Oxford Studies in Medieval Philosophy, Volume 3 [UK ed.] 9780198743798, 9780198743804, 0198743793

Citation preview

OXFORD STUDIES IN MEDIEVAL PHILOSOPHY VOLUME 3

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

Oxford Studies in Medieval Philosophy Volume 3 Edited by

ROBERT PASNAU

1

OUP CORRECTED PROOF – FINAL, 28/5/2015, SPi

3

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries # the several contributors 2015 The moral rights of the authors have been asserted First Edition published in 2015 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2014959430 ISBN 978–0–19–874379–8 (hbk) ISBN 978–0–19–874380–4 (pbk) Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

Contents Articles Augustine on the Varieties of Understanding and Why There is No Learning from Words Tamer Nawar The Logic of Dead Humans: Abelard and the Transformation of the Porphyrian Tree Margaret Cameron The Fate of the Flying Man: Medieval Reception of Avicenna’s Thought Experiment Juhana Toivanen Mind the Gap? The Principle of Non-Repeatability and Aquinas’s Account of the Resurrection Adam Wood Duns Scotus and Divine Necessity Richard Cross Ockham’s Scientia Argument for Mental Language Eric W. Hagedorn

1

32

64

99 128 145

Critical Notice Silence, Skepticism, and Vulgar Theology: On Stern’s Maimonides Daniel Davies

169

Discussion Aquinas on the Beginning and End of Human Life: A Rejoinder to Patrick Toner Fabrizio Amerini

189

Briefly Noted Smith—Pessin—Grellard—Cory—Uncertain Knowledge

197

Notes for Contributors Index

199 201

Augustine on the Varieties of Understanding and Why There is No Learning from Words Tamer Nawar

1. INTRODUCTION This paper examines Augustine’s account of understanding and its attainment in De Magistro. Following Myles Burnyeat,1 it is often held that Augustine is especially concerned with explanatory understanding (a complex cognitive state characterized by its synoptic nature and awareness of explanatory relations) and that a principal thesis of De Magistro—that there is no learning from words—should be construed not as the claim that testimony is deficient in producing justification in the hearer, but rather as the claim that testimony fails to impart explanatory understanding. Against this view, I argue that in De Magistro Augustine is not in fact especially concerned with explanatory understanding but with various other forms of cognizance (which have hitherto not been clearly characterized or sufficiently distinguished) and that Augustine’s claim that there is no learning from words cannot be construed as typically proposed. I begin by offering a brief, critical account of the received view and observe some of the difficulties it faces (section 2). Next, I distinguish between various different kinds of cognizance which Augustine is interested in— what I term ‘conceptual understanding,’ ‘linguistic understanding,’ ‘explanatory understanding,’ ‘spiritual understanding,’ and ‘infallible knowledge’— and examine Augustine’s discussion in De Magistro in the light of this distinction (section 3). I challenge the traditional interpretation(s) on several points and show that Augustine is initially concerned with basic cognitive states, 1 Myles Burnyeat, “Wittgenstein and Augustine De Magistro,” Proceedings of the Aristotelian Society 61 (1987), 1–24.

2

Tamer Nawar

namely conceptual understanding (a cognitive state constituted by concept possession) and linguistic understanding (a cognitive state constituted by an awareness of the literal meaning of linguistic expressions), while clarifying the precise claims Augustine makes concerning these states. I then argue that when Augustine discusses less basic cognitive states he is not—as is often supposed—discussing explanatory understanding but instead infallible knowledge (a cognitive state which is neither synoptic nor explanatory but is instead characterized by possessing infallible justification). Finally (section 4), I turn to why Augustine should think that there is no learning from words. Insofar as Augustine is not concerned with explanatory understanding, the standard view does not make intelligible Augustine’s thesis: it fails to explain why Augustine thinks these other, nonexplanatory forms of cognizance cannot be satisfactorily attained through testimony. In the case of conceptual understanding, Augustine’s point is that while we may, through hearing words and using our imaginations, form a concept of a thing despite never having seen it, such concepts (in contrast to those formed through perception) are not satisfactorily grounded and whether they accurately represent their objects is down to luck; they are—Augustine thinks—epistemically deficient and their attainment does not deserve the name ‘learning.’ With regard to infallible knowledge (the second of Augustine’s principal focal points in the dialogue), the sort of cognizance brought about by testimony also falls short of the kind of knowledge Augustine has in mind. It is attained by a method the reliability of which is less than absolute and, crucially, when one accepts that p on the basis of testimony, one does not give independent reflection to whether p, and so the reflective grasp of reasons for belief (not, as per Burnyeat, an explanation of the fact believed) required for knowing full well is absent.

2. THE STANDARD VIEW In a seminal article, Myles Burnyeat argued that Augustine took the term ‘scientia,’ in its strictest and most proper sense, to refer to understanding. Drawing upon Augustine’s interchangeable use of the terms ‘scire’ and ‘intellegere’ (something which Burnyeat raised with regard to Retractationes 1.14.3),2 Burnyeat claimed: 2 “And when I said . . . ‘What we know, therefore, we owe to reason, what we believe, to authority’ (quod scimus igitur, debemus rationi, quod credimus, auctoritati) [Util. Cred. 11.25], this is not to be taken in such a way as to make us frightened in mere ordinary conversation of saying that we know what we believe on adequate testimony. It is true

Augustine on the Varieties of Understanding

3

If Augustine feels that it makes no odds whether he writes scire or intellegere, that implies that in his view the proper meaning of scire is intellegere. And that in turn explains why he thinks it loose or improper to use ‘knowing’ (scire) in the ordinary way of what we believe on adequate testimony. Intellegere would not here fit at all. Adequate testimony is excellent justification for believing something, but it does not contribute an understanding of the thing believed. (“Wittgenstein and Augustine,” 7)

On this view, Augustine uses ‘scientia’ and ‘intellegentia’ for the same cognitive state, which I shall call ‘explanatory understanding.’ Explanatory understanding seems to be characterized principally by two elements: first, its synoptic nature, it is not directed at single propositions but rather groups of propositions which form, in some sense, a system; and second, it requires an awareness of explanatory relations among said propositions (Burnyeat, “Wittgenstein and Augustine,” 20). The paradigm case of such explanatory understanding seems to be that of understanding, as opposed to merely parroting, a mathematical proof (e.g. Burnyeat, “Wittgenstein and Augustine,” 21) and the “first-handedness”3 of explanatory understanding is also meant to explain why Augustine (at least in De Magistro)4 thinks there is no learning from the words of another: [Since] Augustine thinks of understanding rather than justification as the differentiating ingredient of knowledge, the main thesis of the De Magistro, that no man can teach another knowledge (scientia), can now be glossed as the claim that no man can teach another to understand something. The argument will not be that information cannot be transmitted from one person to another, but that the appreciation or understanding of any such information is a task that each person must work at for himself. (“Wittgenstein and Augustine,” 7–8)

The thought goes that while one may attain (e.g.) justified true belief through accepting the testimony of another, one cannot attain explanatory understanding this way. Testimony is not, on Burnyeat’s view, deficient in transmitting justification but in transmitting explanation; explanatory that when we speak properly (proprie), we say we know only that which we grasp by firm reasoning of the mind. But when we speak in language more suited to common use, as even the Holy Scripture speaks, we should not hesitate to say we know both what we perceive by our bodily senses and what we believe on the authority of trustworthy witnesses, while nevertheless understanding the distance between these and that” (Retr. 1.14.3, tr. Burnyeat, “Wittgenstein and Augustine,” 6). 3 E.g. “first-hand learning” (Burnyeat, “Wittgenstein and Augustine,” 19); “first-hand appreciation” (ibid., 19); “knowledge must be first hand if it is essentially of connections” (ibid., 21); “like empirical vision, it involves seeing things for oneself” (ibid., 22). 4 In later works (e.g. Ep. 147), Augustine is less negative about testimony and focuses on its virtues rather than its shortcomings. For discussion, see Peter King and Nathan Ballantyne, “Augustine on Testimony,” Canadian Journal of Philosophy 39 (2009), 195–214.

4

Tamer Nawar

understanding is something that a person can attain only by seeing how things fit together for him or herself. This is why Burnyeat’s Augustine insists that scientia cannot be attained through testimony: “Adequate testimony is excellent justification for believing something, but it does not contribute an understanding of the thing believed” (Burnyeat, “Wittgenstein and Augustine,” 7).5 The thought attributed to Augustine is attractive and Burnyeat’s view has been widely followed.6 However, there are several problems which deserve attention. First, Burnyeat appeals to Augustine’s interchangeable use of the terms ‘scire’ and ‘intellegere’ (or their cognate nouns) as evidence that Augustine was concerned with explanatory understanding. However, while it is fairly uncontroversial that ‘intellegere’ often means ‘understand,’ it is unclear why the fact that Augustine often uses ‘scire’ and ‘intellegere’ interchangeably should indicate that Augustine conceives of scientia (or intellegentia) as explanatory understanding. It is equally compatible with Augustine thinking that intellegentia is no more than scientia, or that some other sort of understanding is at issue, or with Augustine simply being loose in using the relevant epistemic terms. Secondly, Burnyeat takes Augustine’s scientia to be a cognitive state which is explanatory and admits of no piecemeal cognition (Burnyeat “Wittgenstein and Augustine,” 20). However, elsewhere (notably in his Contra Academicos) Augustine seems to provide us with non-explanatory, piecemeal instances of cognition as paradigm examples of scientia (e.g. I know this seems white to me, C. Acad. 3.10.23ff.)7 and, as we shall see, in De Magistro Augustine also seems to be concerned with instances of nonexplanatory, piecemeal cognition (e.g. my awareness that ‘walking’ signifies walking, my knowing that a flying man has been seen). Thirdly, Burnyeat’s thesis might explain why testimony is deficient in transmitting explanatory understanding. However, it does not explain why one cannot learn (e.g.) what walking is or what ‘walking’ means from the 5 Cf. Peter King, “Augustine on the Impossibility of Teaching,” Metaphilosophy 29 (1998), 180–2, 190–4. 6 E.g. John Rist, Augustine: Ancient Thought Baptized (Cambridge: Cambridge University Press, 1994), 44–7, 57–60, 87–90; Peter King, “Augustine on the Impossibility of Teaching,” 191–3; Gerard O’Daly, “The Response to Skepticism and the Mechanisms of Cognition,” in E. Stump and N. Kretzmann (eds), The Cambridge Companion to Augustine (Cambridge: Cambridge University Press, 2001), 164; Emmanuel Bermon, La signification et l’enseignement: texte latin, traduction franc¸aise et commentaire du De magistro de Saint Augustin (Paris: Vrin, 2007), 493–6. 7 Even in mathematics, the paradigm cases of scientia Augustine invokes are not proofs or arguments (as commonly supposed) but rather simple arithmetical operations (e.g. C. Acad. 3.11.25). Though one may know for sure that three threes are nine, it seems much harder to explain why this is so.

Augustine on the Varieties of Understanding

5

testimony of another. Insofar as Augustine is interested in non-synoptic, non-explanatory kinds of cognizance, the reasons why Augustine thinks they cannot be appropriately attained through testimony is not adequately explained by the standard interpretation.

3. D E M A G I S T R O A N D T H E V A R I E T I E S O F U N D ER S T A N D I N G In attempting to provide a more satisfactory account of Augustine’s discussion in De Magistro, it will be useful to begin by distinguishing the following kinds of cognition which Augustine addresses or might be taken to address:  Conceptual understanding. To have conceptual understanding of A is to have a concept of A such as to be capable of applying it when appropriate. That is to say, to have a basic form of cognition which allows one to categorize kinds of thing, distinguish between kinds of things, and think of things as belonging to that kind.8  Linguistic understanding. In our terms, this cognitive state is constituted by an awareness of the literal meaning of a meaningful linguistic expression. In Augustine’s terms, this is grasping the significate of a sign.  Explanatory understanding (see above).  Spiritual understanding. A hitherto largely neglected cognitive state constituted, in Augustine’s terms, by grasping the spiritual significate of a sign.  Infallible knowledge. This cognitive state is similar to how we speak of (infallible) propositional knowledge. If S (infallibly) knows that p, then it is necessary that, given the method (evidence, reasons) through which S

8 In De Magistro, Augustine speaks of having a concept of A primarily as either having notitia of A (e.g. Mag. 11.34) or of A as being in notitia (e.g. Mag. 11.37). ‘Notitia’ is used for various forms of cognition; it is elsewhere used to speak of concept possession (e.g. Conf. 11.23.29; Trin. 8.6.9; cf. Duab. An. 8.10, 14.22; B. Vita 4.30; Civ. Dei 10.32, 18.37; C. Faust. 22.56), but also of more sophisticated cognitive states (e.g. Gn. Litt. 5.15.33; En. Ps. 49.18). Notice that what is closest to the standard term for concept is ‘notio’ (e.g. C. Acad. 1.4.11, 1.5.15; Quant. An. 32.65; Lib. Arb. 1.6.15, 1.13.27, 2.9.26; Conf. 10.14.22, 17.26; Civ. Dei 8.7 [where it renders ‘
ØÆ’]; Trin. 8.3.4) but Cicero had also used the term ‘intelligentia’ (e.g. Cic. Fin. 3.21). Notice also that there may be different levels of conceptual understanding (e.g. the difference between bare concept possession and conceptual mastery) and different concepts (some more informative others less informative) of the same object.

6

Tamer Nawar came to think that p, p. For Augustine, infallible knowledge is also rationally certain and indefeasible.9

Several points deserve mention. Firstly, this list is not intended to be exhaustive; secondly, while all these forms of cognizance are seen as the result of the activity of the mind (mens, intellectus), they do seem to be distinct kinds of cognizance;10 thirdly, the distinctness of these kinds of cognizance is often neglected in the scholarship;11 and finally, none apart from explanatory understanding seems to require an awareness of explanations.12 Distinguishing between these cognitive states will be helpful for seeing what is at issue in De Magistro and for providing a framework for seeing what could be at issue. In what follows, I shall challenge traditional interpretations on a number of points. I will first (section 3a) show that in the latter part of De Magistro Augustine is initially concerned with conceptual understanding and linguistic understanding. I will then (section 3b) argue that later in the dialogue (Mag. 11.37) Augustine is discussing infallible knowledge (not explanatory understanding).

9 If S indefeasibly believes that p, then S cannot be dissuaded from believing that p (e.g. inconcussum tenere, C. Acad. 1.7.19; cf. C. Acad. 2.9.27; Quant. An. 26.49, 30.58; Ep. 118.2.12.). If S is certain that p, S believes that p with utter strength and does not entertain any doubts whether p (e.g. C. Acad. 3.10.23; Trin. 15.12.21). For Augustine, these features find their basis in a reflective grasp of reasons (i.e. justification construed in an internalist manner). 10 For most of these, ‘Intellectus’ may be used. However, in De Magistro, Augustine most often uses the terms ‘notitia’ (see n. 8) and ‘cognitio.’ In other works, ‘cognitio’ embraces: perceptual beliefs (Quant. An. 24.45); linguistic understanding (Doc. Chr. 2.11.16–14.21); knowing a text (Ep. 118.2.11); historical knowledge (Trin. 12.14.22, 13.1.2); the inner sense’s awareness of what to flee and pursue (Div. Qu. 61.4); contemplation as contrasted with action (Vera Rel. 17.33; cf. Div. Qu. 58.2); selfawareness (Gn. Litt. 4.22.39); knowledge of the sciences and arts (Doc. Chr. 2.29.45–30.47); infallible knowledge (e.g. C. Acad. 3.13.29); and various other cognitive states. 11 For instance, the scholarship often speaks of understanding as if understanding the meaning of a word were the same sort of thing as justifiably believing that p or having explanatory understanding (e.g. Burnyeat, “Wittgenstein and Augustine,” 17). However, even cognitivist accounts of language typically emphasize that linguistic understanding is very different from justified belief or standard accounts of propositional knowledge (e.g. Noam Chomsky, Knowledge of Language: Its Nature, Origin, and Use (New York: Praeger, 1986), 265–73). 12 For instance, one may have linguistic understanding (something which a parrot lacks) without awareness of explanations (or even of rules). At Doc. Chr. 4.3.4–5, Augustine discusses how an awareness of the rules of eloquence (eloquentia) is not needed for speaking eloquently (and may even be counterproductive); something similar holds of grammatica.

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

Augustine on the Varieties of Understanding

7

a. Conceptual Understanding and Linguistic Understanding De Magistro begins as an inquiry into the purpose(s) of speaking (loqui). Augustine’s interlocutor, his son Adeodatus, initially proposes that we speak either in order to teach (docere) or to learn (discere). In response, Augustine says that he himself thinks that there is a certain kind of teaching that occurs through reminding (commemoratio) and suggests that the purpose of speaking is either to teach or to remind others or ourselves (Mag. 1.1; cf. Doc. Chr. 2.2.3). While some objections are raised to this thesis, the issue of which intentions prompt speaking is not explicitly resolved and the dialogue quickly turns to the nature of words (verba) and signs (signa) (Mag. 2.3ff.) and goes on to take a number of twists and turns.13 In the latter half of the dialogue (8.22–13.46) the themes of teaching and learning return to the forefront. Augustine and Adeodatus first give significant attention to ambiguity and something closely resembling the distinction between the use and mention of a term (Mag. 8.22–4). A distinction is also drawn between: signs (signa); things (res); cognizance of signs (cognitio signorum); cognizance of things (cognitio rerum); and how these compare in value (Mag. 9.25–8). In each case the signs discussed are general terms (e.g. ‘human,’ ‘lion,’ ‘filth,’ ‘vice’) and their significates. From there, the discussion turns to whether signs play a role in learning. Adeodatus initially proposes that signs are necessary for learning (Mag. 10.29). He argues towards this conclusion by questioning how someone wishing to learn what walking is (quid sit ambulare) might learn this merely from seeing a manifestation of walking (Mag. 10.29). Why, for instance, does one not mistake walking for a very particular kind of walking (e.g. strolling) or a particular amount of walking (unde cavebo, ne id tantum putet esse ambulare quantum ego ambulavero? Mag. 10.29)? The difficulty is provisionally conceded and it is supposed that a learner cannot learn what A is merely from seeing A. This being so, Adeodatus supposes that to learn

13 Among the principal questions Augustine addresses are how signs signify their significates and what the nature of the signifying relation and relata are. Augustine (perhaps merely dialectically) entertains the thought that all words act as proper names and that the meaning of a term is its extension (the so-called “Augustinian picture of language,” dubbed so by Wittgenstein). However, elsewhere he seems to favor something like a propositional account of meaning or, alternatively, construes meaning (in Gricean fashion) as the intention of the speaker. For discussion, see Mary Sirridge, “Augustine: Every Word is a Name,” New Scholasticism 50 (1976), 183–92; Burnyeat, “Wittgenstein and Augustine,” 11–13; Christopher Kirwan, Augustine (London: Routledge, 1989), 35–59; Rist, Augustine, 314–16; Bermon, La signification et l’enseignement, 51–7, 166–218, 278–322.

8

Tamer Nawar

what (e.g.) walking is, some signs must be provided (Mag. 10.29–30), and it is provisionally concluded that nothing is taught without signs (confectum est igitur et nihil sine signis doceri, Mag. 10.31). However, Augustine immediately gives Adeodatus reason to doubt whether this conclusion is correct (Mag. 10.32). He proposes that someone unfamiliar with bird-catching (quisquam ignarus deceptionis avium) and wanting to learn what bird-catching is can learn what bird-catching is merely from observing some instance(s) of bird-catching (Mag. 10.32). This occurs, Augustine tells us, “without a sign” (nullo significatu) simply through the spectator seeing the thing itself (Mag. 10.32). Adeodatus does not see how this advances on what was previously discussed with regard to walking (cf. Mag. 10.29) and complains that the spectator has not been shown the whole of bird-catching (neque enim video et hic totum illud aucupium esse monstratum, Mag. 10.32). What Adeodatus finds puzzling in both the case of walking and birdcatching is not, as is often thought, how we understand something about a sign. The inquiry is not about how we learn what ‘walking’ or ‘birdcatching’ means or how we cut through ambiguity or otherwise attain linguistic understanding but instead how we learn what walking or birdcatching is (e.g. quid sit ambulare, Mag. 10.29).14 Not only do these cases concern a world-directed form of cognition (concerning what a thing is) rather than a sign-directed form of cognition (concerning what a sign signifies) but the forms of cognition in question also come about, as Augustine himself emphasizes, without signs (sine signo, Mag. 10.29; nullo significatu [ . . . ] nullo signo datur, Mag. 10.32). Adeodatus’ worry centers on the grounding of our concepts; namely, how it is that, from seeing some instance(s) of walking, one forms an appropriate conception of walking which allows one to recognize all instances of walking (“the whole of walking”) as instances of walking (Mag. 10.29, 32). Similarly, the bird-catching case concerns how we form a concept of a craft as a whole (Mag. 10.32) merely from seeing some example(s) of it. How is it that, from seeing only some part of the 14 Augustine has here often been taken to offer a close anticipation of Wittgenstein’s insights (Phil. Inv. }26–32) about the ambiguity of ostensive definition, e.g. Burnyeat, “Wittgenstein and Augustine,” 12–13; Kirwan, Augustine, 53; Gareth Matthews, Thought’s Ego in Augustine and Descartes (Ithaca, NY: Cornell University Press, 1992), 153–8; Gareth Matthews, “Knowledge and Illumination,” in E. Stump and N. Kretzmann (eds), The Cambridge Companion to Augustine (Cambridge: Cambridge University Press, 2001), 181; Bermon, La signification et l’enseignement, 58–9, 220–1, 350–3, 384–7. However, this is slightly misleading. There is no ostension, signification, or linguistic act of any kind in these cases; therefore ambiguity, as usually understood, cannot be at issue.

Augustine on the Varieties of Understanding

9

things which are F, we form an accurate conception of F-ness? To this question, Augustine replies: It’s easy to get rid of your worry. I add that if he is sufficiently intelligent, he recognizes the kind of craft as a whole on the basis of what he has seen (si ille intellegens esset, ut ex hoc quod uidit totum illud genus artis agnosceret). It’s surely enough for the matter at hand that some men can be taught about some things, even if not all, without a sign. (Mag. 10.32)15

Previously, at Mag. 10.29, it was assumed that S ’s forming an adequate conception of walking could not be explained merely by S ’s seeing some instance(s) of walking. Adeodatus further supposed that if S were to form an adequate conception of walking, then S would have to be presented with certain signs. However, in this case, no signs are presented and instead of attempting to appeal to signs to explain how S attains conceptual understanding, Augustine appeals to S’s perceptual acquaintance with A and also S’s being intellegens. Adeodatus finds this agreeable;16 however, neither Augustine nor Adeodatus pause to consider precisely how this invocation of being intellegens resolves the original difficulty in any detail. Here it suffices to note that Augustine is happy to conclude that many things are learned without signs but simply shown through themselves with no sign being given (nullo signo dato per se ipsa monstrentur, Mag. 10.32). At this point in the dialogue, Augustine shifts into a long monologue which occupies the rest of the work (Mag. 10.33–14.46). Having previously examined whether learning can occur without signs (and having established that it can), he now turns his attention to examining whether signs can impart learning and goes on to claim that they cannot. Nothing, we are told, is learned or taught simply by means of its sign (nihil [ . . . ] per sua signa discatur, Mag. 10.33). To press the point Augustine notes that when one first encounters a (puzzling) word like ‘sarabara,’17 “the word does not show me what thing it signifies” (Non enim mihi rem, quam significat, ostendit verbum, Mag. 10.33). As put, the claim is ambiguous. It could mean either: (i) that hearing or reading the linguistic expression ‘A’ does not impart understanding of A (i.e. hearing or reading ‘A’ does not grant conceptual understanding); or

15 English citations of De Magistro follow Augustine, Against the Academicians and The Teacher, tr. P. King (Indianapolis: Hackett, 1995), sometimes with modifications. 16 “If he is sufficiently intelligent, he’ll know the whole of what it is to walk once walking has been illustrated by a few steps” (si enim sit bene intellegens, paucis passibus ambulatione monstrata totum quid sit ambulare cognoscet, Mag. 10.32). 17 For ‘sarabara’ (‘sarabala’ in the Vulgate), see G. N. Knauer, “Sarabara. (Dan. 3,94 [27] bei Aug. mag. 10,33–11,37),” Glotta 33 (1954), 100–18.

10

Tamer Nawar

(ii) that hearing or reading the linguistic expression ‘A’ does not impart understanding of the signifying relation that holds between ‘A’ and A (i.e. hearing or reading ‘A’ does not grant linguistic understanding). This ambiguity (cf. note 18) means that in the relevant sorts of locutions it is often not entirely clear whether Augustine is addressing conceptual understanding (knowing what A is) or linguistic understanding (knowing that ‘A’ signifies A). However, what Augustine goes on to say suggests that at this particular point he has in mind (i): If certain head-coverings are denominated by this name [‘sarabarae’], have I learned upon hearing it what the head is or what coverings are? I knew these things before; my conception of them wasn’t fashioned because they were named by others, but because I saw them (Ante ista noveram neque cum appellarentur ab aliis, sed cum a me viderentur, eorum est mihi facta notitia). (Mag. 10.33)

Augustine’s point is that we do not (and seemingly cannot) learn what (e.g.) head-coverings are simply from hearing the word ‘head-covering.’ Despite considering whether learning takes place accompanied by a sign, Augustine does not seem to have linguistic understanding in mind but rather conceptual understanding.18 We learn what A is (i.e. attain conceptual understanding) not through hearing ‘A,’ but—Augustine thinks—through perceptual acquaintance with A (and, taking into account what was discussed above, the learner being intellegens). In what immediately follows, Augustine again considers whether a sign shows its significate (and again his discussion is ambiguous). However, this time Augustine does seem to have in mind (ii), that is to say, linguistic understanding: Suppose that we hear ‘head’ now for the first time. Not knowing whether that utterance (vox) is a mere noise or also signifies something, we ask what ‘head’ is. (Remember we want to have a conception [notitia] not of the thing signified but of the sign itself [non rei quae significatur, sed ipsius signi], which we surely don’t have as long as we don’t know what it’s a sign of.) (Mag. 10.34)

In claiming to desire cognizance of the sign itself (rather than the thing it signifies), Augustine is interested in the sign qua sign and so is concerned

18 One should distinguish between: (1) the difficulty of learning what A is from perceiving A; (2) the difficulty of learning what A is from hearing ‘A’ (even when accompanied by ostension); and (3) the difficulty of learning what ‘A’ signifies from hearing an utterance of ‘A’ (even when accompanied by ostension). To think that, because a sign is involved, Augustine must be considering linguistic understanding is probably due to thinking that the only options are (1) and (3), and consequently taking the mention of signs to indicate (3) (when in fact it indicates (2) or (3)).

Augustine on the Varieties of Understanding

11

with linguistic understanding.19 Accordingly, his concern is with how we become aware of the meaning of ‘head.’ What Augustine wants to emphasize is that conceptual understanding of A is prior to linguistic understanding of ‘A.’20 One must know what a head is (i.e. one must have the concept head) in order to learn what ‘head’ means. “We learn the meaning of a word—that is, the signification hidden in the sound—once the thing signified is itself known” (vim verbi, id est significationem, quae latet in sono, re ipsa, quae significatur, cognita discimus, Mag. 10.34). Further, one’s (linguistic) understanding of a term or phrase relies upon appropriately applying the relevant concepts (that is what is missing when a parrot utters a phrase). If attaining linguistic understanding of ‘A’ requires having conceptual understanding of A, and (at least for the sorts of cases envisioned here) having conceptual understanding of A requires having had perceptual acquaintance with A, then attaining linguistic understanding of ‘A’ requires having had perceptual acquaintance with A. I wouldn’t learn the thing I was ignorant of by the words that he has spoken, but by looking at it (discam rem quam nesciebam, non per verba quae dicta sunt, sed per eius aspectum). This is the way it came to pass that I know and grasp what meaning the name has. When I learned the thing itself, I trusted my eyes, not the words of another—though perhaps I trusted the words to direct my attention, that is, to find out what I would see by looking. (Mag. 10.35)

Thus, while words may direct us to look somewhere and see a thing, it is from the seeing (rather than the words) that we learn what the thing is. In sum, at Mag. 10.29–34 (and seemingly also 8.22–10.29), Augustine focuses on conceptual understanding and linguistic understanding. Of these, conceptual understanding has a more important role than is typically allowed and it seems to be prior to and necessary for linguistic understanding. Neither of these forms of understanding seems to require explanatory awareness and Augustine emphasizes that while perceptual acquaintance is an adequate stimulus for learning what A is (i.e. attaining conceptual understanding), hearing the words of another is not.

19 It is claimed that we cannot attain linguistic understanding of a sign “as long as we don’t know what it’s a sign of ” (Mag. 10.34). This could mean either: (i*) we cannot attain linguistic understanding of ‘A’ (i.e. become aware that ‘A’ signifies A) as long as we don’t know that ‘A’ signifies A (an uninformative tautology); or else (ii*) we cannot attain linguistic of understanding of ‘A’ (i.e. become aware that ‘A’ signifies A) as long as we don’t know A (i.e. know what A is, have conceptual understanding of A). I take Augustine’s point to be (ii*); this is supported by Mag. 10.34 (see below). 20 The sense of ‘priority’ here may be temporal but might also indicate dependence or grounding.

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

12

Tamer Nawar

b. Going Beyond Conceptual and Linguistic Understanding Until this point in De Magistro (11.36), Augustine has focused on attaining basic forms of cognizance. However, a significant change occurs at Mag. 11.37. Instead of considering how we learn what a thing is or what a term signifies, Augustine turns to a less basic cognitive state: our accepting (accipere) the story of King Nebuchadnezzar (Daniel 3). As the story goes, three boys were thrown into a fire by the king and yet, miraculously, they emerged unscathed due to the strength of their faith. An imaginary objector to the thesis that words do not impart understanding asks: do we not learn this story by means of words? Augustine replies: [1] I will respond that all the things which those words signified were already in our notitia. I’m already familiar with what three boys are, what a furnace is, what a fire is, what a king is, and what being unharmed by fire is, and all the other things that those words signify. [2] Yet Ananias, Azarias, and Misael are just as unknown to me (ignoti mihi) as the sarabarae, and these names didn’t help me at all to become acquainted with them, nor could they help me (nec ad eos cognoscendos haec me nomina quicquam adiuuerunt aut adiuuare iam potuerunt). [3] I confess that I do believe rather than know (credere me potius quam scire confiteor) that everything read in the story happened then just as it was recorded. [4] Those on whom we rely were themselves not unaware of the difference. In fact, the Prophet says: “Unless you believe, you shall not understand” (nisi credideritis, non intellegetis). Surely he wouldn’t have said this if he had thought there was no difference. (Mag. 11.37)

Augustine begins (in [1]) by talking of the same sort of things as before: concept possession and linguistic understanding. He says that he already had the relevant conceptual understanding—he had the concepts boy, fire, etc. (these were in notitia). He also had the relevant linguistic understanding—he was aware of what ‘boy,’ ‘fire,’ etc. signified. Having linguistic understanding of the individual words of which the sentences were composed, Augustine was thereby also able to understand the meaning of the sentences. However, although he had the conception boy and was aware of what ‘boy’ signified, Augustine stresses (in [2]) that he was not acquainted with the particular boys mentioned in the story (ignoti mihi). Further, the names of the boys (‘Ananias,’ etc.) did not help him to come to know (cognoscere) the relevant individuals (i.e. being able to recognize Ananias, knowing what he is like, etc.).21 However, in [3] and [4], Augustine turns to discussing something rather different. In [3], Augustine is no longer concerned with our having a concept boy, or a linguistic understanding of ‘boy.’ Instead, in [3] Augustine talks of his believing that the story 21 Cf. Trin. 8.4.7–6.9; 13.1.2. For discussion, see section 4.

Augustine on the Varieties of Understanding

13

occurred as described (e.g. believing that the king Nebuchadnezzar threw the boys into the fire) and his prospects of knowing (scire) as opposed to merely believing (credere) these details. In [4], Augustine tells us that the authors of scripture were aware of the difference between merely believing (credere) and understanding (intellegere), and that the former is necessary for the latter (but not vice versa).22 On Burnyeat’s proposal, Augustine uses ‘scire’ and ‘intellegere’ interchangeably in [3] and [4] to discuss the same cognitive state, and this cognitive state, which Augustine disclaims in [3], is explanatory understanding. On this view, while hearing the biblical story can impart belief (and even justified belief) that the story occurred as described, explanatory understanding—a cognitive state wherein one must have a synoptic grasp of the relevant domain and be appropriately aware of the whys and the wherefores—cannot be transmitted through testimony (see section 2). Thus, one cannot attain explanatory understanding (presumably of the events described by the story) from hearing the story and that is why Augustine thinks there is no learning from words.23 So goes the standard interpretation. However, it seems to me that it faces two very significant objections. First, such an interpretation tells us little about why conceptual understanding and linguistic understanding (the cognitive states which were Augustine’s focus until this point in the dialogue and which do not involve awareness of explanatory relations) cannot be learned from words. Second, supposing explanatory understanding to be at issue does not fit the example adduced by Augustine in this part of De Magistro at all well. Prima facie, understanding of a biblical passage is not like understanding a mathematical proof (the paradigm case envisioned by Burnyeat) but even if it were and one allows for the differences between understanding a proof and a biblical story (or the historical event the story is about), it does not seem especially plausible that Augustine is pursuing explanatory understanding here. He is not, after all, concerned with explaining why the king threw the boys into the fire; nor with the motivations of the boys or with how or why their faith saved them from the fire, let alone why (e.g.) the king threw them into the fire as opposed to not throwing them into the fire, or why he

22 In employing the term ‘intellegere’ it is not immediately clear whether Augustine is drawing an equivalence between intellegentia and scientia (as Burnyeat proposes), using the relevant terms interchangeably, or in fact discussing different cognitive states (and in any case, we should recall Augustine was using the term ‘notitia’ in [1] and [2]). 23 Cf. Burnyeat, “Wittgenstein and Augustine,” 21; King, “Augustine on the Impossibility of Teaching,” 180, 193.

14

Tamer Nawar

threw them into the fire as opposed to a chasm, etc., or several other matters (such as why God might have ordained things to occur in this manner). If Augustine is not seeking to understand the whys and the wherefores of the story of King Nebuchadnezzar, what then does he disclaim in [3] when he says he believes (credere) but does not yet know (scire)? The answer, I think, must be either spiritual understanding or infallible knowledge (cognitive states which Augustine shows considerable interest in elsewhere). I’ll offer some brief remarks about each and go on to say why I take Augustine to have in mind the latter (i.e. infallible knowledge). For Augustine, spiritual understanding (e.g. intellectus spiritalis, Ep. 148.4.13, 196.3.9–12; En. Ps. 33.1.7) is a cognitive state attained primarily through the (non-literal) interpretation (tractatio) of the Christian scriptures and is constituted by an awareness of the figurative or spiritual meaning (e.g. sensus spiritalis, Doc. Chr. 4.7.15; cf. secreta significatio, Util. Cred. 7.17) of metaphorical signs (signa translata) and figurative expressions (locutiones figuratae, e.g. Doc. Chr. 2.6.23, 3.24.34).24 It is characterized primarily by its ethical aspect and when Augustine discusses it, he emphasizes that the sort of understanding he has in mind is that which enables us to love God and neighbor as we should (e.g. Doc. Chr. 1.36.40, 2.7.10, 3.10.14, 12.18; cf. En. Ps. 31.1.8, 118.11.4, 27.3, 28.7). Accordingly, it has a prominent conative element and certain ethical effects:25 it directly informs our wills and, in so doing, is especially efficacious in guiding action (cf. S. Dom. Mon. 2.11.38; En. Ps. 31.1.8, 118.27.3) and plays an important role in making us more virtuous (e.g. Doc. Chr. 2.5.6–7.9).26 Although spiritual understanding is often neglected in philosophical scholarship, given that Augustine is here (at Mag. 11.37) discussing 24 Consider the sentence: ‘Christ fasted for forty days.’ This has a literal meaning: that Christ fasted for forty days. We attain literal linguistic understanding of this linguistic expression when we are aware of this. However, the expression, as it appears in scripture (e.g. Matthew 4:2), also has a spiritual meaning: that we should abstain from bodily pleasures (Doc. Chr. 2.16.25). We attain spiritual understanding when we become aware of this (cf. Ep. 148.4.13; 157.2.5; 164.2.3). Augustine thinks there is a spiritual meaning (or various spiritual meanings) to be discerned in all of scripture (e.g. Doc. Chr. 3.10.14) and one should note that Augustine is a generous pluralist about the non-literal meaning(s) of scripture (e.g. Conf. 12.31.42–32.43; Doc. Chr. 3.27.38). 25 There are also ethical prerequisites. Augustine stresses that to attain spiritual understanding of scripture our souls must be cleansed (e.g. Doc. Chr. 1.10.10ff., Div. Qu. 48; Util. Cred. 16.34; Civ. Dei 11.2; Trin. 1.13.31; cf. Plot- 1.6.8–9, 3.6.5), and so he dedicates the first book of De Doctrina Christiana to discussing ethical notions (e.g. usus and fruitio, Doc. Chr. 1.4.4ff.). Of course, one also needs careful acquaintance with the biblical texts (e.g. Doc. Chr. 2.8.12, 9.14) so as to use the clearer passages to illuminate the less clear (Doc. Chr. 3.16.24; cf. 3.11.17). 26 Cf. En. Ps. 118.11.4, 118.28.7; Div. Qu. 31.1; Ep. Jo. 8.6; S. Dom. Mon. 2.11.38.

Augustine on the Varieties of Understanding

15

understanding a scriptural story from the Old Testament, that he elsewhere emphasizes that all of the Old Testament must be accepted not only literally but also figuratively (Doc. Chr. 3.22.32), and that qua biblical exegete spiritual understanding often occupies Augustine’s attention, we have some reason to suppose that it may be spiritual understanding that is at issue at Mag. 11.37. If that is right, then what Augustine lacked upon hearing the story and what he is disclaiming in [3] was an awareness— roughly speaking—of figurative meaning (which would have special ethical significance). However, while taking Augustine to have spiritual understanding in mind at Mag. 11.37 cannot be ruled out (and it does, I think, fit the text better than the standard suggestion), it also does not seem to be the best interpretative option. In De Magistro, Augustine does not mention spiritual understanding or dwell upon the exegesis of scripture; nor are there any gestures towards the ethical import of the biblical story or any significant discussion of those features (just now identified) taken to be characteristic of spiritual understanding. In fact, instead of spiritual understanding, it seems more likely that what Augustine has in mind at Mag. 11.37 is infallible knowledge. Let me say something about this cognitive state and then explain why it fits the text better. That we should think of knowledge as factive reveals why the claim ‘Smith knows that p, but not-p’ seems absurd (e.g. C. Acad. 3.3.5, 3.4.10; cf. Div. Qu. 32). That we should think of knowledge as infallible is meant to reveal why it also seems rather odd to say: ‘Smith knows that p, but maybe not-p.’ If S (infallibly) knows that p, then it is necessary, given the method through which (or the evidence upon which) S came to believe that p, that p.27 It is infallible knowledge, rather than (e.g.) explanatory understanding, that Augustine seeks to defend in his (roughly contemporary) Contra Academicos when confronted with the sceptical thesis that humans lack knowledge (scientia, e.g. C. Acad. 2.5.11, 3.4.10), and he is often interested in it elsewhere. Infallible knowledge is characterized by infallible justification (more on this below); in contrast to explanatory understanding, it requires no awareness of explanations and need not be synoptic (in

27 Factivity ensures that the person who knows is not mistaken. Infallibility ensures that the person who knows cannot be (or could not have been) mistaken insofar as the method or evidence upon which the belief that p is based guarantees that p. I labor the point because the scholarship often either conflates infallibility and factivity or else uses the term ‘infallibility’ for factivity (e.g. “La de´finition de la science par l’infallibilite´ fait l’objet d’un consensus chez les Ancients depuis Plato. S’il y a erreur, par de´finition la science fait de´faut” [Emmanuel Bermon, Le cogito dans la pense´e de saint Augustin (Paris: Vrin, 2001), 110]).

16

Tamer Nawar

contrast to infallible knowledge, explanatory understanding does not require infallibility).28 There is good reason to think that it is infallible knowledge that Augustine is disclaiming in [3]. Looking ahead to slightly later in the dialogue, Augustine emphasizes his point that there is no learning from words by asking: “What if I should say that I had seen a flying man? Do my words then make you as certain as if you were to hear that wise men are better than fools? [ . . . ] you know the latter statement with utter certainty (certissime scire)” (Mag. 12.40 [quoted below in full]). His point there is not that he lacked explanatory awareness upon hearing the tale of a flying man (nor that he lacked spiritual understanding) but rather that he cannot know for sure (certissime scire) that it happened. He makes the point often enough with regard to unlikely tales and events of the distant past (e.g. Util. Cred. 11.25; Div. Qu. 48; Ep. 120.2.9–11) and his point at Mag. 11.37 with regard to the biblical story (which seems to fall under both categories) must, I think, be the same. Just as we cannot—on the basis of testimony—know for sure that a flying man has been seen, so too we cannot know for sure that the youths were miraculously saved from the fire (as the biblical story claims). The point is not that we cannot explain the relevant events but that we lack certain proof that they occurred as described. Testimony that p does not eo ipso guarantee that p and thus does not constitute a means of attaining infallible knowledge that p.

4 . W HY T H E R E I S N O L E A R N IN G F R O M W O R D S As we have seen, it turns out that Augustine is not especially concerned with explanatory understanding in De Magistro. Instead, Augustine addresses several kinds of cognizance—namely, conceptual understanding, linguistic understanding, and (at Mag. 11.37, 12.40) infallible knowledge. He seems to claim, of each of these, that they cannot be learned through words. Insofar as these cognitive states are not constituted by explanatory understanding and do not require awareness of explanatory relations, Augustine’s claim that there is no learning from words remains in need of explication.

28 Notice that in his Contra Academicos, Augustine’s discussion of infallible knowledge riffs on the Stoic account of apprehension (ŒÆºÅłØ) (C. Acad. 2.5.11, 6.14; 3.9.18, 21ff.) and that the Stoics took (infallible) apprehension to be attainable but by and large (Posidonius being the notable exception) explicitly distanced themselves from aetiologies and causal explanations (Strabo, Geography 2.3.8).

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

Augustine on the Varieties of Understanding

17

a. Attaining Conceptual Understanding and Linguistic Understanding Immediately after disclaiming infallible knowledge at Mag. 11.37 (see above), Augustine begins discussing why in both perceiving (sentire) and understanding (intellegere) there is no learning from words. Regarding perception, Augustine’s point is simple and straightforward. When we use our bodily senses to perceive, we “consult” (consulere) the bodily senses and their sensible objects (Mag. 12.39). In Augustine’s terms, for S to perceive a thing, that thing must be directly present to S ([de sensibilibus] respondemus, si praesto sunt ea quae sentimus, Mag. 12.39; cf. Ep. 147.1.6–2.7). We would agree with this. Perception seems to be such that appropriate (i.e. direct, nondeviant) causation is essential to it. When S perceives that p, then the fact that p directly and non-deviantly causes S to think that p.29 Testimony precludes such directness. After making these remarks about perception, Augustine (peculiarly) claims that we cannot directly talk about sensible things which we have previously perceived but are not currently perceiving; instead, we can only talk about the images (imagines) those things leave behind in memoria (Mag. 12.39).30 He goes on to claim: [1] We carry these images (imagines) in the recesses of our memory in this way as certain attestation (documenta) of things sensed previously [ . . . ] [2] If anyone hearing me was then present and sensed these things (sensit ea), he doesn’t learn (discere) from my words but knows them again (recognoscere) from the images stored away within himself. [3] If he hasn’t sensed them, isn’t it obvious that he merely believes (credere) my words rather than learns from them? (Mag. 12.39)

While the precise nature of the imagines (mentioned in [1]) is not entirely clear (they may be picture-like representations but needn’t be),31 they seem 29 There is a complication. In some early works (though not explicitly in De Magistro) Augustine claims that the soul is not directly affected during such perceptual processes (Mus. 6.5.8; cf. Quant. An. 24.45–25.48; cf. Charles Brittain, “Non-Rational Perception in the Stoics and Augustine,” Oxford Studies in Ancient Philosophy 22 (2002), 275–7). Insofar as he offers reasons for this claim, Augustine seems to have conflated the soul being under the yoke of the body (a claim he wishes to deny) with the soul being such as to be causally affected by the body (Mus. 6.5.8). Fortunately, he does recognize the obscurity of his thought on this matter (Mus. 6.5.8) and in later works typically emphasizes not so much that the soul is unaffected but (far more plausibly) that it is importantly active in such processes (especially in arranging and uniting information, e.g. conligere, cogere, cf. cogitare, Conf. 10.11.18; Trin. 9.3.3, 11.3.6; Gn. Litt. 12.16.33). 30 For discussion of this (odd) notion, see Gareth Matthews, “Augustine on Speaking from Memory,” American Philosophical Quarterly 2 (1965), 157–60. 31 Augustine says they are picture-like (id est imaginem quasi picturam eius, Trin. 8.6.9) but insofar as we have notiones of emotions (e.g. Conf. 10.14.22) and incorporeal

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

18

Tamer Nawar

to be constituents of thought which are either concepts (notiones), constituents of concepts (e.g. a concept might be constituted by some imago or several imagines), or else play a similar role to concepts. If one is to think of A, then one needs a notio or imago of A.32 The central argument (in [2] and [3]) then seems to be a destructive dilemma.33 Suppose that S (the listener) hears S* (the speaker) talk about A (a sensible item). (i) If S had previous perceptual acquaintance with A, then S does not learn but may merely “recognize” or “recollect” (recognoscere). (ii) If S did not have previous perceptual acquaintance with A, then S does not “learn” but may merely “believe” (credere). But either (i) or (ii). Therefore, when S hears S* talk about A, S does not “learn.”34 Now, the precise sort of cognition which Augustine is claiming words cannot impart is not entirely clear. [2] seems to commit Augustine to the claim that if I am acquainted with (e.g.) cats, then I cannot learn (i.e. attain the relevant form of cognizance) from your words but at most may merely remember (by looking again at my imagines). An unsympathetic reader might see this as absurd. For instance, I understand what a cat is (i.e. have the concept cat) and understand what a parrot is and yet I may certainly learn that Tibbles was a naughty cat from hearing my neighbor utter: “your cat ate my parrot.”35 While the text is not entirely clear, for Augustine’s point to be plausible we should suppose that what is at issue is not so much the listener knowing or accepting some particular proposition but rather

things (such as the soul, justice, etc.) and imagines of sounds and tastes (e.g. Conf. 10.10.17), I am inclined not to take this too literally. He also, famously, speaks of them as “inner words” (e.g. Trin. 8.6.9, 9.10.15). I won’t say much more here about their possible content or otherwise representative nature (for the Stoic view, which might have influenced Augustine, see Cicero, Acad. 2.21). 32 In the Confessiones, Augustine tells us that one needs a concept (notio) of an emotion or sensation if one is to think about or discuss it (Conf. 10.14.22; cf. Lib. Arb. 2.9.26) and the same holds, he thinks, for sensible things and their imagines. If A is something sensible (a stone or the sun are the examples Augustine invokes), then if one is to think about A, one must have an image (imago, cf. descriptum, Trin. 12.14.23) of A in one’s memoria (Conf. 10.15.23). 33 Augustine’s discussion concerns sensible things not currently present. We have already seen that, of present sensible things, Augustine claimed words could advise us where to look but that we learn not from the words but from seeing the thing (e.g. per eius aspectum, Mag. 10.35). 34 Cf. “When words are put forward, either we know (scire) what they signify, or we do not know (nescire); if we know, then it is reminding (commemorari) rather than learning [that words can bring about]; if, however, we do not know, then it is not even reminding, though perhaps there might be some admonishing (admoneri) towards inquiring” (Mag. 11.36). 35 Cf. Norman Kretzmann, “Semantics, History of,” in P. Edwards (ed.), The Encyclopedia of Philosophy (New York: Collier Macmillan, 1967), 366.

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

Augustine on the Varieties of Understanding

19

the listener coming to be acquainted with the object of that speaker’s words (which is indeed suggested by the text).36 Thus understood, Augustine’s point would be that words cannot lead us to satisfactorily attain new concepts (more on this in a moment).37 Supposing Augustine is still discussing the same form of cognition in [3],38 the claim would be that whatever cognizance of the relevant sort one does attain from words is deficient (as mere believing is deficient when compared to knowing). So, why does Augustine think that there is no genuine “learning” from words in such cases? Well, Augustine thinks that some instances of conceptual understanding (those of certain queer properties such as mathematical or moral properties) are not learned from words (or, for that matter, perception) because, quite simply, they do not have to find their way into memoria at all but are instead already there (Conf. 10.10.17, 12.19).39

36 Augustine speaks of representations of things (imagines rerum sensarum, documenta rerum) and it is assumed that having an imago of a sensible thing requires having previously perceived that thing (see above). 37 One might think that the speaker does not assert a proposition but simply utters ‘A,’ where ‘A’ is a variable for a general term (e.g. ‘cat’). Augustine’s point would then be that uttering ‘cat’ is not an adequate stimulus for a listener to learn what cats are (this seems to be how the point is taken by Burnyeat, “Wittgenstein and Augustine,” 16 n. 18). However, while thus construed, Augustine’s point is plausible, it renders his thesis both too obvious and too weak to sustain his claim about the inefficacy of words. 38 The talk of believing in [3] makes it sound like a proposition is being put forward and that the learning in question concerns the listener (S) coming to accept that proposition. It is possible that ‘believing’ refers to a deficient form of conceptual understanding or that there is some fluidity between [2] and [3] in the precise form of cognition or understanding that Augustine has in mind (if the latter, then Augustine is anticipating what I discuss in greater detail below). 39 Cf. Quant. An. 4.5; Lib. Arb. 2.8.20ff.; Gn. Litt. 12.11.22, 24.50; Ep. 147.17.43–4; Trin. 8.6.9. Of sensible objects and emotions, we have imagines and notiones imprinted in our memories. However, in Conf. 10.8.15–9.16, Augustine stresses that we do not have imagines or notiones of the objects of the disciplinae liberales; instead, the objects themselves (res ipsas) are themselves present in memoria (cf. Ep. 7.2.4). This has received little discussion but is not easy to make sense of (cf. Luca Castagnoli, “Liberal Arts and Recollection in Augustine’s Confessions X (ix 16–xii 19),” Philosophie Antique 6 (2006), 107–35). Prima facie it stipulates that numbers, lines, etc. are themselves in memoria (e.g. Conf. 10.12.19) but it is not clear what that means; later in the book Augustine more plausibly suggests not that it is the objects of these disciplinae that are present but rather the artes themselves (e.g. mathematical skill) (Conf. 10.17.26). Further, in De Libero Arbitrio he seems happy to admit that we do have notiones impressae of numbers, justice, etc. in the soul (i.e. notiones of numbers are present in the soul instead of numbers, justice, etc. themselves, Lib. Arb. 1.6.15, 2.9.26, 15.40). De Trinitate captures his puzzlement on the subject (e.g. sed praesens quiddam cerno, et cerno apud me, etsi non sum ipse quod cerno, Trin. 8.6.9; cf. Trin. 12.14.23). However, whether we have (e.g.) justice itself present in our soul (despite it not making us just) or a notio of justice, it is its conceptual role (e.g. enabling us to think about justice) that Augustine focuses on. Accordingly, I shall use the term ‘concept’ in these cases.

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

20

Tamer Nawar

Augustine elsewhere says that, in these cases, words merely remind (e.g. Conf. 10.10.17, he may be adverting to this in [3]). However, the problem which is of concern here is why more mundane concepts (e.g. walking, boy, fire), i.e. those of sensible things—which are Augustine’s principal focus in the earlier part of De Magistro (see section 3a) and are his avowed focus in Mag. 12.39—are such that hearing words from another is not an adequate stimulus to their acquisition and why he should say that we must instead have actually perceived the things themselves (e.g. Mag. 10.33–4).40 Talk of such a perceptual acquaintance requirement is not unique to De Magistro; the claim is made explicitly in the Confessiones (e.g. 10.16.25) and De Trinitate, where Augustine tells us that it is impossible for us to at all conceive (omnino cogitare) of a color or bodily figure (figura corporis) that we have never seen (Trin. 11.8.14). One might think that this stipulates that we can form imagines of only those things we have actually seen. However, this cannot be what Augustine means. After all, memoria is capable of imagining, and in that very discussion in De Trinitate Augustine discusses how the mind has the power to fashion images of things which have not been previously experienced by enlarging, diminishing, and altering those concepts which it has (Trin. 11.5.8; cf. Ep. 7.3.6). Instead, the perceptual acquaintance requirement should be read generously. Augustine is not claiming that we need to be acquainted with (e.g.) flamingos in order to form a concept of or imagine flamingos but that we require acquaintance with birds, or at least, with bird-like animals (cf. Mus. 6.11.32). In short, there is a limit to imagination. The thought is plausible and is not infrequently made in contemporary discussions of concept acquisition (e.g. by Simon Blackburn).41 Augustine’s remarks elsewhere 40 Notice that I use the expression ‘adequate stimulus’ (rather than, e.g., ‘cause’) because some (e.g. King, “Augustine on the Impossibility of Teaching,” 194) seem to assume a sufficiency account of causation (i.e. A causes B iff the presence of A is sufficient for the presence of B) and take Augustine’s claim that there is no learning from words to mean that words do not cause (i.e. are not sufficient for producing) understanding. However, perceiving is also not a cause of (i.e. not sufficient for) understanding (recall the emphasis on the learner’s being intellegens, Mag. 10.32 [discussed above]). What must be explained is why perceiving is an adequate stimulus but words are not. Burnyeat generally seems to construe the talk of perceiving as applying, metaphorically, to explanatory understanding (cf. note 3). However, after dwelling on explanation not being transmitted through testimony, he very briefly suggests that Augustine believed in “epistemic categories” (“Wittgenstein and Augustine,” 17–18), i.e. that certain things are such that if one knows them, then one has perceived them. This gives a label to Augustine’s view but it does not explain why we need to perceive to have the relevant form of cognizance. In what follows, I attempt to improve upon this, explaining why Augustine thought this. 41 “With joy or sorrow, we make a first acquaintance with dogs and carrots, and grow into an appreciation of what a Romantic sonata, or a smile, or an electron, is. Growing

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

Augustine on the Varieties of Understanding

21

(notably in De Trinitate) suggest this train of thought, and the discussion of ‘sarabara’ earlier in De Magistro (10.33) fits this interpretation well.42 Put simply, in order to imagine something complex such as a head-covering, one needs to have concepts of (e.g.) heads and coverings (similarly, imagining a centaur might require the concepts horse and man). This suffices to explain the perceptual acquaintance requirement claim and to render it plausible. However, thus understood, the perceptual acquaintance requirement merely poses limits on imagination; it does not explain why we cannot learn from words. After all, while one might need to have seen birds or animals, one can imagine a flamingo, despite never having seen one, on the basis of another’s words and so come to possess the relevant concept. Why shouldn’t this be considered an instance of learning? Augustine’s answer is not explicitly spelled out in De Magistro but is hinted at elsewhere and seems to turn upon the deficiency of such instances of imagining; they are too vulnerable to epistemic vicissitudes and don’t seem to have the right connection with the world to be regarded as a cognitive success. Augustine draws a distinction between two sorts of imagines. Phantasiae are produced by direct perceptual experience; in contrast, phantasmata are produced by imagination (prompted, for instance, by testimony) acting on experience (Mus. 6.11.32; cf. Trin. 8.6.9).43 Phantasiae are such that the way they are formed gives a good guarantee that they represent things correctly. In contrast, phantasmata lack such a guarantee; they may well not represent their objects as they are (Augustine tells us that this happens only very rarely, Trin. 8.6.7) and whether they do so at all seems to be a matter of luck. into such an appreciation needs an acquaintance with things. [ . . . ] With the acquaintance comes not only an appreciation of what words actually mean, but of the kind of thing they could mean. Acquiring the concept needs learning that there are things which persist in time, which have various clusters of properties, and it requires becoming able to distinguish those things from others on the basis of these properties” (Simon Blackburn, Spreading the Word: Groundings in the Philosophy of Language (Oxford: Oxford University Press, 1984), 56). 42 “If certain head-coverings are denominated by this name [‘sarabarae’], have I learned upon hearing it what the head is or what coverings are? I knew these things before; my conception (notitia) of them wasn’t fashioned because they were named by others” (Mag. 10.33). 43 Aliter enim cogito patrem meum quem saepe vidi, aliter avum quem numquam vidi. Horum primum phantasia est, alterum phantasma. Illud in memoria invenio, hoc in eo motu animi, qui ex iis ortus est quos habet memoria (Mus. 6.11.32). In De Trinitate, Augustine tells us that in order to think of a city one has perceived, such as Carthage, one uses one’s remembered perceptual appearance (phantasia) of Carthage (which one has received through the senses of the body). In contrast, in order to think of a city one has not perceived, such as Alexandria, one uses one’s imagined quasi-perceptual appearance (phantasma) (Trin. 8.6.9).

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

22

Tamer Nawar

That testimony cannot serve as an adequate grounding for concepts (while perception can) would seem to be due to the fact that a picture, as they say, is worth a thousand words, i.e. perceiving is a much richer experience and (contra Burnyeat, e.g. “Wittgenstein and Augustine,” 7–8 [quoted above]) the linguistic stimulus is deficient in transmitting information.44 For instance, I may try to tell you what churidars are (I employ ‘churidar’ to approximate some of the effects of ‘sarabara’) by saying something like: “churidars are long trousers, originally from South Asia, which are baggy at the hips but tight at the ankles.” Now, you may know what trousers are (know what hips are, etc.) and upon hearing these words you may well imagine something, but given the limited information in the linguistic stimulus provided, it is—I think—unlikely you would imagine churidars correctly (so as to be able to recognize them upon seeing them) unless you had in fact already seen them (in which case my words would merely remind you of what you had seen).45 Further, even if, by luck, one does imagine churidars correctly (without having seen them) from this description (i.e. represent churidars as they really are) and so have cognizance of a sort, it is epistemically deficient much like a lucky guess is. Its accuracy arises not through its grounding but through luck and it would, Augustine thinks, be very rash to suppose that one has imagined correctly (Mus. 6.11.32). Thus understood, it is the inadequate grounding and frequent inaccuracy of concepts formed on testimony that accounts for why Augustine thinks there is no “learning” with regard to this basic form of cognizance and, as the churidar example shows, his point seems to have some merit. What then of linguistic understanding? Augustine’s point here is much briefer. Quite simply, he focuses on our learning of individual words (cf. Burnyeat, “Wittgenstein and Augustine,” 16) and, as we have seen, he supposes that conceptual understanding is prior to linguistic understanding. In order to learn what ‘cat’ signifies, one must already have the relevant concept (cat). Accordingly, it seems that learning what ‘cat’

44 This may seem to attribute to Augustine thoughts like those which motivate Jerry Fodor (and which are attributed by him to Augustine [Jerry Fodor, The Language of Thought (Hassocks: Harvester Press, 1975), 63–4]). However, moderns typically argue that the information present in linguistic stimuli is much less than what is required to cognize the syntactic and semantic rules of language (e.g. Noam Chomsky, Aspects of the Theory of Syntax (Cambridge, MA: MIT Press, 1965); Fodor, The Language of Thought). While Augustine might be sympathetic to this, this is not his focus. In De Magistro he is focusing on the acquisition of concepts corresponding to mundane general terms. 45 Augustine thinks that the same point (or a similar point) applies to descriptions of individual persons (cf. Ananias, Azarias, and Misael at Mag. 11.37; the Apostle Paul and Christ at Trin. 8.4.7).

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

Augustine on the Varieties of Understanding

23

signifies is akin to merely associating a sonorous label with a concept (e.g. Mag. 10.33; cf. Jo. Ev. Tr. 37.4) and that is why, with regard to linguistic understanding, Augustine says that from words we merely learn the sound or noise of words (Mag. 11.36).46

b. A More Complex Form of Cognizance At De Magistro 12.40, Augustine moves to consider our acquisition of a different kind of cognizance: When we deal with things that we see (conspicere) by the mind (mens), namely by the intellect (intellectus) and reason (ratio), we’re speaking of things that we look upon immediately in the inner light of Truth, in virtue of which the so-called inner man is illuminated and rejoices. Under these conditions our listener, if he likewise sees these things with his inward and undivided eye, knows what I’m saying from his own contemplation (contemplatio), not from my words. Therefore, when I’m stating truths, I don’t even teach the person who is looking upon these truths. He’s taught not by my words but by the things themselves made manifest within when God discloses them. (Mag. 12.40)

Augustine here seems to envision words not as teaching us but, at most, as either reminding or admonishing (admonere) us (cf. Mag. 11.36; Conf. 10.10.17) or otherwise prompting (but not causing, cf. note 40) us to “see” with our minds (cf. Mag. 11.38).47 What follows strongly suggests that it is infallible knowledge which Augustine has in mind here (witness, for 46 Augustine appreciates that words are composed of sound (sonus) and meaning (intellectus) (S. 28.4; Quant. An. 32.65–6; Conf. 10.26.24). However, when he emphasizes that from words we only learn the sound of words, he often also seems to identify words with their sound(s) (e.g. strepitus, Mag. 11.36; personat, Mag. 11.38; sonus verborum, Ep. Jo. 13.3) or else talks of words qua sounds (e.g. et verba ipsa, quod ad sonum adtinet, Imm. An. 15.23). 47 De Magistro contains both talk of reminding (e.g. commemoratio, commemorare, Mag. 1.1–2, 4.7, 7.19–20, 11.36 [quoted above]; cf. admonere, Mag. 9.25, 10.35–11.36, 11.38; recognoscere, Mag. 12.39 [quoted above]) and illumination (e.g. Mag. 11.38–12.40). However, despite early enthusiasm (e.g. Ep. 7.1.2; Quant. An. 20.34; Sol. 2.20.35), in later works Augustine objects to Platonic recollection and employs such talk less frequently because he takes it to imply pre-existence of the soul (it needn’t though; it could merely imply innateness of the relevant concepts). Illumination, on the other hand, is a consistent theme in Augustine. He appeals to it when discussing how we cognize a priori truths (e.g. Div. Qu. 46; Lib. Arb. 2.8.21); how we attain spiritual understanding (e.g. En. Ps. 118.18.4); the objective (i.e. non-relativist) nature of truth (e.g. Conf. 12.25.35, Lib. Arb. 2.12.33, 2.19); that cognition requires the aid of God (e.g. En. Ps. 118.18.3–4; Gn. Litt. 12.31.59; Ep. 120.2.10; cf. 2 Cor. 3: 5); and also for other reasons. The controversies are many (especially regarding the precise nature of God’s aid) and the details (if indeed such talk can be said to amount to a theory) are difficult to work out. I will say little more here.

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

24

Tamer Nawar

instance, the talk of certainty below) and also sheds some further light onto why such a cognitive state cannot be attained from the words of another: [1] I might say: “the things I’m saying that you admit to be truths, and that you’re certain (certus) of, and that you affirm yourself to know (nosse)—where did you learn them?” Maybe you would reply that I had taught them to you. [2] Then I would rejoin: “What if I should say that I had seen a flying man? Do my words then make you as certain as if you were to hear that wise men are better than fools?” Surely you would deny it and reply that you do not believe the former statement, or even if you did believe it that you do not know it; whereas you know the latter statement with utter certainty (certissime scire). [3] As a result, you would then understand (intellegere) that you hadn’t learned anything from my words, neither in the former case (where you did not know although I was asserting it) nor in the latter case (where you knew quite well [optime scire]). (Mag. 12.40)

In [1], Augustine invites us to consider those things we really know (or at least take ourselves to know) and asks how we came by the relevant cognizance. He points out that it seems natural to think that when we heard another assert that p and consequently come ourselves to accept that p, that we learned that p from the other person’s assertion. However, imagine, Augustine requests in [2], two distinct cases. In the first case, S hears S* assert that p1 (e.g. that a flying man has just been seen) but does not give credence to p1 (and thus neither knows nor believes that p1).48 In the second case, S hears S* assert that p2 (e.g. that the wise are superior to fools) and consequently S comes to accept that p2. Finally, in [3], Augustine claims that the preceding considerations show that neither in the first case (where one hears an implausible proposition asserted) nor in the second case (where one hears a plausible proposition asserted) is there learning from words. Now, to see why there is no learning (i.e. attaining of infallible knowledge, as the talk of certainty strongly suggests) from words, we must, I think, appreciate not only that—due to its being fallible—testimony fails to bring about infallible knowledge, but also that when one relies upon testimony one lacks something more. The first fact—concerning the fallibility of testimony—is simple and plain. As Augustine later points out, people lie, or are deceived, or communicate poorly, or misunderstand, or mishear, and so forth (Mag. 13.41–4), and so the reliability of testimony falls far short of being absolute. However, the point is not just that testimony is fallible (for testimony is hardly distinctive in falling short in 48 Augustine recognizes more explicitly than many that there are degrees of belief or of confidence in a belief that fall short of plena fides (e.g. Civ. Dei 16.31, S. 115.1, Cons. Ev. 2.28.66). However, for the sake of simplicity, I will here mostly talk of belief as an all-or-nothing affair.

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

Augustine on the Varieties of Understanding

25

this way). Rather, it is that testimony does not communicate a reflective grasp of the justification of one’s belief. The infallible knowledge Augustine has in mind is not just a cognitive state formed through a method which is absolutely reliable (an account of which could be given in externalist terms), it is also rationally indefeasible and certain (and these are features which require the agent’s awareness of the reasons for the relevant belief).49 As Augustine elsewhere tells us, this form of cognizance is attained through the firm reasoning of the mind (mentis firma ratione, Util. Cred. 11.25). When one has such infallible knowledge, one is also aware of the evidence for this belief and can even respond to the sceptic and say why it is not possible that one is mistaken in the present case. “Knowledge (scientia) doesn’t consist merely in the matters that are apprehended. Instead, it consists in the fact that they are apprehended in such a way that nobody should be in error about it or vacillate when pressed by any opponents” (C. Acad. 1.7.19). Augustine recalls this in De Magistro when, in speaking of the relevant form of cognizance, he says of its possessor: “if he were questioned, he could give answers even about these matters” (Mag. 12.40). Fallibility notwithstanding, testimony fails to bring about this reflective grasp of the reasons for one’s belief. Now, consider once more the cases Augustine offers. In the first case, an outrageous assertion (that a flying man had been seen) is made. Augustine highlights the fact that we do not simply accept outrageous things on the basis of testimony.50 In this (outrageous) case, the listener would presumably reject the testifier’s testimony and, since they have not come to believe that a flying man has been seen, neither would they learn this. Hence, in the case of an implausible assertion which is rejected, there is no learning from the speaker’s words. However, even if one did put one’s negative reasons aside and did come to believe the speaker’s report, one should—Augustine points out—recognize that one did not thereby come to acquire knowledge of the relevant sort. Why not? Presumably because even if one were willing to give the testifier the benefit of the doubt, testimony’s reliability is less than absolute and one would lack independent positive reasons for believing that p (aside from so-and-so’s say-so). In fact, one’s negative reasons (i.e. reasons against p) would render one’s belief that p especially tenuous

49 That is to say, the (psychological) certainty and indefeasibility of infallible knowledge is not mere stubbornness but rational certainty and indefeasibility. 50 As presented here, the absence of negative reasons (i.e. reasons to disbelieve that p or the testifier’s trustworthiness) is a deciding factor in whether a testifier’s assertion that p is accepted (cf. Ep. 147.2.7). In this case, we would probably be inclined to say that testimony is defeasible testimony and either the proposition asserted or the hearer’s other beliefs acted as defeaters.

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

26

Tamer Nawar

and uncertain in a manner unfitting for the cognitive achievement Augustine has in mind. Consider also the second case where a plausible assertion (that the wise are superior to fools) is made. When S accepts S*’s testimony, Augustine seems to suppose that S’s acceptance that p occurs either: (i) because S accepts what S* asserts simply on S *’s testimony; or else (ii) because S independently reflects over whether p and it seems, on reflection, to S to be true.51 If (i), S ’s coming to accept that p because S* asserted it would not (for Augustine) be a case of infallible knowledge. Why not? Because when S accepts another’s assertion that p simply on the basis of testimony or trust, S accepts that p without deliberation. S does not give independent reflection to whether p and does not weigh the assertion against his or her other reasons for whether p. S is not thereby aware of any independent evidence for thinking that p and does not “consult” (cf. Mag. 11.38) his or her own evidence; instead, S “consults” only the speaker (S*) and if S were to be asked why they believe that p, they would (again) simply say S* told them so. Thus, the firm reasoning of the mind and awareness of justificatory grounds which is required for a more elevated form of cognizance is absent. If (ii), S ’s coming to accept that p after hearing S* assert that p might amount to infallible knowledge (if the other conditions are met) but it would not be learning from words. This is because—if it is indeed to be infallible knowledge—although S * has put forward that p, S’s acceptance that p has little or nothing to do with the fact that p originates in S*’s testimony. Instead, S has instead accepted that p due to independently reflecting upon whether p and considering p on its own merits. S accepted that p not on S*’s say-so, but in the light of S’s own deliberation. It is not the fact that p originated in S *’s testimony that convinces S to accept that p, rather it is simply p, either on its own (as in the case of self-evident truths) or how p fits with S’s evidence (in the case of evident propositions). As Augustine tells us, in such cases the learner is “taught not by my words but by the things themselves made manifest within when God discloses them” (Mag. 12.40 [quoted above]). In accepting p in this way, S is aware of independent grounds for accepting that p and may (if the evidence is good enough) attain infallible knowledge. In Augustine’s terms, when S attains infallible knowledge, S does not do so as a result of “consulting” S* (cf. Mag. 11.38, 12.39). While S*’s testimony 51 Augustine does not here envision cases wherein testimony is treated as evidence to be aggregated with our own reasoning. In such cases, one might accept another’s testimony that p because one had inferred from the fact that the testifier asserted that p, and that this testifier (or testifiers of a similar sort, or testimony in general) usually assert truths, that (ceteris paribus) p is probably true.

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

Augustine on the Varieties of Understanding

27

may “admonish” or “remind” (admonere), i.e. give S an opportunity to reflect on whether p (Mag. 11.36, 38),52 if it is to be infallible knowledge then S “consulted” (consulere, Mag. 12.40) the truth of whether p within themselves (e.g. “he knows what I’m saying from his own contemplation, not from my words,” Mag. 12.40 [quoted above]; cf. Conf. 11.3.5).53 Thus, Augustine tells us that students “consider within themselves whether truths have been stated. They do so by looking upon the inner Truth, according to their abilities. That is therefore the point at which they learn” (Mag. 14.45). As it is portrayed here, the students look within not in order to explain that p, but in order to decide whether to accept that p, i.e. for justification. In those cases wherein genuine learning has occurred, S depends upon his or her own reflection and evidence for justification in the belief that p. Presumably, if they were later questioned about this belief, the reasons for acceptance of the relevant proposition would—Augustine thinks—be readily accessible to them and they could publicly justify it.54 Thus, in a manner similar to Burnyeat, it must be emphasized that in order to attain scientia, the agent must—in some sense—see how things fit together for themselves (see note 3). However, on Burnyeat’s view, Augustine is concerned with explanatory understanding, and what testimony fails to impart is explanation: an appreciation of why something is so. In contrast, the interpretation I offer asserts that for Augustine what testimony fails to impart is the right sort of justification, i.e. the right kind of appreciation of why it is right to believe that something is so (something

52 “[ . . . ] though perhaps words prompt us to consult Him” (Mag. 11.38 [quoted above]); “though perhaps there might be some admonishing (admoneri) towards inquiring” (Mag. 11.36). 53 In the Confessiones, Augustine imagines speaking to Moses about the creation of the world: “If he [Moses] spoke Latin, I would know what he meant. But how would I know whether he was speaking the truth? If I did know this, I could not know it from him (quod si et hoc scirem, num ab illo scirem). Within me, within the dwelling place of thought, a truth which is neither Hebrew nor Greek nor Latin would speak [ . . . ] It would say: ‘What he is saying is true’ ” (Conf. 11.3.5; Lib. Arb. 2.12.33; Trin. 3 proem. 2). Augustine can also appeal to inner truth for the attainment of spiritual understanding (e.g. Conf. 12.13.16–15.18) but in De Magistro, as in Conf. 11.3.5, inner truth performs a justifying function. 54 I focus here primarily on admonishing. However, the talk of reminding (cf. n. 52) may be similarly explained. For instance, the teacher might say the square of the hypotenuse is equal to the sum of the squares of the other two sides of the triangle. However, the words do not teach us, they merely prompt us to examine our concepts (line, triangle, etc.), and if we attain the relevant cognizance it is from this examination. In both cases, the student seems to already “have” the relevant information but merely needs to come to have a different awareness of it or see it under a different mode of presentation (for discussion of this aspect, see King, “Augustine on the Impossibility of Teaching”).

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

28

Tamer Nawar

which Burnyeat explicitly denied).55 People who believe that p on the basis of testimony rather than independent reflection will not, Augustine thinks, be able to give good reasons (to themselves or others) for their belief. If faced with questions they will quickly run out of satisfactory answers (cf. C. Acad. 1.7.19; Mag. 12.40) and the relevant cognizance will prove rationally unstable and lack resilience in the face of opposing views. Further, it should be noted that relying on testimony makes one’s belief(s) especially vulnerable to that pervasive feature of human interaction: disagreement. (If one believes that p because another said that p, what will one do when a different person says that not-p?) Thus, when S hears S* assert that p and S subsequently accepts that p, one of two things may have happened. Either S may simply have accepted S*’s testimony without reflection, or else S independently reflected over whether p and accepted that p due to the evidence available to them (in which case infallible knowledge may be attained). In the latter case, S* makes an assertion that p but when S, in deciding whether to accept it, reflects upon it independently and decides to accept it (owing to this reflection), S did not learn from S*’s words but from his or her own reflection. In the former case, S does not independently reflect over whether p and the most one can attain is fides (aut credere aut opinari, Mag. 12.40).56 This sort of cognizance, brought about by testimony, falls short of more elevated cognitive states in several respects but Augustine’s focus here seems to be on the fact that it does not have sufficient reliability and lacks an appropriate basis in reasons (i.e. justification in an internalist’s sense; cf. quod scimus igitur, debemus rationi, quod credimus, auctoritati, Util. Cred. 11.25) and, thus, rational resilience.57 It is thereby vulnerable to counter-evidence (for instance, something as simple as disagreement among testifiers) and may 55 “What is missing [ . . . ] is something other than justification” (Burnyeat, “Wittgenstein and Augustine,” 7). 56 Fides is justified belief (and often seems to be justified externalistically). Unlike fides, opinio (unjustified belief or wishful thinking) is formed rashly and irresponsibly (through temeritas, e.g. Util. Cred. 11.25). It is the result of an epistemic vice (credulitas) and is usually false and always blameworthy (numquam sine vitio, Util. Cred. 11.25). 57 Christian believers possess warrant and believe truly but their evidence or testimony is not utterly certain proof (indubitata ratio). That is why, in De Utilitate Credendi, they have fides but not scientia (e.g. Util. Cred. 14.32). Being reliant on testimony, we may reliably believe that so-and-so are our parents but we cannot (infallibly) know this (scire, Util. Cred. 10.24–12.26; Div. Qu. 48; cf. F. Invis. 2.4). In later works, Augustine often calls the (fallible and justified in an externalist manner) sorts of cognizance attained through testimony ‘cognitio’ or ‘scientia’ (e.g. Retr. 1.14.3 [quoted above]; Gn. Litt. 12.25.52; Trin. 13.1.2, 15.12.21; for discussion, see King and Ballantyne, “Augustine on Testimony”). However, we need not suppose either incoherence (cf. Burnyeat, “Wittgenstein and Augustine,” 19 n. 23) or a radical change in his conception of scientia (or of testimony). Even in later works Augustine thinks there is a sense of ‘scientia’ which

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

Augustine on the Varieties of Understanding

29

even fall apart under sustained questioning. This is not to say that cognizance brought about by testimony is without worth. Even in early works it enjoys a positive epistemic status, and Augustine often emphasizes both its usefulness (cf. Mag. 13.41, 14.46) and that one should typically accept another’s testimony.58 However, even notwithstanding the fallibility of testimony, cognizance brought about by accepting testimony falls short of knowing full well (optime scire) and its acquisition does not—Augustine thinks—deserve the name ‘learning’ (cf. Mag. 14.45 [quoted above]).59

5 . C O N C LU S I O N Following Burnyeat, it is commonly thought that in De Magistro Augustine is especially interested in explanatory understanding and that his thesis that there is no learning from words should be glossed as the claim that explanatory understanding cannot be transmitted through testimony. I have sought to offer a more fine-grained discussion of Augustine’s account. In the first instance, in De Magistro, Augustine is concerned with a variety of cognitive states. He initially focuses on more basic cognitive states—conceptual understanding (understanding what A is, having a concept of A) and linguistic understanding (understanding what ‘A’ signifies). When he turns his attention to more complex forms of cognizance (e.g. at Mag. 11.37, 12.40), it seems that he has infallible knowledge (rather than explanatory understanding) in mind. Augustine’s thinking as to why there is no learning from words is also more complex than has hitherto been appreciated. With regard to conceptual understanding, there is no learning from words because, in contrast with (e.g.) perceiving, words are insufficiently informative to serve as satisfactory grounding for conceptual understanding. Concepts attained this way are frequently inaccurate and even when they represent their objects accurately this is down to luck. With regard to linguistic understanding, Augustine thinks that words teach us signifies infallible knowledge (e.g. Ep. 120.2.9–11; Retr. 1.14.3; cf. Trin. 15.12.21) and this does not seem to be such as to come about through testimony. 58 It is not entirely clear whether the sense of ‘should’ is strictly epistemic because Augustine seems to take doxastic deliberation to be informed by non-epistemic reasons. For instance, when an unreliable friend speaks, I may have little epistemic reason to believe him (cf. Ep. 147.2.7) but caritas may compel me to (e.g. caritas omnia credit, Conf. 10.3.3; cf. F. Invis. 2.3–4). 59 Even in later works Augustine emphasizes that fides is to act as scaffolding for more elevated cognitive states (e.g. fides enim gradus est intellegendi; intellectus autem meritum fidei, S. 126.1.1, 6.8; cf. S. 139.1.1). At most it marks the start of a learning process; not its completion.

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

30

Tamer Nawar

only the “sound” or “noise” of words (merely attaching a sound to a concept already possessed). Finally, he thinks that accepting that p on the basis of another’s assertion that p fails to produce the rational certainty and indefeasibility required by knowing full well. Being fallible, testimony does not bring about infallible knowledge (one cannot, on the basis of testimony that p, rule out that not-p), and depending upon another’s testimony fails to lead one to the reflective grasp of reasons required by such knowledge.60 University of Oxford

BIBLIOGRAPHY Augustine. Against the Academicians and The Teacher, tr. P. King (Indianapolis: Hackett, 1995). Augustine. Contra Academicos, De Beata Vita, De Ordine, De Magistro, De Libero Arbitrio, ed. W. M. Green and K. D. Daur (Turnhout: Brepols, 1970). Augustine. De Trinitate libri quindecim, ed. J. W. Mountain (Turnhout: Brepols, 1968). Bermon, Emmanuel. La signification et l’enseignement: texte latin, traduction fran¸caise et commentaire du De magistro de Saint Augustin (Paris: Vrin, 2007). Bermon, Emmanuel. Le cogito dans la pense´e de saint Augustin (Paris: Vrin, 2001). Blackburn, Simon. Spreading the Word: Groundings in the Philosophy of Language (Oxford: Oxford University Press, 1984). Brittain, Charles. “Non-Rational Perception in the Stoics and Augustine,” Oxford Studies in Ancient Philosophy 22 (2002), 253–308. Burnyeat, Myles. “Wittgenstein and Augustine De Magistro,” Proceedings of the Aristotelian Society 61 (1987), 1–24. Castagnoli, Luca. “Liberal Arts and Recollection in Augustine’s Confessions X (ix 16–xii 19),” Philosophie Antique 6 (2006), 107–35. Chomsky, Noam. Aspects of the Theory of Syntax (Cambridge, MA: MIT Press, 1965). Chomsky, Noam. Knowledge of Language: Its Nature, Origin, and Use (New York: Praeger, 1986). Fodor, Jerry. The Language of Thought (Hassocks: Harvester Press, 1975). King, Peter. “Augustine on the Impossibility of Teaching,” Metaphilosophy 29 (1998), 179–95. King, Peter and Nathan Ballantyne. “Augustine on Testimony,” Canadian Journal of Philosophy 39 (2009), 195–214. Kirwan, Christopher. Augustine (London: Routledge, 1989).

60 I would like to thank Luca Castagnoli and an anonymous referee for Oxford Studies in Medieval Philosophy for very helpful criticisms and suggestions on earlier versions of this piece.

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

Augustine on the Varieties of Understanding

31

Knauer, G. N. “Sarabara. (Dan. 3,94 [27] bei Aug. mag. 10,33–11,37),” Glotta 33 (1954), 100–18. Kretzmann, Norman. “Semantics, History of,” in P. Edwards (ed.), The Encyclopedia of Philosophy (New York: Collier Macmillan, 1967), 358–406. Matthews, Gareth. “Augustine on Speaking from Memory,” American Philosophical Quarterly 2 (1965), 157–60. Matthews, Gareth. “Knowledge and Illumination,” in E. Stump and N. Kretzmann (eds), The Cambridge Companion to Augustine (Cambridge: Cambridge University Press, 2001), 171–85. Matthews, Gareth. Thought’s Ego in Augustine and Descartes (Ithaca, NY: Cornell University Press, 1992). O’Daly, Gerard. “The Response to Skepticism and the Mechanisms of Cognition,” in E. Stump and N. Kretzmann (eds), The Cambridge Companion to Augustine (Cambridge: Cambridge University Press, 2001), 159–70. Rist, John. Augustine: Ancient Thought Baptized (Cambridge: Cambridge University Press, 1994). Sirridge, Mary. “Augustine: Every Word is a Name,” New Scholasticism 50 (1976), 183–92. Strabo. Geography, tr. H. L. Jones, 8 vols (Cambridge, MA: Harvard University Press, 1917–33). Wittgenstein, Ludwig. Philosophical Investigations: The German Text, with a Revised English Translation (Oxford: Blackwell, 2001).

The Logic of Dead Humans Abelard and the Transformation of the Porphyrian Tree Margaret Cameron

Is a dead human still a human? Is it the same human, or even the same body? What is the correct meaning of the sentence, “Humans die”? These sorts of problems, which are being actively debated today, were also recognized in the early twelfth century in Paris by Peter Abelard and others who adhered to the classical definition of man, or human, as ‘rational mortal animal.’ Their concerns were not, as far as I know, ethical, although they were likely to an extent theological.1 Instead, they recognized a problem with the logic of living and dead humans based on the doctrine of substance found in Aristotle’s Categories and in the division of substance, as outlined by Porphyry to exemplify the logic of genus and species relations in the Isagoge. Abelard (1079–1142) held the view that there is no such thing as a dead human, and this provoked an intense response from his contemporaries. To understand Abelard’s position, we need to see how it is rooted in his theory of what he called “natures” and species-level possibilities, as well as his belief in substance dualism for human beings. Abelard’s contemporaries, including Alberic of Paris (fl. 1140) and other anonymous authors whose views are preserved in a handful of commentaries from this period,2 mounted a number of arguments against Abelard’s 1 The theological implications of the substantial status of human beings are, obviously, far reaching. In addition to debates over the Trinity, there are also implications for the real presence of Christ in the Eucharist and the endurance of Christ’s body in the tomb. 2 The anonymous commentaries are catalogued by John Marenbon, “Medieval Latin Commentaries and Glosses on Aristotelian Logical Texts, before c. 1150,” in C. Burnett (ed.), Commentaries and Glosses on Aristotelian Logical Texts: The Syriac, Arabic, and Medieval Latin Traditions (London: Warburg Institute, 1993). They are given an

The Logic of Dead Humans

33

views. The stakes were high for them: as realists about species and genera, Abelard’s denial that there are dead humans threatened their entire interpretation of the structure of species and genera, and thereby their theory of universals. But to see this only as a problem of the status of universals is to obscure many other interesting issues, including the status of biological entities, their identity and persistence conditions, and the viability of including human beings as individuals and as species in the category of substance. For those of us familiar with Aristotle’s entire corpus, it can be a challenge to project oneself imaginatively into this twelfth-century context in which Aristotle’s Categories and Porphyry’s Isagoge, but not De anima or Metaphysics, were circulating. Their interest in philosophical anthropology was limited to the logical problem presented by how to classify dead humans as humans. Contrast this point of focus with the great debates over the nature of human beings in the thirteenth and fourteenth centuries: as Robert Pasnau puts it, “Nothing in medieval philosophy was more fiercely contested than the topic of human nature.”3 Many of these thirteenthcentury debates were triggered by recently available texts, especially Aristotle’s De anima and commentaries on it, as well as by Council decrees and proscriptions of heretical beliefs.4 Despite the thin philosophical background for discussion, however, these twelfth-century debates are precious for revealing insight into early medieval philosophical anthropology. I will first outline the reasons for Abelard’s claim that there are no dead humans, as well as his theory of what a human being is. There is plenty of evidence that Abelard’s view was faced with criticism, and in section 1 I will run through some of those criticisms. This period is best known for the alphanumeric code. Except for those written by Abelard, all the twelfth-century commentaries on Isagoge are anonymous. Recently a commentary (or set of commentaries) has been attributed to Abelard’s teacher, William of Champeaux (see note 39). There are a few debates preserved between Abelard and another master, Alberic of Paris, excerpts of which are in L. M. De Rijk, “Some New Evidence on Twelfth-Century Logic: Alberic and the School of Mont Ste Genevie`ve,” Vivarium 4 (1966), 1–57. Yukio Iwakuma has recently highlighted some of these debates: “Alberic of Paris on Mont Ste Genevie`ve against Peter Abelard,” in J. Fink, H. Hansen, and A. Mora-Ma´rquez (eds), Logic and Language in the Middle Ages: A Volume in Honour of Sten Ebbesen (Leiden: Brill, 2013), 27–48. Iwakuma’s transcriptions of many anonymous pre-1150 commentaries are found at . I am enormously grateful to Dr Iwakuma’s generosity in sharing his extensive labors. 3 Robert Pasnau, “Human Nature,” in A. S. McGrade (ed.), The Cambridge Companion to Medieval Philosophy (Cambridge: Cambridge University Press, 2003), 208. 4 The De anima was translated in the mid-thirteenth century, followed by commentaries on it by Averroes, Themistius, and Philoponus. See Bernard Dod, “Aristoteles latinus,” in N. Kretzmann, A. Kenny, and J. Pinborg (eds), The Cambridge History of Later Medieval Philosophy (Cambridge: Cambridge University Press, 1982), 76.

34

Margaret Cameron

debates over the status of universals, which were presented to us by Abelard himself in his commentary on Isagoge, and as it turns out, the debates over the implications of Abelard’s view that there are no dead humans were tightly bound up with the debates over the status of universals. The debates found in the anonymous commentaries have to do with how to understand the so-called Porphyrian tree, which was regarded as a fixed and inviolable structure of the category of substance. Specifically, the anonymous commentators try to defend the view known as Material Essence Realism (or MERealism), or at least a version of realism very similar to it, against Abelard’s criticisms of it. By tracking the problems generated by Abelard’s view that there are no dead humans we acquire greater insight into these early twelfth-century debates over the status of genera and species, especially those prompted by the difficulty of trying to make sense of the logic of dead humans.

1 . T HE R E A R E N O D E A D H U M A N S Abelard held the view that there are no dead humans. The reason for his view is that human beings are, by nature, rational, mortal animals, i.e., living things, and thus to be dead is, as Abelard says, repugnant to the nature of a human. A clear illustration of his reasoning can be found in his treatment of modality, specifically regarding his views on possibility and modal conversions of categorical propositions in Dialectica. Abelard does not take possibility as a primitive, but rather grounds possibility in capacities or abilities, which are in turn grounded in species-natures.5 For example, given any human’s species-nature, it is possible for him to walk, run, sing, or be in Paris: this is the broad sense of possibility for a human.6 Any particular human may be better adept at, say, walking, running, or singing because of some particular developments in their lives, for example, acquired skills and practice. While the former are natural in the sense of being in-born, the latter are natural in the sense of being acquired—but

5 For more on Abelard on modality, see Christopher Martin, “Abaelard on Modality: Some Possibilities and Some Puzzles,” in T. Buchhein, C. Kneepkens, and K. Lorenz (eds), Potentialita¨t und Possibilita¨t: Modalaussagen in der Geschichte der Metaphysik (Stuttgart: Frommann–Holzboog, 2001), 97–124; Paul Thom, Medieval Modal Systems (Aldershot: Ashgate, 2003), 58–63. 6 John Marenbon, The Philosophy of Peter Abelard (Cambridge: Cambridge University Press, 1997), correctly emphasizes that possibility according to Abelard must be construed as possibility for, rather than possibility that.

OUP CORRECTED PROOF – FINAL, 28/5/2015, SPi

The Logic of Dead Humans

35

acquired in accordance with one’s species-nature.7 But in no way can a human being have the capacity for being dead, according to Abelard. This point emerges in the context of how to correctly interpret the sentence, “It is possible that a human is dead.” Abelard raises a criticism of his (unnamed) teacher (magister noster), who favors a de dicto interpretation of the sentence, “It is possible that a human is dead.” But, Abelard argues, such a reading is clearly false. On the de dicto reading— What this proposition says is possible: a human is dead [◊ (a human is dead)]— ‘human’ and ‘dead’ are supposed to cohere, but they do not because of the nature of opposites. (Abelard, Dialectica 196, 29–35)

Abelard prefers the de re interpretation— “It is possible that a human is dead” should be read “what is human is able to be made dead (idest id quod est homo, potest mortuum fieri).” (Abelard, Dialectica 196, 40–197, 4)

The reason supporting this reading of the sentence is that ‘to be made dead,’ or ‘to be dying,’ is not the same as ‘to be dead.’ Since the word ‘dying’ is synonymous with ‘living,’ the correct opposition ought to be between the words ‘dying’ and ‘dead.’ Thus, Abelard claims, “[the word] ‘human’ is not the opposite of ‘dying,’ but of ‘dead’ ” (Dialectica 197, 5–10).8 The reason, of course, is that the word ‘human’ is defined as ‘rational mortal animal,’ i.e, ‘rational mortal living-thing.’ Abelard suggests that his teacher ought to have tried to prove, not “It is possible that a human is dead,” but “A human will be dead”—perhaps then he might have seen his error, since as a categorical proposition it is supposed to assert an inherence, the inherence of death in a human (Dialectica 197, 15–25), but this is obviously false. What is a human, according to Abelard? Abelard scholars have noted that the famed logician had very little to say about human nature and the composition of body and soul.9 The human soul is immaterial, a

7 Note that Abelard, Dialectica 96, 4–14, assigns the investigation of what is in fact natural in each case to the natural scientists (physici). 8 On the semantics for the expression ‘dead human,’ see Abelard, Dialectica 116; see also Ire`ne Rosier-Catach, “Prata Rident,” in A. de Libera, A. Elamrani-Jamal, and A. Galonnier (eds), Langages et philosophie: hommage a` Jean Jolivet (Paris: Vrin, 1997), 155–76. 9 William Mann, “Ethics,” in J. Brower and K. Guilfoy (eds), The Cambridge Companion to Abelard (Cambridge: Cambridge University Press, 2004), 282: “If Abelard had a philosophical theory about the relation between soul and body, it is certainly not prominent in his writings.”

OUP CORRECTED PROOF – FINAL, 28/5/2015, SPi

36

Margaret Cameron

characterization supported by Abelard’s account of the creation of humans by God: The Lord God fashioned man from clay, i.e., from moist earth, as though it were tightly packed together rather than loose; and He thus infused the soul into a body that was already created. This passage suggests that the human soul is dissimilar from all other living beings in the very manner of its creation. For in the creation of all other living beings, God was said to have produced “the heavens and the earth” along with body and soul together—which suggests that their souls [i.e., all non-human souls] are made of those very elements.10

Abelard claims that the human soul is not composed of those elements (non de aliquo materiali primordio).11 In fact, he denies that humans are hylomorphic compounds, composed of form and matter. Abelard was aware of the Aristotelian doctrine according to which the soul is the form of the human, but he rejected this view: since forms do not have accidents, he reasoned, but the soul exhibits qualities of its own, then the soul cannot be a form.12 Thus, a human cannot be a form-matter composite. According to Abelard, each human is composed of two substances—a body and a non-body substance, body and soul—but by an act of nature, he claims, these two substances are made one.13 Abelard’s position amounts to a substance dualism for human beings, a view he shares with, and probably inherited from, Augustine. According to Augustine’s view, a human soul is a rational soul that uses a body.14 Body and soul are entirely distinct substances: “If we should define a human being such that it is a rational substance consisting of body and soul, there is no doubt that a human has a soul which is not body and a body which is not soul.”15 (Augustine then draws the point that “For man, as the ancients defined him, is a rational, mortal animal. These three, therefore, excel in man, but they themselves are not man.” The context is Augustine’s discussion of the Trinity.) Augustine is not unaware of the difficulties that accrue to a dual-substance metaphysics. He notes the challenge of deciding whether it is soul or body that is primarily constitutive of a human,16 but 10 Peter Abelard, Hexameron 102, 21–103, 3, commenting on Genesis 2: 7; translated by Peter King, “Metaphysics,” in Companion to Abelard, 76–7. See also Marenbon, Philosophy of Peter Abelard, 125–6, on God’s two creations. 11 King, “Metaphysics,” 126 cites Hexameron 104, 8. 12 Noted by Marenbon, Philosophy of Peter Abelard, 124–5. Abelard did not believe that the view was Aristotle’s: see Abelard, Gl. sup. Praed. 212, 37–213, 5. See also King, “Metaphysics,” for further references. 13 Abelard, Gl. sup. Porph. 48, 38–49, 8. Abelard’s reasons for holding this view will be discussed in section 2. 14 Augustine, De moribus 1, 27, 52: “Homo igitur, ut homini apparet, anima rationalis est, mortali atque terreno utens corpore.” 15 Augustine, De trinitate XV, ch. 7, 11. 16 Augustine, De moribus 1, 4, 6.

The Logic of Dead Humans

37

he nonetheless emphasizes that together they comprise a unity: “A soul which has a body does not make two persons, but one human.”17 Abelard agrees with Augustine, and provides this explanation (to which we will return below) of how the soul and body are one: I say that the human composite is one thing by nature, which is unified alone by the operation of nature in one inner person (in unam penitus personam), by the authority of Boethius in the book On the Trinity.18 Consider also Anastasius: for the unity of the person of Christ out of the conjoined divinity and humanity, he introduces the likeness (similitudo) out of the unity of the composite human body, saying, “For just as the rational soul and the flesh is one man, so God and man are one as Christ.” (Abelard, Gl. sup. Porph. 48, 38–49, 2)

Compared to the detailed treatments of the topic of human nature in later centuries, Abelard’s remarks are indeed scant. But they are also clear: a human is a composite of two substances, body and soul, made one by the unifying operation of nature, which is the work of God. While Abelard maintains that there can be no dead humans, he concedes that there can be dead bodies. Indexing possibility to time, Abelard’s view is that at no time will a human being be dead, although a body will be dead: However, this body, which is the human, will be dead; indeed taking ‘body’ in this way, it does not contradict ‘dead.’ Therefore just as it is granted that this body will be dead, so also we accept that it is possible that this body is dead, but not, for this reason, the human (although this body is [i.e., overlaps with] human). For it is one thing to attend to the body in its nature, and another to consider what is proper to human. For according to the substance which a human is, in the destruction of the human, the body will remain, but not the human. (Abelard, Dialectica 197, 26–31)

This passage is quite difficult to understand at first, and will be explicated in what follows. The concessive clause, “although this body is human,” seems to suggest that there can, in fact, be dead humans, although throughout the passage Abelard otherwise insists that there cannot be dead humans, only dead bodies. (Abelard says nothing further on the question of whether it is one and the same body that survives death, although it might be reasonable to presume this is the case.) Given his substance dualism, note that death is not a substantial transformation for a human being, that is, a transformation from being one type of substance (living) to another (dead). Rather, that which undergoes the change is only the body, and the human is corrupted.

17 Augustine, In evangelium Iohannis XIX, ch. 5, t. 35, col. 1553, n. 15: “Anima habens corpus non facit duas personas, sed unum hominem.” 18 Boethius, De trinitate, ch. 11.

38

Margaret Cameron

The tricky part of the passage quoted above is to understand what Abelard means by saying, “For it is one thing to attend to the body in its nature, and another to consider what is proper to human.” We can explain what Abelard means by looking at what he says elsewhere about change. Moreover, these other discussions of change will help us see the nature of the relationship between the body which can be alive or dead and the human, which cannot be dead. Abelard’s views on change, sameness, and difference have attracted a significant amount of attention from scholars, since they are both sophisticated and original. Compare a case which Abelard does discuss more elaborately, and which has been carefully analyzed by recent scholars, namely the relation between wax, as the material of a statue, and the waxen statue itself.19 According to Andrew Arlig’s analysis, “Abelard’s theory of sameness and difference in number is spelled out in terms of mereological overlap and mereological coincidence.”20 To see how Abelard construes the relation between wax and the waxen statue, we need to make a distinction between coincident and overlapping entities. As Arlig explains,  A coincides with B if and only if every part of A is a part of B and every part of B is a part of A.  A overlaps with B if and only if there is at least one thing which is both a part of A and a part of B.21 In the case of the wax and the statue, they both actually exist and, as such, are essentially the same, i.e., in actuality the same. Note that by ‘essentially’ (essentialiter) Abelard meant ‘in actual reality,’ or ‘really existent.’22 This unusual meaning of ‘essentially’ was common in the twelfth century and is, as Peter King helpfully suggests, akin to “concrete existence.”23 However, the wax, as matter, has properties that are distinct from the statue. What does Abelard mean by identifying differences in properties? Items different in property or definition are those which, although they can be predicated of themselves because they are essentially the same, with respect to their unique conditions (secundum proprios status), the proper characteristic (proprium) of 19 Abelard, Theologia Christiana III }140, 247ff. 20 Andrew Arlig, “Peter Abelard on Material Constitution,” Archiv fu¨r Geschichte der Philosophie 94 (2013), 131. 21 Arlig, “Material Constitution,” 131 n. 34. 22 Abelard scholars routinely point out this idiosyncratic, twelfth-century use of the term, beginning with Jean Jolivet, “Lexicographie Abe´lardienne,” in R. Louis, J. Jolivet, and J. Chatillon (eds), Pierre Abe´lard—Pierre le Ve´ne´rable: les courants philosophiques, litte´raires et artistiques en occident au milieu du XIIe sie`cle (Paris: E´ditions du Centre National de la Recherche Scientifique, 1975), 531–47. 23 King, “Metaphysics.”

The Logic of Dead Humans

39

this is one and the proper characteristic of that is another, and each item is bounded by definitions that are unique and distinct in their sense. (Abelard, Theologia Scholarium II }97, 455)24

For example, the wax existed prior to the statue, and the wax can be reshaped, whereas the statue is posterior and it cannot be reshaped. Reshaping the statue, in fact, amounts to its destruction. The wax thus has properties distinct from the statue and, correspondingly, a definition distinct from that of the statue. Having distinct properties entails that the wax and the statue do not coincide, but rather they overlap. These distinct properties also mark out distinct ways of being, or unique conditions of being, for the wax and the statue. These are what Abelard referred to as their ‘status,’ a technical term that is closer to our expressions ‘fact’ or ‘state of affairs.’ At the same time, the wax and the statue are numerically the same.25 Turning back to the case of the body and the human, we can now explain in Abelardian terms precisely why the body can be dead but not the human, and yet the body and the human, when it exists (i.e., when it lives), are numerically the same. As Abelard said in the passage being explicated, “For it is one thing to attend to the body in its nature, and another to consider what is proper to human.” In its nature, a human is an actually existent (in Abelard’s terms, an essentially) animate substance. Just as the wax has the property to be reshaped, the body is distinguished from the human by having the property that it can either be living or non-living (animate or inanimate). In the same way, then, the body and the human are essentially the same, differing only in properties. When the human exists, it overlaps with, but does not coincide with, the body. As an essentially biological substance, the human, when it is corrupted, i.e., when it lacks its substantial form of living, ceases to be what it was altogether. The body, which is not substantially living, however, continues. The analogy with the wax and statue is helpful for getting clear on what Abelard meant by saying that, “according to the substance which a human is, in the destruction of the human, the body will remain, but not the human” (Abelard, Dialectica 197, 30–1). As it turns out, Abelard’s substance dualism commits him to the view that the human is a composite of body and soul. As long as the human exists, it overlaps, but does not coincide, with the body. Recall that humans, on Abelard’s irrealist view, are not matter-form composites. Abelard’s analysis of the composition of 24 Abelard, Theologia scholarium II }97, 455, as translated in Arlig, “Material Constitution,” 138. For a longer explanation, see Arlig, “Material Constitution,” 133–9. 25 Arlig, “Material Constitution,” 138, puts this point formulaically: “A is different in property/definition from B if and only if A is numerically the same as B, but A’s condition (status) is different from B’s condition.”

40

Margaret Cameron

human beings is, accordingly, mereological. Both body and soul survive their decomposition, but no longer together. But the analogy must not be pressed too far, for a number of reasons. The first is that, as we have seen, the case of the body and the human is more complex because, according to Abelard, humans are dual substances, not simple substances like wax, and are the only exception to the matter-form composition of things in the world. The second is that the analogy makes it seem as if the body, as the matter of the human, is some sort of concrete stuff like wax. This is an interpretation that Abelard firmly (and famously) rejects. And the third reason not to press the analogy too far is that Abelard would not want to talk about human generation and corruption in terms of other types of change. This is because only God is capable of creating and destroying natural kinds (i.e., genuine substances, as opposed to statues and other artificial objects).26 Abelard’s view that there are no dead humans but only dead bodies provoked a flurry of responses. But to appreciate why there was such concern, we need to look more carefully at the philosophical context in which the discussion arose, namely, in the course of commenting on Porphyry’s Isagoge. We turn to this in section 2. In section 5 I will look at some of these objections and show how they construed Abelard’s position to be a threat to their way of interpreting the status of genera and species.

2 . T HE CO N T EX T O F T HE D EB A T E Objections to Abelard’s views—both that there are no dead humans and that humans are dual substances—can be found in a number of commentaries from this period. With one clear exception, the authors of these commentaries are anonymous, although they seem to share a common view about the status of genera and species: they want to defend a realist interpretation of the status of universals. The context for airing their complaints is provided by Porphyry’s Isagoge in the chapter on “Species.” 26 To be precise, the change involved in the coming-into-being of the waxen statue is not creation, but a change in status. Only God can be the author of change itself, i.e., of generation and corruption. Abelard’s views on this topic deserve more attention than I can give here, although there have been able treatments by Calvin Normore, “The Tradition of Mediaeval Nominalism,” in J. F. Wippel (ed.), Studies in Medieval Philosophy (Washington: The Catholic University of America Press, 1987), 201–17, Jeffrey Brower, “Trinity,” in Companion to Abelard, 223–57, and especially Andrew Arlig, A Study in Early Medieval Mereology: Boethius, Abelard and Pseudo-Joscelin (Ph.D. Dissertation, Ohio State University, 2005) and Arlig, “Material Constitution,” to which my discussion here is indebted.

OUP CORRECTED PROOF – FINAL, 28/5/2015, SPi

The Logic of Dead Humans

41

To understand what is at issue, and to see how this led to disputes with Abelard over the status of dead humans, a bit of background will need to be given. First, we will look at Porphyry’s text; the passage should be familiar since it is that from which Porphyry’s readers derived the Porphyrian tree, which is a division of the category of substance from genus to species, and ultimately to individual substances. To illustrate the tree, Porphyry uses the example of human beings, and in his commentary on the Isagoge Boethius follows suit. This choice of example may have inevitably led to debates over the appropriate logical categorization of human beings.27 Second, we will look at how Abelard interpreted this passage and the example of human beings. In section 3, we will turn to the critical responses to Abelard’s position on dead humans. In the Isagoge, Porphyry promised to outline, for the sake of better understanding Aristotle’s logic, what is meant by genera, species, differences, properties, and accidents, famously leaving off the “deep and demanding” questions about their ontological status.28 In the chapter on “Species,” Porphyry illustrates the division of species using the category Substance: Indeed substance is itself a genus. However, under this is body, under body is animate body, under which is animal; under animal is rational animal, under which is human; under human are Socrates and Plato and particular humans. But of these, substance is indeed most general and what is uniquely a genus, but human is most specific and the one that is only a species; body is the species of substance but is the genus of animate body; animate body is the species of body but the genus of animal; however, animal is the species of animate body but the genus of rational; rational animal is the species of animal yet is the genus of human; human is the species of rational animal, however, it is not also the genus of particular humans, but only the species; and everything which is proximate before individuals will be only a species and not also a genus.29

This passage gives us the so-called “Tree of Porphyry,” although twelfthcentury philosophers did not characterize it as such. Rather, they saw it as a

27 A similar debate took place amongst ancient Neoplatonic commentators, although none of this debate was transmitted to the early twelfth century. Boethius, if he knew it, is silent about it. See Porphyry, In Cat. 106, 26–7, following Plotinus, Ennead VI.i.2, 1–8, and Dexippus, In Cat. 40, 14–18. See especially, Pierre Hadot, “The Harmony of Plotinus and Aristotle according to Porphyry,” in R. Sorabji (ed.), Aristotle Transformed: The Ancient Commentators and Their Influence (London: Duckworth, 1990), 125–40. 28 Porphyry, Isagoge I, 1–17. 29 This is a translation of Boethius’ translation of Porphyry’s Isagoge (9, l. 19ff ).

OUP CORRECTED PROOF – FINAL, 28/5/2015, SPi

42

Margaret Cameron

fixed structure of divisions of contrary pairs organized by genus and species relations:

substance / \ body non-body / \ animal non-animal / \ rational non-rational / \ human non-human /// |||| \\\\ individual humans Of concern was the division of substance into body and non-body, given the specific case of human beings. Note that the species that is human is “the most specific and the one that is only a species.” (As we will see, it is this point, characterized by some of the commentators as “a rule of Porphyry,” that the anonymous commentators think that Abelard has violated. Their arguments turn, however, on either a willful or ignorant misreading of that claim.) In his commentary on Isagoge, Abelard remarks that it has become a commonly asked question by contemporary logicians whether the division of substance into body and non-body is sufficient, by which they meant adequate and exhaustive, or not.30 Abelard’s answer is to deny that the division of substance into either body or non-body is adequate because of the exceptional case of human beings. If we consider human beings as composites of body and soul, made one by an act of nature but ultimately two substances, then a human can be neither body alone nor soul alone. Ultimately, this results in Abelard’s decision simply to graft another branch to the Porphyrian tree, so to speak, such that the category of substance is divided into body, non-body, and the combination of body and non-body. Abelard supplies the following series of arguments to demonstrate that a body-soul composite cannot fit on Porphyry’s tree with only body and non-body as divisions of substance. If the body-soul composite is body, then (per impossibile) the body-soul composite as body would be either 30 Questioning the sufficiency of distinctions and definitions in Aristotelian logic was soon to become a commonplace trope in the commentary tradition, and it could well have been raised in the first place merely as an exegetical exercise or teaching questions. But given the ink spilled on it in the commentaries from the period, the question had clearly become a philosophical problem unto itself.

The Logic of Dead Humans

43

living or non-living. This is because, on Porphyry’s tree, body is divided into living and non-living:

body / \ living non-living If the body-soul composite is living, then (by definition of the soul as that which gives life to the body) the body-soul composite would have been made alive by means of the soul. But this is clearly false, since (by definition) the soul gives life only to the body (not to the body-soul composite). Abelard presses on in order to show that at each level or stage of Porphyry’s tree there is a problem accounting for the body-soul composite. If (per impossibile) the body-soul composite is living, it is either sensitive or non-sensitive. On Porphyry’s tree, living substance is divided into either sensitive or non-sensitive:

living / \ sensitive non-sensitive If the body-soul composite is sensitive, then it is animal and thus either mortal or immortal. If it is mortal, it is mortal either in the sense that it dies at some time, or in the sense that it never dies. But if the body-soul composite dies at some point, at some point it will be dead and thus either a dead human or something else. If it is a dead human, it will be either a body-soul composite, or a cadaver, which is clearly false (Abelard, Gl. sup. Porph. 48, 31–8). Therefore, Abelard concludes, the division of substance into body and non-body is, by reductio, inadequate. Abelard’s own answer to the question of the adequacy of the division is to make human beings an exception, and as a consequence, he simply adds another branch to Porphyry’s tree. Abelard felt himself justified in doing so because he regards Porphyry’s so-called division of Substance not as an exercise in dividing substances into contrary pairs, but as a chain of predicaments (“una praedicamenti catena sunt,” Gl. sup. Porph. 49, 9). Thus, in accordance with Abelard’s deflationary metaphysics, it is merely a chain of words that can be used in making predications. As a consequence of Abelard’s adjustment to the traditional Porphyrian tree, Abelard’s contemporaries realized that the structure of the tree as a division of substance into contrary pairs is violated. In particular, they

44

Margaret Cameron

noticed that Abelard’s solution contravenes Porphyry’s claim that the species human is “the most specific and the one that is only a species.” According to them, Abelard’s new tree looks like this:

substance / \ body non-body / \ animate non-animate / \ non-rational rational / \ non-human human ///||||\\\\ individual humans This was found to be unacceptable to the anonymous commentators and in need of being corrected. At the conclusion of his discussion of the question of the sufficiency or adequacy of the division of substance, Abelard makes the following claims, which we mentioned above, about how he classifies humans, which are by nature composite. I say that the human composite is one thing by nature, which is unified alone by the operation of nature in one inner person (in unam penitus personam), according to the authority of Boethius in the book On the Trinity.31 Consider also [what] Anastasius [said]: for the unity of the person of Christ out of the conjoined divinity and humanity, he introduces the likeness (similitudo) out of the unity of the composite human body, saying, “For just as the rational soul and the flesh is one man, so God and man are one as Christ.”

After this explanation, Abelard adds: Since, therefore, composite human is one substance by nature, in no way do I want substance to be sufficiently divided by body and non-body (unless one were to understand, perhaps, ‘non-body’ as ‘infinite body,’ i.e., ‘a-corporeal,’32 which contains the composite human). (Abelard, Gl. sup. Porph. 48, 38–49, 8)

31 Boethius, De trinitate, ch. 11. 32 It is difficult to make this point clearly in English. The Latin text reads: “Cum ergo homo compositus una sit per naturam substantia, nullo modo per corpoream et incorpoream substantiam dividi sufficienter volumus, nisi quis forte incorporeum infinitum corporei intelligat, id est, ‘non-corpus,’ quod hominem compositum continet.”

The Logic of Dead Humans

45

What Abelard means by the final claim that the only way one might make the division of substance into body and non-body, without adding a third member, is to understand ‘non-body’ as ‘infinite body’ or ‘a-corporeal’ is at first not very clear. To render words as ‘infinite’ is a lesson Abelard takes from Aristotle’s On Interpretation, in which, in outlining the characteristics of nouns, Aristotle considers whether infinite, or indefinite, nouns are, indeed, nouns properly speaking. According to Aristotle, ‘Not-man’ is not a name, nor is there any correct name for it. It is neither a phrase nor a negation. Let us call it an indefinite name. (Aristotle, On Interpretation 2, 16a30–1)

It seems from what Abelard has said that ‘infinite body or ‘a-corporeal’ would contain in its signification everything that is not a body, which would then include the soul-body composite.33

3. CRITICAL RESPONSES TO ABELARD Abelard’s views were well known, as the extant anonymous commentaries from this period attest. Most of the Isagoge commentaries from this period document awareness of Abelard’s position that the category of substance divides into (1) body, (2) non-body, and (3) both body and non-body (in the case of the human being), and nearly all of them complain about it! One set of complaints comes from the school of Alberic, who was a contemporary and rival of Abelard’s, and whose views are reported by two anonymous commentaries on this treatise. Both commentaries are products of Alberic’s school at Mont Ste-Genevie`ve, likely produced around 1140 (just after

Since I have chosen to translated ‘incorporeum’ thoughout as ‘non-body,’ I have chosen ‘a-corporeal’ for ‘non-corpus’ to indicate the infinitizing sense of the negation. 33 It bears noting that this strange comment was alluded to by John of Salisbury (1115–76) in his Metalogicon in the chapter on how to teach Porphyry’s Isagoge. The only philosophical issue taken up by John regarding this treatise concerns this very debate over the division of substance. John marshalls evidence from Cicero, Apuleius, Jerome, and Augustine, “as well as many others, both Christians and non-Christians,” that “a human is no more a body than he is a soul, and is rather, in a way, less a body than he is a soul.” John (Metalogicon III, 1) tracks Abelard’s views throughout this chapter and on this issue: “It is, however, equally true that a human is a soul, though this is held only by philosophers, but it does not follow that a human is non-body, although as Abelard used to say, negation is stronger [than affirmation].” John’s comment is not especially illuminating, unfortunately, but it conveys the sense that Abelard was willing to concede that a human can be considered a non-body, only if by non-body one meant that a human is a body-soul composite.

46

Margaret Cameron

Abelard stopped teaching), wherein Alberic’s views are favored and Abelard’s subjected to Alberic’s criticism. We know from John of Salisbury that Alberic had a reputation for constantly questioning everything and for being “a most bitter opponent” of the nominalist sect (John of Salisbury, Metalogicon II, 10). Abelard’s position on the division of substance was known to Alberic, and it prompted him to insist that, at least in the context of doing logic, ‘human’ should be taken to refer exclusively to the body. In what follows, the commentator records the debate on this issue between the two masters. Abelard’s concern throughout, according to this commentator’s report, is on the nuances of the meaning of words. Abelard recognized that the word ‘human’ is ambiguous, admitting of three meanings: For ‘human’ (homo) is taken in three ways, namely: (i) in the designation of body, which is called ‘exterior man,’ which Porphyry put under Body in the order of the categories; (ii) also ‘interior man,’ which is the species; and (iii) the human composed out of body and soul.34 The language of ‘interior’ and ‘exterior’ humanity is biblical, and is probably taken from Augustine, who uses it explicitly in describing how Father and Son are one, and how both are God with the same nature or substance.35 Because Abelard is concerned not with humans and bodies but with the meanings of the words ‘human’ and ‘body,’ he realizes that in different contexts the word ‘human’ can admit any of the three above meanings. For example, when one says, ‘Every human is body,’ it can be interpreted in terms of body only; when one says, ‘A human is consistent with soul and body,’ ‘human’ means ‘soul and body’; and when one says, “A human will be saved,” the meaning of ‘human’ is simply ‘soul’. (P20, f. 50vb; transcribed in De Rijk, “Some New Evidence,” 25)

However, Alberic took the division of substance given in Porphyry’s Isagoge to be adequate, and thereby he struggled to defend the view that humans are bodies, not body-soul composites. He suggests that, for example, in the sentence, ‘Every human is rational,’ ‘human’ can be restricted to refer to the body only. How so? According to Alberic, it is indifferent whether we say, 34 P20: MS Vienna, Oesterreichische Nationalbibliothek, V.P.L. 2486, f. 50vb; transcribed in De Rijk, “Some New Evidence.” 35 For example, see Augustine, De trinitate IV, 3, where he draws on 2 Corinthians 4: 16 for the exterior/interior distinction. See also Augustine, Epistulae 238, vol. 57, 2, p. 541, 9ff. In a report of Abelard’s mature lectures on Isagoge, there is a reference to this tripartition, although it is ambiguous whether it is supposed to be original with Abelard: see LNPS 548, 38–549, 3.

The Logic of Dead Humans

47

“A human is rational (homo rationalis est)” or “A human [i.e., its body] is sustained by the rational soul (homo vegetatur anima rationali).” Similarly, if we say, “A human has literacy (homo habet grammaticam),” ‘human’ can be taken to refer to the body in the sense that “It [i.e., its body] is sustained by a literate soul (vegetatur anima grammatica).”36 Abelard’s own view, however, is that while the soul does require the body’s eyes, ears, nose, and so on as its instruments in order to sense, the soul in no way depends upon the body for its rational activities.37 The commentary records Abelard’s response: it is not by means of the body, but by means of the soul that the human thinks, has reason, is grammatical, and so on. Similarly, if one were to say, “I am (the) animal which speaks,” or “You read wisely,” on Alberic’s restrictive meaning of ‘human’ this would mean that, “I, lacking a soul, speak,” or “You, missing a soul, read” (P20, f. 51vb–ra). For Abelard, we do not, properly speaking, say that the body understands, discerns, sees, or uses reason. Rather, it is by means of the soul that humans do such things. The body does not speak, dispute, or pray. The commentary reports some of Alberic’s criticisms of Abelard’s view, and concludes with a positive endorsement of Alberic’s opinion that ‘human’ ought to be understood as referring only to the body in logic. His primary reason is that everything that Abelard attributes to the activities of the soul can instead be attributed to the body in terms of its being sensible, i.e., both in terms of its having senses and its being able to be sensed.38 Alberic’s insistence on maintaining this meaning of ‘human’ betrays the realists’ insistence on preserving the structure of the Porphyrian tree, since admitting that ‘human’ means either ‘soul’ or ‘body-soul composite’ would violate its integrity. To focus on the body’s ability to sense and be sensed, however, is a serious misdirection on Alberic’s part. Other anonymous commentaries from this period also show familiarity with Abelard’s views. Specifically, there are a number of reproductions of Abelard’s reductio argument that aimed to prove that humans are essentially constituted by body and soul. A recently edited commentary by Yukio Iwakuma, which we will refer to as the ‘Pseudo-Rabanus’ after an incipit from one of its manuscripts, contains the following report. It is 36 P20, f. 50vb; transcribed in De Rijk, “Some New Evidence,” 25. 37 Abelard, LNPS 549. See King, “Metaphysics,” 77 and note 42, who highlights Abelard’s coincidental agreement with Aristotle in De anima 3.5: “he holds that it (i.e., the soul) is non-body for essentially the same reason as Aristotle: the understanding doesn’t need a body in order to think, whether as an instrument or object.” 38 P20, f. 51rb: “Master P. complains when we say, ‘Socrates is a body’ that the body would be Socrates, whether this body or that one, but [in fact] he seems to be neither this or that one. To this I say that Socrates is some body, because he is a sensible thing that is able to sense and also to be sensed.”

48

Margaret Cameron

unattributed to Abelard (only to “quibusdam”), but is unmistakably the same sort of argumentation given by him in his Dialectica to support that the division of substance is inadequate in the case of humans: For this human, which is made of soul and body, is neither only body nor only nonbody, but both. That he is not non-body is obvious. However, that he is not body is proved in this way: because he is neither animate nor inanimate. He is not animate because he is not living. And truly, he is not living because he is not the sort of living thing that is always alive, nor is he a living thing when he dies. That he is not the sort of living thing that is always alive is obvious. But that he is not when he dies, at some time, is proven thus: either he never dies or he dies at some time; but he dies at some time. But if this man conjoined of soul and body dies at some time, then this soul with this body dies at some time (because whatever is predicated of an integral whole is predicated of all its parts taken together). And if this soul with this body dies, then this soul with this body will be dead, which is absurd (for this soul neither on its own nor with something else will ever be dead).39

Clearly this is a version of Abelard’s argument. According to the author of this commentary, Abelard’s argument does not work, and he supplies a number of arguments against the conclusion that a third member of the division of substance needs to be added in order to handle the case of human beings. The arguments, however, are pretty thin. For example, it is merely stated that it is superfluous to add a third division because divisions are, properly speaking, made into two. Also, that composite humans “considered either as a species or a universal thing cannot be called ‘composite’ because universals cannot have constitutive parts (other than matter and form).” In this sense, the author of the commentary wants to restrict the meaning of ‘human’ when considered as a universal to a simple substance, although what it would mean to consider human beings as such is not explained. Finally, the weakest argument is that “the species is called ‘composite man’ not with regard to itself, but because it has composite individuals, since individuals can have parts, such as this man has parts: this

39 P3: MS Paris BnF, lat. 13368, f. 218vb; edited by Yukio Iwakuma, “PseudoRabanus Super Porphyrium,” Archives d’histoire doctrinale et litte´raire du Moyen Aˆge 75 (2008), 43–196. P3 is critically edited from three manuscripts that contain varying contents, likely the result of being reportationes of a master’s lectures. (In addition to the Paris version, there are MS Oxford, Bodl. Libr., Laud. lat. 67, ff. 215ra–23vb (given the title ‘Rabanvs svper Por.’), and MS Assisi, Bibl. Communale, 573, ff. 4ra–15vb.) Iwakuma has attributed these versions to William of Champeaux and his students. In this article, I set aside all attribution questions. Iwakuma, “Alberic of Paris,” mentions that a full reconstruction of all the pre-1150 manuscripts containing the debate on substance and its divisions needs to be done, and this would indeed be extremely difficult but very welcome.

The Logic of Dead Humans

49

body and this soul.”40 It is hard to evaluate these claims given the fact that the reasoning is not further articulated. The objections considered so far have to do not with dead bodies, or the qualities of life and death, but only with Abelard’s claim that humans are essentially body-soul composites, that is, with his substance dualism. The commentators appear most upset because what Abelard says appears to contradict the authority of Aristotle, Porphyry, and Boethius. Indeed, each of the objections raised against Abelard is supported with an authoritative claim, not with any independent-minded philosophical reasoning. Where things get more interesting, however, is when these anonymous commentators confront Abelard’s view that it is only the body, but not the human, that dies.

4. DEA D H U MAN S AG AI N Abelard’s contemporaries were aware of his view that only bodies, but not humans, can die. However, it is not easy to make sense of their criticisms without some careful explication of their commentaries and some important theoretical contextualization. This discussion takes place in the commentaries when commenting on the lemma: “and everything which is proximate before individuals will be only a species and not also a genus,” which is the final sentence in Porphyry’s explication of the division of substance into genus, species, and individuals, i.e., of the Porphyrian tree. The commentators note that an objection to this has been raised, and while they do not name the objector, it will become clear that it is Abelard, specifically Abelard’s claim that only bodies, but not humans, die. The objection reported by the anonymous commentator in “PseudoRabanus” is this: Here it is opposed by someone (quibusdam) in this way. If everything which is proximate before individuals (proximum ante individua) is a most special species, then [the species] body is a most special species. For [the species] body is proximate before individuals, namely, before this [individual] body in which life and death are founded. (P3, Iwakuma, “Pseudo-Rabanus,” 117)

This is indeed the implication of Abelard’s view that bodies, but not humans, die, since the quality of death does not inhere in human beings. The reasoning is reported as follows:

40 P3, Iwakuma, “Pseudo-Rabanus,” 115–16.

50

Margaret Cameron

If someone were to say this [individual] body in which death and life are founded is supposited to [the species] human, he would concede it necessary that death and life are founded in [an individual] human (for whatever are founded in some primary substance are founded in whichever is the secondary substance superior to that primary substance). But this is absurd. For death is never in [an individual] human but is only founded in a dead thing, whence they are called dead things. (P3, “Pseudo-Rabanus,” 117)

To clarify, we need to understand the distinction between something’s being founded in (fundatur) something else versus something’s being supposited to (supponitur) something else. None of the commentaries explain this distinction, but it amounts, it seems, to this. To be founded in something else is to be an actual form of something else, e.g., the whiteness of the wall is founded in the wall when the wall is, in reality, white. To be supposited to something else is not a metaphysical or physical condition; rather, it is a logical relation, which amounts to being said of something else. (Elsewhere, one commentary likens being supposited to to being conceived as.41) The rule that is given in parentheses in the quotation above—that whatever are founded in some primary substance, i.e., in some individual substance, are founded in whichever are its secondary substances, i.e., in its species and genera superior to that primary substance—can be found throughout the commentaries at this time. This rule, sometimes explicitly attributed to Aristotle, is a natural reading of Aristotle’s characterization of primary substances given in Categories 5, 2b15–17: “it is because the primary substances are subjects for all the other things, and all the other things are predicated of them or are in them, that they are called substances most of all.” This rule is accompanied by its converse: “Whatever is founded in a secondary substance is also founded in whichever are the primary substances of that secondary.” This is also supported by Aristotle’s Categories at 3a4–5: “For if you will call the individual man grammatical, then you will call both man and animal grammatical; and similarly in other cases.” However, given both these rules—that whatever is founded in, i.e., really existent in, primary substance is also founded in its secondary substance, and vice versa—there is a real difficulty with regard to the qualities of life and death. For how can death be founded in a human who is, by definition, a rational, mortal living thing? Abelard argued that this would be absurd, such that life and death can be said to be properties only of bodies, not humans. 41 See P3, “Pseudo-Rabanus,” 119: “For they say that rationality is founded in this animal, not in that which is according to the nature of human, but in that in which the nature of animal is conceived.” There is no reason to believe that the verb ‘supponere’ here is connected to the later development of the doctrine of supposition.

The Logic of Dead Humans

51

To understand the arguments given by the anonymous commentators, we must appreciate the nature of their complaint against Abelard’s view. Their complaint about Abelard’s point regarding dead bodies is twofold. First, it appears to violate the division of the category of substance into its inferiors. Second, it violates Porphyry’s rule that, directly above the level of individuals in the hierarchy of genera and species, there is always founded a most specific species (species specialissima) (e.g., human), not a genus (e.g., body). In one attempt to solve the problem, we can see that the commentator has begun to recognize the force of Abelard’s criticism, and thus tries to use the distinction between being founded in and supposited to as a way around it. The commentator explains that the qualities of life and death are founded in individual bodies, “but not always together.” For when life is in it, it is supposited to the most specific species of animate body, but when death is in it, it is supposited to the most specific species of inanimate body. (P3, Iwakuma, “Pseudo-Rabanus,” 117)

We can find the idea in another anonymous commentary from the period: This body, when it sustains life, is supposited in actuality (in essentia) to human or to some other species of animate body. However, when it falls dead, the same body, I say, regarding the nature of the body, is supposited to some species of inanimate body, whether the cadaver (if that is a most special species) or another. Therefore this body, in which death and life are founded, is always contained under a most special species. (P14, f. 7va–vb; from an unpublished transcription by Iwakuma).

The solution appeals to the logical distinction or conceptual distinction, suggesting that an individual can be logically ordered to one species and somehow, while remaining the same individual, be logically ordered to a different species when dead. Not only does this violate the Aristotelian dictum that one and the same individual cannot somehow exchange its species membership and remain the same thing,42 it does not get around the problem of how the qualities of life and death can also be founded in one and the same individual. The difficulties confronting these commentators are exacerbated by trying to make sense of the distinction between being founded in and being supposited to. This distinction tracks that given by Aristotle in Categories 2 between that which is in a subject and that which is said of a subject. By ‘in a subject,’ Aristotle means, “what is in something, not as a part, and cannot

42 According to Aristotle, “[I]t is impossible for a thing still to remain the same if it is entirely transferred out of its species, just as the same animal could not at one time be, and at another not be, a human” (Aristotle, Topics 125b37–9). This treatise was not available to scholars at this time.

52

Margaret Cameron

exist separately from what it is in” (Aristotle, Categories 1a25). Given their adherence to the two principles listed above—that whatever is founded in primary substance is founded in its secondary substances, and vice versa— these commentators were left to determine what it means to say that if life is in an individual substance it is also in its secondary substances, namely human, animal, and body. But in what sense is life and death in these secondary substances, let alone in the primary ones?

5 . T HE Q U A L I T I E S O F L I F E A N D D E A T H A N D T H E PR O B LE M O F U NIV ERS A L S We can have some sympathy for these early twelfth-century logicians trying to make sense of how life and death can be considered qualities of biological substances. Stuck with the static structure of the Porphyrian tree, with its scheme of divisions, it became increasingly difficult to maintain a place for human beings in the category of substance. Abelard’s decision to add another branch to Porphyry’s tree was one way to deal with the problem. Of course, this was not such a problem for him, since Abelard was happy to regard Porphyry’s tree not as a division of actually existent substances, but as a chain of predicaments, i.e., words that can be predicated of other words. In other words, Abelard’s non-realism provided a way around the problem of accounting for the qualities of life and death for humans. But other logicians at this time who resisted Abelard’s anti-realist position found themselves struggling to make sense of how life and death can be qualitatively founded in human beings. It is now evident that the debate over dead humans, initially triggered by Abelard’s well-known views, was bound up in an even bigger debate for which the early twelfth century is probably best known in philosophy, that is, the problem of universals. Let us now briefly set this debate over dead humans into the context of the famous disputes between Abelard and his teacher, William of Champeaux (c.1070–1122), since the answers given by the anonymous commentators discussed in section 4 become clearer in the light of the background assumptions of early twelfth-century realism. The focus of the conflict between Abelard and his former teacher was the problem of the status of universals. According to Abelard, after he destroyed William of Champeaux’s two main theories of universals, material essence realism and indifference realism, William left the profession altogether.43 Material Essence Realism (MERealism), which is the theory 43 Abelard, Historia calamitatum, 65, 81–91: “I pressured him with irrefutable arguments to make him modify, or rather to reject, his old opinion on universals.

The Logic of Dead Humans

53

that concerns us here, is described by Abelard both in his autobiography, Historia calamitatum, and in his commentary on Porphyry’s Isagoge.44 According to Abelard’s account of MERealism, universals are real and exist in individuals and their species as their matter: “Some people take ‘universal thing’ in such a way that they set up essentially the same substance in things diverse from one another through forms” (Abelard, Gl. sup. Porph. 10, 17–18; Spade, Universals, 29). Abelard uses the example of wax to illustrate MERealism: For example, if out of this [same piece of] wax I at one time (modo) make a statue of a man and at another (modo) a statue of an ox, by adapting diverse forms to entirely the same essence (essentia) that remains the same throughout. (Abelard, Gl. sup. Porph. 10, 32–4; Spade, Universals, 30, modified)

How ‘forms’ (formae) should be interpreted on each occasion in Abelard’s critique is tricky, since the term did double-duty for both accidental forms (i.e., those that can come and go from their substance without destroying it) and substantial forms, or differentiae. In this context, the former is meant. This substance is the “material essence” of the singulars it is in. It is one in itself and diverse only through the forms of its inferiors. If these forms happened to be separated from it, there would be no difference at all among the things that are in fact distinct from one another only by the diversity of forms. For the matter is entirely the same essentially. (Abelard, Gl. sup. Porph. 10, 18–23; Spade, Universals, 29–30)

Abelard illustrates MERealism with an example: “in single, numerically distinct humans there is the same substance human, which becomes Plato here through these accidents and Socrates there through those accidents” (ibid.).45 Abelard explains that the same thing holds for that which differs Now, his opinion on the community of universals was that the same thing is essentially, completely and simultaneously present in every one of its individuals, between which there was not a difference in essence, but only variety through the plurality of accidents. But he then modified his view so as then to say that it was the same not in essence but through indifference.” 44 This theory has been analyzed by Martin Tweedale, Abailard on Universals (Amsterdam: North-Holland, 1976 ), Alain de Libera, La Querelle des universaux: de Platon a` la fin du Moyen Aˆge (Paris:E´ditions du Seuil, 1976), Peter King, Peter Abailard and the Problem of Universals, 2 vols (Ph.D. Dissertation, Princeton University, 1982), and Caterina Tarlazzi, “ ‘Iam Corpus’ or ‘Non Corpus’? On Abelard’s First Argument against Material Essence Realism in the Logica ‘Ingredientibus,’ ” Vivarium 52 (2014), 1–22. The label ‘material essence realism’ was given by Tweedale, and I borrow from Tarlazzi the handy way of referring to the theory as MERealism. 45 Explicit reference to this view that things are individuated by their accidents is found in at least one of the anonymous Isagoge commentaries: “Individuals of species and of genera can in one way be considered verisimilar, but are considered in different ways.

54

Margaret Cameron

in species: animals, for example, different by species have “one and essentially the same substance animal.” Thus, individuals of the same species, and species of the same genus, have one and the same (material) essence. The addition of individual accidents to a species (e.g., Socrates’ hair color, position) makes something individually distinguishable from other members of the species. Species of the same genus are distinguished from one another by the addition of specific differentiae (e.g., the addition of rationality to animal makes human).46 Abelard’s objection to this theory is that it confounds physical explanation: “Although most authorities appear to agree with it,” Abelard says, “the philosophy of nature (physica) contradicts it in all respects” (Abelard, Gl. sup. Porph. 11, 10–11; Spade, Universals, 30). In all, Abelard gives four arguments against the theory. The first argument is this: . . . [I]f in actuality (essentialiter) the same thing exists in several individual things, then even though diverse forms are in it, this substance brought about by these forms must be the same as that one occupied by those forms. For example the animal formed by rationality is the animal formed by irrationality. Thus, rational animal is irrational animal, and so contraries reside together in the same thing. (Abelard, Gl. sup. Porph. 11–16; Spade, Universals, 30, modified)

Contraries are not able to occupy the same thing at the same time (lest they turn out not to be contraries after all, as Abelard points out). But in the scenario pictured by Abelard, one and the same individual, the animal, has contrary forms, namely rationality and irrationality. Abelard’s point is made here by applying the principle that we have seen advocated by the anonymous commentators, that “Whatever is in secondary substance will be in the primary substance of that secondary.” Thus, if animate body is divided by rationality and irrationality as its contrary pair:

For Socrates is an individual of [the species] body and of [the species] human. For if we consider Socrates according to which he is animal, rational, mortal, and discrete from others by his accidents, we attend to his being an individual of [the species] human” (P16, MS Munich clm. 14458, f. 88va; my transcription). The author of this commentary links this to a semantic explanation: “And when we hear the word (uocabulum) ‘Socrates,’ we understand these properties, for ‘Socrates’ means (notat) in its signification ‘animal, rational, mortal, and also the accidents by which he is personally discrete from others’ ” (ibid.). 46 As Iwakuma has correctly suggested, most recently in “School of Alberic,” these commentaries are indeed responding to Abelard’s critique of MERealism, a view I once challenged, what I took to be thin evidence. The case here is clear, however, since they explicitly endorse the view that it is by the addition of forms that genera are made species, and by adding substantial forms the species, not the individuals are made (pace to Abelard’s arguments to the contrary).

The Logic of Dead Humans

55

animate body / \ rational irrational e.g., human

e.g., donkey

then rationality and irrationality turn out to be in the same thing in actuality. This is because the rational animal, i.e., a human, is made up of the matter of the animate body, but the animate body contains both rationality and irrationality. Thus, both rationality and irrationality are in the same thing, that is, they make up the matter of the individual human. Similarly, rationality and irrationality make up the matter of the irrational animal, the donkey, since its matter, too, is wholly constituted by animality. Abelard notices that the success of his argument turns on demonstrating that contrary forms are actually (essentialiter) in one and the same individual at the same time, not just in species or genera, i.e., in secondary substances. The argument he uses against MERealism is this: Rationality and irrationality are truly in the same individual, because they are in Socrates. Now, the fact that they are together in Socrates is proved on the grounds that they are together in Socrates and Brunellus [i.e., a donkey]. But Socrates and Brunellus are Socrates. Socrates and Brunellus really are Socrates because Socrates is Socrates and Brunellus—that is, because Socrates is Socrates and Socrates is Brunellus (Abelard, Gl. sup. Porph. 11, 29–12, 3; Spade, Universals, 31, modified).

Note that Abelard is not trying to say that Socrates, who is a human, and Brunellus, which is a donkey, are identical to one another.47 For a MERealist, Socrates and the donkey are individuated by individual forms, such as Socrates’ hair color and height, the donkey’s being four-legged and with a tail, and so on. For this reason, Socrates and the donkey differ in properties, namely, in their individuating properties. Identity is not what is at issue in Abelard’s argument. Rather, he is concerned with substantial sameness and difference, such that Socrates and Brunellus, according to the MERealist, turn out to be the same with regard to their overlapping substance, namely, their animality. They are not, however, entirely coincident, or identical, items. But if they are substantially the same, then Socrates is Socrates and Brunellus. While the MERealist would not be at all troubled by this characterization, since it is after all just a statement of their own view, Abelard points out the 47 In the translation by Spade, Universals, 31, he adds this: “But Socrates and Brunellus are [identical with] Socrates.” The addition of ‘identical with’ is misleading for the reasons given here.

56

Margaret Cameron

absurdity of this conclusion with the following argument against the MERealist: Whatever is in Socrates other than the individual forms of Socrates is what is in Brunellus other than the individual forms of Brunellus. But whatever is in Brunellus other than the individual forms of Brunellus just is Brunellus. Thus, whatever is in Socrates other than the individual forms of Socrates is Brunellus. But if that is so, then since Socrates himself is what is other than the individual forms of Socrates, Socrates himself is Brunellus. (Abelard, Gl. sup. Porph. 12, 4–9; Spade, Universals, 31, modified)

This makes Abelard’s argument against MERealism somewhat clearer. For a MERealist, individuals are individuated by their individual forms, e.g., Socrates’ hair color and Brunellus’ tail, which are unique to the individuals that have them. Thus, if these individuating features are stripped away, what is left is what is common to both individuals as animals. Their animality comprises their matter, and their individuating forms their form. But if this is so, then ultimately MERealism holds that, in the absence of individuating, unique forms, Socrates and the donkey are one and the same. This substantial, or essential, sameness is unacceptable to Abelard, given the prevailing twelfth-century understanding of essentia as a concrete, existing thing. Socrates’ and Brunellus’ essential sameness is as impossible as the same piece of wax somehow at the same time being shaped as a human and as an ox. Abelard recognizes two further implications of MERealism. Whatever is in the donkey, according to this view, other than its individuating forms, actually is the donkey, because as he says, the [individuating] forms of Brunellus are not Brunellus, since in that case accidents would be substance, and neither are the matter and the [individuating] forms of Brunellus together Brunellus, since in that case one would have to grant that body and non-body are body. (Abelard, Gl. sup. Porph. 12, 10–14; Spade, Universals, 31, modified).

The latter part of this sentence has given interpreters some trouble; we’ll return to it below. The first part is clear: the forms that individuate Socrates or Brunellus are accidental, not substantial, forms. But if Socrates’ hair color or Brunellus’ particular tail were the forms that make Socrates and Brunellus what they actually are, then “accidents would be substance,” which is absurd.48 48 One might argue that Abelard has been unfair to the MERealist on this point, since after all what distinguishes the human from the donkey is that the human has the substantial difference of rationality. The human and donkey are substantially the same insofar as they are animals, but are substantially distinct insofar as the human is rational

The Logic of Dead Humans

57

The latter part of the sentence has caused interpretive difficulty because of a paleographical problem. According to the best reading of the single manuscript in which Abelard’s argument is contained, the text reads: cum iam corpus et iam corpus esse corpus necesse esset confiteri. (Abelard, Gl. sup. Porph. 12, 14)

The two editions of this single manuscript both correct the second iam to non: cum iam corpus et iam non corpus esse corpus necesse esset confiteri.

The reasons to accept the emendation have been carefully argued by Tarlazzi in a recently published article.49 What Abelard intends to highlight in this part of his argument is that if the matter of Brunellus, his animality, along with his individuating forms, e.g., his tail, together are what Brunellus actually is, then something which is body (Brunellus’ matter, his animality) and something which is non-body (Brunellus’ individuating forms) would be body, which is absurd. We can now see that the type of argument Abelard makes against the MERealist is the same type of argument he made in order to demonstrate that humans cannot die, but only the body can die. Both are reductio-style arguments that demonstrate the incoherence of maintaining that contrary forms occupy one and the same individual at the same time. In the case of and the donkey not. However, given the MERealists’ commitment to the rule that whatever is founded in primary substances is founded also in their species and genera, then the rationality that is founded in Socrates is also founded in animality (and would thereby be founded in Brunellus, absurdly). In other words, the MERealist conflates divisive and constitutive differentiae. I would like to thank Robert Pasnau for raising this concern. 49 The text was originally edited by Geyer, who suggested the emendation. A recent edition of part of the commentary in Claude Lafleur and Joanne Carrier, “Abe´lard et les universaux: e´dition et traduction du de´but de la Logica ‘Ingredientibus’: Super Porphyrium,” Laval the´ologique et philosophique 68 (2012), 129–210, follows Geyer’s emendation. However, there has been some controversy generated by King’s arguments in Peter Abailard in favor of the original manuscript reading, which is followed by Spade (Universals, 32) in his translation of this passage: “since in that case one would have to grant that body and something already a body are body.” Analysis of this single word in the context of Abelard’s argument has recently been undertaken by Tarlazzi, “Iam Corpus,” who argues for the emended reading. As evidence, Tarlazzi points out that, if King and Spade are right, then the two iams in the sentence would have to be doing very different jobs (i.e., ‘since’ in the first case, ‘already’ in the second). She also claims that using iam to mean ‘already’ is not an attested use of the term at this time. Tarlazzi also draws attention to the frequency of the coupling of ‘body and non-body’ in Abelard’s writings and in some of the anonymous material under examination here in order to support her argument. There is a way to make sense of the passage following either the manuscript or the emendation: see note 50.

58

Margaret Cameron

Socrates and Brunellus, if we remove each of their individuating forms what we are left with is their matter, i.e. their animality. But then, since animality admits of being both rational and irrational, these contrary forms are in each of the individuals, which is impossible. Similarly, in the case of Socrates and dead-Socrates, if we removed each of their individuating forms what we are left with is their matter, i.e., their body. But then, since body admits of being both living and dead, these contrary forms are in each of the individuals, which is impossible. Both arguments turn on Abelard’s rejection of the commitment to the principles, which are articulated by the anonymous commentators, that whatever is in a primary substance is in its secondary substances, and whatever is in a secondary substance is in each of its individuals, if by ‘in’ they mean ‘actually founded in.’ Let us now turn back to the anonymous commentators. We have already seen the report of an argument used to support the position, discussed above, that the substance animal is not proximate before its individuals, but rather body is. This is because, in the case of human beings, only bodies but not humans die. For the anonymous commentators, this conclusion unacceptably disrupts the structure of Porphyry’s tree, specifically the division of substance into contrary dividing pairs:

substance / \ body non-body / \ animate non-animate / \ non-rational rational / \ non-human human ///||||\\\\ individual humans Neither an individual human, nor a rational animal, nor an animate body is able to die, according to Abelard. Thus, ‘death’ cannot be truly predicated of any of them. This conclusion, despite its merits, makes an untidy mess of the Porphyrian structure of substance. A second argument is reported in a number of anonymous commentaries, in this case to argue pace Abelard that the species of animate substance is proximate to its individuals. That is:

The Logic of Dead Humans

59

substance / \ body non-body / \ animate non-animate / \ rational non-rational / \ non-human human ///||||\\\\ individual humans The reason, as reported by a set of anonymous commentaries, makes sense against the backdrop of Abelard’s argument against the MERealist: It is argued that [the species] living [body] is proximate before its individuals, namely this animal and that animal, for this reason: that the divisive differences of animal, rationality and irrationality, would be founded in [the species] living [body]. And thus rationality is founded in a certain individual animal itself. For whatever is founded in secondary substance is founded in whichever primary substance is of that secondary one. But this animal in which rationality is founded seems to be supposited to [the species] living [body] only, not to [the species] human, nor to any other most special species. For if it is supposited to [the species] human, then rationality is founded in [the species] human. For whatever is founded in primary substance is founded in whichever are of its secondary ones. However, that rationality is founded in [the species] human is absurd. For no constitutive difference is founded in that of which it is constituted. (P3, “Pseudo-Rabanus,” 118–19)

With a bit of reconstruction, it becomes clear that Abelard was pressing these logicians to recognize the absurd implications of MERealism. Individual humans are to be supposited, i.e., logically ordered to the species of living body, not to the species human. This is because of the rule, adhered to by MERealists, that whatever is founded in some primary substance which is logically ordered to some secondary substance is founded in that secondary substance. Thus, if rationality is actually in some individual human, then it is actually in its species and genera. However, it makes no sense to say that rationality is actually in, or founded in, the species human, since the very species itself is defined in terms of its being rational. Here, Abelard displays the incoherence of MERealism by forcing its adherents to revise what they mean by their rules about what is founded in primary and secondary substances.

60

Margaret Cameron

There are some recorded efforts by MERealists to respond to this criticism. Here is one example: This [individual] animal, of which rationality is an accident, is considered in one way according to its being, in another way according to the reception of some property. For according to the essence of the thing, it is actually under [the species] human or God, because what this man or this God is is essentially the same [i.e., rational animal]. But according to what has this [accident of ] rationality, it is under [the species] living [body], which is the proper subject of rationality. For just as Socrates from the nature of body sustains whiteness, so this [individual] animal due the nature of living [body] sustains rationality; and yet Socrates in his essence is supposited to [the species] human, others not. (P14, f. 7vb; from an unpublished transcription by Iwakuma)

The commentator tries to answer the objection by distinguishing between two ways in which an individual, rational animal can be considered. The first is in accordance with what it actually is (essentialiter), and what this animal actually is is its species, namely human. The second way to consider this individual animal is “according to the reception of some property,” that is, as a thing that is able to receive the form of rationality. In this case, what is able to receive the property or form of rationality is not the human (since the human is already rational), but the animal.50 To establish this point, a refinement of the rules of MERealism is provided by the anonymous commentator to accommodate Abelardian objections: Whatever is founded in primary substance is founded in whichever is the secondary substance in that respect in which it is founded in the primary, such that rationality is founded in that animal that is Socrates, in regard to which he is animal, not however in regard to which he is human. Or it is explained thus: Whatever is founded in primary substance in adjacency is also founded in [its] secondary substance in adjacency [and not essentially]. (P14, f. 7vb; from an unpublished transcription by Iwakuma)

By ‘in adjacency,’ the commentator means ‘with respect to its accidental forms.’ However, this response is completely unsuccessful on any number of fronts. Most of all, the commentator arbitrarily and without reason jettisons the Aristotelian view that rationality and other differentiae of substance are substantial, not accidental differences! Rationality is not merely accidental to humans. But it seems Abelard was already aware of 50 Putting the argument this way makes it seem as if King’s interpretation of Abelard’s argument against MERealism is entirely defensible. Briefly, King, Peter Abailard, 156, favored the interpretation of the apparently corrupted sentence (see note 49) that preserves the second ‘iam,’ resisting emending it to ‘non.’ For details about this interpretive debate, see Tarlazzi, “Iam Corpus.”

The Logic of Dead Humans

61

an effort such as this one to save MERealism, since he supplies a response to it in his set of arguments against the theory. Abelard, making quick work of this feeble effort, pointedly remarks that “nothing is a human insofar as it is an animal” (Abelard, Gl. sup. Porph. 12, 25–6).

6 . C O N C LU S I O N Abelard’s views on human nature provoked critical response from his contemporaries, as we have seen here. The quality of their critical responses is often very poor, especially since their arguments are not elaborated at length. It might seem odd that Abelard did not use the forms of life and death in order to generate his reductio of MERealism, since the example makes his point so clearly. After all, we need an explanation of how a living thing can die according to a theory in which an individual’s essence is its matter, and for any biological substance this involves its being essentially alive. Moreover, on the MERealists view, given that division into inferiors is made by accidents, not substantial differences, then being alive (or dead) is merely accidental to being a biological substance, indeed, to being a human being, which seems absurd. By rejecting Abelard’s substance dualism for human beings, and by insisting on the fixity of the Porphyrian tree as a division of contrary pairs, they leave themselves no resource to account for something that can endure the transformation of bodily death. University of Victoria

BIBLIOGRAPHY Arlig, Andrew. A Study in Early Medieval Mereology: Boethius, Abelard and PseudoJoscelin (Ph.D. dissertation, Ohio State University, 2005). Arlig, Andrew. “Peter Abelard on Material Constitution,” Archiv fu¨r Geschichte der Philosophie 94 (2013), 119–46. Augustine. Epistulae 185–270 (Corpus scriptorum ecclesiasticorum latinorum 57) (Vienna: Tempsky, 1923). Augustine. De moribus ecclesiae catholicae et de moribus manichaeorum libri duo (Corpus scriptorum ecclesiasticorum latinorum 90) (Vienna: Tempsky, 1992). Augustine. De trinitate (Corpus Christianorum, series latina 50/50A) (Turnhout: Brepols, 1968). Augustine. In evangelium Iohannis CXXIV (Corpus Christianorum, series latina 36) (Turnhout: Brepols, 1954). Boethius. De trinitate, in C. Moreschini (ed.), Opuscula sacra and De consolatione philosophiae (Munich: K. G. Saur, 2000).

62

Margaret Cameron

Brower, Jeffrey. “Trinity,” in J. Brower and K. Guilfoy (eds), The Cambridge Companion to Abelard (Cambridge: Cambridge University Press, 2004), 223–57. de Libera, Alain. La querelle des universaux: de Platon a` la fin du Moyen Aˆge (Paris: E´ditions du Seuil, 1976). De Rijk, L. M. “Some New Evidence on Twelfth-Century Logic: Alberic and the School of Mont Ste Genevie`ve,” Vivarium 4 (1966), 1–57. Dexippus. Commentary on the Categories [In Cat.], in J. Dillon (tr.), On Aristotle Categories (London: Duckworth, 1990). Dod, Bernard G. “Aristoteles latinus,” in N. Kretzmann, A. Kenny, and J. Pinborg (eds), The Cambridge History of Later Medieval Philosophy (Cambridge: Cambridge University Press, 1982), 45–79. Hadot, Pierre. “The Harmony of Plotinus and Aristotle According to Porphyry,” in R. Sorabji (ed.), Aristotle Transformed: The Ancient Commentators and their Influence (London: Duckworth, 1990), 125–40. Iwakuma, Yukio. “Alberic of Paris on Mont Ste Genevie`ve against Peter Abelard,” in J. Fink, H. Hansen, and A. Mora-Ma´rquez (eds), Logic and Language in the Middle Ages: A Volume in Honour of Sten Ebbesen (Leiden: Brill, 2013), 27–48. Iwakuma, Yukio. “Pseudo-Rabanus Super Porphyrium,” Archives d’histoire doctrinale et litte´raire du Moyen Aˆge 75 (2008), 43–196. John of Salisbury, Metalogicon, ed. J. B. Hall and K. S. B. Keats-Rohan (Corpus Christianorum, continuatio medievalis 98) (Turnhout: Brepols, 1991). Jolivet, Jean. “Lexicographie Abe´lardienne,” in R. Louis, J. Jolivet, and J. Chatillon (eds), Pierre Abe´lard—Pierre le Ve´ne´rable: les courants philosophiques, litte´raires et artistiques en occident au milieu du XIIe sie`cle (Paris: E´ditions du Centre National de la Recherche Scientifique, 1975), 531–47. King, Peter. “Metaphysics,” in J. Brower and K. Guilfoy (eds), The Cambridge Companion to Abelard (Cambridge: Cambridge University Press, 2004), 65–125. King, Peter. Peter Abailard and the Problem of Universals, 2 vols (Ph.D. Dissertation, Princeton University, 1982). Lafleur, Claude and Joanne Carrier. “Abe´lard et les universaux: e´dition et traduction du de´but de la Logica ‘Ingredientibus’: Super Porphyrium,” Laval the´ologique et philosophique 68 (2012), 129–210. Mann, William E. “Ethics,” in J. Brower and K. Guilfoy (eds), The Cambridge Companion to Abelard (Cambridge: Cambridge University Press, 2004), 279–304. Marenbon, John. “Medieval Latin Commentaries and Glosses on Aristotelian Logical Texts, before c.1150,” in C. Burnett (ed.), Commentaries and Glosses on Aristotelian Logical Texts: The Syriac, Arabic, and Medieval Latin Traditions (London: Warburg Institute, 1993). Marenbon, John. The Philosophy of Peter Abelard (Cambridge: Cambridge University Press, 1997). Martin, Christopher J. “Abaelard on Modality: Some Possibilities and Some Puzzles,” in T. Buchhein, C. Kneepkens, and K. Lorenz (eds), Potentialita¨t und Possibilita¨t: Modalaussagen in der Geschichte der Metaphysik (Stuttgart: Frommann-Holzboog, 2001), 97–124. Normore, Calvin G. “The Tradition of Mediaeval Nominalism,” in J. F. Wippel (ed.), Studies in Medieval Philosophy (Washington: The Catholic University of America Press, 1987), 201–17.

The Logic of Dead Humans

63

Pasnau, Robert. “Human Nature,” in A. S. McGrade (ed.), The Cambridge Companion to Medieval Philosophy (Cambridge: Cambridge University Press, 2003), 208–30. Peter Abelard. Dialectica, in L. M. De Rijk (ed.), Petrus Abaelardus Dialectica: First Complete Edition of the Parisian Manuscript (Assen: Van Gorcum, 1956). Peter Abelard. Glossae super Praedicamenta [Gl. sup. Praed.] in B. Geyer (ed.), Peter Abaelards philosophische Schriften II: die Logica ‘Ingredientibus’, 2. Die Glossen zu den Kategorien (Beitra¨ge zur Geschichte der Philosophie und Theologie des Mittelalters 21/2) (Munster: Aschendorff, 1921). Peter Abelard. Glossulae super Porphyrium [Gl. sup. Porph.] in B. Geyer (ed.), Peter Abaelards philosophische Schriften II: die Logica ‘Ingredientibus’, 1. Die Glossen zu Porphyrius (Beitra¨ge zur Geschichte der Philosophie und Theologie des Mittelalters 21/1) (Munster: Aschendorff, 1919). Peter Abelard. Hexameron, in M. Romig (ed.) (with D. Luscombe), Expositio in Hexameron (Corpus Christianorum, continuatio medievalis 15) (Turnout: Brepols, 2004). Peter Abelard, Historia calamitatum, ed. J. Monfrin, 4th edn (Paris: Vrin, 1974). Peter Abelard, Logica ‘nostrorum petitioni sociorum’ [LNPS], in B. Geyer (ed.), Peter Abaelards philosophische Schriften: die Logica “Nostrorum Petitioni Sociorum,” Die Glossen zu Porphyrius (Beitra¨ge zur Geschichte der Philosophie und Theologie des Mittelalters 21/4) (Munster: Aschendorff, 1933). Peter Abelard. Theologia christiana, in E. M. Buytaert (ed.), Opera theologica, vol. 2 (Turnhout: Brepols, 1969). Peter Abelard. Theologia ‘scholarium,’ in E. M. Buytaert and C. Mews (eds), Petri Abaelardi Opera theologica (Corpus Christianorum, continuatio mediaevalis 13) (Brepols: Turnhout 1987). Plotinus. Ennead V.I., 1–5, ed. A. H. Armstrong (Cambridge, MA: Loeb Classical Library, 1988). Porphyry. Introduction, tr. J. Barnes (Oxford: Oxford University Press, 2003). Porphyry. Isagoge secundum translationem quam fecit Boethius, ed. L. Minio-Paluello (Aristoteles Latinus I, 6–7: Categoriarum supplementa) (Paris: Descle´e de Brouwer, 1966), 5–31. Porphyry. On Aristotle’s Categories [In Cat.], tr. S. Strange (Ithaca, NY: Cornell University Press, 1992). Rosier-Catach, Ire`ne. “Prata Rident,” in A. de Libera, A. Elamrani-Jamal, and A. Galonnier (eds), Langages et philosophie: hommage a` Jean Jolivet (Paris: Vrin, 1997), 155–76. Spade, Paul Vincent. Five Texts on the Mediaeval Problem of Universals (Indianapolis: Hackett, 1994). Tarlazzi, Caterina. “ ‘Iam Corpus’ or ‘Non Corpus’? On Abelard’s First Argument against Material Essence Realism in the Logica ‘Ingredientibus,’ ” Vivarium 52 (2014), 1–22. Thom, Paul. Medieval Modal Systems (Aldershot: Ashgate, 2003). Tweedale, Martin. Abailard on Universals (Amsterdam: North-Holland, 1976).

The Fate of the Flying Man Medieval Reception of Avicenna’s Thought Experiment Juhana Toivanen 1. INTRODUCTION Thought experiments are nowadays an important part of philosophical discourse. They are used to illuminate theoretical arguments and to reveal philosophically significant features of an argument or a problem in a way that aspires to precision. Although the phrase ‘thought experiment’ itself is quite recent, the use of imagined counterfactual cases which are designed to have a role in philosophical argument can be traced back to Antiquity, and the technique was widely employed in the Middle Ages.1 One of the most well-known medieval thought experiments is the so-called ‘flying man’ of Avicenna. The flying man is an imaginary person whose senses do not provide any input but who nevertheless admits his own existence, or the existence of his soul. Avicenna’s idea was that this thought experiment helps us to understand that the soul is not identical to the body but an immaterial entity. Modern scholarship has discussed the Latin reception of this thought experiment to some extent,2 but there are a couple of issues that have not

1 For the role of thought experiments in the history of Western philosophy, see Katerina Ierodiakonou and Sophie Roux (eds), Thought Experiments in Methodological and Historical Contexts, History of Science and Medicine Library 19 (Leiden: Brill, 2011), 1–142. 2 Dag Nikolaus Hasse, Avicenna’s De anima in the Latin West: The Formation of a Peripatetic Philosophy of the Soul, 1160–1300 (London: The Warburg Institute, 2000), 80–92 and 236; Deborah J. Brown, “Aquinas’ Missing Flying Man,” Sophia 40 (2001), 17–31; E´tienne Gilson, “Les sources gre´co-arabes de l’augustinisme avicennisant,” Archives d’histoire doctrinale et litte´raire du moyen aˆge 4 (1929), 40–2.

The Fate of the Flying Man

65

received as much attention as they deserve. The first is related to the changes in the argumentative role of the flying man.3 Avicenna’s illustration was known to medieval authors in the Latin West through his De anima, which is a translation of the sixth book of his vast scientific and philosophical work Kita¯b al-Shifa¯’. In the first stage of the reception, before the middle of the thirteenth century, medieval authors used the flying man roughly for the same purpose as Avicenna: to show the immateriality of the soul and its existence as something that is distinct from the body. Some authors understood the illustration solely as an argument for the existence of the soul, while others thought that it also revealed the nature of the soul, but it was unquestionably considered to be an ontological argument. The argumentative role of the flying man changed quite radically in the latter half of the thirteenth century, however, when the flying man was brought into an epistemological context and understood as an argument for the soul’s ability to be directly aware of itself. Philosophers started to use the flying man as an epistemological argument rather than as an ontological one. In order to see how this change took place, I shall discuss all the known Latin versions of the flying man: Dominicus Gundissalinus (fl. c.1150), William of Auvergne (d. 1249), Peter of Spain (fl. c.1240?), John of la Rochelle (d. 1245), Matthew of Aquasparta (c.1240–1302), and Vital du Four (1260–1327). The other neglected aspect of Avicenna’s thought experiment is closely related to the change in the argumentative role that medieval authors gave to it. Namely, Avicenna and his early thirteenth-century readers agreed that the flying man is directly aware of his soul, but, at the same time, they needed to argue that he is totally unaware of his body. Lack of awareness of the body is one of the main premises in the ontological version of the thought experiment. When the emphasis shifted to the epistemological aspect of the thought experiment, the distinction between the soul and the body became peripheral to the argument. In order to establish that the soul can be directly aware of itself, it was not necessary anymore to hold on to the presupposition that it would be unaware of the body if it did not perceive external objects. This development is fairly clear in the case of the late thirteenth-century authors. Matthew of Aquasparta does not mention bodily self-awareness or lack thereof at all. Vital du Four mentions it, but he shows some signs of 3 Although Hasse points out that certain Latin authors use the flying man for different purposes than Avicenna, he does not provide a detailed and contextualized analysis of their positions for their own sake. His main aim is to compare the Latin interpretations with Avicenna’s original view. See Hasse, Avicenna’s De anima in the Latin West, 89–90.

66

Juhana Toivanen

willingness to go against Avicenna, as he seems to think that even the flying man would be aware of his body. However, there is a more striking example of this trend, namely, Peter of John Olivi’s (c.1248–98) thought experiment of a ‘man before the creation.’ While developing his theory of perception, Olivi illustrated how the soul’s cognitive attention functions by asking us to imagine a man who existed before the creation of the world. This thought experiment resembles the flying man in many ways, but the most significant difference is that Olivi’s thought experiment requires that the man in this peculiar state is aware of his body and his ability to perceive. In this way, the man before the creation reflects the late thirteenth-century trend of the diminishing centrality of the premise that the flying man is unaware of his body. It also shows how the thought experiment received a new role in the hands of an original philosopher. My objective in the present essay is to analyze the various versions of the flying man and to trace its fate from its first appearance in Latin philosophy to the dawn of the fourteenth century. I shall begin with a concise explanation of the relevant features of Avicenna’s flying man, after which I shall discuss its reception in the Latin West and point out how the various interpretations presented by Latin authors shift the emphasis of the thought experiment and bring new aspects to the fore (sections 2 and 3). Finally, I shall present Olivi’s ‘man before the creation’ and discuss its relation to the flying man tradition (section 4).

2 . O N T O LO G I C A L F L Y I N G M A N The locus classicus of the flying man is at the end of the first chapter of Avicenna’s De anima. The flying man appears in a context where Avicenna is striving to establish the existence of the soul as well as to define it and its relation to the body. By employing the thought experiment involving the flying man, he wants to illuminate his dualistic view that the soul is not the whole body or a part thereof.4 As we are dealing here less with the original flying man and more with his fate in Latin philosophy, let us have a look at the passage as it appears in the Latin translation, which was prepared by Dominicus Gundissalinus and Avendauth:

4 Avicenna does not consider the flying man as a scientific argument in the strict sense. Rather, it is a pedagogical aid or pointer that helps us to grasp the truth without actually proving it. Latin authors often gave it a stronger epistemological role and considered it as a real argument.

The Fate of the Flying Man

67

We say, therefore, that one of us must imagine himself as created all at once and perfect but with his sight veiled from seeing external things, and created as moving in the air or the void so that the density of the air, which he could perceive, would not touch him, and with his limbs separate in such a way that they neither meet nor touch each other. He must then see if he affirms the existence of his essence. For, he would not hesitate to affirm that he exists, but he would not affirm anything external about his members, anything hidden about what is inside him, neither his mind nor his brain, or anything external whatsoever. But he would affirm that he exists without affirming his length, width, or density. If he were able at that time to imagine a hand or some other member, he would not imagine it to be a part of him or necessary to his essence. Now, you know that what is affirmed is different from what is not affirmed and what is granted is different from what is not granted. And because the essence that he affirms to exist is proper to him in that it is himself and in addition to the body, he who is attentive has the means to be awakened to knowing that the being of the soul is different from the being of the body; indeed, he does not need the body in order to know and perceive the soul.5

Modern scholars have presented various interpretations concerning Avicenna’s illustration.6 A particular problem that they have tackled arises from the terminology that Avicenna uses. In the original Arabic version he claims that the flying man is aware of his dha¯t, but because this term means both ‘essence’ and ‘self,’ it is not clear what the main thrust of this claim is. Dag Hasse has argued that Avicenna meant ‘essence,’ and this reading is one of the cornerstones of his analysis of the Latin reception of the flying man: in effect, he evaluates the Latin versions with respect to their closeness to what he calls “the correct interpretation.”7 Yet, it has been suggested that the ambiguity was intentional on the part of Avicenna and that he deliberately employed a term that denoted both the self and the essence.8 If this is truly the case, the solution of Gundissalinus and Avendauth to juxtapose different expressions, some of which make use of the essence (“affirmat esse suae essentiae”) and others that seem to denote self-awareness or awareness of one’s existence (“affirmabit se esse”), may be intentional as well. Latin does not have a single word similar to the Arabic dha¯t, and thus the translation is actually quite good at conveying the ambiguity. 5 Avicenna, Liber de anima 1.1, 36–7. See also ibid., 5.7, 162–3. 6 See e.g. Jari Kaukua, Self-Awareness in Islamic Philosophy: Avicenna and Beyond (Cambridge: Cambridge University Press, 2015), 30–42; Deborah Black, “Avicenna on Self-Awareness and Knowing that One Knows,” in S. Rahman, T. Street, and H. Tahiri (eds), The Unity of Science in the Arabic Tradition (Dordrecht: Springer, 2008), 63–87; Daniel Heller-Roazen, The Inner Touch: Archaeology of a Sensation (New York: Zone Books, 2007), 219–21; Michael Marmura, “Avicenna’s ‘Flying Man’ in Context,” Monist 69 (1986), 383–95. 7 Hasse, Avicenna’s De anima in the Latin West, 87–9. 8 Kaukua, Self-Awareness in Islamic Philosophy, 38–41.

68

Juhana Toivanen

Whatever Avicenna’s original intention was, the thought experiment contains many aspects that are less controversial. The flying man does not have any sensory experiences, he does not perceive any external objects or their qualities, and he does not feel his body through a sense of touch. He is created perfect and, we may assume, without any previous experiences. Even in this state, the flying man would admit his own existence (or the existence of his essence, which is the soul), but he would be completely unaware of his body and would not attribute any bodily quality to himself. In order to understand the later adaptations of this thought experiment, we may reconstruct its argumentative structure in a simplified form as follows: (P1) The senses of the flying man are inactive. (P2) The flying man is not aware of his body (from P1). (P3) The flying man is aware of his existence/soul. (P4) What the flying man is aware of is different from that which he is unaware of. (C1) The soul is different from the body (from P2, P3, and P4). [(C2) The body is not needed in order to become aware of the soul (from P1/P2 and P3).]

The first premise P1 is taken as granted in the thought experiment, but it can be explained on the basis of Avicenna’s theory of perception. Even though he characterizes sense organs as instruments of the soul, and thus rejects certain aspects of Aristotle’s theory in which perception is portrayed as completely passive reception of external stimuli, he nevertheless seems to think that the process of perception is triggered by an external object which acts on the sense organs.9 Avicenna describes the condition of the flying man in such a way that either there are no external objects or they do not act upon the senses, and thus the senses themselves cannot act. The premise P2 presupposes that direct perception of one’s body prior to perceiving external objects is impossible. Avicenna does not provide any further evidence in favor of P3, but his idea is that the initial setting of the thought experiment leads the reader to realize that this kind of self-awareness is part of our experience even though we normally do not notice it.10 He also assumes the

9 See, e.g., Simo Knuuttila, “Aristotle’s Theory of Perception and Medieval Aristotelianism,” in S. Knuuttila and P. Ka¨rkka¨inen (eds), Theories of Perception in Medieval and Early Modern Philosophy, Studies in the History of Philosophy of Mind 6 (Dordrecht: Springer, 2008), 8–9. 10 In accepting this reading of the thought experiment, I disagree with Hasse, who claims that self-awareness does not figure at all in the thought experiment (Hasse, Avicenna’s De anima in the Latin West, 83–6). Most modern scholars think that it does. For a discussion and references, see Kaukua, Self-Awareness in Islamic Philosophy, 30–42.

The Fate of the Flying Man

69

premise P4, which is problematic, because it is not clear whether it is supposed to prove only that the body and the soul are two different entities—in which case, it would prove only the existence of the soul—or whether it proves additionally that the nature of the soul is different from that of the body (i.e. that the soul is incorporeal).11 To be sure, the latter reading is suggested by the flying man’s denial of any corporeal quality to himself, but we shall see below that Avicenna’s Latin readers sometimes used the thought experiment to prove only the existence of the soul. On the basis of these premises, Avicenna concludes that the soul is an immaterial entity which is not identical to the body (the conclusion C1). He uses the thought experiment to suggest to the reader that there is something else in a human being other than just the body. In other words, the flying man supports Avicenna’s substance dualism, which he will later prove by using other means.12 The thought experiment ends with what may appear as an additional conclusion, C2: the body is not needed for becoming aware of the soul. Whether this addition is meant to be a further epistemological conclusion, or just a reminder of the role that selfawareness plays in the thought experiment, is not completely clear. In the original Arabic version Avicenna does not present it as a conclusion—in fact, he writes merely that the thought experiment enables us to realize that we are aware of our souls, and the body is mentioned only with respect to C1. Gundissalinus’ translation, however, permits an alternative reading in which C2 follows after C1 as a further conclusion.13 We shall now see that Gundissalinus himself understands the ending of the thought experiment in the latter way, and thus Latin readers of Avicenna encounter an ambiguity concerning the central outcome of the flying man right from the beginning.

11 It has been argued that the premise is problematic also because it appears to be a fallacy: from the fact that I am aware of X but unaware of Y, it does not follow that X is not Y. See Black “Avicenna on Self-Awareness,” 64–5. 12 Avicenna, Liber de anima 5.2, 81–101. 13 Avicenna’s original text goes as follows: “Thus, he who takes heed has the means to take heed of the existence of the soul as something different from the body—indeed, as different from any body—and to know and be aware of it.” (Avicenna, Avicenna’s De anima (Arabic Text): Being the Psychological Part of Kita¯b al-shifa¯’, ed. F. Rahman (London: Oxford University Press, 1959), 1.1, 16; translated by J. Kaukua in SelfAwareness in Islamic Philosophy, 35.) Gundissalinus translates: “Ideo expergefactus habet viam evigilandi ad sciendum quod esse animae aliud est quam esse corporis; immo non eget corpore ad hoc ut sciat animam et percipiat eam.” (Avicenna, Liber de anima 1.1, 37.) It is possible to understand the word immo as the beginning of an additional conclusion, even though this may not be the most natural reading.

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

70

Juhana Toivanen

a. Dominicus Gundissalinus The flying man is well known in modern scholarship but, perhaps surprisingly, it was discussed relatively little in the Latin West during the Middle Ages. Avicenna’s De anima became highly influential in psychological theories after it was translated into Latin in the latter half of the twelfth century, but there are only a handful of philosophers who repeat the thought experiment.14 Generally speaking, they think that it is reliable: the flying man is aware of himself and his soul, but his body is completely oblivious to him, and therefore the soul cannot be identical to the body or any part of it. Yet, a close reading of even the earlier interpretations reveals important differences from Avicenna. Let us begin with Dominicus Gundissalinus, who was the first Latin philosopher to use the thought experiment. He presents the flying man in his Tractatus de anima, which is a philosophical treatise concerning the nature and powers of the soul. In order to understand what kind of argumentative role Gundissalinus assigns to the flying man, it is necessary to take into consideration the structure of the work. Gundissalinus begins by asking whether the soul exists. After having established its existence, he gives a general Platonic definition of the nature of the soul that is taken from Costa ben Luca’s De differentia animae et spiritus: “The soul is an incorporeal substance, which moves the body.”15 This definition contains three parts (the soul is a substance, it is incorporeal, and it moves the body), and Gundissalinus proceeds to argue for each of them in turn. He begins by presenting two short arguments that purport to show that the soul is a substance, and then he proceeds to demonstrate its incorporeity. The flying man thought experiment appears at this point in the treatise. It is the first argument in a long series of arguments that Gundissalinus presents in order to prove the incorporeity of the soul. In other words, he understands the flying man as an ontological proof, but he does not use it to establish the existence or substantiality of the soul, because these two are already proved by other means. The role that he assigns to it in his overall argument is to prove that this existing substance is not a body. Gundissalinus’ version of the flying man is very close to the Latin translation of Avicenna’s original text—which is not surprising, given 14 Gilson, “Les sources gre´co-arabes,” 41–2, was the first to provide a list of Latin authors who quote Avicenna’s De anima 1.1. It includes William of Auvergne, John of la Rochelle, Matthew of Aquasparta, and Vital du Four. Hasse supplements Gilson’s list by adding Peter of Spain and an anonymous author (see footnote 36). 15 Dominicus Gundissalinus, Tractatus de anima 2, 37. Cf. Costa ben Luca, De differentia spiritus et animae, 166–9.

The Fate of the Flying Man

71

that the translation was prepared by him—and he accepts all the premises that are listed above. There is no need to quote the whole passage, but a couple of points are worth mentioning. First, Gundissalinus makes an addition to Avicenna’s list of things that the flying man is unaware of: “Indeed, he would not then affirm anything external or internal about his members nor that he is living or capable of sensations (animatum vel sensibilem) or anything like that.”16 The original flying man is unaware of his own body and all corporeal qualities (the premise P2), but Gundissalinus adds a further detail by saying that the man in this state would not be aware of his ability to perceive, which means that he would not be aware of those functions of the soul that take place in his body. In addition, he would not be aware of being alive. We may assume that Gundissalinus means biological life: the flying man does not notice his body, and therefore he fails to notice that he is living in the sense of being a living body. Another important point in Gundissalinus’ interpretation is related to the argumentative role that he assigns to the flying man. There is no doubt that he sees it primarily as an ontological argument. Nevertheless, at the end of the passage in which the flying man appears, we encounter an additional epistemological conclusion: “Surely, it can be clearly understood that the soul has a different being than the being of the body, and that a human being does not need the body in order to know and perceive his soul.”17 Unlike Avicenna (and more clearly than the Latin translation18), Gundissalinus draws two conclusions: an ontological one that the soul exists independently of the body and is therefore incorporeal (the conclusion C1), and an epistemological one that the body is not necessary in order to be aware of the soul (C2). Yet, even though C2 is understood as a conclusion, the rationale for making it is somewhat obscure. Arguably, it is nothing but a repetition and reformulation of the premise P3—that the flying man is able to be aware of his existence/soul even though he does not use any cognitive powers that require bodily organs. Gundissalinus accepts this premise in order to establish the ontological conclusion, and the epistemological conclusion does not add anything new to the picture. Thus, even though C2 appears in conjunction with C1 and Gundissalinus seems to consider it as a conclusion, it

16 Gundissalinus, Tractatus de anima 2, 37. 17 Ibid. 18 The differences between the translation and Gundissalinus’ own version are small but important: where the translation has it that the flying man does not need the body in order to know his soul, Gundissalinus makes a general conclusion that no human being needs it. Also, instead of the ambiguous immo, Gundissalinus uses et in order to connect two equally important conclusions.

72

Juhana Toivanen

is difficult to see how it differs from Avicenna’s initial assumptions and especially from the premise P3. One might therefore argue that it is nothing but a necessary premise that leads to C1. True enough, C2 contains an idea which is lacking in the premise P3, namely that the body is not needed for an awareness of the soul. This addition suggests that the former is not identical to the latter because it states both that human beings are capable of cognizing their souls and that they can do so without the body. However, even in this case, it is clear that C2 does not bring anything new to the picture but only repeats the basic assumptions that constitute the thought experiment in the first place. Moreover, C2 contains an ambiguity, because it is not clear whether it is supposed to result from P1 or from P2. In the first case the claim would be that bodily acts pertaining to external objects are not necessary for being aware of oneself; in the latter case the awareness of the body itself would be the crucial element that is not needed for self-awareness. The choice between these readings depends on the way in which we read the expression non eget corpore, which is puzzling because the flying man does have a body, after all. He just does not use it. It would be more precise to say either that bodily acts are not necessary for being aware of the soul (P1 and P3), or that awareness of the body is not required (P2 and P3). We shall see below that late thirteenth-century authors favored the former interpretation, but Gundissalinus may simply think that there is no difference between the two: awareness of one’s soul is possible without bodily acts and without awareness of one’s body. What is certain is that he does not explain his idea further, probably because his interest lies in the ontological consequences of the thought experiment. The epistemological conclusion—if indeed it is a conclusion, given that it does not add anything to the premises that are presupposed in the thought experiment—is more like an afterthought, as the main thrust of the argument is to show that the soul is incorporeal.

b. William of Auvergne We shall see below how the philosophers in the latter half of the thirteenth century emphasized the relation between bodily acts and self-awareness, but the idea that cognitive acts are relevant for self-awareness appeared already during the first half of the century. Namely, William of Auvergne’s (d. 1249) interpretation of the thought experiment is based on the idea that even though the flying man does not perceive anything, he has cognitive acts through which he becomes aware of himself: If we put a man in the air who has his face covered and is without the use of any sense and who has not used any sense, there is no doubt but that it is possible that

The Fate of the Flying Man

73

this man thinks and understands. Hence, he will know that he thinks or understands, and he will know that he himself exists. And if he asks himself whether he has a body, he will undoubtedly say that he does not have a body, and in the same way he will deny of himself each and every part of a human body.19

William argues that the flying man is not deprived of intellectual operations of the soul even though he cannot use his senses. In this way, he is able to give a reason for accepting the premise P3: the flying man can think and understand although his senses are inactive (P1), and this cognitive activity makes him aware of his own thinking and existence. William’s argument proceeds from the acts of the soul to the soul’s awareness of itself, but he paraphrases Avicenna’s thought experiment in order to argue that the soul is not a body and that the soul is completely spiritual. It is therefore clear that his version is an ontological flying man. The aim of the argument is to establish the conclusion C1 and to let us know the nature of the soul. William’s emphasis is on the incorporeity of the soul, and he also conserves the Avicennian idea that perception of external objects is necessary in order to become aware of one’s own body and its powers. However, if we read William’s version of the thought experiment closely and take heed of the context as well, we find a significant difference from Avicenna. The original flying man is incapable of imagining a hand or any other organ, probably because Avicenna thinks that imagination is a bodily function. Even if he per impossibile were to imagine a hand, he would imagine it without attributing it to himself, seemingly in the same way as we imagine external objects that are not parts of us. The crucial thing is that the flying man would acknowledge his existence but not that he has corporeal qualities or bodily members. He would be indifferent with respect to the imagined hand in the sense that he would apprehend it neither as his part nor as something that does not belong to him, as he would just imagine it without recognizing it as a part of himself. To put this point in systematic terms, there are three possible ways in which the flying man might react to the imagined hand: (1) he apprehends it as a part of himself; (2) he apprehends that it is not a part of himself; and (3) he apprehends it without having any prima facie awareness of its being or not being a part of himself.20

19 William of Auvergne, De anima 2.13, 82b–83a. The translation is from William of Auvergne, The Soul, tr. R. J. Teske (Milwaukee: Marquette University Press, 2000), 91. For discussion on Avicenna’s influence on William of Auvergne, see Roland Teske, “William of Auvergne’s Debt to Avicenna,” in Studies in the Philosophy of William Auvergne (Milwaukee: Marquette University Press, 2006), 232–4. 20 Another way of expressing the difference between these three attitudes is as follows: (1) feeling of ownness: this is my hand; (2) feeling of alienness: this is not my hand; and (3) feeling of indifference: this is a hand.

74

Juhana Toivanen

The reason why I emphasize the difference between (2) and (3) is that William’s wording clearly indicates (2),21 whereas the Latin translation of Avicenna’s flying man suggests (3). In William’s version, the flying man is capable of imagining a body but he thinks that the body does not belong to him. In fact, William goes on to argue that the human soul necessarily knows that it is not a corporeal substance. The soul knows its intellectual acts, it understands that it is the subject of these acts, and it knows evidently that these acts are not corporeal. On the basis of this reasoning, William concludes that every soul which happens to be in a situation similar to the flying man is fully aware of not being a corporeal substance.22 Thus, William’s flying man acknowledges, in addition to his existence, that he is not a body. This difference between Avicenna and William may seem insignificant, but it is related to the argumentative roles that these two authors give to the flying man. Namely, Avicenna’s main point (in De anima 1.1) is to make the reader realize that the soul exists and is not a body: the illustration is meant to display the nature of the soul, not that of a human being. In contrast, William’s aim is to argue that a human being is nothing but the soul, and thus his version of the flying man puts forth a stronger anthropological claim. He thinks that the soul is not a body, but he also argues that the body is not a part of us. William’s position becomes clear if we look at another passage in which he mentions the thought experiment: But when you look into Avicenna’s book, On the Soul, you will clearly read in it that a man, without using any bodily sense, but thinking of himself or of his being, grants that he is only spiritual and that he has no body. Avicenna believes that he explains there that a man thinking of himself in that way finds that the body is not a part of him. Hence, he finds that his whole being or whole essence is his soul.23

William’s Platonic conclusion is that the body is a prison and an instrument of the soul. He equates the human being with the soul, and denies that the body is a true part of a human being. His interpretation is not completely faithful to Avicenna’s text. It may be a correct interpretation, but Avicenna says nothing to the effect that the body would not be a part of a human being—let alone that the flying man would explicitly reject that he has a body.24

21 Obviously the exact wording that William uses may be subject to change once we possess a modern critical edition of his De anima. 22 William of Auvergne, De anima 2.13, 83a–b; The Soul, 92. 23 William of Auvergne, De anima 3.11, 101a; The Soul, 141. 24 To be sure, in De anima 5.7, 162–3, Avicenna argues that the bodily organs are nothing but garments of the soul: “haec autem membra non sunt nisi sicut vestes”

OUP CORRECTED PROOF – FINAL, 9/5/2015, SPi

The Fate of the Flying Man

75

c. Peter of Spain The next flying man comes from Peter of Spain’s commentary on Aristotle’s De anima, which was possibly written in the early 1240s.25 Peter uses the thought experiment in two different but interrelated contexts. In the first instance, he discusses Aristotle’s definition of the soul as the perfection of the body and his proof for the existence of the soul, which proceeds from what is better known to less easily knowable things—in the case of the soul, from acts of the soul to the existence of the soul.26 Following this approach, Peter argues that we have to assume that plants are animated by a vegetative soul and animals by a sensitive soul, because we see that their bodies are alive. Plants live, grow, and take nourishment, and animals perceive external objects in addition to their vegetative functions. The observable actions of plants and animals prove that they differ from lifeless bodies, and the difference is due to their having souls. The argument concerning the intellectual soul is somewhat different, because intellectual operations do not manifest themselves in the body. Peter argues that human beings are capable of understanding essences of things without their matter, and this intellectual ability (among other similar powers) shows that we have an intellectual soul.27 The flying man is taken up as an additional argument that leads us to think that we have intellectual souls. Peter explains the condition of the flying man and claims that the man is aware of himself (the premise P3) but not of his body, because external perceptible objects do not act on his body and bodily senses (P1), and because he does not perceive his body (P2). On this basis, Peter concludes, we must assume that the cognitive subject is something other than the human body—namely, the intellectual soul. Peter’s interpretation proceeds from the alleged fact that the soul is capable of being aware of itself without being aware of its body, but there is a slight shift of emphasis with respect to Dominicus Gundissalinus and William of

(p. 162). However, even in this context his main point is to argue that the unity of experience cannot be due to the body. 25 The identity of the Peter of Spain who wrote this treatise is not known for certain. For an overview of the problems in identifying the authors who share this somewhat popular name, see Joke Spruyt, “Peter of Spain,” in E. N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Winter 2012) . The dating of the work to the 1240s is suggested by R. A. Gauthier, “Les commentaires de la Vetus,” in Thomas Aquinas, Sentencia libri De anima, ed. R. A Gauthier, Sancti Thomae de Aquino Opera omnia iussu Leonis XIII P.M. edita 45.1 (Rome: Commissio Leonina, 1984), 239*. 26 Aristotle, De anima 2.2. 27 Peter of Spain, Questiones De anima 2.4.10, 622.

76

Juhana Toivanen

Auvergne, as Peter’s main conclusion is not C1 understood as an ontological claim about the nature of the soul, but C1 understood as an ontological claim about the mere existence of the soul.28 It is somewhat telling that he does not mention the premise P4—what the flying man is aware of differs from that which he is unaware of—because that is (arguably) needed in order to conclude that the soul is incorporeal. Peter may well think that it is, but his flying man does not show any evidence for making that inference. Another notable aspect in Peter’s version of the thought experiment is the argumentative role that he gives to the flying man. The larger context of his discussion is the argument for the existence of the soul, but the exact question that Peter is dealing with when he takes up the flying man concerns the way in which we come to imagine the existence of the soul. He seems to realize that Avicenna did not consider the thought experiment as a proof but as a pointer, which leads us to acknowledge (or, as Peter says, imagine) the existence of the soul without actually proving it in a demonstrative way. In this respect, he differs from those Latin authors who think that the flying man actually proves either the nature or the existence of the soul.29 In the other context where the flying man appears, Peter argues that the intellectual soul differs substantially from the sensitive and vegetative souls in a human being. The first argument he presents in favor of the difference refers to Avicenna’s thought experiment: If some human being were suddenly created or made, and he would not cognize his body or his hand or his foot, then if someone were to ask him (by a spiritual inquiry) whether he exists, he would answer (by a spiritual reply) and say that “I exist.” And if he were asked: “Are you your hand or your body or your senses or your life?”, he would immediately say that he knows nothing about such things—on the contrary, he would be ignorant of all of these, so that he would be none of them. But the power which would say this in him is the intellectual soul. Therefore, the intellectual soul does not have a natural gaze (aspectus) to the body, to the senses, or to the life

28 Peter of Spain, Questiones De anima 2.4.10, 623. 29 “Alia est ratio Avicenne per quam possumus animam intellectivam ymaginari esse in nobis” (Peter of Spain, Questiones De anima 2.4.10, 622). Peter claims that the arguments that he has presented for the existence of the soul are demonstrations even though they proceed from the effects of the soul to the existence thereof, but they prove only the existence of the soul as a perfection, not as a substance. He does not seem to count the flying man among these demonstrations, however. See Peter of Spain, Questiones De anima 2.4.10, 622–3. Hasse argues that because Peter uses the word ratio, he must consider the thought experiment as a proof (Hasse, Avicenna’s De anima in the Latin West, 91). However, it is possible that the expression ratio Avicenne is used in a loose sense to mean “Avicenna’s account” or “Avicenna’s explanation.”

The Fate of the Flying Man

77

given by the vegetative soul. On the contrary, the intellectual soul is ignorant of all these.30

This passage contains several important ideas. The first of them is that here Peter does not care about the flying man’s inability to perceive external objects. The crucial premise is—in addition to the obvious one that the flying man is aware of his existence (P3)—that he does not apprehend his own body. In other words, Peter does not deduce P2 from P1. Avicenna’s view is (or so I have argued) that the flying man is unaware of his body because he does not have any sensations of external objects. By contrast, Peter appears to think that being aware of one’s body is at least theoretically possible without cognizing external objects by acts that take place in bodily organs. Of course, it is difficult to be certain on the basis of a short passage, but it seems that we are witnessing a slight shift of emphasis at the least. Avicenna thinks that the flying man does not perceive external things and is therefore unaware of his body; Peter supposes that the flying man is unaware of his body tout court. At any rate, Peter goes beyond Avicenna by explaining why the flying man is ignorant of his body (that is, he gives a rationale for P2), and his explanation differs from Avicenna’s view. He first identifies the intellectual soul as the experiential subject within a human being and as a kind of experiential center in the soul. The flying man is ignorant of his body and of the sensitive and vegetative parts of his soul, because his intellectual power is not aware of them: the intellect says that it knows nothing of the body or of the other parts of the soul. Further, the flying man (or his intellectual soul) is unaware of his life, senses, and body, because the intellectual soul lacks an aspectus—which apparently means some kind of attention or intentional directedness—to the body. As a consequence, the flying man is aware only of the existence of his intellectual soul. He knows nothing about his body or bodily functions, and he does not even realize that he is alive, at least in the sense of biological life.31 Peter puts the flying man thought experiment to a radically different use from Avicenna32—or any of the medieval authors discussed here, for that matter—as he thinks that it may be used to establish a substantial difference between the intellectual soul and the other kinds of souls within a single human being.33 In this way, the conclusion that Peter draws from the 30 Peter of Spain, Questiones De anima 2.6.1, 650. 31 We shall see in section 4 that certain elements in this explanation are strikingly similar to Peter Olivi’s psychological views. 32 This has been pointed out by Hasse, Avicenna’s De anima in the Latin West, 90. 33 Richard Dales thinks that the author of the Quaestiones wrote another treatise, Scientia libri de anima. He points out that the author of the former accepts the

78

Juhana Toivanen

thought experiment is highly original. Yet, it is still an ontological conclusion. The flying man experiment is not designed to establish an epistemological idea about human self-awareness, as it only points to the existence of the soul and its relation to the other parts of a human being. From this perspective, Peter does not break with the traditional way of understanding the general meaning and the argumentative role of the thought experiment, even though some of the details are different.

d. John of la Rochelle Finally, John of la Rochelle (d. 1245) brings up his version of the flying man in his influential Summa de anima.34 Unlike Dominicus Gundissalinus and William of Auvergne, who present the thought experiment in order to reveal the incorporeal nature of the soul, John understands it as an ontological proof for the existence of the soul. He asks whether the soul exists and begins his answer by pointing out that we see different kinds of actions in plants, animals, and human beings. These actions are caused by some principle which cannot be the body, and therefore the existence of the soul is a necessity. After this argument John brings up the flying man in order to reach his final conclusion C1: “the being of the soul is different from the being of the body.”35 He seems to think that this conclusion concerns the existence rather than the nature of the soul, and in this respect his version is closer to Peter of Spain than to the other Latin authors.36

substantial unity of the soul for the sake of the argument (see Peter of Spain, Questiones De anima 1.4.3, 263), and that the author of Scientia accepts it without qualification. However, on the basis of the present question, the author of the Quaestiones does not accept the substantial unity (see Peter of Spain, Questiones De anima 1.4.3, esp. p. 656). This doctrinal difference casts further doubts on the identification of these authors. 34 Summa de anima was actually written some years before Peter’s and William’s works, as it has been dated to 1235–6 (Jacques Guy Bougerol, “Introduction,” in Jean de la Rochelle, Summa de anima, p. 12). I present these works in reverse chronological order because John’s argument includes certain elements that lead to the epistemological version of the flying man in the latter half of the century. 35 John of la Rochelle, Summa de anima 1.1, 51. 36 John’s use of the thought experiment is also close to an anonymous author, who writes in his Dubitationes circa animam: “Suppose that a man perfect in knowledge is created in the air in such a state that he can perceive nothing through the senses. It is clear that, in this state, he would say that he is something. But if it were possible for him to imagine a hand or brain or other such bodily things, it is clear that he would not say that he is those things or that they are a part of him. Therefore, since that which is asserted is distinct from that which is not asserted, there is something in him beyond a corporeal nature. And this is Avicenna’s argument.” (Anonymous Vaticanus, Dubitationes circa animam, Vat. lat. 175, f. 219ra.) This text is also mentioned by Hasse, Avicenna’s De anima in the Latin West, 89 n. 50.

The Fate of the Flying Man

79

The wording of John’s version is very close to Gundissalinus’ translation of Avicenna, and the structure of the argument is roughly the same. John accepts the premises P1 and P2, although it is not clear that he infers the latter from the former. He also accepts P3 and P4, and uses the latter to establish that the soul and body are different entities. However, he makes a small addition to P3. It may be nothing but an innocent paraphrase, but it points toward the epistemological version of the flying man that appears in the latter half of the thirteenth century. The crucial passage is as follows: “It is certain that a man thus created, cognizing himself, would not hesitate to affirm that he exists. However, he would not affirm anything external about his members or anything hidden about what is inside him, such as the brain and the like.”37 The other versions of the thought experiment state simply that the flying man is aware of his existence (Avicenna, Gundissalinus, and Peter of Spain) or they base the self-awareness on the soul’s ability to apprehend its cognitive acts (William of Auvergne). By contrast, it seems as if John was saying that the flying man admits his existence (or the existence of his soul) because he cognizes himself directly without having any other cognitive acts. Obviously self-awareness is a central premise in the other versions of the thought experiment as well, but John of la Rochelle emphasizes it in a new way. Interestingly, he does not mention anything that could look like the conclusion C2 at the end of the passage in which the flying man is presented. Direct self-awareness functions only as a premise P3, which is used to reach the ontological conclusion C1. By leaving C2 unstated, John, in effect, enables one to see more clearly that the flying man thought experiment does not prove that the soul is aware of itself without bodily acts (or without awareness of the body), but presupposes it. In itself, the addition to the premise P3 may not be particularly significant, but if we take into consideration the other arguments that John provides for the existence of the soul, the situation changes radically. Namely, right after experimenting with the flying man, John goes on to quote two passages from pseudo-Augustine’s De spiritu et anima. The central idea in both of these quotations is that the soul is immediately present to itself and thus knows itself. It is clear that John thinks that these passages confirm Avicenna’s argument. In this way, he highlights the epistemological aspect of the thought experiment—at least his argument is easy to read in such a way. Given the strong influence that he had on the subsequent generations of Franciscan intellectuals, it is likely that the connection which he establishes between the flying man and the (pseudo-) 37 John of la Rochelle, Summa de anima 1.1, 51; emphasis mine.

80

Juhana Toivanen

Augustinian conception of the soul’s direct awareness of itself was taken over by later Franciscans. Nevertheless, the flying man remains an ontological argument for John.

3 . E PI ST E M O L OG I C A L F L Y I N G M A N When we turn to the latter half of the thirteenth century, we encounter a manifest change in the argumentative role that is given to the flying man. Matthew of Aquasparta (c. 1240–1302) and Vital du Four (1260–1327) use the thought experiment, but they do not understand it as an ontological proof. Rather, they use it in an epistemological context and consider it as an argument for the soul’s ability to be directly aware of itself without having any sensory experiences. Although self-awareness had been a part of the thought experiment from the beginning, as a necessary premise P3 for reaching the ontological conclusion C1, the earlier Latin authors did not think that the central meaning of the flying man would have been to prove the soul’s ability to be aware of itself. Self-awareness was mainly considered to be a premise—this is especially clear in the case of John of la Rochelle, but applies to the other authors as well. By contrast, in the latter half of the thirteenth century, self-awareness was taken to be the only aspect of the thought experiment that was worth dealing with. The difference from the earlier usage is so radical that it is appropriate to call this version of the thought experiment ‘the epistemological flying man,’ in contrast to the earlier ‘ontological flying man.’ One of the most important reasons for this change is external to the thought experiment itself. The latter half of the thirteenth century witnessed a philosophical debate over human self-awareness, and the main dividing line in this debate was the soul’s ability or inability to be directly aware of itself without having any preceding cognitive acts. To simplify the issue slightly, Thomas Aquinas defended an Aristotelian view, according to which self-awareness is indirect. We become aware of our souls only by first apprehending something else (say, an external object), which makes us also aware of the act of cognition. After becoming aware of the act we come to know the subject to which this act belongs. This subject is the soul.38 It is not surprising that the authors who follow this line of reasoning do not 38 See, e.g., Thomas Aquinas, Summa theologiae I.87.1; Quaestiones disputatae de veritate, q. 10.8. Aquinas’ view has been analyzed in modern scholarly literature. One may begin with Therese Scarpelli Cory’s recent book, Aquinas on Human Self-Knowledge (Cambridge: Cambridge University Press, 2014); and Brown, “Aquinas’ Missing Flying Man,” 17–31.

The Fate of the Flying Man

81

appeal to the flying man, because they reject the central premise P3 and with it the whole experiment loses its appeal. By contrast, Aquinas’ opponents (mainly from the Franciscan order) were willing to accept that the soul is capable of being directly aware of itself without being first aware of something else.39 Some of them continued to use the thought experiment, but because the discussion centered on the possibility of direct selfawareness, the premise P3 became the whole point of discussion. At the risk of oversimplifying the debate, we may say that the central question turned out to be whether P3 follows, given that P1 is granted. Aquinas and his followers answered in the negative, whereas their opponents accepted Avicenna’s positive answer. In this way, the new philosophical debate concerning human self-awareness subsumed Avicenna’s thought experiment, and this development contributed to the neglect of the other aspects (indeed, the main point) thereof.

a. Matthew of Aquasparta The first author to actually establish a connection between the philosophical debate over human self-awareness and the flying man thought experiment is Matthew of Aquasparta. He employs the thought experiment as one proof of the soul’s ability to cognize itself directly by its essence: But the soul would understand itself even if it did not use any of the senses or have any acts of the senses; therefore, it does not understand itself through the mediation of phantasms. The minor premise is clear on the basis of Avicenna’s De anima [ . . . ] where he says that if a human being were created as suspended in air—in such a way that his members are wholly separated and do not touch each other in any way, and he does not perceive anything at all—he would not hesitate to affirm that he exists. After this Avicenna says that [the flying man] does not need the body in order to know the soul and perceive it.40

Unlike the earlier authors, Matthew concentrates on the very end of the original thought experiment: the body and bodily acts are not necessary for the soul to be aware of itself. In other words, Matthew thinks that the crucial element in the thought experiment is the premise P3 and what he takes to be its epistemological conclusion C2. He does not draw the ontological conclusion concerning the nature of the soul or the human being and, like John of la Rochelle, he connects the flying man with

39 For an overview of the medieval discussion, see Franc¸ois-Xavier Putallaz, La connaissance de soi au XIIIe sie`cle: de Matthieu d’Aquasparta a` Thierry de Freiberg (Paris: Vrin, 1991). 40 Matthew of Aquasparta, Quaestiones de cognitione, q. 5, 295.

82

Juhana Toivanen

Augustinian texts—this time genuine—which he reads as proof for the soul’s direct self-awareness. However, Matthew does not accept the argument: Avicenna wanted that understanding (ratio intelligendi) would be completely from above—namely, from the agent intellect, which is the last in the order of intelligences—and therefore the soul would be able to understand itself without any act of, or service from, the senses; but he erred in this.41

Matthew thinks that the plausibility of Avicenna’s thought experiment depends on the rejection of Aristotelian empiricism. If all knowledge is based on sense experience, the intellectual soul cannot know itself without first perceiving something else. Thus, when Avicenna claims that the flying man is aware of his own existence, he practically suggests that there is one special kind of knowledge that is not empirical. Matthew rejects Avicenna’s view, because he thinks that the soul is initially incapable of having intuitive knowledge of itself. In this regard, he approaches Aquinas’ position, according to which self-cognition proceeds necessarily from the firstorder acts of the soul, which pertain to external objects. However, Matthew also criticizes the Thomistic view. He argues that only the very first act of cognition must pertain to external perceptible objects. After the intellectual part of the soul has been actualized by an act of understanding which is abstracted from the objects perceived by the senses, it is capable of directly apprehending itself: the soul does not need perceptual acts every time in order to have intuitive cognition of itself.42 Thus, the flying man would not be aware of himself because he never had any cognitive act that could bring him to realize himself. By contrast, human beings in normal conditions can be directly aware of themselves, because they have had some sensory input. Matthew’s position with respect to the flying man’s awareness of his body (or the lack thereof ) is not completely clear, because he does not mention the premise P2 at all. He emphasizes instead that the flying man does not have any perceptual acts (the premise P1). However, we may assume that if the flying man does not perceive anything, he does not perceive his body either, and therefore it seems that Matthew accepts the idea that the flying man would not apprehend his own body. He does not say explicitly that this is the case, probably because the awareness of the body is not relevant for his purposes. The flying man functions as an argument for the soul’s direct self-awareness and, in this epistemological 41 Ibid., 312. 42 Ibid., 304. For discussion, see Putallaz, La connaissance de soi au XIIIe sie`cle, 42–50; Helen Marie Beha, “Matthew of Aquasparta’s Cognition Theory,” Franciscan Studies 21 (1961), 423–8.

The Fate of the Flying Man

83

context, the awareness of the body does not play the same central role as it does in the ontological version of the thought experiment. It is nevertheless significant that Matthew is silent about this premise, which is crucial for the ontological flying man.

b. Vital du Four Vital du Four’s version of the flying man is in many ways similar to the one we saw with Matthew. The most obvious similarity is that he understands the flying man primarily as an epistemological thought experiment, which can be used as an argument for the soul’s direct self-awareness—that is, he considers the premise P3 as the most important aspect of Avicenna’s experiment. Vital begins his analysis of self-awareness by expounding Aquinas’ view, according to which the soul cannot cognize itself directly but proceeds from the awareness of its acts to the awareness of itself.43 Vital criticizes this view because it does not recognize the possibility of a direct intuitive cognition that the soul has of itself, and the flying man is one of the arguments that he brings forth to show that Aquinas cannot be right: Again, if the soul were in a body, which it could not use at all, it would not apprehend sensible species. [ . . . ] However, it is certain that such a soul in such a body would apprehend itself, as becomes clear from Avicenna’s Sextus naturalium 1.1, where he says that if a human being were created and placed per impossibile in air in such a way that his members would not touch each other but they all would be separate and he would not perceive anything about his members or about anything else, he would not doubt his existence but he would affirm it.44

Vital’s description of the situation of the flying man does not differ in any relevant way from his predecessors, as he accepts the premises P1, P2, and P3. Yet, he does not mention the premise P4 at all, probably because it is not relevant for his epistemological purposes. Moreover, the way he uses the illustration raises some questions. Unlike Matthew, who presents the thought experiment as a positive argument for a view that he does not fully accept and which he explicitly criticizes, Vital brings it forth as an argument against Aquinas and seems to accept it. Yet, it is difficult to see how the flying man could work in the context of Vital’s own theory. Namely, he argues in the same way as Matthew that we must make a distinction between the beginning of self-awareness and its completion. Self-awareness begins with the acts of the soul, but afterwards the soul can cognize itself 43 Vital does not attribute the view to Aquinas, but there is little doubt that the theory he analyzes comes from him. 44 Vital du Four, Quaestiones disputatae de cognitione, q. 4, 242.

84

Juhana Toivanen

directly.45 The problem is that if direct self-awareness is, at first, impossible without any acts of the soul, not even the flying man can be aware of himself unless he acts—and by definition he does not have any acts which could function as the basis for self-cognition. It seems that there are two possible explanations: either Vital thinks that the flying man is not a good thought experiment (like Matthew did before him) but uses it against Aquinas nevertheless; or he thinks that the flying man has acts, albeit internal ones, which do not require the use of the body. In the first case, the flying man would be nothing but an ad hoc argument, and in the latter case, Vital would be putting forth a radical interpretation of Avicenna’s view—an interpretation which resembles, in some respects, William of Auvergne’s position. Unfortunately, he does not say what his final view is, but the way the text continues immediately after the thought experiment may be taken as a modest support for the latter reading: [Although the flying man] would not perceive anything about his members or about anything else, he would not doubt his existence but he would affirm it. Thus, it is clear that the soul in the body apprehends that it exists (se esse) without receiving any species. And because after apprehending that it exists (quia est 46), the soul asks what it is, it does not necessarily need the species of any external object in order to apprehend what it is. It notices itself (devenit in sui notitiam)—both that it is and what it is—through its internal acts such as living, willing, desiring and the like, which it experiences in itself. It can do the same through the acts that it exercises in relation to external objects, but these are not necessary for this, as the [aforementioned] view stated.47

Thus, Vital argues that the soul knows that it exists (and even what it is) without any acts that pertain to external objects, but self-cognition begins from the acts of the soul nevertheless. It seems that he understands Avicenna’s thought experiment in this light: the flying man is aware of himself because he is aware of these internal acts. Vital seems to think that not only living but also willing and desiring are acts that are not necessarily related to external objects. In the case of living, this seems uncontroversial; the other examples are more problematic.48

45 Ibid., 243–8. 46 Vital uses the expressions quia est and quid est to distinguish two aspects of selfcognition. ‘Quia est’ pertains to the existence of the soul or the intellect, and ‘quid est’ to the knowledge of its essence. 47 Vital du Four, Quaestiones disputatae de cognitione, q. 4, 242. 48 It is worth noting that desiring and willing are not cognitive acts. Medieval philosophers typically thought that these acts are not related to external objects directly but through, or in relation to, cognitive acts of the soul. Thus, in a sense, one might be able to claim that desiring and willing do not require external objects; but then again,

The Fate of the Flying Man

85

Vital is not interested in the other aspects of the thought experiment. He mentions that according to Avicenna the flying man is unaware of his body (the premise P2), but he does not draw any ontological conclusions from this premise and there are reasons to believe that he does not accept it. Namely, he explicitly claims further in the text that the soul is aware of being in the body: “Therefore, there is another view, which says that the conjoined soul cognizes [1] that it exists (quia est), [2] that it exists in a certain kind of body, [3] what it is, and [4] it also regards itself and its dispositions intuitively by an interior sense—that is, by an intellectual eye.”49 Vital accepts this view, as he also writes that “a united soul cognizes by its internal or external acts, and without a species, that it exists and that it is in the body. [ . . . ] And because the soul comes to cognize what it is by cognizing what its acts are, it does not know as certainly what it is, as it knows that it exists or exists in its body.”50 On the basis of these passages, it seems clear that the soul is aware of being conjoined to the body. This addition is quite interesting, although it does not play a significant role in Vital’s discussion. He does not give any argumentative role to it and he does not relate it to the flying man. He just mentions this ability without discussing it at all. The addition may be partly explained by the nature of Vital’s work. Namely, the titles of his disputed questions reveal that they belong to a genre of philosophical literature that emerged in the latter half of the thirteenth century. In this genre, human cognition is discussed in two sets of questions, one dealing with the activities of a soul that is separated from the body, and the other with a soul when it is conjoined to the body. Vital’s questions belong clearly to the latter set, and it is possible that when he mentions—right after performing the flying man thought experiment— that the soul is in the body, he only wants to issue a reminder that he is talking about self-awareness as it happens in a conjoined soul. However, the other quoted passages cannot be completely explained away by appealing to the genre of the work. It seems, therefore, that Vital really thinks that the soul is aware of being in the body, at least in the case of men who do not fly. This idea may explain partially why he does not draw the ontological conclusion C1 from the flying man: if the premise P2 is rejected, the conclusion C1 does not follow. To be sure, the omission of the ontological conclusion is reasonable already on the basis of the context. The existence and nature of the soul are not relevant for Vital’s purposes when he discusses the soul’s awareness of itself. Yet, the rejection of P2 bears philosophical most medieval authors thought that it is not possible to desire or to will without having a cognitive act that provides the object for desiring and willing. 49 Vital du Four, Quaestiones disputatae de cognitione, q. 4, 243. 50 Ibid., 247.

86

Juhana Toivanen

significance, as it shows how one may be able to reject the ontological conclusion that Avicenna draws even if one accepts the crucial premise P3. Matthew and Vital reveal that the function of the thought experiment changes in the latter half of the thirteenth century. They are not particularly interested in the flying man for its own sake, and unlike the earlier authors, they do not give it an important position in their discussions. The illustration is taken up only as one possible argument for the soul’s ability to have some kind of direct and intuitive cognition of itself. One consequence of this ‘epistemological turn’ is that the premise P2, which was central for Avicenna and the earlier Latin authors, namely, that the flying man does not apprehend his body, loses its importance. If the thought experiment is meant to show only that the soul can be aware of itself directly, without having previous cognitive acts that pertain to external objects (the premises P1 and P3), and it is not used to reach the ontological conclusion C1, the lack of bodily self-awareness does not matter anymore.

4 . M A N B E F O R E T H E C R E AT I ON The original flying man thought experiment is obviously based on Avicenna’s Neoplatonic substance dualism, but it also reveals, as a corollary, his position with respect to the role that the body and the sense organs play in perception. Direct awareness of one’s own body is impossible. We become aware of our bodies only by perceiving them through the external senses— we may touch or see our own body just like we touch or see other external objects—or by using the sensory powers that are actualized in bodily organs. In this model the senses function as transmitters of external information, and we apprehend them only when we have sensations through them. If there is no information coming from the external world to the senses, they remain transparent to us. Although this move from P1 to P2 loses its importance as the latter premise becomes less relevant in the epistemological flying man, it is crucial for the ontological version of the thought experiment. By contrast, the latter half of the thirteenth century witnessed a thought experiment which makes sense only if the opposite of the premise P2 is supposed to be true. This thought experiment is Peter Olivi’s ‘man before the creation,’ which he presents while developing his theory of perception.51 He illustrates how the soul’s cognitive attention functions by asking

51 I have briefly discussed this thought experiment in Juhana Toivanen, “Perceptual Self-Awareness in Seneca, Augustine, and Peter Olivi,” Journal of the History of Philosophy 51 (2013), 374–5.

The Fate of the Flying Man

87

us to imagine a man who exists before anything else is created in the world. This man tries to see his surroundings by directing his attention, but he sees nothing simply because there is nothing to be seen. The context in which Olivi presents his thought experiment and the argumentative role he gives to it differ from those of Avicenna and the other Latin authors. He does not relate it directly to the nature of the soul or selfawareness, as its purpose is to explain how intentional directedness of the soul takes place in the perceptual process.52 In particular, Olivi tackles a problem concerning the effect of mirrors, which is especially challenging for him, not only because he rejects the efficient causality between the power of vision and the object, but also because he denies that there is any real connection between them. When someone sees an object, her sight is intentionally directed to it—to use Olivi’s expression, the power has an aspectus which reaches the object virtually—and this intentional relation enables the act of seeing.53 It is fairly easy to understand the strengths of this theory in the case of a direct perception of an object which is in front of the perceiver. However, when perception takes place through a mirror, Olivi’s view seems intuitively less plausible, especially as he thinks that we do not see an image of the object in the mirror but the object itself.54 Olivi admits that the case of mirrors is strange—strange enough to explain why a mirror is called “a miracle” (miraculum) and looking into a mirror “to marvel” (mirari) in his native Occitan language.55 Thus, in order to defend his theory, he needs to explain how the reflection of the intention of the soul takes place. His explanation is rather complicated and we need not go into the details in this connection. The basic idea is that the aspectus of sight that meets the mirror causes the reflected aspectus that reaches the object. In order to make this

52 However, Olivi’s conception of intentional directedness of the soul is central for his theory of self-awareness, and therefore the idea that he defends with his man before the creation is related to the epistemological version of the flying man. 53 My intention here is not to explain Olivi’s theory of perception, which has been discussed adequately elsewhere. See Robert Pasnau, Theories of Cognition in the Later Middle Ages (Cambridge: Cambridge University Press, 1997), 121–81; Dominik Perler, The´ories de l’intentionnalite´ au Moyen Aˆge (Paris: Vrin, 2003), 43–75; Jose´ Filipe Silva and Juhana Toivanen, “The Active Nature of the Soul in Sense Perception: Robert Kilwardby and Peter Olivi,” Vivarium 48 (2010), 245–78; Katherine Tachau, Vision and Certitude in the Age of Ockham: Optics, Epistemology and the Foundations of Semantics, 1250–1345 (Leiden: Brill, 1999), 39–54; Juhana Toivanen, Perception and the Internal Senses: Peter of John Olivi on the Cognitive Functions of the Sensitive Soul, Investigating Medieval Philosophy 5 (Leiden: Brill, 2013), 115–222. 54 Peter Olivi, Summa II, q. 73, 69–70. 55 Peter Olivi, Summa II, q. 73, 67; B. Jansen, “Descriptio codicum—Ratio editionis,” in Summa II, vol. 3, xix–xx.

88

Juhana Toivanen

solution more plausible, Olivi takes up another case in which one aspectus is the cause of another, or in which one aspectus gives rise to another when external conditions change: he asks us to imagine the man before the creation. It should also be said that the powers [of the soul] have a double aspectus. The first of them is indeterminate with respect to the objects—as when we tend outward in such a way that the power of vision is applied to seeing by the will or by nature, and the eye is awake but kept closed, or we are in darkness. Because of the unsuitable condition of the medium, or because of the impediment of some obstacle, the power of sight does not determinately tend toward any object. The other aspectus is a determination of the first aspectus because the first one is related to the second as a root to a branch, and a sensation is caused from the first when an object is present. For example, given that only a man whose eyes are open would have been created before the creation of everything else, and he would strive with all effort to tend to see by his eyes (toto conatu niteretur per oculos intendere ad videndum) as if there were external visible things: it is clear that in that case his aspectus would not be terminated at or determinately carried to any external object. If, after a while, all the external things (which exist now) were created, the first aspectus of the eyes would be thereby fixed at external objects.56

Like the flying man, the man before the creation does not have any sensations, because there are no objects that he could perceive. The first major difference from the flying man, however, appears immediately after this initial similarity in the setting of the thought experiment. Namely, Olivi does not derive the premise P1 from the fact that external objects do not act upon the senses. Avicenna supposes that if external objects do not cause changes in the sense organs and sense powers, the senses remain inactive, but Olivi thinks that the senses may be active even when they do not sense anything, as in the exceptional case of the man before the creation—and we may suppose that he would claim that the senses of the flying man would be similarly active. Olivi defends an active theory of perception in which the perceptual process is based on the soul’s intentional directedness. The man before the creation could not be prevented from using his senses, because if he were unable to use them when there are no objects present, he would never perceive anything. The sudden creation of external objects cannot cause perception, according to Olivi, because that would make perception a passive process. Thus, the difference between the man before the creation and the flying man stems partly from Olivi’s active theory of perception. At the same time, it reveals that Olivi’s attitude toward awareness of one’s own body is

56 Peter Olivi, Summa II, q. 73, 68–9. See also ibid., q. 59, 543–4.

The Fate of the Flying Man

89

radically different from Avicenna’s position. In contrast to the flying man, who is utterly unaware of his body and bodily functions, Olivi’s man before the creation is well aware of his ability to see even though there is nothing to be seen. The man is not only created with his eyes open but he toto conatu niteretur per oculos intendere ad videndum—he directs his attention to the external world and tries to see if there is anything, and he does this by using his eyes intentionally. One might think that the man before the creation is aware only of his ability to see and not of his eyes as bodily organs. However, it seems to me that when Olivi writes per oculos, he really means that the man before the creation is aware not only of his power of sight but also of his eyes. He claims elsewhere that one must pay attention to the senses in order to see (or try to see), and he seems to think that this requires paying attention to the organs in which the powers are seated.57 Thus, in order to use his power of sight, the man before the creation must be aware of his eyes on some level. To be sure, this kind of awareness does not need to be reflexive in the sense that the eyes would be objects of cognition. Rather, it is sufficient that the eyes are cognized as something that can be used for seeing. The man does not need to be aware of the eyes in a similar way as he is aware of other visual objects that he perceives after they have been created—he could perceive his eyes in this way only by looking at them through a mirror—but as instruments that he can use for seeing at will. Without this kind of awareness he could not turn his attention to his eyes and use them in an appropriate way. It is important to note that Avicenna and his Latin followers also seem to commit to the idea that being aware of one’s sensory powers entails awareness of the body. Avicenna’s move from P1 to P2 is based on the idea that the flying man is unaware of his sense powers because they are not acting, which in turn entails a lack of awareness of the body. This reasoning involves a transition from awareness of a bodily function to awareness of the body. To be sure, we may question the validity of this move, but in a way it makes sense: it is difficult to understand how one could be aware of the ability to see without being aware that this ability is located in the eyes, that it can be altered by turning the eyes in different directions, and that it gives a perspective on the external world. This experience is especially clear in the case of touch. When someone touches an object, she experiences the 57 See, e.g., Summa II, q. 58, 510, where Olivi states that “the organ of the common sense is the whole brain, just as the organ of sight is the whole eye with the visual nerves. [ . . . ] Different acts are brought about according to the various aspectus that the common sense has in the brain. For according to the aspectus that it has towards the eyes it apprehends visible things, and according to the aspectus towards the ears it apprehends audible things, and so on for the other senses.”

90

Juhana Toivanen

touching in a certain part of her body. It makes sense to claim that if she is aware of the location of perceptual acts within her body, she must have some sort of awareness of her body and of the bodily functions as located in it.58 My interpretation of Olivi’s thought experiment may sound daring, given that it is based on very scant textual evidence, but, in fact, it is compatible with what Olivi says in other contexts. He thinks that the human body is the immediate object of the sense of touch, and some of the formulations that he uses when he discusses the sense of touch come very close to the modern concept of proprioception, the sense by which we perceive our own bodies.59 Moreover, he thinks that not only human beings but also other animals are immediately and pre-reflexively aware of the functions of their bodily parts and organs: When a dog or a snake sacrifices one of its members in order to save its head or sacrifices some part in order to save the whole, then it prefers the whole over the part and the head over the other member. Therefore, these animals must have some common power which shows both extremes simultaneously, their mutual comparison, and the preference of one over the other—although it does not do this with the same fullness and degree of reflective judgment as does the intellect.60

Even non-human animals are aware of their bodies, of the functions and mutual values of the various parts. A dog may block a strike with its paw in order to save its head because it is aware that its head is more important for its well-being than the paw. I have argued elsewhere that this kind of immediate bodily self-awareness is a necessary condition for the ability to use the body and the senses appropriately, and it is not learnt but innate.61 From this point of view, it is significant that Olivi describes the man before the creation immediately and intentionally using his bodily senses. Even though his purpose is not to use the thought experiment as an argument concerning the nature of the soul or its relation to the body, it is based on the idea that the man before the creation actually uses his eyes and therefore is aware of them and their function. A central presupposition 58 Olivi points out that we are aware that our acts of seeing are located in the eyes in Summa II, q. 73, 94. 59 Mikko Yrjo¨nsuuri, “Perceiving One’s Own Body,” in S. Knuuttila and P. Ka¨rkka¨inen (eds), Theories of Perception in Medieval and Early Modern Philosophy, 101–16. 60 Peter Olivi, Summa II, q. 62, 588. 61 Toivanen, “Perceptual Self-Awareness,” 372–9; Toivanen, Perception and the Internal Senses, 281–92. See also Mikko Yrjo¨nsuuri, “Types of Self-Awareness in Medieval Thought,” in V. Hirvonen, T. J. Holopainen, and M. Tuominen (eds), Mind and Modality: Studies in the History of Philosophy in Honour of Simo Knuuttila (Leiden: Brill, 2006), 153–69.

The Fate of the Flying Man

91

in the thought experiment is that the man is directly aware of his sensory powers, which entails a certain kind of awareness of the bodily organs as well.62 We cannot question this premise without completely undermining the plausibility and effectiveness of the thought experiment; and that is exactly why Olivi’s stance toward the soul’s awareness of the bodily senses can be inferred from it. We can now see in detail the central differences between the flying man and the man before the creation. First, Olivi thinks that external senses are active even when they do not perceive anything, and thus he rejects Avicenna’s premise P1. On the other hand, if we read P1 as referring to perceptual acts (and not to any kind of activity of the senses), Olivi rejects the inference from P1 to P2: one can be aware of the senses (and have a certain kind of awareness of their organs as well) even when one does not perceive anything through them. In other words, Olivi makes a conceptual difference between P1 and P2—something which Avicenna does not seem to acknowledge. It is possible that Olivi would think that P2 requires P1 if the latter is understood as any kind of activity—that is, if the man before the creation was, for some reason, unable to even try to see and did not have active visual attention toward the void, then he would not be aware of his senses and body. Olivi does not develop the scene enough for us to know his position on this imaginary case, but it is important to note that he considers the kind of activity he attributes to the man before the creation as an inseparable feature of the soul and its cognitive powers. Thus, although the man before the creation has, at first glance, much in common with the flying man, there is a fundamental difference between these two experiments. The difference is not so much ontological—both Avicenna and Olivi can be understood as defending a qualified substance dualism63—as phenomenological. It reveals a different understanding of the contents of elementary self-awareness that human beings have: to put it

62 Olivi thinks that animals must be able to apprehend their sensory powers and the state of bodily organs (for instance, that the eyelids are closed) even when they do not perceive anything through them. See Peter Olivi, Summa II, q. 62, 587–9; Toivanen, Perception and the Internal Senses, 275–81. 63 See, e.g., Jon McGinnis, Great Medieval Thinkers: Avicenna (Oxford: Oxford University Press, 2010), 90–5; Toivanen, Perception and the Internal Senses, 25–42. A further similarity is that Olivi accepts the Avicennian idea that the soul is aware of itself directly, without the mediation of the acts of the soul. However, this topic cannot be dealt with in detail here. See Peter Olivi, Summa II, q. 76, 145–9; Peter Olivi, “Impugnatio quorundam articulorum Arnaldi Galliardi, articulus 19,” ed. S. Piron, in C. Ko¨nig-Pralong, O. Ribordy, and T. Suarez-Nani (eds), Pierre de Jean Olivi—Philosophe et the´ologien, Scrinium Friburgense 29 (Berlin: De Gruyter, 2010), 457; Putallaz, La connaissance de soi, 85–133.

92

Juhana Toivanen

simply, the flying man is completely unaware of his body and sensory powers, but the man before the creation is not. It is impossible to say with certainty whether Olivi was inspired by Avicenna’s flying man when he came up with his thought experiment. It is clear that he had a good knowledge of Avicenna’s psychology, and occasionally he refers to Avicenna’s De anima directly.64 The situation is similar with respect to his familiarity with other Latin authors who use the flying man: we do not have sufficient knowledge of Olivi’s use of earlier and contemporary sources, and further research is needed in order to establish historical connections to the flying man tradition. It is nevertheless likely that Olivi was familiar with the works of William of Auvergne and John of la Rochelle, given their importance in thirteenth-century philosophy and especially among Franciscan philosophers. Moreover, it is almost beyond doubt that Olivi knew Matthew’s work. After all, Matthew was Olivi’s contemporary, who helped to rehabilitate Olivi and appointed him a lector at the highly esteemed studium in Florence, when Matthew was acting as the minister general of the Franciscan order.65 Finally, Olivi knew at least some of Vital’s works—but question 73 of Summa II was written in the 1280s and therefore Vital cannot have influenced it.66 What is certain is that Olivi does not indicate any source for his idea in the immediate context of the man before the creation. It is possible that he found a similar argument from some earlier Latin author (Augustine would be the usual suspect here) or from his own contemporaries, but we cannot rule out the option that he invented the thought experiment by himself. At any rate, it is safe to say that at least his way of using the experiment was original, because it makes sense only in the context of his theory of perception. Leaving the historical connections aside, there is yet another doctrinal affinity that is worth mentioning. Peter of Spain’s thought experiment bears a striking resemblance to two central aspects of Olivi’s philosophical psychology, and this suggests a more general similarity between the psychological theories of these two authors. First, as we have seen, Peter of Spain thinks that the intellectual soul functions as an experiential center of the soul. The flying man is unaware of his body because the intellect does 64 E.g. in Summa II, q. 59, 524. 65 David Burr, “The Persecution of Peter Olivi,” Transactions of the American Philosophical Society 66 (1976), 67. 66 For an account of a dispute between Olivi and Vital du Four, see Vincenzo Mauro, “La disputa de anima tra Vitale du Four e Pietro di Giovanni Olivi,” Studi medievali 38 (1997), 89–138. For dating of Olivi’s questions, see Sylvain Piron, Parcours d’un intellectuel franciscain. D’une the´ologie vers une pense´e sociale: l’oeuvre de Pierre de Jean Olivi (ca. 1248–1298) et son traite´ De contractibus (Ph.D. dissertation, E´cole des Hautes E´tudes en Sciences Sociales, Paris, 1999).

The Fate of the Flying Man

93

not know anything about it. Olivi understands the role of the intellectual power of the soul in a similar way. When he discusses the experiential unity of the soul, he argues that we experience different kinds of acts as our own because the intellect apprehends these acts as belonging to us.67 The other similarity is related to the ontological explanation for the ability to cognize various things either within oneself or in the external world. Peter of Spain argues that the flying man does not perceive his body because the intellectual soul does not have a natural aspectus to the body. Olivi’s theory of cognition is based on the same principle, namely that the cognitive processes require intentional directedness from the part of the soul.68 The only difference is that, in Olivi’s case, the soul is capable of paying attention to itself and to the body even in the counterfactual case of the man before the creation. These affinities may be significant, but they alone do not prove that Olivi would have known Peter of Spain’s work. The doctrinal similarities or historical connections between these authors have not been studied in modern scholarship, and before we have a better knowledge of Peter of Spain’s philosophical psychology and Olivi’s familiarity with his work, we cannot make any firm conclusions in this respect. Is the man before the creation a version of the flying man or an independent thought experiment? The answer to this question depends on the criteria that we use in judging whether two ideas are the same, and therefore the answer will of necessity be controversial. Olivi’s thought experiment resembles the flying man so much that it is natural to think that there are connections to the earlier tradition. On the other hand, the context in which Olivi presents his thought experiment and the argumentative role he gives to it differ from Avicenna and the Latin authors. The experiments have different functions, and if the emphasis is laid on the function and the context, it is clear that we are dealing with two different experiments. But then again, the epistemological and ontological flying men serve a different purpose and come up in different contexts as well. One might argue that the question concerning the relation of the man before the creation to the flying man is not very important. By analyzing the man before the creation in connection with the various versions of the flying man, we may arrive at a general picture of medieval variations of a certain kind of thought experiment. Whether or not these experiments are considered to be the same, it is clear that a man who does not perceive 67 Peter Olivi, Summa II, q. 51, 112 and q. 54, 280. 68 Olivi even uses the same concept, aspectus, as we have seen. But then again, this terminological affinity is easy to explain without supposing a historical connection between Peter of Spain and Olivi by appealing to a common source. The concept was not uncommon in medieval philosophical vocabulary.

94

Juhana Toivanen

anything external to him was a part of philosophical imagery in the thirteenth century. It is also clear that medieval philosophers came up with original insights by means of this man.

5 . C O N C LU S I O N The Latin discussion concerning the flying man endured roughly one and a half centuries, from the middle of the twelfth to the beginning of the fourteenth century. During this time it was transformed from an ontological proof to an epistemological remark. Up to the middle of the thirteenth century, Latin authors used it to prove the existence and/or incorporeal nature of the soul. In the latter half of the century it was transferred to an epistemological context and used as an argument for the soul’s ability to be directly aware of itself. The central difference between these two versions of the thought experiment can be presented by using the schematic analysis of the flying man experiment, which I have presented above. The ontological version begins from the inactivity of the senses (the premise P1), which entails a lack of awareness of the body (P2). The flying man does not perceive anything and is therefore unaware of his body. The experiment continues with a supposition that the flying man would be aware of himself or of his soul even in this state (P3). The conclusion that is claimed to follow from these premises is that the soul exists and is incorporeal (C1). When medieval philosophers turn their minds to an epistemological question concerning the soul’s ability to be aware of itself directly, the argumentative role of the thought experiment changes. It turns into an epistemological argument, which is supposed to prove that self-awareness is not based on cognitive acts that pertain to external objects. Even in the absence of acts of perception (P1), the flying man would be aware of himself or of his soul (P3). Late thirteenth-century philosophers presented this argument against Aquinas’ theory of self-cognition, and they thought that its primary function was to show that self-awareness is not grounded on previous cognitive activity of the soul but must be a primitive feature of the human cognitive system. In this discussion the original conclusion C1 is left aside and the emphasis is solely on the premises P1 and P3. One may ask why the authors who accepted the epistemological premise P3 did not continue to make the further ontological conclusion C1. The reason for this omission can be partially explained by appealing to their philosophical interests and the context in which they used the flying man: the ontological conclusion is irrelevant or at least peripheral to the discussion

The Fate of the Flying Man

95

concerning self-awareness. However, this answer begs the question because we would still want to know why they did not use it also in an ontological context, when they were discussing the nature of the soul. One possible explanation is that the rise of Aristotelian natural philosophy and Aristotle’s aversion to substance dualism made Avicenna’s dualistic model seem suspicious and outdated. Even those who considered the flying man as an effective instrument for defending the soul’s direct self-awareness tended to soften the dualistic elements in their anthropology, at least to the extent that it was possible without threatening their commitment to the immateriality and immortality of the soul. By the end of the thirteenth century, the reaction to the premise P2— that one would be unaware of one’s body without cognitive acts that pertain to external objects—was reconsidered. Vital du Four seems to challenge it, and when Olivi presents his own thought experiment that resembles the flying man, he presupposes that even in this peculiar condition we would be aware of our bodily senses, at least as powers which we may actively use and which are located in the organs of the body. According to Olivi, our senses do not have to be actualized by an act of perception before we know how to use them. His thought experiment featuring the man before the creation does not make sense without this presupposition. The difference between Avicenna and Olivi stems partly from their differing descriptions of the initial conditions of the man in the void—Olivi allows him to use his senses, whereas Avicenna denies this ability—but it can be traced to more fundamental differences in their understanding of the perceptual process. Olivi thinks that perception is actively caused by the soul, whereas Avicenna thinks that it is initiated by the objects of perception. Moreover, I have argued that this difference reveals the divergent positions concerning bodily self-awareness that these authors have. However, the emerging idea of an ability to be aware of one’s body cannot be considered as a general or universal development. It may be nothing but an occasional peculiarity. Before reaching any conclusions on this score, more research needs to be conducted. Lastly, we may ask why the flying man disappeared from medieval philosophical imagery. There were enough authors who used it in relevant philosophical discussions that it could have lived on. Yet, it seems that fourteenth-century authors stopped using it. The final answer to this question is beyond the scope of this essay, but one important reason may be inherent in the change in the understanding of the nature of the thought experiment. Viewed from the perspective of the epistemological discussion concerning self-awareness, it becomes apparent that the flying man is not an argument. The reasoning behind the epistemological flying man— supposing that one is deprived of perceptual acts (P1), he is nevertheless

96

Juhana Toivanen

aware of himself (P3)—is supported by Avicenna’s thought experiment, but, within the experiment, the only reason to accept these two premises is Avicenna’s authority. Thus, the thought experiment becomes nothing but an argument from authority in the late thirteenth-century discussions: Avicenna says that the flying man knows his existence or the existence of his soul; therefore, the soul is directly aware of itself. We all know how vulnerable arguments from authority are. Once the authority loses his or her status as an authority, the argument ceases to be appealing. The flying man may have disappeared into the void whence he came due to the combined effect of the waning fame of Avicenna and the intrinsic development of medieval philosophical discourse.69 University of Jyva¨skyla¨

BIBLIOGRAPHY Aristotle. De anima, tr. J. A. Smith, in J. Barnes (ed.), The Complete Works of Aristotle: The Revised Oxford Translation (Princeton: Princeton University Press, 1984), 641–92. Avicenna. Avicenna latinus, Liber de anima seu Sextus de naturalibus, ed. S. van Riet (Louvain/Leiden: Peeters/Brill, 1968–72). (= Liber de anima) Avicenna. Avicenna’s De anima (Arabic Text): Being the Psychological Part of Kita¯b alshifa¯’, ed. F. Rahman (London: Oxford University Press, 1959). Beha, Helen Marie. “Matthew of Aquasparta’s Cognition Theory,” Franciscan Studies 21 (1961), 1–79, 383–465. Black, Deborah. “Avicenna on Self-Awareness and Knowing that One Knows,” in S. Rahman, T. Street, and H. Tahiri (eds), The Unity of Science in the Arabic Tradition (Dordrecht: Springer, 2008), 63–87. Bougerol, Jacques Guy. “Introduction,” in John of la Rochelle, Summa de anima, ed. J. G. Bougerol, Textes philosophiques du Moyen Aˆge 19 (Paris: Vrin, 1995), 9–43. Brown, Deborah J. “Aquinas’ Missing Flying Man,” Sophia 40 (2001), 17–31. Burr, David. “The Persecution of Peter Olivi,” Transactions of the American Philosophical Society 66 (1976), 1–98. Cory, Therese Scarpelli. Aquinas on Human Self-Knowledge (Cambridge: Cambridge University Press, 2014). Costa ben Luca. De differentia spiritus et animae, ed. J. C. Wilcox, in The Transmission and Influence of Qusta ibn Luca’s “On the Difference Between Spirit and the Soul,” Ph.D. dissertation (Ann Arbor: University Microfilms International, 1985), 143–91.

69 This research was financially supported by the European Research Council, the Stiftelsen Riksbankens Jubileumsfond, and the Academy of Finland.

The Fate of the Flying Man

97

Dominicus Gundissalinus. Tractatus de anima, in “The Treatise De Anima of Dominicus Gundissalinus,” ed. J. T. Muckle, Mediaeval Studies 2 (1940), 31–103. (= Tractatus de anima) Gauthier, Rene´ Antoine. “Les commentaires de la Vetus,” in Thomas Aquinas, Sentencia libri De anima, ed. R. A Gauthier, Sancti Thomae de Aquino Opera omnia iussu Leonis XIII P.M. edita 45.1 (Rome: Commissio Leonina, 1984), 1*–294*. Gilson, E´tienne. “Les sources gre´co-arabes de l’augustinisme avicennisant,” Archives d’histoire doctrinale et litte´raire du moyen aˆge 4 (1929), 5–149. Hasse, Dag Nikolaus. Avicenna’s De anima in the Latin West: The Formation of a Peripatetic Philosophy of the Soul, 1160–1300 (London: The Warburg Institute, 2000). Heller-Roazen, Daniel. The Inner Touch: Archaeology of a Sensation (New York: Zone Books, 2007). Ierodiakonou, Katerina and Sophie Roux (eds). Thought Experiments in Methodological and Historical Contexts, History of Science and Medicine Library 19 (Leiden: Brill, 2011). Jansen, Bernhard. “Descriptio codicum—Ratio editionis,” in Peter of John Olivi, Quaestiones in secundum librum sententiarum, vol. 3, ed. B. Jansen, Bibliotheca Franciscana Scholastica Medii Aevi 6 (Florence: Collegium S. Bonaventurae, 1926), xi–xlviii. John of la Rochelle. Summa de anima, ed. J. G. Bougerol, Textes philosophiques du Moyen Aˆge 19 (Paris: Vrin, 1995). Kaukua, Jari. Self-Awareness in Islamic Philosophy: Avicenna and Beyond (Cambridge: Cambridge University Press, 2015). Knuuttila, Simo. “Aristotle’s Theory of Perception and Medieval Aristotelianism,” in S. Knuuttila and P. Ka¨rkka¨inen (eds), Theories of Perception in Medieval and Early Modern Philosophy, Studies in the History of Philosophy of Mind 6 (Dordrecht: Springer, 2008), 1–22. McGinnis, Jon. Great Medieval Thinkers: Avicenna (Oxford: Oxford University Press, 2010). Marmura, Michael. “Avicenna’s ‘Flying Man’ in Context,” Monist 69 (1986), 383–95. Matthew of Aquasparta. Quaestiones de cognitione, in Quaestiones disputatae de fide et cognitione, Bibliotheca Franciscana Scholastica Medii Aevi 1 (Florence: Collegium S. Bonaventurae, 1957), 199–406. Mauro, Vincenzo. “La disputa de anima tra Vitale du Four e Pietro di Giovanni Olivi,” Studi medievali 38 (1997), 89–138. Pasnau, Robert. Theories of Cognition in the Later Middle Ages (Cambridge: Cambridge University Press, 1997). Perler, Dominik. The´ories de l’intentionnalite´ au Moyen Aˆge (Paris: Vrin, 2003). Peter of John Olivi. “Impugnatio quorundam articulorum Arnaldi Galliardi, articulus 19,” ed. S. Piron, in C. Ko¨nig-Pralong, O. Ribordy, and T. SuarezNani (eds), Pierre de Jean Olivi—Philosophe et the´ologien, Scrinium Friburgense 29 (Berlin: De Gruyter, 2010), 453–62.

98

Juhana Toivanen

Peter of John Olivi. Quaestiones in secundum librum sententiarum, 3 vols, ed. B. Jansen, Bibliotheca Franciscana Scholastica Medii Aevi 4–6 (Florence: Collegium S. Bonaventurae, 1922–6). (= Summa II) Peter of Spain. Commentario al “De anima” de Aristo´teles, ed. M. Alonso, Pedro Hispano Obras Filoso´ficas 2 (Madrid: Luis Vives, 1944). (= Questiones De anima) Piron, Sylvain. Parcours d’un intellectuel franciscain. D’une the´ologie vers une pense´e sociale: l’oeuvre de Pierre de Jean Olivi (ca. 1248–1298) et son traite´ De contractibus (Ph.D. dissertation, E´cole des Hautes E´tudes en Sciences Sociales, Paris, 1999). Putallaz, Franc¸ois-Xavier. La connaissance de soi au XIIIe sie`cle: de Matthieu d’Aquasparta a` Thierry de Freiberg (Paris: Vrin, 1991). Silva, Jose´ Filipe and Juhana Toivanen. “The Active Nature of the Soul in Sense Perception: Robert Kilwardby and Peter Olivi,” Vivarium 48 (2010), 245–78. Spruyt, Joke. “Peter of Spain,” in E. N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Winter 2012) . Tachau, Katherine. Vision and Certitude in the Age of Ockham: Optics, Epistemology and the Foundations of Semantics, 1250–1345 (Leiden: Brill, 1999). Teske, Roland. “William of Auvergne’s Debt to Avicenna,” in R. Teske (ed.), Studies in the Philosophy of William Auvergne (Milwaukee: Marquette University Press, 2006), 217–37. Thomas Aquinas. Quaestiones disputatae de veritate, in Opera Omnia Iussu Leonis XIII P.M. edita t. 22 (Rome: Ad Sanctae Sabinae, 1970). Thomas Aquinas. Summa theologiae, ed. P. Caramello (Turin: Marietti, 1948–50). Toivanen, Juhana. Perception and the Internal Senses: Peter of John Olivi on the Cognitive Functions of the Sensitive Soul, Investigating Medieval Philosophy 5 (Leiden: Brill, 2013). Toivanen, Juhana. “Perceptual Self-Awareness in Seneca, Augustine, and Peter Olivi,” Journal of the History of Philosophy 51 (2013), 355–82. Vital du Four. Quaestiones disputatae de cognitione, in “Le cardinal Vital du Four: Huit questions dispute´es sur le proble`me de la connaissance,” ed. F. Delorme, Archives d’histoire doctrinale et litte´raire du moyen aˆge 2 (1927), 151–337. William of Auvergne. De anima, ed. F. Hotot (Paris, 1674; reprinted Frankfurt: Minerva, 1963). William of Auvergne. The Soul, tr. R. J. Teske (Milwaukee: Marquette University Press, 2000). Yrjo¨nsuuri, Mikko. “Perceiving One’s Own Body,” in S. Knuuttila and P. Ka¨rkka¨inen (eds), Theories of Perception in Medieval and Early Modern Philosophy, Studies in the History of Philosophy of Mind 6 (Dordrecht: Springer, 2008), 101–16. Yrjo¨nsuuri, Mikko. “Types of Self-Awareness in Medieval Thought,” in V. Hirvonen, T. J. Holopainen and M. Tuominen (eds), Mind and Modality: Studies in the History of Philosophy in Honour of Simo Knuuttila (Leiden: Brill, 2006), 153–69.

Mind the Gap? The Principle of Non-Repeatability and Aquinas’s Account of the Resurrection Adam Wood

Recent years have seen considerable philosophical interest in Thomas Aquinas’s account of the general resurrection.1 An essential feature of his account, most recent commentators have supposed, is that when humans die, their separated souls carry on with their lives in heaven.2 Aquinas’s appeal to separated souls is necessary, according to these commentators, because he rejects the possibility of “gappy” or intermittent existence for 1 Some notable recent discussions of Aquinas on the resurrection include Jason Eberl, “The Metaphysics of the Resurrection: Issues of Identity in Aquinas,” Proceedings of the American Catholic Philosophical Association 74 (2000), 215–30, and “Do Human Persons Persist Between Death and Resurrection?” in Kevin Timpe (ed.), Metaphysics and God: Essays in Honor of Eleonore Stump (New York: Routledge, 2009), 188–205; Silas Langley, “Aquinas, Resurrection and Material Continuity,” Proceedings of the American Catholic Philosophical Association 75 (2001), 135–47; Robert Pasnau, Thomas Aquinas on Human Nature (Cambridge: Cambridge University Press, 2002), ch. 12; Eleonore Stump, “Resurrection, Reassembly and Reconstitution: Aquinas on the Soul,” in B. Niederbacher and E. Runggaldier (eds), Die menslische Seele: brauchen wir den Dualismus? (Frankfurt: Ontos, 2006), 151–71; Patrick Toner, “Personhood and Death in St. Thomas Aquinas,” History of Philosophy Quarterly 26 (2009), 121–38, and “St. Thomas on Death and the Separated Soul,” Pacific Philosophical Quarterly 91 (2010), 587–99; and Christina Van Dyke, “Not Properly a Person: The Rational Soul and ‘Thomistic Substance Dualism,’ ” Faith and Philosophy 26 (2009), 685–704, and “Human Identity, Immanent Causal Relations, and the Principle of Non-Repeatability: Thomas Aquinas on the Bodily Resurrection,” Religious Studies 43 (2007), 373–94. 2 There is debate in the literature about whether Aquinas thought that separated souls wholly constitute the human persons of whom they were formerly (pre-separation) just parts, as Stump and Eberl hold, or whether separated souls are merely former parts of human persons who no longer exist, as Pasnau, Toner, and Van Dyke all maintain. For proponents of the first view, it is because Aquinas thinks that separated souls constitute human persons, and thus that the humans themselves never really cease to exist, that he is able to secure our pre- and post-resurrection identity. For their opponents, it suffices that

100

Adam Wood

natural things, instead subscribing to what Christina Van Dyke calls “the principle of non-repeatability.”3 This principle states that once something has ceased entirely to exist, even God’s power cannot bring it back into existence numerically the same as it was before. Separated souls continuing on with our lives after our deaths thus play a crucial identity-preserving role between our pre- and post-resurrection selves. Or so most contemporary readers of Aquinas have assumed. This assumption rests, however, on shaky foundations. Some of Aquinas’s ablest late-medieval interpreters denied that Thomas endorsed a nonrepeatability principle strong enough to preclude God’s bringing humans back from the dead, whether or not their souls survived in the interim. Both John Capreolus, the so-called “prince of Thomists,” and Sylvester of Ferrara, the author of a massive commentary on the Summa contra gentiles, read Aquinas as having held that while God does in fact resurrect us by reuniting our separated souls to bodies, he need not have done things this way.4 God is perfectly capable of restoring us to existence numerically the same, even if our souls are corrupted at death along with the rest of us. In support of their interpretation, Capreolus and Ferrara lean heavily on one of Aquinas’s late quodlibetal questions, in which Thomas quite clearly rejects the principle of non-repeatability with respect to certain things, provided divine power is taken into account—humans among them, according to the two late-medieval Thomists.5 This question has gone unnoticed by most of Aquinas’s recent commentators.6 Understood along Capreolus and Ferrara’s lines, however, it poses a significant challenge to the prevailing view regarding Thomas’s principle of non-repeatability and

separated souls carry on with the same lives as the deceased humans formerly possessed. I side with Pasnau, Toner, and Van Dyke myself, and I will occasionally speak in this paper as though human persons cease to exist at death. None of the arguments I advance here ride on this assumption, however. Stump and Eberl agree with their opponents that Aquinas considered separated souls a necessary component of his account of the resurrection. They are free to adapt my reasoning in support of this assumption. 3 Van Dyke, “Human Identity,” 374. 4 Sylvester Ferrara, Commentaria in libros quatuor contra gentiles S. Thomae de Aquino 4.81 (ed. Leonine, 254–9) and John Capreolus, Defensiones theologiae divi Thomae Aquinatis, In quarto Sententiarum 43.1.1 (ed. Paban and Pegues, 1–29). 5 Quaestiones de quolibet (QQ) 4.3.2. All references to Aquinas’s works are to the online Roberto Busa and Enrique Alarco´n edition. All translations are my own unless otherwise noted. 6 There is a brief discussion of this quodlibetal question in Philip L. Reynolds, Food and the Body: Some Peculiar Questions in High Medieval Theology (Leiden: Brill, 1999), 426–7, and a more detailed treatment in Turner Nevitt, “Survivalism, Corruptionism, and Intermittent Existence in Aquinas,” History of Philosophy Quarterly 31 (2014), 1–19 (the conclusions of which differ considerably from my own).

Mind the Gap?

101

the role of separated souls in his account of the resurrection. At the very least, the contemporary consensus stands in need of further defense than has so far appeared in the literature. In this paper I argue that careful analysis of the late quodlibetal question, together with related texts, vindicates the contemporary consensus regarding what Aquinas held about the possibility of gappy human existence. Humans are indeed non-repeatable entities, for Thomas. The continued existence of our separated souls preserving our lives between death and rising again is indeed necessary to explain how God is able to resurrect us. On the other hand, somewhat surprisingly, Aquinas’s principle of non-repeatability turns out to be limited in scope, applying only to ordinary, sublunary material entities, but not to incorruptible beings like angels or celestial bodies. What ultimately underlies this limitation, I propose, is Aquinas’s theory of individuation. The best way of understanding Thomas’s core metaphysical tenet that designated matter is the principle of individuation of ordinary material forms is that such forms are individuated by their spatiotemporal starting points. Accordingly, objects originating at two different places or times are necessarily numerically distinct, even if God’s power is involved. Because angels and celestial bodies have different synchronic individuation conditions from ordinary material entities, the conditions for their diachronic identity differ as well. I defend these claims in the four sections of this paper as follows. I begin by examining Aquinas’s view in the late quodlibet, contrasting Capreolus and Ferrara’s interpretation of the passage with my own. In the second section I provide textual evidence indicating that throughout his career Aquinas endorsed a principle of non-repeatability strong enough to apply to human beings. Third, I show how Thomas’s theory of individuation gives him good reason for endorsing precisely the sort of non-repeatability principle I ascribe to him, and also helps to explain some otherwise confusing features of the quodlibetal question. I conclude with a look at two further scholastics who agree with several key aspects of my interpretation, and some thoughts about the philosophical attractiveness of Aquinas’s position.

1 . AQ U I N A S ’S Q U O D L I B E T 4. 3. 2 : TWO INTERPRETATIONS Scholastic quodlibetal disputations were held during Advent and Lent, and it was most likely during the spring of 1271 that someone asked Aquinas “whether God can restore numerically the same thing which was reduced to

102

Adam Wood

nothing?”7 Our record of the discussion opens with passages from Aristotle and Augustine arguing, respectively, for a negative and positive reply to the question. The passage from Aristotle’s De generatione et corruptione claims that “things whose substance is perishable must ‘return upon themselves’ specifically, not numerically.”8 For instance, if a cloud falls as rain and a new cloud is formed from the rainwater’s evaporation, the second cloud will only be of the same kind as the first, not numerically identical to it. Aquinas and many other scholastics took this passage to be expressing a nonrepeatability principle of some kind, although they disagreed about how far to accept it. The passage from Augustine’s City of God, in contrast, speaks optimistically about an omnipotent God’s ability to piece a human body back together, “even if its flesh had completely disappeared, and none of its material had remained in any cranny of the natural world.”9 For instance, Augustine says, even if I am eaten by cannibals and my flesh becomes theirs, God can keep tabs on my flesh and restore it to me.10 For human bodies, then, Augustine appears to reject the principle of non-repeatability and subscribe instead to what has been called the “reassembly” model of resurrection.11 Since much of what follows will revolve around Aquinas’s own reply to the question, it is worth quoting in full: It must be said that we should note a certain difference among things that can be reduced to nothingness. For there are some whose unity has continuity of duration in its very notion, as is clear in the case of motion and time. Therefore an interruption of such things is indirectly contrary to their numerical unity. And things that imply a contradiction are not contained among the number of things that are possible for God, since they fall short of the notion of being. Therefore if things of this sort are reduced to nothingness, God cannot restore them the same in number. For this would require contradictories to be true at the same time— namely, that an interrupted motion should be one. But there are other things whose unity does not have continuity of duration in its very notion, as is the case

7 For the dating of Aquinas’s quodlibetal questions, see Sandra Edwards’s introduction to Thomas Aquinas, Quodlibetal Questions 1 and 2 (Toronto: Pontifical Institute of Mediaeval Studies, 1983), 6. 8 Aristotle, De generatione et corruptione 2.11.338b16–19. 9 Augustine, City of God 22.20. 10 As for the flesh of the cannibals who have temporarily “borrowed” mine, Augustine says God could “reproduce it from whatever source he wished.” 11 Trenton Merricks discusses the “reassembly model” in “The Resurrection of the Body and the Life Everlasting,” in Michael Murray (ed.), Reason for the Hope Within (Grand Rapids, MI: Eerdmans, 1999), 272–6.

Mind the Gap?

103

with the unity of permanent things, unless per accidens, inasmuch as their act of being is subject to motion. For in this way things of this sort are measured by time, and their act of being is one and continuous according to the unity and continuity of time. And since an agent by nature cannot produce such things without motion, it is for that reason that a natural agent cannot restore things of this sort numerically the same if they were reduced to nothingness, or if they were corrupted substantially. But God could restore things of this sort, and do so without motion, since it is in his power to produce an effect without the intermediate causes. Therefore he can repair some things the same in number even if they have elapsed into nothingness.12

Two important features of this passage are obvious right off the bat. First, Aquinas clearly thinks there are some things that even God’s power could not bring back from nothingness. Once gone, they are gone forever. So Thomas clearly endorses some sort of non-repeatability principle. Second, however, his non-repeatability principle clearly has limits, since there are other things that God’s power could bring back from nothingness. To explain how far the limits of his non-repeatability principle extend, Aquinas introduces several important distinctions. The first is between successive entities and permanent entities. Time and motion are Aquinas’s examples of the former, and he holds that even God cannot restore such things once they are gone. Aquinas holds that God is omnipotent, of course, but he regularly denies that God’s power extends to bringing about contradictory states of affairs, such as that something should both exist and not exist at the same time. Such states of affairs, he sometimes says, “fall short of the notion of being.”13 This is what he says about the prospect of a successive entity being brought back from nothingness. The unity of a successive entity, he explains, “includes continuity of duration in its very notion.” An interruption or discontinuity in its duration directly contradicts such continuity, and hence indirectly contradicts its unity. And since in Aquinas’s estimation unity and being are “convertible” transcendental properties, an interruption in a successive entity’s duration contradicts its very existence as well, causing it to “fall short of the notion of being.”14 To suggest that a successive entity can be interrupted is like saying it can both exist and not exist at the same time in the same respect—something even God cannot bring about.

12 Aquinas, QQ 4.3.2. 13 See Summa theologiae (ST) 1a.25.3 and Quaestiones disputatae de potentia Dei (QDP) 1.3. 14 Aquinas discusses the convertibility of being and unity in Scripta super libros Sententiarum (In Sent.) 1.24.1.3, Quaestiones disputatae de veritate (QDV) 1.1 and ST 1a.11.1 ad 1.

104

Adam Wood

The crucial premise in Aquinas’s argument against the repeatability of successive entities is evidently the claim that the unity of such entities includes continuity of duration intrinsically: “in its very notion.” Aquinas frequently marks the contrast between permanent and successive entities by stating that the former exist totum simul, “all at once,” whereas the latter exist in part at one time and in part at others.15 The permanent/successive contrast thus roughly tracks the contemporary distinction between endurant and perdurant entities. Thomas’s claim is that for a successive entity to be unified, its temporal parts must be contiguous. His brief words concerning successive entities leave several matters unclear. Most importantly, he simply asserts that successive entities must be temporally continuous without explaining why. In the case of motions, at least, this seems open to question. What bars them from being paused and then resumed? If I interrupt my bike ride home from the office with a quick stop at the pub, could not my progress still be considered two halves of the same journey? Aquinas also ignores the question whether an entire motion could be repeated. If an athlete on our favorite team executes a particularly spectacular play in one game, we naturally hope he will repeat the same magic in the next. Could he repeat numerically the same magic? In section three below I will suggest what might underlie Aquinas’s approach to each of these issues concerning the repeatability of successive entities. For their part, however, Capreolus and Ferrara raise neither issue, but accept unreservedly that Aquinas has succeeded in demonstrating the absolute impossibility of bringing back times or motions from nothingness. The second important distinction Aquinas draws in the quodlibet we are considering is between what is possible for natural agents and what is possible for God. Natural agents, Thomas maintains, are incapable of restoring anything to existence that was previously destroyed. Only God’s supernatural power gives him this ability. As Capreolus and Ferrara understand him, Aquinas’s argument in support of this position is an enthymeme. What Aquinas actually states is that natural agents cannot cause anything to happen without some motion being involved. Capreolus supplies the rest of the argument as follows: “a difference in a motion or mutation entails a difference in what is produced by the motion or mutation. But what is produced twice by a natural agent is produced by a twofold motion or mutation. Therefore it will not be the same in number, 15 See for this formulation Summa contra gentiles (SCG) 1.55, 66; In Sent. 1.8.2.2; and ST 1a.7.4 ad 1 and 1a.10.5 ad 2. Robert Pasnau has helpful discussions of the permanent/successive distinction in “On Existing All at Once,” in C. Tapp and E. Runggaldier (eds), God, Eternity and Time (Farnham: Ashgate, 2011), 11–28, and Metaphysical Themes: 1274–1671 (Oxford: Clarendon Press, 2011), ch. 18.

Mind the Gap?

105

but in species.”16 It seems plausible to me that Aquinas’s case against the possibility of natural agents bringing anything back from nothingness does indeed involve these two unstated premises.17 The second of them piggybacks on Thomas’s reasoning regarding successive entities. For a natural agent to bring something into existence always involves a motion: a process of generation or alteration. But since motions, as successive entities, are non-repeatable, a natural agent restoring something to existence would have to do so by means of a numerically distinct process from the one that originally brought it about. The other unstated premise is that two numerically distinct processes of generation or alteration cannot bring about numerically one and the same result. Neither Capreolus nor Ferrara offers clear evidence for attributing this position to Aquinas, but as I’ll explain in section 3, Thomas had good reasons for accepting it. Putting the two unstated premises together, Thomas’s conclusion regarding natural agents’ inability to restore things to existence appears to follow, along with his conclusion that the same restriction does not apply to God. As Aquinas points out, God is able to bring things about without any motions being involved. He does not require processes of generation or alteration to accomplish his aims, but can create ex nihilo. Accordingly, God alone, and not natural agents, is able by a direct act of creation to restore to existence things that were previously destroyed. The important question is which things God is able to restore in this way. Capreolus and Ferrara believe Aquinas’s answer is ‘any permanent entities.’ Capreolus reminds us that God can do anything not implying a contradiction, then states simply that, “it implies no contradiction for a permanent thing that has been annihilated or corrupted to be repaired the same in number,” provided the repair-work is done “without motion or mutation” (which he says in deference to the reasoning discussed in the previous paragraph).18 For most medieval scholastics, Aquinas included, human beings count as clear cases of permanent entities. Accordingly, on Capreolus and Ferrara’s view, God could bring us back from total destruction simply by recreating us at some point after our deaths. Contrary to what most contemporary scholars have supposed, separated souls persisting in the interim between death and rising again turn out to be entirely

16 Capreolus, Defensiones (ed. Paban and Pegues, 2a). 17 Not only do Capreolus and Ferrara understand him this way, but some of Aquinas’s immediate successors, such as Godfrey of Fontaines and Henry of Ghent, offer precisely the same line of argument against the possibility of natural agents returning anything from nothingness. See Godfrey, Quaestiones quodlibetales 6.2 (ed. De Wulf and Hoffmans, 111) and Henry, Quaestiones quodlibetales 7.16 (ed. Wilson, 112). 18 Capreolus, Defensiones (ed. Paban and Pegues, 2a).

106

Adam Wood

inessential to Aquinas’s account of the resurrection. Ferrara explains that while on Aquinas’s view human souls do in fact survive death, “St. Thomas didn’t intend to rule it out that if some essential principle of a [permanent] thing were annihilated, the same thing in number could be repaired by divine power.”19 In other words, even if one of my essential components— such as my soul—should perish entirely at my death, God could still bring me back. Or such is Capreolus and Ferrara’s assessment of Aquinas’s view, guided chiefly by the quodlibet question we have been considering. Despite the excellent Thomist credentials of this interpretation, however, I do not think it is one we should accept. What Capreolus and Ferrara fail to notice is a third crucial distinction Aquinas draws in his quodlibet between two different kinds of permanent entities. Immediately after noting that the unity of permanent entities does not require continuity of duration, Aquinas adds the following: “unless per accidens, inasmuch as their act of being is subject to motion. For in this way things of this sort are measured by time, and their act of being is one and continuous according to the unity and continuity of time.” Thomas makes an important qualification here. It is true that the unity of permanent entities, unlike that of successive entities, does not require continuity of duration per se. The unity of a permanent entity may require such continuity per accidens, however, if the entity in question has an “act of being” that is “subject to motion.” Aquinas distinguishes, then, between permanent entities with an act of being subject to motion, and those that are not subject to motion in this way. It is only the latter sort of permanent entity that God can bring back from nothingness. The former sort, because their act of being is “measured by time,” requires continuity of duration in their act of being just as successive entities do. What exactly is Aquinas claiming here, and how does it bear on his account of the resurrection? The term esse, which I have translated “act of being,” plays a notoriously complex variety of roles in Aquinas’s thought. Yet, in the case of living things, Aquinas frequently says that our act of being is just our life.20 With this in mind, a helpful passage for coming to grips with Aquinas’s 19 Ferrara, Commentaria (ed. Leonine, 257a), is commenting on SCG 4.81.6, in which Aquinas responds to resurrection-critics claiming that since many of our “essential principles” cease to exist at death, not even God can bring us back once we are gone. Thomas’s response is to remind such critics that our essential principles can survive death. Our souls, for instance, can survive in separation from our bodies. Ferrara acknowledges that Aquinas responds this way, but attempts to clarify Thomas’s intent as seen in the translated quotation. See section 2 for more on Aquinas’s reply to this objection and Ferrara’s interpretation of it. 20 Aquinas repeats frequently and approvingly the phrase “for living things, to live is to exist,” which he traces to Aristotle’s De anima 2.4.415b13.

Mind the Gap?

107

distinction between the two different types of permanent entities is his discussion in the Summa theologiae of the way ingenerable, incorruptible, yet created entities such as angels and celestial bodies occupy an intermediate position between God’s eternal existence, on the one hand, and the temporal existence of ordinary, sublunary material entities, on the other.21 As Thomas explains, the measure of any permanent thing’s act of being per se is eternity. But all permanent things other than God “recede from permanence” inasmuch as they are in certain ways subject to change. Since only God is perfectly immutable, only God is eternal. Aquinas recognizes, however, two main ways that permanent entities may recede from permanence. Some do so insofar as their acts of being are subject to change. This is the case with all corruptible things—things that are generated into existence and later corrupted out of it. Insofar as a corruptible entity’s act of being recedes from permanence by being subject to change, Aquinas says, it becomes measured by time in just the same way as motions themselves are. Other created permanent entities recede from permanence in a different way. Angels and celestial bodies are unchangeable with respect to their act of being. They are neither generated nor corrupted. Yet, they are not perfectly unchangeable like God is. Celestial bodies, for instance, change their spatial position constantly as they move in circular fashion through the heavens. Angels, for their part, move locally only in a highly attenuated sense.22 They do change, however, with respect to their mental states of intellect and will. They make decisions, and they are not always thinking about everything they know.23 Insofar as celestial bodies move locally and angels move mentally, Aquinas recognizes that these entities fall short of God’s eternality, and are indeed measured by time. This is true of them, however, only insofar as they engage in these particular activities. Because their acts of being are unchangeable, Aquinas claims that angels and celestial bodies per se are not measured by time, but rather by what he calls the aevum.24 21 ST 1a.10.5. 22 Angels occupy a particular location only in a highly attenuated sense, insofar as they are busy applying their power to some task in that location, and consequently they move locally only in a highly attenuated sense as well. See ST 1a.52–3 and QQ 1.3.4–5 on angelic location and motion. 23 See ST 1a.10.5: “angels . . . combine an unchangeable act of being with changeableness according to choice, insofar as it pertains to their nature, and with changeableness of understanding and desire, and of motion in their own way.” See also ST 1a.58.1 for the claim that angels aren’t always thinking about everything they know and ST 1a.60.2 on the angelic capacity for choice. 24 See ST 1a.10.5: “and therefore things of this sort [i.e., celestial bodies and angels] are measured by the aevum, which is intermediate between eternity and time. The act of being that is measured by eternity, therefore, is neither mutable nor joined to mutability.

108

Adam Wood

Aeviternal existence differs from temporal existence in that it does not unfold part by part, such that some parts are “before” and others are “after.” Recalling, then, that an angel’s act of being is just its life, Aquinas’s claim is that an angel’s life does not unfold part by part. Rather, Gabriel has his life “all at once.” In this and related passages we see Aquinas invoking precisely the same distinction as he introduces in his quodlibet between permanent entities with an act of being subject to motion and those that are not subject to motion in this way.25 In addition, we see him maintaining that insofar as the former sort of permanent entity has an act of being subject to motion, its act of being is measured by time, rather than eternity or the aevum. Now humans clearly fall into the “corruptible” category of permanent entity. We are born and later die. Accordingly, our acts of being are measured by time. Unlike Gabriel’s life, our lives do include “befores” and “afters,” unfolding part by part, as the phenomenal data of our lived experience surely attest. Returning to our quodlibetal question, however, what Aquinas claims is that insofar as our lives are measured by time, they must unfold part by part with uninterrupted continuity. Our lives cannot, then, straddle temporal gaps. Their temporal parts must lie contiguous with one another, just as is the case with successive entities. Importantly, if these claims are correct, then Capreolus and Ferrara are mistaken about Aquinas’s quodlibet showing that God is able to bring back any permanent entities, including humans, from total destruction. When Aquinas concludes his quodlibetal question by stating that God “can repair some things the same in number even if they have elapsed into nothingness,” the “things” he has in mind do not include corruptible entities like us. If our lives ceased entirely at death, it would be just as impossible for God to resurrect us as it is for him to bring any successive entity back into existence once it is gone. It is fortunate for us, then, that our separated souls preserve our lives even after we ourselves are dead and gone. What God could bring back from nothingness, if he so chose, are incorruptible entities like angels and celestial bodies. Because Gabriel’s aeviternal life does not So therefore time includes before and after, but the aevum doesn’t include before and after in itself, though they are joined to it, while eternity does not include before and after at all, nor are they compatible with it.” See also In libros De caelo et mundo expositio 1.6.4, where Aquinas contrasts the acts of being of sublunary and celestial bodies on the grounds that the former, but not the latter, are subject to motion and hence measured by time. QQ 10.3.2, likewise, contrasts angelic acts of being with those of corruptible creatures on exactly the same grounds. 25 See ST 1a.2ae.31.2, QDP 5.1 ad 2, and In Sent 4.17.2.1.3 ad 3 for further cases of Aquinas’s claiming that the acts of being of permanent entities like human beings are measured by time insofar as they are subject to substantial change.

Mind the Gap?

109

unfold part by part, it need not unfold continuously. God could decide to annihilate Gabriel, then bring him back into existence at some point in the future.26 But God could not decide to do this with the likes of you or me. Up to this point I have tried to show that there is an alternative way of reading Aquinas’s quodlibet than Capreolus and Ferrara’s, based on a distinction the text appears to draw between two different kinds of permanent entities. The distinction is one Thomas relied on elsewhere. The reading I have proposed accords with the contemporary consensus that separated souls are essential to Thomas’s account of the resurrection. One thing my discussion so far should have made clear is that there is a range of different non-repeatability principles one might endorse. The strongest such principle would rule out the possibility of anything being brought back from destruction numerically the same as it was before, even by God’s power. Weaker, limited principles might rule out only the possibility of successive entities being brought back, or of entities being brought back by natural causes (as opposed to God’s power). Capreolus, Ferrara, and I agree that Aquinas endorses a limited-strength non-repeatability principle. There are some things that even God cannot bring back once they are gone, and other things that God (though not natural agents) is able to return from nothingness. On my view, however, the things God can bring back do not include corruptible entities such as human beings, while Capreolus and Ferrara think they do. For the remainder of this paper I will call my view that there are many permanent entities even God cannot bring back the ‘strong non-repeatability principle,’ and Capreolus and Ferrara’s view that God can bring back any permanent entities ‘weak non-repeatability.’ In section 2 I survey some textual evidence indicating that Aquinas endorsed strong non-repeatability.

26 Aquinas thinks there are good reasons for denying that God ever would annihilate any creature (see ST 1a.104.4, QDP 5.4, QQ 4.3.1), but believes he is free to do so (ST 1a.104.3, QDP 5.3). It is more difficult for me to see what Thomas might say about God’s ability to annihilate and recreate a celestial body. Aquinas thought of time as the measure of the first thing in motion: the outermost celestial sphere (see Commentaria in libros Physicorum [In Phys.] 4.17.4). So it is not clear that a gap in the existence of that celestial body, at least, could even count as temporal. Similarly, since all sublunary motion has celestial motion as its cause (see ST 1a.115.3), it might be that removing a heavenly body would require starting the whole world over again. Still, it may follow from my interpretation that God could annihilate the sun, for instance, and recreate numerically the same star after a temporal interval. I will explain further in section 3 why I do not think this poses any insurmountable difficulties for my interpretation.

110

Adam Wood 2 . S T R O N G N O N - R EP E A T A B I L I T Y IN A Q U I N A S

One reason for rejecting Capreolus and Ferrara’s understanding of Aquinas’s quodlibet is that in various other passages throughout his career Aquinas seems baldly to assert the non-repeatability of certain permanent entities, even by God’s power. Capreolus and Ferrara were well aware of this, and attempt to explain away some of these passages in accordance with their interpretation. But I think the balance of evidence is still against them. Consider first three objections that Aquinas addresses against the possibility of God’s raising humans from the dead. The first, from his Sentences commentary, runs as follows: our sense powers will not remain in our separated souls, and where there are two distinct sets of sense powers, there are two distinct animals, but where there are two distinct animals, there are two distinct men. Hence, even if God re-embodies my separated soul, the resultant individual will not be me.27 In the Summa contra gentiles a similar objection crops up: nothing can remain numerically the same if its “essential principles” are varied, and what is “returned to nothingness” cannot be restored with numerical identity, yet a human’s essential principles are returned to nothingness at death, so humans themselves cannot come back from the dead numerically the same as they were before.28 Finally, the Compendium theologiae addresses worries along related lines that if numerically the same “corporeity” is not recovered at the resurrection, then neither can the same human beings rise again.29 In each of these three cases Aquinas’s response is similar. Instead of challenging the premise asserting the impossibility of bringing back what was previously destroyed, he points out that because our souls are incorruptible, what really matters for the preservation of our numerical identity is never lost. In the Sentences commentary, for instance, Aquinas appears to accept that our sense powers are lost at death.30 Yet, because on Aquinas’s view the soul’s powers are only accidental forms really distinct from the soul

27 In Sent. 4.44.1.1.2 arg. 3. 28 SCG 4.80.3. 29 Compendium theologiae (CT) 1.154. 30 In Sent. 4.44.1.1.2 ad 3: “This argument holds good against those who posit a divided sensible and rational soul in a human: since according to this [view] the sensitive soul in a human would not be incorruptible, just as it is not in other animals. Whence at the resurrection it will not be the same sensible soul, and as a result nor will it be the same animal, nor the same man. If, however, we posit that the rational and sensible [souls] in a human are the same soul in substance, we need not worry on this account. . . . After death the sensible soul remains in substance just as the rational soul does. But the sensitive powers, according to some, do not remain. But since these powers are accidental properties, their variation does not affect the identity of the whole animal.”

Mind the Gap?

111

itself, our souls can survive without them, preserving our acts of being until the day of resurrection. Gaining new accidental forms on that day need not affect our numerical identity. On the other hand, Aquinas notes, the objection is a real worry for those who hold that sensitive and rational souls are distinct substantial forms. If this were the case, then losing our sense powers would also entail the loss of our sensitive soul. And losing a soul—one of our substantial forms—really would result in the loss of our numerical identity. Because Aquinas subscribes to the unicity of substantial forms, and thinks our rational and sensitive souls are really one and the same, this objection does not bother him. Importantly, however, the fact that he thinks it succeeds against proponents of a multiplicity of substantial forms indicates that he endorses the strong principle of non-repeatability. Otherwise, they could simply retort that God brings back numerically the same sense powers and sensitive soul at the resurrection, even though these had been corrupted at our deaths. In his response to the Summa contra gentiles objection, Aquinas does not explicitly state that without our souls surviving in a separated state, the objection would succeed. He simply points out that our essential principles are not destroyed at our deaths.31 But in the Compendium theologiae he does indicate that if by ‘corporeity’ is meant the particular quantitative dimensions that our body has in this life, then these dimensions won’t be recovered.32 He simply denies that getting numerically different dimensions would affect our post-resurrection identity, since they are mere accidents, not essential principles. Other passages also suggest that Aquinas accepted the strong nonrepeatability principle. In his Quaestio disputata de anima, for instance, one article inquires whether separated souls will have sense powers. Aquinas thinks they won’t, but argues that this doesn’t mean we’ll get new sense organs at the resurrection. That is because even though we will get new sense powers, they will proceed from the same sensitive souls, which is 31 SCG 4.81.6. The latter portion of Aquinas’s response to this objection, in which he claims that our matter under its individuating dimensions remains after our death, has seemed problematic to many recent commentators, since strictly speaking, Thomas’s commitment to the unicity of substantial forms requires him to say that nothing but prime matter remains after a substance’s demise. This means that for Aquinas, when a substance is corrupted, all of its accidents are corrupted as well, as he claims, for instance, in ST 1a.77.8. This includes accidents in the category of quantity, such as a body’s dimensions. So how can Aquinas claim here that our matter remains under the same dimensions as made it ours, even after our deaths? See Langley, “Material Continuity”; Pasnau, Aquinas on Human Nature, 390–3; and Eberl, “The Metaphysics of Resurrection,” 218–22, for three different proposed solutions to this problem. 32 CT 1.154. He goes on to add that the same is true of the soul’s sensitive and vegetative powers, just as we saw him say in his reply to the Sentences commentary objection (n. 30).

112

Adam Wood

enough to secure the numerical identity of the organs in question.33 Again, if Aquinas rejected strong non-repeatability, there would be no need for him to admit that we will get new sense powers at the resurrection. God could simply bring back the old ones. Two further passages have nothing to do with the resurrection, but rather with Eucharistic theology.34 If a Eucharistic host is burnt and ashes are generated, it cannot be said that the ashes are generated from the bread that was there before the consecration, since this substance, having been annihilated, “cannot return again, since what falls into nothingness does not return the same in number.”35 This holds true, Aquinas says, “unless perhaps the aforementioned substance is said to return because God creates a new substance in place of the first.”36 Setting aside the theological issues at stake here, the salient point for our purposes is that Aquinas thinks God could only replace the bread annihilated at the consecration with something similar in kind. Even God couldn’t bring back numerically the same bread as was there before. Capreolus and Ferrara, for their part, suggest several possible strategies for reconciling some of the passages I have discussed with their interpretation. In some of the passages where Aquinas appears to endorse the strong non-repeatability principle, they propose, he is making a point about what is possible for natural agents alone, apart from divine intervention.37 In other passages he is showing that even if, for the sake of argument or out of deference to Aristotle’s De generatione et corruptione, we were to concede strong non-repeatability, the resurrection would still be possible.38 A third possibility is that Aquinas is speaking simply about what is fitting for God to do in his ordained wisdom, rather than what God is capable of doing in his absolute power. As Ferrara puts it, “it would be unfitting for God to return something that had been wholly annihilated to existence, for since

33 Quaestio disputata de anima (QDA) 19 ad 13. 34 In Sent. 4.12.1.2.4 and ST 3a.77.5. 35 ST 3a.77.5. 36 Ibid. 37 This is what both say about the passages on the Eucharist I cited above. Ferrara, Commentaria (ed. Leonine, 256b): “he means that an annihilated thing cannot return the same in number through the action of a natural agent, but not that it cannot return through divine action.” Capreolus, Defensiones (ed. Paban and Pegues, 42b): “his opinion is that although something corrupted cannot return the same in number through a natural change, it can nevertheless return the same through a supernatural action.” 38 This is what Capreolus, Defensiones (ed. Paban and Pegues, 42a), says about the passages on the resurrection from In Sent. and SCG: “and perhaps . . . where he deals with the identity of the things rising again he means only that even having posited and conceded Aristotle’s principle from De generatione 2 . . . this would not impede the identity of the human rising again; since no essential part of the human falls entirely into nothingness, if we hold that there is no other substantial form in a human than the rational soul.”

Mind the Gap?

113

annihilation serves the purpose that something should entirely cease to exist, it would seem superfluous for God to annihilate something then desire to restore it to existence.”39 None of these explanations sounds very plausible to me. Take the first. In none of the passages I discussed above does Aquinas indicate that he is restricting his attention to what is possible for natural agents alone. Furthermore, given that the passages discuss miraculous events like the resurrection and the Eucharist, why would he restrict his attention to natural possibility? The same points apply to their second explanation. Why would Aquinas concede strong non-repeatability for the sake of argument in a discussion of the resurrection? Moreover, why would he do so without explaining that this is what he is up to? As for the last explanation, it is true that Aquinas thinks God never actually annihilates anything.40 But God allows things to fall prey to natural corruption. I am not sure what would be unfitting about his bringing some of them back. The various passages in which Aquinas appears to endorse the strong non-repeatability principle, which span the entire length of his career, are difficult to explain away in any of the ways Capreolus and Ferrara attempt.41 Thus, it would be strange if he rejects it in the late quodlibetal question. Admittedly, the textual case in favor of my interpretation isn’t ironclad. If Aquinas had finished commenting on the De generatione et corruptione, his remarks on the passage I mentioned above might have made his views clearer. But that passage is from near the end of Aristotle’s work, and Aquinas left off commenting long before he reached it.42 In his Physics commentary, Aquinas notes that Aristotle raises the question whether someone who loses then regains her health could be said to regain numerically the same health as she possessed previously. Health is presumably an 39 See Ferrara, Commentaria (ed. Leonine, 256b). See also Capreolus, Defensiones (ed. Paban and Pegues, 42a): “nor nevertheless from his having said these things can it be plainly inferred that it is impossible according to divine power, but rather that [Aquinas’s statement] applies to God’s ordained power.” 40 See the texts cited in n. 26. 41 Jean-Pierre Torrell’s catalogue of Aquinas’s works dates In Sent. to between 1252 and 1256, SCG book four to 1264–5, QDA to 1265–6, CT to 1266–7, and the later questions of the third part of ST to 1273 (Saint Thomas Aquinas: Vol. 1, The Person and his Work, tr. Robert Royal [Washington, DC: Catholic University of America Press, 1996], 330–61). 42 Many later medieval scholastics took De generatione et corruptione 2.11 as a point of departure for discussing God’s ability to return things from nothingness. H. A. G. Braakhuis examines the ways John Buridan, Albert of Saxony, Marsilius of Inghen, and various others did so in “John Buridan and the ‘Parisian School’ on the Possibility of Returning Numerically the Same: A Note on a Chapter in the History of the Relationship Between Faith and Science,” in Stefano Caroti and Pierre Souffrin (eds), La nouvelle physique du XIVe sie`cle (Florence: Olschki, 1997), 111–40.

114

Adam Wood

example of the sort of permanent entity that, on my view, Aquinas held to be non-repeatable. Unfortunately, Thomas does not think Aristotle answered the question about health after raising it, and he does not venture his own opinion on the issue either. Even if textual evidence does not conclusively show that Aquinas endorsed the strong non-repeatability principle, however, Thomas had strong philosophical reasons for doing so, stemming from his theory of individuation. The following section explains why.

3 . N O N - R EP E A T A B L I T Y A N D I N D I V I D U A T I O N The thesis that designated matter (materia signata) is the principle of individuation is about as fundamental to Aquinas’s metaphysics as anything can be. It is well known that Thomas thinks of designated matter as matter under certain dimensions. But there has been considerable disagreement about which dimensions, exactly, are supposed to do the individuating work.43 On one interpretation of his theory, though, the dimensions in question are the spatial and temporal ones a material substance starts out with. I will explain why this is a highly plausible construal of Aquinas’s view. I will then show how it entails the strong principle of non-repeatability and the need for separated souls to explain the resurrection. Finally, I will argue that while the role of Aquinas’s theory of individuation isn’t explicit in the quodlibet passage I discussed above, it best answers some of the questions about this passage I previously left unaddressed. Hence, it is plausible to suppose that the passage tacitly assumes something like the reasoning I outline here. The reason Aquinas needs a theory of individuation in the first place is that he does not think the substantial forms of material substances are intrinsically individual.44 He appeals to common forms or natures to explain how two or more material entities may belong to the same kind. 43 The disagreements about how to understand Aquinas’s view began early on with his later-medieval followers. Linda Peterson and Mauricio Beuchot describe how Cajetan and Chrysostum Javellus championed one line of interpretation, Capreolus and Ferrara another, in chapters 19 and 20 of Jorge J. E. Gracia (ed.), Individuation in Scholasticism: The Later Middle Ages and the Counter-Reformation, 1150–1650 (Albany, NY: State University of New York Press, 1994). 44 His usual way of putting this is to say that a form (or nature or essence) considered absolutely (absolute considerata) is neither one nor many. He reasons that if it did belong to human form (say) considered absolutely (that is, just in terms of what is intrinsic to it as such) to belong to some particular individual, such as Socrates, then it would never be found in any other individual, such as Plato. But because Aquinas wants to account for kind-membership in terms of the co-possession of forms, he must maintain that human form is found in Socrates and Plato alike. The third section of Aquinas’s early treatise, De ente et essentia (DEE), is probably his clearest discussion of these issues.

Mind the Gap?

115

Hence, he cannot directly invoke forms to explain how individuals of the same kind differ from one another. Instead, he assigns this role to matter. Not, of course, to prime matter by itself. On its own, Aquinas says, prime matter is numerically one since “it lacks the dispositions through which numerical differences occur.”45 Rather, matter under certain dimensions plays the individuating role for Aquinas.46 But which dimensions? Various commentators have noted that Aquinas’s terminology for the dimensions in question shifts throughout his career.47 Sometimes he says matter must be determinate with respect to its dimensions to do its individuating work.48 Elsewhere, however, he argues that the designating dimensions must be indeterminate, since the determinate dimensions of a typical material substance change considerably throughout its career.49 Sometimes commentators try to avoid the interpretative difficulties this shifting terminology poses by glossing “designated matter” as “a particular chunk of matter.”50 This phrase may well capture the gist of Aquinas’s meaning, but it is vague. In the case of living things, for instance, it cannot mean the particular grouping of material constituents from which they are composed at any given moment, since new matter is constantly cycled in and

45 De principiis naturae 2. 46 Importantly, it cannot be the dimensions themselves that do Aquinas’s individuating work, since, as accidents, their individuation depends on that of the substances to which they belong. As Thomas puts it in ST 1a.29.1: “a substance is individuated through itself, but accidents are individuated through their subject, which is a substance, for a whiteness is called ‘this’ insofar as it is in this subject.” 47 John Wippel provides a thorough analysis of Aquinas’s shifting terminology with respect to the principle of individuation in The Metaphysical Thought of Thomas Aquinas (Washington, DC: Catholic University of America Press, 2000), ch. 9.4. 48 See, for instance, Aquinas, DEE 2: “it is not matter taken in any way that is the principle of individuation, but only designated matter, which is considered under determinate dimensions.” 49 Aquinas, Expositio super librum Boetium De trinitate (In BDT) 4.2: “for these dimensions can be considered in two ways. In one way according to their termination; and I say that they are terminated according to determinate measure and figure . . . And in that way they cannot be the principle of individuation, since such termination of dimensions varies frequently in the same individual, and it would follow that the individual would not always remain the same in number. In another way they can be considered without such determination in the nature of their dimensions . . . And from these indeterminate dimensions matter becomes this designated matter, and in that way it individuates a form, and so from matter is caused numerical diversity in the same species.” 50 See, for instance, Anthony Kenny, Aquinas on Mind (New York: Routledge, 2003), 138, and Gyula Klima’s notes to his translation of DEE in Klima, Fritz Alhoff, and Anand Jayprakash Vaidya (eds), Medieval Philosophy: Essential Readings with Commentary (Malden, MA: Blackwell, 2007), 237 n. 18.

116

Adam Wood

out of this grouping to maintain homeostatis. So again, which dimensions does Aquinas think are responsible for designating matter into particular chunks? By my lights, the most satisfactory answer Thomas could give is that the spatial and temporal dimensions a material substance starts out with render its substantial form sufficiently unique that it is itself able to maintain the substance’s numerical identity from that point forward. This answer would solve the problem of a substance’s dimensions changing over time. It also fits well with the metaphors Aquinas adopts to illustrate the relationship between a living thing and the grouping of material constituents from which it is composed at any particular time.51 I include both the spatial and the temporal dimensions a material object possesses since Aquinas frequently characterizes what it is to be individual, as opposed to nonindividual, as a matter of being this thing here and now.52 That is why he describes the process of intellectual abstraction, by which we form general, non-individual concepts, as one of cognizing things apart from their individuating, here-and-now conditions. Aquinas frequently tells us that matter is the principle of individuation of forms.53 I am proposing that we take him at his word.54 51 Aquinas adopts the metaphors of a city, whose population comes and goes, and a fire, into which new bits of wood are constantly thrown and which consumes the old ones, to illustrate the relationship between a living thing and the particular material constituents it has at any given time. See In Sent. 2.30.2.1 ad 4, ST 1a.119.1 ad 5, QQ 8.3 ad 2, and In libros De generatione et corruptione expositio 1.15.4. 52 See, for instance, SCG 1.65.8; ST 1a.57.2 and 110.1, 1a.2ae.6.6; QDV 2.6 ad 1 and 8.11; and In BDT 4.2. 53 I am indebted on this point to Pasnau, who writes: “Contrary to what is usually supposed, matter is not Aquinas’s principle of individuation for material substances. Aquinas often does call matter the principium individuationis, but it is important to notice what matter is said to individuate. Matter is not that which individuates the whole substance (at least not directly), but that which individuates form. . . . For material substances, matter is the ultimate principium—the basis or starting point—of individuation. But it is more accurate to say that matter individuates the form, and that the form individuates the substance” (Aquinas on Human Nature, 391–2). Pasnau cites numerous passages in which Aquinas states that forms or natures are what matter individuates. 54 I do not mean to suggest that my account of Aquinas’s theory of individuation solves all the exegetical difficulties raised by his shifting terminology for the dimensions to which designated matter is subject. It would require a thorough case-by-case analysis of his various discussions of individuation to defend any thesis about whether his view changed throughout his career, or whether he eventually settled on some particular position. I agree with Stump, who writes that “one could wish for a great deal more clarity and precision with regard to the notion of matter that Aquinas has in mind when he claims that matter individuates” (Aquinas [New York: Routledge, 2003], 48). But I think Peter King overstates the case when he claims that Aquinas’s theory “founders” on the question “what makes matter ‘designated’?” (“The Problem of Individuation in the Middle Ages,” Theoria 66 [2000], 170). Thomas’s view may face some serious

Mind the Gap?

117

Supposing this understanding of Aquinas’s theory of individuation is correct, an important consequence is that once the substantial form of something individuated in this way is destroyed, even God cannot bring the substance back numerically the same as it was before. Substantial forms that inform matter under particular spatial and temporal dimensions are rendered unique and individual precisely in virtue of doing so. So if a given substantial form informs matter under some particular spatio-temporal dimensions, it is necessarily numerically distinct from any substantial form coming into matter at a different place or time. Hence, the substances resulting from uniting form and matter at two different times or places are also necessarily numerically distinct. No single substance can have two distinct sets of spatial and temporal starting points. An equally important result is that the only way of securing numerical identity between a substance which is corrupted at some point in time and one coming into existence at a future time is by its substantial form continuing to exist in the interim, preserving its numerical identity. Substantial forms are uniquely capable of playing this identity-preserving role since according to Aquinas they possess the same act of being as the substances they inform.55 In the case of a living thing like me, this means that my soul possesses the same life as I do. Accordingly, my soul can preserve my life even if, having died, the life no longer belongs to me. If God then reunites my soul to some matter in the future, the resulting human being will be me.56 If my soul is destroyed at death, however, and

objections, many of which his later-medieval critics pointed out, but it is not simply dead in the water. Indeed, understood the way I’m suggesting, Aquinas’s view bears interesting similarities to the ‘origins essentialism’ proposed by Saul Kripke and defended in recent years by Graeme Forbes, among others. See Kripke, Naming and Necessity (Cambridge, MA: Harvard University Press, 1980), 112–14, esp. 114 n. 56, and Forbes, “Origin and Identity,” Philosophical Studies 37 (1980), 353–62. 55 See Aquinas, QQ 9.2.2, for an illuminating account of the way in which a substance and its form possess the same act of being, but in different modes. A substance has its act of being in a subsistent mode, as that which exists. Its form has the same act of being in an inherent mode, as that by which the substance exists. Aquinas holds that human souls alone of all forms have their acts of being in both an inherent mode, as that by which humans exist, and in a subsistent mode. Because they are subsistent entities, they are able to exist in a separated state after we die. See QDA 1 and ST 1a.75.2. Klima has a helpful discussion of these issues in “Aquinas on the Materiality of the Human Soul and the Immateriality of the Human Intellect,” Philosophical Investigations 32 (2009), 169–71. 56 Van Dyke, “Human Identity,” 388–90, explains how separated souls provide the necessary connections between our earthly and post-resurrection stages of existence.

118

Adam Wood

my life ceases altogether, then even God could not bring it back. Separated souls are necessary after all for God to bring us back from the dead.57 Someone might worry at this point that my reasoning in the previous two paragraphs is inconsistent with the interpretation of Aquinas’s quodlibetal question I offered above. I claimed that Aquinas only thinks corruptible, sublunary material entities are non-repeatable. The non-repeatability principle does not extend to celestial bodies because they do not have an act of being subject to motion and measured by time. Yet, the heavenly spheres are still material substances, and their substantial forms are not intrinsically individuated. One might assume, then, that the argument I have presented in this section would show that the sun, moon, etc. are non-repeatable in just the same way as any other material substances. Can celestial bodies be brought back from nothingness or not? The fact that Aquinas says little about the individuation of the heavenly bodies makes replying to this objection difficult. But it seems very likely to me that Thomas viewed the individuation of the celestial spheres differently from that of ordinary, corruptible, sublunary material entities. In several places Aquinas claims that while it is not impossible for there to be more than one celestial body of a certain kind in the same way as it is impossible to have more than one angel in a given species, it is impossible nonetheless.58 This is because for the sun, moon, or any other heavenly sphere to inform matter at all requires that it “use up” all the matter suitable for it. There is no further matter left over for a second sun or moon to inform. I will not try to unpack all the details of this claim here. But it seems clear

57 A helpful passage from Aquinas’s Sentences commentary appears to endorse precisely the line of reasoning I’ve advanced here in response to an objection against the possibility of the resurrection. The objection simply appeals to the same line from the De generatione et corruptione we saw above (see n. 8) as expressing some sort of nonrepeatability principle. Aquinas replies that such a principle doesn’t threaten the possibility of our resurrection: “the form of other generable and corruptible things isn’t subsistent per se, so that it is able to remain after the corruption of the composite, as is true of the rational soul, which retains the act of being which it acquires in the body even after the body [is gone]. And the body is brought into participation in that act of being at the resurrection, since the act of being of the body and of the soul in the body are not distinct, or else the conjunction of the soul and body would be accidental, and thus no interruption occurs in the substantial act of being of the human, such that he would be unable to return the same in number because of an interruption in his existence, just as happens in other corporeal things, of which the act of being is wholly interrupted, since the form does not remain, while the matter remains under a different act of being. But nevertheless neither is a human repeated the same in number through natural generation, since the body of the human generated is not made from the whole matter of the one generating, whence there is diversity in number, and as a result, of the soul, and of the whole human” (In Sent. 4.44.1.1.2 ad 1). 58 See ST 1a.119.1, SCG 2.93.6, and In Phys. 6.3.774.

Mind the Gap?

119

that it entails different individuation conditions for celestial spheres than for ordinary material entities. The sun and moon are not individuated by informing matter under particular spatio-temporal dimensions. After all, time and place only began to exist with the production of the heavens. Heavenly bodies are individuated instead by informing all the matter of a kind suitable for their nature. Accordingly, I think Aquinas might consistently maintain that they are among the permanent entities God is able to bring back from total annihilation, even if God cannot do so in the case of corruptible material entities such as human beings.59 Even if this reply assuages a potential worry about the consistency of my reasoning in this section with my interpretation of Aquinas’s quodlibet above, one might still wonder how exactly the two are related. Thomas’s discussion in the quodlibetal question of successive vs. permanent entities, acts of being subject to motion, and the abilities of God vs. natural agents to bring things back from destruction includes no explicit mention of his theory of individuation. Furthermore, if the reasoning I have presented in this section is correct, it might seem that it supersedes or renders otiose what he says about non-repeatability in the quodlibet. What I would claim is that Aquinas’s theory of synchronic individuation implicitly underlies much of his argumentation in the quodlibet. This can be seen by recalling some of the questions I left unanswered when I discussed the passage above. Consider first Aquinas’s claim that the temporal parts of successive entities must lie contiguous to one another without any intervening temporal gaps. He offered no explanation why we should accept this view, and in the absence of any explanation, the question arose why a single motion— a bike-ride home from my office, say—could not be interrupted for a spell.60 59 I owe the suggestion that Aquinas recognized different individuating conditions for celestial bodies than he did for sublunary material entities to Giorgio Pini, who writes that “[c]reatures are organized in a hierarchical structure where numerical multiplicity is gradually replaced by specific multiplicity. At the bottom, there are material things; these can be numerically multiplied in the same species, because their form is such that it is received in matter, and their essence is such that it is never fully actualized in any of its instantiations. Just above them are heavenly bodies such as the sun and the moon, whose form is received in matter and whose essence is not multiply instantiated, because all the suitable matter is used up by a single instantiation. At the pinnacle of the created universe there are angels. In angels, the distinction between individual and essence vanishes: their essence is by itself individual and they are just their essence” (“The Individuation of Angels from Bonaventure to Duns Scotus,” in Tobias Hoffman (ed.), A Companion to Angels in Medieval Philosophy [Leiden: Brill, 2012], 92–3). 60 One thing to point out regarding this question is that when it comes to the way we conventionally designate motions into parcels like bike-rides, we may do so however we please. Simply pointing out that we regard an interrupted trip as a single journey does not conclusively support any metaphysical conclusions. As Van Dyke (“Human

120

Adam Wood

His only significant discussion of the unity of motions, in his Physics commentary, sheds little light. He states that “if rest is interposed, there will be many motions and not one, for motions that have rest between them are many and not one.”61 But he does not tell us why this must be so. My proposal, however, is that Aquinas saw no good way of distinguishing one motion from another without insisting on temporal continuity, and that his theory of the way material forms are individuated gave him good reason for believing this to be the case. Suppose I begin to raise my hand above my head, pause for a moment at shoulder level, then continue to extend all the way up. What could settle the question whether I have just described two halves of the same upward motion, or two distinct motions upward? In general, Aquinas holds that the individuation of material forms depends crucially on their spatio-temporal starting points. Accordingly, it would make perfect sense for him to insist that motions too are individuated, at least in part by the time and place they begin. Indeed, it is hard to see what else he might say. Aquinas’s views concerning synchronic individuation thus plausibly provide the theoretical justification for what he says about the diachronic identity of motions. If this proposal is correct, then it also helps to answer two other questions I raised above. Why could not a particular motion, such as a particularly heroic swing of the bat, be repeated anew at some point in the future? Suppose it is the same player swinging the same bat in the same way over the same plate both times. Aquinas does not even raise this issue, much less address it. But possibly this is because he saw it as admitting the same solution as the question about why motions must be temporally continuous. Motions are individuated in part by their temporal starting points. So two swings undertaken at different times must be numerically distinct, even if they are identical in many other respects. Something similar might be said about Aquinas’s reasoning regarding the abilities of natural agents to return things from nothingness. His argument appeared to rely on the claim that two numerically distinct production processes of generation or alteration would necessarily result in two numerically distinct entities being produced. But why must this be so? Well, on the view I’m proposing, for the two processes

Identity,” 377–82) argues, Aquinas’s claims about non-repeatability range only over natural entities. When it comes to artifacts, our intuitions are largely guided by the uses to which we put them. I could disassemble my bicycle, put it back together, and have numerically the same bike. I could not, presumably, crush Michelangelo’s “David” into powder, then reconstruct numerically the same statue, no matter how meticulously I did so. Even if we restrict our attention to natural motions, however, we might still wonder why Aquinas is so insistent upon their temporal continuity. 61 In Phys. 5.7.706.

Mind the Gap?

121

to be distinct, they must have different spatio-temporal starting points. But in that case, supposing it is processes of generation we are talking about, the two processes will also involve substantial forms informing different designated matter—matter under different spatio-temporal dimensions. Hence, the resulting substances will be distinct too. It is certainly possible that there is some other explanation than the one I have offered why Aquinas’s quodlibet says what it does about the diachronic identity conditions for motions, things with acts of being subject to motion, etc.62 But in the cases I have discussed it seems plausible to suppose that Thomas tacitly assumes his best-known theory of synchronic individuation, providing him with a deep-seated philosophical reason for insisting on the non-repeatability of both motions and ordinary, sublunary permanent entities like us. If this is true, then it gives us in turn a good reason for believing, pace Capreolus and Ferrara, that he actually endorsed the strong non-repeatability principle.

4. CAJETAN, SCOTUS, AND CONCLUSION As it turns out, not all of Aquinas’s fellow scholastics interpreted his views on the possibility of God’s bringing us back from total destruction in the same way as Capreolus and Ferrara did. Both John Duns Scotus and Thomas de Vio Cajetan seem to have understood Aquinas as having held that humans are non-repeatable. They also appreciate the connection I stressed in section 3 between the conditions for our synchronic individuation and for our diachronic identity. Consider first a passage from Cajetan’s commentary on Aquinas’s De ente et essentia. Cajetan confronts a Scotist objection against Aquinas’s view that each species of separated substance has just one individual. The objection states: God can annihilate a particular angel in a particular species. Once the angel has been annihilated, God can restore this species in some other individual, but not in the same individual, according to that [i.e., the Thomist] opinion, because according to them man could not rise again numerically the same unless the intellectual soul remains the same in number. Therefore in the same species there will be many individuals, at least successively.63

62 He might, for instance, simply be deferring to the authority of Aristotle, who claims that temporal continuity is required for the unity of a motion in Physics 5.4.228b2–10. 63 Cajetan (Thomas de Vio), Commentary on St. Thomas Aquinas’ On Being and Essence, tr. Lottie Kendzierski and Francis C. Wade (Milwaukee, Wis.: Marquette University Press, 1964), 197.

122

Adam Wood

On Capreolus and Ferrara’s interpretation of Aquinas, the obvious rejoinder would be that Thomas did not require the intellectual soul to remain the same in number for God to raise numerically the same human from the dead. According to the weak non-repeatability principle, God can return things from nothingness even if their essential principles are lost entirely. Instead of replying this way, however, Cajetan points out that there is an important difference between the cases of humans and angels: In the case he mentioned the angel who is restored would necessarily be numerically the same angel, since it is from one and the same principle that Gabriel is an individual and is Gabriel. Nor is it true that we hold God could not restore the same individual, even though we say that if numerically the same soul is not taken up again—which ought not to be annihilated if the notion of resurrection is to be saved—it would not be numerically the same man as the one who has the possibility of being restored by God numerically the same.64

Cajetan points out that it is from one and the same principle that Gabriel is an angel and that he is the particular angel he is. That is because Gabriel is not composed of form and matter—of an essential nature and some designated matter that contracts his nature to some particular dimensions. Gabriel’s substantial form is intrinsically individuated, rather than being individuated by informing matter under certain spatio-temporal dimensions. Accordingly, as we saw above, God could bring Gabriel back from annihilation simply by recreating him. Not so with humans, however. God is, of course, able to raise us from the dead. According to Cajetan, though, this is only true if our souls are not annihilated. In our case, continuity of our souls is required. A human’s soul “ought not to be annihilated if the notion of resurrection is to be saved.” Certainly Cajetan’s brief text falls short of corroborating the entire interpretation I’ve advanced. But Cajetan does appear to recognize that humans are non-repeatable unless our souls survive after our deaths, that angels do not share this requirement, and that this difference between the two stems from their different metaphysical makeup. So at least one formidable representative of the Thomist tradition endorsed several key features of my reading. So too, apparently, did Scotus. The subtle doctor rejects the strong principle of non-repeatability.65 In fact, it is not clear he endorsed any non-repeatability principle at all. He argues, for instance, that there is no good reason why natural agents could not bring things back from total destruction.66 In his discussion of the resurrection he wants to show that God could bring back to life even a human being who had been entirely 64 Ibid., 201. 65 See Scotus, Ordinatio 4.43.1.2–7 (ed. Wadding, 5–8). 66 Ibid., 4.43.1.8 (ed. Wadding, 26–7).

Mind the Gap?

123

destroyed, body and soul. There’s a good reason for this. Scotus repeats almost verbatim the argument we saw above from Aquinas’s Sentences commentary, which claims that if the sensitive soul is a distinct substantial form from the intellective, then the resurrection really is impossible, since the sensitive soul is destroyed at death, and even God could not bring it back.67 Since Scotus himself believed that sensitive and intellective souls are distinct substantial forms, this argument would cut against his own view. Scotus was clearly familiar with Aquinas’s treatment of the resurrection in his Sentences commentary, and appears to agree with me that Aquinas endorses the strong non-repeatability principle there. What is even more interesting, however, is that in the course of arguing against strong nonrepeatability one point Scotus makes is that the time at which a thing comes about plays no essential role in determining its identity.68 That, he says, is why an interruption in something’s act of being need not affect its diachronic identity. Scotus evidently sees that a view like Aquinas’s, on which my starting points do play an important role in who I end up becoming, makes it impossible for one and the same thing to start up its existence twice at two distinct times. So Scotus also seems to recognize the connection I alleged in section 3 between the individuating role of a thing’s temporal origins and the requirement that its act of being unfold continuously if it is to remain the same over time. To be sure, Scotus no more than Cajetan corroborates the whole interpretation I have advanced in this paper. But he does apparently accept at least two of its important components. Now, in bringing up Scotus and Aquinas’s disagreement about individuation, I should make clear that nothing I have said up to this point is meant to give us a reason for preferring Thomas’s view over any of its competitors. My focus in this paper has been solely on explicating Aquinas’s thought against what I take to have been the mistaken interpretations of Capreolus and Ferrara. I have argued that Thomas endorsed the strong non-repeatability principle and insisted on the need for separated souls to explain the 67 Ibid., 4.43.1.2 (ed. Wadding, 5–6); see Aquinas, In Sent. 4.44.1.1.2 arg. 3 and ad 3, cited in nn. 27 and 30 above. Scotus also repeats verbatim the explanation Aquinas offers at In Sent. 4.44.1.1.2 ad 1 how God is able to bring numerically the same humans back from the dead; see n. 57. 68 See Scotus, Ordinatio 4.43.1.5 (ed. Wadding, 7): “diversity of posterior things does not entail diversity of what is essentially prior; but those instants [of a thing’s generation] are essentially posterior to the act of being of a permanent thing; whence that act of being remains the same in whichever succeeding instants; therefore whether there is continuation between those instants or not, no less will the act of being be able to be the same.” Richard Cross has a very helpful discussion of how Scotus rejected the significance of temporal origins for a thing’s identity in “Identity, Origin and Persistence in Duns Scotus’ Physics,” History of Philosophy Quarterly 16 (1999), 1–18.

124

Adam Wood

resurrection, just as most contemporary commentators have assumed. I also hope to have shown that his reasons for doing so are intimately tied to his metaphysics of individuation, such that only entities which get their start and unfold continuously in time are non-repeatable. I take these to be significant findings for our understanding of what Aquinas himself believed. When it comes to the philosophical correctness of Thomas’s stance on non-repeatability and the resurrection, however, it is well known that Scotus had his own highly original theory of individuation, supported both by powerful positive arguments and by compelling negative arguments against the views of his opponents.69 Anyone convinced by Scotus’s arguments is unlikely to find Aquinas’s position philosophically attractive. Similarly, in recent years, philosophers have offered up a variety of attempts to explain how God might resurrect the dead without appealing to separated souls. Dean Zimmerman’s “falling elevator model” of the resurrection, for instance, suggests that God might grant every atom of our body a miraculous fissioning power to project an offspring forward in time to the day of the resurrection.70 Without entering into the details of this proposal, Zimmerman thinks one of its virtues is preserving the “immanent causal connections” between our earlier temporal stages and our later post-resurrection stages that he considers necessary for diachronic numerical continuity in material beings.71 Aquinas, I think, would want to know from Zimmerman whether or not, on his view, one and the same human being is supposed to have two distinct spatio-temporal starting points. If so, Aquinas would challenge it for the reasons I discussed above. If not, Thomas would want to know what exactly is supposed to persist in the interim between death and resurrection, if not a separated

69 See King, “Duns Scotus on the Common Nature and the Individual Differentia,” Philosophical Topics 20 (1992), 51–76, and Allan B. Wolter’s ch. 12 in Gracia (ed.), Individuation in Scholasticism, for two excellent discussions of Scotus’s theory of individuation. Klima finds resources in Thomas of Sutton for defending Aquinas’s account against Scotist objections in “Thomas Sutton on Individuation,” Proceedings of the Society for Medieval Logic and Metaphysics 5 (2005), 70–8. 70 Dean Zimmerman, “The Compatibility of Materialism and Survival: The ‘Falling Elevator’ Model,” Faith and Philosophy 16 (1999), 194–212, and “Bodily Resurrection: The Falling Elevator Model Revisited,” in G. Gasser (ed.), Personal Identity and Resurrection: How to Survive your Death (Farnham: Ashgate, 2010), 33–50. 71 Zimmerman provides an exhaustive account of immanent causation in “Immanent Causation,” Philosophical Perspectives 11 (1997), 433–71. He defends the necessity of immanent causal connections for preserving the diachronic numerical identity of material objects in “Criteria of Identity and the ‘Identity Mystics’,” Erkenntnis 48 (1999), 281–301.

Mind the Gap?

125

soul. That Zimmerman or anyone else should be moved by questions like these, however, is not something I will argue here.72 Wheaton College

BIBLIOGRAPHY Aristotle. The Complete Works of Aristotle, ed. Jonathan Barnes (Princeton: Princeton University Press, 1984). Augustine. City of God, tr. Henry Bettenson (London: Penguin Books, 1972). Beuchot, Maurice. “Chrysostum Javellus and Francis Sylvester Ferrara,” in J. J. E. Gracia (ed.), Individuation in Scholasticism: The Later Middle Ages and the Counter-Reformation, 1150–1650 (Albany, NY: State University of New York Press, 1994), 457–74. Braakhuis, H. A. G. “John Buridan and the ‘Parisian School’ on the Possibility of Returning Numerically the Same: A Note on a Chapter in the History of the Relationship between Faith and Science,” in Stefano Caroti and Pierre Souffrin (eds), La nouvelle physique du XIVe sie`cle (Florence: Olschki, 1997), 111–40. Cajetan, Thomas de Vio. Commentary on St. Thomas Aquinas’ On Being and Essence, tr. Lottie Kendzierski and Francis C. Wade (Milwaukee, WI: Marquette University Press, 1964). Cross, Richard. “Identity, Origin and Persistence in Duns Scotus’ Physics,” History of Philosophy Quarterly 16 (1999), 1–18. Eberl, Jason. “Do Human Persons Persist between Death and Resurrection?” in Kevin Timpe (ed.), Metaphysics and God: Essays in Honor of Eleonore Stump (New York: Routledge, 2009), 188–205. Eberl, Jason. “The Metaphysics of the Resurrection: Issues of Identity in Aquinas,” Proceedings of the American Catholic Philosophical Association 74 (2000), 215–30. Forbes, Graeme. “Origin and Identity,” Philosophical Studies 37 (1980), 353–62. Godfrey of Fontaines. Quaestiones quodlibetales in Les philosophes Belges, textes et ´etudes, tome III, Les quodlibet cinq, sex et sept de Godefroid de Fontaines, ed. M. De Wulf and J. Hoffmans (Louvain: Institut Superieur, 1914). Gracia, Jorge J. E. (ed.). Individuation in Scholasticism: The Later Middle Ages and the Counter-Reformation, 1150–1650 (Albany, NY: State University of New York Press, 1994). Henry of Ghent. Quaestiones quodlibetales, in Henrici de Gandavo Opera omnia, vol. 11, ed. G. A. Wilson (Leuven: Leuven University Press, 1991). John Capreolus. Defensiones theologiae divi Thomae Aquinatis, In quarto Sententiarum, ed. C. Paban and T. Pegues, vol. 7 (Tours: Cattier, 1908).

72 I am grateful to Gyula Klima for invaluable help with this paper’s early stages, to Turner Nevitt for drawing my attention to some key texts, and to the anonymous referee for OSMP for prompting me to think much harder about my central claims.

126

Adam Wood

John Duns Scotus. Ordinatio, in Joannis Duns Scoti Opera omnia, ed. L. Wadding, vol. 20 (Paris: Vives, 1894). Kenny, Anthony. Aquinas on Mind (New York: Routledge, 2003). King, Peter. “Duns Scotus on the Common Nature and the Individual Differentia,” Philosophical Topics 20 (1992), 51–76. King, Peter. “The Problem of Individuation in the Middle Ages,” Theoria 66 (2000), 159–84. Klima, Gyula. “Aquinas on the Materiality of the Human Soul and the Immateriality of the Human Intellect,” Philosophical Investigations 32 (2009), 163–82. Klima, Gyula. “Thomas Sutton on Individuation,” Proceedings of the Society for Medieval Logic and Metaphysics 5 (2005), 70–8. Klima, Gyula, Fritz Alhoff, and Anand Jayprakash Vaidya. Medieval Philosophy: Essential Readings with Commentary (Malden, MA: Blackwell, 2007). Kripke, Saul. Naming and Necessity (Cambridge, MA: Harvard University Press, 1980). Langley, Silas. “Aquinas, Resurrection and Material Continuity,” Proceedings of the American Catholic Philosophical Association 75 (2001), 135–47. Merricks, Trenton. “The Resurrection of the Body and the Life Everlasting,” in Michael Murray (ed.), Reason for the Hope Within (Grand Rapids, MI: Eerdmans, 1999), 261–86. Nevitt, Turner. “Survivalism, Corruptionism, and Intermittent Existence in Aquinas,” History of Philosophy Quarterly 31 (2014), 1–19. Pasnau, Robert. Metaphysical Themes: 1274–1671 (Oxford: Clarendon Press, 2011). Pasnau, Robert. “On Existing All at Once,” in C. Trapp and E. Runggaldier (eds), God, Eternity and Time (Farnham: Ashgate, 2011), 11–28. Pasnau, Robert. Thomas Aquinas on Human Nature (Cambridge: Cambridge University Press, 2002). Peterson, Linda. “Cardinal Cajetan and Giles of Rome,” in J. J. E. Gracia (ed.), Individuation in Scholasticism: The Later Middle Ages and the Counter-Reformation, 1150–1650 (Albany, NY: State University of New York Press, 1994), 431–56. Pini, Giorgio. “The Individuation of Angels from Bonaventure to Duns Scotus,” in Tobias Hoffman (ed.), A Companion to Angels in Medieval Philosophy (Leiden: Brill, 2012), 79–116. Reynolds, Philip L. Food and the Body: Some Peculiar Questions in High Medieval Theology (Leiden: Brill, 1999). Stump, Eleonore. “Resurrection, Reassembly and Reconstitution: Aquinas on the Soul,” in B. Niederbacher and E. Runggaldier (eds), Die menslische Seele: brauchen wir den Dualismus? (Frankfurt: Ontos, 2006), 151–71. Sylvester Ferrara. Commentaria in libros quatuor contra gentiles S. Thomae de Aquino, in Sancti Thomae Aquinatis opera omnia iussu impressaque Leonis XIII P.M. edita, vol. 15 (Rome: Commissio Leonina, 1950). Thomas Aquinas. Quodlibetal Questions 1 and 2, tr. Sandra Edwards (Toronto: Pontifical Institute of Mediaeval Studies, 1983).

Mind the Gap?

127

Thomas Aquinas. Sancti Thomae de Aquino opera omnia, ed. Roberto Busa and Enrique Alarco´n (2000–). Toner, Patrick. “Personhood and Death in St. Thomas Aquinas,” History of Philosophy Quarterly 26 (2009), 121–38. Toner, Patrick. “St. Thomas on Death and the Separated Soul,” Pacific Philosophical Quarterly 91 (2010), 587–99. Torrell, Jean-Pierre. Saint Thomas Aquinas: Vol. 1, The Person and his Work, tr. Robert Royal (Washington, DC: Catholic University of America Press, 1996). Van Dyke, Christina. “Human Identity, Immanent Causal Relations, and the Principle of Non-Repeatability: Thomas Aquinas on the Bodily Resurrection,” Religious Studies 43 (2007), 373–94. Van Dyke, Christina. “Not Properly a Person: The Rational Soul and ‘Thomistic Substance Dualism,’ ” Faith and Philosophy 26 (2009), 685–704. Wippel, John. The Metaphysical Thought of Thomas Aquinas (Washington, DC: Catholic University of America Press, 2000). Wolter, Allan. “John Duns Scotus,” in J. J. E. Gracia (ed.), Individuation in Scholasticism: The Later Middle Ages and the Counter-Reformation, 1150–1650 (Albany, NY: State University of New York Press, 1994), 271–98. Zimmerman, Dean. “Bodily Resurrection: The Falling Elevator Model Revisited,” in G. Gasser (ed.), Personal Identity and Resurrection: How to Survive your Death (Farnham: Ashgate, 2010), 33–50. Zimmerman, Dean. “The Compatibility of Materialism and Survival: The ‘Falling Elevator’ Model,” Faith and Philosophy 16 (1999), 194–212. Zimmerman, Dean. “Criteria of Identity and the ‘Identity Mystics’,” Erkenntnis 48 (1999), 281–301. Zimmerman, Dean. “Immanent Causation,” Philosophical Perspectives 11 (1997), 433–71.

Duns Scotus and Divine Necessity Richard Cross

In what follows, I aim to answer two questions: Does Duns Scotus hold that ‘God exists’ is metaphysically necessary? And, does he hold that ‘God exists’ is, as the medievals would say, self-evident (roughly, a priori)? As we shall see, the answer to both questions is affirmative, and thus Scotus holds that ‘God exists’ is logically necessary. ( Just to be clear: when I use modern senses of metaphysical and logical necessity (as I do in these opening sentences, and will do by and large in what follows), I assume that metaphysical necessity is that the opposite of which includes a contradiction, and that logical necessity is that subset of metaphysical necessity restricted to that the opposite of which can be known a priori to include a contradiction. And equivalently for the other modalities.) Although Scotus believes that it can be shown that there is a concept in virtue of which ‘God exists’ is self-evident, he also believes that human beings in the current run of things can have no access to the content of this concept. An inquiry about Scotus’s views on divine necessity gains some purchase from the fact that, while Scotus has well-developed and explicit accounts of logical and metaphysical possibility that are quite close to our modern ones, his modal proof for God’s existence, which makes use of a notion of divine necessity, does not use the notions of logical or metaphysical necessity, but some other kind of necessity that is probably entailed by these kinds of necessity, but that certainly does not entail them. So I begin with a brief account of some of the varieties of modality in Scotus, including a brief discussion of divine necessity in the proof for God’s existence, and then turn to my two main questions.

1. VAR IETIE S O F M ODALI T IES According to Scotus, logical possibility is a property of propositions, namely, that the proposition includes no logical incompossibility in its content:

Duns Scotus and Divine Necessity

129

Logical potency (potentia logica [elsewhere possibilitas logica]) is a certain mode of the composition made by the intellect, caused by the relation between the terms of that composition, namely, that they are not repugnant [to each other].1

Claiming that things are “not repugnant” to each other is, in this passage, a way of claiming that they are not incompossible with each other. The extension of ‘logical possibility’ here is akin to our notion of logical possibility, but a little narrower than it, because Scotus considers among logical incompossibilities in his sense cases such as ‘a human being is not rational’—something that modern versions of essentialism might incline us to think of as a metaphysical but not a strictly logical impossibility—akin to ‘Water is not H2O’: a posteriori, not a priori.2 In what follows, I generally talk about logical and metaphysical modalities in our modern sense, using ‘metaphysical’ necessity and cognates to refer to the necessary a posteriori. As it turns out, Scotus does have something like this distinction too, though he does not talk about it using quite this terminology. But there are some wrinkles, as we shall see. Scotus’s metaphysical possibility is a little more complex, since he classifies various different kinds of modality under this heading:3 [Metaphysical potency] is understood in three ways. [1] In one way it is opposed to the impossible, not as it implies the way of composition . . . but as it means the disposition of something non-complex. . . . And in this way the possible is co-extensive with the whole of being, nothing is a being whose notion includes a contradiction. [2] In another way potency is understood as it is opposed to the necessary . . . ; and in this way the necessary is said to be what has of itself indefectible entity, and possible being that [has] defectible [entity]. [3] In the third way metaphysical potency is taken most strictly, as it is incompatible with act in the same respect (circa idem).4

1 Quaestiones super libros Metaphysicorum Aristotelis [= In metaph.], IX, qq. 1–2, n. 18 (Opera philosophica [= OP], ed. G. Etzkorn et al., 5 vols (St Bonaventure, NY: Franciscan Institute, 1997–2006), IV, 514). On Scotus’s role in the development of the notion, see Simo Knuuttila, Modalities in Medieval Philosophy (London: Routledge, 1993), 139–49. For ‘possibilitas logica,’ see Ordinatio [= Ord.] I, d. 36, q. un., n. 61 (ed. C. Balic´ et al., 21 vols (Rome: Vatican Press, 1950–2011), VI, 296), and Lectura [= Lect.] I, d. 39, qq. 1–5, nn. 49–50 (Vatican, XVII, 494–5), in both of which passages it is used interchangeably with ‘potentia logica’; also e.g. Lect. I, d. 2, p. 2, qq. 1–4, n. 188 (Vatican, XVI, 176). 2 See e.g. Scotus, Reportatio [= Rep.] I, d. 11, q. 2, n. 28 (ed. Wolter and Bychkov, I, 413–14). For an excellent discussion of the relation between Scotus’s notion of logical possibility and ours, see Calvin Normore, “Duns Scotus’s Modal Theory,” in T. Williams (ed.), The Cambridge Companion to Duns Scotus (Cambridge: Cambridge University Press, 2003), 146–7. 3 For all this, see my “Duns Scotus on Essence and Existence,” Oxford Studies in Medieval Philosophy 1 (2013), 190–2. 4 Scotus, In metaph. IX, qq. 1–2, n. 21 (OP, IV, 515); translation from my “Duns Scotus on Essence and Existence,” 190–1.

130

Richard Cross

The first of these is (as it were) the real counterpart of Scotus’s logical possibility: where logical possibility concerns propositions, the possibility referred to here concerns things. Claiming that it is a “disposition of something non-complex” merely amounts to claiming that it is a property of things, not of syntactic complexes. The second is irrelevant for our purposes here. But the third is not: this kind of metaphysical potency— that which is “incompatible with act in the same respect”—is elsewhere labeled by Scotus “objective potency”: Something is said to be in potency in two ways: in one way, because it is the end term of a power [potentia], or is that to which the power is directed, and this is said to be objectively in potency (as the Antichrist, or whiteness that is to be generated, is said in one way to be a being in potency). In the other way, something is said to be in potency as the subject of a potency, or as that in which the potency is. And in this way something is said to be in potency subjectively, because it is in potency to something but not yet perfected by it (as a surface that is to be made white).5

Subjective potency is the potency had by something that is not but can be ç. Scotus understands this kind of potency to be possessed only by things that have some kind of actuality. The reason is easy to find: what has subjective potency is what can be a substrate for properties—and any such substrate must have some kind of real actuality of its own.6 Objective potency is analogous to this: it is the potency attaching to something that is not at all but that will be. And such a thing lacks any kind of actuality—it is “incompatible with act.” Scotus uses the necessity opposed to this kind of potency in his proof for God’s existence.7 A first cause, if such a thing there be, cannot be brought into existence—it lacks objective potency (and thus is in this sense necessary). But—and here are two crucial modal premises—(major) “nothing can not-exist unless something positively or privatively incompossible with it can exist,” and (minor) any uncausable being (e.g. a first cause) is such that “there cannot be anything positively or privatively incompossible with it.” The conclusion is that such a being cannot not-exist: it is non-existence, not existence, that requires explanation, and there could be no explanation for the non-existence of a first cause. But how should we understand the modalities? Incompossibility here is something like physical or factual incompatibility—so some kind of nomological impossibility. Thus, Scotus attempts to prove the major premise by 5 Scotus, Lect. II, d. 12, q. un., n. 30 (Vatican, XIX, 80); translation from my “Duns Scotus on Essence and Existence,” 191–2. 6 See Scotus, Lect. II, d. 12, q. un., nn. 37–8 (Vatican, XIX, 82). 7 For the whole discussion in the next two paragraphs, see Scotus, Ord. I, d. 2, p. 1, qq. 1–2, n. 70 (Vatican, II, 170–1).

Duns Scotus and Divine Necessity

131

arguing that “no being can be destroyed other than through something positively or privatively incompossible with it.” So the argument is that, if there is no real power incompatible with a first cause, then such a cause exists. The proof of the minor premise (that there is no such power) is that such a power would be either self-existent (a se; lacking objective potency) or “from another” (and thus with objective potency), and that neither option is available. The first option generates a causal paradox: two incompossibles together, or neither, and thus no explanation for the nonexistence of the first cause; and the second something similarly implausible: a self-explanatory object’s being destroyed by something that depends either on it or on some other self-explanatory object.8 Despite its ingenuity, there is probably not much to be said for this as an argument for God’s existence. But the point that I wish to emphasize here is that no part of the argument relies on Scotus’s notion of logical possibility, and the necessity that is concluded at the end is (in our sense) neither logical nor metaphysical—it is the much weaker notion of existing without being the end term of a power, or of being essentially uncaused. I note this as a temporary puzzle, as it were, since, as we shall see in section 3, Scotus is perfectly clear as to why he could not employ the notions of logical necessity/possibility in this context. Basically, he maintains that there is indeed a concept of God, such that ‘God exists’ is both analytic and a priori; but that we have no way, absent special divine revelation, to come to possess the relevant concept. What I have tried to highlight in this first section is that the notions of necessity that Scotus deems to be of utility in the proof for God’s existence—and thus the notions of divine necessity that one might most naturally associate with Scotus’s view—turn out to be notions that fall far short of logical or metaphysical necessity.

2 . I S ‘G O D E X I S T S ’ M E T A PH Y S I C A L LY N E C E S S AR Y ? With this much ground-clearing done, I now attempt to answer the first of the two questions that I set myself in my opening paragraph. For it turns out that, in addition to the accounts of divine necessity just examined, Scotus does indeed hold that ‘God exists’ is metaphysically necessary in our sense. This might seem anomalous. Calvin Normore, for example, once talked of the “Janus-faced character of Scotus’s modal thought,” in which we find the view that “the present is as contingent as the future” (at least from a logical point of view) coupled, perhaps incongruously, with the 8 For a fuller discussion of this, see Normore, “Duns Scotus’s Modal Theory,” 137–40.

132

Richard Cross

discussion of “logical possibilities in a framework in which intellects and real powers are in the background.”9 Now, as we have seen, Scotus clearly has the notion of metaphysical possibility in something like our sense (the non-contradictory). So he ought to have the cognate notion of metaphysical necessity too, even if he has other notions of necessity as well. And it turns out that he does, and that (as far as I can tell) he applies it in the divine case—indeed, as I show in the third section, he believes that ‘God exists’ is logically necessary. The anomaly is neutralized by the simple fact that Scotus does not believe that we currently possess a concept of God serviceable in any ontological argument. The context of the defense of God’s metaphysical necessity is strictly theological. Theologians of the Latin and Greek traditions disagree on the following rather abstruse Trinitarian question: whether the Holy Spirit is produced by the Father and the Son, or merely by the Father. The Western theologians accept the former claim (the so-called filioque); the Eastern theologians the latter. Anselm of Canterbury held that denying the filioque entails denying any distinction between the Son and the Holy Spirit, since what distinguishes the Holy Spirit from the Son is simply the fact that the Holy Spirit proceeds from the Son.10 It seems as though we could test Anselm’s view for cogency by thinking about the following conditional: if the Holy Spirit did not proceed from the Son, the Holy Spirit would not be distinct from the Son. But this conditional is problematic, since (so the Western theologians maintain) the antecedent seems to be not merely counterfactual but also impossible, and it is a basic logical rule that anything follows from an impossible antecedent. So it looks after all as though thinking about this conditional would not allow us to make any progress on the substantive question at all. This view of the logic of the question, indeed, was taken by some of Scotus’s contemporaries—according to William of Nottingham, it was held by John of Berwick and William of Macclesfield.11 Scotus’s clearest summary of the opponents’ view about the relevant impossibility, found in his last discussion of the topic (in the Reportatio), runs as follows: 9 Normore, “Duns Scotus’s Modal Theory,” 155–6. 10 This argument was commonly ascribed to Anselm, and does indeed seem to be presupposed in his argument in favor of the filioque: see Anselm, De processione spiritus sancti, ch. 1 (Opera omnia, ed. F. S. Schmitt (Edinburgh: Nelson, 1946–61), II, 181, ll. 2–4, and p. 182, l. 17–p. 184, l. 15). As Scotus points out, the argument is somewhat undermined by a claim that Anselm later makes in response to an objection, affirming in effect that different modes of production (by generation, by procession) are sufficient to distinguish the Son from the Spirit, irrespective of the filioque (see Anselm, De processione, ch. 1 (II, 151, ll. 3–15), discussed in Scotus, Ord. I, d. 11, q. 2, n. 38 (Vatican, V, 15)). 11 See the evidence offered in the Vatican edition of Scotus’s works, V, 10, footnote F.

Duns Scotus and Divine Necessity

133

Whatever is true in God is maximally necessary; therefore, by opposition, whatever is false is maximally impossible—and this is the incompossible.12

I call this principle Q (from the first word in the Latin: “Quidquid est in divinis verum est summe necessarium; ergo per oppositum quidquid est falsum est summe impossibile, et tale est incompossibile”). It is clear that, as the opponent understands the situation, the incompossible here is in fact the logically impossible in our sense. The opponent wants to defend the view that the entire question is fruitless, since anything follows from an impossible antecedent, and the opponent uses Q to show this. Since the question is fruitless only if the antecedent is logically impossible in our sense, the incompossible is the logically impossible. So the maximally necessary must be the logically necessary—otherwise we could not infer incompossibility “by opposition” to the relevant notion of necessity. What the opponent has in mind is the top row of a modal square of opposition: the contraries necessary and impossible. All of the modalities in such a square must of course be of the same kind. And what is logically necessary is clearly metaphysically necessary too. Now, Q is the opponents’ principle; but Scotus objects not to Q but to the thought that Q renders any kind of reasoning from an impossible premise pointless. Scotus offers two alternative accounts. In one—the only one to appear in his earlier treatment of the issue—he maintains that the premise denying the filioque does not involve a logical impossibility (in our sense)—a semantic inconsistency or explicit contradiction: something that “immediately (statim) includes contradictories in its very concept.”13 The idea, in other words, is that Scotus believes Q to express a metaphysical but not a logical impossibility. And, he avers, only the logically contradictory entails any and every proposition. The second account, proposed only in the later treatment, is slightly different. It allows that Q expresses a logical impossibility, but suggests a relevance criterion that would allow us to reason from such a contradiction: Once there has been removed something that belongs to the essential notion of a thing, it is still possible to enquire about the thing in relation to some predicate which belongs to it not in virtue of the thing that has been removed. For example, man is rational, but rational does not distinguish a human being from a stone. So, if rational were removed from man, it would still be possible to enquire whether a 12 Scotus, Rep. I, d. 11, q. 2, n. 26 (I, 413). 13 Scotus, Ord. I, d. 11, q. 2, n. 29 (Vatican, V, 11); see too Rep. I, d. 11, q. 2, n. 28 (ed. Wolter and Bychkov, I, 413–14). For discussion, see Christopher J. Martin, “Formal Consequence in Scotus and Ockham: Towards an Account of Scotus’ Logic,” in O. Boulnois et al. (eds), Duns Scot a` Paris: actes du colloque de Paris, 2–4 Septembre 2002 (Turnhout: Brepols, 2004), 117–50.

134

Richard Cross

human being were distinguished from a stone in virtue of animal or in virtue of something else.14

The premise is “man is not rational.” Scotus does not believe that we could actually assert such a premise (I give evidence for this claim in section 3 below). But he seems to think that we can at least get as far as conceptually bracketing rational and animal, and reasoning on the basis of this kind of bracketing. Christopher J. Martin labels the procedure Scotus has in mind “radical circumscription,” allowing us to posit, for the sake of argument, without asserting, incompossibilia.15 All this, I take it, goes some way towards confirming my impression that Scotus thinks of ‘God exists’ as a metaphysically necessary truth: it is a truth about God (at least according to Scotus), and thus, according to Q, maximally necessary, and thus metaphysically necessary. The later account allows that it might be logically necessary too, by parity of reasoning. But this seems to be a mistake: as we shall see in section 3, self-evidence is an epistemic criterion, and has nothing to do with the ontological strength, so to speak, of the necessity involved. Elsewhere, Scotus draws the inference to the metaphysical necessity of ‘God exists’ explicitly—though admittedly without referencing Q explicitly. He relies instead on the following related principle: “existence does not belong to anything more perfectly than to this [viz. the divine] essence.” (I label this principle E, from ‘existence’ or ‘esse’, and refer to it again in section 3.) He infers from E that ‘God exists’ is both necessary and self-evident.16 A self-evident proposition here is one that is known though itself, and Scotus thinks of this requirement as being satisfied in the case that a proposition’s truth can be intuited by simple inspection of the concepts or objects involved. As I show in the next section, necessary propositions that are self-evident are logically necessary in our sense. So, according to Scotus, E entails not merely God’s metaphysical necessity, but also God’s logical necessity. (Again, this seems to be a mistake; but we might at least take E as entailing the metaphysical necessity of ‘God exists’.) Why accept Q? Scotus does not really say. He gives no extended treatment of Q: it just crops up in the course of a discussion of a quite different issue. But it looks as though Q could be derived by the Anselmian method of perfect-being theology: identifying pure perfections, things that

14 Scotus, Reportatio I, d. 11, q. 2, n. 29 (ed. Wolter and Bychkov, I, 414). 15 Martin, “Formal Consequence,” 135. 16 Other than when noted, the material in this paragraph derives from Scotus, Ord. I, d. 2, p. 1, qq. 1–2, n. 25 (Vatican, II, 137).

Duns Scotus and Divine Necessity

135

it is unqualifiedly better to be than not to be, and ascribing them to God.17 Thus, Q speaks of the “maximally” necessary—ascribing necessity to something in the highest degree. Scotus sometimes worries about the utility of the principle as a fully reliable way of establishing certain perfections, but never about its utility in showing that God must possess in the highest degree being and features co-extensive with being, including the highest degree of certain disjunctive properties co-extensive with being, among which he includes necessary-or-contingent.18 And why accept E? One reason Scotus has is scriptural: he takes it to be implied by Exodus 3.14, and thus credible at least on the basis of divine revelation.19 But it could, I think, be derived by perfect-being theology in a way analogous to Q—at least in E’s weaker reading (as entailing metaphysical, not logical, necessity). But Scotus in fact presents some kind of sketch of an argument for E by maintaining that ‘God exists’ is the proposition “to which all [those propositions which] enunciate something of God, howsoever conceived, are resolved.”20 Resolution is an analytical method that is supposed to result in the concept of an essence—a definition—so the point of claiming that the proposition is that “to which all” others are “resolved” is that such a proposition will amount to an essential definition of an object or kind.21 And if this does not strictly entail E, it is clearly in the general area, and perhaps could form a premise in an argument for E. But as it stands, this is nothing more than suggestive, not least since Scotus offers no defense of the “resolution” premise. But the argument does, incidentally, make Scotus’s God surprisingly like Aquinas’s, at least in this respect: what the divine essence most fundamentally is is God’s esse—his existence.22 17 See Anselm, Monologion, c. 15 (I, 28, l. 24–p. 29, l. 33), frequently endorsed in Scotus—e.g. Quodlibetum, 1 n. 8 (Opera omnia, ed. L. Wadding, 12 vols (Lyons, 1639), XII, 10). 18 See Scotus, De primo principio 4, n. 8 (ed. Wolter, 87–9). 19 See Scotus, Ord. I, d. 8, p. 1, q. 4, n. 198 (Vatican, IV, 264). 20 Scotus, Ord. I, d. 2, p. 1, qq. 1–2, n. 25 (Vatican, II, 137). 21 The details of the method are of no concern here: for discussion, see Scotus, Ord. I, d. 3, p. 1, qq. 1–2, nn. 80–1 (Vatican, III, 54–6); see too my Duns Scotus’s Theory of Cognition (Oxford: Oxford University Press, 2014), 69–74. We might note in passing that, as we can infer from things I discuss in section 3, Scotus’s claim that ‘God exists’ is logically necessary requires more than that ‘God exists’ is a definition of ‘God’ (other than in a rather specialized sense of ‘definition’—that is to say, tautology). 22 See e.g. Aquinas, Summa theologiae I, q. 3, a. 4 (ed. Caramello, I, 17a–b). There are nevertheless very significant differences. For example, Scotus defends a distinction between the divine essence/esse, on the one hand, and the divine attributes on the other (a version of a distinction that was a commonplace among the Greek Church Fathers and that Scotus has learned from John of Damascus): something that Aquinas would not at all endorse. On this, see my “Duns Scotus on Essence and Existence,” 187–90. And I return to another possible refinement to this account below.

136

Richard Cross 3 . I S ‘ G O D E X I ST S’ S E L F - E V I D E N T ?

Self-evidence, according to Scotus, is an epistemic property. Being selfevident is a property had by (and only by) propositions that can be known to be true by simple inspection of their terms: “a self-evident proposition is one that has evident truth from nothing other than its proper terms which belong to it.”23 In the case that the proposition is just an instance of conceptual analysis, such propositions are (in our sense) logically necessary, and their having the epistemic property of self-evidence is a consequence of this. The terms in such self-evident propositions have such identity that one evidently necessarily includes the other. . . . Just as the composition of the terms cannot obtain without truth, so an intuition (perceptio) of the composite (compositio) and an intuition of the terms cannot obtain without an intuition of the conformity of the composite with the terms, and thus an intuition of the truth [of the composite], because the first things that are intuited evidently include an intuition of that truth.24

Talk of the (syntactic) composition’s “conforming” with the terms is a way of saying that the terms themselves make the proposition true—and that we intuit its truth simply on the basis of our understanding of the terms. Scotus expressly excludes from the category of self-evident propositions those that we would label the necessary a posteriori or metaphysically necessary: definitions such as ‘a human being is a rational animal,’ or ‘water is H2O,’ which in this context he construes as being such that the subject and predicate have or express different intensions: That proposition which is merely known by the definition of the terms is not selfevident, because, since a proposition is self-evident only if it has its evidence from the knowledge of the terms, and since the definition and the word [defined] are different, it follows that it is not self-evident in relation to words that have evidence from the definition of the terms.25

23 Scotus, Ord. I, d. 2, p. 1, qq. 1–2 (Vatican, II, 131). 24 Scotus, Ord. I, d. 3, p. 1, q. 4, n. 230 (Vatican, III, 139). Strictly speaking, Scotus believes that there are contingent self-evident propositions: e.g. those about our own introspectively available mental states: see Ord. I, d. 3, p. 1, q. 4, nn. 238–9 (Vatican, III, 144–6). But in this case, of course, the terms of such propositions are not simply concepts, but rather contingent particular objects. I ignore these cases in what follows, since they make no difference to my argument. For discussion, see Roberto Hofmeister Pich, “Scotus on Contingent Propositions ‘Known through Themselves’ (per se notae),” in L. Honnefelder et al. (eds), Johannes Duns Scotus, 1308–2008: die philosophischen Perspektiven seines Werkes (Mu¨nster: Aschendorff, 2010), 277–306. 25 Scotus, Lectura I, d. 2, p. 1, qq. 1–2, n. 17 (Vatican, XVI, 116).

Duns Scotus and Divine Necessity

137

The idea is that, in definitions such as ‘water is H2O,’ we understand what is defined “confusedly”—under some or all of its propria and genus (wet, drinkable stuff, and so on)—and the definition “distinctly”—as an essence (H2O).26 And these are different concepts—different intensions. Even knowing the genus and all the propria would never be sufficient for an intuition of the essence. Scotus would regard these definitions as metaphysically necessary in our sense; but not a priori. In fact, definitions turn out to be systematically ambiguous between the logically and the metaphysically necessary. Definitions are a case of what the medievals refer to as propositions per se primo modo, and at one point, Scotus distinguishes no fewer than three types of such propositions, sorted according to the degree of conceptual containment involved: The highest [degree of difference] is when a part of the definition is predicated of what is defined; a lesser [degree] when the whole definition [is predicated] of what is defined; and the lowest when the same thing [is predicated] of itself.27

The third is a tautology (e.g. ‘H2O is H2O’), and in the first the predicate is part of the intension of the subject (e.g. ‘man is an animal’). The second is the case relevant here: precisely the same thing (e.g. H2O) is signified by two different words or phrases corresponding to distinct but necessarily coextensive intensions (‘water is H2O’, the terms of which signify respectively the relevant set of H2O’s propria, and H2O itself). The contrast between the second and third here is the contrast between the metaphysically and the logically necessary. The point of all of this is to allow for non-circular syllogistic reasoning in the case that the minor premise is simply a definition. The particular cases that troubled the medievals were ones such as the following: Every rational animal has a capacity for humor. Man is a rational animal. Man has a capacity for humor.

This is the so-called “demonstratio potissima,”28 in which the major premise predicates a proprium of a definition or essence—the so-called per se secundo modo—and the minor premise is the definition of the essence. Scotus’s way 26 See Scotus, Ord. I, d. 2, p. 1, qq. 1–2, nn. 18–19 (Vatican, II, 133–4)—the next displayed quotation. For the distinct/confused distinction, see Ord. I, d. 3, p. 1, qq. 1–2, nn. 71–91 (Vatican, III, 49–60). 27 Scotus, In metaph. IV, q. 2, n. 34 (OP, III, 328). 28 At least according to Aquinas and Simon of Faversham, whom Scotus follows. For the demonstratio potissima in Aquinas, and the controversies surrounding it, see the discussion in John L. Longeway, “Aegidius Romanus and Albertus Magnus vs. Thomas Aquinas on the Highest Sort of Demonstration (demonstratio potissima),” Documenti e

138

Richard Cross

of avoiding circularity is to claim that the minor premise can be genuinely informative in the way that he claims—metaphysically necessary but not logically necessary. In this way, the conclusion is not simply synonymous with the major premise.29 The disambiguation latent in a definition is easily made using the epistemic criterion of self-evidence: any relevant self-evident proposition is logically necessary in our sense. Scotus makes the point in a passage of daunting obscurity: Since a self-evident proposition is one that has evident truth from its terms, and [since] the term of a concept of the quiddity [conceived] distinctly (as it is conveyed by the definition) is different from the term of a concept of the quiddity [conceived] confusedly (as it is conveyed by a word), it follows that a proposition about a quiddity taken confusedly that is not known unless the same thing is distinctly conceived through a definition will not be self-evident. And this conclusion is proved, because otherwise any other proposition which is per se primo modo (such ‘man is an animal,’ and ‘a body,’ right up to substance) will be self-evident, because if the notion of each extreme were assigned from the notions of the extremes conceived distinctly, it would be manifestly apparent that one extreme includes the other.30

The first paragraph here reiterates the claim that propositions that are merely metaphysically necessary are not self-evident. The second paragraph attempts to prove this by arguing that, if it were not the case, there could be no such thing as the necessary a posteriori. Definitions are per se primo modo in virtue of conceptual inclusion; but self-evident only if this conceptual inclusion is known by some kind of immediate intuition (as in the case of the logically necessary). The whole point of the discussion just given is to exclude from the realm of the selfevident those propositions that are genuinely informative. Self-evident propositions are tautologies. Thus, just as with our a priori, Scotus treats self-evidence as a criterion for concept possession: the idea is that any intellect, grasping the terms, would intuit the truth of the proposition— and thus that someone not intuiting the truth of the proposition had

studi sulla tradizione filosofica medievale, 13 (2002), 373–434. See too the discussion in my Duns Scotus’s Theory of Cognition, 78–80, and the literature cited there. 29 Scotus, Ord. I, d. 2, p. 1, qq. 1–2, n. 17 (Vatican, II, 132). For extensive discussion of the role of the self-evident (but not per se) in syllogistic reasoning, see Peter C. Vier, Evidence and its Function according to John Duns Scotus (St Bonaventure, NY: Franciscan Institute, 1951), 66–97. See too Dominique Demange, Jean Duns Scot: la the´orie du savoir (Paris: Vrin, 2007), 77–114, 187–95. 30 Scotus, Ord. I, d. 2, p. 1, qq. 1–2, nn. 18–19 (Vatican, II, 133–4).

Duns Scotus and Divine Necessity

139

simply not grasped the terms. Such a person would be committed to affirming an evident contradiction, and such an affirmation is not possible: The argument is confirmed . . . by Aristotle, in Metaphysics IV, where he claims that the opposite of the first principle (i.e. ‘it is impossible for the same thing to be and not to be’) cannot enter into anyone’s intellect, “for then there would be contrary beliefs in the mind”: and this is certainly true about contrary beliefs, namely, ones that are formally incompatible, because a belief asserting being of something and one asserting non-being of the same thing are formally incompatible.31

By the same token, if it is not possible for one possessing the relevant concepts to intuit the truth of the proposition, then that proposition is not self-evident. We should note too that self-evidence is an intrinsic property of the proposition, had by it irrespective of the existence of things external to it (e.g. minds): We should not distinguish between a proposition that is known per se and one that can be known per se, because these are the same: for it is not because a proposition is known by some intellect that it is said to be self-evident (because in that case if no intellect actually cognized, no proposition would be self-evident); rather, a proposition is said to be self-evident because it naturally has evident truth contained in the terms, with respect to the nature of the terms, whatever intellect conceives the terms. But if some intellect does not conceive the terms, and thus does not conceive the proposition, the proposition is no less self-evident, with respect to itself: and it is thus that we speak of the self-evident.32

The discussion contrasts very strikingly with Aquinas’s. Aquinas seems to understand the self-evident as the analytic: such that “the predicate is included in the notion of the subject.”33 A subset of such propositions consists of those that are self-evident in themselves but “not to us (non quoad nos)”: analytic propositions whose truth can only be known a posteriori. ‘God exists’ is apparently a case in point: The proposition ‘God exists,’ as it is in itself, is self-evident, because the predicate is the same as the subject. . . . But because we do not know of God what his essence is, [the proposition ‘God exists’] is not self-evident to us, but needs to be demonstrated through those things that are more known to us.34

31 Scotus, Ord. I, d. 3, p. 1, q. 4, n. 231 (Vatican, III, 139), quoting Aristotle, Metaphysica , ch. 3 (1005b29–32). 32 Scotus, Ord. I, d. 1, p. 1, qq. 1–2, n. 22 (Vatican, II, 136). I do not want here to discuss the ontological status of the complexa that Scotus is talking about—as propositions, or as sentence-types, for example. 33 Aquinas, Summa theologiae I, q. 2, a. 1 c (I, 10b). 34 Ibid.

140

Richard Cross

So here the notion of self-evidence is in itself a logical notion, while the notion of self-evidence quoad nos is an epistemic one. But it follows straightforwardly from his account that Scotus should reject the distinction between what is self-evident in itself and self-evident to us, since this distinction elides the logical and the epistemic, things that he is at pains to separate. And this is precisely what he does: if a proposition is self-evident in itself, it is self-evident, since what makes it evident are simply the terms involved.35 So, given all this, is ‘God exists’ self-evident? Yes, because it is logically necessary in our sense: “This proposition . . . is per se primo modo and is immediately evident from its terms.”36 This is the (fallacious) argument I commented on briefly near the end of the last section: E entails (according to Scotus) that ‘God exists’ is metaphysically necessary and self-evident (and so logically necessary, in our sense). The relevant subject term—the concept signified by ‘God’—is God “as he is this thing,” or God’s essence and existence “under the most proper concept by which this existence is in God.”37 Anyone who possesses this concept will know ‘God exists’ simply by inspecting the concept: ‘God exists’, or ‘this [viz. the divine] essence exists’ is self-evident, because these extremes naturally make this composite evident to anyone who perfectly apprehends the extremes of this composite.38

Note that, if my account of logical necessity in Scotus is correct, the syntactic complex ‘God exists’ has exactly the same semantic value as the relevant syntactically simple term ‘God’: but this, I suppose, is in some sense just what it would be for a concept to include its own existence. (I return to this in a moment.) Scotus puts his account of self-evidence to use in trying to solve the following conundrum: given that ‘God exists’ is logically necessary, why is it that we do not know it a priori? The answer is that we do not possess the relevant concept. While, if we know both that God exists and that Q, or if we know both that God exists and that E, then we in principle know that ‘God exists’ is logically necessary, we do not know this a priori, since we lack a concept of God, such that we can infer God’s existence simply from that concept. (Rather paradoxically, then, we know only a posteriori that ‘God exists’ is a priori.) Scotus holds that the only concepts that we 35 Scotus, Rep. I, d. 3, q. 2, n. 62 (ed. Wolter and Bychkov, I, 201); see Ord. I, d. 2, p. 1, qq. 1–2, n. 23 (Vatican, II, 136). 36 Scotus, Ord. I, d. 2, p. 1, qq. 1–2, n. 25 (Vatican, II, 137). 37 Ibid. 38 Ibid.

Duns Scotus and Divine Necessity

141

naturally have of God are derived from creatures, and he does not see how any such concepts could be sufficient to allow us to grasp God’s existence simply by conceptual analysis. His specific reason is that concepts of God derived from creatures always involve some kind of conceptual complexity, such that the basic concepts involved are wholly disjoint—e.g. the compositional concept necessary being, neither of which component concepts includes the other, whether wholly or partly—and Scotus holds that mere conceptual analysis cannot tell us whether two disjoint concepts (e.g. necessary, being) are realized together in some real thing: None of these propositions—‘a necessary being exists,’ or ‘an agent actually exists’—is self-evident, because whether the parts that are in the subject are actually united is not self-evident. And when [the objector] says that the opposite of the predicate is incompatible with the subject, I say that it does not follow from this that the proposition is self-evident, or that this incompatibility is evident.39

The crucial thing here is that the concepts necessary being and God are disjoint: ‘God exists’ is self-evident to one possessing the relevant concept; ‘a necessary being exists’ is not.40 However much existence we pack into this kind of concept of God, that concept will be a complex composed of

39 Scotus, Ord. I, d. 2, p. 1, qq. 1–2, n. 33 (Vatican, II, 144). The point of the second of these propositions—‘an agent actually exists’—relates again to a rather remarkable link between Scotus and the Greek theological tradition, via John of Damascus: the thought that ‘God’ signifies not the divine essence but the external activity of the first being. Hence, Scotus claims, “ ‘God exists’ and ‘an agent actually exists’ are the same according to every sense (acceptionem)”—presumably in the case that ‘an agent’ is understood to signify the divine activity: Scotus, Ord, I, d. 2, p. 1, qq. 1–2, n. 32 (Vatican, II, 143). 40 This is not to say that concepts such as necessary being are not concepts of the divine essence. Scotus implicitly makes use of his distinction between the divine essence and the divine attributes to make the point. We can have plenty of concepts of the divine attributes (perfect wisdom, for example), and from these concepts we can infer that there is a being or essence that underlies the various attributes, which we can conceptualize as infinite being. For discussion of this, see Scotus, Ord. I, d. 3, p. 1, qq. 1–2, nn. 25, 58–60 (Vatican, III, 16–17, 40–2). But the concept infinite being, while it is indeed a concept of the divine essence, is not a concept of the divine essence as this essence in itself (Scotus, Ord. I, d. 3, p. 1, qq. 1–2 (Vatican, III, 38–9))—by which Scotus means that it is not the concept that would allow us to see that ‘God exists’ is logically necessary. Is there a reason why the relevant concept of God is inaccessible to us? Scotus believes that there is: God is the kind of object who is accessible only if God wills to make himself so: “God, as he is this essence in itself, is not naturally known by us, because he can be known, under this description, as a voluntary object, not as a natural one (other than in relation to his own intellect)”: Scotus, Ord. I, d. 3, p. 1, qq. 1–2, n. 57 (Vatican, III, 39). So the relevant concept can only be had by revelation. For discussion of the accessibility (or otherwise) of the divine essence to human beings in different states of existence, see my “Duns Scotus on Religious Experience,” in Richard Cross (ed.), The Opera theologica of John Duns Scotus: Proceedings of the “Quadruple Congress” on John Duns Scotus, Part 2 (Mu¨nster: Aschendorff, 2012), 89–111.

142

Richard Cross

disjoint simple concepts (e.g. the concept of an essence and the concept of an existence), and thus it will always be, from an epistemic point of view, a further issue, whether or not the complex concept is instantiated—that is to say, whether the two simple disjoint concepts are in fact jointly realized or not. Scotus’s claim that God is or includes his esse is supposed to show that there is a concept of God that somehow just is, or at least includes, a concept of the relevant kind of real existence.41 Equally, as Scotus notes, it is just poor logical form to argue on the basis of a putative contradiction (‘a necessary being does not exist’) that a necessary being exists. A further related point: given that ‘God exists’ is logically necessary, and that (necessarily) God is a necessary being, it seems that ‘a necessary being exists’ is thus necessary too. But the former is self-evident, and the latter not. And we know this last because we can possess the concepts of (logical) necessity and being but yet not know whether or not ‘a necessary being exists’

41 I say “is or at least includes” because there seems to be an ambiguity in Scotus’s presentation—indeed, an ambiguity of a very significant kind, given the historically fraught question of the relationship between God’s essence and God’s existence. I noted at the end of section 2 Scotus’s claim that ‘God exists’ is the proposition “to which all those which enunciate something of God, howsoever conceived, are resolved,” and thus that ‘exists’ is in some sense the definition of ‘God’. The ambiguity is this: should we in this case construe the term ‘existence’ and the concept existence (esse) as univocal or not? Claiming that ‘exists’ is the definition of ‘God’ pushes in the direction of supposing it to be simply equivocal. But claiming that we can show principles Q and E by natural reason, or even that they are intelligible components of revealed theology—as Scotus’s account seems to imply—pushes in the opposite direction, towards univocity. One proposal: ‘exists’ could be part of the definition of ‘God’ (and thus existence part of the concept God ) and still univocal (at least, given Scotus’s assumptions about the semantics of univocity), and this might be what we should say in order to reconcile the two conflicting trends in Scotus’s account. Existence, after all, is just factuality, and, plausibly, factuality seems univocal over necessarily and contingently factual things—albeit that the things themselves are very different kinds. Factuality is ascribed to things in degrees according to the necessity or contingency of their possession of that factuality—according to whether factuality is contained in the thing’s essence or not. That the concept God is more complex than the concept existence (which it includes as a proper part on this account) would certainly mark a step away from the identification with Aquinas that I adumbrated at the end of the second section. One further question in this regard: could ‘existence’ and existence themselves be analogical? Not, I assume, in Scotus’s technical semantic sense of being overlapping meanings or concepts. After all, what could motivate replicating at the level of the concept existence the conceptual complexity involved in those concepts of God that are available to us in this life (e.g. infinite being)? Doing so would give us a further, core sense of ‘existence’ shared by God and creatures, and modified by different disjunctive concepts in the respective cases (perhaps infinite/finite): and this seems conceptually baroque and otiose. Of course, Scotus assumes that there is a proportion between God and creatures (see e.g. Scotus, Ord. I, d. 8, p. 1, q. 3 (Vatican, IV, 191–2)—I am grateful to Andrew Helms for bringing this passage back to my attention): but that is a metaphysical matter, not a semantic one, and has no bearing on the content of the concept of existence. For Scotus on analogy, see my “Duns Scotus and Analogy: A Brief Note,” The Modern Schoolman 89 (2012), 147–54.

Duns Scotus and Divine Necessity

143

is true. Scotus makes a somewhat similar point in the context of a discussion of a slightly different necessary proposition, ‘an infinite being exists’: This proposition, ‘an infinite being exists,’ is not evident to our intellect from its terms. I prove this: we do not conceive the terms [to be united] before we believe it [viz. the proposition] or come to know it by demonstration. And in that prior time it is not evident to us. For we do not hold it with certainty, from its terms, other than through faith or demonstration.42

Here we have a criterion that allows us to deny that ‘an infinite being exists’ is a priori: we can fully grasp the terms while yet not having sufficient reason to believe that the proposition is true. All this, incidentally, explains why Scotus does not think that the notion of logical necessity is of any service in an attempt to prove the existence of God. That ‘God exists’ is logically necessary is the conclusion of a set of arguments; we lack the relevant concept that would enable it to function as an a priori premise. It might sound wildly implausible to suppose that there could be a concept of God that would allow someone to infer God’s existence simply from that concept, but I think we should say rather just that it is a little mysterious. After all, the strangeness of this form of theism, from a philosophical point of view, is that there could be an analytic existential claim. From this, it seems but a small step to posit that if there is such a claim, it should in principle be possible to know that claim a priori. Certainly, there is no concept known to us which is such that mere inspection of the concept can convince us that the concept is instantiated. But the fact that a bit of perfect-being theology yields a curious result in philosophy is not, by itself, grounds to reject the theology.43 University of Notre Dame

BIBLIOGRAPHY Anselm of Canterbury. Opera omnia, ed. F. Schmitt (Edinburgh: Nelson, 1946–61). Aristotle. The Complete Works of Aristotle: The Revised Oxford Translation, ed. J. Barnes (Princeton: Princeton University Press, 1984). Cross, Richard. “Duns Scotus and Analogy: A Brief Note,” The Modern Schoolman 89 (2012), 147–54. 42 Scotus, Ord. I, d. 2, p. 1, qq. 1–2, n. 28 (Vatican, II, 140). 43 Thanks to Richard Swinburne for commenting on an earlier version, and for getting me to think about the whole issue in the first place. A referee for Oxford Studies gave helpful comments on an earlier draft.

144

Richard Cross

Cross, Richard. “Duns Scotus on Essence and Existence,” in R. Pasnau (ed.), Oxford Studies in Medieval Philosophy, vol. 1 (Oxford: Oxford University Press, 2013), 172–204. Cross, Richard. “Duns Scotus on Religious Experience,” in R. Cross (ed.), The Opera theologica of John Duns Scotus: Proceedings of the ‘Quadruple Congress’ on John Duns Scotus, Part 2 (Mu¨nster: Aschendorff, 2012), 89–111. Cross, Richard. Duns Scotus’s Theory of Cognition (Oxford: Oxford University Press, 2014). Demange, Dominique. Jean Duns Scot: la the´orie du savoir (Paris: Vrin, 2007). John Duns Scotus. De primo principio, ed. A. Wolter (Chicago: Franciscan Herald Press, 1982). John Duns Scotus. Opera omnia, ed. C. Balic´ et al. (Rome: Vatican Press, 1950–2011), 21 vols. John Duns Scotus. Opera omnia, ed. L. Wadding (Lyons, 1639), 12 vols. John Duns Scotus. Opera philosophica, ed. G. Etzkorn et al. (St Bonaventure, NY: Franciscan Institute, 1997–2006). John Duns Scotus. Reportatio I-A, ed. A. Wolter and O. Bychkov (St Bonaventure, NY: Franciscan Institute, 2004–8), 2 vols. Knuuttila, Simo. Modalities in Medieval Philosophy (London: Routledge, 1993). Longeway, John. “Aegidius Romanus and Albertus Magnus vs. Thomas Aquinas on the Highest Sort of Demonstration (demonstratio potissima),” Documenti e studi sulla tradizione filosofica medievale 13 (2002), 373–434. Martin, Christopher J. “Formal Consequence in Scotus and Ockham: Towards an Account of Scotus’ Logic,” in Olivier Boulnois et al. (eds), Duns Scot a` Paris: actes du colloque de Paris, 2–4 Septembre 2002 (Turnhout: Brepols, 2004), 117–50. Normore, Calvin. “Duns Scotus’s Modal Theory,” in T. Williams (ed.), The Cambridge Companion to Duns Scotus (Cambridge: Cambridge University Press, 2003), 129–60. Pich, Roberto Hofmeister. “Scotus on Contingent Propositions ‘Known through Themselves’ (per se notae),” in L. Honnefelder et al. (eds), Johannes Duns Scotus, 1308–2008: die philosophischen Perspektiven seines Werkes (Mu¨nster: Aschendorff, 2010), 277–306. Thomas Aquinas. Summa theologiae, ed. P. Caramello (Turin: Marietti, 1952–6). Vier, Peter C. Evidence and its Function According to John Duns Scotus (St Bonaventure, NY: Franciscan Institute, 1951).

Ockham’s Scientia Argument for Mental Language Eric W. Hagedorn

1. INTRODUCTION William of Ockham famously held that, in addition to written and spoken language, there exists a mental language, a structured representational system common to all thinkers (human and angelic), containing both atomic representations (so-called “mental terms”) and molecular representations (including “mental sentences” and “mental syllogisms”).1 Ockham’s account of mental language has been much studied, but there has been very little discussion of Ockham’s reasons for positing mental language in the first place.2 In what follows, I present a line of argument by which Ockham seeks to establish the existence of mental language, an argument which to this point has been uniformly overlooked by the secondary literature. In the first half of the paper I briefly present Ockham’s account of mental language and examine a set of texts which, when taken together, show Ockham arguing that positing a mental language is the only way a nominalist can meet certain ontological constraints imposed by Aristotle’s account of scientific demonstration. In the second half of the paper, I discuss and evaluate Ockham’s argument in greater detail. 1 Throughout, I use the following abbreviations when discussing Ockham’s works: Ordinatio in librum primum Sententiarum [Ord.], Reportatio in libros II, III, IV Sententiarum [Rep.], Brevis summa libri Physicorum [BrevSumma], Expositio in libros Physicorum [ExpPhys], Expositio in libros Perihermenias [ExpPeri], Summa logicae [SL], Quodlibeta septem [Quod.]. All references to Ockham’s works are to the standard critical editions, Opera theologica [OTh] and Opera philosophica [OPh]. 2 The secondary literature on Ockham’s mental language is quite large; to start with, see Calvin Normore, “Ockham on Mental Language,” in J. C. Smith (ed.), Historical Foundations of Cognitive Science, Philosophical Studies Series 46 (Dordrecht: Kluwer, 1990), 53–70; Claude Panaccio, Ockham on Concepts (Burlington, VT: Ashgate Publishing, 2004); and John Trentman, “Ockham on Mental,” Mind 79 (1970), 586–90.

146

Eric W. Hagedorn 2. A N U N REC O GN IZED AR GU MENT FOR O C K H A M ’S M E N T A L L A N G U A G E

Ockham presents his theory of mental language in a number of seemingly unconnected passages, scattered throughout his corpus; but the main lines of his theory seem to be as follows.3 In addition to conventional human languages such as English, Latin, and Mandarin, there exists a single mental language which is shared by all human beings (and also by non-human thinkers such as angels). Mental language, like other languages, is a representational system; in addition to sentential representations (so-called mental sentences or mental propositions), it also contains both subsentential representations (mental terms or concepts) and even supersentential representations (mental syllogisms). Mental language has its own grammar and syntax, according to which mental terms are combined into mental sentences and mental sentences are combined into mental syllogisms. The atomic elements of mental language essentially signify whatever they signify; the signification of any composite element of mental language, by contrast, is a function of the signification of its components together with its syntax. The utterances of conventional languages, on Ockham’s picture, have the significations they do in virtue of the relationships they bear to the elements of mental language. Furthermore, this representational system serves as an underlying basis for our cognitive activity; thinking, according to Ockham, is just “speaking in mental language.”4 If one searches the secondary literature on Ockham’s mental language, seeking an answer to the question of why Ockham posits this system of representations in the first place, one will discover a common refrain: Ockham’s acceptance of mental language largely rests on a few supposed proof-texts, largely taken from Aristotle, Augustine, and Boethius. Normore accurately represents the consensus of the scholarly literature when he states that Ockham simply does not argue for his view: “Early fourteenth century thinkers like Burley and Ockham do not argue for the [Mental Language] Hypothesis but suggest that it is the natural way to understand such writers as Aristotle and Augustine . . . .”5 Despite this silence among

3 Ockham’s longest sustained discussion of the central doctrines of his theory of mental language is SL I.1–13 (OPh I, 7–47). Here I ignore a number of scholarly controversies regarding the precise details of Ockham’s theory. 4 Quod. I.6 (OTh IX, 36). 5 Calvin Normore, “The End of Mental Language,” in J. Biard (ed.), Le langage mental du Moyen Aˆge a` l’Aˆge Classique (Leuven: Peeters, 2009), 297.

Ockham’s Scientia Argument for Mental Language

147

most scholars,6 I take it that Ockham does take himself to have a reason for positing mental language, and, I believe, this reason is presented several times throughout his corpus. It makes perhaps its most explicit appearances in the prologues to Ockham’s various commentaries on Aristotle’s Physics. In the prologue to the Brevis summa libri Physicorum, for instance, Ockham gives the following account of the subject matter of Aristotelian science: [P1] It must be noted that this science—just as with all others—is about universal [concepts] and non-complex [cognitions], and not of real things. For if it were of real things, these would be either universal real things or particular real things: but it is not of universal real things, because there are no such things (as is proved in Metaphysics VII); nor is it of singular real things (as is also proved in Metaphysics VII, and is frequently proved elsewhere). Thus, this science is of concepts.7

In P1, Ockham briefly lays out an argument of the following form: Aristotelian science must be about either real (i.e., extra-mental) things or concepts. It cannot be about extra-mental singulars, for reasons that would be familiar to Aristotelians: singulars are not necessarily existing entities and do not have the requisite intelligibility to be the subjects of scientific knowledge. Furthermore, Ockham argues, science also cannot be about extra-mental universals, for (as Ockham takes himself to have shown repeatedly and at length in his corpus) there are no extra-mental universals. So, by process of elimination, Aristotelian science must be about concepts. Ockham expands upon this consideration in the Prologue to his much longer Expositio in libros Physicorum, saying a bit more about the system of concepts he claims is required by Aristotelian science. There he states that the strictures of Aristotelian science also require that there be mental 6 Claude Panaccio, by contrast, does claim that Ockham has an argument for mental language in “Mental Language and Tradition: Encounters in Medieval Philosophy: Anselm, Albert, and Ockham,” Vivarium 45 (2007), 262–82. However, it is a rather strange argument on its face. On Panaccio’s account, Ockham argues that one who already accepts the existence of mental sentences on the word of Boethius should thereby posit mental terms as well: “Ockham’s [argument], consequently, is the following: since there are three sorts of complex discursive units, as Boethius says, there must be, accordingly, the same three sorts of simple significant units” (279). As an argument for mental language, this looks to be question-begging: I know of no thinker who accepts that there are (or even might be) sentential mental representations but who rejects the existence of subsentential representations. But perhaps the argument is intended to convince those who accept the existence of both mental sentences and mental terms but do not believe that mental sentences are composed of mental terms. (Thanks to an anonymous referee for alerting me to this possibility.) The argument would then not be circular; yet, it still appears highly inadequate, as the passage in question has nothing at all to say about compositionality or why one who admits the existence of mental sentences should conclude that they are composed of mental terms. 7 BrevSumma, Prologue, ch. 2 (OPh VI, 5).

148

Eric W. Hagedorn

sentences containing subject and predicate terms, and that the terms of those sentences must possess the semantic property of suppositing for other entities: [P2] It must be known that a real science is not of real things, but is of concepts suppositing for real things, because the terms of the mental sentences which are known supposit for real things. Hence, in the known mental sentence every heat is calefactive, a concept common to every heat is made the subject and supposits for every heat . . . .8

Taking both P1 and P2 together shows Ockham claiming that positing a mental language—a representational system of mental sentences with concepts as constituents, concepts which supposit for extra-mental things—is a way the nominalist about universals can fulfill the requirements of Aristotelian demonstrative science. I will argue in what follows that Ockham does not merely take mental language to be one way the nominalist can meet the requirements of Aristotelian science; rather, I take it that throughout many of his discussions of mental language, Ockham asserts that mental language is the only way for the nominalist to meet certain ontological constraints laid down by Aristotelian science, such as the requirement that the subjects of such science be imperishable. Taken as a whole, then, Ockham’s scattered writings on mental language present the following argument for mental language: The Scientia Argument for Mental Language ML1. To fulfill the strictures of Aristotelian science, one must posit either extra-mental universals or a representational system that meets certain constraints.9 ML2. The only representational system that can meet these constraints is a mental language. ML3. All theories of extra-mental universals are incoherent. ML4. Therefore, to fulfill the strictures of Aristotelian science, one must posit a mental language. It is quite possible that Ockham is implicitly drawing on Aristotle when he argues in this fashion; though I’m not aware of any place where Ockham explicitly cites the following passage, in the Posterior Analytics Aristotle

8 ExpPhys, Prologue, sec. 4 (OPh IV, 12). In this essay, I follow a convention of using small capitals to name elements of mental language. 9 The precise nature of these constraints will be explained in section 4.

Ockham’s Scientia Argument for Mental Language

149

himself suggests that Aristotelian science merely requires positing the right kind of representational system, rather than a theory of universals: [P3] There need not be any forms, or some one item apart from the many, in order for there to be demonstrations. It must, however, be true to say that one thing holds of many.10

Ockham’s Scientia Argument does not just appear in his Physics commentaries, though; years before he prepared these commentaries, Ockham drew a similarly close connection between the requirements of Aristotelian science and the need for mental language in his Ordinatio on the first book of the Sentences. To provide further evidence that Ockham does in fact argue this way, it is to this passage that I must now turn.

3 . T H E SC I E N T IA AR GU M E N T IN D I S T I N C T I O N 2 OF THE ORDINATIO In the second distinction of Ockham’s Ordinatio, Ockham attempts to systematically examine every possible account of extra-mental universals that can be given and to show that each such account is ultimately false or incoherent.11 He proceeds by moving from what he sees as the most fullfledged realism about universals—namely, that they are real extra-mental items which are really distinct from particulars and are identical to each of their instantiations12—to increasingly less realist positions: that universals are really distinct from particulars yet are multiplied according to their instantiations;13 that universals are only formally distinct from and so exist in ordinary particulars;14 and, lastly, that universals are in no way distinct 10 Aristotle, Posterior Analytics I.11, 77a5–7. 11 Ockham concludes his discussion by claiming that he has shown “that no real thing outside the soul—either in and of itself, or by means of something else added to it (whether real or merely conceptual), and no matter how it is considered or thought about—is a universal. Thus it is as impossible that some real thing outside the soul be in any way universal . . . as it is impossible for a man to be an ass” (Ord., d. 2, q. 7; OTh II, 248–9). 12 Among medieval authors, such a view seems to have been rare, though Paul Vincent Spade attributes such a view to Walter Burley in Five Texts on the Mediaeval Problem of Universals (Indianapolis: Hackett, 1994), 115 n. 1. 13 Ockham claims that this position was (mistakenly) attributed to Scotus by some of his contemporaries. The editors of the critical edition of Ockham’s Ordinatio note that William of Alnwick, a student of Scotus and contemporary of Ockham, attributed such a position to Scotus in Alnwick’s Sentences II, d. 3, q. 1. (See Ockham, Ord., d. 2, q. 5; OTh II, 154, note 1.) 14 This is the position of Scotus himself. See his Ordinatio, d. 3, p. 1, qq. 1–6 (Vatican VII, 391–494).

150

Eric W. Hagedorn

from ordinary particulars, but nevertheless there is still “something in some way really universal and common in extra-mental reality.”15 Near the beginning of this extended treatise on universals, Ockham considers an indispensability argument that seems to have had a certain amount of purchase for his contemporaries; call this The Indispensability Argument for Universals. The Indispensability Argument is meant to establish that there are real extra-mental universals; as presented by Ockham, the argument goes as follows: [P4] Secondly, some argue that a real science is about true real [universal] things outside the soul, because this is what distinguishes a real science from a rational science. And since no science is primarily of real singular things (as is clear according to the Philosopher in Posterior Analytics book I and Metaphysics book VII), thus there are some real [universal] things outside the soul in addition to the singular real things.16

The “real sciences” [scientia realis] are those that are about extra-mental reality, such as physics or theology; these are contrasted with “rational sciences” [scientia rationalis] such as logic and grammar, which concern themselves with words and other signs. Now, the Indispensability Argument given in P4 alleges that the character of a real science straightforwardly entails the existence of extra-mental universals, since such a science must be about extra-mental universals.17 The word ‘about’ is ambiguous here, though. As the Scholastics understood Aristotle’s Posterior Analytics, a science is a collection of scientific demonstrations; as Ockham and his contemporaries recognize, such a science may be about a given item (or items) in the sense that the item is a subject of that science or in the sense that the item is an object of that science.18 A subject of a given science is an item of which the demonstrations of that science predicate various attributes; in Aristotle’s words, the subject is that “whose attributes—i.e., the items incidental to it in itself—the demonstrations make plain.”19 In other words, a subject is ultimately what the predications of 15 Ord., d. 2, q. 7 (OTh II, 225). 16 Ord., d. 2, q. 4 (OTh II, 103). 17 Cf. ExpPhys, Prologue, sec. 4 (OPh IV, 11–12). 18 My terminology here follows Ockham: “There is a difference between the subject [subjectum] and the object [objectum] of a science, because a subject of a science is the subject of a conclusion, but the object of a science is that which is known and which terminates the act of knowing” (Ord. I, Prologue, q. 9; OTh I, 266). This usage is not universal among Ockham’s contemporaries. For just one example, Aquinas, in his commentary on Boethius’s De trinitate, uses the terms objectum and materia when he is talking about the subject matter of the sciences (see Super Boetium de trinitate, part 3, q. 5, a. 1, resp.). Aquinas does, however, use subjectum in the discussion in Summa theologiae Ia, q. 1, a. 7, when he asks whether God is the subjectum of the science of sacred doctrine. 19 Aristotle, Posterior Analytics I.7, 75b1–2.

Ockham’s Scientia Argument for Mental Language

151

the science are about. An object of a science, by contrast, is an item which actually comes to be known via scientific demonstration; it is whatever item is the direct object of the act of knowledge that results from considering the demonstration. So while a subject of science is what the predications are about, an object of science is what the knowledge of the science is about.20 Given this, it is not hard to see why the objects of the sciences were commonly seen to be some kind of propositional item, while the subjects of sciences would be non-propositional. The objects of science are propositional because propositional entities are the direct objects of acts of knowing; the subjects are non-propositional, however, since the subject of an act of predication is typically some kind of non-propositional item. Given this distinction between the subjects and objects of science, then, the Indispensability Argument quoted in P4 is intended to show that the subjects of the real sciences must be extra-mental universals. This reasoning seems to have had a broad appeal with the Scholastics: Henry of Harclay attributes this very argument to “many,”21 and the conclusion that the subjects of Aristotelian science are extra-mental universals was held in some fashion or other by many Scholastics, including Aquinas, Scotus, and Henry of Ghent. But, of course, as Ockham rejects the existence of any sort of extra-mental universal, he cannot accept the upshot of this argument. Nor is he satisfied with the response given by Harclay, who says that ordinary particulars are the subject of science, but only when considered confusedly and abstractly.22 Rather, Ockham responds to the Indispensability Argument of P4 as follows: [P5] I say that real science is not always23 of real things (speaking of the things which are immediately known), but is of other things that only supposit for the real things. In order to understand this, and because many things have been and will be

20 The need for such a distinction arises in part from the fact that the question of what scientia is about is ambiguous, as the Latin word ‘scientia’ can mean either a scientific discipline or an instance of knowledge. 21 The Indispensability Argument can be found in the Ordinary Questions of Henry of Harclay, written, at most, a handful of years before Ockham’s Ordinatio; Harclay claims that the argument has been given “by many” (Question XIV, n. 18; in Henninger, 604–5). Though Ockham himself presents this argument as a consideration in favor of a very strong version of imminent realism about universals (namely, that universals are really distinct—and thus separable—from their instances), the fact that Harclay ascribes this argument to many indicates that it was seen as establishing a far weaker form of imminent realism, given how few of their contemporaries held such a strong position. Indeed, Harclay himself presents the argument as a consideration in favor of Scotus’s more moderate view. 22 See Henninger, 661–3. 23 Ockham’s wording here suggests that real science is sometimes about real things and sometimes about concepts (“things which supposit for other things”). This seems to me

152

Eric W. Hagedorn

said by some who are untrained in logic, it must be known that every science, whether real or rational, is only of sentences regarding those things which are known, because only sentences are known. A sentence, however (according to Boethius in Book I of his commentary on De interpretatione), has threefold being, namely, in the mind, in speech, and in writing. That is to say, some sentences are only conceived and understood, others are spoken, and others are written.24

Ignore for the present the discussion of which things are “immediately known” and examine the justification Ockham gives for introducing mental sentences here. It may seem that Ockham is just engaging in a bit of prooftexting: it is natural to read Ockham as saying here that “according to Boethius” there are sentences in the mind, and thus Ockham sees it as legitimate to utilize them in his philosophical speculations. But this, I contend, is not the force of this passage at all. Part of the problem with understanding the argumentative force of passage P5 is that the numerous parentheticals and rhetorical asides in the passage can distract from Ockham’s main point; if we strip all that away, we are left with a much more straightforward claim: [P5*] . . . I say that real science is . . . [of] things that only supposit for real things. In order to understand this . . . it must be known that every science . . . is only of sentences . . . [However], some sentences are only conceived and understood, others are spoken, and others are written.

Ockham’s claim here is that all the subjects of science are “things that supposit” (namely, words and/or concepts), while all the objects of science are sentences (and as I will make clear later, mental sentences in particular). But how is this supposed to work? Ockham fills in the answer just a few paragraphs later: [P6] Yet because the terms of some sentences stand and supposit personally, namely for the external real things themselves (as in the mental sentences ‘every mobile thing is partly in the terminus a quo [and partly in the terminus ad quem],’ ‘every human is risible,’ ‘every triangle has three [sides],’ and so on), there is said to be real science of such sentences. The terms of other mental sentences supposit simply, namely for the concepts themselves (as in ‘every demonstration is from true first [principles],’ ‘human is a species,’ and so on); and so there is said to be rational science of such sentences. So then, it does not matter at all for real science whether the terms of the known mental sentence are real external things or are only in the soul, provided that the terms stand and supposit

contrary to everything else he has to say on the matter; one should thus not read too much into the ‘always’ here. 24 Ord., d. 2, q. 4 (OTh II, 134).

Ockham’s Scientia Argument for Mental Language

153

for real external things; and so it is not necessary on account of real science to posit any such universal things really distinct from singular things.25

What distinguishes a real science like physics from a rational science like logic, on Ockham’s view, is not that the real sciences have extra-mental universals for subjects, while the rational sciences have concepts for subjects; rather, the difference between them rests upon the different modes of supposition had by their subjects: The subject (or subjects) of a real science supposits personally in the conclusions of that science, while the subject(s) of a rational science supposits simply in the conclusions of that science. To put it casually, the concepts which are the subjects of physics refer to extramental particulars, while the concepts which are the subjects of logic refer to intra-mental particulars; and this, Ockham thinks, is sufficient to distinguish the real sciences from the rational sciences. I must now take a moment to quickly summarize and make explicit the outlines of the theory Ockham ultimately constructs concerning how mental language is supposed to be an ontology for Aristotelian science. After seeing this fuller description of the theory, we will be better equipped to assess Ockham’s Scientia Argument for mental language. According to Ockham, an Aristotelian science like geometry or physics is a collection of scientific demonstrations. A demonstration is a certain kind of syllogism—namely, a syllogism which has two necessary truths as premises and which brings about knowledge of a previously unknown necessary truth26—and syllogisms will themselves turn out to be elements of mental language that are composed of mental sentences.27 25 Ord., d. 2., q. 4 (OTh II, 137). 26 “In the beginning it must be known that, according to the teaching of Aristotle, a demonstration is a syllogism producing knowing. . . . [Here] ‘knowing’ is taken for evident comprehension of a necessary truth brought about by evident comprehension of two necessary truths (placed in the proper mood and figure), so that those two truths make the third truth (which otherwise would have been unknown) known evidently” (SL III-2, ch. 1; OPh I, 505–6). I ignore here for ease of presentation Ockham’s parenthetical comment, which refers to his view that only syllogisms in the modes Barbara (a syllogism having two universal affirmative premises and a universal affirmative conclusion) and Celarent (a syllogism having one universal affirmative premise, one universal negative premise, and a universal negative conclusion) count as demonstrations. 27 “ . . . passions of the soul . . . as well as mental sentences, syllogisms, and all universals, are nothing other than certain ficta in the soul having only objective being (that is, being cognized), really existing nowhere” (ExpPeri, Prologue, sec. 10; OPh II, 370). Cf. Ord., Prologue, q. 8 (OTh I, 218–19): “I say that there can be a single habit of both principles and a conclusion. This is proved, for of what there is apt to be a single act, there can be one habit. But there can be one act with respect to the premises and conclusions, because it is no more repugnant that a syllogism composed from many sentences be understood by one act than that a sentence composed from many terms [be understood by one act]; but a sentence is understood by a single act; therefore, etc.”

154

Eric W. Hagedorn

With this account of the nature of demonstration, Ockham turns his attention to the subjects and objects of both the scientific demonstrations and the scientific discipline as a whole. An object of scientific demonstration, Ockham decides, cannot be anything other than that which is proved by the demonstration, and what is proved is the sentence that serves as the argument’s conclusion. So, the object of scientific demonstration must be a mental sentence.28 Likewise, the subject of such a demonstration is, Ockham tells us, nothing other than the subject term of the argument’s conclusion, since the subject term is that which is having attributes predicated of it.29 For example, with respect to Aristotle’s canonical example of scientific demonstration (“the planets are not far away, and what is far away does not twinkle, therefore the planets do not twinkle”), the object of the demonstration is the mental sentence the planets do not twinkle, and the subject of the demonstration is just the mental term planets.30 Finally, Ockham extends this line of reasoning to the science as a whole: since a science just is a collection of scientific demonstrations, an object of a science—that at which the science as a whole is directed—cannot be anything other than an object of one of those individual demonstrations. Thus, he concludes that a science such as astronomy has many objects; the objects of astronomy are all the mental sentences demonstrated in that science, such as the planets do not twinkle, the moon is spherical, the stars are far away, and so on. Similarly, to be a subject of a science is nothing other than to be the subject term of the conclusion of one of demonstrations contained in the science; in stark contrast to his predecessors, then, Ockham 28 This is a subtle point that some of Ockham’s interpreters have either missed or at least not made explicit, implying that Ockham’s view is merely that sentences (of whatever kind) are the objects of knowledge. For instances of this, see Gordon Leff, William of Ockham: The Metamorphosis of Scholastic Discourse (Manchester: Manchester University Press, 1975), 320–7: the objects of knowledge are “conclusions” or “propositions” (320–7); Armand Maurer, The Philosophy of William of Ockham in the Light of its Principles (Toronto: Pontifical Institute of Mediaeval Studies, 1999), 142: “the object of a science [is] the whole proposition”; and even the writer of the Ockhamist Tractatus de principiis theologiae (OPh VII, 538): “the known conclusion is the object of a science.” Ockham’s colleague Walter Chatton has it right when he attributes to Ockham the view that the object of scientific knowledge is a “complex in the intellect” (Prologus, q. 1, art. 1; ed. Wey, 21); see also the discussion in Susan Brower-Toland, “Ockham on Judgment, Concepts, and the Problem of Intentionality,” Canadian Journal of Philosophy 37 (2007), 67–110. 29 “ . . . the subject of scientific knowledge is the subject of the conclusion, while the object of a science is that which is known and terminates the act of knowing. This, however, is the conclusion which is known” (Ord., Prologue, q. 9; OTh I, 266). See also ExpPhys, Prologue, sec. 3 (OPh IV, 9): “The object of scientific knowledge is the whole known mental sentence, and the subject is part of that sentence, namely, the subject term.” 30 Aristotle’s example comes from Posterior Analytics I.13, 78a30–b4.

Ockham’s Scientia Argument for Mental Language

155

denies that each science has but a single subject.31 Rather, a science has as many subjects as there are distinct subject terms in the science’s conclusions. Thus, the subjects of astronomy include (but are not limited to) mental terms such as planets, moon, and stars; and where Aquinas might have insisted that the sole subject of geometry is magnitude, Ockham claims that geometry has as subjects the concepts triangle, square, polygon, angle, hypotenuse, line, and point.32 With that summary of his theory in hand, recall that I claimed that these passages from the Ordinatio and from his Physics commentaries (namely, passages P1, P2, and P5) are a manifestation of what I named the Scientia Argument, whereby Ockham is arguing for the existence of mental language: The Scientia Argument for Mental Language ML1. To fulfill the strictures of Aristotelian science, one must posit either extra-mental universals or a representational system that meets certain constraints. ML2. The only representational system that can meet these constraints is a mental language. ML3. All theories of extra-mental universals are incoherent. ML4. Therefore, to fulfill the strictures of Aristotelian science, one must posit a mental language. But given that no other scholars have identified such an argument in Ockham’s works, let alone in such well-known passages as the second distinction of the Ordinatio, why think that this is how Ockham is ultimately arguing here? For several reasons: The first is simply that Ockham seems to indicate that this is exactly what he is doing; he concludes his

31 “In the way in which the Philosopher takes ‘subject’ in the Posterior Analytics, the very same thing is the subject both of the conclusion and of the science; and it is called the subject [of the science] only because it is the subject of the conclusion” (ExpPhys, Prologue, sec. 3; OPh IV, 9). 32 “So I say that the nature of a subject is nothing other than to be made the subject of some predicate in a sentence known by scientific knowledge properly speaking, so that universally the same thing and under the same ratione is the subject of a science and the subject of a scientifically known conclusion” (Ord., Prologue, q. 9; OTh I, 247–8). See also ExpPhys, Prologue, sec. 3 (OPh IV, 9): “The same thing is the subject of a conclusion and of the science; for nothing is called a subject except because it is the subject of a conclusion. And so when there are many conclusions having many subjects, . . . then of that science which is aggregated from all the instances of scientific knowledge of those conclusions, there is not some one subject, but there are many subjects of its many parts.” For Aquinas’s opinion on the subject of geometry, see his Expositio Posteriorum, book I, lectio 15.

156

Eric W. Hagedorn

reply to the Indispensability Argument by stating that what he has done by positing mental language is to provide an alternative ontology for Aristotelian science, one that does not require positing extra-mental universals (thus Ockham in P6: “ . . . so it is not necessary on account of real science to posit any such universal things really distinct from singular things”). Second, Ockham regularly links the theory of mental language to the demands of Aristotelian science, especially in his scientific works (as in passages P1 and P2, for example). In those passages, Ockham reiterates that taking the objects of science to be mental sentences and the subjects of science to be concepts is intended to replace the positing of extra-mental universals in these roles. Lastly, Ockham’s philosophical methodology seems to commit him to provide some reason for positing mental language; his eponymous razor states that entities should not be posited unless they are vindicated by experience, religious authority, or demonstrative argument.33 As such, it would appear to be a significant lacuna in his philosophical project, were he to entirely refrain from providing an argument for mental language. Given that the discussion in the second distinction of the Ordinatio, from which passages P4, P5, and P6 have been taken, is the first time in his corpus that he presents all the core elements of the theory of mental language, this would be a reasonable place to expect such an argument to make an appearance.34 I conclude that Ockham indeed does argue this way; the natural question to consider then is whether the Scientia Argument is sound. Assessing the truth of premise ML3 by determining whether or not there is some coherent realist theory of universals is far beyond the scope of this article.35 Instead, I will focus my attention on ML1 and ML2; my project in what remains will be to discuss the reasons Ockham gives for claiming that only

33 See, among other places, Ord., d. 30, q. 1 (OTh IV, 290). 34 The only extant texts in Ockham’s corpus that are earlier than the Ordinatio are his Reportatio on Books II–IV of the Sentences; but mental language is largely unmentioned in those texts, apart from a few remarks embedded in a discussion of angelic communication (see Rep. II.16; OTh V, 359–81). Mental language also makes a very brief appearance in the third question of the Prologue to the Ordinatio, alongside a promise that a fuller explanation of the theory will come later; this promissory note is presumably being redeemed in Ordinatio d. 2 (see Ord., Prologue, q. 3; OTh I, 134–5). 35 Even determining the extent to which Ockham is successful at disproving the specific theories he considers to be incoherent would be a project in itself. For attempts to do just this, see Marilyn McCord Adams, William Ockham, 2 vols (Notre Dame, IN: University of Notre Dame Press, 1993), chs 1–2, and Martin Tweedale, Scotus vs. Ockham: A Medieval Debate over Universals, 2 vols (Lewiston, NY: Edwin Mellen Press, 1999).

Ockham’s Scientia Argument for Mental Language

157

mental language is the right kind of representational system to fulfill the requirements of Aristotelian science.

4 . T HE O N T O L O G Y O F A R I S T O T E LI A N SC I E N C E Aristotle’s Posterior Analytics sets forth a laundry list of requirements for scientific demonstration. Here I focus on just three requirements that constitute the main constraints on the ontology required for science:36  (C1)The subjects of scientific demonstration are imperishable.37  (C2)In a scientific demonstration, it is demonstrated that some attribute necessarily belongs to the subject of the demonstration.38  (C3)The objects of scientific demonstration are necessary truths.39 The first premise of the Scientia Argument, ML1, claims that these three constraints from the Posterior Analytics can be met only by either a theory of universals or an appropriate representational system. It is not all that difficult to see how these constraints might be satisfied by a theory of universals. On such a scheme, the subjects of science are imperishable universals and a scientific demonstration indicates necessary connections among these universals: for example, by proving that risibility necessarily belongs to humanity. Thus, the object of demonstration—what becomes known—is this necessary connection between the attribute and the subject; the knowledge of this connection is thus knowledge of a necessary truth.40

36 What Aristotle lays down as constraints for scientific demonstration in particular come to be widely seen as requirements for science in general. Thus, Aristotle’s stipulation that the subjects of scientific demonstration must be imperishable is taken to also be claiming that the subjects of science must be imperishable, and similarly it is thought that the objects of science must be necessary truths, just as the objects of scientific demonstrations are supposed to be. Although most Scholastic authors do not postulate any difference among the subjects/objects of a scientific demonstration, the subjects/objects of scientific knowledge, and the subjects/objects of the scientific discipline as a whole, there is at least the logical space to think that these could differ. 37 “There is no demonstration of perishable things . . . because nothing holds of them universally but only at some time and in some way” (Posterior Analytics I.8, 74b24–6). 38 “Since in each kind whatever holds of something in itself and as such holds of it from necessity, it is clear that scientific demonstrations are concerned with what holds of things in themselves . . . ” (Posterior Analytics I.6, 75a29–32). 39 “If there is understanding simpliciter of something, it is impossible for it to be otherwise” (Posterior Analytics I.2, 71b16–17). Also: “What is understandable in virtue of demonstrative understanding will be necessary” (Posterior Analytics I.4, 73a21–3). 40 I do not assert that any Scholastic author held precisely this position on the ontology of science; but the basic idea that the subjects of science are some kind of

158

Eric W. Hagedorn

(The resulting theory is in some ways evocative of David Armstrong’s view that scientific laws express necessity relationships among properties.41 On such a view, even the demonstration itself might be seen, not as a set of linguistic sentences, but rather as some real extra-mental collection of these ontological connections among universals.) It is a bit more difficult, though, to see how a representational system like a language is supposed to be able to satisfy these constraints; explaining this (and showing why Ockham thinks only mental language can ultimately succeed at doing so) will be the aim of the remaining sections of this essay. Before turning to that project, though, one thing must still be done to justify ML1: ML1 claims that either a theory of universals or an appropriate representational system are the only ways to meet the ontological constraints of Aristotelian science; some justification must be given for the exclusion of any other possibilities. The main ontological alternative that Ockham sees as relevant here is a proposal held by his contemporary Walter Chatton, who held that the objects of science were non-sentential objects in the world, rather than linguistic items like sentences.42 Ockham’s response to Chatton’s proposal is rather straightforward: Ockham simply notes that non-sentential objects lack a key property that the objects of science must have. For the objects of science to be necessary truths, they must be the sort of thing that can be true or false, and no non-sentential object can have a truth value. Rather, Ockham insists, only sentences can be true or false.43 Of course, Ockham’s response is not the end of the story here. Many of his contemporaries did, in fact, think that ordinary particulars can be true or false; that is a core component of the slogan that being, truth, and goodness are convertible. Adjudicating this debate would require an essay all its own; but Ockham does have additional considerations to marshal here. Any attempt to provide an ontology for science which does not posit either extra-mental universals or some sort of representational system will have to say that the subjects of science are particulars, and that the predications made by scientific demonstrations indicate relations between universal entity (whether common natures, formal aspects of particulars, ideas in God’s mind, etc.) is one shared by Aquinas, Giles of Rome, Henry of Ghent, and Duns Scotus. 41 See David Armstrong, What is a Law of Nature? (Cambridge: Cambridge University Press, 1983). 42 Thus, Chatton: “An act of believing, as well as an act of knowing and an act of opining . . . have an external real thing for their object, and not a sentence [complexum]” (Prologus, q. 1, art. 1; ed. Wey, 20–1). 43 “[The act of knowing scientifically] is a complex act, having a complex for its object, because this act is one by which something true is known. . . . Therefore, only what is true is an object of scientific knowledge” (Quod. III.8; OTh IX, 234–5).

Ockham’s Scientia Argument for Mental Language

159

substances and their properties, rather than merely being relationships between linguistic elements (as Ockham takes them to be). But in response, Ockham offers two reasons for thinking that the predications in scientific demonstrations must be merely linguistic. First, he claims that predications are sentences in which two terms are connected by the ‘is’ of predication. For predication to be a metaphysical relationship, then, it would have to be that non-linguistic items in the world could enter into the structure of a sentence; but this leads to absurd consequences.44 Second, Ockham points out that at least some of the predications his contemporaries agree with must be merely linguistic, for there are predications which are true of God, but since God is simple, there are no accidents which inhere in God. Thus, a predication such as ‘God is wise’ must be merely a linguistic relationship, rather than a metaphysical one relating God and wisdom. And so, Ockham appears to argue, if theology requires us to state that some of our predications are linguistic, parsimony considerations should lead us to think that all predications are linguistic.45 So then, if an ontology of ordinary particulars cannot meet the ontological constraints of science for the reasons just given, what remains is the disjunction given in ML1: a theory of universals is needed unless an appropriate representational system can be supplied. In what remains, I will at last turn my attention to premise ML2 of the Scientia Argument, explaining why Ockham thinks that conventional spoken and written languages fail to meet the ontological constraints and a mental language must instead be posited to serve as the ontology of Aristotelian science.

44 Ockham provides eight reasons for denying that non-linguistic items can enter into the structure of a sentence in Quod. III.12 (OTh IX, 246–50). These eight arguments vary widely in quality; the most interesting among them are the second and the eighth. Ockham’s second argument claims that if ordinary objects could enter into the structure of a sentence, then a sentence could be a human person, since it could contain a body and soul, and anything composed of a body and a soul is a human being. The eighth argument argues that if ordinary objects could enter into the structure of a sentence, then a sentence could contain God as a part; but necessarily God is not a part of any further object. It should be noted that, in his earliest works, Ockham did believe that ordinary particulars could enter into the structure of a mental sentence, a position he seems to have taken from Walter Burley. Ockham came to reject this view by the time he begin revising his lectures on the Sentences into his Ordinatio on book 1. For a study of Ockham’s about-face here, see Elizabeth Karger, “Mental Sentences According to Burley and to the Early Ockham,” Vivarium 34 (1996), 192–230. 45 So far as I know, Ockham never gives this argument in full. But in several places he seems to provide an enthymeme of this argument; see, for instance, SL I.37 (OPh I, 104–6): “Of God there are predicated attributes proper to him; but there are no other real things inhering in him; and so an attribute is not some real thing inhering in its subject.”

160

Eric W. Hagedorn 5. MENTAL SENTENCES AS THE OBJECTS OF SCIENTIFIC DEMONSTRATION

Ockham has two main lines of argument for his claim that mental language is the only kind of representational system that can meet the ontological constraints for Aristotelian science. The first of Ockham’s reasons against conventional representational systems argues for the claim that mental sentences must be posited to be the objects of scientific demonstration because mental sentences are the only kind of sentences that could be, strictly speaking, necessarily true; the latter argues for the claim that mental terms must be posited to be the subject terms of such demonstrations because mental terms are the only kinds of signs that could ever justifiably be called imperishable. I will begin by considering his case that the objects of science must be mental sentences. First, though, one might well wonder why I keep speaking of sentences rather than propositions. Why does Ockham think that the only representations that can be true or false are sentences, rather than, say, propositions (where propositions are taken to be necessarily existing abstract objects that are the fundamental bearers of truth and falsity)? He thinks this in part due to his thorough-going nominalism: he straightforwardly rejects the existence of any such abstracta.46 Given that his nominalism is already a central presupposition of the Scientia Argument—those who are not nominalists are surely going to balk at premise ML3—perhaps we can just grant Ockham the non-existence of propositions for the sake of argument. With that worry set aside, at last we reach the centerpiece of Ockham’s case for mental language: even if we accept that the objects of science must be sentences, we may still wonder why they must be mental sentences, rather than spoken or written ones. If Ockham cannot show that the constraints of Aristotelian science cannot be met by conventional languages alone, then the argument we have been considering falls apart. So here it is especially important to proceed slowly and carefully. As I mentioned above, Ockham’s 46 In this respect, Ockham differs from some of his later followers such as Adam Wodeham and Gregory of Rimini, both of whom take the objects of knowledge to be some sort of abstracta that provide the content for mental sentences. But Ockham seems not to be aware of the arguments that the thinkers of the next generation will provide for thinking that some additional entity over and above sentences is needed to be a fundamental truth-bearer. Following Gregory, this abstractum has come to be known as a complexe significabile (literally, “something that can be signified complexly”). For more on Wodeham and Gregory’s views, see Wodeham’s Lectura secunda, d. 1, q. 1 (ed. Ga´l and Wood, vol. 1, 180–208) and Gregory’s Lectura super primum et secundum Sententiarum, Prologue (ed. Trapp et al., vol. 1).

Ockham’s Scientia Argument for Mental Language

161

reason for thinking that mental sentences are the only kind of sentences that can be the objects of science is that he thinks mental sentences are the only kind of sentences that can be necessary in a primary sense. Why, according to Ockham, are only mental sentences necessary? This is a view he does not, to the best of my knowledge, ever explicitly defend at length, but I take the following to be an accurate synthesis of some scattered remarks he makes on the topic.47 Let’s begin by considering a common contemporary line of thought, one that Ockham explicitly rejects. Typically, we think that a necessary truth is something that is true in all possible circumstances; but this can be only if (some of) the fundamental truth-bearers are objects that exist necessarily. For if it is necessary that every prime number greater than two is odd, then it must be that the truthbearer with this content exists in all possible worlds; otherwise, there would be a possible world in which it is the case both that every prime number greater than two is odd and that the truth-bearer ‘Every prime number greater than two is odd’ fails to be true (because it does not exist). But this would be a possible situation in which a necessary truth is possibly not true—which seems absurd. This is the line of argument often presented for thinking that propositions must be necessarily existing entities.48 Since Ockham believes that the only necessarily existent object is God— and thus that truth-bearers must be contingent entities—he rejects this argument for the claim that any truth-bearers exist necessarily.49 He evades the argument by rejecting the initial claim that necessary truths are truths that are true in all possible circumstances; in place of this criteria, he proposes that a necessary truth is “[a sentence] that is true and can never be false.”50 That is, a necessary truth is a sentence which correctly 47 For the sources of the reasons I give below, see Quod. II.19 (OTh IX, 193–7), Quod. III.13 (OTh IX, 251–3), Quod. V.9 (OTh IX, 513–18), and Quod. V.24 (OTh IX, 479–85). The question for Ockham, whether a spoken or written sentence can be necessary in a primary sense, is closely connected to the question whether spoken or written sentences are true in any primary sense; and in these passages he offers a number of reasons for thinking that spoken sentences cannot be the primary truth-bearers: among them, that the semantic contents of any spoken sentence is a contingent matter and that there is never a moment of time in which a spoken sentence exists in its entirety, since it is a successive entity, rather than a permanent one. 48 For further discussion on this argument and its place in both historical and contemporary discussions, see Marian David, “Defending Existentialism?” in M. Reicher (ed.), States of Affairs (Frankfurt: Ontos Verlag, 2009), 167–209. 49 “ ‘Necessary’ or ‘cannot be otherwise’ can be taken in two ways; in one way, it means what cannot not-be, and this kind of necessity is not required for scientific knowledge, because nothing is necessary in this way other than God alone” (BrevSumma, Prologue, ch. 2; OPh VI, 6). 50 “In what way then is a conclusion necessary? It must not be thought that it is necessary because it is always actually true in the way in which it is apt to exist in actuality

162

Eric W. Hagedorn

represents the world whenever it exists. Alternatively, a sentence fails to be necessary only if there is a possible circumstance in which the sentence exists and is false. (It should be noted that Ockham’s criterion here is faulty; this cannot be the correct criterion for being a necessary truth. Consider the following mental sentence: at least one mental sentence exists. This is a sentence which is true whenever it exists; there is no possible circumstance in which this sentence exists and is false. According to Ockham’s criterion, then, it would be necessary that at least one mental sentence exists. But this cannot be correct; all mental sentences are contingent entities for Ockham, and God could have refrained from creating any of them at all. Thus, it is false that this mental sentence is a necessary truth, contrary to Ockham’s proposed criterion.51) But let’s ignore this problem for now and assume that Ockham is correct in his definition of necessary truth. Why think that this definition can only correctly attach to mental sentences? Why cannot spoken and written sentences be necessarily true? Well, Ockham thinks, there is no spoken or written sentence which possesses the property of correctly representing the world in any possible circumstance in which it exists; and this is because both spoken and written sentences have their content by convention, while mental sentences have their content essentially.52 Thus, even a spoken sentence such as ‘Every prime number greater than two is odd,’ which we typically take to be a necessary truth, is not, strictly speaking, necessary. Since the sentence has its content merely by convention, there are possible circumstances in which that sentence exists and misrepresents the world, circumstances such as that in which the word ‘odd’ is subordinated to the concept even. (except perhaps in the divine intellect); rather, it is necessary because it is true, and it can never be false” (Ord., Prologue, q. 8; OTh I, 222). Compare SL II.9: “ . . . a sentence is called necessary not because it is always true, but because it is true if it exists and cannot be false” (OPh I, 275). Also, BrevSumma, Prologue, ch. 2: “What is called ‘necessary’ is that which cannot be false, although it could not exist” (OPh VI, 6). 51 This argument succeeds only if there are not mental sentences in God’s mind. I find no reason to think Ockham posits that God has mental sentences, but Paul Vincent Spade disagrees (on a philosophical, if not textual, basis); see his Thoughts, Words, and Things: An Introduction to Late Mediaeval Logic and Semantic Theory (, last modified December 27, 2007), 124. I take it that Buridan is aware of this kind of problem with Ockham’s criterion, and this is at least part of Buridan’s basis for distinguishing between being possible and being possibly true. See Buridan, Sophismata, ch. 8, Third Conclusion (tr. Klima, 954–5). 52 Ockham expresses this point by stressing that “the same mental sentence cannot be both true and false at the same time,” and that this is not the case for spoken propositions, since any given spoken word may have a multiplicity of meanings which varies the truth value of the corresponding sentence. See Quod. V.9 (OTh IX, 517).

Ockham’s Scientia Argument for Mental Language

163

If mental sentences are the only possible bearers of necessary truth, then they must be the objects of scientific demonstration. But does Ockham’s argument work? It seems to me that anyone acquainted with contemporary philosophy of language would point out here that Ockham simply needs to distinguish between the context of utterance and the circumstance of evaluation; ‘Every prime number greater than two is odd’ is necessary, such a philosopher would say, because, according to the meanings those words have in our context, they represent correctly in any possible world. This may seem to be a powerful objection to Ockham’s claim that only mental sentences can be necessary. Even if Ockham were correct on the definition of what it is for a sentence to be necessary (which he is not), he would still be on shaky ground regarding what the bearers of necessary truth must be. But perhaps Ockham’s line of thought is more potent than it seems; after all, the distinction between context of utterance and circumstance of evaluation relies on the notion of the meaning of a word in our context—that is, the distinction demands that we have an account of word meanings, where meanings are entities which can be fixed even when discussing other possible worlds. But Ockham’s point here just is that the only entities which can be necessary in a primary sense must have their content essentially across worlds—and mental sentences (and their component mental terms) just are supposed to be those meanings which are fixed from world to world. So then, he does not actually seem to be disagreeing with the contemporary philosopher of language. His claim that mental sentences are the only representations that could be the primary bearers of necessary truth is one part of Ockham’s case that only mental language meets the ontological constraints C1–C3. But he also has another reason to think that mental language can satisfy the demands of Aristotelian science and that conventional language cannot; namely, that the subjects of science cannot be spoken or written terms because such terms are not imperishable. It is to this final consideration I now turn.

6. M EN T AL T E R M S AS T H E S UB J EC T S OF SCIENTIFIC DEMONSTRATION We have seen why Ockham takes the objects of scientific demonstrations and the sciences themselves to be mental sentences. If we were to accept this view, it would not be unnatural to then conclude that the subjects of scientific demonstrations and scientific disciplines will be the subject terms of those mental sentences. But Ockham offers additional evidence as well: Ockham’s central claim regarding the subjects of the sciences is that

164

Eric W. Hagedorn

they must be mental terms because such terms are the relata of the kinds of predication involved in scientific demonstrations. In a demonstration, Aristotle teaches, an attribute (a passio, in Ockham’s Latin) is predicated of a subject.53 As we have already seen, Ockham argues that predication is a relation that occurs between linguistic terms (whether spoken, written, or mental); thus, the conclusion of a demonstration must predicate one term (the attribute) of another term (its subject).54 But even if we accept Ockham’s reasoning that predication is a relation among terms, why should we accept that the predications involved in scientific demonstrations are predications among mental terms? This move is based on Ockham’s contention that attributes (taking ‘attribute’ for “what is predicated in a scientific demonstration”) have to be mental terms: . . . ‘attribute’ can be taken in multiple ways. In one way, it can be taken for some real thing which is said to coincide with another real thing. In a second way, ‘attribute’ is taken for that which is predicated of something in the second way of per se predication. . . . ‘Attribute’ is not typically taken [in the first way] in the sciences. . . . An attribute [in the second way of taking ‘attribute’] is but a concept predicable of another [concept] which stands for a real thing (and does not stand for itself).55

But this looks to be merely a stipulation on Ockham’s part; he states that ‘attribute,’ in the sense of scientific predications, should be taken to refer to concepts which are only (truly) predicable of other concepts which supposit personally.56 Why though should we accept this usage of the term, rather than taking ‘attribute’ also for spoken and written words which are predicable in this way? Here Ockham seems to offer no specific argument. The section on attributes in the Summa logicae offers no help, beyond telling us that even though logicians commonly apply the term ‘attribute’ to written or spoken terms, this way of speaking is not wholly precise: . . . According to the way ‘attribute’ is used by the logician . . . an attribute is some mental or spoken or written predicable, predicable per se in the second way of that subject of which it is called the attribute. However, properly and strictly speaking, an attribute is nothing other than a mental predicable (and not a spoken or written one); yet secondarily and improperly a spoken or written word can be called an attribute.57 53 Ockham appeals to this, noting that “It is commonly said that a subject is that of which properties and attributes are demonstrated” (Summulae, Preamble; OPh VI, 141). 54 See SL I.32 (OPh I, 94–5). 55 Ord., Prologue, q. 3 (OTh I, 133–4). 56 This is the meaning of the clause that an attribute is “a concept predicable of another [concept] which stands for a real thing (and does not stand for itself).” 57 SL I.37 (OPh I, 104–5).

Ockham’s Scientia Argument for Mental Language

165

Though Ockham does not state his reasoning here, perhaps we can try to infer what his reasoning must be. For Ockham, as for his Scholastic contemporaries, predication is primarily a function of the judicative faculty; it is a judgment which is made by a rational being. But judgments can be made even when one does not have the language to express them; when I walk into an exotic pet store, I can form a judgment in which I predicate a certain color of the exotic bird before me, even if I don’t know the English word for that color (and even if English has no word for that precise color).58 In such a predication, since I don’t know the appropriate English terms (or no such term exists), it must be that the terms being predicated are mental terms. This line of reasoning may indicate to Ockham that predication is primarily a relationship among mental terms; as such, mental terms would seem to be the appropriate items for the predications involved in scientific demonstrations. But even if mental terms are the most suitable choice for the attributes and subjects of the predication in scientific demonstrations, it is not at all clear that mental terms satisfy the first of Aristotle’s three ontological constraints, namely, that the subjects of science must be imperishable. If we accept Ockham’s reasoning up to this point that the subjects of demonstrations are mental terms and that mental terms are contingent things, it seems he cannot accommodate Aristotle’s claim that the subjects of the sciences are imperishable things. Ockham even admits as much: [P7] From this it is clear that—although it contradicts the sayings of Aristotle— according to the truth no sentence made from terms which convey only contingent things, and which is affirmative, categorical, and about the present, can be the principle or conclusion of a demonstration, because such a sentence is contingent.59

Ockham’s precise point in passage P7 is that—strictly speaking—there cannot be scientific knowledge concerning certain terms which supposit for contingent things, but the point transfers to the terms themselves; there are no imperishable things other than God in Ockham’s universe, and so Aristotle’s dictum cannot be preserved according to its literal truth. Yet, even though the literal truth of Aristotle’s first constraint cannot be saved, Ockham nevertheless believes that mental language comes close enough to preserving Aristotle’s vision, since the subjects of science will be concepts such as the concept horse, which would represent horses even if there were

58 Thus, Ockham: “Mental sentences belong to no language in such a way that many people frequently internally form sentences that they nevertheless do not know how to express because of a lack of language” (SL I.12; OPh I, 42). 59 SL III-2, ch. 5 (OPh I, 512–13).

166

Eric W. Hagedorn

horses. Thus, Ockham continues on after passage P7 to affirm that concepts can be the subjects of scientific demonstrations after all: . . . even though genera, species, and every universal distinct from the cognition of God are simply contingent in such a way that they could be nothing, yet of them there can be demonstrations and scientific knowledge, because even though they could be simply destroyed, yet necessary sentences can be formed of them, which can be known by scientific knowledge strictly speaking.60

Thus, Ockham claims, mental terms come close enough to fulfilling Aristotle’s dictum to suffice for science. Unlike spoken and written terms, whose semantic value is fixed by arbitrary convention, mental terms essentially signify whatever they signify. Whenever horse is tokened in a mind, it always and everywhere signifies horses, even if no horses exist. Furthermore, unlike spoken terms, which exist only for a moment when uttered, and unlike written terms, which can be blotted out or erased, mental terms exist so long as there are created minds tokening them. In a world full of humans and angels, then, mental terms are about as imperishable as anything in Ockham’s world could ever be. Thus, there can be scientific demonstrations even when the subject terms of those demonstrations are, strictly speaking, perishable; mental terms are near enough to imperishable that, Ockham claims, they are sufficient to meet the ontological constraints of scientific demonstration.

7 . C O N C LU S I O N To close, I take it that Ockham’s Scientia Argument indicates one of his central purposes in putting forth his theory of mental language: he believes that positing this sort of representational system is the only way to harmonize his nominalism with his commitment to the basic outlines of Aristotle’s account of scientific demonstration and scientific knowledge. In giving this argument, he concurs with what Aristotle said above in passage P3: [P3] There need not be any forms, or some one item apart from the many, in order for there to be demonstrations. It must, however, be true to say that one thing holds of many.61

Ockham agrees with this observation by Aristotle—but only in part. In order for there to be scientific demonstration, it indeed must be true to say that one thing holds of many. But though this is necessary for there to be scientific demonstration, it is not yet sufficient for there to be scientific 60 SL III-2, ch. 5 (OPh I, 513).

61 Aristotle, Posterior Analytics I.11, 77a5–7.

Ockham’s Scientia Argument for Mental Language

167

demonstration. The representational system with which one says “that one thing holds of many” must also be able to accommodate imperishable subject terms and necessarily true objects of scientific demonstration; and this, Ockham thinks, can only be accommodated by a mental language. St Norbert College

BIBLIOGRAPHY Adam Wodeham. Lectura secunda, ed. G. Ga´l and R. Wood, 3 vols (St Bonaventure, NY: Franciscan Institute Publications, 1990). Adams, Marilyn McCord. William Ockham, 2 vols (Notre Dame, IN: University of Notre Dame Press, 1987). Aristotle. Posterior Analytics, tr. J. Barnes, 2nd edn (Oxford: Clarendon Press, 1993). Armstrong, David. What is a Law of Nature? (Cambridge: Cambridge University Press, 1983). Brower-Toland, Susan. “Ockham on Judgment, Concepts, and the Problem of Intentionality,” Canadian Journal of Philosophy 37 (2007), 67–110. David, Marian. “Defending Existentialism?” in M. Reicher (ed.), States of Affairs (Frankfurt: Ontos Verlag, 2009), 167–209. Gregory of Rimini. Lectura super primum et secundum Sententiarum, ed. A. D. Trapp et al., 6 vols (Berlin: De Gruyter, 1979–84). Henry of Harclay. Ordinary Questions I–XIV (Auctores Britannici Medii Aevi XVII), ed. M. G. Henninger (Oxford: Oxford University Press, 2008). John Buridan. Summulae de dialectica, tr. G. Klima (New Haven: Yale University Press, 2001). John Duns Scotus. Opera omnia, ed. C. Balic´ et al. (Vatican City: Vatican Press, 1950–). John Duns Scotus. Questions on the Metaphysics of Aristotle, tr. G. J. Etzkorn and A. B. Wolter, 2 vols (St Bonaventure, NY: Franciscan Institute Publications, 1997–8). Karger, Elizabeth. “Mental Sentences According to Burley and to the Early Ockham,” Vivarium 34 (1996), 192–230. Leff, Gordon. William of Ockham: The Metamorphosis of Scholastic Discourse (Manchester: Manchester University Press, 1975). Maurer, Armand. The Philosophy of William of Ockham in the Light of its Principles (Toronto: Pontifical Institute of Mediaeval Studies, 1999). Normore, Calvin G. “The End of Mental Language,” in J. Biard (ed.), Le langage mental du Moyen Aˆge a` l’Aˆge Classique (Leuven: Peeters, 2009), 293–306. Normore, Calvin G. “Ockham on Mental Language,” in J. C. Smith (ed.), Historical Foundations of Cognitive Science, Philosophical Studies Series 46 (Dordrecht: Kluwer, 1990), 53–70. Panaccio, Claude. “Mental Language and Tradition: Encounters in Medieval Philosophy: Anselm, Albert, and Ockham,” Vivarium 45 (2007), 262–82.

168

Eric W. Hagedorn

Panaccio, Claude. Ockham on Concepts (Burlington, VT: Ashgate Publishing, 2004). Spade, Paul Vincent. Five Texts on the Mediaeval Problem of Universals (Indianapolis: Hackett, 1994). Spade, Paul Vincent. Thoughts, Words, and Things: An Introduction to Late Mediaeval Logic and Semantic Theory (December 27, 2007). Thomas Aquinas. Opera omnia (Fundacio´n Toma´s de Aquino, 2012). Trentman, John. “Ockham on Mental,” Mind 79 (1970), 586–90. Tweedale, Martin. Scotus vs. Ockham: A Medieval Debate over Universals, 2 vols (Lewiston, NY: Edwin Mellen Press, 1999). Walter Chatton, Reportatio et lectura super Sententias: collatio ad librum primum et prologus, ed. J. C. Wey (Toronto: Pontifical Institute of Medieval Studies, 1989). William Ockham. Opera philosophica, ed. G. Ga´l et al. (St Bonaventure, NY: Franciscan Institute, 1967–86). William Ockham. Opera theologica, ed. G. Ga´l et al. (St Bonaventure, NY: Franciscan Institute, 1974–88).

Silence, Skepticism, and Vulgar Theology On Stern’s Maimonides Daniel Davies

Josef Stern’s extensive and sophisticated new book1 covers Maimonides’ biblical and rabbinic exegesis, various methods of communication, arguments about God, the cosmos, human nature, the human goal, and the overall purpose of the Guide for the Perplexed. The result is a wide-ranging, challenging, and rewarding work, which offers original readings of a number of the Guide’s chapters, recasting it as an expression of philosophical and religious praxis designed to help people lead the best life possible. The theme uniting the work is skepticism, which results inevitably from human nature and from the fact that neither matter nor form exist independently of one another. Neither religion nor the medieval period seem to loom large in the history of skepticism. Nevertheless, despite his status as a giant of the Jewish tradition, scholars with varied approaches have claimed that Maimonides’ thought contains skeptical elements, a fact that may partially account for his immense and enduring appeal. Stern places skeptical themes at the center of the Guide, making them more deep seated than most other interpreters and arguing that they steer the entire work. Given the length and detail of the book, it is impossible to touch on all of its themes, so my aim in this review is not merely to survey.2 Instead, I first suggest that Stern’s skeptical Maimonides could take his place in our growing understanding of medieval skeptical traditions, then explain what his skepticism amounts to. Maimonides is skeptical that God can be known, so he advocates silence as proper worship. His own silence leads Stern to raise questions about further issues that are skeptical in more extreme ways. Those questions oppose a straightforward reading of the Guide that 1 Josef Stern, The Matter and Form of Maimonides’ Guide (Cambridge, MA: Harvard University Press, 2013). 2 I have published a brief review of this sort on the H-Judaic list .

170

Daniel Davies

attempts to take Maimonides at his word, and conflict with the religious claims over which he proclaims ignorance. The second half of the article will seek to explain the background in Maimonides scholarship of recent decades to some of these claims and suggest how an alternative, more straightforward reading can incorporate some of Stern’s insights. By likening Maimonides to the ancient Pyrrhonian school, Stern’s purpose is to illustrate by drawing out the parallels rather than claim historical influence. Nevertheless, there may be ways in which Maimonides’ skepticism fits with the wider currents of his time. Paul Heck has recently traced some technical meanings that the word Maimonides uses in his title for “perplexity,” h: ayra, acquired in the writings of some Islamic authors. One of these is the perplexity that arises when faced with anthropomorphic verses in the Qur’a¯n, and this is the meaning that Maimonides gives when he explains why he chose it to name his work.3 The perplexed is someone who is beset by doubts and hesitation because the literal sense of scripture seems to teach that God is corporeal, but that is impossible. Such a person cannot decide whether to accept the scriptural text or to reject it altogether. Maimonides’ goal is like that of other Arabic authors, to ease this perplexity by explaining that parables have a deeper meaning; biblical exegesis is his stated aim in writing the Guide.4 Stern is alive to the literary dimension, and he explores Maimonides’ interpretations of a variety of parables, presenting a new account of what Maimonides understood parables to be and how he used them.5 Every parable that he interpreted or invented is seen to have three layers of meaning, the deepest of which is a skeptical message. Although it is an important part of the book, and constitutes some of the evidence for Stern’s reading, space prevents me from discussing this theory. It might stand independently of the general skeptical interpretation. Among the medieval Arabic writers, al-Ghaza¯lı¯ is unusual, inasmuch as his skepticism is well known and has even been said to anticipate Descartes and Hume. In a famous passage of his autobiography, he seeks certainty and reports going through a stage in which he doubted the reliability and veracity of sense experience or even of reason itself. In his Incoherence of the Philosophers he questions our knowledge of secondary causality, pointing out that the cause itself is never seen; we only observe one event following 3 Paul Heck, Skepticism in Classical Islam: Moments of Confusion (London: Routledge, 2014), 36–46. 4 The Guide of the Perplexed, translated with an introduction and notes by Shlomo Pines (Chicago: University of Chicago Press, 1963), I: Introduction, 6. 5 For an alternative explanation of Maimonides’ use of parables, see Roslyn Weiss, “Four Parables about Peshat as Parable,” in Shlomo Carmy and Yamin Levy (eds), The Legacy of Maimonides: Religion, Reason and Community (New York: Yashar Books, 2006), 111–26.

Silence, Skepticism, and Vulgar Theology

171

another, and then we suppose that the first was a cause of the second.6 On Stern’s reading, Maimonides does not follow this sort of skepticism. He explains that it is crucial to Maimonides’ naturalist thought that the senses are trustworthy, since only with that presumption is it possible to base proofs and beliefs on what exists, and “the correct opinions conform to that which exists,” not vice versa.7 Maimonides’ doubts do not pertain to the premises drawn from sense experience, nor to those the intellect accepts as soon as it understands, but to the conclusions that can be derived from such premises. So Stern explains that “the facts of nature as revealed by science place a strong constraint on metaphysics and divine science. What is metaphysically possible (or necessary) must be compatible with nature as it actually is” (11). Stern’s account of Maimonides emphasizes that he pleads ignorance of a number of important issues. There are limits on what it is humanly possible to know and understand. Skepticism results from recognition that there are good arguments both for and against important metaphysical positions, such as the nature of God’s relationship with the world, and no criterion to adjudicate between them. Here again, it is possible that there is a precedent of sorts, in that there were authors who argued that there is no way to appeal to an independent criterion to adjudicate between rival claims of different religions.8 Acknowledging the intellect’s limitations allows a person to contain her epistemic desire and restrict the areas she investigates to within the possible, as “to fatigue the minds with notions that cannot be grasped by human intellects and for the grasp of which they have no instrument is a defect in one’s inborn disposition or some sort of temptation.”9 Skepticism is the end point of rational inquiry, whose goal is to reach the highest possible level of learned ignorance, acknowledge aporias, and live a life governed by the intellect, to the extent possible.10 Stern traces the debate over Maimonides’ skepticism to an essay that Shlomo Pines published in 1979, in which he argued that although Maimonides states that human perfection consists in perfecting the intellect, his true position was that intellectual perfection is impossible, and the 6 Taneli Kukkonen sets these issues in their historical context in “Al-Ghaza¯lı¯’s Skepticism Revisited,” in Henrik Lagerlund (ed.), Rethinking the History of Skepticism: The Missing Medieval Background (Leiden: Brill, 2010), 29–59. 7 Guide I: 71, 179. 8 Heck, Skepticism, 71–4. 9 Guide II: 24, 327. 10 If this skeptical reading is right, it could make sense of the prevailing current scholarly preference to render the title Guide of the Perplexed, rather than Guide for the Perplexed, as the Guide is understood to be a book written by one who is perplexed in order to lead people into perplexity, or at least from one kind of perplexity to another. See Joel Kraemer, “Maimonides, the Great Healer,” Maimonidean Studies 5 (2008), 20 nn. 28 and 29.

172

Daniel Davies

human goal is therefore civic, to perfect society.11 Maimonides is presented as a Kantian “critical” philosopher, who rejects the possibility of metaphysics and, consequently, considers ethics more important. Pines’s position has many critics and some followers.12 What seems to be most crucial about this essay is the switch from focusing on the relationship between religion and philosophy, which is the most common way to approach the Guide, to an attempt to identify the human goal. While embracing Pines’s claim that the Guide is ultimately about the question of how one ought to live, and accepting that Maimonides places severe limits on the range of human knowledge, Stern offers important qualifications to Pines’s view. Instead of arguing that knowledge of metaphysics is altogether impossible, a conclusion that would call into question Maimonides’ arguments for God’s existence, Stern argues that what cannot be attained is strict scientific knowledge of God and the separate intelligences. Elaborating further on what constitutes scientific knowledge, he uses the Aristotelian distinction between what came to be known as demonstrations propter quid and demonstrations quia. The skeptic argues that proper scientific knowledge is available only through the former and that, since all arguments for God and the intelligences are quia, it is impossible to know such beings scientifically. Knowledge, on this view, is that which results from propter quid demonstrations, and bears “the narrow sense of scientific knowledge, because that is the exclusive cognitive state that, in ideal circumstances, would enable the human to achieve the state of a fully actualized or acquired intellect and to conjoin with the Active Intellect, the true human perfection and happiness” (137). When Aristotle explains the difference between the two kinds of arguments, he distinguishes between one that both proves and explains its conclusion and one that only proves that the conclusion is true. For example, it would be possible to prove that the planets are near by pointing out that they do not sparkle in the following way. The planets do not sparkle. What does not sparkle is near. Therefore the planets are near. This is merely a proof that the planets are near but does not explain their proximity. The middle term shows the conclusion to be true but does not explain why the minor term holds of the major term. It is a proof quia. However, if the proof argued from their proximity to the conclusion that

11 Shlomo Pines, “The Limitations of Human Knowledge according to al-Farabi, ibn Bajja, and Maimonides,” in Isadore Twersky (ed.), Studies in Medieval Jewish History and Literature (Cambridge, MA: Harvard University Press, 1979), 82–109. 12 Stern cites a number of responses on 134–5. Pines’s claim depends also on a particular reading of Alfarabi, which is dismissed by The´re`se-Anne Druart in “Alfarabi and Emanationism,” in John F. Wippel (ed.), Studies in Medieval Philosophy (Washington, DC: Catholic University of America Press, 1987), 23 n. 5.

Silence, Skepticism, and Vulgar Theology

173

they do not sparkle, the demonstration would be propter quid because the proximity would explain why the planets do not sparkle. Another example is the relationship between the moon and its waxing. One can tell that the moon is spherical from the fact that it waxes, but that does not explain the moon’s sphericity. On the other hand, the sphericity does explain the waxing, so a propter quid demonstration reasons from the moon’s sphericity to its waxing. This demonstration propter quid avails proper scientific understanding since the moon’s waxing is understood through its cause, which is that it is spherical. In the case of the proof quia, on the other hand, there is no scientific understanding. The waxing does not explain why the moon is spherical.13 Similarly, God is uncaused and so cannot be understood through a cause. Nothing explains God’s existence because God is prior to all other beings. The world exists, providing evidence that it is created and that God exists, but the world’s existence cannot explain God’s existence, so a propter quid demonstration of God’s existence is ruled out. Stern employs this distinction to offer a novel answer to a major current debate. Maimonides states that the heavenly revolution is “the greatest proof through which one can know the existence of the deity.” Later in the Guide, when discussing what is known about the heavens, he states that “the general conclusion that may be drawn from them, namely, that they prove the existence of their mover, is a matter the knowledge of which cannot be reached by human intellects.”14 Stern addresses two issues in order to come to grips with the apparent inconsistency. The first involves identifying what “the greatest proof ” is, and the second involves Maimonides’ meaning when he appears to suggest that the heavens do not prove their mover’s existence. Most scholars take the “greatest proof ” to be included in the four demonstrations for God’s existence presented at the beginning of part II of the Guide, each of which is based on the premise that the heavens’ motions had no beginning. Stern suggests that there is another contender, which occurs later in part II of the Guide, and is based on the heavens’ disordered motions. Both proofs are valid but the premises on which they are based are uncertain. The first assumes that there was no beginning to time, and that God must exist to account for the perpetual motion, and the second establishes that God creates through purpose which, at least in Maimonides’ stated view, requires an absolute beginning.15 Furthermore, 13 Post. An. 78a–78b. 14 Alternative translations and interpretations have been offered for this sentence, to which most of volume 8 of the journal Aleph (2008) was dedicated. 15 It may also be that the second proof is supposed to indicate that God creates with purpose but that it does not independently demonstrate God’s existence. See Kenneth

174

Daniel Davies

if the heavenly motions seem disordered from our point of view, any explanation of their motions would run counter to the accepted scientific premises that ground our understanding of the sublunar world and of the first contender to be the greatest proof. As Maimonides explains them, the two proofs entail two different understandings of God’s relationship with the world. He is unable to say which of these arguments is true, and in cases of uncertainty he advocates affirming one of the possible positions or suspending judgment.16 It seems that here we might have arrived at one of the cases of equipollence, a situation in which there are good arguments for one position and for its opposite, presuming the two conceptions are opposites, and there is no way of telling which is true. Stern argues that neither of these two arguments can be considered the greatest proof for God’s existence because neither can be known to be sound. He suggests one of Maimonides’ dilemmatic arguments instead. Either the world is everlasting or it is not. If it is everlasting, God must exist to explain its infinite motion. If it is not everlasting, God must exist in order to have created an absolute beginning. Therefore, whichever of the two initial points of view is true, and one must be, God exists. Stern argues that this dichotomous argument covers all possibilities, and so qualifies as the greatest argument for God’s existence. Returning to the problematic statement that the heavenly motions do not provide knowledge of God, Stern’s solution is to argue that the word ‘knowledge’ is used to refer to proper scientific knowledge. Since Maimonides is skeptical about which of the two kinds of arguments from the heavens is true, he is unable to claim that they avail genuine knowledge of God, even though God’s existence is demonstrated. Although they succeed, arguments for God’s existence lead to doubt and to recognizing that the answers to some crucial questions are beyond our reach. Accepting that these questions cannot be answered frees one of the desire to understand what cannot be known. Ultimately, because God cannot be known at all, in the strict scientific sense, Maimonides advocates an extreme negative theology in which no words can be used of God, leaving silence and withholding judgment as the only possible option. Withholding judgment when there is no demonstration prevents the unhappiness that stems from pursuing insatiable epistemic desires. The point of investigation into cosmology and natural theology is “to bring oneself to recognition of the limitations of one’s intellect and, once one

Seeskin, Maimonides on the Origin of the World (Cambridge: Cambridge University Press, 2005), 122. 16 Guide II: 16, 293.

Silence, Skepticism, and Vulgar Theology

175

has recognized the finitude of his intellect, ‘to stop’ where neither demonstration nor refutation is possible. . . . By curing oneself of this major source of intellectual anxiety and unhappiness, one ipso facto restores one’s happiness” (311). But this is not all that Maimonides advocates, and in the final two chapters, which I find the most compelling, Stern argues that he retains the ideal of intellectual contemplation. The Guide’s arguments are therefore taken to be exercises designed to lead someone toward proper, intellectual worship. The perfect person acts in a way that “he is so moved to act by his intellectual judgment that x is rationally fitting,” not because of a passion or emotion. “He does them solely in imitation of God” (345). But “what is the content of intellectual worship of the deity?” (316) As Kenneth Seeskin asks, “how do you imitate something you cannot comprehend?”17 Stern’s answer builds on the reason that humans can know very little about God and the heavens, which is that people are fundamentally material beings. Maimonides states that “matter is a strong veil preventing the apprehension of that which is separate from matter as it truly is,”18 so even the greatest of humans cannot understand the intelligences.19 Matter limits a person’s ability to understand in two ways. It both distracts someone from focusing on what she already knows, and it prevents apprehension altogether. Nevertheless, the perfect person realizes that although intellectual perfection is impossible, it is superior to all bodily perfections, and so she is ashamed of her corporeality. This shame impels her to continue striving to identify as closely as possible with the intellectual part of her nature. In drawing up his arguments, Maimonides is offering exercises that can be used by readers to undertake their intellectual journeys, and which mirror his own. These exercises “exemplify a conception of philosophy according to which it is not, or not primarily, a theory, discourse, doctrine, or the exegesis of a text, but rather a set of practices that constitutes a way of life” (313). Instead of teaching his readers religious and philosophical doctrines, or expecting humans to become divine, Maimonides helps them to understand their natural limitations while constantly striving to be as perfect as humanly possible. There is also an ethical dimension to this teaching. Turning to God as the sole object worthy of worship involves turning away from worshipping

17 Kenneth Seeskin, Searching for a Distant God: The Legacy of Maimonides (New York: Oxford University Press, 2000), 92. 18 Guide III: 9, 436. 19 Maimonides depicts Moses as having reached the peak of intellectual perfection, so that he alone was able to prophesy without using the imagination. If I understand correctly, Stern intimates that Maimonides uses Moses as a metaphor for the active intellect (329).

176

Daniel Davies

everything else. In Maimonides’ view, that does not involve ignoring everything else. On the contrary, recognizing intellectual limits is accompanied by constant thought about what is within those limits and thinking about all things as created beings. Understanding our own natures, as beings made up of form and matter, includes understanding that everything is created and appreciating our own place in the created order, and that requires understanding other beings as far as possible and giving all of them their due importance. Meditating on our own unimportance, and the relative importance of other things, raises obligations toward all those things, as fulfilling our nature involves living harmoniously with other creatures and treating them appropriately.20 This is an aspect of Maimonides’ thought that is often overlooked but toward which Stern directs our attention. Until this point, the skeptical Maimonides is not so far from the traditional view, although the emphasis and details would differ. A major problem that a regular reader would have is with the claim that Maimonides rejected any form of eternal life. Although he at one point appears to oppose belief in an individual afterlife, since what survive of different people’s intellects “are all one in number,”21 he repeatedly argues that knowledge of truths, which are unchanging, constitutes eternal life. The reason that Maimonides would reject even an impersonal afterlife, on Stern’s reading, is that he is not committed to the existence of the intelligences or the active intellect. But in order for people to have eternal knowledge, they have to unite their own intellects with the active intellect. This is an aspect of the skeptical interpretation that extends Pines’s claim that Maimonides thought knowledge of separate beings to be impossible, and presumably involves doubting that there can be demonstrations quia that such beings exist. It is not the case that he presents definitive arguments against their existence, but, in Stern’s view, nor are there arguments proving adequately that they must exist. How could it be possible to know that such immaterial entities exist when all knowledge is based on sense experience and therefore must be knowledge of material beings? Here the skeptical Maimonides is seen at perhaps his most Kantian. Although he is unwilling to admit that the active intellect must exist in order to account for the possibility of human knowledge, he recognizes that denying its existence results in epistemological problems. Without the active intellect, Maimonides’ understanding of knowledge would collapse. Uniting with something that is in the active intellect involves knowing an eternal truth. The active intellect informs objects in external reality and the objects’ forms are then conjoined 20 See also Seeskin, Searching, 115–22.

21 Guide I: 74, 221.

Silence, Skepticism, and Vulgar Theology

177

with the human intellect when the mind abstracts them from the objects.22 So, Stern argues, Maimonides posits an active intellect as a condition for knowledge even though he is reluctant to affirm its existence. However, in order to follow Stern’s conclusions about the limitations of the human intellect and the consequences for living well, I would like to claim that it is not necessary to reject the existence of the active intellect and the intelligences, or the theological positions that are often thought to be aimed only at the vulgar. Furthermore, Stern succeeds in his aim of producing “a unified reading of the Guide,” but at the cost of excluding some of its claims or assigning them a political role. A straightforward reading could offer a slightly different methodology in order to attempt a unified interpretation of the Guide as a whole. Before offering such an alternative reading, I think that it will be helpful to provide some background to Stern’s arguments in another of Pines’s essays. Pines wrote a masterly introduction to his monumental translation of the Guide, published in 1963, outlining the historical influences on Maimonides. There he pointed out that Maimonides seems to propound at least two separate depictions of God. The first is the famous doctrine of negative attributes, which denies that words can be used of God; the second is the claim that God is an intellect, to which Maimonides dedicates chapter I: 68. Of the second doctrine, Pines states, “obviously this view goes counter to negative theology.”23 As Stern explains in detail, Maimonides’ negative theology precludes using any words of God, for doing so would imply plurality in God. Pines argues that “in the face of such statements any attempt to make out a case for God’s cognition having nothing except homonymy in common with man’s cognition may easily seem mere quibbling; and yet only such an attempt can satisfy negative theology.”24 Because they are incompatible, says Pines, both doctrines cannot represent Maimonides’ genuine view. To the claim that Maimonides might not have been aware of the conflict, Pines writes that, “in this particular case this point of view would amount to a grave and, in my opinion, very implausible accusation of muddle-headedness directed against Maimonides.”25 While Pines’s 1979 essay has been widely discussed but not universally accepted, the observations made in 1963 founded an ensuing debate in 22 For this explanation of Avicenna’s account of the doctrine, see Dag Nikolaus Hasse, “Avicenna’s Epistemological Optimism,” in Peter Adamson (ed.), Interpreting Avicenna (Cambridge: Cambridge University Press, 2013), 109–19. Stern collates Maimonides’ statements on knowledge in “Maimonides’ Epistemology,” in Kenneth Seeskin (ed.), The Cambridge Companion to Maimonides (Cambridge: Cambridge University Press, 2005), 105–33. 23 “Translator’s Introduction,” Guide of the Perplexed, xcviii. 24 Ibid., xcvii. 25 Ibid., xcviii.

178

Daniel Davies

Maimonides scholarship. That the two positions are opposed is rarely, if ever, questioned, and scholars tend to discuss which of the two was Maimonides’ real opinion. Stern’s solution is to argue that although the two doctrines are opposed, Maimonides’ own positions are not, because he does not adhere to the view that God is intellect. Instead, the purpose of I: 68 is to cast doubt on the doctrine. It is part of his critical, skeptical program. Unlike previous responses to the question, then, Stern sees no inconsistency in Maimonides’ own approach, even though he holds that the doctrines presented do not accord with one another. “There is no tension between this chapter and those on divine attributes because all of them articulate Maimonides’ sustained argument for skepticism with respect to any knowledge about the deity” (233). As a skeptic, his project is purely critical, to show that there is no irrefutable argument to show that God is an intellect, but nor can it be demonstrated otherwise. Human reason simply cannot stretch that far. Stern’s reading successfully navigates a way to explain Maimonides’ text while respecting what he says of both positions. The first hint that Maimonides might not accept that God can be characterized as an intellect is the chapter’s location. Stern considers the whole of part one of the Guide to be dedicated to incorrect views that certain groups hold. Most of the first fifty chapters are taken up with explaining how to construe anthropomorphic scriptural representations of God. They amount to a critique of those who read biblical texts literally, and they culminate in the section on negative theology, in which Maimonides offers a critique of those who believe in divine attributes. I: 68 reports the philosophers’ doctrine, and continues the critical motif. A second hint occurs when Maimonides states that the doctrine is “generally admitted by the philosophers.” What is widely accepted might not be true; in Stern’s view, by attributing the idea to the philosophers without explicitly endorsing it, Maimonides is here expressing his disagreement. Maimonides also states that “anyone who has not studied the books composed on the intellect, has not grasped the essence of the intellect, does not know its quiddity, and has no understanding of it other than one that resembles his understanding of blackness and whiteness, will have great difficulty understanding this notion.” On Stern’s reading, this is a third hint that the chapter is critical of the doctrine, and he reads it to mean that “no one can understand what is wrong with” identifying subject and object in God’s intellection (234). His interpretation may be based on Pines’s remark that “in order to illustrate this conception Maimonides uses (probably deliberately) a humble example,”26 the form of a piece of wood. Pines seems to indicate that Maimonides’ using 26 Ibid., xcviii.

Silence, Skepticism, and Vulgar Theology

179

wood as an example is a rhetorical device designed to emphasize the similarity he draws between human and divine knowledge. Maimonides thereby draws attention to the anthropomorphic, and therefore false, nature of the claim that God is an intellect. These comments appear in the context of Pines’s discussion of the Avicennan background to the Guide. Despite Pines’s unparalleled qualifications, it might be possible to re-evaluate the question in the light of the half-century of Avicenna scholarship that has since passed.27 Pines states that “as far as I know, nobody beforehand had regarded as closely connected” the Aristotelian texts in Meta. L, in which God is depicted as thought thinking itself, and in De Anima III, which considers human cognition.28 However, Peter Adamson shows that Avicenna argues that all immaterial substances must be intellects and that this claim, established in psychology, is an important part of Avicenna’s metaphysics and theology.29 If other philosophers also considered God to be ungraspable and nevertheless depicted God as a fully actualized intellect, Pines’s assertion that unknowingly adhering to both doctrines would amount to an “accusation of muddle headedness” must be leveled against them as well. That raises the question whether there might be an alternative solution.30 27 Herbert Davidson shows that when Maimonides discusses Aristotle’s metaphysics, he references Avicenna’s views. Davidson suggests that Maimonides either read Aristotle through Avicenna or was not very familiar with Aristotle’s own texts. See Davidson, Moses Maimonides: The Man and his Works (New York: Oxford University Press, 2005), 106. Recently, there has been a suggestion that Maimonides did not read Avicenna. See Freudenthal and Zonta, “The Reception of Avicenna in Jewish Cultures East and West,” in Peter Adamson (ed.), Interpreting Avicenna (Cambridge: Cambridge University Press, 2013), 218. I do not understand the claim that “had he read a work by Avicenna he would have become aware of the difference between Aristotle’s and Avicenna’s doctrines and would not have ascribed Avicennian views to Aristotle.” It seems to me that Maimonides would have needed to read Aristotle to become aware that the doctrines he ascribes to Aristotle are not Aristotle’s. It is possible that he was not aware of the differences between the two. 28 Aryeh Kosman states that Aristotle himself intended Meta. L to concern human cognition as well as divine cognition.“Metaphysics’ ¸9: Divine Thought,” in Michael Frede and David Charles (eds), Aristotle’s Metaphysics Lambda: Symposium Aristotelicum (Oxford: Oxford University Press, 2000), 307–26. See also Myles Burnyeat, Aristotle’s Divine Intellect (Milwaukee: Marquette University Press, 2008). 29 Peter Adamson, “Avicenna and his Commentators on Human and Divine SelfIntellection,” in Dag Nikolaus Hasse and Amos Bertolacci (eds), The Arabic, Hebrew and Latin Reception of Avicenna’s Metaphysics (Berlin: De Gruyter, 2012), 97–122. 30 If Stern’s claim about Maimonides’ attitude to the peripatetic philosophers is right, Maimonides would be a critic of metaphysical doctrines espoused by Avicenna rather than the follower he is usually thought to be. Warren Zev Harvey argues that Maimonides held Avicenna in high regard as a theologian but expressed reservations about other aspects of his philosophy. See Harvey, “Maimonides’ Avicennianism,” Maimonidean Studies 5 (2008), 107–19.

180

Daniel Davies

Maimonides’ statement could indicate that it is difficult to understand how an intellect could be identical with its object of thought without having studied books on the intellect. If one has studied only other aspects of physics, one might think that the way in which the intellect is related to an abstract form resembles the way in which whiteness is related to things that are white. Those who know what the intellect is, on the other hand, will understand that both the intellect and its objects, when abstracted, are immaterial and therefore not necessarily distinguished, unlike a bucket of white paint and a painted fence. That Maimonides chose a piece of wood rather than, say, a horse makes no difference to the claim that an intellect can be identical with the form that is its object. He might simply be claiming that there is no difference between God’s thought and God’s essence, and illustrating that claim with an example designed to show that it is possible for an intellect to be identical with a concept. There remains the qualification that unlike the divine intellect, which is identical in subject and object in every respect, the human intellect does not become the piece of wood that it is contemplating, as it grasps the abstract form of wood, not a physical piece of wood. Had Maimonides likened the human intellect to God’s intellect by using multiple examples, that would pose a greater problem. It is possible to understand how an intellect can be identical with a single object of intellection. But in the case of two objects of thought, wood and horse, for example, drawing a parallel would be more problematic because the form of wood differs from the form of horse, even though both subsist in a single intellectual subject. However, God’s intellect differs from human intellect in this regard. Maimonides draws five distinctions between God’s knowledge and human knowledge, one of which is that God’s knowledge covers multiple objects without itself being multiplied, whereas human knowledge does not.31 Because of these differences, the word ‘knowledge’ is used equivocally of the two. Similarly, if a permanently actualized, selfthinking, and entirely self-subsistent intellect cannot be conceived, and words gain their meanings from our conceptions, the meaning of the word ‘intellect’ will not apply to the permanently actualized intellect. Intellect could only be used in an equivocal fashion which, as Pines states, would satisfy the requirements of negative theology.32 31 The discussion of God’s knowledge appears in part III of the Guide, and Stern distinguishes it from the critical chapter in part I. For more on Maimonides’ account of God’s knowledge, see Charles Manekin, “Maimonides on Divine Knowledge: Moses of Narbonne’s Averroist Reading,” American Catholic Philosophical Quarterly 76 (2002), 51–74. 32 Stern could respond by distinguishing Maimonides’ negative theology from Avicenna’s use of religious language. My aim in this essay is merely to gesture toward an

Silence, Skepticism, and Vulgar Theology

181

The difference between these two approaches might stem from certain ontological presuppositions and from different ideas about what negative theology is supposed to do. Does Maimonides want to cut the link between God and creation altogether, or does the unique relation between creator and creation, which is unintelligible to us, entail that created perfections are caused by a necessary being that possesses them all in perfect actuality? In the latter case, negative theology could be concerned with what can be known and said, without denying that God is permanently actual and therefore lacking the deficiencies that attend potentiality.33 The former, as well as supporting the view that Maimonides thought of human perfection as moral or political, would appeal to many philosophers who consider existence to be an empty property that adds nothing to that of which it is predicated, as negative theology would be denying any content to the necessary being, rather than to our thought and language. Pines raises a methodological question that goes some way toward explaining why his approach is widely adopted: “It may be recalled in this connection that in his Introduction to the Guide Maimonides states that for reasons given by him he deliberately inserted into this work contradictory theses (one false and one correct). Is this an instance of this didactic method, and if so, which of the two doctrines represents Maimonides’ real opinion?”34 There are two ways in which Maimonides says that he contradicts himself in the Guide. One occurs when a student has progressed enough to understand a more advanced position that conflicts with an elementary position learned previously. This is the fifth in the list of reasons for contradictions that Maimonides draws up. The other contradiction that he uses occurs when the author devises a method to “conceal and reveal” while carrying out two different discussions on the basis of contradictory premises. In these cases, “the vulgar should in no way be aware of the contradiction.” This is the seventh in Maimonides’ list, and is often known simply as “the seventh contradiction.” Pines does not specify which of the two kinds he has in mind, although most responses seem to assume that Maimonides is using the seventh.35 What exactly it amounts to is a matter of debate, but the contradiction is often read through the work alternative possible way to approach the Guide, so it is not possible to discuss Stern’s detailed account of Maimonides’ doctrine. 33 At one point Stern too states that “He is always actualized or in act” (341). 34 It might be relevant that Pines did not make any clear assertion here but appears simply to be asking a question. “Translators Introduction,” xcviii. 35 I have argued that Maimonides uses the seventh cause of contradictions in his explanation of certain scriptural passages that the rabbinic tradition forbids expounding in public. If this is right, there is no need to seek a hidden philosophical or theological position in order to account for its presence in the Guide. See my Method and Metaphysics

182

Daniel Davies

of Leo Strauss, whose influence on studies of Maimonides would be difficult to overstate.36 Stern emphasizes that his own interpretations do not follow Strauss, who divided Maimonides’ teachings into exoteric and esoteric positions, the former for the philosophers and the latter for the vulgar (50–1). For present purposes, it is worth noting that the two methodologies differ in at least two important respects. One is that Stern does not posit an inherent tension between religion and philosophy, a foundation of Strauss’s worldview and interpretation of philosophical texts. The second is that Stern refutes the view that when Maimonides presents contradictory opinions, he would have accepted one side of the opposition and rejected the other. This second difference also distinguishes Stern from Pines, when the latter asks which of the two opinions would reflect Maimonides’ true view. By arguing that the contradictions result in equipollence, Stern removes the need to affirm one view but not its contrary. Alongside these differences, there are at least two pertinent similarities. The first is that, like Strauss, Stern considers the hidden message to be a more advanced position than that which is openly expressed. It is important, however, that Stern stresses pedagogical rather than political reasons for not clearly stating the advanced position. A reader’s working through the outer meanings is an exercise that aids her progression toward truth and the good life, by encouraging critical engagement with views held by both theologians and philosophers. The Guide is therefore a teacher, and a guide on a religious and philosophical journey, rather than mainly a political work addressing the relationship between philosophy and religion. The second similarity between Stern and Strauss is that both consider Maimonides’ silences to be crucial. Strauss explains a number of “devices,” which he claims that Maimonides uses in the Guide, one of which “consists in silence, i.e., the omission of something which only the learned, or the learned who are able to understand of themselves, would miss.”37 Similarly,

in Maimonides’ “Guide for the Perplexed” (New York: Oxford University Press, 2011), 20–5. 36 As Kenneth Hart Green explains, Strauss did much to encourage renewed engagement with Maimonides. Despite the strong reactions that his work provokes, even some who oppose his readings can recognize his brilliance. See Green, Leo Strauss and the Rediscovery of Maimonides (Chicago: University of Chicago Press, 2013). Strauss also anticipated Pines’s argument that Alfarabi hinted at disbelief in an afterlife. See Strauss, Persecution and the Art of Writing (Chicago: University of Chicago Press, 1988), 13. 37 Strauss, Persecution, 75. This is only one way in which Strauss claimed that silence is used. He argued that authors drop hints as to their true opinions, and surround those hints with passages that serve to camouflage the clues, although even those later passages may contain further hints relevant to different parts of the argument: “the writer has to

Silence, Skepticism, and Vulgar Theology

183

Stern argues that there are places in the Guide where Maimonides keeps silent about the intelligences and, in doing so, indicates his discomfort with them. “Maimonides’ repeated silences precisely at those moments when we would most expect him to introduce separate intellects and the Active Intellect suggest that he is unwilling to commit himself to their explanatory roles (although he also seems unwilling to deny their roles categorically)” (271). Arguments from silence are of course difficult to prove. At what point can a reader feel confident to count herself among the learned? It is possible to be confused by a perceived silence because of the need to learn and reflect even more. Instead of focusing on what Maimonides misses out, a straightforward reading, one that preserves the theological positions often classified vulgar, would instead ask about what motivated him to make the comments he did, and try to understand those statements. So, in this context, the intelligences might not be mentioned simply because Maimonides is not dealing with them in that chapter. However, the Guide is rarely that simple. Even a traditional reader cannot escape the fact that it is laconic, and that Maimonides expects pupils to fill in gaps through wider reading and independent thought, so he advises to “learn everything that ought to be learned and constantly study this treatise.” He even includes instructions to “connect its chapters one with another, and when reading a given chapter, your intention must be not only to understand the totality of the subject of that chapter, but also to grasp each word that occurs in it in the course of the speech.”38 There is no doubt, for example, that digressions are often important. When Maimonides appears to minimize certain incidental remarks by pointing out that they are not the purpose of the chapter in which they appear, those remarks are relevant to an issue that he discusses elsewhere in the Guide. The question is, then, whether the silence that Stern identifies is a similar sort of indication. One piece of evidence lies in the difference between two accounts of motion that appear in the Guide, the first in I: 72 and the second in several chapters of part II. Stern reports that the usual way to explain such differences is to view the earlier section as a summary introduction of the issues considered in greater length later on. Against this explanation, he argues that there are genuine differences between the two accounts, and that they indicate that one is not an introduction to the other. A major difference is in the use of the separate intelligences to explain heavenly

interrupt his short hints by long stretches of silence, i.e., of insignificant talk” (Persecution, 53). I did not detect Stern making use of this method. 38 Guide I: Introduction, 15.

184

Daniel Davies

motions. In the earlier passage the intelligences are not posited, although Maimonides raises the question whether or not they exist, but they are used in the later chapters. The active intellect is also excluded from the earlier chapter, even though Maimonides refers to a number of ways in which the spheres’ motions influence the sublunar world. Stern concludes that “the reason Maimonides fails to mention separate intellects where we would expect to find them in I: 72 is precisely because of his own objections against them” (257). Maimonides could have mentioned the intelligences in I: 72, but perhaps it is not so surprising that he omits them. It seems reasonable to defend the straightforward view that Maimonides excludes the active intellect from that chapter, and does not argue that the intelligences exist, because it is not the subject of the chapter. The section that Stern points to is concerned, instead, with the different kinds of motions, and is stating that sublunar motion is natural, whereas the spheres are moved by animating principles. Again, Avicenna might be in the background to this passage, as Maimonides’ explanations here are very similar to those in his Physics.39 By saying that the spheres are moved by souls, Avicenna simply means that their motion is not natural, as natural motion has an end point. The elements move with a natural motion toward their natural locations, where they come to rest, and they move unnaturally, or violently, when they are moved away from their natural place of rest. But the spheres have no natural places of rest since they are perpetually in motion without changing place. Furthermore, a natural motion occurs in a straight line, which is the quickest route towards its end point, the place at which a body naturally rests, whereas the spheres have circular motions. Since the spheres do not move directly toward their natural places, as their places do not change when they move, their motion is caused by a soul rather than by nature or force. Like Maimonides, Avicenna makes no mention of the intelligences in these chapters since they are not the subject of his discussion. He is merely explaining the different kinds of motions and arguing that the heavens must be animate because their motion is circular rather than rectilinear. Given that Maimonides’ primary purpose seems to be describing the different motions, it is not surprising that he does not mention the intelligences, and since he is not concerned in this chapter with how humans know the external world, it is not surprising that he does not mention the active intellect either.

39 Avicenna, Kita¯b al-Naja¯t (Salvation), ed. Majid Fakhry (Beirut: Da¯r al-Afa¯q, 1982), 146–7.

Silence, Skepticism, and Vulgar Theology

185

Stern includes a wealth of philosophical arguments and close textual readings to support his claims, and the preceding paragraphs have only dealt with two points of disagreement.40 Moreover, even if these particular pieces of evidence are rejected, a reader could still be challenged by others, including Stern’s analysis of Maimonides’ arguments about the varied motions of the spheres. He states that Aristotle posited the intelligences in order to account for their differences, as he recognized the need for a particularizer, but he then argues that Aristotle “gains nothing by this.”41 Since Maimonides makes this point explicitly, it is not hidden from the vulgar reader, and a straightforward interpretation must also take it into account. The difference between Stern’s skeptical reading and a vulgar reading is that Stern limits this particularization to the superlunar realm, claiming that Maimonides simply “baptizes” our lack of understanding, whereas the vulgar, along with the average believer, also extends it to this world. The reason that particularization is likely, in Maimonides’ view, is that there are matters that cannot be understood by human reason, as they do not follow the scientific laws with which we are familiar. In one of the clearest skeptical comments that Stern cites, Maimonides discusses matters that cannot be demonstrated and states that “the things about which there is this perplexity are very numerous in divine matters, few in matters pertaining to natural science, and nonexistent in matters pertaining to mathematics.”42 This comment raises the question of which matters in physics are also in doubt. Aside from those matters about which Avicenna expresses doubts, Maimonides also connects particularization to a number of topics in physics that reason, he states, cannot decide.43 For example, Avicenna argues in the Physics that there cannot be a beginning to time or motion, but Maimonides claims that there is no way to know whether or not that is true since such a beginning cannot be conceived. In a straightforward reading, the notion of God’s particularization is used to justify arguments for creation at an absolute beginning. Maimonides then 40 I should note that Stern does not only rely on silence. He also builds on the image of a human as a microcosm, which Maimonides uses in I: 72. An alternative reading is given by Tzvi Langermann, who argues that the way in which Maimonides explains the analogy is intended to distance his own views from the astrological opinions of the Brethren of Purity. See Langermann, “Maimonides’ Repudiation of Astrology,” Maimonidean Studies 2 (1991), 144–5. 41 Guide II: 19, 308. 42 Guide I: 31, 66. 43 E.g., Avicenna expresses uncertainty about how the soul attaches to the body. See The´re`se-Anne Druart, “The Human Soul’s Individuation and its Survival after the Body’s Death: Avicenna on the Causal Relation between Body and Soul,” Arabic Sciences and Philosophy 10 (2000), 259–73. Like Maimonides, he also states that “the human faculty falls short of knowing all” that is in the heavens: e.g, Al-Shifa¯’: al-Tabı¯‘iya¯t (The ˙ Healing: Physics), ed. Mahmoud Qassem (Cairo: Da¯r al-Ka¯tib al-‘Arabi, 1969), 49.

186

Daniel Davies

connects belief in an absolute beginning to belief that miracles are possible. Miracles are bound up with God’s knowledge of particulars. If God knows particulars, God can act on particulars, and anomalies that depart from the regular, expected order of events are therefore possible. The problem for this account is that particulars are not the sorts of things that can be known scientifically if, as Stern has shown, Maimonides follows the Aristotelian account of scientific knowledge as one that involves universal judgments. God’s knowing particulars therefore emphasizes that we do not understand God, and that the word ‘knowledge’ can only be used equivocally. As above, only that way can it satisfy Maimonides’ negative theology. Moreover, by arguing that we are unable to rule out anomalies that do not conform to universal scientific knowledge, Maimonides effectively limits the certainty that we can have about the sublunar world, even when our knowledge is of the highest kind available. Perhaps ironically, the straightforward reading limits human knowledge of the sublunar world even more than the skeptical reading.44 It is a pleasure to read this genuinely philosophical book on the Guide, and it deserves to receive wide critical engagement. Some of Stern’s interpretations are motivated by certain debates in current scholarship, but they extend previous streams of thought with longer histories in a new direction. He refines the skeptical reading, adding philosophical rigor and offering new literary analyses. As presented in this book, Maimonides is a serious philosopher whose constant skeptical questioning generates humility by emphasizing what we do not and cannot know, and also encourages scientific inquiry into what we can. He could serve as a role model for philosophers of religion even today. Clare Hall, Cambridge

BIBLIOGRAPHY Adamson, Peter. “Avicenna and his Commentators on Human and Divine SelfIntellection,” in Dag Nikolaus Hasse and Amos Bertolacci (eds), The Arabic, Hebrew and Latin Reception of Avicenna’s Metaphysics (Berlin: De Gruyter, 2012), 97–122. Aristotle. Posterior Analytics, ed. and tr. Hugh Tredennick (Cambridge, MA: Harvard University Press, 1960).

44 Tzvi Langermann explains that Maimonides’ belief in the possibility of miracles limits scientific understanding. See Langermann, “Maimonides and Miracles: The Growth of a (Dis)belief,” Jewish History 18 (2004), 147–72.

Silence, Skepticism, and Vulgar Theology

187

Avicenna. Al-Shifa¯’: al-Tabı¯‘iya¯t (The Healing: Physics), ed. Mahmoud Qassem ˙ (Cairo: Da¯r al-Ka¯tib al-‘Arabi, 1969). Avicenna. Kita¯b al-Naja¯t (The Salvation), ed. Majid Fakhry (Beirut: Da¯r al-Afa¯q, 1982). Burnyeat, Myles. Aristotle’s Divine Intellect (Milwaukee: Marquette University Press, 2008). Davidson, Herbert. Moses Maimonides: The Man and his Works (New York: Oxford University Press, 2005). Davies, Daniel. “Davies on Stern, ‘The Matter and Form of Maimonides’ Guide’ ” . Davies, Daniel. Method and Metaphysics in Maimonides’ Guide for the Perplexed (New York: Oxford University Press, 2011). Druart, The´re`se-Anne. “Alfarabi and Emanationism,” in John F. Wippel (ed.), Studies in Medieval Philosophy (Washington, DC: Catholic University of America Press, 1987), 23–43. Druart, The´re`se-Anne. “The Human Soul’s Individuation and its Survival after the Body’s Death: Avicenna on the Causal Relation between Body and Soul,” Arabic Sciences and Philosophy 10 (2000), 259–73. Freudenthal, Gad and Mauro Zonta. “The Reception of Avicenna in Jewish Cultures East and West,” in Peter Adamson (ed.), Interpreting Avicenna (Cambridge: Cambridge University Press, 2013), 214–41. Hart Green, Kenneth. Leo Strauss and the Rediscovery of Maimonides (Chicago: University of Chicago Press, 2013). Harvey, Warren Zev. “Maimonides’ Avicennianism,” Maimonidean Studies 5 (2008), 107–19. Hasse, Dag Nikolaus. “Avicenna’s Epistemological Optimism,” in Peter Adamson (ed.), Interpreting Avicenna (Cambridge: Cambridge University Press, 2013), 109–19. Heck, Paul. Skepticism in Classical Islam: Moments of Confusion (London: Routledge, 2014). Kosman, Aryeh. “Metaphysics’ ¸9: Divine Thought,” in Michael Frede and David Charles (eds), Aristotle’s Metaphysics Lambda: Symposium Aristotelicum (Oxford: Oxford University Press, 2000), 307–26. Kraemer, Joel. “Maimonides, the Great Healer,” Maimonidean Studies 5 (2008), 1–30. Kukkonen, Taneli. “Al-Ghaza¯lı¯’s Skepticism Revisited,” in Henrik Lagerlund (ed.), Rethinking the History of Skepticism: The Missing Medieval Background (Leiden: Brill, 2010), 29–59. Langermann, Tzvi. “Maimonides and Miracles: The Growth of a (Dis)belief,” Jewish History 18 (2004), 147–72. Langermann, Tzvi. “Maimonides’ Repudiation of Astrology,” Maimonidean Studies 2 (1991), 123–58. Maimonides, Moses. The Guide of the Perplexed, tr. Shlomo Pines (Chicago: University of Chicago Press, 1963).

188

Daniel Davies

Manekin, Charles. “Maimonides on Divine Knowledge: Moses of Narbonne’s Averroist Reading,” American Catholic Philosophical Quarterly 76 (2002), 51–74. Pines, Shlomo, “The Limitations of Human Knowledge according to al-Farabi, ibn Bajja, and Maimonides,” in Isadore Twersky (ed.), Studies in Medieval Jewish History and Literature (Cambridge, MA: Harvard University Press, 1979), 82–109. Pines, Shlomo. “Translator’s Introduction,” in The Guide of the Perplexed (Chicago: University of Chicago Press, 1963), lvii–cxxxiv. Seeskin, Kenneth. Maimonides on the Origin of the World (Cambridge: Cambridge University Press, 2005). Seeskin, Kenneth. Searching for a Distant God: The Legacy of Maimonides (New York: Oxford University Press, 2000). Stern, Josef. “Maimonides’ Epistemology,” in Kenneth Seeskin (ed.), The Cambridge Companion to Maimonides (Cambridge: Cambridge University Press, 2005), 105–33. Stern, Josef. The Matter and Form of Maimonides’ Guide (Cambridge, MA: Harvard University Press, 2013). Strauss, Leo. Persecution and the Art of Writing (Chicago: University of Chicago Press, 1988). Weiss, Roslyn. “Four Parables about Peshat as Parable,” in Shlomo Carmy and Yamin Levy (eds), The Legacy of Maimonides: Religion, Reason and Community (New York: Yashar Books, 2006), 111–26.

Aquinas on the Beginning and End of Human Life A Rejoinder to Patrick Toner Fabrizio Amerini

Despite its announced “relentless” and “pointed” criticism, I sincerely appreciated Patrick Toner’s review. It is the outcome of a close and engaged reading of my book, and for this I feel quite honored. In the first part of his study, Toner expresses interpretative worries about my reconstruction of Aquinas’ embryology, while in the second part, he faults me for not criticizing Aquinas’ embryological teaching in the light of today’s science and, as a consequence, for not replacing Aquinas’ explanation of human embryogenesis with a new one. The exegesis of texts, however, is not the real target of his criticism, since Toner at times acknowledges that my interpretation of Aquinas is plausible and textually grounded.1 So most of his disagreement concerns the correct way one should approach medieval doctrines and which consequences one could draw from one’s investigation of medieval thought for today’s philosophy. In particular, his interest in my book leans in the direction of current philosophical discussions about delayed hominization (p. 213), so his review of the final chapter is actually the core of his criticism. In the present article I will focus on this part of Toner’s criticism; I will discuss elsewhere his remarks concerning my interpretation of Aquinas’ texts.

1 See e.g. p. 218. Unless otherwise noted, page references in this paper are from Patrick Toner, “Critical Study of Fabrizio Amerini’s Aquinas on the Beginning and End of Human Life,” Oxford Studies in Medieval Philosophy 2 (2013), 211–28. The references to my book are from Aquinas on the Beginning and End of Human Life, tr. Mark Henninger (Cambridge, MA: Harvard University Press, 2013); it is the English translation of my Tommaso d’Aquino: origine e fine della vita umana (Pisa: ETS, 2009), with the addition of a final bibliography.

190

Fabrizio Amerini

Toner’s judgment of Aquinas’ embryology is quite negative. He considers it as “to say the least . . . completely bogus” (p. 219), and from this he concludes that there is no reason to follow it today. I cannot say if Aquinas’ embryology is bogus or not. What I know is that it is a metaphysical explanation of human embryogenesis and as such it should be evaluated; in this respect, it could be still helpful today. In particular, it is a philosophically first-rate interpretation of Aristotle’s embryology and this was the point of interest for me in the book. Nevertheless, I can understand Toner’s dissatisfaction with such an embryology. One may agree with him that Aquinas’ teaching contains several approximations and scientifically unfounded beliefs. Among the many others, today’s embryology has definitively proven the contribution, both material and formal, given by the mother to the transmission of the genetic characters to the child, while for Aquinas—as is known—the father is solely responsible for the “formal” identity of the child.2 No doubt, biology is not the right plane on which Aquinas’ embryology should be evaluated. But the true question, to be clear, is not which scientifically correct account of human embryogenesis can be extracted from Aquinas’ texts, but rather the following: “given what we know about embryology, would St Thomas still hold to delayed hominization?” (p. 220) My answer in the book was yes, but Toner retorts that I have not adequately defended the assertion on which my argument ultimately rests, that is, the Aristotelian characterization of the human soul as the form or act of an organic physical body having life potentially (p. 220). I find Toner’s criticism ungenerous. The declared purpose of the book was not to give arguments for defending or rejecting Aquinas’ embryology, but to clarify Aquinas’ embryological teaching through a close inspection of his texts, and his teaching attributes a key role to Aristotle’s characterization of soul. Aquinas takes it as the starting point of his argument. I simply followed him and tried to explain in the book why he could have thought so. The argument to which Toner resorts in his study for rejecting my interpretation is the standard way of reconciling Aquinas’ undeniable refusal of immediate hominization with today’s embryological knowledge. In general, as I wrote in the book, as a historian of philosophy I find it barely promising to speculate about Aquinas’ change of mind on hominization in the light of today’s science, about what he might have thought but never actually thought. How could we verify this change? For the sake of discussion, however, we might ask about it. My position, in this case, is that

2 See Amerini, Aquinas, ch. 8, pp. 212–13; also ch. 1, for a general presentation of Aquinas’ embryology.

Aquinas on the Beginning and End of Human Life

191

given what we know today about human embryogenesis, Aquinas would still hold to delayed hominization. I shall give the reason in a moment. But before, let me clarify a point. Let us suppose that Aquinas would change his mind about delayed hominization if he was acquainted with what we know today about human embryogenesis. In this case, I would expect an interpreter to attribute to Aquinas the view that hominization takes place at conception immediately, i.e., before any bodily organization has started. But as it seems, this is not Toner’s view. It is rather unexpected to find Toner defending immediate hominization and holding, at the same time, that by “organic body” (corpus organicum) Aquinas would intend some kind of primordial body, although not a body with the proper sorts of organs: “Must one reject such a claim [i.e., that the infusion of a human soul presupposes a body with the proper sorts of organs] while endorsing immediate hominization? I would say no. The organs in question, in this case, are the organs that will—unless prevented—bring about the presence of the whole human nervous system: that is to say, the many various cells and/or cell parts that make up the early embryo” (p. 221). Here I find Toner at the very least forcing Aquinas’ view. Cells and/or cell parts cannot be considered “organs” from Aquinas’ perspective; at best, they are matter disposed to become specialized parts of the organism, as Toner implicitly acknowledges when he says that they are “the organs that will—unless prevented—bring about the presence of the whole human nervous system.” In fact, even such specialized matters as flesh, sinew, and bones are not organs for Aquinas, but parts of the vital organs (like heart and brain) and of the functional bodily parts (such as feet and hands)—as the text which Toner quotes from Aquinas’ commentary on De anima clearly states.3 In any case, even conceding that such matters are “remarkably differentiated and specialized and complex” (p. 221) just as organs are, it still remains true that, as they are, they are actually unable to support the superior vital functions, especially the intellective functions, of the human being—unlike organs taken in “the robust way” (p. 221), as Aquinas understands them, can do. But being in the actual condition to be able to support all the vital functions depending on the human soul appears to be crucial, from Aquinas’ perspective, for discriminating the case of true organs from that of parcels of matter (i.e., specialized cells or/and cell parts) that may bring about such organs. 3 “An animal’s foot is composed of various parts (flesh, sinew, bones, and so forth), but the organs of plants do not have such a variety of parts out of which they are composed” (p. 221). The reference is to Sentencia libri de anima II.1 (ed. Leonine, p. 72.345–9).

192

Fabrizio Amerini

Toner, however, replies: “True, the early embryo cannot (let us suppose) actually engage in high level human thought. But neither can the normal human infant or college sophomore. But whether it can actually engage in high level human thought is not the decisive factor, as Amerini concedes the case of anencephalic infants shows. The question is, does the early embryo have the right sort of potentiality so that St Thomas’s philosophical thought would, when combined with modern embryology, lead him to think that the early embryo is capable of rational ensoulment? I see nothing in Amerini that gives me any reason to think not” (p. 222). This way of speaking reveals that not all the qualifications I introduced in the book to specify Aquinas’ position have been retained. It is quite evident that, for Aquinas, anencephalic infants are in a state in which they could engage in intellective operations—or do that again, or for the first time—if the natural process, per impossibile, had gone well, or if, by some sort of medical or even divine intervention, their condition could be repaired. But embryos at the early stages are not actually provided with such organs, nor can they possibly be provided with them, so they cannot or could not actually engage in intellective operations. On the other hand, saying that embryos at the early stages are nonetheless potentially able to engage in intellective operations—in the sense that they, unless prevented, will become able to engage in intellective operations as soon as the proper organs have been formed—is not sufficient for saying that embryos at the early stages are actually rationally ensouled. With respect to this, note that Toner’s counterargument concerning the newborn human infants is discussed by Aquinas too, but it is dismissed as an inconclusive argument. For Aquinas, newborn human infants, although not able completely and fully to exercise intellective operations due to the consistency of the cerebral mass being still too “moist,” completely and fully have the actual possession of the capacity to exercise intellective operations. Here Aquinas’ distinction between first and second act is crucial. In his vocabulary, ‘first act’ designates a thing’s possession of a given capacity for exercising an operation (e.g., the soul’s capacity for thinking), while ‘second act’ designates the actual exercise of this capacity (i.e., the soul’s actual act of thinking). According to this distinction, newborn human infants have rational ensoulment perfectly in first act, but not yet perfectly in second act, however, this is enough to consider them completely and fully human beings. I have discussed this case in the book,4 but Toner fails to note this point. My principal argument for thinking that Aquinas would not change his mind about delayed hominization is that, as I have written in the book, 4 See Aquinas, ch. 7, pp. 192ff.

Aquinas on the Beginning and End of Human Life

193

“Thomas is of the opinion that one must not start from any description of the matter to establish when the rational soul arrives, but, vice versa, one must start from a given definition of the rational soul and from there move to specify the conditions under which the matter can be considered in near potency to being rationally ensouled.”5 On this principle, no scientific refinement of what we know about embryonic matter at conception can give the reason for the moment God waited to infuse the rational soul into the body. Nor can the act of conception as such justify God’s act of infusion. My argument in the book was that, “if one assumes that the occasion is given by conception as such, such a choice seems completely arbitrary on the part of God and entirely detached from any possibility we have of explaining it rationally. From this standpoint, nothing stops one from adopting as the absolute starting point of the embryo’s hominization (and more coherently with the embryology of Thomas) the moment when the semen is detached from the father.”6 I find this argument quite misrepresented in Toner’s study. My point in the book was not to show that today’s embryology would underwrite the argument that, if we try to push rational ensoulment back to the moment of conception, then we should keep pushing it back to the detachment of the semen from the father. This is obviously not the case, for the father’s semen contains only half the genetic code of the embryo. My point was rather that today’s embryology provides detailed information especially about the matter of the embryo, but Aquinas builds his embryology by attributing primacy to form, for, in his metaphysics, it is the form that confers being and identity on the matter. Since the form is passed on to the embryo by the father through the semen, to have such detailed knowledge of what happens to embryonic matter at conception cannot be the reason to predate hominization precisely to conception. Aquinas would consider it as an insufficient reason, that is, insufficient to prevent pushing back hominization all the way to the detachment of the semen from the father—as Toner himself recognizes (p. 224). In any case, Aquinas would consider today’s embryology as insufficient grounds for discarding the (Aristotelian) characterization of the human soul as the form or act of an organic physical body having life potentially. But regardless of how my argument can be assessed, from a bioethical point of view, I cannot see why an interpreter would need to place Aquinas among the advocates of the immediate hominization of the embryo. For the reason that will become clear in a moment, the denial of hominization

5 See Aquinas, ch. 8, p. 217. Toner quotes this principle at p. 221. 6 See Aquinas, ch. 8, p. 217.

194

Fabrizio Amerini

at conception has, of itself, no special effect on the problem of abortion— the one I have at greatest length discussed in the book—so attempts at updating and/or adjusting Aquinas’ teaching (and/or our interpretations of Aquinas’ teaching) in the light of today’s embryology are simply pointless. Such accounts come across as, if I may say, unfruitful ways of approaching Aquinas’ embryology. In the book I suggested a different path, more practicable to my mind. An interpreter who wishes to respect what Aquinas actually says in his writings about the human embryo would detach the question of abortion (and the problem of assessing any decision and/or argument concerning abortion) from the metaphysical explanation of human embryogenesis proposed by Aquinas. The reason is that a decision in favor or against abortion may (and should) be founded on considerations other than those concerning the moment of the hominization of the embryo. With respect to this, in the book I proposed distinguishing between two ways of inquiring into human embryogenesis: (i) the inquiry into the typical properties of the embryo and the moment of their acquisition, and (ii) the inquiry into the identity of the subject-embryo through the process of generation. I developed such a proposal in the central chapter of the book, the fifth, “The Identity of the Embryo.” Toner finds it a “deflating” chapter for, after reconstructing and discussing Aquinas’ texts, I raised objections against any possible way of saving, from Aquinas’ perspective, the transtemporal identity of the embryo. This impression is, however, unfounded. At the end of my discussion of the difficulties of Aquinas’ argument, I offered a positive solution to the problem of the identity of the embryo, and my solution consisted in proving that Aquinas detaches the identity of the subject of generation from the identity of its matter and/or form. Aquinas seems to think that a thing can incessantly change its matter and/or form while continuing to be the same thing, although not the same thing in number. This precisely happens in the case of the human embryo. In the process of generation, only one subject is involved, but its numerical identity does not depend ultimately on the identity of its matter and/or its form. All this induces us to consider the identity of the subject of generation a primitive fact in Aquinas’ metaphysics. The identity of the subject-embryo can be inferred from the non-interruption of its existential story within the process of generation, but it cannot be proved by referring to the identity of any more primitive thing such as its matter and/or form. My conclusion in the book was that Aquinas establishes this primitiveness only with an argument-stipulation: for him, it must be one and the same thing that can be described as passing from potency to act, from an incomplete to a complete state, i.e., from imperfection to perfection. Without entering into the details of Aquinas’

Aquinas on the Beginning and End of Human Life

195

solution,7 what matters here is that the shift of the inquiry from the temporality of the process of hominization to the identity of the subject of such a process is what justifies my proposal of keeping distinct the question of abortion and the question of the immediate vs. delayed hominization of the embryo. One could reject any act of abortion simply by giving arguments for the identity and continuity of the subject-embryo in the process of human generation. What I mean is that the rejection of abortion may be defended while conceding, at the same time, first, that the human embryo does not remain the same subject in number (since neither its matter nor its form remain the same in number), and second, that it becomes humanly ensouled only after some time of gestation. If one takes seriously the above distinction between the two ways of inquiring into human embryogenesis, there seems to be no longer any need to update or adjust Aquinas’ teaching (and/or our interpretations of it) for reconciling it with immediate hominization. University of Parma

BIBLIOGRAPHY Amerini, Fabrizio. Aquinas on the Beginning and End of Human Life, tr. M. Henninger (Cambridge, MA: Harvard University Press, 2013). Amerini, Fabrizio. “The Puzzle of the Embryo: Thomas Aquinas, Hylomorphism, and Identity over Time,” in R. Guerizoli (ed.), Proceedings of the 2nd Rio Colloquium: Hylomorphism in the 13th and 14th Centuries (forthcoming). Amerini, Fabrizio. Tommaso d’Aquino: origine e fine della vita umana (Pisa: ETS, 2009). Thomas Aquinas. Sentencia libri de anima, ed. R.-A. Gauthier (Rome: Commissio Leonina, 1984). Toner, Patrick. “Critical Study of Fabrizio Amerini’s Aquinas on the Beginning and End of Human Life,” Oxford Studies in Medieval Philosophy 2 (2013), 211–28.

7 For an extensive treatment see Aquinas, ch. 5, esp. pp. 133ff. For further discussion, also see Amerini, “The Puzzle of the Embryo: Thomas Aquinas, Hylomorphism, and Identity over Time,” in The Proceedings of the 2nd Rio Colloquium: Hylomorphism in the 13th and 14th Centuries, ed. R. Guerizoli (forthcoming).

Briefly Noted A. D. Smith, Anselm’s Other Argument (Harvard University Press, 2014). Even if you have thought about the ontological argument enough for several lifetimes, it still must be said that this new book on a very old topic is the perfect model of historicalphilosophical criticism. Better known for his work on the philosophy of perception and its early modern history, David Smith has now given us an immaculately researched and argued study of Anselm’s famous proof. The book has almost nothing to say about the main argument of Proslogion 2, but instead focuses on chapter 3 and related remarks in the Reply to Gaunilo, where Smith argues that a second argument for God’s existence can be found, one that he takes to be sound and compelling. This is not the familiar Modal Ontological Argument of Hartshorne, Malcolm, and Plantinga, which Smith argues at length is not to be found in Anselm. Instead, it runs off the key principle that “If anything that cannot be conceived to be caused can be conceived to exist, it actually exists” (150). Although this argument has received little attention (at least not since Scotus), Smith contends that it is the most persuasive version of Anselm’s proof because its plausibility rests on little more than a fairly weak version of the Principle of Sufficient Reason. One might worry that the key principle proves too much—think of all sorts of platonic abstracta—but Smith’s discussion throughout is masterful, ranging over large swaths of secondary literature and drawing on far-flung aspects of Anselm’s thought to illuminate these famous few pages. This is mandatory reading for future work on the topic. Sarah Pessin, Ibn Gabirol’s Theology of Desire: Matter and Method in Jewish Medieval Neoplatonism (Cambridge University Press, 2013). The first Englishlanguage monograph on Ibn Gabirol’s philosophy, this deeply learned study situates the eleventh-century Spanish Jew in his own context. Rather than serving as a stepping stone to Thomistic Aristotelianism, Pessin’s Ibn Gabirol gets read in light of Arabic Neoplatonism. Going beyond the Latinized Fons vitae that is his main surviving philosophical work, Pessin reconstructs the conceptual framework of the Arabic original, and connects it with Ibn Gabirol’s Hebrew poetry. Instead of a universal hylomorphism of prime matter, understood in Aristotelian terms, we are given a Grounding Element that is the expression of divine desire, and that arises out of pseudo-Empedoclean sources. Instead of a Divine Will from which the world is voluntarily created, Pessin sees in Ibn Gabirol a Divine Desire that grounds an emanationist theology. In language that crosses over into the mystical-poetical, Pessin describes “a pulsing field of life” (27) that has God’s desire at its core. Forget Aristotelian matter as the mundane substratum of change. Open yourself up to the Grounding Element of Love, infused with “the breath of sacred yearning.” Christophe Grellard, De la certitude volontaire: de´bats nominalistes sur la foi a` la fin du Moyen Aˆge (Publications de la Sorbonne, 2014). This brief but ambitious volume (156 pp.) attempts to chart the course of a debate through the scholastic era

198

Briefly Noted

on the foundations of faith. The debate centers on whether faith can arise from a pure act of will, or must instead have some kind of cognitive foundation. As with so much recent work in the field, especially in Europe, Grellard treats Aquinas and other thirteenth-century authors as mere precursors to the more sophisticated discussions of the fourteenth century. His thesis is that one can identify a voluntaristic current of thought that gets articulated in its most extreme form by Ockham, and then ultimately restated more modestly by later nominalists such as Peter of Ailly and, later, John Major. For Ockham, faith must, by its very nature, rest ultimately on “an initial decision of free will” (73). This view came under powerful attack by Robert Holcot, who denied that we can ever believe something without cognitive support, just because we will to do so. Later authors, feeling the force of Holcot’s criticisms, offer a more moderate form of voluntarism, and this becomes the consensus nominalist view. The book is much too short on detail to make a wholly compelling case that this is the essential shape of the debate. But the issues discussed are, all the same, exceedingly interesting, with implications that run well beyond the narrow question of religious faith. The topic cries out for further study. Therese Scarpelli Cory, Aquinas on Human Self-Knowledge (Cambridge University Press, 2014). Originally a dissertation at Catholic University of America, Cory’s book offers a glimpse of how the various methodological divides among historians of medieval philosophy might someday be smoothed into a single, global style. Like CUA theses in the grand-old manner, Cory takes us through the relevant texts in chronological order, complete with exhaustive references to the secondary literature. But that’s just the prologue to the real work of the study, which in chapter 3 commences to pose the sorts of hard questions that philosophically trained historians like to ask, e.g., “Does engaging in intellectual activity merely provide the opportunity for me to consider myself, or do I necessarily cognize myself in all my mental acts?” (134). Cory answers questions such as this in creative ways, venturing rather far from the letter of the texts, but always grounding her answers in Aquinas’s own conceptual framework. Yet this is no narrow Thomism. Cory attempts—albeit with mixed results—to situate the discussion both relative to earlier medieval debates, Latin and Arabic, and relative to the better-known early modern discussions of Descartes and Hume. Mixing in a steady stream of plain-spoken anecdotes from ordinary life, the book manages to assimilate the best of various approaches. The results, even if not always persuasive, augur well for the future. Dallas G. Denery II, Kantik Ghosh, and Nicolette Zeeman (eds), Uncertain Knowledge: Scepticism, Relativism, and Doubt in the Middle Ages (Brepols, 2014). This interesting volume attempts to pull together work from various medieval fields on questions pertaining to uncertainty. Most often, the juxtaposition serves only to highlight the disciplinary differences. Sweeney, Perler, and Grellard do the sorts of things that readers of this volume would expect. The historians talk about fairly similar material, but in predictably different ways. The discussions of literature run still further afield. Anyone interested in these topics will find much of interest—my favorite was Helen Swift’s discussion of hope and despair in French romantic poetry. One would like to find a way to shuffle the conversation in such a way as to produce genuine interdisciplinarity. As it is, everyone talks about what they know best. And perhaps, in the end, that is for the best.

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

Index Alberic of Paris 45–7 Anselm of Canterbury 132 Aristotle 33, 45, 50–1, 102, 157, 165, 179n Augustine 1–30, 36–7, 102 Avicenna 64–74, 76–7, 79, 81–3, 89, 91–2, 95–6, 179, 184

Peter Abelard 30–61 Peter John Olivi 66, 86–92, 93, 95 Peter of Spain 75–8, 92–3 Pines, Shlomo 171–2, 177, 179, 181 Porphyry 40–3 Pseudo-Rabanus 49–51, 59

Blackburn 20n Burnyeat, Myles 2–5, 13, 27

Strauss, Leo 182 Sylvester of Ferrara 100, 106, 109, 110

Cajetan (Thomas de Vio) 121–2 Dominicus Gundissalinus 70–2 John Capreolus 100, 104, 105, 109, 110 John Duns Scotus 122–4, 128–43 John of la Rochelle 78–80 Maimonides 169–86 Matthew of Aquasparta 81–3 Panaccio, Claude 147n Pasnau, Robert 116n

Tarlazzi, Caterina 57n Thomas Aquinas 80–1, 99–121, 123–4, 139, 189–95 Toner, Patrick 189–93 Vital du Four 83–6 William of Auvergne 72–4 William of Ockham 145–56, 158–67 Zimmerman, Dean 124