Metaphysical Themes, Medieval and Modern: Proceedings of the Society for Medieval Logic and Metaphysics (The Immateriality of the Human Mind, the Semantics of Analogy and the Conceivability of God) 1443853860, 9781443853866

Citation preview

Metaphysical Themes, Medieval and Modern Volume 11

Also available in the series: The Immateriality of the Human Mind, the Semantics of Analogy, and the Conceivability of God Volume 1: Proceedings of the Society for Medieval Logic and Metaphysics Categories, and What Is Beyond Volume 2: Proceedings of the Society for Medieval Logic and Metaphysics Knowledge, Mental Language, and Free Will Volume 3: Proceedings of the Society for Medieval Logic and Metaphysics Mental Representation Volume 4: Proceedings of the Society for Medieval Logic and Metaphysics Universal Representation, and the Ontology of Individuation Volume 5: Proceedings of the Society for Medieval Logic and Metaphysics Medieval Skepticism, and the Claim to Metaphysical Knowledge Volume 6: Proceedings of the Society for Medieval Logic and Metaphysics Medieval Metaphysics; or Is It "Just Semantics"? Volume 7: Proceedings of the Society for Medieval Logic and Metaphysics After God, with Reason Alone—Saikat Guha Commemorative Volume Volume 8: Proceedings of the Society for Medieval Logic and Metaphysics The Demonic Temptations of Medieval Nominalism Volume 9: Proceedings of the Society for Medieval Logic and Metaphysics Skepticism, Causality and Skepticism about Causality Volume 10: Proceedings of the Society for Medieval Logic and Metaphysics Metaphysical Themes, Medieval and Modern Volume 11: Proceedings of the Society for Medieval Logic and Metaphysics

Metaphysical Themes, Medieval and Modern Volume 11: Proceedings of the Society for Medieval Logic and Metaphysics

Edited by

Gyula Klima and Alexander W. Hall

Metaphysical Themes, Medieval and Modern Volume 11: Proceedings of the Society for Medieval Logic and Metaphysics, Edited by Gyula Klima and Alexander W. Hall This book first published 2014 Cambridge Scholars Publishing 12 Back Chapman Street, Newcastle upon Tyne, NE6 2XX, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2014 by Gyula Klima, Alexander W. Hall and contributors All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-5386-0, ISBN (13): 978-1-4438-5386-6

TABLE OF CONTENTS

Introduction ................................................................................................. 1 Alexander W. Hall Part One: Metaphysical Themes: 1274-1671 Remarks on Pasnau’s Metaphysical Themes: 1274-1671 ............................ 9 Andrew W. Arlig Categories and Modes of Being: A Discussion of Robert Pasnau’s Metaphysical Themes ................................................................................ 27 Paul Symington Response to Arlig and Symington ............................................................. 57 Robert Pasnau Part Two: Substance Ontology, Medieval and Modern An Argument for Hylomorphism or Theism (But Not Both) .................... 75 Travis Dumsday The Rises and Falls of Analysis and Metaphysics: Comments on “An Argument for Hylomorphism or Theism (But Not Both)” by Travis Dumsday.................................................................................... 85 Gyula Klima Response to Gyula Klima’s “The Rises and Falls of Analysis and Metaphysics” ...................................................................................... 89 Travis Dumsday Rejoinder to Travis Dumsday’s Response ................................................. 93 Gyula Klima

vi

Table of Contents

Part Three: The Natural Theology of Thomas Aquinas The Burden of Proof: Aquinas and God Science ....................................... 97 Alexander W. Hall Comments on Alexander W. Hall’s “The Burden of Proof: Aquinas and God Science” .................................................................................... 117 Michael Sirilla Response to Michael Sirilla’s Comments ................................................ 123 Alexander W. Hall Appendix ................................................................................................. 129 Contributors ............................................................................................. 131

INTRODUCTION ALEXANDER W. HALL

The Proceedings of the Society for Medieval Logic and Metaphysics (PSMLM) collects original materials presented at sessions sponsored by the Society for Medieval Logic and Metaphysics (SMLM). Founded by Gyula Klima (Director), Joshua Hochschild, Jack Zupko and Jeffrey Brower in 2000 (joined in 2011 by Assistant Director and Secretary, Alexander Hall) to recover the profound metaphysical insights of medieval thinkers for our own philosophical thought, the Society currently has over a hundred members on five continents. The Society’s maiden publication appeared online in 2001 and the decade that followed saw the release of eight more volumes. In 2011, PSMLM transitioned to print. Sharp-eyed readers of these volumes will note the replacement of our (lamentably copyrighted for commercial use) lions, who guarded the integrity of the body of an intellectual tradition thought to be dead, with the phoenixes that mark our rebirth. Friends of the lions will be happy to note that they remain at their post, protecting SMLM’s online proceedings and announcements at http://faculty.fordham.edu/klima/SMLM/. This volume gathers papers read by SMLM members at the 2012 meetings of The International Congress on Medieval Studies (ICMS) and The American Catholic Philosophical Association (ACPA). Parts I and II (taken from the ICMS and ACPA sessions, respectively) look at important figures in the history of the study of the metaphysics of substance over the last eight centuries. Attention to this period sheds light on contemporary disputes as well as the history of thought that leads into the modern period. The essays in Part III, read at the ACPA’s SMLM satellite session, present vying, contemporary interpretations of what metaphysical and logical presuppositions underlie the natural theology of Thomas Aquinas.

2

Introduction

Part I: Metaphysical Themes: 1274-1671 The ICMS session papers that make up Part I discuss Robert Pasnau’s Metaphysical Themes: 1274-1671 (Oxford University Press, 2011), which treats the metaphysics of substance over the four centuries that separate the deaths in 1274 of Aquinas and Bonaventure from the completion of the first drafts of Locke’s Essay Concerning Human Understanding in 1671, thereby picking up with early critics of classical scholasticism and drawing to a close in the birth of the modern period. The narrative in six parts concerns the material substratum of change, material substances and their properties, modes or accidents, and the persistence of substances over time. Though a single session of papers cannot do justice to Pasnau’s rich study, the essays in this volume speak to central concerns of the work, namely, category theory and medieval mereology. Medieval mereology is the study of parts, wholes and the relations between them that develops out of Plato’s use of the method of collection and division to demarcate natural kinds and Aristotle’s discussions of parts and wholes in Metaphysics and Topics.1 In his “Remarks on Pasnau’s Metaphysical Themes,” Andrew Arlig reflects on Pasnau’s discussions of holenmerism and nominalist approaches to identity over time. Holenmerism is the thesis that one thing can be present as a whole in a multiplicity of discrete things, as most scholastic thinkers2 believe the rational human soul to be present in the body’s integral parts and God in everything. In response to Pasnau’s suggestion that, were it the case that immaterial things exist, holenmerism would be a plausible candidate to distinguish them from material things, Arlig expands upon arguments by Henry More that call into question the possible unity of holenmers (which, like universals, must exist wholly in discrete things) to argue that holenmerism is incoherent. Arlig then turns to Pasnau’s discussion of John Buridan on identity over time, taking issue with Pasnau’s statement that Buridan denies persistence across time for humans and other animals.

1 See Andrew Arlig, “Medieval Mereology”, The Stanford Encyclopedia of Philosophy, first published 2006, revised 2011 (latest version on-line: http://plato .stanford.edu/entries/mereology-medieval/); and “Mereology”, Encyclopedia of Medieval Philosophy, edited by Henrik Lagerlund (Springer, 2011). 2 And not just scholastics. The list of figures who subscribe to the existence of holenmers would include Plotinus, Augustine and Descartes, to name but a few. For references, see Metaphysical Themes p. 337 n. 18.

Alexander W. Hall

3

Agreeing that some arguments against universals will probably work equally well against holenmers, Pasnau notes that by limiting the existence of holenmers to God and the rational soul, one may preserve the unity of holenmers by pointing to the fact that, unlike universals, holenmers have operations beyond those they perform in discrete parts of things. Pasnau then contends that Buridan’s nominalism prohibits him from accepting that two things similar and yet somehow distinct can ever literally be described as identical (as one and the same universal cannot subsist in disparate individuals), and hence, whatever commonsense ontology may suggest, persons (for instance) cannot persist across time in the literal sense that would require that their parts remain the same. As regards the categories, Aristotle describes them as a division of “things that are” (Categories (Cat.) 1a20);3 but whether the things in question are extramental, linguistic or mental entities (or some combination thereof) is unclear.4 On the one hand, Aristotle states that accidental and universal features cannot exist apart from the individual substances whom they characterize (Cat. 2b4-6), thereby inverting a Platonic realism that hypostasizes universal features into unique, immaterial, paradigmatic forms, participation in which renders individuals the types of things that they are (hence many beautiful things would be more or less beautiful to the extent they participate in the form of beauty).5 On the other hand, Aristotle presents the ten categories themselves as a list of “things said without any combination” (Cat. 1b25), thereby giving rise to the impression that the work is about terms or what they call to mind. Mid-thirteenth-century interpretations of the Categories are shaped by the reception in the Latin West of the complete writings of Aristotle, viewed in part through the lens of his Islamic commentators. Responding to (1) the distinction between concepts of first and second intention (i.e., concepts of things in the world, e.g., cats, and concepts about these concepts, e.g., the species cat, respectively); (2) attention to the boundaries 3

All translations from Categories are from Aristotle’s “Categories” and “De interpretatione,” trans. and ed. J. L. Ackrill, Clarendon Aristotle Series (Oxford, 1963). 4 See Jorge J. E. Gracia, “Categories Vs. Genera: Suárez’s Difficult Balancing Act,” in Categories and What Is Beyond, Gyula Klima and Alexander Hall eds., Proceedings of the Society for Medieval Logic and Metaphysics vol. 2 (Cambridge Scholars Publishing, 2011), 7-18. 5 See, e.g., Republic 504e–518c; 596e–597a, Phaedo 100b–102a3, and Phaedrus 247c3–247e6.

4

Introduction

of various Aristotelian sciences (such as metaphysics, physics and logic); and (3) Aristotle’s notion of an isomorphism between things, concepts and words, medieval thinkers enter into a sustained dialog regarding the ontological status of substances and the accidents (making up the other nine categories) that depend on these substances for their existence (as a quality, for instance, cannot exist apart from its subject)6 and wonder whether there exists a method (a sufficientia praedicamentorum) to fix the number of categories, either through attention to predication, with the belief that such corresponds to some extent to ways in which things are (as we see in Albertus Magnus and Thomas Aquinas) or by attention to various modes of being (as with Simon Faversham and Radulphus Brito). These different medieval accounts of what the categories categorize and how to fix the number of categories have led contemporary scholars to class medieval accounts with reference to whether or not they are reductionist or deflationary, with reductionist accounts seeking to reduce the number of irreducibly different types of beings picked out by the accidental categories (as we see in fourteenth-century authors such as Ockham and Buridan) and deflationary accounts interpreting talk of accidents as of ways in which substances exist, with these accounts ranging from a realism that views accidents as beings (though less real than the substances in which they inhere) to eliminitivism.7 Robert Pasnau and Paul Symington are in broad agreement that Aquinas’s account of the accidental categories is deflationary, whereas the two differ over whether it is reductionist, with Symington contending in his “Categories and Modes of Being” that this is the case and that Aquinas develops his account via an approach (idiosyncratic relative to later scholastics) to metaphysics involving the real distinction between essence and existence, the analogy of being and an isomorphism between thought and reality, concluding with remarks on the limitations of Pasnau’s approach to metaphysics in general given what Symington describes as Pasnau’s resistance to viewing conceptual or linguistic structures as useful guides to the practice. 6

Though medievals allow that such is possible through divine agency, as in transubstantiation, wherein the accidents (appearance, smell, etc.) of bread and wine persist absent the substance of the host. 7 Good places to begin sorting out these various issues regarding medieval interpretations of Aristotle’s Categories are the article on this topic by Jorge J. E. Gracia and Lloyd Newton in the Stanford Encyclopedia of Philosophy http://plato.stanford.edu/entries/ medieval-categories/#ThiCen and Newton ed., Medieval Commentaries on Aristotle’s Categories, (Brill, 2008).

Alexander W. Hall

5

In response to Symington, Pasnau speaks to the value of what ordinary talk can tell us of the world, reckoning attention to such one of the valuable features of scholastic metaphysics, while cautioning that it is unlikely language serves as a guide to what fundamentally exists. As regards the claim that Aquinas looks on the accidental categories as each corresponding to an irreducible kind of being, Pasnau contends that for Aquinas this likely holds only for the categories of quantity and quality, marking the lesser accidental categories as mere structures or features of reality, not themselves items over and above the items in the other categories, illustrating this with reference to Aquinas’s treatment of the categories of action and passion.

Part II: Substance ontology, medieval and modern Part II concerns the development of analytic metaphysics and principal, competing theories in contemporary substance ontology literature: (1) substratum theory (2) bundle theory (3) primitive substance theory and (4) hylomorphism. The respective theories construe material objects as (1) bare substratum-attribute compounds (2) aggregates of properties (3) primitively unified and individuated or (4) prime matter-substantial form compounds. Travis Dumsday’s “An Argument for Hylomorphism or Theism (But Not Both)” contends that concerns bound up with (1) force the substratum theorist either to abandon (1) for (4) or to recognize the existence of a causal agent or agents operating outside the laws of nature, and thereby move in the direction of some form of theism or, at the least, a denial of metaphysical naturalism, closing with some reflection on the relevance of substance ontology to natural theology. Gyula Klima’s “The Rises and Falls of Analysis and Metaphysics” sees the substratum theorist’s dilemma as a result in part of a desire for metaphysical novelties that grew out of the logical positivists’ conviction that the study of the history of philosophy offers little of value. Klima judges Dumsday’s attempt to saddle the substratum theorist with a dilemma a success, adding that analysis likewise reveals that the substratum theorist’s notion of bare-particularity is self-contradictory, as the truth of the statement ‘x is a bare particular’ assigns the bare particular a property, viz., bare-particularity, given that, as Klima contends in his rejoinder to Dumsday’s reply (which works on multiple fronts to save the substratum theorist from advancing an open contradiction), ‘bare particular’ must here indicate a common predicable rather than a directly referring term, like a singular term.

6

Introduction

Part III: The natural theology of Thomas Aquinas In Part III the discussion turns from analysis, metaphysics, substance ontology and the importance of the latter to natural theology to the natural theology of Thomas Aquinas in particular, whose Five Ways are generally taken to be demonstrative proofs of the type Aristotle terms ‘that’ or ‘hoti’ (rendered ‘quia’ in Latin) in Posterior Analytics (An.Post). Such proofs are thought to be able to reason from phenomena to the existence of what is their cause. Recent scholarship issuing from Radical Orthodoxy thinkers Catherine Pickstock and John Milbank, however, contends that the proofs are merely probabilistic. Again, Aquinas’s own Commentary on Aristotle’s “Posterior Analytics” rules out quia proof in the ordinary sense as regards God. Alexander Hall’s “God Science” disputes Pickstock and Milbank on the status of the Five Ways as demonstrative and identifies the type of quia proof at work in the arguments as the ‘through the accidents (per accidens)’ variety discussed at An.Post. 2, 8-10, illustrating this thesis with reference to the First Way. Michael Sirilla’s comments on Hall’s paper accept Hall’s criticism of Pickstock and Milbank as well-founded, while noting that if per accidens quia proof of the type Hall describes really is at work in the First Way, it would render the proof invalid as (what Sirilla takes it to be) a demonstration of God’s existence, as the First Way would then show rather only that some first mover exists. Hall’s response acknowledges this difficulty in his account of Aquinas and notes that it arose from his having supposed that the First Way would use the same form as the syllogism that Aquinas uses to illustrate per accidens quia proof. Unfortunately, the proof that Aquinas uses to illustrate per accidens quia proof falls prey to a formal fallacy. Nevertheless, this difficulty is not endemic to per accidens quia proof as such and therefore does not vitiate the First Way, which Hall reconstructs as a valid per accidens quia proof intended by Aquinas to demonstrate the existence of God and not merely some first mover.

PART ONE: METAPHYSICAL THEMES: 1274-1671

REMARKS ON PASNAU’S METAPHYSICAL THEMES: 1274-1671 ANDREW W. ARLIG

Robert Pasnau’s Metaphysical Themes is a bold work, covering a large swath of relatively unknown and sometimes underappreciated material. It is a tremendous contribution to the study of both medieval and early modern philosophy. Pasnau’s subject is material substance as it was understood by thinkers living and working in the period between the fourteenth century and the seventeenth. He surveys the views of both wellknown philosophers (including Ockham, Buridan, Suarez, as well as Henry More, Pierre Gassendi, and René Descartes) and many lesser known figures, some of whom deserve more attention (my votes go to John Wyclif and Anne Conway). By bringing this body of work into view, Pasnau reveals a number of interesting avenues for future research, both in medieval philosophy and in early modern philosophy. When I say that Pasnau’s focus is material substance, I mean that he considers almost every conceivable way of analyzing material substances: their essences, their parts, their properties and modes, their identity and persistence conditions, and so forth. Given the ambitious sweep of this study and the amount of space at my disposal, clearly I can cover only a small part of Pasnau’s book. In the following remarks, I will focus on two issues that arise in the book. I hope that by detaching these topics from the complex whole of which they are parts, I have not distorted Pasnau’s views inappropriately.

1. Holenmerism (chapter 16)1 As Pasnau notes, the standard Scholastic line on immaterial things— including, in particular, the intellective soul of a human being—is that 1

Unless otherwise specified, all page references are to Robert Pasnau, Metaphysical Themes 1274-1671. Oxford: Clarendon Press, 2011. And unless otherwise noted, translations are mine.

10

Remarks on Pasnau’s Metaphysical Themes: 1274-1671

they are present in every region or part of a body as a whole (p. 337). This commitment comes in degrees. Aquinas, for example, holds that every substantial form exists as a whole in each part of the composite substance.2 Many later thinkers, including those in the so called “nominalist” camp (such as Ockham, Buridan, and Nicole Oresme) assert that it is only the intellective, or rational, soul that exists as a whole in the whole body and as a whole in each part of the body (tota in toto et tota in qualibet parte corporis). So, for example, in what is thought to be his first series of lectures on Aristotle’s On the Soul, Buridan clarifies the ways in which an animal or plant soul is wholly in its body: Here then is the first conclusion: If we interpret “whole” categorematically, the whole soul of a horse is in the whole body. This is clear because [the soul] is extended through the whole body. The second conclusion: If we interpret “whole” syncategorematically, it should be accepted that the whole soul of a horse is in the body. This is clear because each part of the soul of the horse is in the body. But if we interpret “whole” syncategorematically, it is false that the whole soul of a horse is in the whole body and in each part of [the horse]. This is clear since it is false that each part of the soul of a horse is in each part of the body. The part of the horse’s soul which is the part in the eye is not in the foot.3

Or, consider this assertion in a question on Aristotle’s On the Soul attributed to Nicole Oresme: This kind of soul is not in every part as a whole, not integrally, not potentially, and not essentially. That it is not integrally is obvious, since the part that is in the foot is not in the hand. Nor is it potentially, since it is not the case that all the powers are in every part. Nor is it essentially, as I explained previously, since the part that is in the hand is not soul, but something belonging to soul. For this reason, Aristotle compares a sensitive soul to a figure. For example, the figure of a quadrangle is not in each part of a quadrangle.4 2

See for example, Summa Theologiae 1a 76.8c. Buridan Quaestiones in Aristotelis De anima (prima lectura) 2.6 (Patar ed., pp. 284-5, lines 122-32). (Text: Benoît Patar (ed.) Le Traité de l’Âme de Jean Buridan [de prima lectura]. Philosophes Médiévaux 29. Louvain-la-Neuve / Longueuil (Quebec): Éditions de l’Institut Supérieur de Philosophie / Éditions du Préambule, 1991.) 4 Oresme Quaestiones in Aristotelis De anima 2.4 (Patar ed., p. 144, lines 108-15). (Text: Benoît Patar (ed.) Nicolai Oresme Expositio et Quaestiones in Aristotelis ‘De Anima’. Philosophes Médiévaux 32. Louvain-la-Neuve / Louvain-Paris: Éditions de l’Institut Supérieur de Philosophie / Éditions Peeters, 1995.) 3

Andrew W. Arlig

11

In short, the soul of a non-human animal or a plant is “extended as its body is extended” (extensa extensione sui corporis), and hence the whole soul is in the whole body only because one part of the soul imbues one part of the body and no part of the soul is not imbuing some part of the body.5 The rational soul of a human, however, “is not extended in the manner that its body is extended”,6 and thus it can be wholly in each part: Again, if in a human there is not some other soul, then it ought to be said without qualification (absolute) that the whole soul of a human and every power is in every part of a human, and hence that the intellective [power], the visual [power], and so on, are in the foot. But if in a human there is another soul and form—such as a sensitive [soul]—then one should say the same thing about those [lower souls] as [it will be said] about the souls of brutes.7

Indeed, if the rational soul is mereologically simple, then there is no other way in which it could animate the whole body: Now we should briefly elaborate upon the intellective [soul] (about which more is to be made apparent in Book 3). I say that [the intellective soul] is not whole properly speaking, because “whole” is not said of anything except a divisible thing having parts (although sometimes it is improperly said of an indivisible thing). And, thus, it should be said that an intellective soul is in each part as a whole in this manner [i.e. improperly] because it informs a body and is not extended as a body is extended (extensa ad extensionem corporis). Thus, it follows that in every part either it or part of it is, and given it does not have a part, it follows that in every part it exists as a whole.8

Later on in the same question, we are given an explanation for difference. The human intellective soul is a form that is not drawn from the potentiality of the matter.9 And it is for this reason that intellective soul is not extended throughout the body in the way that body itself is extended. That is, unlike a body or non-human soul,

5

the out the the the

Buridan Quaestiones in De anima (prima lectura) 2.6 (Patar ed., p. 281, line 68). Ibid., lines 66-67. 7 Oresme Quaestiones in De anima 2.4 (Patar ed., p. 142, lines 50-5). 8 Oresme Quaestiones in De anima 2.4 (Patar ed., p. 142, lines 39-47). See also Buridan Quaestiones in De anima (prima lectura) 2.6 (Patar ed., p. 283, lines 1059). 9 Oresme Quaestiones in De anima 2.4 (Patar ed., p. 145, lines 133-39 and 14347). 6

12

Remarks on Pasnau’s Metaphysical Themes: 1274-1671

intellective soul is not extended by having one part here and another part there. Pasnau sees in the doctrine of holenmerism a “promising” way to demarcate the immaterial from the material: “All and only material things have corpuscular, non-holenmeric structure” (p. 342). Of course, this method for marking the material off from the immaterial can only succeed if holenmerism is intelligible and ontologically principled. In this section, I will try to raise some concerns about holenmerism. Let me restate the doctrine of holenmerism with a little more precision. An entity X is holenmerically present in something else, Y, if and only if, Y has integral parts and for each integral part of Y, the whole of X exists in that part.10 Specifically for my purposes, this fact about a holenmeric soul is important: If a soul is holenmerically present in a body, the soul will be present in at least two parts of the body, P1 and P2, in such a way that the soul is wholly present in P1 and it is wholly present in P2, and P1 is mereologically discrete from P2.11

To say that “the soul is wholly present in some part Pn” is to say that If S1 … Sn are the parts of the soul, then the soul is present in a part Pn only if S1 is present in Pn and S2 is present in Pn and … Sn is present in Pn.12 10

The restriction to “integral” parts is here because medieval authors tend to recognize a host of different kinds of parts, some of which probably cannot be hosts for holenmers. There is no common definition of integral parts, but many medieval philosophers characterize integral parts as proper parts (in contrast to the contemporary mereological notion of an improper part), which comprise some sort of quantity. For more on the types of parts in medieval mereologies, consult Andrew Arlig “Medieval Mereology”, The Stanford Encyclopedia of Philosophy, first published 2006, revised 2011 (latest version on-line: http://plato.stanford.edu/entries/mereology-medieval/). 11 X is mereologically discrete from Y if and only if no part of X is a part of Y and no part of Y is a part of X. Of course, it is not true that every pair of parts of the body that one could pick out are discrete from one another. Some pairs will be overlapping parts. But in the cases we are interested in, there will be noncoincident overlapping parts (say, my forearm and my hand), which entails that there will be at least two parts of the body that do not overlap at all (e.g., the oneinch long part near my elbow and the index finger). 12 If the soul is simple, then the only part of the soul is the soul itself, and then it would follow as Oresme notes that the soul is present in Pn only if the soul’s

Andrew W. Arlig

13

Now let us consider a criticism of holenmerism by Henry More: [I]t is the same as if someone were to say that there is nothing of the soul that is not included within [bodily part] A, and yet that, at the same moment of time, […] the whole soul is in [some distinct bodily part] B, as if the whole soul were outside its whole self. This is clearly impossible in any singular and individual thing. As for universals, they are not things, but rather notions we apply in contemplating things.13

More observes that if a soul were present holenmerically in a body, then it would behave as a universal is often said to behave. But More thinks that there is something metaphysically unprincipled about this. No concrete thing can behave in the manner that a universal behaves. To see the parallel, consider the classical understanding of a universal as a thing that can be wholly present in many discrete particulars at the same time. U is a universal if and only if it can be wholly present in an individual I1 and wholly present in another individual I2 at the same time, where I1 is mereologically discrete from I2.

This notion of a universal was famously problematized by Boethius in his second commentary on Porphyry’s Isagoge: But genera and species cannot be. And this is understood based on these [considerations]. Everything that is common to many at one time cannot be one. For that which is common is of many, especially when one and the same thing is in many at one time as a whole. For no matter how many species there are, in all of them there is one genus, and not because each species grabs from it some, let us say, “parts”. Rather, at one time each [species] has the whole genus. The result of this is that because the whole genus has been posited in many individual [species] at one time, it cannot be one. For it cannot come about, when the whole is in many at one time, that it itself is numerically one. But if that is so, then the genus in particular cannot be one. And the result of this is that [the genus] is altogether nothing. For every thing that is, is precisely for the reason that it is one.14

(improper) part is present in Pn. That is, the soul is present in Pn only if the soul is present in Pn. 13 Enchiridion Metaphysicum 27.12 (translated by Pasnau, p. 342). 14 Boethius In Isagogen Porphyrii commenta, editionis secundae, book 1, c. 10 (CSEL 48, pp. 161-2).

14

Remarks on Pasnau’s Metaphysical Themes: 1274-1671

If some thing—a concrete thing, not a concept—were universal, it would have to be capable of existing wholly in I1 and wholly in I2, where I1 and I2 are mereologically discrete things. But, then, this universal thing would have to be, as More puts it, “outside itself”: All of it would have to be here and, yet, at least some of it is not here but there. But some of it cannot be there, since every one of the potential parts that could be there is already spoken for; they are all here.15 Pasnau sees that More’s argument boils down to the Boethian assertion that no thing can behave like a universal, and he is right to insist that in so far as this is the criticism, More can be accused of begging the question. The behavior attributed to universals—existing wholly in many discrete particulars at the same time—is precisely what would have to set them apart from particulars (p. 342). Thus, if a realist were confronted with Boethius’s argument, he could merely shrug: “Yes,” he could say, “that is how universals work; that is precisely what makes them universal.” Hence, in so far as More tries to undermine the doctrine of holenmerism by asserting that no thing can be wholly present in many discrete particulars, the argument does not have any force. But let me try to help More out here. The real impetus behind More’s argument is that no particular thing can exist holenmerically in some other particular. Of course, to merely assert this is not sufficient, since then too one could be accused of begging the question. But the Moreans could take up a stronger position if they were to call into question whether holenmerism is a principled position. Notice that the advocate of holenmerism is forced to give away the notion that being wholly present in many discrete particulars at the same time is the proprium (or perhaps differentia) of universals. After all, for most of the Scholastics in the period covered by Pasnau’s book, substantial forms—and specifically souls—are particulars. Hence, some kinds of particulars, rational souls, are capable of existing wholly in discrete bodies or regions at the same time. But other particulars cannot exist in this 15 An analogous worry is raised by Anselm concerning God’s presence in the world (Monologion c. 21, Opera Omnia, ed. Schmitt, vol. 1, pp. 36-7): If God were wholly in R1, then it would seem that He could not be wholly in a nonoverlapping region R2. For if He were wholly in R1, then nothing of Him is not in R1. And if He is wholly in R2, then nothing of Him is not in R2. But by hypothesis He is in R1 as well as R2. Hence, it cannot be the case that nothing of Him is in not in R2.

Andrew W. Arlig

15

manner. Why are both kinds of particulars each particular? If there are universal things, what sets them apart from particulars? Considered from this angle, it seems that once we give away the notion that being wholly present in many at the same time is the proprium of universal things, the metaphysician’s nonchalant shrug begins to look rather unprincipled. Here then is my challenge, inspired by More, to the advocate of holenmerism: You, the advocate, should give us some reasons for thinking that a nonuniversal, natural thing can be extended without being divided. At the very least, it seems that either (1) you must concede that my soul is in fact a universal,

or (2) if my soul is a particular, then (2a) if you think that there are universal things, you owe us another distinguishing characteristic (a differentia or proprium) that separates universal beings from particular beings, or

(2b) if you think that there are no universal things, then you owe us a reason—a reason different from the one that Boethius identified—why there can be no universal things.

As we have already noted, no one in the period we are concerned about will concede (1). It should be stressed that (2) spells out the bare minimum that the advocate of holenmerism must provide. But to really satisfy the Morean critic, much more than the bare minimum would be desirable. Let us start with (2a): The realist advocate of holenmerism owes us that distinguishing characteristic. But another plausible candidate does not come to mind; for the other obvious proprium of universals—namely, being predicable of many numerically distinct things—has its own difficulties. Homogeneous wholes (i.e. stuffs and masses) share this behavior. Every portion of the lake is, like the whole lake itself, water.16 16

For this difficulty see Boethius De Div. 879d-880a (Magee ed., p. 14). Boethius’s solution is that while a portion of the lake is the same in substance as

16

Remarks on Pasnau’s Metaphysical Themes: 1274-1671

Moreover, if the whole animal gets its name and definition in virtue of being imbued by a soul of a certain type and this soul imbues each part as a whole, then it would seem that each part can take the name and definition of the whole animal. In other words, it seems that holenmerism implies that each part of a human will be human, and hence that the soul is predicable of the each part in the same way that a universal is predicable of each individual. And, in fact, a common challenge to the doctrine of holenmerism is that, if true, then each part of an animal (e.g. its foot or ear) would be animal, and each part of a man would be man.17 The nominalist advocate of the doctrine runs into trouble further down the path, since he has a whole host of reasons why there cannot be universals anywhere outside the mind.18 Some of the nominalist’s arguments are better than others. But for the present, I will concede that there are no the lake, it is not the same in quantity. Interestingly, several discussions of the manner in which a soul imbues a body employ the distinction between homogeneous wholes and heterogeneous wholes to explain why the parts of some kinds of animals (e.g. worms) and plants can live when the original creature is cut in two: Here are the conclusions: The first is that in certain segmented animals and in certain plants the soul is a homogenous whole and the soul is in every part of discernible size as a whole potentially and essentially. This is clear because the activities of life appear there after the separation of the parts, and as a consequence, a soul is there, since an activity calls for a form. (Oresme In de An. II.4 [Patar ed., p. 143, lines 74-8]) The souls of other animals, Oresme continues, are heterogeneous wholes and these souls exist as integral wholes in their bodies. 17 See, e.g., Oresme QQ. In De Anima 2.4, challenge # 2 (p. 141), and Buridan QQ. In De Anima (ultima lectura) 2.7, challenge # 1 (p. 81, in Peter G. Sobol, John Buridan on the Soul and Sensation. An Edition of Book II of His Commentary on Aristotle’s Book on the Soul with an Introduction and a Translation of Question 18 on Sensible Species, Ph.D. dissertation, Indiana University, 1984). In the replies to this objection, medieval authors try out several strategies for restricting the predication of substance terms to whole animals or humans. Intriguingly, in his later lectures, Buridan seems to concede that there is a sense in which a hoof is animal (Sobol, pp. 94 f.). The reason that we are not entitled to say that a horse’s hoof is a horse is that, in fact, “horse” is a connotative term, not strictly speaking a substantial term (p. 97). 18 For a representative list, see Ockham Summa Logicae part 1, c. 15 (Opera Philosophica 1, pp. 50-4) as well as the mind-numbingly comprehensive discussion in his Ordinatio 1, d. 2, qq. 4-7 (Opera Theologica 2, pp. 99 f.).

Andrew W. Arlig

17

universal things and that there are reasons to hold this that are independent of the Boethian objection. In other words, I will concede that the nominalist can give us an answer to (2b). But recall that I said giving an answer to (2a) or (2b) was the bare minimum. To thoroughly eliminate the suspicion that holenmerism is an unprincipled, perhaps even ad hoc, doctrine, the nominalist advocate of holenmerism should give us some reasons why a soul—which again, I will stress, is a particular thing inhabiting the natural world—can be extended throughout a body without being divided and apportioned part to part. This last demand is all the more urgent for nominalists like Buridan and Oresme, who restrict holenmerism to rational souls, since here in particular it begins to appear that holenmerism is brought in solely to save some cherished doctrine of the faith. To be sure, when it comes to metaphysical disputes, it is not always clear upon whom the burden of proof lies. But in the present case, I think the Moreans have the stronger claim. After all, both the Moreans and the advocates of holenmerism believe that some kinds of particulars, if they have parts, must have these parts spread out part outside of part. There is an even trickier point that needs to be stated as a caveat, and that is that in metaphysical debates reasons and explanations must at some point come to an end.19 Philosophers rightfully complain if their opponents refuse to offer any reasons or explanations, but the tougher part is determining whether someone has said enough. However, on this point as well, I think the Morean has the stronger case. As Pasnau acknowledges, later medieval advocates of holenmeric rational souls generally have nothing substantial to say in response to the Morean (p. 339, note 23). The best that our authors apparently can do is offer an analogy: And about the intellective soul as it is related to a man it should be imagined just as we imagine about God as He is related to the universe: For God is present to each part of the universe in virtue of [His] 19 See for example David Lewis’s observation that in the debate over the existence and status of universals, all sides must resort to some primitive relation or fact: the realist has the primitive instantiation relation, the resemblance nominalist takes it as a primitive fact that some things resemble other things, and the natural class nominalist takes it as a primitive fact that things break down into natural classes (p. 347). David Lewis “New Work for a Theory of Universals,” Australasian Journal of Philosophy 64 (1983): 343-77.

18

Remarks on Pasnau’s Metaphysical Themes: 1274-1671 unmediated assistance and lack of distance (indistantiam), and not by delimiting Himself with respect to some specific state or some specific place. And, thus, [by existing] in the East He is not distant from Himself who is [also] existing in the West. For distance is due to quantity, and God is affected by no quantity. About the intellective soul as it is related to a man it ought to be imagined that [they are related] in a similar manner: that [the soul] is present to each part of the man in virtue of unmediated assistance and lack of distance, and in this way [even though it is] in the head it is not distant from itself, since it exists in the foot.20

This seems to be no more than an acknowledgement that some part of the natural world has supernatural properties, and it leaves the holenmerist with a gap in his account of nature, if not his ontological framework.21 In my view, the explanations have run out too soon and that More is entitled to say, “So much the worse for holenmerism!”

2. Identity over time (chapter 29) The fact that an animal soul is not holenmerically present in the animal body has implications for the permanence and persistence of animals over time and change. If an animal soul imbues a body by having one part here and one part there, then if one of the bodily parts (the one here, say) is removed, it appears that the soul has lost a part as well. This seems to entail that the animal—the hylomorphic composite—does not endure as a whole if it gains or loses parts.22 And, indeed, as Pasnau shows, several fourteenth-century “nominalist” thinkers did conclude just that. Here I think that Pasnau gets the broad outlines of the nominalist view right. But I think some refinements should be noted.

20

Buridan Quaestiones in De Anima (prima lectura) 2.6 (Patar ed., pp. 283-4, lines 109-118). 21 In this respect, Aquinas might have the stronger position, since he asserts that all substantial forms exist holenmerically in their composites. (Of course, he still needs to explain why accidental forms do not have this property.) Looked at from this perspective, one might think that what is really most remarkable in this whole discussion is the later Scholastic position that the souls of plants and animals do not exist holenmerically in their corresponding bodies. 22 This line of thought is elaborated in some detail in an interpolation in the version of Buridan’s Quaestio (prima lectura) 2.6 found in the Turin MS H.III 30 (a transcription can be found in the apparatus of Patar’s edition (1991), pp. 282-4).

Andrew W. Arlig

19

We will focus on Buridan’s theory of persistence over time.23 On several occasions Buridan pondered whether a composite can endure if it gains or loses parts. In answering this question, Buridan reveals his allegiance to a principle embraced by, among others, Ockham and Abelard: a whole is the same thing as its parts taken together all at once. For this reason, only mereological simples and composites that never gain or lose parts (for example, perhaps celestial substances) can persist as numerically the same thing in the strictest sense.24 Since humans do gain and lose parts, they cannot be “wholly” the same in number over time and change. The second conclusion is that the exact thing that is Socrates today is not wholly the same (idem totaliter) with that exact thing which was Socrates yesterday, because some parts have flowed away from that exact thing which was Socrates yesterday and other parts have come in from the outside. But no thing is wholly the same before and after, if anything has been removed or anything has been added. This can be confirmed just as the first [objector] argued:25 that exact thing that was Socrates yesterday will be A, and that which comes to him, given that he is augmented, is called B. It follows that now Socrates is the composite of A and B. Therefore, Socrates [now] is not wholly the same as what is A, and yet yesterday he was wholly the same as that which is A. Therefore, it is clear that Socrates now is not wholly the same as that which Socrates was yesterday.26

If a human or a mundane non-human substance is to persist, it would seem that it can only do so in a less than total sense of numerical sameness. Otherwise, we would be stuck with the absurd consequence that Socrates

23

I will mostly be drawing on Buridan’s discussion of persistence in his questions on Aristotle’s On Generation and Corruption, book 1, question 13. (Text: John Buridan Quaestiones super libros De generatione et corruptione Aristotelis. Edited by Michiel Streijger, Paul J. J. M. Bakker, and Johannes M. M. H. Thijssen. Leiden / Boston: Brill, 2010.) For other discussions of persistence, see Buridan’s Questions on the Physics 1.10, the briefer treatment in his Questions on the Metaphysics 7.12, and the quick summary in his Quaestiones in De anima (ultima lectura) 2.7 (Sobol ed., pp. 100-2). 24 See Buridan In Phys. 1.10, f. 13vb (translated by Pasnau, p. 696). 25 The first objection is this: “The whole is its parts, as it is commonly said. But the parts do not remain the same; rather, they come in and flow away. Therefore, [the proposition under] investigation is false.” 26 Quaestiones super De gen. 1.13 (pp. 112-13).

20

Remarks on Pasnau’s Metaphysical Themes: 1274-1671

today “would have been generated from scratch”.27 For humans, the solution is to fix upon a part of him that does persist in the strictest sense: The third conclusion is that, from the beginning of his life up to the end, a man remains partially the same (idem partialiter), or I should say, [the same] with respect to his noblest and most principal part (that is, with respect to the intellective soul, which always remains wholly the same). And from this we can conclude that speaking in an unqualified way and without anything added that a man remains the same from the beginning of his life to the end. And this is because we customarily denominate, unqualifiedly and without adding anything, a thing by means of its most principal part, and this is especially so if the most principal part is something that manifestly stands out (valde excedens) in the way that the intellective soul stands out from the body.28

Non-human animals and plants, however, cannot even persist in this sense. But I believe that something else should be said about [for example] a horse or a dog. For I believe that this big horse, the exact one here today, even if it were partially the same with that exact one born from a mother’s womb, nevertheless is not the same with respect to its greatest part or even with respect to its most principal one, because in the big horse there is much more of the matter added since he was born than of the matter which was then in him—[and this is true] whether we are speaking of the matter of the head or the heart or the brain or any of the other limbs.29 Moreover, since in the case of material forms (namely, those which are brought out from the potency of the matter) a form does not migrate from [one batch of] matter into [another batch of] matter,30 it follows that there is much more of the substantial form (both in the heart and in the brain) which was not in the newborn than of that [form] which was. And so it follows that if there were partial identity (identitas partialis) between this exact [horse] and that exact [horse], this identity is in virtue of lesser or smaller parts.

27 “from scratch”: de novo (Opp. # 2, p. 112). That is, if Socrates exists today but he did not exist yesterday, then he must have come into existence as if out of thin air. 28 Quaestiones super De gen. 1.13 (p. 113). 29 In all likelihood, Buridan mentions the head, the heart and the brain specifically because these are some of the obvious candidates for a most principal part of an animal. After all, if you remove a horse’s heart (or its head, or its brain) the horse dies, even if the remaining parts are left intact. 30 “migrate from [one batch of] matter into [another batch of] matter”: transeat de materia in materiam.

Andrew W. Arlig

21

And similarly, [in the case of these smaller parts] there is more of diversity than of sameness.31

Brutes and plants, then, persist only in an improper sense, namely, in the way that rivers persist: It follows from this that, in order to see how a horse remains the same in number, we should return to the opinion of Seneca and speak of a horse as [we do] of a river, with this caveat, which Seneca expresses well: A river more rapidly and manifestly passes by and changes (even when considering its greater parts taken all at once), whereas a horse [changes] more slowly and with respect to smaller parts, and hence [it does so] more obscurely, nay, imperceptibly.32 Hence, both the name “Browny” and the name “Seine” are discrete names properly belonging to a quality. And for this reason, it must be conceded that in some manner or other [something] for which [the name] supposits remains the same in number. Moreover, I believe that this [kind of] numerical identity should be considered in virtue of a continuous succession of new parts coming in while previous parts recede, and thus that if I say, “The Seine has lasted for a thousand years,” I mean that some parts have succeeded other parts continuously for a thousand years. And it is thus also the case for a horse or a dog, when this is so: in a succession of this sort, there always remains the same or similar shape. Even if there is not in this case unqualified identity, nevertheless an animal is said, without qualification and anything added, to remain the same by the commoner, to whom the coming and going of parts is not apparent to the senses (especially in the case of living things).33

Buridan is perhaps overstating his case, since the river and the horse in fact seem to exhibit a weaker form of partial identity.34 As the quotation above makes clear, Buridan admits that there may be partial identity between stages in a horse’s life. Our horse Browny does not change parts wholesale from moment to moment. Some of the material parts present at t1 are also present at t2. Thus, in fact, Browny is really a succession of 31

Quaestiones super De gen. 1.13 (pp. 113-14). This is a paraphrase of Seneca. In his Epistulae ad Lucilium, 58.23, Seneca compares a river to a human, not a horse. Both fluctuate, but the river’s fluctuations are more manifest. Hence Seneca is amazed by “our madness”, namely, that “we love the most fleeting of things, the body, and we live in fear that we may at some point die, when every moment is disposed to be a death for the previous one” (OCT p. 158). 33 Quaestiones super De gen. 1.13 (pp. 114-15). 34 See Klima “Buridan on Substantial Unity”, p. 2. (Gyula Klima “Buridan on Substantial Unity and Substantial Concepts.” Paper accessed through Klima’s webpage: http://faculty.fordham.edu/klima/.) 32

22

Remarks on Pasnau’s Metaphysical Themes: 1274-1671

partially identical horses. It also is quite plausible to think that a river does not change parts in a wholesale manner from t1 to t2. Although, in the case of a river, the parts change much more rapidly, and so I grant that it is hard to verify whether we have partial identity between stages or wholesale mereological change.35 At any rate, even if horses are partially identical, they are not identical with respect to the “most principal part”, and hence they do not persist in the way that a human does. Therefore, I will not linger any longer over whether this means that we really have three distinct modes of numerical identity, or merely two, one that is all-ornothing (the kind that corresponds to the standard contemporary interpretation of “ = ”) and one that comes in degrees. No matter whether we have two or three modes of numerical sameness, it is tempting to think that Buridan is drawing a distinction between, on the one hand, a strict and proper sense of “same” and, on the other hand, a “loose and popular” sense.36 Certain things that Pasnau says suggest that he is enticed by this interpretation. What Buridan’s discussion makes clear, however, is that this is one of those instances where the way we talk does not correspond with the 35

This might be why Albert of Saxony suggests that a river is an example of a thing that does change its parts completely from moment to moment. In a fascinating discussion in his questions on Aristotle’s Physics Albert asserts that if God were to create a series of instantaneous Socrateses, rather than create him once and conserve him, then Socrates would be a successive entity (in the way that time is a successive entity, not in Chisholm’s sense): An example of this would be if Socrates were continually made and made again by the First Cause, corresponding to the way in which the Seine continuously flows and flows, so that nothing of the preexisting [river] remains. (Albert of Saxony Quaestiones in Phys. 3.3 [pp. 483-4], translated by Pasnau [p. 393].) Intriguingly, Albert even concedes that a series of Socrateses would be indistinguishable from a permanent substance (p. 484). 36 Both Roderick Chisholm and Donald Baxter have worked with this sort of distinction. In both cases, they take their inspiration from remarks by Bishop Butler. It is unclear whether Butler was inspired by Buridan’s work or other fourteenth-century discussions of persistence. Roderick Chisholm Person and Object: A Metaphysical Study. La Salle, IL: Open Court Publishing, 1976, pp. 92 f. Donald L. M. Baxter, “Identity in the Loose and Popular Sense,” Mind 97 (1988): 575-82, and Donald L. M. Baxter “Loose Identity and Becoming Something Else,” Nous 35 (2001): 592-601.

Andrew W. Arlig

23

metaphysical facts on the ground. It is perfectly legitimate to say, with qualification, that Socrates persists through change—this is legitimate, because our customary idioms allow it. From a metaphysical point of view, however, such claims are liable to mislead, if they are understood as entailing that Socrates wholly survives.37

The last part of what Pasnau says is right. In so far as our idioms suggest to us that Socrates or Browny wholly survives mereologically change, these idioms are deceptive. But I want to suggest that Buridan is not distinguishing between a metaphysically correct sense of “being numerically the same as” and two loose, popular senses of the phrase.38 Now, I grant that Buridan’s choice of terminology is not always helpful. For example, in his treatment of this issue in his Physics commentary, Buridan claims that the third mode of numerical sameness is that something is the same as another “less properly”.39 But I think that on a careful reading of these texts, one will see that Buridan thinks all three senses of numerical sameness are “proper” in the sense that they are rigorously defined notions with more or less precise criteria for application. The only place where custom clearly creeps in is that in many cases it is acceptable to say “This horse is the same one you saw last year” or “This man is the same person you knew as a boy.” That is, it is acceptable to drop the modifier “wholly”, “partially”, or “successively”40 37

Pasnau, p. 698. I have argued for this claim in a recent paper, “Parts, Wholes and Identity”, pp. 457-8. (Andrew Arlig “Parts, Wholes, and Identity,” in John Marenbon ed., The Oxford Handbook of Medieval Philosophy, pp. 445-67. Oxford / New York: Oxford University Press, 2012.) In early drafts of this paper I myself had been tempted to think that Buridan was distinguishing between a strict sense of identity, and looser and popular senses. I was urged to reconsider this notion by Claude Panaccio, Peter King, Henrik Lagerlund, and other members of the audience at Toronto, where I presented a draft of the article. 39 In Phys. 1.10, f. 13vb (Pasnau, p. 696). Compare Albert of Saxony In Phys. 1.8, where the three modes of being the same in number are (1) being the same “properly”, (2) being the same “less properly or partially”, and (3) being the same “improperly on account of the continuous succession of parts in relation to one another” (p. 129). (The text of Albert of Saxony: Benoît Patar (ed.) Expositio et Quaestiones in Aristotelis ‘Physicam’ ad Albertum de Saxonia attributae. Vol. 2. Philosophes Médiévaux 40. Louvain-la-Neuve / Louvain-Paris: Éditions de l’Institut Supérieur de Philosophie / Éditions Peeters, 1999.) 40 Buridan does not use this term to modify the third sense of numerical sameness. I coin it based on the longer qualification that he offers when responding to the opening objections: “…at least that it remains the same when identity is asserted in 38

24

Remarks on Pasnau’s Metaphysical Themes: 1274-1671

when speaking to one’s neighbors or making transactions in the marketplace. One reason to think that Buridan wants to identify more than one legitimately philosophical sense of “being numerically the same” is that, on the one hand, he needs to account for the fact that humans and animals change and so (as even common sense admits) humans are not altogether the same from time to time. But, on the other hand, Buridan must find a way to deny such untoward consequences as ones that are alluded to in the discussion about augmentation and decrease: And in light of these arguments it is not necessary to concede anything more. Nor are certain pronouncements about a human valid, namely, the ones in which it is said that if you are not the same [human] who you had been, you had not been baptized. For it was said that a man does not remain the same unqualifiedly, but [he does] with respect to his most principal part.41

I do not think that Buridan wants to validate the claim that I am the one who was baptized merely by appealing to custom. Rather, as I see it, this claim is true, as are claims about moral responsibility for past actions, for principled metaphysical reasons. Socrates is not the kind of thing that is mereologically changeless. Instead, Socrates is the kind of thing such that, if Socrates’s soul is present, then Socrates is still present and the proper bearer of many important properties. Indeed, to see that Buridan thinks this is a philosophically principled reason, notice that he attributes the position to Aristotle (as well as to the Church): For this reason in books seven and nine of the Ethics Aristotle says that a human is principally an intellect or an intellective soul.42 And thus it is said that a man is your lover if he loves the intellectual part [of you]. And our faith holds this to be true, since we say that Saint Peter is in Paradise (and in the Litany one says, “O’ Saint Peter, light on our behalf…”), even though the bodies of saints have been corrupted and only their souls are in Paradise.43 virtue of [the fact that there is] a continuous succession of parts succeeding one another through a period of time” (Quaestiones super De gen. 1.13 [p. 115]). See also Albert’s formulation (previous note). 41 Quaestiones super De gen. 1.13 (p. 115). 42 Compare Aristotle Nicomachean Ethics, 7.6, 1150a1-4, and 9.8, 1168b31-34. 43 Quaestiones super De gen. 1.13 (p. 113).

Andrew W. Arlig

25

Likewise, I don’t think that Buridan’s view is that metaphysically speaking animals and plants do not persist. Rather, he only needs to say that animals and plants are the kinds of things such that, if an animal of kind K is present at t2 and this animal stands in the right sort of causal and historical relations to an animal of kind K at t1, then the animal at t2 is one and the same (successively speaking) animal as the animal at t1.44 Pasnau is right to point out that Buridan’s theory is at odds with the earlier Scholastic consensus, epitomized by Aquinas and others.45 But I don’t agree with Pasnau that the view is all that much at odds with pretheoretical, commonsense ontology: But here is another cost of the view: one has to say that the dog that grew up from a puppy, or even the man who grew up from a boy, is only partly the same thing that it was. Whereas it seems obvious, at least pretheoretically, that your dog is the very same dog you brought home as a puppy, and your boy the very same boy you brought home as a baby, none of the authors we are considering can allow this.46

Pasnau leans on the notion of being the “very same thing”, suggesting that our commonsense understanding of this relation corresponds more or less with a strict notion of numerical identity. But I suspect that our commonsense notion of being the very same thing is slipperier than that, 44

Gyula Klima and Henrik Lagerlund have drawn a distinction between a potential “ontological problem” and a potential “epistemological problem” for Buridan’s account of non-human, mundane substances. The ontological problem is whether on Buridan’s conception there be any genuine identity over time of animals, plants, and other non-human material substances. The epistemological problem is “whether the ‘toned down’ identity assigned by Buridan to such material substances can serve as an ontological ground for the formation of absolute concepts about them” (Klima “Buridan on Substantial Unity”, p. 1). In Klima’s view, Buridan has a perfectly good response to the ontological problem (which is our concern here), especially once one appreciates the fact that for Buridan, like all medieval authors, “the concept of identity is derivative with regard to the fundamental, transcendental concept of unity” (op cit.). Just as unity comes in degrees, so too sameness comes in degrees. Animals have a lesser degree of unity than humans, but pace Lagerlund (p. 3) it does not follow that animals have a unity no greater than an aggregate (Klima, pp. 2-3). (Henrik Lagerlund “John Buridan’s Empiricism and the Knowledge of Substances,” a paper read at the University of Western Ontario Colloquium (October 9, 2009), accessed through Klima’s webpage: http://faculty.fordham.edu/klima/.) 45 Pasnau, pp. 689-91. 46 Pasnau, p. 701.

26

Remarks on Pasnau’s Metaphysical Themes: 1274-1671

and that it actually is sensitive to another commonsense intuition, material things in fact do change. It is precisely because our notion of “being the very same thing” is fuzzy that puzzles like the Ship of Theseus gain traction. Moreover, I take it that we often do acknowledge without resorting to a whole lot of theory that someone is not precisely the same person she was a year ago, and that the puppy has changed a lot since he was brought home—that is, that he is not exactly the same dog he was back then. Having a fight at the level of intuitions or commonsense is risky, and so I won’t lean too heavily on any specific claim about precisely what our pretheoretical intuitions are. I will note, however, that while a metaphysical system might aim to accommodate as many commonsense intuitions as it can, I can’t think of any respectable system that has managed to retain all of them. Commonsense must give way to theory here or there. As I see it (and as Pasnau concedes), the consensus view keeps some aspects of commonsense ontology, but at the cost of postulating metaphysical parts that are not enshrined in commonsense. (Aquinas’s view is stranger still: when the animal dies, there is not even an animal body left, and a severed hand turns out to not be a hand at all.) Buridan and other so-called nominalists can also claim to capture a good number of our commonsense intuitions about objects and their parts. (For example, once you have shown me all the parts of something, haven’t you shown me the whole? Or, if I sell each parcel of my land, can I still sell the whole that is my land?47) The theory comes into play when we sit down and try to regiment all the various senses in which something is the same or different. And here it does turn out that, perhaps contrary to first impressions, my dog persists like a river.

47

Baxter “Identity in the Loose and Popular Sense”, p. 579.

CATEGORIES AND MODES OF BEING: A DISCUSSION OF ROBERT PASNAU’S METAPHYSICAL THEMES PAUL SYMINGTON

In Metaphysical Themes: 1274-1671, Robert Pasnau describes Aquinas’s and Henry of Ghent’s views on the ontological nature of some of Aristotle’s categories as “structures.”1 Although not based on a medieval term, Pasnau suggests that the term ‘structure’ captures their penchant for reductionism and deflationism about the lesser accidental categories (such as action, passion, and perhaps, position). “Structure” is a useful notion because it is “ontologically innocent: it is an attempt to account for how the world is organized, but without postulating any further items in the world.”2 In this critical paper reflecting on Pasnau’s magisterial and invaluable work, I shall focus on this reductive interpretation of Aquinas’s view of categories. Since there are certainly good textual reasons supporting Pasnau’s interpretation, I shall present a view that focuses on explicit discussions of categories in Aquinas’s corpus. On this basis, I disagree with Pasnau and argue that Aquinas should be considered a nonreductionist and realist regarding categories; or at least Aquinas attempted to achieve this objective. However, to do so, one must grant Aquinas some idiosyncratic approaches to metaphysics (in comparison to later scholastics)—some views which, by the way, I believe may be defendable and philosophically fruitful. This analysis will lead me to close with a broader assessment of Pasnau’s work that may be helpful in thinking about approaches to the history of metaphysics in the later medieval and early modern periods. 1

Robert Pasnau, Metaphysical Themes: 1276-1671 (New York: Oxford University Press, 2011), pp. 229-35. 2 Pasnau, p. 231.

28

Categories and Modes of Being

In his chapter “Real Accidents,” Pasnau identifies Aquinas’s view on accidents as deflationary.3 A deflationary account of accidents maintains that accidents “do not exist in the same sense that substances exist;”4 and that “talk of an accident’s existing is best understood as shorthand for a substance’s existing in a certain way.”5 Although there are stronger formulations of deflationism about accidents that have eliminitivism as their limit,6 Pasnau holds that Aquinas’s view indicates a weaker characterization than eliminitivism because he claims that a thing such as, “whiteness is said to exist not because it subsists in itself, but because by it something has existence-as-white.”7 Yet, Pasnau contends that at least a weakly deflationary account of accidents is appropriately imputed to Aquinas since he held that substance is what properly exists—not accidents—and accidents are ways in which the substance exists accidentally. For example, snow is white because of whiteness.8 Of course, there arises suspicion about any deflationary interpretation due to the seeming metaphysical separability of accidents, even if only under miraculous conditions (such as transubstantiation). Yet, despite Aquinas’s deflationism about accidents, due to the fact that accidents are forms—and in this sense themselves principles of actuality—Aquinas indeed holds that it is metaphysically possible that accidents can be conserved by the power of God even without a subject.9 How this is understood to work is beyond the scope of this paper, but as we shall see, this point about forms as principles of actuality is important for establishing a realist interpretation of Aquinas on categories. Given Aquinas’s deflationary tendency regarding accidents, it may not be surprising that Pasnau interprets him as holding a reductionistic account of 3

Pasnau, pp. 179-99. Pasnau, p. 181. 5 Pasnau, p. 183. 6 Stronger deflationism is encapsulated by Pasnau later when he says the following: “It might seem that either one should endorse accidental forms as metaphysical parts that exist in their own right, as substances do, or else treat them as merely an aspect (a mode?) of the substance. In the latter case, however, it would seem odd to say, as Aquinas seems to, that a substance has multiple existences, substantial and accidental. On a strictly deflationary view, it would seem better to say that only the substance exists,” p. 194. 7 Pasnau, p. 184. Here Pasnau is quoting Aquinas, Quodlibet, 9.2.2. 8 Pasnau, p. 192. This is a partial quote from Aquinas, Summa theologiae, 3.77.1 ad 4. 9 Pasnau, p. 187. 4

Paul Symington

29

some of the categories. As mentioned, Pasnau describes Aquinas’s view of categories as structures. Whereas Aquinas clearly does not have a reductionist view of substance, quantity, quality and perhaps relation, it is possible that any or all of the remaining categories—Place Where, Time When, Position, Having, Action, Passion—do not each pick out any true kind of entity. Along these lines, Pasnau suggests that Aquinas “endorses the idea that each of the categories marks off a distinct kind of being, but without supposing that there is a one-to-one mapping from categories to basic entities.”10 He is led to this from Aquinas’s view that there are cases where one and the same thing can be classified into more than one category. For example, the same change (motus) can fall either under Passion or Action; such as when a single specific event can be expressed either as falling under Action—“Mary built this table”—or under Passion—“This table was built by Mary.” Pasnau suggests that it would be odd for these two sentences to involve different metaphysical commitments since the only difference is between the active and passive voices. Although Aquinas holds that there is a basis in reality between action and passion (“to build” is different from “to be built”), this does not mean there is not some more basic entity that these reduce to: namely, the change itself (which Aquinas sees as being just one actuality for any agent-patient pair). For this reason, Pasnau thinks that the notion of structure is helpful: the lesser categories are ontologically neutral and are fundamental ways in which the world may be arranged without mapping reality at its most fundamental level. In a footnote, Pasnau states that Aquinas holds this structure view for perhaps all categories except Substance, Quantity, and Quality.11 When coupled with the deflationary view of accidents, what Pasnau’s view seems to amount to is that although some categories pick out distinct ways in which the substance exists, between some categories—such as Action and Passion—distinct ways in which the substance exists are not picked out in virtue of a real distinction between them. 10

Pasnau, p. 230. Pasnau, p. 231n. Pasnau cites the following passage from Aquinas to support his view: “[T]he other classes of things are a result of relation rather than a cause of it. For the category when consists in a relation to time; and the category where in a relation to place. And posture implies an arrangement of parts; and having (attire), the relation of the thing having to the things had.” See Aquinas, In duodecim libros Metaphysicorum Aristotelis expositio, ed. M. R. Cathala and R. M. Spiazzi (Rome: Marietti, 1971).

11

30

Categories and Modes of Being

Having introduced Pasnau’s view, I want to broaden the scope of the discussion beyond Metaphysical Themes by examining Aquinas’s ontology of categories. I am doing this in hopes of presenting Aquinas’s view as both deflationary regarding accidents and non-reductionist regarding categories.12 Not only do I think that Aquinas is a nonreductionist about categories, but I also think that such an analysis can serve as a way of offering an assessment of Pasnau’s book. Specifically, I think that a helpful approach to a philosophical interpretation of Aquinas’s metaphysics is found (1) in assuming that he has a lean and yet realist ontology, and (2) that he relies oftentimes on an analysis of cognitive acts to support ontological distinctions. However, in order to accept Aquinas’s view of categories as realist, there are some controversial philosophical points that must be granted to Aquinas, not the least of which is the real distinction between existence and essence, the analogy of being, and a kind of isomorphism between thought and reality. As many of us are aware, each of these themes is subject to misinterpretation and sophistical and incoherent application. The way through which we shall examine categories will be first to reflect on the role that predication plays in Aquinas’s view of categories, and then to discuss how categories are a way of mediating being through distinct essences. The debate over the categories usually orbits around the question about whether they are linguistic, conceptual, or real features of the world. In sizing-up this debate, it is easy to become confused because of linguistic or conceptual approaches that some thirteenth-century scholastics take to identifying the list of categories (and because of the debate over how Aristotle’s Categories relates to the methodological study of logic and metaphysics).13 For example, Aquinas and Albert Magnus both advocate establishing the list of categories by reflecting on various modes of predication. This has led both contemporary and medieval thinkers (such as Scotus) to conclude that as a result of such a technique Aquinas has only succeeded at best in providing a rational distinction of the categories 12

By ‘reductionist’ I mean the view that although a difference can be made among things, this difference does not mark a real distinction in such a way as to pick out two distinct things. 13 See Paul Symington, “Thomas Aquinas on Establishing the Identity of Aristotle’s Categories,” Medieval Commentaries on Aristotle’s Categories, edited by Lloyd Newton, (Boston: Brill Academic Publishers, 2008).

Paul Symington

31

and not a real one.14 Although we shall not take an in-depth look at the role that predication plays in identifying the list and nature of the categories for Aquinas, it is important to identify a previous question about the role that categories play in human cognition. Far from being known in a derivative manner, Aquinas says in De potentia 7.9 that categories are fundamentally ordered to the first things understood by the intellect (prima intellecta), which are things existing extramentally: Because relation is rather weak among all the categories, for this reason, certain men supposed that it was from the second things understood (secundis intellectibus). For the things first understood are things beyond the soul; with respect to cognizing such things, intellect is drawn at first. However, the second things understood are called intentions consequent upon the mode of understanding; in this second (hoc secundo), the intellect understands itself in however much it reflects upon itself, understanding itself to understand, and [understanding] the mode by which it understands. According then to this position, it might follow that relation is not among things beyond the soul, but in the intellect alone, just like the intention of genus and species, and second substances. This, however, cannot be possible. For something is placed in no category unless as a thing (res) existing beyond the soul. For a being of reason is divided against being divided by the ten categories (Meta. 5).15 14 For example, this is John Duns Scotus’s conclusion in his Quaestiones super libros Metaphysicorum Aristotelis, eds. R. Andrews, G. Etzkorn, G. Gàl, R. Green, F. Kelley, G. Marcil, T. Noone, R. Wood, Opera philosophica, Vols. III & IV (St. Bonaventure, N.Y.: The Franciscan Institute, 1997). Agreeing perhaps with Pasnau’s interpretation of Aquinas on the categories, Scotus also seems to hold the notion that establishing only a rational distinction rather than a real one was not done by accident by Aquinas. That is, it is not as if Aquinas desired to provide a real distinction among the categories but only managed to conduct his examination within a rationally distinct scope, but that Aquinas was intending to divide the categories rationally by dividing them via modes of predication. 15 Special thanks to Sarah Wear for essential translation suggestions for the above passage. Cf. Thomas Aquinas, On the Power of God, trans. by English Dominican Fathers (Westminster: The Newman Press, 1952), which continues . . . “Now if relation had no objective reality, it would not be placed among the predicaments. Moreover the perfection and goodness that are in things outside the mind are ascribed not only to something absolute and inherent to things but also to the order between one thing and another: thus the good of an army consists in the mutual ordering of its parts, to which good the Philosopher (Metaph. x) compares the good of the universe. Consequently there must be order in things themselves, and this order is a kind of relation. Wherefore there must be relations in things themselves,

32

Categories and Modes of Being

Here, Aquinas is equating categories with extra-mental things. As first intelligibles, these have cognitive priority to those things that follow upon our understanding. Far from being derivative or dependent on our thought, Aquinas seems to be saying that since categories are related to the first things understood by us, they are grouped-in with that which is cognitively foundational. Importantly, Aquinas also says that “a being of reason is divided against being having been divided by the ten categories.” This is seen in the fact that although a category can be understood as a genus, they are directly predicated of things themselves, whereas the predicate “genus” cannot be. The following is presupposed in the above discussion of Aquinas’s theory of abstraction as regards primary and secondary understandings: that which exists extramentally becomes known by us through the process of abstraction in which the extramental content becomes unified in the mind as independent of the existence conditions of the thing existing extramentally.16 The categories themselves are part of the content contained in the mind upon a primary understanding of things—that which is most general in such an understanding—and as such are identified with whereby one is ordered to another. Now one thing is ordered to another either as to quantity or as to active or passive power: for on these two counts alone can we find in a thing something whereby we compare it with another. For a thing is measured not only by its intrinsic quantity but also in reference to an extrinsic quantity. And again by its active power one thing acts on another, and by its passive power is acted on by another: while by its substance and quality a thing is ordered to itself alone and not to another, except accidentally: namely inasmuch as a quality, substantial form or matter is a kind of active or passive power, and forasmuch as one may ascribe to them a certain kind of quantity: thus one thing produces the same in substance; and one thing produces its like in quality; and number or multitude causes dissimilarity and diversity in the same things; and dissimilarity in that one thing is considered as being more or less so and so than another, thus one thing is said to be whiter than another. Hence the Philosopher (Metaph. v) in giving the species of relations, says that some are based on quantity and some on action and passion. Accordingly things that are ordered to something must be really related to it, and this relation must be some real thing in them. Now all creatures are ordered to God both as to their beginning and as to their end: since the order of the parts of the universe to one another results from the order of the whole universe to God: even as the mutual order of the parts of an army is on account of the order of the whole army to its commander (Metaph. xii). Therefore creatures are really related to God, and this relation is something real in the creature.” 16 See H. Smit, “Aquinas's Abstractionism,” Medieval Philosophy and Theology 10 (2001): 85-118.

Paul Symington

33

the extramental things themselves. This distinguishes categories from logical beings since the latter are secondarily divided against the being that is divided by the ten categories. In this way, there is a priority to the division of the categories in our understanding of the world to the division of our thoughts of them. In conjunction with this prior division, the intellect combines and separates predicates and subjects.17 However, no category identified with extramental things is known in a way fully independent of substance.18 In fact, each accidental category is known concretely in relation to substance, even though each accidental category can be signified independently of it. An accidental essence or form can be distinctly identified from the essence of substance.19 17

See Thomas Aquinas, Sententia libri Metaphysicae, ed. by Robert Busa S. J. (Turin: 1950), lib. 6, l. 4, n. 21 (henceforth, In Met.): “Et alia ratio est, quia utrumque, scilicet ens verum et ens per accidens, sunt circa aliquod genus entis, non circa ens simpliciter per se quod est in rebus; et non ostendunt aliquam aliam naturam entis existentem extra per se entia. Patet enim quod ens per accidens est ex concursu accidentaliter entium extra animam, quorum unumquodque est per se. Sicut grammaticum musicum licet sit per accidens, tamen et grammaticum et musicum est per se ens, quia utrumque per se acceptum, habet causam determinatam. Et similiter intellectus compositionem et divisionem facit circa res, quae sub praedicamentis continentur.” 18 In Met., lib. 9, l. 1, n. 1: “Postquam determinavit philosophus de ente secundum quod dividitur per decem praedicamenta, hic intendit determinare de ente secundum quod dividitur per potentiam et actum. Et dividitur in duas partes. In prima continuat se ad praecedentia, et manifestat suam intentionem in hoc libro. In secunda prosequitur quod intendit, ibi, quod quidem igitur. Dicit ergo primo, quod in praemissis dictum est de ente primo, ad quod omnia alia praedicamenta entis referuntur, scilicet de substantia. Et quod ad substantiam omnia alia referantur sicut ad ens primum, manifestat, quia omnia alia entia, scilicet qualitas, quantitas et huiusmodi dicuntur secundum rationem substantiae. Dicitur enim quantitas ex hoc quod est mensura substantiae, et qualitas ex hoc quod est quaedam dispositio substantiae; similiter in aliis. Et hoc patet ex hoc, quod omnia accidentia habent rationem substantiae, quia in definitione cuiuslibet accidentis oportet ponere proprium subiectum, sicut in definitione simi ponitur nasus. Et hoc declaratum est in praemissis, scilicet in principio septimi.” 19 In Met., lib. 7, l. 1, n. 15: “Quod etiam sit prior ordine cognitionis, patet. Illud enim est primum secundum cognitionem, quod est magis notum et magis manifestat rem. Res autem unaquaeque magis noscitur, quando scitur eius substantia, quam quando scitur eius quantitas aut qualitas. Tunc enim putamus nos maxime scire singula, quando noscitur quid est homo aut ignis, magis quam quando cognoscimus quale est aut quantum, aut ubi, aut secundum aliquod aliud praedicamentum. Quare etiam de ipsis, quae sunt in praedicamentis accidentium,

34

Categories and Modes of Being

Given that predicates are identified with the things of which they are predicated, it is not surprising when Aquinas makes his famous claim in his Commentary on Aristotle’s Metaphysics that categories divide being because “being is said to be in just as many ways as we can make predications.”20 That is, since our concepts are intrinsically ordered by distinct categories, and predicates can be truly predicated of their subjects, Aquinas believed that through a reflection on the essential relations between certain predicates and subjects one can identify the list of Aristotle’s categories; a list which he claims is finite.21 tunc scimus singula, quando de unoquoque scimus quid est. Sicut quando scimus quid est ipsum quale, scimus qualitatem, et quando scimus quid est ipsum quantum, scimus quantitatem. Sicut enim alia praedicamenta non habent esse nisi per hoc quod insunt substantiae, ita non habent cognosci nisi inquantum participant aliquid de modo cognitionis substantiae, quae est cognoscere quid est.” 20 In Met., lib. 5 l. 9 n. 6: “Unde oportet, quod ens contrahatur ad diversa genera secundum diversum modum praedicandi, qui consequitur diversum modum essendi; quia quoties ens dicitur, idest quot modis aliquid praedicatur, toties esse significatur, idest tot modis significatur aliquid esse. Et propter hoc ea in quae dividitur ens primo, dicuntur esse praedicamenta, quia distinguuntur secundum diversum modum praedicandi. Quia igitur eorum quae praedicantur, quaedam significant quid, idest substantiam, quaedam quale, quaedam quantum, et sic de aliis; oportet quod unicuique modo praedicandi, esse significet idem; ut cum dicitur homo est animal, esse significat substantiam. Cum autem dicitur, homo est albus, significat qualitatem, et sic de aliis.” 21 Thomas Aquinas, Expositio libri Posteriorum Analyticorum, Leonine edition (Rome: 1882), lib. 1, l. 34, n. 9 (henceforth, PA): “Circa primum, primo resumit quod de unoquoque possunt aliqua praedicari, quidquid significent: sive sit quale, sive quantum, vel quodcunque aliud genus accidentis, vel etiam quae intrant substantiam rei, quae sunt essentialia praedicata. Secundo, resumit quod haec, scilicet substantialia praedicata, sunt finita. Tertio, resumit quod genera praedicamentorum sunt finita; scilicet quale et quantum et cetera. Si enim aliquis dicat quod quantitas praedicetur de substantia, et qualitas de quantitate, et sic in infinitum; hoc excludit per hoc, quod genera praedicamentorum sunt finita. Quarto, resumit quod, sicut supra expositum est, unum de uno praedicatur in simplici praedicatione. Et hoc ideo inducit, quia posset aliquis dicere quod primo praedicabitur unum de uno, puta de homine animal; et ista praedicatio multiplicabitur quousque poterit inveniri aliquod unum, quod de homine praedicetur. Quibus finitis, praedicabuntur duo de uno: puta, dicetur quod homo est animal album; et sic multo plura praedicata invenirentur secundum diversas combinationes praedicatorum. Rursus, praedicabuntur tria de uno: puta, dicetur quod homo est animal album magnum; et sic semper addendo ad numerum, magis multiplicabuntur praedicata, et erit procedere in infinitum in praedicatis, sicut etiam in additione numerorum. Sed hoc excludit per praedicationem unius de uno.

Paul Symington

35

This allowed Aristotle to generate a logic of categorial predication; such as the rule that two predicates falling under different categories cannot be essentially predicated of each other.22 One of the things that predication shows us is that there are predicates that cannot be predicated essentially of some things that other predicates can be essentially predicated of. For example, ‘color’ can be predicated essentially of ‘whiteness’ but it cannot Quinto, resumit ut non dicamus aliqua simpliciter praedicari de ipsis, quae non aliquid sunt, idest de accidentibus, quorum nullum est aliquid subsistens. De accidente enim neque subiectum neque accidens proprie praedicatur, ut supra dictum est. Omnia enim huiusmodi, quae non sunt aliquid substantiale, sunt accidentia, et de his nihil praedicatur simpliciter loquendo: sed haec quidem praedicantur per se, scilicet de subiectis, vel substantialia praedicata vel accidentalia. Illa vero secundum alium modum, idest per accidens, scilicet cum praedicantur de accidentibus, aut subiecta, aut accidentia. Haec enim omnia, scilicet accidentia, habent de sui ratione quod dicantur de subiecto: illud autem quod est accidens, non est subiectum aliquod; unde nihil proprie loquendo potest de eo praedicari, quia nihil talium, scilicet accidentium, ponimus esse tale, quod dicatur id, quod dicitur, idest quod suscipiat praedicationem eius, quod de eo praedicatur, non quasi aliquid alterum existens, sicut accidit in substantiis. Homo enim dicitur animal vel album, non quia aliquid aliud sit animal vel album, sed quia ipsummet quod est homo, est animal vel album: sed album ideo dicitur homo vel musicum, quia aliquid alterum, scilicet subiectum albi, est homo vel musicum. Sed ipsum accidens inest aliis; et alia, quae praedicantur de accidente, praedicantur de altero, idest de subiecto accidentis; et propter hoc praedicantur de accidente, ut dictum est. Hoc autem introduxit, quia si accidens praedicatur de subiecto, et e converso, et omnia quae accidunt subiecto, praedicentur de se invicem, sequetur quod praedicatio procedat in infinitum, quia uni infinita accidunt.” 22 PA, lib. 1, l. 26, n. 7: “Deinde cum dicit: quod autem contingit etc., manifestat quod supposuerat, scilicet quod, altero extremorum existente in aliquo toto, alterum non sit in eodem, dicens quod manifestum est ex coordinationibus, scilicet praedicamentorum diversorum, quae non commutantur ad invicem. Scilicet quia id quod est in uno praedicamento, non est in altero, manifestum est quod contingat b non esse in toto, in quo est a, aut e converso, quia videlicet contingit unum terminorum accipi in uno praedicamento, in quo non est alius. Sit enim una coordinatio praedicamenti acd, puta praedicamentum substantiae; et alia coordinatio sit bef, puta praedicamentum quantitatis. Si ergo nihil eorum, quae sunt in coordinatione acd, de nullo praedicatur eorum, quae sunt in coordinatione bef; a autem sit in p, quasi in quodam generalissimo, quod sit principium totius primae coordinationis; manifestum est quod b non erit in p, quia sic coordinationes, idest praedicamenta, commutarentur. Similiter autem est si b sit in quodam toto, ut puta in e; manifestum est quod a non erit in e.”

36

Categories and Modes of Being

be essentially predicated of ‘human,’ even though ‘rational’ is predicated of ‘human’ in this way.23 Forgoing a discussion of Aquinas’s derivation of the categories (which I have treated elsewhere24), the more pertinent question for our purposes is whether Aquinas held that the results of such a distinction among the categories yield a division of essences. That is, does he conclude that the categories mark a division of things—a real division—rather than merely a rational division? Scotus, for example, thought that the result of Aquinas’s derivation based on modes of predication is only a rational division at best, because differences in modes of predication are themselves only rationally distinct and do not imply a distinction of essences.25 Pasnau points out that this view is echoed by Ockham’s view of categories: “the linguisticconceptual items that fall into the categories pick out not a distinctive kind of thing, but merely substance and quality in some oblique way.”26 What is implied in denying that Aquinas held that categories are divided essentially is that it is possible for essence x to fall under more than one category. However, Aquinas seems to rule this out when he says that being signifies “the entity of a thing, as divided by the ten categories,” and that 23

PA, lib. 1, l. 33, n. 6: “Deinde cum dicit: quare autem in quod etc. ostendit differentiam praedicatorum per se ad invicem. Et circa hoc duo facit: primo, distinguit praedicata ad invicem secundum diversa genera; secundo, ostendit differentiam praedicatorum; ibi: amplius substantiam quidem et cetera. Dicit ergo primo, quod quia nos praedicari dicimus solum illud, quod praedicatur non secundum aliud subiectum, hoc autem diversificatur secundum decem praedicamenta; sequitur quod omne quod sic praedicatur, praedicetur aut in quod quid est, idest per modum substantialis praedicati, aut per modum qualis, vel quanti, vel alicuius alterius praedicamentorum, de quibus actum est in praedicamentis. Et addit cum unum de uno praedicetur: quia si praedicatum non sit unum sed multa, non poterit praedicatum simpliciter dici quid vel quale; sed forte dicetur simul quale quid, puta si dicam, homo est animal album. Fuit autem necessaria haec additio; quia si multa praedicentur de uno, ita quod multa accipiantur in ratione unius praedicati, poterunt in infinitum praedicationes multiplicari, secundum infinitos modos combinandi praedicata ad invicem. Unde cum quaeritur status in his quae praedicantur, necesse est accipere unum de uno praedicari.” 24 Paul Symington, On Determining What There Is (New Brunswick: Ontos, 2010). 25 See Symington, 47-89. 26 Pasnau, p. 226.

Paul Symington

37

being “is convertible with thing (re).”27 Since the transcendental res signifies the fact that beings have essence,28 what he can be taken to be saying is that any being as divided by the categories is a thing and has an essence. No two essences falling under distinct categories will have any predicates in common (beyond ‘being,’ etc.). In this way, distinct categories are not themselves essences but express things that are essentially distinct. One category is accidental to another, and so one thing cannot result from two, except accidentally.29 However, given the fact that being is analogically predicated of things falling under distinct categories, Aquinas also suggests that things falling under distinct categories have essences in different but related senses. Essence translated into the language of predication is definition. Definition demarks the “whatness” of a thing signified. When we look at how accidental things are understood essentially through predication, we see that a predicate can be concrete or abstract. Whereas the category of substance is predicated of concrete things (e.g., “Socrates is a substance”), accidental categories are predicated of abstract nouns (e.g., “Wisdom is a quality”). ‘Wisdom’ signifies the category of quality independently of its subject of inherence. Thus, quiddity applies to other categories because it 27

Thomas Aquinas, Summa theologiae, Leonine edition (Rome: 1888), 1.48.2 ad 2 (henceforth, ST): “Ad secundum dicendum quod, sicut dicitur in V Metaphys., ens dupliciter dicitur. Uno modo, secundum quod significat entitatem rei, prout dividitur per decem praedicamenta, et sic convertitur cum re. Et hoc modo, nulla privatio est ens, unde nec malum. Alio modo dicitur ens, quod significat veritatem propositionis, quae in compositione consistit, cuius nota est hoc verbum est, et hoc est ens quo respondetur ad quaestionem an est. Et sic caecitatem dicimus esse in oculo, vel quamcumque aliam privationem. Et hoc modo etiam malum dicitur ens. Propter huius autem distinctionis ignorantiam, aliqui, considerantes quod aliquae res dicuntur malae, vel quod malum dicitur esse in rebus, crediderunt quod malum esset res quaedam.” 28 Thomas Aquinas, Quaestiones disputatae de veritate, Leonine edition (Rome: 1970), 1.1 (henceforth, De veritate). 29 De potentia, 2.2 arg. 2: “Sed dicitur, quod significat simul essentiam et notionem. Sed contra, in divinis, secundum Boetium, sunt haec duo praedicamenta; substantia, ad quam pertinet essentia; et ad aliquid, ad quod pertinent notionalia. Non potest autem aliquid esse in duobus praedicamentis, quia homo albus non est aliquid unum nisi per accidens, ut habetur V Metaph. Ergo potentia generandi non potest in sua ratione utrumque complecti, scilicet substantiam et notionem. Also, De potentia, 2.2 ad 2: “Ad secundum dicendum, quod in rebus creatis unum praedicamentum accidit alteri, propter quod non potest ex duobus fieri unum, nisi unum per accidens; sed in divinis relatio est realiter ipsa essentia: et ideo non est simile.”

38

Categories and Modes of Being

makes sense to ask what something is.30 In this way, all ten categories are essentially distinct. However, predicates that signify accidental categories have their concrete form predicated of the primary substances in which they inhere (e.g., “Socrates is wise”). The concrete sense is important since it is directly applicable to, and abstracted from, fundamental things existing outside of the mind. When considering accidents in a concrete sense, there are differences when discussing the whatness of each accidental item per category. This is because the concrete term ‘wise’ signifies a subject insofar as it signifies wisdom after the mode of an accident.31 An accident, although when signified abstractly does not include the subject in its signification, when signified concretely, depends on, and is individuated by, its subject.32 All accidental essences are referred to substance as a primary kind of being because accidents involve the ratio of substance. For example, the ratio of quantity (considered in relation to a concrete predicate) includes the notion that it is the measure of substance and quality includes the notion that it is the disposition of substance.33 In this way, accidents do not have a strict whatness as 30

In Met., lib. 7, l. 4, n. 2: “Quod enim aliquo modo, idest secundum quid aliis conveniat quid est, ex hoc patet, quod in singulis praedicamentis respondetur aliquid ad quaestionem factam per quid. Interrogamus enim de quali sive qualitate quid est, sicut quid est albedo, et respondemus quod est color. Unde patet, quod qualitas est de numero eorum, in quibus est quod quid est.” 31 In Met., lib. 5, l. 9, n. 10: “Nec est verum quod Avicenna dicit, quod praedicata, quae sunt in generibus accidentis, principaliter significant substantiam, et per posterius accidens, sicut hoc quod dico album et musicum. Nam album ut in praedicamentis dicitur, solam qualitatem significat. Hoc autem nomen album significat subiectum ex consequenti, inquantum significat albedinem per modum accidentis. Unde oportet, quod ex consequenti includat in sui ratione subiectum. Nam accidentis esse est inesse. Albedo enim etsi significet accidens, non tamen per modum accidentis, sed per modum substantiae. Unde nullo modo consignificat subiectum. Si enim principaliter significaret subiectum, tunc praedicata accidentalia non ponerentur a philosopho sub ente secundum se, sed sub ente secundum accidens. Nam hoc totum, quod est homo albus, est ens secundum accidens, ut dictum est.” 32 In Met., lib. 7, l. 4: “Quare sic quidem, idest simpliciter per prius, nullius erit definitio nisi substantiae, nec etiam quod quid erat esse. Sic autem, idest secundum quid et posterius, erit etiam aliorum.” I think that it is apt to say that concrete essences are signified as possessing both the formal component and the individuating principle, whereas when signified abstractly what is signified is the form only, which positively excludes its principle of individuation. 33 In Met., lib. 9, l. 1, n. 1: “Postquam determinavit philosophus de ente secundum quod dividitur per decem praedicamenta, hic intendit determinare de ente secundum quod dividitur per potentiam et actum. Et dividitur in duas partes. In

Paul Symington

39

substance does but the nine accidental categories have whatness and essence in an analogous way.34 Despite this condition, each category can be signified abstractly in a way independent of that in which it inheres; its form can be signified independently from its principle of individuation. This indicates that there is some essential content not reducible to the essential content of its subject of inherence. Having discussed the categories as essential divisions of things, we next turn to a discussion of categories as a division of being. The connection between being and essence is of course Aquinas’s view that that by which and through which something exists is its essence. Thus, Aquinas says that “being is divided into ten categories as considered absolutely,”35 and that prima continuat se ad praecedentia, et manifestat suam intentionem in hoc libro. In secunda prosequitur quod intendit, ibi, quod quidem igitur. Dicit ergo primo, quod in praemissis dictum est de ente primo, ad quod omnia alia praedicamenta entis referuntur, scilicet de substantia. Et quod ad substantiam omnia alia referantur sicut ad ens primum, manifestat, quia omnia alia entia, scilicet qualitas, quantitas et huiusmodi dicuntur secundum rationem substantiae. Dicitur enim quantitas ex hoc quod est mensura substantiae, et qualitas ex hoc quod est quaedam dispositio substantiae; similiter in aliis. Et hoc patet ex hoc, quod omnia accidentia habent rationem substantiae, quia in definitione cuiuslibet accidentis oportet ponere proprium subiectum, sicut in definitione simi ponitur nasus. Et hoc declaratum est in praemissis, scilicet in principio septimi.” 34 In Met., lib. 7, l. 4, n. 1: “Hic ponit secundam solutionem propositae quaestionis: et circa hoc tria facit. Primo ponit solutionem. Secundo probat eam, ibi, illud autem palam, et cetera. Tertio removet quasdam dubitationes, quae possent ex praedictis oriri, ibi, habet autem dubitationem. Circa primum duo facit. Primo ostendit quomodo definitio et quod quid est invenitur in substantia et accidentibus. Secundo quomodo de utrisque praedicetur, ibi, oportet quidem igitur intendere. Dicit ergo primo, quod dicendum est, sicut in praedicta solutione est dictum, quod quod quid est et definitio non sit accidentium, sed substantiarum: aut oportet secundum alium modum solvendi dicere, quod definitio dicitur multipliciter sicut et quod quid est. Ipsum enim quod quid est, uno modo significat substantiam et hoc aliquid. Alio modo significat singula aliorum praedicamentorum, sicut qualitatem et quantitatem et alia huiusmodi talia. Sicut autem ens praedicatur de omnibus praedicamentis, non autem similiter, sed primum de substantia, et per posterius de aliis praedicamentis, ita et quod quid est, simpliciter convenit substantiae, aliis autem alio modo, idest secundum quid.” 35 In Met., lib. 5, l. 9, n. 1: “Hic philosophus distinguit quot modis dicitur ens. Et circa hoc tria facit. Primo distinguit ens in ens per se et per accidens. Secundo distinguit modos entis per accidens, ibi, secundum accidens quidem et cetera. Tertio modos entis per se, ibi, secundum se vero. Dicit ergo, quod ens dicitur quoddam secundum se, et quoddam secundum accidens. Sciendum tamen est quod

40

Categories and Modes of Being

each thing falling distinctly under each of the ten categories is a complete being (ens perfectum).36 Sometimes he refers to entities falling under distinct categories as ens secundum se because they exist and have essence not reducible to the essence of substance.37 Since there are ten categories, each with independently signifiable essences that are classed by them, the single act of being that actualizes a substance is diversified qua beings in proportion to these essences. For this reason, any two essences falling under distinct categories will each be called distinct beings. As Aquinas says, accidents “have a proper mode of being in their proper essence…. In view of the fact that all accidents are forms of a sort superadded to the substance and caused by the principles of the substance, it must be that their being is superadded to the being of the substance and dependent on that being.”38 However, we must consider this in relation to the fact that illa divisio entis non est eadem cum illa divisione qua dividitur ens in substantiam et accidens. Quod ex hoc patet, quia ipse postmodum, ens secundum se dividit in decem praedicamenta, quorum novem sunt de genere accidentis. Ens igitur dividitur in substantiam et accidens, secundum absolutam entis considerationem, sicut ipsa albedo in se considerata dicitur accidens, et homo substantia. Sed ens secundum accidens prout hic sumitur, oportet accipi per comparationem accidentis ad substantiam. Quae quidem comparatio significatur hoc verbo, est, cum dicitur, homo est albus. Unde hoc totum, homo est albus, est ens per accidens. Unde patet quod divisio entis secundum se et secundum accidens, attenditur secundum quod aliquid praedicatur de aliquo per se vel per accidens. Divisio vero entis in substantiam et accidens attenditur secundum hoc quod aliquid in natura sua est vel substantia vel accidens.” 36 In Met., lib. 5, l. 9, n. 5: “Deinde cum dicit secundum se distinguit modum entis per se: et circa hoc tria facit. Primo distinguit ens, quod est extra animam, per decem praedicamenta, quod est ens perfectum. Secundo ponit alium modum entis, secundum quod est tantum in mente, ibi, amplius autem et esse significat. Tertio dividit ens per potentiam et actum: et ens sic divisum est communius quam ens perfectum. Nam ens in potentia, est ens secundum quid tantum et imperfectum, ibi, amplius esse significat et ens. Dicit ergo primo, quod illa dicuntur esse secundum se, quaecumque significant figuras praedicationis. Sciendum est enim quod ens non potest hoc modo contrahi ad aliquid determinatum, sicut genus contrahitur ad species per differentias. Nam differentia, cum non participet genus, est extra essentiam generis. Nihil autem posset esse extra essentiam entis, quod per additionem ad ens aliquam speciem entis constituat: nam quod est extra ens, nihil est, et differentia esse non potest. Unde in tertio huius probavit philosophus, quod ens, genus esse non potest.” 37 Ibid. 38 Thomas Aquinas, Summa Contra Gentiles, trans. by Charles J. O’Neil, Book IV: Salvation (South Bend: University of Notre Dame Press, 1989). Cf. Thomas Aquinas, Summa contra gentiles, Leonine edition (Rome: 1961), lib. 4, cap. 14, n.

Paul Symington

41

for Aquinas things are essentially distinct from each other—even within one and the same substance. There is the one being of the individually existing substance but that one being is directed and actualized according to the accidental essences inhering in it. As a result, there arises the deflationary view that accidents are ways in which the substance exists accidentally. However, there is more to the story. Aquinas sees the real distinction between essence and existence as having a strong role for a full understanding of the categories. The content expressed by the predicate itself is derived from abstraction from real things and is in itself independent of existential content. For example, the predicate ‘animal’ is understood independently of how humans exist—such as individuals or as contingent beings, etc. For this reason, Aquinas holds that a quiddity can 12: “Quamvis autem in Deo ponatur esse relatio, non tamen sequitur quod in Deo sit aliquid habens esse dependens. In nobis enim relationes habent esse dependens, quia earum esse est aliud ab esse substantiae: unde habent proprium modum essendi secundum propriam rationem, sicut et in aliis accidentibus contingit. Quia enim omnia accidentia sunt formae quaedam substantiae superadditae, et a principiis substantiae causatae; oportet quod eorum esse sit superadditum supra esse substantiae, et ab ipso dependens; et tanto uniuscuiusque eorum esse est prius vel posterius, quanto forma accidentalis, secundum propriam rationem, fuerit propinquior substantiae vel magis perfecta. Propter quod et relatio realiter substantiae adveniens et postremum et imperfectissimum esse habet: postremum quidem, quia non solum praeexigit esse substantiae, sed etiam esse aliorum accidentium, ex quibus causatur relatio, sicut unum in quantitate causat aequalitatem, et unum in qualitate similitudinem; imperfectissimum autem, quia propria relationis ratio consistit in eo quod est ad alterum, unde esse eius proprium, quod substantiae superaddit, non solum dependet ab esse substantiae, sed etiam ab esse alicuius exterioris. Haec autem in divinis locum non habent: quia non est in Deo aliquod aliud esse quam substantiae; quicquid enim in Deo est, substantia est. Sicut igitur esse sapientiae in Deo non est esse dependens a substantia, quia esse sapientiae est esse substantiae; ita nec esse relationis est esse dependens neque a substantia, neque ab alio exteriori, quia etiam esse relationis est esse substantiae. Non igitur per hoc quod relatio in Deo ponitur, sequitur quod sit in eo aliquod esse dependens; sed solum quod in Deo sit respectus aliquis, in quo ratio relationis consistit; sicut ex hoc quod sapientia in Deo ponitur, non sequitur quod sit in eo aliquid accidentale, sed solum perfectio quaedam in qua ratio sapientiae consistit. Per quod etiam patet quod ex imperfectione quae in relationibus creatis esse videtur, non sequitur quod personae divinae sint imperfectae, quae relationibus distinguuntur: sed sequitur quod divinarum personarum minima sit distinctio.”

42

Categories and Modes of Being

be within a category only if it is not the same as its existence.39 This is because things are contained in a category only with respect to their common nature or essence. This allows for things like material and immaterial substances—things with different modes of being—to be both contained under the category of substance.40 The way in which a material substance exists is fundamentally different from the way in which an immaterial substance—like an angel—exists. This also leaves open the possibility of sorting things not only according to the logic of their common natures—under which the categories fall—but also according to their individual natures or acts of existence: “Two things in the same category can still be diverse in the sense that they have diverse first subjects. He [Aristotle] says that the diversity of the categories from the predication of being is considered by the logician because it is conceptual.”41 This last points us to Aquinas’s view of accidental entities according to the modality of existence and as modes of being. Not only does each accident have its own essential content, but each also has its own way in which it is found to exist or found in reality. Whereas on one hand, a quality is essentially a disposition of substance, on the other hand, it is found in 39

Thomas Aquinas, De ente et essentia, Leonine edition (Rome: 1976), cap. 4 (henceforth, DEE): “Et quia in istis substantiis quiditas non est idem quod esse, ideo sunt ordinabiles in praedicamento, et propter hoc inventiur in eis genus et species et differentia, quamvis earum differentiae propriae nobis occultae sint.” 40 ST, 88.2 ad 4: “Ad quartum dicendum quod substantiae immateriales creatae in genere quidem naturali non conveniunt cum substantiis materialibus, quia non est in eis eadem ratio potentiae et materiae, conveniunt tamen cum eis in genere logico, quia etiam substantiae immateriales sunt in praedicamento substantiae, cum earum quidditas non sit earum esse. Sed Deus non convenit cum rebus materialibus neque secundum genus naturale, neque secundum genus logicum, quia Deus nullo modo est in genere, ut supra dictum est. Unde per similitudines rerum materialium aliquid affirmative potest cognosci de Angelis secundum rationem communem, licet non secundum rationem speciei; de Deo autem nullo modo.” 41 Thomas Aquinas, Commentary on Aristotle’s Metaphysics, trans. by John P. Rowan (Notre Dame: Dumb Ox Books, 1995). Cf. In Met., lib. 5, l. 22, n. 9: “Patet autem ex dictis quod aliqua continentur sub uno praedicamento, et sunt unum genere hoc modo secundo, quae tamen sunt diversa genere primo modo. Sicut corpora caelestia et elementaria, et colores, et sapores. Primus autem modus diversitatis secundum genus consideratur magis a naturali, et etiam a philosopho, quia est magis realis. Secundus autem modus consideratur a logico, quia est rationis.”

Paul Symington

43

reality as existing in substance. For this reason, accidents do not exist independently of the being of the substance.42 However, Aquinas holds that not only does each categorial thing classify distinct sets of essences but distinct sets of modes of being as well. (Each distinction can be used to classify or identify each of the categories.) For example, although relations have their own essence, they also have the mode of being of “being to something.”43 Also, quantity exists in such a distinct way from otherwise categorially classified entities insofar as they can themselves be the subject of distinct categorial entities.44 Cribbing Augustine, Aquinas defines ‘mode’ as “that which a measure determines: wherefore it implies a certain determination according to a certain measure.”45 In this way, modes of being 42

Thomas Aquinas, In libros De generatione et corruptione, Leonine edition (Rome: 1886), lib. 1, l. 6, n. 6: “Secundo ibi: si quidem primum etc., ostendit quod secundum utrumque sensum sequitur inconveniens. Si enim simpliciter dicatur primum ens quod est substantia, ergo et simpliciter non ens dicetur non substantia. Si ergo generatio simplex hoc requirit, quod sit simpliciter entis ex simpliciter non ente, sequetur quod erit substantia ex non substantia. Sed quando ponitur non esse substantiam neque hoc (quod est demonstrativum individualis substantiae), manifestum est quod nullum aliorum praedicamentorum remanebit, idest neque quale neque quantum neque ubi: quia sequeretur quod passiones, idest accidentia, separarentur a substantiis, quod est impossibile. Si autem dicatur quod illud ex quo aliquid generatur simpliciter, sit non ens universaliter, prout ens simpliciter dicitur ens commune, sequetur quod per hoc quod dicitur non ens, intelligatur universaliter negatio omnium entium. Unde sequetur quod illud quod generatur simpliciter, generetur penitus ex nihilo: quod est contra rationem naturalis generationis, et contra sententias omnium philosophorum naturalium, qui scilicet de generatione naturali locuti sunt.” 43 De veritate, 21.1 arg. 3: “Sed dicebat, quod addit respectum ad finem.- Sed contra: secundum hoc enim bonum nihil aliud esset quam ens relatum. Sed ens relatum concernit determinatum genus entis, quod est ad aliquid. Ergo bonum est in aliquo uno praedicamento determinato; quod est contra philosophum in I Ethic., ubi ponit bonum in omnibus generibus.” 44 Thomas Aquinas, Super De Trinitate, ed. by Bruno Decker (Boston: Brill, 1959), 2.4.2 arg. 6: “Praeterea, posterius numquam est causa prioris. Sed inter omnia accidentia primum locum tenet quantitas, ut dicit Boethius in commento praedicamentorum. Inter quantitates autem naturaliter numerus prior est, cum sit simplicior et magis abstractus. Ergo impossibile est quod aliquod aliud accidens sit principium pluralitatis secundum numerum.” 45 ST, 2.1.49.2 co.: “Respondeo dicendum quod philosophus, in praedicamentis, ponit inter quatuor species qualitatis primam, dispositionem et habitum. Quarum quidem specierum differentias sic assignat Simplicius, in commento praedicamentorum, dicens quod qualitatum quaedam sunt naturales, quae secundum naturam insunt, et semper, quaedam autem sunt adventitiae, quae ab extrinseco efficiuntur, et possunt amitti. Et haec quidem, quae sunt

44

Categories and Modes of Being

are both measures of being of the substance as well as a determination of the substance itself. Accordingly, the being of a substance is divided into adventitiae, sunt habitus et dispositiones, secundum facile et difficile amissibile differentes. Naturalium autem qualitatum quaedam sunt secundum id quod aliquid est in potentia, et sic est secunda species qualitatis. Quaedam vero secundum quod aliquid est in actu, et hoc vel in profundum, vel secundum superficiem. Si in profundum quidem, sic est tertia species qualitatis, secundum vero superficiem, est quarta species qualitatis, sicut figura et forma, quae est figura animati. Sed ista distinctio specierum qualitatis inconveniens videtur. Sunt enim multae figurae et qualitates passibiles non naturales, sed adventitiae, et multae dispositiones non adventitiae, sed naturales, sicut sanitas et pulchritudo et huiusmodi. Et praeterea hoc non convenit ordini specierum, semper enim quod naturalius est, prius est. Et ideo aliter accipienda est distinctio dispositionum et habituum ab aliis qualitatibus. Proprie enim qualitas importat quendam modum substantiae. Modus autem est, ut dicit Augustinus, super Gen. ad litteram, quem mensura praefigit, unde importat quandam determinationem secundum aliquam mensuram. Et ideo sicut id secundum quod determinatur potentia materiae secundum esse substantiale dicitur qualitas quae est differentia substantiae; ita id secundum quod determinatur potentia subiecti secundum esse accidentale, dicitur qualitas accidentalis, quae est etiam quaedam differentia, ut patet per philosophum in V Metaphys. Modus autem sive determinatio subiecti secundum esse accidentale, potest accipi vel in ordine ad ipsam naturam subiecti; vel secundum actionem et passionem quae consequuntur principia naturae, quae sunt materia et forma; vel secundum quantitatem. Si autem accipiatur modus vel determinatio subiecti secundum quantitatem, sic est quarta species qualitatis. Et quia quantitas, secundum sui rationem, est sine motu, et sine ratione boni et mali; ideo ad quartam speciem qualitatis non pertinet quod aliquid sit bene vel male, cito vel tarde transiens. Modus autem sive determinatio subiecti secundum actionem et passionem, attenditur in secunda et tertia specie qualitatis. Et ideo in utraque consideratur quod aliquid facile vel difficile fiat, vel quod sit cito transiens aut diuturnum. Non autem consideratur in his aliquid pertinens ad rationem boni vel mali, quia motus et passiones non habent rationem finis, bonum autem et malum dicitur per respectum ad finem. Sed modus et determinatio subiecti in ordine ad naturam rei, pertinet ad primam speciem qualitatis, quae est habitus et dispositio, dicit enim philosophus, in VII Physic., loquens de habitibus animae et corporis, quod sunt dispositiones quaedam perfecti ad optimum; dico autem perfecti, quod est dispositum secundum naturam. Et quia ipsa forma et natura rei est finis et cuius causa fit aliquid, ut dicitur in II Physic. ideo in prima specie consideratur et bonum et malum; et etiam facile et difficile mobile, secundum quod aliqua natura est finis generationis et motus. Unde in V Metaphys. philosophus definit habitum, quod est dispositio secundum quam aliquis disponitur bene vel male. Et in II Ethic. dicit quod habitus sunt secundum quos ad passiones nos habemus bene vel male. Quando enim est modus conveniens naturae rei, tunc habet rationem boni, quando autem non convenit, tunc habet rationem mali. Et quia natura est id quod primum consideratur in re, ideo habitus ponitur prima species qualitatis.”

Paul Symington

45

ten categories according to diverse modes of existence and these diverse modes are the ultimate determination that “this” is “that.”46 This is 46

Thomas Aquinas, Commentaria in octo libros Physicorum, Leonine edition (Turin: 1954), lib. 3, l. 5, n. 15 (henceforth, In Phys.): “Ad horum igitur evidentiam sciendum est quod ens dividitur in decem praedicamenta non univoce, sicut genus in species, sed secundum diversum modum essendi. Modi autem essendi proportionales sunt modis praedicandi. Praedicando enim aliquid de aliquo altero, dicimus hoc esse illud: unde et decem genera entis dicuntur decem praedicamenta. Tripliciter autem fit omnis praedicatio. Unus quidem modus est, quando de aliquo subiecto praedicatur id quod pertinet ad essentiam eius, ut cum dico Socrates est homo, vel homo est animal; et secundum hoc accipitur praedicamentum substantiae. Alius autem modus est quo praedicatur de aliquo id quod non est de essentia eius, tamen inhaeret ei. Quod quidem vel se habet ex parte materiae subiecti, et secundum hoc est praedicamentum quantitatis (nam quantitas proprie consequitur materiam: unde et Plato posuit magnum ex parte materiae); aut consequitur formam, et sic est praedicamentum qualitatis (unde et qualitates fundantur super quantitatem, sicut color in superficie, et figura in lineis vel in superficiebus); aut se habet per respectum ad alterum, et sic est praedicamentum relationis (cum enim dico homo est pater, non praedicatur de homine aliquid absolutum, sed respectus qui ei inest ad aliquid extrinsecum). Tertius autem modus praedicandi est, quando aliquid extrinsecum de aliquo praedicatur per modum alicuius denominationis: sic enim et accidentia extrinseca de substantiis praedicantur; non tamen dicimus quod homo sit albedo, sed quod homo sit albus. Denominari autem ab aliquo extrinseco invenitur quidem quodammodo communiter in omnibus, et aliquo modo specialiter in iis quae ad homines pertinent tantum. Communiter autem invenitur aliquid denominari ab aliquo extrinseco, vel secundum rationem causae, vel secundum rationem mensurae; denominatur enim aliquid causatum et mensuratum ab aliquo exteriori. Cum autem quatuor sint genera causarum, duo ex his sunt partes essentiae, scilicet materia et forma: unde praedicatio quae posset fieri secundum haec duo, pertinet ad praedicamentum substantiae, utpote si dicamus quod homo est rationalis, et homo est corporeus. Causa autem finalis non causat seorsum aliquid ab agente: intantum enim finis habet rationem causae, inquantum movet agentem. Remanet igitur sola causa agens a qua potest denominari aliquid sicut ab exteriori. Sic igitur secundum quod aliquid denominatur a causa agente, est praedicamentum passionis, nam pati nihil est aliud quam suscipere aliquid ab agente: secundum autem quod e converso denominatur causa agens ab effectu, est praedicamentum actionis, nam actio est actus ab agente in aliud, ut supra dictum est. Mensura autem quaedam est extrinseca et quaedam intrinseca. Intrinseca quidem sicut propria longitudo uniuscuiusque et latitudo et profunditas: ab his ergo denominatur aliquid sicut ab intrinseco inhaerente; unde pertinet ad praedicamentum quantitatis. Exteriores autem mensurae sunt tempus et locus: secundum igitur quod aliquid denominatur a tempore, est praedicamentum quando; secundum autem quod denominatur a loco, est praedicamentum ubi et situs, quod addit supra ubi ordinem partium in loco. Hoc autem non erat

46

Categories and Modes of Being

consistent with Aquinas’s view that accidents make the substance to exist accidentally in some way. Aquinas says that “from an accident and a subject follows accidental existence when the accident joins with the subject.”47 We can next address the topic of the composition of categorially distinct entities. That is, by addressing the real distinction between the modes of being and essence of categorial things we can now see how they exist in composition. The logic of composition of accidents and substances is best expressed by Aquinas in his Commentary on the Metaphysics, where he shows us how accidents can be conceived concretely or abstractly: Now a subject is given directly in the definition of an accident when an accident is signified concretely as an accident fused with a subject, as when I say that snubness is a concave nose; for nose is given in the definition of snub as a genus in order to signify that accidents subsist only necessarium addi ex parte temporis, cum ordo partium in tempore in ratione temporis importetur: est enim tempus numerus motus secundum prius et posterius. Sic igitur aliquid dicitur esse quando vel ubi per denominationem a tempore vel a loco. Est autem aliquid speciale in hominibus. In aliis enim animalibus natura dedit sufficienter ea quae ad conservationem vitae pertinent, ut cornua ad defendendum, corium grossum et pilosum ad tegendum, ungulas vel aliquid huiusmodi ad incedendum sine laesione. Et sic cum talia animalia dicuntur armata vel vestita vel calceata, quodammodo non denominantur ab aliquo extrinseco, sed ab aliquibus suis partibus. Unde hoc refertur in his ad praedicamentum substantiae: ut puta si diceretur quod homo est manuatus vel pedatus. Sed huiusmodi non poterant dari homini a natura, tum quia non conveniebant subtilitati complexionis eius, tum propter multiformitatem operum quae conveniunt homini inquantum habet rationem, quibus aliqua determinata instrumenta accommodari non poterant a natura: sed loco omnium inest homini ratio, qua exteriora sibi praeparat loco horum quae aliis animalibus intrinseca sunt. Unde cum homo dicitur armatus vel vestitus vel calceatus, denominatur ab aliquo extrinseco, quod non habet rationem neque causae, neque mensurae: unde est speciale praedicamentum, et dicitur habitus. Sed attendendum est quod etiam aliis animalibus hoc praedicamentum attribuitur, non secundum quod in sua natura considerantur, sed secundum quod in hominis usum veniunt; ut si dicamus equum phaleratum vel sellatum seu armatum.” 47 DEE, cap. 5: “Et hoc ideo est, quia non habent per se esse, absolutum a subiecto, sed sicut ex forma et materia relinquitur esse substantiale, quando componuntur, ita ex accidente et subiecto relinquitur esse accidentale, quando accidens subiecto advenit.”

Paul Symington

47

in a subject. But when an accident is signified in the abstract, after the manner of a substance, then the subject is given in its definition indirectly, as a difference, as it is said that snubness is the concavity of a nose.48

Amplifying this is the following difficult passage from the De ente et essentia where Aquinas brings in the notion of modes of being in the determination of the composition of accident and substance: And because accidents are not composed of matter and form, their genus cannot be taken from matter and their difference from form, as in the case of composed substances. Rather, their first genus must be taken from their way of existing itself, according to which the word “being” is diversely predicated of the ten genera according to a priority and posteriority; for example, an accident is called quantity from the fact that it is the measure of substance, and quality according as it is the disposition of substance, and so with the other accidents, according to the Philosopher in the fourth book of the Metaphysics. But their differences are taken from the diversity of the principles by which they are caused. And because proper attributes are caused by the proper principles of the subject, the subject is placed in their definition to function as the difference if they are defined in the abstract, which is the way in which they are properly in a genus; as when it is said 48

Thomas Aquinas, On Being and Essence, trans. by A. Maurer (Toronto: Pontifical Institute of Medieval Studies, 1968). Cf. DEE, cap. 5: “Et quia accidentia non componuntur ex materia et forma, ideo non potest in eis sumi genus a materia et differentia a forma sicut in substantiis compositis, sed oportet ut genus primum sumatur ex ipso modo essendi, secundum quod ens diversimode secundum prius et posterius de decem generibus praedicatur; sicut dicitur quantitas ex eo quod est mensura substantiae, et qualitas secundum quod est dispositio substantiae, et sic de aliis secundum philosophum IX metaphysicae. Differentiae vero in eis sumuntur ex diversitate principiorum, ex quibus causantur. Et quia propriae passiones ex propriis principiis subiecti causantur, ideo subiectum ponitur in diffinitione eorum loco differentiae, si in abstracto diffiniuntur secundum quod sunt proprie in genere, sicut dicitur quod simitas est nasi curvitas. Sed e converso esset, si eorum diffinitio sumeretur secundum quod concretive dicuntur. Sic enim subiectum in eorum diffinitione poneretur sicut genus, quia tunc diffinirentur per modum substantiarum compositarum, in quibus ratio generis sumitur a materia, sicut dicimus quod simum est nasus curvus. Similiter etiam est, si unum accidens alterius accidentis principium sit, sicut principium relationis est actio et passio et quantitas; et ideo secundum haec dividit philosophus relationem in V metaphysicae. Sed quia propria principia accidentium non semper sunt manifesta, ideo quandoque sumimus differentias accidentium ex eorum effectibus, sicut congregativum et disgregativum dicuntur differentiae coloris, quae causantur ex abundantia vel paucitate lucis, ex quo diversae species colorum causantur.”

48

Categories and Modes of Being that snubnosedness is the turned-up-ness of the nose. But the converse would be the case if their definition were taken according as they are said concretely. For in this way the subject is placed in their definition as a genus because they are then being defined after the manner of composed substance, in which the genus is taken from matter; as when we say that a snub nose is a turned up nose.49

Aquinas argues that insofar as it is a category, every accidental category has a genus-difference-species ordering, in a way similar to the example of how “continuous” and “discrete” are differences of quantity, except that the category itself is composed as a genus-difference union. Between these two passages, there seems to be four ways in which this can occur: in (1) pseudo-defining concrete accidents in which (1a) the subject serves as the genus and the categorial ordered accidental essence serves as its difference (e.g., snub contains the genus “nose” as differentiated by “snubness”); (1b) the mode of being of the accident serves as the genus and the accidental essence serves as the difference (e.g., existing snubness contains the mode of being inesse,50 which serves as the genus as differentiated by “snubness”); (2) pseudo-defining abstract categorially classed accidents in which (2a) the categorially classed accidental essence serves as the genus and the subject serves as the difference (e.g., snubness as differentiated by nose); (2b) the categorially classed accidental essences serve as the genus and the mode of being of the accident serves as the difference (e.g., snubness as differentiated by inesse). Although the abstract signification of an accident is what properly falls under a category, nevertheless, since the accidental essence adds to the notion of the being of the substance, the mode of being of the accident can serve to amplify the intelligibility of the nature of the accident itself. How this is done depends on whether one is considering the accident abstractly (categorially) or concretely (compositionally or existentially). If considered concretely, Aquinas suggests that an accident can be understood in such a way that that which signifies the essence itself should stand as a difference to the subject in which the concretely understood accident exists as its genus. Thus, “snubness” is a concaved nose such that snubness is a concrete property that includes nose as its subject and genus. A similar thing can be done when considering the relationship between 49

Ibid. Aquinas derives the distinct modes of being of each of the ten categories in In Met., lib. 5, l. 9. I am giving the more generic notion of a mode of being as “to be in” 50

Paul Symington

49

how snubness is found to exist (as distinct from its essence) in relation to the substance in which it exists. However, if taken abstractly (and signifying the category itself), the accident will be the genus that has as its difference its subject in which it exists and has its mode of being. For example, concavity of the nose (abstractly conceived) is further specified by snubnosedness in that the latter is that in which the abstract concavity is realized or actualized.51 So, on the one hand, the accidental essence is actualized (as specific difference to genus) by the existing subject (the substance) whereas on the other hand, the being of the substance is further concretely differentiated or determined by receiving a distinct mode of being through the distinct essence of the accident. Fundamentally, by adding the mode of being, there is allowed a fuller and more specific recognition of the distinctness of each categorial entity as its own thing, and ontologically diversified from every other thing, while including the unity that exists among these distinct ontological elements in a single being. The problem of each categorial thing not having a complete essence is resolved by showing how concrete and abstract accidental accidents relate to the being of the substance. So, how does Aquinas conceive of how the categorial entities combine within a single substance? We have the particular challenge of answering Pasnau’s charge of reductionism with respect to some of the categories, especially regarding action and passion. How can Aquinas claim that action and passion are essentially distinct when there is a single event between them? In general, I think that the natural tendency to be reductive about Aquinas’s view of the latter accidental categories stems from the close connections that these categorial entities form, especially with respect to the deflationary order that arises in the subjection of one to another.52 Yet the principle seems to hold for Aquinas that as long as a distinct intelligible principle (hence a distinct essence) is able to be 51 In Met., lib. 7, l. 4: “In recto quidem, quando accidens significatur ut accidens in concretione ad subiectum: ut cum dico, simus est nasus concavus. Tunc enim nasus ponitur in definitione simi quasi genus, ad designandum quod accidentia non habent subsistentiam, nisi ex subiecto. Quando vero accidens significatur per modum substantiae in abstracto, tunc subiectum ponitur in definitione eius in obliquo, ut differentia; sicut dicitur, simitas est concavitas nasi.” 52 This close relationship is especially seen in the relationship between substance and quantity and between quantity and quality. See Paul Symington, “Thomas Aquinas, Perceptual Resemblance, Categories, and the Reality of Secondary Qualities,” Proceedings of the American Catholic Philosophical Association 86 (2012).

50

Categories and Modes of Being

signified in a way distinct from substance (although still understood as ontologically dependent on substance), there arises the articulation of a new mode being within the esse of substance. In what follows, I provide some examples of this, which serve only as a rough sketch. First, Aquinas maintains that among the accidental categories only relation does not imply a habitude to that subject of which it is predicated.53 With the other accidental categories, there is within the grasp of each categorial being (not including substance and relation) an inherent semantic relation to that wherein it exists. As a consequence of this, each is conditioned by the subject in which they inhere. This gives these eight categorial beings the appearance of indistinctness from the subjects in which they inhere or are otherwise related. For example, he identifies body as “quantity having position.”54 Or, one can take the category of habit. In the Summa theologiae, Aquinas says that the notion of having can be understood in a variety of different ways: in one sense quality and quantity is “had” by substance. In another sense, a person “has” a friend. However, these distinct ways of having, which are post-predicamental, do not prevent the proper category of habit from being identified. Specifically, Aquinas identifies habit as follows: And, further, there are some in which there is a medium, not indeed an action or passion, but something after the manner of action or passion: 53

ST, 1.28.2 ad 1: “Ad primum ergo dicendum quod verba illa Augustini non pertinent ad hoc, quod paternitas, vel alia relatio quae est in Deo, secundum esse suum non sit idem quod divina essentia; sed quod non praedicatur secundum modum substantiae, ut existens in eo de quo dicitur, sed ut ad alterum se habens. Et propter hoc dicuntur duo tantum esse praedicamenta in divinis. Quia alia praedicamenta important habitudinem ad id de quo dicuntur, tam secundum suum esse, quam secundum proprii generis rationem, nihil autem quod est in Deo, potest habere habitudinem ad id in quo est, vel de quo dicitur, nisi habitudinem identitatis, propter summam Dei simplicitatem.” 54 ST, 3.76.3 arg. 3: Praeterea, corpus Christi semper veram retinet corporis naturam, nec unquam mutatur in spiritum. Sed de ratione corporis est ut sit quantitas positionem habens, ut patet in praedicamentis. Sed ad rationem huius quantitatis pertinet quod diversae partes in diversis partibus loci existant. Non ergo potest esse, ut videtur, quod totus Christus sit sub qualibet parte specierum. “Further, Christ's body always retains the true nature of a body, nor is it ever changed into a spirit. Now it is the nature of a body for it to be "quantity having position" (Predic. iv). But it belongs to the nature of this quantity that the various parts exist in various parts of place. Therefore, apparently it is impossible for the entire Christ to be under every part of the species.”

Paul Symington

51

thus, for instance, something adorns or covers, and something else is adorned or covered: wherefore the Philosopher says (Metaph. v, text. 25) that “a habit is said to be, as it were, an action or a passion of the haver and that which is had”; as is the case in those things which we have about ourselves. And therefore these constitute a special genus of things, which are comprised under the category of habit: of which the Philosopher says (Metaph. v, text. 25) that “there is a habit between clothing and the man who is clothed.”55

An important aspect in Aquinas’s identification of habit is that the having of one thing by another is mediated, or has the intervening subject (medium) of action and passion. I think that one can interpret ‘medium’ here as that which is involved as a subject of a categorial essence insofar as it is included in its concrete pseudo-definition (although it can be signified abstractly in a way independently of these subjects). So, it seems that Aquinas’s view of habit is that it is an irreducible essence between two things (e.g., a man and some clothing) that is mediated by action or 55

Thomas Aquinas, Summa Theologica, trans. by English Dominican Fathers (Ann Arbor: R. & T. Washbourne, 1920), ST, 2.1.49.1 co. Cf. “Respondeo dicendum quod hoc nomen habitus ab habendo est sumptum. A quo quidem nomen habitus dupliciter derivatur, uno quidem modo, secundum quod homo, vel quaecumque alia res, dicitur aliquid habere; alio modo, secundum quod aliqua res aliquo modo se habet in seipsa vel ad aliquid aliud. Circa primum autem, considerandum est quod habere, secundum quod dicitur respectu cuiuscumque quod habetur, commune est ad diversa genera. Unde philosophus inter post praedicamenta habere ponit, quae scilicet diversa rerum genera consequuntur; sicut sunt opposita, et prius et posterius, et alia huiusmodi. Sed inter ea quae habentur, talis videtur esse distinctio, quod quaedam sunt in quibus nihil est medium inter habens et id quod habetur, sicut inter subiectum et qualitatem vel quantitatem nihil est medium. Quaedam vero sunt in quibus est aliquid medium inter utrumque, sed sola relatio, sicut dicitur aliquis habere socium vel amicum. Quaedam vero sunt inter quae est aliquid medium, non quidem actio vel passio, sed aliquid per modum actionis vel passionis, prout scilicet unum est ornans vel tegens, et aliud ornatum aut tectum, unde philosophus dicit, in V Metaphys., quod habitus dicitur tanquam actio quaedam habentis et habiti, sicut est in illis quae circa nos habemus. Et ideo in his constituitur unum speciale genus rerum, quod dicitur praedicamentum habitus, de quo dicit philosophus, in V Metaphys., quod inter habentem indumentum, et indumentum quod habetur, est habitus medius. Si autem sumatur habere prout res aliqua dicitur quodam modo se habere in seipsa vel ad aliud; cum iste modus se habendi sit secundum aliquam qualitatem, hoc modo habitus quaedam qualitas est, de quo philosophus, in V Metaphys., dicit quod habitus dicitur dispositio secundum quam bene vel male disponitur dispositum, et aut secundum se aut ad aliud, ut sanitas habitus quidam est. Et sic loquimur nunc de habitu. Unde dicendum est quod habitus est qualitas.”

52

Categories and Modes of Being

passion. To put it in terms identified above, habit—although signifiable according to its own principle of intelligibility—is pseudo-defined either abstractly or concretely in relation to modes of being and subjects of inherence. This is distinct from action, which does not include the notion of having. Habit betokens a relationship existing between a body and what is adjacent to it.56 The theme of unifying and explicating categorial essences along with various intervening subjects appears to be similar to the way he presents the other categories. For example, he says that “position is a disposition, which is the order of that which has parts,” but with the further determination, “with respect to place.”57 This runs parallel to Aquinas’s identification of quality as “disposition of substance” cited above in the passage from the De ente.58 This is interesting since although there is expressed a similar concept (“disposition of”), yet they are differentiated in that they are dispositions with respect to different subjects or mediums. 56 In Physic., lib. 5, l. 3, n. 3: “Deinde cum dicit: secundum substantiam autem etc., manifestat conditionalem praemissam. Et primo ostendit quod in aliis generibus a tribus praedictis, non potest esse motus; secundo ostendit quomodo in istis tribus generibus motus sit, ibi: quoniam autem neque substantiae et cetera. Circa primum tria facit: primo ostendit quod in genere substantiae non est motus; secundo quod nec in genere ad aliquid, ibi: neque est in ad aliquid etc.; tertio quod nec in genere actionis et passionis, ibi: neque agentis neque patientis et cetera. Praetermittit autem tria praedicamenta, scilicet quando et situm et habere. Quando enim significat in tempore esse; tempus autem mensura motus est: unde per quam rationem non est motus in actione et passione, quae pertinent ad motum, eadem ratione nec in quando. Situs autem ordinem quendam partium demonstrat; ordo vero relatio est: et similiter habere dicitur secundum quandam habitudinem corporis ad id quod ei adiacet: unde in his non potest esse motus, sicut nec in relatione. Quod ergo motus non sit in genere substantiae, sic probat. Omnis motus est inter contraria, sicut supra dictum est: sed substantiae nihil est contrarium: ergo secundum substantiam non est motus.” 57 ST 2.1.49.1 ad 3: “Ad tertium dicendum quod dispositio quidem semper importat ordinem alicuius habentis partes, sed hoc contingit tripliciter, ut statim ibidem philosophus subdit, scilicet aut secundum locum, aut secundum potentiam, aut secundum speciem. In quo, ut Simplicius dicit in commento praedicamentorum, comprehendit omnes dispositiones. Corporales quidem, in eo quod dicit secundum locum, et hoc pertinet ad praedicamentum situs, qui est ordo partium in loco. Quod autem dicit secundum potentiam, includit illas dispositiones quae sunt in praeparatione et idoneitate nondum perfecte, sicut scientia et virtus inchoata. Quod autem dicit secundum speciem, includit perfectas dispositiones, quae dicuntur habitus, sicut scientia et virtus complete.” 58 De ente, cap. 5.

Paul Symington

53

Namely, the category of position implies the order of parts in place; and place can be considered a subject or medium of position. As a result of this, whatever is moved according to position must be moved according to place.59 This brings us specifically to the categories of action and passion. Regarding these, Aquinas says that, Although motion is one, nevertheless there are two categories which are based on motion depending on the different external things according to which the predicamental denominations are made. For an agent is one thing from which as from something external the predicament of passion is taken; and the patient is some other thing from which something in denominated an agent.60

This is an important passage for addressing Pasnau’s worry about action and passion. The motion (or ‘event’ in Pasnau’s language) is indeed one, 59

In Phys., lib. 4, l. 7, n. 4: “Unde Alexander dixit quod ultima sphaera nullo modo est in loco: non enim omne corpus de necessitate est in loco, cum locus non cadat in definitione corporis. Et propter hoc dixit quod ultima sphaera non movetur in loco, neque secundum totum, neque secundum partes. Sed quia oportet omnem motum in aliquo genere motus poni, Avicenna eum secutus, dixit quod motus ultimae sphaerae non est motus in loco, sed motus in situ, contra Aristotelem, qui dicit in quinto huius, quod motus est tantum in tribus generibus, scilicet in quantitate, qualitate et ubi. Sed hoc non potest stare: impossibile est enim quod motus sit per se loquendo in aliquo genere cuius specierum ratio in indivisibili consistit. Propter hoc enim in substantia non est motus, quia ratio cuiuslibet speciei substantiae consistit in indivisibili, eo quod species substantiae non dicuntur secundum magis et minus: et propter hoc, cum motus habeat successionem, non producitur in esse forma substantialis per motum, sed per generationem, quae est terminus motus. Secus autem est de albedine et similibus, quae participantur secundum magis et minus. Quaelibet autem species situs habet rationem in indivisibili consistentem; ita quod si aliquid additur vel minuitur, non est eadem species situs. Unde impossibile est quod in genere situs sit motus. Et praeterea, remanet eadem difficultas. Nam situs, secundum quod ponitur praedicamentum, importat ordinem partium in loco: licet secundum quod ponitur differentia quantitatis, non importet nisi ordinem partium in toto. Omne igitur quod movetur secundum situm, oportet quod moveatur secundum locum.” 60 In Phys., lib. 3, l. 5, n. 16: “Sic igitur patet quod licet motus sit unus, tamen praedicamenta quae sumuntur secundum motum, sunt duo, secundum quod a diversis rebus exterioribus fiunt praedicamentales denominationes. Nam alia res est agens, a qua sicut ab exteriori, sumitur per modum denominationis praedicamentum passionis: et alia res est patiens a qua denominatur agens. Et sic patet solutio primae dubitationis.”

54

Categories and Modes of Being

but it serves only as the subject or medium (viz., the basis) which yet requires further formal specification or differentiation. This further differentiation comes from both the cause of the action and the receiver of the action. So, the essence of action arises from the agent itself and the motion, whereas the essence of passion arises from the passive subject and the motion. It seems that the agent, the passive subject, and the motion are all merely subjects that serve as the genus of the categorial essences of action and passion themselves. That is, the categories of action and passion are based on the notions of acting cause and of effect,61 coupled with the single event of the change itself. Regarding action and passion specifically, as with all the categories, action and passion have their own subjects, which include quality, quantity, where and when, but are not exhausted by these and have their own distinct forms that outstrip these.62 Does this analysis that I have given prove that Aquinas is not a reductionist about categories even though he is a deflationist about accidental beings? I think that the answer to this question can come only with a broader discussion about ontological methodology and expectations. Of course, Aquinas sees his view as ontologically robust, but in doing so he asks us to approach his conclusions in a certain way. In fact, this brings me to what constitutes a broad assessment of Pasnau’s book. Although I have gone fairly quickly through some complicated issues in Aquinas’s metaphysics, right or wrong, an important lesson can be drawn 61

In Phys., lib. 3, l. 5, n. 17: “Quantum igitur ad id quod in rerum natura est de motu, motus ponitur per reductionem in illo genere quod terminat motum, sicut imperfectum reducitur ad perfectum, ut supra dictum est. Sed quantum ad id quod ratio apprehendit circa motum, scilicet esse medium quoddam inter duos terminos, sic iam implicatur ratio causae et effectus: nam reduci aliquid de potentia in actum, non est nisi ab aliqua causa agente. Et secundum hoc motus pertinet ad praedicamentum actionis et passionis: haec enim duo praedicamenta accipiuntur secundum rationem causae agentis et effectus, ut dictum est.” 62 In Met., lib. 7, l. 3, n. 10: “Deinde cum dicit quoniam vero. Inquirit quorum sit quod quid erat esse. Et primo movet quaestionem. Secundo solvit eam, ibi, at vero secundum se dictorum. Dicit ergo primo, quod sunt quaedam composita in aliis praedicamentis, et non solum in substantia. Quod quidem dicit propter hoc, quod substantiarum sensibilium, quae sunt compositae, quidditatem inquirit. Sicut enim in substantiis sensibilibus compositis est materia, quae subiicitur formae substantiali, ita etiam alia praedicamenta habent suum subiectum. Est enim aliquod subiectum unicuique eorum, sicut qualitati et quantitati et quando et ubi et motui, sub quo comprehenditur agere et pati. Unde sicut quoddam compositum est ignis ex materia et forma substantiali, ita est quaedam compositio ex substantiis et accidentibus.”

Paul Symington

55

from it in relation to the way that Pasnau approaches metaphysics in general in his book. Specifically, Pasnau seems resistant to entertaining the idea that conceptual or linguistic structures can be a valid way of articulating, envisioning, and establishing ontological concepts. This can be best illustrated by highlighting two passages from Pasnau’s book: If the substance-accident ontology does not fall out of the definition of what a substance is, then how does it arise? No doubt, part of its appeal comes from an uncritical reliance on the surface structure of language. Since language attaches predicates to subjects, it is easy to suppose that the world’s structure corresponds. This sort of simple-minded thought should have carried little weight with scholastic authors, however. They had at their disposal a variety of semantic theories that explained predication without any commitment to a substance-accident ontology, such as Ockham’s version of supposition theory, which he formulated in the interests of his own austere ontological program.63

Compare that quote with the following one in which he is discussing a doctrine that is characteristic of nominalists: [The] characterization of the disagreement [between nominalists and realists] focuses on whether the surface structure of language corresponds to the structure of reality, in such a way that distinct terms match up with distinct things in reality. This, however, has little to do with the problem of universals; it refers mainly to a dispute over the categories (see Ch. 12): does every predicate across Aristotle’s categorial scheme—e.g., warm, sixfeet tall, next to, sitting—have corresponding to it a real accidental form?64

I think in these two quotations we see, despite its heroic merits, a limitation in Pasnau’s approach. In not taking seriously metaphysical approaches through language, Pasnau bends his analysis to the side of the nominalist, even if this label is radically deficient. I think that his analysis as a consequence is forced to pass over serious treatments of those historically sympathetic to a (Thomistic?) realism, which envisions, for example, that the best way of conceiving and mediating matter and potency is the propositional subject, form through predicates, existence and actuality through the predication of the copula, and inherence through content expressed in the act of judgment. I think that this is unfortunate since an appealing aspect of Aquinas’s metaphysics is its leanness; a tightness that both gives rise to natural ways of understanding his 63 64

Pasnau, pp. 106, 107. Pasnau, p. 86.

56

Categories and Modes of Being

metaphysical principles and to helpful reductive moves on the part of his interpreters.

RESPONSE TO ARLIG AND SYMINGTON ROBERT PASNAU

1. Response to Andrew Arlig Arlig raises many hard questions about some of the hardest parts of my book, and I am sure that I cannot satisfactorily meet all of his challenges. But let me do what I can. Arlig begins by considering the doctrine that Henry More, in the seventeenth century, dubbed holenmerism, which is roughly speaking the idea that the whole of a thing exists in each of the parts of some other thing. The paradigmatic case is the rational soul, which is standardly said to exist as a whole in each individual part of the body. One finds this view not only throughout the later Middle Ages, but also in earlier figures such as Plotinus, Augustine, John of Damascus, and Anselm.1 I suggest, in my book, that holenmerism is a plausible candidate for demarcating the material from the immaterial. No such principle of demarcation is really needed for the first part of my period—the later scholastic era—because scholastic authors are generally content to think that the distinction can be marked off in terms of the presence or absence of prime matter, making immaterial entities literally those that lack matter. But as increasing doubts arise, toward the end of my four centuries, about the reality of prime matter, it becomes less clear what makes something material or immaterial. Descartes famously said that bodies are essentially characterized by having extension, but this is less clear than it initially seems, because Descartes also thinks that the human soul exists throughout the human body, and he thinks that God exists everywhere. Inasmuch as that would seem, prima facie, to make both soul and God extended, Descartes needs to say something further about the kind of extension that characterizes bodies. Holenmerism—or rather its opposite, being extended partem extra partem—looks like a promising candidate, 1

For references see Metaphysical Themes p. 337 n. 18.

58

Response to Arlig and Symington

and indeed Descartes does describe both souls and God as existing holenmerically, and describes bodies as having true extension, part outside of part.2 Rather than think of holenmerism as the unique way of demarcating the material from the immaterial, it is more plausible to think of it as a sufficient mark of immateriality. It seems to me we can most effectively capture the notion of an immaterial realm of entities by allowing for a variety of possible ways in which the world might be, such that, if there were entities like that, then it would be right to think of the world as carved up into two kinds—the familiar material kind, and a distinct weird kind of entity, so different from what we’re familiar with as to go into a fundamentally different category of being. If there are, for instance, entities that have no spatiotemporal location, or that have no causal interactions with other entities, then it would seem to me reasonable to label such entities immaterial. They would be so radically unlike bodies as to belong in a separate class. Holenmerism, it seems to me, is yet another way of being immaterial. If there are things that can be wholly in one place, and wholly in another place at the same time, then that would be so truly weird as to justify thinking of them as incorporeal or immaterial. This is not to say that they would eo ipso be supernatural, or even that they would be beyond the scope of scientific inquiry—only that such things would seem to be radically different in character from the material things we are familiar with. Authors during my four centuries almost all accept that God exists holenmerically in the world, and that the rational soul exists holenmerically in the human body. That much fits fairly well with my suggestion that we associate holenmerism with immateriality. Admittedly, my suggestion fits less well with the view of someone like Thomas Aquinas, according to whom not just the human soul, but all souls and even the substantial forms of non-living things exist holenmerically.3 Aquinas’s view, however, seems to be in the minority among later scholastic authors. More common is to suppose that only God and the rational soul exist holenmerically in this way. (At least my impression is that this is more common; there has been amazingly little research into the

2

See Metaphysical Themes §16.4. I base this conclusion on Summa theol. 1a 76.8c. It would be interesting to investigate this issue in more detail, to see whether Aquinas is truly committed to this view, and whether many others agreed with him in this respect.

3

Robert Pasnau

59

topic.) Thus John Buridan remarks that the rational soul’s holenmeric existence within the body is “mirabile et super naturam.”4 When one does limit holenmerism to a few exceptional cases, a natural reply emerges to Arlig’s remarks regarding universals. He raises the puzzle of why so many scholastic authors would embrace holenmerism while rejecting universals, even though universals on their face seem hardly distinct from holenmers. The reply would not dispute the similarity between holenmers and universals. The only difference, one might say, is that the rational soul’s ability to be wholly and simultaneously located at multiple places is limited to places within a single substance, whereas universals characteristically cross substances. But this, admittedly, will not do to define the difference. After all, God is said to be holenmeric in a way that does cross substances. And we might imagine a universal, nearlymissing shade of blue that exists in just one single, solitary material substance, but is wholly and simultaneously present at multiple places on the surface of that body. Still, I don’t think it’s particularly pressing to provide a clear demarcation between the holenmeric and the universal. This strikes me as merely a terminology matter. The more pressing question concerns why the nominalist would reject universals but embrace holenmers. And here is where it seems to me the nominalist like Buridan has a clear answer. Both holenmers and universals are unnatural and perplexing. So far as possible, we should posit neither. Certainly, when it comes to the natural realm, and familiar entities like sensible qualities, it would be disastrous to natural philosophy to embrace any such things. If we are to embrace them, we should do so only in supernatural cases— cases that go beyond the scope of natural philosophy. This sort of reply will work only for the nominalist whose arguments against universals are grounded in natural philosophy. The nominalist who has some sort of in-principle logical or metaphysical argument against universals may be harder pressed to remain consistent. This is to say that I think Arlig is right in suggesting there are going to be arguments against the coherence of universals that apply just as well against holenmers. It would be interesting to look in detail at some such arguments. Arlig does mention one argument, due to Boethius, which has as its crucial premise the claim that “it cannot come about, when the whole is in many at one time, that it itself is numerically one” (p. 13). Obviously, this result 4

In De an. II.9, ed. Sobol p. 138.

60

Response to Arlig and Symington

is even more unacceptable in the case of holenmers than it is in the case of universals: whatever we might say about God or the rational soul, we must not deny that it is one thing. It is interesting to consider whether the proponent of holenmers might evade this line of argument. One way around it would be to note that holenmers, unlike universals, exist continuously, everywhere they exist. Whereas ordinary, robustlyinstantiated shades of blue exist in patches here and patches there, the soul occupies a continuous region of space. Something similar can be said of God, inasmuch as God occupies all space. But there seems something feeble about the idea that the soul’s unity turns on its occupying a continuous body—to say nothing of God’s case. After all, we do not ordinarily think that physical continuity is sufficient for unity. And what about that nearly-missing shade of blue, which likewise occupies just a single, continuous body? But there is perhaps a stronger reply to be made in favor of the unity of holenmers like God and the rational soul—namely, that they have operations above and beyond the discrete operations they perform in discrete parts of bodies. The rational soul, in addition to whatever it may do in my liver or kidneys, also thinks, and this is an operation that cannot be attributed to the soul as it is in one organ or another. Something similar is surely true for God. Nothing like that is the case, however, for a shade of blue, which has no operations beyond the discrete roles that it plays on this surface and that one. So there is pressure to think of the rational soul or God as a single thing, in virtue of the operation it performs as a whole—a pressure that seems entirely lacking in the case of ordinary universals. Interestingly, however, this sort of response is available only to someone, like Buridan, who limits holenmerism to God and the rational soul. For someone like Aquinas, who thinks that all substantial forms exist holenmerically, it will be considerably harder to explain why holenmers are allowed but universals are not. For it does seem plausible to suppose that the only operations performed by a tree’s substantial form are the ones it performs in this part of the tree and that part of the tree. Arlig next takes up nominalist approaches to identity over time. These fourteenth-century discussions—in figures like Ockham, Buridan, and Oresme—reveal identity over time to have received a much more nuanced account during the scholastic era than is ordinarily recognized. One might naturally have thought that Aristotelians would have no difficulty with identity over time, inasmuch as they can distinguish between the changeable matter and the unchanging form of a substance, and then insist that the whole remains the same over time, provided the substantial form

Robert Pasnau

61

continues to inform some appropriate sort of matter. But in Chapter 29 I identify two theses that give rise to a distinctively nominalist approach to identity. The first of these is the part-whole identity thesis: that the whole composite material substance is nothing over and above its various parts. The second is the no-transfer principle: when the integral parts of a substance change, the substantial form must also change, at least partially. If one embraces both of these theses, then it becomes extremely difficult to explain identity over change. Part-whole identity entails that any change among the parts is tantamount to a change to the whole. The no-transfer principle entails that not even the substantial form can remain wholly identical through change to the parts. So whereas Aristotelians seemed to have an easy time of accounting for identity over time, these two theses destroy that advantage entirely. The theses are both extremely contentious. Scotus clearly rejects the first, and probably rejected the second too. Aquinas clearly rejects the second, and probably the first too. Aquinas has to reject the no-transfer principle, because, as we have seen, he thinks all substantial forms exist holenmerically. In general, holenmerism is incompatible with the notransfer principle, because holenmerism tells us that the very same soul is in each part of the body, and so its identity will not depend on whether the body gains or loses a part. Since even the nominalists embrace holenmerism with respect to the rational soul, they must reject the notransfer principle in that one special case. This naturally suggests a distinction between three senses of identity over time. And indeed one finds just that in Buridan. Here in brief is how he describes it:

5

x

The first way is by being totally (totaliter) the same—namely, because this is that and there is nothing belonging to the whole of this that does not belong to the whole of the other and vice versa. This is numerical sameness in the most proper sense….

x

In a second way, however, one thing is said to be partially the same as another—namely, because this is part of that (and this is especially said if it is a major or principal part), or else because this and that take part in something that is a major or principal part of each….

x

But in a still third way, less properly, one thing is said to be numerically the same as another according to the continuity of distinct parts, one in succession after another.5

In Phys. I.10, f. 13vb, as translated (in full) in Metaphysical Themes p. 696.

62

Response to Arlig and Symington

The first applies to very little—to the heavenly bodies, but certainly not to animals over any appreciable length of time. The second applies to human beings, in virtue of our persisting rational souls. The third applies to the human body and to other animals; this is the sort of identity possessed by the Seine and other bodies of water. Arlig’s main focus is on just how we are to characterize these lesser sorts of identity. My view is that we have identity of the whole thing only in the first case, and that in the second case we have partial identity (meaning simply that some part of the thing remains identical over time), and that in the third case we have no identity at all. But on my view Buridan wants to couple these metaphysical conclusions with a theory of how we in practice talk about identity. Thus he says that we can say, without qualification, that human beings remain the same throughout their lives, “because we customarily denominate, unqualifiedly and without adding anything, a thing by means of its most principal part.”6 Even further from the actual metaphysical facts of identity, “the Seine is said to be the same river after a thousand years, although properly speaking nothing is now a part of the Seine that was part of it ten years ago.”7 Arlig agrees with me—at least I think he does—about the metaphysical claims being made, but he wants to resist the suggestion he finds in my book, that a strict metaphysical account is being replaced with a “loose and popular” sense of identity. Now in fact I never use Bishop Butler’s famous phrase, but I do repeatedly describe the second and third forms of identity as “looser,” so to that extent Arlig’s characterization is fair enough. Here, though, I want to draw some distinctions. First, I am happy to grant that Buridan develops these different senses of identity with considerable rigor and precision. If there is anything loose here, I do not mean to suggest that the looseness lies on Buridan’s side. Second, I think that to understand the nominalist project, one needs to begin with a clear sense of what identity involves. Things are identical when they are in fact not multiple things at all, but are just one thing. This is the identity of the equal sign, the identity that licenses the indiscernibility of identicals, which is to say that things are identical only if they share all the same features. It is unintelligible to say that things are identical and yet different. Or, rather, such talk can be made intelligible, but only when construed in some looser, less-than-strict sense. That is, to speak of 6 7

In De gen. et cor. I.13, as quoted in Arlig, above p. 20. In Phys. I.10.

Robert Pasnau

63

identity where there is difference requires construing such claims as saying something other than what they seem on their face to say. Hence Buridan offers a perfectly rigorous account of the true propositions that people can be understood to be expressing, when they say—seemingly against all reason—that two different things are the same thing. Third, I do want to insist that, strictly and literally speaking, Buridan is denying that human beings and animals are the same over time. Here I fear Arlig and I are not entirely in agreement. He says, for instance, “I don’t think that Buridan’s view is that metaphysically speaking animals and plants do not persist” (p. 25). I do think exactly that. I think that, speaking strictly and literally, animals and plants and human beings can persist only for as long as their parts remain the same—which is not very long, as Buridan knew, given the constant change at the level of their particles. Now of course Buridan is also explaining how it can be true to say that the same tree has lived in the courtyard for 100 years. But that’s only loosely true—loosely, because for it to come out true it has to be interpreted as a claim about a certain kind of succession of distinct substances. Such paraphrasing allows Buridan to affirm the truth of what the folk say. But this is not enough to count Buridan’s view as consistent with “pretheoretical, commonsense ontology” (Arlig p. 25)—unless one thinks that the folk believe that the tree growing in the courtyard, and the dog barking in the courtyard, and the man walking through the courtyard, are all just continuous sequences of numerically distinct substances. Arlig stresses that commonsense must surely leave room not just for identity over time, but also for change. I quite agree. But it seems to me what commonsense wants is a theory that allows for both without compromise. One way to do that is to reject the part-whole identity thesis, and argue that the whole can remain the same even while there is change at the level of the parts. But this introduces new puzzles, for it requires us to say that the tree, for instance, is something over and above the roots, trunk, branches, and leaves. That certainly does not sound like commonsense. Philosophers pick and choose among the various mysteries here, but one thing they cannot do is have a view that embraces commonsense without any mystery at all. This is what makes metaphysics both interesting and difficult.

64

Response to Arlig and Symington

2. Response to Paul Symington Symington raises many interesting questions about Thomas Aquinas’s conception of the nine accidental categories of being. He and I basically agree on the most fundamental point: on what it is for an accident to exist, on Aquinas’s view. Aquinas takes what I call a “deflationary” approach to accidents, which is just to say that he regards their existence as somehow secondary to the existence of the substance in which they inhere. Consider the following passage, from Quaestiones de virtutibus in communi q. 11c: Many err regarding form because they judge it as if they were judging substance. This seems to happen because forms are signified as substances are, in the abstract, as whiteness or virtue, and so on. As a result, some follow this mode of speech and judge accidents as if they were substances. . . . For they hold that forms are suited to be made just as substances are, and so when they do not find what it is that generates forms, they claim that they are either created or preexist within matter. What they do not notice is that just as existing belongs not to form, but to the subject through the form, so too being made (which culminates in existing) belongs not to form, but to the subject. For just as a form is said to be a being not because it exists—if we are to speak properly—but because something exists by it, so too a form is said to be made not because it is made, but because something is made by it, when a subject is brought from potentiality to actuality.8

It would be possible to push this passage quite far, all the way to some sort of eliminative view regarding accidents, on which accidents, for Aquinas, do not exist at all, and only substances exist. But I think that this is not quite what Aquinas says, even in this very strongly worded passage. The crucial, final sentence does not deny that accidents are beings, but instead tells us the sense in which they are beings—an accident exists inasmuch as “something [a substance] exists by it.” This is a puzzling thing to say, to be sure, but it does not prima facie seem to deny the existence of accidents. Compare this passage from the De ente et essentia: “being absolutely and primarily is said of substances, and is said of accidents in a derivative and relative way.”9 Hence I say not that Aquinas is an eliminativist about accidents, but that he holds a deflationary view.

8

Sicut enim forma ens dicitur, non quia ipsa sit, si proprie loquamur, sed quia aliquid ea est…. 9 “ens absolute et primo dicitur de substantiis et per posterius et secundum quid de accidentibus” (De ente ch. 1, ed. Leo. 43:370)

Robert Pasnau

65

Deflationary views were quite common—perhaps even standard—among thirteenth-century Latin Aristotelians. One finds views of this sort in Richard Rufus of Cornwall, Albert the Great, and Siger of Brabant. (See Metaphysical Themes §10.2 for details.) Interestingly, however, deflationary views seem to go out of fashion around the end of the thirteenth century, as a result (or so I claim) of John Duns Scotus’s influence. Scotus argued that accidents exist in just the way that substances exist, which is the view I characterize as the doctrine of real accidents. From the end of the thirteenth century forward, it becomes standard to suppose that if accidents exist, they really exist, which is just to say that their existence is not fundamentally different from the existence of substances. (See Metaphysical Themes §§10.4-5, 11.1.) In insisting that accidents exist in a different way from how substances exist, Aquinas is seeking to make sense of the Aristotelian dictum that “being is said in many ways” (see, e.g., Meta. IV.2). Just how many ways is precisely the task of the Categories to explain, and Aquinas takes it that the ten categories are setting out ten different modes of being. So far Symington and I are in agreement. It may be that, if I tried to say more about what this deflationary theory amounts to, he and I would start to disagree. But I am not sure that I can say much more about Aquinas’s position, and since I have no interest in making more trouble with my critics, let me continue to focus on where we agree. We agree that accidents somehow are beings. And since the nine categories just do divide the different kinds of accidental being, we agree that those categories pick out different modes of being. Since this is what Aquinas expressly says that the categories are doing—picking out different modes of being—it would be hard to deny that much. Accordingly, there is some sense in which each of the categories is ontologically committing. As Aquinas puts it in his Quaestiones de potentia, “something is put into a category only if it is a thing (res) existing outside the soul” (7.9c). So where then do Symington and I disagree? We disagree over whether Aquinas thinks that each of the ten categories marks off an irreducible category of being. He thinks that they do—that the categories, for Aquinas, set out ten irreducible kinds of being. This seems to me unlikely, simply on the basis of the texts. But before looking at the texts, let me set out the interpretive scheme I offer for the various kinds of views that one finds regarding the categories. Here are five different sorts of positions one could take about how to understand a given category.

Response to Arlig and Symington

66 A.

As a distinct kind of res (a substance or a real accident)

B.

As a distinct kind of mode (a real item in the world, but somehow not a res)

C.

As a distinct kind of structure (a feature of reality, but not an item [res or mode] over and above the items in other categories)

D.

As a distinct linguistic or conceptual kind

E.

As not a distinct kind at all, but wholly eliminable (see Metaphysical Themes p. 238)

We can set aside D, which characterizes nominalist views of the categories like Ockham’s and Buridan’s, and we can set aside E, which characterizes the skeptical attitude of someone like Peter John Olivi toward whether some of the categories carve up anything at all. That leaves A, B, and C. I understand Aquinas as putting only the category of Substance into A, inasmuch as I take him not to adhere to the doctrine of real accidents. I take Aquinas to treat some accidental categories, at least Quantity and Quality, as falling into B. Such accidents are, for Aquinas, in effect modes, in something like the way that Suárez and Descartes would later speak of modes. (There are, however, a great many subtleties here regarding what exactly modes are, which I explore in some detail in Metaphysical Themes Ch. 13, and I will not try even to summarize that material here.) With respect to substance, and with respect to these accidents, I think again that Symington and I are broadly in agreement. He might well hesitate to compare Aquinas’s accidents with Suarezian or Cartesian modes, and I would not blame him in so hesitating, because the issues there are really extremely difficult. Still I think we are on mostly the same page here, with respect to Quantity and Quality. But whereas Symington thinks that all the accidental categories fall roughly into my class B, I want to suggest that some of the categories pick out reality in a different way, marking out neither substance-like res or mode-like dependent entities, but rather features of reality that are not irreducible entities at all, but rather structures of entities—the world so-and-so organized. This class C is the only way I can find to make sense of how Aquinas treats some of the lesser categories of being. I do not argue, in my book, over just exactly how many of the accidental categories get put into class C, for Aquinas. As Symington points out, I suggest in a note that an enthusiast of this approach might take Aquinas’s view to be that only Quantity and Quality are class-B categories. But I

Robert Pasnau

67

don’t feel sure about that by any means, and in particular I don’t feel sure what to say about the very difficult category of Relation. Here, then, let me focus only on a few examples. If I can make my view look persuasive anywhere across the categorial scheme then that’s enough of a response to Symington, because he thinks Aquinas never resorts to anything like classC accidents. To my mind, the clearest example are the categories of Action and Passion. (Where I speak of “Passion” or “patient,” I just mean Being Acted On or the thing acted on. But it is useful to have a one-word label for the passive category that is the counterpart of Action. Hence the stilted language.) Consider this passage, from the Physics commentary, which Symington also quotes: Although there is one motion, there are two categories based on motion, based on how categorial denominations are made from different external things. For the agent is one thing, from which as from something external the category of Passion is taken through its mode of denomination; the patient is another thing, from which the agent is denominated.10

For my purposes, it’s crucial to stress the passage’s initial clause, that “there is one motion.” This is an Aristotelian dictum, set out at Physics III.3. There Aristotle says the following: The solution of the difficulty is plain: motion is in the movable. It is the fulfillment of this potentiality by the action of that which has the power of causing motion; and the actuality of that which has the power of causing motion is not other than the actuality of the movable; for it must be the fulfillment of both…. Hence there is a single actuality of both alike, just as one to two and two to one are the same interval, and the steep ascent and the steep descent are one—for these are one and the same, although their definitions are not one. So it is with the mover and the moved (202a12-20).

There is, then, just one motion—or one action—within agent and patient. With this principle in mind, let us have a case. If one says, This table was built by Mary

then a passio is ascribed to the table. If one says, 10 “Sic igitur patet quod licet motus sit unus, tamen praedicamenta quae sumuntur secundum motum sunt duo, secundum quod a diversis rebus exterioribus fiunt praedicamentales denominationes. Nam alia res est agens, a qua sicut ab exteriori sumitur per modum denominationis praedicamentum passionis; et alia res est patiens a quo denominatur agens” (In Phys. III.5.323).

Response to Arlig and Symington

68 Mary built this table

then an actio is ascribed to Mary. With a concrete example like this in mind, the difference between me and Symington is clear. On his account, the actio and the passio are two distinct, irreducible accidental forms, one inhering in the table and the other inhering in Mary. On my account, there are not two such distinct accidents. What we have here, instead, is one motion and two substances (assuming, for simplicity’s sake, that we allow the table to count as a substance). If we had a conception of motion on which there was one motion inhering in Mary, and another inhering in the table, then I would happily embrace Symington’s understanding of the example. But it seems quite clear both from what Aristotle says and from what Aquinas says that we are not allowed two motions, but just one. Does that motion inhere in Mary or in the table? Well, that is like asking whether the road is in Thebes or Athens, to use Aristotle’s famous example from later in Physics III.3 (202b14). It is just one road, running from one place to the other. Similarly, here, it is just one motion or action, running from Mary to the table. To be sure, there are a lot of questions one might ask about what a motion is, and to get very far on this subject we would need to engage with the thorny problem of entia successiva. But, whatever one might say about motions, it does not look like it will furnish the materials for a realistic, non-reductive story about Action and Passion. Using just the ingredients Aquinas gives us, I do not see how Symington can account for two distinct, irreducible accidents. We cannot very well say that one and the same motion is both the accidental form of action in Mary, and the accidental form of passion in the table. That would make this accident into a universal property, multiply instantiated at once in multiple individuals. Aquinas, we can surely agree, wants no such thing. Moreover, Symington’s notes contain a nice text that explicitly rules out this sort of view, in virtue of the principle that “nothing can be in two categories.”11 So what Symington has to say, so far as I can see, is that the motion in Mary gives rise to some further entity, the action, and that the motion in the table gives rise to yet another further entity, the passion. If I understand him properly, this is exactly what he thinks. Perhaps the most plausible way to develop such an idea would be in terms of some sort of 11 “Non potest autem aliquid esse in duobus praedicamentis, quia homo albus non est aliquid unum nisi per accidens, ut habetur V Metaph.” (De potentia 2.2. arg. 2).

Robert Pasnau

69

relational entities connecting the motion with Mary on one end and the motion with the table on the other. I think that Scotus’s view is something like that. But I can find no evidence whatsoever that Aquinas wants this sort of extra ontological baggage.12 Certainly, there is nothing in Aquinas’s discussions of Action and Passion to encourage the idea that there are actually five basic entities in play when one thing acts on another. Consider the above passage from the Physics commentary. After raising the worry about how there is just one motion, it does not go on to reassure the reader that nevertheless the motion gives rise to a pair of distinct accidental forms at either end. Rather, it attributes the categorial distinction to a difference in “categorial denominations.” What I take this to mean is that the different categories arise in virtue of our different ways of speaking. If we want to talk about the motion from the perspective of the table, then we are invoking the category of Passion. If we take the perspective of Mary, then we invoke Action. A passage from the Metaphysics commentary is still clearer: If Action and Passion are the same in substance, then it seems that they are not distinct categories. But it should be known that the categories (praedicamenta) are distinguished according to the different modes of predicating. Hence the same thing, inasmuch as it is differently predicated of different things, pertains to different categories. For location (locus), inasmuch as it is predicated of that which locates [i.e., the surrounding body], pertains to the genus of Quantity. But inasmuch as it is predicated denominatively of the thing that is located, it constitutes the category of Where. Likewise motion, inasmuch as it is predicated of the subject in which it is, constitutes the category of Passion. But inasmuch as it is predicated of that from which it is, it constitutes the category of Action.13

12

For Scotus see Metaphysical Themes §12.5. It is worth stressing Symington’s observation that Scotus himself does not find this sort of category realism in Aquinas. Nor, later, would Suárez, as I observe at Metaphysical Themes p. 232 n. 15. 13 “Sed si actio et passio sunt idem secundum substantiam, videtur quod non sint diversa praedicamenta. Sed sciendum quod praedicamenta diversificantur secundum diversos modos praedicandi. Unde idem, secundum quod diversimode de diversis praedicatur, ad diversa praedicamenta pertinet. Locus enim, secundum quod praedicatur de locante, pertinet ad genus quantitatis. Secundum autem quod praedicatur denominative de locato, constituit praedicamentum ubi. Similiter motus, secundum quod praedicatur de subiecto in quo est, constituit praedicamentum passionis. Secundum autem quod praedicatur de eo a quo est, constituit praedicamentum actionis” (In Meta. XI.9.2313).

70

Response to Arlig and Symington

Here my translation switches back and forth, rather unfortunately, between using ‘categories’ to translate praedicamenta and using ‘predicate’ to translate the verb ‘praedico.’ It’s crucial to see that the same root is being used in both contexts, because that’s the heart of Aquinas’s conception of what’s going on here. We speak of different categories, at least in cases such as these, not because there are distinct, irreducible entities, but because of linguistic differences in how we describe the situation. The penultimate sentences of the passage tells us that it is the motion itself that is predicated of a receiving subject like a table. This all by itself—no appeal is made to some further accidental form—“constitutes the category of Passion.” Likewise, the final sentence tells us that the motion itself, when predicated of its active source, “constitutes the category of Action.” There is no license here for introducing any further entities beyond the motion, the agent, and the patient. What makes for a categorial difference is our different linguistic usages. (The passage makes similar remarks about the categories of Quantity and Where, but I will set those aside given the complexities concerning the Aristotelian idea of locus, and the even greater complexities that surround the category of Quantity.) At this point, one might well feel that my enthusiasm for these texts has taken me too far, all the way to a class-D reading of the categories Action and Passion, which is to say in effect that I have turned Aquinas into a nominalist, at least with respect to some of the lesser categories. This is not what I want. I took pains to stress my agreement with Symington that the categories mark off distinctions among things in the world. Symington’s paper is full of passages that make it quite clear that the ten categories describe features of the world—the categories are not linguistic. Or, at any rate, they are not wholly linguistic. Here is one revealing way in which Aquinas, in his Metaphysics Commentary, formulates his commitment to category realism: “being is delimited into different genera in accord with different modes of predicating, which depend on different modes of being.”14 The passage signals that the theory of categories depends both on a difference in modes of predication and a difference in modes of being. I hope it is fairly clear, at least for Action and Passion, how differences on the side of predication contribute to categorial difference. But now the worry is whether I have something adequate to say about how differences on the side of reality play a role. Here is where I need class C, structures. On the one hand, I cannot see any encouragement 14

“Unde oportet, quod ens contrahatur ad diversa genera secundum diversum modum praedicandi, qui consequitur diversum modum essendi” (In Meta. V.9.890).

Robert Pasnau

71

in Aquinas for the idea that there are distinct accidental forms corresponding to actions and passions. On the other hand, Aquinas clearly tells us that he takes categorial distinctions to be grounded in distinct “modes of being.” The solution, I believe, is to articulate a kind of ontological commitment that is less than category realism, but more than mere nominalism. Hence I offer the notion of a structure. We have a difference of structure when we distinguish between the table as made by Mary, and Mary as maker of the table. The basic ingredients are the same, in each case, but we are describing the world differently, and that description corresponds to a difference in the arrangement of the entities under discussion. If it is the Action we are interested in, then we are describing the motion as emanating from Mary. When we focus on the Passion, then we are focused on the motion as it unfolds within the table. The whole story involves a single motion running from Mary to the table, but inasmuch as we focus on one part of that story or another, we are picking out the Action or the Passion. In my book, I describe these structures as “ontologically innocent,”15 by which I meant to say that they did not commit Aquinas to anything more by way of basic, irreducible ontology. In the context of my dispute with Symington, I need to stress that these structures are things in the world. When we invoke the category of Action in talking about Mary’s carpentry, we are talking about something real. The action is not to be identified with the motion, because if we say that then—given that there is just one action running from Mary to table—we can no longer distinguish between the action and the passion. So the action is the motion as it emanates from Mary. Call this structure an entity, if you like—or, better, call it a “mode of being”—because it is something real in the world. But it is a wholly reducible entity, in the way that an army can be reduced to its constitutive parts. There are lots and lots of structures in the world, as I am thinking of them, and most are not given a seat at the categorial table. On Aquinas’s approach, this is because they do not have the right sort of status in our predicative practices. Here is where Aquinas’s theory, as I understand it, does depend crucially on language. Although one can find, for each of the ten categories, some sort of corresponding mode of being—whether that be substance, accident, or structure—the ground for the ten-fold division comes out of language rather than ontology. This, as it seems to me, is how it should be. Only someone wholly besotted with the authority of Aristotle could suppose that the world itself divides neatly into the ten-fold 15

E.g., at p. 232.

72

Response to Arlig and Symington

categorial scheme. But it is not absurd to think that Aristotle has set out at least some of the more fundamental ways we have of conceptualizing and talking about the world. This brings me, finally, to Symington’s interesting closing remarks about the relationship between language and ontology. He remarks that I “seem resistant to entertaining the idea that conceptual or linguistic structures can be a valid way of articulating and establishing ontological concepts” (p. 55). He is certainly right to detect a certain amount of ambivalence on my part regarding this project. I do think that it is reasonable to suppose that what we ordinarily say about the world is largely true. To this extent, I accept the program of going from language to world. This is another way of describing a commitment to common-sense ontology—that we want to articulate a story about the world that more or less validates the way we ordinarily talk about the world. It is one of the main themes of my book that the scholastic era—for all its baroque subtleties—is ultimately aimed at making sense of our ordinary ways of talking about what exists in the world. When scholasticism collapsed in the seventeenth century, philosophers like Hobbes and Descartes (and, more egregiously, Spinoza, Leibniz, and Berkeley) were left with almost no capacity to accommodate our commonsense picture of what the world is like. But though I feel the force of our commonsense frameworks, I am at the same time extremely suspicious of the idea that language can serve as a guide to what fundamentally exists. Symington seems to think that we can use our linguistic practices to get insight into the basic, irreducible features of the world. I have no confidence whatsoever that this project works; I can see no reason to think that our language is so metaphysically astute. And it seems to me an attractive feature of Aquinas’s view that he likewise wants to avoid making that assumption.

PART TWO: SUBSTANCE ONTOLOGY, MEDIEVAL AND MODERN

AN ARGUMENT FOR HYLOMORPHISM OR THEISM (BUT NOT BOTH) TRAVIS DUMSDAY

Abstract Substratum theory remains a key competitor in the substance ontology literature. Here I argue that an internal worry for the theory gives rise to an interesting dilemma: Either (1) the substratum theorist should abandon the theory in favor of hylomorphism, or (2) she can keep substratum theory but must add to her ontology a powerful causal agent or agents able to operate outside the laws of nature (which would get us part of the way to theism, and at the very least a denial of metaphysical naturalism).

1. Introduction The principal competing theories in the contemporary substance ontology literature are substratum theory,1 bundle theory,2 primitive substance 1

For readers who may want a bit of a refresher: in substratum theory the basic idea is that a whole material object is a compound of a bare substratum (also known as a ‘bare particular’ or ‘thin particular’) and a set of properties. (Read ‘properties’ here as ‘attributes’ or ‘contingent accidents’—in this context they are not to be equated with the propria of Scholastic ontology.) Some substratum theorists take the properties to be universals, and others tropes. The substratum is seen as unifying the properties and, if one takes the properties to be universals, individuating the substance. Recent advocates include David Armstrong, A World of States of Affairs (Cambridge: Cambridge University Press, 1997); Michael LaBossiere, “Substances and Substrata,” Australasian Journal of Philosophy 72 (1994): 360-370; C.B. Martin, “Substance Substantiated,” Australasian Journal of Philosophy 58 (1980): 3-10; J.P. Moreland, “Theories of Individuation: A Reconsideration of Bare Particulars,” Pacific Philosophical Quarterly 79 (1998): 251-263, and Universals (Montreal and Kingston: McGill-Queen’s University Press, 2001); Timothy Pickavance, “In Defence of ‘Partially Clad’ Bare Particulars,” Australasian Journal of Philosophy 87 (2009): 155-158; and Ted Sider, “Bare Particulars,” Philosophical Perspectives 20 (2006): 387-397.

76

An Argument for Hylomorphism or Theism (But Not Both)

theory (sometimes called the ‘substance theory of substance’),3 and hylomorphism.4 The last is of course a staple of Scholastic thought, but increasingly it is also taking its place as a significant player in debates within analytic metaphysics and philosophy of mind. The relevance of these various substance ontologies to natural theology is not often noted, though there is a long history of cosmological arguments made on the basis of hylomorphism,5 and substratum theory has been put 2

According to bundle theorists, a material object is an aggregate of properties, with some viewing the properties as universals and others as tropes. See for instance Keith Campbell, Abstract Particulars (Oxford: Blackwell, 1990); Albert Casullo, “A Fourth Version of Bundle Theory,” Philosophical Studies 54 (1988): 125-139; Arda Denkel, Object and Property (Cambridge: Cambridge University Press, 1996); L.A. Paul, “Logical Parts,” Noûs 36 (2002): 578-596; David Robb, “Qualitative Unity and the Bundle Theory,” Monist 88 (2005): 466-492; Jonathan Schaffer, “The Problem of Free Mass: Must Properties Cluster?” Philosophy and Phenomenological Research 66 (2003): 125-138 ; and Peter Simons, “Particulars in Particular Clothing: Three Trope Theories of Substance,” Philosophy and Phenomenological Research 54 (1994): 553-575. 3 In primitive substance theory a material object is neither a compound of substratum and attributes nor a bundle of properties. Rather, the category of ‘substance’ is taken to be a primitive. Its properties are primitively unified, and the substance is primitively individuated from other substances of the same kind. See for instance Justin Broackes, “Substance,” Proceedings of the Aristotelian Society 106 (2006): 131-166; John Heil, From an Ontological Point of View (Oxford: Oxford University Press, 2003); Michael Loux, Metaphysics: A Contemporary Introduction 2nd ed. (London: Routledge, 2002); E.J. Lowe, The Four-Category Ontology: A Metaphysical Foundation for Natural Science (Oxford: Oxford University Press, 2006); and Cynthia Macdonald, Varieties of Things: Foundations of Contemporary Metaphysics (Oxford: Blackwell, 2005). 4 By ‘hylomorphism’ I will mean Aristotelian-Thomistic hylomorphism, according to which a physical substance is a compound of two distinct yet inseparable substantial metaphysical co-principles: prime matter (a basic potency for becoming a physical substance) and substantial form (a basic principle according to which prime matter is actualized). I certainly recognize that there are other historically important versions of hylomorphism, some of which retain contemporary defenders, but they will not be under consideration here. 5 For a summary of mediaeval cosmological arguments from hylomorphic composition see Herbert Davidson, Proofs for Eternity, Creation, and the Existence of God in Medieval Islamic and Jewish Philosophy (Oxford: Oxford University Press, 1987), 146-153. For recent applications see Norris Clarke, The One and the Many: A Contemporary Thomistic Metaphysics (Notre Dame: University of Notre Dame Press), 183-184; and Edward Feser, “Existential Inertia and the Five Ways,” American Catholic Philosophical Quarterly 85 (2011): 237-267.

Travis Dumsday

77

to such use as well.6 Here I will be dipping into some of the same waters. I am going to argue that a substratum theorist is driven by internal features of her position either to the abandonment of that theory in favour of hylomorphism, or to the retention of substratum theory but with the addition of theism (or at least something approaching theism). Consequently, my argument will be an argument for hylomorphism or theism but not both. It is an argument whose target audience is thereby somewhat circumscribed, holding no appeal for bundle theorists and primitive substance theorists. Still, if sound it should retain some interest. In the next section I lay out my case, then in section three I take up potential objections, and finally in section four I make some brief concluding remarks.

2. The Argument Substratum theory, like each of the major substance ontologies, faces a number of serious objections. One longstanding worry has to do with the prima facie conflicting characterizations of substrata. On the one hand, substrata are considered to be inherently bare, devoid of any intrinsic attributes. On the other hand, substrata are thought to be receptive of attributes. And this immediately gives rise to a tension within the theory: how can substrata be intrinsically bare, devoid of essential attributes, when they seem to possess, intrinsically, a dispositional property, namely the power to receive properties? If a bare particular is able to receive properties, then it sounds like it has a passive causal power, an ability. That in turn seems something over and above bare particularity. Is a bare particular genuinely bare or not? If it is, it seems unable to receive properties, and therefore unable to ground them. But the notion that properties need grounds, that they cannot exist freestanding, is a (probably the) central point of disagreement between substratum theorists and bundle theorists, and will not be lightly abandoned by the former. So the substratum theorist must affirm that the substratum is inherently able to receive properties. But to do so seems an abandonment of the theory, for now a property is being attributed to the substratum itself—it is no longer regarded as genuinely bare, but rather inherently characterized. 6

See William Vallicella, “Three Conceptions of States of Affairs,” Noûs 34 (2000): 237-259; also see his “From Facts to God: An Onto-Cosmological Argument,” International Journal for Philosophy of Religion 48 (2000): 157-181.

78

An Argument for Hylomorphism or Theism (But Not Both)

How might a substratum theorist reply to this worry? Probably the best option is to claim that even though a substratum can truly be called ‘receptive’ of properties, this receptivity is not explained by any literal causal power on the part of the substratum;7 or, to put it in the lingo of contemporary truthmaker theory, the truthmaker for this dispositionascription is not a dispositional property. Rather, the substratum is itself the truthmaker for the disposition-ascription, and primitively so. There is nothing about the substratum that makes it receptive of properties—it just is receptive of properties, and there is nothing more to be said. Every ontology needs certain primitives, and this is simply one of the (hopefully few) that must be admitted by substratum theorists. This sort of reply is vulnerable to several possible counter-replies. One is that while it may be that any ontology needs primitives, primitive counterfactuals are especially worrisome. To say that substratum x would, if acted upon in manner y, receive property z, is to assert a counterfactual of the substratum. And in this case, it is to do so while claiming that nothing really explains why this counterfactual obtains rather than not. To many, this sort of move will seem unsatisfactory in the extreme. Indeed, contemporary truthmaker theory developed largely out of the desire to formulate a more precise objection against primitive counterfactuals, specifically Rylean disposition-ascriptions.8 To employ a concrete example, take for instance the (roughly stated) counterfactual truth that a bucketful of water would help put out a campfire if dunked on it. Most would find it unsatisfying to be told that this counterfactual holds true not because of the active and passive causal powers of water and fire, or the obtaining of certain laws of nature, etc., but instead that it holds true simply as a matter of primitive fact, subject to no further explanation. Yet we are being asked to accept something comparable of the substratum (capacity to receive properties rather than capacity to put out flames), and so, by extension, of every physical object in existence. Granted, some are already committed to primitive counterfactuals— Molinists, for instance, given their belief in primitively true counterfactuals of creaturely freedom; likewise, some recent views in the metaphysics of science literature posit primitively true counterfactuals, notably Lange’s

7

For clear statements of this reply see LaBossiere, “Substances and Substrata,” 368-369, and Moreland, Universals, 153-154. 8 For the relevant history, see David Armstrong, Truth and Truthmakers (Cambridge: Cambridge University Press, 2004) chapter 1.

Travis Dumsday

79

ontology of laws9 and Whittle’s non-Humean causal nominalism.10 However, it is the very positing of such counterfactuals that constitutes one of the most important criticisms of these theories, such that if a substratum theorist must rely on primitive counterfactuals this will surely count as a blow against her theory. There are primitives, and then there are primitives. Perhaps a more promising reply would be to claim that the substratum is receptive of properties not because it has some dispositional property making it receptive, and not because the receptivity is a primitive fact, but because the substratum just is a power for the reception of properties. To say that a substratum ‘is’ receptive of properties is to employ the ‘is’ of identity, not the ‘is’ of subject / object predication. With that reply, there is no longer a need to think of the substratum’s receptivity as something over and above the substratum, something that would challenge its status as inherently propertyless. However, that reply too is problematic. For, prima facie, powers are dispositional properties, such that now substrata seem to be properties, and substratum theory is reduced to bundle theory. Is there another way out for the substratum theorist? Well, at this point the hylomorphist might interject and note first of all that, of the four main substance ontologies, hylomorphism seems the closest cousin of substratum theory. They are both constituent substance ontologies, unlike primitive substance theory; and, unlike bundle theory, they both reject the idea that properties can exist on their own.11 And hylomorphism seems tailor made to avoid the present problem, for on hylomorphism prime matter just is a capacity for reception—except it’s not a capacity for the reception of properties (at least not directly), but rather of substantial form. Prime matter is basically a capacity to become a substance. It is thus a substantial metaphysical co-principle, and there is no risk of prime matter being seen as a property, and hence of hylomorphism reducing to bundle theory. That risk is further mitigated when one recalls that on AristotelianThomistic hylomorphism, prime matter cannot exist except as actualized 9

See Marc Lange, Laws and Lawmakers: Science, Metaphysics, and the Laws of Nature (Oxford: Oxford University Press, 2009). 10 See Ann Whittle, “Causal Nominalism,” in Dispositions and Causes, ed. Toby Handfield (Oxford: Oxford University Press, 2009), 242-285. 11 I am here prescinding from complications that arise within eucharistic theology, where, for some hylomorphists, a certain exception to the rule applies.

80

An Argument for Hylomorphism or Theism (But Not Both)

by some substantial form. It cannot exist on its own. Hylomorphism thus preserves the commitment of substratum theory that there is some nonqualitative constituent necessarily present in all physical substances which is involved in individuation and which is receptive of attributes.12 It preserves that commitment while avoiding the problems it seems to entail for substratum theory. In other words, one honourable way out for the substratum theorist is to shift allegiance to hylomorphism, a smaller and less traumatic jump than would be involved with a shift to any of the other competing substance ontologies. So one might exploit the present difficulty for substratum theory in order to make an argument for hylomorphism. However, there is at least one other option open to the substratum theorist, and that is to claim that the truthmaker for the substratum’s receptivity is not internal to the substratum, but external. Perhaps there is something wholly distinct from the substratum that is able to bestow properties upon it regardless of the fact that the substratum is genuinely bare, lacking any intrinsic capacities. The substratum would therefore be receptive of properties due to something wholly other than itself. This would have to be an entity able to act on an object even when that object lacks passive powers relevant to the reception of the action. And that would make it a very special kind of causal agent. To see why, let’s first assume, quite plausibly, that dispositionalism is true, that objects possess real, irreducible causal powers.13 If one is a dispositionalist, one holds to some approximation of the following schema: an object A acts on an object B when object A has some active power x to act in a certain way, and that power is actualized in part because object B has some passive capacity y to be affected by the action. Most dispositionalists maintain that this view also provides for an ontology of laws, according to which laws of nature are descriptive of regularities grounded by the dispositional properties of objects. Why does an apple fall 12

Though on hylomorphism this receptivity is indirect, since it is first a receptivity of substantial form, which results in the instantiation of a compound substance, which substance is then the direct ground of accidents. 13 For defenses of dispositionalism see for instance Norris Clarke, The One and the Many; Alexander Bird, Nature’s Metaphysics: Laws and Properties (Oxford: Oxford University Press, 2007); Brian Ellis, Scientific Essentialism (Cambridge: Cambridge University Press, 2001); Stephen Mumford, Laws in Nature (New York: Routledge, 2004); and David Oderberg, Real Essentialism (London: Routledge, 2007).

Travis Dumsday

81

to the ground when dropped? According to the dispositionalist, it is not because of Newton’s law of universal gravitation; the abstract equation in no way literally contributes to the actual falling. Rather, the apple falls because it is a massive body, and mass (very roughly described) is a dispositional property the possession of which enables an object to move toward another massive object proportional to the product of the masses and the distance between them (taking into account too the value of the gravitational constant), ceteris paribus. The law, qua abstract equation, is descriptive of what happens in nature, not prescriptive. It is dispositionladen objects, powerful objects, which do the work, and the resulting regularities can then be captured in the abstract equations we call laws. So, assuming that dispositionalism is true and that the dispositionalist ontology of laws obtains, any causal impact in the material world which occurs in the absence of relevant passive powers on the part of the entities being affected would ipso facto be causal action taking place outside the laws of nature. And in the case of a substratum receiving properties from some causal agent without first having a passive power to receive them, we have a being or beings making things happen without any reliance on passive receptive powers. This is a paradigm example, in fact, of causal action taking place outside the laws of nature. Now, since no selfrespecting metaphysical naturalist would countenance causal activity occurring outside the laws of nature, metaphysical naturalism is false. While hardly a full-blown proof of theism, if the preceding point holds it does show that there is a powerful being or beings able to operate outside the bounds of physical laws, which gets us at least one divine attribute. And one could probably extend the argument a bit further to arrive at another divine attribute, immateriality. For, in order to avoid a vicious regress, it must be supposed that whatever is playing the role of external truthmaker here does not itself have a substratum. That might be thought to constitute an argument against substratum theory (‘look, a causal agent, a substance, lacking a substratum!’), but such would be the case only if the substratum theorist implausibly held that the theory is meant to apply not only to material substances, but to all types of substance. So if one properly restricts the scope of the theory, and if one assumes that the external truthmaker is immaterial, there is no further problem for substratum theory and we conclude to a further divine attribute.14 14 This issue does not arise for Aristotelian-Thomistic hylomorphism, given Aquinas’ explicit teaching that immaterial substances are not matter/form

82

An Argument for Hylomorphism or Theism (But Not Both)

To summarize the argument: Premise 1 According to substratum theory, substrata are receptive of properties. Premise 2 If substrata are receptive of properties, then there must be some truthmaker for that receptivity, i.e., something that explains why they are receptive. Conclusion 1 / Premise 3 Therefore, there must be some truthmaker for that receptivity, i.e., something that explains why they are receptive. Premise 4 Whatever explains why substrata are receptive of properties must either be internal to substrata or (exclusive ‘or’) external to substrata. Premise 5 If one attempts to posit an internal truthmaker, one is led to affirm hylomorphism. Premise 6 If one attempts to posit an external truthmaker, one is led to affirm the existence of a being or beings able to operate outside the laws of nature (where ‘laws’ are understood in a dispositionalist sense), which affirmation is inconsistent with metaphysical naturalism and could be used as part of a cumulative case for theism. Final Conclusion Therefore, substratum theory leads either to the proposition that hylomorphism is true or (inclusive ‘or’) that metaphysical naturalism is false.

Each step of the argument certainly admits of further expansion and defense. But let’s turn now to a few of the many objections looming.

3. Objections (3.1) At this point, a substratum theorist who is convinced by the argument but strongly opposed to hylomorphism and devoted to metaphysical naturalism can simply shift allegiance to bundle theory, or primitive substance theory, or to a thoroughgoing agnosticism about substance ontology.

True. The ideal target of the argument would be a substratum theorist who finds both bundle theory and primitive substance theory deeply countercompounds, but are simple substantial forms. Hylomorphism applies only to the material realm.

Travis Dumsday

83

intuitive, and who is inclined to think that the four major substance ontologies in the current literature are the only options—that no plausible, radically different substance ontology is out there waiting to be discovered. (3.2) If the substratum theorist goes the theistic route, wouldn’t that entail the implausible idea that every instance of property-acquisition on the part of a substratum requires divine intervention? And wouldn’t that result in something like occasionalism?

It needn’t entail full-blown occasionalism—something more like a souped-up concurrentism perhaps. Object A can still causally impact object B via the active power of A and the passive power of B. It’s just that God’s concurrence in the action will go beyond merely conserving objects A and B in existence, and even a bit beyond the further requirements of Thomistic concurrence, to the idea that the substratum partly constituting A requires the external action of God to make it serve as ground of the new properties being introduced via the physical process. This does not eliminate causal action in the world, nor does it eliminate the necessary role of the substratum as bearer of properties, but it does introduce an extra factor, an external factor, needed to make the substratum actually play that role. Of course, if the substratum theorist is really opposed to all this, she can simply turn to hylomorphism. (3.3) Surely it’s impossible for any being, including a substratum, to have something done to it (like being made to receive a property) without first being able to have that done to it. The present argument should just be billed as an argument for hylomorphism, not as an argument for hylomorphism / wild-eyed argument for theism.

I have no major qualms with someone seeing this paper simply as a new argument for hylomorphism. However, I’m not as convinced by the opening intuition the objector is employing. Consider: most would grant that any material object can be moved. Now, in our physical world objects can move each other because they have active and passive powers relevant to attraction and repulsion, like mass and charge. Now imagine that God takes one of these objects, an electron say, and removes all its attractive and repulsive powers. Accordingly, no possible physical object can now move it. For all physicists could ever discover, it is immobile. And yet God could still move it, by fiat—He could simply declare that it move, and it would move, and this regardless of its lack of motive power. How does God do this? In this case it’s probably best seen as a form of creation ex

84

An Argument for Hylomorphism or Theism (But Not Both)

nihilo, but a creation of an event rather than an object. In any case, as long as an omnipotent being exists, we can rest assured that any physical object is mobile, even if it lacks inherent powers relevant to mobility. I would suggest that something analogous is open to the advocate of bare substrata in the case described above. Many more objections could be provided, and I make no claim to having fully dispatched the few just considered. Still, for reasons of space I had best leave things there.

4. Conclusion Aside from any utility of the actual argument I’ve made, I hope that the preceding will help spark some further reflection on the relevance of substance ontology to natural theology. As noted in the introduction, historically this has been fertile ground for cosmological arguments of various kinds, and there is a great deal more work to be done exploring the possibilities opened up by contemporary substance ontologies and their relationships to one another. Hylomorphism will have an important role to play here, in addition to its already growing profile within contemporary analytic metaphysics.15

15

I would like to express my sincere thanks to my commentator, Gyula Klima, and other attendees at the 2012 ACPA meeting for their helpful comments.

THE RISES AND FALLS OF ANALYSIS AND METAPHYSICS: COMMENTS ON “AN ARGUMENT FOR HYLOMORPHISM OR THEISM (BUT NOT BOTH)” BY TRAVIS DUMSDAY GYULA KLIMA

As I was reading this paper, I couldn’t but be reminded of an old little sketch of mine I wrote when I was asked to present something at a graduate student conference at Notre Dame on contemporary analytic metaphysics. Whether it’s the attitude expressed in it that got me fired there in ’98 (giving me a chance to become a full professor at Fordham in record time) or not, here goes: A recipe for modern metaphysics Here is an easy recipe for a “revolutionary” metaphysics. Take any type of entity you like (and normal people dislike—say, universals, tropes, states of affairs, properties, events, etc.), except good old primary substances (this exception is required for making your theory revolutionary). Declare these to be the primary entities (not that you are willing to endorse, or even acknowledge, various senses of existence), and then “construct” everything else out of these, i.e., make the “bold” [a euphemism for “nonsensical”] claim that everything else should be regarded as some sort of a collection or (to sound more eloquent) as a “mereological whole” of the former. Then anticipate the obvious objections of those who still have not completely lost their minds, and show them how they should “interpret” (i.e., twist and torture) ordinary ways of speaking in order to give your claim the appearance of a coherent theory. The rest is just a matter of deft marketing.

To be sure, whereas this sketch targets all sorts of extravagant, nonsubstance metaphysics, the paper to be commented on has its sights on a contemporary substance-metaphysics: substrate-theory (the “substrate” being a “bare particular”), among its competitors, namely, bundle theory,

86

The Rises and Falls of Analysis and Metaphysics

primitive substance theory, and hylomorphism, to reduce it to a version of the last one or to show substrate theory’s otherwise surprising commitment to theism. I believe this only goes to show what a long way analytic philosophy has come from its roots in logical positivism. Then it was all logical analysis without almost any metaphysics, now it is all metaphysics, without almost any logical analysis. Let me explain what I mean by this last remark. Being tired of all apparently endless and meaningless metaphysical debates, inspired partly by Kant’s antinomies, partly by the progress of positive sciences, logical positivists designated as the only meaningful task remaining for philosophy the logical analysis of the language of true positive sciences and thereby the elimination of pseudo-problems generated by the misuse and abuse of that language in metaphysics. As we all know, the logical positivist movement, originating mostly from the Vienna Circle, blending itself with the “ordinary language philosophy” of Oxford, eventually successfully analyzed itself out of existence to yield the stage in the late seventies and early eighties of the last century for “analytic philosophy”, mostly unified by a few slogans (such as “existence is not a predicate”) and a shared distaste for the history of philosophy and for a vaguely defined opposing trend designated as “continental philosophy”, indicating the European continent. (Perhaps, it’s worth noting here that I first heard it on this continent that the European continent was supposed to have a philosophy of its own when I arrived to this continent from that continent in 1991. Well, what a surprise it was to me then!) But by the late eighties, the anti-metaphysicalist edge of analytic philosophy became dulled by the arrival of essentialism in possible-worlds semantics and the corresponding re-emergence of good old metaphysical issues, such as the problem of individuation, persistence over time and personal identity, essentialism, etc., leading to the emergence of a new chimera: “analytic metaphysics”. Indeed, the beast raised its metaphysical head so forcefully that the “analytic” part of its designation soon became a name-tag only. I remember that when I was at Notre Dame (so this happened in the second half of the nineties), I asked several of my colleagues, and even the then visiting David Armstrong, to provide metaphysically non-committal clarifications of the semantics of the language they were using in describing their metaphysical theories. In response, I was given puzzled looks and declarations strongly reminiscent of the way medieval nominalists characterized the attitude of their realist opponents: we don’t care about names; we go right to the things themselves!—Well, just look

Gyula Klima

87

at the history of late-scholasticism and early modern philosophy to see what good that attitude did for them. So, what can we do to avoid the late-scholastic scenario, going on another cycle of endless and more and more meaningless metaphysical debates until the arrival of another Kant declaring the whole enterprise ill-founded and another Carnap declaring it to be meaningless, to launch another antimetaphysical cycle of meaningless search for meaning to be abandoned yet again for metaphysics, etc., etc.? Why don’t we try both in tandem, i.e., analysis and metaphysics at the same time, as the very designation “analytic metaphysics” would seem to demand? For then we could start by laying down our clearly defined semantic principles (instead of making them up and twisting them around as we go) and engage each other in our metaphysical debates according to the same principles, instead of talking past each other, making clear that whoever is talking according to different semantic principles is just playing a different game. So, for instance, if I specify that by ‘property’ I mean whatever it is that is designated by a predicate P, such that the predication ‘x is P’ is true if and only if the individual designated by x has the property designated by P, then it is clear that no individual can be a bare particular. For of a bare particular the predication ‘x is a bare particular’ must be true; whence, the individual designated by x must have the property designated by its predicate, i.e., it cannot be bare; ergo, etc. As far as I’m concerned, then, this would be the end of the story of bare particulars. But that is certainly not the end of the story for Travis, who deftly forces the bare-particularist into the corner of the hylomorphist or the theist, to the bare-particularist’s greatest surprise. To be sure, if my little reductio above is OK, then the bare-particularist position is selfcontradictory, and so, since from a contradiction anything follows, it should be not surprising that this position may lead to such consequences. But then the bare-particularist would be surprised to learn that his position is self-contradictory in the first place. Actually, Travis’ strategy is initially not very different from my simple reductio: he takes what a bare-particularist holds minimally true of a bare particular, and posits its truth-maker, namely, the disposition of propertyreceptivity. The property in question is either intrinsic to the bare particular, in which case the particular is not bare, or it is the same as the bare particular, in which case this bare property is not a particular, but a property, or it is extrinsic, but always present in the bare particular by a

88

The Rises and Falls of Analysis and Metaphysics

miracle, without the bare particular’s having any intrinsic capacity to receive even this property (namely, the property of property-receptivity). The argument is fine and the paper in its conclusion very neatly shows why any other alternatives apparently left open to the bare-particularist would lead to his self-annihilation. All I claim here in addition is that the poor little guy could have spared us all the trouble by annihilating himself from the get-go, by thinking about what he could possibly mean before he started to speak.

RESPONSE TO GYULA KLIMA’S “THE RISES AND FALLS OF ANALYSIS AND METAPHYSICS” TRAVIS DUMSDAY

Klima here provides us not only an interesting commentary on the particular line of reasoning pursued in “An Argument for Hylomorphism….”, but also a broader reflection on proper method in metaphysics and the consequences of ignoring proper method. His witty ‘recipe for modern metaphysics’ seems (alas) to be followed not infrequently in the contemporary literature. Members of the Vienna Circle, scandalized even by talk of real substances, essences etc., are no doubt rolling over in their graves. However, while I agree with Klima in the ultimate rejection of bare substrata, I do not think it warrants being classed with those theories developed seemingly in accordance with the ‘recipe.’ (Though I take it that, insofar as substratum theorists are at least attempting to put forward a realist ontology of substances, Klima might not class it there either.) The notion that there is something like an inherently propertyless, natureless substrate underlying material objects does after all have a long historical pedigree. For instance, it was sufficiently widely held in ancient circles that we see it bleeding into patristic literature. The following, from Origen’s De Principiis (Book II, Chapter I) could easily be mistaken for the summary of a contemporary substratum theorist: “By matter, therefore, we understand that which is placed under bodies, viz., that by which, through the bestowing and implanting of qualities, bodies exist; and we mention four qualities—heat, cold, dryness, humidity. These four qualities being implanted in…matter (for matter is found to exist in its own nature without those qualities before mentioned), produce the different kinds of bodies. Although this matter is, as we have said above, according to its own proper nature without qualities, it is never found to exist without a quality.” Exchange ‘matter’ for ‘substratum’ and we’re there. (Most contemporary substratum theorists grant that substrata can only exist if they possess properties,

90

Response to “The Rises and Falls of Analysis and Metaphysics”

though Ted Sider is a prominent exception.) Now, one can argue that this is a misinterpretation of Aristotelian hylomorphism, but if so it was a misinterpretation with a broad base of subscribers. Moreover it cannot be denied that there is some philosophical motivation for putting forward such a view—hence its frequent recurrence in different philosophical climates. This long pedigree leads me to wonder if Klima’s stated objection, arising jointly as a semantic and metaphysical point, might be a bit too quick. The substratum theorist might perhaps try to pull him/herself out of the bind he’s presenting by reference to a distinction among various kinds of predication. Thus, he/she might claim that predications of identity (ex. Tully is Cicero) cause no problem for the theory, which is what ‘x is a bare particular’ amounts to. What would more evidently cause a problem would be the predication of an intrinsic property not obviously identical with ‘bare particular.’ That’s the route I’ve chosen to take, claiming that the theory’s seeming commitment to predicating a causal power of the substratum drives it into the dilemma I’ve described. In further correspondence (generously undertaken), Klima has responded to this latter point by saying that ‘x is a bare particular’ is importantly different from ‘Tully is Cicero’ insofar as ‘bare particular’ is a common predicable rather than a proper name or singular term. Hence the substratum theorist still seems committed either to (A) treating ‘bare particularity’ as a sort of property (the property of bare particularity), which is destructive of the theory, or (B) giving a non-ad-hoc account of which predicables are directly referential and why, which is more difficult given the substratum theorist’s apparent commitment to the idea that anything predicated of a substratum will be predicated accidentally of it. I think here the substratum theorist does have to grant that prima facie, bare particulars look like universals in certain respects (objects x, y, and z all are bare particulars, so doesn’t that mean they share in the universal ‘bare particularity’?). Perhaps this is more of a problem for their theory than I had initially thought. However, I believe the typical line taken here is to fall back on a claim that with this we have reached a rock-bottom metaphysical brute fact: bare particularity can be commonly predicated without being a universal, and it’s probably the only such predicable. Nevertheless metaphysics demands, for explanatory reasons (having to do with individuation etc.) that there be some such entity, odd as it may be. I don’t think this is an especially plausible answer, but I’m still not quite

Travis Dumsday

91

convinced that the substratum theorist is caught in an open contradiction either.

REJOINDER TO TRAVIS DUMSDAY’S RESPONSE GYULA KLIMA

By way of a quick rejoinder, I should mention just two points. 1. Without questioning Origen's Aristotelian credentials (which may well be questionable, though), the big difference between Aristotelian prime matter and a bare particular seems to be that a bare particular is supposed to be an actually existing individual that may or may not have any and all of the properties it actually has, whereas Aristotelian prime matter in and of itself (perhaps, “according to its own proper nature”, in Origen's text) is neither actual, nor an individual, nor can it exist without all the forms it can take on, although it can take on any material form. Thus, although prime matter always actually exists as long as any material thing exists, it can only exist as a part of an actually existing material substance (which, by the way, also holds for any material substantial form, with the exception of the intellective soul), in contrast to bare particulars, which are supposed to be able to exist without any properties whatsoever. This renders, by the way, all their properties accidental, in diametrical opposition to Aristotelian essentialism, regardless of any maneuvering with "intrinsic" and "extrinsic" properties, unless intrinsic properties are taken to be essential properties in the good old Aristotelian sense, in which case, however, a bare particular (an existing particular having no essential properties) would be a metaphysical impossibility. 2. The bare particularist's putative evasive move in terms of an “identity theory of predication” is just an epicycle on a doomed theory, and one that would not roll well with the rest of the system. ‘Bare particular’ in contrast to ‘Cicero’ is a common predicable, applying to individuals by virtue of signifying their properties and not by way of the sort of direct reference some (Millians) would attribute to singular terms. But then the bare particularist is committed either to bare particulars’ having the property of bare particularity (the inconsistency I pointed out), or to a strategy of systematically (as opposed to ad hoc) explaining which common predicables are non-descriptive, directly referential and why, which he

94

Rejoinder to Travis Dumsday’s Response

hardly can do, given his contention that predicables of bare particulars are on a par (they are equally accidental, again, regardless of any maneuvering with the “intrinsic/extrinsic property”-distinction).

PART THREE: THE NATURAL THEOLOGY OF THOMAS AQUINAS

THE BURDEN OF PROOF: AQUINAS AND GOD SCIENCE ALEXANDER W. HALL

1. Introduction Thomas Aquinas introduces the Five Ways as demonstrative proofs that disclose the created nature of our world through effects traceable only to God as their source (Summa Theologiae (ST) 1.2.1-2). He uses what medieval logicians termed quia (or ‘that’) proofs, which, by means of demonstration through effects whose existence in some way depends upon that of the proof’s subject, reveal either: (1) that something exists in some manner, e.g., that the moon is eclipsed; or (2) that something is, in an unqualified sense, e.g., that there exists an eclipse of the moon (Commentary on Aristotle’s ‘Posterior Analytics’ (In PA) 2.1).1 In keeping with this understanding of quia proofs, the Summa Theologiae, sets the stage for the Five Ways with the comment that: When an effect is better known to us than its cause, from the effect we proceed to the knowledge of the cause. And from every effect the existence of its proper cause can be demonstrated, so long as its effects are better known to us; because since every effect depends upon its cause, if the effect exists, the cause must pre-exist. Hence the existence of God, in so far as it is not self-evident to us, can be demonstrated from those of His effects which are known to us (1.2.2c).2

Despite Aquinas’s professed intent to demonstrate that God exists, scholars associated with the Radical Orthodoxy movement in theology in 1

Strictly speaking, quia proof shows only (1), whereas (2) is had by a proof ‘whether it is (si est)’ (In PA 2.1). However, ST does not recognize this distinction (e.g., ST Ia.2.2c). I adopt this practice and refer to both (1) and (2) as ‘quia’ proofs. 2 Translations of ST are taken from .

98

The Burden of Proof: Aquinas and God Science

particular argue that Aquinas’s Five Ways are not demonstrative in the strict sense that is discussed in Aristotle’s Posterior Analytics. Catherine Pickstock, for instance, notes that: Since Aquinas admits that demonstration in general requires univocity, one can only assume (unless uncharacteristically he overlooked something rather crucial) that for him . . . demonstration of God is not genuinely apodeictic (related to syllogism) but more dialectical (allowing of probable assumptions).3

John Milbank, for his part, recognizes that: Aquinas does indeed say that we can know many things by the light of natural reason without appeal to faith. And he does indeed say that the theology pursued by philosophy is able, by the natural light of reason alone, to know God as first cause.4

Nevertheless, Milbank adds that, “Exegesis is easy; it is interpretation that is difficult, and Aquinas, more than most thinkers, requires interpretation” (ibid.). And thus on the subject of Aquinas’s Five Ways, Milbank contends that: [Since] scientific demonstration proper depends . . . on a univocity of terms . . . for Aquinas . . . there can only be an analogical or not strictly scientific approach to the divine. Hence, for example, his ‘demonstrations’ of God’s existence can only be meant to offer weakly probable modes of argument and very attenuated ‘showings’ (pp. 454-55).

For Milbank, the proofs require that we already know that God exists by means of an inchoate grasp of the divine essence: Aquinas’s . . . Five Ways taken together . . . only work because motion is understood from the outset as being undergone with a purpose, or for a reason, and on account of a goal in accord with a nature . . . If all these motions are themselves unmoved . . . their purposes, reasons, goals and natures . . . are themselves illusory. In knowing motions, therefore, which are all aims towards perfections (while the latter are only knowable as participations of the supreme end, the supreme good), the first mover is really radically presupposed (p. 459). 3

Catherine Pickstock, “Duns Scotus: His Historical and Contemporary Significance,” in Modern Theology 21, no. 4 (2005): pp. 543-574, at p. 570, n. 4. 4 John Milbank, “Intensities,” in Modern Theology 14, no. 4 (1999): pp. 445-497, at p. 447.

Alexander W. Hall

99

Milbank points to Aquinas’s claim that “To know that God exists in a general and confused way is implanted in us by nature” (ST 1.2.1 ad 1). However, we must take careful note of the fact that this quote is taken from an Article wherein Aquinas insists that God’s existence “needs to be demonstrated (indiget demonstrari)” (ST 1.2.1c), as it is not self-evident for us. A statement is self-evident for us when the predicate is included in the essence (ratione)5 of the subject and the essence (quid est)6 of predicate and subject are known to us. We grasp the truth of the statement when we understand the meaning of its terms, as in the statement that a whole is greater than its parts. But, “we do not know the essence of God” (ST 1.2.2c); hence, that God exists is self-evident only in itself, but not for us. The passage Milbank cites responds to an objection that, since knowledge of God is implanted in us, that God exists is self-evident for us. Aquinas responds that, whereas we do have some confused knowledge of God inasmuch as we desire happiness, which lies in beatitude, nevertheless that God exists needs to be demonstrated. Taken as a whole, then, the Article does not suggest any skepticism on the part of Aquinas as regards proof that God exists. Rather the point is to show why we need to engage in the effort, viz., because that God exists is not self-evident for us. Moreover, as Denys Turner points out, on Milbank’s account “the Five Ways would be simply invalid, since they would fail by the fallacy of petitio principii”7 and it would thus be wrong for Aquinas to speak of them as demonstrations. Turner traces the Radical Orthodoxy contention that Aquinas’s proofs are not scientific to a belief that the status of these proofs as such would entail that Aquinas embraces an onto-theology such as Radical Orthodoxy detects in Duns Scotus, one that supposes “some common conception of being” univocal to God and creatures (p. 26). On this view, 5

All Latin texts are taken from the Leonine editions available online at . The Greek of An.Post. is from Aristotle’s ‘Prior and Posterior Analytics’, a revised text with introduction and commentary by W. D. Ross (Oxford: Clarendon Press, 1949). 6 See In.PA 2.1, n. 5; 2.7, nn. 5-7 for Aquinas’s use of this terminology. To know of x quid est is to know the essence (essentia) of x. 7 Denys Turner, Faith, Reason and the Existence of God (Cambridge University Press, 2004), p. 197.

100

The Burden of Proof: Aquinas and God Science Thomas could not consistently have intended the Five Ways to be logically valid proofs of the existence of God, for . . . logically valid proofs of the existence of God are impossible, short of engagement in a form of ‘Scotist onto-theology’ (p. 199).

Questions of motivation aside, the intent of this essay is to demonstrate two closely related conclusions regarding the status of Aquinas’s Five Ways as scientific in the strict, Aristotelian sense: (1) that Aquinas thinks that we have quia proof that God exists, though (2) the proofs are of a weak type discussed in lectiones seven and eight of book two of his commentary on Aristotle’s Posterior Analytics. I focus on Aquinas’s Five Ways because Milbank targets them in his construal (he allows that the proofs in Summa Contra Gentiles may be scientific): It is by no means clear that [Aquinas] . . . ever engages in ‘metaphysics’ . . . as opposed to commenting on metaphysical treatises . . . The nearest he comes to this might be the Summa Contra Gentiles . . . The Summa Theologiae [on the other hand] should be taken as a more reliable guide at once to Aquinas’s purely theological and to his most mature understanding . . . [Here] the ‘preliminary’ role of metaphysics on its own as establishing God as first cause is now barely gestured towards [1.1; 1.5 ad. 2], and instead the focus is upon the need of sacra doctrina itself to deploy philosophical arguments . . . Even the relative certainty proffered by reason is very weak . . . There can only be an analogical or not strictly scientific approach to the divine (pp. 452-55).

Milbank’s claim that Aquinas barely gestures toward metaphysical proof seems too vague to contest, but I do believe Milbank is wrong to state that there is in the Summa Theologiae no “strictly scientific approach to the divine.” To the contrary, Aquinas follows to the letter stipulations pertaining to scientific proof as outlined in the Posterior Analytics. Nevertheless, I will argue that the quia proofs offered in the Summa Theologiae do seem to be of the weakest type discussed in the former work, as they tell us only that the subject of some effect exists but not what it is (In PA 2.7). But this should not trouble us. Immediately before Aquinas proceeds to the proofs, he tells us that we cannot know God as

Alexander W. Hall

101

God is in God’s essence (ST 1.2.2, ad 3). Finally, drawing inspiration from Richard Cross,8 I consider whether it is possible that the Five Ways provide knowledge of God that is merely extensional, the sense fixed entirely by an accident, yet nonetheless scientific.

1. On the Status of the Five Ways as Scientific 1.1 Quia and Propter Quid Proofs A quia proof answers two types of questions: (1) whether this or that is the case and (2) whether something is or is not in an unqualified sense. When we answer (1), we may ask (1‫ )މ‬why it is the case; and when we answer (2), we may ask (2‫ )މ‬what it is. We answer questions (1‫ )މ‬and (2‫ )މ‬with propter quid (‘on account of what’ or ‘why’) and quid est (‘what is it’) proofs, respectively.9 In all of the cases described “when we find the answers, these are the things we know scientifically” (An.Post. 2.1, 89b36, p. 229).10

1.2 Role of the Middle Term Each time we ask one of these questions, we are asking after a middle term, “for the middle term is the cause, and this is what is sought in all questions” (An.Post. 2.1, 90a6-7, p. 230). (1) and (2) ask whether there is a middle term by which we might derive a quia proof that this or that is the case or, rather, is in an unqualified sense. For instance, we have with the discovery of a sound in the clouds a middle term for a quia proof that there is thunder in the clouds (a proof that this or that is the case), hence: There is a rumbling in the clouds. Thunder is a rumbling in the clouds. Hence, there is thunder in the clouds.

8

“Univocity and Mystery,” in Roberto Hofmeister Pich (ed.), New Essays on Metaphysics as “Scientia Transcendens” (Fédération Internationale des Instituts d'Études Médiévales, 2007). 9 Like ‘quia’, ‘propter quid’ is used with varying degrees of strictness, with Aquinas using ‘propter quid’ where one would expect ‘quid est’ (e.g., ST Ia.2.2c) (see n. 1). 10 Page number references and translation for An.Post and In PA are taken from Commentary on Aristotle’s ‘Posterior Analytics’, translated by Richard Berquist, with a Preface by Ralph McInerny (Dumb Ox Books, 2007).

102

The Burden of Proof: Aquinas and God Science

(1‫ )މ‬and (2‫)މ‬, on the other hand, ask what the middle term is, that is, what is the ‘expression that signifies what type of thing the subject is (in the Greek and Latin, respectively, logos tou ti esti / ratio ipsius quod quid est)’ (An.Post 2.10, 93b29). For, to know what a thing is in this sense is to know why it is. For instance, an eclipse is “a privation of light from the moon when the earth is interposed” (An.Post 2.1, 90a16, p. 230). But if we know this, then when the moon is eclipsed we know also why it is eclipsed, namely, because there exists such a privation. For both quia and propter quid proof “the cause is the middle term . . . and to know scientifically is to know the cause of a thing” (In PA 2.1, p. 234). To ask (1) whether this or that is the case or (2) whether something is or is not in an unqualified sense is to ask “whether there is a middle term” (ibid.), i.e., whether some cause or other is active in the requisite sense; whereas to ask (1‫ )މ‬why it is the case or (2‫ )މ‬what it is, is to ask “what the middle term is” (ibid.), i.e., what is the nature of the cause.

1.3 Principles Immediate and Convertible Quia proof proceeds from principles that are immediate or mediate, on the one hand, and convertible or non-convertible, on the other; whereas propter quid proof proceeds through principles (derived from the quod quid est) that are immediate causes (An.Post. 1.23).11 Principles are mediate when “the middle term is placed outside” the genus of the subject (An.Post 1.13, 78b13, p. 108). Here we learn that something is so, but not why. For instance, Whatever breathes is an animal. No wall is an animal. No wall breathes.

11 My analysis of An.Post 1.13 is not intended to provide a comprehensive treatment of the text. In fact, I believe there is more to Aristotle’s discussion of proof where the middle is “placed outside” than what I have gathered. Nevertheless, it is clear to me that this type of quia proof (which has a negative demonstration from remote causes) does not figure in the Summa Theologiae. I recommend Berquist’s excellent commentary to the reader who wishes to look more deeply into the issues at play here and throughout Aristotle’s text.

Alexander W. Hall

103

The proof does not get at the immediate cause of the conclusion, “consequently there is no scientific knowledge of the why, since knowledge of the why is knowledge of the first cause” (In PA 1.23, p. 105). Convertible terms, for their part, are counterpredicatable. A, B and C are counterpredicatable when “A belongs to every B and every C, and . . . B and C belong to each other, so that every B is C and every C is B, and . . . both of these belong to A, so that every A is B and every A is C” (In PA 1.8, pp. 36-37). When a quia proof proceeds through an effect that is immediate, convertible and “more known than its cause . . . there can be a demonstration . . . of the fact [but not the reason why]” (In PA 1.23, p. 106). Aristotle provides an example of this type of quia proof in terms of a refined version of the cosmology of Eudoxus of Cnidus (c.395-c.337), which accounts for the movements of the stars in terms of the motions of numerous geo-concentric spheres in which the stars are embedded (On the Heavens (DC) 2.7-12). Stars in the outermost sphere are fixed or stationary because, unlike the wanderers (Saturn, Jupiter, Mars, Venus, Mercury), they retain the same position with respect to one another (DC 2.12, 14; In DC 2.12.407; In M 7.9.2564). As the fixed stars are farthest out, the eye strains to see them; hence the stars seem to waver or twinkle (DC 2.8, 290a7-24; In DC 2.12.405). The wanderers, though, on account of their relative proximity, do not seem to twinkle. Accordingly, ‘stars that are near’, ‘non-twinkling stars’ and ‘wanderers’ are convertible expressions, which emerges in a quia proof of the wanderers’ relative proximity from an effect better known than its cause, through principles immediate and convertible, hence: The wanderers do not twinkle. Every heavenly body which does not twinkle is near. Accordingly, the wanderers are near.

A demonstration propter quid, by contrast, accounts for an effect in terms of its cause, proceeding through causes and immediate principles, for instance: The wanderers are near. Every heavenly body which does not twinkle is near. Accordingly, the wanderers do not twinkle.

The Burden of Proof: Aquinas and God Science

104

1.4 Scientific Proof in Aquinas’s Natural Theology Returning to Aquinas’s natural theology, we cannot acquire a proof propter quid that shows us what (and hence why) God is. The gap between creator and creation ensures that we lack the requisite verbal expression signifying what God is. We do, however, have quia proof that God exists: Demonstration can be made in two ways: One is through the cause, and is called ‘propter quid’, and this is to argue from what is prior absolutely. The other is through the effect, and is called a demonstration ‘quia’; this is to argue from what is prior relatively only to us. When an effect is better known to us than its cause, from the effect we proceed to the knowledge of the cause. And from every effect the existence of its proper cause can be demonstrated, so long as its effects are better known to us; because since every effect depends upon its cause, if the effect exists, the cause must preexist. Hence the existence of God, in so far as it is not self-evident to us, can be demonstrated from those of His effects which are known to us (ST 1.2.2c).

Again, The existence of God and other like truths about God, which can be known by natural reason, are not articles of faith, but are preambles to the articles; for faith presupposes natural knowledge, even as grace presupposes nature, and perfection supposes something that can be perfected. Nevertheless, there is nothing to prevent a man, who cannot grasp a proof, accepting, as a matter of faith, something which in itself is capable of being scientifically known and demonstrated (ST 1.2.2, ad 1).

Note Aquinas’s use of the expressions ‘can be demonstrated (demonstrabile est)’ and ‘capable of being scientifically known and demonstrated (secundum se demonstrabile est et scibile).’ Aristotle defines a demonstration (apodeixis) as: A scientific syllogism (sullogismon epistƝmonikon), that is, a syllogism by which we know scientifically (epistametha). It is scientific (epistƝmonikon) in the sense that we know scientifically (epistametha) by possessing it (An.Post.1. 2, 71b17-18, p. 15).

The thirteenth century had four Latin translations of Aristotle’s Posterior Analytics, and Aquinas used at least two of these (those of James of Venice and William of Moerbeke). In each of the four texts, the translator selects ‘demonstrationem’ for Aristotle’s ‘apodeixin’ or ‘demonstration’

Alexander W. Hall

105

and some form of ‘scire’ for ‘epistƝmetha’, ‘we know scientifically’.12 Note as well, ‘propter quid’ and ‘quia’ in Aquinas’s text. Hence the technical sense of Aquinas’s terminology is well established at the time he composed the introduction to his Five Ways. When we recall that the Summa Theologiae is an instructional text (ST Prooemium), Aquinas would seem a poor teacher were he to risk so seriously misleading his pupils as, quite carefully and consistently, to employ the language of demonstrative proof to describe arguments whose conclusions are merely probable. Consider as well the parallelism between Aquinas’s discussions of quia proof in the Summa Theologiae and his commentary on the Posterior Analytics: ST 1.2.2, ad 2

IN PA 1.23; 2.8; 2.1; 2.1

1. When the existence of a

1'. Given two convertible terms, one the cause and cause is demonstrated from the other the effect, there is no reason why the an effect, this effect takes the effect might not sometimes be more known than place of the definition of the the cause . . . Consequently, there can be a cause in proof of the cause’s demonstration through an effect which is more existence. known than its cause. 2. This is especially the case in 2'. A definition is a verbal expression signifying regard to God, because, in what a thing is . . . We see, however, that there order to prove the existence is an idea of a thing other than its definition . . . of anything, it is necessary to [for instance] a verbal expression manifesting accept as a middle term the the thing named, but other than a definition, i.e., meaning of the word, and not an expression which does not signify what the its essence, thing is, but perhaps an accident of it. 3. for the question of its 3'. The questions we ask are equal in number to essence (quid est) follows on the things known scientifically (sciuntur). Now the question of its existence there are four things which we seek to know: (a) (an est). Now the names that it is so (quia), (b) why it is so (propter given to God are derived quid), (c) whether it is (si est) and (d) what it is from His effects; (quid est) . . . When we have discovered what we were seeking through (a) and (c) we know (a) that this is this or (c) that something exists . . . When we know (a) that this is this, we ask (b) why this is this, and when we know (c) that something exists [in an unqualified sense], we ask (d) what it is. We are then asking what the middle term is. 12 See L. Minio-Paluello and B. G. Dod, eds., analytica posteriori, Aristoteles latinus, 4, pt. 1-4 (Desclée de Brouwer, 1968).

106

The Burden of Proof: Aquinas and God Science

4. consequently, in

demonstrating the existence of God from His effects, we may take for the middle term the meaning of the word ‘God’.

4'. The cause is the middle term in a scientific

demonstration, and to know scientifically is to know the cause of a thing . . . [When we ask whether something is] in an unqualified sense [we ask] whether . . . [a] subject exists, and then [assume] . . . a middle term to demonstrate this.

(1.) describes the demonstration of an effect from its cause discussed at (1‫މ‬.). ( 2‫މ‬.) describes the nature of the middle referred to in (2.). (3.) references the move from (c) to (d) discussed in (3‫މ‬.), a move ruled out as concerns God. Finally, (4.) and (4‫މ‬.) both draw attention to the role of the middle as the cause of scientific knowledge. Reverting to the theme of univocity and by way of closing this section of the essay, let us note that Aquinas himself raises and dismisses the line of reasoning that Pickstock contends he embraces, namely, that, having conceded that we lack terms univocal to God and creatures, we likewise should concede that we lack demonstrative proof that God exists: Univocal predication is impossible between God and creatures . . . [For] no name is predicated univocally of God and of creatures. Neither, on the other hand, are names applied to God and creatures in a purely equivocal sense . . . Because if that were so, it follows that from creatures nothing could be known or demonstrated about God at all; for the reasoning would always be exposed to the fallacy of equivocation. Such a view is against the philosophers, who proved many things about God, and also against what the Apostle says: “The invisible things of God are clearly seen being understood by the things that are made” (Romans 1:20) (1.13.5c).

In these few, brief lines, Aquinas tells us that, from creatures, we have demonstrative knowledge of God by proof that is not subject to the fallacy of equivocation, despite the fact that we lack terms univocal to God and creatures. Moreover, scripture tells us such proof is possible and philosophers have succeeded at it. Recall that Pickstock states: Since Aquinas admits that demonstration in general requires univocity, one can only assume . . . that for him . . . demonstration of God is not genuinely apodeictic (p. 570, n. 4).

Pickstock does not state where Aquinas makes the claim that demonstration in general requires univocity. I do not know whether Aquinas says this, but he states in the passage just quoted that the

Alexander W. Hall

107

requirement does not hold for the natural theologian. Whatever Aquinas’s thoughts regarding proof in general, Pickstock seems to be mistaken as concerns his natural theology on this point. Milbank, on the other hand, correctly observes that Aquinas disallows “transgeneric ‘science’ in the strictest sense” (p. 455), basing his claim on Aquinas’s comment that “From effects not proportionate to the cause, no perfect knowledge of the cause can be obtained” (1.2.2, ad 3). Aquinas’s commentary on the Posterior Analytics, (not cited by Milbank) offers a lengthier discussion of this theme: We cannot demonstrate by going from one genus to another . . . Therefore, the genus must either be the same without qualification or the same in a certain respect if the demonstration is to go over to it. Otherwise, this is clearly impossible, since the middle terms and the extremes must be from the same genus. For if they are not per se they will be accidents (1.7.75a38-1.7.75b11, p. 67).

A middle belongs per se to the subject when “the cause of what belongs to a subject is the subject itself or something pertaining to the subject” (In PA 1.10, p. 44). When the middle that unites the extremes is of a different genus than the subject, any connection between subject and predicate is accidental, as Turner aptly notes, “Nothing in the physics of color could ever strictly entail conclusions about the blueness of a mood” (p. 202). Yes, Aquinas disallows transgeneric science. But it does not follow that Aquinas’s proofs are thereby “weakly probable,” as Milbank concludes (455); for, as Turner points out, Milbank thus supposes that the gap between God and creatures is just like that between one genus and another: Milbank’s objection . . . appears to rest on the supposition that if transgeneric demonstration is invalid, then an inference which purported to transgress the boundary between any created genus and God, who is beyond every genus, must by at least the same token be invalid. But this is a . . . non sequitur . . . To suppose without more ado that because an inference is invalid by the fallacy of equivocation if it crosses from generic being to God, who is beyond every genus, is to suppose, without more ado, that the gap to be crossed between one genus and another and the gap to be crossed between generic being and God are logically the same kind of gap, only—one supposes—‘bigger’ in the latter case (p. 200).

108

The Burden of Proof: Aquinas and God Science

Hence the univocity objection seems to falter on both fronts. Pickstock’s contention contradicts Aquinas’s statement regarding univocity and natural theology and, contra Milbank, lack of univocity needn’t lead us to construe knowledge of God and creatures on a transgeneric model. The unsatisfied burden of proof thus appears to remain on the shoulders of Milbank and Pickstock, if they construe Aquinas’s comment that God’s existence is scibile et demonstrabile as anything other than an assertion that the existence of God is demonstrable in keeping with Aquinas’s understanding of the parameters of scientific demonstration as outlined in the Posterior Analytics.

2. Weak Quia Proof and Per Accidens Scientia in Aquinas’s Natural Theology In what follows, I turn first to Aquinas’s understanding of scientific knowledge (scientia) and then to the various types of quia proof that he discusses in his commentary on the Posterior Analytics. I contend the Five Ways most resemble the weak quia proof that Aquinas discusses in 2.7-8.

2.1 Scientia and Per Se Belonging Scientia tells of what is true of something owing to itself, or per se and is of individuals or natural phenomena only inasmuch as they fall into natural kinds: Since demonstrations are about sensible things in the universal and not in the particular, it follows that they are only incidentally about corruptible things and that, per se, they are about eternal things . . . There are demonstrations . . . about things which happen frequently, like the eclipse of the moon . . . [only] insofar as . . . they always exist . . . [That is] the effect never fails to take place when the cause is active (In PA 1.16, pp. 7475).

Hence, scientia is necessary because it relates what is true of the subject per se, with this definitional knowledge gathered through induction: Since the knowledge of universals is taken from singulars, the first universal principles must be known by induction. For this is the way in which the senses produce the universal in the soul, i.e., by taking all the singulars into account (In PA 2.20, p. 341).

Alexander W. Hall

109

Aquinas selects from Aristotle’s discussion of four types of per se belonging three that are relevant to scientia, the first, second and fourth: [ps1] Whatever things belong to what something is, are per se, as the line belongs to the triangle and the point to the line . . . Whatever elements are in the definitions signifying what they are (are per se). [ps2] Some of the things that inhere in subjects are also per se, i.e., things which include their subjects in the definitions signifying what they are. Straight and circular belong to line in this way . . . [ps4] In yet another way, that which is in a thing because of the thing itself is per se . . . As when something slaughtered dies by the act of slaughtering it. For it died because it was slaughtered; it is not an accident that something slaughtered dies (An.Post. 1.4, 73a35-b17, p. 42).

The first manner of per se belonging is identified with formal causality. An entity’s form is signified by its definition, hence “we have the first mode of per se whenever a definition or part of a definition is predicated of a thing” (In PA 1.10, p. 44). Aquinas identifies the second mode with material causality, “as when that to which something is attributed is the proper matter or proper subject of that thing” (ibid.). In this way, a subject is the per se cause of its proper attributes, whose definitions must incorporate their subjects, as the definition of ‘aquilinity’ incorporates ‘nose’. Finally, the fourth mode of per se belonging comprises all four Aristotelian causes, taking in any way that a subject can act as a cause with respect to itself: “He [Aristotle] says that whatever belongs to a thing because of the thing itself (quicquid inest unicuique propter se ipsum) belongs to it per se” (ibid., p. 45). Aristotle stipulates the types of per se belonging relevant to the scientific syllogism at An.Post. 1.4, 73b16-18: In things knowable without qualification, what is said is per se either in the sense that the subjects are in the predicates or in the sense that the predicates are in the subjects because of (the subjects) themselves. Such predicates are necessary.13 13 The Latin translation misconstrues the Greek, which likely leads Aquinas to misunderstand Aristotle on the types of per se belonging at play in the conclusion of the scientific syllogism. See my “Aquinas, Scientia, and a Medieval Misconstruction of Aristotle’s Posterior Analytics.” In Proceedings of the Society for Medieval Logic and Metaphysics 3 (2003): pp. 4-14; reprinted in Gyula Klima

110

The Burden of Proof: Aquinas and God Science

What is “knowable without qualification” is the conclusion of a scientific demonstration, which proposes two types of per se belonging. Aquinas believes the conclusion’s predicate term conceives a proper accident of the entity signified by the subject term. Since the predicate term signifies a property and the subject term signifies this property’s ontological subject, the conclusion has reference to what belongs to the property conceived by the predicate, viz., the entity mentioned in (and thus belonging to) the property’s definition. This appears to be the sense of Aquinas’s claim that demonstrable propositions “include (includunt)” the second type of per se predication (In PA 1.10, p. 46), viz., by naming a proper accident and the ontological subject incorporated in its definition. Again, since properties belong to subjects through the subjects themselves, the conclusion also formulates the fourth type of per se belonging. To these criteria, Aquinas appends the following schematism: When an attribute is demonstrated of a subject through a middle term which is a definition, the first proposition—the one whose predicate is the attribute and whose subject is the definition containing the principles of the attribute—is per se in the fourth mode. The second proposition—the one whose subject is the subject is the subject of the demonstration and whose predicate is the definition of the subject—is per se in the first mode. The conclusion, in which the attribute is predicated of the subject, is per se in the second mode (In PA 1.13, p. 60).

2.2 Types of Quia Proof At An.Post. 1.13, Aristotle begins to consider how propter quid and quia proof differ. Over the course of his commentary, Aquinas identifies several variants of quia proof. Before returning to the consideration of what type of quia proof (or proofs) may be at work in the Five Ways, we must narrow the field. We have seen that not every quia proof uses convertible effects (In PA 1.23). Aquinas considers four instances: (1) the middle converts with the major but exceeds the minor, e.g., Venus is known to be near on account of its non-twinkling; (2) the minor exceeds the middle, e.g., non-twinkling cannot show the nearness of all stars, for not all stars are non-twinklers; (3) the major exceeds the middle, e.g., an entity capable of progressive and Alexander Hall (eds.), Knowledge, Mental Language, and Free Will (Cambridge Scholars Publishing, 2011).

Alexander W. Hall

111

local motion must have a sensitive soul; and (4) the middle exceeds the major, e.g., rapid pulse need not be a sign of fever. Only (1) and (3) can furnish “appropriate (conveniens)” quia proof. In (2) the subject term is of wider extension than the middle and hence the minor premise will be false; whereas in (4) “the effect [rapid pulse] is more general than the cause [fever]”14 and thus needn’t signal the inherence of the latter. Here then we have six types of demonstration quia, one from effects immediate and convertible, four from effects immediate and non-convertible and one from principles that are mediate (when the middle is placed outside). In addition to these, there are quia proofs in subalternated sciences, which needn’t concern us, for Aquinas states that philosophy has its own quia proof that God exists and needn’t borrow from another science (ST 1.1.1c).

2.3 Per Accidens Scientia In Book 2, Aristotle discusses quia proof of what is known to exist solely through its accidents: Sometimes we know per accidens that a thing exists, and sometimes we know that a thing exists by grasping something of the thing itself, e.g., when we know that thunder is a kind of sound in the clouds . . . When we know per accidens that a thing exists, we cannot in any way grasp what it is. For we do not know that it exists (oude gar hoti estin ismen / neque enim scimus quia sunt), and to ask what a thing is without knowing whether it exists is to ask about nothing (An.Post. 2.8, 93a21-27, pp. 26465).

Aquinas, in his commentary on this passage and what precedes, notes that: There are two ways in which we can know that a thing exists without knowing perfectly what it is: in one way, when we know one of its accidents—for example, if we suppose that something is a rabbit because of its speed—and, in another way, when we know something of the essence of the thing. This latter is possible in composite substances. For example, we could know that something is a man because of its rationality . . . But in simple substances, this is not possible, since we cannot know something pertaining to the substance of a simple being unless we know the whole of it [Metaph., 9.1051b23-33] . . . When we know of the existence of things through their accidents, we cannot know in any way 14

See Berquist, pp. 382-83.

112

The Burden of Proof: Aquinas and God Science what they are. For by accidents of this kind, we do not truly know even that they exist. We know that the accidents exist, but accidents are not the things themselves; therefore, we do not know whether it exists (In PA 2.7, p. 268, emphasis mine).

The issue resurfaces in the lectio that follows, when Aquinas has occasion to comment on Aristotle’s claim that: When we know that a thing exists, we ask why it exists. But it is difficult to grasp things in this way when we do not know that they exist. The cause of this difficulty was stated earlier, i.e., we do not know whether they exist or not, except per accidens (An.Post. 2.9, 93b32-37, p. 271).

Aquinas comments that: We see . . . that there is an idea of a thing other than its definition, an idea found either in a verbal expression signifying the meaning of a name or in a verbal expression manifesting the thing named, but other than its definition, i.e., an expression which does not signify what the thing is, but perhaps an accident of it . . . When, by means of such an expression, we grasp the fact that a thing exists, we ask why it exists and, in this way, we grasp what it is. But, as stated above, it is difficult to grasp what a thing is when we do not know whether it exists. The cause of this difficulty, also stated above, is that when we do not know the existence of the thing through something of the thing itself, we do not know without qualification whether it exists or not, but only per accidens, as was explained (In PA 2.8, p. 274).

Stepping back, we may gather that, for Aquinas, we know per accidens that a thing exists when our awareness of it stems solely from its effects.15 When we know something per accidens, we cannot know in any way what it is, for we do not truly know even that it exists, rather, strictly speaking, we know only that certain accidents exist. We can have a quia proof that does tell us that something exists absolutely speaking (an est) but this is possible only in the case of composite things, where we can grasp something of the essence of the thing without thereby grasping the entirety of the essence. From this, it seems clear that Aquinas does not believe we 15

The Aristotelian commentary tradition suggests ways to meet the “something of the thing itself” requirement in proofs of God’s existence (an attempt, I am suggesting, Aquinas foregoes). See David Twetten’s “To Which ‘God’ Must a Proof of God’s Existence Conclude for Aquinas?,” in Laudemus viros gloriosos: Essays in Honor of Armand Maurer, edited by R. E. Houser (University of Notre Dame Press, 2007), pp. 146-183.

Alexander W. Hall

113

have demonstration quia that God exists in the sense of a quia proof that tells us something of what God is. 2.3.1 Per Accidens Quia Proof and The Five Ways As (1) there is no composition in the divine essence and (2) knowledge of God is indirect, through effects, Aquinas’s comments rule out a quia proof that God exists wherein the premises are immediate and convertible. Rather, I suggest, the Five Ways are quia proofs of the per accidens variety. In an earlier work, I argue that Aquinas’s quia proofs that God exists comprise sets of analogical ascriptions, and that, as such, they can meet the requirements of quia proof from immediate and convertible effects.16 But, Denys Turner raises a serious objection to such reliance on analogy.17 Just as the Five Ways appear question begging if we suppose at the outset that effects adduced to show that God exists are effects produced by God, so too one cannot suppose the need for analogy absent a proof that reveals God to be an entity wholly unlike any other. If the premises are analogical from the get go, then we have already supposed the existence of an entity very much like the one we have set out to discover. As Turner notes, None of our human rational procedures for inferring knowledge of God’s attributes—and they are all analogical of one sort of another—can stand on their own as inferences demonstrating that there is a God to be thus talking about. You cannot argue to God’s existence by analogy. Hence, we are able to conclude that no proof of the existence or nature of God can depend upon our knowing in advance that some analogy between creatures and God could hold. To put it another way, if an argument for the existence of God is to succeed, it cannot depend upon analogy: it must demonstrate analogy; it will be an argument to, not from, analogy (p. 207).

That said, before I advance my contention that Aquinas’s weak quia proof is at work in the Summa Theologiae, let us turn briefly to Richard Cross’s 16

Thomas Aquinas and John Duns Scotus: Natural Theology in the High Middle Ages (Continuum, 2007), chap. 3. 17 I am uncertain just how one might answer Turner’s objection. At any rate, such resolution would lie beyond the scope of this attempt to show that Aquinas relies on weak quia proofs of the rabbit/speed-blur type and must remain the topic of another investigation, perhaps such as is underway in Dominc D’Ettore’s working paper “Analogy of Proportionality in Thomas Aquinas’ Prima Via,” which the author has been kind enough to share with me.

114

The Burden of Proof: Aquinas and God Science

criticism of Turner’s own attempt to show how Aquinas manages a quia proof that God exists, based on Turner’s contention that “there can be formal contradiction between two analogically related propositions” (p. 208). Hence, analogous terms can have a commonality of meaning sufficient to support deductively valid inference: What is required of any valid inference is not more than that any term occurring more than once in the premises of a valid inference (the so called ‘middle-term’) is used in the same sense (univocally) on every occasion of its occurrence. It is not required that there be a univocity of terms in premises and conclusion . . . There can be no objection to there being a formally valid inference between premises and a conclusion analogically related to them across the ‘gap’ between creatures and God (Turner, pp. 200, 208).

As I understand Cross, he charges that Turner claims that the sense of the terms in the conclusion of the establishment phase of the cosmological argument is fixed merely extensionally, by identity of reference. Hence ‘first mover’ means simply ‘whatever it is that explains all motion’ (p. 138). Accordingly, for Turner, “there is no core of meaning that is held in common between the premises and conclusion” (p. 138). Cross’ objection seems to be that if the sense of the terms in the conclusion is fixed merely extensionally, we cannot move forward to the identification phase of the argument, whereas, if there is some additional content of meaning, univocity between the premises and the conclusion is required. Cross’ nuanced criticism considers the notion of scientific proof wherein the reference of the subject term is fixed by an accident that is nonconvertible with the subject and it seems to me that this is just what Aquinas has in mind when describing the weak type of quia proof that I think may be at work in the Five Ways. Recall that Aquinas uses the example of a rabbit whose existence we infer from a blur of motion: “We suppose that something is a rabbit because of its speed” (In PA 2.7, p. 268). Hence: A speed-blur exists. Rabbits produce speed-blurs. Therefore, a rabbit exists.

The case is parallel to that of the inference from a rapid pulse to a fever. The effect (here a blur of motion) is more general than the purported cause (a rabbit). Something other than a rabbit or a fever may produce a speedblur or a rapid pulse. At any rate, we do not know that either the fever or

Alexander W. Hall

115

the rabbit exists, because in the respective cases our grasp of what causes the blur and what speeds up the pulse is fixed by the immediate reference of the middle term, which is an accident that does not convert with the subject term. For this reason, such proofs will fail to meet the requirement for strong proof outlined above, namely, that the major term should reference a property of the subject. This seems to be the sense of Aquinas’s commentary: When we know of the existence of things through their accidents, we cannot know in any way what they are. For by accidents of this kind, we do not truly know even that they exist. We know the accidents to exist, but accidents are not the things themselves (ibid.).

So too, when the sense of the term ‘God’ is fixed with reference to motion, an effect non-convertible with the divine essence, we don’t truly know that God exists.18 We know motion to exist, but motion is not God. Nevertheless, and this is important, we do know that what explains all motion exists, just as we do know that what produces the blur or speeds the pulse exists and hence the quia proof does disclose the existence of something, along with its effect and the proof likewise discloses that the effect is the effect of this thing. Again, the something is more than ‘something we know not what’, because we know that the thing in question accounts in the various cases for motion, a blur or the quickened pulse. Of this type of quia proof (if we may identify the weak quia proof of An.Post. 2.8 with the quia proof whose middle exceeds its major in 1.23), Aquinas tells us that, “When an effect can arise from a number of causes, we cannot conclude to any one of them. Thus, we cannot conclude that someone has a fever from the fact that he has a rapid pulse” (1.23, p. 107). So too, from a blur, we cannot conclude that a rabbit exists and from motion we cannot conclude that God exists, but, rather, that ‘whatever it is that explains all motion’ exists. It seems, though, that Aquinas does not intend anything more than this for the First Way and these results are generalizable to the other proofs. Likewise, an analysis of the expression ‘what it is that explains all motion’ such as follows in Questions 3-11 may be what reveals the need for analogy and discloses the unmoved mover to be God.

18

I withdraw this suggestion in my rejoinder to Michael Sirilla as Aquinas’s First Way, though a per accidens quia proof, needn’t mimic the form of the speed-blur illustration of such proof. I now am in general agreement with Sirilla the the aim of the First Way is to show that God, and not just a first mover, exists.

COMMENTS ON ALEXANDER W. HALL’S “THE BURDEN OF PROOF: AQUINAS AND GOD SCIENCE” MICHAEL SIRILLA

In reply to Dr. Hall’s paper, I will first make some comments on his response to Professors Catherine Pickstock and John Milbank and then proceed with some remarks on his construal of “science” in Aquinas’s proofs for the existence of God. As Dr. Hall aptly shows from Aquinas’s own claim (in Summa Theologiae [hereafter STh] I.13.5.corpus), though univocal predication is impossible between God and creatures, nevertheless, since names applied to God and creatures are also not predicated purely equivocally, we can make the move from created effects to demonstrative knowledge of God not (strictly) subject to the fallacy of equivocation. Of course, this is done by analogous predication which constitutes an issue unto itself. To be sure, Aquinas states that “effects must be proportionate to their causes and principles” (STh I-II.63.3) in the sense that “whatever perfection exists in an effect must be found in the efficient cause” (STh I.4.2). Nevertheless, there are diverse ways in which the perfection or actuality in an effect exists in its cause. First, in a univocal way in which the perfection existing in the effect exists in the cause in the same mode of being and in the same species (or kind)—in other words, formally, “as when fire generates fire or man begets man” (Aquinas, In Metaphysicam [hereafter In Meta.] VII.8.1444). Second, the perfection in the effect may exist in the cause but not according to the same mode of being nor in the same kind, as when the form of a house exists in the mind of the builder and in the house itself (In Meta. VII.8.1445). Third, the perfection in the effect exists in the cause more excellently as heat is in the sun more excellently than it is in fire (STh I.6.2). Finally, the effect’s perfection may exist in the cause virtually but not actually as when heat is caused by motion (In Meta. VII.8.1448-9). At least in the second and fourth instances (and, arguably, in the third), the perfection is predicated analogously of effect

118

Comments on “The Burden of Proof: Aquinas and God Science”

and cause. Some Thomists, Edward Feser is one, have called this variable extension of causal perfection in effects the principle of proportionate causality. Further, see Aquinas’s remarks in STh I.4.3 on perfection when said of God, where he argues that created effects resemble God analogously not univocally. For these reasons and others, Dr. Hall’s response to Professor Pickstock regarding non-univocal predication in scientific syllogisms is well-founded. Turning to his response to Professor Milbank’s objection, we should grant with Dr. Hall that Aquinas is not attempting to demonstrate the existence of God by moving from one genus—understood as one of the ten categories of Aristotle—to another. In STh I.3.5.corpus, Aquinas argues that God is not in any genus at all. He says there, “A thing can be in a genus in two ways; either absolutely and properly, as a species contained under a genus; or as being reducible to it, as principles and privations. For example, a point and unity are reduced to the genus of quantity, as its principles.” And a little later, in the same article, Aquinas adds That God is not in a genus, as reducible to it as its principle, is clear from this, that a principle reducible to any genus does not extend beyond that genus; as, a point is the principle of continuous quantity alone; and unity, of discontinuous quantity. But God is the principle of all being. Therefore He is not contained in any genus as its principle.

The existence of God is not established in a “transgeneric” fashion; rather, God’s existence is shown licitly as super-generic. And yet, in this case as well, God is posited as an analogous cause. In fact, Aquinas says (STh I.13.ad 1) that all univocal agents—that is, agents producing effects having the same form and modality as themselves—are reduced to “nonunivocal” (in this case, analogous) agents since only a non-univocal agent can be a universal cause. Otherwise, if the universal cause of all were a univocal agent, it would have to be the cause of itself. Now we arrive at the substance of Dr. Hall’s argument: The “science” of Aquinas’s proof for God. No doubt it is true that, in his commentary on Aristotle’s Posterior Analytics [hereafter In PA] 2.7, Aquinas states that when we are dealing with demonstrations terminating in “simple substances”—namely, things not composed of matter and form—we cannot know a part or “something” of the essence of such a thing, “we cannot know something pertaining to the substance of a simple being unless we know the whole of it” (cf. In Meta. IX [1051b23-33]). This is so since simple substances, being non-material, are not composed of parts.

Michael Sirilla

119

Aquinas concludes from this that, although we know that the “accidents” of such a being exist, the accidents in this case being the effects of a simple being, nevertheless since the accidents are not the simple substance itself, “we do not know whether it exists,” simply speaking. This raises the following urgent question: If this kind of proof—namely, one that proceeds from the effects or accidents of a simple substance to establish the existence of that substance itself—does not tell us whether that simple substance exists, then how can it serve as a proof for the existence of that substance at all, in any respect? The answer quickly follows in PA 2.9 (93b32-37) where Aristotle states, “we do not know whether they exist or not, except per accidens.” This seems to hold out the possibility that we can know, per accidens and indirectly, the existence of simple, nonmaterial substances as the cause of certain effects. Aquinas affirms as much in his commentary on this passage, “we do not know without qualification whether it [the simple substance] exists or not, but only per accidens.” The problem, it seems to me, is this: Dr. Hall likens Aquinas’s proofs of God’s existence to per accidens quia proofs of a rabbit from speed-blur and a fever from a rapid pulse, the latter two proofs used as examples in PA. Yet those proofs are not only weak, they are invalid and therefore not “proofs,” strictly speaking, at all. Why? In the case of Aquinas’s first way, it is one thing to mount a proof demonstrating that some “ultimate cause of motion” exists, but it is another thing altogether to demonstrate that “God” is that ultimate cause of motion. To put this in the form of a syllogism, the following argument demonstrates that something exists as the ultimate cause of all motion: x x x

The cause of all motion in the universe is a single, ultimate cause of motion. The prime mover causes all motion in the universe. Therefore, the prime mover is the single, ultimate cause of motion.

This syllogism is valid and, perhaps upon further investigation, sound. But one further step, along with an additional syllogism is required to identify that prime mover with God. And, in my opinion, Aquinas’s first way implies this step and syllogism, which he asserts somewhat more clearly just a little later in the STh. If this is correct, then it seems to me that the first way is at least a valid proof for the existence of a being that can legitimately be called “God.” First, it ought to be established—as Aquinas does in STh I.13.8.ad 2—that the name “God” signifies “something existing above all things, the principle of all things and

Comments on “The Burden of Proof: Aquinas and God Science”

120

removed from all things.”1 This step then enables the additional syllogism to proceed thusly: x x

x

That which exists above all, is the principle of all, and is removed from all is signified by the name “God.” The prime mover exists above all [above all motion—since prior to metaphysics mobile being is the only kind of being we know with certitude], is the principle of all [motion] and is removed from all [motion]. Therefore the prime mover is signified by the name “God”2

(I might add that this seems to be the reason why Thomas ends each of his five ways in the STh with this precisely-worded conclusion: “This is what everyone understands to be [or calls] ‘God’,” rather than “This is God.”) Now if, in the first syllogism above demonstrating the prime mover as the single source of all motion, the middle term, “[the cause of] all motion in the universe” is convertible with the minor term, “prime mover,” then we have a valid syllogism. And this seems to be the case. But if, in the second syllogism, the middle term, “that which exists above all, is the principle of all, and is removed from all” is not convertible with the minor term, “prime mover,” then we have a syllogism that is invalid for the same reason that the “speed-blur” or “rapid pulse” syllogisms from PA (adduced by Dr. Hall) are invalid—namely, on the grounds of an illicit distribution wherein the middle term, exceeds the minor term, “prime mover,” with the result that the prime mover could be a being that could not be called “God.” Put simply, rabbits are not the only causes of speed-blurs, nor fevers of rapid pulses. If some being actually distinct from the prime mover (and not the prime mover under a different name, such as the “first appetible”) is also a viable candidate for that which exists above all (etc.), 1

I am grateful to Dr. David Twetten for sharing with me this largely overlooked insight of Aquinas. See his essay “To Which ‘God’ Must a Proof of God’s Existence Conclude for Aquinas” in Laudemus viros gloriosos: Essays in Honor of Armand Maurer, CSB (ed. by R. E. Houser [Notre Dame, IN: University of Notre Dame Press, 2007]): 146-83. 2 This syllogism certainly seems to represent the argumentation deployed in order to establish that the prime mover exists and can be called “God” in Aristotle’s Physics books VII and VIII as understood by Aquinas in his commentary on the same. It addition, it is highly likely that it is implied in his arguments establishing the existence of God as the prime mover in his Summa Contra Gentiles I.13 and especially in the first way in STh I.2.3, where he concludes his proof, saying, “this everyone understands to be God.”

Michael Sirilla

121

then the second syllogism, above, is likewise invalid. To put all this in more concrete terms: if the prime mover, whose existence is validly demonstrated, is the highest “sphere-soul” (as some argue it could be, given Aristotle’s eudoxian cosmology3), then it does not exist above all and cannot, therefore, be called “God.” To conclude otherwise is fallacious. Nevertheless, it is not the case that it is impossible to mount a valid per accidens quia syllogism where the middle, nominal term is not convertible with the other terms, as Dr. Hall has ably shown. An example of this is given by Aquinas, who mounts a valid argument that something capable of local motion has a sensitive soul, even though the terms in that argument are not convertible.4 But in the case of the “speed-blur,” the inconvertibility yields invalidity. Thus, in one respect, and this appears to be the tenor of Dr. Hall’s argument, the proof from motion succeeds in validly establishing that there is an ultimate cause of motion. But in another crucial respect—if and only if the further step and additional syllogism mentioned above are not admitted—it fails to establish that this ultimate cause is able to be signified by the name “God” and only by the name “God.” While it certainly has been argued that the proof from motion in Aquinas’s STh I.2.3 establishes the existence of God only per accidens with the full force of that term employed according to Posterior Analytics book 2, chapters 7-9 (and this is the way Cajetan argues), and that the prime mover or the ultimate cause of motion is able to be signified by the name 'God' (and only by that name) not properly at the end of the first way but only after the demonstrations of the attributes of the prime mover in questions (3ff. of Aquinas's STh), nevertheless, it appears inescapable that the final claim, namely, that “this everyone understands to be God” is a logically invalid move if (and only if) one denies the convertibility of the term “prime mover” and the term “that which is above all,” etc. But it clear neither that such a denial is warranted, nor that Aquinas himself left his proof vulnerable to such a basic logical error. 3

See, for instance, David Twetten, “Clearing A ‘Way’ for Aquinas: How the Proof from Motion Concludes to God,” American Catholic Philosophical Quarterly 70 Supplement (1996): 259–78. 4 His proof from In PA 2.7: Things capable of local motion have sensitive souls. X is capable of local motion. Therefore X has a sensitive soul. In this case, the major exceeds the middle, since some things with sensitive souls (for instance, some sea creatures) are not capable of local motion.

122

Comments on “The Burden of Proof: Aquinas and God Science”

A clue to how Aquinas himself understood how this proof (and the other four) functions is given in a passage cited by Dr. Hall. In STh I.2.2.corpus, Aquinas says, “And from every effect the existence of its proper cause can be demonstrated, so long as its effects are better known to us.” Here, Aquinas is setting up his five ways to demonstrate God’s existence, the existence of the being who is the proper cause of the various effects that serve as a posteriori starting-points for each way. Aquinas uses the term “proper cause” many times, for instance: Summa contra Gentiles [hereafter ScG] I.94, “Science is the knowledge of a thing through its proper cause”; ScG II.16: “Insofar as a given effect is more universal, so its proper cause is higher”; but especially, see ScG II.21, “The proper cause of being is the first and universal agent [efficient cause], which is God.” And also, we have in ScG II.42, “The order of the universe must be referred to God as its proper cause, whom we have proved to be the highest good.” Thus, it seems that Aquinas is convinced that the five ways all by themselves and apart from the further demonstrations he employs in questions 3 and following (pace Dr. Hall and Cajetan), not only serve to establish the existence of an ultimate cause (for example, of motion, as in the first way), but also that this ultimate cause is known to be and is called “God” properly (as proper cause) and in a formally valid fashion. And since he is aware of the fallacy of the illicit distribution of terms, it would be unfair to quickly judge the first way as failing on that score. As mentioned above, STh I.13.8,ad 2 is the key to understanding what Aquinas is thinking of when he says at the end of each of the five ways: “this is what everyone calls [or understands to be] God.” In that passage, Aquinas says: “The name ‘God’ signifies the divine nature, for this name was imposed to signify something that is above all, is the principle of all, and distinct from all. For this is what people naming God intend to signify.” In the beginning of Aquinas’s science of God in his STh, where he asks the question an sit Deus, he presents proofs that conclude “and this everyone calls God” in which the minor term is indeed convertible with its middle, in his estimation. I am very grateful to Dr. Hall for producing such a stimulating paper, for inviting me to contribute a few comments, and especially for his copious labor in compiling and editing these proceedings of the Society for Medieval Logic and Metaphysics.

RESPONSE TO MICHAEL SIRILLA’S COMMENTS ALEXANDER W. HALL

Aquinas accepts the truth of statements (1)-(4):1 (1) (2) (3) (4)

We cannot know merely a part of the essence of a simple substance. God is a simple substance. We know nothing of the essence of God. We know that God exists.

(3) and (4) may appear inconsistent. Specifically, it may seem that to know that God exists is to know something of the essence of God. But, Aquinas says that, for simple substances, we can know that they exist from “something outside [praeter] the essence of the thing;” whereas, for composite substances, we know they exist “when we know something of the essence of the thing” (Commentary on Aristotle’s ‘Posterior Analytics’ (In PA) 2.7, p. 268).2 Hence, we can know that God exists, but this needn’t (and cannot) be from knowledge of the essence of God; and thus, on Aquinas’s reckoning, (3) and (4) are consistent. But, in the two cases (concerning simple or composite substances), we arrive at the knowledge of existence by a different path and so Aquinas allows that ‘exists’ has a different sense in each case. Knowledge that a simple substance exists is qualified, inasmuch as it arises from an awareness that certain accidents or effects depend for their existence on the existence of the simple substance, in a manner analogous to that in which our knowledge that something is a rabbit may emerge only through a glimpse of the speed blur it produces while in motion:

1

I defend these ascriptions in “The Burden of Proof.” Page number references and translation for In PA are taken from Commentary on Aristotle’s ‘Posterior Analytics’, translated by Richard Berquist, with a Preface by Ralph McInerny (Dumb Ox Books, 2007). 2

124

Response to Michael Sirilla’s Comments We know that a thing exists without knowing perfectly what it is: in one way, when we know one of its accidents—for example, if we suppose that something is a rabbit because of its speed (In PA 2.7, p. 268).

Adapting Aquinas’s description, we can call this a ‘solely through the accidents (or per accidens)’ quia proof.3 I have called this type of quia proof ‘weak’ and Michael Sirilla aptly notes that the speed-blur and rapidpulse proofs that Aquinas uses to illustrate a through the accidents demonstration are invalid. Yet, it is not the case that this type of proof, as such, is invalid (nor does Sirilla suggest that this is so). Let us look again at the speed-blur proof: A speed-blur exists. Rabbits cause speed-blurs. A rabbit exists.

The middle exceeds the minor and a lack of convertibility, in this case, renders the proof invalid, as something other than a rabbit might have caused the speed-blur. Under the false impression that the First Way should mimic the form of the speed-blur proof, I supposed that it discloses only a first mover, later to be identified with God.4 On reflection, it appears that the First Way (though a per accidens quia proof)5 avoids this difficulty. Without spelling out each move in the proof through a series of syllogisms whose premises rely on various tenets of Aristotle’s physics, we may abbreviate the proof as follows: A first cause of motion exists. The unmoved mover is the first cause of motion. The unmoved mover exists.

‘Unmoved mover’ and ‘first cause of motion’ are counterpredicatable, this is, after all, what is at issue in the First Way; the first cause of motion cannot itself be moved. Again, the first mover is immaterial (ST 1.3.1c) and hence unique (ST 1a.3.3c), so there is no chance that there are two such first movers.

3

“We do not know without qualification whether it exists or not, but only per accidens, as was explained” (In PA 2.8, p. 274). 4 I present this view above (pp. 114-15). I discuss what gave rise to this confusion below (pp. 125-26). 5 See Summa Theologiae (ST) 1.2.2 ad 2.

Alexander W. Hall

125

Here the major exceeds the middle and minor terms,6 the first cause of motion is but one existent out of many. Yet, as we have seen,7 lack of convertibility as such needn’t invalidate a proof, as with Aquinas’s example of a per accidens quia proof than some entity has a sensitive soul: If the middle term extends to less than the major term, an appropriate syllogism results. We can prove that something has a sensitive soul, for example, from the fact that it is capable of progressive motion (In PA 1.23).8

Mistakenly holding that the First Way and speed-blur proofs were patterned after the same form, I concluded that the First Way was likely intended to disclose the existence of a first mover, later to be identified with God: From motion we cannot conclude that God exists, but, rather, that ‘whatever it is that explains all motion’ exists. It seems, though, that Aquinas does not intend anything more than this for the First Way . . . An analysis of the expression ‘what it is that explains all motion’ such as follows in Questions 3-11 may be what . . . discloses the unmoved mover to be God (p. 115).

In reply, Sirilla contends that the Five Ways do prove that there exists an ultimate cause known to be, and properly called ‘God’: Aquinas is convinced that the five ways all by themselves and apart from the further demonstrations he employs in questions 3 and following . . . not only serve to establish the existence of an ultimate cause (for example, of motion, as in the first way), but also that this ultimate cause is known to be and is called “God” properly (as proper cause) and in a formally valid fashion (p. 122).

I believe that Sirilla and I are now in general agreement on this point. Before I turn to this, let me speak briefly about why I tried to reconstruct the First Way after the pattern of the speed-blur proof. Simply put, given its illustrative role, I mistakenly saw Aquinas’s speed-blur proof as representative of per accidens quia proof as such. About such proof, 6

Unlike in Sirilla’s reconstruction of the First Way, wherein all terms are convertible (or counterpredicatable). See above, p. 119. 7 Above, pp. 110-111; 121. 8 Aquinas adduces the shellfish as an example of an entity that is incapable of progressive local motion and yet has a sensitive soul (ST 1a.78.1c).

126

Response to Michael Sirilla’s Comments

Aquinas notes that we don’t truly know that the subject of this type of proof exists: When we know of the existence of things through their accidents, we cannot know in any way what they are. For by accidents of this kind, we do not truly know even that they exist (In PA 2.7, p. 268).

Misled by the speed-blur example, I thought that this failure to grasp that the subject of a per accidens quia proof exists stems from the aforementioned lack of convertibility, which allows that the effect may exist absent the purported cause. But it is false that bare lack of convertibility as such must invalidate a per accidens quia proof (witness the sensitive-soul proof). Hence, it now seems that what Aquinas was getting at is far simpler: we do not truly know that the subject exists because we have no first-hand experience of the subject, as the subject is known only through some accident (per accidens). I am, then, grateful for the opportunity Sirilla has given me to emend my account. I still hold that: Aquinas thinks that we have quia proof that God exists, though the proofs are of a weak type discussed in lectiones seven and eight of book two of his commentary on Aristotle’s Posterior Analytics (p. 100).

However, I would retract the suggestion that per accidens quia proof as such (and the First Way in particular) must fail to isolate (in a formally valid fashion) a particular cause (as the proper cause) for some given phenomena. A few issues require further, brief clarification. Mistaken in the belief that the First Way should mimic the pattern of the speed-blur proof, I stated that: When the sense of the term ‘God’ is fixed with reference to motion, an effect non-convertible with the divine essence, we don’t truly know that God exists. We know motion to exist, but motion is not God (p. 115).

Hence Sirilla notes that I argue that (A) the First Way proves that a first mover exists, whereas (B) the First Way fails to show that the first mover must be God. Sirilla, in response, contends that (C) the First Way uses convertible terms to conclude that God exists. I believe (A) to be true. (B) I now accept only with certain qualifications and, as is evident by my reconstruction of the proof, I do not follow Sirilla as regards (C). Reflection on this last point allows me to close with a few comments on the status of the First Way as a series of analogical ascriptions.

Alexander W. Hall

127

As regards (B), the First Way concludes that: It is necessary to arrive at some first mover that is not moved by anything (necesse est devenire ad aliquod primum movens, quod a nullo movetur) (ST Ia.2.3c).9

Aquinas adds “and this all understand to be God (et hoc omnes intelligunt Deum),” but the First Way does not offer any argument to the effect that the first mover is God, and so, strictly speaking, the shift from a first mover to God is a jump, and I cannot unqualifiedly reject (B). Nevertheless, Sirilla’s point that Aquinas sets out to disclose God as the “proper cause” of motion is well taken and Question 3 clearly supposes that the Five Ways establish this identification,10 which, I think, Aquinas sees as per se nota, in the sense that he outlines at ST 1.2.1c. That is, one who grasps the technical, Aristotelian sense of ‘pure actuality’ can, as Aquinas does, identify the first mover as God.11 As concerns (C), Sirilla and I differ as regards the convertibility of the proof’s terms. On my reconstruction, the major would exceed the minor and middle. For reasons I shall soon explain, this supposed lack of convertibility may prove a virtue: A first cause of motion exists. The unmoved mover is the first cause of motion. The unmoved mover exists.

If my reconstruction is accurate, it may not be the case that the First Way uses a series of analogical ascriptions, as analogical terms that signify God refer uniquely to God in the context of their utterance, whereas here the major exceeds both the minor and middle terms and so it appears that the minor and middle, on the one hand, and the major, on the other, signify distinct yet overlapping classes. Still, analogy is meant to get around such difficulties. If the proof is wholly a series of analogical ascriptions, all of its terms refer uniquely to God and the existence referenced in the major is that existence which God uniquely is, and then the major would not exceed the middle and minor terms, 9

Translations of ST are taken from . “It has been shown above (ST 1.2.3c) that God is the first unmoved mover (ostensum est autem supra quod Deus est primum movens immobile)” (ST 1.3.1c). 11 Aquinas’s work offers a variety of paths toward the identification of God with the first mover. I sketch one in Thomas Aquinas and John Duns Scotus (Continuum, 2007), p. 58. 10

128

Response to Michael Sirilla’s Comments

despite appearances to the contrary. In addition, as the subject of the proof is (ultimately) seen to be God, the proof should draw on analogy, as, for Aquinas, all theological discourse is analogical. Still, Turner’s point is yet apt, as it would seem that a First Way that draws on analogy would presuppose what it purports to prove. Hence, if a nonanalogical reading of the proof is a desideratum, my reconstruction may be of some use. Nevertheless, given these competing demands, whether or not the First Way draws on analogy seems difficult to determine. In closing, I would like to express my gratitude to Sirilla for his very helpful comments and generous correspondence and to David Twetten who participated in our session. It is my hope that their thoughts on these matters bore some fruit in this rejoinder.

APPEN NDIX

Volume 11,, 2013 The Proceeedings of thee Society for Medieval Loogic and Meetaphysics (P.S.M.L.M M.) is the pubblication of th he Society forr Medieval Logic L and Metaphysicss, collecting original o materrials presentedd at sessions sponsored s by the Socieety. Publicatioon in the Proceedings consstitutes prepu ublication, leaving the authors’ rightt to publish (aa possibly moodified version n of) their materials elssewhere unafffected. The Society for Medievall Logic and Metaphysics M (S S.M.L.M.) is a network of scholars founded withh the aim of fostering f collaaboration and d research based on thee recognition that t ™ recovvering the proofound metaph hysical insightts of medievaal thinkers for oour own philosophical thou ught is highly desirable, and, despite the vvast conceptuual changes in the intervvening period d, is still possiible; but ™ this rrecovery is onnly possible if we carefullyy reflect on th he logical frameework in whhich those in nsights were articulated, given g the paraddigmatic diffe ferences betw ween medievaal and moderrn logical theorries. The Societty’s web sitte (http://facu ulty.fordham.eedu/klima/SM MLM/) is designed to serve the puurpose of keeeping each otther up-to-datte on our current projeects, sharing recent r results,, discussing sccholarly questtions, and organizing m meetings. If you are innterested in joining, pleasee contact Gyuula Klima (Ph hilosophy, Fordham Unniversity) by e-mail e at: klim [email protected] © Society foor Medieval Logic L and Mettaphysics, 20113

CONTRIBUTORS

Andrew Arlig, Assistant Professor of Philosophy, Brooklyn College of the City University of New York Travis Dumsday, Assistant Professor of Philosophy, Concordia University College of Alberta Alexander W. Hall, Associate Professor of Philosophy, Clayton State University—Assistant Director and Secretary of the Society for Medieval Logic and Metaphysics and Managing Editor of its Proceedings. Gyula Klima, Professor of Philosophy, Fordham University—Founding Member and Director of the Society for Medieval Logic and Metaphysics and Editor of its Proceedings Robert Pasnau, Professor of Philosophy, University of Colorado, Boulder Michael Sirilla, Associate Professor of Theology, Franciscan University of Steubenville Paul Symington, Associate Professor of Philosophy, Franciscan University of Steubenville