Oxford Studies in Medieval Philosophy Volume 5 [1 ed.] 9780198806035, 9780198806042, 0198806035

Citation preview

OXFORD STUDIES IN MEDIEVAL PHILOSOPHY VOLUME 5

ADVISORY BOARD Marilyn McCord Adams, Rutgers Center for Philosophy of Religion ({ 2017) Peter Adamson, Ludwig-Maximilians University, Munich Peter King, University of Toronto Henrik Lagerlund, University of Western Ontario John Marenbon, Trinity College, Cambridge Calvin Normore, University of California, Los Angeles Dominik Perler, Humboldt University, Berlin Eleonore Stump, St. Louis University Editorial Assistant Philip Choi, University of Colorado

Oxford Studies in Medieval Philosophy Volume 5 Edited by

ROBERT PASNAU

1

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 2017 The moral rights of the authors have been asserted First Edition published in 2017 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available ISBN 978–0–19–880603–5 (hbk.) 978–0–19–880604–2 (pbk.) Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY 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, 10/5/2017, SPi

Contents Articles Anselm on Necessity Brian Leftow Avicenna on the Origination of the Human Soul Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi The Flow of Powers: Emanation in the Psychologies of Avicenna, Albert the Great, and Aquinas Charles Ehret Aquinas on the Individuation of Substances Jeffrey E. Brower Peter Auriol on the Intuitive Cognition of Nonexistents: Revisiting the Charge of Skepticism in Walter Chatton and Adam Wodeham Han Thomas Adriaenssen Ockham on the Parts of the Continuum Magali Roques

1 41

87 122

151 181

Text Yah.yaˉ Ibn ʿAdī on a Kalaˉ m Argument for Creation Peter Adamson and Robert Wisnovsky

213

Briefly Noted

241

Notes for Contributors Index

243 245

Anselm on Necessity Brian Leftow

Anselm did not write much about modality, and what he did write is puzzling. Anselm glosses necessity in terms of power, causation, and even compulsion. Yet he uses the concept as if he had alethic necessity in mind. Anselm distinguishes two sorts of necessity: most translate his terms as ‘antecedent’ and ‘consequent,’ but I use ‘prior’ and ‘following.’ As Anselm’s mix of the causal and the alethic seems odd to us, Anselm scholars have tended to pick the two apart.1 They have thought the talk of power etc. can only apply to physical, causal, or natural necessity, and tended to identify this with prior necessity. They have then had to complete the story by finding something in Anselm to express alethic or what we would call “broadly logical,” metaphysical, or absolute necessity.2 Some think Anselm’s “following” necessities do this; others think his talk of what cannot be conceived or thought does it. I think these approaches are incorrect. I now offer a different account. I say that Anselm’s power-based account is his metaphysics of all necessity whatsoever and that both prior and following necessity are absolute. I start with the modal metaphysics. I first dig it out of three texts. Then I discuss further texts to show that Anselm intends this metaphysics to apply specifically to what we would call absolute necessity. Finally I turn to two variations on the rival proposal, which segregates logical or “absolute” modality from the power-based account. To clarify the structure of the argument, I number the sections Tractatusstyle: 3.1 is a subsection under 3, 3.11 a subsection under 3.1, etc. 1 See A. D. Smith, Anselm’s Other Argument (Cambridge, MA: Harvard University Press, 2014); Thomas Williams and Sandra Visser, Anselm (Oxford: Oxford University Press, 2009); Katherin Rogers, Anselm on Freedom (Oxford: Oxford University Press, 2008), 170; D. P. Henry, The Logic of St. Anselm (Oxford: Oxford University Press, 1967), 177–80; and René Roques, Anselme de Cantorbéry. Pourquoi Dieu s’est fait homme (Paris: Cerf, 1963). 2 For an account of absolute necessity, see my God and Necessity (Oxford: Oxford University Press, 2012), ch. 1.

2

Brian Leftow 1. TH E M ETA P H YSIC S

Anselm’s account of necessity is this: All necessity is either compulsion or prevention (prohibitio), which two necessities are converted contrarily just as “necessary” and “impossible.” For whatever is compelled to be is prevented from not being, and what is compelled not to be is prevented from being, just as what is necessary to be is impossible not to be and what is necessary not to be is impossible to be.3

This is an account of what necessity is. Anselm frames it as a claim about all necessity. So he means this account to be universal in scope. Note that, as Anselm sees it, necessity consists in compulsion or prevention. These are “two necessities”: two bases for necessity-ascriptions. So for Anselm, talk of necessity is really talk about these. Anselm nails the point down, as he sees it, by showing that the relations between compulsion and prevention mirror those between necessity and impossibility: they are a viable basis for logical truths Anselm recognizes about the relation between necessity and impossibility. One thing Anselm wants from a modal metaphysics is that it be adequate to the modal logic he knows, and he sees this adequacy as a point in favor of his views. Thus we see at the outset the odd Anselmian mix of causal and apparently alethic concepts. Ordinary literal prevention and compulsion—the sorts of case that come first to mind when we read “compel” and “prevent”—involve at least potential exercises of power.4 (They can involve actual exercises, but they need not.) The presence of a watchful mother can prevent a child from running into the street. She need not actually do anything. It is enough that if the child began to run amok, she would act, and succeed. She prevents because she potentially exercises her powers. It is enough that her powers are there, ready to act. Equally, her presence compels that the child not reach the street. Thus Anselm’s is a power-based approach to necessity. A fact’s

3 Cur Deus Homo (henceforth CDH) II, 17, in S. Anselmi Cantuariensis archiepiscopi opera omnia, ed. F. S. Schmitt (Edinburgh: Thomas Nelson and Sons, 1940–61) (henceforth “Schmitt”) II, 123. 4 We sometimes use ‘compel’ and ‘prevent’ in other ways. We say e.g., that mathematical fact compels a perfect mathematician intending to answer correctly and functioning properly to get ‘4’ as the answer to “what is 2
2?” But this is just loose talk. If anything compels the perfect mathematician, it is the intention to get the right answer and the current well-functioning of the relevant bits of brain. Again, we can say that the fact that 2
2 = 4 prevents me from doubling 2 correctly and getting 5 as the answer. But this is just a roundabout way to say that no other answer is correct. In such cases, talk of prevention does not reflect the underlying metaphysics of the situation.

Anselm on Necessity

3

necessity, for Anselm, consists in facts about powers (which can include facts about how they are used). Anselm’s approach to possibility emerges in this text: before it existed, it was both possible and impossible that (the world) exist. It was impossible to that which did not have it in its power to make it exist. But it was possible to God, in whose power it was to make it be.5

Anselm explains ‘possible’ and ‘impossible’ in terms of ‘possible to’ and ‘impossible to.’ What is possible to x in the sense Anselm means is what x can bring about. For Anselm, then, possibly y exists if it is possible to some existing x to bring y to be. Further, possibly y exists because making it is possible to x. For Anselm, just as the world’s not being able to make itself makes it impossible with respect to itself, God’s being able to make the world makes it possible. For Anselm, there is possibility for things to exist if there is power, because the power accounts for the possibility. The same can and should apply to other possibilities—if I am blind, my seeing is impossible to me, but possible to the surgeon who can give me sight, and so possible. So for Anselm, it seems, many modal facts rest on facts about powers (as we see later, about lacks of power too).

1.1 Intrinsic and Primed Powers Powers can be intrinsic or extrinsic. An attribute F is intrinsic just if necessarily, for any x, whether x has F is settled solely by how things are with x, x’s parts (if any), and F (if there is such an entity), and extrinsic just if not intrinsic.6 I have the power to jump vertically ten feet. The earth’s gravity prevents my doing this, but I nonetheless have this power. I have it just because of the way I and my parts are—my leg-strength, nerve connections, etc. Nothing else is relevant. These things empower me to jump that high if circumstances are right. I have this power no matter whether circumstances let me use it. So my power to jump vertically ten feet is intrinsic, based just on the way I and my parts are. The key sorts of extrinsic power for Anselm’s theory are primed powers. A primed power is one that is in all ways ready to produce its effect. If it needs the co-operation of other powers, they are all present and primed, waiting to help. If circumstances must be propitious, they are. I would have the primed power to jump ten feet only if my circumstances let me use my intrinsic power—e.g., if I were on the moon—and all I had to do was try. My power is primed if and only if 5

De Casu Diaboli (henceforth De Casu) 12, Schmitt I, 253. How to define ‘intrinsic’ is controversial. Other definitions’ differences from this one would not affect the points I want to make, and so the controversy is not relevant here. 6

Brian Leftow

4

it lacks only an act of will or some other non-power triggering condition to operate—e.g., if all I need do is try and the power operates. De Libertate Arbitrii 3 shows us that Anselm distinguishes intrinsic from primed power. There he writes that “We have no power . . . which suffices by itself to act, and yet when those things are absent without which our powers are not at all led to act, we have the powers no less.”7 We have some powers whether or not circumstances let us use them. These are intrinsic. Anselm asserts that we have no powers that by themselves suffice to perform any act. That is, for him, none of our intrinsic powers are just of themselves primed. But he thinks we do have primed powers, powers that suffice to act: he gives the example of sight when light, a visible object, etc. are present.8 As he sees it, our primed powers are extrinsic. But Anselm must also allow that there are intrinsic primed powers, since he thinks that God has powers and has nothing from anything else:9 if so, then if God had no intrinsic primed powers, He could not actually do anything. For Anselm, only a primed power grounds a power-based possibility. For De Libertate Arbitrii 3 states that it takes four powers acting together to produce an act of seeing, and adds that when any of these is absent, “the other three neither singly or all at once can bring something about.”10 Absent the fourth, the other three are collectively powerless to produce sight. They are as powerless to produce sight as a non-existent world is to bring itself to be. But then just as the non-existent world, with no power to bring itself to be, does not ground a possibility that it exist, the three intrinsic powers together do not ground the possibility of sight. They do not, because they do not suffice for a primed power to see. A power-based account of possibility cannot apply to God’s existence or having His nature, because nothing has the power to cause these: not even God can cause Himself to exist or be divine, for He must first exist and be divine to cause anything at all.11 But Anselm adds a clause to his account of possibility that deals with these. He knew that if P, then possibly P. What Anselm adds is that what is actual is possible because it is actual.12 This applies to God: for Anselm, I suggest, God is possible because He is actual. To put it in our terms, His existence in any possible world is grounded on His concrete actual existence.

7

8 Schmitt I, 212. Schmitt I, 212–13. 10 Monologion 1–4. Schmitt I, 213. 11 For discussion of an attempt to make God the cause of His own divinity, see my “God and Abstract Entities,” Faith and Philosophy 7 (1990), 193–217. 12 Schmitt I, 252. The Latin is emphatic on that: omne quod est, eo ipso quia est potest esse. 9

Anselm on Necessity

5

1.2 Proper and Improper Modalities So far, we have seen Anselm admit two bases for necessity—compulsion and prevention—and two bases for possibility, namely actuality and power. Anselm also makes another distinction in the kind of basis a modal ascription can have. Throughout his career, Anselm invoked a distinction between “proper” and “improper” speech: his first work employs it,13 and his last work, the Lambeth Fragments, tries at length to clarify it. In general, Anselm thinks that we speak improperly when our language misrepresents what makes our claims true. Thus e.g., it is improper to say that justice departs from a fallen angel, when what is really the case is that the angel departs from justice,14 and it is improper to agree with a denial by saying “it is as you say it is”; strictly, what one should say is that it is not, as you say it is not.15 Improper speech is not as such false, but it is not metaphysically perspicuous. Further, where S is P improperly, S does not satisfy the full meaning of ‘P.’ If it did, the ascription would be proper. In “justice departs from a fallen angel,” the predicate taken literally predicates a change of justice. As Anselm sees it, justice does not change when an angel departs from it. Nothing happens to justice. Only the angel changes. We might say that justice (if there is such an entity) changes extrinsically and relationally. Anselm would not. For Anselm, justice is identical with God.16 Monologion 25 both allows this sort of extrinsic, relational change in God/justice and calls God/justice immutable. Anselm does both in the same chapter because he thinks that in such cases, properly speaking, God/justice does not change at all. Properly speaking, nothing happens to God/justice. (The sort of property that can arrive without anything really happening to its subject is (says Anselm) only improperly called an accident of that subject.17 The reason, unstated but clear, is that in this case properly nothing accidit God.) So taking all words properly, “justice departs from a fallen angel” says something false. But there is another way to take it. When an angel departs from justice, a proposition involving justice—that the angel is just— changes truth-value. We can read “__departs from a fallen angel” improperly, as asserting not a change of attribute in justice, but this change of truth-value—a change of attribution of justice. So read, “justice departs from a fallen angel” says something true, but misrepresents what makes it true. On this reading, justice satisfies part of the sense of the predicable: yes change, yes involving justice, but not in justice. Thus where there are proper

13 16

See Monologion 65. Monologion 16.

17

14 15 De Casu 27. Proslogion 7. Monologion 25, Schmitt I, 43.

Brian Leftow

6

and improper uses of a term ‘F,’ there will be distinct elements in its sense, and if ‘Fa’ is improper, A will satisfy only part of ‘F’s sense. Anselm distinguishes proper and improper ascriptions of modal terms. As he sees it, “a book can be written by me” is improper and “I can write a book” is proper because the power that grounds the possibility is in me, not the book.18 “S can do A,” Anselm thinks, ascribes a power. In the proper case, the power is where ‘can’ ascribes it, and in the improper it is not.19 A power, just as such, is at least partly inherent (even extrinsic powers have some intrinsic basis) and possibly produces an effect. So I suggest that as Anselm sees it, there are two elements in the sense of ‘can,’ one of inherence and one of possibility. In “a book can be written by me,” the power does not inhere in the book. So the book does not satisfy what Anselm sees as the full sense of ‘can.’ Taking ‘can’ properly, then, the sentence says something false. But we can also read ‘can’ improperly, as asserting merely the possibility of ,20 a possibility grounded on my powers. So read, the sentence says something true. Thus, as just noted, where a use is improper, the subject satisfies only part of the predicate’s sense. As for ‘can,’ so for ‘possible’: in Latin, ‘can’ is just ‘potest,’ a form of ‘posse,’ the word from which ‘possibilitas’ derives. Anselm treats necessity similarly. He writes in Cur Deus Homo that In the end, God does nothing by necessity, because He is in no way compelled to do something or prevented from doing something. And when we say that God does something as if by necessity of avoiding dishonor . . . it is preferably understood that He does this by necessity of preserving His honor. This necessity is nothing other than the immutability of His honor, which He has from Himself alone and not from another, and for this reason it is improperly called necessity.21 ‘Necessity’ is said improperly where there is neither compulsion nor prevention.22

Anselm reasons that if God’s immutably being honorable is entirely from Himself, it is not compelled, and if it is not compelled, it is not properly a case of necessity. For Anselm, as possibility is based on power, necessity is based on compulsion or prevention. Just as for Anselm “S can do A” says something true and uses ‘can’ properly if and only if the relevant power is in S, “S necessarily is P” says something true and uses ‘necessarily’ properly if and only if its grounding compulsion is in or on S or its grounding prevention is of S. In Anselm’s eyes, only so does ‘necessarily’ perfectly reveal the underlying facts. If S is not compelled/prevented, ‘necessarily S is P’ may express a truth, 18

De Casu 12. In this case, it cannot be. A not yet written book lacks even what Aristotelians call passive powers. 20 ‘

’ is a name of the proposition normally expressed by the sentence ‘P.’ 21 22 CDH II, 5, Schmitt II, 99–100. CDH II, 10, Schmitt II, 108. 19

Anselm on Necessity

7

but misleads about the underlying facts. The second text might suggest that Anselm allows cases of improper necessity not grounded in compulsion or prevention at all. But “where there is” could as easily make a point about a subject rather than an overall situation. The text might as easily say that ‘necessity’ is said improperly of a subject on which there is neither compulsion nor prevention, as in “God necessarily maintains His honor.” The point is then not that there is no compulsion or prevention in the picture at all, but rather that if it is present but not in the subject called necessitated, the necessity is of a particular kind. As we see later, on related subjects—the necessity that God be truthful, for instance—Anselm does cash the necessity out in terms of compulsion/prevention on things other than God. If we read this case in parallel, for Anselm, God freely maintains His honor, and so does not do it by necessity/compulsion. But behind the use of ‘necessarily’ stands an immutable will which prevents other things’ making Him act dishonorably. The necessity “is” the immutability of His honor because this is what does the preventing. To tidy Anselm’s theory, note that prevention or compulsion effect a lack of primed power. (If the mother prevents the child from running amok, then because of her presence—an unpropitious circumstance—the child lacks the primed power to do so.) And necessity involves a lack of possibility: if S necessarily is P, it is not possible that S not be P. So Anselm holds that proper possibility has a power-basis in the subject of the possibility, and proper necessity involves lack of possibility and a lack-ofpower basis in the subject of necessity. Let ‘□P’ symbolize proper and ‘□I’ improper necessity. When Anselm says “In the end, God does nothing by necessity,” he speaks of proper necessity. But he then shows himself willing to grant that God necessarily does something—as long as we understand that the necessity is improper. So Anselm’s discussion of God’s honor really makes these claims: Nothing compels or prevents God. So for all acts A, ¬□P(God does A). But for some act A, □I(God does A), because the immutability of God’s honor explains His doing A.

2 . A N S E LM ’S M O D A L IT Y We have Anselm’s metaphysics of modality. Let’s now explore what sort(s) of modality it applies to. Metaphysical bases that P are why in some or all of a class of possible worlds, P. They are either what it is for

to be so in these worlds, or what brings it about that P in them. But it is one thing what the metaphysical basis of a modal fact is, and another what kind of modality that fact involves. That is, it is one thing what makes it the case that P in

Brian Leftow

8

some or all of a class of worlds, and another what worlds that class includes: for physical necessities, for instance, the class is all worlds with our world’s natural laws, while for absolute necessities it is all worlds whatsoever. So having seen Anselm’s metaphysical basis, we now must ask what sort(s) of modality it is a basis for. There are two questions to ask about this: what sort Anselm intends his metaphysics to ground, and whether it can do what he intends. I discuss Anselm’s intention at length, and make just a brief comment about the second question. It is natural to think, as Anselm scholars have tended to,23 that Anselm’s modal metaphysics can ground only physical modalities. Here is why. For Anselm, leaving aside God’s existence and having His nature, it takes an allthings-considered power to have a possibility. So suppose that I am not blind, my eyes are open and I am in Darkworld: there is no light and nothing able to produce light. Then I lack the all-things-considered power to see. And so for Anselm, it is not possible that I see. I necessarily do not see. Now suppose that I am not blind, my eyes are open, I am in the dark and something has the primed power to light my location while my eyes stay open. Then I have all-things-considered power to see. I do not control whether I use this power. The lighter does. The lighter, not I, must trigger my use of it. But I do have it; the lighter need only act and whap—I see. And so for Anselm, it is possible that I see. Now if I am in Darkworld, it is physically impossible that I ever see. But I could have been in a lighted world instead. This is at least absolutely possible. So it is absolutely contingent, not necessary, that I cannot see. Seemingly the necessity in Darkworld Leftow’s “I cannot see” must be merely physical. This argument supposes that Darkworld is possible, though, and we think this only if we ignore God. As Anselm sees it, God necessarily exists. With absolute necessity, there is a cause able to say “let there be light.” If there is, Darkworld is absolutely impossible. If God necessarily exists, then necessarily, if I am not blind and my eyes are open, I have the primed power to see, for necessarily, I co-exist with God, and God’s power is primed to light any darkness. Just as a non-existent world absolutely cannot exist of itself, but absolutely can exist due to God’s power, Leftow in a world without physical sources of light absolutely cannot see of himself, but absolutely can see due to God’s power. There is no Anselmian necessity that I not see in Darkworld, for there is no Darkworld. So the argument that Anselmian necessity must be merely physical fails. With an omnipotent God in the picture, necessities grounded on powers can indeed be absolute. Because omnipotence can actualize or create powers able to actualize the 23

See e.g., those noted in n.1.

Anselm on Necessity

9

whole of absolute possibility, an omnipotent God puts the whole of absolute possibility within the range of actual powers.24 Anselm’s theism lets him maintain that some power grounds every absolute possibility save God’s existence and having His nature. So Anselm can indeed say that where there is no power at all to have it be the case that P and it is not the case that P,25 it is absolutely impossible that P. Anselm need not say that power grounds only physical modality. I think that Anselm means his power-based metaphysical story to be inter alia about necessities we would call absolute, metaphysical, or broadly logical. I say this because he sometimes uses it to ground these. I now discuss four texts that display this.

2.1 First Text: God’s Veracity Consider Cur Deus Homo II, 17: When we call something necessary to be or not be in God, it is not understood that there is in Him a necessity which either compels or prohibits, but it is signified that in all other things there is a necessity prohibiting them to bring about and compelling them not to bring about what is against that which is said of God. For example, when we say that it is necessary for God always to speak the truth . . . nothing else is said than that so great is His constancy for maintaining the truth that it is necessary that nothing can bring it about that He does not speak the truth.26

I first explicate this text, then discuss what sort of necessity is in play in it. As Anselm does not think that God is always speaking,27 he probably means by “it is necessary for God always to speak the truth” that NT.

Necessarily, if God asserts something, what He asserts is true.

24 See my “Omnipotence,” in T. Flint and M. Rea (eds.), The Oxford Handbook of Philosophical Theology (Oxford: Oxford University Press, 2009), 167–98. The wiggle about creating powers is because an omnipotent being cannot bring about libertarian-free agents’ actions, but they can. Thus God’s power to create such agents makes their free actions possibly possible. If Anselmian modality validates S5, as in the Anselmian account of my “The Nature of Necessity” (forthcoming in Res Philosophica), possible possibility gives Anselm possibility. 25 The second clause is needed because Anselm allows that possibly P because P. 26 Schmitt II, 123–4. 27 In the sense that at some times, no one hears God’s voice. If He is atemporal, as Anselm believes, then at all times, it is timelessly true that He timelessly-is doing whatever He ever timelessly-does. For a timeless God cannot change. So at all times, it is timelessly true that God timelessly-is speaking to Moses. In that sense, God is always speaking, for Anselm. But “constancy for maintaining the truth” is a habit manifest in communicating. So Anselm has in mind a virtue that operates when people hear God’s voice.

10

Brian Leftow

What makes (NT) true is that His “constancy for maintaining the truth” is so great. This constancy is just His absolute immutability of willing to tell only truth.28 Anselm gives what might seem to be two accounts of what the necessity in (NT) consists in: that “all other things” are prevented from making Him lie and even from having the power to do so, and that everything is prevented from this (“nothing can bring it about . . . ”). But one way or the other, it consists in prevention. If there are two accounts here, they differ over whether the prevention applies even to God. But perhaps there are not two accounts. The second formulation occurs in an example intended to illustrate the first, and so perhaps the quantifier ‘nothing’ has an implicit restriction based on the first formulation. Anselm writes first that God is not compelled to speak truly. But Anselm does not say that (NT) is false. Instead, he lets (NT) state a truth improperly by basing (NT) on prevention. Let us now ask what kind of necessity (NT)’s is. Anselm calls God “supreme truth” and “truth itself.”29 This characterization of God’s nature may be enough to ground (NT)—how could truth itself speak falsely? In any case, (NT) is part of His goodness, and so His nature. His immutability is also part of His nature. Proslogion 7 argues that inability to lie follows from God’s omnipotence, which is again of His nature.30 For Anselm, then, God’s nature grounds (NT). Necessities expressing natures are absolute or metaphysical. So here Anselm’s powerbased ideas about necessity ground ascription of what we would call absolute or metaphysical necessity. Anselm would surely call the necessity in (NT) absolute if he had an explicit concept of absolute necessity, and not just because it expresses a nature. Cur Deus II, 10 asks how an incarnate God can freely keep justice if He is unable to sin. Had Anselm believed that there was some objective sense of possibility in which possibly God lies, he would surely have made that possibility part of his answer: He can do otherwise than keep it.31 So for Anselm, (NT) has a necessity that rules out all objective possibility of its contradictory. Only absolute necessity does this. If and only if it is absolutely necessary that P, it is not objectively possible that P in any sense, however weak.32 We thus can be confident that if Anselm had an explicit concept of absolute necessity, he would insist that (NT) is absolutely necessary.33

28

29 CDH II, 5, Schmitt II, 100. E.g., De Veritate 4. Schmitt I, 105. 31 This would of course force a weaker-than-absolute reading of ‘unable’ in ‘unable to sin.’ 32 Again, see my God and Necessity, ch. 1. 33 For that matter, on Anselm’s terms, it is not even epistemically possible that God lie: that He lie is not compatible with things we know about His nature, and even know a priori by perfect being theology. 30

Anselm on Necessity

11

So here we have what is in fact an absolute necessity, and what Anselm would have called one, grounded on a fact about prevention—that God’s nature keeps anything else from causing Him to lie.34

2.2 Second Text: Justice and Immortality In his last completed work,35 the De Concordia, Anselm writes that Often we call what is not compelled to be by any force “necessary to be,” and call what is removed by no prevention “necessary not to be.” For instance, we say that it is necessary for God to be immortal and . . . not to be unjust, not because some force compels him to be immortal or prohibits him to be unjust, but because nothing can bring it about that he not be immortal or be unjust . . . 36

These are improper necessities, and if Cur Deus gives two accounts of improper necessity, one of them disappears: nothing at all can bring it about that these things are not so. The underlying proper necessity is explained in power-based terms: lack of power constitutes an impossibility, which gets expressed as a necessity for God.37 The necessities that God be immortal and not unjust are absolute. Anselm would not let exceptions to them have any sort of possibility. Once again, theistic absolute necessities get a (lack of ) power-based parsing.

2.3 Third Text: Following Absolute Necessity In Cur Deus Homo, Anselm explains “following” necessity this way: There is a prior necessity, which is a cause of a thing’s being so, and a following necessity, which a thing causes . . . when I say that you necessarily speak because you 34 As Anselm parses divine necessities in terms of other things being prevented, one might wonder whether, even if God’s nature is intrinsic, its necessity is for Anselm extrinsic: His veracity’s necessity seems to depend on the presence of other things, which are prevented from making Him lie. But a good lock prevents any burglar from stealing a bicycle, even if there are no burglars around. It prevents theft by its own internal structure and strength, if thieves are around and if they are not. So its preventing theft is intrinsic. God’s veracity prevents His being caused to lie if causes that might try to get a lie out of Him exist and if they do not. The prevention arises simply from the way He is internally. So it seems intrinsic to Him. What is extrinsic is whether the prevention generates a proper necessity in addition to the improper one. Of course, for this to represent Anselm, he would have either to let there be an improper necessity with no grounding proper necessity or to let a proper necessity exist if its subject does not. I think he would accept the latter, as I think he holds that not all subjects of true predications exist. But I cannot pursue this here. 35 His incomplete Philosophical Fragments might be later. 36 De Concordia I, 2, Schmitt II, 247. 37 Even God lacks the power to bring these things about. Anselm might explain this lack-talk as improper speech of an underlying fullness of power, as in Proslogion 7.

Brian Leftow

12

are speaking, it is following . . . for when I say this, I signify that nothing can bring it about while you are speaking that you are not speaking.38

I deal first with the prior/following distinction, then with the necessity and the “following” character of Anselm’s example. Some things are so because they must be so. Anselm calls these priornecessary; in these cases, he thinks, a necessity prior in some explanatory order to the fact explains the fact. (Why is it so? Because there was no alternative.) Other things must be so because they are so. In the latter sort of case, necessity follows upon actuality—that you must be speaking follows upon your speaking. Before you spoke, you need not have; as you spoke and thereafter, it could not be otherwise than that you spoke then. Note Anselm’s “cause” and “bring about.” For Anselm, your speaking causes an inability in other things, or prevents their doing something. When Anselm said that all necessity consisted in compulsion or prevention, he was not overlooking his belief that necessity comes in two varieties. He was saying something he meant to apply in some way to both prior and following necessity. We must bear in mind, though, that Anselm uses ‘cause’ for a wide range of explanatory relations—the Lambeth Fragments are one long illustration of this. We would not say that your speaking causes other things’ impotence. But we might well say that it explains it and prevents their bringing the opposite about. I now take up the necessity in following necessity. Your speaking is following-necessary, but you were not and are not compelled to speak. Improper necessity for you is grounded on compulsion or prevention or impotence of other things. For “when I say . . . that you necessarily speak . . . I signify that nothing can bring it about . . . that you are not speaking.” But also, if nothing can bring this about, you cannot bring it about. So if Anselm means ‘nothing can’ seriously, in this case you impose a proper necessity on yourself. What ‘necessary’ signifies—what the necessity consists in—is lack of power. For Anselm, your speaking prevents—“you necessarily speak because you are speaking”—because it creates a lack of power and lack of power prevents. Only if lack of power prevents does nothing’s having the power to bring something about yield an impossibility that it come about. Recall Anselm’s assertion that it was impossible for the world to come to be, considered just in itself, because it did not have the power to bring itself to be: if so, had no existing power been able to make it, the world could not have come to be, and so would have been prevented from existing. Anselm thinks that lack of power prevents because he holds a strong principle of sufficient reason. For Anselm, “nothing is through nothing. It cannot even be thought” that something be, save through something.39 In his eyes, it is just rationally evident that whatever comes to be, is brought to be—that nothing “just 38

Schmitt II, 125.

39

Monologion 3, Schmitt I, 15–16.

Anselm on Necessity

13

happens.” If whatever happens is brought about, lack of power to bring about guarantees not coming about, and so prevents coming about. Lack of power prevents, or compels that not. In Anselm’s example, all things are compelled not to cause you not to speak—prevented from causing this. I emphasize: for Anselm, compulsion not to cause and prevention consisting in lack of power are what ‘necessarily’ signifies. They are what make the necessity-ascription true. They are, again, what the necessity consists in. But what is being prevented is that something bring it about that you both speak and not speak at once. The necessity is that things not cause a contradiction to be true, and so that a contradiction not be true. So the metaphysics of following necessity is the metaphysics of a contradiction’s not being true. Anselm explicates the necessity of a contradiction’s not being true in terms of compulsion, prevention, powers and their lack. The necessity that contradictions not be true is the strongest there is. It is absolute. It is “logical.” Thus the necessity in following necessity: now to its “following” character. By calling this a following necessity, Anselm signifies that your speaking explains your speaking’s being necessary. Anselm does not say just how it does this. He merely states the result, the thing brought about, that nothing can bring it about that you are then not speaking. Let’s try to fill out the story. The most natural first step is this: if you are now speaking, it is too late for anything to bring it about that you are now not. Your now speaking makes it too late. Other things had a chance to keep you from speaking now; once you speak, they have lost their chance. We must next ask how your speaking makes it too late. I first give two stories that will not do, then one that will, then one that is still better. 2.3.1 A Pure Non-Contradiction Option The simplest way to explain this lets the principle of non-contradiction (PNC) do all the work: “if I am speaking, then if I also do not speak, I make a contradiction true. There is no possibility that a contradiction be true. So if I am speaking, there is no possibility that I also not speak. So if I am speaking, my speaking is necessary.” This explication fits neatly with the idea that following necessity is purely logical, i.e. based somehow on logic rather than power-facts. But it won’t do to fill out Anselm. Anselm did not say, “ . . . I signify that while you are speaking you cannot also not speak.” He said, “ . . . I signify that nothing can bring it about while you are speaking that you are not speaking.” The pure PNC option just ignores the bringing-about bit of the text. 2.3.2 A Mixed Non-Contradiction Option Another option mixes PNC with Anselm’s power-based ideas. It runs this way: before you spoke, something could have made you not speak without

14

Brian Leftow

making a contradiction true. Now that you are speaking, it is too late: if something adds your not speaking to reality, it makes a contradiction true. Contradictions cannot be true. So nothing can now add your not speaking to reality. Given PNC’s necessity and your speaking, nothing has the primed power to add your not-speaking to reality. Though things retain the intrinsic power they would have used to do so, circumstances including your speaking and PNC’s necessity don’t let them use it, and so they have lost the primed power. For Anselm, no power, no possibility. So it is now not possible that you not be speaking. So your speaking is necessary. Thus PNC helps explain how your speaking makes it necessary that you speak. We could put the explanation more carefully as follows: 1. 2.

3.

4. 5. 6. 7. 8. 9.

You speak at t. Necessarily, if you speak at t, then if something causes you not to speak at t, something brings it about that you speak at t and you do not speak at t. So Necessarily, if you speak at t, then if something has the primed power to cause you not to speak at t, something has the primed power to bring it about that you speak at t and do not speak at t. Necessarily, no contradiction is true. So Necessarily, nothing has the primed power to bring it about that you speak at t and do not speak at t. So Necessarily, if you speak at t, nothing has the primed power to bring it about that you also not speak at t. So Necessarily, if you speak at t, it is not possible that you also not speak at t. So, Necessarily, if you speak at t, then necessarily, you speak at t. So Necessarily, you speak at t.

Anselm states that the necessity in (9) “signifies” (5). That is, (9) makes a necessity-claim, and its real basis is (5). So the real explanandum here is (5); (6)–(8) just show how (5) grounds the final modal claim via Anselm’s story about modality. If Anselm would appeal to PNC and PNC would have no roots in his power-and-prevention metaphysics, then for him, PNC would be a second source of necessity apart from power and prevention, a source involved in “following” necessity. This would make it plausible that “following” necessity is “logical” in some way prior necessity is not. But for three reasons, I do not think this is how we should fill Anselm out. To begin, the necessity in (5), the thing Anselm wants your speaking to explain, is the necessity of a case of PNC. For in (5), “necessarily, nothing has the primed power to bring it about that” just parses ‘nothing can,’ which

Anselm on Necessity

15

for Anselm is the real freight of ‘□,’ what it signifies, and in (5), this modal operator precedes a case of PNC. So if Anselm would appeal to PNC, he would use the necessity of PNC to explain the necessity of a case of PNC. But in a strong sense, this does not explain at all. The necessity of PNC just is the necessity of all its cases. Thus the explanans contains the explanandum, and so does not explain it. Further, as there is nothing special about your speaking, a good explanation of your speaking’s necessity will apply to any case of PNC at all, and so explain why PNC itself is necessary. It will turn Anselm’s particular case into just an illustration of why PNC is universally necessary. And to my eyes, at least, this is what Anselm’s passage is trying to offer, by basing following necessity in a real event bringing about real inabilities. We want (in effect) to know why all cases of PNC are necessary; Anselm’s illustration is just a way to talk about the broader question. That all cases of PNC are necessary cannot explain why all cases of PNC are necessary. So it is most charitable not to read Anselm as appealing to PNC. Again, explanation via PNC would not explain all that needs explaining here, and so again, it would be most charitable not to ascribe one to Anselm. The following necessity Anselm seeks to explain is that nothing can bring it about while you are speaking that you are not speaking. One way to bring this about would violate PNC. That is one reason I represented the explanandum above as (5). The necessity of PNC rules this way out. But there is another way to bring it about while you are speaking that you are not speaking. This would be to substitute your not speaking for your speaking. This would make it the case while you are speaking, i.e. at the time you are in fact speaking, not that you are both speaking and not speaking, but that you are instead then simply not speaking. Your speaking yields a full explanation of Anselm’s explanandum only in conjunction with something that explains not just why nothing can make a contradiction true but also why nothing can do this further trick. PNC won’t help with this. For if I substitute ¬P for P, I do not make it the case that P & ¬P. Rather, “first” it is only the case that P, “then” it is only the case that ¬P. So a good account of everything Anselm needs to explain cannot proceed by PNC alone—and if the account’s second element can also account for the truth of PNC, PNC drops out as an ultimate explainer of necessity, and the account consists entirely of this second element. I think this does in fact happen. I suggest how below. To put the problem here more pointedly, (2) looks shaky. (2)’s last consequent contains only one disjunct of what should be a disjunction with at least two disjuncts. For violating PNC is not the only impossibility whose impossibility we need to explain. At least two impossibilities could figure in a story of an impossible situation in which someone causes you not to speak then: contradiction, and the “substitution” just noted. We see shortly that this is reason to prefer one particular way of filling out Anselm and grounding PNC.

16

Brian Leftow

Finally, if Anselm did not rest PNC’s necessity on his power-based metaphysics, he would have to account for it some other way. What I now say does not just make a point against the mixed non-contradiction option.40 Rather, it shows more generally why it must be a mistake in reading Anselm to treat following necessity as having anything other than a power basis. Thus it shows why the many Anselm scholars alluded to earlier must be wrong in seeing in Anselm a necessity that is somehow purely logical (whatever that means, exactly). It also shows why Anselm’s modal metaphysics as a whole should be read as power-based. For if (as I now argue) even PNC must be power-based, then we must take the whole story to be power-based. Suppose that Anselm does not rest PNC’s necessity on his power-based metaphysics. Then there is a limited menu of alternate sources of necessity: i. Our actions, as conventionalists suggest. ii. Our or other creatures’ non-power attributes. iii. Platonic entities independent of God. iv. God’s being or ideas as apart from His power. I now argue that Anselm cannot move in any of these directions. As to (i), Anselm gives no hint of even having conceived of conventionalism. If it was in the air at all, it would have been as a consequence of Roscelin’s “flatus vocis” nominalism. Anselm derided that view.41 Anselm’s predilection on non-logical necessity was Platonist. Monologion 1 argues that something plays the role of the Platonic Form of the Good, which is somehow “in” all good things.42 This turns out to be God, and by the time of the Reply, Anselm is clear that God exists necessarily. Thus for Anselm, truths about goodness—and any other attribute in God’s nature—are grounded on the nature of a necessary, simple, timeless entity. Further, perfect being theology—for Anselm a way of characterizing the nature of God43—delivers the conclusion that God is supreme wisdom.44 Anselm says that this entails that He “speaks” the “Word” which contains all natures.45 So all natures are for Anselm somehow in God’s mind46 by His very nature, and so all truths about all natures are grounded on the nature of a necessary entity. So Anselm’s views rule out modal conventionalism about nature-truths. While he says nothing specifically about God’s providing a basis for logic, it would

40

41 My thanks here to a referee. De Incarnatione Verbi 1. Anselm’s account of original sin seems to require a fairly strong reading of this immanence: see de Conceptu Virginali 2. 43 For Anselm, God has no accidents (Monologion 25). So there is nothing else for it to characterize. 44 45 46 Monologion 16. Monologion 32. Monologion 9–11, 34. 42

Anselm on Necessity

17

be incongruous for him to be so hyper-realist about these other things and shift to conventionalism for that. Further, conventionalism is really powerbased: it rests necessity on our powers (and incapacities) of language-use. So (i) doesn’t really leave the power-based realm. Turning to (ii), if Anselm wanted to rest PNC on concrete creatures but not on their actions or powers and incapacities, he would have to rest it on their bare existence or non-power attributes. It was orthodox to hold that God need not have created. So anyone orthodox who rested PNC on concrete creatures’ attributes or existence would have had to hold that PNC is only contingently true. Thus orthodoxy would rule out a creature-based account of PNC’s necessity. Anselm meant to be orthodox. So Anselm’s commitments militate against (ii). As to (iii), Anselm does not believe in Platonic entities independent of God. For Anselm, like Augustine, God or His ideas play the Form roles. Finally, as to (iv), on Anselm’s doctrine of divine simplicity, there are no such things as God’s being or ideas apart from His power.47 Thus it seems that Anselm would have to look to his power-based metaphysics for an account of the necessity of PNC. And he can do so. 2.3.3 A Purely Power-Based Option Suppose that 10.

11. 12. 13.

if all existing causes jointly have the primed power to bring it about that P in the circumstances, and that P plus the circumstances entail that Q, where Q is not one of the circumstances, then existing causes jointly have the primed power to bring it about that Q in the circumstances, but existing causes jointly don’t have the primed power to bring it about that Q in the circumstances, and that P plus the circumstances entail that Q, which is not one of the circumstances. Then existing causes jointly don’t have the primed power to bring it about that P.

This reasoning is in terms of jointly possessed powers. I say that if A individually has the power to bring it about that P, all existing causes jointly have that power in virtue of A’s having it. (This is an extension of the idea that I have the power to think because I have as a part a brain with this power.) But causes can jointly have a power none have individually. The sun cannot melt ice without ice to melt, and ice cannot melt without a source of heat: so the 47

See De Incarnatione Verbi 7.

Brian Leftow

18

sun and an ice cube jointly have the power to bring it about that ice melts, though neither has the power individually. Now suppose that one circumstance is that ¬P. Then that P plus the circumstances entail that P and ¬P, which is not one of the circumstances: so a case of (12) is true. Existing causes don’t have the primed power to bring this about. Everything whatever lacks that power. So a case of (11) is true. So if (10) is true, Anselm can get a case of (13): existing causes jointly lack the power, given this circumstance, to bring it about that P. Again, no power, no possibility. And so we rest PNC’s necessity on lacks of power: for any P, given that ¬P, nothing has or ever will have the primed power to bring it about that P. The key obviously is (10). Consider a simpler principle, that if something has the power to bring it about that P, and that P entails that Q, that thing has the power to bring it about that Q. This is false, for perhaps it cannot bring it about that Q, and can manage the trick with

only because something else takes care of for it. This sort of problem does not arise for all causes jointly. There is nothing beyond them to bring it about that Q. So they can manage

only if they can also manage , unless it can “just happen” that Q precisely when it happens that P. But if all causes jointly depend on it “just happening” that Q, they do not have the primed power to bring it about that P. Nothing in the situation parallels the lighter in my blindness example earlier: no power is present and primed to bring it about that Q. All causes jointly could not “just try” to bring it about that P. So to speak, they would have to both try and pray to nothing that Q. Thus (10) appears true, and so the case of (10)–(13) Anselm requires seems sound. 2.3.4 Another Power-Based Option The second way to fill out how your speaking explains its own necessity runs as follows: you are now speaking. Even if something could add to reality that you are not, thereby violating PNC, there is something else it could not do. It could not simply uproot your speaking and replace it with your not speaking. Adding that you are not speaking would only add to settled reality—it would not subtract from it.48 Nothing has the intrinsic power to uproot settled reality. Nothing can uproot the settled present for the same sort of reason that nothing can uproot the settled past. (We see Anselm’s account of that shortly.) So now that you have settled the present, nothing can settle it otherwise. Things have lost their primed power to do so, because they do not have the intrinsic power to unsettle what is settled. As we have seen, for Anselm, no primed power, no possibility. Thus nothing—not even 48

We see shortly that this is not quite right. But it will do as a first pass.

Anselm on Necessity

19

God—can by “uprooting” bring it about while you are speaking that you are not speaking. As I shortly show, this entails that nothing can make a contradiction true. Lack of this intrinsic power thus entails that nothing can bring it about in any way at all while you are speaking that you are not speaking. So this lack of intrinsic power tells us how your speaking brings about its necessity. It does so due to the universal lack of this intrinsic power. Lack of intrinsic power to uproot settled reality can explain both things Anselm needs to explain. Lacking this intrinsic power explains why I lack primed power to uproot settled reality. Asserting that I lack it does not simply repeat that I cannot do it, since there could be other reasons I lack a primed power, namely circumstances that deny it to me. Further, lacking the power to unsettle what is settled can explain why we cannot violate PNC. For if nothing has the power to uproot what is settled, nothing can add the contradictory of your speaking to reality, since as of now, it is settled that your speaking is present and its contradictory is not. I am of course not suggesting that if only something had acted in time, had gotten in the notspeaking just as you started to speak, something could have made a contradiction true. For every pair of propositions {P, ¬P}, I submit, what makes one of them true is just the absence from reality of what makes the other true.49 If that is correct, the only way to make a contradiction true is to remove the truth-maker for something at the same time it is placed in reality. Now at every instant, this truth-maker either is there or is not. If it is not there, it cannot be removed. If it is there, then if nothing has the power to uproot it, it cannot be removed. Either way, then, it cannot be removed, and so a contradiction cannot be made true. The truth-maker either has been placed, is being placed, or has not yet been placed in reality. If it has not yet been placed, it cannot be removed because it is not there yet. If it has been placed, it is too late to uproot it. It is being placed only at the very instant it is placed. If it is being placed then, this is just the first instant of “too late,” because this is the first instant at which it is settled that it is there. Thus it is not possible to make a contradiction true “by acting just in the nick of time,” and the reason for this again is inability to uproot what is settled. Necessity of a case of PNC is as absolute as necessity gets, and Anselm can explain even this in (lack of ) power and prevention terms. As I see it, his explanation can be in terms of (lack of ) power and prevention all way down. I do not mean this symmetrically. Rather, I take it that one of these will be “really” positive, the other “really” its negation: one will really be true due to a presence, the other due to an absence. Which is which may not correspond to their surface forms (consider “there is a vacuum”). What makes the “real” negation true is the absence of what would make the “really” positive proposition true. 49

Brian Leftow

20

As only this account does the full job Anselm needs to do, it is most charitable to fill Anselm out this way; as the power-and-prevention account of modality is clearly there in Anselm’s text, it is faithful to his text to do this as well. And as noted earlier, once we do this, PNC drops out as an independent element of Anselm’s account of modality. It is powers all the way down.

2.4 Fourth Text: The Necessity of the Past Let’s take up another text. In Cur Deus II, 17, Anselm asserts that all necessity and impossibility are subject to God’s will, then illustrates: it is impossible for what is past not to have happened—that is, necessarily Past.

the past does not change.

But (says Anselm) this necessity is due only to God’s willing that truth be immutable.50 Now Past is an absolute, metaphysical necessity. It is part of time’s very nature. For Anselm, what makes Past necessary is that God immutably wills that Past be true. By so doing, He denies anything the primed power to change the past, since nothing has the intrinsic power to overrule an omnipotent will. God’s immutable willing causes inabilities, and so Past is necessary. For Anselm, then, Past’s absolute necessity consists in a lack of power God’s will effects. Further, truth’s being immutable makes it the case that nothing can uproot settled truth, present as well as past. So Anselm traces this lack of power too to an exercise of divine will. And as we have seen, what explains this lack of power can also explain the truth of PNC. Anselm gave a different story in Proslogion 7. There he suggested that God cannot change the past because He is omnipotent: because ability to do this would be impotence, not power. It would be impotence, Anselm asserts, because changing the past would be either wrong or to God’s disadvantage. If he had his Cur Deus story in mind this early in his career, he could flesh the last claim out this way: God willed that Past be necessary for the best moral reasons (whatever they are), and so it would be wrong of Him to contravene what He willed. Or perhaps the thought is that talk of His inability to change the past is really improper talk of God’s power to cleave to the truth, just as in the same chapter, talk of His inability to sin is really improper talk of His power to cleave to the good. Be this as it may, Proslogion 7 rests the necessity of Past on God’s power. The story is about power, yet (again) the necessity being explained is absolute. 50

Schmitt II, 123.

Anselm on Necessity

21

2.5 A Complicating Text There is one complication to my overall story: while we have seen De Concordia 1 repeat a Cur Deus account of “following” necessity, De Concordia 2 gives what looks like a different account. Anselm writes, If it will be, necessarily it will be. But this necessity neither compels nor prohibits . . . Because a thing is posited to be, it is said to be necessarily . . . this necessity follows the positing of the thing. It . . . signifies only that what will be will not be able at the same time not to be.51

Again, we have improper necessity, but it might look as if Anselm is treating PNC as a source of necessity divorced from power considerations. However, this is not clear. The necessity “signifies” that what will be will lack an ability. We can as easily take that in a thick sense, as ascribing a lack of power—I could have translated the text’s “non poterit” as “will not have the power,” but did not want to be tendentious—as in a thin, which would treat this only as a way to say that it will not be (say) logically possible that it also not be, because contradictions can’t be true. The thick-sense explication is for Anselm just a consequence of one Cur Deus account of following necessity: that nothing at all will be able to bring this about entails that it will not be able at the same time not to be, given that nothing “just happens.” Thus that Anselm has not changed his mind about his Cur Deus story is at least as likely as that he has. It becomes more likely when we recall that De Concordia 1 repeats that Cur Deus account, from which (again) the thick-sense explication follows. Continuing, Anselm illustrates: It is always necessary for white wood to be white, because neither before nor after it is white can it be made simultaneously white and not-white. Similarly . . . it is always necessary for a present thing to be present, because neither before nor after it is present can it be simultaneously present and not present.52

Again, it is not clear that Anselm is treating PNC as a source of necessity divorced from power considerations. This text uses ‘cannot be made’ about being at once white and non-white, and so explicates the necessity in lackof-power terms, and again as in Cur Deus. (“Nor . . . can it be made” is just a passive-voice way to say “nothing can make it.”) ‘Similarly’ may suggest that Anselm would extend this account to the next illustration. Anselm goes on: When we say of what is future that it is future, what is said is necessary . . . just as when we say the same of the same. When we say that every human is human, or that if it is human, it is human . . . what is said is necessary, because something cannot at 51

Schmitt II, 249.

52

Schmitt II, 249.

22

Brian Leftow

once be and not be . . . By necessity, then, every future thing is future . . . but this is “following” necessity, which does not compel anything to be.53

Anselm may not be invoking PNC at all, in a sense: “something cannot at once . . . ” includes ‘potest,’ and so the sense could be “something does not have the power to . . . ” If PNC really does play a role here, it need not be to explain the source of the necessity or what it consists in. It might be merely to make it obvious that this is necessary. Further, Anselm (so I have claimed) gave a metaphysical basis for PNC in Cur Deus, one he can fill out entirely in power-and-prevention terms. If he has done so, he need not repeat it. He can simply advert to PNC. So there is no clear case in De Concordia for a different account of following necessity, nor then for a non-powers basis for some form of necessity in Anselm’s work. I submit, then, that Anselm intends his power-based story to apply to what we would call absolute necessities, and that there is no clear basis in his text for any other account of these. I think his proposal is at least workable: I think it can extend to all absolute necessities. For Anselm’s God is omnipotent. He has power enough to make possible whatever is absolutely possible for creatures. So His powers can ground all absolute possibilities for creatures.54 If God’s power determines all absolute possibilities for creatures, and that they are all, it determines the absolute necessities about creatures, and so a “powers” account can apply to all creaturely absolute necessities. The necessities of His own nature and existence will arise simply because they are so and there never is a power for things to be otherwise.55 Of course, it is one thing to say that Anselm extends his story to what we call absolute necessities, and another to say that Anselm has anything like a clear concept of absolute necessity. Anselm just speaks of “necessity,” full stop. He does not speak of different strengths of necessity. He distinguishes different kinds—prior, following, proper, improper—but does not treat some as stronger than others. So some might contend that we should not ascribe to him modal concepts of any one strength—physical, absolute, etc. They might say that all we have evidence for is his having some (confused) necessity-concept for which the whole question of strength has not arisen. Still, Anselm in using ‘necessary’ without qualification or distinction of strengths may mean to express one concept strong enough to cover all cases that he thinks have any sort of necessity, even if he does not explicitly raise the question of strength at all. Absolute necessity can cover them all. We can 53

Schmitt II, 250. Also all unrealized possibilities for God and the realized ones other than God’s existing and having His nature. 55 This statement has a complex metaphysical back-story. See my “The Nature of Necessity,” and also God and Necessity, ch. 9. 54

Anselm on Necessity

23

obtain weaker necessities by restrictions on it: it is physically necessary that P just if it is absolutely necessary that if there are the actual natural laws, P. (Anselm knew the distinction between conditional and unconditional necessity from Boethius, so at least part of this move would be congenial to him. Perhaps he would prefer it in these terms: it is physically necessary that P just if it is absolutely necessary that if there are exactly the actual natural powers and they operate without supernatural interference, P.) So it seems to me not less plausible than the no-determinate-strength thesis that Anselm operates with a modal concept best construed as absolute. Further, it seems to me more plausible. If one thinks that necessity is just one thing, that it covers truths with what we call absolute necessity, and that it really is just obviously, wholly impossible that those propositions be false, surely one thinks that the one thing necessity is, is as strong as what we call absolute necessity. Thus my account. At the outset, I noted that most Anselm scholars have responded to Anselm’s mix of causal and alethic notions by limiting the application of the power-based story to physical, causal, or suchlike modalities, some treating his prior necessities as purely causal, physical, etc., and his following necessities as logical, alethic, or absolute. I have offered a positive case that prior and following necessity are absolute, and that Anselm’s power-based account applies to both. But the case is not complete until we examine what “limiters” can say for their view. I now examine the two most recent and most careful “limiter” accounts.

3 . LI M I T E D - S C O P E A C C O U N T S

3.1 Ad A. D. A. D. Smith contends that Natural. Anselm’s metaphysical story about power and prevention applies only to “natural” or “causal” modality, and so Tokens. Anselm’s tokens of ‘possible’ and ‘necessary’ often express only natural or causal modality;56 further, Conceive. Anselm uses talk of what is and is not conceivable to express absolute modalities.57

56 “Often” is the strongest claim Smith can make: he gives more than a solid page of citations of passages in which Anselm uses ‘possible’ and ‘necessary’ to express what appear to be absolute modalities (A. D. Smith, Anselm’s Other Argument, 69–70). 57 Ibid., 32–80.

24

Brian Leftow

Natural is just the basic “limiter” claim. Thus I have given my case against it. As far as I can see, Smith never says much in Natural’s favor. Perhaps he just sees no alternative to it. But there is an alternative. If Anselm’s power-based story can apply even to PNC, Natural has no presumption in its favor, and Anselm’s text speaks against it. Tokens is a consequence of Natural— something a limiter must say given Anselm’s text. If so, it does not require separate discussion (and Smith, I think, does not offer it much or any separate support). Conceive has been a popular thought independent of general debate over Anselm’s modalities: for instance, anyone who thinks that Anselm offers a modal argument for God’s existence in Proslogion 3 presupposes it. If Anselm’s power-based story is not his account of absolute modality, then either he has no account of it at all, or something else in him provides one. Given the historic tendency to tie conceivability and possibility and the prominence of conceivability in Anselm’s thought, that Anselm holds some sort of conceivability-based account of absolute modality is a natural conjecture, and so Conceive is a natural complement to Natural. I must discuss the case for Conceive. For if that case is strong, it casts doubt on my own account of Anselm. It would be very odd for someone whose metaphysical basis for modality was as I say to use talk of conceivability to express absolute modal facts. If Conceive is true, my reading of Anselm is very probably false. Natural and Conceive jointly imply that where the “power” metaphysics is in play, the modal facts involved are not absolute or “logical,” as Anselm does not use this metaphysics to undergird claims about what is conceivable. So their conjunction leads Smith to write that What Anselm means when he says that what does not exist is able not to exist . . . is far from expressing any triviality of modal logic. Indeed, (it) expresses a claim that is utterly perplexing from a merely logical point of view . . . that if something is able not to exist (i.e. it is possible for it not to exist), something else does exist! . . . something is required to possess a power so that this (improper) ascription of ability can be properly grounded.58

Smith reads ‘able’ in “able not to exist” as expressing the possession of power. So he thinks Anselm’s “power” metaphysics is in play. He thinks that where it is in play we are not dealing with logical modalities. So he denies that Anselm’s claim is a modal-logical thesis. As Smith sees it, “what does not exist is able not to exist” ascribes to non-existents a power not to exist. Non-existents have no powers. So ascribing a power to them is improper speech, and must be “cashed out” in terms of a power something else has. Thus Anselm’s claim is true only if, for whatever does not exist, something 58

Ibid., 77.

Anselm on Necessity

25

else has the power to bring it to be. For Smith, Anselm’s claim “expresses” this. I now make three points against Smith’s reading. One is quick: even if Anselm’s claim expresses something about only physical or natural modality, it might still be a modal-logical thesis. Any coherent modal concept has a modal logic. The move from “merely natural” to “not a logical thesis” is just too fast. But perhaps Smith really means “not a thesis of the logic of absolute modality.” My next is not as quick: pace Smith, in the strictest sense, the sentence “expresses” the claim it asserts. If token mental states are identical with token brain-states, the truth-condition for “I am thinking of a rose” is still that I am thinking of a rose, and that does not assert anything about brain-states—even if a brain-state makes my claim true. Truth-conditions need not specify ultimate truth-makers. In the same way, truth-conditions for need not assert the full metaphysical story of what makes this true. Still less need reasons to think the proposition true—since a simple impression of self-evidence is a good reason. I submit that what Anselm means to assert is precisely the modallogical thesis that if not-P then possibly not-P. Anselm is willing simply to drop it into his argument without discussion because he sees what it says as obviously, self-evidently true, even if the metaphysics beneath it is complex. That Anselm might see both difficult metaphysical and obvious conceptual reasons for the same truth should not surprise us, for Anselm elsewhere acknowledges more than one level of reason for the same truth. When he writes about the Incarnation, he states that there are deeper reasons for it than he can discover: but, he says, best at least to indicate that it is not without reason by giving some not-so-deep reasons for it.59 For Anselm there are the shallow reasons we can grasp, and the deep reasons only God knows. The shallow reasons let us make sense of the Incarnation, but they might or might not be reasons God acted on. Equally, then, there could well be shallow, self-evident conceptual considerations that let us see that a modal truth is true and a different deep metaphysical basis for it. Smith offers a metaphysics that might lie beneath “what does not exist, can not exist,” ascribes it to Anselm, then writes that “this is in total contrast to the status of the principle ‘if (not-) P then possibly (not-) P’ in modal logic and the modern philosophy it informs. Here the inference is immediate and self-evident.”60 But philosophers routinely treat as “immediate and 59

CDH II, 16. Smith, Anselm’s Other Argument, 115–16. Note, incidentally, that for Smith, Anselm would read this principle as (in effect) that if P, it is not “repugnant to thought” that P- which I take as “not a priori false that P.” This is certainly obvious and trivial. But it translates a modal principle into something entirely non-modal. 60

26

Brian Leftow

self-evident” inferences they also have complex metaphysical or epistemic reasons for deploying as they do. Systems of modal logic treat some inferences as axiomatic and so “immediate and self-evident.” Philosophers who accept the systems also treat the inferences this way. Yet these very philosophers may have complex metaphysical or epistemic reasons for taking the modal system to express the logic of a particular sort of modality, and so deploying the inferences in reasoning about that sort. Modal logics express the consequences of axioms in systematic form. Taking a truth as an axiom means precisely offering (within the logic itself) no justification for it. It is treating the truth as self-evident within the system. Further, modal logics in a sense only partially interpret their ‘◊’ and ‘□’ operators. For we operate with many different modal concepts, and modal logicians just as such are not concerned to map their systems on to any particular one of them. (Many modal systems probably map on to no concept we actually employ.) It takes informal argument for a philosopher to show that any logical thesis is part, or any logical system is all, of the logic of any ordinary or philosophical modal concept. And part of that is informal argument that the axioms of that logic are truths for the modal concept in question. The informal argument might well be metaphysical. If it is, it in effect provides a metaphysical basis for the truths of that logic. So if Anselm has a complex metaphysical rationale behind accepting some modal axiom, even one as simple as that P!◊P, that does nothing at all to show that he does not see it as a “logical triviality.” The metaphysical rationale shows why that logical truth is a truth of the logic of some kind of real modal fact, and thereby provides metaphysical grounding for the logical truth. Anselm’s having a metaphysical rationale behind simple logical truths lines his thinking up with that of any contemporary philosopher who adopts modal theses. My last point against Smith is that Anselm in principle has other ways to explicate possible non-existence, which do not require that anything have a power, or that anything exist. One is quick: recall Anselm’s claim that what is actual is possible because it is actual. This could apply to actual states of affairs, such as the world’s not existing. The world could be possibly nonexistent because actually non-existent, even if there were no cause able to make it, just as God is possibly existent because actually existent, even if there is no cause able to make Him. Another requires more work. Recall Anselm’s reasoning about the world: while non-existent, it has in itself no power to make itself, and so with respect to itself it is impossible that it exist. Anselm knows that ‘impossible to be’ and ‘necessary not to be’ are equivalent. So Anselm would accept that with respect to itself, the world necessarily does not exist. But what is necessarily so is possibly so. So for Anselm, with respect to itself, the world possibly does not exist. Anselm seems

Anselm on Necessity

27

committed to the claim that were there no power to cause it not to exist, it would be possible that the world not exist. Now for Anselm, the antecedent of this conditional is impossible: there necessarily is a God, who is necessarily able to destroy the world. This is a counterpossible, and so trivially true. But for Anselm it is not only trivially true. His metaphysics gives substantive reason for it. He could account for the possibility that the world not exist even if nothing at all existed: in that situation, the world’s intrinsic lack of power to exist would constitute a lack of primed power for it to exist, and so ground a necessity and so a possibility that it not exist. All this has concerned the non-existence of something contingent. Let’s now consider the non-existence of what necessarily does not exist: consider the claim that possibly no round square exists. Again, Anselm could just get it via actuality implying possibility. But we can ask a further question: has anything the power to bring it about that no round square exists? Here various thoughts clamor. Perhaps nothing has the primed power, because it is always too late: it is always already true that no round square exists. Perhaps nothing even has the intrinsic power: what would one do to make no round square exist? On the other hand, perhaps God has both powers. For Anselm, all necessity is due to God’s will, and so to His power.61 God prevents the past changing. It cannot change, but this is His doing. Perhaps then He also prevents round squares, though they cannot exist. But here is another thought. Nothing, not even God, has the power to cause a round square to exist, and for Anselm nothing can come to exist uncaused. So on Anselm’s terms, it is impossible that round squares exist, because lack of power prevents their existing. Just as we express the fact that nothing can cause God to lie as “necessarily, God does not lie,” we express the fact that nothing can cause a round square to exist as “necessarily, no round square exists.” So necessarily, no round square exists. Necessity entails possibility, and so possibly no round square exists. For Anselm, lack of power serves as a basis for necessity-claims. If so, it also grounds this sort of possibility-claim. And a lack of power does not require a subject. Nothing has the power to make a round square if something exists, but so too nothing has this power if nothing exists. So on Anselm’s terms, If nothing exists, possibly no round square exists is another substantively-as-well-as-trivially true counterpossible. Anselm allows “improper” speech. We can treat “possibly no round square exists” as an improper ascription of power to the non-existent: it 61

CDH II, 17.

28

Brian Leftow

really has none, but we can talk that way. Lack of power prevents and does not require a subject. (Not existing is one way to lack power.) So the metaphysical basis for this would be a subjectless lack of power. So Anselm could ground an improper ascription of possibility without requiring that anything possess a power, and perhaps even without requiring that anything exist. I suggest, then, that Smith’s account of Anselm’s modal claim is false. I now argue that Conceive is false. Once I have done so, I consider what follows from its falsity. 3.1.1 Against Conceive Smith thinks Anselm was committed to hold that 14. 15.

necessarily, if something is inconceivable, it is absolutely impossible, and necessarily, if something is conceivable, it is absolutely possible.62

Conceive requires this. If Anselm is not at least implicitly committed to (14) and (15), he is not even implicitly committed to the claim that inconceivability and absolute impossibility are necessarily coextensive. But if he uses ‘inconceivable’ to express ‘impossible,’ he is implicitly committed to their necessary coextension. Now there are many different ways of conceiving, and Smith does not say just which he has in mind. So the precise import of (14) and (15) is unclear. Smith provides a distinction: some things are inconceivable because merely beyond human thought. They are inconceivable by us. Others are inconceivable because they “conflict with thought” quite generally, not just with human thought. They in some way are in “conflict . . . with understanding,” “repugnant to what the intellect understands.”63 I find these phrases hard to interpret, and they are all Smith gives us. It is the latter, more absolute sort of inconceivability that for Smith expresses absolute impossibility.64 This makes his position a bit problematic, for there is reason to think that when Anselm speaks of being inconceivable, he means only inconceivable by us, or at any rate by finite minds. This comes in Proslogion 15. There Anselm argues that God is greater than can be conceived, using his standard verb for conceiving, cogitari.65 Anselm surely does not mean that God is greater than God can conceive. He means only that God is greater than we can conceive, or perhaps than any finite mind can—and as we see in the following, both the rhetoric and the 62

Smith, Anselm’s Other Argument, 33. Ibid., 35. They do, however, suppose some understanding: by understanding them, we see that they “conflict with thought.” 64 65 Ibid., 35. Schmitt I, 112. 63

Anselm on Necessity

29

argument of this passage make it very likely that here he means ‘conceive’ to express the same notion (whatever it is) it does elsewhere in the Proslogion. If this is correct, Anselm quite generally uses ‘inconceivable’ in the sense ‘inconceivable by us’ (or by finite minds) and it is unclear that he even operates with the more absolute notion of inconceivability that Smith ascribes to him. As to what that notion is, very tentatively, I suggest that a proposition is “inconceivable” in Smith’s more absolute sense just if it is a priori false and a state of affairs S is “inconceivable” in this sense just if it is a priori false that S obtains. If that is correct, then Smith means by (14) that necessarily, if something is a priori false, it is absolutely impossible. I doubt that Anselm held (14). As we see in the following, there is good reason to doubt it even if we are not sure what Anselm meant by “conceive.” Smith offers a series of arguments that Anselm held (14). I now consider and reject all of them—and as it happens, doing so does not require us to foist on Smith any particular interpretation of (14). After dealing with Smith, I argue that Anselm did not hold (14). If he did not, Conceive must be false. Smith begins, If Anselm’s claims about what is and is not conceivable had merely psychological import (without) logical or ontological consequences . . . Anselm would never infer anything about an object or a state of affairs or the truth-value of any proposition from the fact that it cannot be conceived to exist or to obtain or to be true. But he does repeatedly. Here is one of the more striking passages: “what does not exist is able not to exist, and what is able not to exist can be thought not to exist” . . . this yields . . . “what cannot be thought not to exist exists.” There are many other passages where Anselm infers falsity or non-existence from inconceivability.66

Perhaps there are, but here Anselm does not infer any such thing. He just states two conditionals, from which Smith, not Anselm, infers a third.67 Further, Smith’s argument rests on a false dichotomy. He takes it that Anselm’s ‘conceivable’ either reports on human psychology or expresses absolute possibility. But there is a third alternative: it might express epistemic possibility. What is conceivable, in this sense, is what is compatible with all we know. Used this way, ‘conceivable’ does not report a psychological fact. It reports an epistemic one. And this epistemic fact can have “logical or ontological” consequences. If it is not compatible with all we know that P, it follows that ¬P, an ontological consequence. And if it is not compatible with all we know that P & ¬Q, it follows that P  Q, a “logical” consequence. 66

Smith, Anselm’s Other Argument, 35, 57, 56. I stand with most Anselm scholars in denying that Proslogion 3 argues from to God’s existence. There is no such argument in the Reply either. So there is no reason to think Anselm would employ that particular instance of Smith’s third conditional. 67

30

Brian Leftow

Smith next points to Monologion 18, where Anselm states that it is inconceivable that truth begin or end, and also that it does not begin or end. Smith asks, “is it supposed to be a coincidence that what cannot be conceived is not the case?”68 Perhaps not, but it does not follow that Anselm makes the suppressed inference by way of (14). He might instead make it by 16.

necessarily, if something is inconceivable, it is not the case.

If ‘conceivable’ asserts epistemic possibility, being inconceivable is being epistemically impossible, i.e. incompatible with something known, and so (16) is really 17.

necessarily, if something is incompatible with something known, it is not the case.

Since what is known is true, whatever is incompatible with something known is incompatible with something true, and only falsehoods are incompatible with truths. So (17) is true, and obvious a priori. Read as I have suggested, (14) asserts that whatever is a priori false is absolutely impossible—a claim controversial since Kripke. So it would be charitable to ascribe (16) rather than (14) to Anselm if we could. If we knew that Anselm always meant by ‘conceivable’ something other than ‘epistemically possible,’ we could not ascribe (16) to him. But in fact, Anselm is clear that ‘conceivable’ has more than one sense.69 He is also clear that in one of them, we cannot both know that P and conceive that ¬P.70 If we cannot, knowing that P prevents also conceiving that ¬P; if conceiving it is prevented, then on Anselm’s terms conceiving it is impossible; so for Anselm, if it is known that P, it is inconceivable (in one sense) that ¬P.71 This comes out true and obvious if inconceivability is epistemic impossibility. So this is a charitable reading. It also accords with Proslogion 3, which seems to me to reason precisely that because (having read P2) we now know that God exists, we cannot (in this sense) conceive that He does not. Thus Anselm might well hold (16). So we have in many places the option of taking Anselm to make inferences via (16). Smith next notes that in Monologion 7, Anselm rules out there being anything other than the “supreme essence” and its creation on the grounds that “nothing . . . can be conceived to exist” beyond these 68

69 Smith, Anselm’s Other Argument, 35, 57. Proslogion 4. Reply 4. 71 Putting it in possible-worlds terms, we might introduce a restricted necessityoperator □k, which a proposition deserves just if it is true in all worlds in which it is known that P: then for Anselm, in this sense of “conceive,” if it is known that P, □k¬(it is conceived that ¬P), and so with “inconceivable” expressing both this sense of “conceive” and □k, if it is known that P, it is inconceivable that ¬P. 70

Anselm on Necessity

31

two. How could Anselm have been sure that there is nothing other than (these two) unless he believed it to be impossible?72

One can be sure that many things are false without believing them impossible, and the inference might rest on (16), not (14). Anselm thinks he has shown that the supreme essence created everything else. So he thinks knows this. If this is known, it is epistemically impossible that anything is not the creator or a creature. This provides a sense in which it is “inconceivable” that anything but these exist, and this, via (16), yields that only these exist. Further, Anselm can be as sure of that as he can that he has shown that the supreme essence created everything else. A case that an inference is by (14) must show this to be more likely or more charitable than alternative readings. Smith has not shown that the (14) option is more likely than the (16), and as (14) is controversial and (16) is not, the (16)-reading is more charitable. Smith continues, If inconceivability does not imply absolute impossibility, Anselm’s core notion of something than which a greater cannot be conceived would lack the power that he manifestly attributed to it. Can it really be supposed that Anselm thought that even if there is something a greater than which cannot be conceived, something greater than this is possible—in any sense? Indeed, if Anselm did not believe that it is simply impossible for there to be something greater . . . it is puzzling how he can regard it as self-evident . . . that something than which no greater can be conceived is, given that it exists, the greatest thing there actually is.73

I agree that for Anselm, being a greatest conceivable entails being a greatest possible. But this entailment does not have to move via a general tie between inconceivability and impossibility. It can be established simply from the content of Anselm’s phrase. Consider the property of being a greatest possible. We can conceive that something has it. This property is clearly more worth having than any candidate perfection incompatible with it. So if it conflicts with any other candidate perfection, it wins. Thus it must be ascribed to a thing than which no greater can be conceived. For if anything does not have it, we can conceive of something still greater, which does have it. So anything which does not have it is not a thing than which no greater can be conceived. Thus being a thing than which no greater can be conceived entails having no possible greater. The self-evidence can be explained in many ways. One is simply that the reasoning just given seems perfectly obvious. Here is another: if A exists and it is conceivable that something be greater, A is not a thing than which no greater can be thought, for something greater than A is conceivable. Hence if a thing than which no

72

Smith, Anselm’s Other Argument, 35, 57.

73

Ibid., 35, 57.

32

Brian Leftow

greater can be thought exists, it is not conceivable that there be a greater actual thing: and if ‘conceivable’ bears the epistemic sense above, this plus (16) yield that it is the greatest actual thing. One can use (16) as well as (14) to get the conclusion Anselm plainly accepts. Smith next points us to De Incarnatione Verbi 4. There Anselm tells us that “every composite thing can be decomposed either actually or conceivably, something that cannot be understood of simple things. For no mind can dissolve into parts that whose parts cannot be conceived.”74 Smith comments, “Anselm is . . . inferring from the fact that God cannot be conceived to be divided that it is absolutely impossible that God should be divided—since such an absolute impossibility manifestly attaches to anything that has no parts at all.”75 Smith takes Anselm’s argument to be this: 18. No simple thing conceivably has parts. 19. What does not conceivably have parts cannot be conceived as divided into parts. So 20. No simple thing can be conceived as divided into parts. 21. What cannot be conceived as divided into parts cannot be divided into parts. So 22. No simple thing can be divided into parts. I think something simpler is going on. It is obvious that a simple thing, as such, has no parts, and so cannot really (another legitimate reading of Anselm’s “actu”76) be divided. So Anselm does not argue that. All that needs arguing is that it is not conceivably divided, and this is all Anselm does argue in the quoted text: his reasoning stops at (20). A move from inconceivability to impossibility never comes in. And if Anselm’s “inconceivable” expresses being impossible, as Smith thinks, he would not infer impossibility from inconceivability. Asserting inconceivability would assert impossibility. So there would be nothing to infer. Again, in Monologion 3, Anselm infers from its being inconceivable that something not be through something that in fact, nothing is through nothing. Smith comments: Anselm certainly did not think it merely contingent . . . that nothing is through nothing. Even . . . God cannot, he asserts, bring it about that something should exist through nothing (M5). An absolute impossibility is here indicated by the impossibility of conceiving something.77 74 Ibid., 58, Smith’s translation modified in light of a suggestion he makes later in the paragraph. 75 76 Ibid., 58. Schmitt II, 17. 77 Smith, Anselm’s Other Argument, 58. Smith’s reference to Monologion 5 is incorrect. I do not know what text he meant to mention. As a result, I do not know whether he is really entitled to claim what he does about Anselm’s assertion of God’s inability.

Anselm on Necessity

33

I reply that Anselm’s inference could as easily be by (16) as by (14). Again, Smith points us to Monologion 32–33, where Anselm infers from its being inconceivable that the Supreme Spirit fail to utter a word that “reason . . . compels that word . . . by necessity to exist.”78 Smith reads this as moving from inconceivability to impossibility, but the Latin (“ratio tamen cogat verbum illud quo se dicit ex necessitate esse”) seems more naturally to express a necessity of consequence than one of consequent. And it would be odd of Anselm to infer the necessary existence of the Son when the Monologion nowhere argues the necessary existence of the Father. If the necessity is one of consequence, (16) is a suppressed premise. Finally, Smith refers to Cur Deus Homo I, 19, where Anselm asks Boso whether he can conceive that a human who sins and makes no satisfaction for it is equal to an angel who never sins. Boso replies that he can think and say the words, but cannot think their sense, just as he “cannot understand falsity to be truth.”79 Smith comments, “the suggestion is . . . that the words bear no coherent sense. There is simply no possibility, in any sense of the word, that they express. For Anselm the impossibility that inconceivability entails is as great as the impossibility of what is senseless being true.”80 But impossibility is simply nowhere in the exchange. Anselm asks Boso whether he can conceive something. Boso replies that in one sense he can and in another he cannot. Proslogion 4 distinguishes those senses. Anselm’s treatment there does not make an explicit connection to alethic impossibility, and a reading of Proslogion 4 in terms of epistemic impossibility is at least as plausible as one in terms of alethic. (More so, I think, in light of the reading of Proslogion 3 I suggested above.) Smith reads Boso’s reply as “that’s not just false, it’s impossible!” But it could also be “that’s not just false, but I can see a priori that it’s false!,” a reading which fits better with Boso’s further remark about falsity and truth. Cur Deus I, 19 goes on to infer a conclusion about what it is fitting for God to do. It does not take a premise about alethic impossibility to get this conclusion; epistemic impossibility will do just as well. Once we see the available alternative, Smith’s reading seems simply tendentious. In Proslogion 4, Anselm says that due to the way it is inconceivable that God not exist, if we try to assert that God does not exist, we wind up using words with an incorrect signification or no signification at all. Either we mean by the words something they do not in fact mean, or we use them without (a certain sort of) meaning. The latter might suggest that we do not manage to assert something. So it might seem that for Anselm, the one thing we cannot do by using “God does not exist” assertively is assert that God

78 80

Monologion 33, Schmitt I, 52. Smith, Anselm’s Other Argument, 61.

79

Schmitt II, 84.

34

Brian Leftow

does not exist. Some might continue in Smith’s direction by adding: what cannot even be asserted cannot be true. There is no possibility of any sort that it be true. It is false with the strongest necessity. Strongest necessity just is absolute necessity. So in Proslogion 4, inconceivability implies absolute impossibility. Smith may have this in mind in discussing Cur Deus I, 19. But lacking what Anselm calls signification is not the same thing as lacking meaning, not saying anything, or being unassertable. It is just not bringing anything in the world to mind—something like not referring.81 Suppose that there are no impossible worlds or states of affairs: possible sentences express possible states of affairs, and impossible ones express none. Then if I say “there is a round square,” I assert something, but either I bring no real state of affairs to mind, or I bring to mind some other state of affairs than there being a round square. This is a fairly plausible thought, and its ontological economy commends it. So it is a more than tenable claim that one can assert things without signifying anything (in Anselm’s sense). Further, it is a tenable claim that necessarily, all minds are finite. If this is how things are, there could well be truths too complex or large for any possible mind to grasp. These could not then be asserted, yet they would be true. So it seems a tenable claim that something that cannot be asserted is true. I claim, then, that whatever precisely Smith means by (14), Smith does not show that Anselm held (14). I now argue that Anselm did not hold (14), with ‘conceivable’ taken in the only sense that should concern us. Here is Proslogion 15: Lord, not only are you a thing than which no greater can be conceived (cogitari), but you are . . . greater than can be conceived (cogitari) . . . since it is possible to conceive something of this kind to be, if you are not that thing, it is possible to conceive something greater than you: which cannot happen.82

God, who exists, is greater than can be conceived. It follows that God’s greatness is actual but its full extent cannot be (fully, adequately) conceived.83 Anselm knows that whatever is actual is possible. On the face of it, then, Anselm is committed to the claim that something is inconceivable but possible. ‘Conceive’ has in Proslogion 15 the sense it has earlier in the Proslogion. For Proslogion 15 infers “greater than can be conceived” from the content of “thing than which no greater can be conceived” (plus other premises), as Anselm did with his earlier conclusions, and it seems clear 81 See Peter King, “Anselm’s Philosophy of Language,” in B. Leftow and B. Davies (eds.), The Cambridge Companion to Anselm (Cambridge: Cambridge University Press, 2004), 85–8. 82 Schmitt I, 112. 83 The first to note this was Gareth Matthews, “On Conceivability in Anselm and Malcolm,” Philosophical Review 70 (1961), 110–11.

Anselm on Necessity

35

from the rhetorical contrast Anselm makes—not just can no greater cogitari, but greater than can cogitari—that he means ‘conceive’ to bear the very sense in his conclusion that it bears in “thing than which no greater can be conceived.” So whatever Anselm means by ‘conceiving’—however we are to interpret his term—he asserts something incompatible with (14) taken in that sense. Smith’s response to the Proslogion 15 problem uses the distinction I noted above: some things are inconceivable because merely beyond human or finite thought, others in some more absolute way. Proslogion 15 (says Smith) deals only in the first sort of inconceivability, but (14) deals in the second. The distinction does not address my point about ‘conceive’s’ continuity of sense. Moreover, while Smith’s distinction may be a way to save the claim that Anselm held (14), if applying it here is not to be simply ad hoc, it requires a criterion for when we have the one or the other sort of inconceivability. Anselm never gives one.84 Nor does he explicitly deploy Smith’s distinction. As he does not, not only is there no basis for the claim that Anselm made whatever this distinction is, there is no textual support even for saying that Anselm’s usage of ‘conceivable’ conforms to it. At best, we might maintain that Anselm holds (14) despite Proslogion 15 by imposing Smith’s distinction on Anselm’s text—if we had independent reason to think that Anselm held (14). But we do not. As I see it, then, this is how matters stand. Proslogion 15 provides a strong prima facie reason to deny that Anselm held (14). One can overcome that reason by introducing a distinction with no basis in Anselm’s text. One can justify doing that only if there is some other textual reason to think that Anselm held (14). Smith fails to give us that sort of reason, inter alia because he has nothing to say against the more charitable and no less plausible claim that Anselm held and used (16). So on balance, it seems best to stay with the obvious reading of Proslogion 15, and so deny that Anselm held (14). If Anselm did not hold (14), it is unlikely that Anselm in fact used talk of inconceivability to express absolute impossibility. Conceive appears false. So my own account of Anselm seems safe on this flank. I now consider a second recent “limiter” account.

3.2 Williams, Visser, and Following Necessity I hold that Anselm’s modal locutions express absolute modality. Williams and Visser hold in effect that if they do, the necessity has a particular source 84 It will not do to submit as a criterion, that Anselm would say in the relevant sense that something is inconceivable “because and only because it is absolutely impossible” (Smith, Anselm’s Other Argument, 34). That would just define Smith’s interpretation of being inconceivable into correctness, with no explicit warrant from Anselm.

36

Brian Leftow

in cases of following necessity. For they think cases of following necessity are cases of logical entailment, and that following necessity thus just is logical necessity,85 by which I take them to mean a necessity rooted ultimately in concepts or propositions, not the extra-mental or -representational world. For them, then, the power-based story is not beneath Anselm’s modal locutions when they express absolute modality: it applies, again, only to physical etc. modality. If they can make a case for their “entailment” account, then, my claim that Anselm’s power-based account is his account of absolute modality is again put in question. Entailment relates only propositions or the like. I have not distinguished the relata of Anselm’s prior and following necessity. I do not think Anselm does so. I think that for Anselm, both prior and following necessity can relate whatever the relata are. Your speaking, for instance, can be prior-necessitated—it can be caused to occur—and Anselm says that it accounts for following necessities. Williams and Visser, however, think that for Anselm, sometimes, “causation . . . holds between concepts,”86 and that we have “subsequent necessity . . . when the causal relation holds between concepts”87 and prior necessity when it does not. Subsequent or following necessity is the sort they think purely logical. Part of their “limiter” account, then, is asserting that there is a special sort of relatum when the necessity involved is purely logical, a matter of entailment. If concepts really are the relata of following necessity, it seems fairly plausible that facts of following necessity are not based on powers, for it would be at best odd to think of concepts as having powers: concepts are some sort of abstract object. If there is no such special sort of relatum in cases of following necessity, conversely, then in these necessity is not a matter of entailment between representations of the world. It is instead (I submit) based in Anselm’s power-based account. To support their special-relatum claim, Williams and Visser translate a passage and comment: Consider the following excerpt from the Lambeth Fragments: If we say, “A human being is an animal,” the human being is a cause . . . that he himself is . . . and is called an animal. For by . . . ‘human being’ the whole human being is signified and conceived, and animal is a part of that whole. And so in this case the part follows from the whole in this way, because where the whole is, the part must also be. So because the whole human being is conceived in the name ‘human being’, the whole human being is himself a cause that he is . . . and is called an animal, since the conception of the whole is a cause that the part is conceived in it and is said of it.

85

Williams and Visser, Anselm, 160.

86

Ibid., 151.

87

Ibid., 150.

Anselm on Necessity

37

In this passage Anselm extends the use of causal language . . . to include what we would think of as relations of entailment. A human being is a cause that he is an animal. This is not to say that the human being produces an animal, but that the human being, just by being a human being, makes it the case both that he is an animal and that ‘animal’ is predicated of him. This sort of “making it the case” is in turn explained by an appeal to . . . the contents of concepts. That is, we can say that the human being brings it about that he is an animal because the concept human being includes the concept animal: “The conception of the whole is a cause that the part is conceived in it.”88

I do not see causation holding between concepts here. I sort this passage out differently. Anselm is explaining two things—that a human is an animal, and that a human is called an animal. So the passage contains two separate explanations, that the whole human causes being an animal and that the whole human causes being called one. The first explanation is this: by . . . ‘human being’ the whole . . . is signified and conceived, and animal is a part of that whole. And so in this case the part follows from the whole in this way, because where the whole is, the part must also be.

The passage starts with us saying “a human is an animal,” and so the point about signification and conception is there to get us from the word ‘human’ or concept , which are involved in what “we say,” to the real order, where the explanation takes place. The explanation is this: a human is a whole. The nature is part of this whole. The whole must include this part. So the human includes it. Because the human does, the human is an animal. In this sort of case, the whole’s presence is prior to and explains the part’s.89 So humans explain their being animals, i.e. the state of affairs constituted by their including as a part. The explanation is part/ whole, but in this sort of case, rather than the part causing the whole to be present, the whole—the concrete human—makes the part, , present, and so explains its presence. Anselm speaks of causation because he is dealing in explanation: that I am human explains it that I am an animal. The whole explanation is in the real order. It appeals only to humans and natures. It does not run by way of relations between concepts, let alone show us causation holding between concepts. The point about conception of the whole and conception of the part figures only in the second explanation. There is nothing about concepts in the first.

88

Ibid., 151. Anselm had no access to Aristotle’s metaphysics, but this sort of explanation would be standard stuff to medieval Aristotelians. 89

Brian Leftow

38

Anselm’s other explanation is of how a human comes to be called an animal. We see it in this: because the whole human . . . is conceived in the name ‘human being,’ the whole human . . . is himself a cause that he is . . . is called an animal, since the conception of the whole is a cause that the part is conceived in it and is said of it.

Here the explanation runs: humans cause us to conceive them as human. Humans have as part of their nature. So they cause us to conceive the concept of a human as including the concept of an animal. Causing us to conceive the whole concept of a human causes us to conceive the concept of an animal as part of it. (Anselm doubtless is assuming the classical definition of a human as a rational animal.) Because we so conceive humans, we call them animals. And so because humans cause us to conceive them as human, a human is a cause that he is called an animal. Here the causation runs from concrete humans to concepts to patterns of human speech. Nothing both causal and concept-to-concept is involved. Williams and Visser claim that following necessity “holds when there is a relation of causal compulsion between concepts.”90 I have suggested that their prime text for the claim that Anselm believes in such relations does not really show that. Now consider this passage again: There is . . . a following necessity, which a thing causes . . . when I say that you necessarily speak because you are speaking, it is following . . . for when I say this, I signify that nothing can bring it about while you are speaking that you are not speaking.91

In this text, your speaking necessitates. Your speaking is not a concept. It is an event or state of affairs. And it necessitates because of how it affects other real entities. That nothing can bring something about, because something else is coming about, is an objective causal fact. It has nothing to do with concepts. But it is a case of following necessity. I hold that Anselm’s modal locutions express absolute modality. Williams and Visser hold in effect that if they do, the necessity is a matter of logical entailment between representations of the world. They point further to texts where Anselm speaks of things as following by necessity (the Latin, (ex) necessitate sequitur, is closely related to Anselm’s term for following necessity, necessitas sequens) and of necessity in cases of entailment: Burning is not the cause of fire; rather fire is the cause of burning. And yet if we posit that there is burning, this causes it to follow that there is fire. For if there is burning, there must be fire.

90

Williams and Visser, Anselm, 158.

91

Schmitt II, 125.

Anselm on Necessity

39

From the unity of the Godhead it follows by necessity that the Father is eternal, the Son is eternal and the Holy Spirit is eternal.92

I agree that these are cases of following necessity. They do not involve an explanatorily or causally prior state of affairs producing what necessarily follows. The first uses “causes it to follow,” I think, because it involves explanation: that there is burning explains its following that there is fire, because if there is burning, there must be fire, as its proper subject. These texts can easily involve the account of following necessity Anselm actually gives in the text above, which is located entirely in the real order. Nothing can bring it about while there is burning that there is not fire, for burning is a proper accident of fire, an accident that can have only fire as its (immediate) subject. The proper-accident relation is why if there is burning, there must be fire, and the entailment from fire to burning, rather than having some distinctively conceptual source, records the real necessity this relation accounts for, which consists in a lack of power, an impossibility to bring about. Perhaps Anselm also has it in mind that positing the burning causes us to infer that there is fire. Again, given the unity of the Godhead, nothing can bring it about that the Persons are not co-eternal, and this real impossibility, this lack of power, explains its following that they are. I submit then that for Anselm, necessity is always in or grounded in the real order, and is the sort we call absolute. Anselm has a powers-based metaphysics of absolute necessity. It is all he needs; it can account for even the necessity of basic logical truths like PNC. There is no case for a non-powers source of necessity in Anselm. Oxford University

BIBLIOGRAPHY Anselm. S. Anselmi Cantuariensis archiepiscopi opera omnia, ed. F. S. Schmitt (Edinburgh: Thomas Nelson and Sons, 1940–61). Henry, D. P. The Logic of St. Anselm (Oxford: Oxford University Press, 1967). King, Peter. “Anselm’s Philosophy of Language,” in B. Leftow and B. Davis (eds.), The Cambridge Companion to Anselm (Cambridge: Cambridge University Press, 2004), 84–110. Leftow, Brian. “God and Abstract Entities,” Faith and Philosophy 7 (1990), 193–217. Leftow, Brian. “Omnipotence,” in T. Flint and M. Rea (eds.), The Oxford Handbook of Philosophical Theology (Oxford: Oxford University Press, 2009), 167–98. 92 De Casu 3 and De Processione Spiritus Sancti 1. The translation is theirs (Williams and Visser, Anselm, 160).

40

Brian Leftow

Leftow, Brian. God and Necessity (Oxford: Oxford University Press, 2012). Leftow, Brian. “The Nature of Necessity,” Res Philosophica (forthcoming). Matthews, Gareth. “On Conceivability in Anselm and Malcolm,” The Philosophical Review 70 (1961), 110–11. Rogers, Katherin. Anselm on Freedom (Oxford: Oxford University Press, 2008). Roques, René. Anselme de Cantorbéry. Pourquoi Dieu s’est fait homme (Paris: Cerf, 1963). Smith, A. D. Anselm’s Other Argument (Cambridge, MA: Harvard University Press, 2014). Williams, Thomas and Sandra Visser, Anselm (Oxford: Oxford University Press, 2009).

Avicenna on the Origination of the Human Soul Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

Ontological problems with Avicenna’s view on the nature of the human soul run deep.1 Here we shall focus on one aspect of his view, namely, the temporal origination of the human soul. According to the common wisdom, among both contemporary scholars and classic interpreters, Avicenna is committed to: (1)

The human soul is temporally originated with the human body.

We shall refer to (1) as the thesis of ‘Co-origination.’ In section 1, we will introduce Co-origination. In section 2, against the common wisdom, we will try

1 See, for some contemporary discussions, Herbert A. Davidson, Alfarabi, Avicenna, & Averroes on Intellect: Their Cosmologies, Theories of the Active Intellect, and Theories of Human Intellect (Oxford: Oxford University Press, 1992), 81; Thérèse-Anne Druart, “The Human Soul’s Individuation and its Survival after the Body’s Death: Avicenna on the Causal Relation between Body and Soul,” Arabic Sciences and Philosophy 10 (2000), 259–73, particularly 273; Robert E. Hall, “Intellect, Soul and Body in Ibn Sina: Systematic Synthesis and Development of the Aristotelian, Neoplatonic and Galenic Theories,” in J. McGinnis with the assistance of D. C. Reisman (eds.), Interpreting Avicenna: Science and Philosophy in Medieval Islam, Proceedings of the Second Conference of the Avicenna Study Group (Leiden: Brill, 2004), 62–86, particularly 70; Michael Marmura, “Some Questions regarding Avicenna’s Theory of the Temporal Origination of the Human Rational Soul,” Arabic Sciences and Philosophy 18 (2008), 121–38, in particular 121–2; and Dimitri Gutas, “Avicenna: The Metaphysics of the Rational Soul,” The Muslim World, Special Issue: The Ontology of the Soul in Medieval Arabic Thought 102 (2012), 417–25, particularly 417. Here, for instance, is how Hall summarizes the problem: Ibn Sina’s cosmology with its dual hierarchy of Intellects and Souls is perhaps ontologically too complex. [ . . . ] Still, there are unresolved difficulties in the ontological aspects of the ensoulment of the embryo, the intellectual ascent (in this life), the interrelations with the celestial souls, and—least clear of all—the ontology of the afterlife. The ontological problems run deep. (“Intellect, Soul and Body,” 70)

42

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

to argue that Co-origination is ambiguous and vague and thus its attribution to Avicenna is in need of clarification and precisification. Next, in section 3, we will highlight the significance of the problem by exploring the logical space of ‘possible’ solutions and introduce our methodology. The problem is broken down into two sub-problems: First, in section 4, we will consider the problem of the origination of different souls/powers, namely the vegetative, animal, and rational, in humans, and second, in section 5, we will discuss the problem of the relationship between these souls/powers. Based on our solutions to these two sub-problems, in section 6, we will offer our own reading of Co-origination according to which Avicenna is not committed to the view that the human soul is originated with the ‘human body’ in its ordinary sense. In section 7, we will try to derive two corollaries of our reading: First, we will attempt to show where an influential argument by Dag Hasse2 and Dimitri Gutas3 against a form of ‘supernal knowledge’ by the faculty of imagination (al-hayal) goes astray, and second, we will offer a ̆ solution to the notorious problem of the integrator and retentive factor of the fundamental elements of the embryo’s body, a matter of significant disagreement between ar-Raˉ zī and at. -T. ūsī. In section 8, we will briefly explain how the seemingly ‘contrary’ textual evidence may be handled. We will end, in section 9, by touching upon some open questions that our study engenders.

1 . C O - O R I GI N A T I O N There are pieces of ‘explicit’ textual evidence throughout Avicenna’s works of different periods that (apparently) support Co-origination. Here are two examples: Text 1. Thus, it is evident that human souls (al-anfus al-insaˉ niyya) are originated with the origination (h.udūtˍ) of human bodies (al-abdaˉ n al-insaˉ niyya).4 Hence, it is shown that when a human body, with a temperament specific to a human being, is 2 See Dag N. Hasse, Avicenna’s De Anima in the Latin West: The Formation of a Peripatetic Philosophy of the Soul, 1160–1300 (London: Warburg Institute, 2000), particularly 159. 3 See Dimitri Gutas, “Imagination and Transcendental Knowledge in Avicenna,” 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), 337–54. 4 ‘Being originated’ or ‘origination’ (h.udūtˍ), in medieval Arabic philosophy, has various uses, relative vs. absolute (something might be relatively originated, given a point of reference, without being absolutely originated) as well as temporal vs. essential (e.g., something might be essentially originated without being temporally originated). Here, not trying to distinguish between these uses, Avicenna presupposes that everything originated has a cause.

Avicenna on the Origination of the Human Soul

43

born, a human soul whose cause is the active intellect, since everything originated has a cause, is simultaneously born with.5

The above passage is striking in that Avicenna explicitly refers to the human body, not the material body, nor the body simpliciter, and to the human soul, not just the soul and is keen to mention ‘simultaneity’ between the origination of the human body and that of the human soul with a specific preposition in Arabic, i.e. ma ʿa (simultaneously with).6 Or, consider: Text 2. Thus it is correct that the soul is originated just as the body appropriate for its use is originated, to be employed by it [i.e. the soul] or to employ it, and the originated body will become its domain of governance and its tool. What the body naturally needs in terms of being occupied and employed, as well as being concerned with its states, may be found in the substance of the soul originated with the body.7

Here Avicenna seems to be engaged in explaining the close connection between the human soul originated with the ‘body appropriate for its use,’ on the one hand, and the human body as the ‘domain of the governance’ of its soul, on the other hand. The explanation unfolds in a reciprocal manner: the origination of an individual human soul is described by reference to the individual body that possesses the soul, as the appropriate object to be employed by the soul, and the origination of the corresponding individual human body is described by reference to the soul that ensouls the body, as the appropriate object to be occupied with the body.8 It is worth emphasizing that though Avicenna might have changed his view on some details of the ontology of the human soul, he adheres to two principles determinedly: (some version of ) Co-origination and the principle 5 Ibn Sīnaˉ , Avicenna. Al-Mabdaʾ wa-l-maʿād, The Provenance and Destination, ed. Abdallaˉ h Nūraˉ nī (Tehran: The Institute of Islamic Studies, 1363/1984), 108. In what follows, all translations are ours unless otherwise specified. 6 We use italicization ambiguously. In some occasions, we use it to generate stress effect, e.g., ‘human’ in the above sentence. In other occasions, we use it to produce a name of something, particularly when we refer to Arabic terms, e.g., ‘ma ʿa’ in the same sentence. Obviously, we use other means, e.g., single or double quotation, to produce names as well, e.g.,the last two occurrences of single quotation in this very footnote. We try to save double quotation for cases in which quotation of some kind is involved. We hope, perhaps naively, that context clarifies how we use italicization and quotation. 7 Ibn Sīnaˉ [Avicenna], Ah.waˉ l an-nafs [States of the Soul], ed. A. F. al-Ahwaˉ nī (Cairo: Daˉ r ih.yaˉ ʾ al-kutub al-ʿarabiyya, 1371/1952), 97. 8 The explanation is not viciously circular since the human soul and the human body are not ontologically interdependent; each has its own principle of existence, though the human body contributes to the individuation of the human soul and the human soul manages the human body. This kind of contribution to individuation does not necessitate ontological interdependence; an accident might contribute to the individuation of a substance when they have different principles of existence.

44

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

of ‘No-imprint,’ as we call it: “The rational soul, according to Avicenna, is a substance subsisting by itself which is imprinted neither in a human body nor in anything else corporeal; it is completely separable and abstracted from matter” (Gutas, “Metaphysics of the Rational Soul,” 417). Thus, Co-origination seems to constitute the heart of Avicenna’s ontology of the human soul. Contemporary Avicenna scholars, almost with no exception,9 interpret Avicenna as being committed to Co-origination. For instance, An individual rational human soul comes into existence the moment a human body, suitable for its reception and for becoming its instrument, is formed. It comes into existence as an emanation necessitated by the celestial intellects. (Marmura, “Some Questions,” 122)

Likewise, according to Druart, “In De anima, V, 4, Avicenna shows that though the human soul and its body originate simultaneously (ma ʿa), they do not corrupt simultaneously [ . . . ]” (Druart, “The Human Soul’s Individuation,” 262–3). Although Gutas does not closely follow Marmura’s interpretation of Avicenna’s philosophy of the soul, his view on the origination of the human soul is similar: “The rational soul comes into existence together with the human body, not before, and it maintains a certain association with it as long as a person is alive” (Gutas, “Metaphysics of the Rational Soul,” 417–18).10

2. THE PROBLEM: CO-ORIGINATION IS A M B I GU O U S A N D V A G U E Co-origination, (1), is ambiguous and vague and thus its attribution to Avicenna is in need of clarification and precisification, or so we will argue. Disambiguation and precisification of (1) has far-reaching consequences for Avicenna’s view on the origination of the human soul and engenders a host of open questions ranging from Avicenna’s philosophy of mind and 9 As far as we can see, the only exception is: Joseph Kenny, “Ibn Sīnaˉ and the Origin of Human Life,” in K. D. Crow (ed.), Islam, Cultural Transformation and the Reemergence of Falsafah: Studies Honoring Professor George Fransic McLean on His Eightieth Birthday (Tehran: Iranian Institute of Philosophy Press, 2009). Available online at: . We find Kenny’s interpretation of Avicenna’s ontology of the soul, however, problematic for other reasons. He concludes that “Ibn-Sina is the only medieval thinker I know of who, in so many words, maintains the humanization of the human fetus right from conception” (17). We disagree. See section 5. 10 All emphases on ‘human’ and ‘its’ in the above quotes are ours.

Avicenna on the Origination of the Human Soul

45

metaphysics to his moral philosophy (see section 9). Now, we shall argue for the ambiguity and vagueness of (1) in two steps.

a. Co-origination is Ambiguous (1) is ambiguous since the term ‘human body’ is ambiguous. In one sense, the term may refer to a body with all major human organs that occupy their natural location and function properly. In this sense, the ‘human body’ is the ‘body’ of a human individual in its ordinary sense. Alternatively, it may refer to a material body that does not have all the major human organs or whose organs do not occupy their natural location or function properly. In this sense, the ‘human body’ is not the ‘body’ of a human being in its ordinary sense. The ‘human body,’ as used in (1), is ambiguous between the ordinary sense of the term and its non-ordinary sense. (We will come back to the ambiguity embedded in the non-ordinary sense later.) We will explain why. Let’s begin with Avicenna’s view on the chronology of embryonic development in the fifth chapter of the ninth treatise of al-Ḥ ayawaˉ n of aš-Šifaˉ ʾ: Text 3. And the frothy/foamy period is six days. And the red lines and dots appear after another three days. That would be nine days after the beginning [i.e. the conception], and this may be predated a day or delayed for a day. And then after another 6 days, this would be the fifteenth after the conception, the hematic [element] penetrates throughout the whole, and it becomes a blood clot. And this may be predated a day or two or delayed two days. And by 12 days after that, it will become flesh. [Subsequently,] the segmentation of the flesh becomes distinct and the three [major] organs are set apart and then the moist [/membranes] surrounding the spinal cord spans, this may be predated or delayed two or three days. Then after nine days the head gets separated from the sides [/shoulders] and the limbs from the ribs and the abdomen, such a distinction is sensible [/perceivable] in some and concealed in others till it will be sensible [/perceivable] by 4 days after that, and this completes the forty days. This is rarely delayed to 45 days. And the earliest, in this regard, is thirty days.11

For Avicenna, the human body is completed over a period of time, between thirty and forty-five days. Otherwise put, the human body does not come into existence all at once. This is consistent with what we know from Muslim physicians’ consensus, as well as natural philosophers’, over

11 Ibn Sīnaˉ [Avicenna], Aš-Šifaˉ ʾ, at. -T. abī ʿiyyaˉ t, al-Ḥ ayawaˉ n [The Healing, Natural Sciences, The Animal], ed. A. Muntas.ir, S. Zaˉ yid, and A. Ismaˉ ʿīl (Cairo: al-Hayʾa almis.riyya al-ʿaˉ mma li-t-taʾlīf wa-n-našr, 1970), 172–3.

46

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

Table 2.1 Avicenna’s view on the chronology of embryonic development Stages

The length of each stage

The total time from conception

The frothy/foamy period Red lines and dots Blood clot Flesh and the three major organs Distinction of some organs Distinction of all organs

6 days 3 days  1 day 6 days  2 days 12 days  3 days 9 days 4 days

6 9 15 27 36 40  5 (min: 30 days)

epigenesis, the view that “the organs are not all formed at once, but one after the other.”12 Epigenesis is endorsed in the Quraˉ n as well.13 An almost identical chronology of embryonic development may be found in Avicenna’s al-Qaˉ nūn fī t. -t. ibb (569).14 Thus, he is committed to epigenesis. Table 2.1 summarizes his view on the chronology of embryonic development.

12 See Basim Musallam, “The Human Embryo in Arabic Scientific and Religious Thought,” in G. R. Dunstan (ed.), The Human Embryo: Aristotle and the Arabic and European Traditions (Exeter: University of Exeter Press, 1990), 32–48. Particularly, here: The belief that the fetus passes through different stages, that the organs are not all formed at once, but one after the other, is called epigenesis. On this issue Muslim religious thinkers had no quarrel with Aristotle, and both scientific traditions of medicine and natural philosophy emphasized epigenesis as the standard idea of fetal development. The Quran left no doubt that the fetus undergoes a series of transformations before becoming human. (38) 13 For instance, as Musallam (ibid.) refers to, consider the Quraˉ n (24, 12–14): (12) We created man of a quintessence of clay. (13) Then we placed him as semen in a firm receptacle. (14) Then we formed the semen into a blood-like clot; then we formed the clot into a lump of flesh; then we formed out of that lump bones and clothed the bones with flesh; then we made him another creation. So blessed be God the best creator. 14 Ibn Sina [Avicenna], Al-Qaˉ nūn fī t. -t. ibb [The Canon of Medicine]. F1953. Saab Medical Library, American University of Beirut, 2002–2007. A point of difference between what Avicenna says on the chronology of embryonic development in aš-Šifaˉ ʾ and that in al-Qaˉ nūn is his emphasis on the role of the faculty of form-bearing (al-mus.awwira) in the first stage of embryonic development in al-Qaˉ nūn: Text 4. And the frothy period is six days or seven, and during these days the form-bearing [faculty/power] (al-mus.awwira) acts on the semen (an-nut. fa) without the help of the uterus, and after that it needs its assistance [ . . . ] (Al-Qaˉ nūn fī t. -t. ibb, Book 3, Part 21, Treatise 1, chapter on genesis of the embryo, 569) Apparently, Avicenna is talking about the form-bearing faculty of the mother’s soul and its active role on the evolution of the semen after conception. This, nonetheless, should be verified elsewhere.

Avicenna on the Origination of the Human Soul

47

By the ‘three major organs,’ in the fourth stage, Avicenna means the heart, brain, and liver (ibid., 568–9). In the fifth stage, the organs that become distinct before all the rest are the head and the limbs. The human body, then, is completed about forty days after conception. Given (1), if by the ‘human body,’ we mean the human body in its complete form, i.e. in its ordinary sense, then, the human soul should come into existence at about forty days after conception. Some Avicenna scholars, e.g., McGinnis, might be read as arguing that, for Avicenna, the animal soul is originated with the perfected animal body, and thus the human soul is originated with the perfected human body: First, there is the initial substance, the semen (minan), which remains for a while, and then all at once a new substance appears, namely, the embryo (or blood clot, ʿalaqa). Similarly again, after a while the embryo is replaced by a new substance that comes to be all at once, in this case, the fetus (mud.g.a). This state is followed by the generation of the various organs and limbs. Finally, the perfected animal itself comes to be, which is yet a different substance.15

Jon McGinnis, “On the Moment of Substantial Change: A Vexed Question in the History of Ideas,” in J. McGinnis with the assistance of D. C. Reisman (eds.), Interpreting Avicenna: Science and Philosophy in Medieval Islam, Proceedings of the Second Conference of the Avicenna Study Group (Leiden: Brill, 2004), 42–61, at 54 (emphasis added). A minor terminological point may be worth mentioning. We use the ‘human embryo’ and the ‘human fetus’ in their contemporary senses: In humans, the term ‘embryo’ is currently “applied to the unborn child until the end of the seventh week following conception; from the eighth week the unborn child is called a fetus” (Encyclopedia Britannica, , retrieved Nov. 27, 2014). Medieval terminology, and in particular Avicenna’s, is different. In Avicenna’s language, ʿalaqa applies to the human embryo (as we use the term) when it is between the second and third week after conception (see Table 2.1), and not much further. Moreover, the term ʿalaqa is not semantically characterized to convey temporal information, though it refers to the human embryo between the second and third week. The word ʿalaqa is semantically characterized to convey information about the matter, namely blood, and the form (in a loose sense), namely a clot, that ‘constitutes’ the human embryo; ʿalaqa literally means ‘coagulated blood,’ see Hans Wehr, A Dictionary of Modern Written Arabic (3rd edition), ed. J. M. Cowan (New York: Spoken Language Services Inc., 1976). Hence, we shall prefer to transcribe/transliterate the term or, depending on the context, translate it by a description, e.g., ‘a blood clot.’ Regarding mod.g.a, one may have similar concerns. For Avicenna, the term refers to the human embryo (as we use the term) after the stage of being a blood clot and before the formation of the first major organ, i.e. the heart (al-Qaˉ nūn fī t. -t. ibb, 568) (we shall come back to this reference and discuss it in more detail later). This, according to Avicenna, should occur between the third and the fourth week (not after the seventh or eighth week, as we know). The word mod.g.a is semantically characterized to convey information about the matter, i.e. flesh, and the form (in a loose sense), namely a small chunk or a property of it, that “constitutes” the human embryo; mod.g.a literally means ‘a small chunk of meat’ or something ‘to be chewed’ (Wehr, Dictionary). Hence, we shall prefer to transcribe/transliterate the term or, depending on the context, translate it by a description, e.g., ‘a chunk of flesh.’ 15

48

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

McGinnis might be interpreted as suggesting that, first, embryonic development, as a biological process, involves some stages of substantial change, as a metaphysical event, and second that at the end of these substantial changes and after the generation of various organs and limbs, the perfected animal, as a different substance, is originated. What makes the perfected animal a different substance is its possession of the animal soul. Therefore, the animal soul and the perfected animal should be co-originated. Applying this analysis to the origination of the human body, it might be concluded that the human soul is originated with the perfected human body. Likewise, among the classic interpreters of Avicenna, Nas.īr-ad-Dīn at. T. ūsī has been read as arguing that the rational soul is co-originated with the completed human body:16 Text 5. [ . . . ] Thus the body gets completed and evolves to become capable of receiving a rational soul, or to bring about reason besides all the previous [acts]. And this soul will remain as the administrator in [/of] the body until the death [of the body] arrives.17

In fact, if one takes the ‘human body’ to be exclusively used in its ordinary sense, the above interpretation may be the most straightforward reading of Avicenna’s Texts 1 and 2, and all the like. The interpretation, however, does not go well with what Avicenna says elsewhere in the same book. In the 16th chapter of the 8th volume of aš-Šifaˉ ʾ, after considering different possibilities regarding the origination of the human soul inasmuch as it is sensitive or inasmuch as it is rational, Avicenna concludes as follows: Text 6. And whatever the case, as soon as the heart and the brain come into existence inside [the embryo] the rational soul is attached to it and from it the sensitive [power] follows. But the rational [power] is distinct and immaterial though not yet operative, it is like what it is in the drunk or the epileptic and shall acquire its perfection from something external [to it] that bestows intellect [/intellection upon it]. However, the other powers acquire their perfection by means of the body and the bodily factors. And if the newborn were first sensitive and then became a human being by [possessing] rationality, then by its perfection it would be transferred from one species to another species.18

16 See, for example, Seyed Hasan [Saadat] Mostafavi, “A Problem in Avicenna’s View on the Origination of the Soul and a Reply to It,” Sophia Perennis: The Journal of Sapiential Wisdom and Philosophy 2/4 (2011), 19–29. 17 See Nas.īr-ad-Dīn at. -T. ūsī, al-Išaˉ raˉ t wa-t-tanbīhaˉ t li-Abī ʿAlī ibn Sīnaˉ , ma ʿa Šarh. Nas.īr ad-Dīn at. -T. ūsī [/Šarh. al-Išaˉ raˉ t], ed. S. Dunyaˉ , 4 vols. (Cairo: Daˉ r al-maʿaˉ rif, 1971), 355. Emphasis added. 18 The Healing, Natural Sciences, The Animal Bk. 16 ch. 1, 403.

Avicenna on the Origination of the Human Soul

49

The first sentence of Text 6 describes the origination of the rational soul as early as the formation of the first major organ(s) of the body, namely the heart and/or the brain. The formation of these organs, according to Avicenna, occurs after the second week following conception (see Table 2.1) and long before the perfected human body. Turning to the last sentence of Text 6, and taking it at face value, it reads that the newborn does not become first an animal and then a human being. The middle parts of the quote explain that the sensitive power of the soul is something that follows from the rational soul even if the rational power of the soul is inactive. The passage seems to suggest that the origination of the human soul is at least as early as the origination of the first major organ(s) or at least as early as the activity of sensitive faculty of the human embryo. The idea that the human soul initially belongs to the heart, as the first major organ, reappears at the end of Avicenna’s De Anima: Text 7. Thus it should be the case that the soul’s first attachment is to the heart and it is not permissible [to say] that it is attached to the heart and then to the brain since when the soul is attached to the first organ, then the whole body becomes ensouled but [regarding] the second [organ being called the ‘second’ organ ensouled it is because] it [i.e. the soul] acts upon it [i.e. the second organ] via [acting upon] this first [organ]. Thus the soul animates the animal via the heart [ . . . ].19

One way to read Text 7 is that the first attachment of the soul is to the heart and thus as soon as the embryo’s heart is originated, it necessitates the origination of the soul to manage it. Assuming that the embryo’s heart cannot be managed by a soul other than the human soul, it follows that the human soul is originated as early as the origination of the embryo’s heart. Given epigenesis, it can be derived that the human soul is originated long before the origination of the completed human body. In Text 7, Avicenna also explains that calling the second organ ensouled the ‘second’ is not because the soul first, in a temporal sense, belongs to the heart and then, in a later instant, belongs to the second organ, e.g., the brain; rather, as soon as the first major organ is originated the soul is originated and the whole body of the human embryo is ensouled. The ‘second’ organ ensouled is called the ‘second’ because, metaphysically speaking, it is acted upon by the soul via the first organ. The above pieces of evidence show that Avicenna does not uniformly use the ‘human body’ (or the ‘animal body’) to refer to the completed or perfected human (or animal) body. Coming back to (1), the thesis reads 19 Ibn Sīnaˉ [Avicenna], [Aš-Šifaˉ ʾ, at. -T. abī ʿiyyaˉ t, an-Nafs] Avicenna’s De Anima (Arabic Text): Being the Psychological part of Kitaˉ b al-Shifa’, ed. F. Rahman (London: Oxford University Press, 1959), Bk. 5 ch. 8, p. 264.

50

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

that the human soul is temporally originated with the human body. The ‘human body,’ as used in (1), is ambiguous; it may refer to the perfected or completed human body (namely, the ‘human body’ in its ordinary sense) or to the first major organ or to the material body from which the perfected/ completed human body is formed (what we may call the ‘non-ordinary’ sense of the term).

b. Co-origination is Vague (1) also leads to a borderline case and, thus, is vague. By the ‘human soul’ we mean an individual human soul in its Avicennian sense. To wit, “a substance subsisting by itself which is imprinted neither in a human body nor in anything else corporeal; it is completely separable and abstracted from matter” (Gutas, “Metaphysics of the Rational Soul,” 417). Note that Avicenna explicitly argues that all substantial changes occur instantaneously or all at once (duf ʿatan).20 His argument for this latter statement, in The Physics of the Healing (II.3), goes like this:21 if substantial changes were not instantaneous, substances would undergo either intension/intensification (ištidaˉ d ) or remission/deterioration (tanaqqus.). However, substances are not susceptible to intension or remission (something is not more or less a substance). Therefore, substantial changes are instantaneous. The details of the argument aside, it indicates that for Avicenna substances come into existence instantaneously. Putting these two premises together, namely that the human soul is a self-subsisting substance and that substances come into existence instantaneously, it follows that the human soul comes into existence instantaneously. Given (1), that the human soul is temporally originated with the human body, it may be concluded that, the human body and the human soul come into existence simultaneously and instantaneously. Thus, Avicenna is committed to the view that there is an instant (in a metaphysical sense) in which the human body, either the first major organ or the material, from which the completed human body is formed or the completed human body itself, is originated. But which instant? This is a borderline case. It is true that the seminal fluid does not have a human body and that, in normal circumstances, the newborn has a human body. The instant in which the human body is originated, however, falls in the ‘gray area.’ If no such instant is determined, the application of the term ‘human For an analysis of Avicenna’s argument for the above claim, see McGinnis, “On the Moment of Substantial Change.” 21 Ibn Sīnaˉ [Avicenna], [Aš-Šifaˉ ʾ, at. -T. abī ʿiyyaˉ t, as-Samaˉ ʿ at. -t. abī ʿī] The Physics of the Healing, ed. and trans. Jon McGinnis (Provo, UT: Brigham Young University Press, 2009), 138–40. 20

Avicenna on the Origination of the Human Soul

51

body’ to the material that constitutes the embryo’s body does not have a definite truth-value. Recall, “[a] borderline case is a situation in which the application of a particular expression to a (name of a) particular object does not generate an expression with a definite truth-value.”22 Hence, the application of the term ‘human body’ leads to a borderline case. But a “borderline case, in the logical sense” is “a case that falls within the ‘gray area’ or ‘twilight zone’ associated with a vague concept.”23 Therefore, the term ‘human body’ should have a vague concept. Hence, (1) is vague.

3 . T H E S I G N I F IC A N C E O F T H E P R O B L E M : A M E T H OD O L OG I C A L N O T E The problem of ambiguity and vagueness of Co-origination is deep since both concepts involved, namely the human soul and the human body, call for explanation. To show the significance of the problem, we may explore the logical space of ‘possible’ solutions.24 Let us distinguish two dimensions of the problem: temporal and substantial. We will explain these dimensions. First, consider the following question on the temporal dimension of Co-origination: (Q1) Is the human soul temporally originated with (or after) the completion of the embryo’s body? Let’s call any view that answers (Q1) affirmatively a ‘late ensoulment’ (L) solution and any one that answers (Q1) negatively an ‘early ensoulment’ (E) solution. Co-origination requires one to characterize the material body ensouled by the human soul, as the ‘human body.’ Thus, if an E-solution fixes the origination of the human soul at the moment of conception, for example, the ‘human body’ in its non-ordinary sense will refer to the ‘fertilized egg’ or the ‘coagulated seminal fluid’ at conception. Or, if an E-solution fixes the origination of the human soul at some time between the moment of conception and the completion of the embryo’s body, say as early as the vegetative soul/power of the embryo is activated, then the ‘human body’ will refer to ‘a properly disposed humoral mixture’ that fulfills

22 Francis Jeffry Pelletier and István Berkeley, “Vagueness,” in R. Audi (ed.), The Cambridge Dictionary of Philosophy, 2nd edition (New York: Cambridge University Press, 1999), 945–7. 23 John Corcoran, “Borderline Case,” in Audi, Cambridge Dictionary, 96. 24 Single quotes around ‘possible’ are scare quotes; by ‘possible’ we mean apparently or epistemically possible.

52

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

the vegetative functions in question.25 Similarly, if an L-solution fixes the origination of the human soul at the moment of perfection of the embryo’s body in the mother’s womb, then the ‘human body’ will refer to the completed embryo’s body. Or, if an L-solution fixes the origination of the human soul at some later stage, say when the brain is fully developed and normally functioning, then even the completed embryo’s body in an earlier stage is not the human body. Thus what the ‘human body’ refers to calls for explanation. Second, consider the following question on the substantial complexity of the human soul: (Q2) Is the human soul, at its origination, ‘composed’ of some substantial forms? Let us call any view that answers (Q2) affirmatively a ‘composite soul’ (C) solution and any one that answers (Q2) negatively a ‘simple soul’ (S) solution. Historically speaking, a C-solution may resonate with Robert Kilwardby’s view, for instance, according to which “the human soul is a composite of three forms: vegetative, sensitive, and intellective,”26 and an S-solution may resonate with Aquinas’s view, for example, according to which “the rational soul is the body’s only soul and its only substantial form.”27 Note that our question is similar but considerably different from the problem of the ‘unicity or multiplicity of the substantial forms in a human being’ as discussed in the Latin tradition; for Avicenna, the human soul is not the form, simple or composite, of the human body; rather, it is a self-subsisting abstract object, the perfection of the human body. Table 2.2 The logical space of solutions

Early ensoulment

Late ensoulment

Composite soul

CE

CL

Simple soul

SE

SL

25 The potential sense of the ‘human body,’ i.e. a material body that is potentially a ‘human body’ (in its ordinary sense) or a body suited to being the body of an actual human being, is subsumed under the non-ordinary sense of the ‘human being’ (as the potential sense of the ‘tree,’ i.e. a material body that is potentially a tree, is subsumed under the non-ordinary sense of the term; a seed is a tree, only if ‘tree’ is used non-ordinarily). 26 José Filipe Silva, “Robert Kilwardby,” The Stanford Encyclopedia of Philosophy (Summer 2016 edition), ed. Edward Zalta, . 27 Robert Pasnau, Thomas Aquinas on Human Nature: A Philosophical Study of Summa theologiae Ia 75–89 (Cambridge: Cambridge University Press, 2002), 126.

Avicenna on the Origination of the Human Soul

53

By combining the above options, we may describe the logical space of ‘possible’ solutions as in Table 2.2. C-solutions, i.e. CE and CL, picture a layer-cake model of the human soul as being composed of some substantial forms (vegetative, animal, and rational soul). Accordingly, the vegetative, animal, and rational soul come to existence, most likely consecutively, and none is corrupted or degenerated. They all co-exist as three different components of an individual human soul. Such views can account for the plurality of different souls/powers managing the embryo’s body more straightforwardly but have a hard time explaining the unity of the human soul. S-solutions, i.e. SE and SL, picture a simplesubstance model of the human soul, as not being composed of other substantial forms.28 Accordingly, the vegetative and animal soul/power either are not generated as some substantial forms or, if generated as such, get corrupted (/degenerated) before the origination of the rational soul/power. Such views can account for the unity of the human soul more straightforwardly but have a hard time explaining the sense in which the vegetative and animal soul/power, not as substantial forms, manage the embryo’s body. L-solutions, i.e. CL and SL, well accord with the common wisdom of Muslim peripatetics and solve the problem of ambiguity of the ‘human body’ more naturally (they preserve the natural sense of the term). E-solutions, i.e. CE and SE, make ‘vagueness’ a distinctive problem for Co-origination, however. If the human soul is originated before the completion of the embryo’s body, then Avicenna uses the ‘human body’ in its non-ordinary sense. Most concepts expressed by ordinary language terms are both ambiguous and vague; the ‘human body’ is no exception. The ‘human body’ may be ambiguous between the non-ordinary and ordinary sense but vagueness is more of a problem if the term is used non-ordinarily. The ‘human body,’ in this usage, covers a range of dissimilar objects that are not ordinarily categorized under this concept. The issue is not only that there is a ‘gray area’ associated with a concept. In fact, the non-ordinary sense of the term is not a sense/concept in the first place; worse, it is not clear how many senses are at stake. Otherwise put, the ‘sense’ of the ‘human body,’ in its non-ordinary use, is lost. In this case, (partial) precisification is necessary for understanding the meaning of the term. Likewise, ‘hairy,’ in its ordinary use, is vague. However, if one uses ‘hairy’ non-ordinarily and applies it to relatively hairless people, then vagueness will become a distinctive problem and (partial) precisification will be necessary for understanding the meaning of the term.

28 The human soul may be complex, by having different faculties and powers, without being composite, namely being composed of some substantial forms.

54

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

Now we may briefly describe and justify our methodology. We break down the problem of ambiguity and vagueness of (1) into two sub-problems: First, the problem of the origination of different souls/powers, namely the vegetative, animal, and rational, in humans, and second, the problem of the relationship between these souls/powers. Based on our solution to these two sub-problems, we suggest a disambiguation and precisification of (1). We may justify this methodology as follows. The most significant characteristic of the material undergoing embryonic development is its possession of vegetative and animal soul/power. An E-solution may be vindicated if it turns out that the vegetative and animal soul/power in humans are originated pretty early and, moreover, they cannot be originated without the origination of the human soul. An L-solution may be vindicated if it turns out that the origination of the vegetative and animal soul/power in humans is dissociated from the origination of the human soul. On this account, only a completed body, in which the vegetative and animal soul/power are actualized, is “capable of receiving a rational soul” (see Text 5). Moreover, the relationship between the vegetative, animal, and rational soul/power is the key to the question of (substantial) complexity of the human soul. If the vegetative and animal soul/power, as they exist in humans, have their own substantial forms, then a C-solution is vindicated, otherwise, an S-solution. Thus, we propose that we discuss these two sub-problems in order: the origination of different souls/powers in humans, and the relationship between these souls/powers.29

29 Avicenna’s view on the soul contrasts, at least, with three main views: Platonic, Aristotelian, and Trans-categorical. A Platonic view, according to which the human soul pre-exists the human body, obviously clashes with Avicenna’s commitment to Coorigination. An Aristotelian view, according to which the human soul is the form of the human body, violates Avicenna’s commitment to No-imprint (roughly that the human soul is a self-subsisting abstract entity imprinted in no object). A ‘Trans-categorical’ view (as we call it), according to which the human soul is corporeal at its origination and incorporeal at its perfection, though it may vacuously satisfy Co-origination and Noimprint, contravenes Avicenna’s opposition to the idea of motion in substance. A Transcategorical view suggests that the human soul at its origination is identical with a material/ human body and then evolves via substantial motion to the human rational soul as an incorporeal object. Hence, the identity of the human soul is not to be explained by appeal to its Aristotelian substance category. This solution may have odd consequences; for instance, the human soul would be contingently corporeal, at its origination, and contingently incorporeal, at its perfection. A view along this line may be attributed to Mulla Sadra, The Wisdom of the Throne (al-Ḥ ikmaal- ʿAršiyya), trans. J. W. Morris (Princeton, NJ: Princeton University Press, 1981), 131–2: The human soul has many levels and stations, from the beginning of its generation to the end of its goal; and it has certain essential states and modes of being. At first, in its state of connection (with the body) it is a corporeal substance. Then it gradually becomes more and more intensified and develops through the different stages of its natural constitution

Avicenna on the Origination of the Human Soul

55

4. THE ORIGINATION OF DIFFERENT SOULS/POWERS Let us begin by clarifying Avicenna’s view on the origination of the vegetative soul/power in humans.

a. The Vegetative Soul/Power The vegetative soul/power has three major powers, the first one being the nutritive (al-g.ˉadiya) (Avicenna’s De Anima, 40). This is the power that is ˉ active throughout the individual’s lifetime: Text 8. The nutritive power, which is a power of the vegetative soul, acts throughout the person’s lifetime and as long as it exists [and] performs its actions [/functions], the plant and the animal will be staying alive and if it gets ineffective the plant and the animal will not be staying alive [ . . . ]30

That the nutritive power as a power of the vegetative soul is active during the person’s lifetime does not tell us when the nutritive power comes to activity. However, assuming that a power does not exist unless its principle exists, it may show that the vegetative soul and the nutritive power co-exist. We should note that Text 8 does not say that when the person is not alive, the nutritive power of the corresponding body does not exist, though it may be the case; rather, it only implies that if the person is alive, the nutritive power is active. Since the nutritive power seems to be necessary for the first major organ, we may suppose that this power exists at least as early as the origination of the first organ. Thus, the vegetative soul, as the source of this power, should also be originated as early as the first organ. Avicenna’s discussion in al-Ḥ ayawaˉ n confirms this reading: Text 9. Then the semen (al-manī) moves something else, namely the seminal fluid of the woman. Firstly, it moves toward forming the principle [/the source], thus a power that is the principle [/source] leading to the formation of other organs, in [a particular] order, will arise from the first organ, and then the coagulated seminal fluid comes to the possession of the soul in virtue of the penetration of the male power into it. So, the spirit seems to be born from the male seminal fluid and the

until it subsists by itself and moves from this world to the other world, and so returns to its Lord (89:27). Thus the soul is originated in a corporeal (state), but endures in a spiritual (state). 30 Avicenna’s De Anima Bk. 2 ch. 1, 52–3.

56

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

body from the female seminal fluid. Then when that is ensouled, the soul takes action through it to complete the organs. And this soul, at the time, is a nutritive [/vegetative] soul or [equivalently] has no other action [besides fulfilling nutritional tasks], though it has the potentiality for other things.31

Text 9 contains many interesting points; for now, we will focus on seven. First, Avicenna is explicit that the coagulated seminal fluid is ensouled or ‘comes to the possession of the soul’ even before it develops into a human body. Accordingly, some kind of ensoulment occurs with the origination of the first organ within the coagulated seminal fluid. Second, this ensoulment is “in virtue of the penetration of the male power into it.” Hence, the mother’s soul does not play any active role in this ensoulment, though it may contribute to the formation of the embryo’s body at later stages. Third, Avicenna apparently does not use ‘soul’ (an-nafs) and ‘spirit’ (ar-rūh.) interchangeably. Thus we may suppose that he distinguishes between the human soul and the human spirit, as he does elsewhere:32 The human soul is an incorporeal substance that ontologically depends on higher or abstract principles; the human spirit, in contrast, is a corporeal substance that initially emerges from the mixture of elements. Fourth, the male semen is responsible for the origination of the human spirit and the female seminal fluid for the origination of the human body, or more accurately for the origination of the material from which the complete human body is formed. Fifth, the soul originated at this stage is responsible for the formation of the organs of the human body and thus completing it by acting upon the coagulated seminal fluid. Sixth, this soul has one operative power, namely the nutritive power. Therefore, this soul should be a vegetative soul. Seventh, and most importantly, this vegetative soul has potentiality for other activities. Note that a vegetative soul of an animal, though it may actually be the same as the vegetative soul of a plant, in the sense of performing the same nutritional tasks, potentially differs from that since it has the potentiality of performing some acts that the vegetative soul of the plant lacks (we will come back to this point in section 5). To clarify the distinction between the human spirit and the human soul, consider: Text 10. This spirit is a kind of divine body, it relates to the semen and to the organs as the intellect relates to the powers of the soul. Thus, the intellect is the supreme incorporeal substance and the spirit is the supreme corporeal substance. This

31

The Healing, Natural Sciences, The Animal Bk. 16 ch. 1, 401–2. See Ibn Sīnaˉ [Avicenna], al-Adwiya al-qalbiyya [Remedies for the Heart], in Min Mu ʾallafaˉ t Ibn Sīnaˉ at. -T. ibbiyya, ed. M. Z. al-Baˉ baˉ (Aleppo: Maʿhad at-turaˉtˍ al-ʿilmī alIslamī, 1404/1984), 221–94, particularly 226. 32

Avicenna on the Origination of the Human Soul

57

substance is inseparable from the semen as long as it is [kept] sound and proper in the womb [ . . . ]33

The ‘human spirit,’ in the above use, refers to the supreme corporeal substance. It is not separable from the sound semen. This characterization of the human spirit is repeated elsewhere. In al-Qaˉ nūn fī t. -t. ibb (The Canon of Medicine), where he distinguishes between three kinds of spirit, a similar characterization of ‘spirit’ may be found: Text 11. And when the uterus contains the semen (al-manī), the first state that occurs there is the foaming of the semen, and that is an act of the form-bearing faculty/power. And the truth regarding that state of foaming is [that it results from] the stimulation [initiated] by the form-bearing faculty/power when there is [something] from the human [spirit],34 natural [spirit] and animal spirit in the semen [and it runs each] toward its own principal location35 to reside there. And from that each organ is created in accordance with the way we have explicated and clarified [ . . . ]36

Hence, the utmost attention should be given to the distinction between different kinds of the soul, namely the vegetative, animal, and rational, on the one hand, and different kinds of the spirit, namely the natural, animal, and human, on the other hand.37 To sum up, in humans, the nutritive power, and thus the vegetative soul, is originated much earlier than the formation of all major organs and the vegetative soul, with the mediation of the spirit, contributes to the process of completing the human body.

b. The Animal Soul/Power The animal soul is characterized by two main sets of power: motive and perceptive.38 The latter, in turn is divided into two categories: internal and external, depending on the object of perception (or sense).39 Sensation or sensitive power is restricted to animals and humans. The first sensation in virtue of which an organism becomes an animal is the sense of touch, one of 33 The Healing, Natural Sciences, The Animal Bk. 16 ch. 1, 404. Our translation of the first two sentences follows Kenny, “Ibn Sīnaˉ and the Origin of Human Life,” 11. 34 Note that the order of consecutive adjectives in Arabic is different from that in English. 35 In the original text, Avicenna uses the word ma ʿdin. In modern Arabic ma ʿdin means mineral or metal (Al-Mawrid: A Modern Arabic English Dictionary, ed. R. Baʿalbaki (Beirut: Dar el-ilm lilmalayin, 1995)). Here, however, we suspect Avicenna has the Persian use of the word in mind, in which it means center, principle, or principal location (Dehkhoda Lexicon, available at: ). 36 The Canon of Medicine Bk. 3, pt. 21, tr. 1, 568. 37 We expect some variation in Avicenna’s terminology as well. 38 Ibn Sīnaˉ [Avicenna], [An-Najaˉ t, at. -T. abī ʿiyyaˉ t, an-Nafs] Avicenna’s Psychology: An English Translation of Kitaˉ b al-Najaˉ t, Book II, Chapter VI, ed. and trans. F. Rahman (London: Oxford University Press, 1952), 25–6; Avicenna’s De Anima 41. 39 Avicenna’s Psychology 25–6.

58

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

the external senses.40 Again assuming that a power of the animal soul does not exist unless its principle exists, the origination of the animal soul in humans can be traced back to the origination of sensation: Text 12. Thus, in summary, the semen is the foamy [/extract of the]41 substance, and because of this, Venus42 is called ‘Aphrodite’43 [from the Greek word ‘aphros’ meaning foam],44 since it is the origin of lust and the source of semen production.45 And coldness does not freeze the semen, as long as it is semen.46 And when the nutritive power in the seminal fluid is prepared to receive the acts set for the sensitive soul [namely, specific to it], then the power to receive the soul inasmuch as it is sensitive will come about therein, though in the ones [i.e. the organisms] that possess rationality (an-nut. q), the sensitive [power] and the nature (at. -t. abī ʿa)47 are one and the same, and this is because the organs for sensation and the ones for rationality are completed concurrently, and this is not the case for the nutritive [power] and 40

Avicenna’s De Anima 67. The Arabic term may be read in two ways: zabadī and zubdī, the first means foamy and the second extract or something related to it. 42 Again the Arabic term may be read in two ways: az-zahra and az-zuhra, the first means the flower (‘the’ shall reflect the definite article ‘al’ at the beginning, transcribed as ‘az’) and the second means the planet Venus. Pace Joseph Kenny, “Ibn Sīnaˉ and the Origin of Human Life,” 7, we find the first reading absolutely wrong. 43 Like above, the Arabic term has two readings: zabadiyya and zubdiyya, each being a relative adjective of the corresponding term, namely zabadī and zubdī. The two meanings are interrelated but distinct. Translation of the names in this context may raise significant questions regarding their semantics. Addressing such concerns, however, goes beyond our purposes here and thus we shall put them aside. 44 See , retrieved May 29, 2015: Aphrodite, ancient Greek goddess of sexual love and beauty, identified with Venus by the Romans. The Greek word aphros means “foam,” and Hesiod relates in his Theogony that Aphrodite was born from the white foam produced by the severed genitals of Uranus (Heaven), after his son Cronus threw them into the sea. 45 Here is the translation of the first two sentences in Kenny, “Ibn Sīnaˉ and the Origin of Human Life,” 7: In summary, semen is naturally foamy. That is why a flower is called foamy, because it is a source of pleasure and a source of semen production. 46 Elsewhere Avicenna reports that Venus bestows upon human souls the ‘predisposition’ or ‘preparedness’ (isti ʿdaˉ d) for reproduction: Text 13. But Venus, from which a power is bestowed upon corporeal bodies that confers coldness and concurrence and [from which another power is bestowed] upon souls [that confers] the predisposition for reproduction power [ . . . ] (Fī l-Ajraˉ m al- ʿulwiyya [Treatise on the Supernal Bodies], in Tis ʿRasaˉ ʾil fī l-h.ikma wa-t. -t. abī ʿiyyaˉ t, Treatise 2 (Cairo: Mat. baʿa hindiyya, 1326/1908), 59). 47 We suspect that ‘nature’ here means neither substance nor essence. Regarding this use of the term, one may consult the first book of The Physics of the Healing, under the titles On defining nature and On nature’s relation to matter, form, and motion (McGinnis, “On the Moment of Substantial Change,” 37–50). We will come back to this point shortly. 41

Avicenna on the Origination of the Human Soul

59

its organs. Also, sensation does not generalize to all organs of the animal but attaining [/receiving] nourishment does. And it is not unlikely that the seminal fluid is in this state when the nutritive power therein is acquired from the father and then another [nutritive power] is coming afterwards. It may also be permissible that the nutritive soul coming from the father stays until the temperament transforms, via some kind of transformation, [into another] and then a particular nutritive soul is attached to it, the one [i.e. the nutritive soul or power] acquired from the father does not have the power to completely manage [the embryo] to the end, and it’s only adequate for some [stages of] managing, [and] needs a powerful principle. As though the one got from the father is just changed from what is necessary for it, thus it does not belong to the kind of absolute (al-mut. laqa) nutritive [power] that was in the father (al-abb), [nor does it belong to] that [which] will be in the child, nonetheless this change does not drive that [power] out from performing the appropriate act required for action.48 And whatever the case, as soon as the heart and the brain come into existence inside [the embryo] the rational soul is attached to it and from it the sensitive [power] follows.49

The above text is pretty complex and carries several key points. Avicenna begins by explaining the origination of the power (qūwa) to receive the soul inasmuch as it is sensitive. There are two points here: first, he talks about the power to receive the soul, not the reception of the soul or the soul itself. Note that Avicenna does not introduce the soul as the form of the body. It is essentially distinct from the matter. The body, however, can receive the soul if it has the adequate humoral temperament. So, there are some stages of preparation before the origination of the human soul. Otherwise put, the human soul is not originated at the moment of conception (thus, Kenny’s interpretation (“Ibn Sīnaˉ and the Origin of Human Life”) is false, as we shortly explain). Second, what Avicenna means by the ‘sensitive soul’ can be determined by what he says in the same sentence, namely the soul inasmuch as it is sensitive. Recalling that sensation is a distinctive feature of animals, including humans, the sensitive soul should be a power of the animal soul that is responsible for sensation.

48 This sentence in Arabic is pretty hard to read, translate, and interpret. We have read the concatenation of ‘k,’ ‘ ʾa’ or ‘aˉ ,’ and ‘n’ as ka ʾanna (not kaˉ na), meaning as though. So, we do not take the term to be a verb. Accordingly, and for literary reasons, we have read the qad before tag.ayyara as qad of taqrīb (approximation), not of tah.qīq (realization). Hence, we have added ‘just’ before the first occurrence of ‘changed’ in our translation. Moreover, after al-abb we have supposed, added in brackets above, that the previous negation is repeated. Finally, we have, perhaps with no convincing justification, tried to fill in the references of the pronouns in brackets to give a full sense to the sentence. 49 The Healing, Natural Sciences, The Animal Bk. 16 ch. 1, 402–3.

60

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

Next, we may turn to the point that in humans, the sensitive power and the nature are one and the same since the organs for sensation and the organs for rationality are formed in parallel. ‘Nature’ has many different uses in Avicenna’s philosophy and he discusses and distinguishes them on different occasions for different purposes. Let us first review a summary of the three main uses of the term in The Physics of the Healing: Text 14. Still, the term nature might be used in many ways, three of which we shall mention [as] deserving that title the most. So [(1)] nature is said of the principle, which we mentioned;50 [(2)] nature is said of that by which the substance of anything subsists; and [(3)] nature is said of the very essence51 of anything.52

It is not clear in what sense Avicenna uses the term ‘nature’ in Text 12. Kenny reads him as arguing for ‘hominization at conception’: Ibn-Sīnaˉ is the only medieval thinker I know of who, in so many words, maintains the hominization of the human fetus right from conception. (Kenny, “Ibn Sīnaˉ and the Origin of Human Life,” 17)

This reading, however, encounters various problems. We will discuss two. First, as far as Text 12 is concerned, Avicenna emphasizes, as another possibility (note “it is not unlikely”), that “the nutritive soul coming from the father stays until the temperament transforms [ . . . ].” The fact that Avicenna raises this possibility should give us pause in ascribing hominization at conception to Avicenna. According to this possibility, the temperament of the embryo should undergo a transformation to obtain its own nutritive power. So, the nutritive power coming from the father would stay for a while after conception. Thus, for this period, the human embryo does not possess its own nutritive power. But if the embryo does not possess its own nutritive power, then it does not possess its own vegetative soul/power either. In fact, in the subsequent sentences, Avicenna attempts to fill in the details of this possibility by considering a nutritive power that is neither of the kind that is in the father nor of that which is in the child. At the end of Text 12, he tries to explicitly refer to the point at which he undoubtedly takes

50

We will come back to this meaning shortly. McGinnis uses the term ‘being’ instead of ‘essence’ here. The Arabic term is dˉat. Admittedly, there are occasions in which dˉat should be translated as being. However,ˉ in ˉ this context, we find essence a more appropriate translation. For one thing, if something does not exist, it may still, in this sense of ‘nature,’ have a nature or essence, but it may be odd to say that ‘nature’ is applicable to the very being of that, given that it is non-existent. Alternatively, one may translate dˉat to ‘itself ’ or ‘the thing itself.’ If so, the last sentence of Text 14 should be rendered intoˉ “[(3)] nature is said of the very thing itself, of anything.” 52 The Physics of the Healing Bk. 1 ch. 6, 47. Emphasis added. 51

Avicenna on the Origination of the Human Soul

61

the human embryo to be ensouled: “whatever the case, as soon as the heart and the brain come into existence inside [the embryo] the rational soul is attached to it and from it the sensitive [power] follows.” This suggests that Avicenna withholds belief in ‘hominization of the human fetus right from conception.’ Second, in The Physics of the Healing, when Avicenna is trying to argue against the possibility of motion in substance, he describes the reception of the form of life by the human embryo as occurring when the bones, the nervous system, and the veins are getting completed, which in turn is in the third and the fourth week after conception (see Table 2.1): Text 15. Still, on the basis of observing semen gradually developing into an animal and the seed gradually into a plant, it is imagined that there is a motion here [namely, with respect to substance]. What should be known is that, up to the point that the semen develops into an animal, it happens to undergo a number of other developments [/genesis]53 between which there are continuous qualitative and quantitative alterations; and so, all the while, the semen is gradually undergoing alteration. In other words, it is still semen until it reaches the point where it is divested of its seminal form and becomes an embryo. Its condition [remains] like that until it is altered [into] a fetus,54 after which there are bones, a nervous system, veins, and other things that we do not perceive, [remaining] like that until it receives the form of life. Then, in like fashion, it alters and changes until it is viable and there is parturition. Someone superficially observing the transformation imagines that this is a single process from one substantial form to another and therefore supposes that there is a motion with respect to the substance, when that is not the case and, instead there are numerous motions and rests.55

Avicenna explains the reception of ‘the form of life’ as occurring after some genesis and changes in which ‘bones, a nervous system, veins and other things that we do not perceive’ are formed. This provides strong evidence that the idea of “hominization of the human fetus right from conception” cannot be attributed to Avicenna, if by ‘hominization’ we mean reception of the form of life.

53 Here we should like to slightly depart from McGinnis’s translation. The Arabic expression for ‘other developments’ is takawwunaˉ tun uhraˉ . We find the content of ̆ ‘development,’ in one sense, broader than that of takawwun. If Avicenna uses the latter in the technical sense of ‘generation’ as in ‘generation and corruption’ (or ‘generation and degeneration’), then ‘development’ is not the best word. We suggest ‘genesis’ (or ‘transformation’) instead; we hope that this translation represents the difference between kawn and takawwun. It is noteworthy that we do not find any Arabic term corresponding to McGinnis’s term ‘transformation’ used in the last sentence of Text 15. 54 Pace McGinnis, we do not translate ʿalaqa to ‘embryo’(or mod.g.a to ‘fetus’). Please see note 15. 55 The Physics of the Healing Bk. 2 ch. 3, 141.

62

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

Coming back to Avicenna’s use of ‘nature’ in Text 12, we propose to interpret ‘nature’ along the following lines: Text 16. (1) Every body has a nature, form, matter, and accidents. Its nature, again, is the power that gives rise to its producing motion and change, which are from [the body] itself, as well as its being at rest and stable. Its form is its essence by which it is what it is, while its matter is the thing bearing its essence. Accidents are those things that, when [the body’s] form shapes its matter and completes its specific nature, either necessarily belong to it as concomitants or accidentally belong to it from some external agent.56

In the above quote, the first meaning of ‘nature,’ referred to in Text 14, is unpacked. Avicenna is explicit in this discussion that only “in the case of the simples [that is, the elements], the nature is the very form itself.” Therefore, in humans, ‘nature’ should not be identified with the form. Given that ‘nature’ is used in accordance with Text 16, what Avicenna says in Text 12 about the identity of the sensitive power and the nature means that the soul inasmuch as it is sensitive is the power that gives rise to the production of motion and change in the human embryo, and the organs responsible for motion and change are completed concurrently with the organs responsible for cognition and intellection, though they may become active at different times. To sum up, since the organs responsible for motion and change are originated before the completion of the human body, it follows that the animal soul/power is so too.

c. The Rational Soul/Power The situation of the rational soul, however, is not as clear. A passage like Text 15 might be read as suggesting that the rational soul is originated with the origination of the perfected human body. In contrast, a passage like Text 12 describes the origination of the rational soul as early as the origination of the first major organ, i.e. the heart, and hence long before the completion of the human body. Trying to avoid a premature solution to the problem, we shall keep both possibilities alive at this stage. It is, however, worth repeating that Avicenna does not consider the human soul as the form of the human body imprinted therein, whereas he takes the vegetative soul as it exists in a plant to be the form imprinted in the body of the plant. The same story goes with the animal soul of non-human animals. The immediate question, then, is what is the relationship between the vegetative and animal soul/power in humans, on the one hand, and the rational soul/power, on the other hand? 56

The Physics of the Healing Bk. 1 ch. 6, 45.

Avicenna on the Origination of the Human Soul

63

5 . T H E RE LA T I O N S H I P B E T W E E N D IF F E R E N T S O U L S / P O W ER S Recall that according to C-solutions (composite soul), Avicenna is committed to the view that the human soul is composed of some substantial forms. These forms should be the vegetative, animal, and rational soul, as they exist in an individual human soul. If so, it follows that the vegetative and animal soul exist in one and the same sense, namely as substantial forms, in plants and non-human animals, on the one hand, and in humans, on the other hand. However, if it is shown that this is not the case, namely the vegetative and animal soul/power exist in an individual human soul not as substantial forms, then C-solutions should be avoided. Thus digging into the relation between different souls/powers is crucial for our purposes. Let us begin by looking at the first book of Avicenna’s De Anima where he tries to argue that “the soul is a substance” and in particular “a form [that is] not in a subject” (Avicenna’s De Anima Bk. 1 ch. 3, 27). After arguing that the subsistence of the proximate subject of the soul is due to the soul (ibid., 28), he turns to the following objection and rebuttal: Text 17. But one might say: We indeed accept that the vegetative soul is the form of that [i.e. its matter] and thus it is a cause of the subsistence [/constitution] of its proximate matter; however, the animal soul seems to be such that the vegetative soul [first] constitutes its matter and then the animal soul shall follow this and constitute it [i.e. the matter already constituted by the vegetative soul], thus the animal soul [/animality] is factual (mutih.as..silatun) in a matter that is [already] subsisted by itself and that is the cause of the subsistence of what is occurring to it, namely the animal soul [/animality], therefore there is no animal soul [/animality] except [as being something that is] subsistent in a subject. In reply to that, we say: The vegetative soul inasmuch as it is a vegetative soul does not necessitate anything except a nutritive body in an absolute sense, and the vegetative soul in the absolute sense has no existence except as the existence of a generic meaning (ma ʿnaˉ jinsī), and that is only in the [faculty of] estimation; and what exist among concrete particulars (al-a ʿyaˉ n) are [different] species of it. And what should be said [goes like this]: The cause of the vegetative soul [in this absolute sense] is also something general ( ʿaˉ mm) and universal (kullī), not factual, and that is the growing nutritive body in its absolute generic sense, [which is] not speciesdetermining. But the principle of the body with the means [/tools] for sensation, discrimination, and voluntary movement, is not the vegetative soul inasmuch as it is a vegetative soul, rather this [i.e. the vegetative soul inasmuch as it is a vegetative soul] becomes a distinct nature in virtue of a distinct differentia being attached to it, and that will not be, except by becoming an animal soul.57 57

Avicenna’s De Anima Bk. 1 ch. 3, 29–30.

64

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

The objection Avicenna addresses here may be reformulated as follows: one may accept that the vegetative soul is ‘a form not in a subject’ but reject that the animal soul is like that because the animal soul only occurs when the nutritive power is operative and this implies that the vegetative soul is originated (given that the origination of the vegetative soul is a necessary condition of the operation of its power). Thus, the animal soul occurs to (/is realized in) a matter that already subsists by itself in virtue of the origination of the vegetative soul. Therefore the animal soul is ‘a form in a subject.’ Avicenna’s reply includes five main points: First, ‘the vegetative soul’ in its absolute sense, namely, the vegetative soul inasmuch as it is a vegetative soul, requires a nutritive body in its absolute sense, not any (specific or) particular nutritive body. Second, the vegetative soul, in this sense, exists as a generic meaning, it has no existence as a particular object or individual or a species. The vegetative soul in its absolute sense is not part of the ‘external’ world; it exists nowhere as an external fact, it only exists in our faculty of estimation. Third, if the vegetative soul in its absolute sense only necessitates a body in its general sense and only exists in estimation, then it determines no species. Fourth, the principle of the ‘body with the means [/tools] for sensation, discrimination, and voluntary movement,’ is not the vegetative soul in its absolute sense, rather it is the animal soul in its specific sense [as a species]. Fifth, the vegetative soul inasmuch as it is a vegetative soul does not contribute to the subsistence of the matter constituting the body of the animal. In other words, the animal soul does not occur to (/is not realized in) an already subsisting matter because the ‘vegetative soul’ in its absolute sense fails to contribute to the subsistence of any matter whatsoever. So far Avicenna has identified the generic meaning of the ‘vegetative soul’ or its absolute sense. Accordingly, the ‘vegetative soul’ is like a theoretical term for an ideal object, e.g., ‘ideal gas’ and ‘frictionless surface,’ a term that does not refer to something in the external world but is necessary for our understanding of it and only exists in theory, in Avicenna’s language “in estimation.”58 Then, he moves on to explain some other senses of the term: Text 18. Nonetheless, it is necessary to begin [once more] and add further explanation here. Thus we say: By the ‘vegetative soul’ either [1] we mean59 the specific soul [as a species]60 that is specific to a plant, not an animal, or [2] we mean a general

58 We do not intend to imply that ‘existence in theory,’ in our language, has the same content as ‘existence in estimation’ in Avicenna’s language. 59 The original text has this, and the following two cases numbered (2) and (3) above, in a passive form, namely ‘it is meant.’ 60 In the literature, ‘specific’ is used as a translation of both hˉa.s.s and an-nawʿiyya/an-naw ʿī ̆ (see, for instance, McGinnis, “On the Moment of Substantial Change,” 74–5, for the first use and McGinnis, ibid., 76, 180, for the second one). To distinguish these, we add ‘as a species’

Avicenna on the Origination of the Human Soul

65

meaning that generalizes over the vegetative [in the first sense] and the animal soul so far as it attains nourishment, reproduces and grows, this is sometimes called the ‘vegetative soul,’ but this is a metaphorical [way of] speech. The vegetative soul does not exist but in the plant; however, the meaning that generalizes over the vegetative and the animal soul exists in animals as well as in plants and its existence is like the existence of a general sense in objects. Or, [3] by this [i.e. the ‘vegetative soul’] we mean a power/faculty of the powers/faculties of the animal soul that the acts of attaining nourishment, cultivation, and reproduction are issued by it. If [by the ‘vegetative soul’] we mean the vegetative soul that in relation to the soul effective in nourishment is a species, that only exists in plants and not in any other thing, and [particularly] not in animals. And if [by the ‘vegetative soul’] we mean the general meaning [of the term], then it is necessary to attribute a61 general meaning to it, not a specific meaning; thus a general craftsman[as a genus] is the one to whom a general craftwork[as a genus] is attributed, and a specific craftsman [as a species], like a carpenter, is the one to whom a specific craftwork [as a species] is attributed and a determinate (al-mu ʿayyan) craftsman is the one to whom a determinate craftwork is attributed. And this is something the enquiry of which has [already] been presented to you, and what is attributed to the general vegetative soul, vis-à-vis bodily affairs, is [also being] generally growing, it is growing inasmuch as it may suit the reception of sensation or it may not suit [it], and that [i.e. any of these two] is not attributed to the vegetative soul inasmuch as it is general, nor this meaning follows from it. But [if by the ‘vegetative soul’ we mean] the third case, then it would be impossible, according to what is suspected [by the critic], for the vegetative power/ faculty [of the animal soul] to come [into existence] alone and render the body animal-like and [it is so, in turn, since] if that power/faculty were alone in managing [the body] it would complete the body as a plant, and this is not true, rather it completes the body as an animal, with the means [/tools] for sensation and motion. Thus it should be a power of the soul which has other powers and this power, among others, conduct oneself with the acts that paradigmatically contribute to the tool’s preparedness for [the realization of] the secondary perfections of the soul with this power, and this soul is the animal soul, and it will be clarified afterwards that the soul is unique and these powers in the organs stem from it.62

In this long quote, Avicenna distinguishes three meanings of the ‘vegetative soul.’63 In the first sense, the ‘vegetative soul’ refers to the specific soul (as a or ‘the species’ parenthetically, depending on the structure of the sentence and context, whenever the second use is intended. 61 Bearing in mind that there is no syntactic device to indicate stress effect in Avicenna’s text, we hope that our emphasis here is consistent with the content of the text and adequately represents it. 62 Avicenna’s De Anima, 30–1. Our reading, and particularly our parsing of this passage, is different from Fazlur Rahman’s. 63 Avicenna, here, seems to be relying on The Physics of the Healing Bk. 1 ch. 12, in which he discusses similar distinctions vis-à-vis causes and effects (McGinnis, “On the Moment of Substantial Change,” 74–80, particularly 75–6). Therein Avicenna apparently

66

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

species) of a plant. He explicates this sense in the second book of Avicenna’s De Anima in more detail. The palm soul is a vegetative soul and “in its vegetative-ness is palm-ish”;64 it is not the case that it is a vegetative soul and then something else should be added to it to become the palm soul, or palmish. The palm vegetative soul, in its vegetative-ness realizes all the powers required for the palm tree. Otherwise put, the vegetative soul of the palm is not a genus in need of a differentia to become a natural kind; it is itself the specific soul (i.e. the species) of the palm and specifically exists therein. The same story goes with every species of plant, e.g., the grape has a grape soul that in its vegetative-ness is grape-ish.65 In this sense, “the vegetative soul does not exist but in the plant” (Text 18). In the second sense, introduced in Text 17 as the “absolute sense,” the ‘vegetative soul’ refers to a genus, a meaning that generalizes over plants and animals. This is the generic sense in which the vegetative soul does not exist among concrete particulars (al-aʿyaˉ n) and determines no species. The vegetative soul, in this sense, may also be associated with the body. Since the vegetative soul is responsible for growth, the ‘growing body’ attributed to the general vegetative soul should also be understood in a generic sense: this ‘body’ is open to reception of sensation, in the case of animals, and is not so, in the case of plants. Such a generic body is a general meaning that does not exist among concrete particulars; the actual bodies are not indifferent to reception of sensation. In the third sense, the ‘vegetative soul’ refers to a power of the animal soul, a power that is responsible for the acts in common between animals and plants, namely “attaining nourishment, cultivation, and reproduction.” The ‘vegetative soul’ as referring to this power exists among concrete particulars and is the ground to attribute the ‘vegetative soul’ in the second sense to animals, i.e. the absolute or generic meaning. The key point is that the vegetative soul, in this sense, does not come into existence alone to manage the body; rather, it manages the body as a power of a unique animal soul that has other powers as well. Avicenna’s argument for this claim is short: if the vegetative soul as a power of the animal soul were alone in managing the body, the body would be completed as a plant. However, the body of an animal is not completed like that; instead it is completed as an animal. Hence, other powers of the animal soul also manage the body, though they may become operative at different stages and at different times. draws on Aristotle’s discussion of modes of causation in Physics 2.3. 195b25ff in The Complete Works of Aristotle: The Revised Oxford Translation (Sixth Printing, with Corrections), ed. J. Barnes (Princeton, NJ: Princeton University Press, 1995), 739–40. 64 Avicenna’s De Anima, 48–9, 56–7. 65 We will come back to this passage and discuss it further shortly.

Avicenna on the Origination of the Human Soul

67

To conclude, there are three different senses of the ‘vegetative soul’: as a species, as a genus, and as power of another soul. In the first sense, it only exists in plants as species. In the second sense, it only exists in our estimation, as an abstraction that generalizes over plants, animals, and humans. In the third sense, it exists in animals as a power of their souls.

6 . C O - O R I G I N AT I O N P A R T I AL L Y D I S A M B I G U A T E D AN D P R E CI S I F I E D The conclusion of section 4 reads that the vegetative and animal soul/power (in humans) are originated much earlier than the completion of the ‘human body’ in its ordinary sense. We now have different senses of the ‘vegetative’ and ‘animal’ soul/power handy (section 5). Thus, back to the problem of the ambiguity and vagueness of Co-origination, we may clarify in what sense the ‘vegetative soul’ is originated before the formation of all major organs in humans. The vegetative soul that the coagulated seminal fluid comes to possess and which arises from the first organ and leads to the formation of other organs (Text 9) should be something that exists among concrete particulars and thus cannot be the ‘vegetative soul’ in its second use, the generic sense, since that only exists as a meaning in estimation. It is noteworthy that the ‘vegetative soul’ in its second sense may be attributed to humans in the sense that a genus may be attributed to particulars falling under a species under that genus. Thus being an animal is truly attributed to every individual human. But being an animal, in this sense, does not exist among concrete particulars. It only has the existence of a general meaning in objects. The vegetative soul that the coagulated seminal fluid comes to possess cannot be the ‘vegetative soul’ in its first meaning, i.e. as a species, since that only exists in plants, not in animals and, a fortiori, not in humans. Therefore, it should be the ‘vegetative soul’ in its third sense, namely as a power of a unique soul that has other powers as well, that comes into existence before the formation of all major organs or the complete human body. In this sense, the ‘vegetative soul’ that exists in a plant is different in species from the ‘vegetative soul’ that exists in a human: Text 19. [ . . . ] And that is because the vegetative soul that exists in the palm [tree], as a species, clearly does not contribute to the power of growth [/augmentation] (an-naˉ miyya) existing in a human, and that power is not such as to be qualified to be associated with the animal soul at all, nor is the power of growth existing in an animal qualified to be associated with the palm soul. Nonetheless, a single meaning [/sense] embraces both and that is [that] each one of them attains nourishment, grows and

68

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

reproduces, though each is differentiated [from the other] by a species-determining differentia that is constitutive [to it], not just by an accident. And the meaning that entirely exists in both is the genus of the vegetative power which belongs to the human and that is distinct [from both] in the manner in which a generic meaning is distinct [from what falls under it]. And we do not deny that the genus of these powers may be found in other things [as well] and nothing [from what we have said] implies that these powers cannot be conjoined and belong to a unique soul of a human [ . . . ]66

We read Avicenna as explaining that the vegetative soul existing in a plant, as something that exists in a concrete particular, is different from the vegetative soul existing in a human (or in an animal). The sense in which the ‘vegetative soul’ is applicable to both plants and humans (or animals) is the sense in which a genus, as a common description, is applicable to its species. We take Avicenna’s last sentence to point toward his own view on the unity of the human soul: the vegetative soul that exists in a human, namely the vegetative power, belongs to a unique soul that possesses different powers and faculties. So, a human soul, as a self-subsisting substance, has different powers, among them the vegetative power, which may be called the ‘vegetative soul.’ The ‘vegetative soul,’ in this sense, as Avicenna explains shortly afterwards, is predicable of the ‘human soul’ since it is an instantiation of its power; there is no human soul unless it is capable of attaining nourishment, reproduction, and growth (thus it has nutritive, generative, and augmentative power). This vegetative soul/power cannot exist independently of the human soul. The idea that the vegetative soul or power of an animal is different from the vegetative soul as a genus applicable to both plants and animals is discussed earlier in Avicenna’s De Anima in more detail: Text 20. The vegetative power that is in the animal reproduces an animal body and that is because it is a vegetative [power] to which an animal power belongs, and this [i.e. the animal power] is the differentia that has, among other things, what contributes to its possession of the nutritive power and growth and thus [it] mixes the basic [ingredients] and elements to a temperament appropriate for the animal. And its temperament does not manage a power common (muštaraka) to the plant and animal inasmuch as it is common; therefore, inasmuch as it is common, it does not necessitate any specific temperament. Rather, it necessitates a specific temperament in it [i.e. in the animal] because in addition to being a nutritive [power] it is an animal [power] that in its nature is to sense and move if it obtains the [necessary] means [/tools], and that [power] by its very reality is a preserver of that aggregate and the temperament [in some form of preservation], and when this [i.e. the harmony] is brought to the things that are aggregated it should be by force because it is not in the 66

Avicenna’s De Anima Bk. 5 ch. 7, 259–60.

Avicenna on the Origination of the Human Soul

69

nature of opposing bodies and elements to get aggregated in virtue of their essences, rather by nature they are prone to move toward different directions and this is a specific soul that aggregates them. For example, in the palm, [this specific soul] is a palm soul and in the grape, is a grape soul, and in general it is the soul that is the form of that matter. And when the soul becomes palm-ish, there is something more and above the very reality of growth to it, and that is the palm soul and [likewise] in the grape, [that is] the grape soul, and the palm is not in need of a soul in addition to its vegetative soul to become a palm, though it [i.e. its vegetative soul] has no acts besides the acts of the plant, in fact its vegetative soul in its vegetative-ness is palmish. But the vegetative soul that is in the animal exceeds the creation of the animal toward the acts that are distinct from the vegetative acts, and this manager [/administrator] is the animal soul, and in fact it is not the same as the vegetative soul, except that it is said that it is the vegetative soul in the meaning that we discussed, namely the general [meaning of the term].67

So what exists, among concrete particulars, as a vegetative soul or power in an animal or in a human is entirely different from the ‘vegetative soul’ in its specific sense (as a species) that only exists in a plant. The vegetative soul, as a power of a unique animal soul, has a different causal power than the ‘vegetative soul’ in its specific sense: the former manages the body to evolve and become the body of an animal whereas the latter manages the body to evolve and become the body of a plant. Likewise, the ‘animal soul’ may have three different meanings, as a species, as a genus, and as a power of another soul. We may use it to refer to (i) the specific soul (as a species) that is specific to each animal: in this sense the animal soul only exists in animals and every species of animal has its own soul. Or, by the ‘animal soul,’ we may mean (ii) a general meaning that generalizes over all animals, including humans: the ‘animal soul’ in this generic sense is a meaning that applies to what the ‘animal soul’ in the first sense refers to. This general meaning can be expressed by “what contributes to motion and/or perception as the common characteristics of animals.” The ‘animal soul’ in the generic sense exists in the faculty of estimation. Or we may use the ‘animal soul’ to refer to (iii) a power of the rational soul that is responsible for the acts of motion and perception. In this sense the ‘animal soul’ is a power of the human soul which has other powers such as reasoning. At this point, we may disregard some ‘possible’ solutions in Table 2.2. Note that a substantial form contributes to the subsistence of its matter. Therefore, if the vegetative soul, as it exists in animals (humans included), were a substantial form, it would “contribute to the subsistence of the matter constituting the body of the animal.” However, Texts 17–20 suggest otherwise. Hence, the vegetative soul, as it exists in animals, is not a substantial 67

Avicenna’s De Anima Bk. 2 ch. 1, 56–7.

70

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

form. Therefore, the prospect for C-solutions (composite soul: the human soul is composed of vegetative, animal, and rational soul as substantial forms) is dim. More particularly, versions of CL-solutions (a Composite soul originated Late) according to which the embryo first possesses a vegetative soul, as a substantial form, and then comes to the possession of the animal soul, as a distinct substantial form, and finally receives the rational soul, as its perfection, when none of these souls gets corrupted in between, jeopardize the sense in which they are substantial (as each subsequent soul should come into existence in a subject already in-formed by the previous soul). Of S-solutions, versions of SL-solutions (a Simple soul originated Late) might be developed in two main forms: Either the management of the embryo’s body at the earlier stages, i.e. before the origination of the human soul, is attributed to no souls, of any kind, or to some souls, in some sense. Both options are untenable. The former ignores numerous pieces of textual evidence (e.g., Texts 8, 9, and 12) in which Avicenna explains the development of the embryo’s body as being managed by the vegetative and animal soul/power. The latter, namely taking some souls to be responsible for the early embryonic development, might posit them either as some substantial forms or as powers belonging to a different substance. Again, both alternatives are unacceptable. The first one needs to further assume that through successive ensoulment, at each new stage, the substantial form originated at the previous stage gets corrupted (otherwise this solution collapses to a version of C-solutions). This implies that the embryo would be transferred from one species to another (from a plant to an animal and from an animal to a human), a transition that Avicenna implicitly denies: “And if the newborn were first sensitive and then became a human being by [possessing] rationality, then by its perfection it would be transferred from one species to another species” (Text 6). The second alternative postulates a third substance, neither the human body nor the human soul, which has specific powers, i.e. the vegetative and animal, to manage the embryo’s body before the late origination of the human soul. This solution finds no textual support and violates Avicenna’s soul–body dualism (we will come back to this issue in section 7, b). If C-solutions (in both CE and CL versions) and SL-solutions are not promising, as the above observations suggest, an SEsolution (a Simple soul originated Early) should be favored. Having these points in mind, let us return to the problem of ambiguity of Co-origination. We may now say that the ‘vegetative soul’ as a power of a unique soul is originated before the formation of the ‘human body’ in its ordinary sense. This power, however, cannot exist without the existence of the soul itself. But the soul that belongs to a human, as it exists among concrete particulars, is the specific soul (as a species), i.e. the human soul.

Avicenna on the Origination of the Human Soul

71

Therefore, the human soul is originated before the formation of the ‘human body’ in its ordinary sense. More particularly, “a power that is the principle [/source] leading to the formation of other organs, in [a particular] order, will arise from the first organ, and then the coagulated seminal fluid comes to the possession of the [vegetative] soul in virtue of the penetration of the male power into it” (Text 9). The same argument shows that the existence of the ‘animal soul’ as a power of a unique soul necessitates the existence of the human soul before the formation of the ‘human body,’ again in its ordinary sense: “as soon as the heart and the brain come into existence inside [the embryo] the rational soul is attached to it and from it the sensitive [power] follows” (Text 12). Therefore, since the vegetative and the animal power of the human soul are originated as early as the first major organ and the sensitive power ‘follows’ from the rational soul, then the rational soul, in pure potentiality, is originated as early as the first major organ. So, the nonordinary sense of the ‘human body,’ namely the material from which the completed human body is formed (after undergoing some stages of substantial change), should be the intended sense of the term, as it is used in Co-origination. Note that according to this reading, the vegetative soul/power, as it exists in an individual human soul, is firstly a power of the human soul and secondly distinct from the vegetative spirit. The vegetative soul, in this sense, is the realization of a power of an abstract entity, namely the human soul, and that abstract entity has other powers, e.g., the animal and rational, that may be inactive at the stage in which the vegetative soul is activated. How, for Avicenna, an abstract entity can have some inactive power is a different issue. Avicenna is anyway committed to such a view since the human intellect at its origination has some inactive powers and only acquires perfection through objects external to itself, e.g., the active intellect or the souls of celestial spheres. The vegetative spirit as it exists in a human, in contrast, is a corporeal entity. It can move and circulate in the human body, or in the material constituting the body, and assist the soul in the formation of other organs. None of these two, i.e. the vegetative soul and the vegetative spirit, should be conflated with the vegetative soul as it exists in a plant, which in fact is the realization of the plant form as a species. The vegetative soul of the plant is corporeal too but it is distinct from the vegetative spirit as it exists in a human since first, it is the form of the plant whereas the vegetative spirit is not the form of the human body, and second, it does not require administration by an abstract entity, whereas the vegetative spirit requires that. Finally, consider the ‘problem’ of vagueness of Co-origination. We explained that ‘vagueness’ may be a distinctive problem, if the ‘human body’ is used in its non-ordinary sense. Moreover, above, we argued that

72

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

this is the case. Thus to establish a meaning for the ‘human body,’ we need to first precisify it. As Avicenna argues, the human soul is a substance and substances come into existence instantaneously (as they do not accept intension (ištidaˉ d) or remission (tanaqqus.)). Therefore, in whatever sense the ‘human body’ and the human soul are co-originated, they are originated instantaneously.68 The above disambiguation of Co-origination also suggests a precisification of it: The instant in which the human soul comes into existence is the instant in which its vegetative power is originated. This, however, does not fully determine the instant in question because on some occasions Avicenna is indeterminate about this instant and on some other occasions he talks about it indirectly by referring to the origination of the first organ. As an example of the former case, consider the middle part of Text 12: And it is not unlikely that the seminal fluid is in this state when the nutritive power therein is acquired from the father and then another [nutritive power] is coming afterwards. It may also be permissible that the nutritive soul coming from the father stays until the temperament transforms, via some kind of transformation, [into another] and then a particular nutritive soul is attached to it, the one [i.e. the nutritive soul/power] acquired from the father does not have the power to completely manage [the embryo] to the end, and it’s only adequate for some [stages of] managing, [and] needs a powerful principle.

As an example of the latter case, consider the beginning of Text 9 where Avicenna asserts that by the origination of the first organ the vegetative power is originated. In the same text, Avicenna explains “when that [i.e. the body] is ensouled, the soul takes action through it to complete the organs. And this soul, at the time, is a nutritive [vegetative] soul or [equivalently] has no other action [besides fulfilling nutritional tasks], though it has the potentiality for other things.” At the end of Text 12, Avicenna makes it utterly clear that the sensitive soul and the rational soul are originated with the origination of the heart, though the rational soul is inactive. Therefore, if we disambiguate the ‘human body’ by taking it to be used in its non-ordinary sense, vagueness will become a distinctive problem since the term, used this way, has no established meaning. To narrow down the range of possible meanings, then, we propose a descriptive and partial precisification: the ‘human body’ refers to whatever material body that comes to possess the vegetative power of the soul. For Avicenna, the first major organ undoubtedly possesses this power, as textual evidence witnesses. Whether, at an earlier stage after the conception and before the origination of the first major organ, the material body undergoing embryonic development possesses the ‘vegetative power of the soul’ is not fully determined. 68

See McGinnis, “On the Moment of Substantial Change.”

Avicenna on the Origination of the Human Soul

73

7 . T W O CO R O LL A R I E S Our reading of Co-origination and the different senses of the ‘vegetative’ and ‘animal’ soul introduced above may also shed light on some related problems in Avicenna’s philosophy of the soul. Here we will consider two such problems, one in the contemporary literature and the other in the classic.

a. The Problem of Receiving Knowledge of Particular Events from Celestial Spheres Dimitri Gutas’s study, “Imagination and Transcendental Knowledge in Avicenna,” begins with a distinction between two kinds of ‘supernal knowledge’ in terms of their objects: one is about ‘intelligible universal concepts’ whereas the other is about ‘forms of particular’ events on earth. Gutas explains that conflating the two kinds of knowledge, their natures and sources, has led to a false interpretation of Avicenna: Most unfortunately, these problems contributed to a false understanding of Avicenna’s basic position and led to the belief in modern scholarship that Imagination . . . may enter conjunction with the active intellect and receive the emanation of the active intellect directly [Herbert Davidson, Alfarabi, Avicenna, and Averroes on Intellect, 123]. But this cannot be correct, as Dag Hasse pointedly remarks: the animal soul cannot receive knowledge from the supernal realm and it cannot have this prophetic property because, “if this were correct . . . animals could become prophets as well” [Hasse, Avicenna’s De Anima in the Latin West, 159]. (Gutas, “Imagination and Transcendental Knowledge,” 341)

In a footnote, Gutas provides further evidence for this alleged misinterpretation of Avicenna by Davidson and Michael Marmura as well.69 Independently of the problem of the nature of ‘supernal knowledge’ and whether Marmura’s and Davidson’s interpretation contain any grain of truth, we suspect that Hasse’s main reason, which is Gutas’s as well, for holding that the ‘animal soul’ of a human “cannot receive knowledge from the supernal realm,” is inconclusive. The claim that “if this were correct . . . animals could become prophets as well” is based on conflating the ‘animal soul’ in its specific sense (as a species) with the animal soul as a power of the rational soul. As was argued, the animal soul in its specific sense exists as the form of 69 See Davidson, Alfarabi, Avicenna, and Averroes on Intellect, 119 and 121, and Michael E. Marmura, Ibn Sīnaˉ . Fī Ithbaˉ t al-nubuwwaˉ t (Beirut: Daˉ r an-nahaˉ r, 1968), xiii.

74

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

the matter that constitutes a non-human animal species; it is material and mortal. In contrast, the animal soul of a human being is not the form of the matter that constitutes a human being. It is a power of the abstract entity, i.e. the human soul, that manages the human body. To wit, the two things we refer to by the ‘animal soul’ have different causal powers: for example, whereas the former, the animal soul of a non-human animal, completes its body as an animal, and cannot exceed that boundary of perfection toward a more perfect body, the latter, i.e. the animal soul of a human, completes its body as a human. For Avicenna, these are entities of two different categories. The animal soul, as a power of the human soul, is not necessarily mortal. For one thing, Avicenna approvingly considers the possibility of the activity of imagination after the death of the human body, for theological purposes, in aš-Šifaˉ ʾ and al-Išaˉ raˉ t wa-t-tanbīhaˉ t (see the following), among other places: Text 21. It also seems likely that what some scholars have said is true, namely: If these [imbecile souls] were pure [of vice] and separated from the body when a mode of belief in [consequences] in the hereafter had been embedded in them (which, for their like, would be like [something] akin to that which one can address the commonality), and this is represented in their souls; and if, then, [these souls] separate from [their] bodies, having no meaning attracting them to the direction above them [ . . . ] ([at the same time] there being no prohibition on the part of the celestial materials [for these souls] to be subjects for the action of a soul on them), then, [these scholars] say, [their souls] will imagine all that they had believed in regarding the states of the hereafter. The instrument [these scholars go on to explain] by means of which [such souls] are enabled to imagine would be something that belongs to the celestial bodies. They thus experience all that they have been told in the [terrestrial] world about the states of the grave, the resurrection, and the goods in the hereafter. The bad souls [these scholars continue] also witness punishment according to what had been portrayed to them in the [terrestrial] world, and they suffer it. For the forms in the imagination are not weaker than the sensible [forms] but are greater in influence and clarity, as one sees in sleep.70

To whom Avicenna is referring here, Faˉ raˉ bī or not, is not at issue; rather, his reasoning matters. If imagination, as a faculty of the animal soul which in turn is a power of the human soul, were necessarily mortal, and were corrupted upon the death of the human body (like the faculty of imagination of the animal soul (as a species) which is corrupted upon the death of the animal body), then it would be impossible for the human soul to imagine anything in the hereafter. Avicenna, however, does not consider this to be impossible. Thus, imagination, as a faculty of the human soul, 70 Ibn Sīnaˉ [Avicenna], [Aš-Šifaˉ ʾ, al-Ilaˉ hiyyaˉ t] The Metaphysics of the Healing, ed. M. E. Marmura (Provo: Brigham Young University Press, 2005), Bk. 9 ch. 7, 356.

Avicenna on the Origination of the Human Soul

75

is ontologically different from imagination as a faculty of the animal soul, in its specific sense. Correct, they both are corporeal in some sense: they need some corporeal object to be active. For this reason, and to make sense of the possibility he envisages, Avicenna hypothesizes that “the instrument by means of which” these human souls are “enabled to imagine would be something that belongs to the celestial bodies.” Again, Avicenna may not endorse this position for other reasons, but not for the corporeality of the animal soul as a power of the human soul. The same explanation may be found in al-Išaˉ raˉ t wa-t-tanbīhaˉ t.71 Thus, we are not sure how to interpret Gutas on occasions like this: They [i.e. the five internal senses] are all material faculties and located in the brain, front to back, in the order listed, with imagination placed in the centre, between imagery and estimation. As such the animal soul is mortal and dies with the death of the body, just like the vegetative soul. Humans have this latter soul in common with animals and plants, and the animal soul only in common with animals. It is only the rational soul, an immaterial substance which characterizes humans exclusively, that is immortal. (Gutas, “Imagination and Transcendental Knowledge,” 338)

The passage apparently conflates different senses of the ‘animal soul’ and is misleading. It is true that ‘humans have this latter soul,’ namely the animal soul, ‘in common with animals’ but they possess the ‘animal soul’ in a different sense. In non-human animals, the animal soul is the form of the body and exists in its specific sense (as a species); in humans, it is a power of the soul, which is not the form of the human body, and does not exist in its specific sense. Humans and animals even differ in their animal souls.

b. The Problem of the Integrator and Retentive Factor of the Fundamental Elements According to Avicenna, the human soul is distinct from the human body, its corporeality and its temperament. In al-Išaˉ raˉ t wa-t-tanbīhaˉ t, under the remark titled “on proving that the human soul is not the corporeality [of the body] nor the temperament,” Avicenna concludes that “the fundamental 71 See Ibn Sīnaˉ [Avicenna], [Al-Išaˉ raˉ t wa-t-tanbīhaˉ t, vol. 4] Ibn Sīnaˉ and Mysticism. Remarks and Admonitions: Part Four, ed. and trans. Shams C. Inati (London and New York: Kegan Paul International, 1996). For example: Text 22. As regards the unalert, if they raise themselves above imperfections, they will be set free from the body and will reach the happiness that befits them. But perhaps even in this state of relative happiness, they do not dispense with the assistance of a body which is the subject of their imaginations (778). It is not impossible that this body be a celestial body or the like. Perhaps this leads them eventually to the preparation for the happinesscausing conjunction that the knowers enjoy (779).

76

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

powers that move, apprehend, and retain the temperament are something other than [the temperament]. This you may call the ‘soul’ ” (Inati, Ibn Sina’s Remarks and Admonitions, 97).72 On this point, in his extensive commentary on Avicenna’s al-Išaˉ raˉ t wa-t-tanbīhaˉ t, at. -T. ūsī explicates Avicenna’s view by adding that the soul manages different parts of the body, i.e. the organs, prior to the body. Thus not only the temperament of the body but also the temperament of each organ is managed by the soul. At. -T. ūsī, then, considers an issue, discussed by Fahr-ad-Dīn ar-Raˉ zī, regard̆ ing the priority of the soul over the temperament. At. -T. ūsī’s summary of the problem goes as follows: Text 23. And on this position a famous question is raised and it goes like this. You said that the composites become prepared to receive their forms from their principles in accordance with their different temperaments. And from that, the priority of the temperaments over those forms [necessarily] follows. And now [in contrast] you are saying that the soul, which is the form of the animal, is the integrator of its fundamental elements and the integrator of the fundamental elements should be prior to the temperament. But this is a contradiction.73

The ‘problem of the integrator and retentive factor of the fundamental elements’ may be reformulated as follows: On the one hand, a composite receives its form only if it is prepared (or has the potentiality) to do so. And it realizes such preparedness (or potentiality) only if it has the right temperament. Therefore, the temperament of the composite (the animal body or the animal organ) is prior to its form (the animal soul). On the other hand, it is said that the animal soul manages the animal body and its organs. Therefore, the animal soul is the integrator and retentive factor of the fundamental elements of the animal body. Moreover, the fundamental elements form the temperament of the animal body (or that of its organs). Hence, the temperament of the animal body is posterior to the animal soul. But this is contradiction: the temperament of the animal body cannot be

72 I have followed Shams Inati’s (2014) translation, except that I have added a pair of single quotations around ‘soul’: Ibn Sīnaˉ [Avicenna], [Al-Išaˉ raˉ t wa-t-tanbīhaˉ t, vols. 2 and 3] Ibn Sina’s Remarks and Admonitions: Physics and Metaphysics, an analysis and annotated translation, ed. and trans. S. C. Inati (New York: Columbia University Press, 2014), third class ch. 5. 73 Al-Išaˉ raˉ t wa-t-tanbīhaˉ t li-Abī ʿAlī ibn Sīnaˉ , ma ʿa Šarh. Nas.īr ad-Dīn at. -T. ūsī [/Šarh. al-Išaˉ raˉ t], 354. The translation is ours. To distinguish the four elements, i.e. al- ʿanaˉ sur alarba ʿa, from the fundamental constituents of objects, i.e. ust. uqiss, we have used ‘fundamental elements’ as a rendering of the latter. This is unfortunate, since ‘fundamental elements,’ predicatively read, may refer to the elements that are more fundamental than the four elements. This, however, is not the intended reading. Fundamental elements, in relation to the ordinary objects, are the fundamental constituents of them, and in relation to the four elements, are made out of them.

Avicenna on the Origination of the Human Soul

77

both prior (as the first horn suggests) and posterior (as the second horn suggests) to the animal soul, in one and the same sense. Before considering our solution, let us briefly review at. -T. ūsī’s and arRaˉ zī’s solution, as reported by at. -T. ūsī in his commentary.74 First, ar-Raˉ zī’s solution: Text 24. And the meritorious commentator replies to that [problem] as follows: the integrator of the parts of the semen is the soul[s] of the parents, then that temperament will be under the administration of the mother’s soul until it is prepared to receive a soul. Then, after its origination, it [i.e. the soul] will become its retentive [factor] and the integrator of other parts [of it] by providing nourishment [for it].75

Ar-Raˉ zī distinguishes between, at least, three different stages: First, the stage in which the parts of the semen are integrated by the soul[s] of the parents; second, the stage in which the same temperament is managed by the soul of the mother; and third, the stage in which the embryo is ensouled and therefore it has its own soul which is the integrator and retentive factor of its temperament. At. -T. ūsī rejects this solution for the following reason: Text 25. If the mother’s soul were the manager of the [embryo’s] temperament, how could it delegate the administration after awhile to the rational [soul], such a thing [i.e. delegation of power] happens in non-natural agents that act in accordance with [their] renewing wills.76

Here is our formulation of at.-T.ūsī’s reasoning: (a) if at one stage the mother’s soul is the integrator and retentive factor of the fundamental elements of the embryo’s body and at a later stage this task is given to another thing, e.g., the embryo’s soul, then some kind of delegation of power should occur. (b) This kind of delegation of power only occurs in “non-natural agents that act in accordance with their renewing wills.” (c) The mother’s soul and the embryo’s soul are natural agents, namely they act in virtue of their natures. (d) Therefore, there is no delegation of power in such agents. (e) Thus, by modus tollens, it is not the case that at one stage the mother’s soul manages the embryo’s body and at a later stage the embryo’s soul takes up the job. We believe that at. -T. ūsī’s reasoning is not sound since (a) is false. (a) would be true, if what is integrated and preserved, i.e. the fundamental elements of the embryo’s body, at different stages, were the same. We believe this is not the case for Avicenna. Let us consider the problem in the case of the human embryo.

At. -T. ūsī reports that ar-Raˉ zī offers two different solutions to the problem. At. -T. ūsī rejects both. Here we only consider the first solution and accept at. -T. ūsī’s reason against the other, being ar-Raˉ zī’s or not. 75 76 Šarh. al-Išaˉ raˉ t vol. 2, third class ch. 5, 354. Ibid., 355. 74

78

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

As we interpret Co-origination, it suggests that the human soul is originated long before the completion of the human body and in fact as early as the origination of the first major organ. We may not be able to determine the exact instant; however, the human soul and the first major organ (or whatever appropriate material body) are instantaneously co-originated as two substances. Therefore, the material from which the first major organ is formed undergoes substantial change, not motion, and then the first major organ is originated. Hence, what is integrated before the origination of the first major organ is a substance different from what is integrated and preserved later on: it is the fundamental elements of the first major organ, or the ‘human body’ in its non-ordinary sense. Therefore, if what is integrated and preserved at different stages is not one and the same thing, no delegation of power follows. Otherwise put, there is no single task of integrating and preserving the fundamental elements of one and the same object: one task is to integrate and preserve the fundamental elements of the mineral material from which, via substantial change, the first major organ is formed, and the other is to integrate and preserve the fundamental elements of the first major organ. These are different tasks involving different objects, and thus may be taken by different natural agents. Hence, at. -T. ūsī’s reasoning against ar-Raˉ zī fails. After rejecting ar-Raˉ zī’s solution, at. -T. ūsī offers his own: Text 26. What the philosophical rules suggested by Avicenna and others imply is that the soul[s] of the parents gather nutritive ingredients by the attracting power and then turns them into mixtures and subsequently extract the matter of seminal fluid from them, by means of the procreative power, and make it ready to receive a potentiality whose significance is to prepare the matter to become a human. Thus, in virtue of this potentiality it becomes seminal fluid and that potentiality would be the form that preserves the temperament of the seminal fluid, like the mineral form. Then, the seminal fluid attains more perfection in the womb in accordance with the predispositions it acquires therein; [this will go on] till it becomes prepared to receive a more perfect soul that executes vegetative acts in addition to preserving the matter, thus it gathers nutrition and supplements that matter and makes it grow, hence the matter evolves under its administration. So, that form becomes the origin of the [new] acts in addition to what is already originated. And like this [it evolves] to become prepared to receive a more perfect soul that originates, in addition to all previous acts, animal acts. Thus, these acts are also originated by it and then the body is completed and evolves to become prepared to receive a rational soul, or to bring about reason [/speech] besides all the previous [acts], and [this soul] will remain as the administrator of the body until the moment of death arrives.77

77

Ibid.

Avicenna on the Origination of the Human Soul

79

Then at. -T. ūsī appeals to a renowned analogy in the peripatetic tradition to further explicate his solution.78 A charcoal juxtaposed to a piece of fire initially gets (i) warm, and then becomes (ii) hot, and when it is heated enough it turns to a (iii) glowing charcoal and finally (iv) it flames. The charcoal’s warmth is analogized to the mineral form, the following hotness to the principle of vegetative acts, the glowing phase to the principle of animal acts and the flaming to the rational soul. At. -T. ūsī’s solution might be interpreted in two ways. Here is the first interpretation. The fundamental elements of the embryo’s body have different integrating and retentive factors: at the first stage, the souls of the parents are the integrators of the fundamental elements of the seminal fluids and the resulting mineral form would be the preserver of its parts. At the second stage, the vegetative soul is the integrative and retentive factor of the fundamental elements of the body, and at the third stage, the animal soul and finally at the fourth stage the rational soul (that will remain the administrator of the human body to the moment of death). Hence, different ‘souls’ are understood as different entities.79 We do not follow this interpretation for reasons discussed against ‘successive ensoulment’ according to which different souls (in humans) come into existence as different entities, particularly as different substantial forms (see section 6). The second interpretation goes like this. There are only two stages: As before, at the first stage the integrator is the souls of the parents and the preserver is the mineral form. The further three stages mentioned above constitute one phase in which a single soul, namely the human soul, is the integrator and retentive factor of the fundamental elements of the embryo’s body, though by means of different powers, namely the vegetative, animal, and rational. So, these are not different ‘souls’ as distinct entities; after the first stage, the human soul is originated, although only its vegetative power is active. The following passage supports this latter interpretation: Text 27. Thus all these powers are like one thing, heading toward a boundary of perfection from a boundary of deficiency. The name ‘soul’ is attributed to the last three [powers] and these, with the differences in their levels, are the soul of the body of the newborn. And from that, it will be clear that the integrator of the nutritive parts constituting the semen is the soul[s] of the parents, and this is not the retentive factor. And the integrator of the parts added to that [i.e. to the semen] until the perfection of the [human] body and up to the end of life and the retentive factor of the temperament [of the human body] is the soul of the newborn.80 78

Ibid., 356. For an interpretation along this line see [Saadat] Mostafavi, “Problem in Avicenna’s View.” 80 Šarh. al-Išaˉ raˉ t, 356. 79

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

80

This interpretation is compatible with our reading of Co-origination. After conception, some substantial changes are necessary for the embryo to get prepared (by developing veins, neural system, etc.) to receive the human soul (Text 16). These changes, nonetheless, do not push the ‘coagulated blood’ to get in-formed by some vegetative or animal soul, as a substantial form. Some substantial changes happen but they are not transitions from one species to another. As soon as the ‘coagulated seminal fluid’ comes to possess the vegetative power of the soul, which certainly occurs by the origination of the first major organ, the human soul is originated. Hence, the human soul is originated long before the completion of the embryo’s body. It is a simple substance with various powers that may get activated at different times, as it is the case with the rational power indisputably. This is an SE-solution (a Simple soul originated Early, see Table 2.2).

8 . O N T H E ‘ C O N T R A R Y’ E V I D E N C E The textual evidence seemingly ‘contrary’ to our reading may be categorized under four headings: directly contrary, ambiguously contrary, apparently contrary, and allegedly contrary. We will discuss each category separately and try to show that they do not undermine our reading. A case of directly ‘contrary’ textual evidence may be found in Text 1. In this passage, and similar ones, Avicenna is explicit that “human souls are originated with the origination of human bodies.”81 We foresee two ways to handle such passages. First, Avicenna might use the ‘human body’ in an ambiguous manner and what he really means is that the ‘human body’ in its non-ordinary sense is co-originated with the human soul. Second, if we note that these texts are mainly coming from treatises like al-Mabda ʾ wa-l-ma ʿaˉ d, where Avicenna is not in the position of formulating his mature analysis of the soul, then even if the ‘human body’ is used unambiguously in its ordinary sense, we may still suppose that he is reporting the dominant peripatetic view on the origination of the human soul, not his own. The fact that Avicenna’s formulation of Co-origination in aš-Šifaˉ ʾ is considerably different from what we find in the above works, supports this latter explanation (see below). A witness of the second category, ambiguously ‘contrary’ evidence, may be found in Text 2: “thus it is correct that the soul is originated just as the body appropriate for its use is originated.”82 In such cases, Avicenna does not use the term ‘human body’; rather, he employs the ‘body appropriate for 81

The Provenance and Destination, 108.

82

States of the Soul, 97.

Avicenna on the Origination of the Human Soul

81

its use,’ namely appropriate for the use of the human soul. We do not find such texts to be directly contrary to our reading of Co-origination. If one takes the ‘body appropriate for its use’ to refer to what the ‘human body’ ordinarily refers to, then the text provides contrary evidence. However, if one takes the ‘body appropriate’ for the human soul’s use to refer to whatever material body that suits specific roles, e.g., individuating the human soul and initiating the human spirit as a medium for the formation of the human organs, then the text does not contradict our interpretation. We speculate that all such ambiguously contrary pieces of evidence may be disambiguated in accordance with our reading. Of the third category, apparently ‘contrary’ evidence, we may consider this: Text 28. [ . . . ] Thus every body, with the origination of the temperament of its matter, demands the origination of its soul, and it is not the case that a body demands [the origination of the soul] and another body does not demand, because the particulars [falling] under [natural] kinds do not differ in what in virtue of which they are constituted [ . . . ]83

Here, Avicenna talks about the “origination of the temperament of its matter,” namely the matter of the body. Such a passage may go contrary to our reading only if it is presupposed that by the term ‘body’ here Avicenna means the ‘human body’ in its ordinary sense and that the temperament of the matter constituting the human body is not prior to the origination of the perfected human body. Both presuppositions are questionable, however. It is consistent with Avicenna’s metaphysics to assume that the temperament of the material constituting the embryo’s body exists long before the completion of the ‘human body’ in its ordinary sense. Accordingly, such a passage may support our reading, rather than hindering it. (Note: presupposing that the term ‘body,’ in such a use, refers to what the ‘human body’ ordinarily refers to is begging the question.) Finally, we shall consider allegedly ‘contrary’ evidence. For instance, Gutas (“Imagination and Transcendental Knowledge,” 339–40) complains about Avicenna’s not giving ‘further specification’ in his use of ‘soul’ on occasions like the following: Text 29. As for the reason for the knowledge of [future] events (al-kaˉ ʾinaˉ t), it is the contact of the human soul with the souls of the celestial bodies which [ . . . ] know what happens in the world of the elements [ . . . ]. For the most part, these [human] souls come into contact with them [the souls of the celestial spheres] by virtue precisely of a congeneric similarity (mujaˉ nasa) between them. The congeneric similarity is that

83

Avicenna’s De Anima Bk. 5 ch. 4, 233.

82

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

thing (al-ma ʿnaˉ ) which, there [in the heavens], is close to the concerns of these [human souls]. So most of what is seen [by the human souls] of what is to be found there [in the heavens] is congeneric to the states of the bodies of these [human] souls or to the states of one who is close to these bodies. And although the contact [of the human souls with the celestial souls] is total, the majority of the influence they receive from them is for the most part close to just their [own] concerns. This contact comes about on the part (min jiha) of the estimation and imagination (al-hayal) and ̆ through their use, and concerns particular things, whereas contact by means of the 84 intellect is something else, which is not our subject here.

Accordingly, the human soul via estimation and imagination, namely two of its animal powers, comes into contact with the celestial souls. As was discussed earlier, Gutas rejects this view: First, there is the obvious contradiction in the two statements: is it the human or animal soul that receives knowledge from the souls of the celestial spheres? Second, and most importantly, if indeed it is the animal soul that receives this knowledge, how is it possible for a material and mortal substance to be in contact with divine and immortal supernal world? (Gutas, “Imagination and Transcendental Knowledge,” 340)

Note that Text 29 prima facie supports the view that Gutas rejects. The view in question, however, is consistent with our understanding of Co-origination. Thus, if Gutas is right, the prima facie reading of Text 29 is misleading and it needs to be reinterpreted. The allegedly correct reinterpretation of Text 29 should not support our understanding of Co-origination. We agree with Gutas’s distinction between the two kinds of supernal knowledge, to wit, one that has ‘intelligible universal concepts’ as its object and the other that has the ‘forms of particular events’ on the earth. Also, Gutas may be right that in the first kind of supernal knowledge, it is the rational soul that is primarily connected to the active intellect. None of this, however, shows that the imagination cannot be in contact, or be acted upon, by the celestial souls. Gutas asks, “How is it possible for a material and mortal substance to be in contact with divine and immortal supernal world?” The question sounds anachronistic and puzzling to us for three reasons. First, the imagination, as an internal sense of the animal soul that is a power of the human soul, is not necessarily material and mortal; it is the imagination as a power of the ‘animal soul,’ in its specific sense (as a species), that is material and mortal since the ‘animal soul’ in this sense is the form of the animal body. Second, given that for Avicenna the souls of celestial

84 The Provenance and Destination Bk. 3 ch. 17, 117. The translation is Gutas’s and is quoted from Gutas, “Imagination and Transcendental Knowledge,” 339.

Avicenna on the Origination of the Human Soul

83

spheres as well as animal souls are corporeal,85 why should the property of being ‘mortal’ or ‘immortal’ matter at all? Corporeality might be enough for the possibility of the ontological connection between such entities. Third, what does ‘connection’ mean here? After all, the human soul, as an immaterial and immortal entity is ‘in contact with’ the animal soul, as a material and mortal entity, in humans, as Gutas reads Avicenna. Gutas’s question neither matches his own presuppositions nor accords with our reading of Avicenna. Thus, we conclude, the prima facie reading of Text 29 is not misleading and hence the text is only ‘allegedly’ against our reading.

9 . O P E N Q U E ST I O N S Let us end this study by turning to some open questions that it engenders. On the moral philosophy side, we would like to mention two: If the human soul is originated at least as early as the first major organ, what will be Avicenna’s view on the ethics of abortion? And, as a related problem, does the human embryo with the human soul but without a human body, in its ordinary sense, count as ‘a person’?86 On the metaphysical side, if the human soul is originated prior to the completion of the human body, what will be the role of the human soul in the formation of the human body, if any? And, as a follow up, what does our solution to the previous problem tell us about Avicenna’s theory of causation, in particular about causation between abstract and concrete objects?87 On the philosophy of mind side, if the human soul has three powers, the vegetative, animal, and rational, and these are different from the ‘vegetative’ and ‘animal’ soul, in the specific sense, as existing in plants and non-human animals, then do these powers contribute to the human cognitive processes in a different manner? And finally, more on the epistemological side, is it possible for the active intellect to act upon the imagination, as a faculty of the human soul, and

Gutas, “Imagination and Transcendental Knowledge,” 344. This question should be taken with extra care, bearing in mind that the distinction between the ‘human individual’ and the ‘human person’ may be a modern innovation, although with some historical roots. 87 The particular corporeal body co-originated with a particular human soul is metaphysically necessary for the individuation of that human soul, as we read Avicenna. However, it is not metaphysically essential to the human soul, in the sense that it does not contribute to the essence of the human soul. Accordingly, the particular corporeal body is the accidental cause of the particular human soul. ‘The,’ here, should imply uniqueness; no other corporeal body individuates that human soul. 85 86

84

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

work on a particular form therein in the process of abstraction of universal concepts? We shall not try exploring these open questions here. We only mentioned them to highlight the far-reaching and wide-ranging possible implications of our reading.88 University of Gothenburg (Sweden) and Institute for Research in Fundamental Sciences (IPM) (Iran) Imam Sadiq University (Iran)

BIBLIOGRAPHY Aristotle. The Complete Works of Aristotle: The Revised Oxford Translation (Sixth Printing, with Corrections), ed. Jonathan Barnes (Princeton, NJ: Princeton University Press, 1995). Baʿalbaki, Rohi (ed.). Al-Mawrid: A Modern Arabic English Dictionary (Beirut: Dar el-ilm lilmalayin, 1995). Corcoran, John. “Borderline Case,” in R. Audi (ed.), The Cambridge Dictionary of Philosophy, 2nd edition (New York: Cambridge University Press, 1999), 96. Davidson, Herbert A. Alfarabi, Avicenna, &Averroes on Intellect: Their Cosmologies, Theories of the Active Intellect, and Theories of Human Intellect (Oxford: Oxford University Press, 1992). Druart, Thérèse-Anne. “The Human Soul’s Individuation and its Survival after the Body’s Death: Avicenna on the Causal Relation between Body and Soul,” Arabic Sciences and Philosophy 10 (2000), 259–73. Gutas, Dimitri. “Avicenna: The Metaphysics of the Rational Soul,” The Muslim World, Special Issue: The Ontology of the Soul in Medieval Arabic Thought 102 (2012), 417–25.

88

This work is done within the research program Representation and Reality: Historical and Contemporary Perspectives on the Aristotelian Tradition, funded by Riksbankens Jubileumsfond. It is also partly supported by the Center for Medieval Islamic Philosophy (MIP) at the Institute for Research in Fundamental Sciences (IPM), Tehran, Iran. We find ourselves indebted to our audience at The Second Annual Conference of the British Association for Islamic Studies/BRAIS 2015 (School of Advanced Study, University of London, London, UK), the IRIP Philosophy Colloquium Series 2014–2015 (Iranian Institute of Philosophy, Tehran, Iran), and the Representation and Reality Research Seminars 2015 (Department of Philosophy, Linguistics, and Theory of Science, Gothenburg, Sweden). We would like thank David Bennett, Börje Bydén, Véronique Decaix, Sten Ebbesen, Jakob Leth Fink, Katerina Ierodiakonou, Taneli Kukkonen, Ana Maria MoraMárquez, Filip Radovic, Ayman Shihadeh, and Christina Thomsen Thörnqvist. We owe special thanks to Sten Ebbesen who read two earlier drafts of the paper thoroughly and made a number of helpful comments. Also, we are thankful to two anonymous referees for numerous instructive suggestions and corrections. In completing this research, we have benefited from the database of the Computer Research Center of Islamic Sciences. The second author’s contribution is limited to section 7b.

Avicenna on the Origination of the Human Soul

85

Gutas, Dimitri. “Imagination and Transcendental Knowledge in Avicenna,” 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), 337–54. Hall, Robert E. “Intellect, Soul and Body in Ibn Sina: Systematic Synthesis and Development of the Aristotelian, Neoplatonic and Galenic Theories,” in J. McGinnis with the assistance of David C. Reisman (eds.), Interpreting Avicenna: Science and Philosophy in Medieval Islam, Proceedings of the Second Conference of the Avicenna Study Group (Leiden: Brill, 2004), 62–86. Hasse, Dag N. Avicenna’s De Anima in the Latin West: The Formation of a Peripatetic Philosophy of the Soul, 1160–1300 (London: Warburg Institute, 2000). Ibn Sīnaˉ [Avicenna]. [Al-Išaˉ raˉ t wa-t-tanbīhaˉ t, vols. 2 and 3] Ibn Sina’s Remarks and Admonitions: Physics and Metaphysics, an analysis and annotated translation, ed. and trans. S. C. Inati (New York: Columbia University Press, 2014). Ibn Sīnaˉ [Avicenna]. [Aš-Šifaˉ ʾ, at.-T.abīʿiyyaˉ t, as-Samaˉ ʿ at.-t.abīʿī] The Physics of the Healing, ed. and trans. J. McGinnis (Provo, UT: Brigham Young University Press, 2009). Ibn Sīnaˉ [Avicenna]. [Aš-Šifaˉ ʾ, al-Ilaˉ hiyyaˉ t] The Metaphysics of the Healing, ed. M. E. Marmura (Provo, UT: Brigham Young University Press, 2005). Ibn Sīnaˉ [Avicenna]. Al-Qaˉ nūn fī t. -t. ibb [The Canon of Medicine]. F1953. Saab Medical Library, American University of Beirut, 2002–2007. Last modified: Mon Mar 26 17:52:34 2007 BL, at: . Ibn Sīnaˉ [Avicenna]. [Al-Išaˉ raˉ t wa-t-tanbīhaˉ t, vol. 4] Ibn Sīnaˉ and Mysticism. Remarks and Admonitions: Part Four, ed. and trans. S. C. Inati (London: Kegan Paul, 1996). Ibn Sīnaˉ [Avicenna]. Al-Mabda ʾ wa-l-ma ʿaˉ d [The Provenance and Destination], ed. Abdallaˉ h Nūraˉ nī (Tehran: The Institute of Islamic Studies, 1363/1984). Ibn Sīnaˉ [Avicenna]. Al-Adwiya al-qalbiyya [Remedies for the Heart], in Min Mu ʾallafaˉ t Ibn Sīnaˉ at. -T. ibbiyya, ed. M. Zuhayr al-Baˉ baˉ (Aleppo: Maʿhad atturaˉ tˍ al-ʿilmī al-Islamī, 1404/1984), 221–94. Ibn Sīnaˉ [Avicenna]. Aš-Šifaˉ ʾ, at. -T. abī ʿiyyaˉ t, al-Ḥ ayawaˉ n [The Healing, Natural Sciences, The Animal], ed. ʿAbdalh.alīm Muntas.ir, Saʿīd Zaˉ yid, and ʿAbdallaˉ h Ismaˉ ʿīl (Cairo: al-Hayʾa al-mis.riyya al-ʿaˉ mma li-t-taʾlīf wa-n-našr, 1970). Ibn Sīnaˉ [Avicenna]. [Aš-Šifaˉ ʾ, at. -T. abī ʿiyyaˉ t, an-Nafs] Avicenna’s De Anima (Arabic Text): Being the Psychological part of Kitaˉ b al-Shifaˉ ’, ed. F. Rahman (London: Oxford University Press, 1959). Ibn Sīnaˉ [Avicenna]. Ah.waˉ l an-nafs [States of the Soul], ed. A. Fuʾaˉ d al-Ahwaˉ nī (Cairo: Daˉ r ih.yaˉ ʾ al-kutub al-ʿarabiyya, 1371/1952). Ibn Sīnaˉ [Avicenna]. [An-Najaˉ t, at. -T. abī ʿiyyaˉ t, an-Nafs] Avicenna’s Psychology: An English Translation of Kitaˉ b al-Najaˉ t, Book II, Chapter VI, ed. and trans. F. Rahman (London: Oxford University Press, 1952). Ibn Sīnaˉ [Avicenna]. Fī l-Ajraˉ m al- ʿulwiyya [Treatise on the Supernal Bodies], in Tis ʿRasaˉ ʾil fī l-h.ikma wa-t. -t. abī ʿiyyaˉ t, Treatise 2 (Cairo: Mat. baʿa hindiyya, 1326/ 1908). Kenny, Joseph. “Ibn Sīnaˉ and the Origin of Human Life,” in K. D. Crow (ed.), Islam, Cultural Transformation and the Re-emergence of Falsafah: Studies Honoring Professor George Fransic McLean on His Eightieth Birthday (Tehran: Iranian Institute of Philosophy Press, 2009).

86

Seyed N. Mousavian and Seyed Hasan Saadat Mostafavi

McGinnis, Jon. “On the Moment of Substantial Change: A Vexed Question in the History of Ideas,” in J. McGinnis with the assistance of D. C. Reisman (eds.), Interpreting Avicenna: Science and Philosophy in Medieval Islam, Proceedings of the Second Conference of the Avicenna Study Group (Leiden: Brill, 2004), 42–61. Marmura, Michael E. “Some Questions regarding Avicenna’s Theory of the Temporal Origination of the Human Rational Soul,” Arabic Sciences and Philosophy 18 (2008), 121–38. Marmura, Michael E. (ed.) Ibn Sīnaˉ . Fī Ithbaˉ t al-Nubuwwaˉ t (Beirut: Daˉ r an-nahaˉ r, 1968). Mulla Sadra. The Wisdom of the Throne (al-Ḥ ikmaal- ʿAršiyya), trans. J. W. Morris (Princeton, NJ: Princeton University Press, 1981). Musallam, Basim. “The Human Embryo in Arabic Scientific and Religious Thought,” in G. R. Dunstan (ed.), The Human Embryo: Aristotle and the Arabic and European Traditions (Exeter: University of Exeter Press, 1990), 32–48. Pasnau, Robert. Thomas Aquinas on Human Nature: A Philosophical Study of Summa theologiae Ia 75–89 (Cambridge: Cambridge University Press, 2002). Pelletier, Francis Jeffry and István Berkeley. “Vagueness,” in R. Audi (ed.), The Cambridge Dictionary of Philosophy, 2nd edition (New York: Cambridge University Press, 1999), 945–7. [Saadat] Mostafavi, Seyed Hasan. “A Problem in Avicenna’s View on the Origination of the Soul and a Reply to It,” Sophia Perennis: The Journal of Sapiential Wisdom and Philosophy 2/4 (2011), 19–29. Silva, José Filipe. “Robert Kilwardby,” The Stanford Encyclopedia of Philosophy (Summer 2016 edition), ed. Edward Zalta, . At. -T. ūsī, Nas.īr-ad-Dīn. Al-Išaˉ raˉ t wa-t-tanbīhaˉ t li-Abī ʿAlī ibn Sīnaˉ , ma ʿa Šarh. Nas.īr ad-Dīn at. -T. ūsī [/Šarh. al-Išaˉ raˉ t], ed. S. Dunyaˉ , 4 vols. (Cairo: Daˉ r al-maʿaˉ rif, 1971). Wehr, Hans. A Dictionary of Modern Written Arabic, 3rd edition, ed. J. M. Cowan (New York: Spoken Language Services Inc., 1976).

The Flow of Powers Emanation in the Psychologies of Avicenna, Albert the Great, and Aquinas Charles Ehret

I’m at the cinema trying to watch a movie, but I’m hungry, and I’m wondering if I have enough money left in my wallet to pick up something to eat on my way home. Quite naturally, I ascribe all these processes to ‘me.’ But on what grounds? And what might ‘I’ be, that such different processes as seeing, feeling hungry, imagining, and calculating may all be ascribed to ‘me’? It is at least feasible that what is hungry and what is calculating is not the same thing. But if that’s true, then what I refer to everyday as ‘me’ might simply not exist. Medieval philosophers were well aware of this problem. According to them, the processes that occur in a living being all have a share in the soul, which may at least partially be described, if not defined, as the actuality of the body (actus corporis), i.e., what makes the body a living organism, as opposed to a corpse. The problem at hand, then, is to understand how the soul itself is one, despite the many and multifarious activities that take place in various parts of the body, and despite the fact that intellectual operations, at least in some sense, do not take place in any particular organ. A common medieval response to this problem is first to ascribe each process to a determinate power, and then to ascribe each of these powers to the soul, powers acting as buffers between the various operations, in need of various principles, and the soul itself. To capture this view, it is often said that the powers of the soul ‘emanate’ or ‘flow’ from its essence.1 Such 1

The medieval doctrine of soul powers has received sustained attention in recent scholarship; see Dominik Perler, “Faculties in Medieval Philosophy,” in D. Perler (ed.), The Faculties: A History (New York: Oxford University Press, 2015), 97–139; Peter King, “The Inner Cathedral: Mental Architecture in High Scholasticism,” Vivarium 46 (2008), 253–74; and Joël Biard, “Diversité des fonctions et unité de l’âme dans la psychologie

88

Charles Ehret

phrasing became widely used after Albert the Great who, as Sander de Boer writes, “supplied what became the most popular way of speaking about the relation between soul and powers, namely that the powers flow forth from the soul.”2 The general idea is that the various powers are parts of the soul, that they depend on the soul’s essence for their continued existence, but that they are nonetheless extrinsic to that essence, thus not involving the soul itself in the various operations these powers perform. This allows a living being to perform many activities through its various powers, without breaking up into multiple agents. Now the flow of powers strongly suggests two claims. First, it suggests that the soul is a distinct substance from the body: insofar as it does not engage in any specific operation, each accounted for by an emanated power, the soul itself remains essentially unaffected by what goes on in a living organism, and may thus be thought to exist independently, as an immaterial substance.3 Second, it suggests that the soul has a distinct essence from its powers: because a power is determined as the principle of a specific operation, and because the soul’s essence is not such a principle, it may be thought of as essentially something else than power. Following these suggestions, one can say that the different operations of a human being relate, through a variety of powers, to something more fundamental than these powers, an immaterial substance, which can give rise, because it is immaterial, to intellectual operations, which don't actualize any specific organ. The human soul should thus be thought of as an immaterial substance essentially distinct from its powers. This is Albert’s view and, for him, to say that powers flow from the soul simply means that the soul is such a substance. However, Albert did not invent the phrase, which he explicitly borrows from Avicenna. It is commonly held that Albert is true to Avicenna when he quotes him on this point. Dag Nikolaus Hasse thus notes:

péripatéticienne (XIVe–XVIe siècle),” Vivarium 46 (2008), 342–67. Few studies, however, give emanation a prominent role. For a notable exception, see Thérèse Bonin, “The Emanative Psychology of Albertus Magnus,” Topoi 19 (2000), 45–57, who offers a broad historical survey, going back to Plotinus, as well as an interpretation of emanation in the psychologies of Albert and Aquinas. According to the author, emanation has a rich and consistent meaning up until Albert (included), but ceases to be truly relevant in Aquinas. My account will differ substantially. 2 Sander W. de Boer, The Science of the Soul: The Commentary Tradition on Aristotle’s De anima, c.1260–c.1360 (Leuven: Leuven University Press, 2013), 227 (n. 49). 3 This view is not identical to ‘substance dualism’: that the soul may exist independently from the body does not imply that the body may exist independently from the soul. Indeed, medieval authors unanimously hold that a body without a soul is only homonymously a ‘body.’

The Flow of Powers

89

Ontologically [soul and faculties] are distinct. On this matter, Albert adopts Avicenna’s thesis that the organic and non-organic faculties emanate from one substance, the soul, which exists independently of both its actions and its body.4

I will argue that such an account is both inaccurate and misleading. It is inaccurate, because Avicenna does not hold that non-organic powers (namely, intellectual powers) emanate from the soul. Misleading, because an ontological distinction between the soul and its powers obtains only if all powers emanate. But because Avicenna only says organic powers flow, the implied distinction is only between the soul qua immaterial substance and the organic powers by which it actualizes the body, and this does not entail that the essence of the soul is something else than power, i.e., intellectual power. Therefore, Albert introduced, under the name of Avicenna, an altered thesis, according to which powers flow without specification from the soul’s essence. In Albert, this thesis not only entails that the soul is a distinct, immaterial substance (as in Avicenna), it additionally entails that its essence is ontologically distinct from its powers. Surprisingly, this addition is what made its success, since it is generally no longer considered true, after Albert, that the soul is an immaterial substance. Thus Thomas Aquinas famously holds the soul to be no more than the form of a body. For him, therefore, emanation only implies that the soul’s essence is distinct from its powers; it no longer implies that the soul is a distinct substance. For this reason, Sander de Boer writes that the formula has become “a slightly awkward metaphor.”5 But would it not be strange that what became “the most popular way of speaking” was in fact an “awkward metaphor”? It is assumed here that it only makes sense to say that powers ‘flow’ if there is a gap to be crossed between two distinct substances, an immaterial soul and a material body. But I do not think this is true. On the contrary, it is precisely because the soul is no longer considered as a separate substance, but as a bodily form, that it becomes crucial to picture all its powers as distinct from its essence, so that

4 Dag Nikolaus Hasse, “The Soul’s Faculties,” in R. Pasnau (ed.), The Cambridge History of Medieval Philosophy (Cambridge: Cambridge University Press, 2010), 305–19, 306. See also: Meryem Sebti, Avicenne: l’âme humaine (Paris: Presses Universitaires de France, 2000), 37: “L’âme humaine ne doit être identifée à aucune de ses puissances opératives. Ces puissances procèdent de son essence et en sont entièrement distinctes. Son exposé de la question aura une influence importante sur les maîtres latins. [ . . . ] La notion d’émanation se trouve à l’origine de la distinction avicennienne entre l’essence de l’âme et ses puissances”; Édouard-Henri Wéber, La Personne humaine au XIIIe siècle (Paris: Vrin, 1991), 205: “Albert le Grand prise fort la notion qui fonde chez Avicenne la distinction entre l’essence de l’âme et ses puissances opératives, celle de flux ou d’émanation des puissances de l’âme.” 5 De Boer, Science of the Soul, 228 (n. 49).

90

Charles Ehret

intellectual power may be understood as immaterial, and thus account for the human soul’s incorruptibility. This is what happens in Aquinas. So, although the flow of powers suggests that the soul is an immaterial substance essentially distinct from its powers, it does not entail such a thesis, which only holds true in Albert’s doctrine. In Avicenna, the human soul is a distinct substance, but is not something distinct from power, because it is identical to intellectual power. In Aquinas, the soul is something distinct from power, but is not a distinct substance, because it is the form of a composite. Furthermore, I believe that the flow of powers only amounts to an intelligible position in Avicenna and Aquinas, not in Albert. Indeed, to meaningfully say that the powers of the soul flow from its essence, you must be able to give an account of the soul’s essence without referring to its emanated powers. Avicenna, since he only sees organic powers as emanated, can meet this challenge by saying that the soul’s essence is the material intellect, namely the power to receive intelligible forms. Aquinas can meet the challenge by saying that the soul is in essence a bodily form, and that informing the body is more fundamental than endowing the composite with various powers. But because Albert holds that the soul is a separate substance (contrary to Aquinas) and pictures all powers as emanated (contrary to Avicenna), he cannot give any positive account of the soul’s essence. Albert’s emanative psychology is therefore best understood as a moment of transition between the two consistent psychological doctrines of Avicenna and Aquinas. The aim of my paper, in establishing this, is to show how emanation successfully amounts to an ontological distinction between the soul and its powers only in Aquinas, because it does not successfully amount to such a distinction in Albert, and because it is not intended to express such a distinction in Avicenna. So although flow is, broadly speaking, a ‘Platonizing’ theme, which is favored by authors such as Avicenna and Albert, whose metaphysics are also driven by emanation theories,6 the thesis that powers as such flow from the soul only holds true in an Aristotelian doctrine, namely Aquinas’s. I will first make good on my claim that in Avicenna only organic powers flow from the soul’s essence, which is identical to intellectual power (}1). I will then show how and why Albert alters Avicenna’s claim to make all powers flow (}2). I will finally show how Aquinas makes use of Albert’s 6

For emanation in Avicenna’s metaphysics, see Olga Lizzini, Fluxus (fayd.): Indagine sui fondamenti della metafisica e della fisica di Avicenna (Bari: Edizioni di Pagina, 2011); for emanation in Albert’s metaphysics, see Thérèse Bonin, Creation as Emanation: The Origin of Diversity in Albert the Great’s On the Causes and the Procession of the Universe (Notre Dame, IN: Notre Dame University Press, 2001).

The Flow of Powers

91

emanation in his own doctrine, giving it the coherence it lacks in its Albertinian context (}3), before concluding (}4).

1. AVICENNA Avicenna opens the Psychology of The Healing with a definition of the soul as that from which operations emanate, so that the soul is first understood as identical to power. He does considerably refine this opening definition, but at no point in this refinement are powers understood as derivative of the soul’s essence, which continues to appear as constituted by its powers, as I will first explain (}1.a). Powers are said to flow only when Avicenna establishes, at the end of his treatise, the soul’s unity; here I will insist that only organic powers are concerned, and I will defend that the human soul is, according to Avicenna, a bundle or an array of coordinated powers, organic ones flowing into various organs to actualize the body, and others belonging exclusively to the soul’s essence (}1.b). Finally, when Avicenna establishes that the human soul is in essence a separate substance, it becomes clear that the essence of the soul is in fact identical to intellectual powers, and more specifically to the material intellect (}1.c).

a. The Definition of the Soul: Powers as Constituents In the opening lines of his Psychology, Avicenna defines the soul as “anything that is a principle of emanation of actions.”7 And insofar as the soul is thus the source of emanating actions, it is power: “it is right to call the soul ‘power’ as related to acts that emanate from it.”8 First, we must note that, with this opening move, Avicenna immediately takes a step back from Aristotle’s restrictive definition of the soul as perfection of a body. Avicenna’s opening definition is indeed more comprehensive and allows one to conceive of a soul as a separate substance. For a separate substance may well initiate outward operations, and be a ‘soul’ in Avicenna’s sense, although it is not the perfection of a body, a ‘soul’ in Aristotle’s sense. Avicenna will thus prove in book V that, because humans are defined by the activity of rational thought, the human soul must be a separate substance.

7

Psychology, I.1, 4. I am responsible for all translations. Psychology I.1, 6. For an account of Avicenna’s opening definition of the soul, see Jon McGinnis, Avicenna (Oxford: Oxford University Press, 2010), 89–95. 8

92

Charles Ehret

Second, we must note that, when he defines the soul, Avicenna says that actions proceed from the soul qua power, not that powers proceed from the soul qua something more fundamental than power. This last idea (that powers proceed from the soul) will appear further down in the treatise. So, as far as powers are concerned, there are not one, but two types of emanation in Avicennian psychology: powers flow from the soul and operations emanate from the soul’s powers.9 To these corresponds a rigorous distinction in terminology. To indicate the emanation of operations, Avicenna uses ‘s.udūr,’ a non-technical term, which refers to the carrying out of an action by an acting agent. To indicate the flowing of powers, he uses ‘fayd.,’ which originally refers to the overflowing of water, and is the main technical term Avicenna uses to refer to the metaphysical flow that originates from God, and from which separate intelligences, intellectual forms, and the soul itself all emanate. I will translate ‘s.udūr’ as ‘emanation,’ and ‘fayd. ’ as ‘flow.’10 This terminological distinction will be of particular importance to contrast Avicenna’s own doctrine with Albert’s understanding of Avicenna’s doctrine, since the Latin translation of Avicenna’s De anima renders both ‘s.udūr’ and ‘fayd. ’ as ‘emanatio,’ allowing for a confusion between emanating operations and flowing powers. Now, Avicenna’s first attempt at a definition of the soul, according to which it is the source of emanating actions, falls short. The reason is that all the powers of the soul do not emanate actions: some receive forms, sensible or intellectual. So the soul cannot be properly defined as ‘power’ without equivocation: “the meaning [of the word ‘power’] does not encompass the essence of the soul inasmuch as it is soul unconditionally.”11 It might be thought that the soul should be defined as a form. But Avicenna also rejects ‘form’ as a proper definiens, because form is correlative to matter.12 Therefore it cannot account for intellectual operations and it

9 At a more fundamental level, the soul itself flows from the separate causes. This is the most studied aspect of Avicenna’s emanative psychology; see Olga Lizzini, “L’âme chez Avicenne: quelques remarques autour de son statut épistémologique et de son fondement métaphysique,” Documenti e studi sulla tradizione filosofica medievale 21 (2010), 223–41; for emphasis on the soul’s celestial provenance, see Michael E. Marmura, “Some Questions Regarding Avicenna’s Theory of the Temporal Origination of the Human Rational Soul,” Arabic Sciences and Philosophy 18 (2008), 121–38; for emphasis on the soul’s relation to the body, see Thérèse-Anne Druart, “The Human Soul’s Individuation and its Survival after the Body’s Death: Avicenna on the Causal Relation between Body and Soul,” Arabic Sciences and Philosophy 10 (2000), 259–73. 10 For detailed analyses of Avicenna’s emanative terminology, see Lizzini, Fluxus, 546–70 and Jules Janssens, “Creation and Emanation in Ibn Sînâ,” Documenti e studi sulla tradizione filosofica medievale 8 (1997), 455–77. 11 12 Psychology I.1, 8. Psychology I.1, 7.

The Flow of Powers

93

rules out a priori that a soul may be a separate substance, an option Avicenna wants to keep alive. To both ‘power’ and ‘form,’ Avicenna prefers ‘perfection,’ which captures the soul as what gives a being its species.13 Perfection as actualizing principle of being is thus chosen (a) over form qua actuality of a body and (b) over power qua principle of operation or reception. Contrary to form, perfection can make an immaterial being what it is, and, contrary to power, it covers both productive and receptive grounds: an eagle’s being an eagle explains both the way it moves and the way it sees.14 One may feel, however, that perfection is undescriptive. Avicenna indeed writes that “perfection means that which, by its existence, makes the animal become animal in actuality.”15 But to say the soul of a dog is what makes this animal a member of the species dog does not give much information— less, in any case, than to say that the dog’s soul is its power to bark, bite, and scent. This feeling is comforted by the fact that power is not abandoned, nor is emanation. Indeed, emanation resurfaces only a few pages after being dismissed, in a reworked version of Aristotle’s definition of the soul: “the perfection of the natural body from which emanate its secondary perfections.”16 These “secondary perfections” have just been defined as being “among [something’s] actions or passions.”17 If the soul is thus understood as primary perfection in relation to secondary perfections, which emanate from it and are either actions or passions, it does seem that soul is still defined as power, i.e., as principle of operation. ‘Perfection,’ therefore, is not the name for something more fundamental than power. Rather, it is a comprehensive name for power. This is not the case with form, which is essentially related to matter, and thus exclusive of immaterial operations and separate souls. ‘Perfection’ covers more than ‘form,’ but it does not cover more than ‘power.’ ‘Power’ might be imprecise, but it is not restrictive. If this is true, Avicenna’s opening moves do not define the soul as perfection rather than power or form, but as power or perfection rather than form.18 13

Psychology I.1, 7. Psychology I.1, 8: “the soul, in regard to the power by which is perfected the animal’s perception, is perfection, and in regard to the power from which actions proceed, it is also perfection.” 15 16 Psychology I.1, 8. Psychology I.1, 12. 17 Psychology I.1, 11: “A secondary perfection is something that, with respect to its acts or its passions, follows the thing’s species, like cutting for the sword, or discriminating, understanding, sensation and motion for man.” 18 For the general meaning of ‘perfection’ in Psychology I.1, see Meryem Sebti, “La signification de la définition avicennienne de l’âme comme ‘perfection première d’un 14

94

Charles Ehret

This is confirmed by a massive textual fact: the problem of the soul’s unity, raised by the variety of its operations and powers, runs along the whole treatise, framing its entire development. First raised in anticipation after the soul’s definition is settled, in I.3, it is the main issue of the penultimate chapter V.7. The equivocation held against ‘power’ in I.1 has not been dispelled by ‘perfection’: the diversity of powers continues to be an issue, and a major one. Although the soul might be best defined as perfection, it is still constituted by powers, which require a stronger unity than a comprehensive definition.

b. The Soul as a Whole: Interconnected Powers When defining the soul in I.1, Avicenna explains that it is constituted by powers of two types, active and receptive, which can all be understood as constituting the perfection of a being, i.e., what makes a given being, material or immaterial, what it is. In V.7, however, where Avicenna addresses the problem of the soul’s unity, he seems to picture powers as extrinsic to the soul’s essence. Indeed, he starts by giving the three available options on the issue, favoring the last one, according to which: The soul is one essence; from it these powers flow, and each power has its own action, and it [i.e., the soul] does what it does, concerning what was just mentioned, only by these intermediate powers.19

According to the way Avicenna presents this option, the unity of the soul stems from the unity of the soul’s essence, from which its various powers flow. The flowing of powers means that they are related to the soul’s essence, which itself does not immediately carry out any action. Avicenna then argues this position. He does so by proving first that “all these powers must have a bond (ribaˉ t. ) that conjoins them all and in which they are all conjoined,”20 and second that this bond cannot be the body, so that it must be the soul. So Avicenna argues in favor of powers flowing from the soul’s essence by proving all powers are conjoined in the soul. Why? corps naturel organique’ dans le livre I du Traité de l’âme du Shifâ,” Bulletin d’études orientales 51 (1999), 299–312. The little attention the author gives to the relation between perfection and power, and her emphasis on the distinction between perfection and form, both confirm our reading. For historical background on Avicenna’s definition, see Fazlur Rahman, Avicenna’s Psychology: An English Translation of Kitâb al-Najât, book II, chapter VI (London: Oxford University Press, 1952), 3–12; for an in-depth analysis of how Aristotle’s original notion evolves, through the Greek and Arabic commentators, into Avicenna’s own notion of ‘perfection,’ see Robert Wisnovsky, Avicenna’s Metaphysics in Context (Ithaca, NY: Cornell University Press, 2003), 113–41. 19 20 Psychology V.7, 251. Psychology V.7, 253.

The Flow of Powers

95

To understand this, we must first note that powers flow from the soul’s essence, and not from the soul as a whole, of which all powers are parts. That powers flow from the soul’s essence does not preclude powers from being part of the soul as a whole, even though they are not part of its essence. Avicenna thus argues that the powers of the soul are all part of the same whole, which is what he means when he says they are all conjoined. And if we ask why these different parts do indeed form a whole, Avicenna answers that they all flow from one essence. So, for Avicenna, the flow of powers first expresses a connection between the powers of the soul and its essence, such that they are all part of the soul as a whole. However, he never mentions the essence of the soul when arguing for his position: he only proves a bond or connection between powers. The best way to make sense of this is to understand that the soul’s essence is itself constituted by certain powers, in such a way that from these powers some other powers flow. According to this reading, to say that powers flow from the soul’s essence is to say that some (unessential) powers flow from some other (essential) powers, which in turn is to say that powers are all bonded together. That Avicenna sees the soul’s essence as constituted by certain powers explains that he argues for the first proposition by proof of the second. This can be confirmed if we look at how Avicenna proves the second proposition, according to which powers are all bonded together. The point entirely rests on one simple psychological datum: power interaction. Let us say that I am thinking and, at the same time, focusing on what is appearing on my computer screen. Both my powers of thought and of sight are at work. But some unexpected image might pop up on the screen and interrupt my thoughts, sight thus interfering with intellection. To explain this, Avicenna gives the following account: If there was not a bond that uses [these powers], so that [sometimes] some are distracted by others, and therefore it [i.e., the bond] does not use those nor rule them, then some powers would not impede in any way the action of others, nor turn them away from it.21

Avicenna reasons thus: sight may distract and impede the intellect only because they are initially bonded. This initial bond means that when I am both thinking and focusing on my computer screen, my powers are actually cooperating. Now, this bond is a fragile one: it ‘uses’ powers, but ceases to do so as soon as one distracts or impedes others. So the bond does not itself control how powers interact. Rather, it describes the usual situation where some powers (e.g., the intellect) control others (e.g., sight), so that all

21

Psychology V.7, 253.

96

Charles Ehret

cooperate. This, says Avicenna, must be presupposed to account for the common experience where a power stops operating because of another (as when I see something that interrupts my thinking). Now, in this picture, there is nothing distinct over and above powers that accounts for their cooperation by ruling them all. Appealing to such a distinct essence would be not only pointless but counter-productive, implying that powers depend on this common origin rather than on each other. Power interaction implies a bond construed as a link between powers. For only such a bond may be broken, as Avicenna describes, by any one power breaking loose. If a link in a chain is broken, the chain is broken, because the chain just is the links. The same goes for the soul: its unity may be compromised by any one power, because the soul just is its powers.22 For further confirmation of our claim that the unity of the soul does not require an essence distinct from its powers, we can look at the early I.3 passage where Avicenna announces the solution to the unity-question we have just examined. There, he makes it explicit that only some powers, i.e., organic ones, are extrinsic to the soul’s essential unity: It shall be made clear later that the soul is one, and that these powers branch off (insha‘aba) from it into organs.23

As we shall shortly see, the same specification is made when Avicenna argues that the body cannot be the source of flowing powers. So one must not jump to conclusions when Avicenna says unqualifiedly that powers flow from the soul: only organic powers are in fact concerned, the other powers constituting the soul’s essence. At this point, Avicenna has established that all powers are conjoined in the soul, insofar as organic powers flow from the soul’s essence. The problem now is that organic powers are also in the body, where they are exercised.

22 The account of power interaction is more detailed in the Kitâb al-Najât (Rahman, Avicenna’s Psychology, 64–8). The problem there is to account for the truth of the statement ‘I perceived and became angry,’ considering that “the faculty of anger, insofar as it is the faculty of anger, does not perceive, and the faculty of perception, insofar as it is the faculty of perception, does not become angry.” Avicenna’s solution is to posit “a faculty by which it [i.e., the soul] is capable of combining both these things. This faculty, not being a physical one, must be the soul itself. Thus the substratum in which both these qualities [i.e., anger and perception] inhere is [ . . . ] the soul itself or the body inasmuch as it possesses soul, the combining substratum in the latter case really being the soul, which itself is the principle of all these faculties.” This passage confirms our reading, because the soul is itself a combining faculty, and because anger and perception both inhere in the soul itself: the soul is thus pictured as a whole containing the various faculties or powers, this whole existing qua whole in virtue of one particular power, namely a combining, non-organic power, the intellect. 23 Psychology I.3, 31.

The Flow of Powers

97

So Avicenna must prove that the link between powers is not, in fact, the body. In other words, Avicenna has presented a coherent picture where all (organic) powers are united in virtue of their common relation to the soul’s (intellectual) essence; he must now show that the rival picture, where all powers are united because they are exercised in one and the same body, is not a coherent one. This point is made with many arguments. But only one eliminates the body without qualification, and involves the flow of powers. It starts from the premise that there are corporeal and incorporeal powers, the former flowing into organs: What is not a body may be the source of powers; from it some flow into the organ, some belong to its essence, and all go [back] to it in a certain way.24

Avicenna is saying that only an incorporeal soul can account both for corporeal and incorporeal powers: the first flow, the second belong to its essence. The argument seems to be this: if the body were the principle uniting the powers, intellectual powers would be out on their own, so to speak, disconnected from organic powers, whereas if the soul’s essence, constituted by incorporeal powers, is the unifying principle, it can link all of them together. Here, it is made crystal clear that only organic powers flow, as opposed to non-organic powers, which belong to the soul’s essence. Now, as they flow from the soul, organic powers, which do not belong to the soul’s essence, are correlatively received in organs. So an organic power is extrinsically connected to the soul’s essence from which it flows, but it intrinsically belongs to the body to which it flows, i.e., is part of its essence. An organic power is therefore best conceived as a shared part: a non-essential (separable) part of the soul and an essential (inseparable) part of the organic body. And because the body is essentially determined by organic powers, it cannot be the source of these powers, which is what Avicenna wants to prove. He thus writes: It is impossible that all these powers flow from it [i.e., the body], for powers relate to the body not in the way of flowing, but in the way of reception; flowing can be in the way of separation of flowing from that from which it flows; reception, however, cannot be such a separation, nor can it be made thus.25

Flowing and receiving are incompatible. But consider a fountain: it shoots up into the air gushes of water, which fall back down into the basin. Why cannot powers likewise flow from the body and back into it? Avicenna says that flowing entails separation, while reception excludes it. This means that the body does not ‘receive’ the powers in the way the fountain basin ‘receives’ water: the powers received in the body are constitutive of what it 24

Psychology V.7, 254.

25

Psychology V.7, 254.

Charles Ehret

98

is, i.e., are essential to it qua organic. This is why separation, in this case, is disallowed, and why the ‘receiver’ may not be the source. To say otherwise would amount to a case of self-causation. According to Avicenna, then, it is necessary that powers flow from a distinct source into a distinct receiver. This is why he only ever says that organic powers flow: the non-organic powers of the soul lack a distinct receiver, and, precisely, flowing means ‘to be received in something else.’ If non-organic powers flowed from the soul’s essence, they would have to flow from and be received by the soul. But this is impossible, as it would amount to a case of self-causation, just as the materialist’s case for a body source of its own powers. So, just as the body cannot be the principle of its organic powers, the soul cannot be the source of its non-organic powers, which therefore cannot be said to flow from its essence.

c. The Soul as Substance: Intellectual Power Having described the different actions that proceed from the human soul, finishing with the activities of the theoretical and practical intellects, Avicenna closes Psychology V.1 by what is surely the most important passage in the whole Psychology on the relation between the soul and its powers. It kicks off with what seems to be a clear-cut distinction between the soul and its powers: The soul is what has these powers, and it is, as we have said, a separate substance, possessing a readiness for actions, some of which are only perfected by organs and by the fact that, in general, [the powers] engage in them; some others require organs [only] for a specific purpose, and, finally, some do not require them in any way.26

This passage is the main piece of evidence in favor of an ontological distinction between the soul and its powers in Avicenna.27 And indeed, it does say that the soul is not any of its powers, because each power, which is “a readiness for actions,” is possessed by the soul, the context making it perfectly clear that non-organic powers are also concerned. Avicenna’s aim, here, is to prove that the soul is a separate substance. The general idea is that a power is dependent upon the performing of an action, whereas the soul is a being per se, which exists as itself whatever actions are actually carried out. So, insofar as it exists independently from the fact that such or such action is carried out, the soul is something else than its powers. However, if we keep on reading, it becomes clear, once again, that not all powers are, in truth, concerned by this distinction. 26 27

Psychology V.1, 208. See Hasse, “Soul Faculties,” 306 (n. 1) and Sebti, L’âme humaine, 36–8.

The Flow of Powers

99

So the substance of the human soul is of itself ready to be perfected by some kind of perfection; concerning what is above it, it does not require what is below it. And this readiness it possesses is due to what is called the ‘theoretical intellect’; and it is [also] ready to take precautions against the harm that comes from [its] community [with the body], as we will show in due course, and to have, acting as it chooses, freedom of action on [its] community [with the body], according to its convenience. This readiness it holds from a power called the ‘practical intellect,’ and it is the main power that [the soul] possesses in [its] connection with the body.28

In the preceding text, “readiness for actions” was the defining characteristic of powers distinct from the soul itself, a separate substance disengaged from outward operation. But now it appears that the soul “is of itself ready to be perfected,” insofar as it is a substance. So the soul qua substance is not, after all, something ontologically distinct from power: it is also a readiness, which is precisely what defines a power. And, as the text goes on, it is made clear by Avicenna that the soul as substance is ontologically distinct only from the lower active powers. Positively, it is constituted qua readiness by a higher receptive power, the theoretical intellect, through which it can be perfected by receiving “that which is above it,” namely immaterial forms. Negatively, it can protect itself from bodily influences by the practical intellect, a “power” in connection with the body only to prevent the soul’s general connection with the body from being intrusive and harmful. Now the theoretical and the practical intellects are among the many powers of the soul. The soul, therefore, is constituted as a substance, i.e., as independent from the body, by its intellectual powers. So the soul is not distinct, as substance, from power as such: the distinction is between the intellectual powers that make it a substance, and the organic powers that accidentally make it a part of a composite. This is why only the latter can be pictured as flowing outside the soul. This is what Avicenna immediately points out: As regards the things below this [i.e., the practical intellect], they are powers that pour out (inba ʿat ̱a) from it, because of the body’s readiness to receive them and of the benefit it receives from them.29

Triggered by the body, powers flow from the practical intellect, which itself has just been called a ‘power.’ They do not flow from the soul’s essence as something distinct of all its powers. That powers flow does not imply an ontological distinction between the soul’s essence and its powers, it only implies that organic powers qua organic are outside the soul’s essence, as opposed to the intellectual powers that are constitutive of the soul’s essence, insofar as it is a separate substance.

28

Psychology V.1, 208–9.

29

Psychology V.1, 209.

100

Charles Ehret

However, this may seem to leave the unity question open. For one may still want to hold that intellectual powers, because they differ from one another, must be traced back to one ‘essence of the soul,’ distinct from its powers, all its powers, even intellectual. But this is not the case. Indeed, here is what Avicenna writes next: Each of these two [intellectual] powers has a readiness and a perfection. On both sides, the pure readiness is called ‘material intellect,’ whether considered as theoretical or practical. Then, after this, it happens only to each of these two that for each the principles occur by which it accomplishes its acts (for the theoretical intellect, the first premises and things of the sort; for the practical intellect, the probable propositions and other dispositions), and, at that moment, each of them becomes a habitual intellect, and then for each of them occurs the acquired perfection.30

Avicenna’s point, here, is that the contemplative and active intellect are only different ex parte obiecti, but are initially one power, since their readiness is the same, and is the material intellect. So, as far as intellectual powers go, there is no need for something else to unify them. And we have seen that the perfection of this one readiness is the perfection of the soul itself qua separate substance. So the human soul is in essence the material intellect, which names intellectual power as such and as one, i.e., prior to its reception of theoretical propositions (“the first premises and things of this sort”) and of contingent facts (“probable propositions”). The same point was made when Avicenna conjured up the image of the soul’s two faces: one that is turned upwards facing the higher principles, the other turned downwards and facing the body.31 An obvious point of the metaphor is that our soul is not split in two because of its two faces, just as a coin is not split in two between its head and tail. The problem of the soul’s unity is thus solved once we reach the intellectual powers that constitute the soul’s essence, since the different kind of objects the material intellect receives does not compromise its basic unity. There is thus no need to posit a mysterious essence of the soul underneath all powers to unite them. The soul simply is the material intellect, of which all other intellects, so to speak, are specifications, and from which all organic powers flow because of its relation to the body.32 30

31 Psychology V.1, 209. Psychology I.5, 47. For a contrary reading, see Meryem Sebti, “La distinction entre intellect pratique et intellect théorique dans la doctrine de l’âme humaine d’Avicenne,” Philosophie 77 (2003), 23–44. The author first acknowledges that Avicenna’s way of talking might lead us to believe that for him the rational soul just is the intellect (“une imprécision dans le vocabulaire peut laisser croire qu’Avicenne identifie l’âme rationnelle à l’intellect”); however, she goes on to prove that Avicenna actually distinguishes the material intellect from the soul. In order to achieve this, the author looks at the argument where Avicenna rejects that knowledge is the assimilation of the soul to the known form it receives 32

The Flow of Powers

101

Let us sum up. The soul, for Avicenna, is an array of powers, some receptive, some active, which can all be gathered and understood as a being’s perfection, i.e., what makes a being the being it is. In the case of human souls, some of these powers are organic, but some not. Insofar as powers are organic, they flow into various organs, from the soul’s essence. This means that they are all connected to that essence, and thus to each other. It also means that organic powers are non-essential parts of the soul, and that they are received in the body, of which they are, on the contrary, essential constituents. Now the human soul is in essence a separate substance, and what constitutes it as a separate substance are its non-organic, intellectual powers. By perfecting them, it perfects itself qua substance, freeing itself from its dependence on the body. Finally, these intellectual powers need not be united by reference to something else, because they are multiplied only by the multiple kind of objects that intellectual power is faced with. Initially, intellectual power is just one receptive power: the material intellect.33 (Psychology V.6, 240). Avicenna’s problem is this: if what receives is assimilated to what it receives, then it becomes unreceptive, and is destroyed; on the other hand, it is unconceivable that what receives does not change when receiving. Sebti’s reading (Sebti, L’âme humaine, 26) is that distinguishing the material intellect from the soul’s essence is a solution to this problem. The distinction would preserve the rational soul qua substance from destruction by assimilation, while allowing the material intellect qua distinct power to assimilate to the agent intellect: “l’affirmation de la transcendance de l’âme rationnelle vis-à-vis de ses puissances permet à Avicenne de soumettre son activité à l’intermittence de la jonction avec l’intellect agent tout en préservant sa substantialité.” However, the assimilation problem is exactly the same for the material intellect, as Avicenna carefully notes (Psychology V.6, 241): the material intellect must either be unchanging qua disposition of the soul, or destroyed qua assimilated to the known object. In no case can it be both, as an intellectual power seems to require. So distinguishing the rational soul from the material intellect is of no use, as it just shifts the problem from here to there. The solution lies in Avicenna’s later refusal of intellectual memory, i.e., his refusal of the thesis that to know is to receive and acquire forms, to which he substitutes the thesis that to know is to acquire a power, namely the power to join with the agent intellect, where forms are stored (Psychology V.6, 247). Now, because to acquire such a power is to learn and thus be perfected, this power obviously does not proceed from our own soul, in which case we would be perfected from the start, and would never learn anything. Knowledge qua power thus comes from the agent intellect, and Avicenna writes that this intellectual power flows from the agent intellect (Psychology V.6, 247). This last point confirms that the idea of intellectual powers flowing from the soul is alien to Avicenna, and, in fact, goes head on against his general metaphysical picture, in which what is intellectual flows to the soul, not from it. 33 For a formulation that concurs with my reading, see Robert E. Hall, “Intellect, Soul and Body in Ibn Sînâ: Systematic Synthesis and Development of the Aristotelian, Neoplatonic and Galenic Theories,” in J. McGinnis (ed.), Interpreting Avicenna: Science and Philosophy in Medieval Islam (Leiden: Brill, 2004), 62–88, 65: “According to Ibn Sînâ, the Active Intellect emanates the human soul, which is essentially the rational soul, to the foetus that becomes its particular body. There it emanates its own sub-rational (animal, vegetative) faculties.”

102

Charles Ehret

The idea that the essence of the soul is ontologically distinct from its powers as such is simply nowhere to be found in Avicenna. Admittedly, there are many suggestions that the soul is distinct from each of its powers. But when we look closely at Avicenna’s arguments, it always appears that he is only distinguishing between the soul’s essence and its organic powers, this implying that its intellectual powers make up its essence. And indeed, if we look at what the soul essentially is, we see that it is a substance, a being per se independent of the body, which it can only be as a material intellect receiving immaterial content. As a matter of textual fact, Avicenna does not say that the intellect flows from the essence of the soul, and as a matter of doctrine, Avicenna could not have said so, because, for him, to flow from something means to be received in something else, and there is nothing else intellectual powers could flow to if they flowed from the soul. If we are nonetheless prone to mistakenly read into Avicenna’s psychology a distinction between the essence of the soul and all its powers, intellect included, it is because we perceive it in the light of its Latin reception, and notably of its main promoter and most in-depth reader, Albert the Great.34

2 . AL B E R T Albert uses Avicennian material most extensively and explicitly in his early De homine. There, he first deals with the soul’s existence (q. 1), then with its definition (qq. 2–4). In q. 4, Albert explains Aristotle’s definition of the soul. As he does, he confidently defers to Avicenna, and intertwines his reading of Aristotle with the opening lines of the Persian De anima, giving emanation a prominent role. This synthetic effort makes Albert misread Avicenna’s emanation (}2.a), a misreading that will lead to all powers flowing from the soul, whether or not the body is present (}2.b). I will show that Albert makes all powers flow in order to have the body actualized by the whole set of powers, intellect included, while preserving the separate status of their source (}2.c).

a. From Emanating Operations to Emanating Powers For Albert, as for Avicenna, the human soul is in essence a separate substance. In De homine q. 4 a. 1, Albert starts explaining Aristotle’s definition of the soul, and answers the question “how is the soul an actuality?” An important 34 On Albert’s preeminence among Latin readers of Avicenna’s De anima, see Dag Nikolaus Hasse, Avicenna’s De Anima in the Latin West: The Formation of a Peripatetic Philosophy of the Soul 1160–1300 (London: The Warburg Institute, 2000), 60.

The Flow of Powers

103

danger here, in Albert’s view, is to think that the soul is no more than the perfection of a body, because then it would have to cease to exist as soon as the body ceases to exist.35 In order to avoid such a consequence, Albert must make Aristotle’s ‘actuality’ cover the case of a separate substance, such as the incorruptible human soul. To this end, Albert uses Avicenna’s definition of the soul as a ‘principle of emanation.’ Albert thus understands ‘principle of emanation’ as equivalent to actuality, and Avicenna’s definition as giving the substance of Aristotle’s: . . . when the soul is defined as it is by Aristotle, “its being is asserted only as principle of emanation from itself of affections which are not of one kind and are voluntary.”36

Roughly, this means that the soul is not the actuality of a body in itself, but only insofar as it emanates “affections.” A quick note immediately follows the solution: “ ‘perfection’ and ‘actuality’ are the same according to the commentators.”37 So for Albert, ‘principle of emanation’ makes explicit how soul is actuality, and ‘actuality’ is synonymous with ‘perfection.’ Albert thus implicitly writes the following equation: actuality = perfection = principle of emanation From an Avicennian perspective, this is wrong. For as we have seen, soon after proposing his opening definition, Avicenna shows that ‘principle of emanation’ is in fact an incomplete definition of the soul. He then shows how the soul is more than a principle of emanation, because it is also a principle of reception. When Avicenna redefines the soul as ‘perfection,’ he is giving a comprehensive term that covers both active and passive powers. The soul as perfection is thus inclusive of powers, only some of which are principles of emanating operations. Albert, however, as he immediately equates it with actuality and perfection, understands ‘principle of emanation’ as a complete definition of the soul. Because of this, the idea that ‘power’ is a deficient definiens suggests that Avicenna’s emanating principle is something more fundamental than active power. And in turn this distinction between the emanative source and power naturally brings to mind that powers are what the source is emanating, an idea alien to the text of Avicenna. This shift—from active power emanating operations (in Avicenna) to the soul emanating powers (in Albert)—is confirmed when Albert examines power as a candidate for 35

DH q. 4 a. 1 (5), 30. DH q. 4 a. 1 (c.), 31; for Albert’s quote, see DA I, 1, 15. The Latin ‘affectiones’ correspond to ‘af ʿaˉ l ’ in Avicenna’s Arabic (Psychology I.1, 4, translated at n. 7), which means ‘actions’ rather than ‘affections.’ 37 DH q. 4 a. 1 (1–2), 31. 36

104

Charles Ehret

defining the soul. Indeed, after giving Avicenna’s argument that ‘power’ is equivocal, as it refers either to active powers or to passive powers, Albert adds another argument of his own, which goes much further than Avicenna’s: it shows powers not only to be divided between active and passive ones, but as being derived from the soul’s essence. It is possible to supply another reason for the same [thesis],38 because capacity is caused by matter or form, and especially capacity to operate, and so capacity, faculty or power (potentia vel vis sive virtus)39 are amongst what follows being, and because of this, since the soul is here defined insofar as it is the primary actuality which makes [something] be, it cannot be defined by the term ‘capacity.’40

Admittedly, Albert makes no mention of emanation at this point. Clearly, however, saying power “follows being” and thus follows “the primary actuality which makes [something] be,” namely perfection, implies, according to his equation, that power follows from the soul qua emanative principle. All is now set to say that the soul is the source of emanating powers. This stage-setting had in truth begun as soon as Albert introduced Avicenna’s emanating principle in the solution of q. 4 a. 1. Indeed, what emanates from the soul had not been defined: all we had were obscure “affections” (affectiones), a term Albert cuts off from Avicenna’s illustrating examples, never picks up, never explains, and which thus acts as a mere placeholder for the forthcoming powers. So it seems that, from the start, Albert reads Avicenna’s opening definition as meaning that the soul is the source of emanating powers, even though Avicenna is saying something significantly different. Why such a misunderstanding? It might very well be prompted by the translation Albert is working with. Indeed, as noted above, the Latin De anima renders both ‘s.udūr’ (the process of carrying out actions) and ‘fayd. ’ (the ontological flow) as ‘emanatio.’41 My hypothesis is that the partial 38 The Cologne edition understands “alia posset assignari ratio eiusdem” as “it is possible to supply another reason from the same [author],” and indicates as Albert’s source Avicenna Latinus, Philosophia prima IV.2, 195–6. However, the cited text has very little in common with the passage, and seems highly implausible as a source for Albert’s argument. My reading is that eiusdem means ‘for the same thesis,’ and that Albert is supplying an argument of his own against ‘power’ as definiens. For an unequivocal use of the same phrase, in the sense I understand it (no author is mentioned), see for instance Aquinas, De veritate q. 13 a. 3, co. 39 Because the Latin terms potentia, virtus, and vis are synonyms, I will uniformly speak of ‘power’; for the sake of accuracy, however, I will translate these terms, respectively, as ‘capacity,’ ‘power,’ and ‘faculty.’ 40 DH q. 4 a. 1 (10), 34. 41 The Latin translator himself seems to have misunderstood Avicenna’s psychological flow, and to have thought, as Albert did, that it concerns all powers, intellectual ones included. Indeed, when Avicenna writes that the power to join the agent intellect flows

The Flow of Powers

105

definition of the soul as principle of emanation (s.udūr), in its Latin translation, brings to Albert’s mind Avicenna’s later talk of flow (fayd. ), according to which powers flow from the soul into organs. Because he believes he is dealing with the same concept, Albert converts a definition of the soul as what initiates actions into a definition of the soul as what emanates powers. This would be why, in dismissing ‘power’ as an adequate definiens for the soul, he does not simply hold, as Avicenna, that ‘power’ is equivocal, but claims that powers are derivative of the soul’s essence. And this, in turn, would have led Albert to hold that powers as such flow from the soul’s essence, so that they must all flow, even though Avicenna, who he claims to be following, actually limits flow to organic powers. Now loss in translation might well have triggered Albert’s thesis that all powers flow from the soul, but it obviously does not give his philosophical motivation for it. Before asking what it might be, though, we must first see how he actually develops it.

b. Emanation Disembodied Because, unlike Avicenna, he keeps on board Aristotle’s definition of the soul, Albert must face a crucial objection: from a soul that is actus corporis cannot follow an intellectual operation, which is sine corporis. To answer, Albert uses his ‘actuality of the body’ = ‘principle of emanation’ equation, and it is now made explicit that powers are what emanate from this emanative principle: If the rational soul is taken as soul, then it will be the actuality of a body, from which emanate some powers affixed to organs and some that are not affixed. And that this presents no inconvenience will be shown later.42

The objector complained that an immaterial operation cannot come from a material soul (from the soul qua actuality of the body). Albert replies that the immaterial operation of intellection comes from an immaterial power— which in turn, however, comes from a material soul. The objector will now want to know how non-organic powers can themselves emanate from the soul qua actuality of a body. Albert is aware he is just pushing back the problem here, and so reassures: there is no inconvenience. But he will explain later. from the agent intellect (see n. 32 in fine), the Latin version has this intellectual power emanate from the soul itself (see DA 149). The De anima’s confusion is not in the Latin translation of the Ilâhiyyât (Philosophia prima), which systematically gives fluxus for fayd. and emanatio for .sudūr. 42 DH q. 4 a. 2 (7), 37.

106

Charles Ehret

This postponing is not without reason: Albert simply cannot explain how an immaterial power comes from soul qua actuality of a body, as the question requires. To account for intellectual power, he must wait until he has finished dealing with Aristotle’s definition, for an opportunity to consider the soul qua substance. This opportunity comes when Albert asks whether the vegetative, the sensitive, and the rational are one or different substances (q. 7 a. 1). The major objection here is that what is rational is incorruptible, whereas what is vegetative or sensitive is not, so that the ‘rational soul,’ namely the human soul, cannot be a single substance, since such a substance would have to be both corruptible and incorruptible. Albert replies: Of what is corruptible and incorruptible per se, in itself, and in the same sense, there is never one substance, and thus the rational soul is not corruptible and incorruptible. Indeed, the rational soul is one substance existing only incorruptibly, from which flow some powers that operate in bodily organs and some that operate without bodily organs.43

Powers flow from “one substance existing only incorruptibly”: the formula is strikingly powerful—as well as off the expected mark. Indeed, until now, Albert has always said, as Avicenna, that the soul is a principle of emanation in its relation to the body, as actus corporis. But here he shifts to the soul qua substance, and says that the soul is an emanating principle independently of its relation to the body. The reason for this shift is that Albert has understood Avicenna’s incomplete “emanating principle” as a complete definition of the soul, so that all powers now flow, non-organic ones included. Because of this, Albert cannot understand emanation as a process that simply describes the way the soul diffuses throughout the body, as is the case in Avicenna. Albert must understand emanation as a process that concerns the soul at a more fundamental level, namely as a separate substance. This shift—from the soul emanating powers in its relation with the body to the soul emanating powers qua separate substance—had already been made in one of the arguments in favor of the soul’s unity. There, Albert’s own position is presented as an Avicennian doctrine: Avicenna says in the De anima that from a soul that is one according to substance emanate various powers, some of which are tied with the body, like sense perception, and some not, like the intellect.44

43

DH q. 7 a. 1 (1), 85. DH q. 7 a. 1, 82. For Albert’s general idea that both organic and non-organic powers emanate, and its Avicennian credentials, see Dag Nikolaus Hasse, “The Early Albertus Magnus and his Arabic Sources on the Theory of the Soul,” Vivarium 46 (2008), 232–52. The author gives as Albert’s sources on this point the passages quoted in n. 23 and n. 26. However, as we have insisted, the first passage only says powers branch off into 44

The Flow of Powers

107

Avicenna never goes so far. As I have argued, the flow of powers refers in his Psychology to the soul imparting the body with organic powers: as such, it does not pertain to the soul qua substance, which is constituted by intellectual power, which does not flow. Now Albert’s misquote is not a one-off slip of the tongue: almost each time he writes powers flow, Albert makes it clear that both organic and non-organic ones do, and nearly always attributes the idea to Avicenna.45 So, Albert’s distortion of Avicenna in the opening of the De homine becomes a firmly held doctrine, one that drives Albert’s general understanding of the soul–powers relation, and one that makes Albert repeatedly misread a key point of Avicenna’s psychology. Now the immediate point of this misreading is to unite all powers. And Albert does so, spectacularly, by making all powers incorruptible, with help from an Aristotelian example: Therefore corruption is not on the side of the soul, but on the side of organs, and it is the corruption of operation, not of operating power. [ . . . ] Indeed, this corruption happens on the side of the body and not on the side of the parts of the soul, as also the feebleness of operation comes from the feebleness of the organ and not from the feebleness of the organ’s power. For the philosopher says that if an old man were given a young man’s eye, he would see as a young man.46 (DA I.4, 408b21–22)

Albert understands Aristotle as saying that the old man, although he cannot see because his eye is corrupted, still possesses the power of sight, in such way that, given a new eye, he would see anew; this means that the power of sight, although it is exercised through a corruptible organ, is not itself corruptible, as is true for all the powers of the human soul. This general incorruptibility of powers implies that even separated from its material counterpart, the soul remains with all its powers proceeding from it. As Thérèse Bonin writes, “the soul will overflow into its powers whether or not

organs, and the second does not speak of emanation, which is mentioned only later (n. 29), where it concerns, once more, only organic powers. 45 See DH q. 78 a. u., 572; Super ethica 1.8, 38, as well as 1.15, 81 and 6.2, 407; De anima 3.2.12, 193; De causis 2.3.16, 153; Summa theologiae I, 3.13.1 (9), 39. More rarely, we can find Albert ascribing emanation simply to organic powers (De anima 3.5.4, 249). For a broader list that includes passages where Albert simply states that powers flow, see Hasse, Avicenna’s De anima, 239, from which I have borrowed much. 46 DH q. 7 a. 1 (1), 85. It is not clear how the putative case should be read in Aristotle himself: it obviously has no empirical ground, and, because it implies that all the soul is immortal (as Albert is happy to accept), it seems to imply in turn that all souls are immortal (something Albert would certainly not accept). This is what Aquinas thinks, and the reason why he contextualizes the passage by saying Aristotle is borrowing from a famous opinion of his time, according to which every soul is immortal (Sent. libri De an. I, l.10). According to Albert’s pupil, then, the blind man’s case should not be taken at face value, as in Albert.

108

Charles Ehret

it still has his body.”47 This prompts Albert’s way of talking about the relation between organic powers and their corresponding organs, to which they are “affixed” (affixae) or have a “tie” (colligantia), this prudently indicating that the powers themselves are not organic, nor corporeal: they exist, subsist, and emanate because of the soul’s essence, not because of the soul’s relation to the body. For Avicenna, powers flow in the body, and flow because the body is present. For Albert, powers flow to the body, but flow whether or not the body is there. In his 1943 article, Étienne Gilson argues that general incorruptibility is what drives Albert in saying so often that powers flow from the soul: Albert would want the human soul, after the body’s demise, to remain with all its parts.48 This surely is an implication, and maybe a desirable implication, of Albert’s doctrine. I do not think, however, that it can hold as an explanation. For it only explains that organic powers flow independently of the body. But if Albert sees organic powers flowing independently of the soul’s actual relation to the body, it is only because he sees non-organic powers flowing from the soul, and thus sees the emanation of powers in general as a disembodied process. And there would be no sense in thinking that Albert has non-organic powers flowing from the soul in order to have them subsist when the soul is deprived of its material counterpart (as Gilson’s explanation would have it), since non-organic powers are in any case independent of matter. So Albert’s emanation is still in need of an explanation.

c. The Set of Emanated Powers We have seen how Albert disorganizes Avicenna’s flow of powers, cutting them off from their organic destination, so that the flow of powers has become a purely psychological process, powers proceeding from the soul without emanation being triggered by the soul’s relation to an existing body. We must now ask why. Why does Albert claim, repeatedly, that nonorganic powers flow, and does so in the name of Avicenna’s De anima, where Albert cannot explicitly read such a claim? What is his philosophical motivation for reading it into Avicenna’s text? My answer is that, by setting, on one side, a purely incorruptible substance and, on the other, all the powers, intellect included, Albert’s aim is to 47 Bonin, “Emanative Psychology,” 53. The claim is made with reference to DH q. 8 a. 2 (4), 96. 48 Étienne Gilson, “L’âme raisonnable chez Albert le Grand,” Archives d’histoire doctrinale et littéraire du Moyen Âge 14 (1943), 1–80, 55 (our emphasis): “Pourquoi saint Albert met-il cette insistance à relier les virtutes, vires, ou potentiae, à l’âme dont elles découlent? Pour assurer l’immortalité totale de l’âme lors de la mort du corps.”

The Flow of Powers

109

involve the intellect in the soul’s relation to the body. Because organic powers flow with the intellect, they are as incorruptible as the intellect; but what drives Albert is the converse truth that, because the intellect flows with organic powers, it actuates the body as do organic powers. Power emanation is thus a piece of a constant and developed doctrine of Albert’s, according to which the rational soul is qua rational the immediate actuality of the body. This immediacy is vigorously upheld by Albert, who claims that the contrary would be “against all philosophy and reason,” since “nothing is more immediate to something than that from which it holds its being and perfection.”49 Albert’s problem, on this point, is that he cannot consistently say that the rational soul is the immediate actuality of the body, while still maintaining that it is a separate substance that exists only incorruptibly. His way out is to have intellectual powers flow along with organic ones, and to have the set of emanated powers actualize the body: because a member of this set is the intellective power, the whole set may be referred to as the ‘rational soul,’ in the Aristotelian sense. Albert can thus say, referring to the set of powers, that the immediate actuality of the body is the ‘rational soul,’ but because all the powers are emanated from an immaterial substance, he can also say the ‘rational soul’ is in essence separate and incorruptible. ‘Rational soul’ is thus equivocal. For the sake of clarity, I will only use the phrase in reference to what actualizes the body, and which must be identified as the set of emanated powers. The rational soul is composed of three main parts: the rational, sensitive, and vegetative powers. So it might seem that the rational soul is a whole composed of three mutually exclusive parts. If one sees things this way, one will hold that the rational soul actualizes the body only through its sensitive and vegetative powers, the intellect being left out, as it does not require any bodily organ for its operation. But if this is the case, then the rational soul does not immediately actualize the body, because it does not do so qua rational, but only qua sensitive and vegetative. This is Philip the Chancellor’s and Pseudo-Robert Grosseteste’s doctrine, which Albert emphatically rejects.50 To hold that the rational soul is as such the actuality of the body, Albert must hold that the powers are not mutually exclusive parts of a distinct whole. And indeed, for him, the powers of the soul are like Russian dolls: lower powers are contained in higher powers. So if you look at the highest of all powers, you are actually looking at the whole set. This view is framed by

49

DH q. 4 a. 1 (8), 33.

50

DH q. 4 a. 1 (8), 33.

110

Charles Ehret

the Aristotelian analogy of De anima 414b20–32, between powers of the soul and figures of geometry: Any species of soul is the actuality of a body. Indeed, the rational soul has in itself the vegetative and sensitive power, according to which it is itself the actuality of a body. In the same way a triangle is in a rectangle [as] power and not by its own being, thus the vegetative and sensitive are in the rational [as] power without specific being except that of the rational soul.51

The important point here is that, on this view, there is no way of distinguishing the highest power, the intellective power, in which all lower powers are contained, and the rational soul as a whole. So the immediate actuality of the body is in fact the intellect. Albert discreetly but strikingly makes this point, as he continues: However, there is no need that, because the rational soul is the actuality of a body, each of its powers be attached to the body. And thus says the Philosopher “nothing prevents some parts of the soul of being separate because they aren’t the actuality of any body,” that is, of any body part.52

The passage is striking, because it reads Aristotle as only referring to body parts, where Aristotle is, quite unambiguously, referring to the body simpliciter. This tinkering is a clear sign that Albert is making a point of his own. And the point is that the part of the soul unattached to the body, although it is not the actuality of any body part, still is the actuality of the body, that is, of the body as a whole. This means that the intellective power is not, strictly speaking, a part of the soul. Rather, it is the soul as a whole, in the Aristotelian sense of what actualizes the body. So the intellect is not a distinct part of a rational soul actualizing the body by its other parts, the organic powers. If such were the case, the rational soul would not be the actuality of the body qua rational, as Albert firmly wants it to be. To this end, he holds that the intellect contains the sensitive and vegetative powers, so that the human body is actuated by an intellective soul, which makes it a human body. Now I believe this is the philosophical reason why Albert sees all powers emanating from the soul’s essence: organic powers must be embedded in the intellect for the human body to be a human body. If only organic powers flowed, as in Avicenna, and if, as in Avicenna, the intellectual powers belonged exclusively to the soul as a separate substance, then the human body would not be a human body, it would be a remote-controlled animal. This is what Albert wants to avoid, 51 DH q. 4 a. 1 (8), 32–3. On Aristotle’s analogy, see Julie K. Ward, “Souls and Figures: Defining the Soul in De anima ii 3,” Ancient Philosophy 16 (1996), 113–28. 52 DH q. 4 a. 1 (8), 33.

The Flow of Powers

111

and this is why he has non-organic powers flowing alongside organic powers, so that the actuality of the body may be described as the ‘rational soul.’ So the picture in Albert is this: there is a separate, wholly incorruptible substance, from which emanates a set of powers, organic and non-organic. Organic powers are all contained in the highest, non-organic power, which is the intellect. The intellect, as it contains the lower organic powers, is the immediate actuality of the human body, and as such is referred to as the intellective or rational soul, in a strictly Aristotelian sense. And although the emanated powers constitute the actuality of the corruptible body to which they attach, they do not cease to flow when the body is gone, because their source is immaterial and incorruptible, and flows as such, not because it relates to an existing body, which would only trigger the emanation of organic powers.

3 . F R O M AL B E R T T O A Q U I N A S

a. A Bare Substratum? Is Albert’s emanative psychology, when fully developed, an intelligible doctrine? I don’t think so. The problem I see is this: because the soul as source is distinct from all its powers, it appears as a bare substratum of sorts, i.e., something that cannot be described except by reference to what is supposedly distinct from it, namely the powers that it emanates. It indeed seems impossible to predicate anything of the soul’s essence, as Albert conceives it, without identifying it with one of its powers: the human soul is ‘intellectual’ insofar as it is considered with its intellectual power, ‘sensitive’ insofar as it is considered with its sensitive power, and so forth. Nothing, it seems, can relevantly be predicated of the soul taken in itself. And if there is no way to describe the soul qua distinct from its powers, there is no way to understand the distinction itself, which ends up collapsing. An account of Albert’s psychology will thus be tempted to leave emanation aside, and disregard the flow of powers, insofar as it implies a distinction between the soul’s essence and its powers, as an obscure, unintelligible doctrine. This is precisely what Gilson does. He acknowledges that Albert posits a distinction between the soul and its powers, but stresses that Albert never really explains what this distinction amounts to.53 Because 53 Gilson, “L’âme raisonnable,” 53–7: “Il faut donc dire avec lui que l’âme est distincte de ses facultés, et que les facultés ont par conséquent une existence propre. De quelle distinction, et de quel mode d’existence s’agit-il? C’est un point sur lequel il est assez

112

Charles Ehret

there is no way of grasping the soul apart from its powers, there is no real sense in picturing the soul’s essence behind them, as their source. So we are left with a separate substance that just is a set of powers, all contained in the highest one, the intellect. In this reading, Albert is a perfect Platonist, for whom the human soul is an incorruptible and intellectual substance that simply moves a separate body.54 A response to this criticism, and a way to avoid dropping emanation from Albert’s psychology, as Gilson does, would be to say that, although the soul is indeed distinct from its emanated powers, it can nonetheless be truly described by them, since for Albert the soul is their subject.55 The soul would thus be a substance distinct from its powers, but an intellectual substance nonetheless, because it is the subject of the intellectual power of which it is the source. But this answer only makes things worse. To say that the soul is an intellectual substance in virtue of its emanated intellectual power requires that this intellectual power be both emanated and received by the soul itself. Avicenna had warned us against this: he argued that the body could not be the source of organic powers, because it is itself organic. For the same reason, the soul cannot emanate intellectual powers and be intellectual in virtue of these same powers. To say otherwise would amount to a case of selfcausation. And, in the end, it seems that Albert’s emanative psychology amounts to just that. Another defense of Albert’s doctrine might insist that the soul is the source of powers prior to being their subject. Talking of a ‘bare substratum’ would thus be misleading: if the soul were a bare substratum, the argument goes, nothing would flow from it. So we should hold that intellectual powers flow from the soul precisely because the soul is in the first place an intellectual substance. Now it must be stressed that this would give ‘flowing’ a meaning perfectly alien to its original meaning in Avicenna. In Avicenna, if what flows is p, then its source is not p: the flowing of organic powers from the human soul means that it is not itself organic. If you give up on this to hold that an

difficile d’obtenir des precisions [ . . . ]. Il est diffile d’obtenir de notre philosophe qu’il aille plus loin sur ce point et nous dise quel genre de distinction il reconnaît entre le pouvoir central de l’âme et ces ‘potentiae naturales ab ipsa fluentes, quae dicuntur proprietates.’ ” 54 Gilson, “L’âme raisonnable,” 26: “Nous revenons ainsi à la formule platonicienne classique: substantia incorporea movens corpus.” On the Albertinian soul as mover, and for a recent account very close to Gilson’s, see Steven Baldner, “St. Albert the Great on the Union of the Human Soul and Body,” American Catholic Philosophical Quarterly 70 (1996), 108–17. 55 Summa theologiae II, tr. 12, q. 70 c.; see Gilson, “L’âme raisonnable,” 57.

The Flow of Powers

113

intellectual power emanates from an intellectual substance, you give up Avicenna’s notion of flow. Could this nonetheless be a solution to save Albert’s emanative psychology? I do not think so. If you hold that the emanated intellectual powers are the outward manifestation of something essentially intellectual, then you must also hold that, because organic powers flow from the soul, the soul is something essentially organic, so that its relation to the body will be part of its essence. In other terms, if emanation is the unraveling of an essence, and if both organic and non-organic powers flow, then the human soul can indeed be posited as something intellectual, but it cannot be an intellectual substance, and must be essentially related to the body, hence the part of a composite. This, of course, Albert could never accept, because he is committed to the idea that the human soul is not essentially related to the body, and because this commitment is what motivates his appeal to emanation in the first place. The idea that the human soul is part of a composite substance is an idea Albert’s emanative psychology is precisely built to resist. On the contrary, it is an idea that Aquinas will come to embrace. Consequently, Aquinas can say what Albert cannot, namely, that all powers flow from the soul’s essence.

b. Emanation in Aquinas: Revealing the Essence In Aquinas, the human soul is a form: it makes a lump of matter a human body, and with it composes a human being.56 So to be human is not simply to have a soul, says Aquinas, it is also to have flesh and bones. Man’s essence thus includes in addition to the soul what Aquinas calls common matter, matter that is not individuated: flesh and bones, hands, etc.57 But these parts of man become particular parts of a particular man through the soul, which informs a particular piece of matter. So the soul is the part of the essence responsible for the realization of the whole essence. For instance, insofar as to have a hand is part of man’s essence, my soul makes a certain chunk of matter the hand that it is, so that I have a hand, and am a man. Now when Aquinas says that powers are consequent upon the soul’s essence, he makes perfectly clear that they are consequent upon the soul qua form.58 So in Aquinas, the thesis that powers flow from the soul means that empowerment follows information: I do not have a hand because I have the 56

STh Ia q. 75 a. 5. For a contextualized account of Aquinas’s position, see Bernardo Carlos Bazán, “The Human Soul: Form and Substance? Thomas Aquinas’ Critique of Eclectic Aristotelianism,” Archives d’histoire doctrinale et littéraire du Moyen Âge 64 (1997), 95–126. 57 58 STh Ia q. 75 a. 4 c. STh Ia q. 77 a. 1 c.

114

Charles Ehret

power to grasp, but I have the power to grasp because I first have a hand.59 In the same sense that the bed in the bedroom is consequent upon the independent fact that the room is a bedroom, the power to grasp in my hand is consequent upon the independent fact that my hand is a hand, a fact that obtains because my soul is the form of my body. Although Aquinas never goes so far, his doctrine does suggest that if I lose my power to grasp, because of, say, brain damage, I do not lose my hand, and this in turn would explain that I can learn to use it again. On this view, Aquinas can be perfectly consistent in holding that all powers flow from the soul’s essence. First, because the soul’s essence is to inform the body, and because information may be considered more fundamental than the endowment of powers. Second, because his doctrine establishes a one to one correspondence between the powers (e.g., the power to handle) and the constituents of man’s essence (e.g., the hand), which the soul actualizes. It is thus not only unproblematic, but perfectly in line with Aquinas’s psychology to say that all powers flow from the essence of the soul, because the soul is the substantial form, which accounts for all the constituents of a being. In this picture, to flow, as I have already stressed, has an opposite meaning to the one we find in Avicenna. In Avicenna, if what flows is p, then the essence from which it flows is not p. In Aquinas, if what flows is p, then the essence from which it flows is p. For instance, the human soul makes one’s skin sensitive, and it is indeed part of man’s essence to be sensitive. In Aquinas the flow of powers thus reveals the essence. That Aquinas actually sees things this way can easily be shown. For example, in

59 Here, it is important to note that by ‘hand’ I do not mean ‘a certain chunk of matter’ but ‘a certain body part’: my claim is not that my hand has the power to grasp because it is a certain chunk of matter; rather my claim is that my hand has the power to grasp because it is a part of my body, i.e., because it is informed by my soul. When Aquinas writes, in STh Ia q. 78 a. 3 c., that “powers do not exist on account of organs, but organs on account of powers,” ‘organ’ means ‘chunk of matter,’ and Aquinas’s general idea is that matter does not determine form, but form determines matter. Perler, “Faculties,” 111, thus comments: “the faculties have priority because they give a specific function to bodily parts, which, taken in themselves, are nothing but chunks of matter.” I agree with such a formulation, in the sense that a cut-off hand is nothing but a chunk of matter, but insofar as it is a bodily part, it receives a power which gives it a specific function. This, however, implies that a certain power is received by a certain chunk of matter because it is part of a certain body. This means that an organ is in essence a body part, and that it follows from that essence that it is endowed with a certain power. I thus think it is too strong to claim, with Perler, 112, that “an organ lacking a faculty is only homonymously called an organ— it is no longer a real organ”: this suggests that a power is what makes an organ what it is, which contradicts Aquinas’s thesis that powers are consequent upon the soul qua form. On my understanding, the lack of power can only be, at most, the sign that an organ is no longer a real organ, insofar as it indicates that it is no longer informed by the soul, hence no longer a part of a body.

The Flow of Powers

115

the De veritate, he argues that the mind isn’t the essence of the soul, but one of its powers. One argument contra is that since powers are accidents of the soul, then the intellect, or mind, isn’t a power, because rationality is part of man’s essence. If the De veritate’s argument is of particular interest, it is because it mentions Avicenna: An accident cannot be the principle of a substantial distinction. But man is substantially distinguished from beasts insofar as he has a mind. So the mind is not an accident. But the capacity of a soul is its property, according to Avicenna (potentia animae est proprietas eius, secundum Avicennam), and thus falls under the genus of accident. Therefore, the mind is not a capacity, but is the essence of the soul itself.60

The reference to Avicenna is odd, as the phrase used, if I am not mistaken, is neither in his De anima nor in his Metaphysics.61 Aquinas is surely picking up the general idea, promoted by Albert, that powers are extrinsic to the soul itself. But why does he not say that powers ‘flow’ or ‘emanate’ from the soul’s essence, as Albert so often repeats, and as we may indeed read in the Avicenna latinus? The reason is that Aquinas uses ‘flow’ to answer the objection, and to show how the mind may be a power of the soul and pertain to man’s essence: According to the Philosopher in Metaphysics VIII [1043a11], because the substantial differences of things are unknown to us, definitions sometimes make use of accidents instead, insofar as these accidents indicate or reveal the essence, as proper effects reveal the cause. This is why sensitive, insofar as it is the constitutive difference of an animal, is not taken from sense inasmuch as it names a capacity, but inasmuch as it names the essence of the soul itself, from which such a capacity flows. And the same goes with rational, or with what has a mind.62

The argument Aquinas is responding to rests on the genuinely Avicennian idea that if something proceeds from the essence of the soul, then it is in no way part of that essence. Aquinas uses the flow of powers against this Avicennian argument, and accordingly, his response gives to flow an opposite meaning to the one it has in Avicenna. Whereas Avicenna’s flow expels the unwanted powers from the essence of the soul, Aquinas’s flow allows tracing back what is true of emanated powers to the soul’s unknown essence. Avicenna reasons: organic powers flow from the essence of the soul, hence

60

De veritate q. 10 a. 1 arg. 6. The reference occurs again at De veritate q. 27 a. 6 arg. 3. Hasse, Avicenna’s De anima, 311, relates these occurrences to DA 171–2, but neither ‘potentia’ nor ‘proprietas’ are used; the author reads away this discrepancy by referring back to the affectiones of DA I.1, 15, implying that these ‘emanating affections’ are powers, as in Albert. 62 De veritate q. 10 a. 1 ad 6. 61

116

Charles Ehret

the soul’s essence is purely immaterial. Aquinas reasons the other way round: the rational power flows from the soul’s essence, hence the essence of the soul is rational. This role played by emanation is key. Indeed, the De veritate argument is but a variant of one Aquinas raises each time he deals with the distinction between the soul and its accidental powers, and which represents the most serious objection against it: if ‘rational’ qualifies an accidental power, the argument goes, then it does not qualify the essence. Aquinas’s answer never varies: ‘rational,’ when predicated of man, is predicated of his essence, although it is taken from power.63 But this inference from accidental powers to the essence is justified only because the former proceed from the latter. So even if Aquinas is more prone to say that powers are ‘accidents’ of the soul than to say that they ‘flow’ from its essence, he must say both, because if the accidental powers didn’t flow from the essence, the essence would remain completely unknown to us.64

c. Avoiding Self-Causation Now it is clear that the reasoning Aquinas applies to ‘rational’ can be applied to all powers, since, as we have seen, the soul is a substantial form, which as such accounts for all the constituent parts of a being. However, Aquinas has his eye on the intellect, and this is a good thing. Indeed, the intellectual powers are not received in the composite, but in the soul itself, which is its source,65 so it seems that Aquinas, like Albert, is presenting us with a case of self-causation. But unlike Albert, Aquinas sees the problem. In the Summa theologiae, for the first time, Aquinas writes a full article on “whether the capacities of the soul flow from its essence.” There, we find the following argument contra: ‘Emanation’ names a kind of motion. But nothing is moved by itself, as is proved in Physics VII [241b24–242a14], unless perhaps because of a part, as an animal is said to be moved by itself, because one of its parts moves and the other is moved. And what is more, the soul is not moved, as is proved in De anima I [408a34–b31]. So the soul does not cause in itself its own capacities.66

Aquinas replies: 63 In Sent. l. 1, d. 3 q. 4 a. 2 ad 5; STh Ia q. 77 a.1 ad 7; Q. d. de anima a. 12 ad 8; De spiritualibus creaturis a. 11 ad 3. 64 On the unknowability of essence, see Robert Pasnau, Thomas Aquinas on Human Nature: A Philosophical Study of Summa theologiae 1a 75–89 (Cambridge: Cambridge University Press, 2002), 164–70. The author concludes at 169 that, for Aquinas, “we do reach an understanding of essences; we simply do not do it directly.” 65 66 STh Ia q. 77 a. 5 c. STh Ia q. 77 a. 6 arg. 3.

The Flow of Powers

117

The emanation of proper accidents from a subject does not proceed by a change, but by a natural outcome, in the way something naturally comes from another, as color from light.67

When power flows, nothing changes. So some powers may well have the soul itself as subject, this does not amount to a case of self-causation, because it does not amount to a case of causation tout court. To justify this, Aquinas does not say much. The only clue he gives is an analogy: powers emanate from the soul in the same way colors come from light. Now the doctrine of light and perception is a complex one, so we may wonder what part of it the analogy points to.68 To get it right, we must keep in mind that the aim of the analogy is to explain how the flow of powers does not imply change. So the question is this: what part of the theory of colors describes a process that is not a change? I believe the key point here is that color is produced by light as a species, which has a spiritual or intentional being, both in the medium and in the eye where it is received.69 What this means is that the intermediate air as well as the eye, although they receive a likeness of color, are not colored because of this.70 So to receive color as it comes from light is not a change in the relevant physical sense, from which argues the objection. By ‘physical change’ I mean change as it is it defined in Aristotle’s Physics, namely as the process by which something potentially p becomes actually p.71 Aquinas’s point, then, is this: in the same way the color 67

STh Ia q. 77 a. 6 ad 3. A reading is proposed in Pasnau’s commentary on Aquinas’s Treatise on Human Nature, 269–70: “Aquinas takes for granted that colors are qualities belonging to objects. A source of light does not give existence to color, in Aquinas’s view, but instead actualizes (illuminates) the medium (the intervening air), enabling color to produce a likeness of itself in that medium, and thereby to enter the eye. Color, then, is not produced by the motion of a light source.” According to this reading, color is not produced by the motion of a light source because it is not produced by a light source at all; but Aquinas is saying that color is produced by a light source, only not by change. 69 The doctrine is developed in Sent. libri De anima II l. 14. For the theoretical background Aquinas is working from, see Lawrence Dewan, “St. Albert, the Sensibles, and Spiritual Being,” in J. A. Weisheipl (ed.), Albertus Magus and the Sciences: Commemorative Essays 1980 (Toronto: Pontifical Institute of Mediaeval Studies, 1980), 291–320. 70 In Sent. l. 1 d. 1 q. 1 a. 4 qc. 4 c. 71 Whether intentional or spiritual change in the medium or the senses is physical has been a matter of much recent discussion, where ‘physical’ takes on various meanings: Richard Sorabji, “From Aristotle to Brentano: The Development of the Concept of Intentionality,” in H. Blumenthal and H. M. Robinson (eds.), Aristotle and the Later Tradition, Oxford Studies in Ancient Philosophy, Supplementary Volume (Oxford: Clarendon Press, 1991), 227–59, contends that receiving an intention is for Aristotle a physical=physiological process, and Robert Pasnau, Theories of Cognition in the Later Middle Ages (Cambridge: Cambridge University Press, 1997), 35–47 argues that it is for Aquinas a physical=material process (in the modern sense of ‘material’); contra, Myles Burnyeat, “Aquinas on ‘Spiritual Change’ in Perception,” in D. Perler (ed.), Ancient and Medieval 68

118

Charles Ehret

produced by light (the likeness of color) does not determine what receives it, the emanated powers do not determine their subjects. So, when the soul emanates a rational power, this emitted power does not make the soul itself rational. And we can easily understand why this is so: the soul is already rational prior to the emanation of rational power, because it is rational in essence. In the same way, when my soul emanates a digestive power, this does not make my intestines digestive organs, because to digest is part of my essence qua animal, and because my body is informed by my soul prior to powers flowing from it, so that my intestines are digestive organs prior to them receiving a digestive power: they receive a digestive power because they are digestive organs, not the other way around. The analogy may be pushed one step further. Indeed, the likenesses of colors allow us to perceive the colors of objects themselves precisely because they do not color the intermediate air. In the same way, emanated powers allow us to perceive what things are precisely because they do not give them any distinct feature on top of their essence. I do not see the color red floating in the air, I see a red cloth; in the same way, I do not see an animal who has the power to eat meat, I see a carnivorous animal. As colors in the air, powers do not present a distinctive feature, they just make manifest what things are. So they do not change anything, and there is thus no risk of self-causation. The idea that all powers flow comes to fruition in Aquinas, for three reasons. First, because the soul is a substantial form that accounts for all the constituents of a being. Second, because the soul in Aquinas can be described as giving the body its form prior to giving the composite (and itself) its powers. Third, because emanation is taken into account as a specific kind of process, namely a non-physical process that does not imply change, and thus avoids any risk of self-causation.

4 . C O N C LU S I O N Albert wants to have it both ways. On the one hand, he wants the human soul to be an intellectual substance, because he is committed to the idea that, for it to be incorruptible, it cannot be essentially related to the body, i.e., the Theories of Intentionality (Leiden: Brill, 2001), 129–53, argues that, for Aristotle and Aquinas, it is a formal process, ‘physical’ nonetheless because form is a basic principle of Aristotelian physics; contra, Paul Hoffman, “Aquinas on Spiritual Change,” Oxford Studies in Medieval Philosophy 2 (2014), 98–103, notes that ‘physical’ in the mere sense of ‘involving form’ is too loose to be relevant. Here I take ‘physical’ to mean (with Burnyeat) what falls under Aristotelian physics, but (contra Burnyeat) I take Aristotelian physics to deal with form qua actuality of matter, so that mere form falls outside its realm.

The Flow of Powers

119

part of a composite. To make it a substance, he makes its organic powers, which relate it to the body, flow from it, so that they are expelled from its essence. But on the other hand, Albert wants the human soul to be an intellectual substance, even though its intellectual powers flow from it as well, in order to actualize the human body as a whole, so that it may be human and not simply animal. This makes emanation an equivocal notion, because what can truly be said of emanated powers, that they are ‘intellectual’ or ‘organic,’ both can and cannot be ascribed to the soul’s essence. The flow of powers amounts to a coherent doctrine only in Avicenna, where Albert picks it up, and in Aquinas, who uses it following Albert. But in Avicenna and in Aquinas the flow of powers has opposite meanings, as each author sides with the opposite half of Albert’s equivocal notion. In Avicenna, only organic powers flow from the soul’s essence, and this means that the soul’s essence is not organic, as it is constituted by intellectual powers. In Aquinas, both organic and intellectual powers flow from the soul’s essence, and this means that the soul’s essence is not power, but form, while it does imply that the human soul is organic and intellectual, as it is the substantial form of a material being. The flow of powers in Avicenna describes an intellectual substance pouring down, as it were, so that lower organic functions, which cannot be ascribed to its essence. The flow of powers in Aquinas describes a substantial form pouring out, so to speak, so that the functions may be performed that correspond to its essence and manifest it.72 University of Paris I-Sorbonne

BIBLIOGRAPHY Albert the Great. Opera Omnia, ed. A. Borgnet, 38 vols. (Paris: L. Vivès, 1890–99). Cited works: Summa theologiae II (32). Albert the Great. Opera Omnia . . . edenda . . . curavit Institutum Alberti Magni Coloniense . . . (Münster i. W.: Aschendorff, 1951–). Cited works: De anima (7.1), Super ethica (14), De causis (17.2), De homine (27.2) [= DH], Summa theologiae I (34.1). Avicenna [Ibn Sînâ]. Avicenna’s De Anima (Arabic Text): Being the Psychological part of Kitâb al-shifâ’, ed. F. Rahman (London: Oxford University Press, 1959) [= Psychology].

72

I am very grateful to Yamina Adouhane and Jean-Baptiste Brenet for their help with Avicenna’s Arabic. I would also like to thank the anonymous referees whose criticisms have been invaluable in improving earlier versions of my text. I benefited from presenting the topic at a workshop organized by Studium in Paris, and I wish to thank Véronique Decaix for her encouraging remarks on what was to become the first draft of this paper.

120

Charles Ehret

Avicenna [Ibn Sînâ]. Liber de anima seu Sextus de naturalibus, ed. S. van Riet, 2 vols. (Louvain: Éditions Orientalistes/Leiden: Brill, 1968–72) [= DA]. Avicenna [Ibn Sînâ]. Liber de philosophia prima, ed. S. van Riet, 3 vols. (Louvain: Peeters, 1977–83). Baldner, Steven. “St. Albert the Great on the Union of the Human Soul and Body,” American Catholic Philosophical Quarterly 70 (1996), 108–17. Bazán, Bernardo Carlos. “The Human Soul: Form and Substance? Thomas Aquinas’ Critique of Eclectic Aristotelianism,” Archives d’histoire doctrinale et littéraire du Moyen Âge 64 (1997), 95–126. Biard, Joël. “Diversité des fonctions et unité de l’âme dans la psychologie péripatéticienne (XIVe–XVIe siècle),” Vivarium 46 (2008), 342–67. Bonin, Thérèse. “The Emanative Psychology of Albertus Magnus,” Topoi 19 (2000), 45–57. Bonin, Thérèse. Creation as Emanation: The Origin of Diversity in Albert the Great’s On the Causes and the Procession of the Universe (Notre Dame, IN: Notre Dame University Press, 2001). Burnyeat, Myles. “Aquinas on ‘Spiritual Change’ in Perception,” in D. Perler (ed.), Ancient and Medieval Theories of Intentionality (Leiden: Brill, 2001), 129–53. De Boer, Sander W. The Science of the Soul: The Commentary Tradition on Aristotle’s De anima, c.1260–c.1360 (Leuven: Leuven University Press, 2013). Dewan, Lawrence. “St. Albert, the Sensibles, and Spiritual Being,” in J. A. Weisheipl (ed.), Albertus Magnus and the Sciences: Commemorative Essays 1980 (Toronto: Pontifical Institute of Mediaeval Studies, 1980), 291–320. Druart, Thérèse-Anne. “The Human Soul’s Individuation and Its Survival after the Body’s Death: Avicenna on the Causal Relation between Body and Soul,” Arabic Sciences and Philosophy 10 (2000), 259–73. Gilson, Étienne. “L’âme raisonnable chez Albert le Grand,” Archives d’histoire doctrinale et littéraire du Moyen Âge 14 (1943), 1–80. Hall, Robert E. “Intellect, Soul and Body in Ibn Sînâ: Systematic Synthesis and Development of the Aristotelian, Neoplatonic and Galenic Theories,” in J. McGinnis (ed.), Interpreting Avicenna: Science and Philosophy in Medieval Islam (Leiden: Brill, 2004), 62–88. Hasse, Dag Nikolaus. Avicenna’s De Anima in the Latin West: The Formation of a Peripatetic Philosophy of the Soul 1160–1300 (London: The Warburg Institute/ Turin: Nino Aragno Editore, 2000). Hasse, Dag Nikolaus. “The Early Albertus Magnus and his Arabic Sources on the Theory of the Soul,” Vivarium 46 (2008), 232–52. Hasse, Dag Nikolaus. “The Soul’s Faculties,” in R. Pasnau (ed.), The Cambridge History of Medieval Philosophy (Cambridge: Cambridge University Press, 2010), 305–19. Hoffman, Paul. “Aquinas on Spiritual Change,” Oxford Studies in Medieval Philosophy 2 (2014), 98–103. Janssens, Jules. “Creation and Emanation in Ibn Sînâ,” Documenti e studi sulla tradizione filosofica medievale 8 (1997), 455–77. King, Peter. “The Inner Cathedral: Mental Architecture in High Scholasticism,” Vivarium 46 (2008), 253–74.

The Flow of Powers

121

Lizzini, Olga. “L’âme chez Avicenne: quelques remarques autour de son statut épistémologique et de son fondement métaphysique,” Documenti e studi sulla tradizione filosofica medievale 21 (2010), 223–41. Lizzini, Olga. Fluxus (fayd.): Indagine sui fondamenti della metafisica e della fisica di Avicenna (Bari: Edizioni di Pagina, 2011). McGinnis, Jon. Avicenna (Oxford: Oxford University Press, 2010). Marmura, Michael E. “Some Questions Regarding Avicenna’s Theory of the Temporal Origination of the Human Rational Soul,” Arabic Sciences and Philosophy 18 (2008), 121–38. Pasnau, Robert. Theories of Cognition in the Later Middle Ages (Cambridge: Cambridge University Press, 1997). Pasnau, Robert. Thomas Aquinas on Human Nature: A Philosophical Study of Summa theologiae 1a 75–89 (Cambridge: Cambridge University Press, 2002). Perler, Dominik. “Faculties in Medieval Philosophy,” in D. Perler (ed.), The Faculties: A History (New York: Oxford University Press, 2015), 97–139. Rahman, Fazlur. Avicenna’s Psychology: An English Translation of Kitâb al-Najât, book II, chapter VI (London: Oxford University Press, 1952). Sebti, Meryem. “La signification de la définition avicennienne de l’âme comme ‘perfection première d’un corps naturel organique’ dans le livre I du Traité de l’âme du Shifâ,” Bulletin d’études orientales 51 (1999), 299–312. Sebti, Meryem. Avicenne: l’âme humaine (Paris: Presses Universitaires de France, 2000). Sebti, Meryem. “La distinction entre intellect pratique et intellect théorique dans la doctrine de l’âme humaine d’Avicenne,” Philosophie 77 (2003), 23–44. Sorabji, Richard. “From Aristotle to Brentano: The Development of the Concept of Intentionality,” in H. Blumenthal and H. M. Robinson (eds.), Aristotle and the Later Tradition, Oxford Studies in Ancient Philosophy, Supplementary Volume (Oxford: Clarendon Press, 1991), 227–59. Thomas Aquinas. Opera Omnia (Rome: Commissio Leonina, 1882–). Cited works: Summa theologiae Ia (4–5) [= STh], De veritate (22), Quaestiones disputatae de anima (24.1), De spiritualibus creaturis (24.2), Sentencia libri De anima (45.1). Thomas Aquinas. Scriptum super libros Sententiarum, ed. P. Mandonnet and M.-F. Moos (Paris: Lethiellieux, 1929–47) [= In Sent.]. Thomas Aquinas. The Treatise on Human Nature: Summa Theologiae 1a 75–89, trans. R. Pasnau (Indianapolis: Hackett, 2002). Ward, Julie K. “Souls and Figures: Defining the Soul in De anima ii 3,” Ancient Philosophy 16 (1996), 113–28. Wéber, Édouard-Henri. La Personne humaine au XIIIe siècle (Paris: Vrin, 1991). Wisnovsky, Robert. Avicenna’s Metaphysics in Context (Ithaca, NY: Cornell University Press, 2003).

Aquinas on the Individuation of Substances Jeffrey E. Brower

Aquinas addresses a number of issues in his writings that can be grouped under the heading ‘the problem of individuation.’1 But the issue that is of most importance to him, and does the most to shape his understanding of all the others, concerns the individuation of substances. We can state it as follows: The Chief Problem of Individuation What accounts for the numerical distinctness of substances belonging to the same species?

1 For an introduction to some of these issues, as well as those discussed by other thirteenth- and fourteenth-century philosophers in the context of individuation, see Jorge J. E. Gracia (ed.), Individuation in Scholasticism: The Later Middle Ages and the CounterReformation, 1150–1650 (New York: State University of New York Press, 1994); Peter King, “The Problem of Individuation in the Middle Ages,” Theoria 66 (2000), 159–84; and Martin Pickavé, “The Controversy over the Principle of Individuation in Medieval Quodlibeta (1277–1320): A Forrest Map,” in C. Schabel (ed.), Theological Quodlibeta in the Middle Ages: The Fourteenth Century (Leiden: Brill, 2007), 17–79. For Aquinas’s works, I rely on the following abbreviations: DEE De ente et essentia DPN De principiis naturae In BDT Expositio super librum Boethii De Trinitate In CA Super librum De Causis expositio In DA Sententia super De anima In Sent. Scriptum super libros Sententiarum QDA Quaestiones disputatae de anima QDSC Quaestio disputata de spiritualibus creaturis ST Summa theologiae For editions of these works, see the bibliography. All translations are mine, though I have consulted other English translations and in some cases adopted their wording without significant changes.

Aquinas on the Individuation of Substances

123

As this statement makes clear, Aquinas thinks of the problem of individuation as arising primarily in contexts in which we have substances sharing a common nature or essence, and hence in which we must appeal to something distinct from their nature or essence to account for their individuality. Indeed, it is precisely for this reason that Aquinas takes the chief problem to be that of identifying what he often refers to simply as “the principle of individuation” (principium individuationis).2 Certain details of Aquinas’s solution to this problem show evidence of change over time. But its overall structure does not. Indeed, for the purposes of this paper, we can formulate Aquinas’s solution in terms of the following account: Aquinas’s Account of Individuation Prime matter and quantity jointly account for the numerical distinctness of substances belonging to the same species.

As this formulation helps to make clear, Aquinas takes the principle of individuation to involve two things—what he often refers to simply as ‘matter’ (materia) and ‘dimensions’ (dimensiones). But as this formulation is also intended to emphasize, Aquinas takes dimensions to be a type of quantity and he takes the relevant sort of matter to be prime matter. Indeed, he often refers to matter in the primary sense (i.e., prime matter) simply as ‘matter.’3 Of the two, moreover, Aquinas clearly thinks prime matter is the 2 Although Aquinas often speaks of the principle of individuation as that which accounts for the “numerical distinctness” of substances belonging the same species, he also speaks of it as that which accounts for their “multiplication” or “division” (more on this latter way of speaking in §3). As we shall also see (esp. n. 17), there are contexts in which Aquinas is willing to speak of individuation in a broad enough sense to include anything that accounts for a substance’s numerical distinctness or division from other things, even if this is nothing apart from the substance itself (or some primitive fact about it). For more on Aquinas’s views about common natures, and their connection to individuation, see my “Aquinas on the Problem of Universals,” Philosophy and Phenomenological Research 92 (2016), 715–35. For a systematic overview of the various ways in which Aquinas speaks about both individuation and the principle of individuation, see John F. Wippel, The Metaphysical Thought of Thomas Aquinas: From Finite Being to Uncreated Being (Washington, DC: The Catholic University of America Press, 2000), esp. 296–312 and 351–75. Finally, for evidence that it was perfectly standard, in the thirteenth and fourteenth centuries, to distinguish broader vs. narrower senses of individuation, and to focus only on the latter, as Aquinas himself does, see Pickavé, “Controversy over the Principle of Individuation.” 3 Although my formulation represents what we might think of as the standard interpretation of Aquinas’s account, it is not accepted by everyone. For a very different understanding of Aquinas’s use of ‘matter’ in the context of individuation, see Christopher Hughes, “Matter and Individuation in Aquinas,” History of Philosophy Quarterly 13

124

Jeffrey E. Brower

more important, explicitly referring to it in one context as “the primary principle of individuation” (primum principium individuationis) and relegating dimensions in this same context to the status of “a secondary principle” (secundarium principium).4 Aquinas’s account of individuation has received a great deal of attention from commentators, both past and present. Some of this attention has focused on subtle exegetical questions concerning certain of its specific details.5 But much more attention has been given to some basic philosophical questions concerning the coherence of the account as a whole.6 For as commentators since Aquinas’s own time have been quick to point out, it is not at all obvious that either prime matter or quantity can be coherently said to have any role to play in the individuation of substances. For one thing, Aquinas is notorious for identifying prime matter with ‘pure potentiality.’ But how can something that is in pure potentiality be said to exist, much less play a leading role in individuation? Again, Aquinas accepts the standard medieval view that quantities are accidents, and hence derive their being or identity from the substances to which they belong. Indeed, it is for precisely this reason that Aquinas insists that all accidents are individuated by substances.7 But if that is right, in what sense could quantities or dimensions possibly be said to “individuate” these same substances? I think these questions about the coherence of Aquinas’s account can all be answered, though not in the way that commentators have typically

(1996), 1–17. For a detailed defense of the standard interpretation, see Wippel, Metaphysical Thought of Aquinas, esp. 351–75. 4 In Sent. 4.12.1.1.3 ad 3. 5 Of special concern here is the question of what specific type of dimensions Aquinas takes to be involved in individuation, since he describes such dimensions in different ways at different places in his writings—sometimes as “terminate” (terminatae), other times as “determinate” (determinatae), and yet other times as “indeterminate” (interminatae). In what follows, I take no stand on questions of this sort. But see again Wippel, Metaphysical Thought of Aquinas, esp. 351–75, for detailed discussion, including a survey of the relevant texts and issues, as well as the different positions taken by commentators regarding the consistency and development of Aquinas’s views over time. 6 I should emphasize that when I speak of Aquinas’s account as a whole, I am thinking of it only insofar as it provides an account of what we might call synchronic individuation—that is, the individuation of substances at a given time. Whether Aquinas intends this account to extend to what we might call diachronic or modal individuation—that is, the individuation of substances across times or across possible worlds—are among the exegetical questions I want to set aside. But for relevant discussion, see again the references in nn. 3 and 5. For some purely philosophical grounds for thinking that matter cannot account for either diachronic or modal individuation, see Kathrin Koslicki, “Essence and Identity,” in M. Dumitru (ed.), Metaphysics, Meaning and Modality: Themes from Kit Fine (Oxford: Oxford University Press, forthcoming). 7 See, e.g., ST 1.29.1.

Aquinas on the Individuation of Substances

125

assumed. Indeed, as I see it, the standard difficulties that arise for the coherence of Aquinas’s account all stem from a failure to appreciate the precise nature of the problems that he means to be addressing. In what follows, I attempt to correct this state of affairs, beginning with a fresh examination of Aquinas’s views about prime matter (§§1–2). After a brief survey of some of the most familiar and elementary aspects of these views, I argue that there is a clear, but often overlooked role for prime matter to play in individuation, one that is perfectly consistent with its status as pure potentiality. With these results in hand, I then turn to the proper understanding of Aquinas’s views about quantity or dimensions in individuation (§3). Here again, I argue that once we are clear about the precise nature of the problem to be addressed, and in particular the phenomenon he is attempting to explain, we can see that Aquinas’s account is not only perfectly coherent, but well motivated within the context of his broader philosophical commitments, many of which have yet to be fully appreciated.8

1. PRIME MATTER AS THE PRIMARY P R I N C I P LE O F IN D I V I D U A T I O N In order to understand Aquinas’s views about the role of prime matter in individuation, we must first be clear about how he is thinking of prime matter itself.

a. Prime Matter as a Special Type of Substratum Aquinas develops his views about prime matter as part of his hylomorphism—that is, his broadly Aristotelian views about matter (hyle) and form (morphe). In this context, Aquinas treats the notions of matter and form as correlative, not only with each other but also with the further notion of a hylomorphic compound. In order to clarify all three notions, as well as to bring out their correlativity, Aquinas associates them with two other notions—namely, potentiality and actuality. Consider, for example, the 8 My discussion in what follows builds on ideas first presented in my “Matter, Form, and Individuation,” in B. Davies and E. Stump (eds.), The Oxford Handbook of Aquinas (Oxford: Oxford University Press, 2012), 85–103 and further developed in my Aquinas’s Ontology of the Material World: Change, Hylomorphism, and Material Objects (Oxford: Oxford University Press, 2014). But with respect to Aquinas’s views about individuation and individuality, my discussion here is intended to supersede both.

Jeffrey E. Brower

126 A Being in Actuality

A Being in Potentiality

Compound

Matter

Inherence

Form

An Actuality

Composition

Figure 4.1 Matter, form, and compound

following passage, which highlights the precise nature of this association via some technical Aristotelian terminology: There is a difference between matter and form: matter is a being in potentiality, whereas form is an entelechy—that is, an actuality—by which [the potentiality of] matter is actualized. For the same reason, the compound itself is the being in actuality. (In DA 2.1.113–17)

As this passage indicates, Aquinas conceives of matter, form, and compound as follows: matter is “a being in potentiality” (ens in potentia)—that is, something in potentiality for receiving form; form is “an actuality” (actus)—that is, something that actualizes the potentiality of matter; and a hylomorphic compound is “a being in actuality” (ens actu)—that is, something that exists when some form actualizes the potentiality of some matter. For the sake of clarity, we can illustrate this conception of matter, form, and compound, as well as the technical terminology that Aquinas uses to express it, as in Figure 4.1. As Figure 4.1 indicates, Aquinas takes matter, form, and compound to be three distinct types of being, bearing two distinct types of relation to one another.9 I have used a solid arrow to stand for the relation that a form bears to the matter that serves as its subject or substratum. Aquinas refers to this relation as “inherence” (inhaerentia).10 By contrast, I have used dotted arrows to stand for the relation that matter and form jointly bear to the

9 Although Aquinas often speaks of matter, form, and compound as different types of being, he is also happy to speak of them as beings or entities having different modes of being. Elsewhere, I have argued that there is a systematic correspondence between types and modes of being, for Aquinas, and that modes are in fact explanatorily more basic or prior (see Aquinas’s Ontology of the Material World, 42–54). Even so, since Aquinas’s views about modes introduce a host of unnecessary complications associated with his doctrine of the analogy of being, I shall hereafter speak only of types of being. 10 See, e.g., ST 1.75.2 ad 2.

Aquinas on the Individuation of Substances

127

compound of which they are constituent parts. Aquinas refers to this relation as “composition” (compositio).11 The three central notions of Aquinas’s hylomorphism—namely, matter, form, and compound—bear a striking resemblance to three notions familiar from contemporary metaphysics—namely, substratum, property, and complexes of each. Admittedly, contemporary metaphysicians do not conceive of substrata in terms of any kind of matter, but rather conceive of them as particulars of a distinctive type (namely, bare particulars). And the contemporary notion of a property would seem not to apply to one of the types of entity that Aquinas thinks of as a form—namely, individual human souls. Even so, if we focus only on the structural relations between these two sets of notions, the resemblance would appear to be exact. Indeed, if we set aside the complications associated with the human soul (which are not relevant for our purposes), and leave open the precise nature of the beings or entities playing the role of substrata, we can think of Aquinas’s hylomorphic compounds in terms of substratum–property complexes (or what are now commonly called facts or concrete states of affairs). So far so good. But how exactly are we to understand the place of prime matter in this broadly Aristotelian form of substratum theory? In order to answer this question, it must be noted that Aquinas thinks we can distinguish two different types of matter—and corresponding to them, two different types of form and compound. To see what these distinctions come to, let us start with what Aquinas calls “material substances” or “bodies.” These are compounds of prime matter and substantial form. Aquinas refers to their matter as “prime matter” to emphasize that he takes it to be matter in the primary or proper sense (more on this shortly); and he refers to their forms as “substantial forms” to emphasize that they are part of the very nature or substance of the compounds they compose. Now as it turns out, Aquinas thinks that substances themselves can serve as substratra for forms of a further type and hence play the role of matter for larger compounds. Since the forms that combine with substances to compose larger compounds fall outside their nature, Aquinas refers to them as “accidental forms” or “accidents.” And he refers to compounds composed of substances and accidents as “accidental unities.”12 11

See, e.g., ST 1.85.5 ad 3. See, e.g., the following passage in which Aquinas distinguishes all of these various types of matter, form, and compound: Just as a substantial form does not exist separately from that in which it inheres, neither does [the substratum] in which the form inheres—namely, [prime] matter—exist separately. What results from their union, therefore, is a type of being in which something subsists through itself, and what is produced from them 12

Jeffrey E. Brower

128

A Being in Actuality

A Being in Potentiality & Actuality

A Being in Potentiality

Prime Matter

Accidental Unity

Material Substance

Accidental Form

Substantial Form

An Actuality

An Actuality

Figure 4.2 Hierarchy of compounds in the material world

To illustrate these different types of matter, form, and compound, consider one of Aquinas’s favorite examples of an artifact—namely, a bronze statue.13 According to Aquinas, such a statue is an accidental unity, composed of a specific type of material substance (namely, a lump of bronze) and a specific type of accidental form (namely, a shape). What is more, he takes the lump of bronze that serves as the matter for the statue to be itself a compound of matter and form—more precisely, a compound of prime matter and substantial form (namely, the form of bronze). As this particular example helps us to see, Aquinas thinks that there can be nested hierarchies of compounds in the material world involving each of the various types of matter, form, and compound that he distinguishes. For the sake of clarity, we can bring out the structure of such hierarchies, as well as their connection to the technical notions of matter, form, and compound introduced earlier, using the diagram in Figure 4.2. In the context of Figure 4.2, we can see why Aquinas would regard prime matter as matter in the primary or proper sense. (I have used dotted lines to emphasize its special place and status in the hierarchy.) Like all matter, prime matter serves as the substratum for a form or property, and hence enters into a larger compound as a proper part or constituent. But unlike all other matter, prime matter is not itself composed of any substrata and properties. On the contrary, it functions merely as a substratum, never as is an essential unity [or substance] . . . By contrast, that in which an accident inheres is a complete being in itself, subsisting in its own being. Indeed, such a being naturally precedes the accident that comes to it. Hence, what is produced from an accident and its subject is not an essential unity [or substance], but an accidental unity. (DEE 6.24–49) 13 See, e.g., DPN 2.

Aquinas on the Individuation of Substances

129

a compound, and hence must be said to possess forms or properties only via inherence, never via composition (or constituency). We might put all this by saying that, for Aquinas, prime matter is the ultimate substratum for things in the material world. Aquinas often cites the fact that prime matter functions merely as a substratum, never as a compound, as the reason it qualifies as matter in the primary or proper sense.14 But in fact there is more to it than this. Aquinas also thinks that prime matter is precisely that which accounts for the fact that certain substances qualify as material objects or bodies. (More on this in §3.b.) Indeed, I think it is clear that insofar as substances serve as substrata, they qualify as ‘matter’ only in a purely functional sense. For Aquinas thinks that even wholly immaterial substances, such as angels, can serve as the substrata for accidental forms and hence qualify as matter in the technical sense of “a being in potentiality.” But as he himself makes clear, this should not be confused with saying that such substances are in any sense material objects or bodies (as opposed to spirits): If the term ‘matter’ is used in the sense [of prime matter]—which is its proper and common sense—it is impossible for there to be matter in spiritual substances . . . On the other hand, if the terms ‘matter’ and ‘form’ are used for any two things which are related as potentiality to actuality then there can be no objection (unless it is a mere verbal dispute) to saying that there is both matter and form in spiritual substances. (QDSC 1.300–2, 357–60)

In short, prime matter, for Aquinas, is not merely the ultimate substratum for things in the material world, but also that which accounts for their materiality in the ordinary sense of the term. These points might seem so basic or elementary as to hardly warrant mentioning. But as we shall see shortly, their significance for Aquinas’s views about individuation continues to be overlooked.

b. Prime Matter and Individuation Seeing the place that prime matter occupies in Aquinas’s hylomorphism, and in particular its status there as a special type of substratum, helps to make sense of why he appeals to it in the context of individuation. Indeed, philosophers often invoke substrata in this context precisely because substrata appear to be uniquely suited to function as individuators. For example, contemporary substratum theorists standardly reject the view of their chief rival, the bundle theorists, precisely because it provides nothing to play the role of individuator. Bundle theorists take objects to be 14

See, e.g., DPN 2.70–8, which I return to in §2.a.

130

Jeffrey E. Brower

complexes (or “bundles”) of properties, and hence would seem to have no way of allowing for distinct objects to share all the same properties. By contrast, substratum theorists take objects to be complexes of both substrata and properties, and hence can easily allow for distinct objects to share all the same properties—namely, by saying that such objects differ merely with respect to their substrata. Of course, contemporary substratum theorists do not conceive of substrata in the same way Aquinas does—namely, as a distinctive type of matter. On the contrary, as noted earlier, they conceive of substrata as individuals or particulars of distinctive type—namely, as bare particulars. As we will see (§2.b), this difference turns out to be a significant one. Even so, I think it is clear that Aquinas intends prime matter to play the same sort of individuating role that bare particulars play for contemporary substratum theorists. After all, Aquinas assumes that substances belonging to the same species must possess the same common nature or essence, and hence the same substantial form. For the same reason, he often insists there is no way to distinguish multiple substances of the same type without appealing to the distinctness of their substrata or prime matter.15 Indeed, it is precisely because Aquinas thinks that incorporeal substances or spirits lack prime matter that he famously embraces the controversial doctrine that there can be no more than one angel per species. For insofar as angels lack prime matter, he thinks there is nothing to play the role of substratum for their substantial forms, and hence no way to generate more than a single instance of their specific type.16

15 It might be wondered why Aquinas conceives of substrata as a type of matter rather than as a type of particular (say, bare particulars). The short answer is this: Aquinas thinks that if substrata were particular, they would themselves qualify as substances. But since he denies that substances can have other substances as proper parts or constituents, it would follow that no complex of which such substrata are a part could qualify as substances. Thus, if substrata were particular, Aquinas thinks that it would be impossible to identify ordinary material substances with hylomorphic compounds (at least insofar as such compounds are conceived of as substratum–property complexes). For further development and defense of these claims, see my Aquinas’s Ontology of the Material World, esp. 22–7 and the references cited there. 16 Aquinas denies that angels can possess prime matter on the grounds that this would make it impossible for them to engage in their characteristic sort of cognition, and he denies that they can possess substrata of any other sort for the reason given in the previous note. In light of all this, Aquinas thinks we have no choice but to insist that angels are sui specie. For helpful discussion of Aquinas’s views about the individuation of angels, including relevant texts and historical context, see Giorgio Pini, “The Individuation of Angels from Bonaventure to Duns Scotus,” in T. Hoffmann (ed.), A Companion to Angels in Medieval Philosophy (Leiden: Brill, 2012), 79–115. According to Pini (79), “It is not an exaggeration to say that later medieval discussions on this topic were largely debates over Aquinas’s position.”

Aquinas on the Individuation of Substances

131

Admittedly, what I have just said about Aquinas’s views about common natures or essences glosses over some important complications. In particular, when Aquinas speaks of distinct substances—say, two lumps of bronze—sharing the same common nature or essence, this is not to be understood in terms of their literally sharing numerically one and the same form or property. On the contrary, like most other medieval philosophers, Aquinas emphatically denies the existence of any universals outside the mind, insisting instead that every individual substance has its own distinct forms or properties. At the same time, however, and in contrast to some other medieval philosophers, he insists that the sort of sameness that holds between substances belonging to the same species is such that their forms or properties cannot be regarded as distinct from one another in and of themselves. On the contrary, they must be regarded as merely derivatively distinct—that is to say, distinct solely in virtue of their relation to some primitively distinct individuators. And here again, Aquinas appeals to prime matter to serve as the ultimate source of their distinctness. Far from detracting from prime matter’s role as an individuator, therefore, such complications merely serve to reinforce it.17

17 See my “Aquinas on the Problem of Universals” for a detailed discussion of Aquinas’s views about common natures, including both their textual basis and what they imply about the different ways in which something can be distinct or individual. See also the following passage, in which Aquinas connects these different ways of being individual with prime matter’s role in individuation: [S]omething is said to be individual insofar as it not disposed to exist in many things (for what is disposed to exist in many things is universal). But there are two ways in which something can fail to be disposed to exist in many things. The first is by being disposed to exist only in the single thing to which it belongs. For example, even though whiteness is disposed to exist in many things on account of its species, a particular whiteness can only exist in the particular subject to which it belongs. But this way [of failing to be disposed to exist in many things] cannot go on forever, since as is proved in Metaphysics 2, it is impossible to go on forever with respect to formal and material causes. Hence, we must [eventually] arrive at something that is not disposed to belong to anything and from which other things get their individuation—just as, in the case of corporeal things, we must [eventually] arrive at prime matter, which is the principle of their singularity. Hence, when something is not disposed to belong to anything, it must for that very reason be individual. And this brings us to the second way in which something can fail to be disposed to exist in many things—namely, by not being disposed to belong to anything. For example, if whiteness were a separated [Form], existing without a subject, then it would be individual in this way. Moreover, this is the way in which there is individuation both in separated substances [or angels], which are forms having being, and in the First Cause itself, which just is subsisting being. (In CA 1.9.65–6)

132

Jeffrey E. Brower 2 . P R I M E M A T T E R AS P U R E P O T E N T I A L I T Y

So far I have been trying to motivate Aquinas’s views about the role of prime matter in individuation by highlighting its role as an ultimate substratum within the context of his hylomorphism. But even if everything I have said is correct, serious questions remain about how this conception of prime matter is supposed to cohere with Aquinas’s standard characterization of prime matter as “pure potentiality.” There are really two problems to be distinguished here—what I shall refer to as the problem of actuality and the problem of individuality. The first problem is familiar, and often thought to be the undoing of Aquinas’s hylomorphism as a whole. The second is more subtle and has yet to receive the attention it deserves. Since both strike at the very heart of Aquinas’s views about prime matter and individuation, I will briefly consider each.

a. The Problem of Actuality Few aspects of Aquinas’s hylomorphism have caused more difficulty for commentators than his insistence that prime matter must be understood as pure potentiality. Nor is it hard to see why. For to say that something qualifies as pure potentiality is, presumably, just to say that it is completely lacking in actuality. Aquinas himself seems to confirm this presumption when he says things such as the following: “nothing that is a being in actuality can be called prime matter” (DPN 2.119). At the same time, however, Aquinas also insists that nothing that lacks all actuality can exist. Indeed, like other medieval philosophers, he takes for granted that for something to have being or existence just is for that thing to possess actuality.18 In short, Aquinas’s characterization of prime matter, together with his views about actuality, present us with a straightforward argument for the non-existence of prime matter. The relevance of this argument for Aquinas’s views about individuation should be obvious. For nothing can be a principle of individuation—or for that matter, play any substantive role in one’s ontology—unless it exists.19 18 Compare ST 1.5.1 ad 1: “The term ‘being’ (ens) properly expresses that something exists in actuality.” 19 This is an instance of what Paul Spade calls the Principle of Fair Play. See Paul V. Spade, “Degrees of Being, Degrees of Goodness: Aquinas on Levels of Reality,” in S. MacDonald and E. Stump (eds.), Aquinas’s Moral Theory: Essays in Honor of Norman Kretzmann (Ithaca, NY: Cornell University Press, 1999), 254–75, esp. 271. It is also a consequence of the contemporary doctrine of serious actualism. See, e.g., Alvin Plantinga, “On Existentialism,” Philosophical Studies 44 (1983), 1–20.

Aquinas on the Individuation of Substances

133

Hence, if prime matter does not exist, it cannot be a principle of individuation. The problem of actuality, therefore, is just that prime matter does not seem to have any. The Argument for Non-Existence 1. 2. 3. 4.

Prime matter is pure potentiality. If something is pure potentiality, it is not a being in actuality. Something exists if and only if it is a being in actuality. Hence, prime matter does not exist.

Although something like the above argument has often been thought to be the undoing of Aquinas’s hylomorphism as a whole, I think its force has been greatly exaggerated. Indeed, if we recall what we have already seen of Aquinas’s views about matter, form, and compound, I think it is clear that any apparent force this argument has owes to an equivocation on the term ‘a being in actuality.’ Although it is true that Aquinas sometimes uses the term ‘a being in actuality’ in a broad sense to refer to anything that has being or existence, we have also seen that, in the context of his hylomorphism, he uses it in a much narrower or technical sense to refer to a hylomorphic compound—that is, to a type of being that possesses form or actuality as a proper part or constituent.20 In this sense, the term ‘a being in actuality’ contrasts with ‘a being in potentiality,’ where the latter refers to matter or a substratum—that is, a type of being that possesses form or actuality via inherence. Now if we recall that Aquinas allows for a type of secondary matter (or substance) that can be said to qualify as both matter and compound, and hence as both a being in potentiality and a being in actuality, it becomes clear that his characterization of prime matter as “pure potentiality” is intended to indicate only that prime matter is not a being in actuality in the narrow or technical sense.21 Indeed, Aquinas himself emphasizes this point using his statue example: It must be known that some matter—for example, bronze—is composed of [matter and] form. For even though bronze is matter in respect of a statue, it is itself composed of matter and form. Now because bronze has matter [and form], it is not called prime matter. On the contrary, only that matter which is understood to be without any form or privation but also to be a substratum for form or privation is called prime matter. For only it has no other matter [and form] before it. And such matter is also called hyle. (DPN 2.70–8)

20 21

See again Figure 4.1. See again the place of material substance in the hierarchy at Figure 4.2.

134

Jeffrey E. Brower

Admittedly, Aquinas does not explicitly characterize prime matter here as “pure potentiality.” Even so, it clear that what he takes to be distinctive about prime matter is not the lack of form or actuality as such (since otherwise prime matter could not be understood “to be a substratum for form or privation”), but only the lack of form or actuality “as a proper part or constituent.” Moreover, if we look to passages in which Aquinas does explicitly characterize prime matter as “pure potentiality,” we can see that he often makes the same point by emphasizing that prime matter is a type of being that lacks being or actuality not as such, but only “through itself ”: Prime matter can never exist through itself (per se). Indeed, since it does not have any form in its nature, it does not have being in actuality, but only in potentiality (est solum in potentia), since being in actuality comes only through a form. For the same reason, nothing that is a being in actuality can be called prime matter. (DPN 2.114–19) Because prime matter is not a being in actuality (ens in actu), but merely in potentiality (potentia tantum), it does not exist in reality through itself (per seipsam). (ST 1.7.2 ad 3)

In each of these passages, Aquinas is attempting to highlight the distinctive way in which prime matter possesses being or actuality. And he does so by way of a comparison with material substances or compounds. Thus, whereas a compound possesses being or actuality “through itself ” (per se)—that is, by having a form or actuality as a proper part or constituent—prime matter possesses being or actuality only “through another” (per aliud)—that is, by having a form or actuality inhering in it. As all of this goes to show, we must distinguish for Aquinas two different senses of the term ‘a being in actuality’:

Two Senses of ‘A Being in Actuality’ • Narrow sense: A being that possesses actuality as a proper part or constituent. • Broad sense: A being that possesses actuality in some way or other.

The narrow sense of ‘a being in actuality’ applies only to compounds or beings that possess actuality in the distinctive way that a compound does (namely, via constituency or “through itself ”). By contrast, the broad sense of this term applies to all beings, regardless of how they possess actuality. Although it is clear that prime matter cannot be said to qualify as a being in actuality in the narrow sense (since it does not possess actuality in the way a compound does—namely, as a proper part or constituent or

Aquinas on the Individuation of Substances

135

“through itself ”), it is equally clear that it must be said to qualify as a being in actuality in the first sense (since it does possess actuality in the way that matter or a substratum does—namely, via inherence or “through a form”).22 If we return to the argument that raises the problem of actuality with the foregoing in mind, we can see that there is no consistent reading of this argument that preserves the truth of all of its premises: The Argument for Non-Existence 1. 2. 3. 4.

Prime matter is pure potentiality. If something is pure potentiality, it is not a being in actuality. Something exists if and only if is a being in actuality. Hence, prime matter does not exist.

For the second premise is true only if the argument is read with the narrow sense of ‘a being in actuality,’ whereas the third premise is true only if the argument is read with the broad sense of this term. I conclude, therefore, that the argument is unsound, and hence does not by itself raise any insuperable difficulties, either for Aquinas’s hylomorphism in general or for his specific views about individuation.23

22 As all of this helps to make clear, Aquinas has a tendency to use terms in a multiplicity of senses (narrow vs. broad, proper vs. improper, qualified vs. unqualified, etc.). Although this tendency significantly complicates the explication of his views, it will be important to keep in mind as we proceed. For some discussion of the sources of this tendency, which is characteristic of medieval philosophy as a whole, see my Aquinas’s Ontology of the Material World, 7, n. 10. 23 It is, perhaps, worth mentioning here another difficulty that is often associated with the problem of actuality—namely, that of how prime matter can be said to possess any nature or character of its own. For insofar as prime matter lacks a form or actuality through itself, it might seem to lack any characteristics or properties whatsoever. I cannot deal with this difficulty in detail here, but will simply note that it rests on the mistaken assumption that things cannot possess their nature or character primitively, for Aquinas, but must instead derive it from some distinct forms or properties that they possess. That this assumption is mistaken seems clear from the fact that Aquinas countenances many things that possess their nature or character independently of such forms or properties— including both God, who has no distinct forms or properties, and accidents, which can possess distinct forms or properties in certain contexts (in particular, the Eucharist) but do not derive their nature or character from them. For a detailed discussion of this difficulty, one that locates it within a systematic treatment of Aquinas’s views about the role of forms or properties in characterization, see my Aquinas’s Ontology of the Material World, esp. §§6.3–5 and 7.1–2.

136

Jeffrey E. Brower

b. The Problem of Individuality The considerations discussed so far seem to me to put to rest the most common concern that has been raised in the literature for the role that Aquinas assigns to prime matter in individuation, as well as for his hylomorphism as a whole. There is, however, another concern that can be raised for prime matter’s role in individuation, much less familiar but also closely connected to Aquinas’s conception of pure potentiality. For the sake of parallelism, we can refer to this concern as “the problem of individuality” and state the argument that gives rise to it as follows: The Argument for Non-Individuality 1. 2. 3. 4.

Prime matter is pure potentiality. If something is pure potentiality, it is not a being in actuality through itself. Something is individual if and only if it is a being in actuality through itself. Hence, prime matter is not individual.

Although prime matter’s status as pure potentiality does not entail its nonexistence, for Aquinas, we have seen that it does entail its lack of being or actuality through itself (per se). As it turns out, however, Aquinas takes the possession of being or actuality through itself to be a necessary condition for individuality. That is to say, he takes prime matter’s status as pure potentiality to entail its non-individuality. But it would seem to be a conceptual truth that only something individual can individuate. For presumably an individuator just is something that accounts for individuality. But if that is right, and prime matter really is non-individual for Aquinas, then it is hard to see how it could play any role in individuation. Once again, the problem is that prime matter seems to lack something necessary to serve as a principle of individuation—though in this case its individuality rather than actuality. Unlike the previous argument we considered, I think there is a reading of this one on which it is sound. The real question, therefore, is how something non-individual can qualify as a principle of individuation for Aquinas, and hence how the apparent conceptual truth mentioned above can be resisted. To answer this question, however, we must first be clearer about how Aquinas is thinking of individuality itself. Aquinas uses the term ‘individual’ (individualis) in different ways at different places in his writings. Sometimes he uses it so broadly as to apply to anything that exists. This is not perhaps surprising, since the term can just mean ‘non-universal’ (non universalis) and Aquinas thinks

Aquinas on the Individuation of Substances

137

everything that exists is non-universal.24 More often than not, however, Aquinas uses it in a narrower sense to apply only to those types of nonuniversal that possesses being or actuality through themselves. Consider, for example, the following passage, which contains perhaps his clearest and best-known account of individuality: Two things belong to the notion of an individual—namely, (a) that it is a being in actuality, either in itself or in another, and (b) that it is a being divided from all other things which are (or can be) in the same species, though it is undivided in itself. (In Sent. 4.12.1.1.3 ad 3)

In this passage, Aquinas identifies two conditions on individuality. First, to be an individual something must qualify as a being in actuality either in itself or in another (in se vel in alio). Second, this same being must belong to a species in such a way that it possesses a kind of unity distinct from other (possible) members of the same species. There are a number of things that could be said about this passage. But here I want to make just three points. The first, and perhaps, most obvious point has to do with the first condition on individuality—and in particular, Aquinas’s distinctive use in this condition of the term ‘a being in actuality.’ As we have seen, in hylomorphic contexts Aquinas uses this term to mark the contrast between compounds, which possess actuality through themselves (per se), and matter, which possesses actuality only through another (per aliud ). In the present context, however, Aquinas is using the term ‘a being in actuality’ in a broader sense, to contrast not only compounds but also forms with matter. For although forms do not possess actuality in precisely the same way that compounds do (namely, as a proper part or constituent), there is a clear sense in which they too possess actuality through themselves (per se) rather than through another (per aliud )—for forms just are actualities. Indeed, they are actualities that inhere “in” other things. In short, when Aquinas speaks of a being in actuality “either in itself or in another,” I think he means to be distinguishing two different ways in which something can be said to possess actuality “through itself”—namely, the way in which compounds do (i.e., via proper parthood or constituency) or the way in which forms do (i.e., via identity or improper parthood).25

24 For a discussion of this sense of ‘individual’ in Aquinas, see Lawrence Dewan, “The Individual as a Mode of Being According to Thomas Aquinas,” The Thomist 63 (1999), 403–24. See also the first sentence of the text from In CA 1.9 cited in n. 17 earlier. 25 In order to emphasize that both forms and compounds possess actuality through themselves, it will be useful in what follows to speak of forms as possessing actuality via improper parthood rather than identity (where x is an improper part of y just in case x = y).

138

Jeffrey E. Brower

If I am right about this, then there is for Aquinas a sense of ‘a being in actuality’ that stands midway between the narrow and broad senses distinguished earlier: Three Senses of ‘A Being in Actuality’ • Narrow sense: A being that possesses actuality as a proper part or constituent. • Middle sense: A being that possesses actuality as any sort of part or constituent (whether proper or improper). • Broad sense: A being that possesses actuality in some way or other.

As should be clear, the different senses listed here are in increasing order of generality. Thus, the narrow sense is one that applies only to compounds (or beings that possess actuality only via proper parthood or constituency); the middle sense is one that applies to both compounds and forms (or beings that possess actuality via any sort of parthood or constituency); and the broad sense is one that applies to all beings (or things that possess actuality via any sort of relation whatsoever, including not only parthood or constituency but also inherence). As should also be clear, the middle sense is the only one that is relevant for understanding the first of Aquinas’s two conditions on individuality. So much for the first point. The second point I want to make about the passage above is related to the first, and also has to do with the scope of Aquinas’s first condition on individuality. In particular, I want to note that insofar as this condition must be understood in terms of the middle sense of ‘a being in actuality,’ it does not apply to prime matter. Indeed, as we can now see, to say that prime matter lacks actuality through itself (per se) is just to say that it is not a being in actuality in either the narrow or the middle senses, but only a being in actuality in the broad sense. The final point I want to make about the passage above has to do with Aquinas’s second condition on individuality—and in particular its relationship to the first. Evidently, to qualify as an individual in the strict and proper sense, something must satisfy both conditions. And although Aquinas thinks that all forms and compounds satisfy the first condition, he does not think that they all satisfy the second. On the contrary, he insists that only a proper subset of such beings satisfies the second condition—namely, all and only those that are “divided from all other things which are (or can be) in the same species.” In adding this qualification, Aquinas seems to be requiring that individuals, in the strict and proper sense, must belong to a species that can possess more than one member. And indeed, in many contexts, he is happy to accept this requirement, and hence to speak as if

Aquinas on the Individuation of Substances

139

only things individuated by matter qualify as individuals.26 But as Aquinas also recognizes, this requirement has the odd consequence that incorporeal substances or spirits—such as God and the angels, which do not belong to a species admitting of more than a single member—are neither universal nor individual. In certain contexts, therefore, Aquinas is prepared to relax his strict conception of individuality, to allow for anything that both satisfies its first condition and is distinct from all other beings to count as individual.27 Even so, the point I want to make here and that will be important in what follows is this: the reason that Aquinas associates both of his two conditions with individuality is that he takes the paradigmatic notion of an individual to involve something individuated by matter—that is to say, something that belongs to a maximally determinate kind of which there can be more than one instance. All of these points have direct bearing on Aquinas’s conception of prime matter. Indeed, taken collectively they guarantee that prime matter satisfies neither of his two conditions on paradigmatic individuality, and hence that it does not even qualify as an individual in the relaxed sense that God and angels do. For insofar as prime matter is pure potentiality, it cannot be said to possess actuality in either of the senses necessary to satisfy the first condition. And insofar as the second condition is satisfied only by a proper subset of beings possessing actuality in the relevant senses, it cannot satisfy this condition either. In short, unlike all other beings, prime matter is not individual in anything but the thin or trivial sense of being non-universal—a sense that we can hereafter ignore. The non-individuality of prime matter certainly complicates the proper understanding of Aquinas’s hylomorphism. But I do not think it renders it unintelligible. On the contrary, I think this conception of prime matter can be understood in terms of the contemporary category of stuff—that is, a type of being that is not itself individual, but is often said to be incapable of existing apart from individuals. Indeed, I think Aquinas’s hylomorphism as a whole is best thought of as a mixed ontology of stuff and things—one in which prime matter is identified with non-individual stuff and said to serve as the ultimate substratum for two different sorts of thing (namely, forms and compounds).28 26 For a systematic overview of such contexts, see Wippel, Metaphysical Thought of Aquinas, 351–73. 27 See the second paragraph of the text from In CA 1.9 cited in n. 17, as well as ST 1.29.4, where Aquinas describes an individual as “that which is undivided in itself and distinct from others.” See also the brief discussion of this more relaxed notion of individuality, and the contexts in which it occurs, in Wippel, Metaphysical Thought of Aquinas, 373–5. 28 For further development of this understanding of Aquinas’s hylomorphism, as well as some comparison of it with the views of Ned Markosian, perhaps the chief contemporary proponent of a mixed ontology, see my Aquinas’s Ontology of the Material World.

140

Jeffrey E. Brower

But does not all of this simply bring us back to the problem of individuality—that is, the problem of how something non-individual can play the role of substratum or individuator? No, it does not. In fact, I think we are now in a position to see that the appearance of a problem arises from a tendency to conflate different issues associated with individuation. For the sake of clarity, let us distinguish, on Aquinas’s behalf, two different questions: Two Questions about Individuation Q1 What accounts for the numerical distinctness of substances belonging to the same species? Q2 What accounts for the individuality of substances belonging to the same species?

Although Q1 and Q2 both raise issues of individuation, it is really only Q1 that Aquinas means to be answering when he identifies prime matter as the principle of individuation. Initially, it might seem odd to distinguish questions about numerical distinctness from questions about individuality. But in the context of a mixed ontology such as Aquinas’s, this makes perfect sense. For one thing, we have seen that mere numerical distinctness is not sufficient for individuality in this context. On the contrary, Aquinas thinks that something qualifies as individual only if it also possesses being in actuality through itself. But even if prime matter (or stuff) lacks being in actuality through itself, and hence fails to quality as an individual, it does not follow that the prime matter (or stuff ) associated with distinct substances is numerically the same. On the contrary, it is natural to suppose that the prime matter (or stuff ) associated with distinct substances is itself distinct. Indeed, we might speak here of such substances as possessing distinct portions of prime matter (or stuff ).29 But again, even if such portions are numerically distinct, for Aquinas, they are still non-individual beings. For the same reason, if they are to play a role in individuation, it must be as a principle of numerical distinctness, not individuality. It is important to see that there is nothing absurd about something non-individual playing the role of an individuator in this sense. Indeed, provided we conceive of prime matter (or stuff) as coming in portions that are themselves primitively distinct—that is to say, distinct in and of themselves—there would seem to be nothing to prevent us from explaining the numerical distinctness of substances in the same way that the 29 This talk of portions is introduced merely for the sake of convenience and could easily be dispensed with in favor of proper names or descriptions, along the lines suggested by Ned Markosian in his “Simples, Stuff, and Simple People,” Monist 87 (2004), 405–28, esp. 409.

Aquinas on the Individuation of Substances

141

contemporary substratum theorists explain the numerical distinctness of objects sharing the same properties—namely, by appealing to the primitive distinctness of their substrata. But this appears to be precisely how Aquinas himself is conceiving of prime matter.30 Indeed, on his view we might say that, whereas the portions of prime matter associated with distinct material substances are primitively distinct, the substances themselves, as well as their substantial forms, are derivatively distinct—that is to say, distinct solely in virtue of their relationship to the primitively distinct portions of prime matter that serve as their substrata or individuators.31 Of course, Aquinas cannot appeal to prime matter alone to answer Q2. But as far as I can tell, he never meant to. Indeed, Q2 seems not to be a question that he devotes much explicit attention to. Still, given that he takes individuals to be a type of being in actuality, and also traces the actuality of substances to their substantial form, we would expect his answer to Q2 to make reference to substantial form.32 But given that he also takes the substantial forms of substances belonging to the same species to be themselves distinct solely in virtue of their relation to prime matter, we might expect that even when it comes to answering Q2, prime matter will still have an important role to play. Indeed, it may even be that this role is one of the things that Aquinas had in mind when he describes prime matter as “the primary principle of individuation.” For immediately after stating the two conditions on individuality quoted above, Aquinas draws the following conclusion: Hence, matter is the primary principle of individuation. For it is [only] by matter that something acquires being in actuality—where the being in actuality is associated with any type of form, substantial or accidental. (In Sent. 4.12.1.1.3 ad 3)

In any case, insofar as Aquinas’s appeal to prime matter is mainly intended to form part of his answer to Q1, I think that the problem of individuality ultimately raises no more of a threat to the coherence of his views than does the problem of actuality. 30 See the last sentence of the first paragraph of the text from In CA 1.9 cited in n. 17. See also my Aquinas’s Ontology of the Material World, esp. 113–29, for evidence that Aquinas’s views about change require primitively distinct portions of prime matter. 31 All of this helps, I think, to resolve one of the puzzles that Kathrin Koslicki raises for Aquinas’s view (see Koslicki, “Essence and Identity,” n. 23). 32 The need to appeal to some form of actuality to answer Q2 may go some distance toward justifying the suggestion of some recent commentators that the source of individuality, for Aquinas, is bound up with his doctrine of being or esse. This suggestion traces to Joseph Owens, “Thomas Aquinas (b. ca. 1225; d. 1274),” in Gracia (ed.), Individuation in Scholasticism, 173–94. But see also the discussion in Kevin White, “Individuation in Aquinas’s Super Boethium De Trinitate, Q. 4,” American Catholic Philosophical Quarterly 69 (1995), 543–56. For critical response, see Dewan, “Individual as a Mode of Being” and Wippel, Metaphysical Thought of Aquinas, 373–5.

142

Jeffrey E. Brower 3 . Q U A N T I T Y A S T H E SE C O N D A R Y P R I N C I P LE O F IN D I V I D U A T I ON

If my argument to this point has been successful, Aquinas’s motivation for appealing to prime matter as part of his solution to the chief problem of individuation should be clear. For he takes the problem itself to be that of accounting for the numerical distinctness of substances belonging to the same species, and he takes prime matter to be the ultimate source of such distinctness. But if all this is correct, it might be wondered why he appeals to anything more than prime matter in his solution. Commentators have often regarded Aquinas’s appeal to quantity as puzzling. But the problem they have typically focused on has to do with its status as a type of accident. For insofar as accidents derive their being or identity from the substances to which they belong, they are logically posterior to such substances and hence must exist “after” they are already individuated.33 As I see it, however, the real problem is a more basic one—namely, that insofar as prime matter alone can account for the numerical distinctness of substances belonging to the same species, Aquinas’s appeal to any further principle, accidental or otherwise, seems superfluous. If Aquinas were thinking of the problem of individuation in terms of what we now typically mean by ‘numerical distinctness,’ then I think his appeal to prime matter would suffice for a complete solution. But as it turns out, this is not how Aquinas is thinking of the problem. On the contrary, for reasons I shall now explain, when he thinks about the problem of individuation, he almost always has in mind a phenomenon that can only be plausibly explained in terms of the conjunction of prime matter and quantity.

a. Two Kinds of Numerical Distinctness We have seen that, when it comes to understanding certain aspects of Aquinas’s views about individuation, it is important to distinguish questions about numerical distinctness from individuality. Something similar is true, I think, when it comes to understanding certain aspects of Aquinas’s views about numerical distinctness. Consider the following two variations on our earlier question about numerical distinctness: 33 For a clear statement of the problem, as well as the most common strategy suggested for resolving this problem on Aquinas’s behalf (namely, one that appeals to different orders of causality or explanation), see Wippel, Metaphysical Thought of Aquinas, esp. 364–5.

Aquinas on the Individuation of Substances

143

Two Questions about Numerical Distinctness Q1a What accounts for the non-identity of substances belonging to the same species? Q1b What accounts for the spatial distinctness (or separation) of substances belonging to the same species?

When Aquinas addresses the problem of individuation, it is tempting to think that he means to be addressing Q1a. Indeed, this is precisely what makes his appeal to quantity (or dimensions) seem so puzzling. For insofar as prime matter is sufficient by itself to account for the non-identity of substances belonging to the same species, the appeal to quantity (or for that matter, anything else) seems wholly superfluous. What I want to suggest, however, is that the question Aquinas really means to be addressing in the context of the chief problem of individuation is not Q1a but Q1b. If I am right about this, then his appeal to both prime matter and quantity is just what we would expect. For insofar as spatially distinct substances are nonidentical, it will still be necessary to appeal to prime matter as their principle of non-identity. But insofar as the appeal to prime matter cannot by itself explain their spatial distinctness, an appeal to something further will also be necessary—and in the medieval context, this further something will have to be very much like quantity or dimensions.34 Of course, the main difficulty for this suggestion is just the implausibility of saying that numerical distinctness has anything in particular to do with spatial distinctness or separation (as opposed to mere non-identity). It seems obvious that objects are numerically distinct just in case they are distinct in number. But is it not equally obvious that objects are distinct in number just in case they are non-identical? Although an affirmative answer might now seem all but inescapable, a little historical perspective reveals that this was not always the case. In the broadly Aristotelian context in which Aquinas was working, it was standard to associate number, and hence the notions of one and many, with quantity. Nor is it hard to see why. To say that there is a number of Fs just is to say there is a quantity of Fs. But ‘quantity’ is the name of one of Aristotle’s 34 For some discussion of the relation of quantity to spatial location in the medieval context in general, and Aquinas in particular, see Joseph Bobik, “Matter and Individuation,” in E. McMullin (ed.), The Concept of Matter in Greek and Medieval Philosophy (Notre Dame, IN: University of Notre Dame Press, 1965), 277–88 and “Dimensions in the Individuation of Bodily Substances,” Philosophical Studies 4 (1954), 60–79. See also Robert Pasnau, Metaphysical Themes 1274–1671 (Oxford: Oxford University Press, 2012), esp. chs. 14 and 17.

144

Jeffrey E. Brower

nine accidental categories of being. Aquinas accepts the standard medieval association of number with Aristotelian quantity, along with the view that Aristotelian quantity is an accident that can only be possessed by corporeal beings. He recognizes, however, that this raises special difficulties when it comes to making sense of the “multitude” (multitudo) of incorporeal beings required by Scripture and tradition (in particular, God and the angels). In response to such difficulties, Aquinas adopts the standard medieval strategy for resolving them—namely, to insist that although number, in the strict and proper sense, must be understood in terms of the Aristotelian category of quantity, there is an extended notion of number—or better, an extended notion of unity or oneness—that can be applied to anything that is both identical to itself and not identical to anything else. Aquinas refers to the latter sort of unity or oneness as transcendental, since it transcends the sorts of limitation associated with particular Aristotelian categories of being. Thus, speaking of the possibility of a multitude of angels in particular, he says: There are two senses of unity. In one sense, unity is interchangeable with being. And since being is not limited to any particular category, neither is the sort of multitude following on such unity. This is the sort of multitude that is found in angels. And in their case it follows not on any division of quantity or matter, but only on the distinction of their natures. It is in another sense, however, that unity is the principle of number—that is, discrete quantity. This sort of unity is caused by the division of matter or something continuous and is found only in material beings. (In Sent. 2.3.1.3 ad 1)

Although Aquinas allows for a type of distinctness that follows on the sort of unity that is “interchangeable with being,” he does not take it to be the same as that which follows on the sort of unity that is “the principle of number.” On the contrary, as passages such as this one make clear, he takes properly “numerical” distinctness to be that which follows on the “division of quantity or matter” and hence “is found only in material beings.”35

35 As such language indicates, there is a close connection for Aquinas between quantity and divisibility. Indeed, when Aquinas speaks of prime matter as lacking divisibility apart from quantity (e.g., In Sent. 2.3.1.1), I think that he means to be indicating that such matter lacks parts that are spatially extended (or separated), not that such matter lacks parts as such. For a defense of the view that prime matter not only has parts, for Aquinas, but is such that each of its parts has proper parts, see my Aquinas’s Ontology of the Material World, 113–19. For a very different understanding of Aquinas’s views, see Pasnau Metaphysical Themes, chs. 4 and 14. (To adopt Pasnau’s own terminology, I think Aquinas holds something like “the extensionless parts view” of matter together with “the B theory of quantity,” whereas Pasnau himself attributes to Aquinas “the simple view” of matter together with “the A theory of quantity.”)

Aquinas on the Individuation of Substances

145

In light of all this, I think it is quite plausible to suppose that what Aquinas regards as the chief problem of individuation concerns something more than the mere non-identity of substances. But in order to see exactly how the “numerical distinctness” of substances relates to their spatial separation, I need to say a bit more about how Aquinas is thinking of prime matter and quantity in relation to one of the distinguishing features of material substances—namely, their corporeality.

b. Corporeality and Spatial Separation We have already noted that Aquinas takes the possession of prime matter to be that which distinguishes material substances (or bodies) from incorporeal substances (or spirits). It is not surprising, therefore, that he would take there to be a special connection between the possession of prime matter and corporeality. Thus, speaking of matter in the primary or proper sense, he says: Matter can be found only in those things that have a potentiality with respect to place (ad ubi). But the only things of this sort are corporeal objects (corporalia), which are circumscribed by place (loco circumscribuntur). Hence, matter can be found only in corporeal objects. (QDA 6)

In this passage, as elsewhere, Aquinas emphasizes the close connection that he takes to exist between prime matter and corporeality. But here he also makes especially clear how he is thinking of corporeality itself. For corporeal objects, he tells us, are the only things that “have a potentiality with respect to place”—or more precisely, the only things that have a capacity for being “circumscribed by” place. Like other medieval philosophers, Aquinas appeals to the notion of “circumscription” to clarify the special way in which corporeal objects are related to place. Aquinas is happy to speak of any substance, whether corporeal or incorporeal, as being located at the place to which its activity extends. Indeed, he thinks of God as omnipresent in this sense.36 Even so, Aquinas insists that only corporeal substances “possess extension in three dimensions” (habent trinam dimensionem) and hence “are circumscribed by place” (loco circumscribuntur).37 As all of this helps to make clear, what Aquinas has in mind by corporeality is capacity for spatial extension (or extension with respect to place).38

36

37 See, e.g., ST 1.8.2. See, e.g., ST 1.3.1 obj. 1; 1.52.1–2; and In BDT 4.3. For some further complications regarding Aquinas’s understanding of corporeality, see my Aquinas’s Ontology of the Material World, 204, n. 11. 38

146

Jeffrey E. Brower

Although the possession of prime matter is what gives material substances their capacity for spatial extension, Aquinas thinks that the capacity itself cannot be actualized apart from the possession of quantities (or dimensions). Again, speaking of prime matter, he says: Matter is [actually] related to place [only] insofar as it is subject to dimensions. And so it is from the nature of matter as subjected to dimensions that multiple bodies are prevented from being in the same place. (In BDT 4.3)

In the previous passage, Aquinas emphasized the nature of the capacity that prime matter confers on its possessors. In this passage, he emphasizes the conditions under which this same capacity is actualized—namely, when the matter that does the conferring is “subjected to dimensions.” Aquinas’s talk of “subjection” here might suggest that he takes dimensions to be forms inhering directly in prime matter. But that is misleading. For Aquinas, prime matter is a substratum only for substantial forms, whereas quantities (or dimensions) are accidental forms, and hence can only have substances for their substrata. Evidently, therefore, when Aquinas talks of matter as “subjected to dimensions,” this is shorthand for its being “part of a material substance that possesses dimensions.” As all of this helps to make clear, Aquinas takes there to be a close connection between corporeality and both prime matter and quantity. Corporeality itself is to be understood in terms of a capacity for spatial extension. And whereas prime matter is what confers this capacity, quantities (or dimensions) are what actualize or manifest this same capacity. Indeed, since Aquinas takes spatial extension to consist in the possession of quantitative dimensions, it would be better to say that the capacity which is conferred by prime matter, and in terms of which corporeality is to be understood, just is the capacity to possess quantitative dimensions. Finally, it must be noted that, for Aquinas, the quantities or dimensions in virtue of which things are spatially extended are always associated with distinct and incompatible locations. Indeed, as he tells us in the final sentence of the passage just quoted, it is precisely insofar as distinct material substances or bodies possess such quantitative dimensions that they are “prevented from being in the same place.”39 But as this just goes to show, the possession of quantitative dimensions by a material substance guarantees its spatial separation or distinctness from all other corporeal things.40 39

Compare also ST 1.8.2. Insofar as Aquinas takes quantities to be the source of all spatial extension and distinctness, neither material substances nor the matter of which they are composed can be said to possess such extension or distinctness in and of themselves. Does this imply that 40

Aquinas on the Individuation of Substances

147

In the end, therefore, I think there can be no doubt that what Aquinas regards as the chief problem of individuation concerns not merely the nonidentity of substances but their spatial separation as well. Indeed, if we return to this problem with Aquinas’s views about “numerical distinctness” in mind, we can see that the problem itself can be stated most perspicuously as follows: The Chief Problem of Individuation—Restated What accounts for the spatial distinctness (as opposed to mere non-identity) of substances belonging to the same species?

Once the problem is stated in this way, however, the coherence of Aquinas’s solution, as well as its motivation, becomes perfectly clear: Aquinas’s Account of Individuation—Clarified Prime matter accounts for the non-identity of substances belonging to the same species, and prime matter and quantity (or dimensions) jointly account for their spatial distinctness.

As this clarification of Aquinas’s solution suggests, prime matter counts as the primary principle of individuation at least partly because it has more work to do than quantity when it comes to individuating substances. For whereas both prime matter and quantity together explain the spatial distinctness of substances belonging to the same species, prime matter alone explains their non-identity. As it turns out, however, there is another respect in which prime matter counts as the primary principle of individuation. For insofar as quantities are accidents, and accidents derive their being or identity from the substances to which they belong, prime matter must do some of its individuating work (namely, that of explaining the non-identity of substances) “before” quantity can even exist to do its work (namely, that of partially explaining the spatial distinctness of these same substances).

material substances, along with their matter, could in principle exist without any spatial extension or separation? Not necessarily. Elsewhere, I have argued that neither material substances nor their matter can come into existence, for Aquinas, without quantities (see my “Form, Matter, and Individuation”). But whether such substances and their matter could continue in existence without any quantities (say, by the absolute power of God) is a further question on which I do not mean to take a stand here.

148

Jeffrey E. Brower 4 . C O N C LU S I O N

This completes my discussion of Aquinas’s account of individuation. If my interpretation is correct, then prime matter and quantity have very different roles to play in this account. Indeed, prime matter has a principal or primary individuating role to play—namely, that of accounting for the non-identity of substances belonging to the same species—whereas quantity has a subordinate or secondary individuating role to play—namely, that of accounting for the spatial distinctness (or separation) of these same substances. But insofar as these roles are perfectly compatible, both with each other and with the main features of Aquinas’s broader metaphysics and natural philosophy, their proper understanding enables us to appreciate the coherence of his account of individuation as a whole. In the course of developing my interpretation, I have attempted to explain those features of Aquinas’s account that have made it seem so puzzling to previous commentators, and hence given rise to the standard questions about its coherence. Part of the explanation, I have suggested, owes to Aquinas’s association of prime matter with “pure potentiality,” and part of it owes to his association of individuation with “numerical distinctness.” For it is easy to be misled by both associations. Indeed, if we follow previous commentators in thinking of pure potentiality as the lack of being or actuality as such (rather than as the lack of being or actuality via constituency), then we will struggle to find any role for prime matter to play in individuation. And if we follow these same commentators in thinking of numerical distinctness primarily in terms of non-identity (rather than spatial distinctness), then we will struggle to see how an accident such as quantity could account, in any way, for the “division” or “multiplication” of what is prior to it in being (namely, material substances or their matter).41 In each case, however, I have argued that these ways of thinking, and the puzzles to which they give rise, result from a misunderstanding of Aquinas’s views. Although my main purpose has been to show that the standard questions surrounding the coherence of Aquinas’s account of individuation can all be answered, once we approach them from the proper perspective, I have also attempted to establish along the way a number of claims that are important 41

Interestingly, these considerations have led some commentators to express puzzlement as to why Aquinas ever appealed to anything other than quantity in his account of individuation. See, e.g., Bobik, “Matter and Individuation,” esp. 285, where he asks: “why not say that quantity alone is the principle of individuation?; why say, as some do, that matter is the principle of individuation along with quantity?”

Aquinas on the Individuation of Substances

149

quite apart from their connection to this account. In particular, I have attempted to put to rest a longstanding worry about the coherence of Aquinas’s views about prime matter (namely, its identification with “pure potentiality”), as well as to draw out certain features of his broader metaphysics and natural philosophy that have often been overlooked (namely, the overall structure of his hylomorphism, the non-individuality of prime matter, and certain ambiguities associated with his understanding of numerical distinctness, division, and multiplication). Even if I have not succeeded in establishing all of these claims, I hope to have done enough to show that Aquinas’s views about individuation, and in particular his understanding of the roles that prime matter and quantity play in them, are more subtle and nuanced than he is often given credit for.42 Purdue University

BIBLIOGRAPHY Bobik, Joseph. “Dimensions in the Individuation of Bodily Substances,” Philosophical Studies 4 (1954), 60–79. Bobik, Joseph. “Matter and Individuation,” in E. McMullin (ed.), The Concept of Matter in Greek and Medieval Philosophy (Notre Dame, IN: University of Notre Dame Press, 1965), 277–88. Brower, Jeffrey E. “Matter, Form, and Individuation,” in B. Davies and E. Stump (eds.), The Oxford Handbook of Aquinas (Oxford: Oxford University Press, 2012), 85–103. 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 92 (2016), 715–35. Dewan, Lawrence. “The Individual as a Mode of Being According to Thomas Aquinas,” The Thomist 63 (1999), 403–24. Gracia, Jorge J. E. (ed.). Individuation in Scholasticism: The Later Middle Ages and the Counter-Reformation, 1150–1650 (New York: State University of New York Press, 1994).

42 Earlier versions of this paper were presented at the University of Delaware, the 2014 Toronto Colloquium in Medieval Philosophy, and the 2016 Morris Colloquium on Metaphysics and its History at the University of Colorado, Boulder. I am grateful to audiences on all three occasions for stimulating comments and criticism, and especially to Jorge Gracia and Tom Donaldson who served, respectively, as my commentators on the second and third occasions. I am also grateful to Michael Bergmann, Susan BrowerToland, Lindsay Cleveland, and two anonymous referees for detailed comments, criticism, and advice.

150

Jeffrey E. Brower

Hughes, Christopher. “Matter and Individuation in Aquinas,” History of Philosophy Quarterly 13 (1996), 1–17. Hughes, Christopher. “Matter and Actuality in Aquinas,” Revue Internationale de Philosophie 52 (1998), 269–86. King, Peter. “The Problem of Individuation in the Middle Ages,” Theoria 66 (2000), 159–84. Koslicki, Kathrin. “Essence and Identity,” in M. Dumitru (ed.), Metaphysics, Meaning and Modality: Themes from Kit Fine (Oxford: Oxford University Press, forthcoming). Markosian, Ned. “Simples, Stuff, and Simple People,” Monist 87 (2004), 405–28. Owens, Joseph. “Thomas Aquinas (b. ca. 1225; d. 1274),” in J. J. E. Gracia (ed.), Individuation in Scholasticism: The Later Middle Ages and the Counter-Reformation, 1150–1650 (New York: State University of New York Press, 1994), 173–94. Pasnau, Robert. Metaphysical Themes 1274–1671 (Oxford: Clarendon Press, 2011). Pickavé, Martin. “The Controversy over the Principle of Individuation in Medieval Quodlibeta (1277–1320): A Forrest Map,” in C. Schabel (ed.), Theological Quodlibeta in the Middle Ages: The Fourteenth Century (Leiden: Brill, 2007), 17–79. Pini, Giorgio. “The Individuation of Angels from Bonaventure to Duns Scotus,” in T. Hoffmann (ed.), A Companion to Angels in Medieval Philosophy (Leiden: Brill, 2012), 79–115. Plantinga, Alvin. “On Existentialism,” Philosophical Studies 44 (1983), 1–20. Spade, Paul V. “Degrees of Being, Degrees of Goodness: Aquinas on Levels of Reality,” in S. MacDonald and E. Stump (eds.), Aquinas’s Moral Theory: Essays in Honor of Norman Kretzmann (Ithaca, NY: Cornell University Press, 1999), 254–75. Thomas Aquinas. S. Thomae Aquinatis Doctoris Angelici Opera Omnia (Rome: Commissio Leonina, 1882–). Cited works: Vols. 4–12: Summa theologiae; Vol. 24.2: Quaestio disputata de spiritualibus creaturis; Vol. 43: De principiis naturae, De ente et essentia; Vol. 45.1: Sentencia libri De anima; Vol. 50: Super Boetium De Trinitate. Thomas Aquinas. Scriptum super libros Sententiarum, ed. P. Mandonnet and M. F. Moos (Paris: P. Lethielleux, 1929–47). Thomas Aquinas. Super librum De Causis expositio, ed. H. D. Saffrey (FribourgLouvain: Société Philosophique, 1954). White, Kevin. “Individuation in Aquinas’s Super Boethium De Trinitate, Q. 4,” American Catholic Philosophical Quarterly 69 (1995), 543–56. Wippel, John F. The Metaphysical Thought of Thomas Aquinas: From Finite Being to Uncreated Being (Washington, DC: Catholic University of America Press, 2000).

Peter Auriol on the Intuitive Cognition of Nonexistents Revisiting the Charge of Skepticism in Walter Chatton and Adam Wodeham Han Thomas Adriaenssen

Recent years have seen a welcome revival of interest in the role of skeptical arguments in later medieval philosophy.1 As studies have shown, scholastic treatments of knowledge and cognition show a keen interest in the skeptical problems, and by the turn of the thirteenth and fourteenth centuries, skepticism had become a ‘main focus’ of philosophical inquiry.2 One key figure in this context was the fourteenth-century Franciscan philosopher and theologian Peter Auriol (1280–1322).3 Although Auriol never thought of himself as a skeptic, two of the signature claims of his philosophy soon came to be seen as major risks to certain knowledge of the world. The first is a claim about the objects of cognition. According to Auriol, the objects of cognition have a special kind of subject-dependent being. They have what he calls apparent being. The second is a claim about sensory cognition, or what he calls intuitive cognition. According to Auriol, although intuitive cognition normally comes with a real and existent object, 1 In particular, see Dominik Perler, Zweifel und Gewissheit: Skeptische Debatten im Mittelalter, 2nd edition (Frankfurt: Vittorio Klöstermann, 2012). Also Christophe Grellard, “Comment peut-on se fier à l’expérience? Esquisse d’une typologie des réponses médiévales au scepticisme,” Quaestio 4 (2004), 113–35. 2 Henrik Lagerlund, “A History of Skepticism in the Middle Ages,” in H. Lagerlund (ed.), Rethinking the History of Skepticism: The Missing Medieval Background (Leiden: Brill, 2010), 1. 3 On Auriol’s role in the history of skepticism, see Charles Bolyard, “Medieval Skepticism,” in The Stanford Encyclopedia of Philosophy (Spring 2013 Edition), ed. Edward Zalta, §4 and Russell Friedman, “Peter Auriol,” The Stanford Encyclopedia of Philosophy (Fall 2015 Edition), ed. Edward Zalta, §3.

152

Han Thomas Adriaenssen

it need not do so. God can bring about intuitions without existent objects, and even without his intervention natural factors can cause intuitive cognitions to lack existent objects. In the eyes of early readers like fellow Franciscans Walter Chatton (1290–1343) and Adam Wodeham (1295–1358), these two claims were closely linked. In particular, both believed that the ontology of apparent being was a major part of why Auriol believed that there could be natural intuitions of nonexistents. But the claim that there could be natural intuitions of nonexistents directly led to skepticism, they found. On this reading of Auriol, the ontology of apparent being put him on the road to the natural intuition of nonexistents, and thus to skepticism. Modern commentators have offered similar readings. According to Rega Wood for example, the theory of apparent being was what ‘forced’ Auriol and his readers to consider the natural intuition of nonexistents.4 And according to Richard Cross, the natural intuition of nonexistents has “understandably been seen to usher in some kind of skepticism.”5 The apparent being opens up the possibility of natural intuitions of nonexistents, and that possibility at its turn opens the door to doubt. But this paper argues that the kind of reading of Auriol we get in Chatton and Wodeham is problematic. In particular, the paper will make two claims. The first is that the theory of apparent being does not give any special reason to think that there could be natural intuitions of nonexistents. So even if Chatton and Wodeham are right to think that his theory of intuition leads to skepticism, it was not his account of the objects of cognition that first led Auriol down this road. The second claim is that the charge in Chatton and Wodeham, namely that Auriol’s theory of intuition leads to skepticism, is problematic. To be sure, his account of intuition and its object is different from the one they offer. Yet his account does not at all put at risk the kind of certain knowledge about external objects that, according to Chatton and Wodeham themselves, intuition affords. When it comes to the challenge of skepticism, Auriol is not at all worse off than his critics. The paper proceeds as follows. First, section 1 will take a closer look at Auriol’s notion of apparent being. Section 2 introduces the concept of intuitive cognition, and section 3 argues that the apparent being is not the source of the natural intuition of nonexistents. To address the charge of skepticism in Chatton and Wodeham sections 4 and 5 first look at the kind 4 Rega Wood, “Adam of Wodeham,” in J. Gracia and T. Noone (eds.), A Companion to Philosophy in the Middle Ages (Oxford: Blackwell, 2003), 77–85, 78. 5 Richard Cross, The Medieval Christian Philosophers: An Introduction (London: I. B. Tauris, 2014), 219.

Peter Auriol on the Intuitive Cognition of Nonexistents

153

of empirical certainty they themselves believe intuition affords. Section 6 then argues that there is no good reason to think that Auriol’s account of intuition puts in jeopardy this kind of certainty.

1. THE APPARENT BEING Suppose you are traveling by boat, and that, standing still on the deck, you look at a row of trees on the riverside. Because of the motion of the boat, every single tree that you look at will gradually pass out of sight. But because you are standing still on the deck, it will appear to you that it is the trees, rather than the boat, that are in motion. This is a common illusion, Auriol believed, and he wanted to know what happens when we fall prey to it. The starting point of his analysis is that, for every cognitive event, there is an act and an object of cognition. Here, the act is an act of vision, and the object of that act is motion. But that leads to a problem: After all there is no actual motion where you are looking. And for Auriol, that raises the question of what the ontological status of the motion that you see can be. In the following passage, he sorts through some options: When someone is transported over water, the trees that exist on the shore appear to move. This motion, which is in the eye as an object, cannot be said to be the vision itself; otherwise a vision would be the object of sight, and a vision would be seen, and vision would be a reflexive power. But it cannot be said to really exist in the tree or in the shore either, because then they would really move. Neither can it be said to be in the air, because motion is not attributed to the air, but to the tree. Therefore it only exists intentionally, not really, in seen being and in adjudged being.6

Here, in the second line, we are told that the motion you see is “in the eye as an object.” By this Auriol means that it appears to the eye, or that it appears as an object of vision. But although it is in the eye in this sense, the motion you see is not some inner state of your visual system. This I take to be the general point of his remark that “this motion . . . cannot be said to be the vision itself.” The motion you see cannot be the act of vision you engage in when looking at the shore, because then the object of vision would be an inner act of cognition rather than outward motion. And by the same token, the object you see cannot be some kind of species, or image, in the eye or in the power of vision. For if it were, the object of cognition would not be outer motion,

6

Scriptum 3.14 (Buytaert, 696).

154

Han Thomas Adriaenssen

but an inner species or image.7 But if the motion you see is not some inner state of the visual system, it needs to be something external. Auriol briefly goes over a few options. The first is that the motion you see somehow is in the air. But this cannot be true, for then the air rather than the trees would appear to move. Indeed, the motion you see can be nowhere but in the trees. If it were anywhere else, then something other than the trees would appear to be in motion. But surely, it cannot have a real being in the trees. If it did, the trees would really move. Hence, the remaining option is that it has a kind of less than real being in the trees. And this is what Auriol in the above passage calls ‘intentional being.’ According to Auriol, this being is causally dependent upon your act of vision, which “puts the thing in intentional being.”8 It is the production by your act of vision of motion in intentional being that makes the immobile trees on the shore appear to be in motion. And similar analyses are offered for other perceptual illusions. When a burning stick is quickly whirled in a dark night, the object of cognition is a fiery circle in the air. But that fiery circle is neither some inner cognitive state nor a real external object. Hence it needs to be an intentional object that is produced by your act of vision. Instead of speaking of intentional being, Auriol also speaks of apparent being. And although he introduces this being to provide a relevant object for sensory illusions like the ones above, he also holds that the apparent being is the being of objects that appear to us in general. When I see the coffee mug on my desk, for example, the mug appears to me. And according to Auriol, this means that my act of vision puts the mug in apparent being. Thus seeing the mug involves at least three beings: a. An act of vision b. The mug in apparent being c. The mug in real being9

7 A bit later on, Auriol rejects species as objects of the imagination for a similar reason. When I imagine my father, he reasons, the object of cognition is my father—not a species (Scriptum 3.14; Buytaert, 697–8). 8 Scriptum 3.14 (Buytaert, 696). 9 According to Perler, the mug in apparent being and the mug in real being are not two beings. To say that the mug has apparent being just is to say that it appears to me, but it is not to introduce a new category of being. See Dominik Perler, “What Am I Thinking About? John Duns Scotus and Peter Aureol on Intentional Objects,” Vivarium 86 (1994), 72–89. But this is hard to square with the claims in Auriol that objects as they appear “non haberent esse verum et principale, sed intentionale et deminutum,” where a strained reading seems required not to take him as marking out a special kind of being of what appears in cognition. See, for example, Scriptum 36.2 (Electronic Scriptum, ll. 287–8).

Peter Auriol on the Intuitive Cognition of Nonexistents

155

But this account provoked strong reactions from Auriol’s contemporaries. Thus according to Chatton and Wodeham, if our access to the outer world is mediated by apparent beings in this way, we will never see the things themselves. According to Chatton, the apparent being was like a ‘sign’ that must lead us to some other thing than itself.10 And Wodeham claimed that if all access to things is mediated by apparent beings, “there will as it were be an intermediary veil” between us and the world.11 It is not clear that these readings are entirely fair to Auriol. For on his account, the apparent mug is the same thing as the real mug, albeit in a different kind of being. The relation between apparent and real beings is thus different from the kind of sign relation that obtains between smoke and fire. And according to Auriol himself, the fact that it is one and the same mug that has both apparent and real being means that the former in fact unveils rather than veils the latter.12 Even so, Chatton and Wodeham are correct insofar as Auriol has moved away from a simple kind of direct realism where we have unmediated access to objects in real being. And by taking this step, they argued, he invited a host of skeptical problems. In particular, they found a close connection between the apparent being on the one hand, and the idea in Auriol that there can be natural intuitive cognitions of nonexistent objects on the other. And with this last idea, they claimed, Auriol directly put us on the road to skepticism.

2. THE INTUITION OF NONEXISTENTS The term ‘intuitive cognition’ appears to have been introduced into the vocabulary of scholastic psychology by Duns Scotus. According to Scotus, the simple or nonpropositional apprehensions that we have of things come in two basic kinds. On the one hand, there are those simple cognitions that represent existent objects as existent, or that represent present objects as present.13 These Scotus terms intuitive cognitions, and paradigmatic examples include acts of vision and of the other external senses.14 On the other hand, there are simple cognitions that as it were ‘abstract’ from the

10

Reportatio et lectura, prol. 2.2 (Wey, 87). Lectura secunda, prol. 4 (Wood, 92). 12 See Scriptum 27.2 (Electronic Scriptum, ll. 598–9). 13 Quodlibet 6.1 (Wadding XXV 243–4). 14 Scotus also appears to have allowed for an intellectual kind of intuition. But throughout this paper the focus will just be on the standard, sensory variety. 11

156

Han Thomas Adriaenssen

existence or presence of their objects, and these Scotus calls abstractive cognitions. These are cognitions that either pertain to absent or nonexistent objects or which, alternatively, pertain to objects that, though present or existent, are not represented as such. Paradigmatic examples of such abstractive cognitions include acts of the imagination: Some cognition is in itself of an existent object, like that which attains an object in its proper actual existence. For instance, the vision of a color, and, generally, acts of the external senses. The other cognition is also of an object, but not as it exists in itself. Rather, either the object does not exist, or at least that cognition is not of that object insofar as it actually exists. For instance, the imagination of a color.15

Scotus’s distinction between intuitive and abstractive cognition had a profound influence on medieval cognitive psychology, and after Scotus, many philosophers and theologians would adopt some version of his distinction. Indeed, as Katherine Tachau has pointed out, the history of medieval epistemology and psychology from Scotus on can in important respects “be traced as a development of this distinction.”16 Auriol was one of the thinkers who, in the wake of Scotus, categorized our simple apprehensions of objects into intuitive and abstractive cognitions. And again like Scotus, he took vision and imagination to provide the paradigmatic examples of intuition and abstraction. “Vision,” he wrote, is “maximally intuitive,” and abstractive cognition is perhaps best described as “imaginative cognition.”17 But although Auriol thus agreed with Scotus on the usefulness of a general distinction between two branches of simple apprehension, he was critical of the way in which Scotus had drawn the boundaries between intuition and abstraction. In particular, Auriol was critical of the claim in Scotus that all intuitions have existent objects. To be sure, Scotus was right to think that my intuitive cognition of a tree will normally be caused by an actual tree in my visual field, but what he had failed sufficiently to appreciate, Auriol thought, is that whatever a creature can do, God can do too. Thus, even when there is no actual tree around to be seen, it is possible for God to bring about in me the very act of intuition that would normally have been caused by a tree in front of me. Intuitive cognition of a nonexistent object is a supernatural possibility, and so, Auriol concludes, Scotus’s claim that intuition is “in itself of an existent object” cannot go through.

15

Quodlibet 13.2 (Wadding XXV 521). Katherine Tachau, Vision and Certitude in the Age of Ockham: Optics, Epistemology, and the Foundations of Semantics (Leiden: Brill, 1988), 81. 17 See Scriptum, prol. 2 (Buytaert, 204–5). 16

Peter Auriol on the Intuitive Cognition of Nonexistents

157

Auriol was not alone in allowing for supernaturally induced intuitions of nonexistents. Thus as we will see later on, Chatton and Wodeham will do the same. But where Auriol differs from them, is that for him the intuitive cognition of nonexistents is not just a supernatural, but also a natural possibility. Independently of any sort of divine intervention, that is, Auriol thinks that our external senses can by nature engage in acts of seeing, hearing, smelling, tasting, and touching absent and nonexistent objects.18 Why did he think this? According to some of his early readers, this idea in Auriol was linked with his ontology of apparent being. And indeed as we have seen above, Auriol believed that the apparent being made possible the cognition of nonexistent things in a way that other theories could not. The theory of apparent being provided relevant objects for such cognitions, and thus succeeded where competitors like the species theory failed. As a result the theory of apparent being and the natural intuition of the nonexistent are linked at least in the sense that if there are going to be natural intuitions of nonexistents, they will have apparent beings as their objects.19 But in early readers like Chatton and Wodeham, we also get the suggestion that they are linked in the further sense that if the objects of cognition are going to be apparent beings, there will be natural intuitions of nonexistents. Thus when Chatton inquires “whether there naturally is intuitive cognition without the presence of the object seen,” the first argument that he presents in favor of an affirmative answer is that there is one opinion according to which . . . an intuitive cognition constitutes the thing in some kind of apparent, objective being.20

As Tachau has commented, Chatton here took the apparent being to be ‘the crucial premise’ for the natural possibility of intuitions with nonexistent objects.21 Again, when Wodeham asks “whether sensory or intellectual intuition can naturally be caused or sustained without the existence of the thing that is

18

Scriptum, prol. 2 (Buytaert, 200). Insofar as apparent being is perhaps itself some kind of existence, strictly speaking we may have to say that the objects of these intuitions are unreal rather than nonexistent. But I will throughout this paper stick to Auriol’s and his critics’ own parlance, and speak of intuitive cognitions of nonexistents. 20 Reportatio et lectura, prol. 2.2 (Wey, 86). 21 For this qualification, see Tachau, Vision and Certitude, 186. 19

158

Han Thomas Adriaenssen

seen,” the first argument that gets cited for an affirmative answer is that the objects of vision have apparent being: It seems to suffice for an act of vision that its object have seen being. But seen being is not the same as the existent object, because it is separable from it.22

Like Chatton before him, Wodeham is here suggesting that if the objects of cognition are apparent beings, this gives us some kind of reason to think that there will as a matter of natural fact be intuitive cognitions of nonexistent objects.23 But the idea that there could be natural intuitions of nonexistent objects, according to Chatton and Wodeham, will lead to skepticism. Ever since Scotus, intuition was seen as a major source of certainty about the world.24 Abstractive cognition of concepts sufficed to grasp such principles as that the whole is greater than its parts, but to come to know with certainty that some whole or part exists, intuition was needed. As Scotus had put it: without intuitive cognition, “the intellect would not be certain of the existence of any object.”25 So if intuitive cognition should turn out to be unreliable, this would be a serious blow to our certainty about the world we live in. Such, at any rate, appears to have been the reasoning of Chatton. Perhaps taking his cues from Scotus, Chatton declared that intuitive cognition was our main source of certainty about the world, writing that “our greatest certainty about sensible things we have because we are aware of our perceptions.”26 But if as a matter of natural fact, intuitive cognition could fail to trace existent objects in the world all of our certainty about the sensible things in the world “would perish,” Chatton feared.27 In the same vein, Wodeham claimed that intuition is “necessarily required” for all knowledge of matters of fact.28 Without it, we could perhaps grasp all kinds of conceptual truths, but not know that the sun has risen, or that it is being eclipsed. But if natural intuition could fail to trace existent objects, it would dry up as a source of certainty, and we would still have no way of knowing anything about the current behavior of the sun.

22

Lectura secunda, prol. 3 (Wood, 65). Tachau (Vision and Certitude, 287) points to Chatton as a possible influence on Wodeham here. 24 See Stephen Dumont, “John Duns Scotus,” in J. Gracia and T. Noone (eds.), A Companion to Philosophy in the Middle Ages (Oxford: Blackwell, 2003), 353–69, especially 364. Also Richard Cross, Duns Scotus’s Theory of Cognition (Oxford: Oxford University Press, 2014), 47. 25 Ordinatio 4.45.2 (Vaticana XIV 157–8). 26 Reportatio et lectura, prol. 2.2 (Wey, 89). 27 Reportatio et lectura, prol. 2.2 (Wey, 89). 28 Lectura secunda, prol. 2 (Wood, 44). 23

Peter Auriol on the Intuitive Cognition of Nonexistents

159

If natural intuitions could lack existent objects, “then there would be no certainty in such cognitions, and then all knowledge that is acquired through experience would perish.”29 In the critical treatment of Auriol that we get in Chatton and Wodeham, we thus find two basic ideas that will need to be looked at in more detail. The first is that the apparent being gives us a reason to believe that there could be natural intuitions of nonexistent objects. The second is that if there can be natural intuitions of nonexistent objects, skepticism looms large. On the picture that we get in these Franciscans then, it was via the natural intuition of nonexistents that the apparent being had put Auriol on a dark path to skepticism. Similar ideas have been put forth by modern commentators. Thus Tachau has described the natural intuition of nonexistent objects as a ‘consequence’ of the view that the objects of cognition are apparent beings.30 And according to Rega Wood, it was with his claim that “the objects of cognition are apparent beings, not things themselves” that Auriol “forced his contemporaries to consider the possibility of naturally produced cognition of nonexistents.”31 As for the idea that the possibility of such cognition puts us on the road to skepticism, Richard Cross in a recent book has argued that, if natural intuitions can come without existent objects, the question will arise how we will tell true from false intuitions. But Auriol does not give us any way to pick out intuitions that lack a real object. As a result, Cross claims, it is understandable that Auriol’s theory of intuition has been perceived as a source of skeptical questions ever since the early reception of his thought.32 But even if Auriol’s readers ever since the fourteenth century have found it plausible that the apparent being gives us reason to think that there could be natural intuitions of nonexistent objects, and that this possibility at its turn puts us on the road to skepticism, the following sections will argue that both of these ideas need correction, or at least qualification. Thus section 3 will argue that, as such, the theory of apparent being does not give us a special reason to think that there could be intuitive cognitions of nonexistent objects. And as sections 4 to 6 will go on to point out, the natural intuition of the nonexistent in Auriol does not leave him any more vulnerable to skepticism than critics like Chatton and Wodeham are themselves.

29

Lectura secunda, prol. 3 (Wood, 66). Katherine Tachau, “The Response to Ockham’s and Aureol’s Epistemology (1320–1340),” in A. Maierù (ed.), English Logic in Italy in the 14th and 15th Centuries (Naples: Bibliopolis, 1982), 185–217, 200. 31 32 Wood, “Adam of Wodeham,” 78. Cross, Christian Philosophers, 219. 30

Han Thomas Adriaenssen

160

3 . A U R I O L O N T HE IN T U I T I O N OF NONEXISTENTS In this section, I will argue that the theory of apparent being is neither necessary nor sufficient to arrive at the conclusion that there could be natural intuitions of nonexistent objects. And the best starting point to do that, is to look at Auriol’s own defense of the natural intuition of such objects. As we found before, Auriol treats visual cognition as the paradigm case of intuitive cognition. But, he claims, there are a number of experiences teaching us that vision can in natural circumstances come without existent objects: Afterimages. According to Auriol, the case of afterimages shows that the vision of a luminous object can remain for some time after its disappearance. Dreams. Following Averroes, Auriol holds that dreams are cases where the dreamer sees objects that are not really there. Apparent horrors. When we are in the grips of intense fear, we hear the screams of dead spirits and see ghosts that are not there. Illusions. When we are tricked by illusions, Auriol thinks we see things that are not really there, like broken sticks. Distorted color vision. Due to an indisposition of the eye, we may see a shade of red when no red body is around.33 On the basis of these cases, Auriol concludes that intuition cannot be defined as the simple apprehension of an existent object as existent. Rather, the hallmark of intuition is that, independently of whether or not its object exists, its object is made to appear in an unmediated way as actual, present, and existent.34 Auriol’s case for the natural intuition of nonexistents thus crucially hinges on the idea that visions come in two kinds: those that do, and those that do not have real objects. Though the former may be the default, the latter kind of vision tends to occur in dreams and fevers, or when we are in the grips of intense emotions. Other than philosophers for whom ‘see’ is a success verb in that just as one cannot know what is false one cannot see what is not there, then, for Auriol seeings can lack existent objects without ceasing to count as seeings. 33 34

Auriol cites these cases in Scriptum, prol. 2 (Buytaert, 199). See Scriptum, prol. 2 (Buytaert, 205).

Peter Auriol on the Intuitive Cognition of Nonexistents

161

Auriol was aware, however, that an opponent might object that illusions and the like are not cases in which the eye sees, but merely cases where it seems to the subject that the eye sees: People who are deceived in this way do not see, but it appears to them that they see, because the common sense so judges.35

But in response to this objection, Auriol points out that the common sense will need some kind of trigger to judge that an object x is seen. And the only thing that will trigger the common sense to judge that x is seen, he reasons, is the fact that x appears as an object of vision. But whenever x appears as an object of vision, it is in fact seen: If the common sense judges that the eye sees, it is necessary that there be something in the eye that it judges, namely the appearance of a thing. But the appearance of a thing in sight is the vision itself.36

Auriol repeats the point that whenever x appears as an object of vision, x is seen later on: Vision is nothing but some kind of appearance.37

But if x is seen whenever it appears to me as an object of vision, this has repercussions for the distinction between real and seeming cases of seeing. Indeed it seems the distinction between real and seeming seeings will collapse. After all, we only ever seem to see an object when it appears to us in the way of an object of vision. So if x is seen whenever it appears as an object of vision, we will see an object whenever we seem to. At this point, we begin see more clearly why Auriol believes intuitive cognitions of nonexistents are a natural possibility. An object x is seen when it visually appears. Hence when in dreams and fevers we get the visual appearance of nonexistent objects, we do not merely seem to see them, but are seeing them indeed. And since all acts of vision are intuitive cognitions, this means that some intuitive cognitions come without actually existent objects as a matter of natural fact: 1. All acts of vision are cases of intuitive cognition. 2. When x appears as an object of vision, x is seen.

35

36 Scriptum, prol. 2.2 (Buytaert, 200). Scriptum, prol. 2 (Buytaert, 201). Scriptum 3.14 (Buytaert, 713). When Auriol says that vision is ‘nothing but’ an appearance, his point is that objects are seen whenever they appear as objects of vision. He does not mean to deny that vision also involves an act of the sensory power to produce an object in apparent being. It is important to distinguish between the act and the appearance it produces. For although in normal cases of vision the latter is an object of cognition, the former is not. 37

162

Han Thomas Adriaenssen

3. It is naturally possible for x not to exist and to visually appear to me. 4. Therefore, it is naturally possible to see nonexistent things. 5. Hence, it is naturally possible to intuit nonexistents.38 And what this reconstruction tells us, is that the apparent being is neither necessary nor sufficient for the conclusion in Auriol that there can be natural intuitions of nonexistents. To see that the ontology of the apparent being is not necessary to arrive at the natural intuition of nonexistents, consider a philosopher who accepts (3), with the caveat that the natural cases where a nonexistent object appears to us can with equal right be described as cases where an existent object appears to be otherwise than it is. More precisely, they can be described either as cases where an existent external object appears to be otherwise than it is, or as cases where an existent inner object seems to be something that it is not. The illusion of the broken stick provides an example of the first kind. Under one description, this is an appearance of a broken stick that is not really there. But under an alternative description, it is a real stick appearing to be shaped otherwise than it is. Likewise the haunting shadow that appears to me in a dark night under one description is an unreal spirit, but under another is just a pale mist or vapor looking like a ghost. Dreams are typical examples where it is an inner object that appears to be otherwise than it is. Thus my dream of a chimera under one description is the appearance of a nonexistent animal. But under another, it is just a case of an inner image looking like an original. Without going into details, it seems plausible that all natural appearances of unreal objects can be redescribed in this way as cases where an actual inner or outer object appears to be otherwise than it is.39 Indeed, it seems that the only appearances of nonexistents that cannot also be described as cases where a real object appears to be otherwise than it is, are cases where God intervenes by bringing about a cognition independently of any existent object internal or external to its subject. Yet even if God can bring about such events, those would not be the natural cases that the above argument is about. But if we can think of all natural appearances of nonexistent objects as cases where an existent object seems to be otherwise than it is, it becomes plausible to think that we could account for (3) without invoking the apparent being. The appearance of a bent stick that is not there just is a case of a real stick looking bent, and the real stick looks bent because of the

38 The same argument could be made for other modes of sensory cognition. But vision being the paradigm case of intuition, the focus will here be on visual cognition. 39 Chatton and Wodeham take this line, and we will look at some of the details of their accounts in what follows.

Peter Auriol on the Intuitive Cognition of Nonexistents

163

refraction of light that results in the retinal image of a bent stick. The appearance of a chimera is but a case of an image that looks like an original. And it does so because the brain processes the image the way it also does visual input. Again details apart, it seems likely that we can give analyses of these cases in a way that makes no special demands on our ontology. And if we can accept the natural appearance of nonexistent objects without invoking the apparent being, it becomes easier to see how one could arrive at the natural intuition of unreal objects independently of the apparent being. It suffices to accept that all visions are intuitions and that an object is seen whenever it visually appears to us. This will get us to the conclusion that we can naturally see and thus intuit the nonexistent, without in any way relying on the kind of ontology of apparent being that we get in Auriol. If the apparent being is not necessary to arrive at the conclusion that there can be natural intuitions of nonexistents, it is not sufficient either. To see this, suppose that you accept that the objects of cognition are apparent beings, and that you also accept that things can appear to us as objects of vision even if they do not exist. Now if you follow Scotus in saying that vision always comes with a real object, you will deny that (2) all cases where an object visually appears are cases where the object is seen.40 But if you do deny that, the fact that it is possible as a matter of natural fact for nonexistent things to appear as objects of vision will not entail that it is possible as a matter of natural fact for nonexistent things to be seen, and hence to be intuited. Accepting Auriol’s ontology of cognition is thus not sufficient to also accept naturally caused intuitions of nonexistents. What about the further claim in Chatton and Wodeham, that Auriol’s allowance of natural intuitions of nonexistents puts in jeopardy certain knowledge about the world? As we have seen in this section, it seems plausible that the examples of such cognitions that we get in Auriol can be redescribed as ordinary cases where an existent object appears to be otherwise than it is. But surely such cases need to be acknowledged by everyone. Hence to the extent that they limit the certainty intuition can give us, this is a point that all of us will need to grant, the critics of Auriol included. The following sections will further develop this. Sections 4 and 5 explore what kind of certainty intuition affords according to Chatton and Wodeham themselves, given that natural conditions can cause real objects to appear 40

In the early fourtheenth century, John Rodington would make the point that the presence of a real object is implied in what it means to see an object in the first place: “The word ‘vision’ necessarily connotes an existent object that is seen. And when it is asked how the object relates to vision, the answer would be that the object is intrinsically included in all that is meant by the word ‘vision’.” I Sententiarum, prol. 2 (Tweedale, 329).

164

Han Thomas Adriaenssen

to be otherwise than they are. Section 6 then argues that there is no good reason to think that Auriol’s account of intuition puts in jeopardy this kind of certainty.

4 . C H A T T O N O N IN T U I T I O N A N D CE R T A I N T Y According to Chatton, intuitive cognition comprises the acts of the five external senses. Generally speaking, indeed, intuitive cognition is nothing but ordinary seeing, hearing, smelling, tasting, or touching. Part of what this means is that every act of intuition makes its object appear as present to us. Accordingly, intuitions always go hand in hand with an affirmative judgment about the existence of their objects.41 But how certain can we be of the judgments that intuitive cognitions cause us to form? A first obstacle to absolute certainty comes from God’s absolute power. For even though in the natural course of events intuitions are always caused by real objects, Chatton admits God “could bring about an act of seeing without the presence of the thing” and make me err as a result.42 For example, God could make me intuit a nonexistent sunflower, and make me judge that it exists. And this possibility raises the question of whether Chatton offers us any means to recognize these divinely caused intuitions of nonexistents. Although Chatton has been criticized for failing to offer such a means,43 Dominik Perler has recently argued that, in Chatton’s philosophy we do in fact get a mechanism allowing us to detect divinely caused intuitions of nonexistents. For after citing the possibility of divine intervention, Chatton goes on to say that “we understand a thing not to be present via argumentation.”44 And according to Perler the point here is as follows. Imagine that you are walking in a field and intuit a blooming sunflower. Because intuitions bring about affirmative judgments about the existence of their objects, you will judge that the sunflower exists. But at this point, you need to ask yourself how that judgment fits with the facts that it is winter, that the field is covered with snow, and that sunflowers do not bloom under these conditions. This

41 See Reportatio et Lectura, prol. 2.3 (Wey, 102). On the affirmative judgments that accompany acts of intuition, see Rega Wood, “Adam Wodeham on Sensory Illusions with an Edition of Lectura Secunda, Prologus, Quaestio 3,” Traditio 38 (1982), 213–52, 220 and Dominik Perler, “Skepticism,” in R. Pasnau (ed.), The Cambridge History of Medieval Philosophy, 2nd edition (Cambridge: Cambridge University Press, 2014), 384–96, 389. 42 Reportatio et lectura, prol. 2.2 (Wey, 92). 43 See Luciano Cova, “Francesco di Meyronnes e Walter Catton nella controversia scolastica sulla ‘notitia intuitive de re non existente’,” Medioevo 2 (1976), 227–52. 44 Reportatio et lectura, prol. 2.3 (Wey, 103).

Peter Auriol on the Intuitive Cognition of Nonexistents

165

will reveal that there could be no sunflower where you are looking. Your intuition of it must hence be caused by God.45 But this line of argument overlooks at least one option, namely that rather than intervening in nature by making you intuit a sunflower that does not exist, God could just as easily have intervened in nature by making a real sunflower for you to intuit in spite of the cold season and the snow. Once we are allowing divine intervention, this option is on the table no less than God causing a intuition with no actual object. And it is hard to see how information about sunflowers or climate conditions could help you rule this option out. A further problem for this reading is that, right after his claim that “we understand a thing not to be present via argumentation,” Chatton proceeds as follows: The absence of a thing the presence of which is cognized intuitively when it is present is cognized via argumentation when it is absent. For someone can perceive that he has no intuitive cognition of such a thing, and argue from this that this thing is not present, because since there is no obstacle that would prevent it from being seen in case it were present, and since the thing is not seen, therefore it is not present. Consequently, he cognizes via argumentation that the thing is not present.46

In this passage, Chatton introduces the principle that whenever I do not experience x but there are no obstacles that would prevent me from experiencing x if x were present, I can conclude that x is not present. Applying this principle is what he refers to as ‘cognizing that x is absent via argumentation.’ But if this principle tells us what to conclude when we do not experience x, it tells us nothing about what to do when we do experience x. That is, it does not offer us a means to decide whether a given cognition of x is a divinely caused intuitive cognition of an absent object, or rather a naturally caused perception of an object that is actually present. Despite appearances, then, Chatton does not give us an argumentative means of classifying determinate acts of cognition as divinely caused, and so, the judgments that we come to form on the basis of our intuitions all come with the proviso that we might have been supernaturally led astray. Absolute certainty on the basis of intuition is impossible, therefore, and in fact this is something Chatton is happy to admit: I say that we can have no certainty through external sensations such that God, an angel, and perhaps some inferior causes too could not deceive us.47 45

Perler, Zweifel und Gewissheit, 279. Reportatio et lectura, prol. 2.3 (Wey, 103). Reportatio 2.11 (Wey and Etzkorn, 280). This claim is repeated in Reportatio 3.2.2 and 4.4 (Wey and Etzkorn, 47 and 284). 46 47

Han Thomas Adriaenssen

166

In this passage, Chatton does not specify which ‘inferior causes’ he has in mind, but from his discussion of intuitive cognition, it becomes clear that the certainty intuitive cognition affords is further limited by a number of rather ordinary factors. Indeed, even without divine intervention, it is not entirely impossible, Chatton believes, for intuitive cognitions to occur without existent objects. For even though they cannot naturally be caused without the existence and presence of their objects, they can nevertheless for a short while outlast their objects. For example, when we intensely look at a flame or at a strong lamp, our perception of light may last for some moments even after the candle has been blown out, or after the lamp has been switched off. To be sure, because intuitive cognitions can ‘linger on’ in this way for a very short time only, we can be certain that things we perceive for a longer period of time do indeed exist for most of that time. Nevertheless, in the short timespan in which we might be dealing with a perception that outlasts its object, we cannot be certain of the actual existence of what we intuit. As Chatton himself puts it: “I concede that during that time in which a vision can remain after the visible object has disappeared, no certainty is had.”48 A final factor qualifying the certainty that intuitive cognition can give us, is the fact that, according to Chatton, intuitive cognitions of real objects can under special circumstances elicit false judgments. When looking at a trompe l’oeil, for example, the intuition of the picture of a rose may lead me to judge that I am having an actual flower in front of my eyes.49 And in a poorly lit room, say in a Madame Tussauds museum, the intuition of a wax sculpture may make me judge that I am in the presence of another human being. Now, Chatton is confident that such natural illusions occur but infrequently, and that, moreover, they never “subject us to invincible error.”50 Chatton appears to believe that, at least on reflection, we are able to discern good and bad conditions for intuitive cognition. So even if we may not be able to avoid all error, we will often be able to correct mistakes, as when in retrospect we come to the conclusion that some given act of intuition took place under poor conditions, and that we need to revise the beliefs we came to hold as a result of it. Even so, the limits on the certainty that intuition affords are clear. After all, God could deceive us no matter what the natural conditions, so reflecting on the conditions under which we saw an object will not help us to detect these cases. And some natural limits remain too. For not only may we err before we have had the time to assess natural conditions, but even given 48 49 50

Reportatio et lectura, prol. 2.2 (Wey, 92). See Reportatio et lectura, prol. 2.2 (Wey, 97). Reportatio et lectura, prol. 2.2 (Wey, 92).

Peter Auriol on the Intuitive Cognition of Nonexistents

167

time, the fact that we have the means to correct natural error does not mean that we always succeed in using these means. The fact that no natural error is invincible in principle does not mean that none will ever slip through as a matter of infelicitous fact. As a result, intuition will give us but a kind of conditional certainty.51 It gives us certainty about its objects as long as God does not deceive us, it does not outlast its object, and conditions are not such as to make it come with an error of judgment. And this notion that perception can give us conditional rather than absolute certainty was more explicitly spelled out by Wodeham. In the following section, we will look at Wodeham’s concepts of certainty and knowledge. This will eventually help us to see how we should appreciate the claims in Chatton and Wodeham to the effect that all certainty and knowledge about external reality perish if Auriol is right that intuitive cognitions can naturally occur without existent objects.

5 . W O D E H A M O N IN T U I T I O N A N D C E R T A I N K N O W L E D GE For Wodeham, intuitive cognitions are those simple apprehensions that can in principle give us knowledge about the contingent present.52 Hence all acts of seeing, hearing, smelling, touching, and tasting count as intuitions. This is not because they never make us err: they do. But an act of vision that makes us err still is the kind of act that in principle could have given us knowledge about the contingent present, if only conditions had been different. In this it differs from an act of fantasizing, which no matter what the conditions, by nature just is not the kind of act that is conducive to knowledge about matters of fact that hold here and now. Like Auriol and Chatton, Wodeham allows that God can supernaturally bring it about that we have intuitive cognitions of nonexistents. Yet unlike both, he is adamant that under no circumstance do we naturally have an intuition of something nonexistent. Even in the case of afterimages, Wodeham argues, what we see is something that actually exists. For when we look at a strong light, a species will get impressed upon the eye, which for a while remains present there even after the light has died out. And afterimages See also Dominik Perler, “Can We Trust Our Senses? Fourteenth-Century Debates on Sensory Illusions,” in D. G. Denery, K. Ghosh, and N. Zeeman (eds.), Uncertain Knowledge: Scepticism, Relativism, and Doubt in the Middle Ages (Turnhout: Brepols, 2014), 63–90, especially page 70. 52 Lectura secunda, prol. 2 (Wood, 35 and 37). 51

168

Han Thomas Adriaenssen

result when these species then become objects of vision in themselves, while the cognitive power judges that what we see is still the outer light.53 All intuitive cognitions are firmly tied to some actually existent object then. But this does not make intuition an infallible source of information about the world. For what the case of afterimages illustrates, is that the intuition of a real object may under special conditions lead to an error of judgment. This typically happens when our intuition is impeded by poor external conditions, or by defects in the organs of perception. But even though intuition is not infallible, Wodeham remains optimistic that it can yield certainty, as well as knowledge about the external world. To see how Wodeham envisions this, and in what way intuitive cognition is conducive to certainty and knowledge, it will be useful to briefly look at his concepts of certainty and knowledge one by one.

a. Certainty For an intuition to give us certainty about the contingent present, is for that intuition to yield a judgment about the contingent present that is certain. And for a judgment to be certain, according to Wodeham, first of all is for that judgment to be true: the concept of a certain judgment that is false “includes a repugnancy.”54 But whereas truth is necessary for certainty, it is not sufficient. For if I correctly judge that John exists without seeing him, Wodeham would say that my judgment, though true, fails to be certain. As he also puts it, my judgment that John exists in this case will amount to ‘estimation’ rather than certainty. But just what does Wodeham think is needed for certainty in addition to truth? This is a question that he himself does not explicitly address in any detail. What he does tell us, however, is that his distinction between certainty and estimation has its roots in “the Commentator and the Philosopher.”55 More precisely, Wodeham’s distinction between estimation and certainty traces back to Averroes’s commentary on book VII of Aristotle’s Metaphysics. According to Aristotle, opinion deals with “what is capable of being otherwise,” but knowledge is about what cannot be otherwise.56 Knowledge, in other words, is about what is necessary, and what falls short of necessity is in the domain of opinion. Or, as Averroes put it: “it is

53

54 Lectura secunda, prol. 3 (Wood, 77). Lectura secunda, prol. 2 (Wood, 37). Lectura secunda, prol. 1 (Wood, 14). 56 Metaphysics 1040a1. For David Bostock’s helpful exposition of Aristotle’s position, see Aristotle, Metaphysics Book Z and H, ed. D. Bostock (Oxford: Clarendon Press, 1994), 217. 55

Peter Auriol on the Intuitive Cognition of Nonexistents

169

impossible for knowing and not knowing to be of that which is not necessary. Of such things, there is only estimation.”57 When Wodeham traces back his distinction between opinion and certainty to Aristotle and his Commentator, then, the suggestion is that certain judgments, for him, are judgments that necessarily hold true. At first this makes it hard to see how intuition could possibly yield certainty. The very point of intuition after all is that it gives us access to contingent matters of fact about the here and now, so if certainty is defined in terms of necessity, there seems to be no way in which intuition could ever yield more than estimation. But at the same time, Wodeham makes it very clear that intuition yields at least some kind of certainty, writing that intuition is that mode of cognition in virtue of which the soul with certainty naturally assents that such a thing exists unless a miracle or some other impediment should stand in the way.58

Intuitive cognition makes it certain that a thing exists, provided that we are not subject to divine deception and that conditions are good—your senses function well, you are not ill, and the light is clear. I propose we read Wodeham as follows here. Suppose that you have an intuitive cognition of John, and judge that John exists. God has not interfered with your intuition, and natural circumstances are good. Hence, your judgment that John exists is certain. But by this, Wodeham does not mean that it deals with ‘what cannot be otherwise.’ What he does mean, however, is that, as long as these conditions obtain, your judgment is infallible. Given these conditions, it could not be false.

b. Knowledge Knowledge, according to Wodeham, is evident judgment. All evident judgments are certain.59 But apart from that, they are also beyond doubt.60 To see what that means for Wodeham, consider a case where as a matter of fact, i. ii. iii.

you intuit John and judge that he exists, God does not deceive you, and natural conditions are good.

57 58 59 60

In Metaphysicam, VII.17 (Venice, 95). Lectura secunda, prol. 3 (Wood, 65). Lectura secunda, prol. 6 (Wood, 173). Lectura secunda, prol. 6 (Wood, 163–4).

Han Thomas Adriaenssen

170

This is a case where, as we have seen above, the judgment that John exists is certain. Now suppose that we add to the above that iv.

you feel fully confident that God does not deceive you and that natural conditions are good.

In this case where you feel no cause for disbelief, we could say that it is subjectively past doubt that John exists. But now contrast this with a situation where v.

you have a way to rule out that God deceives you and that natural conditions are poor.

In this case, your judgment would be past doubt in what we could call an objective sense, which is stronger than the subjective sense above. This objective sense is stronger, in that your judgment that John exists could not be objectively past doubt without also being certain. After all, once it is ruled out that God or natural conditions deceive you, on the account that we have seen Wodeham develop, it could not be false that John exists. But even if it were uncertain that John exists, it could still be subjectively past doubt to you that he does. Thus if light conditions were in fact poor in a way that you had failed to notice, it could well turn out false that John exists. But because you had not evaluated the relevant conditions correctly, you could still be as confident about them as ever. A factual lack of certainty can go hand in hand with a subjective absence of doubt. Now what Wodeham demands of an evident judgment is at least that it be beyond doubt in the first, subjective sense. He writes that an evident judgment will “fully assure the mind,” and I take this to mean that it will only be evident to you that John exists if you feel no cause for disbelief.61 But that John exists also needs to be beyond doubt in the second, objective sense. For no matter how confident you feel, if you cannot rule out that God or natural conditions deceive you, it would not be fully evident. This at any rate seems to be Wodeham’s point when he writes that I concede what is inferred about a judgment that regards a contingent truth about an external thing. For no such judgment is simply evident with an evidentness that excludes all possible doubt. Because if all possible cognitions and judgments would be caused by either God or nature, it would be possible that in virtue of God’s absolute power, things are not such as they would be signified to be by such an apprehended cognition.62

61 62

Lectura secunda, prol. 6 (Wood, 164). Lectura secunda, prol. 6 (Wood, 169).

Peter Auriol on the Intuitive Cognition of Nonexistents

171

In light of this, Wodeham concludes that it is never absolutely evident to you that John exists, or that he has certain properties. The kind of evident judgment that intuition affords is that John exists, unless God miraculously intervenes here, or there is an impediment due to the imperfection of that cognition or due to some indisposition on behalf of the object, the medium, the power, or the organ.63

Now in some cases, Wodeham leaves out the natural impediments and simply says that intuition affords evident knowledge that John exists unless God intervenes: Although a categorical evident judgment is not had . . . yet a conditional evident judgment is had, namely that it is so unless God deceives.64

Perhaps the reason the natural impediments are left out here is that Wodeham thinks that, at least in some cases, we can rule out that poor conditions blur our view. But even in those cases where we can rule out natural grounds for doubt, the divine cause for doubt will always remain on the table: There is no intuition that affords an evident judgment that a thing exists and that God could not uphold the intuition without its object.65

And with this claim, Wodeham has put a limit on what intuition can afford that seemed far away in Scotus. In an early question of his Ordinatio, Scotus aims to show against Henry of Ghent that knowledge can be had solely by natural means. And part of this task here is to show how we can know the objects of vision. In his discussion, Scotus takes for granted that we know our own inner acts. Thus when we are awake we know that we are awake and when we see we know that we see.66 But when we see, there must also be a real object for our act. For as we have seen him claim above, an act of vision ‘attains an object in its proper actual existence.’ So far there is no ground for doubt then. But what remains to be ruled out is that the object is not in fact the way we judge it to be. Scotus in his discussion with Henry takes it for granted that God will not tamper with the way we judge, and the only serious ground for doubt he sees is that natural conditions could make us judge an object of vision to be otherwise than it is. And of course he knows that this happens often enough, as for example when we forget to consult the sense of touch when a stick looks bent to the eye. But at least in some cases Scotus thinks this kind of error can be ruled out. For if vision is not contradicted by any of the other senses, then, 63 65 66

Lectura secunda, prol. 2 (Wood, 35). Lectura secunda, prol. 2 (Wood, 41). See Ordinatio 1.3.1.4 (Vaticana III 145).

64

Lectura secunda 1.2 (Wood, 222).

172

Han Thomas Adriaenssen

Scotus claims, “there is certainty about the truth of what is so cognized by the senses.”67 Here we seem to come close to full evident sensory cognition. To be sure, we may have questions about the way Scotus makes his case. (Do we really know our own acts of cognition in the way Scotus claims we do?) But what matters for now is that it seems Wodeham could agree with Scotus that, at least in some cases, we can rule out that natural conditions make us err. Yet where he and Scotus part ways is that for Wodeham, even if we know for sure that natural conditions are as good as they could ever be, there always remains at least one ground for doubt, namely that our vision could lack a real object and was made by God alone. With this, he builds in a barrier to full evident cognition that, for Scotus, just does not exist. At this point it may seem as if Wodeham is painting a rather bleak picture of our knowledge of the world. But even if his appeal to divine deception may be a traditional skeptical argument,68 Wodeham did not use this argument to get at a traditional skeptical conclusion. For even without absolutely evident knowledge, Wodeham believed, science remained possible. Thus even if the reality of a solar eclipse cannot be made absolutely evident to us, the ‘hypothetical’ knowledge that it is real unless God deceives us, Wodeham at one point claimed, was enough to support the scientific conclusion that the sun is a celestial body that can be eclipsed.69 Less than absolute evidentia, for Wodeham, may thus be evidentia enough.

6. REVISITING THE CHARGE OF SKEPTICISM As we have seen above, Chatton and Wodeham feared that, with Auriol’s admission of naturally caused intuitions of nonexistents, all certainty and knowledge about the external world would perish. In this section, however, I will argue that in fact there is no good reason to believe that the kind of certainty and knowledge they think intuition affords is in jeopardy on Auriol’s theory of intuition. As the previous sections have made clear, for both Chatton and Wodeham intuition yields a conditional rather than absolute kind of certainty. Specifically, as we have seen, Chatton thinks that the certainty an intuition affords is conditional upon the possibility that 67

Ordinatio 1.3.1.4 (Vaticana III 146). For this qualification, see Elizabeth Karger, “Ockham and Wodeham on Divine Deception as a Skeptical Hypothesis,” Vivarium 42 (2004), 225–36, 229. 69 See Lectura secunda 1.2 (Wood, 220). 68

Peter Auriol on the Intuitive Cognition of Nonexistents A1 A2 A3

173

it is a divinely caused intuition of a nonexistent, it is a naturally caused intuition outlasting its object, or it is an intuition of a real object that is judged to be otherwise than it is.

And similarly, we have seen that for Wodeham, the certain and evident cognition that an intuition yields is conditional upon the possibility that B1 B2

it is a divinely caused intuition of a nonexistent, or it is an intuition of a real object that is judged to be otherwise than it is,

where the last class encompasses the case of afterimages that Chatton took to be natural intuitive cognitions of nonexistent light. What about Auriol? Although he does not offer us the kind of discussion of conditional certainty and evidence that we get in Wodeham, it seems clear that for him, the certainty of a given intuition will be conditional upon the possibility that C1 C2

it is a divinely caused intuition of a nonexistent, or it is a naturally caused intuition of a nonexistent.

The first thing to note here is that all three Franciscans allow that God could bring about the intuitive cognition of a nonexistent object. So here Chatton and Wodeham side with Auriol against Scotus, for whom this particular barrier to certain knowledge does not exist. But on the point of natural errors, Chatton and Wodeham see an important difference between them and Auriol. For according to Chatton, intuitions do but rarely outlast their objects, and it is only under special circumstances that the intuition of a real object comes with an error of judgment. Likewise, Wodeham thinks that the cases where intuition leads to error present exceptions to the rule, and occur only under special conditions. But at the same time, both Chatton and Wodeham suggest that, in Auriol, the category of naturally caused intuitions of nonexistents covers a rather more substantial number of cases. Thus Chatton claims that in Auriol, intuition will become pretty much like abstraction, in that like abstraction it is “by nature equally disposed to being, whether its object exist or not.”70 And according to Wodeham, in Auriol there is no greater reason why a given intuitive perception be caused by the object that is perceived through such an act rather than by other causes.71

The suggestion in both authors thus is that, whereas their own cases of natural errors are exceptions to the rule, Auriol’s natural intuitions of 70 71

Reportatio et lectura, prol. 2.2 (Wey, 90). Lectura secunda, prol. 3 (Wood, 66).

174

Han Thomas Adriaenssen

nonexistents are not, but occur with frequency. But if that is correct, it seems that the certainty intuition affords will be more heavily qualified on Auriol’s account than it is on their own. However, to suggest that natural intuitions of nonexistents are more than exceptions to the rule misconstrues what Auriol had in fact said. For in his Scriptum, he had explicitly said that, under normal conditions, our intuitive cognitions will trace existent objects, providing us with reliable information about our environments: In the natural order of things, an intuitive cognition will be impressed on the intellect from the object, and be conserved by it as light is by the sun. And so it does not make the intellect err because when the object is absent, the intuition immediately stops to be.72

For Auriol, then, intuitive cognitions of nonexistents present exceptions to the rule. And that means that, on his account, intuition yields a certainty conditional on divine deception plus a small number of special cases. At this point, it begins to look as if on Auriol’s account, we are in pretty much the same epistemic condition as we are with Chatton and Wodeham. And the way in which Auriol marshals support for the natural intuition of nonexistents confirms this impression. As we have seen, according to Auriol, the following five experiences illustrate the natural possibility of intuitions without objects: Afterimages Dreams Apparent horrors Illusions Distorted color vision But all of these experiences are covered by the last two conditions on certainty that we get in Chatton. As we have already seen, the phenomenon of afterimages is covered by his category A2 of intuitions outlasting their objects. And although Chatton does not discuss the case of distorted color vision, the remaining three cases he thinks are best understood as intuitive cognitions of real objects that in an error of judgment get mistaken for others. As for dreams, Chatton makes it clear that they must not be understood as visions of nonexistent objects. Rather, what happens in dreams appears to be that we see inner species or representations, but judge that what we see are outer objects. And even if Averroes is right, and if dreams are indeed the 72 Scriptum, prol. 2 (Buytaert, 209). Auriol’s discussion here concentrates on intellectual intuitive cognition, but the point applies to standard perceptual intuition as well.

Peter Auriol on the Intuitive Cognition of Nonexistents

175

result of visual activity in the eye, still this vision not pertain to anything unreal. More precisely, it will be the vision of a state of our visual organ, which we mistake for something else: If it is the case that some vision is caused in the sleeper’s sight, this is not the vision of anything nonexistent, but it is the vision of some quality that has just been effected in that organ. Yet it is such an act of vision that a superior power does not judge that it differs from the vision of something white. And thus it mistakenly judges that it sees whiteness.73

The case of dreams thus goes into category A3 of intuitive cognitions of real objects going hand in hand with an erroneous judgment. The same holds for apparent horrors. Those who are in the grips of intense fear do not hear or see anything that is not really there, but rather mistake the real objects that they do perceive for others. Aroused by fear, for example, someone may hear his own heartbeats, but mistake them for the approaching steps of a murderer: With regard to the third case of the frightened person, I say that someone who is deeply afraid judges that he hears things that he does not really hear.74

Likewise, subjects who are tricked by illusions do not actually see unreal objects. Rather, they see real things, but through an act of judgment mistake them for others: I say that someone tricked by an illusion does not see nonbeings, but that he sees present things. A superior power, however, not being able to tell clearly which things it sees, errs in judging that it sees thing that in fact it does not see.75

The illusions that Auriol had cited in support of the apparent being provide clear examples of this. Thus, when the trees on the shore appear to move, what we see are immobile trees. But from our position on the moving boat we are led to judge that what we see are trees in motion. All of the cases Chatton discusses are covered by his own categories of intuitions outlasting their objects and errors of judgment that follow upon the intuition of real objects. And the case of Wodeham is much the same in this respect. According to Wodeham, in afterimages and dreams, what we do in fact see, are species in the eye or in the imagination. What we are led to judge that we see, however, is an external light or some other object. The frightened see real things, but an error of judgment makes them believe that what they see are things that in fact are not there, and when the world seems 73 74 75

Reportatio et lectura, prol. 2.2 (Wey, 96–7). Reportatio et lectura, prol. 2.2 (Wey, 97). Reportatio et lectura, prol. 2.2 (Wey, 97).

Han Thomas Adriaenssen

176

covered by a shade of red, what the perceiver in fact sees is just a state of his own eyes, like the blood of a broken retinal vein, although through an error of the power of judgment he believes that what he sees is external redness.76

Finally, in illusions, we see real and present things, but the power of judgment mistakes them for fictions or absent objects. This, indeed, is how Wodeham analyses all of the illusions Auriol had cited in support of the apparent being. When the trees on the shore appear to move, for example, we do not see real motion in the trees. What we do see, is just the trees on the shore. But because our visual input is distorted by the continuous change in our and the trees’ relative position, we judge that the trees are in motion: But this appearance is not a vision, but it is a wrong judgment that is caused by vision.77

All of Auriol’s purported examples of natural intuitions of nonexistents, then, are covered by Wodeham’s category B2. This leaves us with the following situation. According to Chatton and Wodeham, the certainty that intuition can give us is conditional upon divine deception, plus a small set of natural yet special cases of perception. In Auriol, it is conditional upon divine deception, plus the set of natural intuitions of nonexistents. But contrary to what Chatton and Wodeham suggest, that set is small, and consists of exceptions to the rule. More to the point, the cases it includes appear to be precisely the kind of natural yet special cases of perception upon which the certainty of intuition is conditional according to Chatton and Wodeham themselves. And this casts new light on their worry that with the natural intuition of nonexistents Auriol had made trouble for our main source of certainty about the world. For as we now see, the class of natural intuitions without real objects covers just a set of ordinary errors that Auriol and his critics alike need to take into account. To the extent that these errors create trouble for our main source of certainty then, it is a kind of trouble that both Auriol and his critics need to deal with. In fact if we put to one side the divine case for the moment, it becomes hard to see how intuition would be any more fallible on Auriol’s account than it was on Scotus’s. After all, Scotus knew as well as anyone that under poor conditions, intuition will come with error. But the natural intuition of nonexistents in Auriol just is a way to account for these errors—not a source of new ones. 76 77

Lectura secunda, prol. 3 (Wood, 82). Lectura secunda, prol. 4 (Wood, 97).

Peter Auriol on the Intuitive Cognition of Nonexistents

177

To be sure, Auriol does not give us any way to pick out these natural errors. And indeed it may seem difficult for him to do so. For on a basic level, true vision and illusion are similar in that both are pairs of an act of vision and an apparent being. The only difference is that in one case the apparent object does, and in the other does not, match a real object. But hard as it may be to pick out errors, Auriol does not seem to be that much worse off than his critics. For on their account too, the illusion of a bent stick is similar in a basic way to the true vision of a stick that is bent indeed. Both are visions of a real stick that come with a judgment about its shape. The only difference is that the judgment is true in one case but false in the other. So far then, picking out errors seems just as hard for Auriol as it is for his critics. Or in a more positive key, to the extent that his critics are optimistic that natural errors will get detected most of the time and that none is ‘invincible’ in principle, it is not clear that Auriol could not share that kind of optimism.78 What might appear to complicate matters here, is the fact that for Auriol we only ever see the world via the apparent being. But if that is the case, we can never compare the way an object appears to us with the way it is in itself. And if we can never compare the way an object appears to us with the way it is in itself, it is hard to see how we could tell apart the good and bad cases. It is as if we were given a collection of paintings and had to pick out the ones that are most realistic without ever seeing their originals. This is the kind of ‘veil of perception’ problem that is perhaps best known from the early modern reception of indirect realism. But again I do not see that Auriol is any worse off than his critics. For they too will need to admit that when we see an object, it will always appear to us in one way or another. And just as Auriol cannot ‘look behind’ his apparent being, they too lack the kind of neutral point of view it would take to see things as they are in themselves, independently of the way they appear to us. So even though for them things appear the way they do because of how we judge them to be and not because of the special being they get as objects of vision, the basic task for them is the same as it is for Auriol and the rest of us: to find a way to tell good and bad cases apart despite the fact that we cannot 78 It is true that according to his critics, some illusions happen because we see a species in a way that we do not in veridical vision. But that does not make them easier to detect. Species can be mistaken for outer objects, and that is why we find some dreams to be indistinguishable from true visions. On a more general note, the species theory was the result of philosophical analysis, and to the best of my knowledge, no one ever claimed that the role of species in cognitive processes was evident from introspection. It seems unlikely, then, that we could identify some act of cognition as an illusion just on the ground that it is mediated by a species. At least, the role played by species in these illusions does not seem to give us an advantage in recognizing them for what they are.

178

Han Thomas Adriaenssen

compare the way things appear to us with the neutral way they are in themselves that perhaps only a God’s eye view would reveal.79

7 . C O N C LU S I O N Auriol’s account of appearances provoked strong reactions among contemporaries, who felt that the apparent being would foreclose proper access to the outer world. In particular, his fellow Franciscans Walter Chatton and Adam Wodeham felt that Auriol’s ontology of appearances was what had led him to claim that intuitive cognitions can naturally come without existent objects. And this claim, at its turn, they believed opened the gates to skepticism. But although Auriol’s readers from the fourteenth century to our own days have linked his ontology of apparent being to the intuition of nonexistents, this paper has argued that the former is neither necessary nor sufficient for the latter. Moreover, we have seen that the charge of skepticism in Chatton and Wodeham is problematic. For Chatton and Wodeham, the certainty that intuition affords is conditional rather than absolute. But for Auriol, as we have seen, the certainty of intuition is conditional upon the exact same cases as it is in his critics. So his theory of intuition may have been unusual, and it may have left us without a means to tell truth from error in all cases. But that being said, it did in no way put at risk the kind of certainty that was enough for two of his hardest critics. The charge of skepticism that we get in Chatton and Wodeham, then, lacks a fair ground.80 University of Groningen

BIBLIOGRAPHY Adam Wodeham. Lectura secunda in librum primum Sententiarum. Prologus et distinctio prima, ed. R. Wood (St. Bonaventure: The Franciscan Institute, 1990). Aristotle. Metaphysics Book Z and H, ed. D. Bostock (Oxford: Clarendon Press, 1994). 79

For a detailed argument that indirect realism is not conducive to skepticism in any special way, see John Greco, Putting Skeptics in their Place: The Nature of Skeptical Arguments and their Role in Philosophical Enquiry (Cambridge: Cambridge University Press, 2007), 77–106. 80 I would like to thank the participants of the 2016 Peter Auriol Workshop at the University of Leuven, where an earlier version of this paper was presented. Also, I am grateful to an anonymous referee for a detailed set of helpful comments.

Peter Auriol on the Intuitive Cognition of Nonexistents

179

Averroes. Aristotelis Metaphysicorum libri cum Averrois commentariis (Venice: Apud Iuntas, 1552). Bolyard, Charles. “Medieval Skepticism,” in The Stanford Encyclopedia of Philosophy (Spring 2013 Edition), ed. Edward Zalta, . Cova, Luciano. “Francesco di Meyronnes e Walter Catton nella controversia scolastica sulla ‘notitia intuitiva de re non existente’,” Medioevo 2 (1976), 227–52. Cross, Richard. Duns Scotus’s Theory of Cognition (Oxford: Oxford University Press, 2014). Cross, Richard. The Medieval Christian Philosophers: An Introduction (London: I. B. Tauris, 2014). Dumont, Stephen. “John Duns Scotus,” in J. Gracia and T. Noone (eds.), A Companion to Philosophy in the Middle Ages (Oxford: Blackwell, 2003), 353–69. Duns Scotus. Opera omnia, ed. L. Wadding (Paris: Vivès, 1891–5). Duns Scotus. Opera omnia, ed. C. Balić (Civitas Vaticana: Typis Polyglottis Vaticanis, 1950–). Friedman, Russell. “Peter Auriol,” in The Stanford Encyclopedia of Philosophy (Fall 2015 Edition), ed. Edward Zalta, . Grassi, Onorato. Intenzionalità: La dottrina dell’esse apparens nel secolo XIV (Milan: Marietti, 2005). Greco, John. Putting Skeptics in their Place: The Nature of Skeptical Arguments and Their Role in Philosophical Enquiry (Cambridge: Cambridge University Press, 2007). Grellard, Christophe. “Comment peut-on se fier à l’expérience? Esquisse d’une typologie des réponses médiévales au scepticisme,” Quaestio 4 (2004), 113–35. Karger, Elizabeth. “Ockham and Wodeham on Divine Deception as a Skeptical Hypothesis,” Vivarium 42 (2004), 225–36. Lagerlund, Henrik. “A History of Skepticism in the Middle Ages,” in H. Lagerlund (ed.), Rethinking the History of Skepticism: The Missing Medieval Background (Leiden: Brill, 2010). Perler, Dominik. “What Am I Thinking About? John Duns Scotus and Peter Aureol on Intentional Objects,” Vivarium 33 (1994), 72–89. Perler, Dominik. Zweifel und Gewissheit: Skeptische Debatten im Mittelalter, 2nd edition (Frankfurt: Vittorio Klöstermann, 2012). Perler, Dominik. “Can We Trust Our Senses? Fourteenth-Century Debates on Sensory Illusions,” in D. Denery, K. Ghosh, and N. Zeeman (eds.), Uncertain Knowledge: Scepticism, Relativism, and Doubt in the Middle Ages (Turnhout: Brepols, 2014), 63–90. Perler, Dominik. “Skepticism,” in R. Pasnau (ed.), The Cambridge History of Medieval Philosophy, 2nd edition (Cambridge: Cambridge University Press, 2014), 384–96. Peter Auriol. Scriptum super primum Sententiarum, ed. E. Buytaert (Louvain: Nauwelaerts, 1952–6). Peter Auriol. Scriptum super primum Sententiarum, in R. Friedman, L. Nielsen, and C. Schabel (eds.), The Electronic Scriptum .

180

Han Thomas Adriaenssen

Tachau, Katherine. “The Response to Ockham’s and Aureol’s Epistemology (1320–1340),” in A. Maierù (ed.), English Logic in Italy in the 14th and 15th Centuries (Naples: Bibliopolis, 1982), 185–217. Tachau, Katherine. Vision and Certitude in the Age of Ockham: Optics, Epistemology, and the Foundations of Semantics (Leiden: Brill, 1988). Tweedale, Martin. “John of Rodynton on Knowledge, Science, and Theology” (PhD dissertation, University of California, 1965). Walter Chatton. Reporatatio et lectura super Sententias, ed. J. Wey (Toronto: Pontifical Institute of Mediaeval Studies, 1989). Walter Chatton. Reportatio super Sententias, ed. J. Wey and G. Etzkorn (Toronto: Pontifical Institute of Mediaeval Studies, 2002–5). Wood, Rega. “Adam Wodeham on Sensory Illusions with an Edition of Lectura Secunda, Prologus, Quaestio 3,” Traditio 38 (1982), 213–52. Wood, Rega. “Adam of Wodeham,” in J. Gracia and T. Noone (eds.), A Companion to Philosophy in the Middle Ages (Oxford: Blackwell, 2003), 77–85.

Ockham on the Parts of the Continuum Magali Roques INTRODUCTION In this paper, my aim is to explain how William of Ockham understood what he takes to be the definitional property of the continuum, namely its infinite divisibility. According to Aristotle, a continuous magnitude of finite size can be infinitely divided into parts which can in turn be further divided.1 Moreover, the parts arrived at by this process of division can be added and will constitute another potential infinite.2 With the geometers and against the atomists, Aristotle holds that any given line segment, no matter how small, can be bisected, thus producing a smaller segment, and similarly for surfaces and volumes. There is therefore no smallest line segment, nor is there a smallest surface or volume in a continuous body. However, Aristotle insists that a bisection produces a point in the middle of the segment. That is, according to Aristotle, we cannot say that the point was on the line segment before the bisection. In this sense, the “infinity” of points on a line segment is not actual but only potential. Similarly, the infinity of parts in a continuous three-dimensional body is not actual but only potential, since the parts do not exist before the division of the continuum. Ockham, a well-known representative of fourteenth-century Aristotelian anti-atomism, seems to depart from Aristotle by claiming that the parts of the continuum exist before any division of the continuum. Indeed, he claims that (T1) “whether a part exists in the continuum or is separated from it, it exists properly speaking by an existence and actuality of its own, not of the whole.”3 T1 seems to imply quite straightforwardly that (T2) 1

2 Aristotle, Physics, VI, 1, 231b15–16. Aristotle, Physics, III, 6, 206b3–6. Ockham, QP 69, OPh VI, p. 589, ll. 10–12. The texts of Ockham that I will comment on are taken from the Summa Logicae (abbreviated as SL), the Ordinatio 3

182

Magali Roques

since a part exists in act in the continuum by an actuality of its own, it can be separated from any other part of the continuum, and therefore any part of the continuum can be separated from the continuum, that is, can be divided in act from it. However, according to John Murdoch, “to say, with Ockham, that there is an infinity of actually existing parts is not in any way to say that there is an actual infinity of parts.”4 Ockham uses the term ‘infinite parts’ only in the weaker sense that there can always be more parts, rather than in the stronger sense that there are infinitely many parts taken conjointly. By contrast, Robert Pasnau thinks that Ockham accepts an actual infinity of parts in the stronger sense, on the ground that these parts are actual.5 He calls this position ‘actualism.’6 How then can we decide between these two interpretations of Ockham’s account of the nature of continuity? Ockham also claims that (T3) “for the truth of the proposition ‘A continuum can be divided infinitely,’ the truth of the following proposition is required as its exponent: ‘There is no part of the continuum that cannot be divided’ and, similarly, the truth of the proposition ‘Some part of the continuum is not divided in act.’ ”7 T2 and T3 seem to contradict each other. Since Ockham does not claim T2 explicitly, there must be something, in his view, that blocks the inference from T1 to T2. Murdoch’s interpretation of T1 does block the inference from T1 to T2: the parts of the continuum are infinite in the weaker sense that there are always more parts to be divided in a continuum, as T3 seems to say. Thus, T3 seems to give him good reason to claim that Ockham is not an actualist. However, I will not follow Murdoch’s interpretation, which, I believe, does not take seriously enough Ockham’s rejection of Aristotle’s concept of the infinite in potency. Thus, in this paper I will find another way to deal with the compatibility problem I presented above. I will not try to block the inference from T1 to T2; rather, I will give a different analysis of T3 which makes it compatible with T2. (abbreviated as Ord), the Expositio in librum Physicorum Aristotelis (abbreviated as ExpPhys), the Tractatus de Quantitate (abbreviated as TQ), the Quaestiones in librum Physicorum Aristotelis (abbreviated as QP), the Quaestiones variae (abbreviated as Quaest. Var.), and Quodlibeta (abbreviated as Quodl). I will refer to the critical edition of Ockham’s Opera omnia, edited under the direction of G. Gàl (St. Bonaventure, NY: Franciscan Institute, 1967–88), and abbreviated in the present paper as OPh (Opera Philosophica) and OTh (Opera Theologica). 4 John Murdoch, “Ockham and the Logic of Infinity and Continuity,” in N. Kretzmann (ed.), Infinity and Continuity in Ancient and Medieval Thought (Ithaca, NY: Cornell University Press, 1982), 165–206. 5 Robert Pasnau, Metaphysical Themes 1274–1671 (Oxford: Clarendon Press, 2011), 612. 6 Pasnau, Metaphysical Themes, 613. 7 Ockham, QP 66, OPh VI, 584, ll. 47–51.

Ockham on the Parts of the Continuum

183

I believe that this reconstruction takes into account the crucial place of Ockham’s new concept of the infinite in his analysis of the infinite divisibility of the continuum. Like many of his fellow anti-atomists, Ockham puts great stress on the fact that the concept of a potential infinite seems to contradict Aristotle’s modal logic, in particular the central assumption that there is no potency that will never be realized. In other words, fourteenthcentury anti-atomists believed that they not only had to refute atomism, but also had to propose an analysis of the infinite divisibility of the continuum that is not incompatible with their modal logic. Ockham is a representative of this new trend. On my reading, Ockham defends the view that the parts of the continuum exist in act in the continuum: they are already there in the continuum before any division of the continuum. Moreover, they are infinitely many in the sense that no division of the continuum will exhaust all the existing parts of the continuum taken conjointly. They are not infinitely many in Murdoch’s weaker sense that a part can always be divided into two parts that did not exist before the division. Thus, by focusing on Ockham’s defense of the consistency of Aristotle’s anti-indivisibilism—an aspect of his anti-atomism that has not raised as much scholarly interest as his refutation of atomism—I will be able to give a reading of Ockham’s texts that confirms Pasnau’s interpretation. To defend my reading, I will proceed in three steps. First, I will explain the problem raised by the claim that the continuum is divisible into parts which are always divisible and how Ockham and Adam Wodeham respond to it. This will allow me to present the reason why Ockham rejects Aristotle’s concept of an infinite in potency and how he understands T3. Second, I will present Ockham’s metaphysics of the parts of the continuum. In this section, I will explain what Ockham means by T1 and T2 and I will give my solution to the compatibility problem between T1, T2, and T3. The last section of the paper is dedicated to Ockham’s proof that the possibility of a complete division of the continuum does not imply a kind of infinite indivisibilism such as the one defended by Henry of Harclay. By explaining this proof, we will see how Ockham understands the categorematic infinite to which he appeals in his analysis of the infinite divisibility of the continuum.

1 . A L O G I C A L P R O B LE M

a. Adam Wodeham’s Radical Solution Aristotle thought that place, time, and at least some changes are infinitely divisible into parts, or intervals, which are bounded by indivisibles. A preference for a simpler ontology might lead one to attempt to eliminate either the intervals

184

Magali Roques

or the indivisibles which figure in Aristotle’s account. Both of these routes were explored in the early fourteenth century. Atomists such as Walter Chatton, William Crathorn, and Henry of Harclay proposed that intervals can be constructed out of indivisibles,8 while William Ockham, Adam Wodeham, and John Buridan advocated banishing indivisibles altogether.9 Ockham occupies an important place in the fourteenth-century indivisibilist controversy, offering the most influential account of the ontological status of mathematical indivisibles to emerge from fourteenth-century nominalism.10 As is well known, Ockham does not think that there could be such things as indivisibles at all. He thus accepts what has been labeled ‘non-entitism’: indivisibles do not exist, either as component parts of a magnitude or as the limits of a magnitude.11 It is easy enough to gloss talk of indivisibles in terms of a body’s being ‘no further extended.’12 Mathematicians do speak of points and lines, but these terms are fictions of the mind that are convenient abbreviations to describe nothing other than physical reality.13

8 Medieval finitist atomism (especially as represented by William Crathorn, Walter Chatton, and John Wyclif) has been studied extensively by Aurélien Robert. See especially “Atomisme et géométrie à Oxford au XIVe siècle,” in S. Rommevaux (ed.), Mathématiques et connaissance du monde réel avant Galilée (Paris: Omniscience, 2010), 15–85. For Walter Chatton and Nicholas Bonet, see John Murdoch, “Infinity and Continuity,” in N. Kretzmann, Anthony Kenny, and Jan Pinborg (eds.), The Cambridge History of Later Medieval Philosophy (Cambridge: Cambridge University Press, 1982), 576, n. 36. For these two together with Henry of Harclay and Nicholas of Autrecourt, see Rega Wood’s introduction to her 1988 edition of Adam of Wodeham’s Tractatus de indivisibilibus (Dordrecht: Kluwer, 1988), in which she gives a survey of positions in the indivisibilist controversy. For Gerald Odonis, see S. W. de Boer, “The Importance of Atomism in the Philosophy of Gerard of Odo (O.F.M.),” in C. Grellard and A. Robert (eds.), Atomism in Late Medieval Philosophy and Theology (Leiden: Brill, 2009), 85–106. On the difficulty of explaining “why this current of indivisibilism arose in the first third of the fourteenth century or what function it was held to serve,” see Murdoch, “Infinity and Continuity,” 576–7. 9 For Ockham’s anti-atomism, see especially C. Delisle Burns, “William of Ockham on Continuity,” Mind 25 (1916), 506–12; T. Bruce Birch, “The Theory of Continuity of William Ockham,” Philosophy of Science 3 (1936), 494–505; John Murdoch, “Ockham and the Logic of Infinity and Continuity”; Eleonore Stump, “Theology and Physics in De Sacramento Altaris: Ockham’s Theory of Indivisibles,” in N. Kretzmann (ed.), Infinity and Continuity, 165–206. For John Buridan and Adam Wodeham, see the references given all along this paper. 10 See especially Murdoch, “Infinity and Continuity,” 574–5. 11 On this subject, see Wood, “Introduction,” in Adam Wodeham, Tractatus de indivisibilibus, p. 11. 12 Ockham, TQ, q. 1, OTh X, 22. On this subject, see Marilyn McCord Adams, William Ockham (Notre Dame, IN: University of Notre Dame Press, 1987), vol. 1, 169–214; Stump, “Theology and Physics,” 207–30; Magali Roques, “Introduction,” in Guillaume d’Ockham, Traité sur la quantité et traité sur le corps du Christ, ed. and trans. M. Roques (Paris: Belles Lettres, 2014), LXXVI–LXXXVI. 13 Ockham, ExpPhys VI, 1, OPh V, 461–2, ll. 304–15.

Ockham on the Parts of the Continuum

185

Given this, Ockham infers instead that a magnitude is composed of parts that are themselves magnitudes. Basically, the extended parts that Ockham has in mind are proportional parts, i.e., the sort of parts that can be obtained by dividing the whole and each of its parts into halves, or thirds, and so on. This is why Ockham seems to be content with the common dictum that infinite divisibility is “the nature (ratio) of the continuum.”14 On the other hand, he does not accept without further elaboration the other definition of the continuum that Aristotle gives in his Physics, namely that two things that share an extremity are continuous.15 Indeed, Ockham’s denial that limits (points, lines, surfaces) are things implies that this definition must be paraphrased: two things are said to be continuous when they are contiguous and constitute one thing.16 This aspect of Ockham’s theory is well known, however, and is not what I am concerned with in this paper. Rather, I will focus on Ockham’s understanding of Aristotle’s dictum that infinite divisibility is the “nature” (ratio) of the continuum. Indeed, when Aristotle says that two things whose limits are one are continuous, he defines continuity as a relation between entities rather than as an attribute appertaining to a single entity; in other words, he does not provide an explicit definition of the concept of the continuum. By contrast, the dictum that infinite divisibility is the nature of the continuum might well serve as a definition of the continuum, but it calls for elaboration.17 Since talk about boundaries is not at the center of the analysis of the continuum understood in this sense, we might think that this is a question not of abstract mathematics but of natural philosophy, inasmuch as we are asking about the real composition of continuous material bodies. We might also think that it is a matter of mere conceptual analysis, an interpretation that I tend to prefer. What does the dictum that infinite divisibility is the “nature” of the continuum mean? Aristotle says that any permissible infinite is not that beyond which there is nothing, but rather that beyond which there is always something.18 With this distinction, he thinks he is able to defend the idea that a continuum is never completely divided, that is, that it is not composed of indivisibles, i.e., parts that are not further divisible. However, this claim raises a problem that is spelled out very clearly by Ockham’s former 14 Ockham, ExpPhys VI, 1, §2, OPh V, 453, ll. 22–3: “Unde etiam haec est ratio continui, scilicet quod dividitur in semper divisibilia.” 15 Aristotle, Physics, V, 3, 227a10–12. 16 Ockham, ExpPhys VI, 1, OPh V, 460, ll. 249–55. For an explanation of this definition, see Roques, “Introduction,” LXXVI–LXXVIII. 17 On this subject, see especially John L. Bell, The Continuous and the Infinitesimal in Mathematics and Philosophy (Milan: Polimetrica, 2006), 30–9. 18 Aristotle, Physics, III, 6, 207a1, 7–8.

186

Magali Roques

secretary Adam Wodeham: according to the principles of Aristotle’s modal logic, the inference from the infinite divisibility of the parts of the continuum to their actual complete division is valid. In other words, the principles of Aristotle’s modal logic, Adam Wodeham believes, run counter to the very possibility of an infinite in potency. This is confirmed by the priority of actuality over potentiality in Aristotle’s theory of the modalities. Following a long tradition of reflection upon the definitions of the modalities,19 Wodeham quotes two principles of Aristotle’s modal logic which, when combined, have been identified by Jaako Hintikka as a formulation of the “principle of plenitude.”20 The first is that with the possible posited, nothing impossible follows.21 This principle is at the basis of what Simo Knuuttila has called ‘the model of possibility as contradictoriness.’22 In speaking about the assumed non-contradictory actualization of a possibility, Aristotle thinks that it is realized in our one and only history. Moreover, an argument given in De caelo I.12 excludes those possibilities which remain eternally unrealized from the set of genuine possibilities.23 This is confirmed by the second principle, which states that what is impossible to have been, is impossible to be. This principle underlies what has been called “the statistical or temporal-frequency model” of the modalities, which is an extensional interpretation of modal terms relative to time, that is, it is a tool for classifying types of individual or event from the point of view of their frequency.24 If these two principles are granted, it follows that every potentiality must eventually be realized. Consequently, Wodeham claims, the proposition that “the continuum is divided into everything into which it is divisible” does not entail anything impossible.25 If a continuum is infinitely divisible, there exists an instant at which it will be infinitely divided, which contradicts Aristotle’s idea of an infinite in potency, i.e., something which is never fully actualized. Thus, if one follows Aristotle’s modal principles, nothing prevents the continuum from being completely divided.

19 See, for example, John Duns Scotus, Ordinatio II, d. 2, p. 2, q. 5, ed. K. Balic et al. (Vatican City: Typis Polyglottis Vaticanis, 1973), n. 288, 279–80 and nn. 354–5, 311–12. 20 Jaakko Hintikka, Time and Necessity: Studies in Aristotle’s Theory of Modality (Oxford: Clarendon Press, 1973). The term ‘principle of plenitude’ comes from Arthur Lovejoy, The Great Chain of Being (Cambridge, MA: Harvard University Press, 1936). 21 Aristotle, Prior Analytics, I, 13, 32a19–20. 22 Simo Knuuttila, Modalities in Medieval Philosophy (London: Routledge, 1993), 31–2. 23 24 Aristotle, On the Heavens, I, 12, 283a5–30. Knuuttila, Modalities, 6. 25 Wodeham, Tractatus de indivisibilibus, q. 4, a. 1, 212, ll. 7–15.

Ockham on the Parts of the Continuum

187

Wodeham goes one step further. Aristotle’s modal logic implies that as soon as a division takes place in the continuum, all the other possible divisions occur. Indeed, the principle that every potency must be reduced to act has as a consequence that all the possible divisions of the continuum take place at the same time, since any division implies that a thing is divided from some second thing, the second from a third, and so on.26 Thus, as soon as a continuum is divided into two halves, all the possible divisions occur. But if this is so, the ultimate parts of the continuum are indivisibles, and this contradicts Aristotle’s anti-indivisibilism.27 Aristotle considers this argument as one of the arguments that Democritus used to defend his atomism.28 Thus, the conclusion at which Wodeham arrives is an unfortunate result that calls for a reworking of Aristotle’s anti-indivisibilism. If Wodeham wants to be faithful to Aristotle’s anti-indivisibilism, either Aristotle’s modal logic must be revised or the idea of an infinite in potency must be given up. This is why Wodeham insists that the continuum cannot be infinitely divided: Here I hold that [the infinite divisibility of a continuum] cannot be reduced to actuality, in such a way that it could truly be said to hold that a continuum is actually divided into everything into which it was divisible.29

Since being continuous is essential to being a continuum, and since having been divided is incompatible with being continuous, it is impossible for a continuum to have been divided even once.30 How does Adam Wodeham understand this strong claim? Democritus and other ancient atomists seem to have been concerned mainly with corporeal atoms. But since the argument appears to invoke no physical facts and to apply to lines, squares, and cubes as well as to bodies, Wodeham has to deal with it as an argument for the necessity of indivisible magnitudes generally (atoms of space, time, velocity, heat, or any other magnitude). Adam Wodeham’s answer, however, seems to have a narrower scope. It deals with the real composition of continuous material bodies and is based on mereological principles. Indeed, Wodeham supports his claim that the continuum cannot

26

Wodeham, Tractatus de indivisibilibus, q. 4, a. 1, 228, ll. 27–32. Wodeham, Tractatus de indivisibilibus, q. 4, a. 1, 212, ll. 17–18. For a commentary on Wodeham’s text, see Norman Kretzmann, “Adam Wodeham’s Anti-Aristotelian AntiAtomism,” History of Philosophy Quarterly 1 (1984), 381–98. 28 Kretzmann, “Adam Wodeham,” 2–3. 29 Wodeham, Tractatus de indivisibilibus, q. 4, a. 1, 212, ll. 27–8. The English translations of Wodeham’s texts come from Rega Wood, with slight modifications. 30 Ibid. 27

188

Magali Roques

be infinitely divided by appeal to the mereological principle that a continuum cannot be divided into its parts without ceasing to be: I hold that although the halves of a continuum can be divided from each other and can have been divided, yet literally speaking a continuum can neither naturally be divided nor have been divided into its halves, nor indeed into any other of its parts whatever.31

The metaphysics of continuity supports such a modal analysis. By definition, in order that it be [a continuum, it must be] truly one and a whole arising from all its compoundable parts having been drawn together. But this would not be consistent with its being actually divided.32

To be a continuous whole is to have a certain kind of unity that cannot be broken unless the whole ceases to be. Indeed, “if a part perishes, the whole will not exist.”33 Wodeham’s solution to the problem of the infinitely divisible continuum is thus to deny that any division at all can occur. Norman Kretzmann calls such a position ‘macro-indivisibilism.’34 A few years before Wodeham, Ockham dealt with precisely the same problem. He asks whether the proposition “A continuum can be divided ad infinitum” is true properly speaking.35 He remarks that the negative answer would run counter to Aristotle’s principle that “any natural potency can be reduced to act.”36 Like Wodeham, he gives a negative answer to this question.37 But Ockham’s position on this issue is not quite the same as Wodeham’s. Ockham distinguishes two senses in which a continuum can be divided, namely by a real division or by a mental division. If we speak of a real division, Ockham might be claiming, like Wodeham, that the continuum no longer exists after division because it is no longer one. If we speak only of a mental division or of a division by God’s power, however, Ockham affirms that a continuum can be divided infinitely.38 Ockham’s solution is much less radical than Wodeham’s, since he seems to believe that Aristotle’s idea of an infinite in potency is not intrinsically incoherent. He appeals to a thought experiment showing that the division of a continuum is conceivable.

31

Wodeham, Tractatus de indivisibilibus, q. 4, a. 1, 214, l. 34–16 (l. 4). Wodeham, Tractatus de indivisibilibus, q. 4, a. 1, 212, ll. 29–31. 33 Wodeham, Tractatus de indivisibilibus, q. 4, a. 1, 216, l. 24. 34 Kretzmann, “Adam Wodeham,” 389. 35 Ockham, QP 66, OPh VI, 582, ll. 2–3. 36 Ockham, QP 66, OPh VI, 582, ll. 4–5. 37 Ockham, QP 66, OPh VI, 582, l. 8. 38 Ockham, ExpPhys III, 13, OPh IV, 550, ll. 140–8. See also Ockham, QP 66, OPh VI, 583, ll. 20–1. 32

Ockham on the Parts of the Continuum

189

Ockham’s appeal to conceivability raises more questions than it answers. Indeed, it is based on the claim that a real division could not be completed because of a natural impediment, such as minima naturalia. However, this physical fact does not impinge on the metaphysics of modalities: what is at stake is not the practical possibility of the process but the metaphysical status of a continuum. Our practical inability to perform the operation is no valid objection against its theoretical soundness.

b. Ockham’s Moderate Solution In fact, Ockham’s appeal to mental division remains in draft form. His answer to the problem relies rather on a logical analysis. Indeed, he proceeds to a logical analysis of the proposition ‘A continuum is infinitely divisible’ that aims at showing that the infinite divisibility of the continuum is compatible with the principles of Aristotle’s modal logic. It appears that this analysis does not yield the desired result and must be completed with a conceptual analysis of the very notion of infinity. Let me explain first what this analysis consists in. Ockham begins by uncovering the logical form of the proposition ‘A continuum is infinitely divisible.’ He reminds his readers that some propositions which seem to be non-modal are in fact modal. In particular, the proposition ‘A continuum can be divided infinitely’ could be considered as modal, in which case it would be a proposition de possibili. This means that by the proposition ‘A continuum can be divided infinitely’ it is denoted that a proposition is possible, namely the proposition ‘This—a continuum being designated—is infinitely divided.’39 Indeed, for propositions de possibili, there is a rule stating that “the predicate appellates its form,” which means that, for the truth of a proposition de possibili of the form ‘S can be P’ it is required that a proposition of the form ‘This (some being S being designated) is P’ can be true.40 For instance, for the truth of the proposition ‘Socrates can be white’ it is required that the proposition ‘This (Socrates being designated) is white’ can be true. Ockham then tries to neutralize this interpretation of the proposition ‘A continuum can be divided infinitely,’ since it leads to the unfortunate result that the division of a continuum stops at indivisibles. He explains that the rule of the appellation of form is valid only for a subclass of propositions For the definition of ‘modal proposition,’ see Ockham, SL II, 1, OPh I, 242–3, ll. 44–8. 40 Ockham, QP 66, OPh VI, 582, ll. 8–13. See also Ockham, SL I, 72, OPh I, 216 and, for a short summary of the meaning of the rule, Ockham, SL II, 7, OPh II, 269–70, ll. 9–15. 39

190

Magali Roques

de possibili, namely propositions de possibili in the divided sense, as opposed to propositions de possibili in the compounded sense.41 Roughly speaking, in modal propositions in the compounded sense the scope of the modal operator ranges over the whole corresponding assertoric proposition, while in modal propositions in the divided sense the scope of the modal operator ranges only over the copula.42 For instance, the following proposition de possibili is true in the divided sense: “ ‘Socrates is killed’ and ‘Socrates is not killed’ can both be true.” But the corresponding assertoric proposition is never true, namely “ ‘Socrates is killed’ and ‘Socrates is not killed’ are both true.”43 Thus, for propositions de possibili in the divided sense, we can never infer the corresponding categorical proposition. This is what happens with the proposition ‘A continuum can be divided ad infinitum’: the use of the rule gives a result which is “nonsensical,” namely that the proposition “This proposition is possible: ‘This line is divided ad infinitum’ ” is true.44 The result of the application of the rule is nonsensical only if we presuppose that the division of a continuum can never be wholly completed, which is precisely the point at issue. It is less clear that this is nonsensical when we understand it as referring to mental division. Of course, this depends on what a mental division requires. Since the result of the application of the rule is said to be nonsensical, it is clear that Ockham is speaking of a real division of the continuum. But then Ockham seems to be begging the question. This is confirmed in the second step of his argument. In this step Ockham wants to block the consequence drawn from the modal principle according to which any possible can be reduced to act, a principle that he accepts explicitly.45 In this way, he can justify his claim that the rule of the appellation of form does not hold for some propositions de possibili 41 The practice among medieval logicians of distinguishing the compounded and divided senses of certain propositions was a well-recognized, systematic way of investigating various logically permissible subjects for a single proposition. See Norman Kretzmann, “Sensus Compositus, Sensus Divisus, and Propositional Attitudes,” Medioevo 7 (1981), 195–229. Wodeham gives a similar analysis of the proposition ‘A continuum can be divided ad infinitum’ in his Tractatus de indivisibilibus, q. 4, a. 1, 218, ll. 20–2. 42 Ockham, SL II, 9, OPh I, 273, ll. 12–25. For a much more detailed analysis than the one presented here, see Ernesto Perini Santos, “La structure de la proposition modale ockhamienne,” in A. Maierù and L. Valente (eds.), Medieval Theories on Assertive and Non-Assertive Language (Florence: Olschki, 2004), 355–75. 43 Ockham, QP 66, OPh VI, 584, ll. 53–6. Buridan gives a similar analysis in his Questions on Aristotle’s Physics, III, q. 19, in John Buridan’s Tractatus de Infinito, ed. J. M. M. H. Thijssen (Nijmegen: Ingenium, 1991), 77. For an analysis of Buridan’s text, see John Murdoch and J. M. M. H. Thijssen, “Buridan on Infinity,” in J. M. M. H. Thijssen and J. Zupko (eds.), The Metaphysics and Natural Philosophy of John Buridan (Leiden: Brill, 2001), 137–8. 44 Ockham, QP 66, OPh VI, 582–3, ll. 13–16. 45 Ockham, Ord, d. 38, q. un., OTh IV, 578, ll. 10–12.

Ockham on the Parts of the Continuum

191

taken in the divided sense. In his own terms, the principle implies that it would not be impossible to posit that a continuum is completely divided.46 Ockham answers that the proposition ‘A continuum can be divided ad infinitum’ does not fall within the scope of this principle. For its truth it is required that two propositions be true, namely, ‘There is no part of the continuum that cannot be divided’ and ‘Some part of the continuum is not divided in act.’ The second proposition is an assertoric negative proposition. This blocks the application of the modal principle, which is valid only for affirmative propositions de possibili.47 Thus, Ockham clearly wants to avoid the unfortunate consequence that the possibility of the infinite divisibility of the continuum implies its complete division in act. His position on the question is revealed by the alternative analysis that he gives of the proposition ‘A continuum can be divided ad infinitum.’ He explains that the proposition must be analyzed into two propositions, namely ‘A continuum can be divided’ and ‘After the division of any part another part can be divided.’48 This means first that the division of the continuum is possible, at least in thought, and second that it has no end, in the sense that any part of the continuum can always be further divided. In other words, to say that a continuum can be infinitely divided is nothing more than to say that any part can be divided, the resulting parts further divided, and so on ad infinitum. How is this analysis justified? The burden of proof lies with Ockham: he owes us an explanation of why his analysis of the proposition ‘A continuum can be infinitely divided’ is the right one. If he does not agree with the radical position of Wodeham and thinks that Aristotle’s idea of infinite in potency is not intrinsically contradictory, how does he understand it?

c. A New Concept of Infinity Given that logical analysis is not sufficient to answer the question whether Aristotle’s concept of the infinite in potency is compatible with his modal logic, a conceptual analysis of the very notion of the infinite is required. Ockham’s basic insight consists in substituting a kind of infinite in act for Aristotle’s infinite in potency. There are several ways to understand the idea of an infinite in potency. According to Murdoch, writers of the fourteenth century tended to replace Aristotle’s distinction between the potential infinite and the actual infinite with a more rigorous distinction between the syncategorematically infinite 46 47 48

Ockham, QP 66, OPh VI, 583–4, ll. 42–3. Ockham, QP 66, OPh VI, 584, ll. 46–52. Ockham, QP 66, OPh VI, 583, ll. 16–20.

192

Magali Roques

and the categorematically infinite. The syncategorematically infinite is such that for any finite number there exists something greater than that number; the categorematically infinite is such that there exists something greater than any finite number.49 A natural way of making sense of Ockham’s position would be to say that he argues, like Aristotle, that the parts of the continuum are infinitely divisible, but gives a much more extensive analysis of infinite divisibility. In particular, he says that there are syncategorematically infinitely many parts in a continuum in the sense that for any finite number of them there are always more, but not categorematically infinitely many, in the sense that the process of division can never be completed. So, just as with Aristotle’s potential infinity, the categorematically infinite in Ockham would be conceived in process-like ways, and Ockham would remain faithful to Aristotle. This is the gist of Murdoch’s interpretation of Ockham’s position.50 Is this interpretation justified? Ockham claims that: In response to the second objection, I concede that an infinity of parts in act would have been divided from the continuum, if the hypothesis had been posited, and that the division would stop at an infinity of parts in act, and that these parts would still be infinitely divisible, and that the division would not stop at an infinity of indivisible parts.51

Ockham says that his method of proportional division of the continuum will yield actually infinitely many parts that actually exist before the division. The idea is that the process of division will reach an end only if we reach parts that are indivisible. But since we never reach indivisibles, the process has no end. Is this infinity of parts actually existing in the continuum before the division syncategorematically infinite, as Murdoch has it, or categorematically infinite, as Pasnau thinks? Ockham begins by reminding his reader that to claim that a magnitude has infinitely many parts syncategorematically is to claim that for any finite number n of the parts of the magnitude, the 49 See, for example, Rimini, Sent. II, d. 2, q. 2, art. 1, ed. A. Trapp (Berlin: de Gruyter, 1979), vol. IV, 294, ll. 7–16. On the distinction between categorematic and syncategorematic words, see Norman Kretzmann, “Syncategoremata, Sophismata, Exponibilia,” in Kretzmann, Kenny, and Pinborg(eds.), Cambridge History, 211–45. On categorematic and syncategorematic uses of the term ‘infinite,’ see Anneliese Maier, Die Vorläufer Galileis im 14. Jahrhundert (Rome: Edizioni di storia e letteratura, 1949), 157; Norman Kretzmann, William of Sherwood’s Treatise on Syncategorematic Words (Minneapolis: University of Minnesota Press, 1968), 41–3. 50 Murdoch, “William Ockham,” 189: “Thus, in spite of his conviction that the parts actually exist in continua, there is no question at all but that Ockham firmly believed that the divisibility of a continuum was only potential.” 51 Ockham, Quaest. Var. III, OTh VIII, 78, ll. 329–33.

Ockham on the Parts of the Continuum

193

number of the magnitude’s parts is greater than n. But this does not imply that the number of the magnitude’s parts is infinite, i.e., greater than any finite magnitude.52 To speak of a syncategorematic infinite is to speak of a process that cannot have an end.53 This observation would confirm Murdoch’s interpretation. However, Ockham also claims that the following consequence is necessary: “This proposition is possible: ‘The proposition “For any finite magnitude there is a greater magnitude” is true. Therefore this proposition is possible: “A magnitude is infinite.” ’”54 Suppose that the totality of parts in a magnitude is syncategorematically infinite. On this supposition, there would not be, for any finite magnitude, a greater magnitude, because even if the series of natural numbers has no end, it has a limit. Ockham claims that this is impossible: there must be a number greater than any finite number if Aristotle’s concept of the potential infinite is to be consistent. Therefore, if a continuum has syncategorematically infinitely many parts, it has categorematically infinitely many parts.55 Ockham’s point relies on the possibility of taking the expression ‘the number of the continuous magnitude’ collectively. This is impossible in the first standard sense of the syncategorematically infinite, according to which the syncategorematically infinite has to be conceived as a process, since by definition the parts of a process cannot be taken all at once. Thus, Ockham distinguishes two cases. In the case of a process of division of the parts of the continuum, it is not valid to infer a categorematic infinite from a syncategorematic infinite. But if we consider the parts of the continuum all at once, it is valid to infer a categorematic infinite from a syncategorematic infinite. In other words, an infinite series has no last member, and therefore it is impossible to come to the last member of the series. But it does not follow that it is impossible to finish the series, i.e., to come to a state in which no member remains outstanding. Consequently, Murdoch is right in saying that, according to Ockham, the division of a continuum has no end and is in this sense infinite in a syncategorematic sense. Unlike Pasnau, however, he does not insist on the fact that the parts of the continuum exist in act before any division and that before any division they are categorematically infinite. Gregory of Rimini gives a similar analysis of the meaning of ‘infinite’ in the expression ‘infinite parts of the continuum.’ Like Ockham, he defends a thoroughly divisibilist position based on the following three theses: 52 53 54 55

Ockham, QP 67, OPh VI, 585, ll. 14–18. Ockham, QP 67, OPh VI, 585, ll. 18–22. Ockham, QP 67, OPh VI, 586, ll. 29–35. Ockham, QP 67, OPh VI, 586, ll. 35–45.

194

Magali Roques

As for the first point, I propose three conclusions. The first is that no magnitude is composed of indivisibles, from which it follows that any magnitude is composed of magnitudes. The second is that any magnitude is composed of infinitely many magnitudes. The third is that no magnitude has anything indivisible intrinsic to it.56

How should these three theses be understood? Rimini claims that if the totality of the parts of a continuum were not infinite categorematically then we would “have to give a maximum to the number of parts.” But in this case a continuum would be composed of ultimate indivisible parts, which is false.57 The division of a continuum cannot be completed in the sense that it never comes to things that have not been divided. This means that if by God’s potency the division were completed, the collection of divided things would be made up of a multitude of divisible parts and not of indivisibles.58 Wodeham makes the very same point about what he takes to be Aristotle’s view: Speaking according to the Philosopher’s understanding, a continuum can at once be divided and also at once have been divided into parts totally distinct from each other, infinite in multitude, even taking ‘infinite’ categorematically. That is, infinite parts of a continuum, totally distinct from one another, can be divided and have been divided from each other; indeed, infinitely many times infinite.59

What distinguishes the Philosopher’s view from his own is that Aristotle understood the proposition ‘A continuum can be divided ad infinitum’ as a categorematic proposition, while Wodeham understands it as a modal proposition, namely a proposition de possibili.60 Speaking according to the Philosopher’s understanding, all the parts of a continuum may be referred to together as parts of a single continuum, and as such they may be categorematically infinite. In fact, however, they are never divided at one and the same time. Thus, speaking not according to Aristotle’s understanding, no continuum can be divided at the same time into all its categorematically infinitely many parts.61 As Kretzmann remarks about Wodeham, “because his responses commit him to the position that the parts divided from each

56

Rimini, Sent. II, d. 2, q. 2, art. 1, ed. A. Trapp, vol. IV, 278, ll. 13–17. Rimini, Sent. II, d. 2, q. 2, art. 1, ed. A. Trapp, vol. IV, 296, ll. 17–25. For a commentary on this text, see Richard Cross, “Infinity, Continuity, and Composition: The Contribution of Gregory of Rimini,” Medieval Philosophy and Theology 7 (1998), 100. 58 Rimini, Sent. II, d. 2, q. 2, art. 1, ed. A. Trapp, vol. IV, 293, ll. 5–15. 59 Wodeham, Tractatus de indivisibilibus, q. 4, a. 2, 224, ll. 26–32. For a commentary on this text, see Edith Dudley Sylla, “God, Indivisibles, and Logic in the Later Middle Ages: Adam Wodeham’s Response to Henry of Harclay,” Medieval Philosophy and Theology 7 (1998), 84. 60 Wodeham, Tractatus de indivisibilibus, q. 4, a. 2, 226, ll. 7–11. 61 Wodeham, Tractatus de indivisibilibus, q. 4, a. 2, 226, ll. 2–3. 57

Ockham on the Parts of the Continuum

195

other in the division of a whole must be conceived of as infinitely many preexisting wholes that are separated on the occasion of the division, they commit him to holding that those parts exist before (as well as during and after) the division of part from part.”62 Thus, William of Ockham, Adam Wodeham, and Gregory of Rimini all share the view that Aristotle’s concept of the potential infinite has to be complemented by a kind of actual infinite. For all three theologians, the parts of the continuum exist in the continuum before any division of the continuum and are categorematically infinitely many in the continuum. In this sense, Murdoch’s interpretation cannot be maintained: the categorematically infinite in Ockham is not conceived in process-like ways. Rather, it is a property of the parts of the continuum before any division of the continuum. The question now is why Ockham claims that these parts that are categorematically many in the continuum exist in act and not in potency. After all, they could be said to exist in act only after the division of the continuum. As I will show in the second section, there must be something in the metaphysics of the parts of the continuum that allows us to consider the parts of the continuum all at once as already given before any division of the continuum. This is the existence in act of the parts of the continuum before any division. Once this has been explained, I will be able to give an answer to the compatibility problem I raised at the beginning of this paper.

2. THE METAPHYSICS OF THE PARTS OF THE CONTINUUM

a. Existence In Aristotle’s view, the parts of a continuum usually do not all actually exist. Until a line is divided there actually exist only the line and its endpoints; its various segments and their endpoints exist only potentially. The same applies to two- and three-dimensional magnitudes. Ockham does not read Aristotle in this way. He believes that Aristotle responded to the atomists precisely by saying that the parts of the continuum are in act and that they are infinite.63 Ockham specifies what he means as follows: It must thus be said that the Philosopher means that infinitely many parts are actually existing in nature. However, the Philosopher and similarly the Commentator say that 62 63

Kretzmann, “Adam Wodeham,” 391. Ockham, QP 70, OPh VI, 591, ll. 8–11.

Magali Roques

196

they are in potency and not in act, because they are not all separated, not constituting something one, nor is there a first part to be given, and a second, a third, and so on, but before any part there is another part to be given. And thus he uses equivocally in different places the words ‘to be in act’ and ‘to be in potency.’64

When Aristotle says that the parts of the continuum are in potency in the continuum, he does not say that they do not exist in act. For the parts of the continuum to be in potency means that they are ontologically dependent on the continuum: they are not separated from each other or from their whole, and they contribute to form a continuous whole that does not depend on anything else to be one and thus has a ‘per se’ unity. The parts of the continuum are therefore constitutive parts of the continuum that they contribute to form. Moreover, the parts of the continuum are infinite in the sense that a part is always divisible into two parts in such a way that none of them can be said to be the first or the last part. Ockham supports his interpretation with the following distinction between two ways of being in act: For somethings to be in act can be understood in two ways: in the first way, they are in act separate from each other in such a way that they do not form something one per se. In the second way, things are in act if they exist in nature and exist truly and really outside the soul.65

In the broader sense, something is in act when it exists in nature. In the narrower sense, something is in act when it is not part of a continuous whole that is infinitely divisible according to an order such that there is no absolute beginning of the division. The first sense is broader than the second because it includes it. Thus, the parts of the continuum will be said to exist in the same way in which anything is said to exist: they satisfy the first sense of ‘to exist’ but not the second. Indeed, the opposite of ‘to exist’ is not ‘to be ontologically dependent’—in Ockham’s terms, “not to be actually separate”— but “to be a possible object”: The Philosopher says that the parts of the continuum are in potency because they are not actually separate. But he does not mean that the parts of the continuum are in potency in the same way as the soul of the Antichrist is now in potency, i.e. when he is nothing, because if this were the case, they could never really compose [with each other] in a continuum; therefore, he means that they are something outside their cause. But they are in potency because any part in a continuum is not a certain separate whole existing per se.66

64 65 66

Ockham, ExpPhys VI, 13, §6, OPh V, 564, ll. 79–85. Ockham, QP 67, OPh VI, 588, ll. 12–15. Ockham, Quaest. Var. III, OTh VIII, 78, ll. 321–8.

Ockham on the Parts of the Continuum

197

To be in act in the broader sense is opposed to being a possible object, i.e., being nothing but being able to exist.67 Since it can exist, it is not nothing in the way that impossible objects are nothing, i.e., in the sense that being is incompatible with the sum of their defining properties.68 A possible being does not have a way or mode of being distinct from actual being: when it is not actual it is simply nothing. Indeed, just as there is no such entity as the possible being of a thing, distinct from the thing itself, similarly there is no such entity as a thing’s intelligible being.69 In contrast with possible beings, existing beings have a cause of being, which is at least partly really distinct from it. Since the parts of the continuum do not satisfy the clause of independent existence, why does Ockham bother to find a sense in which the parts of the continuum exist? He could just claim, like Aristotle, that they are things that are possibly separate, thus possibly existing in the most proper sense of ‘to exist.’ Ockham answers that if the parts of the continuum did not exist, their whole would not exist: No real being that exists in act is composed of some non-being; but the continuum is a real being that exists in act and it is composed of all its parts; therefore etc.70

Ockham’s choice is motivated by a basic mereological principle that also governs his argument against the existence of successive beings such as time and motion.71 The basic idea is that since the parts of the continuum contribute to make the continuum the whole that it is, they must exist. However, Ockham does not specify in what sense the parts of the continuum, which are ontologically dependent on their whole—and thus do not exist in the stricter sense of the term—are also ontologically prior to their whole. Does Ockham mean that if one of its parts ceased to be a part of it the continuum would cease to be? In this case, his position would be quite similar to Wodeham’s, according to which a continuum ceases to be as soon as it is divided into two parts. Nevertheless, Ockham tries to defend his view with the following argument: There is no more reason for one part of the continuous body to actually exist in nature than for another to exist; but some part of a continuous body does actually exist; therefore each part does.72 67

Ockham, SL I, 38, OPh I, 108, ll. 54–6. For a commentary on this text, see Elizabeth Karger, “Would Ockham Have Shaved Wyman’s Beard?,” Franciscan Studies 40 (1980), 246–7. 68 Ockham, Ord, d. 36, OTh IV, q. 1, 547, ll. 11–19. 69 Ockham, Ord, d. 36, OTh IV, q. 1, 550, ll. 13–20. 70 Ockham, QP 68, OPh VI, 587, ll. 6–8. For a commentary on these arguments, see Pasnau, Metaphysical Themes, 616–18. 71 Ockham, Exp.Phys. IV, 18, §3, OPh V, 199, ll. 118–24. For a commentary, see Adams, William Ockham, 799–827; Pasnau, Metaphysical Themes, 374–400. 72 Ockham, QP 68, OPh VI, 588, ll. 27–9. The translation is from Pasnau, Metaphysical Themes, 618.

198

Magali Roques

The minor premise is backed up by appeal to the fact that parts can be the subject of accidents; they can be perceived and they can act.73 The argument thus seems to run as follows. Let us grant, as Ockham apparently does, that some parts of physical bodies, such as integral parts, are continuous. Then, in a Moorean way, just as I cannot deny that I have two hands, I cannot deny that at least the integral parts that I perceive exist. Therefore there are at least some parts of the continuum that exist. Since there is no more reason why one part would exist rather than another, they all exist. This is confirmed by another argument, which states that “an integral part of a whole really exists no less than does an essential part; but an essential part actually exists, as is clear for matter and form; therefore every integral part actually exists.”74 Finally, Ockham argues as follows: “half of a whole continuous body really exists, as is clear to the senses; therefore every half really exists, because the same reason holds for one and for all.”75 Of course, these arguments are not really persuasive, since the concept of continuum is at least not physical in the same way as a living being or my hand is: it is not a biological reality. Indeed, even though Ockham does not elaborate this point, in principle, there should be a distinction between the parts of continuum and the integral parts of the body, since not every continuous thing is a living body. The difference between these two kinds of parts appears at the phenomenological level. I never saw my hand qua part of a continuum. If I ever saw it qua part, I saw it qua integral part of my body. From this point of view, the Moorean is entitled to doubt that a part of the continuum is really as palpable as Ockham wants it to be.

b. Actuality One could therefore object that the distinction between two ways of being in act is ad hoc. Ockham does not justify his core argument in favor of the existence of the parts of the continuum, namely that the continuum is ontologically dependent on its parts. It seems that nothing prevents us from claiming that the parts could be said to be in potency. After all, they are said to exist in a weaker sense of ‘to exist,’ which is strikingly similar to the sense of ‘to be in potency’ when it is predicated of ‘the parts of the continuum.’

73

Ockham, QP 68, OPh VI, 588, ll. 19–36. Ockham, QP 68, OPh VI, 588, ll. 36–40. The translation is from Pasnau, Metaphysical Themes, 618. 75 Ockham, QP 68, OPh VI, 588–9, ll. 41–3. The translation is from Pasnau, Metaphysical Themes, 618. 74

Ockham on the Parts of the Continuum

199

This is how Chatton argues. He claims that the (indivisible) parts of the continuum are in potency and not in act in the continuum because they are not separated from each other.76 Chatton holds that a continuum is composed of a finite number of indivisibles and that it is not infinitely divisible.77 We might thus think that the assumption that a continuum is composed of an infinity of parts existing in act in the continuum is a consequence of Ockham’s anti-indivisibilism. Ockham is not the only anti-indivisibilist to claim that the parts of the continuum exist in act in the continuum. According to Gregory of Rimini, the parts of the continuum are commonly said to be in potency not because they do not exist or because they are not really distinct from each other, but because none of them exists separately from the others in the continuum: they are not individuals (hoc aliquid) properly speaking. However, the parts of the continuum “actually exist by the actuality of presence.” In this, they are opposed to mere possible objects, which are nothing, as in Ockham.78 The parts of the continuum together form a collection that is an individual, as opposed to a mere heap that is nothing more than a plurality. A continuum cannot be broken by a physical division without ceasing to be. However, it can be divided conceptually or, in Gregory’s terms, by “a division which is made by the signification of the distinguishing intellect, which division is compatible with real continuity in the continuum thus divided.”79 This description is very close to Ockham’s description of the mode of being of the parts of the continuum, as has been noted by Richard Cross.80 For both Ockham and Gregory, the parts of the continuum exist in act in the continuum by an actuality of their own. Gregory’s description suffers from the same ambiguity as Ockham’s: he does not specify what he means by the ontological dependence of the parts of the continuum on their whole, nor does he explain what he means by the ontological dependence of the continuum on its parts. Thus, either Ockham and Gregory are both actualists—in the sense in which Pasnau uses the term—or they are not. As I explained in the introduction, Pasnau thinks that Ockham accepts an actual infinity of parts in the stronger sense, on the ground that these parts are actual. He calls this position “actualism.”

76 Walter Chatton, Reportatio II, d. 2, q. 3, ed. J. Wey and J. Etzkorn (Toronto: Pontifical Institute of Mediaeval Studies, 2004), 129, ll. 20–9. 77 Chatton, Reportatio II, d. 2, q. 3, 118, ll. 26–9. 78 Rimini, Sent. I, d. 24, q. 1, ed. D. Trapp and V. Marcolino (Berlin: de Gruyter, 1984), vol. 3, 18 (l. 27)–19 (l. 5). 79 Rimini, Sent. I, dd. 35–6, q. 1, vol. 3, p. 223, ll. 16–25. 80 Cross, “Infinity, Continuity and Composition,” 97, n. 35.

Magali Roques

200

Ockham thinks he is an actualist. Indeed, he attempts to go one step further in his argument and claims that the parts of the continuum are in act by their own actuality, and not from the actuality of the whole. He thus tries to prove that a part has the same actuality when it exists in a continuum as when it exists separated from it.81 This amounts to saying that the division of the continuum does not make any ontological difference in what there is. This is a strong claim. It goes far beyond the mere claim that the parts exist in the continuum, but instead is mainly about the relation of dependence of the parts on the continuum. Let us examine the arguments that Ockham puts forward in order to explain what he means. His first argument relies on the claim that actuality is not a property of a thing really distinct from it.82 Since the actuality of a part is nothing other than the part and the actuality of the whole is nothing other than the whole, and since the part and the whole are really distinct from each other, the actuality of the part does not depend on the actuality of the whole.83 Of course, the inference does not hold: although the part and the whole are really distinct from each other, at least one of them depends on the other to exist, which blocks the inference. Ockham thinks that this relation of mutual dependence does not change anything for the way of being of its relata. To prove this, he relies on the idea that the division of a continuum is not a real Aristotelian change, in the sense that nothing new is produced when a continuum is divided.84 This means that nothing changes regarding the actuality of the part of the continuum when the continuum is divided. Therefore, when separated from the continuum, the part has the same actuality as when it existed in the continuum. Therefore its actuality in the continuum was not dependent on the actuality of the whole.85 Is this second argument any more convincing than the first? In Ockham’s distinction between two senses of being in act, it makes a difference whether the part of a continuum is separated from the continuum or not, since when a part exists in the continuum it is not in act in the narrower sense. How is this claim compatible with the claim that the part has the same actuality when it is separated from the continuum as when it exists in the continuum? Ockham’s motivation for distinguishing between two ways of being in act is not so clear after all. This distinction might have a purely exegetical motivation: it would allow Ockham to avoid contradicting Aristotle explicitly 81 82 83 84 85

Ockham, QP 69, OPh VI, 589, ll. 10–12. Ockham, QP 69, OPh VI, 590, ll. 23–7. Ockham, QP 69, OPh VI, 590, ll. 19–21. Ockham, QP 69, OPh VI, 590, ll. 28–37. Ockham, QP 69, OPh VI, 590, ll. 30–2.

Ockham on the Parts of the Continuum

201

when he says that the parts of the continuum exist only in potency in the continuum. It might also have a philosophical motivation, though I do not see what that would be, except that it might be related to part–whole talk or, more precisely, to the relation of mutual dependence between a continuum and its parts. In any case, Pasnau is right in saying that Ockham defends an actualist position: there is an infinity of parts existing in act in the continuum, even though Ockham’s point is not so clear. We are now in a position to propose an answer to the puzzle that was presented in the introduction. Let us recall that Ockham claims that (T1) “whether a part exists in the continuum or is separated from it, it exists properly speaking by an existence and actuality of its own, not of the whole.” T1 seems to imply that (T2) “since a part exists in act in the continuum by an actuality of its own, it can be separated from any other part of the continuum, and therefore any part of the continuum can be separated from the continuum, that is, can be divided in act from it.” However, T2 seems to contradict another explicit claim made by Ockham, namely that (T3) the infinite divisibility of the continuum requires that there is always a part that can be divided from the continuum. Ockham’s solution consists in saying that T2 does not contradict T3 but supports his reading of it, according to which all the parts of the continuum are already there before the division of the continuum. They are already there in the sense that in the continuum the parts have a strong ontological status: their existence does not depend, in any sense of the term ‘depend,’ on the existence of their whole. The process of dividing a continuum is something Ockham ascribes to our minds, but the mind only takes the already existent parts involved in this division; it does not create them. But this is not the end of the story. Ockham has not yet proven that his new analysis of the infinite divisibility of the continuum is consistent. Does the concept of the categorematic infinite he proposes imply, as Aristotle thought, that a continuum is a magnitude that is infinite in extension? In this case, Ockham should give up his analysis. Ockham does not think that he has to, as I will show in the last section of the paper.

3 . T H E M A T H E M A T IC S O F T H E I N F I N I T E

a. Consecutiveness without Adjacency To make his point, Ockham proceeds in two steps. First, he has to prove the assumption that a complete division of the continuum does not result in an infinite number of indivisible parts which are adjacent to each other and which, once combined, would yield a continuum that is infinite in

202

Magali Roques

extension. Second, since he accepts the existence of infinities in act, he has to show how it is possible that infinites are unequal. Let us begin with the claim that a complete division of the continuum does not result in an infinite number of indivisible parts that are adjacent to each other. As we saw when examining his interpretation of T3, Ockham wants to claim, unlike Wodeham, that the infinite division of a continuum is possible, at least in thought. In this, he agrees with Henry of Harclay, who defends the claim that the last level of a complete division of a continuum should include infinitely many parts.86 The central claim in Harclay’s indivisibilist position is that if a continuum is composed of an actual infinity of parts, then it will be composed of parts, each of which is immediately adjacent to at least one other and at most two others.87 In order to answer Aristotle’s most powerful objection to the indivisibilists, namely the paradox of touching, Harclay redefines the concept of adjacency.88 Indivisibles either have a magnitude or they do not. If they have magnitude, they are divisible; but then their parts are divisible, and so on ad infinitum. But if they do not have any magnitude, they cannot form a continuous magnitude because they cannot be in contact. If two things are in contact they touch part to part, part to whole, or whole to whole. The first two possibilities must be rejected for atoms, because by definition they do not have parts. But if they touched whole to whole, they 86 On this subject, see especially John Murdoch, “Henry of Harclay and the Infinite,” in A. Maierù and A. Paravicini-Bagliani (eds.), Studi sul XIV secolo in memoria di Anneliese Maier (Rome: Edizioni di storia e letteratura, 1981), 219–60; Anneliese Maier, “Diskussionen über das aktuell Unendliche in der ersten Hälfte des 14. Jahrhunderts,” Divus Thomas 25 (1947), 147–66, repr. in Maier, Ausgehendes Mittelalter (Rome: Edizioni di storia e letteratura, 1964), vol. 1, 40–85; Maier, “Das Problem des Kontinuums in der Philosophie des 13. und 14. Jahrhunderts,” Antonianum 20 (1945), 331–68, repr. in Maier, Die Vorläufer Galileis (Rome: Edizioni di storia e letteratura, 1949), 155–79. Ockham does not refute Harclay’s infinitist atomism, unlike Wodeham (Tractatus de indivisibilibus, q. 1, a. 3, 103–22; for a commentary on this text, see Sylla, “God, Indivisibles, and Logic”). However, in QP 71, Ockham constructs his own position in opposition to an infinitist indivisibilist position like Harclay’s, which shows that he was clearly aware of such positions. 87 On the immediate adjacency of Harclay’s infinite indivisibles, see Murdoch, “Henry of Harclay and the Infinite,” 230; also Murdoch, “Superposition, Congruence and Continuity in the Middle Ages,” in I. B. Cohen and R. Taton (eds.), L’aventure de la science: Mélanges Alexandre Koyré I (Paris: Hermann, 1964), 430–1; Murdoch, “Infinity and Continuity,” 577, n. 39. 88 Aristotle, Physics, VI, 1, 231a26–b6. This presentation is based on Aurélien Robert, “Atomism,” in H. Lagerlund (ed.), Encyclopedia of Medieval Philosophy (Dordrecht: Springer, 2011), 122–5. For an extensive discussion of the fate of this argument in the Middle Ages, see Murdoch, “Superposition, Congruence and Continuity,” 416–41. For the reception of this argument in Crathorn, see Magali Roques, “Crathorn on Extension,” Recherches de philosophie et théologie médiévales 83 (2016), 423–67.

Ockham on the Parts of the Continuum

203

would be superimposed and therefore could not constitute a magnitude; in other words, two adjacent indivisibles would occupy no more space than one, whether the indivisibles were point-like or extended. By extension, no finite number of adjacent indivisibles would occupy any more space than one would. In responding to the paradox of touching, Harclay defends the idea that a magnitude is composed of indivisibles such that each one of these indivisibles touches at least one other, and at most two others. They do so by being in places that are immediately adjacent to each other.89 For two atoms to be adjacent means that they are situated in two adjacent places. Ockham is not an indivisibilist.90 His aim is to develop a new account of the division of a continuum that does not leave room for the possibility of ultimate indivisible parts that are adjacent because they are in adjacent places. Unsurprisingly, when he explains just what kinds of infinitely numerous parts can obtain in a continuum, he reworks the notion of adjacency or consecutiveness.91 To do this, Ockham relies on his claim that the parts of the continuum are not created during the process of division. He starts with a simple case in which the extended parts of a continuum overlap in the sense that each contains other parts. He then complicates the case step by step until he reaches a situation in which some parts of the continuum are neither overlapping, as they would be on the common view, nor adjacent to each other, as they would be on Harclay’s view. In this way, the possibility of merely adjacent ultimate parts of the continuum is neutralized and the possibility of a complete division of the continuum is warranted. On Ockham’s account, a body is not composed of adjacent parts, but of extended parts, each of which includes infinitely many smaller parts. In this way, Ockham thinks, he is able to claim both that the division of a continuum can be completed, at least conceptually, and that the parts resulting from the division are not indivisible. For instance, in a quantity of two feet there can exist an infinity of parts such that each of them is a quantity of one foot, on the condition that these parts are ordered in a certain way, as described in the five conclusions of question 71 on the Physics. In other words, a continuum can be divided into proportional parts infinitely many times, but none of these divisions will result in a level which 89 Henry of Harclay, Ordinary Questions XV–XXIX, ed. M. Henninger (Oxford: Oxford University Press/British Academy, 2008), 1058, ll. 737–9. For a commentary on this text, see Murdoch, “Henry of Harclay and the Infinite,” 230; and Richard C. Dales, “Henry of Harclay on the Infinite,” Journal of the History of Ideas 45 (1984), 295–301. 90 Ockham, QP 70, OPh VI, 591, ll. 8–12. 91 For an analysis of the parallel passages from ExpPhys III, see Murdoch, “William of Ockham,” 186–7.

204

Magali Roques

contains infinitely many equal parts in every dimension. Thus, between proportional parts there is a first division but no first part, since there will always be a part from a possible former division.92 Consequently, in a continuum there is an infinity of parts that are not consecutive (or adjacent) to each other. These parts, of which none is first and which are not consecutive, are “totally distinct from each other,” just as my foot is totally distinct from my hand.93 By contrast, an infinity of parts all of which are separate from one another and of which there is an absolutely first part will not yield a finite continuum.94 However, if there is no absolutely first part and the parts overlap or are parts of each other, then such an infinity of parts composes a finite continuum.95 Moreover, if there is no absolutely first part, a continuum can be composed of an infinity of parts all of which are wholly separate from one another.96 Going one step further, if there is no absolutely first part, a continuum can be composed of an infinity of parts which are of the same quantity according to one or two dimensions, and are wholly separate from one other.97 Consequently, proportional parts are defined not so much by overlapping as by the fact that before any given part there is always another part.98 Moreover, since real distinction is sufficient for countability, the parts of the continuum are in a sense countable, even if they are categorematically infinite. How should we understand this? In a continuum there are infinitely many overlapping parts, each of which actually exists. We can locate some of these parts by counting the parts that are obtained by dividing infinitely many times approaching one of the limits of the body, such that each part includes the limit of the body. If we do this, we will obtain infinitely many parts, each half the size of the previous one, and each one including the limit of the body. Thus, each one of the infinitely many parts will be immediately adjacent to any body which touches the limits of the divided magnitude. So touching a body entails touching all of the infinitely many proportional parts that are obtained by dividing ever closer to the limit of the body. Now we see the importance of Ockham’s claim that “no part is first,” i.e., that before any part there is always some other part. Murdoch thinks that Ockham elaborates this new method of division of the parts of the continuum in order to prove that an infinity of actually 92

Ockham, QP 71, OPh VI, 596, ll. 28–30. Ockham, QP 71, OPh VI, 595, ll. 7–9. Ockham, QP 71, OPh VI, 596, ll. 40–4. 95 Ockham, QP 71, OPh VI, 596, ll. 11–12. 96 Ockham, QP 71, OPh VI, 596, ll. 17–19. 97 Ockham, QP 71, OPh VI, 596, ll. 31–7. 98 Ockham, QP 70, OPh VI, 593–4, ll. 93–9. See also Ockham, Quaest. Var. III, OTh VIII, 75. 93 94

Ockham on the Parts of the Continuum

205

existing parts is not categorematically infinite, which would yield a magnitude that is infinite in extension.99 Indeed, in a continuous magnitude, Ockham claims, there is by definition no first part in the sense that another prior part could always be picked out. This suffices to block Aristotle’s famous argument according to which, if there were categorematically infinitely many parts existing in the continuum, the continuum would be a magnitude that is infinite in extension. However, if Murdoch thinks that Ockham does not accept that the parts of the continuum are categorematically infinitely many, his reading of Ockham’s texts is not correct. Given any continuum, we are also given all its parts, which are actually existent. In this way, we can take all the parts of the continuum all at once or collectively. And it is in this sense, as we saw before, that the parts of the continuum are categorematically infinite. Ockham’s contemporaries seem to have understood him in this way. An unknown objector says that, given this analysis of the relation between the parts of the continuum, there are unequal infinites, which is impossible. Since there is an infinity in act of existing parts in a continuum, and since not every continuum has as many parts as any other, there are unequal infinites in nature. But this runs directly counter to Aristotle. How does Ockham answer this objection? He simply grants it and gestures toward a new mathematics of the infinite, as Harclay did before him.

b. The Inequality between Infinites Clearly, on Ockham’s account, the number of parts in a continuum is countably infinite: each part can be paired off with a natural number. Ockham thus believes conceptual divisibility to be in some sense sufficient for countability. This is a consequence of his very notion of infinite divisibility. If a continuum is made up of actually infinite parts, there is a sense in which a number can also be said to be actually infinite, namely, the number which corresponds to those infinite parts, which are not of any definite number and which are really distinct from one another. Ockham gives a positive interpretation of the conception of the actual infinite and of a mathematical or numerical infinite. This represents a shift in the history of the idea of the infinite. Ockham explains his ideas when he responds to several objections that are raised against his analysis of the infinite divisibility of the continuum. All are based on the so-called paradoxes of the infinite, i.e., the paradox of infinites 99 Murdoch, “William Ockham,” 189: “Thus, in spite of his conviction that the parts actually exist in continua, there is no question at all but that Ockham firmly believed that the divisibility of a continuum was only potential.”

206

Magali Roques

that are parts of, and yet at the same time equal to, other infinites. Murdoch explains that the structure of these paradoxes was simple in Ockham’s time: if one concedes the existence of actual infinites, then some infinites will be greater than others that are parts of the former; but it is axiomatic that all actual infinites are equal; therefore, in this instance a part is not less than its whole, but equal to it, which is absurd.100 Ockham answers that it makes sense to say that one infinite quantity is smaller or greater than another, but we have to be careful to specify in what sense. He clarifies his answer when he deals with the paradox of the infinitely large, which shows that in the infinite case it is possible for one collection to be contained within another and for their members to be susceptible to being paired off. In such circumstances, Ockham claims, the former collection would be smaller than the latter in one sense but not in another. In the stricter sense, for some things to be greater than other things, it is required that both be of a determinate number and that the first be of a greater number than the second. In this sense, one infinite cannot be greater than another. In the broader sense, however, for a collection of things to be greater than another, it is sufficient that both collections be paired off and that the first collection contains the second and something more. In this broader sense, one infinite can be greater than another.101 Ockham’s distinction is quite similar to the distinction that Harclay uses in his answer to the paradox of the infinitely large. Harclay states that, for a whole to be greater than another, it suffices that the first includes the other and contains something more.102 However, Ockham’s formulation is more precise. He distinguishes between two ways of counting, namely, by means of a determinate number and by means of a number which is not determinate.103 In the stricter sense, no number can be infinite in act, because in the parts of the continuum the process of division has no end, and consequently the number of all the parts of the continuum taken divisim has a limit such that any number picked out at the end of a division is necessarily

100 On the history of the paradoxes from Bonaventure to Gregory of Rimini, see John Murdoch, “Mathematics and Infinity in the Later Middle Ages,” Proceedings of the American Catholic Philosophical Association 55 (1981), 52–5. See also Murdoch, “Infinity and Continuity,” 569–73. 101 Ockham, QP 70, OPh VI, 593, ll. 75–83. 102 Henry of Harclay, Quaestio II, C, 4 as quoted by Murdoch, “Henry of Harclay and the Infinite,” 238, n. 45: “Igitur ista propositio est magis nota: ‘Quod omne quod continent aliud et addit super illud est magis illo’ quam ‘Omne totum est maius sua parte.’ ” For Gregory of Rimini’s formulation, see Cross, “Infinity, Continuity, and Composition,” 106. 103 Ockham, QP 71, OPh VI, 593, ll. 75–83. See also Ockham, Quodl II, q. 5, OTh IX, 131–2, ll. 72–9.

Ockham on the Parts of the Continuum

207

finite.104 In the broader sense, there is an infinite number in act, because any part of the continuum is really and totally distinct from another and exists in act in such a way that there are categorematically infinitely many parts in the continuum before any division.105 This infinite number in act is such that it cannot be grasped by means of a determinate number.106 As Murdoch puts it, in differentiating two meanings for ‘plura’ Ockham had come close to formulating, on the one hand, a notion of unequal cardinality (which is not applicable to infinites) and, on the other hand, a notion of set-subset (which is applicable to infinites). In this he appears to have marked an advance over his predecessors who also attempted to delineate new notions of excess and defect for infinites.107

Moreover, Ockham’s answer to the paradox of the millet seed clearly shows that he accepts equal infinites. In its medieval version, the paradox states that if infinites can be equal to each other, then one must admit that there are as many parts in a millet seed as in the sky.108 Ockham boldly accepts that this can be the case. He answers that, even though smaller magnitudes can be seen as subsets of larger ones, all continuous magnitudes have the same number of parts in the sense that the proportional parts of any magnitude can be paired off with the proportional parts of any other.109 This runs counter to Murdoch’s claim that Ockham provides no sense in which one infinite is equal to another.110 Ockham clearly recognizes the requirement that any number of “normal” mathematical properties be applicable to infinites. Infinites are in principle susceptible to addition, subtraction, and multiplication, and to excess, defect, and other related properties. Ockham’s analysis of the categorematically infinite therefore represents an important advance on that of his predecessors and contemporaries. He is the first thinker to provide a satisfactory rejection of Aristotle’s account of composition from a potential infinity of parts, and the first thinker to provide a clear solution to the paradoxes of inequality that seem to arise if we accept the infinitist conception of the composition of a continuum. Equally, he is the first to provide an account of the view that there are 104

Ockham, QP 109, OPh VI, 688–9, ll. 17–23. Ockham, QP 109, OPh VI, 689, ll. 30–4. 106 Ockham, QP 109, OPh VI, 689, ll. 47–9. For the same distinction, see Ockham, ExpPhys III, 15, OPh IV, 575. 107 Murdoch, “William Ockham,” 170. 108 Aristotle’s version of the paradox seems to concern the reliability of the senses. See Aristotle, Physics, VII, 5, 250a19. 109 Ockham, QP 70, 594, ll. 110–20. 110 Murdoch, “William Ockham,” 170. 105

Magali Roques

208

actually infinitely many parts in a continuum. In this sense, Ockham anticipates many insights that are usually attributed to Gregory of Rimini. All in all, however, even though Ockham claims explicitly that infinites can be unequal and that the relation ‘greater than’ can be applied to infinites, he does not propose a mathematics of the infinite in the sense of treating of mathematical operations that bear upon the infinite. He himself acknowledges that he does not have the logical or mathematical tools necessary to state more precisely which infinite is greater than which other one or how it is greater.111 This is why he claims that parity and imparity are properties only of finite numbers. A number is a collection of units measured by the one; but no infinite collection can be measured by the one.112 In other words, Ockham acknowledges that he is unable to determine the cardinality of infinites. In this sense, he is no forerunner of Cantor.

CONCLUSION Aristotle’s way of talking about the divisibility of a continuum seems to contradict two modal principles at the core of his metaphysics of modality. Various answers were given to this problem at the beginning of the fourteenth century. Adam Wodeham defends a view that Kretzmann called “macro-indivisibilism.” On this view, the continuum is not divisible properly speaking. William of Ockham and Gregory of Rimini defend a view that is in some respects closer to Aristotle’s: a continuum can properly speaking be divided infinitely, at least in thought. How is this claim compatible with Aristotle’s modal logic? Ockham answers that the parts of the continuum are all already there before any division, be it real or mental; the parts must therefore be conceived of not as created in the process of the division of a continuum but as infinitely many pre-existing ordered parts that are separated on this occasion. This means that the parts are in act in the continuum by an actuality of their own. Consequently, there is a sense in which Ockham is committed to the actual existence of infinities, which he explicitly acknowledges. He finds a way to spell out the relations between the parts of the continuum that does not imply either that there is any end to the division (i.e., indivisible entities that are infinitely many in the continuum) or that a continuum is an infinitely extended magnitude. In this sense, Ockham’s anti-indivisibilist position is based on a concept of the categorematically infinite which is completely 111 112

Ockham, Quaest. Var. III, OTh VIII, 80–1, ll. 364–76. Ockham, QP 70, OPh VI, 593, ll. 84–92.

Ockham on the Parts of the Continuum

209

foreign to Aristotle. He goes beyond the advances of the former generation, who, in the wake of Duns Scotus, developed an arsenal of mathematical arguments against indivisibilism.113 Indeed, he sets up a conceptual analysis of the idea of infinite divisibility that will have a long-lasting impact on his contemporaries, since we find parts of this analysis in Wodeham and Rimini. From this perspective, Murdoch’s claim that Ockham “does not bear witness to these new factors in anything near the measure one might expect” must be qualified.114 Ockham’s nominalistic stance on the ontological status of indivisibles was very influential. His conceptual analysis of the “nature of the continuum,” i.e., infinite divisibility, was also very influential. It marks a shift from Aristotle’s concept of the infinite as incomplete and boundless to the more modern conception of the infinite as something actual and complete. What is interesting is to see this shift happening not in a theological context, but in a strictly philosophical one, since in his analysis Ockham deliberately abstains from any appeal to God’s absolute power or to any theological hypothesis. Since the parts of the continuum are already there before any division, the parts of the continuum are not defined or identified by means of a division, as was the case in Aristotle. Ockham, Wodeham, and Rimini substitute for the Aristotelian idea of a process without an end the idea of a construction of the whole from the parts. In this sense, there is an important change taking place in the fourteenth century in the very concept of continuity; nevertheless, some aspects of the metaphysics of the parts of the continuum remain unclear, especially the idea that the parts exist in act in the continuum. University of Geneva

BIBLIOGRAPHY Adam Wodeham. Tractatus de indivisibilibus, ed. R. Wood (Dordrecht: Kluwer, 1988). Adams, Marilyn McCord. William Ockham, 2 vols. (Notre Dame, IN: University of Notre Dame Press, 1987).

113 For these arguments, see especially Cecilia Trifogli, “Scotus and the Medieval Debate about the Continuum,” Medioevo 29 (2004), 233–66; Jean-Luc Solère, “Scotus geometres: The Longevity of Duns Scotus’s Geometric Arguments against Indivisibilism,” in M. Dreyer, E. Mehl, and M. Vollet (eds.), La posterité de Duns Scot/Die Rezeption des Duns Scotus/Scotism through the Centuries. Proceedings of “The Quadruple Congress” on John Duns Scotus, Part 4 (Münster-in-W.: Aschendorff/St. Bonaventure, NY: Franciscan Institute, 2013), 139–54. 114 Murdoch, “William Ockham,” 167.

210

Magali Roques

Bell, John L. The Continuous and the Infinitesimal in Mathematics and Philosophy (Milan: Polimetrica, 2006). Birch, T. Bruce. “The Theory of Continuity of William Ockham,” Philosophy of Science 3 (1936), 494–505. Burns, C. Deslile. “William of Ockham on Continuity,” Mind 25 (1916), 506–12. Cross, Richard. “Infinity, Continuity and Composition: The Contribution of Gregory of Rimini,” Medieval Philosophy and Theology 7 (1998), 89–110. Dales, Richard C. “Henry of Harclay on the Infinite,” Journal of the History of Ideas 45 (1984), 295–301. de Boer, Sander W. “The Importance of Atomism in the Philosophy of Gerard of Odo (O.F.M.),” in C. Grellard and A. Robert (eds.), Atomism in Late Medieval Philosophy and Theology (Leiden: Brill, 2009), 85–106. Gregory of Rimini. Lectura super primum et secundum Sententiarum II, ed. A. Trapp (Berlin: de Gruyter, 1979), vol. IV. Gregory of Rimini. Lectura super primum et secundum Sententiarum I, ed. D. Trapp and V. Marcolino (Berlin: de Gruyter, 1981), vol. III. Henry of Harclay. Ordinary Questions XV–XXIX, ed. M. Henninger (Oxford: Oxford University Press, 2008). Hintikka, Jaakko. Time and Necessity: Studies in Aristotle’s Theory of Modality (Oxford: Clarendon Press, 1973). John Buridan. Tractatus de infinito, ed. J. M. M. H Thijssen (Nijmegen: Ingenium, 1991). John Duns Scotus. Ordinatio II, ed. K. Balic et al. (Vatican City: Typis Polyglottis Vaticanis, 1973). Karger, Elizabeth. “Would Ockham Have Shaved Wyman’s Beard?” Franciscan Studies 40 (1980), 244–64. Knuuttila, Simo. Modalities in Medieval Philosophy (London: Routledge, 1993). Kretzmann, Norman. William of Sherwood’s Treatise on Syncategorematic Words (Minneapolis: University of Minnesota Press, 1968). Kretzmann, Norman. “Sensus Compositus, Sensus Divisus, and Propositional Attitudes,” Medioevo 7 (1981), 195–229. Kretzmann, Norman. “Syncategoremata, Sophismata, Exponibilia,” in N. Kretzmann, Anthony Kenny, and Jan Pinborg (eds.), The Cambridge History of Later Medieval Philosophy (Cambridge: Cambridge University Press, 1982), 211–45. Kretzmann, Norman. “Adam Wodeham’s Anti-Aristotelian Anti-Atomism,” History of Philosophy Quarterly 1 (1984), 381–98. Lovejoy, Arthur. The Great Chain of Being (Cambridge, MA: Harvard University Press, 1936). Maier, Anneliese. “Das Problem des Kontinuums in der Philosophie des 13. und 14. Jahrhunderts,” Antonianum 20 (1945), 331–68; repr. in Maier, Die Vorläufer Galileis im 14. Jahrhundert (Rome: Edizioni di storia e letteratura, 1949), 155–79. Maier, Anneliese. “Diskussionen über das aktuell Unendliche in der ersten Hälfte des 14. Jahrhunderts,” Divus Thomas 25 (1947), 147–66; repr. in Maier, Ausgehendes Mittelalter (Rome: Edizioni di storia e letteratura, 1964), vol. 1, 40–85.

Ockham on the Parts of the Continuum

211

Maier, Anneliese. Die Vorläufer Galileis im 14. Jahrhundert (Rome: Edizioni di storia e letteratura, 1949). Murdoch, John. “Superposition, Congruence and Continuity in the Middle Ages,” in I. B. Cohen and R. Taton (eds.), L’aventure de la science: Mélanges Alexandre Koyré I (Paris: Hermann, 1964), 416–41. Murdoch, John. “Henry of Harclay and the Infinite,” in A. Maierù and A. Paravicini-Bagliani (eds.), Studi sul XIV secolo in memoria di Anneliese Maier (Rome: Edizioni di storia e letteratura, 1981), 219–60. Murdoch, John. “Mathematics and Infinity in the Later Middle Ages,” Proceedings of the American Catholic Philosophical Association 55 (1981), 40–58. Murdoch, John. “Infinity and Continuity,” in N. Kretzmann, Anthony Kenny, and Jan Pinborg (eds.), The Cambridge History of Later Medieval Philosophy (Cambridge: Cambridge University Press, 1982), 564–91. Murdoch, John. “Ockham and the Logic of Infinity and Continuity,” in N. Kretzmann (ed.), Infinity and Continuity in Ancient and Medieval Thought (New York: Cornell University Press, 1982), 165–206. Murdoch, John and J. M. M. H. Thijssen. “Buridan on Infinity,” in J. M. M. H. Thijssen and J. Zupko (eds.), The Metaphysics and Natural Philosophy of John Buridan (Leiden: Brill, 2001), 127–49. Pasnau, Robert. Metaphysical Themes 1274–1671 (Oxford: Clarendon Press, 2011). Perini Santos, Ernesto. “La structure de la proposition modale ockhamienne,” in A. Maierù and L. Valente (eds.), Medieval Theories on Assertive and Non-Assertive Language (Florence: Olschki, 2004), 355–75. Robert, Aurélien. “Atomisme et géométrie à Oxford au XIVe siècle,” in S. Rommevaux (ed.), Mathématiques et connaissance du monde réel avant Galilée (Paris: Omniscience, 2010), 15–85. Robert, Aurélien. “Atomism,” in H. Lagerlund (ed.), Encyclopedia of Medieval Philosophy (Dordrecht: Springer, 2011), 122–5. Roques, Magali. “Introduction,” in Guillaume d’Ockham, Traité sur la quantité et traité sur le corps du Christ, introd., trad. et notes M. Roques (Paris: Belles Lettres, 2014), LXXVI–LXXXVI. Roques, Magali. “Crathorn on Extension,” Recherches de philosophie et théologie médiévales 83 (2016), 432–67. Solère, Jean-Luc. “Scotus geometres: The Longevity of Duns Scotus’s Geometric Arguments against Indivisibilism,” in M. Dreyer, E. Mehl, and M. Vollet (eds.), La posterité de Duns Scot/Die Rezeption des Duns Scotus/Scotism through the Centuries. Proceedings of “The Quadruple Congress” on John Duns Scotus, Part 4 (Münster-in-W.: Aschendorff/St. Bonaventure, NY: Franciscan Institute, 2013), 139–54. Stump, Eleonore. “Theology and Physics in De Sacramento Altaris: Ockham’s Theory of Indivisibles,” in N. Kretzmann (ed.), Infinity and Continuity in Ancient and Medieval Thought (New York: Cornell University Press, 1982), 165–206. Sylla, Edith. “God, Indivisibles, and Logic in the Later Middle Ages: Adam Wodeham’s Response to Henry of Harclay,” Medieval Philosophy and Theology 7 (1998), 69–87.

212

Magali Roques

Trifogli, Cecilia. “Scotus and the Medieval Debate about the Continuum,” Medioevo 29 (2004), 233–66. Walter Chatton. Reportatio II, ed. J. Wey and J. Etzkorn (Toronto: Pontifical Institute of Mediaeval Studies, 2004). William of Ockham. Opera omnia, edited under the direction of G. Gàl (New York: Franciscan Institute, 1967–88).

Yah.yaˉ Ibn ʿAdī on a Kalaˉ m Argument for Creation Peter Adamson and Robert Wisnovsky Of all the figures associated with the group of Aristotelians that has come to be called the “Baghdad school,” by far the most famous today is al-Faˉ raˉ bī (d.950/951). Yet in the middle of the tenth century it was not al-Faˉ raˉ bī, but Yah.yaˉ ibn ʿAdī (d.974), whose star shone most brightly. The historian alMasʿūdī remarked that Ibn ʿAdī was the only man to whom one could turn for instruction in philosophy,1 and Ibn al-Nadīm’s Fihrist also singles him out as a leading faylasūf.2 Despite the fact that fundamental work on Ibn ʿAdī was already done decades ago,3 it is only in recent years that a fuller picture of his intellectual profile has begun to emerge.4 This process should be aided by the evaluation of newly discovered texts by Ibn ʿAdī contained in a Tehran manuscript.5 The present contribution offers an analysis, 1 Kitaˉ b al-Tanbīh wa-l-is.raˉ f, ed. M. J. de Goeje, 2 vols. (Leiden: Brill, 1893–4), vol. 1, 122. For the passage, see further Dominique Urvoy, “Abū Bakr al-Raˉ zī and Yah.yaˉ ibn ʿAdī,” in P. Adamson (ed.), In the Age of al-Faˉ raˉ bī: Arabic Philosophy in the Fourth/Tenth Century (London: Warburg Institute, 2008), 63–70. 2 Kitaˉ b al-Fihrist, ed. G. Flügel (Leipzig: V. C. W. Vogel, 1872), 264. 3 Notably Gerhard Endress, The Works of Yah.yaˉ ibn ʿAdī: An Analytical Inventory (Wiesbaden: L. Reichert, 1977) and the edition of numerous works in Sah.baˉ n Khalīfaˉ t, Maqaˉ laˉ t Yah.yaˉ ibn ʿAdī al-falsafiyya (ʿAmmaˉ n: al-Jaˉ miʿa al-Urdunniyya, 1988). 4 Two of the most useful overall assessments have come in the context of discussing the authorship of the Harmony of the Two Philosophers ascribed to al-Faˉ raˉ bī. See Damien Janos, “Al-Faˉ raˉ bī, Creation ex nihilo, and the Cosmological Doctrine of K. al-Jamʿ and Jawaˉ baˉ t,” Journal of the American Oriental Society 129 (2009), 1–17; Marwan Rashed, “On the Authorship of the Treatise On the Harmonization of the Opinions of the Two Sages Attributed to al-Faˉ raˉ bī,” Arabic Sciences and Philosophy 19 (2009), 43–82. 5 Robert Wisnovsky, “New Philosophical Texts of Yah.yaˉ Ibn ʿAdī: A Supplement to Endress’ Analytical Inventory,” in F. Opwis and D. Reisman (eds.), Islamic Philosophy, Science, Culture, and Religion: Studies in Honor of Dimitri Gutas (Leiden: Brill, 2012), 307–26, and “MS Tehran—Madrasa-yi Marwī 19: An 11th/17th-century Codex of Classical Falsafah, Including ‘Lost’ Works by Yah.yaˉ ibn ʿAdī (d. 363/974),” Journal of Islamic Manuscripts 7 (2016), 89–122. (A facsimile edition of the manuscript, including

214

Peter Adamson and Robert Wisnovsky

translation, and edition of one of these texts, which sheds light on a key issue concerning Ibn ʿAdī’s general stance as a thinker: his attitude towards Islamic theology, or kalaˉ m. The previously available evidence on this question is rather ambiguous. One directly relevant passage is found in the Muqaˉ basaˉ t of al-Tawh.īdī, as part of a discussion of the difference between the mutakallimūn and the falaˉ sifa.6 Here he is quoted as mocking the theologians for styling themselves as “masters of speech (arbaˉ b al-kalaˉ m)”: As if other people do not also speak (yatakallamūna)—or are they not people with something to say (ahl al-kalaˉ m)? Perhaps the theologians suppose they are mute and silent? Or is nothing said by the practitioners of law, grammar, medicine, geometry, logic, astrology, physics, divinity, h.adīth, and Sufism?

On the other hand, we know that Ibn ʿAdī was willing to engage seriously, albeit critically, with the theologians. A striking example is his epistle on the kalaˉ m idea that humans “acquire” their actions, so as to be morally responsible for them, even though the actions are created by God.7 One might be tempted to see this as an exception that proves the rule, given that Ibn ʿAdī indices, is now available: Robert Wisnovsky, ed., A Safavid Anthology of Classical Arabic Philosophy. Facsimile Edition of MS Madrasa-yi Marwī 19 [Tehran: Markaz-i Daˉ ʾirat almaʿaˉ rif-i buzurg-i islaˉ mī—Intishaˉ raˉ t-i kitaˉ b-i rayzaˉ n/Montreal: McGill Institute of Islamic Studies, 2017].) For previous studies of treatises found in this manuscript, see Robert Wisnovsky, “Yah.yaˉ ibn ʿAdī’s Discussion of the Prolegomena to the Study of a Philosophical Text”, in M. Cook, N. Haider, I. Rabb, and A. Sayeed (eds.), Law and Tradition in Classical Islamic Thought (Basingstoke: Palgrave Macmillan, 2012), 171–85; Stephen Menn and Robert Wisnovsky, “Yah.yaˉ ibn ʿAdī’s Essay on the Four Scientific Questions Regarding the Three Categories of Existence: Divine, Natural and Logical. Editio princeps and English Translation, with Historical-Philosophical Notes,” Mélanges de l’Institut dominicain d’études orientales du Caire (MIDEO) 29 (2012), 73–96; Peter Adamson and Robert Wisnovsky, “Yah.yaˉ Ibn ʿAdī on the Location of God,” Oxford Studies in Medieval Philosophy 1 (2013), 205–28; David Bennett and Robert Wisnovsky, “A Newly Discovered Yah.yaˉ ibn ʿAdī Treatise Against Atomism,” in D. Janos (ed.), Ideas in Motion in Baghdad and Beyond (Leiden: Brill, 2015), 298–311. The latter volume contains several useful studies on Ibn ʿAdī. 6 Section 48, which is at al-Muqaˉ basaˉ t li-Abī H . ayyaˉ n al-Tawh.īdī, ed. M. T. H . usayn (Baghdad: Jamīʿat Baghdad, 1970), 204–5, or in the more recent edition by H . . al-Sandūbī (Kuwait: Daˉ r Suʿaˉ d al-S.abbaˉ h., 1992), 223–4. We owe the reference to Endress, The Works of Yah.yaˉ ibn ʿAdī, 6, where Endress also mentions a report of al-Qift.ī to the effect that Ibn ʿAdī refused to debate publicly with theologians because they would be unable to understand one another. For a complete translation of the passage, see Gerhard Endress, “Theology as a Rational Science: Aristotelian Philosophy, the Christian Trinity and Islamic Monotheism in the Thought of Yah.yaˉ ibn ʿAdī,” in Janos (ed.), Ideas in Motion, 221–52, particularly 225–6. 7 Shlomo Pines and Michael Schwarz, “Yah.yaˉ Ibn ʿAdī’s Refutation of the Doctrine of Acquisition (Iktisaˉ b),” in J. Blau, S. Pines, M. J. Kister, and S. Shaked (eds), Studia Orientalia Memoriae D. H. Baneth Dedicata (Jerusalem: Magnes, 1979), 49–94.

Yah.yaˉ Ibn ʿAdī on a Kalaˉ m Argument for Creation

215

prefaces his remarks by saying that he takes up the issue only out of deference to his addressee, and despite the fact that there would be other topics “more suitable (alyaq)” for him to investigate. Indeed a study of this epistle has commented that it is, “as far as theology is concerned, a curious exception, for it contains his considered opinion on a point of Islamic kalaˉ m, a system of thought to which, as he clearly indicates, he felt no personal commitment.”8 Yet the short treatise to be discussed in the present contribution provides us with another case where Ibn ʿAdī’s wields the weapons of philosophy against the mutakallimūn. At stake is the so-called “proof from accidents,” which Herbert Davidson has called “the kalaˉ m demonstration par excellence of creation.”9 Perhaps because of its fame, or simply because of the truncated nature of the discussion (which fits on one and a half folios), Ibn ʿAdī does not explicitly lay out the argument he is attacking. But it is evidently the proof from accidents that is in his sights, especially in light of a section which critiques one of the argument’s premises (§28),10 namely that “accidents are created.” The would-be proof conjoins this premise to the claim that bodies cannot exist without accidents, and infers that bodies are created. Hence the title that appears in the manuscript: “treatise in dialectical refutation of one who says that bodies are created.” In particular, it is the body of the universe that Ibn ʿAdī’s opponent would have in mind (this is the way the argument was historically used in kalaˉ m texts, and implicitly recognized by Ibn ʿAdī: see §36). The argument is meant to show, then, that the universe is created in the sense of coming into existence after not having existed, just as does an accident such as motion or rest. The earliest mutakallim known to have presented such an argument is Abū l-Hudhayl (d.849). Mankdīm reports that he argued as follows: “Bodies do not lack (lam tanfakk) originated features (h.awaˉ dith), nor do they precede them. But what cannot be without something originated (muh.dath), preceding it, must also be originated.”11 Notice the emphasis Pines and Schwarz, “Yah.yaˉ Ibn ʿAdī’s Refutation,” 49–50. Herbert A. Davidson, Proofs for Eternity, Creation and the Existence of God in Medieval Islamic and Jewish Philosophy (New York: Oxford University Press, 1987), 134. He provides a wealth of references to texts that use the argument, some of which we discuss in what follows. 10 We refer to our translation and edition by section number. 11 Sharh. al-us.ūl al-khamsa, ed. ʿA. ʿUthmaˉ n (Cairo: Maktabat Wahba, 1965), 95. Mankdīm then adds, “This proof is built on the following four suppositions. First, that there are features (maʿaˉ nī) in bodies, namely composition and separation (al-ijtimaˉ ʿ wa-liftiraˉ q); second, that these features are originated; third, that body can neither lack them nor proceed them; fourth, given that [bodies] cannot lack [these features] or precede them, it follows that they are likewise originated.” Cf. Josef van Ess, Theologie und Gesellschaft im 2. und 3. Jahrhundert Hidschra: eine Geschichte des religiösen Gesellschaft im Denkens im frühen Islam, 6 vols. (Berlin: Walter de Gruyter, 1991–5), vol. 5, §XXI.42. 8 9

216

Peter Adamson and Robert Wisnovsky

here on the idea that a body cannot precede its accidents. The idea on which Abū l-Hudhayl relies is that any body must necessarily have one of two contrary accidents; let us focus on the case of motion or rest, since this is the pairing discussed by Ibn ʿAdī. Abū l-Hudhayl assumes that such an accident must be “originated (muh.dath),” which here needs to mean existing after not existing. This premise has a certain amount of plausibility on any conception of motion and rest—motions, especially, seem like fleeting phenomena that come to be and pass away. But it would go through even more easily if we imagine that motion or rest is a momentary affair. If an accident of motion belongs to a body at an instant, then there is no question of any motion having existed eternally. But of course, on that construal, the premise would be unacceptable from an Aristotelian perspective, as Ibn ʿAdī will not be slow to point out. A similar argument can be found in the Jewish thinker Saadia Gaon (d.942), who writes in his Book of Doctrines and Beliefs: “I found that bodies are never free of accidents (aʿraˉ d.) that befall every single one them, either from themselves or from something else (min dhaˉ tihī aw min ghayr dhaˉ tihī).”12 He provides examples: animals grow and mature, the earth itself is covered with changing plants and animals, and even the celestial bodies have the accidents of motion and various sorts of luminosity. “So,” he adds, “having found that originated features (h.awaˉ dith) are applicable in their case [sc. that of the celestial bodies], and that [the celestial bodies] do not precede them (tasbiquhaˉ ), I was sure that whatever does not precede the originated is like it, because it falls under its definition.” As in Abū l-Hudhayl, the focus is on the impossibility of a temporal priority of body to its accidents. If body cannot exist without an accident, and if every accident comes into existence after not existing, then so too does the body come into existence after not existing. The further history of the argument is a long one. It is invoked by numerous theologians of both the Muʿtazilite and Ashʿarite traditions, for instance ʿAbd al-Jabbaˉ r and al-Baˉ qillaˉ nī, but criticized by philosophers including Averroes.13 More immediately relevant for Ibn ʿAdī are his fellow Baghdad Peripatetics al-Faˉ raˉ bī and Ibn Suwaˉ r.14 Al-Faˉ raˉ bī gives a version Kitaˉ b al-Amaˉ naˉ t wa-l-i ʿtiqaˉ daˉ t, ed. S. Landauer (Leiden: Brill, 1880), 35. For details see Davidson, Proofs, 136–41. 14 Two more philosophers who apparently engaged with the proof are al-Kindī and his associate al-Sarakhsī. We know from medieval bio-bibliographers that both wrote works on the question whether created things must be either moving or at rest at the moment of their creation. For al-Kindī see Peter Adamson and Peter E. Pormann, The Philosophical Works of al-Kindī (Karachi: Oxford University Press, 2012), lvi, and for al-Sarakhsī, see Ulrich Rudolph (ed.), Grundriss der Geschichte der Philosophie: Philosophie in der islamischen Welt, Band 1: 8.–10. Jahrhundert (Basel: Schwabe, 2012), 152. 12 13

Yah.yaˉ Ibn ʿAdī on a Kalaˉ m Argument for Creation

217

of the proof from accidents as an example argument in his Epitome of the Prior Analytics. But closer is the version given by Ibn ʿAdī’s student Ibn Suwaˉ r (d. after 1017), also known as Ibn al-Khammaˉ r.15 He presents the proof from accidents in the following terms: To show the flaws in an incorrect argument for a given conclusion (mat. lūb) proves neither the falsity nor truth of that conclusion. So if it is shown that the argument by which the mutakallimūn seek to show the origination of bodies is incorrect, it does not follow from this that bodies are eternal—or that they are originated. The argument they use to show this goes as follows. They say that body does not lack originated things, nor does it precede them (laˉ yanfakku min al-h.awaˉ dith wa-laˉ yataqaddamuhaˉ ). But whatever does not lack originated things and does not precede them is itself originated. Therefore body is originated.16

As Ibn Suwaˉ r is at pains to stress, the fact that he will go on to reject the kalaˉ m proof of creation does not mean that he denies the proof ’s conclusion. To the contrary, his point is to show the superiority of Philoponus’ argument for creation—which demonstrates the non-eternity of the body of the universe from its finiteness—over that offered by the mutakallimūn.17 Nonetheless, Ibn Suwaˉ r spends most of the treatise complaining about the kalaˉ m proof rather than exploring the better demonstration found in Philoponus. He understands the claim that “body does not lack accidents” (note that he now switches to aʿrˉad. rather than h.awaˉ dith), as indicating that motion and rest succeed one another in body. But the Aristotelian will not admit this, since he holds that the heavens have always been in motion, and the earth always at rest.18 Furthermore, the proof seems to suppose a symmetrical relation of dependence between accidents and body. To the contrary, “bodies are in no need of accidents to exist; and if this is so, then the fact that accidents are originated does not show that bodies are originated.”19 The body would only need accidents if they were constitutive of the essence

15 ʿAbdurrah.maˉ n Badawī, Aflat. uniyya al-muh.datha ʿinda l-ʿarab (Kuwait: Wakaˉ lat al-Mat. būʿaˉ t, 1977), 243–7. For a translation and discussion of the treatise see Bernhard Lewin, “La Notion de muh.dat dans le kalaˉ m et dans la philosophie,” Orientalia Suecana 3 (1954), 88–93. 16 Badawī, Aflat. uniyya, 243. 17 On the Arabic reception of Philoponus’ arguments against eternity, see e.g., Herbert A. Davidson, “John Philoponus as a Source of Medieval Islamic and Jewish Proofs of Creation,” Journal of the American Oriental Society 89 (1969), 357–91; Shlomo Pines, “An Arabic Summary of a Lost Work of John Philoponus,” Israel Oriental Studies 2 (1972), 320–52; reprinted in his Studies in Arabic Versions of Greek Texts and in Medieval Science, vol. 2 (Jerusalem: Magnes, 1986); and Muhsin Mahdi, “Alfarabi against Philoponus,” Journal of Near Eastern Studies, 26 (1967), 233–60. 18 19 Badawī, Aflat. uniyya, 244. Ibid.

218

Peter Adamson and Robert Wisnovsky

of body. Even if bodies did depend on accidents, though, the kalaˉ m proof would still fail: If one were to admit to them that body does not lack accidents, and that each of the accidents that successively come to it [sc. body] were originated, it would not follow that body is originated. For it is possible20 that this successively comes to it forever (daˉ ʾiman) and without interruption, so that it is not originated, even though each one of those accidents taken individually is originated. For it is possible that a given motion occurs in it, and then departs from it, and that rest occurs in it, then departs from it, and then a different motion arrives in it so that rest departs, and so on like this forever without cease, so that this repeats eternally (abadan). Thus [body] will not be without originated accidents, yet will itself be unoriginated.21

As Davidson has pointed out, it was to block this sort of response that later mutakallimūn, especially al-Juwaynī (d.1085), built into the proof a premise that rules out an infinite succession of accidents.22 A final point worth noting from Ibn Suwaˉ r’s treatise is his disambiguation of the claim that body does not “precede” its accidents. In a sense, this is true: you will never find a body that is neither moving nor at rest. But, he points out, the very name ‘accidents’ (aʿrˉad.) indicates that these belong or “occur (ʿarad.a)” to the body that is their subject, so that the body is in a sense prior to them. Hence, “the connection of the accidents to [body] is a proof of its priority (taqaddum) to them, but not of its origination (h.udūth).”23 We might say that body is causally, though not temporally, prior to its accidents. As Ibn Suwaˉ r goes on to point out, this notion of causal dependence can also be expressed using the verb h.adatha. It is legitimate to use the word ‘originated’ (muh.dath) to mean that one thing is caused by another, even if the cause is not temporally prior to the effect. He mentions the traditional example of illumination coming from a light source, and states that in such cases “the causes are not prior to the effects in time, but rather are prior to them in nature, rank, and nobility.”24 In this sense, we can even ascribe to Aristotle the view that the universe is muh.dath, since he holds that it is (eternally) caused by God.25 20 Reading qad yajūzu for laˉ yajūzu, in parallel with the following construction at 245 line 3. 21 Badawī, Aflat. uniyya, 244–5. 22 Davidson, Proofs, 144–6. His observation that this premise was absent in earlier versions is borne out by the fact that neither Ibn ʿAdī nor Ibn Suwaˉ r reckon with it in their refutations. For an anticipation of this see Wilferd Madelung “Abū l-H . us.ayn alBas.rī’s Proof for the Existence of God,” in J. E. Montgomery (ed.), Arabic Theology, Arabic Philosophy (Leuven: Peeters, 2006), 273–80. 23 24 Badawī, Aflat. uniyya, 245. Badawī, Aflat. uniyya, 247. 25 This should be contrasted to the discussion of Aristotle’s stance on creation in the Harmony of the Two Philosophers. There, Aristotle’s claim that “the universe has no temporal beginning” is interpreted as meaning that “it is not generated one part after

Yah.yaˉ Ibn ʿAdī on a Kalaˉ m Argument for Creation

219

Now let us turn to the new text of Ibn ʿAdī. The title given in the manuscript is: Treatise in Refutation of One Who Says that Bodies are Created, Using a Dialectical Method (Maqaˉ la fī l-Radd ʿalaˉ man qaˉ la bi-anna al-ajsaˉ m muh.datha ʿalaˉ t. arīq al-jadal).26 It is unclear whether the last phrase announces that Ibn ʿAdī himself will be using a dialectical method, or that the opponent who is being refuted has offered only a dialectical proof. It may be tempting to adopt the latter reading, given that kalaˉ m is standardly characterized by Aristotelians as merely dialectical. But, while this is grammatically possible, it seems more natural to take ʿalaˉ t. arīq al-jadal with radd than with qaˉ la. And there are indeed signs that Ibn ʿAdī has chosen to argue dialectically in the treatise. At §15, he accepts a premise for the sake of argument (“according to the doctrine of whoever holds this to be possible”). More significantly, the treatise defends a thesis that Ibn ʿAdī does not himself accept. He shows how an Aristotelian eternalist could respond to the kalaˉ m proof and uses standard dialectical tactics in doing so. For instance, he points out a line of response that has not been ruled out (see §37), and shows that the sort of reasoning of the kalaˉ m proof would lead to an absurdity in an analogous case (§§38–43). Yet we know from other another,” and that God did not create it “in time” because time only comes to be with the motion of the heavens. See Abū Nas.r al-Faˉ raˉ bī, L’Harmonie entre les Opinions de Platon et d’Aristote, texte et traduction, ed. and trans. F. W. Najjar and D. Mallet (Damascus: Institut français de Damas, 1999), §55; and in the newer and superior edition, al-Faˉ raˉ bī, L’armonia delle opinioni dei due sapienti il divino Platone et Aristotele, ed. and trans. C. M. Bonadeo (Pisa: PLUS, 2008), 63. (The attribution of this work to al-Faˉ raˉ bī has been seriously challenged; see, for example, Marwan Rashed, “On the Authorship of the Treatise On the Harmonization of the Opinions of the Two Sages Attributed to al-Faˉ raˉ bī,” Arabic Sciences and Philosophy 19 [2009], 43–82.) Unlike the author of the Harmony, Ibn Suwaˉ r does not try to obscure the fact that Aristotle did accept the eternity of the universe, even if he does note that Aristotle believed it to be muh.dath in the sense of “caused.” On the basis of this treatise, Damien Janos even takes Ibn Suwaˉ r to have “believed in the infinity of time and conceived the world as being without a temporal beginning” (Janos, “Al-Faˉ raˉ bī, Creation ex nihilo,” 6). But in fact, Ibn Suwaˉ r seems to be saying that Aristotle—and Proclus, whom he also mentions here—could accept the universe’s “origination” (i.e., causal dependence on God) despite their false belief in its everlastingness, a belief contradicted by the preferable proof against eternity offered by Philoponus. As we will see, this coheres well with the position of his teacher Ibn ʿAdī. 26 Endress, The Works of Ibn ʿAdī, §4.41, working only with bio-bibliographic reports, suggested mujtaliba or mujallaba as possible alternatives for muh.datha, and ʿalaˉ t. arīq albadal as a possible alternative for ʿalaˉ t. arīq al-jadal. Endress’s suggestions are interesting, especially in the case of badal. The rationale here is that the opponent’s proof is one that works by invoking successive replacement (Greek: antiperistasis, cf. Aristotle, Physics VIII.10, 267a16–17). This is not wrong, but it does not really identify the core point of the argument: that accidents are created so bodies must be also. What is more, the word badal does not actually appear anywhere in our text, and ʿalaˉ t. ariq seems to go better with a remark about methodology. In any case, the title of the treatise, as recorded in the manuscript, is clear in reading muh.datha and jadal.

220

Peter Adamson and Robert Wisnovsky

works that Ibn ʿAdī expressly rejected the eternity of the universe.27 Of course, the title may in any case be a later addition. But if the title is Ibn ʿAdī’s, then its mention of dialectical method is most likely a caution offered in the same spirit as Ibn Suwaˉ r’s caveat at the opening of his own treatise. Both refute the kalaˉ m proof for the creation of the world, but neither would reject the conclusion reached by that argument. Ibn ʿAdī’s refutation may be divided into three main sections. In the first section (§§1–11), Ibn ʿAdī disambiguates two possible ways to understand the claim that “body does not lack motion and rest” (the phrasing is close to Abū Hudhayl’s and Ibn Suwaˉ r’s, using the verb yanfakku). On the one hand (§§2–9), it could mean that body cannot lack both motion and rest at the same time, which is true but leads the proponent of the kalaˉ m proof into the mistaken inference that body is not prior to its accidents. On the other hand (§§10–11), it could mean that it must have both motion and rest simultaneously, but this is impossible. In the second section (§§12–26), Ibn ʿAdī returns to the question of whether a body is prior to its accidents. Because he has not explicitly laid out the kalaˉ m proof like Ibn Suwaˉ r will do, the point of this might not be immediately obvious. But we have seen enough versions of the argument to realize that he is anticipating Ibn Suwaˉ r’s strategy of denying a premise used in the kalaˉ m proof, namely that “body does not precede its accidents.” Like Ibn Suwaˉ r, Ibn ʿAdi wants to draw our attention to the fact that body is indeed prior to accidents, though this priority is not in respect of time but in respect of “nature,” as Ibn ʿAdī puts it (§18). A particular puzzle that arises here (§§19–26) is the case of the heavenly bodies and the earth, since these respectively always move and are always at rest. Again, this point will also appear in Ibn Suwaˉ r’s treatise, albeit in the different context of outlining a possible Aristotelian response to the kalaˉ m proof. Ibn ʿAdī instead focuses on showing that even in these cases where an accident is permanently possessed by a body, the body is prior to the accident. In the third section (§§27–43), he poses two objections to the proof from accidents, first (§§28–37) by attacking the premise that all accidents are created. This, argues Ibn ʿAdī, is a contentious and unproven premise, whose plausibility depends on a part–whole fallacy. Finally (§§38–43), Ibn ʿAdī offers a parallel argument that uses the reasoning of the kalaˉ m proof to generate an absurd contradiction, namely that God is created. One might compare Gaunilo’s famous objection to Anselm’s ontological argument for the existence of God, namely that one could adapt Anselm’s proof

27 See the references in Janos, “Al-Faˉ raˉ bī, Creation ex nihilo,” 7–8, which concern especially Ibn ʿAdī’s theological works.

Yah.yaˉ Ibn ʿAdī on a Kalaˉ m Argument for Creation

221

to establish the existence of a perfect island. The absurd consequence shows that something must be amiss with Anselm’s reasoning, without identifying where exactly the reasoning goes awry. Ibn ʿAdī is aware of this. He notes at §38 that with the absurd parallel case, he is refuting the argument “as a whole” rather than undermining one of its premises, or “parts” as at §§28–37. Now let us look in more detail at each section. Returning to the beginning of the treatise, we find Ibn ʿAdī first discussing the construal of the claim that “body does not lack motion and rest.”28 It is pretty obvious that the opponent does in fact mean that body must always have one or the other of these, but not both at the same time (as the second construal would have it). While the first construal is in itself unobjectionable, Ibn ʿAdī cautions us against the conclusion drawn by the opponent, namely one we have seen in every version of the kalaˉ m proof: that “body does not precede its accidents.” The opponent is thinking that: (a)

Body must be either moving or at rest.

implies (b)

Body is not prior to motion and rest.

But, Ibn ʿAdī argues in §5, if this is a valid inference then the negation of (a) should also imply the negation of (b). The logical rule applied here is that if x implies y, then not-x implies not-y. This is illustrated with parallel cases at §6; for instance “if health implies balance,” i.e., of the bodily humors, then “illness implies imbalance.” So in the case we are interested in, if (a) implied (b) then: not-(a):

Body cannot be both moving and at rest

would imply not-(b):

Body is prior to motion and rest.

Why should this be problematic for the opponent? Presumably because not(a) is just as true as (a). Thus we can use the opponent’s line of reasoning to infer opposed conclusions, namely both (b) and not-(b), from true premises (as stated at §7). Clever though this is, Ibn ʿAdī’s argument is rather dubious. At §5, he formulates the contrary of (a) incorrectly as “the impossibility of its not lacking both together,” which is what we have tried to capture above in more straightforward language as not-(a). The contrary of

28 Notice that he also mentions the pair “separation and combination (al-iftiraˉ q wa-lijtimaˉ ʿ)” (§7), which appears along with motion and rest in Mankdīm’s report of the proof in Abū l-Hudhayl.

222

Peter Adamson and Robert Wisnovsky

(a) should rather be something like “body need not be either moving or at rest,” which unlike his version of not-(a) is of course false. The second section of the treatise argues more positively that body is indeed prior to its accidents. Ibn ʿAdī argues for this in two ways, and then considers the problematic case of things that do not alternate between motion and rest. His first argument (§§12–14) is that body can exist when either of the two contrary accidents does not exist. That is, when a body is moving, rest is non-existent for it, and when a body is at rest, motion is non-existent for it. Thus body as such must be prior to both motion and rest. His second argument (§§15–18) follows almost the same line of thought, but with the added point that the same body can at first be moving and then at rest, or vice versa. This shows that the body is prior to both contraries, given that at different times it exists without one and then without the other. On the assumption that body counts in this context as the substance that underlies the accidents of motion and rest,29 Ibn ʿAdī’s argument is effectively just an application of points made in Aristotle’s discussion of substances in the Categories. Aristotle explains that substances are primary because they “underlie all other things (hupokeisthai panta ta alla),” including that which is “in them (en autais),” meaning the accidents that belong to the substances (Categories 5, 2b15–17, cf. 3a1–2); and also that it is distinctively characteristic of substances that they are receptive of contraries (tôn enantiôn dektikon, 4a11). There is one other point from this section worth dwelling on: Ibn ʿAdī remarks that, when body is for instance moving and not at rest, it still has the possibility (imkaˉ n) to be at rest (§13, cf. §§25–6). In light of this, it actually does not matter whether a given body is really going to swap motion for rest at some future time, or vice versa. The mere fact that a currently moving body could survive despite coming to rest shows that the body is ontologically independent of its motion. The reverse is not the case: the body’s motion cannot survive without the body. This gives us a rationale for saying that the body is prior to its currently possessed contrary, with the priority in question being in respect of ‘nature’ and not time. Thinking about it in this way sharpens the difficulty that emerges at §19: what about 29 This assumption is less uncontroversial than it may seem, and in fact, Ibn ʿAdī was called upon to adjudicate a debate about it between his brother Ibraˉ hīm and an unnamed opponent (Endress, The Works of Ibn ʿAdī, §3.33). The original dispute, Yah.yaˉ ’s adjudication, Ibraˉ hīm’s response, and a final word by Yah.yaˉ are recorded in another “lost” treatise found in the Tehran manuscript (Wisnovsky, “New Philosophical Texts,” 313–14). For an edition, translation, and study, see Stephen Menn and Robert Wisnovsky, “Yah.yaˉ ibn ʿAdī and Ibraˉ hīm ibn ʿAdī on whether Body is a Substance or a Quantity: Text, Translation and Commentary,” Arabic Sciences and Philosophy 27/1 (2017), 1–74.

Yah.yaˉ Ibn ʿAdī on a Kalaˉ m Argument for Creation

223

bodies that do not have the possibility to change from motion to rest, or vice versa? As Ibn ʿAdī puts it, for the heavens and the earth as a whole “only one of these [contraries] exists,” since the heavens never cease moving and the earth is always at rest. And we can add that the heavens and earth do not even have the possibility to change in respect of motion and rest (see §§22–3). What Ibn ʿAdī has to say by way of solution is frustratingly opaque: “in the case of these bodies, there is no distinction between the essences (dhawaˉ t) of their motions (or of their rest) and the motions of these (or their rest). It follows that the essences of these bodies are also prior to the essences of their motions and rest” (§20). Probably his point is that heavenly motion, qua motion, shares its essential features with other cases of motion (and likewise for the rest of the earth). Since other motions are essentially dependent on the bodies in which they occur, so will the motion of the heavens depend on the body of the heavens.30 This interpretation is borne out to some extent by what Ibn ʿAdī goes on to say at §§21–6: Aristotle distinguished between that which is always moving, that which is always at rest, and that which is at times moving and at times at rest. The third class provides strong evidence (adallu dalīlin) that the essence of body is in itself independent of the essence of motion and rest. If so, the eternity and even necessity of the accident of motion or rest in a given body will not impugn that body’s essential priority to that accident. A further confirmation for our reading is the fact that Ibn Suwaˉ r makes a similar point in his treatise, by contrasting the accidents of motion and rest to the essence of the body. “Only if the accidents constituted the essences of body” would their creation imply the createdness of the body, whereas “if the accidents do not enter into the existence of body, then the accidents’ being a certain way (bi-h.ˉal) will not imply that the body is that way.”31 If our interpretation is indeed correct, then there is a certain irony in what happens next. Ibn ʿAdī moves to attacking a tacit inference that underlies the kalaˉ m proof, namely that if some accidents of motion and rest are created (again, in the sense of existing after not existing), then all accidents of motion and rest are created. This is not obvious, argues Ibn ʿAdī, since there could be exceptions to the rule. In particular, even if most motions are created this may not be so in the case of celestial motion (§36). Here we have yet another resonance between Ibn ʿAdī and Ibn Suwaˉ r. As we saw, the 30 In another treatise preserved in the Tehran manuscript, devoted to the question of whether the heat in fire is a substance (see Wisnovsky, “New Philosophical Texts,” 313), Ibn ʿAdī adopts an analogous line of reasoning. He claims that if heat is ever an accident—as it is in iron, for example—then it is always an accident. Our thanks to Fedor Benevich for pointing out the similarity of the two arguments. 31 Badawī, Aflat. uniyya, 244.

224

Peter Adamson and Robert Wisnovsky

latter also proposed a response on behalf of the Aristotelian to the effect that the heavens’ motion and the earth’s rest are both permanent. Ibn ʿAdī’s explanation of the mistake made by the kalaˉ m proof reads like a philosophy professor teaching students about the part–whole fallacy. He gives vivid examples: as every philologist knows, scribes occasionally skip a word when copying manuscripts, but one cannot generalize to imagine a scribe skipping the entire book by mistake. The problem, and the irony, is that we took Ibn ʿAdī to be making an inference of the same sort at §20. For he seemed there to move from the dependence of some accidents on body to the dependence of all accidents on body. In fact, it would even be the same case of heavenly motion that threatens to be a counterexample. Having said that, Ibn ʿAdī did also seem to allude to the idea that accidents are essentially or by their very nature dependent on their subjects. This would license the generalization. To reverse Ibn ʿAdī’s example, consider that if a scribe essentially and not just normally copied the words in front of him without mistake, his entire output will be error-free. But of course, Ibn ʿAdī’s stipulation that motion and rest are essentially dependent on body threatens to be question-begging. In the final section (§38–43), Ibn ʿAdī offers the aforementioned “equivalent example (mathal )” or “case for refutation (muʿaˉ rad.a).” Suppose that God has created a human called Zayd. Clearly, God cannot exist without either creating Zayd or not creating him. Therefore, God’s relation to creating Zayd and not creating Zayd is just like body’s relation to the contrary pair of motion and rest. If the opponent were right to say that body has no priority to its accidental contraries, then neither would God be prior to creating Zayd and not-creating Zayd (§41). Worse still, if the successive nature of motion and rest really implied the createdness of body, then the successive nature of God’s act of creation (i.e., the fact that He creates Zayd after not creating him) will imply that God is created (§§42–3)! This point provides an emphatic ending to the treatise. To avoid the absurd consequence, the proponent of the kalaˉ m proof would need to show that there is a difference between the God case and the body case. Several possible disanalogies do suggest themselves. For instance, the act of creating Zayd is not something that some other cause creates in God, the way that motion or rest is created in body. Furthermore, it is far from obvious that God’s not creating Zayd should be treated as if it were an accident on a par with motion or rest— or, for that matter, on a par with creating Zayd, as if not creating Zayd and creating Zayd were two contrary attributes. Though we have identified several apparent shortcomings in Ibn ʿAdī’s argumentation, it should be borne in mind that the treatise is not only very compressed but also labeled as a dialectical work. Perhaps, then, we should not have expected watertight demonstrations. Instead, Ibn ʿAdī may intentionally be offering his kalaˉ m opponents the only sort of response they

Yah.yaˉ Ibn ʿAdī on a Kalaˉ m Argument for Creation

225

deserve, namely the sort of merely persuasive proofs they themselves bring forth. But would these proofs persuade a mutakallim, and for that matter, are they even intended to do so?32 On the one hand, Ibn ʿAdī does at least allude to ideas that have their home in the Aristotelian system. Would his theological opponents be moved by an argument that invokes the priority of bodily essences to accidents (§20)? Probably not. But one could easily imagine a mutakallim producing an argument like the one we find in the third section: that God must either create Zayd or not create Zayd. Indeed, the example presupposes a non-Aristotelian understanding of divine causation, according to which God directly creates individual substances. Here and more generally, it would seem that Ibn ʿAdī is exerting himself to refute the theologian on shared ground, and not just making question-begging appeals to Aristotelian doctrine. Be that as it may, the treatise is a small but not insignificant piece of evidence for reconstructing Ibn ʿAdī’s intellectual project, and that of the Baghdad school as a whole. It is instructive, though hardly surprising, to see that Ibn Suwaˉ r was in fact following his master when he chose to critique the proof from accidents. More significant is what both treatises tell us about the attitudes of the Baghdad school towards contemporary thinkers. Along with other, aforementioned evidence, their repeated efforts to refute the proof from accidents show that the Christians of the Baghdad school were, like al-Faˉ raˉ bī, interested in exposing the shortcomings of contemporary mutakallimūn. Ibn ʿAdī’s treatise on acquisition thus comes to seem like less of an isolated exception, and more part of a larger pattern of rivalry between Aristotelian falsafa and kalaˉ m. If this is the chief historical interest of the treatise, its main philosophical interest lies in Ibn ʿAdī’s insistence that body has a non-temporal sort of priority in relation to its accidents. Ibn Suwaˉ r will use this idea to argue that an Aristotelian can assert God’s priority to an eternal universe. Ayman Shihadeh has recently given us an illuminating discussion of an important point of contrast between Avicenna and classical kalaˉ m.33 For some mutakallimūn (this applies both to Ashʿarites and also Muʿtazilites, to whom Avicenna especially responds), a thing’s being “created” consists solely in its being brought to be after not being. On this reckoning, God would not necessarily be needed to sustain a thing’s further existence once it has been

32 Our thanks to the anonymous referee of the journal for pressing us to consider this question. 33 Ayman Shihadeh, Doubts on Avicenna: A Study and Edition of Sharaf al-Dīn alMasʿūdī’s Commentary on the Ishaˉ raˉ t (Leiden: Brill, 2016), 88. It should, however, be noted that other mutakallimūn, including some Muʿtazilites, do assert the need for God to “sustain” things in existence.

226

Peter Adamson and Robert Wisnovsky

created. For Avicenna, by contrast, God is precisely a sustaining cause of the things He creates, which depend on Him for their existence at every moment they exist. This is why he saw no incompatibility between the eternity of the universe and its createdness. Our two Christians of the Baghdad school anticipate Avicenna’s stance by distinguishing natural or causal priority from temporal priority—and this despite their opposition to the eternity of the universe. It seems likely that on their view, the universe is “created” in both the temporal sense and the sense of ontological dependence.34 This allows them to defend the ‘philosophical’ position, even if only dialectically. Ibn ʿAdī and Ibn Suwaˉ r may have parted ways from orthodox Aristotelianism over the question of eternity, but they were not about to let Aristotle be refuted by the likes of Abū l-Hudhayl.35 LMU Munich and McGill University

BIBLIOGRAPHY Adamson, Peter and Peter E. Pormann. The Philosophical Works of al-Kindī (Karachi: Oxford University Press, 2012). Adamson, Peter and Robert Wisnovsky. “Yah.yaˉ Ibn ʿAdī on the Location of God,” Oxford Studies in Medieval Philosophy 1 (2013), 205–28. Badawī, ʿAbdurrah.maˉ n (ed.). Aflat. uniyya al-muh.datha ʿinda l-ʿarab (Kuwait: Wakaˉ lat al-Mat. būʿaˉ t, 1977). Bennett, David and Robert Wisnovsky. “A Newly Discovered Yah.yaˉ ibn ʿAdī Treatise Against Atomism,” in D. Janos (ed.), Ideas in Motion in Baghdad and Beyond (Leiden: Brill, 2015), 298–311. Davidson, Herbert A. “John Philoponus as a Source of Medieval Islamic and Jewish Proofs of Creation,” Journal of the American Oriental Society 89 (1969), 357–91. Davidson, Herbert A. Proofs for Eternity, Creation and the Existence of God in Medieval Islamic and Jewish Philosophy (New York: Oxford University Press, 1987).

34 The nature and ramifications of the distinction between h.udūth zamaˉ nī and h.udūth dhaˉ tī would become a central point of debate between post-Avicennian mutakallimūn and muh.aqqiqūn such as Fakhr al-Dīn al-Raˉ zī and Nas.īr al-Dīn al-T. ūsī. 35 The authors would like to thank the two OSMP referees for their helpful comments on and corrections to an earlier version of this article. We are also grateful for support of this research from the Leverhulme Trust, and the input of David Bennett and Rotraud Hansberger, who worked together with Peter Adamson on a project funded by the Trust. Peter Adamson is also grateful for useful discussion of the argument with members of his group at the LMU in Munich: Hanif Amin, Fedor Benevich, Matteo Di Giovanni, Andreas Lammer, and Bethany Somma. Rob Wisnovsky would also like to thank Naser Dumairieh for his help in preparing the initial transcription of the Arabic text.

Yah.yaˉ Ibn ʿAdī on a Kalaˉ m Argument for Creation

227

de Goeje, Michael Jan (ed.). Al-Masʿūdī: Kitaˉ b al-Tanbīh wa-l-is.raˉ f, 2 vols. (Leiden: Brill, 1893–4). Endress, Gerhard. The Works of Yah.yaˉ ibn ʿAdī: An Analytical Inventory (Wiesbaden: L. Reichert, 1977). Endress, Gerhard. “Theology as a Rational Science: Aristotelian Philosophy, the Christian Trinity and Islamic Monotheism in the Thought of Yah.yaˉ ibn ʿAdī,” in D. Janos (ed.), Ideas in Motion in Baghdad and Beyond (Leiden: Brill, 2016), 221–52. Flügel, Gustav (ed.). Kitaˉ b al-Fihrist (Leipzig: V. C. W. Vogel, 1872). H . usayn, Muh.ammad Tawfīq (ed.). Al-Muqaˉ basaˉ t li-Abī H . ayyaˉ n al-Tawh.īdī (Baghdad: Jamīʿat Baghdad, 1970). Janos, Damien. “Al-Faˉ raˉ bī, Creation ex nihilo, and the Cosmological Doctrine of K. al-Jamʿ and Jawaˉ baˉ t,” Journal of the American Oriental Society 129 (2009), 1–17. Khalīfaˉ t, Sah.baˉ n (ed.). Maqaˉ laˉ t Yah.yaˉ ibn ʿAdī al-falsafiyya (ʿAmmaˉ n: al-Jaˉ miʿa alUrdunniyya, 1988). Landauer, Samuel (ed.). Saadia Gaon: Kitaˉ b al-Amaˉ naˉ t wa-l-iʿtiqaˉ daˉ t (Leiden: Brill, 1880). Lewin, Bernhard. “La Notion de muh.dat dans le kalaˉ m et dans la philosophie,” Orientalia Suecana 3 (1954), 88–93. Madelung, Wilferd. “Abū l-H. us.ayn al-Bas.rī’s Proof for the Existence of God,” in J. E. Montgomery (ed.), Arabic Theology, Arabic Philosophy (Leuven: Peeters, 2006), 273–80. Mahdi, Muhsin. “Alfarabi against Philoponus,” Journal of Near Eastern Studies 26 (1967), 233–60. Martini Bonadeo, Cecilia (ed. and trans.). Al-Faˉ raˉ bī: L’armonia delle opinioni dei due sapienti il divino Platone et Aristotele (Pisa: PLUS, 2008). Menn, Stephen and Robert Wisnovsky. “Yah.yaˉ ibn ʿAdī’s Essay on the Four Scientific Questions regarding the Three Categories of Existence: Divine, Natural and Logical. Editio princeps and English Translation, with Historical-Philosophical Notes,” Mélanges de l’Institut dominicain d’études orientales du Caire (MIDEO) 29 (2012), 73–96. Menn, Stephen and Robert Wisnovsky. “Yah.yaˉ ibn ʿAdī and Ibraˉ hīm ibn ʿAdī on whether Body is a Substance or a Quantity: Text, Translation and Commentary,” Arabic Sciences and Philosophy 27/1 (2017), 1–74. Najjar, Fawzi W. and Dominique Mallet (eds. and trans.). Abū Nas.r al-Faˉ raˉ bī: L’Harmonie entre les Opinions de Platon et d’Aristote (Damascus: Institut français de Damas, 1999). Pines, Shlomo. “An Arabic Summary of a Lost Work of John Philoponus,” Israel Oriental Studies 2 (1972), 320–52; reprinted in his Studies in Arabic Versions of Greek Texts and in Medieval Science, vol. 2 (Jerusalem: Magnes, 1986). Pines, Shlomo and Michael Schwarz. “Yah.yaˉ Ibn ʿAdī’s Refutation of the Doctrine of Acquisition (Iktisaˉ b),” in J. Blau, S. Pines, M. J. Kister, and S. Shaked (eds.), Studia Orientalia Memoriae D. H. Baneth Dedicata (Jerusalem: Magnes, 1979), 49–94. Rashed, Marwan. “On the Authorship of the Treatise On the Harmonization of the Opinions of the Two Sages Attributed to al-Faˉ raˉ bī,” Arabic Sciences and Philosophy 19 (2009), 43–82.

228

Peter Adamson and Robert Wisnovsky

Rudolph, Ulrich (ed.). Grundriss der Geschichte der Philosophie: Philosophie in der islamischen Welt, Band 1: 8.–10. Jahrhundert (Basel: Schwabe, 2012). al-Sandūbī, H. asan (ed.). al-Muqaˉ basaˉ t li-Abī H . ayyaˉ n al-Tawh.īdī (Kuwait: Daˉ r Suʿaˉ d as.-S.abbaˉ h., 1992). Shihadeh, Ayman. Doubts on Avicenna: A Study and Edition of Sharaf al-Dīn alMasʿūdī’s Commentary on the Ishaˉ raˉ t (Leiden: Brill, 2016). Urvoy, Dominique. “Abū Bakr al-Raˉ zī and Yah.yaˉ ibn ʿAdī,” in P. Adamson (ed.), In the Age of al-Faˉ raˉ bī: Arabic Philosophy in the Fourth/Tenth Century (London: Warburg Institute, 2008), 63–70. ʿUthmaˉn, ʿAbd al-Karīm (ed.). Sharh. al-us.ūl al-khamsa (Cairo: Maktabat Wahba, 1965). van Ess, Josef. Theologie und Gesellschaft im 2. und 3. Jahrhundert Hidschra: eine Geschichte des religiösen Denkens im frühen Islam, 6 vols. (Berlin: Walter de Gruyter, 1991–5). Wisnovsky, Robert. “New Philosophical Texts of Yah.yaˉ Ibn ʿAdī: A Supplement to Endress’ Analytical Inventory,” in F. Opwis and D. C. Reisman (eds.), Islamic Philosophy, Science, Culture, and Religion: Studies in Honor of Dimitri Gutas (Leiden: Brill, 2012), 307–26. Wisnovsky, Robert. “Yah.yaˉ ibn ʿAdī’s Discussion of the Prolegomena to the Study of a Philosophical Text,” in M. Cook, N. Haider, I. Rabb, and A. Sayeed (eds.), Law and Tradition in Classical Islamic Thought (Basingstoke: Palgrave Macmillan, 2012), 171–85. Wisnovsky, Robert. “MS Tehran—Madrasa-yi Marwī 19: An 11th/17th-century Codex of Classical Falsafah, Including ‘Lost’ Works by Yah.yaˉ ibn ʿAdī (d. 363/ 974),” Journal of Islamic Manuscripts 7/1 (2016), 89–122. Wisnovsky, Robert (ed.). A Safavid Anthology of Classical Arabic Philosophy. Facsimile Edition of MS Madrasa-yi Marwī 19 (Tehran: Markaz-i Daˉ ʾirat al-maʿaˉ rif-i buzurg-i islaˉ mī—Intishaˉ raˉ t-i kitaˉ b-i rayzaˉ n/Montreal: McGill Institute of Islamic Studies, 2017).

OUP CORRECTED PROOF – FINAL, 11/5/2017, SPi

Translation Treatise in Dialectical Refutation of One Who Says that Bodies are Created 36 (1) The statement of someone who asserts that body does not lack motion and rest might be understood to mean two things: (2) First, that it [body] does not lack both [motion and rest] together at one and the same moment (fī h.ˉal waˉ h.ida), such that it would exist neither in motion nor at rest. (3) This is true, but need not imply that body must be prior to both [motion and rest]. (4) For just as it is impossible for body to exist37 while lacking both [motion and rest] together, so will it be impossible for it [body] to have38 both [motion and rest] together. (5) If the impossibility of its [body’s] lacking both simultaneously were to necessarily imply its not being prior to both, then the impossibility of its having39 both together would necessarily imply its being prior to both. (6) This is because if one contrary entails some other contrary, then its [i.e., the former’s] contrary will imply the contrary [of the latter].40 For instance, if health implies balance, then illness will imply imbalance; if the dispersion of heterogeneous things (al-mutabaˉ yyinat al-dhawaˉ t) implies heat, then their contraction (jamʿ) will imply cold; and if contraction implies black, then dilation will imply white.41 (7) The combination (ijtimaˉ ʿ) of these two implied conclusions—that is, body’s being prior to motion and rest, or to separation and combination; and [body’s] not being prior to them—is absurd. (8) Furthermore, what this absurdity implies is [also] absurd, because the absurd implies only the absurd—what is implied by this absurdity being the supposition that body not lack [both contraries simultaneously]. (9) For each of the opposed pairs from among these four would imply that body is not prior to them [sc. to an opposed pair] simultaneously, 36

Based on the sole known manuscript, Madrasa-yi Marwī 19, fol. 27b–28a. Literally “the existence of body”: wujūd al-jism. Literally “not to lack,” but the sentence is complicated enough without the double negation. 39 Again, literally “not lacking.” 40 In other words: “This is because if A implies B, then the contrary of A will imply the contrary of B.” 41 Cf. Timaeus, 67DE. 37 38

232

Peter Adamson and Robert Wisnovsky

Yah.yaˉ Ibn ʿAdī on a Kalaˉ m Argument for Creation

233

and this is absurd—I mean that its [body’s] not lacking both together will necessarily imply that [body] not be prior to both simultaneously. (10) The second thing that would be understood of the statement of someone who says that body does not lack motion and rest, is that it [body] does not lack either one of the two; but this is manifestly and evidently false and absurd. (11) [This is] because at the moment of motion, it lacks rest, and at the moment of rest, it doubtless lacks motion—given that it is impossible for either one of these two to exist at the moment of its contrary’s existence in it [the body]. (12) It is something manifestly evident that it [body] is prior to the one of the two that is non-existent while its contrary exists in it [the body], since it [sc. body] exists while that is non-existent. (13) [What] is also manifest, upon initial reflection, is its [body’s] priority to the contrary that does exist in it, since [that existing contrary] is such that it might not exist in it [the body]. This is because when it [body] lacks it [the contrary], it will be evident that [body] itself will not need it in order to exist, given the possibility of its [the body itself ’s] existing devoid of it. (14) This is already sufficient to prove the priority of its [sc. the body’s] essence to the essences of both [sc. motion and rest].42 (15) Something else that proves very manifestly the priority of body itself to motion and rest themselves, is the fact that whichever body is taken [into account], it must, in the first moments of its existence, be either resting, moving, not resting, or not moving, according to the doctrine of whoever holds this to be possible. (16) For it is absurd that it [the body] exist as [both] resting and moving at a single moment. If it [the body] is resting, then it will doubtless be prior to motion, in which case it must be either one of the bodies that are such as to be in motion after being at rest, or one of those that are not such as to be in motion after being at rest. If it is one of the bodies that are in motion after rest, then its priority to rest will be manifest, just as its priority to motion will be manifest.43 (17) In this there is an obvious proof that body itself, for its existence, does not have to be either at rest (since the body itself existed after the state of rest’s separation from it and the motion existed in it) nor in motion (since it [the motion] existed while the body was before it). (18) Given that this is manifest and clear, it is evident that these bodies, which are in motion some times and at rest other times, are independent of motion or rest, and are thus prior to them by Emending the text’s ʿalaˉ taqaddumi dhaˉ tayhimaˉ to ʿalaˉ taqaddumi dhaˉ tihi dhaˉ tayhimaˉ , in accordance with the structure of the following sentence. 43 For the reason already mentioned at §13: it must be (ontologically) prior to rest if it can continue to exist when it is no longer at rest. This is made explicit in the next sentence. 42

234

Peter Adamson and Robert Wisnovsky

Yah.yaˉ Ibn ʿAdī on a Kalaˉ m Argument for Creation

235

nature, since [bodies] exist while [motion and rest] are non-existent, whereas their [motion’s and rest’s] existence is impossible without them [bodies]. (19) Yet there are some bodies that are not such as to be in motion after being at rest: those for which only one of these two things exists, either motion alone, like the celestial sphere, or rest alone, like the earth taken as a whole. (20) In the case of these bodies, there is no distinction between the essences of their motions (or of their rest) and the motions of these (or their rest). It follows that the essences of these bodies are also prior to the essences of their motions and rest. (21) He [Aristotle] furthermore says that bodies are classified by the existence of motion and rest in them, into three types.44 (22) [First,] that which never exists without being in motion, like the celestial sphere; in such a case, it will not be permissible to claim that it has rest, nor will the coming-to-be of its motion be shown by its [the motion’s] absence from it [the body]. (23) [Second,] that which never exists without being at rest, like the earth taken as a whole; in such a case, it will not be permissible to claim it [such a body] has motion, nor will the comingto-be of its rest be shown by its [the rest’s] absence from it [the body]. (24) [Third,] that which is in motion at one moment and at rest in another. [In] this class, its [the body’s] priority to its being in motion and at rest is shown just as the coming-to-be of the [motion and rest] is shown. (25) This is because its absence of motion at the moment of rest necessarily entails that its essence is independent of [motion], while its absence of rest at the moment of motion entails that its essence is independent of [rest]. (26) In this there is a most powerful proof of the possibility that its essence exists devoid of each one of these two, and that in order to exist, their essences require its [the body’s] existence, since neither [motion and rest] will exist except in it [body], and since when it [body] disappears, the two of them will disappear, and when one of these two exists, body will doubtless exist, these two being properties of that which is by nature prior to what is posterior. (27) This [point] undermines one of the bases of this [sc. the opponent’s] proof and shows its unsoundness (fasaˉ d ). Part of what will make its falsehood more evident and its untruth more disgraceful is exposing the mistake and the sophism contained in it [the opponent’s position]. (28) [This mistake and sophism consists in] taking judgments about particulars to imply universal conclusions. This is found in their statement that accidents are created. (29) This is because if this statement is understood as leading to what they think follows from it, and as implying what they want it to, when they say that every accident is created, then this should not be granted to them, there being no demonstration for it from them. 44

Aristotle, Physics VIII.3, 253a24–30, 254a18–19 and 254b4–6.

236

Peter Adamson and Robert Wisnovsky

Yah.yaˉ Ibn ʿAdī on a Kalaˉ m Argument for Creation

237

[ms 28a] (30) This is because they have no proof in view of the fact that some particular motions are created,45 because whatever holds for each of the particulars or parts of a universal or a whole need not necessarily be the case for that universal or whole. (31) In fact, we may find many things that hold of each of the parts or particulars of some whole or some universal, without being the case regarding the whole of [the parts] or the universal of [the particulars].46 (32) [For instance] if we were to take ten-thousand pounds of lead47 (or something else that has weight), and divide it up into [individual] pound-weights, there would be the possibility of one person lifting each of them, whereas this [possibility] would not exist for all of them [at once]. (33) Likewise, if one were to separate out the water of the entire sea cup by cup, there would be the possibility of someone drinking it, for each [cup taken out], yet there is no [such possibility] for the whole. (34) Likewise, whichever part of a book one chooses, it is possible that the book’s scribe omit or miss out [that part] while transcribing it, but it is absurd that this occur for the whole [book]. (35) There are countless similar examples, but what we have mentioned about it is sufficient to prevent an inference from what is the case for particulars and parts to universals and the whole. (36) But if they are prevented from this [making an inference from the parts to the whole], they will have no way of distinguishing between the case in which the motion of the celestial sphere (about whose createdness they disagree) is created, and the case in which Zayd’s motion (or some other particular’s motion) is created. (37) Unless they make [some further] effort, they will not draw as a necessary conclusion the creation of all bodies. (38) Having shown that this argument is unsound in its parts, let us now show that it is unsound as a whole by comparing it to an equivalent example, regarding which there would be no disagreement between people, concerning His eternity—may He be glorified and exalted above that which the heretics say on this topic! (39) Let us take as a case for refutation Zayd, for example, being created in actuality or not being created in actuality, letting this be analogous to motion or rest in the proof, and take the Eternal— blessed be His name!—as analogous to body in [the proof]. (40) Now just as Literally, “the fact that some particulars of the motion are created.” Emending the text’s fa-innaˉ qad najidu ashyaˉ ʾa kathīratan yalzamu kulla waˉ h.idin min ajzaˉ ʾin aw juz ʾiyyi kullin maˉ aw kullin maˉ laˉ yūjadu fī kullihaˉ wa-laˉ kullihaˉ minhaˉ to fainnaˉ qad najidu ashyaˉ ʾa kathīratan yalzamu kulla waˉ h.idin min ajzaˉ ʾin aw juz ʾiyyaˉ ti kullin maˉ aw kulliyyin maˉ laˉ yūjadu fī kullihaˉ wa-laˉ kulliyyihaˉ minhaˉ in order to retain the previous sentence’s parallelism between part/particular and whole/universal. 47 mannan ras.ˉa.san; a mann (pl. amnaˉ n) is equivalent to two rat. l (pl. art.ˉal); both terms can be used to measure weight (in which case a rat. l is roughly equivalent to a half-pound and a mann to a pound) or to measure volume (in which case a rat. l is roughly equivalent to a cup and a mann to a pint). 45 46

238

Peter Adamson and Robert Wisnovsky

Yah.yaˉ Ibn ʿAdī on a Kalaˉ m Argument for Creation

239

body cannot lack both motion and rest together, so likewise the Eternal— great and glorified—cannot lack creating Zayd and not creating Zayd together. (41) If the “not lacking” were proof that body cannot be prior to motion and rest, then in the case for refutation this “not lacking” would be proof that the Eternal is not prior to the creation of Zayd in actuality or [Zayd’s] not being created. (42) So if motion and rest are created due to their succeeding one another in the body, and [due to] the non-existence of each one at the moment of its contrary’s existence, then likewise it is necessary that the creation of Zayd in actuality, and his not being created, are both created due to their succeeding one another, and [due to] the nonexistence of each one at the moment of its contrary’s existence. (43) If this is proof of the creation of bodies, then in the case for refutation, there necessarily follows the creation of the Eternal! This is obviously absurd and an appalling heresy. Therefore what is equivalent to it is clearly false.

Briefly Noted Peter Adamson, Philosophy in the Islamic World: A History of Philosophy without Any Gaps, Volume 3 (Oxford University Press, 2016), 511 pp. Peter Adamson, Philosophy in the Islamic World: A Very Short Introduction (Oxford University Press, 2015), 133 pp. Borges taught us about a library without any gaps, the limitless, total Library of Babel, in which readers squandered their lives and lost their minds. The indefatigable Peter Adamson aims at something more manageable, both in hundreds of podcasts (historyofphilosophy.net) and in a series of books that has now, with volume 3, arrived at Adamson’s own home turf, Arabic philosophy. He prefers to speak of the Islamic world, which allows him to cover as well, in some detail, Jewish and Christians philosophers living within Moslem society. In keeping with his ambitions, this big book has something of everything, extending to Kabbalah and Sufism, and all to way to contemporary Islamic thought. According to OUP marketing, the history-without-gaps series makes for “the most readable and entertaining history of philosophy.” Others might nominate Anthony Kenny’s New History of Western Philosophy (also with OUP), even with its admitted gaps, but there is no doubt that Adamson is a relentlessly charming guide. Even experts will benefit from having this volume on a nearby shelf, in view of its comprehensiveness, and casual readers will benefit both from its jargon-free prose and its useful supplemental features, such as maps and a list of further readings. But Adamson seeks not just to fill the gaps, but also to corner the market on Arabic philosophy with his Very Short Introduction. Here there are, of course, gaps aplenty, but this is, if anything, an even more useful book, because it is just the right length for the casual reader who wants a solid but engaging overview of the field. Adamson’s witty and informal style really comes into its own here, along with a well-honed sense of his audience, as when he remarks near the start that “If you are willing to memorize just one date while reading this book, let it be AD 622.” If you are looking for gifts for students and colleagues next Eid al-Adha, you could hardly do better than this. Master Richard Sophista. Abstractiones, edited by Mary Sirridge and Sten Ebbesen, with E. J. Ashworth (British Academy/Oxford University Press, 2016), 366 pp. The Abstractiones is a mid-thirteenth-century logical treatise that substantially influenced English philosophy into the fourteenth century. The author is certainly English, and is identified in several manuscripts as “Ricardus Sophista,” but no further solid information about his identity has been found. The work can be confidently dated to sometime between 1240 and 1270, in Oxford. It consists of around 300 sophisms, organized around twenty syncategorematic terms. Each sophism is quite brief, leaving little room for theoretical speculation, but the logical principles articulated in the treatise would be cited, decades later, by leading figures such as Scotus and Ockham. It is surmised that the title reflects the work’s origins as a summary collection of earlier treatises in this genre. This edition, the first ever made, has been

242

Briefly Noted

forthcoming for a long time. Work began in 1977, and various scholars lent a hand to the project at various points, until Sirridge and Ebbesen took charge. The edition is done to the highest editorial standards, which are thoroughly described in the extensive Introduction. One might have hoped for an accompanying translation as well, as is becoming customary with the British Academy’s Auctores series, but after waiting nearly forty years for the edition, it was doubtless prudent to bring the enterprise to a close. Blake D. Dutton. Augustine and Academic Skepticism: A Philosophical Study (Cornell University Press, 2016), 273 pp. Augustine’s Contra academicos might lay claim to this dubious distinction: no book in philosophy generates such a gap between the expectations created by its title and the rewards offered within. But even if this youthful dialogue hardly does much to cement Augustine’s reputation as a great philosopher, still it is one of the few places where Augustine focuses at extended length on exclusively philosophical questions. With this monograph, Blake Dutton offers a detailed investigation of the arguments of the Contra academicos. This is history of philosophy in the old-school analytic style, with long sequences of numbered premises and careful consideration of how one or another vulnerable step might be better reformulated. As befits its method, the book is tightly focused on the Contra academicos, with only a modest amount of attention to Augustine’s other work, and even less attention to the work of the skeptics themselves (mostly Cicero). But Dutton does offer as careful an account as one could want of the various arguments that run through this work. And his discussions have particular value because Dutton makes for such a stern critic. One comes away from this study with a precise sense of just what Augustine did and did not achieve. Dragos Calma (editor). Neoplatonism in the Middle Ages: I. New Commentaries on the Liber de causis (ca. 1250–1350). II. New Commentaries on the Liber de causis and Elementatio theologica (ca. 1350–1500) (Brepols, 2016), 562 + 417 pp. This remarkable collection offers a progress report on a long-term initiative to explore the full range of later medieval Latin commentaries on the Neoplatonic tradition. Calma’s research team at the Babeş-Bolyai University in Romania has compiled a list of 74 unpublished Latin commentaries on the Liber de causis, most previously unknown. Thirteen of these are edited here, either in full or in part, along with one of the few known commentaries on the Elements of Theology. The authors are mainly either anonymous or obscure, and run from the later thirteenth into the fifteenth century. As it happens, most of what is edited here comes from Paris, though in fact the works discovered have their origins in universities all over Europe, east and west. The decision to offer samples of a large number of works is surely a wise one, given the fragmentary state of our knowledge about this tradition. This will doubtless facilitate scholarship much more than would complete editions of only a few works. The process of digestion only barely begins with this volume, however, which is largely focused on fundamental historical questions of manuscripts, chronology, and sources. Although the various contributors to this volume do, in their introductions to the texts, occasionally engage with the philosophical ideas, this is very far from being a philosophical study. Still, it is perhaps exactly what the field needs at the present time.

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

OUP CORRECTED PROOF – FINAL, 23/5/2017, SPi

Index ʿAbd al-Jabbār 216 Abū l-Hudhayl 215–16, 226 Abū Hāshim al-Jubbāʾī 225 Adam Wodeham 151–2, 155, 157–9, 162 n39, 163, 167–76, 178, 183–4, 186–8, 190–1, 194–5, 197, 202, 208–9 Albert the Great 87–92, 102–13, 115–19 Anselm 1–39, 220–1 Aristotle 6 n19, 37 n88, 46 n12, 54 n29, 66 n63, 91, 93, 93–4 n18, 102–3, 105–7, 109–10, 117, 117–18 n71, 125–7, 143–4, 168–9, 181–9, 191–7, 200–2, 205, 207–9, 213, 216–20, 222, 224–6 Augustine 17 Averroes 160, 168–9, 174, 216 Avicenna 41–50, 52–66, 68, 70–8, 80–3, 87–108, 110, 112–15, 119, 226 al-Bāqillānī 216 de Boer, Sander 88 Boethius 23 Bonin, Thérèse 107 Boso 33 Cantor, Georg 208 Cross, Richard 152, 159, 199 Davidson, Herbert 73, 215, 218 Democritus 187 Druart, Thérèse-Anne 44 Fakhr al-Dīn al-Rāzī 42, 76–8, 226 n34 al-Fārābī 74, 213, 216, 225 Gaunilo 220 Gilson, Étienne 108, 111–12 Gregory of Rimini 193–5, 199, 208–9 Gutas, Dimitri 42, 44, 73, 75, 81–3 Hasse, Dag Nikolaus 42, 73, 88 Henry of Ghent 171 Henry of Harclay 183–4, 202–3, 205–6 Hintikka, Jaakko 186

Ibn al-Nadīm 213 Ibn Suwār 216–18, 218–19 n25, 220, 223, 225–6 John Buridan 184, 190 John Duns Scotus 154 n9, 155–6, 158, 163, 171–2, 176, 209 John Rodington 163 n40 al-Juwaynī 218 Kenny, Joseph 44 n9, 56 n32, 58 n44, 59–60 al-Kindī 216 n14 Knuuttila, Simo 186 Koslicki, Kathrin 141 n31 Kretzmann, Norman 188, 194, 208 Kripke, Saul 30 Mankdīm 215 Markosian, Ned 139 n28, 140 n29 Marmura, Michael 44, 73 al-Masʿūdī 213 McGinnis, Jon 47–8, 60 n50, 61 n53 Murdoch, John 182–3, 191–3, 195, 204–7, 209 Nicholas of Autrecourt 184 Pasnau, Robert 116 n64, 117 n68, 144 n35, 182–3, 192–3, 199, 201 Perler, Dominik 154, 164 Peter Auriol 151–7, 159–64, 167, 172–8 Philip the Chancellor 109 Philoponus 217 Plato 16, 54 n29, 90, 112 Plotinus 87–8 n1 Pseudo-Robert Grosseteste 109 Robert Kilwardby 52 Roscelin 16 Saadia Gaon 216 al-Sarakhsī 216 n14

246

Index

Smith, A. D. 23–6, 28–35 Spade, Paul Vincent 132 n19 Tachau, Katherine 156, 158 n23, 159 Thomas Aquinas 52, 87–90, 107 n46, 111, 113–19, 122–49 al-T. ūsī, Nas.īr al-Dīn 42, 48, 76–79, 226

Walter Chatton 151–2, 155, 157–9, 162 n39, 163–7, 172–6, 178, 184, 199 William Crathorn 184 William Ockham 181–5, 188–93, 195–209 Williams, Thomas 35–6, 38 Wood, Rega 152, 159

Visser, Sandra 35–6, 38

Yah.yā Ibn ʿAdī 213–26