhaps, and in the morning, when I woke up, I started having thoughts and perceptions the moment I opened my eyes. My mind changes and it undergoes discrete mental episodes. Medieval authors called such episodes cognitive or mental acts. In this article, I will look at a debate that occurred at the turn of the fourteenth century over the ontological status of such mental acts—what is a mental act?—with a special focus on Durand of St.-Pourçain. The dominant view holds that a mental act is or at the very least requires a quality added to the mind and that episodic mental change is or at the very least requires a kind of qualitative change to the mind.¹ When I newly perceive the color on the wall, my mind is affected and undergoes a kind of qualitative change resulting in a new quality added to it; this quality either is or is required for a mental act. Such a view I will call the quality-theory of mental acts (QTMA). Philosophers as different as Godfrey of Fontaines, who maintained that thought and perception are ¹ In what follows, unless the context demands a more fine-grained term, I will use “mind” and “mental acts” to characterize what medieval authors would call the intellective and sensitive powers of the soul and their intellective and sensitive acts, or more generally, cognitive powers and their cognitive acts. As well, my focus will be on what medieval authors called “direct” mental acts—acts directed at some item outside the mind—and not “reflexive” mental acts— acts directed at some further mental act or the mind itself.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 -   

187

both completely passive,² Thomas Aquinas, who maintained that at least thought is an action, even if perception is a passion,³ and Peter John Olivi, who maintained that all cognition is active in character,⁴ still agreed that mental acts either are or at least require qualities added to the mind. The alternative view—very much a minority position—rejects QTMA: a mental act is not nor does it require a new quality added to the mind, and episodic mental change is not nor does it require a qualitative change to the mind. However, this is just a negative thesis, and there are at least two ways of developing the positive alternative. On the one hand, one might suppose that mental acts are not at all really distinct from the mind—at best a mental act is conceptually distinct, and mental change is not a kind of real change at all. We might call this theory the identitytheory of mental acts (ITMA): on this view, mental acts are really identical with the mind. Such a view, while applicable to divine cognition, seems on the surface to be untenable if applied to human minds, for non-divine mental acts are really distinct from our minds, and mental

² On Godfrey, see Antoine Côté, “L’objet et la cause de la connaissance selon Godefroid de Fontaines,” Freiburger Zeitschrift für Philosophie und Theologie 54 (2007), 407–29 and Peter Hartman, “Causation and Cognition: Durand of St.-Pourçain and Godfrey of Fontaines on the Cause of a Cognitive Act,” in A. Speer, T. Jeschke, F. Retucci, and G. Guldentops (eds.), Durandus and His Sentences Commentary: Historical, Philosophical and Theological Issues (Leuven: Peeters, 2014), 229–56, and the references therein, especially Quodl. (ed. Hoffmans) 9.19. See also footnote 14. ³ On Aquinas’s views about the ontological standing of mental acts and mental change, see especially Giorgio Pini, “Two Models of Thinking: Thomas Aquinas and John Duns Scotus,” in G. Klima (ed.), Intentionality, Cognition and Mental Representation in Medieval Philosophy (New York: Fordham University Press, 2015), 81–103 and Jeffrey Brower and Susan BrowerToland, “Aquinas on Mental Representation: Concepts and Intentionality,” The Philosophical Review 117 (2008), 193–243, and the references therein. In general, a proponent of the speciestheory of cognition, according to which cognition requires the reception of a species, will subscribe to the QTMA, as defined here. Some might hold a further stronger form wherein the mental act is some quality in addition to species—such as John Duns Scotus—but for our purposes here both views amount to the same thing. ⁴ For Olivi, see especially Sent. 2 (ed. Jansen) q. 72, 36–7, q. 58, 470–3, q. 25, q. 16, and q. 54, 274. Olivi maintains that mental acts are distinct from his notion of “virtual attention”: they are non-relational accidental qualities really distinct from and produced by the mind. For discussion, see Juhana Toivanen, Perception and the Internal Senses: Peter John Olivi on the Cognitive Functions of the Sensitive Soul (Leiden: Brill, 2014), esp. chs. 1 and 8, Han Thomas Adriaenssen, “Peter John Olivi on Perceptual Representation,” Vivarium 49 (2011), 324–52, José Silva and Juhana Toivanen, “The Active Nature of the Soul in Sense Perception: Robert Kilwardby and Peter Olivi,” Vivarium 48 (2010), 245–78, and Robert Pasnau, Theories of Cognition in the Later Middle Ages (Cambridge: Cambridge University Press, 1997), 130–4, 168–80, 236–46, 271–6.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

188    change is a real kind of change.⁵ The other way of developing the alternative to QTMA, however, allows that a mental act is a real entity really distinct from and added to the mind, and that mental change is a real kind of change. However, it is not a new non-relational entity, such as a quality, but rather a new relative entity of some sort, having as much ontological standing as certain real relations do. On this view, mental acts are wholly relations and mental change is best analyzed as a kind of relational change. This is the view—call it the relation-theory of mental acts (RTMA)—that Durand of St.-Pourçain defends. After sketching the historical backdrop of this debate (§1), I will present a brief overview of Durand’s core arguments in defense of RTMA (§2). I will then raise two objections put to the view: the first from John Duns Scotus (§3) and the second from Adam Wodeham and an anonymous Thomist (§4), assessing along the way the prospects and pitfalls of the view.

1. The Historical Background Durand puts forward his version of RTMA in the earliest (A) redaction of his commentary on Peter Lombard’s Sentences, especially Book 2, Distinction 3, Questions 5 through 8, and Distinction 38, Questions 1 and 3, as well as in a quaestio entitled “Is thinking something added to the intellect making a real composition with it?” (Utrum intelligere sit aliquid additum intellectui cum eo faciens compositionem realem; henceforth: Quaestio “Utrum intelligere . . . ”).⁶ Durand’s lectures on Book 2, ⁵ Some of Richard Drayton’s critics pinned this view on him. On Drayton, see footnote 11. For discussion of Drayton’s view as a kind of ITMA, see especially Walter Chatton, Lect. (ed. Etzkorn) 1.3.1, a. 1, Rep. (ed. Wey and Etzkorn) 2.5.1, dub. 2, 235–6, Wodeham, Lect. secunda (ed. Wood) 1.1.4, 255, Lect. oxoniensis (Paris, Bibliothèque de l’université Sorbonne 193), 1.1.2, f. 16rb. ⁶ For Durand’s first (A) and second (B) redactions I will use the critical edition unless otherwise noted (Book 2, dd. 1–5: ed. F. Retucci; Book 2, dd. 22–38: ed. F. Retucci and M. Perrone). For the third (C) redaction, I use Venice, 1571. On the dating of Durand’s various redactions, see Chris Schabel, Russell Friedman, and Irene Balcoyiannopoulou, “Peter of Palude and the Parisian Reaction to Durand of St Pourçain on Future Contingents,” Archivum Fratrum Praedicatorum 71 (2001), 183–300 and William Courtenay, “Durand in His Educational and Intellectual Context,” in Speer et al. (eds.), Durandus and His Sentences Commentary, 13–34. In Quaestio “Utrum intelligere . . . ,” which can be dated to the same period, Durand features as the opponens against an anonymous respondens. I use J. Koch’s edition of this text in Josef Koch, Durandi de S. Porciano O.P. Quaestio de natura cognitionis

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 -   

189

Distinction 3 (between 1303 and 1308) seem to have generated an almost immediate reaction from his contemporaries in the Dominican order, especially Hervaeus Natalis, who, upon hearing them or reading a pirated copy of them,⁷ dedicated an entire quodlibetal question (Quodlibet 3, Question 8) against the view, likely during Lent or Advent 1309 (and certainly no later).⁸ Durand’s position is also reproduced in and criticized by Peter of Palude in his commentary on Lombard’s Sentences Book 2, Dist. 3, Q. 4 as well as Thaddeus of Parma in Q. 15 in his commentary on Book 3 of Aristotle’s De anima, and a few years later by, among others, Peter Auriol and Gregory of Rimini in their commentaries on the Sentences: Book 2, Dist. 35, Q. 1 and Book 2, Dist. 7, Q. 2, Art. 1 respectively.⁹ The view had its proponents too. Prosper de Reggio Emilia, for instance, writing in Paris around a decade later, explicitly cites Durand as a source.¹⁰ As well, Richard Drayton defends a version of the view at

(II SENT. [A] D. 3, Q. 5) et Disputatio cum anonymo quodam necnon Determinatio Hervei Natalis O.P. (QUOL. III Q. 8) (Münster: Aschendorff, 1935), 33–42. For a discussion of this text, see Josef Koch, Durandus de S. Porciano O.P. Forschungen zum Streit um Thomas von Aquin zu Beginn des 14. Jahrhunderts (Münster: Aschendorff, 1927), 143–50. ⁷ See Durand’s comment at the end of his third redaction, f. 432rb, where he tells us his early draft was “stolen (subreptum)” from him. For discussion on this episode, see Koch, Durandus de S. Porciano, 68–9 and Isabel Iribarren, Durandus of St. Pourçain: A Dominican Theologian in the Shadows of Aquinas (Oxford: Oxford University Press, 2005). ⁸ Hervaeus appears to make a brief reference to the view in Quodl. 2.8 as well. As to the dating, Hervaeus became regent master in Easter 1307 and he was no longer regent master in Easter 1310. J. Koch argues that Quodl. 1 was delivered during Christmas 1307, Quodl. 2 during 1308, and Quodl. 3 during 1309. A. Guimarães suggests that Quodl. 2 should be later: Easter 1309 (and P. Stella agrees) and so Quodl. 3 should be Christmas 1309. However, the evidence isn’t decisive. For the status quaestionis on the dating, see Russell Friedman, “Dominican Quodlibetal Literature, ca. 1260–1330,” in C. Schabel (ed.), Theological Quodlibeta in the Middle Ages: The Fourteenth Century (Leiden: Brill, 2007), 2: 401–91. ⁹ See also Durandellus, Evidentia contra Durandum (ed. Stella) 2.10 and 2.51, Anonymous, Quaestio “Utrum actus intelligendi aliquid reale absolutum addat super potentiam intellectivam” (Paris, Bibliothèque nationale de France 14 572) f. 159v, and Peter Schwarz aka Petrus Negri, Clipeus Thomistarum (Venice, 1504) 2.46. ¹⁰ Sent. (Vatican lat. 1086) Prologus, pars 1, q. 5, qla 1, ff. 30va–34ra, especially ad 8, f. 33va–vb and ibid., pars 3, q. 3, qla 1, ff. 59rb–67rb, especially ad 1, f. 63ra. On Prosper, see William Courtenay, “Reflections on Vat. lat. 1086 and Prosper of Reggio Emilia, O.E.S.A.,” in Schabel (ed.), Theological Quodlibeta in the Middle Ages, 2: 345–58 and Peter Hartman, “Are Cognitive Habits in the Intellect? Durand of St.-Pourçain and Prosper de Reggio Emilia on Cognitive Habits,” in M. Roques and N. Faucher (eds.), The Ontology, Psychology, and Axiology of Habits (Habitus) in Medieval Philosophy (Berlin: Springer, 2018). Prosper likely started his work in the early 1310s and revised it later in 1318.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

190    Oxford around 1324, and William Crathorn, a decade after that.¹¹ Drayton’s works are not extant, but we can piece together some of his views from what his opponents report, notably Walter Chatton, Adam Wodeham, Gregory of Rimini, and John of Mirecourt.¹² It is not clear if Drayton picks the view up from Durand, although Gregory of Rimini runs together Drayton’s arguments with Durand’s arguments, suggesting a connection.¹³ Crathorn, an Oxford Dominican writing in the 1330s, uses several of Durand’s argumentative strategies in his commentary on the Sentences, Book 1, Question 1 in defense of the view—notably those that focus on the nobility principle as well as those that appeal to the separability of qualities by divine power—although he does not quote Durand directly.¹⁴

¹¹ On Drayton’s dates, see the editorial footnote in Adam Wodeham, Lect. secunda (ed. Wood) 1.1.4, 252 as well as William J. Courtenay, Adam Wodeham: An Introduction to His Life and Writings (Leiden: Brill, 1978), 63. ¹² In Chatton: Rep. et Lect. (ed. Wey) Prologus, q. 2, a. 1, Rep. (ed. Wey and Etzkorn) 2.4.5, a. 1, dub. 3, ibid., 2.5.1, dub. 2, and Lect. (ed. Wey and Etzkorn) 1.3.1, a. 1. In Wodeham: Lect. secunda (ed. Wood) 1.1.4 and Lect. oxoniensis (Paris, Bibliothèque de l’université Sorbonne 193) 1.1.2, both of whom report Chatton’s objections as well. In Rimini: Lect. (ed. Trapp) 2.7.2, a. 1, likely relying on Wodeham. In Mirecourt: Sent. (ed. Parodi) 1.19, who reports the position verbatim from Wodeham. ¹³ However, the marginalia in at least two manuscripts clearly identifies them as separate sources: one Durand’s as recited by Auriol (“hanc opinionem si bene meminit tenuit Durandus in secundo opere libro 2o sic recitat Aureolus / Durandus in secundo opere libro 2o Aureolus”), the other Drayton’s as recited by Wodeham (“sic recitat Adam / Adam”). See the apparatus criticus in Rimini, Lect. (ed. Trapp) 2.7.2, a. 1, 85 and 86. ¹⁴ Sent. (ed. Hoffman) q. 1, conc. 1. On Crathorn, see Pasnau, Theories of Cognition, 64–6, 89–100, 229–35 and Aurélien Robert, “William Crathorn on Predication and Mental Language,” Analytica 14 (2010), 227–58. For Crathorn’s arguments, see footnotes 18 and 19. Before Durand, two authors in particular—James of Viterbo and Godfrey of Fontaines—are worth mentioning in this context, both because of their influence on Durand in other aspects of his philosophy of mind, and because both might be interpreted as having rejected QTMA. James of Viterbo in Quodl. (ed. Ypma) 3.5 (ca. 1294–6) argues that mental acts are not distinct from mental powers, and so seems to defend a kind of identity-theory of mental acts. However, James still demands that mental acts require a quality added to the mind—it is just that for James this quality does not come from outside, but is innate. For discussion, see Antoine Côté, “Simplicius and James of Viterbo on Propensities,” Vivarium 47 (2009), 24–53 and John Wippel, “The Dating of James of Viterbo’s Quodlibet I and Godfrey of Fontaines’ Quodlibet VIII,” Augustiniana 24 (1974), 372–86. As well, Godfrey of Fontaines in Quodl. (ed. Hoffmans) 12.1 (circa 1296/7) might be interpreted as endorsing RTMA. As Giorgio Pini, “Can God Create My Thoughts? Scotus’s Case Against the Causal Account of Intentionality,” Journal of the History of Philosophy 49 (2011), 39–63 and Rega Wood, “Intuitive Cognition and Divine Omnipotence: Ockham in Fourteenth-Century Perspective,” in A. Hudson and M. Wilks (eds.), From Ockham to Wyclif (Oxford: Blackwell, 1987), 51–61 point out, Godfrey seems to hold in Quodl. 12.1 that mental acts are relational entities insofar as he maintains that they are passions, which are themselves relative entities referring to objects as efficient causes. While it is true that Godfrey holds that a mental act necessarily includes such a relation, so much so that even God could not separate it from its object (79–80, 82), he also holds in Quodl. (ed. Hoffmans) 9.19 that mental

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 -   

191

One final historical point: RTMA (or at least the denial of QTMA) came to be condemned three times, and it would seem that its third condemnation took place in complete ignorance of its first two condemnations. Durand’s Dominican order condemned his version of RTMA as against brother Thomas in 1314 and again in 1316, along with several other propositions.¹⁵ However, just over thirty years later, in 1347, Mirecourt’s plagiarized presentation of Wodeham’s presentation of Drayton’s version of the view comes to be condemned—and this time with no reference that I can find to the earlier condemnations.¹⁶

2. Durand’s Motivations Durand’s defense of RTMA might best be viewed as motivated by two primary concerns. First, mental acts seem to be relational in character owing to the fact that all mental acts have a kind of aboutness to them: a mental act is always directed at or about something else. As Durand puts it in Quaestio “Utrum intelligere . . . ”: A non-relational form (forma absoluta) [e.g., a quality] that serves as the foundation for a relation can be thought about without that relation; but, according to the respondent [i.e., Durand’s opponent], a mental act (intelligere) is like this, since it serves as the foundation for a relation to the object; therefore, a mental act can be thought about without this relation to its object. But if this is so, then I can think about acts are in themselves non-relational qualities (272, 275, 279), albeit special qualities, like light in the air, that exist only as long as their efficient causes exist (279–80). It is worth noting that Durand explicitly quotes Godfrey’s Quodl. 9.19 in Sent. (A) 2.3.5, e.g., pp. 151 and 156, and treats his view as a kind of QTMA. For more on Godfrey and Durand, see Hartman, “Causation and Cognition.” ¹⁵ 1314 Condemnation (ed. Koch) n. 18, 58; 1316 Condemnation (ed. Koch) n. 58, 83–4. ¹⁶ For some discussion of the Condemnations of 1347, see Robert Pasnau, Metaphysical Themes 1274–1671 (Oxford: Clarendon Press, 2011), ch. 19, Stefano Caroti, “Modi Rerum and Materialism: A Note on a Quotation of a Condemned Articulus in Some Fourteenth-Century Parisian de Anima Commentaries,” Traditio 55 (2000), 211–34, Stefano Caroti, “Les Modi Rerum . . . Encore une fois. Une source possibile de Nicole Oresme: Le Commentaire sur le livre 1er des Sentences de Jean de Mirecourt,” in S. Caroti and J. Celeyrette (eds.), Quia inter doctores est magna dissensio (Florence: Olschki, 2004), 195–222, and Jack Zupko, “John Buridan’s Philosophy of Mind: An Edition and Translation of Book III of His Questions on Aristotle’s de Anima (Third Redaction) with Commentary and Critical Interpretative Essays” (PhD Dissertation, Cornell University, 1989), 3.11.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

192    a mental act without thinking about its object, which is false. Hence, a mental act is merely a relation, and so it does not make a real composition with the intellect, nor is it some non-relational thing (res absoluta) added to the intellect. (38–9)¹⁷

Not only does Durand think that objectless thoughts are inconceivable, but he also thinks that they are metaphysically impossible: even God cannot make a thought exist on its own without a relation to an object, a metaphysical possibility that, he thinks, follows from the quality-theory of mental acts.¹⁸ A second kind of objection Durand raises against QTMA concerns the causation of mental acts. This objection can be best viewed as a combination of two metaphysical principles. On the one hand, there is what we might call the nobility principle, according to which what is less noble cannot bring about a more noble effect or act upon and affect what is more noble—a principle that can be found in both Augustine (for instance, De musica [ed. Migne] 6.5, n. 8, 1167) and Aristotle (for instance, De anima 3.5 430a18), and one to which many medieval philosophers at least paid lip service.¹⁹ On the other hand, Durand also takes seriously what has come to be called the act-potency axiom, according to which nothing one and the same can be in both act and potency at the same time with respect to the same thing, or, in other

¹⁷ A variation of this argument also shows up in Peter of Palude’s otherwise verbatim presentation of Durand’s Sent. (A) 2.3.5 (Vatican Lat. 1073) f. 20vb, and so it was included in Koch’s 1935 edition; however, it is not included in any of the recognized manuscripts containing A, and so it is not included in Retucci’s 2012 critical edition. ¹⁸ See especially Sent. (A) 2.3.5, 155–6: “God can produce the effect of any given secondary efficient cause without it; but God cannot produce a mental act (intelligere) wherein nothing is thought about (nihil intelligeretur).” Versions of this argument are present in Hervaeus Natalis, Quodl. (ed. Koch) 3.8, 43, Peter Auriol, Sent. (ed. Buytaert) Prooemium, sec. 2, art. 3, 201, and Rimini, Lect. (ed. Trapp et al.) 1.3.1, 321. Durand makes a similar argument in Sent. (C) (Venice 1571) Prologus, q. 3 and Sent. (B) (Paris, Bibliothèque nationale de France 14 454) Prologus, q. 1. A related line of attack is also present in both Durand and Crathorn, namely, that it follows from QTMA that God might separate the mental act from the mental power, and thus it is a metaphysical possibility at least that there could be a mental act without a mental power or a subject of the act, or one in the wrong subject, such as a rock. In Durand, see Sent. (A) 2.3.5, 149–50 and Quaestio “Utrum intelligere . . . ,” 33. In Crathorn, see Sent. (ed. Hoffman) q. 1, conc. 1, 74–8. ¹⁹ For discussion of this principle, see Hartman, “Causation and Cognition,” and Jean-Luc Solère, “Sine Qua Non Causality and the Context of Durand’s Early Theory of Cognition,” in Speer et al. (eds.), Durandus and His Sentences Commentary, 185–227. In Durand, see, for instance, Sent. (A) 2.3.5, 149, 152–5. In Crathorn, see Sent. (ed. Hoffman) q. 1, conc. 1, 78–9, 80–1.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 -   

193

words, nothing can affect itself or change itself. The act-potency axiom, too, was often appealed to by medieval philosophers.²⁰ Now, according to the nobility principle, material objects in the world cannot affect or change the mind, for they are less noble than the mind and its mental acts; but, according to the act-potency axiom, the mind cannot affect or change itself. Hence, it follows that either the mind does not change at all or that the change involved is a special kind of change; Durand opts for the latter.²¹ Neither the nobility principle nor the act-potency axiom is violated if we suppose that thinking and perceiving are a matter of the mind’s entering into new relations to objects in the world, and so mental change is a kind of relational change. Hence, on Durand’s view, a mental act is a new relation that the mind enters into when an object of the right sort is suitably present to it.²² In sum, then, Durand is motivated to maintain that mental acts are wholly relations and mental change relational change because of intuitions he had about the relational character of thinking and very specific metaphysical concerns about causation.²³ It is also worth noting that, ²⁰ For a general discussion of the act-potency axiom, see James Weisheipl, “The Principle Omne Quod Movetur Ab Alio Movetur in Medieval Physics,” Isis 56 (1965), 26–45. For a discussion of the principle in John Duns Scotus, see Roy Effler, John Duns Scotus and the Principle “Omne Quod Movetur Ab Alio Movetur” (St. Bonaventure, NY: The Franciscan Institute, 1962) and Peter King, “Duns Scotus on the Reality of Self-Change,” in M. Gill and J. Lennox (eds.), Self-Motion from Aristotle to Newton (Princeton, NJ: Princeton University Press, 1994), 227–90. For a discussion of the principle in Godfrey of Fontaines, see John Wippel, “Godfrey of Fontaines and the Act-Potency Axiom,” Journal of the History of Philosophy 11 (1973), 299–317. In Durand, see, for instance, Quaestio “Utrum intelligere . . . ,” 33, Sent. (A) 2.3.5, 157–60, and ibid., 2.3.8, 188, 193. ²¹ Durand also rejects co-causation views which attempt to bridge the nobility gap by supposing that the mental power together with the object bring about the mental act, for either this still will not overcome the gap, or this will collapse into a violation of the act-potency axiom. See, for instance, Sent. (A) 1.3.4 (Erfurt, Allgemeinbibl. der Stadt, Ampl. F 369), f. 77rb and Sent. (A) 2.3.5, 147–50. ²² See Sent. (A) 2.3.5, 159–63. ²³ The relation-theory of mental acts is distinct from the relation-theory of mental content or intentionality, according to which the specific content or aboutness of our mental acts is fixed owing to extrinsic, usually causal, relations to the object: I am thinking about cats and not dogs because my thought was caused (somehow) by cats and not dogs. One is a thesis about the ontology of acts, the other about their content. While Durand maintains both RTMA and the relation-theory of intentionality, his arguments for RTMA are quite independent from his arguments against what he saw as a kind of internalism about mental content in his opponents, according to which the content of a mental act is fixed by a quality intrinsic to the mind. For Durand’s relation-theory of intentionality, see Peter Hartman, “Thomas Aquinas and Durand of St.-Pourçain on Mental Representation,” History of Philosophy Quarterly 30 (2013), 19–34 and Peter Hartman, “Direct Realism With and Without Representation: John Buridan and Durand of St.-Pourçain on Intelligible Species,” in G. Klima (ed.), Questions on the Soul by John Buridan and Others (Berlin: Springer, 2017), 107–29. For discussion of the relation-theory of

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

194    for Durand, RTMA applies to all the soul’s cognitive powers: both sensation and intellectual thought are wholly relational in character, the former occurring when certain (non-relational) impressions on the bodily organs are made, and the latter occurring when certain internal physiological (non-relational) conditions, such as the presence of a phantasm in a certain inner organ, are met.²⁴

3. John Duns Scotus and the Change Objection John Duns Scotus provides us with one of the earliest and most direct attacks on RTMA in his only set of quodlibetal lectures, dedicating the whole of its thirteenth question to precisely our question: “Are acts of cognizing and desiring essentially non-relative (absoluti) or essentially relative (relativi)?” His answer is that a mental act is not wholly a relation, despite linguistic practice and even Aristotle’s own remarks to the contrary.²⁵ Scotus does admit that mental acts “include” relations to the object, but the mental act itself, if we are being precise, is a quality added to the mind upon which such relations are founded. While it has not been established who Scotus has in mind as his opponent, the Parisian lecture usually dated to 1306 or 1307 certainly overlaps with Durand’s early redaction (1303–8).²⁶ intentionality more broadly, see Calvin Normore, “Primitive Intentionality and Reduced Intentionality: Ockham’s Legacy,” Quaestio 10 (2010), 255–66. For discussion of medieval theories of intentionality in general, see Peter King, “Rethinking Representation in the Middle Ages: A Vade-Mecum to Mediaeval Theories of Mental Representation,” in H. Lagerlund (ed.), Representation and Objects of Thought in Medieval Philosophy (Aldershot: Ashgate, 2007), 81–100 and Pasnau, Theories of Cognition, among many others. ²⁴ On the extension to sensory forms of cognition, see especially Sent. (A) 2.3.5, 161–2. ²⁵ On Aristotle, see footnote 32. On linguistic practice, see especially Quodl. (ed. WaddingVivès) 13, n. 32. ²⁶ On the view as found in Quodl. 13, see Richard Cross, Duns Scotus’s Theory of Cognition (Oxford: Oxford University Press, 2014), ch. 5, and Pini, “Two Models of Thinking.” Two historical documents attest to Durand’s being in Paris in 1303 and again in 1307. See Courtenay, “Durand in His Educational and Intellectual Context.” On the dates for Scotus, see Timothy B. Noone and H. Francie Roberts, “John Duns Scotus’ Quodlibet: A Brief Study of the Manuscripts and an Edition of Question 16,” in Schabel (ed.), Theological Quodlibeta in the Middle Ages, 2: 131–98 and (with caution) Antonie Vos, The Philosophy of John Duns Scotus (Edinburgh: Edinburgh University Press, 2006). While there is no strong evidence that Scotus had Durand in mind (or Durand had Scotus in mind), the opening argument quod non in Quodl. 13 is precisely Durand’s argument from Quaestio “Utrum intelligere . . . ,” 33.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 -   

195

I will call Scotus’s primary objection to RTMA the basic change objection. He formulates it in two different ways in Quodlibet 13, and it is also, in one form or another, an objection that can be found in a number of other authors who responded to RTMA: Peter of Palude and Hervaeus Natalis both adduce it against Durand, and Walter Chatton raises it against Drayton.²⁷ The argument is straightforward: a mental act cannot wholly be a relation, because a relation must exist when its relata exist; but a mental power and the object (the relata) might exist and yet the mental act about that object might not exist.²⁸ Put another way, relational change requires a non-relational change in one or both of the relata beforehand; but mental change does not require an antecedent non-relational change in either my mental power or the object.²⁹ When I start to think about Felix, a new relation comes about between me and Felix—I am now thinking about Felix whereas before I was not. But in order for this new relation to come about, something non-relational must change beforehand, either in me (my mental power) or in Felix, or in both of us. However, clearly nothing need change in Felix in order for me to newly think about Felix. Hence, something must change in me, a non-relational change that results in a new quality added to my mental power upon which such a new relation to Felix is founded. As simple as the basic change objection is, so too is Durand’s answer to it: not all relational change requires an antecedent non-relational change in one or both of the relata, or, to put it another way, it is sometimes the case that both relata exist and yet the relation does not exist, at least for some sorts of relations. Sure, color similarities must exist when the colors exist, and a change in color similarity requires an antecedent change in one or both of the relata: in order for Socrates to become like Plato in ²⁷ Palude, Sent. (Vatican Lat. 1073) 2.3.4, qla 4, f. 20va; Hervaeus, Quodl. (ed. Koch) 3.8, 47–8; Chatton, Rep. et Lect. (ed. Wey) Prologus, q. 2, a. 1, 83, 85 and Lect. (ed. Wey and Etzkorn) 1.3.1, a. 1, 2–3. See also the anonymous respondent’s argument in Durand, Quaestio “Utrum intelligere . . . ,” 33–4. ²⁸ Scotus, Quodl. (ed. Wadding-Vivès) 13, n. 5: “All real relations necessarily follow from or are concomitant with [the existence of] their proximate foundations (or reason for founding) when the term of the relation exists. However, in the case at hand, an actual relation to the object is not necessarily concomitant with the [existence of] the mental (operativa) power [when the object exists].” ²⁹ Ibid., n. 4: “A relation, in its strict sense, is not new without a newness of some prior nonrelational entity (aliquid absolutum) in either the subject or the term. But a mental act (operatio) can be new without a newness of any prior non-relational entity in the subject (operans) and also without a newness of any non-relational entity in the term.”

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

196    terms of color, either Socrates must acquire a new color, or Plato must acquire a new color. However, color similarity is a special kind of relation. Durand calls these intrinsic relations. There are other sorts of relations, which Durand calls extrinsic relations, and these come about after the existence of the relata. One common example of an extrinsic relation is a spatial relation. If I move a ball, say, from one side of a column to the other, the ball undergoes a relational change—its spatial location changes relative to the column—even though neither its nor the column’s non-relational properties changed at all, that is, neither it nor the column underwent a qualitative or quantitative change beforehand.³⁰ A mental act, conceived of as a relation between the mental power and the object, is an extrinsic not an intrinsic relation, and so Scotus’s argument misses its mark. To be fair to Scotus, he too distinguishes between intrinsic and extrinsic relations elsewhere,³¹ and he tells us that his argument applies only to intrinsic relations. However, Scotus thinks that his opponent either is, or at least must be, committed to the thesis that a mental act is an intrinsic relation on the sole grounds, it would seem, that this is how Aristotle had classified mental relations (n. 5).³² Scotus’s basic change objection, then, rests upon the assumption that a mental act, if it is a relation and not a quality upon which a relation is founded, must be an intrinsic relation, an assumption Durand at least rejects.³³ Another form of the change objection—found in Hervaeus ³⁰ See especially Durand, Quaestio “Utrum intelligere . . . ,” 35. For the distinction between intrinsic and extrinsic relations, see especially Quodl. parisiensis (ed. Takada) 1, qq. 1–4. For Durand’s “modalist” theory of relations more generally, see Thomas Dewender, “Der ontologische Status der Relationen nach Durandus von St.-Pourcain, Hervaeus Natalis und Petrus Aureoli,” in S. Brown, T. Kobusch, and T. Dewender (eds.), Philosophical Debates at Paris in the Early Fourteenth Century (Leiden: Brill, 2009), 287–307 and the references therein. ³¹ See, for instance, Ordinatio (ed. P. Hechich) 4.13.1, nn. 42–55, 450–3 and Quaestiones super secundum et tertium De anima (ed. Bazán et al.) q. 7. For discussion, see Cross, Duns Scotus’s Theory of Cognition, 116. ³² He repeats the assertion that the mental relation must be intrinsic again at n. 11 although no argument is given why this must be so. In Aristotle, see especially Metaphysics V.15 1020b30–3, cited at nn. 5 and 22 in Quodl. 13; Physics VII.3 247b2–4 and 5.2 225b1–12, cited at nn. 4, 7, 19–21, 26, 30; and Categories ch. 8, cited at n. 25. ³³ Prosper de Reggio Emilia does seem to have endorsed the view that mental acts are relations in the category of Relation. See Sent. (Vatican Lat. 1086) Prologus, pars 1, q. 5 qla 1, ad 8, f. 33va–vb and pars 3, q. 3, qla 1, ad 1, f. 63ra. But Durand does not. While Durand is somewhat circumspect as to the appropriate category into which one should place mental acts, in his discussion of mental habits (habitus), where he maintains a similar view about the ontological standing of habits as external relations, he is explicit: even though habits are

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 -   

197

Natalis’s Quodlibet 3.8—concedes that a mental act might be a kind of extrinsic relation, like a location. However, this too won’t do because such extrinsic relations, while not requiring an antecedent quantitative or qualitative change on the side of one or both of the relata, still require an antecedent movement on the side of one or both of the relata. For instance, in order for the ball to change its spatial location relative to the column, I must move it, or the column, first. But with mental change, once more, the mind can change even if neither it nor the object changed or moved beforehand.³⁴ This form of the change objection seems more robust. Is there a kind of relational change that requires neither an antecedent non-relational (qualitative or quantitative) change nor an antecedent movement in one or both of the relata? Hervaeus Natalis can’t make sense of such a change.³⁵ But Durand thinks there is. In response to precisely this sort of objection raised to his account by an anonymous opponent, Durand writes in Quaestio “Utrum intelligere . . . ”: It seems that it is not true that something changes from not such-and-such to such-and-such only through a reception [e.g. of a quality in it] or through a production [e.g. of a quality in something else] or through its being applied [i.e. moved] to something else or something else to it. For something [can] change from not moving to actually moving, for instance, when a light object [begins to] move up, and this does not come about through applying [i.e. moving] something to it, since all that is needed in this case is the removal of an impediment; nor does it come about through its producing something in something else, as is obvious, nor through its receiving something new which enters into composition [with it, i.e., a qualitative or quantitative change]. And yet it changes, from not moving to moving in reality. Therefore, the intellect will be able to change from not

relative entities—indeed, extrinsic relative entities, which he calls modes—they can still be considered qualities in the category of Quality. See Sent. (A/B) (Paris, Bibliothèque nationale de France 12 330) 3.23.1, a. 1, edited in Peter Hartman, “Durand of St.-Pourçain on Cognitive Habits (Sent. III, D. 23, QQ. 1–2),” in M. Roques and J. Pelletier (eds.), The Language of Thought in Late Medieval Philosophy (Berlin: Springer, 2017), 331–68. ³⁴ See Hervaeus Natalis, Quodl. (ed. Koch) 3.8, 47–8. See also the anonymous respondent’s version of this argument in Quaestio “Utrum intelligere . . . ,” 33–4. ³⁵ See, for instance, Hervaeus Natalis, Quodl. (ed. Koch) 3.8, 47–8.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

198    thinking to actually thinking without its reception of something [nonrelational] that enters into composition with it. (34–5)

Consider a rock that sits upon a plank. When we remove the plank, the rock changes—before it was stationary and now it is in motion. Such a change did not require an antecedent motion on the side of the rock, nor did it require any qualitative or quantitative change on the side of the rock. All that is required in this case is that one removes the plank which was acting as an impediment to the rock’s natural operation, its downwards motion. Durand will go on to extend this intuition to mental change: the change from not thinking to thinking about Felix—a relational change—will be owing to the removal of an impediment, and not owing to either a qualitative change or a movement on the side of the mind or the object. He writes in Sent. (A) 2.3.5: Just as a heavy thing obtains its location at the same time that it obtains its heaviness, unless something is impeding it, so too someone who has an intellect immediately thinks with it, unless there is a defect on the side of the intelligible object. (159)

With the rock the impediment is something that impedes its natural operation, its downwards motion; with the mind it is something that impedes its natural operation, thinking and sensing, namely, the absence of the object. Once this impediment is removed, and so the object is suitably present—through whichever means, and this will vary case by case, be it via impressions upon our sensitive organs, the lighting conditions, or phantasms in the brain—a mental act, this relation of presence, will occur, and when the object comes to be absent, the mental act will go away. Consider Durand’s appeal to Augustine’s theory of cognition as set out in De musica here: The same is the case with an act of sensing, a view St. Augustine advanced in De musica, Book 6 [6.5 (ed. Migne), n. 10, 1169]. He writes: “Any given corporeal item outside that is present does not affect the soul but the body.” And later: “And, in brief, it seems to me that the soul in the body when it perceives is not affected by the body but it acts more attentively upon the affections of the body, and these actions, be they easy due to fit or hard due to lack of fit, do not lie hidden from it;

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 -   

199

and this is what we call sensing.” His meaning, as is clear from what he says in Book 6, is that the sense object (sensibile) does not act upon the sensitive [i.e., mental] power but upon the organ by reason of the qualities disposing it. This action, since it is present to sense, is not hidden from sense, and so [the sense object] is sensed. Nor is sensing anything but a present sense object not being hidden from sense. (Sent. [A] 2.3.5, 161–2)

In brief, then, Durand maintains that a mental act, conceived of as a relation, does not depend for its new existence upon a non-relational change in the relata—in the mental power or the object—even though it does require a non-relational change: the removal of an impediment. Like a curtain lifted, certain physiological and physical states in the organ that before impeded the presence of the object are removed, and so when conditions are such and so (and this is an empirical question) a new mental act comes about and when they are not, it does not.

4. The Objection from Agency: Adam Wodeham and an Anonymous Thomist Durand’s ultimate answer to the change objection, then, is that there is a kind of relational change that does not require an antecedent nonrelational change on the side of the relata, even though it still requires an antecedent non-relational change somewhere in the world, and mental change is precisely this kind of relational change. However, such a view entails a kind of dependence thesis: mental change—as with any kind of relational change—is dependent upon a change outside the mind, a change either in the impediment or the object. It is upon this dependence thesis that the final objection—from agency—pushes. Before I turn to the objection, it is worth highlighting a difference between two forms of the quality-theory of mental acts. On the one hand, there is a kind of QTMA that is just as much committed to the dependence thesis as RTMA is. On this view—defended for instance by Godfrey of Fontaines (see footnote 14)—all mental change is passive qualitative change brought about by the object outside the mind as efficient cause. Godfrey rejects the nobility principle (see §2) according to which what is

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

200    less noble cannot act upon or affect what is more noble, or at least he does not think it applies here. However, in terms of the dependence thesis, the two views are the same: something outside the mind must change in order for the mind to change, for the object must come to be present such that it can then act upon and bring about a qualitative change in the mind. Let us call this the passive quality-theory of mind (P-QTMA). However, there is another form of QTMA—this is the view defended by John Duns Scotus, for instance—which rejects the dependence thesis. On this view, mental change does not always require an antecedent change outside the mind: some mental changes are selfcaused. Call this the active quality-theory of mind (A-QTMA). Whereas P-QTMA rejects the nobility principle, this view rejects the act-potency axiom (see §2), according to which nothing can bring about a qualitative change in itself. For A-QTMA, the mind is able to change itself even if all other conditions outside the mind are the same. It is the proponent of the active quality-theory of mental acts (A-QTMA) who, then, will raise the final objection to Durand’s account (and, indeed, such an objection will apply mutatis mutandis to Godfrey’s passive quality-theory of mental acts). In what follows I will look at two versions of the objection from agency: the first from an anonymous Thomist critic of Durand’s view, the second from Adam Wodeham. In a quaestio edited by Josef Koch in 1930 contained in one of the manuscript witnesses to Durand’s Tractatus de habitibus, an anonymous author—whom Koch calls “a certain anonymous Thomist (Thomista quidam anonymus)”³⁶—interprets Durand as defending precisely the dependence thesis outlined above: there can be no change or variation to the mind without some change or variation outside the mind, for instance, on the side of the phantasms present in the imagination. (In this context, the imagination and its phantasms are purely corporeal entities located in the brain.)³⁷ Against this thesis, he writes: ³⁶ Quaestio “Utrum habitus acquisitus intellectualis vel moralis sit ponendus in illa potentia subiective cuius actum primo et immediate respicit” (henceforth: Quaestio “Utrum habitus”). This text is edited in Josef Koch, Durandi de S. Porciano O.P. Tractatus de habitibus. Quaestio quarta [De subiectis habituum] addita quaestione critica anonymi cuiusdam (Münster: Aschendorff, 1930), 70–80. The anonymous author self-identifies as a follower of Thomas Aquinas on p. 73. For more on this author, see Koch, Durandus de S. Porciano, 142–3 and Hartman, “Are Cognitive Habits in the Intellect?” ³⁷ See especially Durand, Tractatus de habitibus 4.8 (ed. Koch) 58.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 -   

201

Variation to the intellect is not totally explained by appeal to the imagination presenting to the intellect its object, for if this were totally explained by the imagination, then there could be no variation to the intellect without an antecedent variation to the imagination. But the consequent is false. (Quaestio “Utrum habitus” [ed. Koch] 76)

According to the anonymous Thomist, the intellect can change—it can elicit a new mental act—even if nothing on the side of the imagination and the phantasms within the imagination changes. In other words, even if everything outside the intellect remains the same, the intellect can still change. As support for this independence thesis, our author points to two phenomena where this seems to happen. First, the intellect is able to perform further deductions (mental acts) on the phantasms already present in the imagination and thus strengthen its intellectual habits associated with such deductions (“Utrum habitus,” 76). Second, the intellect must be able to organize and structure these phantasms in the first place; it must be able to elicit a mental act in virtue of which it changes the phantasms, and this mental act must be prior to and independent from the phantasms themselves (“Utrum habitus,” 77–8).³⁸ Wodeham’s version of the objection makes much the same point. He is responding to Richard Drayton’s view, which he interprets as a kind of RTMA—noting that on Drayton’s view mental change is relational change (relativa transmutatio).³⁹ His example appeals to acts of the will: With all other things involved being the same, experience teaches us that a human being can freely (libere) will and nil and not-will the very same thing. (Lect. secunda [ed. Wood] 1.1.4, sec. 3, 255)

And later: I can will and nil the very same thing without any variation whatsoever—present or future—over and above a change on the side of the will. (Lect. secunda 1.1.4, sec. 4, 261)

With all other things being held the same, I can change my mind: I can stop wanting to eat the pie and then start wanting to eat the pie.

³⁸ For further discussion of these two arguments, see Hartman, “Are Cognitive Habits in the Intellect?” ³⁹ Lect. secunda (ed. Wood) 1.1.4, sec. 4, 256.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

202    What the anonymous Thomist and Wodeham are driving at is mental agency. The dependence thesis denies that there can be a mental change without an antecedent change in the world, because mental change is just relational change, albeit of a very qualified sort. For the anonymous Thomist and Wodeham, however, there are at least some cases of mental change that can come about over and above the, so to speak, physical changes in the world and brain (or “imagination”): both at the level of the will and at the level of the intellect I can change my mind independently from any change whatsoever outside my mind. But if mental acts are mere relations dependent upon some change outside the mind, then this would be impossible, because relations, even extrinsic relations, are dependent upon some antecedent non-relational change. Therefore, a mental act must be a quality, a new quality that the mind brings about in itself.

a. Durand’s Reply to the Objection from Agency RTMA renders mental change a kind of passive affair: it is not up to me, as it were, to think about Felix when Felix or his phantasm is present, and it is not up to me to desire the pie when the pie is presented as desirable, just as it is not up to the fire to burn the branch presented to it, or for Socrates to be like Plato when Plato comes to acquire a new color.⁴⁰ Is there any way out for Durand? One might be tempted to opt for a kind of disjunctive analysis of mental acts: some mental acts—simple apprehensions and perceptions— are such that their occurrence is not up to me and occur once the impediments have been removed. Other mental acts—intellectual deliberations and acts of the will—are up to me, and so some mental acts are purely relations whereas others are qualities (together with relations). However, Durand elsewhere (Sent. 2.38) tells us that RTMA applies to all ⁴⁰ On the analogy with fire burning the branch, see Durand, Sent. (A) 2.3.8, 193: “An angel’s intellect is reduced to a [mental] act about those things it does not actually consider from the Creator as the per se cause and the present object as the per accidens cause, in the same way that fire newly burning the branch does this because of what generated it as the per se cause and from the posited branch as a sine qua non cause. Nor is there any difference between the two, except that burning passes into outside matter [ . . . ] whereas thinking does not pass into an outside matter [ . . . ].”

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 -   

203

immanent acts—including acts of the will.⁴¹ Hence, a disjunctive account seems to be off the table. Could Durand deny the dependency thesis? Could he suppose that mental acts, conceived of as extrinsic relations, can change even if nothing else changes beforehand? When the curtain has been lifted, as it were, and so Felix is suitably present to me, it could still be up to me to elicit a mental act conceived not as a new quality impressed upon the mind, but rather as a new relation to Felix. Or the apple pie: when conditions are the same, the mind can still change such that it is now related to the apple pie in the manner of a wanting whereas before it was not. After all, this seems to be, in part, what the active quality-theory of mental acts holds: when conditions are right, the mind changes itself and produces in itself a new quality upon which the relation to the object is founded, independent of any changes outside the mind. Perhaps Durand could hold the same thing, although the change here is not qualitative change but mere relational change. However, the problem with this move is that it fails to avoid the main thrust of the earlier change objection: if a mental act is a matter of the mind’s taking on a new relation, a different posture, so to speak, then this posture will still require a non-relational change beforehand, either in the mind or outside the mind, in the object or on the side of the impediment. If my hand, for instance, comes to take on a new grip such that it can now grip an object which before it could not, this new grip comes about thanks to the rearrangement of qualities already in the hand; but the intellect has no such qualities, and so any new relation that the intellect enters into requires some non-relational change beforehand. Assuming nothing else changes on the side of the object or outside the mind, this non-relational change will result in a new quality in the mind itself, upon which this new relation to the object is founded. The only option, then, seems to be to bite the bullet: all mental acts (even higher-order acts, such as acts of willing and assent) are dependent upon changes (physiological or physical) outside the mind. I cannot, for instance, change my mind and want the apple pie unless something else changes inside my body or in the pie to make it seem more desirable. ⁴¹ See Sent. (A) 2.38.3, 284: “But this is because acts of the will and other powers with regard to their own objects are not said to be a nature added to the power but rather they arise out of the relation (habitudo) between the power and the object.” See also Sent. (A) 2.38.1, 273.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

204    Indeed, Durand seems to suggest as much in his Tractatus de habitibus (ed. Koch) Question 4, Article 5. He writes: The sensitive power’s judgment about a sense object that ought to be desired or avoided can only change if there is a change either to the object or to the whole subject. (25)

For instance, to use Durand’s own examples, a dog can be trained to judge that taking bread from the left hand is to be avoided if we smack it on the nose each time it goes for bread held out in the left hand. In this case the object—the bread—comes to have something bad associated with it. We, too, can be trained in a similar fashion to view stealing or adultery as something to be avoided. Durand even suggests that dogs (and we too) can be trained to view something as impossible: if every time it goes after the bread in the left hand we take it away, the dog will decide that taking bread from the left hand is not possible. Such are cases where there is variation on the side of the object (ed. Koch, 26–8). As to variation on the side of the whole subject, Durand points out that if you sit in a cold room long enough, your sex drive drops and so certain objects will not appear as desirable as they might have otherwise (ed. Koch, 29). The mind enters into different relations—or even better a mental act is a relation—to objects outside it, but these relations depend upon and arise from states of the body and the world. Such a view is not as odd as it might seem: it is, in fact, a kind of cognitive determinism when applied to acts of the will; and it is a kind of determinism when applied to acts of the intellect. This might well explain a cryptic comment from the Dominicans in Paris who, in condemning the view, write (1314 Condemnation [ed. Koch], n. 19, 58): “We think this is dangerous to freedom of choice.”

5. Conclusions So here’s where we are. Durand defends a minority theory about the ontology of mental acts according to which a mental act is not a quality added to a mental power but rather just the relation that obtains between a mental power and a suitably present object. Richard Cross notes that this theory is “highly plausible on the face of it,”⁴² and even Chatton, ⁴² Cross, Duns Scotus’s Theory of Cognition, p. 111.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 -   

205

Wodeham, and Mirecourt, who maintain that ITMA—the view that mental acts are nothing more than the mind—is “illicit,” find RTMA to be a philosophically plausible (philosophice probabilis) theory, and one that is hard to refute.⁴³ After all, mental acts do look relational—it seems absurd to talk of a mental act that has no object or to think of a mental act without also thinking about its object, and both Aristotle and linguistic practice suggest that mental acts are relational in some way. However, on the other face of it, so to speak, the view is absurd: sometimes I think, sometimes I do not: thinking is a change; but at least according to Aristotelian views about change, this would seem to mean that it must be a qualitative change: the reception of a new quality in my mental power. This was the thrust of Scotus’s argument from change. What is more, even if thinking is a mere relational change and not a qualitative change, it is still entirely dependent upon changes outside of my mental power, and so, at least for Adam Wodeham and the anonymous Thomist, it will not guarantee a robust enough form of mental agency, according to which I can change my mind even if nothing else changes in the world. Durand seems to hold that this is exactly right: thinking is passive inasmuch as mental change is relational change and so depends upon changes outside my mental power. At least in this life, this might just be enough to explain what needs to be explained.⁴⁴ Loyola University Chicago

Bibliography Abbreviations: Lect. Lectura super Sententias Quodl. Quaestiones de quolibet Rep. Reportatio super Sententias Sent. Commentarium in Sententias

⁴³ Wodeham, Lect. secunda (ed. Wood) 1.1.4, sec. 3, 253; Chatton, Rep. (ed. Wey and Etzkorn) 2.5.1, dub. 2, 234; Mirecourt, Sent. (ed. Parodi) 1.19, n. 65. ⁴⁴ I would like to thank Francesco Pica, for providing me with an early draft of his edition of Scotus’s Quodlibet 13, and for helpful comments and feedback: an anonymous reviewer, Simona Vucu, and the audiences at the St. Louis—Chicago Area Medieval Research Group and the Cornell Summer Colloquium in Medieval Philosophy.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

206    Adam Wodeham. Lectura oxoniensis. Paris, Bibliothèque de l’université Sorbonne, Ms. 193. Adam Wodeham. Lectura secunda in librum primum Sententiarum: prologus et distinctio, ed. R. Wood (St. Bonaventure, NY: Franciscan Institute, 1990). Adriaenssen, Han Thomas. “Peter John Olivi on Perceptual Representation,” Vivarium 49 (2011), 324–52. Anonymous. Quaestio “Utrum actus intelligendi aliquid reale absoutum addat super potentiam intellectivam.” Paris, Bibliothèque nationale de France, Ms. Lat. 14 572. Anonymous. Quaestio “Utrum habitus acquisitus intellectualis vel moralis sit ponendus in illa potentia subiective cuius actum primo et immediate respicit,” in J. Koch (ed.), Durandi de S. Porciano O.P. Tractatus de habitibus. Quaestio quarta (De subiectis habituum) addita quaestione critica anonymi cuiusdam (Münster: Aschendorff, 1930). Augustine. De musica, ed. J.-P. Migne (Paris: Migne, 1841). Brower, Jeffrey, and Susan Brower-Toland. “Aquinas on Mental Representation: Concepts and Intentionality,” The Philosophical Review 117 (2008), 193–243. Caroti, Stefano. “Modi Rerum and Materialism: A Note on a Quotation of a Condemned Articulus in Some Fourteenth-Century Parisian de Anima Commentaries,” Traditio 55 (2000), 211–34. Caroti, Stefano. “Les Modi Rerum . . . Encore une fois. Une source possibile de Nicole Oresme: Le Commentaire sur le livre 1er des Sentences de Jean de Mirecourt,” in S. Caroti and J. Celeyrette (eds.), Quia inter doctores est magna dissensio (Florence: Olschki, 2004), 195–222. Côté, Antoine. “L’objet et la cause de la connaissance selon Godefroid de Fontaines,” Freiburger Zeitschrift für Philosophie und Theologie 54 (2007), 407–29. Côté, Antoine. “Simplicius and James of Viterbo on Propensities,” Vivarium 47 (2009), 24–53. Courtenay, William. Adam Wodeham: An Introduction to His Life and Writings (Leiden: Brill, 1978). Courtenay, William. “Reflections on Vat. lat. 1086 and Prosper of Reggio Emilia, O.E.S.A.,” in C. Schabel (ed.), Theological Quodlibeta in the Middle Ages: The Fourteenth Century (Leiden: Brill, 2007), 2: 345–58. Courtenay, William. “Durand in His Educational and Intellectual Context,” in A. Speer, T. Jeschke, F. Retucci, and G. Guldentops (eds.), Durandus

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 -   

207

and His Sentences Commentary: Historical, Philosophical and Theological Issues (Leuven: Peeters, 2014), 13–34. Cross, Richard. Duns Scotus’s Theory of Cognition (Oxford: Oxford University Press, 2014). Dewender, Thomas. “Der ontologische Status der Relationen nach Durandus von St.-Pourcain, Hervaeus Natalis und Petrus Aureoli,” in S. Brown, T. Kobusch, and T. Dewender (eds.), Philosophical Debates at Paris in the Early Fourteenth Century (Leiden: Brill, 2009), 287–307. Durandellus. Evidentia contra Durandum, ed. P. Stella (Tübingen: A. Francke Verlag, 2003). Durand of St.-Pourçain. Quaestio “Utrum intelligere sit aliquid additum intellectui cum eo faciens compositionem realem,” in J. Koch (ed.), Durandi de S. Porciano O.P. Quaestio de natura cognitionis (II SENT. [A] D. 3, Q. 5) et Disputatio cum anonymo quodam necnon Determinatio Hervei Natalis O.P. (QUOL. III Q. 8) (Münster: Aschendorff, 1935). Durand of St.-Pourçain. Quodlibet parisiensis I, qq. 1–4, ed. T. Takada (Kyoto: [s.n.], 1968). Durand of St.-Pourçain. Scriptum super IV libros Sententiarum (Venice, 1571). Durand of St.-Pourçain. Scriptum super IV libros Sententiarum: distinctio 23 libri tertii. Paris, Bibliothèque nationale de France, Ms. Lat. 12 330. Durand of St.-Pourçain. Scriptum super IV libros Sententiarum: distinctiones 1–5 libri secundi, ed. F. Retucci (Leuven: Peeters, 2012). Durand of St.-Pourçain. Scriptum super IV libros Sententiarum: distinctiones 22–38 libri secundi, ed. F. Retucci and M. Perrone (Leuven: Peeters, 2013). Durand of St.-Pourçain. Scriptum super IV libros Sententiarum: prologus. Paris, Bibliothèque nationale de France, Ms. Lat. 14 454. Durand of St.-Pourçain. Tractatus de habitibus: quaestio quarta, ed. J. Koch (Münster: Aschendorff, 1930). Effler, Roy. John Duns Scotus and the Principle “Omne Quod Movetur ab Alio Movetur” (St. Bonaventure, NY: The Franciscan Institute, 1962). Friedman, Russell. “Dominican Quodlibetal Literature, ca. 1260–1330,” in C. Schabel (ed.), Theological Quodlibeta in the Middle Ages: The Fourteenth Century (Leiden: Brill, 2007), 2: 401–91. Godfrey of Fontaines. Quodlibet IX, ed. J. Hoffmans (Louvain: Institut supérieur de philosophie de l’université, 1928). Godfrey of Fontaines. Quodlibeta XI–XII, ed. J. Hoffmans (Louvain: Institut supérieur de philosophie de l’université, 1932).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

208    Gregory of Rimini. Lectura super primum et secundum Sententiarum: super secundum, distinctiones 6–18, ed. A. D. Trapp (Berlin and New York: Walter de Gruyter, 1979). Gregory of Rimini. Lectura super primum et secundum Sententiarum: super primum, prologus et distinctiones 1–6, ed. A. D. Trapp, V. Marcolino, and W. Eckermann (Berlin and New York: Walter de Gruyter, 1981). Hartman, Peter. “Thomas Aquinas and Durand of St.-Pourçain on Mental Representation,” History of Philosophy Quarterly 30 (2013), 19–34. Hartman, Peter. “Causation and Cognition: Durand of St.-Pourçain and Godfrey of Fontaines on the Cause of a Cognitive Act,” in A. Speer, T. Jeschke, F. Retucci, and G. Guldentops (eds.), Durandus and His Sentences Commentary: Historical, Philosophical and Theological Issues (Leuven: Peeters, 2014), 229–56. Hartman, Peter. “Direct Realism With and Without Representation: John Buridan and Durand of St.-Pourçain on Intelligible Species,” in G. Klima (ed.), Questions on the Soul by John Buridan and Others (Berlin: Springer, 2017), 107–29. Hartman, Peter. “Durand of St.-Pourçain on Cognitive Habits (Sent. III, D. 23, QQ. 1–2),” in M. Roques and J. Pelletier (eds.), The Language of Thought in Late Medieval Philosophy (Berlin: Springer, 2017), 331–68. Hartman, Peter. “Are Cognitive Habits in the Intellect? Durand of St.-Pourçain and Prosper de Reggio Emilia on Cognitive Habits,” in M. Roques and N. Faucher (eds.), The Ontology, Psychology, and Axiology of Habits (Habitus) in Medieval Philosophy (Berlin: Springer, 2018). Hervaeus Natalis. Quodlibeta (Venice, 1513). Hervaeus Natalis. Quodlibet III quaestio 8, in J. Koch (ed.), Durandi de S. Porciano O.P. Quaestio de natura cognitionis (II SENT. [A] D. 3, Q. 5) et Disputatio cum anonymo quodam necnon Determinatio Hervei Natalis O. P. (QUOL. III Q. 8) (Münster: Aschendorff, 1935). Iribarren, Isabel. Durandus of St. Pourçain: A Dominican Theologian in the Shadows of Aquinas (Oxford: Oxford University Press, 2005). James of Viterbo. Disputatio tertia de quolibet, ed. E. Ypma (Würzburg: Augustinus Verlag, 1973). John Duns Scotus. Quaestiones quodlibeta 1–13, ed. L. Wadding (Paris: Vivès, 1895). John Duns Scotus. Quaestiones super secundum et tertium De anima, ed. C. Bazán, K. Emery, R. Green, T. Noone, R. Plevano, and A. Traver (St. Bonaventure, NY: Franciscan Institute, 2006).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 -   

209

John Duns Scotus. Ordinatio: liber IV, distinctiones 8–13, ed. P. Hechich (Vatican City: Typis Polyglottis Vaticanis, 2010). John of Mirecourt. In primum librum Sententiarum, ed. M. Parodi, , accessed June 2018. King, Peter. “Duns Scotus on the Reality of Self-Change,” in M. Gill and J. Lennox (eds.), Self-Motion from Aristotle to Newton (Princeton, NJ: Princeton University Press, 1994), 227–90. King, Peter. “Rethinking Representation in the Middle Ages: A Vade-Mecum to Mediaeval Theories of Mental Representation,” in H. Lagerlund (ed.), Representation and Objects of Thought in Medieval Philosophy (Aldershot: Ashgate, 2007), 81–100. Koch, Josef. Durandus de S. Porciano O.P. Forschungen zum Streit um Thomas von Aquin zu Beginn des 14. Jahrhunderts (Münster: Aschendorff, 1927). Koch, Josef. Durandi de S. Porciano O.P. Tractatus de habitibus. Quaestio quarta [De subiectis habituum] addita quaestione critica anonymi cuiusdam (Münster: Aschendorff, 1930). Koch, Josef. Durandi de S. Porciano O.P. Quaestio de natura cognitionis (II SENT. [A] D. 3, Q. 5) et Disputatio cum anonymo quodam necnon Determinatio Hervei Natalis O.P. (QUOL. III Q. 8) (Münster: Aschendorff, 1935). Koch, Josef. “Ein neuer Zeuge für die gegen Durandus de S. Porciano gerichtete Thomistische Irrtumsliste,” in K. Bormann (ed.), Kleine Schriften (Rome: Edizioni di Storia e Letteratura, 1973), 2: 119–25. Noone, Timothy B., and H. Francie Roberts. “John Duns Scotus’ Quodlibet: A Brief Study of the Manuscripts and an Edition of Question 16,” in C. Schabel (ed.), Theological Quodlibeta in the Middle Ages: The Fourteenth Century (Leiden: Brill, 2007), 2: 131–98. Normore, Calvin. “Primitive Intentionality and Reduced Intentionality: Ockham’s Legacy,” Quaestio 10 (2010), 255–66. Pasnau, Robert. Theories of Cognition in the Later Middle Ages (Cambridge: Cambridge University Press, 1997). Pasnau, Robert. Metaphysical Themes 1274–1671 (Oxford: Clarendon Press, 2011). Peter Auriol. Scriptum super primum librum Sententiarum, ed. E. Buytaert (St. Bonaventure, NY: Franciscan Institute, 1952), 2 vols. Peter Auriol. Commentarium in primum librum Sententiarum (Rome, 1596).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

210    Peter John Olivi. Quaestiones in secundum librum Sententiarum, ed. B. Jansen (Quaracchi: Collegium S. Bonaventurae, 1922–6), 3 vols. Peter of Palude. Commentarium in secundum librum Sententiarum. Vatican Lat. 1073. Peter Schwarz. Clipeus Thomistarum (Venice, 1504). Pini, Giorgio. “Can God Create My Thoughts? Scotus’s Case Against the Causal Account of Intentionality,” Journal of the History of Philosophy 49 (2011), 39–63. Pini, Giorgio. “Two Models of Thinking: Thomas Aquinas and John Duns Scotus,” in G. Klima (ed.), Intentionality, Cognition and Mental Representation in Medieval Philosophy (New York: Fordham University Press, 2015), 81–103. Prosper de Reggio Emilia. Commentarium in Sententias. Vatican Lat. 1086. Robert, Aurélien. “William Crathorn on Predication and Mental Language,” Analytica 14 (2010), 227–58. Schabel, Chris, Russell Friedman, and Irene Balcoyiannopoulou. “Peter of Palude and the Parisian Reaction to Durand of St Pourçain on Future Contingents,” Archivum Fratrum Praedicatorum 71 (2001), 183–300. Silva, José, and Juhana Toivanen. “The Active Nature of the Soul in Sense Perception: Robert Kilwardby and Peter Olivi,” Vivarium 48 (2010), 245–78. Solère, Jean-Luc. “Sine Qua Non Causality and the Context of Durand’s Early Theory of Cognition,” in A. Speer, T. Jeschke, F. Retucci, and G. Guldentops (eds.), Durandus and His Sentences Commentary: Historical, Philosophical and Theological Issues (Leuven: Peeters, 2014), 185–227. Thaddeus of Parma. Quaestiones de anima, ed. S. Rovighi (Milan: Vita e Pensiero, 1951). Toivanen, Juhana. Perception and the Internal Senses: Peter John Olivi on the Cognitive Functions of the Sensitive Soul (Leiden: Brill, 2014). Vos, Antonie. The Philosophy of John Duns Scotus (Edinburgh: Edinburgh University Press, 2006). Walter Chatton. Reportatio et Lectura super Sententias: collatio ad librum primum et prologus, ed. J. Wey (Toronto: PIMS, 1989). Walter Chatton. Reportatio super Sententias: liber II, ed. J. Wey and G. Etzkorn (Toronto: PIMS, 2004). Walter Chatton. Lectura super Sententias: liber I, distinctiones 3–7, ed. J. Wey and G. Etzkorn (Toronto: PIMS, 2008).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 -   

211

Weisheipl, James. “The Principle Omne Quod Movetur ab Alio Movetur in Medieval Physics,” Isis 56 (1965), 26–45. William Crathorn. Quaestiones super librum primum Sententiarum, ed. F. Hoffman (Münster: Aschendorff, 1988). Wippel, John. “Godfrey of Fontaines and the Act-Potency Axiom,” Journal of the History of Philosophy 11 (1973), 299–317. Wippel, John. “The Dating of James of Viterbo’s Quodlibet I and Godfrey of Fontaines’ Quodlibet VIII,” Augustiniana 24 (1974), 372–86. Wood, Rega. “Intuitive Cognition and Divine Omnipotence: Ockham in Fourteenth-Century Perspective,” in A. Hudson and M. Wilks (eds.), From Ockham to Wyclif (Oxford: Blackwell, 1987), 51–61. Zupko, Jack. “John Buridan’s Philosophy of Mind: An Edition and Translation of Book III of His Questions on Aristotle’s de Anima (Third Redaction) with Commentary and Critical Interpretative Essays” (PhD Dissertation, Cornell University, 1989).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

Walter Chatton’s Rejection of Final Causality Kamil Majcherek

1. Introduction a. Plan of the Paper My main goal in this paper is to examine the views on final causality presented by a fourteenth-century scholastic, Walter Chatton.¹ Chatton makes an extraordinary claim for his time, according to which there is no final causality.² As we will see in more detail below, Chatton’s position is quite complex: for whereas he claims that there are only three kinds of causality: material, formal, and efficient, and that the so-called “final causality” posited by his contemporaries should in a way be reduced to efficient causality, he also thinks that it is legitimate to still speak of four kinds of cause, that is: including final causes. My main claim, which I intend to develop below, is that Chatton’s solution consists in the thesis that ends can only be treated as causes in a metaphorical sense, in contrast to the remaining three kinds of Aristotelian cause, which are real causes. That is why we can retain the talk about final causes, as long as we are aware of the fact that they are called “causes” only metaphorically; but we must reject final causality, precisely because ends cannot cause anything in the proper sense. Talking about final causality is

¹ For an introduction to Chatton’s life and works, see Rondo Keele, “Formal Ontology in the Fourteenth Century: The Chatton Principle and Ockham’s Razor” (Doctoral Dissertation, Indiana University, 2002), 9–16; and Rondo Keele, “Walter Chatton,” in E. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Spring 2014 Edition), ² In Rep. II.1.1.1 (ed. Wey: 4.4), Chatton says emphatically: “If you ask about the fourth causality [i.e. final causality], I say that nothing of that kind should be posited.”

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 ’    

213

misleading, because it can make one believe that the production of an effect depends on its end, which is not true. Chatton’s own views on final causality are presented most extensively in the opening question of the second book of his Reportatio, based on the lectures he probably delivered in Oxford in 1321–3. Another important place in Chatton’s works is Question 7 of his Prologue to his Reportatio and Lectura, where he sketches his views in a more scattered manner.³ Chatton distinguishes three groups of views on final causality that he argues against. The textual evidence that we possess enables us to identify the most important representatives of these views: the first view is defended by Peter Auriol, the second view by John Duns Scotus and Henry of Ghent, and the third view by William of Ockham. However, delving into the details of these three positions will not be necessary for us here, because all of Chatton’s important objections against them have in fact the same form and are aimed at certain theses shared by the representatives of all three groups. Therefore, below I will only outline these aspects of the three views which I consider to be necessary to understand Chatton’s objections and Chatton’s own position. Out of all the authors criticized by Chatton, I will pay special attention to Ockham, because (for obvious chronological reasons) he was the only one of them who actually had the opportunity to respond to Chatton’s objections. My plan is the following. First, by way of further introduction, I will sketch the general structure of the medieval controversy regarding final causes and causality, and I shall try to specify what the fourteenthcentury authors take final causes and final causality to be. Then, I will focus on Chatton’s theory. I will begin by making some clarifications regarding what he takes final causes and causality to be. Then I will discuss his critique of the competing theories, as well as his own position, at the core of which lies the claim that final causes can only be called “causes” metaphorically. Then I will also briefly address the issue of why Chatton still claims that, in a sense, final causes might be retained in our

³ For the dating of Chatton’s works, see Joseph C. Wey, “Introduction,” in Walter Chatton, Reportatio et Lectura super Sententias: Prologus [Sent. Prol.], ed. J. C. Wey (Toronto: Pontifical Institute of Mediaeval Studies, 1989), 1–2; and Keele, “Walter Chatton.”

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

214   philosophy. In the final section of this paper, I will turn to William of Ockham’s response to Chatton’s criticism.

b. The Problem of Final Causes and Causality: First Approximation As I have said, I think that it is worth beginning with a quick sketch of what the controversy regarding final causes and causality consists in. First, it is a view commonly shared by medieval authors that: [T1] To understand (scire) something means to grasp its cause.⁴

This conviction, which has as its most important source the works of Aristotle, can be developed into a more detailed claim: [T2] To understand something completely means to grasp its ultimate causes in all genera of cause there are.⁵

If this is the case, then it is understandable that great importance must be attached to the enterprise of establishing what these basic genera of cause are. Here, the medievals follow again the lead of Aristotle: [T3] There are as many kinds of cause as there are basic kinds of question that may be asked with respect to something.⁶

Now, the scholastic authors are convinced that there are four such fundamental kinds of question, which can be summarized in the following way: (1) out of what? (ex qua), (2) by what? (per quam), (3) from ⁴ Cf., e.g., Aristotle, Post. An. I.2.71b8–11; Phys. II.3.194b16–18; Auct. Arist. (ed. Hamesse: 116.14; 119.54; 140.2); Aquinas, In Phys. I.1.5 (ed. Leon. II: 4–5), In Met. I.1.24–35 (ed. Marietti: 10–11); Duns Scotus, Ord. I.3.1.4 (ed. Vaticana III: 124–6); Ockham, Brevis summa prol. (OPh VI: 8–9); Exp. Phys. I.1.1 (OPh IV: 15–21). What I am offering here is obviously by no means an all-encompassing list of references, but simply a more or less haphazard set of examples. For a broader treatment of the connection between causes and explanation in scholasticism, see esp. William A. Wallace, Causality and Scientific Explanation, vol. 1: Medieval and Early Classical Science (Ann Arbor, MI: University of Michigan Press, 1972). ⁵ Aristotle, Met. I.2.982a4–982b11, I.3.983a24ff; Auct. Arist. (ed. Hamesse: 115.11; 117.29; 118.43; 140.2); Aquinas, In Met. I.2, I.4.70 (ed. Marietti: 12–15); Ockham, Exp. Phys. I.1.1 (OPh IV: 15–21). ⁶ Cf. Aristotle, Phys. II.3.194b; Aquinas, In Phys. II.5 (ed. Leon.: 69–71); In Met. V.2.771 (ed. Marietti: 212–13); Ockham, Brevis summa II.6.70–3 (OPh VI: 38); Summula II.4.5–10 (OPh VI: 220–1).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 ’    

215

what? (ex quo), and (4) for the sake of what? (propter quid or cuius causa). All these questions are taken to be subtypes of the question of why something is the case. In virtue of what are all these four kinds called “causes”? When responding to this question, medieval authors take their inspiration not from an Aristotelian, but from a Neo-Platonic text: the Liber de causis. The general description of cause that they derive from this work has usually the following form: [T4] The cause is that from the being of which something else follows.⁷

However, if A counts as a cause when B follows from its being, then it would also seem that: [T5a] For A to cause B, A must possess being at the moment of causing B.⁸

The medievals often take (T5a) to imply the claim that: [T5b] Non-being cannot be a cause of anything.

This leads us to problems springing from positing final causes as a genuine (and basic) kind of cause. For it is often acknowledged by the scholastics that, among other kinds of final causes, there are also the effects produced by our operations, which we undertake for the sake of those effects. For instance—and to use a commonly invoked example, dating back to Aristotle—we say that some people fight with one another because they want to gain rule over the rest; and this domination, as an effect which they intend to achieve by their actions, is their final cause, which we posit in our account of what is taking place.⁹ However, the desired domination is clearly something in the future from the perspective of the people fighting; and in fact it might never become present and actualized—if, for instance, it turns out that none of these people has

⁷ Cf., e.g., Aquinas, De princ. nat. 3.76–81 (ed. Leon. XLIII: 43); In Phys. II.5.8 (ed. Leon. II: 70); Ockham, Rep. III.7 (OTh VI: 200.21–201.2); III.11 (OTh VI: 381.1–4); Ord. I prol. 8 (OTh I: 222.3–4); Summa log. III.3.26.37 (OPh I: 691). ⁸ Cf., e.g., Aquinas, In An. Post. II.10.2; Duns Scotus, Q. Met. V.1.2 (OPh III: 395); Henry of Ghent, Summa II.6 (Opera XXI: 235.32); Ockham, Summula II.4.11–15 (OPh VI: 221). See also Anneliese Maier, “Das Problem der Finalkausalität um 1320,” in Anneliese Maier, Metaphysische Hintergründe der spätscholastischen Naturphilosophie (Rome: Edizioni di Storia e Letteratura, 1955), 282–3. ⁹ Cf. Aristotle, Phys. II.198a18–19.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

216   ultimately managed to dominate the rest. The situation I have described exemplifies the following scheme: [T6] At least some final causes of A’s actions undertaken for their sake do not have real being in the moment of A’s acting.¹⁰

Or, to put it differently: [T7] At least some final causes of A’s actions undertaken for their sake either (a) begin to really exist after A’s actions, or even (b) never begin to really exist.

Now, (T6) and (T7) might be problematic for a medieval proponent of teleology because of his previous commitment to (T5a). In other words, (T6) and (T7), according to which some final causes become really existent only after the actions that a given agent undertakes for their sake, might seem irreconcilable with (T5a), according to which for something to be a cause, it must exist when it causes something. However, there might also seem to be an easy way out of this problem: for it can be noticed that whereas (T6) and (T7) rely on the notion of real being, (T5a) only mentions being as such as a condition sine qua non for something to be a cause. Indeed, one of the classical solutions of the problem of final causality that the medievals are acquainted with, and which they sometimes attribute to Avicenna, consists in the claim that for an end to cause something, it is necessary for it to have not real, but intentional being— that is, being in the soul of the agent desiring it and then acting for its sake.¹¹ All authors criticized by Chatton—Auriol, Scotus, Henry of Ghent, and Ockham—agree that the causality of the end consists in its being loved and desired by an agent acting for the sake of this end.¹² As Scotus puts it in the Prologue of his Ordinatio, “the end is a cause only insofar as it moves an agent to act as being loved and desired.”¹³ It is clear,

¹⁰ Cf. Aquinas, In Phys. II.5.6 (ed. Leon. II: 70); In Met. V.3.782 (ed. Marietti: 215); Henry of Ghent, Quodl. XIII.11. ad arg. (Opera XVIII: 98); Ockham, Summula II.4 (OPh VI: 220-25); Quodl. IV.1.27–39 (OTh IX: 294). ¹¹ Cf. Avicenna, Met. V.6 (AL: 326–48). ¹² Cf., e.g., Auriol, Scriptum I.3.12.80.11–19 (ed. Buytaert: 633–4); Scotus, Ord. Prol.5.2 (ed. Vaticana I: 170.10); Henry of Ghent, Quodl. XIII.11 (Opera XVIII: 87–135); Ockham, Ord. Prol. 11 (OTh I: 306.7–22); De fine 91–7 (OTh VIII: 103), 188–227 (OTh VIII: 108–9); Summula II.4.16–28 (OPh VI: 221–2); Quodl. IV.1.13–39 (OTh IX: 293–4), IV.2.15–25 (OTh IX: 301–2). ¹³ Scotus, Ord. Prol.5.2 (ed. Vaticana I: 170.10). All translations are my own.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 ’    

217

though, that the end cannot move someone in a literal way: for that is proper to efficient causes. Hence, if the end were to move someone literally, its causality would collapse into efficient causality. Therefore, as we shall see in more detail below, later medieval authors often employ an expression that they find in Aristotle by stressing that the end is said to “move” an agent in a metaphorical way; and that means nothing more than that it is loved and desired in such a way that an agent does something because of the end and because of her love and desire. Apart from final causes (causae), we can also see authors such as Auriol, Scotus, and Ockham talk about final causality (causalitas). How are these two related? It is a view common to them all that causality— including final causality—is a relation connecting a cause to its effect. Hence, final causality just is a relation holding between a final cause and its effect. Now, the status of relations in general being hotly debated in our period, one would not be surprised to see a divergence of views regarding the status of final causality. And indeed, Auriol and Ockham declare explicitly that final causality is indistinct from the relata (termini) of the causal relation. Hence, to them, final causality just is a final cause and its effect, and nothing over and above these two.¹⁴ In contrast, Scotus—even though he does not really say anything on this issue explicitly—because he is in general a realist about relations, considering them to be really distinct from their relata, would probably embrace a realist view on final causality. What is also important is that Chatton, too, is a realist about relations. Therefore, to him, causality is a thing over and above a cause and an effect and really distinct from them.¹⁵ Hence, Chatton’s arguments have slightly different repercussions for the two groups of theories: realism and non-realism about causality. First, let me repeat that Chatton’s main claim is twofold: that (a) final causality cannot be posited, and that (b) final causes are not real causes, but can only be called “causes” in a metaphorical sense. When applied to a realist theory of final causality such as Scotus’s, that should lead us to the conclusion that (a) final causality, which the realist theory takes to be ¹⁴ Cf., e.g., Auriol, Scriptum I.3.12.80.19–38 (ed. Buytaert: 634); Ockham, De fine 188–97 (OTh VIII: 107–8); Quodl. IV.1.13–26 (OTh IX: 293–4), 57–64 (OTh IX: 295–6). ¹⁵ For Chatton’s view on relations, see esp. Rondo Keele, “Can God Make a Picasso? William of Ockham and Walter Chatton on Divine Power and Real Relations,” Journal of the History of Philosophy 45 (2007), 395–411.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

218   a real thing, cannot be posited; and that (b) final causes are not real causes. These two conclusions follow from each other, as Chatton is convinced that a real cause must always possess its proper causality. What about non-realist theories of final causality, such as Auriol’s and Ockham’s? The thesis that they share with the realists is that something counts as a real cause only when it possesses its proper causality. But, in contrast to the realist position, they think that this causality is really identical to an effect and a cause. Hence, they do not posit final causality understood as a real thing, as the realists do. That could initially seem to pose a problem: how can Chatton’s claims that there is no final causality affect the non-realist theory, the proponents of which do not posit final causality as a distinct thing? The problem might even seem to be aggravated by the fact that (as we will see later) at least some of Chatton’s arguments seem to presuppose the realist understanding of final causality. However, two things need to be noted here. First of all, the final part of this paper will give us the chance to see that Ockham, who is unambiguous about his commitment to non-realism about final causality, nevertheless treats Chatton’s objections as challenging his theory. Secondly, it must be noted that it would not be true to say that the proponents of non-realist theory do not posit final causality at all. As I have just said, for them, to posit final causality is just to posit an effect and its final cause—but that still counts as positing final causality. And indeed, Chatton himself treats non-realist theories as also positing final causality (as we will see in passage (T10) below). Along Auriol’s and Ockham’s interpretation, to say that something possesses its proper causality means to say that it is a real cause. Hence, to say that something possesses final causality is equal to the statement that this thing is a real final cause. When applied to theories of this kind, Chatton’s objections lead us to the conclusion that the ends for the sake of which agents are said to act cannot possess their proper causality and cannot count as real causes.

2. Chatton’s Position a. Final Causes In order to gain a better understanding of Chatton’s own position, we must begin by trying to clarify what he means by (a) final causes and (b)

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 ’    

219

final causality, and especially (c) why, and how, he thinks we should reject the latter while retaining the former. In order to accomplish this task, it will also be necessary to examine what he means when he says that (d) final causes are called “causes” only in a metaphorical sense, and that (e) the causality of final causes is in a sense reducible to efficient causality. Accordingly, let us begin by taking a look at what Chatton understands final causes to be. Chatton recalls Aristotle’s enumeration of the different senses in which the term “end” is used. Out of them all, the two most important senses are, first, the end as a terminus ad quem of movement or production; and second, the end as a final cause in the proper sense.¹⁶ What that means is that not every end is a final cause, because not every terminus of an operation must be a final cause.¹⁷ What are final causes, then? The most general nominal definition of the end understood as a final cause that, in Chatton’s view, can be found in the works of Aristotle has the following form: [T8] ( . . . ) the end is that for the sake of which (propter quod) an effect is produced, and as such extrinsically concurring to the production of the effect.¹⁸

As such, it is classified as one of the four basic genera of cause.¹⁹ Chatton also distinguishes between rational (or, as he calls them, moral) ends— that is, the ends that rational agents act for the sake of, and natural ends, which are ends attained by natural (i.e. non-rational) agents.²⁰ Regarding moral final causes, he agrees with the view shared by most medievals, ¹⁶ Chatton, Rep. II.1.1.1 (ed. Wey: 1), referring to Aristotle, Met. V.17.1022a5–14. Aristotle is in fact talking here about the senses of the term “limit” (“πέρας” in Greek and “terminus” in Latin), but the medievals quite often treat it as interchangeable with the term “end” (finis). ¹⁷ Cf. Chatton, Lectura I.6.33 (ed. Etzkorn-Wey: 376): “ ‘End’ is an equivocal term. First, it can mean a terminus of production ( . . . ). Second, it can mean a final cause, the love of which moves [an agent] to act.” Cf. also Ockham, De fine 23–9 (OTh VIII: 99). A commonly recalled medieval example of a terminus which is not a final cause is the death of a given organism. A locus classicus here is Averroes, In Phys. II. 4.87 (ed. Juntina IV: 39ra). ¹⁸ Chatton, Rep. II.1.1.1 (ed. Wey: 2). Let me stress again, though, that this definition concerns only ends as final causes, and not ends in some other sense, e.g. ends as mere termini or effects. ¹⁹ Cf. Chatton, Rep. II.1.1.1 (ed. Wey: 6). ²⁰ For this distinction, see, e.g., Chatton, Sent. Prol. 7.3.56–7, 77–83 (ed. Wey: 379, 380); Rep. II.1.1.2 (ed. Wey: 12.12, 15.1). Moral ends are so called, because they (or, strictly speaking, the act of choosing them) are subject to moral evaluation. Cf. Chatton, Prol. 7.3.56–7 (ed. Wey: 379).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

220   according to which they are that for the sake of which as loved or desired something is done. However, he adds that if we wish to keep talking about natural final causes as well, as Aristotle clearly seems to have wished to do, then we must describe them in a different way—because natural beings are devoid of cognition and love.²¹ Hence, he states that natural final causes are simply the results of the operations of natural beings, following from the natures (substantial forms) of these beings: [T9] First, I make a distinction: for there is a difference between intending the end naturally and intending the end morally. To intend the end naturally means to intend it in such a way that nature intends it. And that consists in nature producing an action from which such an end follows by nature or is attained either immediately or mediately (that is, through the medium of natural actions). But to intend the end morally means to love it by the love in the proper sense, in virtue of which such an action [i.e. an action by which the end is produced or attained] follows.²²

b. Rejection of Final Causality and Real Final Causes As far as I can tell, Chatton takes it as obvious (and hence not in need of an argument) that if something does not possess real causality, it cannot count as a real cause. He presents his rejection of final causality as opposed to what he considers to be the opinio communis of his predecessors and contemporaries: [T10] All commonly agree that such causality [i.e. final causality] should be posited in things. They also all agree that there are four kinds of causality and four kinds of cause. I shall not embrace this conclusion insofar as regards the four kinds of causality; still, I will posit four kinds of cause for the reasons which I will give and discuss below.²³

²¹ Cf., e.g., Chatton, Rep. II.1.1.2.35.18–20: “But they [i.e. thinkers describing final causes as things loved or desired] are incapable of providing an account of nature’s acting for an end, because this end does not move a natural agent according to its diminished being, that is its being cognized and desired, because [such an agent] does not love nor desire [anything].” ²² Chatton, Sent. Prol. 7.3.85–90 (ed. Wey: 380). Cf. also Sent. Prol. 7.3.63–5 (ed. Wey: 380). ²³ Chatton, Rep. II.1.1.1 (ed. Wey: 6). Cf. also Rep. II.1.1.4 (ed. Wey: 26.28).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 ’    

221

Chatton presents his view as opposed to the dominant position of other authors: [T11] To this question [i.e. what is the causality of the final cause] I respond differently [than all authors quoted before], that is: that the causality of the end, in virtue of which we say metaphorically that the end causes something, is the causality of the efficient cause, and that it is the same causality by which each of these [i.e. the end and the efficient cause] causes something.²⁴

Since Chatton distinguishes between moral and natural final causes, it might also be expected that he will make a distinction between at least two different senses of final causality. In the following passage, Chatton asks his opponent what he understands by the causality of natural ends, and enumerates what he thinks are the available options: [T12] By ‘the natural end causing [something]’ you can either understand that [A] this manner of speaking is real and literal, so that the sense of this phrase would be that the end is something real and it causes an effect by its proper and real causation. In such a case, this sentence would be clearly false. Or you can mean that [B] this is a metaphorical manner of speaking, so that what it really means is that when the end is said to cause [something], the cognition and love of it move [an agent] to act. This sense is not relevant here, because this is not how nature acts for the sake of an end, but is how an end is morally intended, in the way in which a free agent acts for the sake of an end regarding which she deliberates. Or you can understand it in a third way, according to which [C] ‘nature acting for the sake of an end’ is a metaphorical locution, which equals the statement that this is efficient causation, which an effect presupposes according to the natural order of forms, and from which the effect follows by nature.²⁵

As we know, option (A) is the one Chatton wants to reject: he thinks that we cannot ascribe real causality not only to natural, but also to moral

²⁴ Chatton, Rep. II.1.1.2 (ed. Wey: 9). ²⁵ Chatton, Sent. Prol. 7.3.136–47 (ed. Wey: 383).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

222   ends. Option (B) works well, Chatton thinks, for rational agents: for they can cognize their ends (even if these ends do not yet exist), love them, and act for their sake after choosing them (and the means necessary for attaining them) in the process of deliberation. Still, it is not applicable to natural agents, because they are deprived of rational cognition, love, deliberation, and volition. In the case of natural agents, then, one must opt for a different description of their acting for the sake of ends. Their ends are the things produced or attained by these agents, determined by their forms, with no cognition or love coming into play. The causation that is at play here is again only efficient causation, to which our talk of the “causality of natural ends” refers metaphorically. We can now begin our attempt to unfold Chatton’s view, and we shall begin by focusing on his main arguments against positing final causality and real final causes. As I understand Chatton, he articulates three main— and complementary—arguments against positing real final causality: (1) an argument based on parsimony, (2) an argument based on the possible non-existence of ends, (3) an argument based on the efficient character of love and desire. Let us begin with argument (1), based on the principle of ontological parsimony. Chatton is usually associated with the defense of an opposite principle, namely the principle of ontological plurality, often used specifically against Ockham and his razor. Especially important for us is the fact that Chatton uses the principle of plurality to defend the necessity of positing the relation of efficient causality as distinct from the efficient cause and its effect. The principle tells us that: [T13] When three things do not suffice to make a proposition true, a fourth must be posited, and so on.²⁶

And Chatton argues that indeed, a proposition such as “Heat produces heat” cannot be made true only by positing the first heat as an efficient cause and the second heat as its effect. For instance, God can produce the

²⁶ Chatton, Rep. I.30.1.4 (ed. Wey: 237.1-2). For more on Chatton’s understanding and use of the principle of plurality (also in relation to Ockham’s principle of parsimony), see esp. Keele, Formal Ontology.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 ’    

223

two heats by himself, in which case they would both exist, but the first heat would not be a cause of the second one.²⁷ Now, in the case of final causality and causes, we are in fact dealing with the same principle, but this time used to argue for ontological parsimony: [T14] Where positing two things is enough to make it impossible for a given proposition to be false, another, third thing should not be posited to make this proposition true.²⁸

Upon reflection, that can hardly be surprising: the two formulations of the principle follow from each another. If positing A and B is sufficient to explain a given proposition, then positing C is not necessary. But if positing A and B is not sufficient, then C must be posited, too. Straight after (T14), Chatton says: [T15] But it is impossible that when an efficient cause and its efficient causation are present, the efficient cause does not act for the sake of an end, because it cannot not produce an action from which – when posited – such an effect follows. And that is what it means for the end to really cause something: for it cannot have any other real causation, because the end might not exist when the effect is caused.²⁹

This argument works equally well for both moral and natural final causality. It is worth noting here that it treats causes as conditiones sine quae non of effects, so that from the perspective of the effect, its relation to its cause could be described as a relation of ontological dependence: A is a cause of B (and hence B is an effect of A) if B depends for its being on A. This is hardly surprising if we recall the traditional characterization of cause quoted in (T4): “The cause is that from the being of which something else follows.” Quite often, it is spelled out in the form of ²⁷ Cf. Chatton, Rep. I.30.1.4 (ed. Wey: 234). Cf. also Armand Maurer, “Ockham’s Razor and Chatton’s Anti-Razor,” in Armand Maurer, Being and Knowing: Studies in Thomas Aquinas and Later Medieval Philosophers (Toronto: Pontifical Institute of Mediaeval Studies, 1990), 431–46. ²⁸ Chatton, Rep. II.1.1.2 (ed. Wey: 9–10). A well-known example of another case in which Chatton uses the principle of parsimony is his critique of Ockham’s fictum-theory of mental representation. Cf., e.g., Katherine H. Tachau, Vision and Certitude in the Age of Ockham: Optics, Epistemology, and the Foundations of Semantics, 1250–1345 (Leiden: Brill, 1988), 180–208. ²⁹ Chatton, Rep. II.1.1.2 (ed. Wey: 9). Cf. also Rep. II.1.1 (ed. Wey: 1) and II.1.1.2 (ed. Wey: 7); Lectura I.2.2.31 (ed. Etzkorn-Wey: 234).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

224   counterfactual dependence: A is a cause of B if when A exists, B exists, and if A does not exist, B does not exist. But the converse does not hold: A can exist without B.³⁰ Chatton takes over this understanding of causality and uses it to argue against positing final causality and real final causes, as can be seen in (T15). Chatton argues here that in order to explain why A does something else, say X, for the sake of B, it is enough to posit efficient causality as connecting it to what it does, and hence to what is attained as a result of such an action—thus making A an efficient cause. Chatton claims that nothing more needs to be posited here—because for the causality of the efficient cause to be actually present means precisely to produce an effect.³¹ Hence, positing final causality and real final causes is superfluous; and the real meaning of the words of someone saying that “the end (really) causes something” is merely that it is either produced by or attained by the activity of the efficient cause, which is indeed a real cause.³² Presumably, then, Chatton also thinks that the arguments that can be effectively used to prove that efficient causality must be posited as a distinct thing do not apply here. He does not discuss this issue himself, but we may, by way of example, take his previous test: say that we have a (true) proposition “I walk for the sake of my health,” where “I” designates the agent, “walk” his action, and “health” the end of this action. Is it possible that God creates me, my action of walking and my efficient causality, through which my health is attained, but the proposition “I walk for the sake of my health” is still false? Chatton thinks not, as we already have everything needed for the end to be attained: there is the agent, his action, and the efficient causality by which the end is attained. At the end of (T15), Chatton already hints at another reason for why— apart from the preference for parsimony—final causality cannot be

³⁰ From among the thinkers whom Chatton invokes as his main opponents, it is Ockham who makes the most extensive use of such expressions. Cf., e.g., Ockham, Rep. II.12–13.19–21 (OTh V: 276). Ockham applies this characterization to final causes, too. Cf., e.g., Ockham’s Summula II.4.77–85 (OPh VI: 223–4); Ord. Prol. 11 (OTh I: 307.10–14). For more on Ockham’s view on causal explanations, see esp. Aurélien Robert, “L’explication causale selon Guillaume d’Ockham,” Quaestio 2 (2002), 241–65. Cf. also Auriol, Scriptum I.3.12.9–15 (ed. Buytaert: 652). ³¹ Cf. Chatton, Rep. II.1.1.14–18 (ed. Wey: 1): “Secondly, there is no real causality of the end in things; therefore etc. I prove the assumption: because everything else set aside, when only a sufficient efficient cause is present in the patient – for example, when fire is present in a combustible material – together with God’s concurrence, an effect can be caused.” ³² Cf. quote (T20) below.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 ’    

225

posited and final causes cannot be considered to be real causes, which constitutes his argument (b): because it is possible that the end for the sake of which an agent is said to be acting (for instance, my health from the previous example) might not exist when this action is taking place, final causality cannot be posited, and ends must be treated as causes in a merely metaphorical sense. To see an objection of this kind is hardly surprising once we realize that Chatton takes the proposition of the form “A really causes B” to be equivalent to “A is real (i.e. really existent) and is a cause of B.” Hence, Chatton treats it as evident that if the relation of final causality were to be a real relation, then its foundation—a final cause—would have to be really existent, as it is a commonly accepted claim that there can be no real relations which do not have real foundations.³³ This claim, and the entire argument, holds equally well both as applied to moral and natural ends. The common medieval solution to this problem, but only with regards to the ends of rational agents, is to insist that the causal role of final causes is in a sense mediated through something which is indeed already existent: love and desire, thus permitting ends to perform their causal role. To some authors—for instance, to Scotus—the end is a final cause according to its intentional being—that is, it is health-as-present-in-mymind that is a final cause of my taking a walk.³⁴ Other authors—most notably Ockham—refrain from making this claim, and say instead that even though being thought of and desired is a necessary condition of something’s being a final cause, still it is health-as-present-in-the-world, and not in my mind, that is a final cause. Ockham thinks that this will become obvious once we consider what kind of health one desires when one acts for its sake: for it is clearly really existent health, and not mentally existent health, that one wants to acquire.³⁵

³³ Interestingly, this problem is explicitly raised a bit earlier by another Oxford thinker, Thomas Wylton, who—being a realist about relations and final causality—argues for the rejection of the widespread claim that real relations require real terms. In the case of final causality, argues Wylton, we are dealing with a real relation the one term of which is not (or at least might not be) really existent. Cf. Cecilia Trifogli, “Thomas Wylton on Final Causality,” in A. Fidora et al. (eds.), Erfahrung und Beweis: Die Wissenschaften Von Der Natur Im 13. und 14. Jahrhundert (Berlin: Akademie Verlag, 2007), 249–64, and references thereof. ³⁴ Cf. Scotus, Q. Met. V.1.20 (OPh III: 400): “( . . . ) the end, insofar as it exists in the intention of the agent, moves her as desirable and good – because if it is in no way apprehended and does not move the agent’s affection, it does not possess the character of end and cause.” ³⁵ Cf., e.g., Ockham, De fine 343–70 (OTh VIII: 115–16).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

226   However, it is here that Chatton’s final argument (c), based on the efficient character of love and desire, comes into play.³⁶ For Chatton presents his opponents with a dilemma: either (a) the end causes something by its own reality (propria realitas), or (b) it causes something by the reality of something else, which in a sense supplements its power (aliquid supplens vicem eius). And, as we have already seen, Chatton claims that (a) cannot be accepted, because the end might not exist when it is said to cause something.³⁷ At this moment, Chatton’s opponents are likely to take recourse to the claim that the end is a real cause in virtue of its being cognized and loved by the agent. To Chatton, that equals choosing his option (b), according to which the deficient reality of the end is supplemented with the power of cognition and love. And Chatton has a response to such an argument: [T16] Either the end causes [something] by its proper reality – and this is not the case, because when the effect is caused, the end does not exist, or at least it is possible that it does not exist. Or [it causes something] by something else, which is supplementing its power – for instance, by love and cognition. But their causation is reducible to efficient, material, or formal causality. Therefore etc.³⁸

Chatton correctly recognizes that—according to the consensus view— the end plays the role of a final cause in virtue of being cognized and loved. On Chatton’s formulation, the end needs to be supplemented with the causal power of cognition and desire to be a final cause. But now it begins to look as if it was the cognition and love, and not the end itself, which should be counted as genuine causes here. And in his objections against the view of Duns Scotus, Chatton makes it even more explicit that this is what he thinks:

³⁶ Chatton always uses this argument only with regards to moral final causes, obviously because he thinks that it is only rational beings which can cognize and love non-existent things. ³⁷ Interestingly, Scotus—while considering the problem of the same sort—seems to try to bypass it by claiming that being cognized and loved by an agent just is real being. Cf. Scotus, Q. Met. V.1.77 (OPh III: 411): “Therefore, choosing the first way, according to which [an end] is a [final] cause insofar as it is in the intentions of the agent, it should be noted that it is there according to an objective and formal being. Objective being is real being; and formal being is that in virtue of which it is now intended – and that is being in the intention” (italics mine). Cf. also Robert Pasnau, “Intentionality and Final Causes,” in D. Perler (ed.), Ancient and Medieval Theories of Intentionality (Leiden: Brill, 2001), 306. ³⁸ Chatton, Rep. II.1.1.2 (ed. Wey: 7).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 ’    

227

[T17] ( . . . ) the end is not posited here in reality, but only as that because of the love of which the one who loves does something. However, in such a situation it is not the end, but the love itself that is causing [something]. What needs to be said is that the active causality of love, or the motion of love or cognition consists in producing (efficere) something, and therefore the efficient activity of love suffices.³⁹

The claim that love and desire are efficient causes is an uncontroversial claim, accepted by Scotus himself—so Chatton’s rebuttal applies to his theory.⁴⁰ Ultimately, then, to say that an end causes an action of a cognizant agent is nothing more than a metaphorical locution, which is in need of translation: [T18] ( . . . ) for the end to cause and move something morally and be that on which an effect depends means nothing more than that the love of the will by which something is loved causes something.⁴¹

c. Final Causes as Causes Metaphorically Speaking However, as we already know, Chatton insists that final causes can be retained in our philosophy on the condition that we treat them as causing an effect merely metaphorically, and not really. We will consider Chatton’s justification of this claim in the next section. First, however, let us examine in more detail the meaning of the phrase that the end “is metaphorically said to cause something.” We have already seen above, in (T11), that it plays a crucial role in Chatton’s own solution. He says that “the causality of the end, in virtue of which we say that the end causes something metaphorically, is the causality of the efficient cause, and that it is the same causality by which each of these [i.e. the end and the efficient cause] causes something.” This is a change with respect to what the earlier scholastics usually say: for according to them, the end is metaphorically said to move, but not to cause. This earlier phraseology can be found, for example, in Scotus, for whom it means that the end ³⁹ Chatton, Rep. II.1.1.1 (ed. Wey: 4.16–22). Cf. also Rep. II.1.1.2 (ed. Wey: 12.8–9) and Sent. Prol. 7.3.175 and 182 (ed. Wey: 384 and 385). ⁴⁰ Cf., e.g., Scotus, Ord. IV.49.I.2 (ed. Vaticana XIV: 293–7). ⁴¹ Chatton, Sent. Prol.7.3.203–7 (ed. Wey: 385).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

228   does not really induce any change in the agent (that is proper to the efficient cause), but is the object of her love (hence moving the agent metaphorically by attracting her).⁴² Chatton at times uses this more traditional expression, but on other occasions he replaces it with his new expression, according to which ends are metaphorically said to cause. It is also clear that he treats the two (i.e. moving and causing in a metaphorical sense) as coextensive. We can first take a look at his explanation of the former phrase: [T19] Again, why do we metaphorically say that the end moves [someone]? It is a metaphor either [A] because we ascribe real existence to a cause or causation which do not in fact really exist. That would be open falsehood, because [the end] would not cause anything with real causation. Or [B] because a given effect is caused by some real causation, which only denominates an end metaphorically (this is what we claim); and when the thing which this metaphorical locution concerns is set aside, what remains is only that there is such a motion by which the end is naturally attained.⁴³

Chatton argues that when we metaphorically say that the end moves someone, by this statement’s “being metaphorical” we can either mean (A) that the statement is simply false, because it posits a cause or causation (i.e. final causation) which do not exist, and therefore they do not really cause anything (again, because only what really exists can really cause anything). Or we can mean (B) that there is in fact real causation involved here, in virtue of which some effect is produced. However, this real causality is not final, but efficient causation. The end is simply a thing attained or produced by this other causation; and “it only denominates an end metaphorically.” We can supplement this with another passage, this time explaining why it is a metaphor to say that the end causes something:

⁴² Cf. Duns Scotus, De primo princ. 2.4 (ed. Roche: 14–16); Q. Met. V.1.39 (OPh III: 404). Cf. also François-Xavier Putallaz, “Efficience et finalité dans le Traité du premier pincipe de Jean Duns Scot,” Revue de théologie et de philosophie Lausanne 116 (1984), 135. The same explanation is given by Ockham, De fine 188–97 (OTh VIII: 107–8), 201–6 (OTh VIII: 108). ⁴³ Chatton, Rep. II.1.1.2 (ed. Wey: 10.30–11.5).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 ’    

229

[T20] Therefore, what seems to me should be said regarding this issue is that the causation by which the end is metaphorically denominated is real causation; and it is the same causation which by proper denomination denominates the efficient cause causing something. And what it really means is nothing more than that the causation of the efficient cause is that through which the effect is produced or the end is attained.⁴⁴

To put it simply, then, when we say that “the end causes something,” what that means is only that this end is attained through the operations of some efficient cause. Therefore, speaking about ends as causing something actually refers us to the activity of efficient causes: for it is these which are here denominated properly, whereas ends are only said to “cause” metaphorically. Chatton proposes several different accounts of what it means to be acting for the sake of an end, depending on what kind of being we have in mind. The three descriptions that he gives concern (a) God, (b) natural agents, and (c) rational agents. All three of them have as their crucial element the statement that the end is said to be a cause only metaphorically: [T21] However, it should be noted that there is a difference between how different beings, [a] the First Agent, [b] nature, and [c] rational agents [c1] with respect to their first acts and [c2] with respect to their acts following deliberation, should be said to act for the sake of an end. For [a] the First Agent to cause something for the sake of an end means nothing more than to want an effect for the sake of himself, so that by the same causation by which the efficient cause is properly said to cause, the end is metaphorically said to cause. In [b] the actions of nature, the end is metaphorically said to cause by the same action by which the agent is properly said to cause, so that there is such an action through the medium of which the effect is posited or the end is attained. In [c1] the causation of our first acts, the same should be said as with regards to the actions of natural agents.

⁴⁴ Chatton, Rep. II.1.1.2 (ed. Wey: 11.19–22).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

230   If we speak about [c2] the causation of the effects following deliberation, diverse causations concur here: that is, the pre-existent desire, and then the decision to carry out [the effect], and finally carrying out [the effect]. And by every action of an agent in its order, by which the agent is properly said to cause [something], the end is metaphorically said to cause [something].⁴⁵

d. Why Keep Final Causes? After all that has been said, it might seem puzzling just why Chatton insists that we can still use the term “final cause,” even if only in a metaphorical sense, instead of simply giving it up altogether. My suspicion here is that the genuine reason is that he simply does not want to reject openly Aristotle’s fourfold classification of causes.⁴⁶ However, he also makes it clear that final causes cannot function at the same level as the remaining three kinds of cause: for in contrast to the latter, which are genuine, real causes, termed “causes” in a proper way, final causes are only called “causes” in a metaphorical way.⁴⁷ And it is only in this improper way that we can retain them, as in a sense derivative from efficient causes.⁴⁸ [T22] Hence, when you ask whether the causality of the end should be posited in beings, either you want to ask whether a fourth causation or causality should be posited in beings, over and above the three causalities of the remaining three causes – and to that I respond that no. Or [you want to ask] whether from the efficient causation an end can be derivatively said to cause (an a causatione effectiva possit finis denominari causare) by metaphorical causation or causality – and to that I respond that yes, because such is the causation of the efficient ⁴⁵ Chatton, Rep. II.1.1.2 (ed. Wey: 11–12). ⁴⁶ In a passage belonging to a part of Chatton’s treatment of final causality which we are not examining closely here, raising the question of God’s status of a final cause (Rep. II.1.1.3 (ed. Wey: 19.27–30)), Chatton frankly remarks: “To the arguments given above against the second conclusion, I respond that it is difficult to explain what Aristotle says on this matter, as well as on many others. For there are equally evident passages leading in one way, as there are passages leading in another, as is clear from the defenders of opposite opinions concerning the same conclusion, all relying on his words.” ⁴⁷ Cf. Chatton, Rep. II.1.1.2 (ed. Wey: 15): “[the end] is metaphorically said to be a cause and is metaphorically [said] to move.” ⁴⁸ Cf. Chatton, Rep. II.1.1.2.26–9 (ed. Wey: 12).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 ’    

231

cause, from which the effect follows by nature. But regarding the fourth causation, I say that it has no causality proper to it in the world, but it is only efficient causality which must be posited in beings, from which the end is said to cause metaphorically.⁴⁹

Chatton argues that the traditional Aristotelian arguments for positing final causes can all be incorporated into his own theory. For instance, as we already know, Aristotle argues that the four kinds of cause can be distinguished on the basis of four fundamental kinds of question that can be asked about something. In Chapter 7 of Physics Book II, Aristotle summarizes his fourfold distinction and gives examples of each of these questions. The example illustrating our positing of final causes is the question “Why did they go to war?,” to which the response is “That they may rule.”⁵⁰ Chatton presents the following reconstruction of Aristotle’s reasoning behind his positing of final causes: [T23] That is how Aristotle’s view regarding the necessity of positing ends can be explained. For these men are moved to fight because of their desire for power. Therefore, this appetite moves them properly and actively to fight; but this active motion denominates the power metaphorically, because it is through this motion that the power is naturally attained either apparently or in reality.⁵¹

Does all this suggest that ends lose in Chatton’s hands some part of their explanatory role that they have in the view of Aristotle or the authors Chatton criticizes? Chatton does not really present any fuller theory of the explanatory function(s) of different causes, and especially final causes, so it is not easy to say. However, judging from his practice (so to speak) in the rest of his Reportatio, as well as his Lectura, it seems that he feels free to invoke ends (of course, without ascribing to them any proper causality) in his account of our intellectual and moral life. It does make sense to ask about ends, says Chatton, in these situations in which the agent’s intention does not eventuate in producing a given effect: for then one can ask about the next elements of the chain of her intentions.⁵² ⁴⁹ Chatton, Rep. II.1.2 (ed. Wey: 12.30–13.9). ⁵⁰ Cf. Aristotle, Phys. II.198a18–19. ⁵¹ Chatton, Rep. II.1.2.10–15 (ed. Wey: 13). ⁵² Cf. Chatton, Lectura I.6.34 (ed. Etzkorn-Wey: 377): “[The presence of a] final cause is proved in beings in the following way: when the effect produced is not an ultimate end, then

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

232  

3. Ockham’s Reaction to Chatton As I have already said before, out of the authors criticized by Chatton, it is only William of Ockham who had the opportunity to respond. The sequence of their exchange seems to have been the following: Ockham’s position expressed in his De fine was criticized by Chatton in his Reportatio II.1.1. In the face of Chatton’s criticism, Ockham then reiterated his view in Quodlibet IV.1–2, where he insisted that final causes should be retained as real causes—at least insofar as the operations of rational agents are concerned.⁵³ In fact, it is precisely Chatton who seems to be the main opponent whom Ockham has here in mind. Chatton’s arguments are omnipresent: most of them are rejected, although with some of them (especially the ones for the rejection of natural final causes) Ockham ultimately agrees. Ockham’s response to Chatton targets all three of Chatton’s most important arguments: (a) the parsimony argument, (b) the argument based on the possible non-existence of ends, and (c) the argument based on the efficient power of love and desire. It will be convenient here to alter the order in which Chatton’s arguments are discussed in order to follow Ockham’s own reasoning

there can arise a reasonable question of for the sake of what it was produced.” Still, I think that— at least insofar as natural philosophy is concerned—one can detect a certain shift in the way in which Chatton presents his own teleological explanations, or paraphrases the explanations given by others. For instance, in Lectura I.2.2.16–25 (ed. Etzkorn-Wey: 227–32), Chatton responds to the teleological arguments for God’s existence given by Peter Auriol. Chatton rejects the particular arguments given by Auriol, but what is most important for us here is that his own account of why certain natural phenomena are taking place is much more focused on their natures as efficient causes of their regular operations, rather than on their ends. Cf., e.g., the following passage: “To the first proof it is responded that it is not naturally evident that a heavy thing moves upwards in such a case principally to avoid vacuum. The reason why it does so is principally to preserve the natural properties of its nature, because to move upwards in such a case is a natural property of its nature. ( . . . ) To the second and third proof, the response is given for the same reason: for a phoenix and a pelican do such things principally not to preserve the universe, but because to do so is their property following from their nature. Hence, they do so principally to preserve their natural properties, without which their natures could not be preserved.” ⁵³ For some valuable remarks on the traces of Chatton’s position on final causality in Ockham’s works, see Stephen F. Brown, “Ockham and Final Causality,” in J. Wippel (ed.), Studies in Medieval Philosophy (Washington, DC: The Catholic University of America Press, 1987), 249–72. Marilyn McCord Adams, “Ockham on Final Causality: Muddying the Waters,” Franciscan Studies 56 (1998), 1–46, presents a review of all important places concerning teleology in Ockham’s works.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 ’    

233

more closely. Accordingly, we shall begin with (b) the argument based on the possible non-existence of ends. Ockham argues that final causes are special among the Aristotelian causes, because only they can cause something while not existing themselves. They owe this special status to the fact that their causality consists in being loved and desired, which is possible even when they do not exist: [T24] From that [i.e. that the end is that which is loved or desired by an agent] it follows further that the end is sometimes a cause when it does not exist, because it is sometimes desired when it does not exist (for being a final cause is nothing more than being loved or desired in the way described above). Therefore, what is special in the final cause is the fact that it can cause [something] when it does not itself exist.⁵⁴

Ockham seems to treat Chatton’s objections to be based on a misunderstanding of what the causality of the final cause consists in. As I have already mentioned before, Ockham recalls an objection very much like Chatton’s: [T25] If you say: “what does not exist is not a cause of anything,” I respond to this that this statement is false. For it to be true it would have to be added that it is also neither loved, nor desired – and only then would it follow that it is not a cause. However, the end can be loved and desired when it does not exist; and therefore it can be a final cause even though it does not exist.⁵⁵

However, even if Ockham manages to convince us that the ends can indeed be said to cause something, there still seems to be the major worry (shown by Chatton) lingering here, which is Chatton’s argument (c): if ends are causes as loved or desired, and if love and desire are efficient causes, then it seems that final causes do not have any causality of their own. Here Ockham quotes verbatim an objection put forward by Chatton (which we have seen before in (T15)), to which he responds by clarifying what he means when he says that “the end causes something by its proper reality” (rather than by the reality of something else):

⁵⁴ Ockham, Quod. IV.1.28–34 (OTh IX: 294). ⁵⁵ Ockham, Quodl. IV.1.35–9 (OTh IX: 294).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

234   [T26] To that I say that the end causes something by its proper reality in such a way that its proper reality is being desired; and it is not necessary that this reality exists when the effect is being caused, as has been said before.⁵⁶

In the face of the dilemma sketched by Chatton, Ockham wants to choose its first horn: that the end can indeed be said to cause something by its own reality. However, what it means is not that the end must really exist when it causes something, but that “its proper reality is being desired.” In other words, when I take a walk for the sake of my (not yet existent) health, I desire health according to its real being, because I want it to exist in reality. Finally, take Chatton’s (a) parsimony argument, which Ockham explicitly recalls and responds to: [T27] If you say that the proposition ‘an effect is produced’ is sufficiently made true (sufficienter verificatur) by positing an agent and a patient, so that it is superfluous to posit another cause: I respond: the existence of the end is not necessary for the production of an effect. However, in rational agents (in agentibus a proposito) it is necessary that the end is efficaciously loved and desired.⁵⁷

Take some agent A and an effect B produced by it. Ockham responds that it is not necessary to posit a final cause as really existent to explain why A has produced B. However, in the case of rational agents, it is necessary to posit a final cause—not as really existent, but as loved and desired by the agent according to the real being of this cause. Why is the case of rational beings different from all other cases? Because, as we will see in more detail in a moment, it is only them which have the ability of self-determination, consisting in choosing their own ends. Therefore, because there is a genuine variation in their actions, depending on what ends they love, we must posit final causes as an important part of our explanation of why they act as they do. It seems to me that Ockham’s insistence on retaining final causes and final causality has ultimately to do with the explanatory value that he

⁵⁶ Ockham, Quodl. IV.1.112–4 (OTh IX: 298). ⁵⁷ Ockham, Quodl. IV.1.124–9 (OTh IX: 298–9).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 ’    

235

assigns to them in the case of human actions. Ockham’s final position is that in all other cases apart from human actions, the remaining three kinds of cause (material, formal, and efficient) are enough to provide a satisfactory explanation of why something is the case. The sphere of human activity stands out here insofar as, in Ockham’s view, we need an additional, fourth kind of cause—final cause—to account for it. Ockham’s main reason for insisting on the necessity of positing human final causes is experiential: everyone, he thinks, can experience that they are motivated to act by their ends; and what everyone can experience as well is that (in contrast to the operations of nature) there is a genuine variation and unpredictability in our actions, which springs from our ability to choose different ends (which, in Ockham’s view, encompasses also our final ends) and means we use to attain them. He makes it clear when, for example, he gives up defending natural final causes in Quodlibet IV.1, while at the same time insisting that the same cannot be done with the final causes of free agents: [T28] ( . . . ) the question “for the sake of what?” cannot be asked with regards to natural actions, because it can be said that there is no question like “for the sake of what is fire generated?” This question can only be properly asked with regards to voluntary actions. Hence, the question of why these people are fighting is well asked, because they fight voluntarily to dominate one another. And it can be proved by experience – and in no other way – that a free agent acts for the sake of an end [italics mine].⁵⁸

In contrast to Chatton, who treats rational and natural final causes on a par by claiming that (even though each of them should be described in a different way) it should be denied for the same reasons regarding both of them that they are real causes, Ockham views them as two different cases, requiring different explanations. And it is the experience of our being motivated by ends which we love and desire that Ockham seems to think Chatton’s theory cannot satisfactorily account for. This is very ⁵⁸ Ockham, Quodl. IV.1.ad2.138–47 (OTh IX: 299). Cf. also Quodl. IV.2.66–71 (OTh IX: 303), where Ockham argues that it can be proved that God is a final cause of our actions: “Thirdly, I say that it can be evidently known that God can be a final cause of the effects produced by the free agents under the moon, because everyone experiences that they can produce their effects for the sake of the glory of God or for the sake of God as their final cause.”

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

236   important, because having experience of something is one of the basic conditions enumerated in one of the formulations of Ockham’s principle of parsimony, so that whenever we possess such experience, we can posit some new thing: [T29] No plurality should be posited unless it can be shown by [a] reason [in other passages: what is known per se/by demonstration] or [b] by experience, or [c] by some infallible authority.⁵⁹

Hence, in response to Chatton’s deployment of the principle of parsimony, Ockham tries to show that it should in fact lead us to positing final causes, because it is obvious to all of us that we choose our ends and act for their sake. Ockham never seems to have any doubt about the existence of the final causes of rational beings. In the Summula, for instance, he claims that Aristotle’s arguments for positing final causes in fact concern only voluntary agents: [T30] What has been said about the final causes should be understood as applying to these effects which come from cognizant and free agents, regarding whom it is manifest that they act for the sake of an end and that there is some final cause of their actions. However, regarding other beings, which do not act through cognition and will, there is a greater doubt whether they act for the sake of an end and whether a final cause should be posited in their case [italics mine].⁶⁰

This leads us to my second point: not only does Ockham think that the motivational role of ends is clear to all of us on the basis of our experience, but he also thinks that the presence—or absence—of ends as objects of our love and desire clearly makes a difference in our actions. If we change our ends, our actions are going to change as well. This makes us special among material beings, because any other material creature always acts in a uniform way. True, there might seem to be some variation present in the activities of animals, which seem to act differently in different circumstances. However, on closer inspection it turns out that this variation is merely superficial: every animal of a given

⁵⁹ Cf., e.g., Ockham, Ord. I.30.1 (OTh IV: 290.1–3); Rep. III.9 (OTh VI: 281.6–9); Rep. IV.3–5 (OTh VII: 51.26–52.3); Tract. de corp. 29.9–13 (OTh X: 157–8). ⁶⁰ Ockham, Summula II.6.1–6 (OPh VI: 227).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 ’    

237

kind always acts in the same way in the same circumstances. In other words, the change in the operations of a given animal depends only on the change of the stimuli reaching it, and not on some internal power of self-determination. Hence, a given animal will always produce a given effect, unless there is some impediment. But if this obstacle is removed, the normal operations of this creature will immediately resume.⁶¹ Natural agents are completely determined, in contrast to voluntary agents, which have the power of self-determination. Speaking about Aristotle’s argument according to which final causes must be posited, because we say that certain things happen by chance, and that is only possible if they fail to reach their ends (understood as final causes), Ockham states: [T31] This argument only concerns a free agent, who by her nature is not more inclined to one effect rather than another. However, it does not concern a natural agent, because such an agent is by its nature inclined to one determinate effect in such a way that it cannot cause an opposite effect.⁶²

The principle of parsimony returns here, but this time it is used to argue against positing natural final causes, as Ockham argues that we do not have a sufficient basis for that: [T32] ( . . . ) it cannot be sufficiently proved either by the principles known per se or by experience that an agent acting by the necessity of nature acts for the sake of such a final cause established by the will [of God]. The reason for that is that the actions of such an agent only vary when the agent, or patient, or something concurring to the actions vary. Without that, the action always follows in a uniform way. Hence, it cannot be proved that such an agent acts for the sake of an end.⁶³

Whereas Chatton’s influence with regard to Ockham’s position on rational final causes is mostly negative, in that Chatton’s view is rejected

⁶¹ Cf. Ockham, Quodl. IV.1.148–51 (OTh IX: 300). ⁶² Ockham, Quodl. IV.1.152–7 (OTh IX: 300). ⁶³ Ockham, Quodl. IV.2.27–34 (OTh IX: 302). Curiously enough, the only example that Ockham gives here and in one other response is fire, which is a non-living being. Does this mean that the case of living beings, such as animals, should be approached differently in his view? I doubt it, because he clearly presents here a dichotomy between voluntary and non-voluntary beings, which does not leave room for animals as a further kind of agents. It might just be that Ockham decided to pick the most obvious example of a fully determined being, provided by fire.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

238   and criticized, Chatton seems to ultimately play a more positive role in Ockham’s final approach to natural final causes. Indeed, some of the responses that Ockham gives to an objector positing natural final causes are literal quotations from Chatton.

4. Conclusions Over the course of our discussion, a doubt might have arisen in the mind of an attentive reader: just how radical is Chatton’s conception, which he presents in opposition to the views of others? In other words, how far is it different from these criticized views? However disappointing it can be, I must confess that I find myself incapable of making a final judgment on this issue, and so I leave it to my readers. I think that the most uncontroversial claim to make is that Chatton’s own theory is more parsimonious than any realist theory of final causality, such as Scotus’s, and hence the difference between these two seems to be most obvious. As I have already signaled before, though, the difference is less obvious between the position of Chatton and of these authors who (like Auriol and Ockham) do not posit real final causality. To be sure, we know what implications Chatton and Ockham take Chatton’s theory to have: that final causes are not real causes, and that ends cannot have their proper causality. But what further significance does this twofold claim, put forward by Chatton and rebutted by Ockham, have for the status of ends? In other words, how far does the disagreement between these two authors reach? Unfortunately, Chatton’s text, at least in its preserved form, does not help us here, since it is very enigmatic. I also find myself unable to determine what Ockham’s response to these questions would have been. Therefore, without making any definitive judgment myself, I leave the response to my readers. What is much clearer to me, though, is why—to take again Chatton and Ockham—these two authors disagree with each other. They seem to prioritize very different objectives and (to use this tricky word) intuitions when approaching the issue of final causes and causality. Chatton begins with metaphysics: with the formula that “what does not exist cannot be a real cause,” which he treats as axiomatic. This formula, always taken in its literal form, and combined with the principle of parsimony, leads him

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 ’    

239

to reject real final causality and causes. Ockham, on the other hand, seems to focus on the explanatory role of final causes, which, as he thinks, impels us to consider them as genuine causes in the case of human activity. That this is so is even clearer when one sees that it is precisely this point—the difference that the ends can make in the course of our actions—that Ockham insists on in the face of Chatton’s criticism. However, does that mean that Chatton’s theory would be incapable of providing a satisfactory account of our experience as free agents? Chatton surely does not see too much of a problem here, as he stresses the role of our acts of love, desire, and volition (which are all efficient causes) rather than the end per se, in determining our actions. The problem, then, remains unresolved—and perhaps even irresolvable, given the disagreement about some more fundamental principles which seems to be involved here.⁶⁴ University of Toronto

Bibliography Adams, Marilyn McCord. “Ockham on Final Causality: Muddying the Waters,” Franciscan Studies 56 (1998), 1–46. Aristotle. Metaphysics [Met.], tr. W. D. Ross (Oxford: Clarendon Press, 1924). Aristotle. Physics Books I and II [Phys.], tr. W. Charlton (Oxford: Oxford University Press, 1984). Aristotle. Posterior Analytics [An. Post.], 2nd edition, tr. J. Barnes (Oxford: Oxford University Press, 1994). Averroes. Aristotelis Stagiritae de physico auditu libri octo, cum Averrois Cordubensis variis in eosdem commentariis [In Phys.], vol. 4 of Aristotelis opera (Venice: apud Iunctas, 1550).

⁶⁴ Some ideas for this paper were initially born during the preparation of one of my MPhil essays at the University of Cambridge. This work was supervised by John Marenbon, to whom I am indebted for his help. I would also like to thank Martin Pickavé for his very helpful remarks on the penultimate draft of this paper. I owe a special debt of gratitude, though, to the anonymous referee, whose insightful comments forced me to rethink many of my initial theses, and thus (I hope) helped me improve this paper.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

240   Averroes. Aristotelis Metaphysicorum libri XIV cum Averrois Cordubensis in eosdem commentariis et epitome [In Met.], vol. 8 of Aristotelis opera (Venice: apud Iunctas, 1552). Avicenna. Liber de philosophia prima sive scientia divina V–X [Met.], ed. S. van Riet (Leiden: Brill, 2012). Brown, Stephen F. “Ockham and Final Causality,” in J. Wippel (ed.), Studies in Medieval Philosophy (Washington, DC: The Catholic University of America Press, 1987), 249–72. Henninger, Mark G. “John Duns Scotus and Peter Auriol on the Ontological Status of Relations,” Quaestio 13 (2013), 221–42. Henry of Ghent. Quodlibet XIII [Quodl.], ed. J. Descorte, vol. 18 of Opera omnia (Leuven: Leuven University Press, 1987). Henry of Ghent. Summa (Quaestiones ordinariae), art. I–V [Summa], ed. G. A. Wilson (Opera omnia vol. 21) (Leuven: Leuven University Press, 2004). John Duns Scotus. The De Primo Principio of John Duns Scotus. A Revised Text and a Translation [De Primo Princ.], ed. E. Roche (St. Bonaventure, NY: Franciscan Institute, 1949). John Duns Scotus. Ordinatio. Liber primus, Prologus [Ord.], ed. C. Balic et al. (Opera omnia vol. 1) (Vatican: Commissio Scotistica, 1950). John Duns Scotus. Ordinatio. Liber I, Distinctio 3 [Ord.], ed. C. Balic et al. (Opera omnia vol. 3) (Vatican: Commissio Scotistica, 1954). John Duns Scotus. Quaestiones super libros Metaphysicorum Aristotelis. Libri I–V [Q. Met.], ed. R. Andrews et al. (Opera Philosophica vol. 3) (St. Bonaventure, NY: Franciscan Institute, 1997). John Duns Scotus. Ordinatio. Liber quartus. Distinctiones 43–49 [Ord.], ed. B. Hechich et al. (Opera omnia vol. 14) (Vatican: Commissio Scotistica, 2013). Keele, Rondo. “Formal Ontology in the Fourteenth Century: The Chatton Principle and Ockham’s Razor” (Doctoral Dissertation, Indiana University, 2002). Keele, Rondo. “The Early Reception of Peter Auriol at Oxford. Part I: From Ockham to the Black Death,” Recherches de théologie et philosophie médiévales 82 (2005), 301–61. Keele, Rondo. “Can God Make a Picasso? William of Ockham and Walter Chatton on Divine Power and Real Relations,” Journal of the History of Philosophy 45 (2007), 395–411.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

 ’    

241

Keele, Rondo. “Walter Chatton,” in E. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Spring 2014 Edition). . Löwe, Can L. “Peter Auriol on the Metaphysics of Efficient Causation,” Vivarium 55 (2017), 239–72. Maier, Anneliese. “Das Problem der Finalkausalität um 1320,” in Anneliese Maier, Metaphysische Hintergründe der spätscholastischen Naturphilosophie (Rome: Edizioni di Storia e Letteratura, 1955), 273–99. Majcherek, Kamil. “O dwóch średniowiecznych koncepcjach celowości natury: Duns Szkot i Chatton,” Roczniki Filozoficzne 65(1) (2017), 43–62. Maurer, Armand. “Ockham’s Razor and Chatton’s Anti-Razor,” in Armand Maurer, Being and Knowing: Studies in Thomas Aquinas and Later Medieval Philosophers (Toronto: Pontifical Institute of Mediaeval Studies, 1990), 431–46. Pasnau, Robert. “Intentionality and Final Causes,” in D. Perler (ed.), Ancient and Medieval Theories of Intentionality (Leiden: Brill, 2001), 301–23. Peter Auriol. Scriptum super primum Sententiarum, d. II–VIII [Scriptum], ed. E. M. Buytaert (St. Bonaventure, NY: Franciscan Institute, 1956). Putallaz, François-Xavier. “Efficience et finalité dans le Traité du premier pincipe de Jean Duns Scot,” Revue de théologie et de philosophie Lausanne 116 (1984), 131–46. Robert, Aurélien. “L’explication causale selon Guillaume d’Ockham,” Quaestio 2 (2002), 241–65. Tachau, Katherine H. Vision and Certitude in the Age of Ockham: Optics, Epistemology, and the Foundations of Semantics, 1250–1345 (Leiden: Brill, 1988). Thomas Aquinas. Commentaria in octo libros Physicorum Aristotelis [In Phys.], vol. 2 of Opera omnia issu Leonis XIII (Rome, 1884). Thomas Aquinas. In duodecim libros Metaphysicorum Aristotelis expositio [In Met.], ed. M. R. Cathala and S. M. Spiazzi (Taurini-Rome: Marietti, 1971). Thomas Aquinas. De principiis naturae ad fratrem Sylvestrum [De princ. nat.], ed. H.-F. Dondaine, vol. 43 of Opera omnia issu Leonis XIII (Rome: Editori di San Tomasso, 1976). Thomas Aquinas. Expositio libri Posteriorum [In An. Post.], 2nd edition, ed. R.-A. Gauthier, vol. 1*/2 of Opera omnia issu Leonis XIII (Rome-Paris: Commissio Leonina-Vrin, 1989).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

242   Trifogli, Cecilia. “Thomas Wylton on Final Causality,” in A. Fidora et al. (eds.), Erfahrung und Beweis: Die Wissenschaften Von Der Natur Im 13. und 14. Jahrhundert (Berlin: Akademie Verlag, 2007), 249–64. Wallace, William A. Causality and Scientific Explanation, vol. 1: Medieval and Early Classical Science (Ann Arbor, MI: University of Michigan Press, 1972). Walter Chatton. Reportatio et lectura super Sententias: Prologus [Sent. Prol.], ed. J. C. Wey (Toronto: Pontifical Institute of Mediaeval Studies, 1989). Walter Chatton. Reportatio super Sententias: super Librum II [Rep.], ed. J. C. Wey and G. J. Etzkorn (Toronto: Pontifical Institute of Mediaeval Studies, 2002). Walter Chatton. Lectura super Sententias: Liber I, Distinctiones 3–7 [Lectura], ed. G. J. Etzkorn and J. C. Wey (Toronto: Pontifical Institute of Mediaeval Studies, 2008). Wey, Joseph C. “Introduction,” in Walter Chatton, Reportatio et lectura super Sententias: Prologus [Sent. Prol.], ed. J. C. Wey (Toronto: Pontifical Institute of Mediaeval Studies, 1989), 1–8. William of Ockham. Opera theologica et philosophica [OTh and OPh], ed. G. Gal et al., 17 vols. (St. Bonaventure, NY: Franciscan Institute, 1967–88).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

Aquinas on the Sources of Wrongdoing Themes from McCluskey Thomas Williams

Over the last twenty years or so, philosophers have increasingly recognized the fruitfulness of approaching both the history of philosophy and contested issues in contemporary philosophy by way of an investigation of evil. To take just two examples, Susan Neiman’s muchdiscussed Evil in Modern Thought: An Alternative History of Philosophy takes evil as the organizing principle for her historiography of modern philosophy,¹ and Claudia Card explores the notion in a contemporary context in The Atrocity Paradigm: A Theory of Evil.² The approach has borne fruit in medieval philosophy as well: a new six-volume history of evil includes The History of Evil in the Medieval Age: 450–1450 CE,³ edited by Andrew Pinsent, and Bonnie Kent has written on “Evil in Later Medieval Philosophy.”⁴ Colleen McCluskey’s Thomas Aquinas on Moral Wrongdoing⁵ is a worthy and impressive entry in this literature. McCluskey examines the metaphysical foundations of Aquinas’s ethics, his account of evil as a privation, the powers of the soul from which wrongdoing originates, and evil habits, or vices. I shall focus on what I take to be McCluskey’s most significant contribution, the three chapters devoted to a careful exploration of the three sources of moral wrongdoing—defects in the

¹ ² ⁴ ⁵

Princeton, NJ: Princeton University Press, 2002. Oxford: Oxford University Press, 2002. ³ New York: Routledge, 2018. Journal of the History of Philosophy 45 (2007), 177–205. Cambridge: Cambridge University Press, 2017.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

244   intellect (covered in Chapter 3), defects in the sensory appetite (Chapter 4), and defects in the will (Chapter 5)—and treat the other chapters somewhat briefly. In Chapter 1, “Metaphysical Foundations,” McCluskey gives an overview of Aquinas’s account of human nature and his theory of human action. She discusses the powers of the soul, including the sensory appetite and its passions, the intellect, and the will. Crucially, she devotes considerable attention to the ways in which the passions can affect the intellect’s judgment and, thereby, the will. For (controversially, though quite correctly in my view) McCluskey insists that “Aquinas argues for a priority of intellect over will in the process of action” (29); on her interpretation, the will has no independent power to shape human action. This interpretation, as McCluskey notes, “raises worries about necessitation in the will” (29). But the will would be necessitated only if the intellect were necessitated, and McCluskey argues that for “much of practical reasoning” (31) there is no necessitation in the intellect on Aquinas’s view. I shall have more to say about this issue when I discuss McCluskey’s account in Chapter 3 of wrongdoing that originates in the intellect. In Chapter 2, McCluskey explores Aquinas’s account of the ontological status of evil as a privation, “a lack of the goodness that should otherwise be there” (41). At the end of the chapter she acknowledges that “Aquinas’s discussion of wrongdoing does not rest on the privation account in any substantive way” (72), but it is nonetheless illuminating to consider the metaphysics of goodness in Aquinas as a background for his account of moral goodness in particular, and the privation theory is a corollary of that metaphysics of goodness. Everything is good to the extent that it actualizes the potentialities that mark it out as a thing of a particular kind. Nothing can exist at all without actualizing its kinddefining potentialities to some degree, so everything is good to some extent. Evil has no positive reality; it is a deficiency in the actualization of a thing’s potentialities. The human potentialities for moral action, practical intellect and will, are aimed at human flourishing, and so human actions are good to the extent that they lead toward human flourishing, bad to the extent that they lead away from flourishing. As McCluskey puts it, a morally wrong action “lacks the fullness of being that the action ought to have (ST I-II.18.1). The privation is located (so to speak)

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

245

ultimately in the action’s departure from what reason (functioning correctly) dictates is constitutive of or conducive to flourishing” (65). In a certain sense there is something with positive ontological status that goes into making a bad action bad: “in moral matters something is called ‘bad’ in a positive sense insofar as the will’s act is denominated bad from the object willed, but that bad object can only be willed under the aspect of good” (De malo 1.1 ad 12).⁶ So the fact that there is some positive good, or apparent good, willed by the wrongdoer does not mean that wrongdoing itself is a positive reality; the wrongness of a wrong act is the privation of the order of reason in an act. Critics of the privation account who “are looking for a positive account of evil qua evil” (72) will accordingly not find what they want in Aquinas, though other criticisms of the privation theory can be met. McCluskey notes, however, that the privation theory by itself is not enough to “preserve the goodness of God and ensure that God is not responsible for the existence of evil” (49). On McCluskey’s account, these theological motivations for the privation theory are not paramount for Aquinas anyway; the privation theory emerges out of his Aristotelian metaphysics of goodness as sketched in the preceding paragraph (45). As I have said, the centerpiece of Thomas Aquinas on Wrongdoing is a careful exploration of Aquinas’s understanding of the three sources of moral wrongdoing: defects in the intellect, defects in the sensory appetite, and defects in the will. Here the exposition is particularly acute, textually sensitive, and philosophically adept. It is not easy to make everything Aquinas says about these matters cohere. Indeed, on the face of it, the division looks hopeless. For one thing, every instance of moral wrongdoing has to involve some defect in the will: there is no moral action without willing, and if one’s will is not deficient, the willing is not an instance of wrongdoing. For another, every instance of moral wrongdoing likewise has to involve some defect in the intellect, since (again) the will has no independent power to shape human action. What, then, are we to make of this division? McCluskey gives a helpful overview at the outset, in Chapter 3, “Aquinas’s Account of Moral Wrongdoing: General Points and Defects in the Intellect.” Although “Aquinas is willing to grant that ignorance plays a role in all forms of ⁶ All translations are mine.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

246   wrongdoing” (84), “the role that knowledge plays in each kind of wrongdoing is importantly different” (85). In sins of ignorance, which derive from a defect in the intellect, the act depends on a simple lack of knowledge. In sins of passion, which derive from a defect in the sensory appetite, “the passions affect one’s access to the knowledge that one possesses” (85). In sins from deliberate wrongdoing (ex certa malitia), which derive from a defect in the will, the ignorance is more complex still: “the agent fails to understand that one ought not acquire this particular good because of the evil that necessarily accompanies it” (85). Because ignorance plays a role in all forms of wrongdoing, McCluskey turns to Aquinas’s account of voluntariness, according to which some, but not all, forms of ignorance are themselves voluntary, so that actions deriving from such ignorance are likewise voluntary and culpable. The discussion is brief, and properly so, since this material is well known (though Aquinas’s puzzling account of concomitant ignorance receives its due). McCluskey then turns to “wrong actions that originate in the intellect,” which Aquinas classifies as “sins of ignorance” (91): Three conditions must be fulfilled for wrongdoing of this type (ST I-II.76.2): the ignorance involved must be of knowledge that one is obligated to obtain (the obligatoriness condition); (2) it must be true of the agent that it is possible for him to acquire the requisite knowledge (the voluntariness condition); and (3) the missing knowledge would have affected what the agent chose to do (the causal condition). (91–2)

The voluntariness condition is the most interesting for McCluskey’s purposes, because it seems to locate the origin of the resulting wrongdoing in the will rather than in the intellect, contrary to Aquinas’s threefold classification. McCluskey asks, “what about the agent who misses a meeting because she is engrossed in her work and fails to notice the entry on her calendar or the time on her clock?” (93) She could have remembered the meeting, and it was part of her job that she do so. Her ignorance stems from a negligent failure to consider what she could and should have considered. In De malo 1.3, Aquinas attributes such negligence to the will: Pleasure and everything else in human affairs should be measured and governed in accordance with the rule of reason and of the divine law; accordingly, a disordered choice implies a prior failure on the part of

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

247

the will to make use of the rule of reason and of the divine law. Now there is no need to look for some other cause of this failure to make use of the aforementioned rule: the will’s very freedom to act or not act is sufficient for it.

McCluskey argues, however, that “the will’s activity of use presupposes an activity in the intellect” (94), so “the negligence involved in this case can be traced through the will to a defect in the intellect’s judgment, that is, a failure to judge that taking these steps to avoid missing meetings is a good thing to do” (95). Thus wrongdoing attributable to negligent nonconsideration does after all stem ultimately from a defect in the intellect. But is Aquinas right to assimilate all wrongdoing that originates in the intellect to sins of ignorance? What about mistakes of reasoning, McCluskey asks? Here the discussion could have been greatly enriched by considering Aquinas’s discussion of the three acts of prudence (ST II-II.47.8, with related virtues at II-II.51 and vices at II-II.53–5), which are deliberation (consilium: I-II.14), judgment, and command (I-II.17).⁷ Noting that an agent might fail to make dispositional knowledge occurrent (ST I-II.74.5, De malo 3.9) does not do justice to the intricacies of practical reasoning, the many things that an agent has to get right, and consequently the many ways in which practical reasoning can go awry. It is not clear to me that such an investigation would have changed McCluskey’s bottom-line conclusion—that Aquinas does understand mistakes in reasoning as instances (or perhaps causes) of ignorance—but it would certainly have made her account much more thorough and nuanced. This observation prompts me to make a more general point. Aquinas on Moral Wrongdoing is an exceptionally disciplined book. McCluskey keeps tightly to her subject, avoids repeating accounts that she has worked out in detail elsewhere, and declines to get into the weeds of scholarly controversies that don’t matter for her project (while giving thorough and spirited attention to those that do). This is all to the good: a lean book is admirable, and far better than a bloated one. But there are certainly places where one might want more. For me, McCluskey’s account of errors of practical reasoning is one such place. Her chapter on ⁷ There is a brief discussion of the three acts of prudence in Chapter 6, “The Vices in Aquinas’s Moral Psychology.”

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

248   the metaphysical foundations of Aquinas’s ethics is another: McCluskey herself says there that her “discussion . . . will no doubt feel rather thin” (13). Still, these are matters of judgment, and perhaps McCluskey has come at least very close to the mean between excess and deficiency. In Chapter 4, McCluskey turns from sins of ignorance to sins of passion—acts of wrongdoing that have their origin in defects of the sensory appetite. One topic that belongs here is incontinence, particularly of the kind that Aquinas calls weakness (debilitas), in which “the agent gives in to passion while recognizing the cogency of the intellect’s judgment” (108) that the passion should be resisted. She quotes from De malo 3.9 ad 7: “Both the continent and the incontinent are moved in two ways: by reason [they are moved] toward avoiding sin, but by desire [they are moved] toward committing it.”⁸ According to Aquinas: both make use of a syllogism with four propositions but arrive at opposite conclusions. The continent person syllogizes as follows: “No sin is to be done”—and he puts forward this premise in accordance with the judgment of reason, but in accordance with the movement of desire his heart is occupied with the thought, “Everything pleasurable is to be pursued.” But because the judgment of reason prevails in him, he accepts the first [as his major premise] and derives his conclusion from it: “This is sin; therefore, this is not to be done.” In the incontinent person, by contrast, the movement of desire prevails. He accepts the second [as his major premise] and derives his conclusion from it: “This is pleasurable; therefore, this is to be pursued.” (De malo 3.9 ad 7)

McCluskey rightly comments, “There is, of course, a great puzzle here” (109)—indeed, more than one puzzle, I think. Her puzzle is this: “Why does the judgment of reason prevail in one agent and not in the other?” Her answer is that “Aquinas seems unconcerned about the answer to this question, and perhaps it cannot be answered . . . . Ultimately, Aquinas might have to concede that there is no explanation for the difference in ⁸ McCluskey’s translation here is mistaken: she has the continent agent being moved in accordance with reason and the incontinent under the influence of concupiscence (109). But Aquinas’s point is that both the incontinent and the continent have conflicting motivations; both, as Aquinas goes on to say, possess (in some sense) the proposition “no sin is to be done” as well as the proposition “everything pleasurable is to be pursued.” The difference lies in which of these propositions each selects for his practical syllogism. As far as I can see, however, this error does not vitiate her exposition.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

249

choices” (109, 114) between the continent and the incontinent. She quotes Bonnie Kent as saying that agents “have reasons for doing what they did, but not for failing to do what they should have done.”⁹ If I lash out at a colleague because I’m feeling angry, even though I know I shouldn’t behave in such a way, I have a reason for lashing out—in my anger, it seemed good to me to act that way—but no reason for not controlling myself. I could have controlled myself, for I retained the use of my will and intellect; I was not literally out of my mind with anger. I knew, at least dispositionally, that I shouldn’t act that way, but I lashed out nonetheless. Leaving aside for the moment the issue of dispositional versus occurrent knowledge raised by the previous sentence, I see nothing in this account that is incompatible with Aquinas. That is, I don’t see why Aquinas shouldn’t grant that there is no explanation for the incontinent person’s acting incontinently (and surely, by parity of reasoning, that there is likewise no explanation for the continent person’s acting continently). But it is the sort of conclusion that understandably drives some people to voluntarism. Incontinent actions are blameworthy, and blameworthiness presupposes voluntariness. But (one might argue) if my outburst is truly voluntary, then I did it: it was a human action, a product of my intellect and will. To say that I would have acted differently if my reasoning had gone differently suggests, though, that my outburst was something that happened to me (or in me), rather than something I did. Granted, reasoning is something I do just to the extent that I will to think about my alternatives in a particular way; but on Aquinas’s view I will to think about my alternatives in a particular way only in accordance with the intellect’s judgment that I ought to think about them in that way, and it is precisely that kind of judgment for which we are ruling out an explanation. So one might understandably come to think that the only way to secure the conclusion that I am responsible for my outburst—that it is something I do—is to ascribe to the will an independent power to shape human action. I willed my outburst, or at least I willed not to entertain the considerations that would have kept me from it.

⁹ Bonnie Kent, “Aquinas and Weakness of Will,” Philosophy and Phenomenological Research 75 (2007), 70–91, at 89.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

250   Lashing out at my colleague may not be the right sort of example, however, because in a typical case I wouldn’t have occurrent knowledge of the right sort of proposition, such as “Anger should be mitigated” or “Colleagues should always be treated with respect.” I would act before I had time even to entertain such a thought. Such a case looks more like what Aquinas calls unrestrained or unbridled incontinence (irrefrenata incontinentia) (ST II-II.156.1), in which “passion . . . takes the intellect by surprise” (102), rather than delibitas, the kind of incontinence that corresponds more closely to weakness of will as we usually talk about it nowadays, and which Aquinas was describing in the passages from De malo cited above.¹⁰ Debilitas gives rise to the voluntarist worry in just the same way as unrestrained incontinence, and it adds a further puzzle. The “four propositions” account of the practical syllogisms of the continent and the incontinent is very odd. In the standard case of incontinence, the incontinent person “makes use of a syllogism” with four propositions, Aquinas says: (1) Sin is not to be done. (2) Whatever is pleasurable is to be pursued. (3) This is pleasurable. (4) This is to be pursued. Saying that the incontinent person makes use of such a syllogism (or, as at ST I-II.77.2 ad 4, that “the incontinent person’s syllogism has four propositions”) suggests that the incontinent person has occurrent knowledge of all four, at least at some point in his deliberation. But that can’t be right. For unless the incontinent person entertains (10 )

This is sin

there would be no reason to entertain (1), any more than some other general proposition unrelated to the situation at hand. (Except for purposes of illustration when doing ethics, I don’t entertain “Angry ¹⁰ Crucially, however, Aquinas does not speak of it as weakness of will. Will for Aquinas is basically either on or off: sane people have the use of will and reason, insane people do not. Despite his use of the term debilitas, Aquinas does not envision that incontinent people suffer from a lack of will-power (even on a particular occasion, let alone chronically) that continent people enjoy. For him, sinning from debilitas is the same thing as sinning from passion. See Kent, “Aquinas and Weakness of Will,” 78–81.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

251

people should be mollified with kind words” when there’s no angry person about.) And Aquinas expressly denies that the incontinent person has occurrent knowledge of (10 ): “although he does have knowledge regarding the universal, he does not have knowledge regarding the particular, because he takes (assumit) his major premise not in accordance with reason but in accordance with desire” (De malo 3.9 ad 7). In fact, “passion hinders reason” from entertaining (1) and drawing the appropriate conclusion that the act is not to be done (ST I-II.77.2 ad 4). Hinders, not prevents: it bears repeating that passion prevents the exercise of reason only in those who are insane, and their acts are not voluntary and accordingly not blameworthy. The incontinent has dispositional knowledge of (1), but under the influence of the passion he fails to make that knowledge occurrent, though he retains the power to do so. These formulations with the Latin gerundive, translated as “to be done” or “to be pursued” and so forth, also involve important ambiguities. “Whatever is pleasurable is to be pursued” (omne delectabile est prosequendum) might appear to mean that one ought always to pursue pleasure, but someone entertaining that strong claim would be selfindulgent or intemperate, not incontinent (let alone continent). So it must mean something like “Pleasure is worth pursuing” or “That an action is pleasant counts in favor of performing it.” And the conclusion “This is to be done” need not mean “This is what I ought to do,” or even “This is, all things considered, the best thing for me to do,” but simply “This has some attractive (good) feature that makes it worth doing.” It is often said—I have said it myself—that for Aquinas the will chooses what the intellect judges is best; but as McCluskey notes, Aquinas never says this: “agents need not choose what they regard as best, they can choose a lesser good as long as there is something about that alternative that they can describe as good” (111–12). The most complex sort of wrongdoing is precisely that in which an agent chooses a lesser good, not in ignorance or under the influence of a passion, but knowingly. Such sins have their origin in a defect of the will. Aquinas calls them sins ex certa malitia. When she turns to such sins in Chapter 5, McCluskey immediately takes the essential and right step of not translating malitia as “malice.” This may sound like a small thing, but a lazy preference for cognates is a bane of scholarship in medieval philosophy, and reading malitia as “malice” results in all sorts of

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

252   misunderstandings, as indeed McCluskey shows. Malice is “the intention or desire to do evil or cause injury to another person; active ill will or hatred” (OED). Though malitia can have that meaning—Cicero defines it in one place as “a cunning and deceitful intention to do harm” (De natura deorum 3.75)—in the phrase ex certa malitia it is nothing more than the abstract noun from malus, “bad,” and thus means “badness.” McCluskey does well, then, to translate certa malitia as “deliberate wrongdoing.” “Deliberate wrongdoing,” she explains, “arises when the agent chooses a lesser good over a greater good and would rather be deprived of the greater good than lose the lesser good” (117). In a sin of passion, the agent chooses something that looks good to him only because the passion colors his judgment; in deliberate wrongdoing, the agent knows he is choosing what is bad in itself, but goes ahead and chooses it anyway. He does not choose it because it is bad—Aquinas is quite consistent that we can choose only what appears good to us in some way—but rather for the sake of some other good that he loves more than he loves “the order of reason or of divine law, or God’s charity, or something of that sort” (ST I-II.78.1 in corp.). Such choices can spring from a vicious habit, which makes it second nature for someone to “regard what is objectively bad as if it were good and suitable for choice” (123). Aquinas says that all wrongdoing that springs from a vicious habit is deliberate wrongdoing, though there can be deliberate wrongdoing that does not spring from habit.¹¹ Aquinas at least implies (at ST I-II.73.6 and 78.3) that deliberate wrongdoing is more serious than sins from ignorance and sins from passion, because “insofar as the agent has become the kind of person who sees what is bad as something desirable, her moral orientation is itself defective” (131): Someone whose will is inclined to sinning by intending a worse end sins more seriously . . . . Causes that diminish the judgment of reason, such as ignorance, or that diminish the free movement of the will, such as weakness, violence, fear, or something of that sort, diminish sin. (ST I-II.73.6 in corp.) ¹¹ McCluskey wrestles astutely with the difficulties posed by Aquinas’s scattered and not altogether consistent remarks about non-habitual deliberate wrongdoing; for the sake of space I will pass by that discussion.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

253

But has Aquinas drawn the category of deliberate wrongdoing too broadly for it to be plausible that sins from deliberate wrongdoing are the most serious? McCluskey offers the example of deciding to eat toaster pastries for breakfast rather than something healthier (135). If I know perfectly well that this is an unhealthy choice but go ahead and make it anyway, I have committed an act of deliberate wrongdoing. But is it really all that serious? Here I think McCluskey is too reluctant to give an unequivocal yes on Aquinas’s behalf. For there is something wrong with my will in such a case. I’m not ignorant and I’m not under the sway of passion; my will is simply bent on unhealthy pleasure and does not value health and well-being in the way it should. That, surely, is a very serious matter indeed. We can pity the passionate and instruct the ignorant, but what are we to do with those whose wills are bent on what is, in fact if not in appearance, self-destruction? Note, however, that the relative seriousness of sins as Aquinas discusses them in ST I-II.73.6 is a matter of the degree of moral blame we should attach to them.¹² That is, agents are more blameworthy when they sin ex certa malitia than when they sin from ignorance or from passion. Moreover, this is a ranking of the blameworthiness of particular sins, not of the gravity of types of sins. (So one need not say that the deliberate choice of unhealthy breakfast foods is a particularly grave type of sin, though by Aquinas’s lights acting out of contempt for bodily health isn’t exactly trivial either.) For these reasons the category of deliberate wrongdoing is clearly ill-suited to serve as an account of evil in the sense in which some recent writers distinguish evil from “ordinary garden-variety wrongdoing” (128), and McCluskey argues compellingly against John Langan’s reading of Aquinas as attempting (unsuccessfully, in the end) to use the category of sins ex certa malitia to explain the “dark and troubling range of moral phenomena . . . among which are malice, hatred of God, envy of others, and attraction to evil itself ”¹³ (130). But she concedes too much, I think, when she says that if Aquinas’s moral psychology “cannot distinguish between evil and wrongdoing, that makes his account less plausible” (130). Why would one expect a moral ¹² Bonnie Kent, “Aquinas and Weakness of Will,” 74–5, is very good on this point. ¹³ John Langan, “Sins of Malice in the Moral Psychology of Thomas Aquinas,” in D. M. Yeager (ed.), The Annual of the Society of Christian Ethics (Washington, DC: Georgetown University Press, 1987), 179–98, at 180.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

254   psychology—an account of the psychological faculties, mechanisms, and processes that produce voluntary action—to provide a distinction between ordinary wrongdoing and particularly heinous wrongdoing? That seems like a task for a theory of value, perhaps an account of the human good. Moreover, the contrast between evil and wrongdoing in the contemporary discussion seems to be a distinction in search of a rationale: Is the distinction quantitative or qualitative? Are evils particular kinds of harms, or harms perpetrated from particular kinds of motives? The distinction is too nebulous for me to worry much about whether Aquinas has the resources to make it. But McCluskey is not as cavalier as I am about entering the contemporary discussion on Aquinas’s behalf, and she notes various ways in which Aquinas can “distinguish between varying degrees of severity among immoral actions,” though in the end she concedes that she has not provided a way “to distinguish evil from ordinary wrongdoing and to determine what would make an action an instance of evil as opposed to something less grave” (134). In her final chapter McCluskey examines vices, which are habits— more or less firmly rooted dispositions acquired by repeated action—that prompt (but do not necessitate) particular acts of wrongdoing. She provides a helpful overview of Aquinas’s treatment of the vices both in ST II-II, where the vices are discussed in relation to the virtues to which they are opposed, and in the contemporaneous De malo, questions 9–15, in which Gregory the Great’s list of seven capital vices provides the structuring principle. What Aquinas says about the vices seems to stand in tension with his claim that all wrongdoing stemming from habit is deliberate wrongdoing, which derives from defects in the will. The problem is easy enough to see: there are vices in the intellect (such as undue haste in deliberation: praecipitatio or temeritas) and vices in the sensory appetite (such as gluttony in the concupiscible part and excessive daring in the irascible part). So doesn’t wrongdoing that stems from a vice in the intellect or the sensory appetite derive from a defect in those powers rather than from a defect in the will? In the case of intellectual vices, there is an easy solution available. One cannot possess the intellectual virtue of prudence without a correct orientation of the will, and so “a defect in the will (a disordered orientation to the good) gives rise to a vicious habit in the intellect” (173). In the case of vices located in the sensory appetite, however, the story is

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

     

255

more complicated. McCluskey appeals to the fact that although passions are located in the sensory appetite, there are analogues to passions in the intellectual appetite or will. And “whatever stimulates delight in the sensory appetite can also stimulate a concurrent state in the will, a state Aquinas identifies as joy (gaudium) rather than delight (delectatio)” (172).¹⁴ So over time a vice in the sensory appetite can lead to an analogous vice in the will: “The agent’s character is now deformed with respect to this particular passion . . . . The vice might originate in the sensory appetite, but once it involves concomitant habituation in the will, the resulting wrongdoing is no longer a sin of the passions but rather a case of deliberate wrongdoing” (173). But Aquinas doesn’t say that delight in the sensory appetite must produce joy in the will, only that it can; and he certainly doesn’t say that vices in the sensory appetite must be accompanied by (or even eventually lead to) corresponding vices in the will. So this way of saving the claim that all wrongdoing stemming from habit counts as deliberate wrongdoing, and therefore as wrongdoing that derives from a defect in the will, seems to fail. I am inclined to think that nothing much hinges on whether Aquinas can make good on the claim that all wrongdoing stemming from habit is sin ex certa malitia and therefore derives from a defect in the will. The threefold division of wrongdoing into sins of passion, sins of ignorance, and sins ex certa malitia obviously serves a useful purpose, as McCluskey’s work amply shows. But in the end Aquinas’s moral psychology is too complex to permit the distinction to be entirely clean and without ragged edges. All sins will involve ignorance in some way; all sins will involve some disorder, whether transitory or more firmly rooted, in the will. Many sins (not all, but many) will involve passions that affect either our judgments or our orientation to the good or both. The threefold distinction has provided a framework in which McCluskey astutely examines and defends Aquinas’s complex account of wrongdoing; and that is all that we, or Aquinas, could reasonably expect of it. University of South Florida

¹⁴ This is not quite right, since in the very same sentence Aquinas refers to this analogue to passion as “rational delight” (delectatio rationis), but the regimentation of Aquinas’s vocabulary is welcome.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

256  

Bibliography Card, Claudia. The Atrocity Paradigm: A Theory of Evil (Oxford: Oxford University Press, 2002). Kent, Bonnie. “Aquinas and Weakness of Will,” Philosophy and Phenomenological Research 75 (2007), 70–91. Kent, Bonnie. “Evil in Later Medieval Philosophy,” Journal of the History of Philosophy 45 (2007), 177–205. Langan, John. “Sins of Malice in the Moral Psychology of Thomas Aquinas,” in D. M. Yeager, (ed.), The Annual of the Society of Christian Ethics (Washington, DC: Georgetown University Press, 1987), 179–98. McCluskey, Colleen. Thomas Aquinas on Moral Wrongdoing (Cambridge: Cambridge University Press, 2017). Neiman, Susan. Evil in Modern Thought: An Alternative History of Philosophy (Princeton, NJ: Princeton University Press, 2002). Pinsent, Andrew (ed.). The History of Evil in the Medieval Age: 450–1450 CE (New York: Routledge, 2018).

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

Briefly Noted Han Thomas Adriaenssen, Representation and Scepticism from Aquinas to Descartes (Cambridge University Press, 2017), 279 pp. For a generation now, scholars of medieval thought have been working intensively on later medieval cognitive theory. At the heart of that work lies the theory of species, as endorsed by Thomas Aquinas and subsequently criticized, in various ways, by Henry of Ghent, Peter of John Olivi, Peter Auriol, and William of Ockham. Adriaenssen’s book builds on this past research to provide a state-of-the-art synthesis of these discussions, looking closely at the relationship between the species and the cognitive act, on one hand, and the species and the object of cognition, on the other. What is particularly valuable about this study, however, is that Adriaenssen turns in the second part to show, in detail, how these medieval discussions reappear in the seventeenth century. That discussion begins with Descartes’s deployment of ideas to play something like the role that species had played in scholastic thought, and then works through the development of that theory among later Cartesians, and its criticisms at the hands of Arnauld, Desgabets, and John Sergeant. Although Adriaenssen does not even pretend to cope with the two-century gap that lies between these two periods, his narrative makes such a persuasive case for continuity that one might never know he is crossing back and forth between two ostensibly different arears of academic specialization. John Duns Scotus, Notabilia super Metaphysicam, edited by Giorgio Pini (Brepols, 2017), lxxii + 256 pp. Scotus has long been known to refer, in various places, to an expositio of the Metaphysics, but it was not until 1996 that Giorgio Pini announced the news that this work is extant in two manuscripts. Now Pini has published an edition. Given the large number of cross-references back and forth between this work and Scotus’s known oeuvre, Pini’s introduction is able to show in detail that there can be no doubts about its authenticity. Readers may feel some disappointment, however, to learn that the work is not a comprehensive expositio in the manner of, say, Aquinas’s commentaries (although Scotus does make 69 explicit references to Aquinas’s commentary, referring to him as the expositor). Instead, the Notabilia is a series of brief and disconnected notes on specific passages. Pini takes his chosen title from the heading of one of the manuscripts, and it indeed conveys a better sense of the content than would the label expositio. The notes are, moreover, often arcane, even by Scotus’s standards, and there is no easy way to draw on this work for quick takes on one or another lofty metaphysical issue. But in the hands of a patient reader the work is bound to be illuminating, because it contains such a rich collection of comments on so many different sorts of topics. Indeed, although Pini doubts whether the work was ever intended to circulate outside the classroom, he thinks these notes must have been written as part of the same lectures that produced his Quaestiones super Metaphysicam (xlvi–xlvii), that the text of the Notabilia was likely to have been physically attached to Scotus’s personal copy of Aristotle’s Metaphysics, and that

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

258   Scotus is very likely to have expanded these notes over the years whenever he returned to Aristotle’s text (xliii). So, although the Notabilia can hardly be regarded as a lost masterpiece, it provides a unique insight into Scotus’s personal engagement with one of the most foundational texts for his thought. Michael Gorman, Aquinas on the Metaphysics of the Hypostatic Union (Cambridge University Press, 2017), 177 pp. Christ is God and so has a divine nature. Christ is man and so has a human nature. And yet Christ is one person. The question of Gorman’s book is how one person or hypostasis can have two utterly disparate natures united within it. It is hard to imagine anyone writing a more beautifully clear and insightful work on such a notoriously vexatious topic. Although Gorman’s interests are unabashedly theological, he approaches the issues in the most clear-headedly philosophical of ways—just as does Aquinas himself, one might say. The book begins with a lovely précis of the theological issue and its historical context, and then offers a summary of Aquinas’s background metaphysics that is a marvel of compression and precision. From here the book cleverly structures itself around pursuing first what is most clear about Aquinas’s position, and then progressively working in subsequent chapters toward what is most dark and doubtful. Throughout, Gorman is clear-eyed enough to recognize the difficulties that arise. If, for instance, Christ has a divine nature, then Christ should be perfectly simple, but then how can Christ have two natures? And, if Christ has a human nature, then does this not entail that Christ is a human person, and so a union of two distinct persons, contrary to orthodoxy? Although Gorman is resourceful in responding to these and other worries, he is also frank about admitting when Aquinas’s account seems not to provide clear answers. In such places, however, Gorman suggests how Aquinas’s implicit suggestions might be developed more explicitly in a way that is more adequate to handling various puzzles that arise. John Wyclif, De scientia Dei, edited by Luigi Campi (Oxford University Press, for the British Academy, 2017), cxlix + 168 pp. The appearance of this volume brings us close to the finish line of the long quest to edit Wyclif ’s opus magnum, his so-called Summa de ente. Of these thirteen extant treatises, organized into two books, all but a handful have now been published, and those that remain are in progress. Here we have Book II tract 2. Topics include the status of knowledge as a divine attribute, its relationship to creatures, intelligible and actual, and familiar puzzles regarding determinism, future contingents, and divine grace. Wyclif ’s sympathies are with Aquinas and Bradwardine, rather than the voluntarists, and his discussion comes right up to the brink of determinism, if not crossing that brink, as later generations would charge. Among the many innovative features of Wyclif ’s discussion are his pan-propositionalism, his conception of intelligible being, and his peculiar doctrine that all finite beings are essentially identical to God. Campi’s edition contains a very extensive introduction in English, providing a philosophically sophisticated guide to this tract’s contents, a detailed report on the manuscript tradition, and a state-of-the-art account of how this treatise fits within the Summa. The edition includes a full textual apparatus and detailed notes on sources but, unlike many recent British Academy volumes, offers no translation.

OUP CORRECTED PROOF – FINAL, 1/8/2019, 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]. Please do not send queries directly to the editor, since this compromises the process of blind review.

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

OUP CORRECTED PROOF – FINAL, 1/8/2019, SPi

Index of Names Alberic of Paris 45, 62 Abū’l-Hudhayl 11n Adam Wodeham 188, 190, 191, 199, 201–2, 205 Albert the Great 121 Arlig, Andrew 44, 59, 63 Aristotle 4–9, 39, 53, 67–9, 74–8, 96, 98, 119–21, 126, 135–6, 177, 192, 194, 205, 214–15, 217, 219–20, 230–1, 236–7 al-Ashʿarī, Abū’l-Ḥasan 11 Augustine 192, 198 Avicenna 1–5, 7–8, 10, 14–19, 23–5, 34, 36–9, 216 al-Bāqillānī, Abū Bakr 11n Boethius 46, 53, 125–6, 134, 136 Brower, Jeffrey 71, 77–8 Chrysippus 58, 60n Cicero 252 Cross, Richard 125n, 167, 177, 204 Durand of St.-Pourcain 186, 188–200, 202–5 al-Ghazālī, Abū Ḥāmid 12, 32n Godfrey of Fontaines 186, 199–200 Gregory the Great 254 Gregory of Rimini 189–90 Henry of Ghent 213, 216 Hervaeus Natalis 189, 195–7

John of Salisbury 45 al-Juwaynī 28 Kent, Bonnie 249 Koch, Josef 200 McCluskey, Colleen 243–55 Peter Abaelard 44–64 Peter Auriol 189, 213, 216–18, 232n Peter John Olivi 187 Peter Lombard 150, 188 Peter of Palude 189, 195 Philoponus 5n, 25 Plotinus 136 Prosper de Reggio Emilia 189 Pseudo-Dionysius 136 al-Rāzī, Abū Bakr 7n Richard Drayton 188n, 189–91, 195, 201 Rudolph, Ulrich 11n Sabra, A. I. 10 Simplicius 5n, 124n Stump, Eleonore 73 Thaddeus of Parma 189 Thomas Aquinas 67–111, 116–43, 147–67, 171–5, 179–83, 187, 191, 243–55 Thomas Wylton 225n Van Inwagen, Peter 60n

Ibn Fūrak 11 John Duns Scotus 138, 147–9, 152–5, 164–84, 188, 194–6, 200, 205, 213, 216–17, 225–7, 238 John of Mirecourt 190–1, 205

Walter Chatton 190, 195, 204, 212–14, 216–39 William Crathorn 190 William of Ockham 138, 213–14, 216–18, 222, 225, 232–9