sign in sign up
<<
Home , Actus primus, Differentia, Isagoge, Medieval philosophy, Nominalism, Sophismata

ASHGATE STUDIES IN MEDIEVALPHILOSOPHY Ockham on

Series Editors

John Marenbon, Trinity College, Cambridge, UK Scott MacDonald, Cornell , USA Christopher J. Martin, University of Auckland, New Zealand Simo Knuuttila, Academy of Finland and the University of Helsinki, Finland

The study of medieval is flourishing as never before. Historically precise and philosophically informed research is opening up this large but still relatively unknown part of philosophy's past, revealing - in many cases for the first - the nature of medieval thinkers' and the significance of1teir philosophical CLAUDE PANACCIO achievements. Ashgate Studies in presents some of the best of University of Quebec in Montreal, Canada this new work, both from established figures and younger scholars. Chronologically, the series stretches from c.600 to c.1500 and forward to the scholastic of sixteenth and early seventeenth-century Spain and Portugal. The series encompasses both the Western tradition, and the Byzantine, Jewish and Islamic traditions. Authors all share a commitment both to historical accuracy and to careful analysis of arguments of a kind which makes them comprehensible to modern readers, especially those with philosophical interests.

Other titles in the series:

Theology at Paris, 1316-1345 Peter Auriol and the Problem of Divine Foreknowledge and Contingents Chris Schabel ISBN 0 7546 0204 4

Medieval Modal Systems Problems and Concepts Paul Thom ISBN 0 7546 08336

ASHGATE © Claude Panaccio 2004 .. Contents 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, electronic, mechanical, photocopying, recording or otherwise without the prior permission of the publisher. ix The author has asserted his moral right under the Copyright, Designs and Patents Acknowledgements Act, 1988, to be identified as the author of this work. Abbreviations xi 1 Published by Introduction Ashgate Publishing Limited Ashgate Publishing Company Notes 3 Gower House Suite 420 5 Croft Road, Aldershot 101 Cherry Street 1: , and Mental Hampshire GUll 3HR Burlington, VT 05401-4405 1..Intuitive and abstractive cognitions 5 England USA 2. Mental language 8 IAshgate website: http://www.ashgate.com I 3. Abstraction and universals 9 11 British Library Cataloguing in Publication Data 4. Intuition and singular terms Panaccio, Claude, 1946- 5. Mixed cognitions and singular terms of the third type 14 Ockham on concepts. - (Ashgate studies in medieval philosophy) 15 1. , ca. 1285-ca. 13492. Concepts 6. Ockham's representationalism 1. Title Notes 17 189.4 2: Intellectual Acts 21 Cataloging-in-Publication Data 1. Mental acts and habitus 21 Panaccio, Claude, 1946- 23 Ockham on concepts / Claude Panaccio. 2. From the fictum to the actus p. cm. - (Ashgate studies in medieval philosophy) 3. The rejection of intelligible species 27 Includes bibliographical references and index. 3.1 Species as preconditions for intellectual acts 28 ISBN 0-7546-3228-8 (alk. paper) 1. William, of Ockham, ca. 1285-ca. 1349. 2. Philosophy, Medieval. 3.2 The razor against intelligible species 29 3. Concepts. 31 1. Title. II. Series. 4. Combining acts 4.1 Simple and complex units 31 B765.034P36 2004 32 121 '.4-dc22 2004008560 4.2 Propositional acts 4.3 Iudicative acts 35 ISBN 0 7546 32288 Notes 36 Printed and bound in Great Britain·by Antony Rowe Ltd, Chippenham, Wiltshire

Typeset by DC Graphic Design Ltd, Swanley, Kent ,.-.

vi OCKHAM ON CONCEPTS OCKHAM ON CONCEPTS vii

3: Concepts as 45 7: Concepts as Similitudes 119 1. The problem: how can a ever be a ? 45 1. Similitude sustained 119 2. The two meanings of signum 47 2. Acts and similarities 122 3. Conceptual roles 51 3. Varieties of conceptual representation 125 4. or propositionalism? 53 3.1 Absolute specific quality concepts 126 5. Types and tokens 55 3.2 Specific concepts of substances 126 Notes 58 3.3 Simple connotative concepts 129 4: Connotative Terms in Mental Language 63 3.4 Simple generic concepts 131 1. Connotative terms 64 4. Two problems about absolute concepts 133 2. Mental connotation 66 4.1 Klima's objection 133 3. Synonymy and nominal 69 4.2 Brown's puzzle 136 4. Connotative terms and exponible 73 Notes 139

Appendix: A reply to Richard Gaskin 74 H: Logical Concepts 145 Notes 77 I. The earlier theory: logical words internalized 146 5: The Role of Nominal Definitions 85 2. Logical constants in the actus-theory 151 1. Four theses about nominal definitions 86 3. Prepositions and non-standard copulas 155 2. What defining amounts to: a reconstruction 89 Notes 158

3. Some consequences 93 l): The of Words 165 3.1 Definitions and abbreviations 93 I. Subordination 165 3.2 Possession of concepts and of definitions 94 2. Types and tokens again 171 3.3 Real orderings 95 3. Reverse subordination? The instructive case of proper names 173 Notes 97 Notes 176 6: Cognition and Connotation 103 Conclusion 181 1. Spade's questions 103 Notes 186 2. The acquisition of simple connotative concepts 106 Bihliography 187 3. The adequacy of nominal definitions 110 Index 195 4. Ockham and the Classical 113 Notes 116 Acknowledgements

Most of this book was written between the winter of 2002 and the fall of 2003, as I was releaSed from my regular teaching load; first by a sabbatical leave from the University of Quebec in Trois-Riyieres, and then by a Killam Scholarship from the Canadian. Council for the Arts. The Social Sciences and Humanities Research Council of Canada also greatly contributed to the fulfilmel)t of the project by awarding it a three-year research grant from 2000 to 2003. I wish to express my sincere gratitude to all three organizations for their generous support. It should be mentioned that some of the developments to be read here partly correspond to previously published papers of mine:

• Chapter 1 is closely based on an article originally written in French (,Intuition, abstractionet langage mental dans la theorie occamiste de la connaissance', Revue de Metaphysique et de Morale, 97/1,1992, pp. 61-81). •A preliminary version of chapter 4 has circulated for quite some time among specialists in an informal 'prepublication' format ('Connotative terms in Ockham's mental language', Cahiers d' epistemologie, no. 9016, Montreal, Universite du Quebec a Montreal, 1990, p. 21), and a slightly different one later appeared in French 'in an Italian journal ('Guillaume d'Ockham, les connotatifs et Ie langage mental', Documenti e studi sulla tradizioneJilosojka medievale, 11,2000, pp. 297-316). • Chapter 5 freely makes use of about nominal definitions that I originally presented at the XIIIth European Symposium for Medieval and held in Avignon in June 2000, in a contribution that was subsequently published in the acts ('Connotative concepts and their definitions in Ockham's ', in La tradition medievale des , Joel Biard and Irene Rosier-Catach, eds, Leuwen: Peeters, 2003, pp. 141-55). • Section 1 of chapter 9 is a revised version of a recent paper in French (,Guillaume d'Ockham et les syncategoremes mentaux: la premiere theorie', Histoire, Epistemologie, Langage, 25/2, 2003, pp. 145-60).

I am grateful to the editOJ;s and publishers involved for their role in the original dissemination of this material. The is that this book found its motivation in the extraordinarily stimulating - if sometimes critical - reactions these publications, especially the first two ones, elicited, both in print and in private conversations, from a of outstanding scholars whose names be found again and again in the following pages. Let me single out at this point those of Joel Biard, Elizabeth Karger, Calvin Normore and Paul Vincent Spade in particular, to whom this whole work, while disagreeing with them on certain points (and sometimes crucial ones), remains in the end deeply indebted, Several portions of the nook have neen orally presented to various audiences in . . . Ukraine. the US and Canada. and have greatly nenefited from x OCKHAM ON CONCEPTS 'the numerous remarks, questions and objections that were raised on these occasions Abbreviations by colleagues and students. Chapters 3, 5, 6 and 7, most saliently, are elaborations on the Pierre-Abelard Lectures I gave at the 4-Sorbonne in March 2001. Special thanks are due, in this respect, to Cyrille Michon, who not only organized these lectures, but who also provided useful and penetrating commentaries on each one of them when they were delivered. Exp. in Perih. Expositio in Librum Perihermenias Aristotelis Expositio in Libros Physicorum Aristotelis Most of all, finally, I should like to thank Cecile Juneau for her remarkably Exp. in Phys. Exp. in Porph. Expositio in Librum Porphyrii de Praedicabilibus patient and professional secretarial work upon the whole manuscript over the years, Exp. in Praedic. Expositio in Librum Praedicamentorum Aristotelis and my beloved companion Claude-Elizabeth Perreault for her constant and Exp. sup. Elench. Expositio super Libros Elenchorum affectionate support day after day whileI was - often single-mindedly! - working at OPh Guillelmi de Ockham Opera Philosophica it. Ord. Ordinatio. Scriptum in Librum Primum Sententiarum OTh Guillelmi de Ockham Opera Theologica Quaest. in Phys. Quaestiones in Libros Physicorum Aristotelis Quaest. Var. Quaestiones Variae Quodl. Quodlibeta Septem Rep. Reportatio. Quaestiones in Libros II, III, IV Sententiarum SL Summa Logicae Summ. Phil. Nat. Summula Philosophiae Naturalis Tract. de Corp. Chr. Tractatus de Corpore Christi Tract. de Quant. Tractatus de Quantitate Introduction

This book is an exercise in interpretation. My background conviction is that William of Ockham's nominalism, even if elaborated in the d,istant context of the early fourteenth century, still provides a challenging and fruitful body of theory to be in dialogue with today. In philosophy of and language, in particular, Ockham's obstinate refusal to countenance anything but concrete singular in the real , and the way he knits together on this basis an intriguing array of theses and arguments about mental language, , and reference, could have, I suspect, a healthy counterbalancing effect against the prevailing fondness in these fields for abstract entities of all sorts: general properties, natural kinds, linguistic types, Fregean propositions, and what not ... But surely, the prerequisite for these appealing to be actualized is that the theory be well understood! And there still exists, lam afraid, wide disagreements among specialists as to what exactly Ockham's nominalistic programme amounts to, especially with regards to cognition. This is what I want to address here. The focus will be on concepts - conceptus in Ockham's vocabulary :- taken by him as the basic units of . Several discussions have been going on around Ockham's theory of concepts in the last fifteen years or so, in the aftermath of the remarkable critical edition of his philosophical and theological writings completed at S1. University in the late 1980s, and of the publication in 1987 of Marilyn Adams's landmark synthetical study, William Ockham. My aim in the present work 'is to propose a thorough presentation and defence of how I understand Ockham's positions on the matter in the light of these recent developments. The first three chapters will deal with the basics of Ockham's approach: that concepts are the ultimate components of the language of thought, and that they are normally acquired as a result of natural causal processes triggered by direct empirical encounters with real individuals (chapter 1); that concepts are to be identified with intellectual acts, rather than with purely intentional objects (chapter 2); and that they are signs in the logical sense, and can be studied, consequently, with the technical apparatus of grammar and logic (chapter 3). The next three chapters will be devoted to one salient and far-reaching debate in recent Ockhamistic studies, in which the very viability of the theory turned out to be at stake: is our whole stock of concepts supposed to be constructed by the mind out of a limited number of simple non-relational natural kind concepts, by assembling definittons? In other words: how reductivistic is Ockham's doctrine of concepts, with respect in particular to relational terms? It will be argued that contrary to a widespread , Ockham l'un, and does, accept whut he calls 'connotative' terms - relational ones, especially - umong the basic units of the language of thought (chapter 4). His conception of what definitions arc expected to accomplish in such cases will he made explicit (chapter 5); how such concepts are supposed to he acquired will he explained (chapter 6). In the last chapters, finally, I will discuss three more questions that come 3 2 OCKHAM ON CONCEPTS INTRODUCTION out as crucial for Ockham's theory of concepts: in what sense could he subscribe to particular causal links with concrete episodes of direct acquaintance; this holds for the traditional saying that concepts are similitudes of things (chapter 7)? How are the (at least some) relational concepts as well as non-relational ones. logical constants implemented in the mind (chapter 8)? And how do spoken and Many striking similarities with recent ideas in analytic and written words receive their meanings (chapter 9)? language will become more and more ,apparent, as we go on, to those who are The emphasis all along will be on currently controversial issues of interpretation, familiar with the writings of people like , , or - most of and the developments will be supported by careful analysis of numerous relevant all - , that is, a resolute commitment to the language of thought passages and detailed discussions of alternative readings, with some assessment, hypothesis and to semantical atomism, a prominent concern for direct reference, a here and there, of criticisms addressed to Ockham, in so far, at least, as they rest on causal approach to , a strong innatist component, and even some what I take to be misunderstandings. My hope, in so doing, is that the book be useful externalistic tendencies. Yet the really distinctive feature of Ockham's theory of to medievalists, of course, but also to philosophers at large and cognitive scientists concepts, as will be stressed all along, is how it manages to harmonize all these as well, as an in-depth exploration of one of the truly great theories of intellectual interesting trends within the rigid constraints of a radically nominalistic . representation in the of Western thought. It so happens that the main points There lies in the end its deepest lesson. over which disagreements emerged in the Ockham of the last couple of decades have to do, ultimately, with some of the most central - and most difficult ­ issues in the current hectic search for a satisfactory theory of the mind: what are the Notes basic components of human thought? And how can they be representations of anything? I will not engage here in a direct philosophical discussion of Ockham's I. English translations will always be provided for quotes occurring within the main body doctrine on these points, or in a systematiccomparison ofit with today's approaches; of the text, with the Latin original versions to be found in the footnotes. Unless otherwise but my working assumption is that a precise and accurate understanding of this wide-ranging theory has now come to be highly desirable, and that the best way of indicated, the translations are mine. achieving just that at this point is by critically reviewing the detailed arguments of 2. Margolis 1998, p. 549. See also Fodor 1998. recent Ockham scholarship. In accordance with the usual requirements in the field, the account I will propose heavily rests on text analysis, and many of the relevant passages will be quoted and scrutinized in the exposition process. I I know of no better way to proceed in such affairs: historical monographs in philosophy are, first and foremost, a matter of reading texts and understanding them. With Ockham, especially, we are lucky enough to be dealing with such a rigorous writer that most of the current interpretative disagreements about his theory of concepts can be settled, I contend, on the basis of what he actually wrote. In many cases, admittedly, the reconstruction needs to be developed beyond what Ockham himself cared to, but I tried to stick, in so doing, to what he is committed to, given what he explicitly holds; such elaborations, anyway, will be clearly indicated when they occur. The theory that will emerge is surprisingly rich and unified, but quite different on the whole from what it has often been taken to be. The prevailing in the last decades was thatOckham's approach is a canonical - and especially stringent ­ variant of the so-called 'Classical' (or 'Definitional') view of concepts, according to which most of our intellectual representations are in fact complex constructs made out of a limited number of primitives; and that the mental language he postulates should be structured like a logically ideal language a la Frege-Russell. One of my main points in the book will be that this standard reading is seriously misleading on many counts. What Ockham proposes. will turn out to be closer in the end to what we call today an 'atomistic' theory of concepts, as characterized, for example, by the American Eric Margolis: a theory, namely, 'according to whiich what makes a concept the very concept that it is is not how it is related to certain other concepts but how it is related to the world.'2 However general they can get. simple concepts, in Ockham 's approach. ultimately depend. for what they represent, on their Chapter 1 Intuition, Abstraction and Mental Language

On the matter of concepts, Ockham, it is striking, has two distinct terminologies. On the one hand, he regularly resorts to the vocabulary ofcognitio or notitia, and to a distinction he adapts from John DunsScotus between cognitio intuitiva and cognitio abstractiva. On the other hand, he takes more seriously than anybody before him the that conceptual thought is a sort of inner .., an oratio mentalis, to which he systematically applies the semantical vocabulary of terminist logic (significatio, suppositio, connotatio, etc.), which was in use since the twelfth century for the analysis of spoken or written . How .are these two apparatuses linked to each other? This is, curiously enough, a point about which there is 110 v~ryelaborate discussion in Ockham's own texts. Most of the passages which make use of one of the terminologies either entirely ignore the other, or give it but a secondary place; those in which both can be found do not organize them in a very explicit theory. Yet this is something we need to be clear about if we want to understand Ockham's approach to concepts with any precision, since concepts, in his most mature theory, are simultaneously identified both with cognitive acts l and with mental signs susceptible of various referential roles in propositional contexts. 2 Chapters 2 and 3 will come back in more detail to these ideas of cognitive acts and mental signs respectively, but it will. be useful to elucidate first, in the present chapter, how the Ockhamistic doctrine can coherently equate the intuitive and abstractive acts of cognition with significant terms, capable of occurring, in person so to speak, within mental propositions. For the reader unfamiliar with the field this development will serve as a general introduction toOckham's epistemology of ~oncepts,and the first two sections of the chapter will be devoted to short, non-controversial presentations of the doctrine of intuitive and abstractive cognitions (section 1) and of the doctrine of mental language (section 2). The main goal of the chapter, however, will be to establish a number of non-trivial points about how Ockham's theory of concepts maps the cognitive states corresponding to intellectual and into logical or semantical categories (sections 3:---5).An interesting upshot of all this will be that, contrary to what several recent commentators haye suggested, (kkham's theory of concepts can legitimately be labelled as a brand of rcprcsentationalism (section 6).

I. Intuitive and abstractive cognitions

Ockbam's theory of intellectual acts is expounded mainly in the Prologue of his On/inatio and. in a much shorter version. in some of his Quodliht,ftll Qucstions.' The

~ 6 OCKHAM ON CONCEPTS INTUITION, ABSTRACTION AND MENTAL LANGUAGE 7 precise notion of an act which is at work there obviously needs further scrutiny and has been so widely discussed among modern commentators.? Such a cognition we will come back to it in chapter 2, but let us be content at this point with a would not abstract from and non-existence and would indeed cause an preliminary understanding of mental acts, in Ockham's sense, as those mental evident and true judgement about contingent matters; it would not, therefore, be an states that are actualized whenever an agent is actually thinking, feeling, willing or abstractive apprehension, in Ockham's sense, but an intuitive one. Since a perceiving something. This is basically what contemporary philosophers of mind miraculous intervention of this sort cannot be ruled out as intrinsically contradictory, call 'mental episodes'. Some mental acts in this sense are acts of the will, some are the general characterizations of intuitive and abstractive cognitions must take it into acts of desire, some are perceptual acts, and so on. The ones we are interested in are account. Under purely natural conditions, however, such occurrences never happen the intellectual acts. Ockham distinguishes quite a number of them. The .act of and I will therefore ignore them in the following discussion, confining myself in judgement (actus iudicativus), by which an agent judges a certain to be practice to the cases where intuitive cognition is concomitant with the real presence true or false, is contrasted with the act of apprehension (actus apprehensivus), which of the intuited singular thing. 8 is a mere intellectual grasping. The latter in turn can be either complex or incomplex. It must be stressed that intuitive and abstractive cognitions, as I have just The most salient example of a complex apprehensive act is the formation of a mental characterized them, are both of them acts of the . Ockham rejects the proposition, on the value of which the agent does not yet commit himself, but division of functions that was commonly accepted - especially by Thomists ­ which is usually presupposed by the judgement. The incomplex apprehensive act, on between sensation, seen as the sole mode of direct acquaintance with singular the other hand - which always precedes the complex apprehension and makes it objects in this terrestrial life, and intellection, seen as monopolized in principle by possible - further subdivides into intuitive cognition and abstractive cognition; this generality. He does adm.it, of course, that sense never reaches but is the distinction that will now detain us. individual things, and he is quite happy to allow for a sensory variety of intuitive The intuitive intellectual cognition of a thing, William explains, is this simple cognition.9 His point, however, is that the intellect too can reach the individual thing awareness of that thing in of which the intellect can judge as evident that the in its concrete presence, and that this is precisely what happens in the intellectual thing exists, if it exists, orthat it does not exist, if indeed it does not exist.4 More intuitive cognition. lO Such an intellectual intuition is required in addition to the generally, an intuitive cognition of something is this awareness of that thing which, sensory one, in Ockham's view, in order to account for the formation of evident under normal circumstances, causes the agent to form true and evident present­ existential judgements, the argument for this being the following: judgements are tensed contingent judgements about that very thing: my intuitive knowledge of Mary intellectual acts and as such belong within the intellect; but they presuppose, both as when Imeet her, for example, makes me judge as obvious that Mary presently exists, their parts and as their partial cause:s, incomplex cognitions belonging to the same that she is brown-haired, that she is sitting in front of me, that she is talking, and so 'cognitive subject' as themselves. The relevant simple apprehensions capable of on, none of which I could evidently know without such an immediate grasping of causing evident existential judgements, then, must also belong to the intellectual part her. Abstractive knowledge, by contrast, does nothing of the sort: 'through of the mind, and not merely to the senses. II Ockham's position here partially hangs abstractive cognition, Ockham writes, no contingent truth, especially in the present upon his (antithomistic) doctrine of the plurality of substantial forms: a human being tense, can be evidently known.'5 Abstractive cognition is that in virtue of which it for him - as for most of his Franciscan predecessors - is a hylemorphic compound cannot evidently be known, in particular, whether the thing presently exists or not. of a given piece of matter with a number of different substantial forms, including at Thinking of Mary in her absence does not normally allow me to evidently know least a corporeal form, a sensitive form, and an intellective form, the last two being whether she is still alive or not at this very moment, whether she is sitting or not, seen as two different 'cognitive subjects', or distinct compartments of a given mind, talking or not, and so on: this is a typical case of abstractive cognition. The situation if one prefers. '2 Ockham's argument for intellectual intuition, however, also - and is pretty much similar when I use a general concept to think about chickadees, let's even more crucially - rests on the view that the incomplex cognition which causes say, or daisies, when none of them are actually there before me. Abstraction, in this an evident existential judgement must itself be a part of that very judgement. We will sense, as Ockham makes it clear, should - at first - be understood as abstraction come back to this in some detail in section 4 below. 13 from existence or non-existence, rather than from singularity or non-singularity. 6 What we have at this point, to sum up, is that the overall cognitive process, as In the natural order of things, intuitive cognition always corresponds to the simple (kkham describes it, takes the form of a rather complex causal sequence. The and immediate grasping of something which is actually present to the agent. If the external sensible first causes a sensory intuition in the agent, and then, with thing is not there in front of me, my cognition of it can only be abstractive. That the help of this sensation, also cauSes an intellectual intuitive act, which itself brings Ockham does not simply express himself in this way is due to the fact that he is ahout (under normal circumstances) two things within the intellect: a singular and anxious, for theological , to preserve 's omnipotency, which he sees as evident true existential (or at least contingent) judgement on the one hand, and a limited only by the principle of non- - a merely logical rather than real simple ahstractive act on the other hand. This latter abstractive cognition, in turn, limitation - and that, consequently, he wants to allow for the supernatural possibility causes the formation of what Ockham cal1s a habitus - a disposition, if one prefers that God, through some special intervention, should provide me with a peculiar - which wil1 later on enahle the agent to reactivate the abstractive act even when the cognition of something, in virtue of which I would evidently know that this very ohject is not present anymore. and which thus constitutes a brand of intel1ectual thing docs not presently exist. This is the famous intuition of non-existents. which memory.14 General propositions can then he formed. and science can llourish. The 8 OCKHAM ON CONCEPTS INTUITION, ABSTRACTION AND MENTAL LANGUAGE 9

question I want to raise here is the following: how and where exactly do mental signs Ockham realized at some point - and this is a very deep intuition in his mature fit into this process? system - that the· various functions he wanted to attribute to concepts, in.so far precisely as they were semantical functions, could adequately be fulfilled by the cognitive acts themselves: 'for an act of understanding can. signify something and 2.~entallanguage can supposit for something just as well as any sign; and there is no point therefore in postulating anything over and above the act of understanding'.22 Such an act can The idea of an oratio mentalis - or oratio concepta - composed of signs is another be classified, within the framework of an Aristotelian ontology, as a quality of the major theme in Ockham's philosophy of mind, and it is also closely associated with mind. Reducing the concept to it allows Ockham's nominalism to dispense with the his views on concepts. The Summa Logicae, most notably, gives it a prominent role, special mode of existence that was required to accommodate the fictum, and to stick, but the subject is frequent in Ockham's other philosophical and theological writings in psychology as well as in , to a very simple naturalistic ontology with as well. IS Picking up a famous distinction from , William explains in the nothing but singular substances and qualities. Here is a salient case where Ockham very first chapter of the Summa that there are three sorts .of , ,and explicitly invoked the famous Razor Principle.23 correspondingly three sorts of terms: written, spoken and mental. 16The written ones are perceptible to the eye, and the spoken ones to the ear, but the mental ones are hidden to public perception. Concepts, which are the basic units of this mental 3. Abstraction and universals language, are nowhere but in the private of the intelligent agents, even if they are quite similar in each of them, as had noted in chapter 1 of the Restricting ourselves to Ockham'slater theory, weare now in a position to . These concepts, just like the words of spoken and written reformulate our original question as having to do only with mental acts: what, we are languages, .are. signs, but their signification is natural and primitive with respect to asking, is the relation between on the. one hand those mental acts which Ockham that of spoken and written words, which is conventional and derivative - or views as. conceptual signs and Which can occur as parts of mental propositions and 'subordinated' (subordinata). have asuppositio, and on the other hand the intuitive and abstractive acts that were This conceptual discourse, which is equated with thought itself, is endowed with described in section 1 above? a syntactical structure, very much like that of the conventional external languages. Let us begin with the case of abstraction. The answer here is very simple. It holds Most of the standard grammatical categories of or Donatus are borrowed by in the following identification: all simple abstractive acts of cognition are general Ockham for the analysis of inner discourse. There, he claims, we have nouns, verbs, terms in the mental language. The mental of abstraction, in other words, is itself adverbs, conjunctions and prepositions; we have singUlar or plural phrases, case­ a categorematic conceptual term, capable in principle of being the subject or declensions for nouns, tenses and modes for verbs, and so on ---'in short everything predicate of a mental proposition, and of standing for a plurality of distinct singular which is 'necessary for signification' (propter necessitatem significationis).'7 Even entities. It is, literally, a mental common noun. more importantly, the so-called terminist logic, which had been developed in the Concepts, which are often described by Ockham as the basic units of mental twelfth and thirteenth centuries, is now being applied not only to spoken or written language,24 are indeed directly identified in several passages with abstractive acts.2s languages, as in , William of Sherwood, and the other early We don't have, as in many medieval or modem ,two distinct entities contributors to the logica modernorum,18 but also - and even primarily - to mental here, one of which would be. the result or the object of the other, but only one. Once terms and propositions. Concepts are said to have a signification (sign(ficatio) or a the ,fictum-theory of concepts is abandoned in favour of the actus-theory, there is no connotation (connotatio) when they are considered in themselves,.and a referential point, in Ockham's eyes, in distinguishing between the concept and the abstractive function (or suppositio) when they occur within propositional contexts. It is one of act anymore. Ockham's major innovations to have systematically transposed the terminist logic he Why, however, should we say that simple abstractive acts are all general terms? had learned at sC]lOolinto a theory of discursive thought.19 Ockham, after all, distinguished very clearly in the Prologue of the Ordinatia With respect to the mode of existence of conceptual signs, it is now well known between two senses of notitiaabstractiva: abstraction from singularity, and that Ockham has changed his mind in the course of his career.20In the first redaction abstraction from existence or non-existence;26 it was; he said, in the second of these of his Commentary, around 1317-19, he hesitantly subscribed to the so­ senses, not in the first, that he wanted to contrast abstractive cognition with intuitive called fictum-theory, according to which a concept in the mind is a purely ideal cognition. Isn't it possible, we may ask, to think about somebody in her absence? object, which has no other than that of being the intentional correlate of a And isn't such a cognition, then, both singular and abstraCtive in Ockham's sense? simple cognitive act of apprehension.21After a short period of oscillation - probably Thinking about Mary while she is not here does not allow me to evidently know around 1320-21 - he finally adopted in his Questions on Aristotle's Physics, in the whether she still exists or not, whether she is sitting or not, and so on; this is Quodlibetal Questions and in the Summa Lo~icae.a very different view, known as precisely what Ockham calls an abstractive cognition, isn't it? Well, Ockham's the actus-theory, according to which concepts, rather than oeing the oojects of position with regard to singular ahstractive cognition varied across his writings, but certain cognitive acts, are directly identified with these very acts themselves. his final conclusion, in the Quol/liheral QUl'stiollS in particular, is that we cannot 10 OCKHAM ON CONCEPTS INTUITION, ABSTRACTION AND MENTAL LANGUAGE 11 have in this life an abstractive cognition which is both simple and properly singular. and nothing else. In Ockham's words: 'the concept of the species can be abstracted The for this is that a concept for Ockhamis always an intellectual image of from a single individual' .33Its formation does not require, as in some later empiricist some sort, a similitudo of what it represents. But what represents anything in virtue philosophers, that different individuals be first compared to one another and their of being an image of it also represents to the same degree whatever is exactly - or common features extracted. What it presupposes is an innate mental mechanism of relevantly - similar to that thing. My mental image of Mary in her absence is to the\ schematic representation, capable of being triggered by a single encounter. same degree an image of any other being who happens to look like her in therelevant This is not so, on the other hand, for generic concepts, which, Ockham says, are respects. In so far as it functions as an image, this representation, then, is 'never abstracted from just one individual'.34 I must already have at least two intrinsically general: it is equally capable in principle of representing a plurality of concepts of distinct species in order to'form on that basis the concept of the genus different entities. All simple abstractive cognitions, for Ockham, are like that: they which includes them both. There is, in this way, a genealogical hierarchy of are all of them general in principle, in so far as their representational function hangs abstractive acts which does not correspond to the order of logical implication. The upon some sort of .27 simple generic concept of bird, for instance, logically implies no particular species The possibility of a properly singular abstractive cognition of something is left concept, and it is not directly implied by any of them either, but it could not grow open, nevertheless, but it can only be reached, Ockham says, through an appropriate within a particular mind unless it was preceded in it by at least a couple of distinct combination of simple terms, none of which would itself be properly singular. I can lower-level specific concepts such as that of chickadee, blue jay, or sparrow. think about Mary after all, rather than about any twin she might have, but the concept Ockham's concepts - the universals - are neither mere abbreviations for lists of I should form in order to do so will have to be a mental compound rather than a proper names,35 nor structured definitions. They are, in their most elementary form, simple intellectual image.28Properly speaking, there are no proper names among simple abstractive acts, which, in virtue of abstracting from existence and non­ abstractive simple concepts. existence, are directly constituted into intrinsically general natural signs, This amounts to saying that a simple abstractive act is always in itself a . representing in the agent's mind a plurality of things, all of them singular of course Universals in the Porphyrian sense - genus, species, , proprium and but nevertheless objectively similar to one another.36 This similarity among its - were traditionally defined as what can be predicated of many. But, as significates is maximal in the case of the concept of a specialissima species, and Ockham frequently insists, only a sign can be predicated of anything.29Predication more and more relaxed as concepts get more general. In all cases, the universal thus occurs only within propositions, and propositions, in Ockham's mature theory, can generated - which is all at once a concept, an abstractive cognition, and a general only be composed of signs. Nothing therefore is a universal 'except in its categorematic mental sign - can supposit within mental propositions for all or any. signification, in its signifying many things'.3o This indeed is the core of Ockham's of these singular things - real or possible - which it signifies. nominalism. And the signs par excellence, those that are at the very basis of the whole semiotical process, are precisely the concepts (or 'intentions of the ', as Ockham also calls them), which, as we have just seen, are identified with abstractive 4. Intuition and singular terms acts. Thespecies 'man', for example, is nothing extramental, it is a general concept in the mind, in other words, an abstractive cognition. The terminology of the mental What about intuitive acts? Can they also be assimilated with some sort of signs and sign appropriately merges at this point with that of the notitia or cognition, to play logical and referential roles within propositions, such as those which are provide a clear-cut answer to the old . described by supposition-theory? Marilyn Adams, in her William Ockham, does not It follows from this that the process of formation of the universal is exactly the hesitate to answer this question in the affirmative: 'In the case just described - where process that leads to abstraction, as it was described above at the end of section 1. In [ have an intuitive cognition of and his whiteness and thereby formulate the the case of species, in particular, a single encounter suffices for their original judgement 'Socrates is white' - the intuitive cognition of Socrates and whiteness formation in the mind. Suppose I have never seen a blue jay before, and I see one themselves serve as the subject and predicate terms of the mental proposition.'37 now under optimal conditions; The intelleGtual intuitive act, as we have seen - which Since subjects and predicates always have supposition, the intuitive intellectual act is the immediate grasping of that singular bird by my intellect - generates at once a should be seen, in this view, as a semantic and syntactic unit of the language of first simple abstractive cognition, the prima abstractiva, which is straightaway a thought: a mental sign in the fullest sense. general, rather than a singular, representation,3l What will it be common to? This reading, as I will argue, is basically correct. It raises however a number of Ockham's answer is that it will indifferently represent to my mind everything in the difficulties and calls for some nuances. For one thing, the predicate of 'Socrates is world which is maximally similar (simillimus) to this singular bird I am now looking white' being a general term, it is hardto see how it can be identified with an intuitive at. But maximal similarity for Ockham precisely is this similarity which holds cognition. Contrary to what Adams says, the predicate of the mental proposition between all the singular members of a given specialissima species (a species which corresponding to 'Socrates is white' must, in any analysis, be a general concept, and does not naturally subdivide into further subspecies).12 This means, in the present therefore an abstractive act. The interesting question, however, is: what about the example, that the first simple abstractive act which will be caused in my mind by this subject in such cases'! Can my intuitive cognition of Socrates serve as the very intuitive cognition of a blue jay. will be a general concept representing all blue jays. subject of a mental proposition such as .Socrates is white'.• Socrates exists', or 12 OCKHAM ON CONCEPTS INTUITION, ABSTRACTION AND MENTAL LANGUAGE 13

'Socrates is here'? And a similar query arises about the predicates of propositions external object an immediate (albeit partial) cause. In so far, then, as they are parts such as 'this is Socrates' or 'some man is Socrates'. To these questions, Adam's of mental language, intuitive acts, unlike the abstractive ones, are singular terms. answer is indeed the right one, at least for Ockham's later theory. What kind of singular terms? Ockham, in the Summa, distinguishes three It is true that in the original redaction of the Ordinatio, Ockham readily spoke of varieties of singular terms: proper names such as 'Socrates', demonstrative pronouns the 'extremes' of a mental proposition (that is, its subject and predicate) as being such as 'this', and common nouns accompanied by demonstrative pronouns such as objects of intuitions, rather than intuitions themselves.38But these passages were 'this man'.49 Let us leave the third aside for a moment - we will come back written while he was still subscribing to thefictum-theory of concepts. He would to it in the next section - and let us ask the following:. are intuitive. cognitions, seen then agree with his senior colleague that singular mental propositions as signs in the mental language, proper names or demonstrative pronouns? Marilyn can be said to have as their subjects- and in some cases, as their predicates - the Adams opts; without discussion, for the former.50 This, however, meets with an very external things they are about.39At some later point, however, he corrected his important obstacle: proper names, as we know them, can very well be uttered in the own text: wherever he had alluded to the intuitive cognition of the extremes of a absence of their referents. Intuitive acts, by contrast, always require - at least in the proposition, he would now add a corrective formula such as ' ... or rather, of the natural order - the actual presence of their referents, .behaving in this more like things signified by the extremes'.40What these corrections indicate is that the subject demonstratives than like proper names. of a mental proposition like 'Socrates exists' or 'Socrates is white', as formed by the On the other hand, demonstratives in the usual sense do not readily lend agent in the presence of Socrates, is now seen as a sign..And what else could this themselves to the role we would like to attribute to intuitive acts. I am not referring sign be but the intuitive cognition itself, just as Adams says? It is not the real Mister here to the fact that Ockham leans toward excluding all pronouns, whether Socrates obviously, the rejection of this possibility being the very reason for the demonstrative or not, from the mentallanguage;51 this exclusion might not have'to corrections, nor an abstractive cognition, the hypothesis of which would be entirely be taken too seriously, since, as I have tried to show elsewhere, Ockham,in these superfluous in the context. passages, widely underestimates the importance of demonstratives for his own Ockham, actually, gets explicit about that in question 7 of his Questions on the logic.52 What I have .in mind, rather, is that a salient feature of ordinary Physics ofAristotle: demonstratives, as David Kaplan has insisted, is that they have, so to say, two levels of meaning, which Kaplan respectively calls a 'character' and a 'content'.53The When the intellect apprehends a singular thing by intuition, it forms· in itself an intuitive 'character' of a demonstrative, in this vocabulary, is its general capacity to refer to cognition, which is a cognition of this singular thing only, andis capable by its very nature different things indifferent contexts, or - to speak like Kaplan ~,tohave different to supposit for this singular thing .... And just as a spoken word conventionally supposits 'contents' in different contexts: 'this' can refer to any objectinthe world, according for its significate, similarly this [intuitive] intellection naturally supposits for the thing it is to what the speaker designates at the moment of utterance. Ockham's intuitive acts, an intellection of,41 however, do not present any such 'character'; Each one of them is inescapably attached to a particular referent. A given intuition cannot, like a demonstrative, be Here is a case where the epistemological terminology of the cognitio intuitiva is reused againlateron by the agent to refer to II new object. Whether I use 'this' to clearly connected with the semantical vocabulary of supposition-theory. And speak about a chickadee ora car, it stilhs the sameword. My corresponding intuitive supposition, for Ockham, belongs to significant terms in so far only as they are taken acts, by contrast, are not, in my mental language, tokens of the same sign. as subjects or predicates within propositions.42 It follows that the intellectual What is required here is a special semantical and syntactical. category, which intuitive act is an integral part of the language of thought; and that it can be subject presents strong similarities with what used to oall 'proper names in or predicate of mental propositions, where it can display s.emantical properties such the logical sense',54but which is even more stringent. A logically proper name, in 43 as a supposition. . ' Russell's view, is never given but toa single referent, which the speaker must at This is an intriguing idea, ilnd although.. Ockham did not work it out in detail, it some point have been directly acquaintedwith, this episode of direct acquaintance deserves a bit of elaboration. The intuitive act, forOckham, is a realthing (a res having fixed once and for allthe referent for this particular designator. Ockham's absoluta), namely a quality.44 This thing is a sign, in the very sense of the word intuitive cognitions are direct designators too: they do not have descriptive contents, which interests logicians: 'anything which brings something to mind and can any more than Russell's logically proper names do. And they also presuppose a supposit for that thing'.45 Just like the abstractive cognition which wilUollow upon direct acquaintance of the agent with the object. The requirement of acquaintance, it, the intuitive act resembles its object in some way46, but unlike abstractive however, is even stronger in the case of Ockham's intuitive cognitions, since such cognition, it is not thus rendered general. The reason for this is that in the case of cognitions simply cannot occur (in the natural order) in the absence of their objects. intuitive .cognition, it is rather than similarity which determines the Once the speaker has been in direct contact with a given referent, she can, in 47 signification: an intuitive intellection is a cognition - and a sign - of exactly this Russell's view, use its logically proper name, even when the thing is not there singular thing which (partially) causes it. Such a causal account of signification was IInymore; but this is not so with the intuitive cognition. Ockham's intuitive acts must not available to Ockham in the case of abstractive cognitions, because the only true correspond to an even more demanding category of signs, which contemporary causes for him arc the immediate ones,4~and the causal link between an abstractive docs not seem to have cared much about. act and its external object is always indirect. Only in the cilse of intuition is lhl' 14 OCKHAM ON CONCEPTS INTUITION, ABSTRACTION AND MENTAL LANGUAGE 15

I will call them 'rigid deictics', since they literally show their objects (this is the always considered to be true in Ockham's semantics.58 The only other possibility deixis aspect) and never change them (this is the rigidity). They refer to their objects available would be that 'this white thing', in the example under consideration, without the help of any form of description, of any general concept, or of any should supposit for the non-white thing designated by the demonstrative; if it were intermediary whatsoever. They are not repeatable at will, but they nevertheless fully so, however, the sentence 'this white thingis not colored' would not be true for any belong to the language of thought in so far as they can be subjects or predicates of designated object: it would be false, in particular, if the speaker was pointing at a red mental propositions, and in so far as their presence within these propositions directly object, since it is plainly false that a red object is not colored! influences their truth-conditions and inferential roles. They can be identified neither That the supposition of the complex subject-term turns out to be empty in such with ordinary proper names, nor with spoken or written demonstratives, and not deviant cases does not keep the demonstrative from designating something. even, strictly speaking, with Russell's logically proper names. Ockham's semantics Ockham's use of 'quocumque demonstrato' clearly indicates that something is tum out to be suggesting here - without the Venerabilis Inceptor being clearly designated in the example he is then considering; this becomes quite explicit when conscious of it himself - a non-trivial enrichment of the general syntax of language, he imagines someone saying 'this man is an animal' while designating a donkey. The a new 'part of speech', which could exist only in the language of thought. situation, therefore, is the following: the demonstrative, in those cases as in the normal ones, does designate a singular thing, and the common noun signifies the things it applies to (if such things exist). Combined to one another, they yield a 5. Mixed cognitions and singular terms of the third type complex term, which supposits for nothing in the propositions where it occurs if the object designated by the demonstrative is not one ofthe significates ofthe noun. This Let us come back now to the third sort of singular terms listed by Ockham, those that is precisely what happens for all other complex terms, where each component are composed of a demonstrative accompanied by a common noun, such as 'this 'restricts' the scope of the other one: 'black horse'can supposit only for those beings man'.55 If, as we are now entitled to admit, the utterance of a spoken demonstrative which are signified both by 'black' and by 'horse'; if there were none, the complex is normally associated with an intuitive cognition in the mind of the. speaker, such term 'black horse' would supposit for nothing. complex phrases as 'this man' must also correspond to mental acts involving This result can be directly transposed into mental language. The singular mental intuitions in one way or another. Are we to grant, then, that a general term can be term corresponding to 'this man' must be seen not as an intuitive act internally somehow incorporated within an intuition, and that some intuitive cognitions can endowed with a conceptual content, but as a mixed cognition, a complex mental term have a descriptive content after all? The question is important for a correct composed of an intuitive cognition (corresponding to the demonstrative) and an understanding of how intuitions and concepts are connected with each other in abstractive cognition (corresponding to the common noun). Such a compound can Ockham's epistemology. And it has to be answered in the negative: the mental occur within a mental proposition, either as subject or as predicate. If it does, it will cognition corresponding to a phrase like 'this man' must be a compound, just like then precisely supposit for the one singular significate of the general concept which the spoken phrase is. What it is composed of is an intuitive act on the one hand also happens to be the object of the intuitive cognition; if there is none, its (corresponding to 'this') and an abstractive act on the other hand (corresponding to supposition will be empty. Since it is thus possible that such a complex term 'man'). This is what I will call a mixed cognition. Its conceptual content ~which is represents nothing in certain contexts, even in the natural order; it cannot as a whole its abstractive part - stays utterly external· with respect to the intuitive· act which be identified with an intuitive act: a mental proposition of the form 'this F exists' constitutes its other part. Both are associated with each other in a contingent way. could tum out to be false! Theintuitive and the abstractive cognitions, therefore, can The intuitive cognition in itself is, in this case as in the previous one, devoid of any combine with each other to form a complex singular term in the mental language, generality. but they do not merge into one another. Each one keeps its identity and function. In order to see it, we must pay attention to Ockham's developments on the Intuitions and concepts are irreducibly distinct mental signs. semantics of phrases of the form 'this F'. The first conclusion to be drawn from them is that a complex singular term of this sort can sometimes be used although the general term which is part of it does not truly apply to the object designated by the 6. Ockharn's representationalism demonstrative. 56 William mentions at one point the case of 'this man is an animal' uttered when the designated object is in fact a donkey, or that of 'this white thing is The Ockhamistic mapping of semantics into epistemology thus yields the following not colored' uttered when no white thing at all exists in the worldY In this latter results: situation, Ockham surprisingly says, the uttered sentence would be true, whatever object is designated (demonstrato quocumque). This apparently strange diagnosis ­ ( I) All simple abstractive acts are general terms in the mental language; this is which Ockham doesn't care to explain much - becomes intelligible only if we grant precisely what a universal amounts to. that the subject 'this white thing', in such a case, supposits for nothing, even though (2) Intellectual intuitive acts are singular terms in the mental language; they can be a real object is designated by the speaker with the help of the demonstrative. What subjects or predicates of mental propositions, and they can, as such, supposit we have here is a negative proposition with an empty subject. and such sentences are for their objects. 16 OCKHAM ON CONCEPTS INTUITION, ABSTRACTION AND MENTAL LANGUAGE 17

(3) There are mixed cognitions, composed of at least one irituitive .act and one Notes abstractive act; they are complex singular terms in the mental language. 1. See, for instance, Quaest. in Phys., q. 1-7 (rightly called questions de conceptu in the All of these are signs naturally representirig in the mind certain objects other than critical edition), OPh VI, pI>.397-412. themselves. A cognitive act, whether intuitive, abstractive or mixed, is a distinct 2. See, forinstance, SL I, 1, OPhI, pp. 7-9. singular thing, namely an internal quality of the mind, which represents something 3. See Ord. I, Prol., q. 1, OTh I, pp. 3-75; and Quod!. V,q. 5-6, OTh IX, pp. 495-503; and also Rep. II, q. 12-13, OTh V, pp. 251'-310. Many commentators have ably discussed virtue certain external has with its objects: causality, the caSe in of it in of these passages; see in particular: Day 1947; Adams 1987, chap. 13-14, pp. 495-629; intuitive acts, or similarity in the case of abstractive acts. These acts are of such a Alferi 1989, especially pp. 147-214; Michon 1994, chap. 2~3,pp. 53-142; and, most nature, moreover, that they can be parts of mental propositions and influence their recently, Karger 1999a, Stump 1999. truth-conditions and logic, as well as their epistemic attractiveness for the agent. 4. Cf. Ord. I, Prol., q. 1, OTh I, p. 31: ' ... notitia intuitiva rei est talis notitia virtute cuius There is, it seems to me, a deep wisdom in this doctrine, which contemporary potest sciri utrum res sit vel non, ita quod si res scit, statim intellectus iudicat earn esse, philosophy of mind·and language should ponder. It provides a reasonably simple nisi forte impediatur propter imperfectionem illius notitiae'; and p. 70: ' ... dico quod per picture of the cognitive process and of its relation to language, without requiring any notitiam intuitivam rei potest evidenter cognosci resnon essequando non est vel si non special ontological enrichment: no real universals out there in the world, no ideal sit'; see also Quod!. V, q. 5, OTh IX, pp. 495-500. . objects, and no pure consciousness either, mysteriously open to otherness. The 5. Cf. Ord. I, Prol., q, 1, OTh I, p. 31: 'Notitia autem abstractiva est ilIa virtute cuius dere approach is both nominalistic and basically naturalistic. It conjoins the idea of an contingente non potest sciri evidenter utrum sit vel non sit ... Similiter, per notitiam abstractivamnulla veritas contingens, maxime de praesenti, potest evidenter cognosci.' intellectual intuition with that of a mental language, and provides intriguing 6. Ibid., p. 32: 'Et per istum modum notitia abstractiva abstrahit ab exsistentia et non suggestions for a semantical approach to our understanding of singular things and its exsistentia, quia nec per ipsam potest evidenter sciri de re exsistente quod exsistit, nec integration within discursive thought. de non exsistente quod non exsistit, per oppositum ad notitiam intuitivam.' Ockham's epistemology, at this point, turns.out to be far less 'anti­ 7. See in particular: Boehner 1943, 1945; Adams 1970; Streveler 1975; Tachalj 1988, pp. representationalist' than what several recent commentators have taken it to be.59 Of 113:....53;and Karger 1999a. course, we have to be careful with such polymorphic labels. Marilyn Adams, for 8. See Quod!. VI, q. 6, OTh IX, pp. 604-7: 'Utrum cognitio intuitiva possit esse de obiecto example, contrasts two approaches to epistemology, which she respectively calls non existente.' Ockham's answer is twofold: 'In ista quaestione pono duas conclusiones: 'direct realism' and 'representationalism', and she unreservedly associates Ockham prima est quod cognitio intuitiva potest esse per potentiam divinam de obiecto non with the former. 6o Given what she means by those terms, I have no quarrel with that. existente' (p. 604); 'Secunda conclusio est quod natura1iter cognitio intuitiva non potest What Adams calls 'representationalism' is the doctrine according to which the causari nec conservari, obiecto non existente'. (p. 606), Let me streSs - because it has been the source of many misunderstandings - that the intuitive cognition of non-existent immediate object cognition is always a representation rather than the external of beings, forOckham, is not misleading: the agent in such cases correctly judges that the thing itself; and 'direct realism', in her vocabulary, simply is the negation of this things in question do not exist. This, outhe other hand, in rio way implies that Godcould thesis. It is straightforwardthatOckham's later theory of cognition is a case ofdirect not deceive us if He should choose to do so; it would simply be a different sort of case realism in this sense, and not of representationalism, since both intellectual (see on the latter point Quod!. V,q. 5, pp. 498-500; along with Karger 1999a's insightful intuitions and concepts are identified, in this theory, with cognitive acts having (in commentaries on this passage). most cases) external things as their immediate objects. These cognitive acts, 9. See for example Ord. I, Prol., q. 1, OTh I, pp. 25-7. however, are themselves seen by Ockham as representations. 10. Cf. Ord. I, Prol., q. 1, OTh I, p. 64: ' ... habet [intellectus] aliam notitiam per quam A mental representation in of mind is usually taken to concernit hic et nunc ...: et illa est notitia intuitiva'. Also Quod!. I, q. 15, OTh IX, pp. be a symbolic mental token capable of playing a causal role within the mind and of 83-'--6:'Utrum intellectus noster pro statu isto cognoscat intuitive sensibilia.' On the referring to something other than itself, which it can stand for in mental history of singular intellection in medieval philosophy, see Berube 1964. II. This argument is spelled out, for example, in Ord. I, Prol., q. I,DTh I, pp. 25-7,and in computations. Cognitive acts, for Ockham, are precisely like that: both intuitive and Quod!. I, q. 15, OTh IX, pp. 83-6. abstractive acts are real things within the mind, both are similitudes of their objects, 12. For Ockham's theory of the plurality of substantial forms, see in particularQuod!. II, q. both are signs capable of suppositing for them in mental propositions, and both are to-II, OTh IX, pp. 156-64. Adams 1987, chap. 15, provides a detailed presentation and elements of causal chains in the mental life of human agents;·If representationalism discussion of the theory. For the historical background of the.controversy; see Zavalloni is defined - as it seems natural to do - as any epistemological theory that attributes 1951. an indispensable and crucial cognitive role to such mental representations, then 13. Stump 1999 holds, in a Thomistic vein, that Ockham's doubling of intuitions is Ockhamism is a full-fledged form of representationalism in this sense. Although it superfluous and that 'the operations of the two cognitive faculties seem completely does not, in its mature version, postulate intermediate mental objects for cognitive redundant' in his account (p. 194). Her brief discussion of thepoint, however (p. 202, n. acts, its account of cognition always requires, nevertheless, a third entity between the 75), disregards Ockham's requirement that the relevant intuitive act should be apart, and mind and the external things: the cognitive act itself, which is indeed, whether not merely an exlemaI cause, of the existential proposition which the agent evidently judges to he true. More on this helow. intuitive or abstractive, a semantical representation. 18 OCKHAM ON CONCEPTS INTUITION, ABSTRACTION AND MENTAL LANGUAGE 19

14. On Ockham's theory of intellectual habitus, see in particular Rep. III, q. 7, OTh VI, pp. 32. On maximal similarity as cospecificity in Ockham, see Panaccio 1992a, pp. 258-67. 192-219, and Quodl. III, q. 20-21, OTh IX, pp. 281-8; also chapter 2, section 1 below. 33. Quodl. IV, q. 17, OTh IX, p. 385: ' ... conceptus speciei potest abstrahi ab uno individuo' The theory of memory, on the other hand, has been ably studied by Adams 1987, pp. (transl. Freddoso and Kelley 1991, p. 317). Similarly, the short summary of the cognitive 515-25, who shows, in particular, its evolution in Ockham's works. process given in SL III-2, 29, OPh I, p. 557, requires the meeting of only one man for 15. For recent presentations of Ockham's theory of mental language, see, among others: the original formation of the concept of man. Tabarroni 1989, Normore 1990, Karger 1994, Panaccio 1992a, 1992b, 1999a, 1999b. 34. Quodl. I, q. 13, OTh IX, p. 77: ' ... dico quod conceptus generis numquam abstrahitur ab 16. SU, 1 OPh I, pp. 7-9. uno individuo' (transi. Freddoso and Kelley 1991, p. 67). See also SL III-2, 29, OPh I, 17. See SL 1,3, OPh I, pp. 11-14; and Quodl. V, q. 8, OTh IX, pp. 508-13. Ockham raises p.557. a doubt as to the presence of participles and pronouns in mental language. Interjections, 35. As was suggested by Kneale and Kneale 1962, pp. 267-8. For a detailed critical on the other hand, are ignored, and grammatical genders are explicitly excluded as discussion of this interpretation, see Panaccio 1992a, pp. 247-50. devoid of semantical . 36. See Ord. I, dist. 30, q. 5, OTh IV, p. 385: ' ... similitudo dicitur relatio realis propter hoc 18. See, for example, the various treatises of the late twelfth and early thirteenth centuries quod unum album ex natura rei est simile alteri albo, et ad hoc quod unum sit simile edited by de Rijk 1967. alteri non plus facit intellectus quam facit ad hoc quod Sortes sit albus vel quod sit 19. See Panaccio 1999a. albus'. 20. See Boehner 1951, Leff 1975, and Adams 1987, esp. chap. 3. 37. Adams 1987, p. 530. 21. A good presentation of Ockham's version of this fictum-theory is provided by Karger 38. See, for example, Ord. I, ProI., q. 1, OTh I, p. 50: ' ... dico quod ... nulla propositio 1994. contingens possit evidenter cognosci nisi ex notitia intuitiva alicuius extremi ... ' (the 22. SL I, 12, OPh I, p. 43: ' ... supponere pro alio et signijicare aliud potest competere actui italics are mine). intelligendi sicut alii signa. Igitur praeter actum intelligendi non oportet aliquid aliud 39. The point is made in detail by Karger 1994 and 1996. ponere' (transi. Loux 1974, p. 74; the italics are mine). 40. In the final redaction, for example, the passage quoted above in note 38 is followed by 23. Ibid.:' ... frustra fit per plura quod potest fieri per pauciora. Omnia autem quae salvantur 'vel significati per extremum'. The St. Bonaventure critical editiOn of the Ordinatio ponendo aliquid distinctum ab actu intelligendi possunt salvari sine tali distincto ... '. On conveniently identifies Ockham's later additions to his own text. Most of the time, they the place of the Razor Principle in Ockham's nominalism, see Maurer 1978 and 1984. have to do with his shift from the fictum to the actus-theory of concepts. 24. See, for example, SL I, 1, OPh I, p. 7: 'Terminus conceptus est intentio seu passio 41. Quaest. in Phys. q. 7, OPh VI, p. 411: 'intellectus apprehendens intuitione rem animae aliquid naturaliter significans vel consignificans, nata esse pars propositionis singularem elicit unam cognitionem intuitivam in se quae est tantum cognitio illius rei mentalis, et pro eodem nata supponere.' singularis, potens ex natura sua supponere pro illa re singulari '" Et ita sicut vox 25. See, for example, Quod/. V, q. 7, OTh IX, p. 504: 'Et loquor de conceptu qui est cognitio supponit ex institutione pro suo significato, ita ista intellectio supponit naturaliter pro re abstractiva'; or Quodl. I, q. 13, where Ockham indifferently uses conceptus or cognitio cuius est.' abstractiva (OTh IX, esp. pp. 77-.8). 42. See SL I, 63, OPh I, p. 193: ' ... restat dicere de suppositione, quae est proprietas 26. See Ord. I, ProI., q. 1, OTh I, pp. 30-31: 'Sciendum tamen quod notitia abstractiva potest conveniens termino sed numquam nisi in propositione Et sictam subiectum quam accipi dupliciter: uno modo quia est respectu alicuius abstracti a multis singularibus ... praedicatum supponit'; and SL I, 69, OPh I, p. 208: ' solum categorema, quod est Aliter accipitur cognitio abstractiva secundum quod abstrahit ab exsistentia et non extremum propositionis, significative sumptum, supponit personaliter.' exsistentia et ab aliis condicionibus quae contingenter accidunt rei ... '. 43. This is precisely why Ockham's intellectual intuition is not redundant, after all, with the 27. See Quodl. V, q. 7, OTh IX, p. 506: ' ... dico quod intellectus noster de nulla creatura sensory intuition (pace Stump 1999; see above n. 13): only a properly intellectual act potest habere aliquem talem conceptum simplicem proprium ...; et hoc quia quaelibet could be a proper part of a mental proposition formed within the intellect. talis cognitio sive conceptus aequaliter est similitudo et repraesentat omnia individua 44. See Ord. I, ProI., q. 1, OTh I, p. 39: 'Sed visio intuitiva, tam sensitiva quam intellectiva, simillima, et ita non plus est conceptus proprius unius quam alterius'; and Quodl. I, q. est res absoluta, distincta loco et subiecto ab obiecto.' 13, OTh IX, p. 74: ' ... nulla cognitio abstractiva simplex est plus similitudo unius rei 45. SL I, 1, OPh I, p. 9: ' ... accipitur signum pro illo quod aliquid facit in cognitionem singularis quam alterius sibi simillimae ...; igitur nulla talis est propria singulari sed venire et natum est pro ilIo supponere' (transi. Loux 1974, p. 50). More on this quaelibet est universalis'. On Ockham's evolution in this respect, see Adams 1987, pp. in chapter 3 below. 534-6, and Panaccio 1992a, pp. 121--4. 46. See Rep. II, q. 12-13, OTh V, p. 287: 'Dico tunc quod intellectio est similitudo obiecti.' 28. See Quod/. I, q. 13, OTh IX, p. 77: ' ... habeo aliquam cognitionemabstractivam The context shows this to apply to intuitive as well as abstractive intellection, sinc;e propriam, sed illa non erit simplex sed composita ex simplicibus'. Ockham is then speaking ofthe intellectual apprehension of singular things and replying 29. See SL I, 15, OPh I, p. 53: ' ... omne universale, secundum omnes, est de multis to an objection having to do with both intuitive and abstractive cognitions. praedicabile; sed sola intentio animae vel signum voluntarie institutum natum est 47. Ibid., p. 288: ' ... licet intellectus assimiletur omnibus individuis [eiusdem speciei) praedicari et non substantia aliqua; ergo sola intentio animae vel signum voluntarie aequaliter ..., tamen potest unum determinate cognoscere et non aliud. Sed hoc non est institutum est universale.' propter assimilationem, sed causa est quia omnis effectus naturaliter producibilis ex 30. SL I, 14, OPh I, p. 48: ' ... non est universale nisi per significationem, quia est signum natura sua determinat sibi quod producatur ab una causa efficiente et non ab alia ... '. plurium' (transi. Loux 1974, p. 78). 41<. See Ord. I, dis!. 45, q. unica, OTh IV, p. 665:' ... omnis causa proprie dicta est causa 31. See Quodl. I, q. 13, OTh IX, p. 74:' ... dico quod cognitio prima abstractiva primitate immediata'. generationis et simplex non est propria singulari, sed est cognitio communis aliquando, 49. See SL I, 19, OPh I, p. 66. immo semper'. 20 OCKHAM ON CONCEPTS

50. See Adams 1987,p. 319, n. 9. Adams even doubts that pronouns are counted as Chapter 2 categorematic terms by Ockham. Yet this is what he explicitly does in Quodl. II, q. 19, OTh IX, p. 195: 'Et ideo dico quod ex primaria institutione [pronomen] est vox categorematica.' See also SL III-4, 10, OPh I, p. 798, where Ockham counts pronouns Intellectual Acts among the signs having a determinate signitication, which is precisely how a categorematic term is defined when the notion is first introduced in SL I, 4, OPh I, p. 15. 51. SeeSL I, 3, OPh I, p~11, where after having raised a doubt about the presence of participles in the mental language, as distinct from verbs, he laconically adds: 'Et de pronominibus posset esse consimilis dubitatio.' In the corresponding discussion of Concepts, then, are intellectual acts in Ockham's mature theory. This is an intriguing Quodl. Y, q. 8, he doesn't mention pronouns at all among the grammatical categories that idea. For one thing, the sense in which Ockham uses the term 'act' is an unfamiliar are needed,·according to him,.for the analysis of the language of thought. one for modem readers and has to be explained; this will be the object of section 1 52. See Panaccio 1980, 1992a, pp. 177-81. 53. Kaplan 1978, 1989. of the present chapter. The reasons which led him to drop the fictum-theory of 54. Russell 1918, pp. 200 et seq. concepts in favour of the actus-theory will then be reviewed (section2), as well as 55. SDI, 19, OPh I, p. 66. the reasons why he rejected, from the very start, the so-called species-theory, held in 56. Karger 1978 has judiciously called attention to cases of this sort. particular by (section 3). Once this is done, we will be in a position 57. SL III-3, 32, OPh I, p. 709. to reach a better understanding of how conceptual acts fit into human thought in 58. See SL II, 2, OPh I, p. 255, where the thesis is made explicit for the case of particular Ockham's view (section 4). negative propositions. It also holds, a fortiori, for universal negative propositions, as specified in SL III~3,32, OPh I, p. 709. 59. For example: McGrade 1986; Tachau 1988, chap. 5; and Alferi 1989. 1. Mental acts and habitus 60. Adams 1987, pp. 495-550. The first thing to notice concerning Ockham's general notion of actus is that it is very different from today's idea of action as it occurs, say, in the philosophy of action. An action in the modem sense roughly corresponds to 'someone's doing something intentionally'. I Acts in the medieval sense, by contrast, are not always done intentionally, and are not always someone's acts either. One could speak of heating as being the act of a fire, or of shining as the act of the Sun. The eruption of a volcano or the ringing of an alarm-clock would be good examples too. The relevant background here is Aristotle's idea of actuality (entelecheia), as opposed to mere potentiality. The act of something, in this vocabulary, is its actual operation, what it does, that is, in virtue of its internal powers being somehow into activity. Such operations can in many cases be triggered by external factors, yet they will be considered as acts in the relevant sense as long as they are typical realizations of certain internal powers the thing has in virtue of its , rather than merely accidental occurrences: growing, for example, is an actus of a tree, but its falling down as a result of being struck by lightning is not. [n the case of human beings, Ockham distinguishes between external and internal acts. 2 External acts are publicly observable and directly affect something outside the agent. Most intentional actions in the modem sense, such as walking, cutting a tree, or speaking aloud, are typical external human acts in this sense; but so is breathing or sweating, although these are not intentionaL Internal acts, on the other hand, are II/ental actualizations of various human psychological powers or faculties (potentiue): there are appetitive acts, for example, corresponding to the sensitive appetitive faculty, such as acts of desire, fear, or repulsion;3 there are volitive acts or acts of the will, such as decisions;4 and, of course, there are cognitive acts too, whether sensitive or intellectual (among which concepts). None of these, let me insist, need be intentional in the sense in which an action is usually taken to be.

21 22 OCKHAM ON CONCEPTS INTELLECTUAL ACTS 23

Many intellectual acts, in particular, are not under the control of the will, according Ockham as real singular qualities of singular minds. 14 His nominalistic ontology, to Ockham,5 and are not, consequently, the sort of things that can be done remember, admits of only two sorts of real things in the world: singular substances intentionally. and singular qualities. Since the existence of certain psychological acts is known by A further distinction which was unanimously considered in the thirteenth and (we know, for instance, that we are joyful or sad),15 the reality of such fourteenth centuries as crucially relevant for a correct understanding of human acts is above suspicion. Being accidental and transient states of substances (I am mental acts is the one Aristotle had allusively drawn in Book II of the De Anima: sometimes sad, sometimes not), they cannot be identified with the substances 'But actuality; he said there, is so spoken of in two ways, first as knowledge is and themselves. They have, therefore, to be qualities. If this is true of some second as contemplation is.'6 In the medieval vocabulary, these came to be called psychological acts, such as pleasure or sadness, there is no reason why it shouldn't respectively the primary act () and the secondary act (actus secundus),7 be true of all of them in so far as they are real. And if it is true of psychological acts or again - as in Ockham - the habitus and the act tout court. Aristotle's point was in general, it must be true of psychological habitus as well. The upshot of it all for that the actualization of a capacity often comes in two degrees, especially in human our present purpose is that concepts, if they are to be identified with the abstractive beings. The example he points at in the relevant passage of the De Anima is that of acts of human minds, are real episodic singular qualities of such minds, generated a given scientific truth. At first, one is utterly ignorant of it, but has the capacity to either by intuitive acts (when such acts occur) or by previously acquired cognitive learn it: this is a case of mere potentiality. When the person has learned this dispositions or habitus, and causally producing, in turn, new cognitive dispositions particular truth, but is not currently contemplating it, this knowledge is said to be within the mind or, at least, strengthening previously existing ones. actualized in her as a primary act - or habitus. When she does actually contemplate the truth in question, it thereby gains a stronger form of actuality within her, and is said to be there as a secondary act - or act tout court. In Ockham's vocabulary, then, 2. From the fictum to the actus the habitus is an acquired disposition, while the mental act is an actual psychological episode. But why should concepts be identified with cognitive acts? Many medieval authors, Ockham makes abundant use of this distinction in epistemology as well as in after all, thought otherwise. Thomas Aquinas, for one, had it that a concept is the . In the end, the general pattern of connection he establishes between mental object and product of the intellectual act, rather than the act itself. 16Ockham, in his habitus and the corresponding acts comes down to the following two principles: . earlier teaching, had favoured a similar approach himself, known in the literature as the /ictum-theory of concepts. 17 According to this doctrine, concepts are not real (a) A psychological habitus is normally caused by an original psychological act entities belonging to the Aristotelian categories of substance or quality, but mere leaving a trace within the mind;8 objects of thought, having no other existence than that of being cognized (which (b) The habitus in turn is a partial efficient cause for further acts of the same sort Ockham then called 'intentional being' [esse intentionale] or 'objective being' [esse as the original one.9 ohiective]).18 The Venerabilis Inceptor probably thought for a while that the postulation of such special unreal objects was necessary to avoid admitting common In order to avoid an infinite regress, this requires, as Ockham acknowledges, that not natures into being, as his senior colleague Walter Burley had been led to. Burley, in all mental acts should (even partially) be caused by a habitus. 1O In the case of effect, having rejected Aquinas's theory of the concept as the intentional object of knowledge, in particular, the intuitive grasping of an object is not normally caused the intellectual act, was quite happy with having ontological universals, instead, as by an acquired habitus: the J;Ilindsimply has an innate capacity to grasp - sensitively ohjects for such acts. 19 Ockham, on the other hand, thought from the start that the and intellectually - objects that are physically present in its environment. When this admission of universals or common natures as extramental beings was one of the is done, though, the intuitive act, as we have seen in chapter I, causes an abstractive worst possible errors in philosophy, leading to all sorts of incongruities and act, which in turn causes a habitus to be implemented within the mind. This habitus inconsistencies.20While he was ready to admit real individual things themselves as will then serve, under favourable circumstances, as a partial cause for further ohjects for singular cognitive acts, he believed at first that general thoughts required abstractive acts, numerically distinct from the one that caused it, but of the same some special sort of objects and he reverted, therefore, to flcta, rather than common ll sort. This causal effect of the habitus is described by Ockham as an 'inclination'. natures, to play this role. As Elizabeth Karger ably writes: 'It seems that Ockham A mental habitus, he typically says, inclines the mind towards certain sorts of acts: 12 must have introduced /icta as substitutes for common natures ... without the when the other relevant factors are present, acts of this sort will tend to occur within drawhacks of an objectionable ontology.'21 The question is: why did he change his the mind of whoever is endowed with such a habitus. And when a new act is thus mind? engendered by a psychological habitus (whether cognitive, volitive or appetitive), a Many recent commentators have surmised that Ockham's evolution on this point new habitus, similar to the previous one, will in turn be engendered, or - more often, was due to the ohjections raised in the early I320s against his theory of concepts as presumably - the previously existing habitus will he strenghtened.1\ Acts and Jicta hy his Franciscan confrere Walter Chatton.22The matter, however. requires a habitus thus form a rich cluster of causal chains within the mind. closer look. In his most complete listing of reasons for dropping the jictum-theory. Ontologically speaking. all these psychological acts and hahitus are seen hy in question I of the Qum·stion!;.\' in Lihras Physicorum Aristoteli.\·, Ockham provides 24 OCKHAM ON CONCEPTS INTELLECTUAL ACTS 25 seven arguments, only two of which are directly borrowed from Chatton's original contrary to what Marilyn Adams suggests, this objection is considered critique.23The first clear signS of Ockham's evolution on the matter, moreover, occur relevant by Ockham only in so far as it has to do with general abstractive in the Prologue of his commentary on the Perihermeneias, which, according to the cognition, rather than with intuitive cognition, since he never had supposed chronology proposed by the editors, was written before Chatton's intervention,24 and anyway that the latter involved a fictum. Adams takes the argument to be two of the arguments against the .fzctum-theorymentioned with approbation in the confused because she thinks it rests on the implicit premiss that flcta, Quaestiones in Libros Physicorum do come from this earlier text.25It might very according to the flctumctheory, 'are always the immediate objects of thought well be that the discussion with Chatton was the occasion for Ockham to rethink the and awareness',31 which, as she rightly remarks, Ockham had never accepted question for himself andthat some of Chatton's points really impressed him, but on in the case of intuitive cognition. But Ockham, of course, knew that very the whole, as we are about to see, his motivation for shifting from the flctum to the well. What he must mean here is that the fictum-theory threatens the actus-theory of concepts was quite distinctive.26 adequacy of general knowledge. The key reason for his change of mind is revealed by Ockham,not so much in the (1.4) Endorsing the flctum is. not easily compatible with the idea (which was arguments he offers against thefictum-theory, as in the one positive consideration he generally accepted in Ockham's ) that the representational function of brings forward in favour of his new position, both in the Quodlibeta and in the general concepts rests on some similarity they are supposed to have with the Summa Logicae: 'whatever is preserved by appeal to a fictive entity can be preserved thIngs they represent: being an unnatural and unreal sort of entity, the flcta by appeal to an act of understanding' .27Seeing this as the central point invites for a would seem to differ more than any real thing from whatever it is that they reorganization of Ockham's whole argumentation on the matter in the form of the are expected to represent.32 following modus ponens: (1.5) The postulation of ficta, in so far as it implies the eternal and necessary (1) it would be preferable to dispense with the fictum if we can; existence of a realm of purely intelligible objects, runs counter to the (2) it is indeed possible to dispense with the flctum,since all the functions we want principle of God's omnipotency and to the radical contingency of all created to attribute to concepts can be adequately fulfilled by intellectual acts; beings.33 therefore: (3) we should dispense with the.fzctum in favour of the actus. Ockham, it seems to me, must not have taken these five points as decisive by themselves against the flctum-theory. After all, he could have lived with the All of Ockham's particular objections against the flctum-theory, whether in his complications pointed out by the first two objections, had he thought them necessary commentary on the Perihermeneias, in Quodlibeta IV, 35 or in the Quaestiones in for a correct account of cognition. And the introduction of ideal intermediates Libros Physicorum 1, can be associated with either premiss (1) or premiss (2) of this hetween minds and things, that gives rise to objection 1.3, would certainly have been master argument, some of them enumerating the drawbacks of an appeal to the preferable to him than the acceptance of common natures as direct external objects fictum (without, however, being decisive in Ockham's eyes), while the others insist for scientific knowledge, even at the price of tempering his direct realism a bit. As on its actual dispensability. The whole scheme can be reconstructed in the following to the last two difficulties, Ockham, no doubt, could have found ways of alleviating manner. them. As he had already remarked in his commentary on the Perihermeneias when First, the drawbacks. There are five of them, the first two - in the order adopted reacting to objection 1.4, the only sort of similarity which is relevant for an account here - having to do with complications induced by the introduction of the flctum, and of human cognition is the one that can hold between a mental representation ­ the last three with conflicts it tends to generate with well-accepted theoretical whatever that is - and what it represents. In theflctum-theory, that would be the sort principles: of similarity an intelligible object can have with real things - intentional similarity, let's call it - and there would be no point in comparing it in terms of degrees with (1.1) The acceptance of the fictum as the object of cognition requires a similarities among real things: a mule might look pretty much like a horse, yet it counterintuitivecomplexification of the ontology, since it cannot be cannot serve as a mental concept for horses, simply because it is not the sort of entity identified with a natural entity, whether a substance or an accident.28 a mind can do computations withp4 The existence offlcta, finally, could have been made contingent upon the existence of the corresponding acts of thought in human (1.2) It also induces a special complexification of the epistemological process, minds, which would have sufficiently countered objection 1.5. since the position of such aflctum is not analytically required by the idea of The five problems mentioned up to now are not innocuous for all that - far from an intellectual act of cognition (even a general one).29 it - hut they could have force against thefictum-theory in Ockham's eyes only in so (1.3) The insertion of the fictum as the object of general cognition represents a fllr as an acceptable alternative was available, which could avoid these complications threat to direct realism in epistemology, otherwise favoured· by many and drawhacks without reintroducinf? common natures in the external world. a medieval authors including Ockham. since it introduces an intermediate requirement Chatton's approach could hardly satisfy, being very close to that of John between the act of cognition and the thing itself. \()It must be stressed that. ()unsScotwi in . 1\ A second argumentative step was needed in order to 26 OCKHAM ON CONCEPTS INTELLECTUAL ACTS 27 show that the cognitive functions of concepts could be fulfilled by intellectual acts drawbacks mentioned in 1.1-1.5 can be given all their weight, and the general without undesirable ontological consequences. Some of Chatton's remarks could be master argument smoothly runs through. revived in the process, no doubt, but the crucial point, as we shall see, was original. How Ockham came to see this as possible without reintroducing common natures Three more of Ockham's arguments are relevant here: in the ontology is revealed, interestingly enough, by the one aspect of argument 2.3 which he substantially modified with respect to the original version of it he had (2.1) All the uncontroversial propositions we want to accept about conceptual given in this' commentary on the Perihermeneias. He was explicit in this earlier activity, propositions such as 'horses are conceptually cognized' or 'horse is passage that this argument, in his view, could not be invoked in favour of identifying a predicate' and so on, could very well be true even if no fictum existed at all, concepts directly with intellectual acts, because he found it difficult to see what the but only cognitive acts and real external things. If, however, the truth of a objects of such acts would then be, what it is, in other words, that such cognitive acts proposition requires only two things, it is inappropriate to postulate a third would be cognitions of.43What he was presupposing at the time is that every act of one on this sole basis. 36 This objection, as can readily be seen, explicitly rests cognition must have one and only one object into which it 'terminates'. This is why on a version of the famous Razor Principle which came to be associated with he could see argument 2.3 as capable of beingused in favour of identifying concepts Ockham's name.3? The crux of it, nevertheless, is the assertion thatjicta are with special mental qualities serving as objects for intellectual acts, but not yet with not needed to account for the sort of things we normally want to say about the intellectual acts themselves.44Dropping this is precisely what concepts. But this amounts to little more than to what I called premiss (2) opened the way forhis final adoption of the actus-theory. above. How it can be sustained remains to be explained. As I said earlier, it might very well be that the debate with Chatton helped him in this regard, but none Chatton's arguments could have convinced him as long as he (2.2) Concepts are expected to be subjects and predicates in mental general of didn't see how he could avoid the postulation special mental objects for propositions. But this is something the cognitive act could do just .as well as of intellectual acts without having to revert to common natures as what these acts the fictum.38This argument, closely inspired by Chatton,39 brings to the fore a particular case with respect to the previous one by stressing one particular should terminate in. The breakthrough must have come when he realized that simply function that concepts should fulfill, that of being the subjects or predicates dropping an unexamined and unwarranted assumption about cognitive acts could do of mental propositions. What it does not do, however, is to show how exactly the trick. Not that he came to accept that cognitive acts did not need any object at all, an intellectual act could come to serve as a subject or a predicate. In or did not need to 'terminate' in anything at all. What he realized instead is that each Ockham's logical doctrine, what is required for anything to be the subject or cognitive act does not need exactly one object. This must have been greatly predicate of a proposition is that it should be able to have a supposition (or stimulated by his coming to take with utmost seriousness the idea that a general concept is but the sign of many singular things at once, a doctrine he had already referential function).40 The argument, then, presupposes in Ockham's hands 45 that intellectual acts can in general supposit for something. But it does not sketched in his commentary on the Sentences. Clearly realizing that the yet show how this is possible. identification of the cognitive act with the concept amounts to identifying it with a sign of many singular things opens the way to seeing how it can still be a cognition (2.3) An intellectual act can represent something outside the mind, it can signify of something without this something being either a fictum or a common nature: if such a thing, and supposit for it, just as much·as afictum was expected to. conceptual acts are seen as general signs, only their singular significata are needed This, at last, is the heart of the matter. The functions now enumerated­ to serve as their objects. representation, signification, and supposition - are the basic ones concepts are supposed to fulfil. If intellectual acts can do that, then ficta indeed are dispensable without loss. This general argument had first been sketched by .t The rejection of intelligible species Ockham in his commentary on the Perihermeneias as an argument in favour of positing real mental qualities within the mind, which, while distinct from Another doctrine which Ockham is famous for having rejected concerning mental the intellectual acts, could serve as objects for these acts,41 a position he representation is the so-called species-theory. In its full-fledged version, as it had never actually subscribed to. Later on, however, he realized that these been developed by thirteenth-century 'perspectivists' like , John functions could be directly assigned to intellectual acts. Both in the Peckham and Witelo (all heavily influenced in this regard by the Arab philosopher Quodlibeta and in the Summa Logicae this is his main point in favour of the Alhazen),46 this theory held that in order for human cognition to take place, the actus-theory.42 forms of external things had to penetrate the mind somehow. This was thought to re4uire three successive patterns of representative emanations - called species ­ The whole argumentation,. as developed by Ockham, ultimately rests on his llowing out from external things: first, there was the species in medio, which was realization that some intellectual acts can serve as mental s;~ns,be endowed, that is, supposed to earry the thing's image through the ambient environment (called the with significations of their own and be capable of referential functions within mental fI/('(Jium); second. a sensible .lpc'de.I' was said to be formed within the sensitive organ propositions. Once this is admitted. considerations 2.2 and 2.1 directly follow. the of the knower; and third. an intelligible ,Ipecic's was postulated within the intellect. 28 OCKHAM ON CONCEPTS INTELLECTUAL ACTS 29

Ockham, from the very start, rejected all three of these.. I will focus here on his mental word - which he sees as the result of the cognitive act, while the intelligible discussion of the intelligible species, which is the more directly relevant in the species is, in all cases, supposed to be its starting point.55 present context.47This is a theme that has loomed large in the recent literature on That this is where the disagreement lies between Ockham and Aquinas is fourteenth-century epistemology.48 Yet a number of misunderstandings still circulate something several recent commentators, even among the best, seem to have missed. about it. There seems to be a consensus, in particular, that Ockham misinterpreted Robert Pasnau, for one, bluntly asserts in the course of his discussion of Ockham's the theory he was criticizing in this instance49and that his own scheme of intellectual rejection of the intelligible species, in his otherwise remarkable book on Theories of acts and habitus comes pretty much to the same. Which, as far as I can see, is hardly Cognition in the Later , that 'this portion of Ockham's account ... isn't the case. What we need to do to settle the point iS,first, to correctly identify what it different from Aquinas's', since both admit that the contact with external things puts was exactly that Ockham wanted to reject (section 3.1), and second, to recall what the intellect 'into a certain state such that cognition with a certain intentional content his main critique was (section 3.2). follows',56 Ockham's point, however, is precisely that there is no such state, as distinct.from both the cognitive act and the habitus that follows upon it. Pasnau 3.1 Species as preconditions for intellectual acts thinks that these 'states', in Aquinas's theory, are not to be reified as distinct entities with respect to the intellect itself;57 but if the intelligible species has an accidental As to the first point, everybody agrees that Ockham's main target on this was being within the intellect for Aquinas (as Pasnau readily, and correctly, Thomas Aquinas's doctrine (or some version of it). Aquinas, of course, was far from acknowledges),58 then iUs certainly not to be identified with the intellect itself, even being the sole promoter of intelligible species (and Ockham, in fact, explicitly though it is not independent from it. The intelligible species is not a substance of its mentions Scotus in this regard),50 but as Leen Spruit correctly points out, Aquinas own, of course, in Aquinas's view; nevertheless,it is not ontologically reducible to had offered what 'may legitimately be called the "canonical" theory of intelligible either the intellect or the cognitive acts and habitus. species' ,5\ and Ockham was well aware of this.52 Aquinas thought that, since universalintelligible forms couldnot be posited as actually existing outside the mind 3.2 The Razor argument againstintelligible species (as Aristotle was taken to have shown against Plato), they had to be actualized within the mind by a special intellectual power, called the 'agent intellect', which extracted Why Ockham rejects the postulation of such special entities is straightforward: they them, so to say, from the sensible species deposited within the imagination as a result are,he thinks, entirely superfluous. His whole argument, here, rests upon the Razor of sense perception. This process of 'abstraction', in his view, led to the formation Principle: since intellectual cognition can be accounted for with acts and habitus of intellectual general representations -the intelligible species, precisely - which alone, intelligible species, seen as extra entities, are not needed. In the relevant could then serve as the means for general knowledge.53 passage of the Reportatio, he successively considers five cognitive functions one What Ockham objected to in this account is not that it involved intellectual might want to attribute to the intelligible species, and purports to show that all five representations stored in the mind for future use - he needed that too himself in the of them can be sufficiently fulfilled without it. 59 guise of the habitus - but that the intelligible species it postulated should always be The first one is cognitive assimilation. It was commonly accepted in the Middle prior to the intellectual acts of cognition, and that they were seen as presupposed by Ages that cognition, whether sensitive or intellectual, requires the knower to develop these acts: within himself some sort of similarity with the known object. Ockham agrees, but insists that the relevant sort of assimilation is sufficiently explained by assuming the I take it that the species [according to this theory] is a precondition for the act of act of intellection to be originally caused by the object itself.60It has to be stressed, intellection and can remain both before and after intellections, evenin the absence of the pace Tachau, that this argument in no way hangs upon some anti-representationalist thing. And it is distinct, therefore, from the habitus, since the intellectual habitusJollows tTendin Ockham's thought. Tachau wrongly assumes that Ockham 'rejects the upon an act of intellection, while the species precedes both the act and the habitus.54 notion that intellectual cognition is a process of assimilation of intellect to object'.61 He does accept it, quite to the contrary, for every variety of simple cognitions, Ockham's whole discussion of the matter in the Reportatio makes it clear that what whether intuitive or abstractive. His point, rather,is that the existence of preparatory he wants to dispense with is precisely this antecedence of the stored intellectual intelligible species is no more a necessary precondition for abstractive assimilations representation over the intellectual act of cognition. It has to be understood that the than it is for intuitive ones.62 process of abstraction which leads to the formation of the intelligible species is not, A second reason one might want intelligible species for, Ockham says, is to in Aquinas's doctrine, the intellectual act of cognition itself, but a preparatory phase, account for mental representation; but, once more, the object itself and the cognitive postulated as a condition of possibility for intellectual acts to occur: the intelligible act arc sufficient for this. Ockham's starting point, here, is that the intuitive grasping species resulting from this somewhat mechanical - and unconscious - process of of an object does not require any intervening representation between the object and abstraction is, in Aquinas's words, 'the source fprincipium] of the action of the the act, the direct causing of the act by the object being obviously enough in this intellect', and it must be carefully distinguished both from the cognitive act itself, h case. ' If this is so for intuitive. cognition. he argues, then there is no reason why it which comes after it, and from what Aquinas calls a concept - or conception, or should he otherwise for the ahstrat'tive cognition that immediately follows upon it: 30 OCKHAM ON CONCEPTS INTELLECTUAL ACTS 31

'just as the object sufficiently represents itself in the one cognition, it does so too in intelligible species between Ockham and his opponents, there lay another - more the other, which immediately follows upon the intuitive cognition'.64 Again, the idea fundamental - divergence. In the last analysis, the point of resorting to intelligible is not to evacuate mental representation altogether. The argument, on the contrary, species - in Scotus most notably, but also in Aquinas - was that the external singular assumes that such a representative function is already at work at the level of intuitive things were thought to hide some universal features within them, which had first to cognition, from which it concludes that mental representation in general does not be isolated - or extracted - in order for general cognition to take place. Ockham, of necessarily require a species pre-existing to the cognitive act. course, energetically resisted this realist assumption and argued against it on many The causation of intellection was a third motive sometimes invoked in favour of occasions. That he shouldn't do so in the context of his discussion of intelligible positing intelligible species, on the basis of the alleged impossibility for material species, however, reveals that he took the postulation of such species to be things to causally affect the spiritual realm. But if there was such an impossibility, superfluous even if these universal features were to be admitted: whichever way it Ockham argues, how could the formation of the intelligible species itself ever be was with ontology, his point is that if abstraction could be reached in intelligible explained without falling into an infinite regress: 'just like you accept, he says to his species considered as preconditions for intellectual acts of knowledge, then there is opponent, that a corporeal thing can be the partial cause of a species in the spiritual no reason why it could not be reached directly in the intellectual acts themselves, and realm, I accept that a corporeal thing is the partial cause of the intellection in the in the habitus that follow upon them. Whether Ockham correctly understood the spiritual realm.'65 This again is basically an argument from parsimony: since the theory he was criticizing there, or not, solely depends on whether this theory did in causal impact of material things upon the mind has to be acknowledged anyway, the fact postulate intelligible species as prerequisites for intellectual acts of knowledge. introduction of species is useless in this respect. And the argumentative pattern is the There seems to be little doubt that it did. same for the last two functions Ockham discusses in this context: how the cognitive faculty is determined by its object, and how it is moved by it.66Nothing in his whole critique of intelligible species hangs upon either an anti-representationalist on 4. Combining acts Ockham's part or on his wrongly supposing that the intelligible species were intermediate objects of thought in the eyes of their supporters (which, obviously, Both species and ficta, as we now see, were renounced by Ockham for reasons of they were not).67 Ockham,in fact, does not even stress any special drawback of economy. In the case of species, he was never even tempted to countenance them, accepting intelligible species (as he will do later on in the case of the fictum) , except because he couldn't imagine any special job they could be endowed with. Ficta were for useless complications in epistemology. Parsimony, here, is his sole manifest different in his eyes, because he thought for a while that, despite all their drawbacks, preoccupation. they provided the sole alternative to. common natures for serving as objects of This is also the case in his short critique of any temptation to identify concepts general abstractive acts of cognition. This is a notion he came to abandon when he with intelligible species, in the Prologue to his commentary on the Perihermeneias.68 realized that the only aspect that matters for something to be a concept is its capacity He mentions three objections there. First, such species are superfluous anyway, the to fulfil certain semiotic functions in mental computations. From then on, he was left reader being referred to a demonstration of the point elsewhere in Ockham's works with only acts and habitus within the intellect. Not all of them, however, were (presumably the one we have just been discussing). Second, only acts and habitus identified with concepts. Even leaving aside habitus, conceptual acts in his view are should be admitted as real entities within the mind according to Aristotle. And third, but one variety of intellectual acts. They must be distinguished, on the one hand, it would follow that concepts - and the mental propositions made up from them ­ from intuitive acts, as was explained in chapter 1, and, on the other hand, from would remain within the soul even when it is not actively thinking, just like the propositional and judicative acts, which is what I want to insist upon now. The mind, species are supposed to. All three arguments can be read as pointing to unnecessary for Ockham, is basically a combinatorial device. Locating the role of concepts in epistemological complications, and have nothing to do, at any rate, with alleged human thought calls for an elucidation of how intellectual acts combine with each threats to direct realism. other into more complex units, such as propositions and judgements. I will first Eleone Stump mentions still another argument Ockham uses against species: 'No briefly present Ockham's distinction between simple and complex items within the one sees a species intuitively, and therefore experience does not lead us to this mind (section 4.1), and then examine in some details the structure of mental [account of cognition]';69 which she interprets as (wrongly) presupposing that in propositional acts, of which concepts can normally be parts (section 4.2). We will order for the species to be vindicated, its mental reception should be a conscious see, finally, how judicative acts fit into the picture (section 4.3). cognitive act in Ockham's view, as if his reasoning was the following: we have no intuitive grasping of species, therefore they don't exist. The complete argument, 4.1 Simple and complex units however, is quite different, and it is, once more, an economy argument. Ockham's point is that no natural entity should be countenanced without either a good Ockham frequently distinguishes between simple and complex mental items. He theoretical reason or some empirical . 711 The sentence Stump quotes is docs it explicitly. for example, at the outset of the Ordinatio: 'among the acts of the simply the part where he is denying that the latter holdPi in this case. intellect. he says there.... one is the act of apprehension, and it pertains to whatever It is true, as Dominik Perler has inPiisted,71that behind the disagreement over l:an tcrminate an act of thc intcllective power. whether this should ht' ('omph'x or 32 OCKHAM ON CONCEPTS INTELLECTUAL ACTS 33

incomplex'.72 This passage, admittedly, was written from the point of view of the From which he concluded that mental propOSItions have to be simple acts fictum-theory of concepts, which Ockham still held at the time: the distinction drawn themselves: human thought, in Gregory's view, is not intrinsically compositioha1.82 here between complex and incomplex items separates different kinds of possible This is a problem Ockham was well aware .of. He discussed it in some detail in his objects for intellectual apprehensive acts. When he moved to the actus-theory, he commentary on the Perihermeneias and in his Quaestiones in Libros Physicorum, kept the distinction, except that from then on, it was the intellectual acts themselves some twenty years before Gregory's lecture on the Sentences.83 that he saw as either complex or incomplex: when he speaks of incomplex In both texts - which are very· dose to each other - he sees the difficulty as conceptual terms in the Summa Logicae, or of simple or composite concepts in the stemming not from the mereological structuration in itself, but from the apparent Quodlibeta,73 he is now referring to intellectual acts. The criterion for the distinction, necessity for an order among the parts of a mental proposition if such parts are to be nevertheless, remains the same. It corresponds to the one Ockham gives in his admitted. How could the mind distinguish, otherwise, between, for example, 'all commentary on the Categories when explicating the strict sense of the term men are animals' and 'all animals are men', which, apparently, have the same parts, 'incomplex' as applied to words or signs (dictiones in this case): 'Strictly speaking, but obviously not the same truth-conditions? A spatial ordering of parts, such as the an incomplex item is a simple word, one single word, that is, without the addition of one we find in written sentences, is excluded, of course, since the mind is not itself other words, such as 'man', 'runs', 'lion', 'goat'. And we call 'complex', by contrast, extended in space. And so is a temporal ordering, such as the one we have in spoken whatever is composed of several words.'74 A simple sign is a sign no part of which sentences, since the parts of a mental proposition, whatever they are, should be is itself a sign. A simple- or incomplex - intellectual act, by transposition, is an simultaneously present to the mind somehow. Ockham's solution to the riddle is intellectual act no part of which is itself an intellectual act. A complex one is an radical: it amounts to renouncing linear ordering altogether within mental intellectual act that has other intellectual acts as proper parts.75 propositions. It must be insisted that this way of drawing the distinction does not prevent This, he thinks, can be done in two ways. First, we could say that a mental simple conceptual acts from having other kinds of complexity, as long as they don't proposition is in fact a simple cognitive act (just as will), but that have intellectual acts as their proper parts.76It does not prevent them, in particular, it is equivalent to an ordered combination of terms such as the corresponding one in from having some sort of semantical complexity. Ockham, indeed, is quite ready to spoken language.84And second, we could admit that the parts of such non-equivalent admit that a simple intellectual act, whether intuitive or abstractive, can have a propositions as 'all men are animals' and 'all animals are men' are not the same, plurality of different objects all at once,77 or that a simple concept can since a part of the former is the complex act corresponding to 'all men', which is not simultaneously signify different things in d!fferent ways.78 This is a point that will a part of the. latter, and a part of the latter is the complex act corresponding to 'all turn out to be of utmost importance in the discussion of connotative concepts that animals', which is not a part of the former.85 will occupy us in chapters 4 to 6. It might seem at first glance that these two possibilities are exclusive alternatives for Ockham, among which he simply was not ready to choose when he wrote his 4.2 Propositional acts commentary on the Perihermeneias and his Questions on the Physics. A closer reading of the two passages, however, reveals that he took the two alternatives to be Some composite mental acts are non-propositional: the combination of the concept simultaneously acceptable. His point is that' some mental propositions are composed 'white' with the concept 'horse', for example, yields a complex conceptual act which of a subject, a predicate, and a copula, while some are equivalent to such is not yet a mental proposition. Other combinations, by contrast, involve a mental combinations'.86The admission of the former sort of mental propositions neatly copula and are - paradigmatically - true or false. These are the mental propositions demarcates Ockham's approach from the one Rimini will later favour. Human proper. In their most elementary form - corresponding to what Ockham calls thought, for Ockham, is basically compositional, as he explains in detail in the 'categorical' propositions, tqey are composed of a subject, a copula, and a Summa Logicae.87That mental propositions, for him, can sometimes occur as simple predicate,79 each one of which being either complex or incomplex.80More complex acts within the mind is a handy - and fascinating - possibility, no doubt, but it propositional acts can then be produced by combining such categorematical remains a merely parasitic device: we can form, so to say, unanalysed abbreviative propositions with each other with the help of connecting syncategoremata such as acts which, for some purposes, are functionally equivalent to more complex ones (in 'and', 'or', 'if', 'when', 'because', and so on.81 The mind, in this doctrine, is taken truth-conditions, for example), but the outreaching strength of human knowledge to be innately endowed with a capacity for some simple acts (intuitions and simple and reasoning - the possibility of a full-fledged science in particular - ultimately concepts, basically), and with a further set of recursive capacities for combining in hangs upon this fundamental and remarkable capacity we have for combining mental various (and non-arbitrary) ways whatever intellectual acts it produces into more acts into more complex ones in various ways. complex ones. The order-problem is avoided, nevertheless, since composite mental acts do not A problem that was intensely discussed in the fourteenth century about need, in Ockham's eyes, to be linearly ordered. Various intellectual acts of the same propositional acts is how they can have ordered parts. Gregory of Rimini, for one, is ~eneralsort (propositional acts, for example) are differentiated from one another by well known for having argued, in the early D40s, that this is simply impossible, the nature of their parts, not by their order. The relevant feature for Ockham's since a mental propositional act is not supposed to be spread out in space nor in time. solution here is Ihal an immediate part of a complex menial aci clIn itself be 34 OCKHAM ON CONCEPTS INTELLECTUAL ACTS 35 complex: the complex concept 'white horse', for example, is a distinct intellectual 4.3 Judicative acts act which can occur as such as an immediate part of a mental proposition. And the same is true when mental syncategoremataare involved.88 All this supposes, While the main function of conceptual acts in human thought is to contribute in admittedly, that the various parts of a complex mental act - and even the parts of its specific ways to the truth-conditions of mental propositional acts, the latter in tum parts - can all be simultaneously present to the mind. But Ockham never saw that as are characterized by their functional roles with respect to mental units of a further a problem. He thought, on the contrary, that the possibility of simultaneous sort, which are more immediately linked with behaviour: judicative acts. These are intellectual acts had to be acknowledged by any sound theory of mental activity,89 usually contrasted, in Ockham's texts, with merely apprehensive acts, the latter not and that it was, at least implicitly, by Aristotle's.9o His point was simply that the yet involving any special commitment of the thinking agent toward the truth-value complexity of mental discourse does not require a linear ordering of the sort we of propositions.94At the time. of his lectures on the Sentences, when Ockham was observe in spoken or written discourses. still supporting the fictum-theory of concepts, this distinction was drawn in a quite The relevant idea is that some parts of 'all men are animals', for example, have straightforward manner: the apprehension of a complex propositional content was no equivalent in 'all animals are men'. The former, in effect, involves six parts identified with its very formation within the mind,95 and it was taken to be altogether, according to Ockham's complete analysis: presupposed by the judicative act of assentor dissent,96 In the Quodlibeta, however, the subject gets more complicated. Ockham now distinguishes between two sorts of (1) 'all' (a simple syncategorematic act), apprehensive acts having to do with propositions, and two sorts of judicative acts as (2) 'men' (a simple conceptual act), well. (3) 'all men' (a complex conceptual act having (1) and (2) as its parts), The first variety of propositional apprehension is, as in the earlier approach, the (4) 'are' (a simple copulative act),9I mere formation of the mental proposition, in other words the propositional act itself; (5) 'animals' (another simple conceptual act), in the actus-theory now favoured, it usually has conceptual acts as its intrinsic parts. (6) 'all-men are animals' (a complex propositional act having (3), (4) and (5) as its And the second one is the act of self-consciously considering - or cognizing - a parts). previously formed such mental proposition;97 it is, in Ockham's vocabulary, a reflexive act, that is, a mental act with another mental act as its object,98 The first The whole structure can be represented as a tree: form of judicative act, on the other hand, consists in the ordinary non-reflexive acceptance that so and so is the case. Not being reflexive, it does not, properly (6) speaking, have a propositional act as its object, but it simply follows upon the formation of some propositions, without any conscious grasping of them.99It can be ffi correctly described as having a propositional content ~ since it is, indeed, an assent (3) (4) (5) that so and so is the case - but not a propositional object: it is not an assent to a proposition. loo In so far as it can be said to have objects at all, these will be the cxternal things that the judgement pertains to. 101 The judicative act is a simple act in A this case, but it can be said to be functionally equivalent in many respects to a (1) (2) complex one in that it involves somehow an apprehension of a plurality of external things. 102 Judicative acts of the second sort, by contrast, presuppose the reflexive The important thing, however, is that some of its components do differ from those of apprehension of a propositional complex, and are reflexive therefore, in so far as 'all animals are men', which does not have (3) as a pl;lrt, and involves, therefore, their objects are mental acts (propositional acts in this case).103 nothing like (6). The underlying thesis is that when two complex intellectual acts are The highest sort of such judicative act, for Ockham, is the act of scientific of the same sort (two propositional acts, for example), their non-equivalence - if any knowledge (actus scientiae or actus sciendi) in the strongest sense of the phrase, this -ultimately depends on some of their respective parts not being equivalent to each knowledge, that is, by which some necessary truth is evidently known as such as the other. A non-linear arborescent order ultimately results from the very identity of the result of a valid from necessary premisses. 104 Since Oc\

36.0-61. It is true that he had then considered possible replies (see OPh II, p. 370: 'Et 36. This is argument 1 in Quodl. IV, q. 35, OTh IX, p. 472, and Quaest. in Phys., q. 1, OPh qui vellet tenere istam opinionem [that is, thefictum-theory], posset respondere ...'), but VI, p. 397. since he repeats the objections, but not the replies, in Quaest: in Phys., q. 1, he must 37. In SL I, 12, OPh I, p. 43, Ockham uses a more familiar version of the Razor in support ultimately have found these.replies unconvincing. of the actus-theory: 'frustra fit per plura quod potest fieri per pauciora'. On Ockham's 26. It should also be noted that Adam Wodeham, who must have personally followed the general use of this principle, see Maurer 1978. debate between Ockham and Chatton when he was a student, and who had been quite 38. This is the gist of argument 4 in Quodl. IV, q. 35, OTh IX, p. 473 and in Quaest. in close to Ockham at the time, later held that, although the concept is indeed to be Phys., q. 1, OPh VI, p. 397; in the latter's formulation: ' ... sine tali ficto potest identified with a cognitive act rather than with a .fictum, none of Chatton's original sufficienter haberi subiectum et praedicatum'. arguments against Ockham's former fictum-theory was conclusive (see his Lectura 39. See Gal 1967, p. 198 and Weyand Etzkorn 2002, pp. 233-4. Secunda in I Sent., dist. 8, q. 1, ed. Wood and Gal 1990, vol. III, pp. 20-33). 40. Note that a term can have a supposition, for Ockham, without actually suppositing for 27. Quodl. IV, q. 35, OTh IX, p. 474: ' ... per actum potest sa1vari quidquid sa1vatur per something; it might simply purport to refer. See SL 1,72, OPh I, pp. 218-19. fictum' (transl. Freddoso and Kelley.1991, vol. I, p. 390). See also SL 1,12, OPh I, p. 41. Exp. in Perih., Prooemium, 7, OPh II, p. 361. 43: 'Omnia autem quae salvatur ponendo aliquid distinctum ab actu intelligendi possunt 42. See Quodl. IV, q. 35, OTh IX, p. 474: ' ... per actum potest salvari quidquid salvatur per salvari sine tali distincto ... ' fictum, eo quod actus est similitudo obiecti, potest significare et supponere pro rebus 28. This is the gist of argument 6 in Quaest. in Phys. q. 1, OPh VI, p. 398: ' ... difficile est extra, potest esse subiectum et praedicatum in propositione, potest esse genus, species, imaginari aliquid posse intelligi intellectione reali et tamen quod nee ipsum nee aliqua etc., sicut fictum'. The corresponding passage in Summa Logicae I, 12, OPh I, p. 43, pars sui potest esse in rerum natura nee esse substantia nee accidens'. The same mentions only signification and supposition, these being, in the last analysis, the really argument - in pretty much the same words - is the first one adduced against the fictum­ basic functions of the concept: '... supponere pro alio et significare aliud ita potest theory in Exp. in Perih., Prooemium, 7, OPh II, p. 360. competere actui intelligendi sicut alii signo. Igitur praeter actum intelligendi non oportet 29. This is argument 5 both in Quodl. IV, q. 35 and in Quaest. in Phys., q. 1. As was usual aliquid aliud ponere.' Biard 1989, pp. 106-8, has.rightly insisted on the centrality of the in Ockham's times for arguments having to do with what was later called analyticity, it idea of the concept as a sign in Ockham's final adoption of the actus theory. is formulated in terms of what God could do without contradiction; in the Quodlibeta's 43. Exp. in Perih., Prooemium, 7, OPh II, p. 361: ' ... praedicta [that is, representing, formulation, OTh IX, p. 473: ' ... non est contradictio quod Deus faciat cognitionem suppositing, being part of a proposition ... ] propter alias rationes non possunt universalem sine tali ficto ... ' The argument in this case is directly inspired by Chatton competere intellectioni, quia difficile est salvare quid intelligam tali intellectione .. .'; (see Gal 1967, p. 199, and Weyand Etzkorn 2002, p. 234). what Ockham calls 'intellectio' in these lines is the act of intellection. See also Ord. I, 30. This is argument 2 in Quodl. IV, q. 35 and in Quaest. in Phys., q. 1. In the former's dist. 2, q. 8, OTh II, p. 268, where the two arguments adduced against the actus-theory formulation: '... tale fictum impediet cognitionem rei ... quia '" est ... quoddam basically amount to the same point. tertium medium inter cognitionem et rem'. A similar argument had been standardly used 44. Ibid.: 'tamen omnia ilIa [that is, representing, etc.] possunt verius competere alicui since in the polemics against Aquinas's theory of the mental word (see qualitati exsistenti in anima, quae terminat actum intelligendi' (the italics are mine). Olivi's Tractatus de Verbo 6.2.3, ed. in Pasnau 1993, p. 144: '[tale verbum] potius esset 45. See Ord. I, dist. 2, q. 4, OTh II, pp. 138-40. ad impedimentum'; on the use of the argument by other authors between Olivi and 46. See Lindbergh 1976, Tachau 1988, chap. 1. Ockham, especially William of Ware and Walter Burley, see Panaccio 1999a, pp. 47. Ockham's main discussion of the intelligible species theory is in Rep. II, q. 12-13, OTh 192-9). Chatton puts forward a similar argument againstPeter Aureoli in the Prologue V, pp. 251-310. For his discussions of the species in medio and of the sensible species, ofhis Lectura (Wey 1989, p. 87), but only in connection with intuitive knowledge, which see Rep. III, q. 2 and 3, OTh VI, pp. 43-129. is irrelevant for Ockham's theory of the jictum which had to do only with general 48. For instance in Tachau 1988, Tweedale 1990, Spruit 1994-95 (Vol.1, 1994), Perler concepts. 1996a, Pasnau 1997, or Stump 1999. 31. Adams 1987, p. 85. 49. See Spruit 1994-95 (VoLl, 1994), p. 295, Pasnau 1997, pp. 189-94, and Stump 1999, 32. This is argument 7 in Quaest. in Phys., q. 1, OPh VI, p. 398: ' ... tale idolum plus passim. Perler 1996b, p. 280, approves ofthis diagnosis: 'Spruit', he writes, 'correctly points differret a re quacumque quam quaecumque res differtab alia ... et ideo tale idolum out that Ockham's critique [of the intelIigible species] was based on a misunderstanding.' minus assimiletur rei ... ' It had first been put forward by Ockham in the Exp. in Perih., 50. Rep. II, q. 12-13, OTh V, p. 256. Prooemium, 7,OPh II, pp. 360-61, but is not found in. the Quodlibeta version. 51. Spruit 1994-95 (VoLl, 1994), p. 156. 33. This is argument 3 in Quodl. IV, q. 35 and in Quaest. in Phys., q. 1. In the Quodlibeta's 52. In Rep II, q. 12-13, Ockham t1rst presents Aquinas's version of the theory (probably, as formulation (OTh IX, p. 473): ' ... ab aeterno erat coordinatio tot entium fictorum quot the editors point out, through the exposition of it by some disciple - OTh V, pp. 253-4, possunt esse diversae res intelligibiles, quae fuerunt ita necesse est, quod Deus non n. 2), and then briefly remarks that Scotus too subscribed to it, although for different potuit ea destructere ...' It is also found in Chatton's Prologue to the Lectura (Wey 1989, reasons, which Ockham does not even care to mention or discuss in this context (OTh V, p. 87), but who devised it first is unknown. p.256). 34. See Exp. in Perih., Prooemium, 10, OPh II, pp. 370--71. 5~.See in particular Aquinas's Summa Theologiae I, q. 84-6, and his Expositio in Aristotelis 35. On Chatton's connection to Scotus, see Gal 1967, p. 191, and Way and Etzkorn 2002, p. De anima IX. Some of Chatton's dicta on concepts are telling in this regard, for example in Rep. I, 54. Rep. II, q. 12-13, OT!I V, p. 253: 'Hic primo suppono quod species sit illud quod est dist. 3, q. 2 (in Gal 1967, p. 2(9): ' ... conceptus iste speciti<:us [that is. the concept of praevium actui inlelligendi et potest manere ante intellectiones et post, etiam re ahsente. man] hahet humanitatem pro ohiecto primo.' This is something Ockham could not very Et per consequens dislinguitur ah hahitu. quia hahitus intellectus sequitur actum well accept. intelligendi. sed species prael'edit tam aclum quam hahitum.' 40 OCKHAM ON CONCEPTS INTELLECTUAL ACTS 41

55. Thomas Aquinas, Quaest. disp. de potentia, q. 8, art. 1: 'Differt autem [conceptio] a 74. Exp. in Praedic., 4, 1, OPh II, p. 148: 'Striete dicitur incomplexum simplex dictio, hoc specie intelligibili: nam species intelligibilis, qua fit intellectus in actu, consideratur ut est una dictio sine additione alterius dictionis, sieut hie: homo, currit, leo, capra. Et per principium actionis intellectus ...;' the same point is made, for example, in Summa oppositum dicitur complexum omne compositum ex diversis dietionibus.' contra Gentiles I, 53, and in Quodl. V, 5, 2. 75. In the context of a discussion with Ockham, Walter Chatton gets explicit about this, by 56. Pasnau 1997, p: 191. assuming that a simple act, for both of them, is one 'that does not include several acts' 57. Ibid., p. 193. (actus simplex non ineludens plures actus); this is in the Prologue of his Lectura, q. 1, 58. Cf. Thomas Aquinas, Summa contra Gentiles I, 46: 'Species intelligibilis in intellectu art, 1, ed. Wey 1989, p. 36. praeter essentiam eius existens esse accidentale habet.' 76. Ockham makes a similar point about simple sentences in his Exp. in Perih. I, 4, OPh II, 59. Rep. II, q. 12-13, OTh V, pp. 272-6. p. 395: 'Intelligendum quod enuntiatio categorica dicitur simplex non per carentiam 60. Ibid., p. 273: 'Sed isto modo assimilatur intellectus suffieienter per intellectionem cuiuscumque compositionis et quarumcumque partium ...; sed dieitur simplex quia non causatam ab obiecto et receptam in intellectu, igitur non requiritur species.' componitur ex pluribus enuntiationibus.' 61. Tachau 1988, p. 132. 77. See, for example, Rep. II,q. 12-13, OTh V, pp. 28D-81, where Ockham admits that one 62. See Rep. II, q. 12-13, OTh V, p. 273: 'Item, tanta assimilatio requiritur in notitia can have a simple intuitive grasping of a whiteness and a blackness simultaneously, and intuitivaquanta in abstractiva; sed in intuitiva non requiritur 'aliquid, praevium then form a simple abstractive act pertaining to the same objects. cognitioni, assimilans; igitur nee in abstractiva.' On Ockham's acceptance of 78. See, for exemple, Ord. I, dist. 3, q. 2, OTh II, p. 405: 'Tertio, dico quod Deus potest cognitive assimilation for both intuitive and abstractive knowledge, see also OTh V, cognosci a nobis in conceptu simplici connotativo et negativo sibi proprio ..., et iste pp.287-9. conceptus est simplex, quamvis distincta significet, sive principaliter sive secundario, 63. Rep. II,q; 12-13, OTh V, p. 274; the point about intuitive cognition had been made a few hoc est, vel in recto vel in obliquo ... ' pages before (p. 268). 79. See SL II, 1, OPh I, p. 241: 'Propositio categorica est illa quae habet subiectum et 64. Ibid., p. 274: ' ... sicut obiectum sufficienter repraesentat se in una cognitione, ita in alia praedieatum et copulam, et non includit plures tales propositiones'; see also Exp. in quae immediate sequitur intuitivam.' Perih. I, 4, OPh II, p. 396. 65. Ibid., p. 275: ' ... sicut tu ponis quod corporale potest esse causa partialis ad causandum 80. That the copula, and not only the subject and predicate, can sometimes be complex, speciem in spirituali, ita ego pono quod corporale est causa partialis ad causandum according to Ockham, can be gathered, for example, from his treatment of such phrases intellectionem in spirituali.' as 'begins to be' (incipit esse) or 'ceases to be' (desinit esse) (SL I, 75, OPh I, pp. 66. Ibid., pp. 275-6. 231-3), and from his discussion of truth-conditions for modal propositions without 67. Aquinas, for one, frequently insists that the intelligible species are not normally the dictum (SL II, 10, OPh I, pp. 276-9). On Ockham's admission of such special copulas objects of intellectual cognition, but the means of it, not the id quod but the id quo of as 'is believed by you to be', see Panaccio, forthcoming, section 2. intellection. See, for example, In De Anima III, 2: 'Manifestum est etiam quod species 81. Ockham calls all such complex propositions 'hypothetical' (see SL II, 1, OPh I, p. 241: intelligibiles quibus intellectus possibilis fit in actu, non sunt obiectum intellectus. Non 'Propositio hypothetica est illa quae ex pluribus categoricis est composita'); they include enim se habent ad intellectum sieut quod intelligitur, sed sieut quo intellectus intelligit conjunctive, disjunctive, conditional, causal and temporal propositions. ... ' This is a point Ockham could hardly have missed. He uses it actually in Ord. I, dist. 82. See Gregory of Rimini, In I Sent. Prol., q. 1, art. 3, ed, Trapp et al. 1978, pp 33-5. 2, q. 8, OTh II, p. 269, to show that intelligible species, if they were admitted, could not 83. See Exp. in Perih. I, Prooemium, 6, OPh II, pp. 354-8; and Quaest. in Phys., q. 6, OPh be identified with universals, since universals are generally supposed to be the objects VI, pp. 407-10. of general intellection while the species are not. 84. See Exp. in Perih., Prooemium, 6, OPh II, p. 356: 'Ad istud potest dici quod propositio 68. Exp. in Perih., Prooemium, 5, OPh II, pp. 35D-51. It should be noted that the thesis potest esse actus intelligendi aequiva1ens toti uni propositioni compositae ex realiter Ockham discusses in this passage - namely that concepts are intelligible species - is not distinctis, si talem ordinem haberent qualem habent in voce.' Aquinas's- who draws, on the contrary, a clear distinction between concepts and 85. Ibid.: 'Aliter posset dici quod ... in propositione in mente correspondente isti intelligible species (see above n. 55), but nothing suggests that Ockham thought it was. propositioni prolatae "omne animal est album" correspondet unus actus tamquam pars Ockham's editors mention Roger Marston as a possible candidate here (OPh II, pp. propositionis isti toti "omne animal" et alius actus isti "album", sed in propositione in 350-51, n. 1). mente correspondente isti propositioni "omne album est animal" correspondet unus 69. Rep. II, q. 12-13, OTh V, p. 268: ' ... sed nullus videt speciem intuitive; igitur experientia actus isti toti "omne album" et alius isti termino "animal".' non inducit ad hoc'; the sentence is quoted in translation and discussed in Stump 1999, X6. Quaest. in Phys., q. 6, OPh VI, p. 410: 'Propositio autem mentalis aliqua componitur ex p.179. subiecto et praedicato et copula, et aliqua erit aequivalens tali composito' (the italics are 70. Ibid.: 'Item, nihil est ponendum necessario requiri naturaliter ad aliquem effectum nisi mine). The same is said, albeit a little more tentatively, in Exp. in Perih., VPh II, p. 358: ad ilIud inducat ratio certa procedens ex per se notis vel experientia certa; sed neutrum 'Sed de propositione concepta, quae tantum est in mente, potest dici quod aliq!1a ....' istorum inducit ad ponendum speciem, igitur etc.' This passage almost immediately X7. See in particular SL I, 1-4, where the real composition of mental propositions is precedes the sentence quoted in the previous footnote. explicitly stated and analysed. 71. Perler 1996a, especially pp. 243-50. xx. This approach can, of course, be further exploited. Spade 1996, pp. 128-9, quotes an 72. Ord., Prol., q. 1, OTh I, p. 16: ' ... inter actus intellectus ... unus est apprehensivus et est anonymous laIc fourteenth-century text mentioning a doctrine according to which for respectu cuiuslibet quod potest terminare actum potentiae intellectivae, sil'c sit something 10 be the suhjecl of a mental proposition, it does not have to OCt'ural a special complcxum sil'c incomplexum' (italics are mine). position in a given order. hut it must. instead. he accompanied hy a special mental 71. For example. in .'II, I. 3. OPh I. p. II; or Quodl. II..l. OJ''' IX. p. 121. ,n'/I(·/III'X0/'('lI/a. whidl identities it as Ihe suhject. and so 100 for the predicate, I'a/'l' 42 OCKHAM ON CONCEPTS INTELLECTUAL ACTS 43

Spade, this notion appears to be but a rather natural elaboration on Ockham's ideas in 103. Ibid., p. 234: 'Sed loquendo de secundo actu sciendi vel assentiendi, dico quod ille the texts we have just been discussing. proprie est actus complexus habens pro obiecto complexum'; and Quodl. IV, q. 16, OTh 89. See, for instance, Ord. I, Prol., q. 1, OTh I, pp. 19-20, where Ockham argues that Plato IX, p. 377: 'Sed secundus assensus, naturaliter loquendo, necessario praesupponit can all at once love Socrates and know that he does, which requires the simultaneity of apprehensionem complexi ... ' at least three mental acts: loving Socrates, knowing Socrates, and knowing about one's 104. See Exp. in Phys., Prol., 2, OPh IV, p. 6: 'Quarto modo dicitur scientia notitia evidens love for Socrates, two of which are cognitive acts. Also: Quodl. IV, 17: ' ... falsum [est] veri necessarii nata causari ex notitia evidenti praemissarum necessariarum quod partes conceptus compositi succedunt sibi in mente sicut in voce; quia in mente applicatarum per discursum syllogisticum.' The term 'science', of course, can also be conceptus illi simul existunt in subiecto indivisibili, in voce autem non.' taken in less stringent ways, as Ockham explains in the same text, to designate, for 90. See Exp. in Perih., Prooemium, 6, OPh II, p. 358: 'Nec umquam invenitur ab Aristotele example, the certain knowledge of any old truth, or the evident knowledge of a truth, or quod negaret duos actus intellectus posse simul esse in intellectu ...' again the evident knowledge of a necessary truth (ibid., pp. 5-6).

91. In Rep. II, q. 12-13, OPh V, p. 280, Ockham holds that the intellectual act corresponding 105. Ibid., p. 9: 'Nam obiectum scientiae est tota propositio nota ~.. ' Please note that this was to the copula is always complex: 'Sed actus terminatus ad copulam est complexus most probably written when Ockham had already turned to the actus-theory of concepts, quatenus terminatur immediate ad totum complexum.' But this was written when and drawn the distinction we commented above between two sorts of judicative acts in Ockham favoured the fictum-theory: the mental act 'terminating' in the copula is not Quodl. III, 8. here identified with the copula itself (see also his early discussion of the copula in Rep. 106. See above nn. 99-100. II, q. 1, OPh V, pp. 17-23; Karger 1996, pp. 221-7, offers a good study of these 107. See Exp. in Phys., Prol., 2, OPh IV, p. 6: ' ... scientia ... aliquando acclpltur pro passages). His later formulations, in the Summa Logicae in particular, generally collectione multorum habituum ordinem determinatum et certum habentium. ... Et presuppose the simplicity of (at least some) copulative acts. See, for instance, SL I, 3, scientia isto modo comprehendit tamquam partes aliquo modo integrales habitus OPh I, p. 11, where he admits that there are verbs (and hence copulas) among incomplex principiorum et condusionum, notitias terminorum, reprobationes falsorum mental acts (but see above n. 80). argumentorum et errorum, et solutiones eorum. Et sic dicitur metaphysica esse scientia 92. This condition, however, is not necessary since Ockham, as we saw, accepts the et naturalis philosophia esse scientia, et ita de aliis.' possibility of simple propositional acts that are merely 'equivalent' to complex combinations of terms (see above n. 84 and 86). 93. For a further discussion of this point, see below chap. 8, sect. 2. 94. See Ord. I, Prol., q. 1, OTh I, p. 16: 'Alius actus potest dici iudicativus, quo intellectus non tantum apprehendit obiectum, sed etiam illi assentit vel dissentit.' 95. As Karger 1996, p. 226, n. 105, remarks, this identification is riot explicitly asserted in Ockham's earlier works, but it is presupposed. See, for example, Rep. III, q. 2, OTh VI, p. 85, where he refers to 'the apprehension or formation of a complex unit' (apprehensionem sive formationem complexi). 96. See Ord. I, Prol., q. 1, OTh I, p. 17: ' ... actus iudicativus respectu alicuius complexi praesupponit actum apprehensivum respectu eiusdem.' 97. Quodl. V, q. 6, OTh IX, p. 501: ' ... duplex estapprehensio: una quae est compositio et divisio propositionis sive formatio; alia est quae est cognitio ipsius complexi iam formati ...' 98. See Quodl. II, q. 12, OTh IX, p. 165, where Ockham distinguishes reflexive acts from direct ones: ' ... vocatur actus rectus quo intelligimus obiectum extra animam, et actus reflexus quo intelligitur ille actus rectus.' 99. See Quodl. III, q. 8, OTh IX, pp. 233-4: 'Loquendo de primo assensu, dico quod ille actus non habet complexum pro obiecto; tum quia ille actus potest esse per solam formationem complexi et sine omni apprehensione complexi ...; tum quia laicus sciens quod lapis non est asinus, nihil cogitat de propositione ... Licet assentiat et sciat quod sic est in re vel non est, mediante propositione formata in intellectu, tamen hoc non percipit ...' 100. Ibid., p. 234: ' ... dico quod proprie loquendo, non debet dici quod aliquid scitur isto actu, sed quod isto actu scitur quod lapis non est asinus .. .' See also Quodl. IV, q. 16, OTh IX, pp. 376-7, and Quodl. V, q. 6, OTh IX, p. 500, where the same distinction is drawn. 101. Quodl. III, q. 8, OTh IX, p. 234: ' ... iste actus [namely the non-reflexive judgement that a stone is not a rabbit] habet res extra pro obiectis, pula lapidem et asinum ... ' 102. Ibid., p. 234: ' ... et iste actus aequivalet quantum ad multa ali<.:uicomplexo quo aliquid scitur.' Chapter 3 Concepts as Signs

Generality for Ockham is but a matter of signification, and concepts are the prime bearers of it. That concepts should be signs, therefore, is a crucial requirement of Ockham's nominalism. What it means, however, is not self-evident. In his 1994 book, Nominalisme, the French scholar Cyrille Michon has forcefully argued that Ockham's position on this point is ultimately unintelligible. My aim in the present chapter will be to take up Michon's challenge. I will first explain what the problem is with concepts being labelled as signs in the Ockhamistic framework (section 1). Ockham's definition of a sign in the Sum of Logic will then be scrutinized (section 2), and how it applies to concepts will be spelled out (section 3). I will explore, finally, the consistency of taking concepts to be signs in this precise sense with two central Ockhamistic tenets: conceptual atomism (section 4) and nominalism itself (section 5).

1. The problem: How can a concept ever be a sign?

The standard conception of a sign, in Ockham'stimes, was directly dependent upon Augustine's famous definition: 'a sign is a thing which, in addition to the impression it produces upon the senses, brings by itself something else in our thought'. I Taken literally, this venerable definition squarely excluded intellectual entities such as concepts from the realm of possible signs, since it required all signs to produce some 'impressions' of their own upon the senses. Already in the thirteenth century, however, the habit had been taken to sometimes classify intellectual representations as signs too and to drop, consequently, the Augustinian clause about the impression upon the senses. Roger Bacon, for one, was explicit about this in the 1260s: 'not all signs', he wrote, 'are given to the senses as a popular description of sign supposes, but some are given to the intellect only'.2 The rest of the Augustinian definition came to be taken as what really mattered: saying that a concept is a sign was, prima facie, to say that it brought to the mind something different from itself. One intuitive way of seeing how this raises a special difficulty for Ockham's later view that concepts are intellectual acts is to remark that the very notion of a sign as bringing about the cognition of something else seems to involve aternary structure where Ockham would admit of only two items. For we should have, apparently, not only the sign itself (that which brings about something) and the external thing (that about which a cognition is activated by the sign - its 'significate' in Ockham's vocabulary), but also a third item: the cognition - or recognition - that the sign is supposed to bring about within the agent who understands it. Describing the concept as a sign seems to require the postulation of some intellectual interpretation of it within the mind, which interpretation, of course, could not in tum be identified with

4~ 46 OCKHAM ON CONCEPTS CONCEPTS AS SIGNS 47

a sign without the threat of an infinite regress. Ockham, on the other hand, simply such a general sense. In another sense a sign is that which brings something to cognition equates the conceptual sign with the intellectual act, leaving no room for a third item and can supposit for that thing, or be added to a sign of this sort in a proposition (for which could serve as interpretation for the concept. Without interpretation - or example, syncategorematic expressions, verbs, and other parts of speech lacking a understanding - one could ask, how can we still speak of signs? How, in other words, determinate signification), or again that which can be composed of signs of either sort (for could Ockham literally apply to concepts the key-clause of the Augustinian example, propositions). And taking the term 'sign' in this sense the spoken word is not the definition of a sign, namely that a sign should 'bring something in our thought'? natural sign of anything.6 There is indeed a passage of the Ordinatio where the Venerabilis Inceptor discusses the meaning of a closely related phrase: ducere in notitiam (to lead to the In the first sense of signum, according to this characterization, a sign, just like an cognition [of something]). It does nothing much, unfortunately, to alleviate our image or a trace, is typically reminiscent of something which has previously been difficulty. Ockham there distinguished two senses of ducere in notitiam, none of known in some other way. But, Ockham says, this is not the meaning he wants in the which can be applied to the conceptual acts of his later theory. Here is the text: context of his logical treatise; it is the second one. Since conceptual signs are central to logic in his view, we can safely conclude that a concept, in his theory, is a sign in That something leads to the cognition of something else can be understood in two ways. the second sense. This, no doubt, is Ockham's considered answer to the question we Either it causes the cognition of this other thing by means of being cognized itself, in such are discussing in the present chapter. a way that the cognition of it should be the cause of the cognition of the other thing. Or Michon, however, takes this answer to be unsatisfactory. All signs other than immediately, without being cognized, just as the intellect leads as a cause to the cognition concepts, he remarks, are signs in the first sense. The spoken word, the barrel-hoop of any intelligible object. 3 in front of the tavern, even smoke with respect to fire, all of these are, he says, 'recordative' signs: they serve as signs in so far as they recall for us something that Conceptual acts - as Michon rightly points out4 - cannot 'lead to the cognition' of we have an antecedent knowledge of; and they do so by themselves the their significates in the first of these two senses, since they do not have to be objects of a new cognition.? The second sense, consequently, is purely ad hoc. In themselves the objects of a cognition to function as signs, in Ockham's view. But the extending the ordinary scope of the term ~ignumto make it include intellectual acts, second sense will not do either. A concept could 'lead to the cognition' of something Ockham, according to Michon, renders the notion intolerably 'equivocal'S and in this sense, only if it caused some other cognition within the mind. Which brings introduces a new meaning for it that has nothing left to do with the customary use of us back to our original riddle about interpretation and understanding: a third item ­ signum: 'The primary notion of a sign has been forgotten.'9 The new definition still something like an interpretation - seems to be needed for the concept to be a sign, boasts the Augustinian authority, since its first clause is that a sign, even in the in addition to the concept itself and its external significate, something that Ockham, logical sense, should 'bring something to cognition' but this traditional phrase has s surely, cannot accept. now to be taken in a completely new sense: a conceptual sign, in Ockham's mature I do agree with Michon, then, that this passage from the Ordinatio does not shed theory, cannot bring something to cognition in the same sense as a recordative sign much light on the general idea of a sign which is required by Ockham's mature does. A wholly new understanding of the cognitive scheme is required. But Ockham, theory. This is not too surprising, though. Not only did Ockham still subscribe to the Michon says, always explains conceptual knowledge on the basis of signification. So fictum-theory at the time, but he had not yet systematically developed the semiotical we face an apparently impenetrable circle: ' ... until now', Michon writes, 'the sign approach to concepts which came to be central in his later logical works and in the was defined on the basis of cognition, but if cognition is itself a signification, the Quodlibetal Questions. The context of the passage, moreover, is a discussion of the resulting circle renders both notions unintelligible'.l0 ideas of image (imago) and trace (vestigium); with no explicit reference being made Such is the 'paradox' of the concept as sign withinOckham's actus-theory of in the surrounding pages to the term signum. The explicitation of ducere in notitiam concepts, according to Michon. In order to assess this critique, we have to take a found in the Ordinatio simply does not meet the needs of Ockham's iater theory of close look again at the passage of the Summa Logicae I, 1 to which it pertains. signs and concepts; but, let me insist, it was not supposed to. In order to understand in what sense exactly a concept can be said to be a sign in Ockham's mature theory, we must tum, instead, to the first chapter of the Sum of 2. The two meanings of signum Logic, Where a full-fledged definition of signum is provided: The first thing to note about this passage is that it is a reply to an objection. Ockham Nevertheless, to silence hairsplitters, it should be pointed out that the word 'sign' has two in these lines is answering some 'hairsplitters' (protervos). To whatthey would have different senses. In one sense a sign is anything which when apprehended brings objected, he does not explicitly say, but we can find it out ourselves from the very something else to mind, even though it does not bring about a first-time cognition of the structure of the passage and its place in the chapter to which it belongs. As it thing, as has been shown elsewhere, but merely an actual cognition following ,upon a happens, this aspect is quite relevant for the ongoing discussion. Let me show how. habitual cognition of the same thing. In this sense of 'sign' a spoken word is indeed a in natural sign, just like any effect signifies at leasl its cause; and in this sense too is lhe Ockham had just distinguished the preceding paragraphs between three sorts barrel-hoop a sign of wine in the tavern. However. I am not here using the tenn 'sign' in of terms: spoken, written and conceptual. And he had identified two differences 48 OCKHAM ON CONCEPTS CONCEPTS AS SIGNS 49 between the first two on the one side, and the third one on the other, two differences clauses, the second one being a disjunction. A sign in the second sense, according to which are closely related to one another: first, the concept naturally signifies this reading, should: whatever it signifies, 'whereas the spoken or written term signifies only conventionally';11 and second, the signification of spoken and written words can be (A) bring something else to cognition, changed at will by the speakers, while that of concepts can't. and This is where the (real or imagined) hairsplitter breaks in. What he must have either (B1) be able to supposit for that thing in a proposition, asked, I take it, is this: aren't the spoken words natural signs too? Looking at the text or (B2) be able to act as a syncategorematic term in determining the logical and again with this in mind, its organization and progression become clear. Ockham semantical roles of other terms within the proposition, introduces the first sense of signum and then concedes that in this sense, yes, 'a or (B3) be composed of such signs, like a complex expression or a sentence. spoken word is indeed a natural sign'. But, he adds, this is not the sense he is interested in. Having given the second ~ and more relevant - meaning, he finally It would follow that all signs in the second sense, like all signs in the first sense, concludes that 'taking the term "sign" in this sense the spoken word is not the natural would 'bring to cognition' something different from themselves. This, however, can sign of anything'. The distinction between the two senses of signum is introduced to hardly be what Ockham had in mind, since syncategorematic terms at least - which settle a difficulty concerning spoken words, and not especially a problem about are explicitly counted as signs according to sense number two - do not satisfy clause concepts. (A) in his view: there are no particular objects that syncategoremata such as 'if', It is true, of course, that it does clarify, in the process, the precise meaning in 'and', 'all', or 'not', specifically bring to mind. They lack, as Ockham stresses in this which we can say that a concept is a sign in Ockham's epistemology. But it does so very passage, 'a determinate signification'. 16 by spin-off, the strategical point of the distinction being to neutralize the In all consistency with the rest of the semantical theory, the definition, instead, hairsplitter's objection that spoken words, after all, are natural, rather than should be read as a disjunction of three clauses, the first one of which is a conventional, signs. It so happens that both questions are simultaneously settled by conjunction. A sign in this sense must: the clarification of the second sense of signum. But, as the concluding sentences indicate at the end of each definition, the point of the move, first and foremost, was either (A) bring something else to cognition, and (B 1) be able to supposit for that to elucidate the status of spoken words, not that of concepts. This in itself tells thing (in a proposition) against Michon's claim that the second sense of signum was an ad hoc invention of or (B2) be able to be added to the signs of the precedent category within well Ockham in view, solely, of including concepts among signs. formed sentences in some language, Sticking to the first - and admittedly more traditional - meaning of 'sign', or (B3) be composed in an appropriate way of signs of the previous two nothing would have prevented him from saying, with the hairsplitter, that spoken categories. sounds are natural signs. 12 Doesn't a laughter, or a groan, or even a simple intonation of the voice, reveal, by some natural connection, the mental state of whoever It thus becomes clear that the second sense of signum does not merely correspond to produces it? Ockham agrees with this, and goes even further: spoken speech, just a particular case of the first, which would then have been artificially enlarged simply like anything else, naturally evokes its cause, exactly like smoke calls fire to the to include concepts, as Michon suggests. Signs which specifically belong to category mind. So there is a case for saying that a spoken word is a natural sign of something: (B2) in particular (syncategoremata, that is), do not (in so far as they are of interest it can, like anything else, recall to mind whatever it has a natural connection with (a to logicians) 'bring something else to mind'. And this is true not only of conceptual, causal connection, for instance). But the sense in which this is true is of no special but also of conventional syncategoremata. Ockham simply takes it for granted that interest for logic as Ockham conceives of it. '3 The second sense - the one he favours the term signum as it is used in logic, must apply to syncategoremata. This has in the context - corresponds to a completely different notion; Michon is certainly nothing to do with his postulation of an oratio mentalis. The medieval logicians' right that there is an equivocity here for signum. This equivocity, however, does not interest for syncategoremata and the habit of calling them signa were well depend - as Michon thinks it does - on Ockham's eagerness to include concepts entrenched long before himP His definition here simply takes due account of the among signs. 14 It follows, rather, from his attempt to delineate, besides the usual practice and terminology of his fellow logicians: a technical sense of signum was sense of signum, another one - a technical one - which could be of service for logic needed to cover - at least - the case of syncategoremata, whether spoken, written or as a science of arguments, propositions and their components, whether spoken, mental. written or mentaL's The point, admittedly, is not as obvious for signs of category (B3), that is, A precise understanding ofthis second definition is needed to catch its motivation sentences and other complex expressions, but in this case too, the advantage of the and relevance, and to see how exactly the two meanings of si/?num are connected disjunctive reading of the definition, as I just reconstructed it, is that it does not force with each other. Ockham's formulation is rather complex and punctuated with 'and's us to countenance propcr signiticatcs for such special kinds of signs. Sentences as a and 'or's. Even the of the definition is not immediately perspicuous. whole, in this perspective, do not need. any more than .I'YflCllfegoremala, to bring The surface structure of the sentence apparently suggests a conjunction of two somcthing distinctive to mind in order to be correctly described as signs in the 50 OCKHAM ON CONCEPTS CONCEPTS AS SIGNS 51 logical sense. This is in harmony with Ockham's general policy of leaving it to the (BI), to be able to play syncategorematic roles (B2), or to be composed of such signs sole categorematic terms to have proper significates. (B3) ~ correspond to derivative functions which all have to do with the combination Even limiting ourselves to categorematic terms, the clause of definition number ofsigns with each other within sentences. Ockham's technical definition restricts the two which allows us to count them as signs - the conjunction of conditions (A) and domain of signs which are of interest for logic, to those that can occur, in some (B I) - does not correspond to a particular case of the first sense of signum. The first capacity or other, in the well-formed sentences of a language. The primary definition of signum - the general and non-logical one - had three clauses, only one advantage of this move is to provide at the outset, for students of logic, a precise of which matches clause (A) of the second definition. A sign in Ockham's first sense characterization of signs in terms of the semantical and syntactical functions which must not only 'bring something else to cognition' - the Augustinian condition - but will be studied in detail in the subsequent chapters of the Sum of Logic. In so far as it must do so, in addition, by being itself cognitively apprehended (which could be this strategy produces something new, it is not only to legitimize the classification of called the apprehension condition) and by reviving a previously possessed concepts as signs; it is, mainly, to identify a well-delineated network of semantical knowledge of the thing (the reminiscence condition). Now, these last two conditions and syntactical features which are basic for the study of logic. These features do not are utterly absent from Ockham's second definition of signum. His use of the phrase especially depend upon Ockham's particular theory of concepts. The point, 'to bring something to cognition' does not activate by itself the apprehension originally, is to explain in what sense of signum spoken words can be said to be condition nor the reminiscence condition: some things can 'bring something else to conventional, rather than natural, signs. And this leads in the process to the mind' without functioning as 'recordative' signs in Michon's sense. formulation of a new definition, based on what we call today semantical and Is this the case only for concepts, as Michon takes it to be? This is doubtful. The syntactical relations. This definition is suitably tailored to the needs of logic as a correct use of a demonstrative in ordinary language, for example, does not science (terminist logic, in particular). It also suits Ockham's epistemology, of presuppose on the speaker's or the hearer's part any previous knowledge of its course; but this was not its sole - or even primary - intent. There is no objectionably referent. The reminiscence condition, at least, is irrelevant here. And learning new ad hoc manreuvre here, after all. names, whether proper or common, cannot presuppose either, in Ockham's view, a previous acquaintance with all of their significates: 'horse' signifies all horses for me, even those I have never seen; 'baobab' in my spoken language signifies all 3. Conceptual roles baobabs even though I have never seen any. Nothing implies - or even suggests - in the context of Ockhamism that all linguistic signs other than concepts should satisfy To say that internal thought is made up of signs comes down, in this perspective, to all three conditions for being a sign in sense number one; quite to the contrary. crediting it with a propositional structure, the very subject predicate structure indeed In so far as Ockham introduces a new sense for signum, it is to delineate the sort that the terminist logicians had been studying in spoken languages for decades of objects that interest logic as a scientific field. These objects are characterized by before Ockham. The basic idea for this language-minded approach to thought was certain functions - such as bringing something to the mind, suppositing for already in Aristotle, for whom concepts (noemata), although prior to linguistic something within a proposition, or being able to act as syncategoremata - which are words, could combine, just like words, into propositions, true or false. 19 The whole not specific to concepts (even though, of course, concepts are their prime bearers in Aristotelian tradition had acknowledged this requirement, even when it had not Ockham's epistemology). made use of the vocabulary of signification to talk about concepts. Thomas Aquinas, The old Augustinian expression 'to bring something else to mind', which is still for one, tended to reserve the term signum for recordative signs,20 but he accepted used by Ockham to introduce the first sub-category of signs in the second sense ­ without qualms the idea of true or false mental propositions, structurally composed 21 the categorematic terms - cannot receive, as we have seen in section 1, any of the of 'concepts' representing external things. This amounted in effect to an acceptance meanings Ockham had identified a few years before in the Ordinatio for a related in principle ofthe oratio mentalis model. Ockham's originality was to systematically phrase, ducere in notitiam.18 This, however, in no way entails that his use of it at the draw the consequences of the approach, and to mobilize the whole apparatus of beginning of the Summa is out of place, arbitrary, or especially ad hoc. It does leave terminist logic - significatio, suppositio, and so on - for the fine-grained description aside, admittedly, the causal connection that might have been suggested prima facie of thought. by 'facit venire' (to bring to): the cognition brought about by a sign does not, in Identifying the concept with the cognitive act (as in Ockham's mature theory) Ockham's new view, have to be something different from the sign itself. But it rather than with its object (as in his former theory and in that of Aquinas) does not retains what Ockham takes to be really essential in the Augustinian phrase for render this terminological transposition less intelligible; on the contrary. Once the characterizing categorematic terms in general (whether spoken, written or mental): sign is defined by its semantical and syntactical functions, nothing in principle the referential drive, the fact that they relate the mind to something different from prevents these functions from being attributed to mental states (or 'acts') rather than themselves. The role of clause (A) in Ockham's technical definition of siRnum is to 10 mysterious ideal objects such as the ficta of the former theory or the verha introduce the semantical connection proper: the reference of (some) signs to objects n1cntalia of : 'for an act of understanding can signify something and can in the world. supposit for something just as well as any sign', Ockham remarks.22What is required The other three clauses in the definition - to he able to supposit in a proposition to legitimize a systematical extension of the terminist vocabulary to the study of 52 OCKHAM ON CONCEPTS CONCEPTS AS SIGNS 53 thought is the general validity of a certain functional description of human mind, basis of such functions,did not have as its main objective to cover the case of according to which some mental units with a referential import can be combined concepts, but to delineate with some precision - in a purely functional way - the set with each other into true or false propositions. This picture in effect was quite of units which are required by logic as.a science. generally accepted in Ockham's times. This is not to say that it is a trivial picture. The precise functional model Ockham finally adopted supposes a rather complex· conception of the working of the mind. 4. Atomism or propositionalism? Let us consider again, to bring it out, the functional roles that are mentioned in one way or another in the new technical definition of signum. There are four of them: A number of interesting problems still remain about Ockham's treatment of concepts as signs in the second sense. One of them is whether Ockham's semantical theory is (RI) bringing something else to cognition (this is clause (A) in the presentation I best labelled as 'atomistic' or 'propositionalistic', given his technical definition of gave above of Ockham's definition number two), signum, a debate that went on not too long ago in.the literature.23It comes out, at this (R2) suppositing for this in the context of a proposition (this is clause (Bl», point, that propositional roles loom large in the logical definition of a sign. Yet it (R3) being added within a proposition to signs satisfying both (Rl) and (R2) (this must be stressed that this definitional insistence on propositional combinations is by is clause (B2», no means incompatible with the theory of conceptual meaning as being of an (R4) being a well-formed combination (in one language or other) of signs atomistic sort basically. A correct understanding of how concepts can function as satisfying either (RI) and (R2), or (R3), a complete sentence for example signs is at stake here. Simple concepts, for Ockham, receive their signification one (this is clause (B3». by one, on the basis of the intuitive acts that originally triggered their formation. As we have seen in the previous chapters, getting intuitively acquainted, say, with a (Rl) corresponds within Ockham's semantics to signification in the strong sense ­ particular chickadee is both sufficient and necessary, under normal conditions, for the significatio jinita, (R2) to supposition, (R3) to the syncategorematic functions causing the formation of a simple natural-kind concept for chickadees. More general (that of the copula, for instance, or of quantifiers, or connectors), and (R4) to the concepts.- genus concepts - presuppose the antecedent formation of several species functions of the proposition itself, that, in particular, of being the bearer of a truth­ concepts; their meanings, nevertheless, are not given to them by a process of value. Ockham's definition exploits these four types of roles to differentiate three syntactical - let alone propositional - combination, but by a causal chain.24 kinds of signs: those that fulfil (Rl) and (R2), they are the categorematic terms; Structured propositional thoughts become possible only after their categorematic those that fulfil (R3) only, they are the syncategoremata; and those that fulfil (R4), components - the concepts - have, each of them, independently received a such as propositions (or sentences) and other complex expressions. signification. Three out of these four roles - (R2), (R3) and (R4) - directly have to do with the A number of commentators have suggested, in the 1960s and 1970s, a formation of propositions. Taken together, they constitute the distinctive element in 'propositionalist' reading of Ockham's semantics, according to which the the technical sense of signum Ockham wants for logic. Human beings are depicted proposition rather than the term is 'the primary complete linguistic unit' and the as capable of coherently entertaining true or false propositionally structured beliefs 'primary reality' of cognition in Ockham's system.25Their main argument, precisely, about things other than themselves. The basic role of logic, in this perspective, is to rested on the technical definition of 'signum' we have just been considering. Look, explain what it is for a proposition to be true or false, and what it is for a proposition they would say, Ockham defines the very idea of a sign on the basis of propositional to be incompatible with another one, or to follow from it. The. technical notion of roles such as supposition; doesn't this show that signification as a of single sign Ockham thinks appropriate for the fiel\i is built up with these tasks in view. terms must presuppose the formation of some propositional combinations?26 We are Admittedly, this model of the mind also calls for an epistemology; Ockham does now in a position to see with some precision why thisis not so in Ockham's theory. provide one, which is discussed in some detail in other parts of this book. But his Among the four defining functions for signs (Rt) - (R4), three can befulfilled only distinctive epistemological doctrines about concepts, intellectual acts, intentional within the context of propositional complexes: supposition (R2), syncategorematic similarity and so on, are not presupposed by the new definition he proposes for roles (R3) and properly propositional roles (R4). But this is not so for (Ri). Which signum. is no detail. Categorematic terms are the fundamental stuff out ofwhich·thought and The identification of the concept with a sign on the one hand, and with an language are made in Ockham's semantics, and (Rl) identifies their primary intellectual act on the other hand does allow, within Ockham's mature theory, for a function: that of calling something else to mind. This is signification proper. It must wealth of connections between ontology, logic, epistemology and philosophy of helong to single categorematic terms before any proposition involving them can be mind. But this, on the whole, is quite legitimate. Terminist logic had been interested .limned. And this is especially true for conceptual signs: 'All judicative acts in semantical and syntactical functions such as supposition and syncategorematic presuppose ... the incomplex cognition of the terms, since it presupposes an roles long before Ockham, and the general intelligibility of such notions docs not apprehensive act, and all apprehensive acts with respect to complexes presuppose the depend upon how Ockham develops his epistemology in the Ordinatio or in the incomplex cognition of the terms.' 27 In other words, a mental judgement presupposes Quoe/liht'tal Questions. The second definition of siRnum, which is built up on the the formation of a menlal proposition, and mental propositions, in so rar as they are 54 OCKHAM ON CONCEPTS CONCEPTS AS SIGNS 55 complex arrangements of terms, presuppose that their categorematic components plus his central contention that concepts are signs in this very sense. All concepts, in have already been endowed with cognitive signification: they should already, virtue of their being signs, must be able to contribute something to the truth­ independently of any proposition, refer the mind to some objects in the world. conditions of some mental propositional complexes, and to do so as constitutive Functions (R2) and (R3) can be fulfilled in the language of thought only if some parts of these complexes. terms - the categorematical concepts, namely - antecedently fulfil (RI). Such is Conceptual atomism, in Ockham's theory, is the thesis that concepts acquire their Ockham's brand of conceptual atomism. signification one by one before entering any proposition whatsoever. This is a central That the idea of supposition occurs in the definition of 'signum' in chapter I of tenet. But it goes hand in hand with a truly compositional view of the working of the the Summa Logicae, and in the definition of 'significare' in chapter 33, in no way intellectual mind. The combination of these two elements - conceptual atomism and reflects a priority of supposition over signification. The definitional order, here, does a compositional theory of thought - is indeed the gist of Ockham's doctrine of the not reveal the order of priority among the functions themselves. Why should it? The oratio mentalis, just as it is today of Jerry Fodor's doctrine of the language of ideas of signification (as a prepropositional function) and supposition (as a thought.32 propositional function) are intimately connected in Ockham, to the point of interdefining each other.28 The corresponding functions, however, are clearly hierarchized: signification is prior and more fundamental, supposition is derivative, S. Types and tokens and so. are the other propositional roles in their own way. The notion of a propositional act that would be primary - or even simultaneous - either Yet this raises another important - and delicate - question about the integration of chronologically or logically with respect to the signification of its categorematic concepts among signs within the framework of Ockham's nominalism. It has to do components ultimately finds no place in Ockham's semantics. And the with the inevitable type/token distinction. One might ask, shouldn't a signin the (chronological or logical) primacy of a mental proposition over the signification of logical sense be repeatable? Ockham's intellectual acts are not. They are singular its conceptual components is incompatible with Ockham's most distinctive tenets in and transient qualitative episodes in the life of the mind. The repeatability epistemology.29 requirement, on the other hand, seems implied by the very wording of Ockham's On the other hand, there is an important kernel of truth in the propositionalist technical definition: in order to be a sign in this sense, as we insisted, a concept must interpretation that must not be minimized. All three sorts of signs in the logical sense be able to enter some propositions. It must be the same concept, consequently, that - categorematic terms, syncategoremata, and complex phrases - must, by virtue of first receives a signification atomistically, and that enters, afterwards, different the definition of signum favoured by Ockham, be able to play a part within a propositional acts. How is this consistent with Ockham's nominalism? Repeatability, proposition. This has to be true of concepts, in particular, in so far as they are signs prima facie, is a feature for universals. But Ockham, of course, would tolerate no in the technical sense. Nothing is a concept unless it can be combined with other universal in his ontology. Concepts, in the mature theory, are identified with singular concepts and mental syncategoremata into mental propositions. The propositionalist qualities of individual agents. How can such entities ever serve as repeatable signs? commentators went too far, no doubt, when they suggested that a sign 'has no In , most of the time, it is the types that seem to matter, and not the meaning or validity, but in the process of a proposition'.30 But what they were rightly singular episodes of their instanciations. Standard statements are of the form 'x is a pointing at is that any unit of thought or language, for Ockham, must be either a name' or 'x is combinable with y in a certain way', or 'p follows from q', etc., which proposition, a combination of propositions, or an element capable of occurring make more sense about types prima facie than about tokens. Ockham himself, within a proposition. indeed, often speaks as if it was types he was talking about. Think of a principle like Note that in the latter case, it needs not ever actually appear in any propositional the following, which is central in his semantics: a given term can have different context in order to be a sign: the definition of signum does not require the actual suppositions in different contexts.33It looks more readily applicable to types than to occurrence of the sign in a proposition. But under normal conditions all conceptual (okens, no doubt. signs soon find their way into propositions. It is their normal· fate. The original The Venerabilis Inceptor never tackled the issue directly. But what he is formation of a general concept F, in so far as it is based on intuitive cognition in committed to on the topic can be reconstructed safely enough from what he Ockham's picture, would normally be immediately followed by the formation of at explicitly says. For one thing, Ockham was well aware that sucha distinction as the least one mental proposition of the form 'this is an F' or 'this G is an F'. Such type/token distinction can be drawn about mental acts. He thought of it as a special propositions only come after their component general categorematic terms have case of Aristotle's familiar distinction between numerical identity on the one hand, been endowed with significations, and as the effects of such endowments (effects and specific or generic identity on the other hand.34Speaking of mental acts in a which, let me insist, special circumstances could impede);3l yet in the vast majority quite different context - that of a discussion on virtues - Ockham states that an act of cases a newly formed concept will immediately enter some elementary mental which is numerically one 'cannot first occur without circumstances, and then with

proposition, and will most probably reoccur later on in the context of new - and circumstances' .1~Strictly speaking, a singular act of cognition is a transient episode possibly more complicated - propositional combinations. At any rate, it must he able indissoluhly linked to its temporal and spatial context of occurrence. Yet this should to. This requirement directly follows from Ockham's technical definition of 'sign', not keep us, as Ockham makes clear in the same passage, from speaking sometimes 56 OCKHAM ON CONCEPTS CONCEPTS AS SIGNS 57 of 'the same act generically, rather than numerically'.36 Having in view generic (or identity among the conceptual acts of a given agent. Let me propose the following specific) sameness, it is acceptable to say that the same act - the act of loving as a plausible candidate: two singular conceptual acts within a given agent belong to somebody, for example - reoccurs at different times and places in a given process. the same conceptual type if and only if (1) they both belong to a single causal chain The remark, of course, is directly relevant for the case of conceptual signs, in so far, of the kind indicated above, and (2) they are semantically equivalent to each other.39 precisely, as they are supposed to occur first outside any propositional combination, Of course, the satisfaction of condition (2) should follow on, normally, from that and then within various such contexts successively. What the passage shows at any of condition (1). But since we have not yet examined how this link is secured, we rate is, that Ockham's frequent speaking of types in psychological and semiotical better at this point mention both conditions, simply keeping in mind, for the matters was not inconsiderate. He felt ccmfident that talk which is apparently about moment, that condition (2) might turn out to be superfluous.40 types in such matters can be rendered legitimate within his nominalistic system, as However that may be, once we have clear and stringent identity conditions such a special case of resorting to generic or specific identities. as those - with or without condition (2) - there is no harm, presumably, in making In other words, the status of semiotical types such as the word 'horse', the use of a type-idiom, with such typical phrases as 'the term a' or 'the proposition p'. concept 'animal', or the proposition'a horse is an animal' , is but a special case with Such phrases canlegitimatelybe taken as distributively referring to certain singular respect to the general question of universals. Ockham's approach to it, then, should acts of the same kind, just as 'horse' distributively refers to singular substances of be expected to follow his preferred strategy for dissolving universals and abstract the same kind, and 'whiteness' to singular qualities of the same kind. entities: that of reinterpreting an apparent reference to a single universal into a Type-phrases, of course, do not usually occur in the plural. We do not say 'the distributive reference to several singular entities at once. 'Horse' signifies singular names "equus"', 'the concepts "horse'" or so on, in the way we say 'horses' or horses, and when taken in propositional contexts normally suppasits for singular 'animals'. But this can be seen as a surface feature of our spoken and written horses; no common horseness is required at any stage. This, surely, is the approach languages,probably reflecting the fact that what is said of a given mental act when Ockham would want to extend to the treatment of apparent references to semiotical we treat it as part ofa semiotical system should equally be said of a number of other types. Speaking of the word 'horse' or the concept 'horse' should be understood, in tokens connected with it in some determinate way. this perspective, as a way of distributively referring to singular speech or thought The same approach can be extended to the interpersonal situation. In order to episodes. legitimize the grouping of conceptual acts into types across agents, we need to How this is supposed to work obviously calls for some spelling out. But we can devise a similarly inspired set of precise conditions. We can follow for this the very build progressively on Ockham's own texts. Let us start with type identity within a same policy we have been following in the intrapersonal case: that of combining a single agent. The case is easier to figure out. How, let us ask, can different mental causal condition with a semantical one (the latter, maybe, being ultimately episodes in the life of a given agent be grouped into legitimate types? How is it that dispensable), What this approach yields is this: two singular conceptual acts a1 and several numerically different conceptual acts' of mine can be correctly counted as az of two distinct agents can be said to be of the same conceptual type if and only if: instances of the same concept? The obvious place in Ockham to look for an anSWer

to such a question is the theory of causal chains within the mind. As we have seen (1) the intrapersonal causal chain to which a 1 belongs has been set in motion (in above in chapter 2, a singular conceptual act would normally belong, according to the right way) by the intuitive apprehension of an individual i1which is of the Ockham's epistemology, to a definite causal chain of acts and habitus. Such chains same natural kind as the individual iz which triggered (in the right way) the normally start up with intuitive acts, and are made, afterwards, of abstractive acts (or intrapersonal causal chain to which az belongs; concepts) and their corresponding habitus in alternance: 'a habitus is the cause of an and act, but not of the same act numerically as the one it was caused by, but it is the cause (2) a l anda z are semantically equivalent. of a further act. Similarly, an act is the cause ofa habitus, but not of this habitus by which it was caused, but of another habitus or of another degree in the habitus.'3? Condition (1) presupposes, of course, that two different individuals in theworld can A typical chain has the following form: intuitive act ~ first abstractive indeed be correctly said to be 'of the same natural kind'. But Ockham has no conceptual act (prima abstractiva) ~ habitus] ~ conceptual actz ~ habitusz ~ problem with this. He willingly admits that some things are, as a matter offact, of conceptual act , ~ habitus ... There is, in this picture, a natural causal connection the same natural kind as certain others, independently of the human mind: 'that 3 3 between conceptual acts one, two, three and so on. And this causal connection is something is similar to another thing is not brought about by the intellect,he writes, taken to implement semantical equivalence among them. Bow exactly it does that is any more than that Socrates is white, or that Plato is white'.41 Ockham's rejection of something we will have to discuss in some detail later on. 38 Let us only assume, for universals as distinct entities was not intended to compromise all natural groupings. now, as a crucial part of Ockham's epistemology, that if a conceptual act is caused Specific and generic identities are OK, for him, as long as we do not imagine that in the right way by a certain. habitus, then it inherits the semantical features of the special extra entities are needed to serve as kinds. The legitimacy of general terms, act that had caused (in the right way) this very habitus. Causal chains of the relevant in Ockham's view, docs require individual things to he naturally ordered somehow kind must be semantically conducive. independently of the human mind. 42 What his nominalism resolutely wants to avoid This appropriately provides us with the possihility of a clear criterion for ·\Pl'c!!;c is Ihe reification of such orderings into special entilics of their own. The semantics 58 OCKHAM ON CONCEPTS CONCEPTS AS SIGNS 59

and the epistemology are devised to achieve just this. The criteria· for semiotical 6. SL I, 1, OPh I, pp. 8-9: 'Propter tamen protervos est sciendum quod signum dupliciter types, then, can legitimately make use of specific or generic groupings, especially if accipitur. Uno modo pro omni ilia quod apprehensum aliquid aliud facit in cognitionem they can be independently ascertained. venire, quamvis non faciat mentem venire in primam cognitionem eius, sicut alibi est In the end, the characteristic type-idiom of semiotics can be legitimized, within ostensum, sed in actualem post habitualem eiusdem. Et sic vox naturaliter significat, Ockham's system, by changing, as he often invites us to do, an apparent reference to sicut quilibet effectus significat saltern suam causam; sicut [= sic?] etiam circulus an abstract entity into a distributive reference to several well-determined singular significat vinum in taberna. Sed tam generaliter non loquor hic de signa. Aliter accipitur entities. The salient particularity of semiotical typing, as we have reconstructed it signum pro ilia quod aliquit facit in cognitionem venire et natum est pro ilIo supponere here, is that it connects somehow semantical features with causal history and natural vel tali addi in propositione, cuiusmodi sunt syncategoremata et verba et ilIae partes kinds. Ockham's conviction, obviously, is that certain singular conceptual acts are so orationis quae finitam significationem non habent, vel quod natum est componi ex talibus. cuiusmodi est oratio. Et sic accipiendo hoc vocabulum "signum" vox nullius est naturally connected with each other that semantical equivalence between them is signum naturale.' The English translation given here is very much inspired by that of (normally) secured. Loux 1974, p. 51, but it differs from it in many details. Among other things, I favour the On the whole, then, it is quite true that treating concepts as signs raises a number reading of sic instead of sicutat the beginning of the sentence about the wine in the of delicate questions for Ockham's mature theory of concepts. Some, as Cyrille tavern (thus following a number of good manuscripts; see the critical apparatus for SL I, Michon's, have to do with the exact sense of 'sign' that is used in such a context. line 58 in OPh I, p. 9). Others pertain to the consistency of this semiotical approach with Ockham's 7. Michon 1994, pp. 40-3. conceptual atomism or with his strictly nominalistic ontology for conceptual acts. 8. Ibid., p. 40. 9. Ibid., p. 156. What I have tried to show in this chapter is that the theory has fruitful resources at 10. Ibid., pp. 42-3. its disposal for dealing with these puzzles. The technical definition of 'sign' Ockham 11. SL I, 1, OPh I, p. 8: ' ...conceptus seu passio animae naturaliter significat quidquid provides at the very beginning of the Summa Logicae turns out to be an unusually significat, terminus autem prolatus vel scriptus nihil significat nisi secundum elaborate one. It harmonizes, in particular, the Augustinian semantical requirement voluntariam institutionem' (transl. Loux 1974, p. 50). of 'bringing something to mind' with the special interest of logic for syntactical 12. In what sense spoken words could be said to be natural signs had been discussed, for combination. Its motivation, mainly, comes from the needs of terminist logic as it example, by Roger Bacon in paragraphs 8 to 14 of his De signis (ed. Fredborg et aL had developed in the previous century. But the definition does inform us, at the same 1978, pp. 83-6). time, as to the exact sense in which a concept is said to be a sign in Ockham: in virtue 13. It should be noted that signs in the first sense of signum are not all of them 'natural' in of its causal history, it represents within the mind several individuals of the world in Ockham's view. The barrel-hoop in front of the tavern, which he also mentions as an such a way that it can stand for them in various mental propositional combinations, example of signs in the first sense, was a paradigmatic example of a conventional sign the truth-conditions of which it will then determine in definite ways. Ockham's in medieval discussions (see, for instance, Rosier 1994, p. 101). The barrel-hoop evokes distinctive nominalistic claim is that a semiotical system of this sort is indeed wine for whoever has already drank wine (or even only seen some I), but this is due to realized within the mind through causal chains of singular mental acts and habitus. a conventional connection. This is why the choice of sicut by the editors in line 58 of the above quoted passage (from OPh I, p. 9) seems unhappy. I would favour sic, along with many good manuscripts. The barrel-hoop, according to this reading, is given as an example, not of a natural sign, but of a recordative sign in general. The point, anyway, Notes is minor. 14. See Michon 1994,pp. 36---40. 1. Augustine, De doctrina christiana, II, 1.1, ed. Martin 1962, p. 32: 'Signum est enim res 15. See SL I, 1, OPh I, p. 7: 'Omnes logicae tractatores intendunt astruere quod argumenta praeter speciem, quam ingerit sensibus, aliquid aliud ex se faciens in cogitationem ex propositionibus et propositiones ex terminis componuntur.' This is the very first venire.' sentence of the treatise proper. 2. Roger Bacon, De signis 2, ed. Fredborg et aI. 1978, p. 82: ' ...non omne signum offertur 16. William uses this very same phrase (non habentfinitam signijicationem) to characterize sensui ut vulgata descriptio signi supponit, sed aliquod soli intellectui offertur ... ' On the syncategoremata in general in SL I, 4, OPh I, p. 15. progressive acceptance of purely intellectual signs in the thirteenth and early fourteenth 17. On the special importance of syncategorematic terms in medieval logic, see Kretzmann century, see Meier-Oeser 1997, chap. 2, and Panaccio 1999a, chap. 7. 1982. 3. Ord. I, dist. 3, q. 9, OTh II, p. 544: 'Sed tamen aliquid ducere in notitiam alicuius potest 18. See the text quoted above in n. 3. intelligi dupliciter: vel tamquall causativum notitiae alterius mediante sua notitia, ita 19. See Perihermeneias, 1, 16a9-13. quod notitia ipsius sit causa notitiae alterius. Vel immediate sine notitia, sicut intellectus 20. See Thomas Aquinas, Quaest. Disp. de Veritate IX, 4, ad. 4; and S. Theol. III, 60, 4, ad. 1. ducit tamquam causa in notitiam cuiuslibet intelligibilis.' 21. This is the Thomistic theory of the 'mental word' (verbum mentis) as it is presented. for 4. Michon 1994, p. 42. example, in Aquinas's commentary on John's Gospel (Super Evang. S. Ioannis 1) and in 5. This problem, actually, also arises in connection with the tirst sense of ducerl' in his commentary on the Pcrihermeneias. I have examined this doctrine in some detail notitiam: both meanings, as characterized in the passage quoted above. require the elsewhere (Panaccio 1l)lJ2b, Il)l)l)a, chap. 6, 200 I). causation of some cognition. 22. SL I. 12. OPh I. p. 4]: .... supponere pro alio et signiticare aliud ita potesl competere aclui intclligcndi Sleul alii sigllo' (lrans!. Lnux 1974. p. 74).

to

61

the pp.

also

facit

for

See

plus

respect

Ockham's

1992a,

equivalence'

non

679.

with

by

albus'.

relevant

p.

sit

alteri

as

Panaccio

IX,

285.

in

p.

Plato

'identical

simile

OTh

II,

'naturalistic'

sit

be

Ockham

detail

25,

quod

to

as

OTh

by

q.

vel

8,

unum

said

some

VI,

q.

in

labeled

2,

quod

albus

then

sit

are

hoc

Quodl.

SIGNS

dist.

I,

ad

acknowledged

AS

literally

and

discussed

these

3.3.

'et

Sortes

be

Ord.

is

316;

sect.

385:

see

quod

p.

tokens:

p. 5,

could

');

explicitly

IV,

hoc

IV,

CONCEPTS

is

chap.

ad

OTh

distinct

system

cospecificity

OTh

in

1,

of

5,

facit

q.

q.

below

aequivalentiam

natural

30,

equivalence

30,

features.

quam

Ockham's

together

this

on

per

dist.

dist.

on

I,

I,

did,

it

eadem

247-67. intellectus More

Ord.

semantical

Ord. ('

position

grouping Semantical

If

42.

39.

40.

41.

a

I,

it

q.

of

non are

and

de

not

287.

and

pro

same

term

De

uno

must

and

primo

verb

actus

p.

that simple

close

OTh primo

potest

detail

sed

It

realiter

senses

between

strongly

by

exists'), Prol.,

numero

any

of

1,

The notitiam

in -

the

IX,

Et

I,

acts

intellectus

'Nam

nisi

thought

very

whatsoever.

q.

supposition'

...

...

74-96,

propositionis

'

definition

of

'this

modo

is

numero

said

occasions,

which

OTh

incomplexam

these

basic

supponere

Ord.

habitus,

is

de

quod

his

pp.

habitu.'

,propositione

eiusdem

pars

Prol.,

proposed

21,

'significative'

even

human circumstances,

est

distinction

actus

in

3,

in

some

197),

I, which

(see -

q.

more

distinguuntur

or material

vera

of

non

causa

patet

p.

esse

on

proposition

The

supposition:

aliqua

following:

III,

propositional

discussion

Ord

idem

...

notitiam

and

discussed

'

cst

1998).

sed

' chap.

natum

adverb

praesupponit

other

gradu

apprehendere, special

mutandis,

of

-

in

any

...

apprehensivum.

the

1974,

in

picture

exhibit

sitting',

esse

vel

the

potest

iudicioconsequente~).

to

have

iudicium

as

each

actus

two

approach.,.-.

simple

Quodl.

I

is

page:

actus,

1989,

(Fodor

can

in

altcrius

propositione

case(see Loux

mutatis

sine

both

supposition independently

sequens

the

(significatively)

illud

Ockham's

first Junction

potest

actum

modo numero

from

-

iudicativus

1981. supposition.

respect happen,

same

et

vel

the

causa

acts

con

Biard

supponit

and

the

uses

praesupponit

287.

non

illidsignum

same,

to

transl. aliqua

Similiter,

the

dico

'Socrates

est

in

passim.

statements

with

spirit,

concepts

p.

been

per

he

...

intuitiva

actus

Biard in

could

'

personal

distinct

of

The

habitus

proposition

non

and

where

intuitiva

IX,

personal

former.

quando

vel propositionalist

nec

down

actus.

such

normally

and

a

205-6;

-

have

applied

art; same

quocumque

351:

quando

CONCEPTS

exists', habitus

of

have

cognition

propositional

'significative'

is

omnis

illa

notitia

the

complexi

praesupponit

by

of

p.

OTh

chapter,

pp.

are, theory circumtanciis'.

cognition

95-6,

207.

developed

as

the

also

they

the ...

can

ON

by

233-6,

1979,

cognition

I,

alterius

illa

'

quod

further

p.

is

is

21,

alterius

'Dnde

in

of

pp.

occasions.

part

51-4.

genere, 1016b32-3.

next

form

moderate

VIII,

praesenti

cum

adverb

notitia

quia I,

compositional-propositional

-

that

Ockham

vel

q.

significare

intuitive

21:

I,

a

term

recent

fieri 'Socrates OPh

6,

A

Loux

de

pp.

lines

the

of

incomplexa.' alicuius

the

p.

example, 205:

caused

the

lll,

the OTh

as sed

causa

be

other

terms

219-20,

OPh

here

intuitive quod

"significare"

cuilibet

3.

in I, the definition

identity

abstractive

OPh is

p.

saltern

vel

potest postea

est

3, judgement

same

of

few

1,67,

95).

for

in

OCKRAM

on and

est

1972,

potest

aliquid

an

in

mention,

uses

pp.

68,

1997,

1992a.

a

et

id

the maintain

Fodor's

way

VI,

dico

SL

(such

art.

sect.

(po

OTh I, any

latter

Quodl.

very 190.

1,33,

case

sed to

7.

of actus

readings

that for

...

and

7,

1,

futuro

1,

' And

quod

respectu

by SL

in

any

terminorum,

general;

p.

1969,

Metaphysics

OTh

'

SL Biard

accipitur

the

the

q.

defending

de mind

q.

materialis

numerical

causabatur,

signum

in

usually

Rochart

atomism 8,

in

in ...

Ockham

69:

in in

basic

heart

c:hap.

eager

potest',

idem

1984

dico

chap.

q.

that

the that

vel supposition

am

significare'

admits

p.

made

singularia, and

is

1972, Var.,

quo

presupposed,

...

the

fortiori,

characterized

the

...

I

can

propositionem

supponere'

Aliter

I,

in Prol.,

is

'

Andres

a

1969,

a

III,

followed

1,63,

is dicitur

in

circumstanciis

causabatur,

at and

would

i"ji·/I.

Aristotle,

I,

are

...

verb'

illo

pe

supra,

so,

70: be

'),

SL

OTh

it

what

Panaccio

See

point

illius

quo

specific

sine Ibid.:' Rep.

See Quaest.

supposition

material suggests

('suppositio

competere

which

This

both

Conceptual

be

...

to

Ockham

1, p.

Rochart

Ockham

causes

'suppositio'

Even

the

In

intelligat

terminorum

nullam

apprehensivus

incomp1exam Ord.

pro

illo

praeterito

'significare'

of modo

See

See

They

summarized

to

'Propositionalist'

in

Andres

JH.

37.

36.

34. 35.

33.

32.

31.

30.

29.

28.

27.

26.

25.

24.

23. 60

a

to

of

in

do

be

in

in

a.\'

to

of be

of

buC.

and

non­

has.

and

have

0lIl.

to

thus

time

point.

simple

whole

only

it

or

to

general

internal

many

'father'.

tried

his

detailed

defended

relational

terms

should

allows

are

admission him

no

a

objects

and

of

Ockham

by

his

philosophy

said

have

question,

all

simple

an

fine-grained

language,

concepti; impoulbll

have

ably

2

simple

by

in

had

for

intriguing

I

are

technical

legitimately

theory

'white',

non-connotative

allows

the

duality all

they

has

Ockham's of

such

such

As

(significatio) as

primary role

sciences

Western can

they

accepted

hierarchized

best basis

reason:

mental

for

particular,

of

literature,

a

The

what

One

Language

that

to

logically

relational

the

out1il1ed

such

Spade

in

language

the

discussion.

is

been

does

major

number

me

apparently

are

with

language;

as

non-relational

a

a

it

exactly

'absolute';

semantical

ones

this

language

to

terms

has

simple

Ockham

Ockham's

following

this

general categorematic

terms?

logic,

thought.

to

propositions.

precisely

mental

the

Vincent

framework

signification

no

this

signification,

in

from

the

so,

of

plays

is

raises

the

a mental

This

of

the

called

contributions

understanding

of Ockham,

Logicae

in endowed

are

following:

4 as

in

Paul

Russell, seemed

Mental

as

works.!

and

say

then,

for

stake categories

mental

devoted

simplifications. are

(this

for

are

the

connotative

the

6.1

terminist

If,

connotative

at

primary

seen

there

-

long

be

in

him

ideas terms.'3

connotative

for

1975,

terms

is

language

Logicae,

the

original

is

Summa

apparatus

of

ideally

that

are

and

in

-

forefront

hand,

exclusively

by

has

that

well-aware,

will

Bertrand their

the

explicitly

the

Chapter

it

others

form, to

concepts,

the

out

simple

mental

semiotical

of

mind

tools

signs

not was

in

nominalism

of

connotative used signification)

Summa

other

thought

to

ontology).

least

ontological

connotation.

But

technical

technical

absolute

Ockham's

4

lot

connotative

Concepts

are

at

Terms

published

concepts the

the no

sense,

does

and

study

the

a

of

the

turned

which

serve

the include

human

addition

simplest

come

absolute Certain

of

he

truth-conditions

on

in

and

would follow

it,

to Ockham

major

in

him,

and

following two

one

the

its

since

paper

simple

Ockham's

of it

secondary are

introduced

and

10,

scholars.

between

as

as

is

to

conceptual logical

thought. turned

logical

in

the

to

no

for

it

how

relational

(or

terms,

recently

developed

the

the

although

'time',

are and

terms,

approach.

of

relations

theory

about

(suppositio)

in

newly

structure:

all

fatal

to

surprisingly,

human

has

chapter

concepts

Ockham

cases

influential

signification

understands

that,

of

approach,

be questions

language,

according

problem,

distinction

well-known

with

inventively

The

an

signs

theory

'There

elsewhere,

one,

chapter

best

is

a

idea

he

In

The

The

This

connotation

Connotative

as

away connotative would

terms,

the

complex

that:

the

primary

categorematic

'movement',

a crucial

semantical

nominalistic

Part Ockham

This

somewhat

programme

transferred particular,

fruitful

into

that

show

supposition

nominalistic

mental analysis

systematically

logic. Being 64 OCKHAM ON CONCEPTS CONNOTATIVE TERMS IN MENTAL LANGUAGE 65

Ockham's nominalistic doctrine of thought and relations would therefore be doomed nominalism. Every sign, be it linguistic or mental, must, if it is to be truly significant, to failure.s refer the mind in some way to singular substances or singular qualities, and to This worry prompted me to re-examine carefully a large number of relevant nothing else (since nothing else exists). Butit cando so through connotation as well passages in Ockham, and this led me to the surprising conclusion, which I Will as through primary signification, and this duality accounts for much of the richness defend here, that Spade and all these other commentators whom I deeply respect are of language without jeopardizing ontological economy. Ockham, for example, just wrong on this: Ockham consciously thought that there are simple connotative avoids being committed to the ontological acceptance of relations by analysing concepts in mental language (section 2 below) and nothing he says commits him to relational terms as signifying nothing but individuals, some of which they primarily the contrary (sections 3 and 4). signify and some of which they connote. And quantities are disposed of in a similar fashion. Frege's puzzle about identity, which led him to introduce intensional entities such 1. Connotative terms as senses besides denotations, can also be solved in an Ockhamistic framework through connotation theory. How is it that expressions such as 'the Morning Star' Let us first recall some of Ockham's main theses about the distinction between and 'the Evening Star' are not synonymous even though they have the same absolute and connotative terms. It is presented by him as a division among denotation? Ockham's answer would be that while they have the same primary categorematic terms, which are those signs - such as 'man', 'white', 'mother', significate, they are connotative terms with different cannotata. And of course, we 'concept', etc. - which are normally used to refer (or purport to refer) to individual will have a similar solution to Quine's puzzle about 'renate' and 'cordate' which are things in the world or in the mind. 6 In a nominalistic vein, any such term is said to supposed to have the same extension while not being synonymous. Very simple, primarily signify each individual of which it is true. A general term like 'man', for Ockham would say! No need here for ontological eccentricities: 'renate' and instance, primarily signifies not a general entity such as human nature, nor an 'cordate' primarily signify the same animals, but the former connotes kidneys while abstract Platonistic idea or a Fregean sense, but only men themselves, those the latter connotes hearts and that is precisely why they are not synonymous. individuals, that is, of which it would be true to say 'this is a man'. Now, some of So it is not surprising to find out that Ockham himself calls upon his notion of the categoremata - but not all - have, in. addition to this primary signification, a connotation to solve a large array of theoretical problems, not only in logic, but in secondary sign(fication: they direct the mind not merely toward their primary and in as well. The particularities of theological significates, but also toward some other things. Take, for example, a relational term discourse, for example, are in large part explained by connotation. How is it, like 'father'. Its primary significates in the strictest sense are at any time those Ockham asks, that we have in theology different non-synonymous concepts all individuals of which it is at that time true to say 'this is a father'. But the term applicable to God (such as 'good', 'wise', 'creator', etc.) if He really is a perfectly 'father' in virtue of its very meaning also turns the mind toward those individuals simple being? His answer, of course, is that connotation allows in such a case for a that have a father, namely the children. These are its secondary significates, or, as multiplicity of concepts primarily signifying the same thing, but connoting different Ockham also says, its cannatata. Whenever a categorematic sign has secondary other beings;8 And in many occasions does William, with the help of connotation, significates, it is a connotative term. Otherwise, it is an absolute one.? subject the terminology of natural philosophy and of mathematics to reductive An absolute term, then, treats all of its signijicata on a par. 'Man' - which is, for analysis in view of showing that neither , nor .geometrical points or lines, Ockham, a paradigmatic case of absolute term - signifies men and nothing but men, nor movement, void, place or time have any reality distinct from that of singular and it does not signify any man differently than any other. Such terms roughly substances and qualities. 9 Ockham's nominalism, in short, relies heavily on correspond to what is nowadays called 'natural-kind terms'. They are relatively rare. connotation theory. Since only singular substances and singular qualities (singular whitenesses for Now, why would anybody think that such a useful and crucial device only instance) are admitted in Ockham's ontology, the only absolute terms there are are belongs to the superficial level of conventional languages and that it is utterly absent the natural-kind terms from the categories of substance (for example, 'man', 'horse', (at least as a property of simple terms) from the deep structures of mental language? 'animal', 'tree') and quality (for example, 'whiteness', 'colour'). All other I will, in section 3 and 4, scrutinize in some detail the arguments used by Spade and categorematic terms are connotative. This includes all relational terms such as others to sustain such an exclusion. But details apart, it must at this point be said that 'father', 'owner', and so on, all concrete accidental terms such as 'white' (which, their interpretation rests. for a large part on a particular thesis which is clearly and according to Ockham, primarily signifies the individual substances that are white, frequently asserted by William himself: a connotative term always has a nominal and secondarily signifies their singular whitenesses), all categorematic terms from definition (a definition quid nominis), while strictly speaking no absolute term the Aristotelian categories of quantity, action, passion, time, place, position and does. lo Since absolute terms uniformly signify whatever they signify, the only sort of habitus, as well as all negative or privative expressions like 'immaterial' or definition they are liable to receive is a quid rei definition, which is a description of 'blindness', and also such philosophically significant terms as 'true', 'good', the essential features of their significates but in no wayan explicitation of their 'intellect', 'will', and so on. I,('/"hal meaning. Connotative terms, on the contrary, both havc a primary and a One can easily see the extreme importance of connotation theory for Ockham's seclmdary signification. and this internal hierarchizcd structure, Ockham thinks, can 66 OCKHAM ON CONCEPTS CONNOTATIVE TERMS IN MENTAL LANGUAGE 67 always be rendered explicit with the help of a complex expression, which will known by us through a simple connotative and negative concept which is proper to precisely be the term's nominal definition. Him.' Such a concept, he adds, 'is simple,even though it signifies distinct things, Spade's idea is that ifthis is so, then for every connotative term, there exists what principally or secundarily, directly, that is, or obliquely'.16 Concepts, as we already he calls a fully expanded nominal definition which is strictly equivalent with that know, are the basic terms of mental discourse. So what is unambiguously term and in which the only occurring.non-complex categorematic terms are absolute presupposed here is that there are simple mental connotative terms and that their ones. II The point is that if a connotative name occurs in the nominal definition of internal hierarchized semantical structure does not preclude them from being simple. another connotative name, it can be replaced in this definition by its own nominal The notion of a simple term which is involved in this passage is quite clearly, it definition; the process can be repeated until only absolute terms are left, along with seems to me, the same as in Spade, namely: the old Aristotelian idea of a sign no part syncategoremata. In other words, connotative terms, since they are definable, are all of which is itself a sign. It is sufficient to prove it to remark that simple concepts are, logically dispensable. And since, according to Ockham himself, there is no all along the surrounding pages, contrasted with composite concepts, which are synonymy in mental language, all connotative terms, Spade concludes, should be themselves characterized as having different parts, each of which is independently represented in mentalese by their fully expanded nominal definitions. The problem 'abstractible' from things, which comes pretty much to the same as saying that each here, I am afraid, is that this conclusion runs head on against Ockham's own of these parts has an independent signification. I? considered beliefs about connotative terms. This is what I will now show in section In the following question of the same book, we find another very clear assertion: 2 by going through some relevant texts of Ockham. And I will return in section 3 to 'I say', Ockham writes, 'that of the same thing there can be a plurality of different Spade's synonymy argument in order to find out what exactly is wrong with it. simple denominative concepts, and this is due to the diversity of their connotata.'18 Denominative terms are paronyms, terms such as 'white', 'just', 'courageous' which are used to attribute to individual things certain properties for which there exist 2. Mental connotation corresponding abstract names such as 'whiteness', '' or 'courage.'19As is clear from the above quotation(as well as from many other passages), denominative terms Let us first tum to Quodlibeta V, question 25, where the problem under discussion is are all seen by Ockham as connotative. So once more the conclusion is unavoidable: precisely the following: is there in the mind a real distinction between absolute, there are simple connotative terms in the mind. connotative, and relative concepts? Ockham's answer is an unequivocal yes: One can still resist, nevertheless, and remind us that these passages from the Ordinatio were written early in Ockham's career and that he could have changed his I reply that, according to philosophers, the [affirmative answer to this question] is certain. mind on this as he did on other subjects. This is true, of course: he could have. But For the concept human being is absolute, the concept white is connotative, and the concept as far as I can see, he didn't. Let us look at Summa Logical' I. Here the evidence is father is relative. And [the three types of concepts] do not overlap except as superior and admittedly a little more indirect, but still, it seems to me, overwhelming. On the one inferior, since every relative concept is connotative, but not vice versa. 12 hand, Ockham announces at the beginning of chapter 3 that he is now about to introduce a number of distinctions among 'incomplex terms';20 this is certainly So, there are connotative terms in qlentallanguage. Of course, Spade's thesis does meant to include the absolute/connotative distinction drawn in chapter 10. On the not directly contradict that, since in its most careful formulation, it is only that there other hand, we find, at the beginning of chapter 11, the following : 'All are no simple connotative terms in mentalese. But if this is what Ockham thought, divisions we have considered so far apply both to terms which naturally signify and the least that can be said is that he missed a good chance of saying so! His examples to terms which are merely conventional signs.'21 This impliesthat the distinction of here are all simple expressions such as 'man', 'white', and 'father'. If he thought that the immediately preceding chapter is among those which pertain to mental as well such written expressions really corresponded to mental complexes containing only as to conventional language. That some mental terms are both simple and absolute terms and syncategoremata, his way of speaking here has to be seen as very connotative neatly follows from these two considerations taken jointly. misleading. A similar argument can also be derived from the way Ockham understands the Moreover, if we should both accept Ockham's explicit thesis in this passage (that Aristotelian division of the ten categories. This, he says, is a classification of there really are .connotative terms in mentalese) and Spade's thesis (that no incomplex terms, mental as well as verbaP2 And it is well-known, on the other hand, connotative mental concept is simple), there would follow, as Spade clearly saw that all terms from the categories of relation and quantity, as well as many terms 13 himself, that mental connotative terms are identical with their nominal definitions. from the other categories, are, according to him, connotative.23 And this directly contradicts something Ockham frequently repeats: 'a definition is A doubt might be raised, at this point, about what exactly Ockham means in these not identical with what is defined' .14 His of the point is precisely that the passages from the Summa by 'incomplexum'. Couldn't it be, after all, something definition is a longer expression than the defined term. That this is not an accidental different than what was meant by 'simplex' in the previously quoted passages from feature follows from something Ockham considers as a commonplace about the Ordinatio'! In chapter 4 of his commentary on Aristotle's Categories, Ockham definition, namely that it is more explicit than what is defined. I> had distinguished two senses of 'incomplexum': 'In the strict sense, an incomplex is Other passages are even more direct. Speaking about the knowability of God in a simple word. a word. that is. to which no other word is added. such as: man. runs. Book One of his commentary on the S/'nt/'1/('/'s. Ockham writes that 'God can he 68 OCKHAM ON CONCEPTS CONNOTATIVE TERMS IN MENTAL LANGUAGE 69 lion, goat. And by contrast, a complex is any expression composed of different and the complete definition of 'double' has to include the correlative term 'half'.28 words.'24 In this sense, then, an incomplex sign is precisely a sign of which no part Whether these examples, which are from Ockham himself, are well-chosen or not, is itself a sign. If Ockham is using this strict sense in the passages we have looked the important point here is that some relational terms (if not all) cannot be defined at, then the conclusion we have drawn from them does contradictSpade's thesis. But without recourse to other relational terms; therefore, since all relational terms are there is for Ockhamanother meaning of 'incomplexum': connotative, Ockham is directly committed to denying the thesis Spade attributes to him that all connotative terms are liable to fully expanded nominal definitions, 'each In a second sense, an incomplex is the term of a proposition, whether it is a single word or of whose constituent non-complex categorematic terms is abso~ute'.29The resulting composed of different words; 'white man', for example, is an incomplex in this sense, and circularity in nominal definitions (since, for example, 'child' must occur in the so is 'white musician' and all similar expressions. And by contrast a complex is an definition of 'father', and 'father' in the definition of 'child') is not a problem for expression composed of a name and a verb. 25 Ockham: if two concepts are really correlative to each other, they have to be acquired simultaneously and there is no difficulty in admitting that they mutuallydefine each In other words, an incomplex term in this sense is whatever can be the subject, the other - nominal definitions do not necessarily proceed from the better known to the predicate, or the copula of a proposition. In this sense a nominal definition can itself lesser known. 30 Ockham, then, just like Bertrand Russell some six centuries later, be said to bean incomplex term and holding,. therefore, that there are incomplex explicitly denied that relational terms are all logically constructible from absolute connotative terms in mental language in this large sense would not preclude them to ones. They are not dispensable in mental language as he understands i\)' be identified with their fUlly expanded definitions, just as Spade wants it. So, what we have now is that Spade's interpretation contradicts at least three of Assuming that no other sense of 'incomplexum' is relevant here, the question we Ockham's explicit and repeated statements: now have to face is this: which one of these two senses is Ockham using in the passages I have quoted? An important clue is provided by Ockham's distinction in • Definitions are always distinct from the terms they define. Summa Logical' I, chapter 2, of different senses of the word 'terminus'. In the first • Some connotative concepts are simple. of these, a terminus is any expression that can be the copula, the subject, or the • Relative concepts cannot all be completely defined with the sole help of predicate of a proposition. Please note that this corresponds exactly to the second absolute terms and syncategoremata. sense of 'incomplexum' in the commentary on the Categories. But 'terminus' also has another meaning according to which, Ockham says, 'all incomplexes are called But this is still not enough to settle the matter. Spade's point is that Ockham is terms'.26 My point here is that the word 'incorrtplexum' in this last sentence can only committed to the denial of each one of these three theses. And of course, he could be be used in its first sense (that iS,a sign no part of which is a sign), since otherwise without realizing it. His theory of connotation, after all, could be inconsistent! What there would be no between the first and the second meanings of is left for us to scrutinize, then, is the arguments which led such renowned 'terminus'.27 If this is so, then we have a strong presumption that Ockham keeps commentators as Spade, Adams and Normore to believe that William is so using 'incomplexum' in the same sense in the other passages we have referred to committed. from Summa Logical' T, especially in the one from chapter 3, which occurs only a page down. To sustain the rival interpretation, one would have to suppose not only that 3. Synonymy.and nominal definitions Ockham changed his mind between the Ordinatio and the Summa about connotative concepts - and without ever saying so ...,-but also that he changes the meaning of the The crucial argument we have to discuss can be summarized in the following way: word 'incomplexum' from chapter 2 to chapter 3 of Summa Logical'. Such a daring interpretative hypothesis, I would say, should only be admitted in the face of strong (1) each connotative term has a nominal definition; evidence, which, as we shall see in the next sections, we do not have in this case. (2) a term which has a nominal definition is synonymous with it; But before we get to that, I would like to show that not only did Ockham (3) there is no synonymy in mental language. consciously think that some simple concepts were connotative (as I hope to have Therefore: established by now), but also that this is not, in his system, an isolated thesis which (4) connotative terms and their definitions cannot exist as distinct units in mental he could simply have dropped without harm: the thesis, I maintain, is directly language. implied by at least one other important philosophical doctrine which Ockham explicitly holds. Premiss (I), as we have seen in the first part of this paper, is very clear in Ockham, This doctrine has to do with the nominal definitions of correlative terms such as it is often repeated and nowhere denied. Premiss (3) is also quite explicit: 'For there 'father' and 'child', or 'double' and 'half'. None of these, Ockham says, can be is no multitude of concepts corresponding to the multitude of synonymous [spokenl completely defined without the help of its correlative. Thus a complete nominal names.' I! Mental language, according to Ockham, is maximally economic: if a definition of 'father' would have to be something like 'male animal having 1I child', distinction between words has no semantical relevance, then there normally is no 70 OCKHAM ON CONCEPTS CONNOTATIVE TERMS IN MENTAL LANGUAGE 71 corresponding distinction in mentalese. What is wrong with the argument is premiss correctly added to the other ... as if it was totally certain that such a principle was true, (2): as we shall see, it is not the case in Ockham that a connotative term is exactly since in fact it is· not true although it seems to be ...44 synonymous with its definition. It is true that Ockham says in different passages that a connotative term and its Where exactly do the semantical differences lie between the definiendum and the nominal definition 'signify the same'.33 But, as Spade readily admits, it is not definition? Ockham does not say. But maybe we can find it by ourselves. To do that, sufficient, in Ockham's vocabulary, that two terms signify the same individuals in we first have to take a look at which modes of signification there are according to order for them to be synonymous. 'Parent' and 'child' for example both signify Ockham. Many passages suggest a basic' distinction between grammatical and exactly the same individuals: 'parent' primarily signifies the parents and secondarily logical modes.45The grammatical modes of signification of an expression are wholly the children, while 'child' primarily signifies the children and secondarily the determined by its grammatical features: number, case, gender, and so on.46 The parents. But, of course, they are not synonymous for the obvious reason that they do logical modes - which are certainly more relevant for synonymy - must be those not signify the same individuals under the same modes. 34 In Ockham's sense of which are listed in Summa Logical' 1II-4, chapter 10. According to this passage, a 'synonyma', two terms are synonymous if and only if whatever is signified by one term can signify: of them under a certain mode is also signified by the other one under the same mode.35 • in recto or in obliquo (that is: primarily or secondarily); Spade clearly realizes this, but he thinks that for Ockham, a connotative term and • affirmatively or negatively; its definition always signify exactly the same things under the same modes, • categorematically or syncategorematically; although, as he acknowledges, Ockham never explicitly says so. Spade reasons as • as a proper name or as a common name.47 follows. A term like 'blind' is supposed, according to Summa Logical' I, chapter 33, to signify sight negatively. 'And this', Spade remarks, 'seems to be exactly the way It is to be noted that these logical distinctions are independent of the grammatical it is signified by its nominal definition, which would presumably be something like features of the term under consideration: 'father', for example, signifies the children "animal not possessing sight".' 36 If this is so, 'there seems', he concludes, 'to be no obliquely even when it is not taken in one of the oblique cases (genitive, dative, reason not to generalize this to all connotative terms'.37But, on the contrary, we now ablative and accusative), and 'blind' signifies sight negatively even if its external know that there is a very good reason to avoid such a generalization: Ockham's grammatical form does not display any negation. doctrine of connotation would be rendered inconsistent by it! This being clarified, we can readily notice that there are in fact some obvious Moreover, there are a number of passages which strongly suggest that Ockham differences between the logical modes of signification of a connotative term taken consciously refused the general principle of a total synonymy between the definition alone and those which are found in its nominal definition. First, there are cases and the defined terms.38I have quoted above a passage where he says that definitions where the definition itself includes a connotative term. We have seen that relational are more explicit than what they define,39 and this probably expresses his most basic terms are not ultimately dispensable. A term like 'father' is defined as 'male animal intuitions about the matter. But there is much more. In Quodlibeta V, question 19, having a child'. In such a definition, the term 'child' is itself connotative. What does which is wholly about the distinction between quid rei and quid nominis definitions, it secondarily signify? The parents, of course. So the fathers are connoted (or he holds that verbs, adverbs and conjunctions are all liable to nominal definitions secondarily signified) in the definition of 'father', but they are certainly not connoted and his examples are the following: ' ... someone who wants to define "where" will by the term 'father' itself, of which they are the primary significates! say that it is an interrogative adverb of place; likewise, he will saythat "when" is an Attention has been drawn to another sort of case by John Boler in a very interrogative adverb of time'.40 The definitions in these cases are metalinguistic interesting paper about 'Connotative terms in Ockham'.48 Let us consider the while the defined terms are not, and consequently they cannot be synonymous with definition of 'father' again. It includes the expression 'animal' which occurs there in them. Even if they are not about connotative terms, these examples show at least that the nominative case and which, of course, signifies in recto all animals. Should we Ockham did not subscribe to a general principle .of total synonymy between say, then, that 'father' itself signifies in recto all animals? Certainly not, because that definitions and defined terms.41 would amount to saying that it primarily signifies all animals, which it does not, The most explicit passage I know of is in Book One, chapter 20, of Ockham's since it is not true of each animal that it isa father. commentary on Aristotle's . Here he insists that a nominal We have found so far two major differences between the modes under which definition cannot always be substituted for its definiendum without harm.42 The certain things are signified in a nominal definition and the modes under which the definition of 'pug', for example, is 'concave nose', but if we substitute this definition defined term signifies them (if it does signify them at all): some things are connoted for 'pug' in 'pug nose', we get the unacceptable result 'concave nose nose'.43 More by certain parts of the definition which are not connoted by the defined term; and generally, Ockham warns the reader against uncritically using in the idea some things are signified in recto by certain parts of the definition which are not signified in recto by the defined term. ... that a name and its definition signily exactly the same thing and that the one can he Haven't we gone astray, however, in taking into consideration the modes under suhstituted to the other and that whatever can he correctly added to lhe one can also he which certain things are signified by the parIs of the definition rather than the modes 72 OCKHAM ON CONCEPTS CONNOTATIVE TERMS IN MENTAL LANGUAGE 73 under which they are signified by the definition itself taken as a whole? After all, course, he sometimes asserts, as we saw, that a definition and its dejiniendum Spade's thesis is that a connotative term is synonymous with its nominal definition as 'signify the same'. But a close look at the texts reveals that he has a tendency to a whole. But the trouble with the idea of considering the global signification of the make a distinction between 'signifying the same' and 'signifying exactly the same'. definition is that we do not find in Ockham any clear indication about how the A good example of this is found in the Quaestiones Variae, where, at one point, he signification of such complexes as definitions should be constructed out of the writes that "'man" and "animal" do signify the same, although not exactly, since significations of their constituent parts. Spade proposes with regard to this a principle "animal" signifies whatever "man" signifies, and more ..;'54And remember that in which he calls the Additive Principle: 'A complex expression signifies just exactly the his commentary on the Sophistical Refutations, he clearly warns the reader against sum total of what is signified by its constituent non-complex categorematic terms.'49 taking as true the principle that a name and its definition signify exactly ('omnino' Theleast that can be said is that such a principle - which, as Spade remarks, can in this case) the same thing.55So it might very well be that what he means when he be found in Buridan50- is far from being explicit in Ockham himself, who rarely says that the definition and the definiendum signify the same (without the speaks of the global signification of such complex wholes as nominal definitions. qualification: exactly) is only that whatever is signified by the definiendum is also But even if it could plausibly be attributed to Ockham,it still would not teach us signified under some mode in the definition (while the converse might not anything about the modes of signification of the definition as a whole, since it has to necessarily hold). do only with the total list of significates, and not with the modes under which each Be that as it may, we can at least conclude from this discussion that, as far as we of these significatesis signified. Are the modes of signification of a complex know, nothing in Ockham's writings requires that there be a total synonymy between expression different from those of its constituent parts (as the objection we are now a connotative term and its definition. Many considerations, on the contrary, point in considering requires)? And if so, what are the rules which lead from the modes of the other direction. Since, in particular, the admission of the principle leads to the parts to the modes of the whole? Ockham is totally silent about these questions, inconsistency, it should, it seems to me, definitely be discarded. As to what role and we should. avoid attributing to him some theses on this that would render his exactly the doctrine of nominal definitions plays in Ockham's scheme, we will come theory inconsistent! The best attitude here is to limit ourselves to considering the back to this in detail in chapter 5 below. modes of signification which are involved in the definition withouttrying to attribute some other modes to the definition as a whole.51 On the other hand, we do have some clear hints in Ockham about certain rules 4. Connotative terms and exponible propositions which could lead from what is going on in the nominal definitions to conclusions about the modes of signification of the defined connotative terms. The following two What we have said so far also disposes of a somewhat different argument which rules, for example, are plausible candidates: might be tentatively designed in order to exclude connotative terms from mental language. Although I have never seen this argument in print, it is nevertheless clearly (Rl) A connotative term T negatively signifies an object 0 if and only if 0 is one suggested here and there. It is based on Ockham's thesis that any proposition in of the primary significates of a term T' which occurs within the scope of a which there occurs a connotative term is exponible.56 To say that a proposition is negation in the nominal definition of T.52 exponible - or that it has exponentes - is to say that it is equivalent in truth­ conditions to a conjunction. In order for a proposition such as 'something white is (R2) A connotative term T signifies in obliquo an object 0 if 0 is one of the running' to be true, for example, it is necessary that the following be true: something primary significates of a term T' which occurs in one of the oblique cases in is running and in this thing there inheres a whiteness. Each part of such a conjunction the nominal definition of T.53 is called an exponens of the original proposition. The argument, then, would run as follows: Such rules account for the that a term like 'blind' is said to signify sight negatively (its definition being something like 'animal not having sight') or that a (5) any simple proposition in which there occurs a connotative term has at least term like 'father' is said to signify the children obliquely (its definition being 'male two exponentes; animal having a child', where 'child' is in the accusative case). And other such rules (6) in the final analysis, the exponentes should not themselves include any could be constructed (in order, especially, to yield the primary signification of the connotative term; defined connotative terms). But even taken jointly, they do not (and should not) (7) the conjunction of the exponentes is synonymous with the exponible imply the complete synonymy between the definitions and the dejinienda. We have proposition; seen, on the contrary, that certain semantical features of (parts of) the definition have (X) there is no synonymy in mental language; no equivalent in the defined term. therefore: It might even be admitted that in some cases, certain things are signified by some (9) an exponible proposition and the conjunction of its cxponcntc.l' cannot both parts of a definition which are not, under any mode, signitied by the dctinicndum. exist as distinct units in mental language. Nowhere, for example, does Ockham say that 'father' signif1es in any way all the animals there are, although the general term 'animal' occurs in its def1nition. Of 74 OCKHAM ON CONCEPTS CONNOTATIVE TERMS IN MENTAL LANGUAGE 75

What has to be rejected here is premiss (6) and premiss (7). Premiss (6) has pieces of evidence and purports to show that none of them overwhelmingly forces us sometimes been asserted by very good commentators,57 but our previous discussion to accept simple connotative terms in the mental language of Ockham's later theory. now shows it to be unacceptable to Ockham. At least one of the exponentes in his Gaskin concedes that the point has been well-established 'as far as the early theory is built out from the nominal definition of the original connotative term: Ockham is concerned',62but he thinks William changed his mind on this between the 'white', for example, is defined as 'something in which there inheres a whiteness', writing of the Ordinatio and that of the Summa Logicae and the Quodlibetal and that is why the proposition 'in (this) thing there inheres a whiteness' has to be Questions. Gaskin concedes, moreover, that 'Ockham nowhere signals a change of counted among the exponentes of 'something white is running'. But we now know mind' on the matter, but he thinks that 'if challenged on the issue, [Ockham] would that in certain cases, connotative terms cannot be eliminated from complete nominal agree that his mature position was that there are, in fact, no simple connotative terms definitions. There is no reason, therefore, to think that they can be eliminated from at the mental level' .63 the exponentes. The question, of course, is: why would that be? Not only is it the case that If that is true, then premiss (7) also has to fall. As we have seen, some things are Ockham never says anything of the sort, but as I mentioned above, there are a sometimes signified in a definition which are not signified under the same modes by number of passages from the Quodlibetal Questions and the Summa Lqgicae which the definiendum. If such a definition (or something very close to it) occurs in the strongly testify to the contrary. They correspond to arguments (1) to (4) in Gaskin's exponentes, then some things will be signified in the exponentes which are not enumeration: signified under the same modes in the original exponible proposition and hence they will not be synonymous. (l) Quodlibetal Questions V, question 25, where Ockham argues that there is a real It is interesting to notice that this point has been explicitly discussed by some of difference in the mind between absolute and connotative concepts;64 Ockham's successors from the end of the fourteenth century (Peter of Ailly, for (2) Quodlibetal Questions V, question 15, which explicitly posits that a definition example) up to the sixteenth. The question was raised as to whether the exponible differs from the defined term (contrary to the identification required by Spade propositions were really distinct in mental language from the conjunctions of their between Ockham's connotative concepts and their mental definitions);65 exponentes. And the usual answer, as we learn from Jennifer Ashworth's study of (3) Summa Logicae I, chapters 3 and 11, from which it follows, when taken jointly, this discussion, was an unequivocal 'yes'.58 Many arguments were put forward in that the distinction between absolute and connotative terms drawn in Summa favour of this conclusion, for example, that in some cases people clearly understand Logicae I, 10 obtains among simple concepts;66 an exponible proposition without knowing what its exponentes are.59In particular, an (4) Summa Logicae I, chapters 10 and 41, where the Aristotelian ten categories are illuminating distinction was drawn, which, I think, was already implicit in Ockham, presented as a division among simple terms, whether mental or spoken: since 'between two kinds of equivalence, equivalence in significando and equivalence in most of the categories contain only connotative terms, according to Ockham, it inferendo, or in truth and falsity' .60 It is certainly true that Ockham views the follows, once more, that he is straightforwardly committed to the acceptance of conjunction of the exponentes as equivalent to the original exponible proposition. simple connotative concepts.67 But he says so only in relation with truth-conditions. Since, as we have seen, the synonymy in such cases cannot always be total, a distinction is strongly suggested To all of these, Gaskin replies in basically the same way. He agrees that Ockham here between synonymy (equivalence in significando, according to the later expresses himself in these passages 'as if there were simple connotative terms '.68But terminology) and equivalence in truth-conditions. it can all be seen, he contends, as mere far,/ons de parler. In none of these caseS, In short, the argument from exponibles and exponentes fails for the very same according to Gaskin, is the question of whether there are or not simple connotative reason as the Spade-Adams-Normore argument: connotative terms cannot always be concepts really central to Ockham's ongoing purpose in the context. Ockham, seen as strictly synonymous with their nominal definitions. The principle according therefore, might be using convenient 'shorthands' there, having no need, in these to which there is no synonymy in mentalese simply does not apply in such cases. The precise contexts, for nuances and complications about the simplicity. or complexity conclusion of section 2, then, has to be confidently maintained: there are simple of concepts. connotative terms in Ockham's mental language. His nominalistic theory of This strategy calls for several remarks. For one thing, the implication that there connotation, so crucial for his whole philosophical programme, in nO way requires are indeed simple connotative concepts in the mind is not that immaterial in all the the (impossible) elimination of relational terms. Their indispensability, on the passages invoked. When Ockham neatly asserts in Summa Logicae I, 41 that 'the contrary, is explicitly recognized and consistently taken into account. terms subsumed under the categories are simple terms' and that 'this holds for simple terms of both the mental and the vocal sort',69he is well aware that he is thus taking a stand on one of the most controversial issues in the history of philosophy, and one, Appendix: A reply to Richard Gaskin moreover, that crucially matters for his own thought: what are the ten Aristotelian categories, categories of? The quoted sentences from Summa LOf.:icae I, 41 provide Richard Gaskin has recently proposed a detailed critical discussion of the case for nothing less than Ockham's considered answer to this age-old riddle; it is difficult to simple connotative concepts as I have just presented it.'>! He hreaks it down into six think that he doesn't express himself with great care in this context. The very way in 76 OCKHAM ON CONCEPTS CONNOTATIVE TERMS IN MENTAL LANGUAGE 77 which he chose to formulate his general position about the status of the ten The situation at this point is the following. We all agree that: categories strongly suggests that there are simple mental concepts from all ten categories within the human mind. This suggestion, therefore, should be taken prima (A) Ockham accepted simple connotative concepts in his earlier theory. facie with utmost seriousness. If Ockham had thought that all simple terms at. the (B) he never signalled afterwards a change of mind on the question. mental level belong to the categories of substance and quality - as Gaskin wants him (C) he kept expressing himself 'in many passages' of his later works as iihe still to - 70he would normally have seized the occasion of this general discussion 'on the admitted simple connotative concepts. distinction of the Categories' to say so. Secondly, in some cases, Gaskin's eagerness to make the quoted passages Given this basis, a strong reason is needed if we are to accept Gaskin's supposition compatible with Ockham's supposed rejection of simple connotative concepts leads that Ockham did in fact change his mind on simple connotative concepts and finally to devices which are quite implausible, interpretatively speaking, such as the excluded them.78 But the only reason we are given for this crucially rests on the (distinctively non-Ockhamistic) postulation of an 'infinitely expanded nominal attribution to Ockham of another thesis (the synonymy of the nominal definition definition' in the discussion of argument 2,71 and a convoluted reading of the first with the term it defines) that he not only never asserted, but that he, in fact, explicitly sentence of SummaLogicae I, 11 in the discussion of argument 3.72 rejected. Not a very good prospect, it seems to me.79 Thirdly (and the most important aspect), Gaskin's strategy on the issue makes The only way out for Gaskin's interpretation would now be to hold that even sense only if there is strong counter-evidence available to neutralize the prima facie though Ockham inadvertently rejected in so many words the synonymy-thesis, he is literal reading of the quoted passages. This, indeed, is the very basis of Gaskin's nevertheless strongly committed to it for yet other reasons, that have to do with what hermeneutical method as he precisely explains it itself: nominal definitions are supposed to accomplish in the framework of his theory of thought and language. This will be the object of the next chapter. The general principle here ... is that if, in a given passage, Ockham writes as though there were simple connotative terms at the mental level, but if that simplicity plays no role in the argumentation of the passage - if his purpose would have been equally well served had Notes Ockham substituted the nominal definitions of the simple terms for those terms themselves - then, given that there is pressure on Ockham from elsewhere in his philosophy to deny 1. Panaccio 1999a. the existence of simple connotative terms at the mental level, we cannot be sure that in 2. The con;;of the present chapter follows Panaccio 1990a, with a few minor amendments. writing as though there were simple mental connotative terms Ockham does not intend that Some of the questions and objections this paper raised in the literature are addressed in as mere shorthand which could and would be eliminated were one to insist on strict the footnotes and the Appendix, while others will be discussed in chapters 5 and 6. accuracy.73 3. Spade 1975, p. 67. See also, along the same lines, Spade 1980. 4. Two salient examples are Adams 1987, pp. 319-27, and Normore 1990. On the other Gaskin's reinterpretations of the passages referred to in arguments (1)-(4) overtly hand, several scholars, in recent years, have expressed their general agreement with the depend for their plausibility on the assumption 'that there is pressure on Ockham interpretation I am arguing for here; see in particular Tweedale 1992, Goddu 1993, and from elsewhere in his philosophy to deny the existence of simple connotative terms Beretta 1999, pp. 155-77. at the mental level '. 5. Spade did in fact reach the conclusion that Ockham's reductionist programme is doomed But what pressure? That's the question. Gaskin's answer is that the rejection of to failure, and for closely related reasons. See on this Spade 1990 and 1998. A similar simple connotative concepts neatly follows from Ockham's rejection of synonymy point is made by Mertz 1997, who also thinks that the ontological rejection of relations within mental language along with the assumption that a nominal definition is as distinct entities goes hand in hand in Ockham's thought with the logical elimination of all polyadic predicates (for example, pp. 64-7). always synonymous with the defined term. Which, as the reader will recall, was 6. Syncategoremata, by contrast, do not by themselves denote any object neither in the exactly Spade's original argument. My point against it was that the synonymy of the world nor in the mind. They can only be used within complex expressions to modify the definition with the defined term is simply not to be admitted in Ockham's theory, and syntactic and semantic functions of the categoremata. See SL I, 4, OPh I, pp. 35-8. that he did in fact explicitly reject it.74 Gaskin, obviously, is not convinced by this 7. The locus classicus for the explication of the distinction is SL I, 10, OPh I, pp. 35-8. refutation, which he discusses independently in a long footnote of the same paper.75 8. Ord. I, dist. 3, q. 3, OTh II, p.425. . A lot is now seen to depend on which of us is right about this: whether or not 9. See for example, Exp. in Praedic. 10, OPh II, pp. 203-30, Tract. de Corp. Chr. 32, OTh nominal definitions and the defined terms were thought of as synonymous by X, pp. 175-84, and SL I, 44--c6,OPh I, pp. 132--49. Goddu 1984 is particularly explicit Ockham turns outto be central to the whole discussion about connotative concepts. about the central role of connotation in Ockham's analysis of the language of physics. For my part, I must say, I still take the textual dossier I provided above on the On its use in mathematics, see also Goddu 1993. matter to be decisive. Gaskin's dismissal of its impact is based partly on points made 10. See SL I, 10, OPh I, p. 35: 'Immo, proprie loquendotalia nomina [= absoluta] non habent definitionem exprimentem quid nominis ...;' and p. 36: 'Et tale nomen [= by Paul Vincent Spade and CyrilleMichon, to which I have, I think, sufficiently connotalivum I proprie habet definitionem exprimentem quid nominis ... ' replied;7o and partly on an original suggestion of his own for constructing complex II. See Spade 1975. pp. 66-71. signitications out of simple ones, which is not found in Ockham.77 78 OCKHAM ON CONCEPTS CONNOTATIVE TERMS IN MENTAL LANGUAGE 79

12. Quodl. V, q. 25, OTh IX, p. 583: 'Respondeo: conclusio est certa secundum philosophos, qualitate sunt nomina connotativa' (italics are mine). It should be noted that Ockham is nam conceptus hominis est absolutus, conceptus albi est connotativus, et conceptus using 'nomina' here in a wide sense since he acknowledges elsewhere that the categories patris est relativus. Et non coincidunt nisi sicut superius et inferius, quia omnis of action and passion contain only verbs, and no nouns in the strict sense (SL 1,56-7, conceptus relativus est connotativus, et non econverso' (trans!. Freddoso 1991, p. 486). OPh 1, pp. 183--4 and 187), and that the categories of time and place contain only 13. See Spade 1975, p. 68: ' ... then all connotative mental terms, if indeed there are any, are adverbs (SL I, 59-60, OPh I, pp. 188-91). Martin Tweedale has surmised that the complex terms and are their own nominal definitions' (italics are Spade's); see also his Ockhamistic distinction between absol.ute and connotative terms. might not pertain to note 41 on the same page: '[connotative mental terms] would be "synonymous" with verbs, adverbs and prepositional phrases at all (Tweedale 1992, pp. 438--41). This, their nominal definitions only in the degenerate sense of being identical with them'. however, neglects the wide use of 'nomina' that Ockham frequently allows himself: after 14. Quodl. V, q. 15, OTh IX, p. 541: ' ... definitio non est eadem cum definito, quia definitio all, verbs, according to Aristotle, 'are really names and have a determinate meaning' est sermo longus mentalis, vocalis vel scriptus, et per consequens non est eadem realiter (perihermeneias 3, 16bI9-20), and this must apply to adverbs as well, and to cum re extra nec cum uno termino definito' (italics are mine; trans!. Freddoso 1991, p. prepositional phrases such as 'on a horse', 'rapidly', etc. Ockham is entirely explicit that 453). As Cyrille Michon has rightly remarked (1994, pp. 374-5), this particular sentence connotation is at work even in the categories such as those of action, passion, time and occurs in the context of a discussion about real - rather than nominal- definitions. The space, which contain no nouns in the strict sense; see the passage from SL I, 10 quoted point it makes, however, is certainly meant to hold for both. Ockham repeats it almost above in the present note; see also Exp. in Praedic. 12, OPh II, pp. 245-6: ' ... relatio verbally in SL III-3, 22 (see OPh I, p. 680: 'definitio non est idem realiter cum non importat aliquam rem quae non sit de genere substantiae vel qualitatis, et tamen de definito'), a passage which, contrary to what Michon claims (1994, p. 375, n. 1), nulla re praedicatur in quid ... et hoc quia connotat aliquid praeter ipsum de quo explicitly pertains to 'all definitions', the distinction between real and nominal praedicatur. Et sicut est de quantitate et relatione, ita est de omnibus aliis sex generibus' definitions being introduced in the following chapter only. In SL 1,26, Ockham insists (italics are mine). His analysis of the verb 'to heat' (calefacere), for example - a verb that a nominal definition - just like a real one - is always a complex phrase ('oratio') which is assigned to the category of action - clearly comes down to treating it as a explicating a single term (see OPh I, p. 88: 'Definitioautem exprimens quid nominis est connotative (SL 1,57, OPh I, p..184). oratio explicite declarans quid per unam dictionem importatur'). 24. Exp. in Praedic. 4, OPh II, p. 148: 'Stricte dicitur incomplexum simplex dictio, hoc est 15. See Summ. Phil. Nat. 1,3, OPh VI, p. 162: 'communiter dicitur quod definitio significat una dictio sine additione alterius dictionis, sicut hie: homo, currit, leo, capra. Et per illud idem explicite quod definitum significat implicite'. The sentence that follows oppositum dicitur complexum omne compositum ex diversis dictionibus.' makes it clear that Ockham is speaking here of nominal definitions. In Panaccio 1990a, 25. Ibid.: 'Secundo modo dicitur incomplexum terminus propositionis sive sit una dictio I had also quoted a passage from SL III-2, 14, OPh I, p. 530: 'definitio explicat plures sive plures dictiones, sicut "homo albus" dicitur incomplexum et similiter "album res quam definitum'; but, as was remarked by Spade 1996, p. 232, this passage has to musicum" et huiusmodi. Et per oppositum dicitur complexum compositum ex nomine et do with real rather than nominal definitions, while our present discussion is about verbo ...' nominal definitions only. 26. SL I, 2, OPh I, p. 9: ' ... et sic omne incomplexum vocatur terminus'. 16. Ord. I, dist. 3, q. 2, OTh II, p. 405: ' ... dico quod Deus potest cognosci a nobis in 27. Richard Gaskin claims that my reading here rests on a misinterpretation of Ockham's conceptu simplici connotativo et negativo sibi proprio ... et iste conceptus est simplex, second sense for 'terminus' in SL I, 2 (Gaskin 2001, p. 258). His point is that 'terminus' quamvis distincta significet, sive principaliter sive secundario, hoc est, vel in recto vel in this sense is explicitly contrasted by Ockham with 'oratio' (which Gaskin interprets in obliquo ... ' Ockham seems to have considered, at this time, these simple connotative as meaning 'sentence' in this context) and that, consequently, a term in the second sense concepts as conventional; see ibid., p. 403: ' ... dico quod [Deus] est cognoscibilis a of 'terminus' could very well be complex (like 'white man' or 'the man I saw yesterday') nobis in conceptu aliquo modo simplici ad placitum instituto ad significandum, et hoc as long as it is not itself a complete sentence. This is wrong-headed, however. 'Oratio', in conceptu connotativo et negativo sibi proprio'. More on this in chapter 6. here, is not to be taken as meaning 'sentence' as Gaskin presupposes, but 'complex 17. See, for instance, Ord. I, dist. 3, q. 2, OTh II, pp. 402-3: ' ...dieo quod essentia divina phrase'. Ockham is explicit about this distinction between these two usual senses of vel quidditas divina potest cognosci a nobis in aliquo conceptu sibi proprio, composito 'oratio'; see Exp. in Perih. 3, OPh II, p. 390; ' ... oratio dupliciter accipitur: Uno modo tamen, et hoc in conceptu cuius partes sunt abstrahibiles natwaliter a rebus' (italics are large, et sic omnis congeries dictionum est oratio ... Isto modo hoc quod dico "homo mine). See also p. 404, where Ockham is even more explicit about his notion of a albus," et similiter "album animal" et huiusmodi sunt orationes. Aliter accipitur "oratio" composite concept. stricte, et sic oratio est congrua dictionum ordinatio, ubi verbum congruit, et nomen vel 18. Ord. Ii dist. 3, q. 3, OTh II, p. 425: ' ... dico quod eiusdem rei possunt esse plures aliquid loco nominis ...' As I read him, Ockham in SL I, 2 is contrasting 'terminus' in conceptus simplices denominativi,et hoc propter diversitatem connotatorum ... ' the second sense with 'oratio' in the wide sense: no complex phrase, he wants to say, is 19. See Aristotle, Categories 1, laI3-15; and SL I, 13, p. 47. a term in this sense. Gaskin's alternative reading of this passage, by contrast, would 20. SL I, 3, OPh I, p. 11:. ' ... prosequendum est de divisionibus termini incomplexi'. leave us with a contrived and bizarre way of distinguishing the different senses of 21. SL I,ll, OPh I, p. 38: ' ... Positis divisionibus quae possunt competere tam terminis 'terminus' . naturaliter significantibus quam etiam terminis ad placitum institutis ... ' (trans!. Loux 28. See SL III-3, 26, OPh I, p. 690: 'Et est sciendum quod relativum potest definiri vel 1974, p. 72). complete vel incomplete. Incomplete potest definiri sine suo correlativo ... Sed 22. See SL I, 41, OPh I, p. 116: ' ... illa quae sunt in praedicamentis sunt incomplexa ex complete non potest definiri sine suo correlativo; ut definiatur sic "pater est substantia quibus sunt propositiones natae componi. Et hoc est verum tam de incomplexis sensibilis habens filium".' The same thesis is found in the Exp. in Porph. 2, OPh II, pp. mentalibus quam vocalibus ... ' 31-2: 'Est autem notandum quod quando aliqua sunt correlativa, ilIa tantum habent quid 23. See SL 1, 10, OPh I, pp. 37-X: 'Immo, qui ponunt quod quaclibet res est substantia vel nominis et non quid rei, et ideo non est inconveniens quod utrumque per alterum qualitas, habent ponere quod omn;u ('(}lIt('fltU ;n uli;s ''''urt!;l'uml'flt;s II .whsllllll;a ('I detiniatur detinitione exprimentc quid nominis: immo hoI' ('st I/('l'('s.wr;um, cum 80 OCKHAM ON CONCEPTS CONNOTATIVE TERMS IN MENTAL LANGUAGE 81

correlativa sint simul in intellectu ...' (italics are mine). The matter is discussed in some language: 'Nam verbum et participium verbi sumptum cum hoc verbo "est" semper details in Quodl. VI, q. 24, pp. 673-8. See also Exp. sup. Elench. II, 16, OPh III, p. 296, videntur in significando aequivalere. Propter quod ... videtur quod non oportet and Ord. I, 30, q. 3, OTh IV, pp. 354-5; participiis vocalibus distinctos conceptus in mente correspondere' (SL 1,3, OPh I, p. 11). 29. Spade 1975, p. 70. See also Quodl. V, q. 8, OTh IX, pp. 512-13: ' ... verbum semper et participium verbi 30. See SL III-3, 26, OPh I, p. 690: 'Et tali definitione relativa mutuo se definiunt. Nee est sumptum cum hoc verbo "est" in significando aequipollent et sunt synonyma' (italics are hoc inconveniens, quia sicut relativa sunt simul in intellectu, ita simul imponuntur, et mine). It is true, however, that a difference in structure between two expressions often propter hoc non est inconveniens si mutuo definiant se. Unde sciendum est quod ista comes together, for Ockham, with some differences in meaning. But we have to definitio non est per notiora, sed sufficit quod sit per aeque nota.' understand why. That is precisely what I am trying to do in the next paragraphs. 31. Tweedale 1992, pp. 434-5,has rightly remarked that not only does Ockham admit, for 39. See above, n. 15. each of two correlative terms, of the possibility of an incomplete nominal definition 40. Quodl. V, q. 20, OTh IX, p. 556: ' ... sed definitio exprimens quid nominis est verborum, making no use of the other correlate, but that he considers, in addition, this kind of adverbiorum, coniunctionum, quia qui vult definire "ubi" dicet quod est adverbium definition, even if it is sometimes incomplete, as being more properly a definition than interrogativum loci; similiter dicet quod "quando" est adverbium interrogativum the one that contains the correlate (ibid.: ' ... est magis proprie definitio, quamvis non sit temporis, et sic de aliis, ubi praedicatur definitio de definito supponente materialiter' semper ita completa'). Tweedale wrongly concludes, however, tliat the best definition of (trans!. Freddoso 1991,.p. 465). a relational term is the one that does not contain its correlate, and that, consequently, all 41. Spade 1996, p. 233, rejects this argument of mine by remarking thatif the definition is relational terms are supposed to be eliminated in the end from the definitions of other metalinguistic in such cases, so are the defined terms themselves since they are taken in relational terms. It is important, here, to precisely understand Ockham's own vocabulary. material supposition, which, he contends, re-establishes semantical equivalence between Ockham distinguishes between definitions 'properly speaking' and complete definitions. the two. But this is off the mark. The question is whether the definition and the defined The former is the one that goes from the better known to the lesser known; it can be used term are synonymous, and synonymy is a matter of signification, not of supposition. to teach a new term to somebody who does not know it yet and this, no doubt, must be Surely the metalinguistic phrase 'an interrogative adverb of place' signifies (in the reason why it is said to be a definition 'properly speaking'. But the second one is Ockham's sense) certain things that the adverb 'when' does not signify even when taken said to be 'more complete'. Why would that be? I can see only one possible explanation: in material supposition (for taking a term in material supposition does not change its this definition alone unfolds in a complete way the meaning of the definiendum. signification), namely adverbs. Hence they are not synonymous. Ockham's thesis is that the meaning of a relational term cannot be completely given 42. Exp. sup. Elench. I, 20, OPh III, pp. 129-33. See also on the same theme Exp. sup. without the help of its correlate. Contrary to what Tweedale thinks, then, the elimination Elench. II, 16, OPh III, pp. 296-301. of all relational terms is ultimately. impossible without semantical impoverishment. 43. Ibid., OPh III, p. 131: ' ... exemplum est de simo, nam definitio exprimens quid nominis 32. Quodl. V, q. 9, OTh IX, p. 518: ' ... quia multitudini nominum synonymorum non ipsius est "naris cava"; si igitur quidquid potest addi nomini potest addi definitioni, correspondet multitudo conceptuum' (trans!. Freddoso 1991, p. 432). See also Quodl. V, sequitur quod sicut convenienter dicitur "naris sima", ita convenienter dicetur: naris q. 8, OTh IX, pp. 510-13, and SL 1,3, OPh I, pp. 11-12. More on this below in chap. 7, naris cava.' The example is that ofAristotle himself (Soph. Ref 173b9-1O). sect. 4.2. 44. Ibid., OPh III, p. 133: ' ... si opponentes non petant in principio sibi dari ilIam 33. See for example SL III-3, 22, OPh I, p. 680: 'Tamen non obstante quod definitio et propositionem, scilicet quod nomen et sua de/initio idem omnino significentet quod loco definitum non sint idem realiter, tamen significant idem realiter' (italics are mine). unius possit alterum poni et quod quidquid convenienteradditur uni potest convenienter 34. See Quaest. in Phys., q. 25, OPh VI, p. 461: ' ... illa nomina "pater" et "filius" significant addi alted ... quasi omninosit certum quod praedicta propositio sit vera, cum tamen non omnino idem, et tamen haec estfalsa "filius est pater".' See also Exp. in Phys. III, 6, OPh sit vera quamvis videatur esse vera ... ' (italics are mine). Michon 1994, p. 369, claims VI, p. 480: ' ... quidquid importatur per unum correlativum, importatur per reliquum, that this remark occurs 'in the context of a study of real definitions', but the text is tamen non eodem modo, quia ilIud quod importatur per unum in recto, importatur per explicitly intended to cover nominal as well as real definitions (see Exp. sup. Elench. I, reliquum in obliquo et e converso ... sicut quidquid importatur per hoc nomen "pater" 20, p. 129). The redundancy ~or nugatio - problem (as in 'concave nose nose') typically importatur per hoc nomen "filius" et e converso, quamvis non eodem modo.' The arises, Ockham says, with nominal definitions (ibid., p. 130: 'sed etiam [causa example is much better, of course, if we use 'parent' and 'child' rather than 'father' and deceptionis] habet locum in aliis nominibus in quorum definitionibus exprimentibus 'son', quid nominis necessario ponitur subiectum'). Since the rejection of the total synonymy 35. See SL I, 6, OPh I, p. 19: ' ... dicuntur ilia synonyma quae simpliciter idem significant between the definition and the defined term is meant to solve the nugatio problem omnibus modis, ita quod nihil aliquo modo significatur per unum quin eodem modo precisely, it better applies to nominal definitions! See also Exp. sup. Elench. I, 20, pp. significatur per reliquum ...' 132-3. 36. Spade 1975, p. 66. 45. See, among many other examples, Quodl. II, q. 7, OTh IX, pp. 143--4' about the 37. Ibid., p. 67. synonymy between 'esse' taken as a name and 'essentia': 'Tamen "esse" aliquando est 38. Martin Tweedale thinks I am needlessly complicating things in trying to demonstrate nomen, et tunc significat omni modo grammaticali et logicali idem cum "essentia'" this point, since the relation of synonymy, according to Tweedale, never connects but (italics are mine). simple terms, in Ockham's vocabulary (Tweedale 1992, pp. 436-7). To which it must be 46. See, for example, Ord. I, dist. 2, q. 8, OTh II, pp. 285-6: 'Verbi gratia, isti voci "homo" replied, first, that there is at least one case where Ockham does consider the possibility competit talis modus grammaticalis quod est singularis numeri, nominativi casus, of a perfect semantical equivalence between a simple term (any verb, in this case) and a masculini generis, et sic de aliis' (italics are mine). Also Quodl. II, q. 7, OTh IX, p. 144: complex phrase (a participle with the verb 'to be'). Such an equivalence, he says, would 'Aliquando l"esse"l est verbum, et tunc signiticat idem verbal iter quod "essentia" suffice - in conformity with premiss (3) above - to exclude this redundancy from mental signiticat nominaliter.' 82 OCKHAM ON CONCEPTS CONNOTATIVE TERMS IN MENTAL LANGUAGE 83

47. SL III--4, 10, OPh I, p. 817: ' ... necesse est cognoscere quas res tennini significant et 68. Gaskin 2001, p. 257, ad 1 (italics are mine). See also ibid., p. 260, ad 4: 'In many qualiter significant eas, scilicet an in recto vel in obliquo,et an affinnative vel negative, passages Ockham writes as if there were simple mental tenns ... ' (with Gaskin's own et antamquam categorematici vel syncategorematici, et an tamquam nomina propria vel italics this time). communia ... ' 69. SL. I, 41, OPh I, p. 116: ' ... illa quae sunt in praedicamentis sunt incomplexa ... Et hoc 48. See Boler 1985, esp. pp. 27-8. est verum tam de incomplexis mentalibus quam vocalibus ... ' (transi. Loux 1974, pp. 49. Spade 1975, p. 58. 129-30; italics are mine).

50. Ibid.., n. 11. On Buridan's Additive Principle, see also King 1985, p. 13. 70. Gaskin 2001, p. 260, ad 4. 51. Michon 1994, pp. 367-8 and Gaskin 2001, p. 239n. Both suggest that the signification 71. Ibid., p. 257, ad 2. This is how Gaskin ultimately accommodates Ockham's idea that a of a complex expression - such as a definition - should not be seen within Ockham's nominal definition should differ from the term it defines (which was argument 2 above). framework as the sum of the significations of its components (as required by the 72. Ibid., p. 259, ad 3. According to Gaskin, Ockham's assertion that the distinctions apply 'Additive Principle'), but that some sort of special semantical restriction should be both to mental and to conventional signs (as quoted above in n. 21) should not be taken supposed to operate in such cases to exclude from the significates of the complex as pertaining to each individual one of the distinctions drawn in the Summa Logicae expression as a whole some of the significates of its components. I can see, however, no before that point. This, however, would be quite strange. The whole intent of this textual support in Ockham for such a strategy. Since it is inconsistent with some of the particular sentence is to serve as a transition marker within Book One of the Summa things he explicitly holds (as in the passage quoted above in n. 44), it should not, as a Logicae in order to separate those distinctions among simple tenns that do apply to both matter of principle, be attributed to him. Moreover, it is difficult to see, in the case of concepts and words (chapters 4 to 10, including the absolute/connotative distinction mental expressions, how the natural signification of their constituent simple concepts treated in chapter 10) from those that don't (chapters 11 to 13). could be cancelled by the surrounding tenns. If, for example, the concept 'animal' 73. Gaskin 2001, p. 260, ad 4; italics are mine. naturally calls to mind every animal there is, why should it cease to do so when it is 74. See in particular the passage from Exp. sup. Elench I, 20 quoted above in n. 44. followed by the concept 'having a child'? This, actually, is the very reason why Spade, 75. Gaskin 2001, pp. 239--40, n. 60. contrary to Gaskin, thinks that Ockham is committed to the Additive Principle. He does 76. See above nn. 15,41,44. not acknowledge, however, that this principle is incompatible, in many cases, with the 77. See my discussion ofthe point in n. 51 above. synonymy of the definition with the defined tenn. 78. Gaskin even describes this as what Ockham came to be 'officially maintaining' (2001, 52. This fonnulation is directly inspired by one of Spade 1975, p. 67. But his is meant to p.240). yield the mode of signification of the definition as a whole rather than that of the defined 79. In addition to (1)-(4) above, Gaskin discusses two more arguments of mine in favour of tenn, which he only afterwards infers from that of the definition. I don't see any reason simple connotative concepts, which he numbers as (5) and (6) (2001, pp. 260-61). for this detour, unless we presuppose that the defined term and its definition are wholly Argument (5) is that correlative tenns, such as 'parent' and 'child' or 'half' and 'double' synonymous, which, as we saw, we should not do if we care to avoid attributing a fatal are said by Ockham to be ultimately ineliminable from each other's complete nominal inconsistency to Ockham. definitions (see the texts referred to in nn. 28 and 30 above). Gaskin's reaction to this is 53. I say 'if' in this case, rather than 'if and only if', because I want to leave open, at this to point out, as Tweedale (1992) had done, that they might nevertheless be eliminable stage, the question of the necessary conditions for oblique signification. R1 and R2 are from the incomplete definitions. This is true, but of no help, I am afraid, for Gaskin's merely given here as examples of rules, and not, of course, as a complete theory. general interpretation. See my discussion of the point in n. 31 above. As to argument (6), 54. Quaest. Var. q. 6, art. 5, OThVIII, p. 231: ' ... licethomo et animal significent idem, which is that Ockham never explicitly rejected simple connotative concepts, Gaskin quamvis non adaequate quia quidquid significat homo et animal et plus ... ' grants it, but thinks there is indirect evidence available. Which brings us back to his 55. See above n. 44. discussion of arguments (1)-(4)! 56. This doctrine of the exponibilia is expounded in SL II, 11, OPh I, pp. 279-82. Attempts to eliminate connotative terms from the mental language on the basis of it are to be found in Michon 1994, pp. 371--4, andYrjonsuuri 1997, pp. 19-21. But, as we shall see, nothing in Ockham indicates that a complete 'exposition' of any given 'exponible' sentence can always be entirely devoid of connotative tenns. Quite to the contrary. 57. See for example Pinborg 1976, p. 266. 58. See Ashworth 1973, pp. 138--42. 59. Ibid., p. 141. 60. Ibid. 61. Gaskin 2001, pp. 256-61. 62. Ibid., p. 240. 63. Ibid., p. 261. This is Gaskin's reply to what he counts as my sixth argument: Ockham never explicitly rejected simple connotative concepts. 64. The text is quoted in n. 12 above. 65. See above n. 14. 66. See above nn. 20-21. 67. See above nn. 22-3. Chapter 5 The Role of Nominal Definitions

Many, as we have seen, have counted Ockham as a resolute supporter of what we call today the 'Definitional View' of concepts, also labelled sometimes as' the 'Classical Theory'. 1 According to this doctrine, most of our mental concepts are actually definitions. They are, in other words, complex mental representations encoding, on the basis of some primitive terms, necessary and sufficient conditions for their own application. In Paul Spade's original interpretation of Ockham, the primitive basis was supposed to be provided by a limited range of simple absolute concepts, acquired by the agent as a consequence of direct encounters with individual exemplifications; all the connotative ones, by contrast - the bulk of our intellectual equipment, certainly - were identified with their nominal definitions. The discussion of the previous chapter now allows us to renounce this picture of Ockham's theory of thought, but its demise, admittedly, leaves us with a pressing puzzle: given that all - and only - connotative terms have a nominal definition, according to Ockham, what is the connection exactly, in his epistemology, between simple connotative concepts and their nominal definitions? Paul Spade himself has forcefully raised the point. If connotative names are not synonymous with their nominal definitions, he asks, then what would the criterion be of a 'correct' nominal definition for a given concept? In other words, 'if it's not synonymy, then what is it?'2 In a 1998 paper in Franciscan Studies, Spade is even more explicit, insisting that his main reason for attributing to Ockham a reductionist programme with respect to connotative terms, had been that such a programme fits so well with Ockham's theory of nominal definition: 'if he did not accept the moderate reductionist programme I have described', Spade asks again, 'then what is the point a/the theory a/nominal definition? '3 , This is an entirely relevant question and I would like to address it in this chapter. It has to do, let me insist, not with definitions in general, but only with nominal definitions (de/initiones exprimeYJ;tes quid nominis), which Ockham sharply distinguishes from real ~ or essential - definitions (definitiones exprimentes quid rei).4 Real definitions, in the strict sense, are descriptions of the intrinsic of things. A man, for example,can be 'really' - or essentially - described as a rational animal. Nominal definitions, on the other hand, primarily say something about names (whether spoken, written or mental). What it is exactly that nominal definitions say about names, and what role they play in Ockham's general nominalistic programme is what we will now be concerned with. The strategy will be the following. I will first recall four characteristic - and non­ trivial - theses of Ockham about nominal definitions (section 1). I will then explain what I take the function of these definitions to be and show how the four theses fit with this conception (section 2). A numher of interesting consequences will finally he drawn about Ockham's epistemology and metaphysics (section 1).

H~ 86 OCKHAM ON CONCEPTS THE ROLE OF NOMINAL DEFINITIONS 87

1. Four theses about nominal definitions • Thesis 3: Each connotative term has only one nominal definition. Ockham explicitly says so: 'properly speaking, a name that has a nominal definition has only The theses r have in mind form at first sight a disparate collection. Yet they are very one such definition'.15 There can be variants, to be more precise, but Ockham's precise and they are clearly and repeatedly stressed by Ockham. Any understanding of pointis that they cannot differ from each other more than by the substitution of a what his theory of nominal definition is about should certainly take them into account. synonym for a synonym.16Nominal definitiol}s for Ockham are always complex phrases. If this is so, it can routinely happen in some spoken or written languages • Thesis I: Some nominal definitions irreducibly contain connotative terms. that at least one of the terms occurring within the correct nominal definition of a Correlative terms, Ockham often repeats, interdefine each other.5 These certain term T has a synonym in the same language. The phrase which differs at correspond, in his vocabulary, to such pairs as 'father' and 'child', 'double' and most from the nominal definition of T by the substitution of this synonym, then, 'half', 'larger' and 'smaller', and so on. They are, in modemwords, pairs ofnon­ will clearly be a correct definition of T as well. The discrepancy, however, is not synonymous converse relational terms such that if one of them is true of two allowed to go further. Any two correct nominal definitions for a given term must be individuals in a given order, the other is true of the·same individuals.in the isomorphical to each other and such that their corresponding simple components reverse order (for example, if x is larger than y, y is smaller than x, and vice are all strictly synonymous with each other. Since there is supposedly no such versa). Since correlative terms are relational, they are ipso facto connotative. synonymy in mental language, there cannot be for any connotative term more than Saying that these terms interdefine each other, therefore, amounts to hold that one appropriate nominal definition of that term within the mental language of a the nominal definitions of certain connotative terms irreducibly contain at least given agent. Ockham, on the other hand, also holds that all connotative terms ­ one other relational connotative term. Nominal definitions in such cases cannot independently of whether they are spoken, written or mental - do have a nominal be expected to eliminate all connotative terms. definition. 17 It follows from both considerations taken together that any connotative term has exactly one correct definition in the language of thought. • Thesis 2: A nominal definition is not always synonymous with the defined term. That Ockham subscribes to this thesis is also something I take to have· been • Thesis 4: The nominal definition of a connotative term is a complex phrase one established in the previous chapter.6 Since, however, the point has been in component of which normally is a term 'in recto' and at least one other dispute in the recent literature,? it will be useful to quickly review what the component a term 'in obliquo'. See, among others, this passage from Summa main arguments are for this ascription: Logicae III-3: 'Connotative terms are defined by their subjects taken in recto and by the names of their connotata taken in obliquo, or by verbs.' 18 (i) Ockham says himself that the principle of a total synonymy between a Disregarding the clause about verbs - to which I will soon come back - what nominal definition and the defined term is not true. 8 the passage says is that a typical nominal definition for a connotative term (ii) He nowhere asserts that such synonymies generally hold;9 contains at least two component terms, one taken in recto and one in obliquo, (iii) Crediting him with the total synonymy principle between nominal and that the former is expected to be the' subject' of the connotative term, while definitions and defined terms introduces a prima facie tension in his the latter is supposed to be the 'name of its connotata'. This is crucial for any account of mental language;, since he is committed to accepting on the understanding of what a correct nominal definition is, according to Ockham, one hand, the possible coexistence in the mind of simple connotative and we must, accordingly, pause a bit to see what it means exactly. The concepts and their complex nominal definitions,lO while he rejects, on the terminology 'in recto'I'in obliquo' also standardly occurs in Ockham to other hand, total synonymy from the mental language.11 characterize the semantical duality of connotative terms in general. (iv) In the case of verbs, adverbs and conjunctions, Ockham clearly admits of Connotative names, he would say, 'signify something in recto and something nominal definitions which he could not have taken as synonymous with else in obliquo' .19Yet his use of the same phrases in the theory of definitions, the defined terms;12this shows at least that he does not accept synonymy although not unrelated, is markedly different. It is relevant for our present as a general constraint on nominal definitions. purpose to grasp both how these two uses are related and how they differ. (v) Ockham holds that nominal definitions signify more explicitly than the defined terms; 13 this suggests that they signify differently, and hence that When Ockham says that a connotative term signifies something and they are not synonymous with them in the strong sense. in recto (vi) Some things are usually signified in certain ways by parts of the nominal something in obliquo, he is talking about the modes of signification of that term. To signify in is the same, in this vocabulary, as to signify primarily, and to signify definition, which are not signified in the same way by the defined term)4 recto in obliquo is the same as to signify secondarily. 'White', for example, is said to Taken jointly, these considerations are decisive, it seems to me, for attributing thesis signify in recto - or primarily - all the white things, and in obliquo - or secondarily 2 to Ockham. The possibility is not entirely excluded that some nominal definitions - all their whitenesses; 'father' is said to signify in recto - or primarily - all the might, in particular cases, be taken as synonymous with what they are used to define, fathers, and in ohliquo - or secondarily - all their children. The distinction, here, is but such situations cannot be but exceptional. purely semantical; it has to do with two permanent semantical features of every 88 OCKHAM ON CONCEPTS THE ROLE OF NOMINAL DEFINITIONS 89 connotative term. By contrast, when Ockham uses the couple 'in recto'l'in obliquo' the negative term 'non-man', the definition of which, he says, is 'something which in the description of nominal definitions, he is speaking of grammatical features. A is not a man'.26 In this particular case, 'something' (aliquid) is the main term in recto term is said to be taken (sumptus) in recto, according to this use, when it is in the - the subject of the connotative- and 'm.an' (homo) names the connotata of the nominative case, and in obliquo when it is in one of the so-called 'oblique' cases: the defined term - namely all men- because, although it is also in the nominative, it genitive, the dative, the ablative, or the accusative. This, indeed, was the standard occurs nevertheless in a subordinate grammatical position and within the scope of a terminology of grammar at the time. It primarily pertains to languages with negation. declensions like Latin or Greek. Yet, what Ockham uses it to say about nominal As to the verbs themselves, when Ockham says - without explanation- that they definitions exceeds the limits of such languages. A term taken in obliquo in a certain too can occur in nominaldefinitions,27 he is probably thinking of such verbs as 'to 28 context simply is a term which plays in this context the grammatical role of a be' or 'to have', ~nso far as they can be treated as syncategorematic copulas. The complement, while a term taken in the nominative or in.recto is a term which is not, definitions of 'father' and 'white', for example, both contain an occurrence of 'to in the relevant context, playing the role of a complement. In the end, it all amounts have'. Other verbs, such as 'to run' are taken to be analysable into a categorematic to this: a nominal definition, according to Ockham, is a grammatically structured component plus a syncategorematic copula:29 the connotative term 'runner', for phrase that normally contains one categorematic term which is not, in this very example, is presumably best de.fined in mentalese, according to this view, not as i a phrase, playing the role of a grammatical complement, and at least one term which running creature' (which contains no term taken in obliquo),but as something like is. In 'animal having a child', which is supposed to be the nominal definition of 'a creature having (or making) a run', where the verb 'having' (or 'making:) is a 'father',20 'animal' is not a grammatical complement (it is in recto), while 'child', syncategorematic copula, and where the grammatical complement';l run' being the complementof the verb 'to have' ,is in obliquo. designates, exactly like in the previous cases, the connotata of the defined term, Now, the term taken in recto,as we have seen, should be the 'subject' of the namely the acts of running, considered here as real qualities. connotative term.21What this means, I take it, is that it should circumscribe the domain of individuals among which the defined term is . It should indicate, in other words, a natural group - presumably the smallest one - to which all the primary 2. What defining amounts to: a reconstruction significates of the defined term belong: the group of animals, for example, is the smallest natural one in which the term 'father' has its primary significates. It is easy, To this cluster of theses must be added, of course, the two more familiar ones which then, to see the connection between Ockham's two uses of 'in recto': the term which formed the basis for Spade's reductionist reading of Ockham: is grammatically taken in recto within the nominal definition of a connotative term circumscribes the domain where what the defined term signifies in recto is to be found. (5) Every connotative term has a nominal definition. 3D As to the terms taken in obliquo within the definition, they are said by Ockham (6) No absolute termdoes.31 to be the names of the connotata of the defined terms (nomina connotatorum).22 Not the proper name, of course. What Ockham means by 'the name of the connotata' is These six tenets are the pieces of the puzzle that any sound reconstruction of a term which applies exactly to the connotata of the defined term. 'White', for Ockham's theory of nominal definition should assemble. The redllctionist example, is defined as 'a body having a whiteness'.23'Whiteness', here, is the term interpretation, by contrast, is incompatible with (I), (2) and (3) taken separately,32 in obliquo; it is true exactly of all and only the connotata of the defined term, namely and it provides ho account for Thesis (4). The reason for this inadequacy is that the the whitenesses. The same, to give only one more example, holds for 'father': 'child' interpretation in question simply projects on Ockham's idea of a nominal definition occurs as a complement in its definition and applies exactly to its connotata, namely the Frege~Russellconception of what an 'explicit definition' should be. An explicit all the children. This isto be generalized, thus revealing what the relationis between definition for a certain term '1', according to this modem tradition, is a phrase which Ockham's two uses of the phrase 'in obliquo': the terms taken in obliquo in the is wholly synonymous with 'I' and substitutable for it in all contexts (except those, of nominal definitions of connotative names should have as their extensions exactly course, where T is mentioned rather than used). Thus, when a given expression is what the defined terms signify in obliquo. We may add, I suppose, that if the defined 'explicitly' defined with the help of some other terms, it appears as a mere term has more than one group of connotata - if, that is, it connotes some things abbreviation for its own definition and becomes eliminable from the language in under a certain mode and sOme under another mode -, then the correct nominal which it is so defined,. without any semantical loss. This, however, was not how definition for such a term will normally include more than one term taken in obliquo, Ockham conceived of nominal definitions in general. 24 each one occurring with,ln the definition as a distinct grammatical complement. The key to what he expected from such definitions is given in a cameo It is still possible in some definitions, as Ockham admits, that this role of naming formulation by the following sentence from the chapter 'On definition' of the Summa the connotata of the defined term be assigned to a term in the nominative case. But Logicae: 'A nominal definition is a complex phrase (oratio) explicitly indicating this does not change the general picture since it happens only when the term in what is importl'd (importatur) by a certain single term (per unam dictionl'nI).'11 This question occurs as a grammatical attribute introduced within a relative clause by a is what we now must explicate. verb such as 'to be', sometimes accompanied by a negation.2~Ockham'scxample is A signiticant feature of this characterization is that it contrasts the nominal 90 OCKHAM ON CONCEPTS THE ROLE OF NOMINAL DEFINITIONS 91 definition as a complex phrase with the ·defined term, which is normally taken to be the definition one and only one grammatically subordinate categorematical term ­ a simple expression.34The nominal definition of a connotative name, in Ockham's normally a term in obliquo in the medieval vocabulary - having as its own primary system, is expected to unfold the meaning of the defined term. Now, the complete significates exactly these connotata ofthe defined term. Take 'white'. It has a single meaning of a connotative term lies in its signifying some things under a certain group of connotata, according to Ockham, namely all the whitenesses there are, and mode, and some under another mode. What the nominal definition should do, then, its definition is said to be 'a body (or something) having a whiteness'. It can be seen in order to unfold such a meaning, is to explicitly - and separately - identify the that the term 'whiteness', which occurs as the sole grammatical complement in this primary significates of the defined term and its secondary significates, while definition, has as its own primary significates exactly the individuals which happen disclosing in each case the mode under which they are signified by the defined term to be, according to Ockham, the connotata of the defined term, namely the (in recto or in obliquo, affirmatively or negatively, and so on). The point of the whitenesses. Or take 'father'. Its nominal definition is supposed to be 'animal having operation is revealed by Ockham's use ofthe verb 'importare' in the above quotation a child', where it is observed again that the term occurring in obliquo, 'child' namely from Summa Logicae I, 26: 'Importare' is the general verb Ockham standardly uses (which, in this case, is itself a connotative term), has as its own primary significates to indicate the ontological import of an expression: a term is said to 'import' exactly those individuals that are taken to be the secondary significates of the whatever objects of the world it signifies in whatever way.35The primary goal of a defined term, namely the children. More complicated cases could be devised if we nominal definition, for Ockham, is to make explicit the ontological commitment turned to terms with several series of connotata,37 but the general principle would be which is to be associated with the normal use of a given term. the same: each grammatically subordinate name within a good Ockhamistic nominal That is the whole secret: it is all a matter of ontology. Which is why nominal definition should primarily signify exactly those individuals which are connoted definitions are so important within the framework of Ockham's nominalism. Their under a determinate mode by the defined term. function is to make it perspicuous that the things that are referred to in one way or (C3) The nominal definition should connect its categorematic term in recto with another by a meaningful connotative term are but singular things, and more the complements it contains, so as to reveal the modes under which the secondary precisely, singular substances and singular qualities. The role of nominal definitions significates of the defined term are connoted, and to make it perspicuous in this way is not to eliminate connotative terms, but to show, on the contrary, that they are how the primary significates of the term in recto have to be related in reality with the ontologieally innocuous, even when they are ineliminable. Nothing in Ockham's primary significates of the term(s) in obliquo, in order to be the primary significates nominalistic ontology prevents us from having simple connotative concepts in our of the defined term as well. Whether in mentalese or in external language, this is basic mental apparatus. What the nominal definitions must show, precisely, is that secured by grammatical structure and by the use of appropriate syncategorematic such simple connotative concepts require no undesirable complexification of the phrases (including copulas such as 'being' or 'having'). Let us return to the ontology. definition of 'white' once more: 'a body having a whiteness'. As suggested by Paul How is that to be done? We can, I submit, extract from Ockham's own practice Spade, 'having', here, might be seen as a syncategorematic copula connecting the and with respect to the nominal definitions of connotative names, a term in recto 'body' with the complement 'whiteness'}8 What is required from it in relatively precise set of conditions he wanted such definitions standardly to satisfy. this circumstance is that its logic be adequate to express how singular bodies should The following three can be proposed as a fair approximation. stand with respect to whitenesses if they are to be among the primary significates of (Cl) The nominal definition should contain - usually as its first component - a the defined term 'white'. It makes little doubt that Ockham's conception of such term which is not in a grammatically subordinate position within this definition. This syncategorematic copulas remains largely undeveloped,39 but it would certainly be is the term in recto ofThesis (4) above. It should have among its primary significates relevant for the correctness of a definition with 'having' as a copula that 'having' is all the primary significates of the defined term. Its extension, in other words, should asymmetrical in a way that 'being' is not. Since syncategorematic terms, according include the extension of the defined term. Consider our familiar examples of to Oc~ham,do not signify special determinate things of the world, their contribution nominal definitions again. In 'animal having a child', the term 'animal' occurs in to the definition can only rest upon such logical and grammatical properties. recto: it is a categorematical name not occurring here as a complement. Its primary What we end up with, in short, is this: the correct nominal definition of a significates - all the animals there are - include all the primary significates of the connotative name should precisely delineate (through the term(s) in obliquo) the defined term 'father'. In 'a body having a whiteness', it is 'body' that plays this role: connotata of the defined term, while revealing the modes under which they are its primary significates are all the bodies there are, including, of course, the primary connoted; it should provide, moreover, through the syncategoremata and the significates of the defined. term 'white', namely the white bodies. The general idea grammatical structure, a kind of algorithm for picking out the primary significates is that the term in recto should delineate, at the outset of the nominal definition, a ­ of the defined term among the primary significates of the term in recto.40 Whatever usually wider - range of things among which the primary significates of the defined does that job for a given term T will be a good nominal definition of T. term are to be looked for: that of animals in the case of 'father', and that of all the This reconstruction fits well with the passages where Ockham is explicit about bodies there are in the case of 'white'. A passage of the Ordinatio strongly suggests the role of the nominal definitions of connotative names,41 and with most of his that this term in recto should in general be an absolute name. 1fi actual examplesY It accounts, in particular, for the otherwise surprising Theses (C2) For each group of ('onnot(J((J of the defined term, there should occur within (I )-(4) above. which we can now return to (in the reverse order). 92 OCKHAM ON CONCEPTS THE ROLE OF NOMINAL DEFINITIONS 93

First, it is easily seen in the light of CI-C3 why a nominal definition should prefer, with the significata of an absolute term (or combination of absolute terms). normally contain a teFIll in recto and at least one grammatically subordinate Only a connotative, in many cases, willbe able to play the role expected from a term categorematical term,as stated in Thesis.(4): the latter identifies the connotata of the taken in obliquo within a correct nominal definition. Take a relational term like defined term (by C2), while the former delineates the group of things among which 'smaller', for example. Its connotata, presumably, are all the things that are such that its primary significates will be found (by CI). Taken together, they indicate what something else is smaller than they are. This, however, does not correspond to any must exist in the world for the defined term to be significant. natural kind, and there is no absolute concept signifying precisely these individuals. Why now should there be only one correct nominal definition for any given If the complete nominal definitionof sucha,term is to satisfy C2, it will have to use connotative name, in the sense indicated by Thesis (3)? Well, this is precisely as a complement its own correlative, namely 'larger', which is a connotative too. because all the terms occurring within a nominal definition have, by CI-C3, a very Theses (5) and (6), finally, also naturally follow from conditions CI-C3: every precise job to do: the term in recto normally corresponds to the smallest natural connotative name will have a nominal definition of the required form, since every group of things among which the defined term is predicable, and the terms in obliquo such name, as Ockham frequently insists, signifies something in recto and must have as their extensions exactly the connotata of the defined term. The role of something in ohliquo. No absolute term, by contrast, can be given such a definition: revealing in the indicated way the ontological commitment associated with the use having no connotata, it could neveF receive a definition complying with C2 and C3. of a connotative name leaves virtually no room .for manoeuvre, at least at the .level The whole network of these apparently disparate theses, then, is seen to be highly of mental discourse. One might object that since, according to CI and C2, what cohesive when we realize that the primary function of a nominal definition for matter for the correctness of a nominal definition are only the primary significates Ockham is to explicate the ontological import of the defined term. (that is, the extension) of its categorematic components (plus of course,the syncategoremata), nothing in our set of rules prevents any particular categorematic term occurring in obliquo within a definition. to be replaced in this definition by 3. Some consequences some other coextensive, but non-synonymous, term (contrary to what Ockham says himself).43 But note that mere actual coextensiveness will not do: a nominal Three consequences can be drawn, which are of special interest for the ongoing definition should hold for all possible . And given Ockham's ontological discussions on Ockham's theory of concepts. atomism, .there. can hardly be, in his' epistemology, necessary coextensiveness between any two non-synonymous simple concepts" whether absolute or 3.1 Definitions and abbreviations connotative.44 Let us tum now to Thesis (2): a definition is not always synonymous with the Since a connotative name is not in general synonymous with its nominal definition, defined term. At least two reasons can be given why a nominal definition satisfying it cannot be seen as a mere abbreviation for it, whether in mental, spoken or written CI-C3 will not, in general, be synonymous with the defined name. First, the languages. Ockham, on the other hand, does admit of the possibility of spoken or definition is supposed to contain, in virtue of CI, a general term in recto, the primary written conventional abbreviations for any complex sequence of terms whatsoever: significates of which will include - and usually exceed ~those of the defineq teffi1. 'For the speakers of a language can,if they wish, use one locution in place of several. Thus more individuals are normally signified by the definition than by the defined Thus, in place of the complex expression "every man", I could use "A"; and in place term, and hence the two of them cannot be synonymous in Ockham's sense. A of the complex expression "man alone", I could use "B", and so on with other second reason is that,as I have repeatedly stressed, the terms in obliquo which are expressions.'45 Such conventional abbreviations, admittedly, .will be entirely called for by condition C2 need not be absolute tenlls. What is required from them equivalent in meaning with the abbreviated phrases and they will be substitutable for is that their primary significates precisely correspond to the secondary significates them in every context (short of material supposition), since that is what they are of the defined terms. Whether or not they have secondary significates themselves, instituted for. But this should not be conflated with the having of nominal definitions and whatthese secondary significatesare, is not directly relevant for the correctness by normal connotative names. ofthe definition. It is bound to happen, then, that some parts of the definition will For one thing, abbreviations can .be used in place of any sequence, and not only connote certain things that are not Gonnoted by the defined term (as in the definition of those that contain, in conformity with Thesis (4), a term in recto and at least one of 'father' as 'animal having a child'), and this brings about another excess in the gmmmatically subordinate term. Take Ockham's second example in'the passage meaning of the definition with respect to that of the defined term. from Summa Logicae 1,8 just quoted: 'b' in this case is introduced as an abbreviation Why, finally,is it impossible to completely eliminate all connotative terms from for 'tantum homo', which does not have the required form for a regular nominal such nominal definitions, as stated by Thesis (I)? This is also something we can now definition. The abbreviating procedure, indeed, can be used even ifthe abbreviated clearly understand by reflecting upon the role of the terms in ohliquo within the phrase has no logical or grammatical unity. Another of Ockham's examples, just a nominal definition according to C2. What the complements should do, we have said, few lines further, is that of an abstract term like 'humanitas'. In some of its uses, is to delineate exactly the ('onnotata of the defined term. But there is no reason why Ockham says, it is a mere abbreviation for' homo necessario' ('man necessarily'),4tl the ('of/f/otala of any connotative term should coincide with a natural kind. or, if you Yet the phrase 'homo f/c(,c.I',\'(/rio' has in itself no logical or grammatical unity: 94 OCKHAM ON CONCEPTS THE ROLE OF NOMINAL DEFINITIONS 95 except when taken in material supposition, it cannot be the subject or the predicate like 'white'; it is, rather, that they misunderstand the ontological import of these of a well-formed sentence and cannot, properly speaking, supposit for anything. The concepts. Which amounts to say that they are wrong about their nominal definitions. two components 'homo' and 'necessario' simply happen to follow each other in One could object that there are a number of passages where Ockham insists that some sentences (as in 'homo necessario est animal': a man necessarily is an animal), the knowledge of the correct nominal definition of a given connotative name is and only in such sentences can they be replaced by the abbreviation. presupposed by any sound use of this name in demonstrations or in .53 In cases such as the latter, where the abbreviated phrases are not in themselves This, however, does not entail that the possession of the corresponding concept is significant units, the resulting abbreviations are not connotative names. Strictly impossible without the express knowledge of the definition. Disputations and speaking, they are not names at all, as comes out from Ockham's lengthy analysis of demonstrations are scientific activities conducted, in general, in some public the abstract terms 'punctus', 'linea', 'instans', 'mutatio', 'generatio', and so on in conventional language, and their scientific and interpersonal character do require, no his Tractatus de Quantitate. 47 Such words, he insists, are introduced 'in place of an doubt, that all the participants have a common and clear understanding of the words expression or of something composed of several words', but 'they do not have they use. But the point, here, is that we might possess certain simple concepts which precisely the value of a name';48 properly speaking, they cannot be the subject or the we are not ready to make use of in such scientific arguments, precisely because our predicate of well-formed sentences and they cannot supposit for anything.49These grasping of them has not yet been rendered explicit enough. When Ockham writes, limitations, of course, do not hold in all case~of abbreviations: if a single letter, let's for example, that 'the significate of a word cannot be known without the nominal say, was introduced as an abbreviation for an absolute or a connotative name, it could definition, if it has one',54he presumably uses a strong sense of the verb 'to know' properly be used as a subject or predicate and have a personal supposition just as the (scire), and he uses it in connection with what the significate of the connotative term original name could. The general point, however, is that the relation of an is. In other words, what he is saying is that we cannot have a clear knowledge of abbreviation with the abbreviated sequence is entirely different, in Ockham's view, what things exactly are signified by a given connotative term without knowing the from the relation of a connotative name with its nominal definition. The abbreviation nominal definition of that term. This should not surprise us: it is precisely what a process merely aims at a gain in brevity, in elegance or in poetic rhythm.50A nominal nominal definition is for, according to the theory reconstructed above. What should definition, on the other hand, even if it can be used too as a procedure for introducing be concluded is not that we cannot have a concept without knowing its definition, new terms in conventional languages, normally has a completely different function but rather that if we do have a concept·without the definition, our speculative in Ockham: that of revealing the ontological import of a given term, whether mental mastering of this concept is imperfect. We might, for example, applyit correctly to or conventional. its most typical instances in everyday life, without being able to explicate it, and without, consequently, being able to make a rigorous use of it (or of the associated 3.2 Possession of concepts and knowledge of definitions spoken or written words) in scientific demonstrations or academic disputations. It thus turns out - even if he was not quite clear about it himself - that Ockham's It has often been assumed that a nominal definition, for Ockham, is what a person theory of connotative concepts and nominal definitions challenges the idea that the should know if she is to be attributed the corresponding concept.51We can now see conceptual content of the mind is always immediately perspicuous to the thinking that this cannot be generalized. Although Ockham is not explicit on the point, it agent. follows from the theory of nominal definition presented above that, in the case of simple connotative concepts at least, to have the concept is not the same as to know 3.3 Real orderings what its nominal definition is. You could have naturally acquired a simple relational concept such as 'similar' or 'smaller', without being able to tell at the first blush and According to condition C3 above, a good nominal definition for a given connotative with what its correct nominal definition is, even if such a definition exists, name should show how the primary significates of this name are connected in reality and even if you do possess all the terms that are required to build it. This follows with its secondary significates. This implies that real singular things are ordered in from the very notion of what a nominal definition is for, as we have reconstructed it. the world in certain precise ways independently of the human mind, and that the The objective of such a definition, we have said, is to make explicit the ontological human mind can sometimes acknowledge such connections. Ockhamindeed is quite import of the defined term. Now, you could be mistaken about the exact ontological explicit that ordering does not result, in general, from a merely creative work of the import of one of your own simple concepts. This happens to realist philosophers all intellect: 'it is not the intellect', he writes, 'which brings it about that one thing is the time, according to Ockham! They typically think that a concept like 'white', similar to another, any more than it brings it about that Socrates is white or that Plato which is presumably a good example of a simple connotative concept directly is white'.55 In the Summa Logicae, he enumerates a number of ways in which acquired on the basis of intuition,52refers in some way to some single universal something can be prior to something else in reality: 'it can temporally precede the property of whiteness. By Ockham standards, they are mistaken about the other, or it can exist without the other but not conversely, or it can be more perfect significates of the simple concept 'white', and consequently, about what its correct than the other'.56 This is an incomplete list, no doubt. Things can be really ordered nominal definition is. Yet they can hardly be denied the possession of this concept. for Ockham in many ways, without the human mind having anything to do with it: The trouble with realist philosophers is not that they lack certain ordimlryconcepts temporally, spatially, causally, axiologically, rnereologically (as parts to wholes), 96 OCKHAM ON CONCEPTS THE ROLE OF NOMINAL DEFINITIONS 97 hylemorphically (as fonns to chunks of matter), or inherently (as accidents to in many other parts of his system, was with nominalism, let me repeat, and he was substances). more interested, consequently, in the role of categorematic tenns within definitions What is required for thought and knowledge to take place is that the mind be (conditions Cl and C2) than with the role of syncategoremata or grammatical organized in such a way as to be able to recognize in certain cases that .certain things structure (condition C3), for it is the categorematic terms in the definition that are ordered in certain ways with respect to certain others, that the one, for example, ultimately reveal, in his view, the neat ontological import of the defined tenns, by is larger than the other, or that the· one inheres in the other, and so on. When such a identifying both their primary and their secondary significates. As to the remaining reccignitionoccurs on the basis of an intuitive grasping, it involves, as we shall problems about the ways in which nominal definitions are supposed to reveal how develop in the next chapter, the fonnation of a' simple connotative concept the primary and secondary significates should be connected with each other, what associating in some precise way its primary significates with its secondary Ockham is committed· to is that they can be solved in principle without further significates. This is what the nominal definition is supposed to express, in virtue of ontological commitments. This, it seems to me, is not obviously doomed to failure, C3, by its grammatical structure and its syncategorematic components. Ockham is contrary to the reductionist programme Spade - anachronistically - read into the committed to admit, consequently, that the human mind has innate capacities for Venerabilis Inceptor's theory of nominal definitions·. grasping, in at least some respects, how things are ordered outthere in the world, and to map these orderings into the very fonnation of its connotative concepts. This is no breach to Ockham's distinctive brand of nominalism,. since, as he sees Notes it, it does not require the positionof orders or relations as distinct things in addition to singular substances and qualities. A relation, he.often says, is not an additional 1. For a good account of the so-called 'Classical Theory' ,'see Laurence and Margolis 1999, 'small thing' (parva res) in between the relataY The spatial order, for example pp. 8-14. 'It would be difficult', th,e authors write, 'to overstate the .historical 'involves only absolute things' such that one of them is more distant than another predominance of the Classical Theory' (p. 10). Fodor 1998 devotes two chapters (pp. from the same thing, or closerthan another, without any spatial connection inhering 40-87) to the criticism of such approaches. 2. Spac!e1996, p. 214. in any of them as a distinct entity. 58 An order is not like a bond between two bodies, 3. Spade 1998, p. 355 (the italics are Spade's). What he calls here the 'moderate 'as if these bodies would not be ordered .,. without such an intermediary, as reductionist program' holds that 'any statement containing connotative terms can be Simplicius has imagined in his commentary on the Categories' .59 In general, then, paraphrased by (and so in that sense "reduced to") a strongly equivalent (perhaps even Ockham's theory of connotative concepts and nominaldefinitions.implies that strictly synonymous) statement that does not contain connotative terms, but only singular things are ordered in certain ways, but this does not add to the ontologiCal absolute categorematic terms plus syncategoremata' (ibid., p. 350). commitments he associates with such concepts since the orders themselves 'are not 4. See SL I, 26: 'De definitione: quot modis dicitur definitio?', OPh I, pp. 84-9. countenanced as extra entities. What should be said, strictly speaking, is not that 5. See the texts quoted above chap. 4, nn. 28, 30 and 31. The interdefinability of certain there are orders in the. world, but that things are ordered. The difference, here, being terms was accepted, long before Ockham, by ; see Porphyrii lsagoge. between the correct use of a complete true sentence (for example, 'things are Translatio Boethii, ed. Minio-Paluello 1966, p. 9: 'Nosse autemoportet, quoniam et spatially ordered') andthe - strictly speaking, incorrect - use of a pseudo-name such genus alicuius est genus et species alicuius est species, idcirco necesse est et in utrorumque rationibus utrisque uti..' Porphyry's point is that 'genus' and 'species' are as 'order'. That both theses - things are ordered, yet no order exists - can be held correlative terms and therefore interdefine each other. together rests on how Ockham understands what the truth-conditions are for 6. See above chap. 4, sect, 3. relational sentences: these do not require any absolute things in the world other than 7. See Michon 1994, pp. 368-9; Spade 1996, pp. 231-5; Gaskin 2001, p.239, n. 60. singular substances and singular qualities, precisely because connotative names 8. See Exp. sup. Elench. I, 20, OPh III, p. 133. The text is quoted above in chap. 4, n. 44 (including relational ones) signify nothing else, whether primarily or secondarily.60 and Michon's objection to its use in this context is rebutted in the same footnote. Of course, condition C3, as I have already remarked, is not explicit enough and it 9. It is true that Ockham sometimes writes that the definition and the defined term 'signify remains mysterious to some degree how the grammatical and logical structure of a the same' (see, for example, the passage from SL I1I-3, 22,OPh I, p. 680 quoted above definitional phrase is supposed to'" mirror reality. Should mental language .be in chap. 4, n. 33), but, as we have seen, this should not be taken to mean that they are endowed, for example, with a special syncategorematic copula for each type of synonymous in. the strong sense. He also says, sometimes, that the definition' and the ordering thatthe mind is inflately able to recognize?6J Or should we be content with defined term 'are equivalent in signification' (aequivalent in signiflcando), but normally the statement is either qualified with an 'aliquo modo' (see Ord. T, Pro!., q. 3, OTh.1, p. definitions' being incorrigibly. imprecise in this.regard (with, for instance, only one 139), or limited to the special case of conventional abbreviations (as in Quaest. in Phys., copula being available for all asymmetrical orderings)? Those are questions that are q. 63, OPh VI, p. 568). Beretta 1999, pp. 162-4 makes much with a passage from SL I, raised by Ockham's approach, but that he never directly addressed. That he didn't do 5l, where all relative terms are said to be equivalent in signification to the longer phrases so, however, is not surprising. They would have required an in-depth into the which are their nominal definitions (OPh I, p. 171: 'Non enim quaerendum est omnibus structure of the human mind, which he simply did not have the tools to achieve. And ten11inisquid rei scd tantum quid nominis in multis, quales sunt omnes termini relativi however important they may he in themselves, they were not, on the whole, that ct nonnulli alii, qllOl"ll1llqlli/ihct al'qlliva/ct ill sigllifh:ando /ongal' orationi', italics arc central to his own purpose. His main concern in the lheory of nominal definition, as mincJ, hut this vcry chaptcr has hccn dccmcd inauthcntic hy thc cditors of Ockham's 98 OCKHAM ON CONCEPTS THE ROLE OF NOMINAL DEFINITIONS 99

Summa (see the Introduction by G. Gal and S. Brown in OPh I, pp. 40*-41*); even be', but any other connecting verb such as 'inest' (see, for example, SL II, 11, OPh I, p. though they believe that there is nothing in it which Ockham would not have approved 281 where 'inest' is treatedas the copula of such sentences as 'Sorti inest albedo'), of, not too much weight should be given to the details of the formulations found in it. 'incipit esse' (see SL 1,75, OPh I, pp. 231-3), and even verbs like 'vidit' (see SL II, 21, 10. See above chap. 4, sect. 2, with, in particular, the texts quoted in nn. 12, 14-17, and OPh I, p. 321). Some among these, like 'esse', and maybe 'habere', turn out to be purely 22-3. syncategorematic terms in the sense that they do not have any special significata of their 11. See, for example, Quodl. V, q. 8, OTh IX, p. 511: '... nec nominibus synonymis own (see Quodl. VI, q. 29, ad 1, OTh IX, p. 695: ' ... dico quod unio extremorum correspondet 'pluralitas conceptuum'; as well as the other texts mentioned above in propositionis in mente est conceptus syncategorematicus verbi copulativi sive copulantis chap. 4, n. 32. subiectum cum praedicato'; or again Exp. in Perih. I, 2,4, OPh II, p. 389: 'hoc verbum 12. See the passage from Quodl. V, q. 20, OTh IX, p. 556 quoted above in chap. 4, n. 40, as "est" neque est signum rei'). More on this below in chap. 8, sect. 3. well as the discussion of Spade's objection tothis argument in n. 41. 29. Some verbs, Ockham says, signify certain things determinately (Exp. in Perih. I, 2, 2, 13. See the passage from Summ. Phil. Nat. I, 3, OPh VI, p. 162 quoted above in chap. 4, n. OPh II, pp. 387-8). 'To run' is a good example: it primarily signifies the runners and 15. connotes their running acts. Such verbs are normally equivalent to a combination of 'to 14. The point was developed in details in chap. 4, sect. 3. be' plus a categorematic participle (SL 1,3, OPh I, p. 11: 'Nam verbum et participium 15. Quodl. V, q. 19, OTh IX, p. 555: ' ... proprie loquendo unius nominis habentis verbi sumptum cum hoc verbo "est" semper videntur in significando aequivalere'). definitionem exprimentem quid nominis est tantum una definitio explicans quid nominis When it comes to logical analysis, the composite form 'to be + participle' is more ... ' (trans!. Freddoso 1991, p. 464, with a minor amendment). fundamental (see, for example, SL II, 21, OPh I, p. 321, where Ockham discusses the 16. Ibid.:' ... talis nominis non sunt diversae orationes exprimentes quid nominis habentes case of propositions with adverbial determinations: 'Et tunc in tali casu ... opportet partes distinctas, quarum una pars significat aliquid quod non eodem modo importatur resolvere verbum in suum participium et in hoc verbum" est "'). per aliam partem alterius orationis'. 30. See, for example, the passage quoted above in n. 17. 17. SL I, 10, OPh I, p. 36: 'Et tale nomen [connotativum] proprie habet definitionem 31. See SL I, 10, OPh I, p. 35: 'Immo, proprie loquendo talia nomina [= nomina absoluta] exprimentem quid nominis ...' non habent definitionem exprimentem quid nominis ... ' 18. SL III-3, 26, p. 691: 'Connotativa definiuntur per sua subiecta sumpta in recto et per 32. The incompatibility of Spade's original interpretation with Theses (1) and (2) is nomina connotatorum sumpta in obliquo, vel per verba.' See also Exp. sup. Elench. II, straightforward; Spade held both that what he called the 'fully expanded nominal 16, OPh III, pp. 300-301. As Adams remarks (1987, p. 321), Ockham in SL I, 10, OPh definition' of a connotative term contains no simple connotative term (Spade 1975, pp. I, p. 36, says that the nominal definition of a connotative frequently contains a term in' 69-70), which is the negation ofThesis (1), and that every connotative term is synonymous recto and a term in obliquo, leaving it open that it might not always be the case CEt tale with its nominal definition (ibid.), which is the negation of Thesis (2). As for Thesis (3), nomen proprie habet definitionem quid nominis, et frequenter oportet ponereunum the case is more indirect. A connotative term, according to Spade, frequently has more than illius definitionis in recto et aliud in obliquo', italics are mine); this possibility will be one. correct Ockhamistic nominal definition, since not all nominal definitions, he thinks, accounted for below (see in particular the passage quoted in nn. 25 and 26). are 'fully expanded'; some of them, that is, are taken to contain connotative terms which 19. Quodl. V, q. 19, OTh IX, p. 554: 'Definitio autem exprimens quid nominis proprie est de have not yet been replaced by their own nominal definitions (Spade 1975, p. 69). This is nominibus connotativis et relativis quae significant unum in recto et aliud in obliquD, not enough, though, to contradict Thesis (3) as Ockham understands it, since it is explicitly sicut "album", "calidum", "pater" et "filius" sunt huiusmodi nomina.' held by Ockham that there can be several variants for the nominal definition of a given 20. See for example SL III-3, 26, OPh I, p. 690. connotative name, on the condition that they all be strictly synonymous with each other, a 21. See the text quoted above in n. 18. condition which Spade readily admits (Spade 1975, p. 66). The problem, rather, is that 22. Ibid. Spade's notion of synonymy among complex phrases differs from Ockham's. According 23. See SL III-3, 26, OPh I, p. 691: 'Album autem definitur sic "album est corpus habens to Ockham, two different nominal definitions are synonymous variants ofeach other if and albedinem".' Ockham also loosely defines 'album' sometimes as 'aliquid habens only if none of them contains a part which signifies under a given mode something which albedinem' (for example, in SL I, 26, p. 88), but the definition with 'corpus' is more is not signified under the same mode by some corresponding part of the other (see, for precise. example, the passage from Quodl. V, q. 19, quoted above in n. 16). Now, two nominal 24. You mightthink, although this is not one of Ockham's examples, of a term like 'gift', definitions one of which contains a connotative term which is replaced in the other by its which connotes, presumably, both the donators and the receivers, but under different own nominal definition will not always be synonymous in this sense, since, as we have modes. Its Ockhamistic nominal definition, then, should include two distinct previously.stressed, some part of the replacing nominal definition might very well signify grammatical complements. It would read a bit like this: 'something that pa'sses from a something under a certain mode which is not signified under the same mode by any part donator to a receiver'. of the original definition. Suppose, for example, that a connotative term like 'white' occurs 25. See SL II, 12, OPh I, p. 283: ' ... omnes tales termini [= negativi, privativi, infiniti] sunt in some not yet 'fully expanded' nominal definition for a given term T. When 'white' is vere connotative, eo quod in eorum definitionibus exprimentibus quid nominis debet poni replaced within this definition of T by its own nominal definition ('a body having a aliquid in recto et aliquid in obliquo, vel in recto cum negatione praevia' (italics are mine). whiteness', according to Ockham), there will be a part of the resulting definition for T 26. Ibid.:' ... et definitio istius termini "non-homo" est ista "aliquid quod non est homo", et (namely 'body' in this case) which will signify under a certain mode certain things that sic de aHis.' might not be signified at all by any part of the original definition (namely the bodies that 27. See above n. 18. arc not while). And IIIis contradicts Thesis (3) as Ockham formulates it. 2R. In SL I, 31, OPh I, p. 94, the copula is defined as 'verbum copulans praedicatum cum .n SI. I. 26, OPh I. p. XX:'[)efinitio autcm exprimens quid nominis est oratio cxplicite subiecto'. It can be galhered from Ockham's own practice thaI this includcs nol only 'to declarans quid per unam diclioncm im[1l1rl(//III'.... (thc translation and italics arc minc). 100 OCKHAM ON CONCEPTS THE ROLE OF NOMINAL DEFINITIONS 101

In what immediately follows, Ockham mentions that a term like 'white' 'signifies the 45. SL I, 8, OPh I, pp. 29-30: 'Possunt enim utentes, si volentes, uti una dictione loco same' (significat idem) as its nominal definition. As we have previously seen, however plurium, sicut loco istius totius "omnis homo" possem uti hac dictione a et loco istius (in chap. 4, sect. 3), this does not mean that they should be synonymous in the strong totius "tantum.homo" possem uti hoc vocabulo b, et sic de aliis' (transl. Loux 1974, p. sense, but only that every significate of 'white' must be signified in some way by (some 65). part of) the definition, which is a much weaker condition. 46. SL I, 8, OPh I, p. 30. 34. See above chap. 4, n. 14. 47. See Tract. de Quant., q. 1, OTh X, pp. 5---45,esp. 20-39. See on this text (also called De 35. See for example SL I, 33, OPh I, p. 96: 'Aliter accipitur "significare" communissime Sacramento Altaris) Stump 1982, esp. pp. 221-30. quando aliquod signum ... aliquid importat, sive principaliter sive secundario, sive in 48. Ibid., p. 24: ' ... una dictio ponitur loco orationis vel alicuius compositi ex multis recto sive in obliquo, sive det intelligere sive connotet illud, vel quocumque alio modo dictionibus ...;' and a little further down the same page: ' ... non sunt praecise habentia significet .. .' (the italics are mine). virtutem nominis.' See also p. 30: '[...] hoc nomen "instans" et hoc nomen"generatio" 36. See Ord. I, dist. 2, q. 9, OTh II, p. 316: [N.B. - this passage is about denominative and ethuiusmodi non sunt nomina finita habentia finitas significationes, quasi significent quidditative terms, but the former, for Ockham, are all connotative, and the latter are praecise res quasdam pro quibus habeant supponere, ad modum quo talia nomina absolute]: ' ... omnis conceptus denominativus habet definitionem exprimentem quid "honlD", "asinus", "albedo", "nigredo", "calidum", "frigidum" habent finitas nominis, in quo ponitur aliquid in recto et aliquidin obliquo. Tunc quaero de una parte significationes.' Note that the pseudo-names in this last passage are contrasted not only [or: de prima parte, according to ms. E] illius definitionis: aut habet definitionem with absolute names such as 'homo' or 'albedo', but also with connotative ones, such as consimilem exprimentem.quid nominis, aut non. Si non, habetur propositum, quia talis 'calidum' ('hot') and 'frigidum'('cold'). Note also that not all pseudo-names, in necessario est quidditativus; aut habet talem definitionem exprimentem quid nominis, et Ockham's view, are precise abbreviations. In some cases,. they are only loosely quaero tunc de partibus sicut prius, et ita vel erit processus in infinitum vel stabitur ad connected with the complex expressions which they come from and they can be aliquem conceptum quidditativum praedicabilem de illo de quo primus conceptus expanded, then, only in the context of whole sentences, with maybe various explications denominativus praedicabatur' (italics are mine). That this argument has to do only with in various contexts. See, for example, how Ockham deals with the term 'generatio' in the first part of the definition, the term in recto - as explicitly proposed by manuscript Tract. de Quant., q. I, OTh X, p. 31: 'Sed ista propositio "generatio est in instanti" debet E- becomes clear in the last sentence of the quotation (which I italicized). What. recipi sub isto intellectu: .quando aliquid generatur, non generatur pars ante partem, sed Ockham requires is that this term in recto should ultimately be an absolute name totum simul generatur, ut ista propositio brevis "generatio est in instanti" ponatur causa predicable of everything of which the defined term is itself predicable (such as 'body' is brevitatis loco istius propositionis longae ... ' Ockham, in such analysis, comes very with respect to 'white', or 'animal' with respect to 'father'). close to the modem idea of a 'contextual definition'. 37. See above n. 24. 49. Ibid., p. 24: ' ... sine omni figura et improprietate sermonis non possunt esse extrema 38. Spade 1990, p. 605. propositionis distincta a copula'. 39. See below chap. 8, sect. 3. 50. Ibid., p. 30: 'Sed sunt quaedam nomina derivativa a verbis vel aliis partibus orationis, 40. We now see how it can happen, in some cases, that a connotative term ends up with no vel consimilia talibus, quae non sunt instituta nisi causa metri vel ornatus vel brevitatis primary significates at all. Take the figment term 'chimera', for example. Its definition locutionis.' See also Exp. in Phys. III, 2, OPh IV, pp. 425-6, and Quaest. in Phys., q. according to Ockham is 'animal composed of a goat and a cow' (SLI, 26, OPh I, p. 88: 63, OPh VI, pp. 568-9. (1999, pp. 84-6) has argued that 'in the 'animal compositum ex capra et bove'). The grammatical complements 'goat' and 'cow' interests of efficient functioning of the mental system', Ockham should also admit the in this definition respectively designate the two series of connotata ofthe defined term, and possibility of 'internal abbreviations' for complex mental phrases (p. 85; the italics are 'animal' plays the role of the term in recto, the extension of which is expected to include Chalmer's). This might very well be so, and whether ot not it would be a case of that of the defined term. In this case, however, there are no animals in the world which are synonymy in the strong Ockhamistic sense, the Venerabilis Inceptor actually seems to related in the required way with both goats and cows. The defined term, therefore, has no do so when he considers the possibility that certain simple mental acts are 'equivalent' primary significates. Note that, contrary to what Spade sees as a problem in this case to complex propositional acts (see the passages from Quaest. in Phys., q. 6, OPh VI, p. (1975, p. 72), nothing here requires that all the significates of 'animal' -namely all the 410 and Exp. in Perih., Prooemium, 6, OPh II, p. 358, quoted above in chap. 2, n. 86, animals there are - should be signified in any way by the defined term 'chimera'. Spade and the corresponding discussion in the body of the text). only thinks so because he (wrongly) assumes that whatever is signified by some parts of 51. See, for example, Brown 1996, p. 85 (about Ockham): 'Nominal definitions are, the definition should also be signified in some way by the defined term. however, essential to grasp the meaning of a connotative term. We could not conceivably 41. Most notably in SL I, 10, OPhI,pp. 35-8, SL I, 26, OPhI, pp. 88-9, SL III-3, 26, OPh be said to have the concept of a bachelor if we did not know that .bachelors are never I, pp. 689-91, and Quodl. V, q. 19, OTh IX, p. 553-7. married men or the concept of a father without knowing that a father is a male parent.' 42. Spade's interpretation, by contrast, could not account for the fact that many among the 52. More on this in chap. 6 below. most paradigmatic examples of Ockhamistic nominal definitions still include 53. See, for example, SL III-2, 28, OPh I, pp. 555-6: ' ...sed ignoratadefinitione exprimente connotative terms without Ockham expressing uneasiness about it in any way. quid nominis non potest quis cum alio disputare. Et ideo quando quis addiscit significata 43. See the passage from Quod/. V, q. 19 quoted above in n. 16. vocabulorum, tunc addiscit definitiones exprimentes quid nominis ...;'or SL III-2, 34, 44. Among absolute concepts, two necessarily coextensive terms would ipso facto be p. 570: 'Et est dicendum quod generaliter definitio exprimens quid nominis non potest synonymous. In the case of simple connotative concepts, if they do not have the same demonstrari de definito, sed ista praesupponitur omni demonstratione et omni connotata, God could always bring it about that their primary signitkates ditrer. And the syllogismo.' same is true, a fiJrtiori. for any pair composed of an ahsolute concept and

55. Ord. I, dist. 30, q. 5, OTh IV, p. 385: ' ... ad hoc quod unum sit simile alteri non plus Chapter 6 facit intellectus quam facit ad hoc quod Sortes sit albus vel quod Plato sit albus'. See Ord. I, dist. 30, q. 1, OTh IV, pp. 316-17: 'Et ideo concedendum est quod intellectus nihil facit ad hoc quod universum sit unum vel quod totum sit compositum vel quod Cognition and Connotation causae approximatae causent vel quod triangulus habeat tres etc., et sic de aliis, non plus quam faeit ad hoc quod Sortes sit albus vel ad hoc quod ignis sit calidus vel aqua frigida.' See also Quod!. VI, q. 25, OTh IX, p. 679. 56. SL 1II-2, 14, OPh I, p. 529: 'Nam inter res dicitur una prior alia: vel quia praecedit earn tempore, vel quia potest esse sine ea et non e converso, vel quia est perfectior ea.' As to what is going on in the mind, the result of the preceding chapter, admittedly, 57. See for example Quod!. VI, q. 16-20, OTh IX, p. 639-62. For detailed presentations of is negative: nominal definitions, contrary to what Paul Spade and others had thought, Ockham's theory of relations, see Adams 1987, chap. 7, Henninger 1989, chap. 7, are not meant by Ockham to reveal in general how connotative terms are mentally Beretta 1999, Michon 2000. 58. See Quod!. VII,q. 8, OTh IX, p. 729: 'Sed ille ordo [= spatial order] importat solum ipsa constructed out of more basic conceptual units. They are tools, instead, for the absoluta quae non faciunt unam rem numero, inter quae unum ab eodem plus distat et elucidation of the ontological import of such terms, and this role can be fulfilled aliud minus, et unum propinquum alteri et aliud distare plus vel minus sine omni quite independently of how the connotative names are acquired. The epistemological respectu inhaerente, ita quod inter aliqua sit medium et inter aliqua non.' It is remarkable significance of simple connotative concepts, then, still has to be brought out: how are that Ockham's rendering of the spatial order in this passage involves at least three they implemented? And what contribution are they supposed to make to our distinct ordered things (one of which being closer to - or further from - the other than knowledge of reality? Such will be the object of the present chapter. the third). The starting point, once again, will be provided by Paul Spade's stimulating 59. Ibid., pp. 728-9: ' ... dico quod ordo et unitas universi non est quidam respectus, quasi formulation of the problem (section 1). The basic idea for sorting out the quoddam ligamen ligans corpora ordinata in universo ad invicem, quasi illa corpora non Ockhamistic answers to Spade's questions, as we will see, is that the Venerabilis essent ordinata nec universum vere esset unum sine tali respectu secundum Inceptor accepts, in a variety of cases, an intuitive apprehension of ordered n-tuples, imaginationem Simplicii super Praedicamenta'. 60. On the truth-conditions of propositions with connotative names in them, see SL II, 11, which, given the way our mind is built, triggers the formation of simple connotative OPh I, p. 279-82. Such truth-conditions, Ockham explains, can be given by a concepts (section 2). How concepts thus acquired are liable to adequate nominal conjunction of exponentes. This is not, as we have seen, because only these explicating definitions of the sort described in the preceding chapter, will then become apparent propositions - the exponentes - are really present in the mind when such a proposition (section 3). This explanation, however, cannot hold good for all connotative terms. is uttered, but because, just like the nominal definitions in the case of isolated names, Many of them - such as 'father' - are not ascertainable on the direct basis of they make explicit the ontological import of the proposition as a whole: ' ... quaelibet experience and must, presumably, be introduced in some derivative fashion on the propositio quae habet talem terminum [= connotativum], est habens exponentes basis of more primitive terms. The question will have to be discussed, consequently, exprimentes quid importatur per talem propositionem' (p. 281; the italics are mine). whether Ockham's theory of concepts remains or not, in the end, a variety of the so­ 61. See below chap. 8, sect. 3. called Classical View, according to which, remember, the vast majority of our concepts are present in the mind in the guise of complex definitions (section 4).

1. Spade's questions

Although Paul Spade has recently granted that Ockham did in fact countenance simple connotative concepts, he still sees this as a mistake on William's part, given the rest of his theory. One of his main motivations for thinking so has to do with nominal definitions; this is what I have examined in detail in chapter 5. But it is not the only one. A second important reason he gives stems from the theory of knowledge. Here is, in his own words, what he takes the problem to be:

... there is another reason to be worried about the possibility of simple connotative concepts: they threaten to interfere with the epistemology of mental language ... Obviously there are many questions to ask here. What exactly does the mind do to produce these connotative concepts'! [fit does not usc absolute concepts as parts of the connotative concepls it produces. then why arc the absolute concepts required for this mental operation al all'! And exactly what sense are we 10 make out of the 'semanlic cOl1lpleltily' of un 104 OCKHAM ON CONCEPTS COGNITION AND CONNOTATION 105

metaphysically simple concepts? ... In my view, any full account of Ockham's theory of signifying concepts. Although we still have to explain how this is possible, there is connotation must deal with these epistemological issues, since Ockham himself certainly nothing deeply metaphysical in the issue. uses connotation-theory for epistemological purposes.' Let us now tum to Spade's second question. If simple connotative concepts do not have absolute concepts as their parts, he asks, why are absolute concepts Spade is undoubtedly right: these are questions that should not be avoided. Before I nevertheless required for the formation of connotative ones? Spade, it must be said, set out to answer them, let me explain a little bit more what their relevance is. raises this difficulty in connection with a particular hypothesis he . examines, Three different related questions surface in the passage just quoted: according to which simple connotative conceptswould be mere mental abbreviations for their complex definitions. They would indeed be simple intellectual acts, (1) How is the mind supposed to produce simple connotative concepts? according to this conception, but would be semantically equivalent to some complex

(2) Why are absolute concepts needed in the process? expressions - their nominal definitions namely -'.-..in which no connotative would (3) What sense are we to make of the semantic complexity of allegedly simple occur. In view of what we have seen in chapter 5, this hypothesis does not appear concepts? any more tenable than Spade'soriginal rejection of all simple connotative concepts, since it continues to assume that all connotative terms are ultimately .eliminable Let us start with the third one. Why does it arise? Well, Ockham's connotative terms, according to Ockham (contrary to what we have labelled as Thesis 1 in chapter 5),4 whether spoken, written or conceptual, do have in effect a certain semantic and that there is normally no interesting semantical discrepancy between a nominal complexity, in the ~ense,precisely, that they have a primary signification (they are ­ definition and the corresponding connotative term (contrary to Thesis 2). Yet the usually - true of certain things in the world) and a secondary signification or problem Spade raises with this second question is indeed a real one, for Ockham connotation (they - always - obliquely refer the mind to some other things). If we does consider thatthe formation ofa connotative concept presupposes in general the are to speak of simple connotative concepts, consequently, just as we speak of simple formation of at least one absolute concept. connotative words (like 'white' or 'horseman'), we have to reconcile this alleged A clue to this is given in distinction 2 of the Ordinatio, when Ockham strives simplicity with the semantic complexity which is taken to be characteristic of to show, in Scotus's footsteps, that there must be some quidditative concept connotative terms. The reason why Spade speaks of 'metaphysical' simplicity in this univocally applicable both to God and the creatures. One of the arguments he uses context is that Ockham, in his later theory at least (which is the one I am interested in the process explicitly incorporates the claim that the possession of a in here), treats concepts asreal things. They are seen as intellectual acts, and such denominative concept presupposes in general the possession of at. least one acts, for Ockham, are real singular qualities of singular minds. My concept of quidditative concept. Since denominative terms such as 'courageous', 'white', or 'horse' .is something real in me, and your concept of 'horse', dear reader, is 'creative' (which, in the spoken language at least, are concrete terms something real in yourself. Ockham, in fact, does not use any such phrase as morphologically related with corresponding abstract ones such as 'courage', 'metaphysically simple', but he does contrast simple concepts in the mind with 'whiteness', or 'creation') are all connotative according to him, and since complex sequences such as propositions or phrases. How, then, can semantically quidciitative terms are all absolute, this amounts to saying that the possession of complex concepts still be simple objects? some of the most typical connotative concepts presupposes that of some absolute This question, while relevant, is the least difficult of thethree and I might as well concepts. Here is part of the passage: indicate at once what the answer will be. To say that a concept is simple in Ockham's sense is not to say that it is devoid of any internal complexity. As we saw in chapter For example, I have this concept, that of 'creative being' namely, which I know to be 2, to say that a concept isa simple sign, in Ockham's vocabulary, is merely to say denominative, and therefore a certain concept must be prepossessed to which I attribute that it is not composed of several parts each one of which is independently endowed this one, saying, for example, that a Gertainbeing is Gfeative;and it is Gertainthat this with a signification.2 Compare withspoken words. A wordlike 'father', let's say, is Gonceptto which this one is attributed is not denominative, or if it is, ~itherthe process will said to be simple, in Ockham's Aristotelian terminology, because it is not composed go on infinitely or it will stabilize at some quidditative concepts.' of several other spoken words. This does not prevent it from having an internal structure. For one thing, it certainly displays some phonological complexity. And Anybody who has an applicable denominative concept such as 'white' or 'creative' even some sort of semantical complexity in so far as it can be broken down into a must also have an absolute concept of which the former is predicable within an stem and an ending, each one of which having a distinctive semantical role to play. affirmative particular proposition (such as 'body' is in the c;lse of 'white', or 'being' Yet it is simple if we contrast it with such phrases as 'the father of Socrates' or in the case of 'creative'). Spade's second question, then, is entirely appropriate: why complete propositions such as 'my father had a moustache', where several should it be so? What, in other words, is the rationale for this epistemological components independently have a representative value. The situation needs not be priority of absolute concepts over connotative ones? different with concepts. The expression 'metaphysical simplicity', which Spade uses As to the first question, finally - how does the mind produce simple connotative concepts'! - it is obviously the crucial point. Any account of Ockham's theory of here, is a bit unfortunate in the context.' To say of a connotative concept that it is concepts has to provide a detailed explanation for it. Fortunately, such an simple is only to say that it is not composed of several otherindependenlly 106 OCKHAM ON CONCEPTS COGNITION AND CONNOTATION 107 explanation is available on the basis of Ockham's own texts, and, as we shall see, the The whole process naturally starts when the agent gets acquainted with a physical answers to Spade's questions (2) and (3) smoothly follow from it. object - a horse, let's say - through the senses: this is the sensory intuition. The same singular objectis immediately grasped by the agent's intellect: this is the intellectual intuition, which allows the agentto evidently know certain contingent about 2. The acquisition of simple connotative concepts the object in question (that it presently exists, for example). A general concept is then formed: the specific concept of horse in our example, which requires only one How simple connotative concepts originally enter the human mind is one of these encounter with a horse. When the agent later meets with other animals, such as a cow issues on which Ockham changed position as he moved from the fictum-theory of or a dog, he or she forms a generic concept, that of an animal, for instance, or concepts to the actus-theory. What I want to present here, as in most of this book, is something in between like the concept of a mammal, all this, let me insist, being the the later - and final - position he adopted. But let me first say a word about the result of natural causal processes. 13 The question that now faces us is this: how can earlier one. It is very briefly expounded in a single passage of the Ordinatio, the formation of simple connotative concepts fit into such a picture? And the answer distinction 2, which has - to say the least - surprised the commentators who have I want to put forward is that in many cases - though not in all, admittedly - the noticed it. 6 Ockham there is discussing an objection raised against thefictum-theory process, for Ockham, will be very much the same as the one I just described, except of concepts: how can this approach account for the formation of syncategorematic that the intuitive grasping which is required as a starter should be an intuitive concepts, connotative concepts and negative concepts? the objector asks.? Ockham's cognition not merely of a single individual, as in the example of the horse, but of a answer is the same for all three categories: what the fictum-theorist has to say, he plurality of individuals simultaneously present to the agent's perception as a thinks, is that all of these are implemented in the mind as a result of its plurality. internalization of external language. The mind naturally produces simple intellectual Although often neglected, some of Ockham's texts are quite telling on the matter. representations of external words, as it does of any other objects it comes in intuitive Let us carefully read, for example, the following passage from the Prologue of the contact with; then it can, in addition, conventionally (ex institutione, Ockham says) Ordinatio: endow these particular mental units with the very grammatical and semantic'll features of the words they represent, and use them accordingly within mental Similarly, an intuitive cognition is such that when some things [please note the plural here) sentences.8 are cognized, one of which inheres in the other, or is distant from the other or is standing Strikingly enough, this conception shares one of the most salient features of in some other particular relation with the other, then straightaway in virtue of this Spade's original interpretation of Ockham: no simple connotative concept, incomplex cognition [note the singular] of those things [note the plural], it is known according to it, is naturally produced by the mind on the direct basis of intuition.9 whether the thing inheres or not in the other, or is distant or not, and so on for other Yet it ma,rkedly differs from Spade's reconstruction on several counts. For one thing, contingent truths ... For example, if Socrates really is white, then this cognition of Socrates and of his whiteness in virtue of which it can be evidently known that Socrates is it does not identify connotative concepts with their mental complex definitions, as white, is said to be an intuitive cognition. 14 Spade proposed, but with certain inner representations of words, conventionally used \ in special ways. And it does not require,. consequently, that a connotative term be The starter, in the latter example, is a simple intuitive grasping of both Socrates and strictly synonymous with its nominal definition. Whether ornot this approach is his whiteness simultaneously, and this is said to be enough for the cognizer to know compatible with the ineliminability of relational terms from certain nominal that Socrates is white. This strongly suggests that the simple intuitive grasping of the definitions (as stated by what we called Thesis I in chapter 5) is not obvious, but it two objects suffices to bring about the formation of the simple connotative concept might very well be after all, if the phrase 'connotative concept' in this passage is 'white'. taken - as seems probable to me - in what Ockham calls its na,rrow sense, according A related passage from the Quodlibetal Questions is even clearer: to which it excludes relational terms. lO If so, the explanation given in this passage would not be meant to account for the acquisition of rela,tional concepts, and the ... a relative concept is caused by both extremes, posited simultaneously, prior to conventional connotative words which are supposed to serve as starters for the composition and division ... Therefore, the order is as follows: When two whitenesses are implementation of simple connotative concepts might be thought to be originally seen, then, first, the specific concept of a whiteness is caused in the intellect; second, the introduced in the external language on the basis of nominal definitions including in concept of a similarity is naturally caused through the mediation of that specific concept. some cases both absolute and relational terms. II and, I claim, this happens immediately, from the whitenesses themselves, or from the Be that as it may, this theory was never developed l2 and Ockham eventually cognitions of them; and only after that, at least in the order or nature, is a proposition renounced it in favour of a more unified approach, according to which the formulated. 15 acquisition process for simple connotative and relational concepts is much more intimately related to the acquisition process for simple absolute concepts. This is I grasp two ohjects simultaneously, two whitenesses let's say. As usual, I naturally what we will now be interested in. form the relevant ahsolute concept, the concept of whiteness in this case. The Let us recall how we normally acquire ahsolute concepts, according to Ockham. important point. however. for our present discussion. is thai prior to any intellectual

'a

in

so

the

are

the

the the

the

109

this

this

two

part

than

It

This

ones

even

some

a

in

of

kinds.

of

of

which

nicely

like

always

merely

theory.

on

finding

this,

general

ordered

absolute

concepts

acquired

But

least

to

and

view,

which taller

one

compatible

'whiteness' of

more

connotative connotative

parts

at

time

of

whitenesses,

'white'.

And

one

natural are

being

way

of

no

n-tuples

relational obliquoin

should

able

simultaneously

simultaneously

the

the

concepts

each

concepts

of definition

is

produce,

real

things

them

hard

syncategorematic

Ockham's Ockhamreadily in

to

something

signification.

point

'black').

stock that

form

of

the

of be

By

that

naturally

least

a

-

their

on)

a

of

other

of

concepts

connotative

be

of

such

as

case

certain

at

concept.

able

and

are

so

a

make

then

all

to

concept.

of

parts

have

with

(those

basic

Because

I of

about

nominal

this

connotative,

might

in

are

to

naturally

group

are,

not

and connotative

they

simple

in

occurring

context

not

involved,

especially

concepts?

to

combination we

one

we definitions

our

concepts.

straightforward:

dog;

the

-

would

definition

the whitenesses

things

,Its

the

are

things

that·

does

particular

concepts complex

they

connotative

If correct

supposed

are endowed

.

epistemological

it:

that

process?

a

we

in

it

least is.

is

the

speculate

'simplicity'

that

why

is

formation

produce

formal

the

other

at world

white

projectible

this

on

be

concepts

in

the

some

or

coloured,

the

nominal

us

these

to

a

the

'whiteness'

then,

in

concepts,

the

'dog'

the

of

are

connotative

in

and

of

course,

-

however,

connotative

the

absolute

some

some

CONNOTATION

Ockhamism)

n-tuples.

see

of

'white'

From

occur

absolute,

things.

'white'

cognition.

reason

of

I that warns

definition,

to of

insisted

fortunately,

ability

of

absolute

of

particular,

significam;e

all

concept

not

to

similarly Because

from

simple

independently

AND

signifying

needed

triggers

the

simple

the

our that

concept

absolute in

ways.

questions,

is

out

Ockham

parts

the

or

not

of

concepts,

to

of

were

ordered

simply significates

to

does respectto

onhe

the

'whiteness',

complexity

to

from

which,

why

result

Suppose

individual

happens,

three But

but

them

of

even

repeatedly

(according

definition produce

require

or

one

of

similar

reason,

parts

so

and

with

understanding

concepts

than of

'dog'

involved

theory

them

simplicity?

clear

in

It

way.

'white~

not

be

not

representations

COGNITION

limited

no

absolute

itself.

and

there,

it

(the

example,

that

concepts

have

way

mind

kind,

things,

part

of

epistemological

to

'dog'

is

constructed

Spade's

secondary

are

according

semantical

contribution

correct

for

We

does

things

no

was

out plurality

'white'

the

- to

case

world!

alleged

make

a

be

absolute

18

whiteness').

another

a

These the

ordered

things

preclude

example,

general

the

Ockham's

a

major

concept

certain

concept

There natural

that

in

this

else

for

this

happen

all are

a singular

is

a

are

does view,

a

not

indeed,

of

their

(for

representations

concepts, definition

in

occurs

in

world

in

-are

answers

concept

answers

that

distinctive several

certain categorematic

they

gist

having have

does How

How

with Why ones. others grasping means

individual

concept.

is

way

needed

the

the

The

cannot

the

our

absolute world

when

concepts

is

connotative

Ockham's

(1) connotative

happens

nominal hasis.

of namely; that

(2) concepts,

If body

our

(3)

acknowledges.

something

are

must

general These

categorize

connectors.

in ordered

grasp

Their representations constitute

-

a

I

­

it

is

to

the

a

a

the

if

do

(or the

all

with just

see

one

then

and,

so

have

is

those is

parts

one

I

must with

them

black

dog

for

I

to

have

to'

It

'black'.

you

trigger ahsolute

a

ahsolute

If

always

form concept

whatit

that

'a

concept

it for

'taller'

Socrates,

of

relevant

not

primary

primary

16

dog;

least

naturally

example,

This

(if that

least

connotative

a

apprehend

the

and

yet

at

will the

ways

are

its

meeting

and

one

Similarly,

at

for

sentence,

even

one.

simultaneous such

And

is

already

the similar

meeting

of

of

anchored,

black

by

mind

concept

dog

relevant

concept.

occa,\'ion.l'

is

concept must

'blackness',

of

concept absolute a

sentence

same

on.

cases each by

Plato

be acquire

The

it

standardly

as

other

example.

the

example,

among

the

significate.

connotative

concept;

which

didn't

the

so

and

the

'being

part

one

black

examples.

a

with

I of

will

'similar',

Ockham,

such

for

with the

human

also

then

a a

for

which

concept

in

of

if

affirmative

absolute

serve

formation

formation

'black'

that

in

to

whitenesses.

to

and

quoted,

whitenesses,

and

consequently,

do

than the

chosen

see

part

will

latter

'similar', simple

way itself

then,

'horse'

Iwill

the

a

connotative

true

whiteness.

connotative

two

a

two is

refers

aran

just

or

a

the

secondary

meeting

no

concept

one,

they

the

meeting

you

of

whitenesses

it

concept

Suppose,

it

normal

merely

a

more

Plato,

not

member

in

similar in

of

When

in

relational

'black'

concept

by

17

In

as

concepts

the

of

situation,

with

'dog',

is

occasion,-

is

is

be

'black',

according

The

the

that

simple

of

Each

simple

of

and

the

of orit

'white'

addition,

first

passage

present,

is

a

just

this

Socrates-

same

part

on.

here.

concept

as

formation

in

'dog'

another,

same

similar' the

even

them

member

'black'

part

in

:man'

of

the

'dog'. concept. the

simultaneously

form,

perceived so

no

passage.

a

absolute

connotata;

this meeting

in

CONCEPTS than of

the

concept

the

possession

quidditative

of

preclude

another.

such

will

the

concepts

its

you

by

and

to

that

Socrates not

predicts

concepts acquired

a

form

such,

about,

from

ON

on

'being

predicable

concepts,

second

dog,

not

the

is of

involved

absolute

concept is

one quoted

taller

concept

or

concept

of

couples is

that

of

absolute

dog,

is

there

and

naturally

as

latter

leads

the have

acquired

formation

sense

member

the

one

contept

simultaneously

among

that

concept

theory

does

The

brings

black

from

that

one

absolute

the

presupposed

as

connotative

example

and

that

also

the

a

me acquired

one

yetthe

black

concept

have

Plato

Plato one

this OCKHAM

cases, I

connotative.

could

least

a

ordered

the

the

in

question

them

relation

the

example,

'whiteness'

I

previously

least

connotative

at

see Whiteness,

standard

that

this

with of

but why

possession connotative

a

in

have

and

of

in

concept.

at

or

are

distinct

for

other

objects

simple

could

concept,

'man';

process

in

many

particular,

or

acquired

tne

I

as

concept,

his

quidditative

connotative

them)

about

of

simple

the

coloured');

as see

with

to·

will

Ockhamistic

such

two

division,

In

the

in

in

simple The

I

'dog';

concept

simple

the

instead

this

the

have

and

concept

that

the

or

are

concept, stated

the

of

have

thus

distinct

part-whole

and

instead,

natural

bring

them

Socrates

intuitively

of

mind

simple

concept

Suppose,

the

however,that

true. 'taller'.

I

of

ones.

a

example, implies

to

who

that

connotative

No

concept, is

similarly

horse

two

what

can

concept

connotative

the

concept

follows,

apes

absolute

both

through

whitenesses,

already

one

Note,

the

normal

It

addition,

Ockham

Socrates

the

We

least

formation

in

ones.

black

two

absolute concept

see

beyause

agents

of The

say,

black' automatically

acquired

as

-

always

connotative

'being

in

not

horse;

absolute involved.

grasping

connotative

suffices

two

the apprehend

or

at

acquires

significate

presupposes

simple

refers

couples significates

general

simple

108 composition COGNITION AND CONNOTATION 111 110 OCKHAM ON CONCEPTS

3. The adequacy of nominal definitions (C) The nominal definition should be compatible with a sound epistemology. In the case of simple connotative concepts, in particular, this means that the How such concepts lend themselves to nominal definitions, even though they are not connection which the definition sets down between the primary significates of introduced by them, can now be more fully understood. The main thing, once more, its term in recto and that of its term(s) in obliquo, should be ascertainable on is that the triggering intuitions for these simple connotative concepts are graspings the basis of direct experience, at least in some cases. Otherwise, the abstraction of ordered n-tuples. The Ockhamistic picture of the mind supposes that a particular process could never leave the ground! couple, let's say, is naturally treated as a sample for all other couples that are similarly connected; Plato and his whiteness are taken as a (complex) sample for all Taken together, these three constraints make it clear that the adequacy of a proposed couples composed of a body and its whiteness; Socrates and Plato are taken as a definition for a naturally acquired concept cannot be established by simply sample for all couples composed of a taller individual and a shorter one, and so on. confronting it with the defined term, or with what the cognizing agents What a nominal definition should do in such cases is to make it explicit, in the spontaneously say about it. Given what the definition is for, it is bound to be heavily required way, to which couples exactly does a concept which is so acquired, theory-laden, especially with respect to semantics, ontology, and epistemology. Yet naturally extend. it is plain that the three criteria formulated so far do not suffice for establishing in More specifically, Ockham's theory is committed to the idea that it should be any particular case the adequacy of some unique definition. A different sort of possible in principle, for each simple connotative concept: condition is still needed: (D) The nominal definition should generally agree with how the defined concept is 1) to identify the natural group of individuals that all the primary significates of this concept belong to (this corresponds to condition Cl of a good nominal used in fact. Since concepts are not available for public observation, it has to definition, in the analysis put forward in chapter 5); be supposed for this test to be conducted, that a simple concept is usually 2) to delineate exactly the connoted group (condition C2); linked, for a particular agent, with a given spoken word or linguistic phrase. How the agents apply the word in concrete situations and what inferences they 3) to express, by means of the syntax and the syncategoremata, the connection that holds between the primary significates of the considered connotative make with it will count as evidence for how they use the concept. This indeed concept and the corresponding secondary significates (condition C3). seems to be what Ockham does in practice.

The definition that does that must not necessarily be known - or even recognized ­ What is required by this condition, however, can on~ybe general agreement. A as such by every cognitive agent who happens to have the relevant connotative subject can be mistaken in the application of anyone of her own concepts in some concept. 19 The role of an Ockhamistic definition, remember, is not to reveal what is particular occasions - or even in many occasions - and in the inferences she makes going on in the mind of the cognizer, but to tell us how the world should be for a with it. Condition D has to be used in constant conjunction with the other three. given concept to apply. What is required by the theory is merely that it be possible Ockham, 'as I believe, is committed to all of this, but none of it is very explicit in to construct such definitions. his works. So I will not pursue the matter any further at the theoretical level. Let me The real question that arises, then, is how the adequacy of the definition can be simply illustratehow it can work with a couple of examples. assessed in each particular case. If t\Ie cognitive agent cannot be expected to Take, first, the relational concept 'taller'. It is not one of Ockham's own spontaneously give us the nominal definitions of his own simple connotative examples, but it is quite revealing, for 'taller' is, presumably, one of those relational concepts, how are we to know that we get them right? This is not something Ockham connotative concepts that can be acquired on the basis of direct experience according has elaborated upon, but a number of criteria do emerge from what we have to the doctrine sketched in the previous section. In Ockham's perspective, 'taller' established so far. primarily signifies everything which is taller than something, and secondarily First, there are some general conditions to be considered: signifies - or connotes~everything with respect to which something is taller. In so far as it is an asymmetrical relational term, it must have a correlate, 'shorter' let's say (A) Formally, the definition should conform to the structure identified in chapter 5 (or something like it); since correlative terms interdefine each other according to and (normally) contain at least one term in recto and one (or more) Ockham,20 'shorter' should be one of the key categorematic terms in the complete grammatically .subordinate term(s}. nominal definition of 'taller'. The definition of 'taller', consequently, would be something like 'a body with respect to which there is something shorter'. Let us see (B) The nominal definition should be compatible with a sound ontology. In order how this formulation faces criteria (A)-(D). to play its role of clarifying the ontological import of the defined term, all of its For one thing, it does agree with criterion (A) in having the general form Ockham categorematic terms, whether in recto or in obliquo, should clearly apply to expects from a nominal definition: it includes a categorematic term in recto - 'body' entities which are acknowledged as such by a good ontological theory. In the - which applies to all the primary significates of the defined term, and another context of Ockham's doctrine, this means, of course, that they should clearly categorematic term in a grammatically suhordinate position - 'Isomethingl shorter' apply to individual entities. 112 OCKHAM ON CONCEPTS COGNITION AND CONNOTATION 113

- which applies exactly to the connotata of the defined term; the other components with their whitenesses) can intuitively be grasped as such according to him,27 and of the definition being syncategoremata.~l this would typically be the kind of experience that can bring about the acquisition of With regard to criterion (B), the proposed definition for 'taller' clearly agrees a concept such as 'white' (criterion C). The definition, finally, can be tested against with. what Ockham takes to be a sound ontology: the primary significates of its two the intuitions of the speakers, through the observation of how they apply such words categorematic terms are singular concrete things. Bodies certainly are at any rate. As as 'white', 'blanc', or 'albus' in concrete situations, and what inferences they are to shorter things, one can tum, if in doubt, to the definition of 'shorter', which, prone to make with them (criterion D). presumably, would be something like: 'a body with respect to which there is The simple connotative concepts acquired on the basis of direct experience are something taller'. Despite the striking circularity here, the recourse to this latter not identified, within Ockhamism, with their nominal definitions; they are not definition has the significant advantage of making it dear· that the primary introduced by way of them, and they do not have to be synonymous with them. Yet, significates of 'shorter' - and, consequently, the connotata of 'taller' - are also because of their naturally hierarchized semantical structure, they do have such singular bodies. A complete Ockhamistic nominal definition, remember, is not definitions, and these can, in·general,.be made explicit with the' help of a· good intended as a reductive or a learning device, but as a way of making conspicuous the semantical, ontological, and epistemological framework. What has to be assumed for ontological import of the defined term. Thatthiscan best be done in this case with this - as Ockham certainly does -is that the possession of a given simple concept is the help of two correlative definitions .is not a problem for Ockham.2~What the usually linked somehow with certain dispositions to apply it in concrete situations, example nicely reveals is that when two connotative terms are interdefinable in the and to use it within inferences. This does not mean that the simple concept itself is way that ~taller'and 'shorter' are taken to be, it is the terms in recto in the two reduced to a recognitional capacity or to an inferential role. Given its most important definitions that ultimately indicate the sort of things. that must be signified and function, which is to contribute in determinate ways to propositional acts, the connoted by the two terms (since the connotata of each one of them should be the specific identity of any given concept, whether connotative or absolute, is partly primary significates of the other, and the primary significates of each should be determined by the way its (primary and secondary) significates are fixed. In the case among the significates of the term taken in recto within its own nominal of simple connotative concepts of the type we have discussed so far, as in that of definition).~3These terms in recto, moreover, should normally be absolute terms,24 simple absolute concepts, this is done by the natural working of the abstractive and belong, consequently, to the categories ofsubstance or quality. process on the basis of experiential' intuitive graspings; The distinctive· feature of Criterion (C) also applies to the example, in so far as the case is in accordance simple connotative concepts, from this point of view, is that they are specifically with Ockham's epistemology. What is supposed, in particular, is that there can be in triggered by the grasping of ordered n-tuples. This is why, ultimately, anyone of normal human experience a simultaneous intuitive grasping of two things which is thern can in principle be defined along the lines set down in chapter 5. such that one of them· is seen as taller· than the other one, and the latter· as shorter than theformer. It is obv:iously the kind of case Ockham has in mind when he speaks of relational concepts that are 'caused by both extremes, posited simultaneously, 4. Ockham and the Classical View prior to composition and division'.25In such a situation, the two correlative concepts, as he says, 'simultaneously enter the intellect'26 and can legitimately interdefine each It is clear, however, that not all connotative terms can be acquired on the sole basis other. of intuition. Many of them, surely, correspond to connections between things that The proposed definition, finally, can be tested against criterion (D) if we suppose cannot be ascertained in direct experience, even when the relevant objects are all that the defined concept 'taller' is usually associated 'with a certain word, for physically present to the cognizing subject. Such is the case, saliently, for one of example, the word 'taller' for speakers of English. We can check that the speakers Ockham's favourite examples of connotative term: 'father'. Even if I simultaneously normally apply the relevant word in concrete situations to something which is such grasp President George W. Bush and the former President George Bush, let's say, that &omething else appears to them as shorter than it, and that they are prone to infer this cannot be enough for me to acquire a simple concept of 'father' if Ido not from 'xis taller than y' that y is shorter than x, and that x and yare bodies. already have one, since this simultaneous grasping of the two men is not sufficient Another sort of example is provided by the standard case of 'white', with its (by for me to evidently know that one of themis the father of the other. If most terms in now familiar) Ockhamistic definition: 'a body. haying a whiteness'. What is special our external languages are connotative, as comes out from Ockham's treatment of here with respect to the previous example is that the defined connotative concept is the Aristotelian categories, then this situation must hold for a great many· of them. In non-relational, and that it cannot be defined, consequently, by means of a correlate other words, the acquisition process described in section 2 can account, in the (since it doesn't have any). The adequaoy of the proposed definition, nevertheless, context of Ockhamism, for the introduction of only a portion of all connotative can be assessed in the light of criteria (A)--'-(D)just as in the previous case. The terms. This is not to say that it is not important. We have seen, on the contrary, that definition, obviously, has the required form (criterion A). It agrees with Ockham's such a mechanism is indispensable, in Ockham's framework, for the development of nominalistic ontology in requiring only singular bodies as primary significaies and appropriate mental representations of the world as it is ordered. But the question stilI singular whitenesses as connotata (criterion B). It also agrees with Ockham's arises: how are the other connotative terms acquired? And how are they represented epistemology, since the sort of couples that arc descrihed in the definition (hodies in the mind'! Must they nol, in particular. he introduced hy complex arrangements of 114 OCKHAM ON CONCEPTS COGNITION AND CONNOTATION 115 more primitive concepts? And if so, aren't we led back to an interpretation of The outcome is that whether they are introduced by abbreviations or by Ockham's theory of concepts as a variant of the so-called Classical View, according definitions, simple spoken or written connotative words of the sort we are now to which the vast majority of our concepts ate in fact complex definitions? discussing should be represented in the mind in the guise of complex conceptual What we have to find out, to answer this, is what other modes of introduction combinations. Although a bit more complicated, this comes very close to what the there can be for connotative terms within Ockhamism, besides the intuitive grasping so-called Classical View is taken to hold, except that when nominal definitions are of n-tuples. As far as I can see, at least two different - but noncexclusive - such involved, strict synonymy is not required between the simple external word and the procedures can be accepted by Ockham: abbreviation and definition. corresponding mental sequence. Despite this non-neglectible difference, it still will As to abbreviation, Ockham admits that the speakers of a language can always, if be supposed in the Ockhamistic framework that a large part of the simple they wish, 'use one locution in place of several'.28The result of this procedure will not categorematic terms of our external languages receive their meanings from complex in every case be a connotative term, but it can be ifthe abbreviated phrase is originally constructions out of more primitive simple concepts. structured in such a way as to be capable of occurring in personal supposition as the We are not brought back to Spade's picture of Ockhamism for all that. The main subject or predicate of a well-formed sentence. If a complex phrase such as 'man problem with this picture was that the only categorematic concepts it accepted as necessarily' is abbreviated into a single word (as Ockham admits it can be),29 the primitives in the Ockhamistic constructional processes were the absolute ones. It resulting abbreviation will not be a true connotative name since the original complex would have followed that the meanings of all relational terms should be constructible phrase is not capable by itself of being subject or predicate of a well-formed sentence; on the sale basis of absolute terms (plus syncategoremata). Which was doomed to such expressions, as Ockham says, 'do not have precisely the value of a name'.30 A failure. What we arrive at now is entirely different. Among the simple concepts phrase such as 'black horse', on the other hand, can occur as subject or predicate; if naturally implemented in the human mind on the direct basis of experience, there are we should abbreviate it into a single word, the resulting neologism -'blorse', let's say also some connotative terms for Ockham, relational ones especially. And this allows - would be a perfectly good connotative term. When Ockham speaks of this for a much more powerful natural basis than Spade and his followers thought, with abbreviative device in Summa Logicae I, 8, he presents it as holding between linguistic a much better prospect for Ockham's nominalistic program thus reinterpreted. units: a complex spoken or written phrase, he submits, can always be abbreviated into In a recent paper, one of Spade's pupils, Yiwei Zheng, has expressed the worry a single word. But there is no reason why it could not hold between a complex mental that as a result of my acceptance of simple connotative concepts in Ockham's phrase (for example, the complex mental expression corresponding to the English epistemological basis, I'd be led to abandon 'the idea that Ockham's mental words 'black horse') and a single spoken or written word. In such cases, the external language has a recursive semantics', an outcome which Zheng sees as 'clearly connotative term would be represented inside the mind in the guise of a complex undesirable' .32But why should I be so led? What a recursive semantics requires is concept, and it would be strictly synonymous with this concept. that there be both primitive and derivative terms in the considered language and that Nominal definitions, on the other hand, can also be accepted within Ockhamism the semantical properties.of the derivative ones be obtained from that ofthe primitive as a device for introducing connotative terms. Nothing prevents us from constructing by means of a limited number of (indefinitely) repeatable constructional rules. This complex expression having the required Ockhamistic form for a good nominal is indeed what we arrive at. The difference with Spade's original proposal in this definition and, then,. to conventionally strike a word of which this complex respect is that Ockham's primitive basis, as I see it, is extended to include a number expression would be the definition. Such cases would importantly differ from the of connotative terms (those that can be acquired on the basis of intuitive graspings) previous ones in that the introduced term would not necessarily be synonymous with in addition to·· the absolute ones. Even though they are definable, these basic the complex phrase that is used to introduce it, since, as we repeatedly insisted, an connotative terms are not reducible to more primitive concepts, since their Ockhamistic nominal definition is not synonymous, usually, with the term it defines. definitions can circularly involve certain terms which they themselves serve to The definition indicates what the primary and secondary significates of the newly define. But many other connotative terms are still taken to be engendered through introduced· term are, but some of its components can signify or connote certain constructional devices (abbreviations or nominal definitions) which can be governed things that are neither signified nor connoted by the defined term. 'Father' is a case by recursive rules. Just how many simple conrtotative concepts can there be, or what in point. Since it cannot be acquired on the basis of direct intuition, and since it has, their proportion should be, is left open in Ockham's own texts. All we can say is that according to Ockham, a good nominal definition ('[male] animal having a child'), there should be at least one for each sort. of connection between individual things we may conjecture that it can be introduced in a language by way of this definition. which is independently ascertainable ort the basis of direct intuition; a finite quantity If so, it will not signify every single thing that is signified by at least one component presumably. The other ones are to be constructed somehow, and, as far as we know, of the definition (it will not signify all animals, for example), and it will not, an of them might very well be possible ... consequently, be synonymous with the definition in Ockham's strong sense of The important point for Ockham's programme is that the admittance of simple syrtonymy. It· will not be said, in particular, to be subordinated to its mental connotative concepts into the basis of the epistemological system does not definition, in Ockham's technical sense of 'subordination'.31 Yet it would have a complexit'y the ontology, not more at any rate than what is acceptable for him. determinate meaning in the external language, while being represented in the mind Simple connotative terms have primary and secondary significates, but given the by a non-synonymous complex phrase. way these terms are acquired. all of these arc singular things. Spade and many other 116 OCKHAM ON CONCEPTS COGNITION AND CONNOTATION 117 commentators have been presupposing that the ontological elimination of special relative terms (see, for example, the Pseudo-Carnpsall, Logica 9, ed. Synan 1982, p. 102: entities from the categories other than substance and quality was only possible, in ,... differencia est inter connotativa et relativa ...)'. It is relevant to note that this the context of Ockhamism, if all the terms in these categories '- and especially all the particular passage of the Ordinatio was written early in Ockham's career and that the relational ones - could be reduced through constructional devices such as technical word 'connotativum' is introduced in it via an objection. 11. definitions, to absolute terms from the categories of substance and quality.33 But this In Exp. sup. Elench. II, 16, OPh III, p. 302, Ockham makes it explicit that relational terms occur in the nominal definitions of some non-relational connotative terms: 'Ex is not how Ockham's nominalism works. The ontological import of a given term, for quo patet distinctio inter nomina relativaet connotativa. Sequitur autem ex isto quod non the Venerabilis Inceptor, depends on what entities it is supposed to signify, whether omne nomen in cuius definitione exprimente quid nominis ponitur nomen relativum est primarily or secondarily Irreducible connotative concepts, therefore, can be nomen relativum.' admitted if it can be shown ~through their nominal definitions - that they signify or 12. This conventionalist theory of connotative and syncategorematic concepts is allusively connote nothing but singular substances or qualities. referred to again in a few other passages. See in particular Ord. I, dist. 3, q. 2, OTh II, p. 403: 'Tertio, dico quod [Deus] est cognoscibilis a nobis in conceptu aliquo modo simplici ad placitum instituto ad significandum, et hoc in conceptu connotativo et Notes negativo sibi proprio'; and - more surprisingly, since this is supposed to be a later text, but also more restrictively, since it bears only upon mental syncategorematic terms - in 1. Spade 1996, pp. 238-9. Quodl IV, q. 35, OTh IX, p. 471: ' ... signa mentalia ad p1acitum significantia ..., puta 2. See above chap. 2, sect. 4.1. syncategoremata mentalia'. 3. See also Zheng 1998, whose discussion of Ockham's theory of connotative concepts 13. See the summary of the process given in SL III-2, 29, OPh I, p. 557, with, in this case, equally builds upon this distinction between metaphysicaland semantic simplicity. the example of a man as the known object: ' ... sed iste est processus quod primo homo 4. See above chap. 5, sect. 1. cognoscitur aliquo sensu particulari, deinde ille idem homo cognoscitur ab intellectu, 5. Ord. I, dist. 2, q. 9, OTh II, pp. 315-16: 'Verbi gratia, habeo istum conceptum, puta esse quo cognito habetur una notitia generalis et communis omni homini. Et ista cognitio creativum, quem scio esse denominativum, et ideo oportet praehabere unum conceptum vocatur conceptus ... Deinde apprehenso alio animali ab homine vel aHis animalibus, cui istum attribuo, puta dicendo quod aliquid ens est creativum; et certum est quod iste elicitur una notitia generalis omni animali ... ' conceptus cui iste attribuitur non est denominativus, vel si sic, erit processus in infinitum 14. Ord. I, Prol., q. 1, OTh I, p. 31: 'Similiter, notitia intuitiva est talis quod quando aliquae vel stabiturad aliquem conceptum quidditativum' (italics are mine). res cognoscuntur quarum una inhaeret alteri vel una distat loco ab altera vel alio modo se 6. See Ord. I, dist. 2, q. 8, OTh II, pp. 285-86. Normore 1990, p. 59, finds the doctrine habet ad alteram, statim virtute illius notitiae incomplexae illarum rerum scitur si res expounded there simply' astonishing'. See also Adams 1987, p. 301. inhaeret vel non inhaeret, si distat vel non distat, et sic de aliis veritatibus contingentibus '" 7. Ord.. I, dist. 2, q. 8, OTh II, p. 282: 'Quartum dubium est de conceptibus Sicut si Sortes in rei veritate sit albus, illa notitia Sortis et a1bedinis virtute cuius potest syncategorematicis et connotativis et negativis: unde possunt sumi vel abstrahi?' evidenter cognosci quod Sortes est albus, dicitur notitia intuitiva' (italics are mine). 8. Ibid., pp. 285-6: 'Possunt autem tales .conceptus imponi vel conceptus abstrahi a 15. Quodl. IV, q. 17, OTh IX, p. 386: ' ... nam conceptus relativus causatur ab utroque vocibus ... Tunc ab istis. vocibus sic significantibus abstrahit intellectus conceptus extremo simul positis ante compositionem et divisionem ... Unde iste est ordo, quod communes praedieabiles de eiset iinponit istos 'conceptus ad significandum illa eadem visis duabus albedinibus, primo causatur in intellectu conceptus specificus albedinis; quae significant ipsae voces extra. Et eodem modo et de talibus format propositiones secundo, naturaliter mediante ilIo conceptu specifico causatur conceptus similitudinis, et consimiles et habentes consimiles proprietates quales habent propositiones prolatae.' For hoc dico immediate ab ipsis albedinibus vel a cognitionibus earum; et post saltern ordine a detailed analysis of this approach in so far as it applies to syncategorematic concepts, naturae formatur propositio' (transl. Freddoso and Kelley 1991, pp. 317-18, slightly see below chap. 7, sect. 1. modified). 9. Ibid., p. 285: 'Ad quartum dubium dicerent [or dico, according to ms. A and F] quod 16. In the particular example under consideration, the relevant connotative concept - that of conceptus syncategorematici et connotativi et negativi non sunt conceptus abstracti a 'being similar' - happens to be symmetrical, so that, presumably, all of its primary rebus ex sua natura supponentes pro rebus ... Et ideo dicerent [or dico, according to ms. significates are also found among its connotata (everything which is similar to A, B, C and F] quod nullus [est] conceptus syncategorematicus nec connotativus nec something is such, that something is similar to it), but this cannot be seen as a necessary negativus, - nisi tantum ex institutione ... ' feature of all simple relational concepts. 10. See Exp. sup.. Elench. II, 16, OPh III, pp. 301-2: 'Large omne nomen vocatur 17. See the text from Ord. I, dist. 2, q. 9, quoted above in n. 5. connotativum in cuius definitione ponitur aliquid in recto et aliquid in obliquo, vel 18. See above chap. 5\ sect. 3.3. verbum vel alia pars orationis ... Isto modo accipiendo hunc terminum "nomen 19. See above chap. 5, sect. 3.2. connotativum", est in plus quam "nomen relativum"; quia omne nomen relativum est 20. See SL III-3, 26, OPh I, p. 690: 'Et tali definitione relativa mutuo se definiunt.' See nomen connotativum ... Aliter accipitur nomen connotativum stricte ... Omnia autem above chap. 4, nn. 30-31, and chap. 5, sect. I, Thesis 1. nomina connotativa, large accipiendo connotativum, praeter ista quae stricte vocantur 21. 'In respect to which there is' can be treated in the proposed formulation as a single connotativa, sunt nomina relativa' (italics are mine). I deem it probable that Ockham syncategorematic asymmetrical preposition or copula, just like 'having' in Ockham's uses the narrow sense in the passage from the Ordinatio we are currently considering, detinitions of 'white' ('a body having a whiteness') and 'father' ('an animal having a because it is the simplest interpretation in the conlexl. Except in Ockham's own logical child'). This point, of course, requires further elaboration, but, as I said earlier, Ockham works, where he remodelled the notion of a connotative lerm and gave it both a wider docs not provide any detailed theory for such syncategorematic connectors. More on this in chap. X, sect. .1hclow. extension and a greater weight, it was usual al Ihe time 10 distinguish l'onnolalive from 118 OCKHAM ON CONCEPTS

22. Cf. SL 111-3,26, OPh I, p. 690: ' ... non est inconveniens si mutuo definiant se'. 23. Thisfollows from conditions Cl and C2 in chap. 5, sect. 2 above. Chapter 7 24. See above chap. 5, n. 36. In the examples presently under consideration, the relevant absolute term in recto is 'body' in both the definition of 'taller' and that of 'shorter'. It Concepts as Similitudes is true that Ockham sometimes lists 'body' (corpus) as a connotative term (for example, in SL I, 10, OPh I, p. 37, SL I, 44, OPh I, p. 139, SL I, 45, OPh I, p. 142), but he uses it elsewhere as an absolute term (for example, in SL 111-2,14, OPh I, p. 530, where 'corpus' is attributed a real definition - 'res composita ex materia et forma' - typically suitable for an absolute term). The word, therefore, has two different senses in his Faithful to the Aristotelian tradition, Ockham holds that a concept is a likeness of writings. When it occurs in recto within a nominal definition (such as that of 'albus'), it should be taken in the second of these. whatever it represents, a similitude.! Many commentators have seen this as a 25. Quodl. IV, q. 17, OTh IX, p. 386; the text is quoted above in n. 15. problem: what can it mean in the context of Ockhamism? Is it a mere far,;on de 26. SL 111-3, 26, OPh I, p. 690: ' ... relativa sunt simul in intellectu'. parler? Or even an incdnsistency on William's part? Or is there some philosophical 27. See the text from Ord. I, Prol., q. 1, quoted above in n.14. interest, within his nominalism, to this notion of the concept resembling its 28. SL 1,8, OPh I, p. 29-30. See above chap. 5, sect. 3.1. significata in some sense? 29. Ibid. Marilyn Adams, for one, devoted an important development to the question in her 30. Tract. de Quant., q. 1, OTh X, p. 24. See above chap. 5, nn. 47-50. landmark book of 1987: 2 'I do not see', she concluded, 'how Ockham can specify, 31. A word is subordinated to a certain concept in Ockham's vocabulary when it is imposed either on the objective-existence theory or on the mental-act theory, a similarity to signify the very things that are signified by this concept, and to signify them in exactly relation that can constitute the natural signification relation for general concepts the same way as the concept does. See SL I, 1. More on this in chap. 9 below. such as "animal" and "man".'3 32. Zheng 1998, p. 255. A couple of years later, Pierre Alferi went further and claimed that strictly speaking, 33; See, for example, Spade 1998, p. 356. Ockham is inconsistent in accepting the vocabulary of similitude within his mature theory of the concept, and that he yields in so doing to an unrigorous way of speaking, a mere remnant of an old doctrine that he had himself rendered obsolete.4 Joel Biard even suggested in 1989 that Ockham, in fact, abandoned the vocabulary ofresemblance when he turned to the actus-theory.5 While admitting in a later work that this was somewhat of an exaggeration, he still strongly stressed the prominence ofcausality over similitude in Ockham's final theory of natural signification. 6 Both Cyrille Michon and I have discussed the point in some detail in the 1990s, and conc1udedthat similitude was always considered indispensable in Ockham's epistemology;7 but we still underestimated, I am afraid, the interest and importance of the idea in Ockham's own eyes. I would like to revisit the whole issue in the present chapter. As in most of this book, 1 will concentrate on Ockham's mature theory, where concepts are identified with mental acts. We will see first that the notion of conceptual similitude is still importantly present in this phase of his thought, with significant philosophical roles to play (section 1). In the absence of explanations from Ockham himself about what this means exactly, I will propose a tentative interpretation (section 2), and apply it to the main categories of general concepts countenanced by the actus-theory (section 3). This will put us in a position, finally, to discuss in some details two intriguing puzzles recently raised about Ockham's theory of absolute concepts (section 4).

1. Similitude sustained

That the concept is a likeness of the thing represented is an integral part of Ockham's former theory. How can a purely ideal object such as a jictum represent anything within a given mind? The answer was that it is a kind of intellectual picture:

I III 120 OCKHAM ON CONCEPTS CONCEPTS AS·SIMILITUDES 121

... seeing something outside the· soul, the intellect fonns something similar within the or to be a component of a proposition, or to be common to external things etC., and all of mind, in such a way that if it had a productive capacity, just as it has a representative this suits the intellection better than the ideal picture. This ideal picture or fictum, capacity, it would produce such a thing outside itself in real being, which would be therefore, is superfluously posited.14 numerically distinct from the prior one ... And this [= thefictum in the mind] can be called a universal, for it is a portrait and it indifferently refers to all the [relevant] external The possibility for a general·concept to stand for real external things within a individuals; and it is in virtue of this similitude in objective being that it can supposit for proposition is here said, once more, to hang upon the resemblance of the concept external things ...8 with the things in question. But an act,.Ockham says, can be more similar to any real thing than afictum can, since it is itself a real thing. Elizabeth Karger, in commenting on this passage, suggestively says that Ockham's It is true, as Michon stresses, that Ockham had put forward a reply to this argument fictum is 'a sort of purely ideal blueprint of the thing'.9 It has what could be called in a previous discussion of the fictum- versus the actus-theory.'5 This was in the an intentional resemblance with what it represents, and. this resemblance accounts Prologue to his commentary on Aristotle's Perihermeneias, a text in which he still for the semantical features of the concept, its capacity for supposition in particular. seemed to hesitate between the two approaches. His rejoinder, then, was that a real This seems unproblematic within the fictum-theory because the concept, in this resemblance is not needed in order to account for the conceptual functions; a framework, is supposed to have no other mode of existence, precisely, than that of resemblance 'in intentional being' is what is relevant for the concept to be able to an intelligible portrait. When Ockham abandoned the fictum- in favour of the actus­ supposit for a thing, or to be common, etc. 16 Yet, the conclusion he reaches in the theory, however, he did not renounce the idea that the concept is a similitude. Several Questions on the Physics, is that a real resemblance is even.better suited. And the texts of the mature period testify to it. whole discussion, anyway, presupposes on both sides that the likeness of the concept In Quodlibet IV, 35, saliently, the Venerabilis Inceptor, having criticized the with the thing - whether it is a real or a merely intentional similitude - is the condition fictum-theory, reaches the following conclusion: of possibility for the referential functions of general concepts to be fulfilled. In the Quodlibetal Questions, moreover, the idea that the concept is a similitude Therefore, I claim that both a first intention and a second intention are in reality acts of is· the basis for another important philosophical argument. Ockham uses it as the understanding, since whatever is preserved by appeal to a fictive entity can be preserved main to show that there are no simple singular concepts in the human mind. by appeal to an act of understanding. For like a fictive entity, an act of understanding, (i) Since the representational function of the concept depends upon its resemblance is a likeness ofan object, (ii) is able to signify and supposit for things outside the soul, (iii) with certain real things, he argues, any given concept must equally represent all the is able to be the subject or the predicate in a proposition, (iv) is able to be a genusor a individuals which are maximally similar to each other, and it cannot, consequently, species, etc. to singularize any of them in particular. See the following extract: The argument is straightforward: all the representational and semiotical functions we might want to attribute to a concept can be fulfilled by the act of intellection as well Fourthly, I claim that our intellect cannot have any such proper and simple concept with respect to any creature, either with or without a vision of the creature. And this is because as by the fictum; since the act is indispensable anyway, as everybody admits, the 11 each such cognition or concept is equally a likeness of, and equally represents, all exactly Razor Principle requires the elimination of the fictum. Now, among the conceptual similar individuals, and so it is not more a proper concept of the one than of the other. 17 functions, the first one Ockham mentions is similitude: an act, he says in so many words, can be a similitude of the object represented. . The same argument is found in another passage of the Quodlibetal Questions, which This is not a slip or an isolated statement. One can turn, for example, to another is even more interesting for our present purpose: passage where Ockharn elaborately discusses the fictum- and the actus-theory and clearly decides in favour of the latter: the Questions on Aristotle's Physics. Third, I claim that the first (by a primacy of generation) simple abstractive cognition is not Questions 1-7 of this work are dedicated to the ontological status of the concept. a cognition proper to the singular but is sometimes, indeed always, a common cognition Ockham, there too, explicitly subscribes to the idea that 'the concept is a similitude ... [This assumption] is evident from the fact that no simple abstractive cognition is more of the external thing' .12 As Cyrille Michon has remarked, he even turns the point into a likeness of one singular thing than of another exactly like it; nor is it caused by, or apt an argument·in favour of the actus-theory.13 Of the seven arguments against the by nature to be caused by, [just one] thing. Therefore, no such cognition is proper to the .fictum listed in question 1, the seventh - and longest - is the following: singular thing; rather, each such cognition is universaP8

Seventhly, such an ideal picture [idolum] would differ more from a thing, than whatever The 'first abstractive cognition', in Ockham's vocabulary, is the one that is generated thing from another one, since a real being and a rational being differ more from one in the mind on the heels of the singular intuitive cognition. I see a nuthatch for the another than any two real beings; this is why such a picture would less assimilate to a first time, let's say. Once the intellectual intuitive cognition occurs, a concept is thing, and consequently be less able to. supposit for a thing, than the intellection [or immediately generated, which abstracts from the existence of this concrete small intellectual act] which assimilates more to a thing; and the ideal picture will less be bird in front of me. What Ockham is now telling us is that this first abstractive common to external things, and it will less have the status of a universal than the cognition is already 1(enem/. It is not a singular representation of this particular intellection. But such a picture is called for for no reason except for suppositing for a lhing.

a

a

in

of

­

an

to

this

for

as

be

for

this

not

state

idea

the this

not

two

123

they

order mind.

these

Other

or

action

thing,

in

act,

thing.

simply

to

is

at

colour sense,

object.

to

idea

or

in

the

suspect,

general

quality

it.

dog,

a

an

dynamic.

the

an

although

I represent

a does

a

the

one

between

a

what

applies

Ockham's

can

'act'

said

this

discussion

superlative

looking of

is is

for

photographs

be

'maximally

not intrinsically

as

aside

sense

either,

with

accidental

like

It

all,

to

animals;

in

not

mind

classified

For

be

similar

and example, will

is

actualized

to

the are

term

discussing

or

like

at, put

be

the

both

This

are

not

for

an

Ockham's

usually

because, after

genus

-

being represented

does

nuthatches,

model

present

we

quality

be

the

it

horse

by

similitude,

have action.

crucial a

can

look

thing!

mammal they

like redness,

technical when

A

the

both

the at

by

that

resemblance

men,

two

Ockham

encounter

is

could

a

a

representation as

soul,

directly upon

same

an

presupposes

right

object.

a other,

features,

of

for

and

red

-

things

be

it

photograph?

qualitatively

this

for

not

a

concept

or

relation

they

more the

but

Two

not

the

an

mind

a

and

from

could

elsewhere,24

the

a

in

is

be

25

misled

qualitative

looked of

is

act

does

each

two

is

or

in

relation.

of

itself

the

other,

particular, through

to

scholars

could

vocabulary:

not It

What like

it

one

imperatively

it

to

be

be

of

of

have,

classified

similarity

it

likeness

is

understanding

ontology

in

since

not

to

gesture

a

perceptual

is

said

'what

but

each

not

or

hand

that

mental

excluded

a

have

graded species many

must

for

whiteness explained

as

to

a

this

above.

however,

be

acquire

similarity

a

not, act

might

any

represented:

grasping

be

to

speaks

degree

the

need

A

anything

2,

matter.

specialissima,

as

Ockham's

a

similarity

horse,

essential

that

specialissima.

movement. should

same have

he

image

other,26

be

representation

itself

not

they

SIMILITUDES

model

in

a

Ockham's

does

I a

obviously

with that

thing

seem

question:

represents,

not

general

we

the

but

should

act

difficulty

accepted

in

It

not,

AS or

is

manual

similarity

As

it

maximal

that

identified

not

chapter

each

for

in

conclude,

necessarily

the

does

to is

This

term

when

mind

species

it

essences.

moreover,

all,

or

species

intellectual

in

the

to

that

essential

it

the

by

not

been

mind

the

understood

not

in

Yet

incisive

takes.

sense

the

horse

identical

conceptual

resemblance that

perceptible

same

gesture, quality

is

same

a

should

it the

muddled,

And

essential

as

a

picture.

Above of

does a

display, whatever

of

have

role

intellectual

similar

not

Ockham,

technical

substance

could

of

CONCEPTS

is

the

it

the

belonging

simillimae.

explored.

any

maximal

(singular)

or

commentators of

we

a

a

explained

is

similarity

an

whitenesses:

just

not

Biard's

to

for to

of

sort and

presupposes

be

forms

to

Nor

It

similarity,

of

could

exemplified

to

it

since

posture

mean?' comparison,

their

to

non-maximal idea was

upon

realistic

concept however.

two even

some

other:

act,

Aristotelian

What quality

does

sense,

is

a hand

a

a

could

a

act

third

epistemology.

act

a

virtue

As

it

are

substances.

the

Whether

belong

maximally

are, the not

sense

a

belong

the

or

similitude

sense.

the

when

do.

is

noncmaximal all,

for

concept

thing,

in

even

return

an

each

Ockham

amounts

is

to

is

in

is

a

and

be

have

essential

a

anything.

depend

of

which

not

and mind,

non-neglectible

to

us It.

red

Or

modern

to

mental

although

of

to

and

a

which

one

grasping. after

not

a

do

second

which

only

have

intellectual

sort

the

colours.

statues Let

the

There

represent;

-

said

A

a

the

The

quality

Consider, Ockham's

of

in

formulation.

thing

conceptual

possibilities

or

the to

that How,

photograph issues. stage, plays

animal,

something

Not

also between

are ontology

intrinsically

does

similarity they

example,

notion

restricted

horses.

things be

similar'

simillimus

is

of

similar become

act represented

.

to

to

a

a

of

he

of

in

the

be

in all,

was

is

is

th~

one,

since

with

was

The

much

horse

to

point at

idea

which

The

mental

green

similar

both

'strictly

quality,

is

virtue

a

habit

concept example,

respect after

similar

the

use,

it

be

Biard

this

'the

could

though

in

and

disturbing

takes

Ockham,

which

between Sentences.

to

for

at the

the with

is,

be

concept

times.

yet

in

quality

-

(qualitas)

and resemblance

with

Joel

it

a

in

to

us

to

are

a

supposed

even

acceptation

the

irrelevant

language,

How

a

15-19:

simultaneously

nuthatch.

common said

Aristotelian

-

it;23

by

fortiori,

understanding

of

as

a

own

be

on

is not

redness

shirt

of

thesis said

having

A

thing

contends,

restrictive

have

lla

This

upon

does

is

obviously

something

our

uses

These

'quality'

whiteness,

mental be

world?

.22

qualities

out

should

sh:;tre

8,

raised

he

This

he

this

resemblance

contexts,

would

A

standard

both

import

of

of

while

what

a

Biard

that

the

can sense.

'nothing

to

discarded

concept

knowledge

the

is in

possible

corresponding

mind

Ockham's

sort.

such

andB

but

whiteness,

they

further

(ens)

Since since

quality'

be

is

commentary

incumbent sharply

A

a

represents in

theory

frequently admit

act

in

strict

hammer?

the

if

is

concept

that

qualities.

and to

his

a

have

same

elephant,

Categories

every

B must

in

general

quality

well-thought

in

been

submit,

this

'being'

in

according

substances

a

that

a

an

or

I

in

might

not

actus-theory,

the

have

conceptual

raised

him, mental

thispassage,

is

in

instance,

of himself

has

thing

even

resembling

obiectivum,

general

representational cognition

do

every of

inadequate.

other

we

in

the

a

task

and

to

on

Ockham's

Only

the

for

quality

-

1330s

short

this

to according

CONCEPTS

to to

sense

a the

of

been

the

orchid

the

esse

For

still

by'similitudo'

in

each

suggest,

similar

himself.

-

resemblance

see,

concept

ON

whiteness,

Ockham

is

an

19

could

question

early

an

quality

for

apply,

matter

substances

that

of

naturally

be

qualities,

except

a

could

butterfly similar

the

similar

concern.

fictum-

sense.

and

similitude

abstractive now

the

actus-theory.

a means

not

critique

The

much

add,

with that

is

already to

as

only

everything

with

commenting

the

else

of

its as

an

it

A

we

mean?'

he

Aristotle's

how

resemble

another

have

is,

primarily

in

for

quality

generality

notions

of later

of

does

OCKHAM

Ages,

said similitude

is accounts

As to

harsh act

present

a

might

that

role

than

from

naturally

soon

a

And

thing this nuthatch, what

for

his

of

both they

says

be

themselves

there

an

is complained,

resemble

be a

endowed

2o

we explain

our

as

on

some

sense,

of

the

to

This

which

out

What

example,

he

understood?

to

something

familiar

Middle

move

and

idea

this

concept similar

they

for

indeed, had

what

quality

this

being

it.

of to

could

causality.

be

for

a

How,

which

point

account

the

loses

the

be

for

equally sense

in

Ockham

similarities

'colour'

Crathorn,

concept to

figure

own

to

Yet

doctrine.

context similar

to

in

thing

not

In

bother least,

a

the

everything

but

to

they

a

of

resemblance carelessly

not

mental

One

this

at

and

the

whiteness.

seriousness

other

on

given

number

not

a

once?21

and

try

restrictive

impossible,

is

in

a

if,

dissimilar

A

at

'animal'

to

Objections,

similitude

is

resembling

Acts

which

whitenesses,

would

each

too common

or speaking',

is

Bucephalus,

or

according

onset.

does

is supposed

great

writing

Ockham's

questions.

greater

all

of

concept

Ockham's itself

resemble

It

ironic

Dominican

thing of between

particular. How

2.

similitude, even

of

maximally

reason

122 nuthatch, 124 OCKHAM ON CONCEPTS CONCEPTS AS SIMILITUDES 125 correspond to a static position of the hand, which makes it easier to understand how foot, but not conversely. And it differs from the photographical representation by it can be a likeness of the thing grasped. being a wider and looser relation, more akin to what John Buridan later called a Even a dynamic gesture, actually, can mimic an object in certain cases. I can similitude 'by fitness' (secundum convenientiam), which he described as an sketch the form of a human body, or of a house, with a movement of my hands. A appropriate fitting· of something on to something else.28Photographicallikeness is a fortiori, nothing prevents a static actualized state of my hand to be seen as a likeness special case of this, admittedly, but the relevant appropriateness can vary widely, of some external thing. Suppose I grasp a ball, ora pen, and let the grasped object according to the sort of effect which is involved: the appropriate fitness of a foot be removed without any change in the position of my hand. What is left is a likeness track to a foot is quite differentfrom that of an offspring to its parents, or of a heated of the ball or of the pen. In Ockham's vocabulary, the remaining act of my hand- its thing to the fire that heats it. Yet these are all cases of this asymmetrical relation of actualized state or position - is a similitude of the object it previously grasped. It is, similitude we are now trying to circumscribe, and which holds presumably, in one admittedly, a rough similitude in this case, but a similitude nevertheless. This manual form or another, between any effect and its cause.29 act, moreover, is to the same degree a similitude of any other object which is Although asymmetrical, however, and closely related with causality, this sufficiently similar to the original ball or pen. In so far as it is a similitude, this connection is not to be identifiyd with causality, since it relates something not merely posture of my hand pertains to a multiplicity of possible objects. If it could play the to its particular cause, but to everyotherthing which is relevantly equivalent with role of a sign within a proposition, it would be a general sign, able to supposit for all this cause as well. It is not itself a transitive relation for all that (a second-order the objects that have the relevant shape. This comparison, I believe, brings us very conceptual representation of the concept of 'horse', for example, is not a conceptual close to Ockham's point about the similitude ·of the concept with the represented representation of horses), but its co-domain is openly generated by another - and thing. What we have in the case of the grasping hand is a physical analog of the quite different - similarity relation, which is both transitive and symmetrical, an distinction between the intuitive and the abstractive acts. As long as the ball is there, equivalence relation in the technical sense. A foot track, for instance, is a similitude, the act of my hand can be seen as the singular grasping· of this singular ball that in the required sense, of every (possible) foot which is relevantly equivalent with the causally shapes the posture of the hand. This is the equivalent of an intuitive grasping. one that caused it; the posture of a hand is a similitude, in the required sense, of Yet I can also consider the position of the hand in abstraction from the actual presence every (possible) singular object which is relevantly equivalent with the one that of the ball (whether it is still there or not), and then I have an analog of the abstraCtive originally caused it. This is, I believe, Ockham's basic intuition on the matter. act: a general representation of any object that would tightly fit into the hand when it is so placed. This could legitimately be described as a manual concept! If my hand was then productive, it would engender a ball similar in size and shape with the 3. Varieties of conceptual representation original one, which is how Ockham originally characterized intentional similarity.27 In short, to say, in Ockham's vocabulary, that an act resembles a thing raises no It remains to be seen how this model can be applied to intellectual acts in such a way special difficulty. The act of apprehension by which the mind grasps something must as to yield the required results within the framework of Ockhamism. Ockham holds, be a similitude of this particular thing and of all the other ones that are sufficiently in particular, that a single encounter suffices for the formation of a general concept similar to it, exactly like the act of apprehension by which my hand grasps a ball corresponding to a species specialissima, while several are needed for generic must be a similitude of this particular ball and of all the other objects that are concepts.30 The idea of similitude just developed does account for the relevantly like it. We can see why similitude in such cases is the condition of representational generality of the conceptual act, and for the recognitional capacities possibility for generality. In the direct grasping corresponding to the intuitive act, of the intellect, but will the resulting generality always be the right one? Given this what is relevant is the actual causal interaction with a given object. But causality by understanding of intentional similitude, how are we to explain, in particular, that our itself is a link between determinate singular objects or events. What transforms the intellect reaches specific concepts at one sitting? And how can our comparison with grasping act into a general representation is that it fits the shape of the grasped the manual act shed any light on the formation of more general concepts such as that object, thus resembling it to some extent; in so far as it is a likeness of the singular of 'animal', or 'colour', or 'being'? These are difficult questions. I propose to tackle object which is grasped, it is also a likeness of any other object which is relevantly them in a progressive way by successively reviewing the main sorts of simple like it. While causality is strictly singular, generality naturally comes along with categorematic concepts countenanced by Ockham: absolute specific qualitative likeness. concepts such as 'whiteness' or 'redness' (section 3.1), specific concepts of The required notion of similitude, however, is quite distinctive. It is neither substances such as 'horse', 'nuthatch' or 'man' (section 3.2), connotative concepts qualitative resemblance in the Aristotelian sense (as between two red things, or two such as 'white', 'taller' or 'similar' (section 3.3), and generic concepts such as round things), nor essential similitude (as between two nuthatches, or between a dog 'colour', 'coloured', 'animal' or 'being' (section 3.4). We will try to assess in each and a horse), nor perceptual likeness (as between a photograph or a statue and the case how the signification of the concept - the things it signifies - can be connected pictured object). The relevant Ockhamistic connection typically follows upon a with its being an intentional similitude in the sense put forward in the previous causal link and thus differs from both qualitative resemblance and essential section. This. in turn. will shed it revealing light on what the functions of concepts similitude by being asymmetrical: a foot track, in this sense. is a similitude of the are supposed to be in Ockham's theory. 126 OCKHAM ON CONCEPTS CONCEPTS AS SIMILITUDES 127

3.1 Absolute specific quality concepts structure· is not disclosed at first sight to the human mind. On the other hand, substantial concepts are not inferred for Ockham; they are not constructed as Let us start with the most specific qualitative concepts, like that of a precise colour, complex descriptions or nominal definitions of the form: 'whatever is the substratum a specific brand of red, let's say - which, for the sake of commodity, I will simply for such or such perceptible accidents'. Ockham makes it clear that we have, in his call 'redness' in the rest of this discussion. What is striking here is that we cannot view, simple intuitive and abstractive cognitions of substances as well as of appeal to an isomorphism between the concept and the singular rednesses it accidents. 33 His whole epistemological theory could not fly without this assumption. represents, since a plain redness, typically, has no phenomenal structure. When the The acknowledged impossibility of grasping a substance in itselfis not seen by him intellect forms its first concept of 'redness', it has to do so, in Ockham's doctrine, on as a complete impossibility of intuiting the substance: even if through the accidents, the basis of a single encounter with something like a simple object, a given instance we do have simple intuitions of substances, and simple substance concepts are of redness. Similitude in this case cannot be structural isomorphism. And of course, acquired on this basis. it cannot be qualitative resemblance either: the mind does not really become red. The theory of the cognition of substances Ockham supposes is thus In Ockham's doctrine, the first intellectual analog of redness that occurs within approximately the one John Buridan will make explicit a few years later. 34 the mind must be the posture the intellect takes on when the agent really apprehends Everybody admits, Buridan says, that the cognition of accidents is necessary in this some redness. This is the intuitive act. And it is a similitude in the sense identified world for the cognition of substances. It should not be concluded from this, however, above. 3l It is normally caused in the mind by a direct encounter with some redness; that substances are not apprehended by the senses, the imagination or the intellect. yet in so far as it is a similitude, it pertains to all rednesses if - as Ockham is The substance is grasped, indeed, but 'confusedly', along with its accidents: 'it is by committed to suppose - a distinctive intellectual posture is adopted whenever the a direct representation that a substance with its accidents confusedly represents itself agent comes across a redness. This original act, according to the theory, leaves in the to the sense and, through the sense, to the intellect, which can extract it' .35 When I mind a derivative trace which is also an intellectual analog of all rednesses, but perceive a nuthatch, according to this theory, what I grasp - both by the sense and which can be reactivated in the absence of any redness. This one is the concept. It is the intellect - is not merely a certain shape and a certain arrangement of colours, but also a posture of the intellect: this precise posture, namely, that it becomes .able to the substance which is so shaped and coloured. This substance, admittedly, is not take on if and only if it has previously met with at least one redness. Since any apprehended in its bare purity ('in se', Ockham says), but it is nevertheless present redness will do as its trigger, this concept can be said to be a similitude of all of them in person to the receptive mind; it is the immediate object of an intuitive act and of equally, and it will thus enable the agent to correctly re-identify new rednesses when the abstractive act that follows upon it. they occur: the relevant equivalence relation between the represented things in this This abstractive act, then, truly is a simple concept of some sort of substances. case is the cospecificity - or maximal essential similarity - of all rednesses among Yet, if it is to be a similitude, it cannot be a similitude of the internal structure of the themselves. We thus reach the required result: the specific concept of redness is a perceived individual (which is still hidden to the mind), but only of its outward similitude pertaining in the same way to all rednesses, and to nothing else; it can be appearance. Since Ockham takes this concept to be a likeness not only of the caused in the mind by a single encounter with some particular redness in optimal particular individual that triggered its formation, but of all those who belong to the conditions. So far so good! same species specialissima as well, he must presuppose that the perceptible appearance is in general a reliable indicator of the species. The shape and colour of 3.2 Specific concepts of substances a given bird or of a given flower, for example, usually are reliable indicators of the species they belong to. Ockham's presupposition is that this is the general and The case of substantial specific concepts, like 'horse' or 'nuthatch', is more delicate. normal situation with respect to perceptible substances. The natural working of our The problem is that while qualities are immediately grasped in themselves according mind must rest on this relative regularity in the connections between substances and to Ockham; this is not so for substances: 'it is by the accidents that individual the typical accidents which shape their outward appearance. Without such substances are known to us, because the accidents are cognized directly and in regularities, the human mind could not form reliable intellectual similitudes of themselves by our intellect, but not the substances'.32 How, then, can an adequate external substances, at least not on the basis of a few encounters. general concept of a given species of substances be formed upon a single encounter? If this is so, the idea of similitude which is required for specific substantial One thing is sure: we cannot expect, on such a doctrine, that the intellect which concepts is very closely related to what we have found in the previous case. It forms the concept of 'nuthatch', let's say, .should ipso facto develop in so doing a corresponds to the assimilation of the intellect to external objects by means of its deep understanding of the essences of nuthatches, of their internal structure or of the capacity to recognize perceptible features, and it is rooted in the typical posture the necessary conditions something must meet in order to be a nuthatch. Ockham is intellect takes on whenever it comes across a substance of a certain specific sort. The certainly not supposing that the mind is endowed with some sort of metaphysical X­ agent, for example, sees a nuthatch. Her intellect, then, takes on a posture of ray which allows it to grasp the intern;l1 essence of a thing on the basis of a single assimilation with respect to the perceptible features of this hird. Once ahstracted perceptual encounter. In so far as the general concept of nuthatch is a similitude, from the concrete presence of this particular nuthatch, the remaining intellectual therefore. it cannot he a similitude of the internal structure of the nuthatch. since this posture is supposed to he the one that would typically result from encountering any 128 OCKHAM ON CONCEPTS CONCEPTS AS SIMILITUDES 129 other nuthatch. The difference with simple qualitative concepts is that the ones we powers a nuthatch normally has. Intentional similitude, on the other hand, is what are now considering pertain not to the qualities as such, but to the things that have helps us recognize, on the spot so to say, whether something falls or not under a these qualities. It must be supposed, accordingly, that our intellect is innately given conceptY What the case of the nuthatch-robot reveals is that intentional capable of (at least) two sorts of different assimilating postures, one that is actualized similitude can in special conditions be misleading. And if it can be misleading, then when it attends to one or more qualities, and the other one when it reacts to the things it cannot constitute the sole criterion for what belongs in the extension or not. that have these qualities. Those must be something like a priori categories of our Another ~and more fundamental- function of substantial concepts must be at play intellect. This is not said by Ockham in so many words, of course, but I do not see here: the correct guiding of expectations with respect to the causal powers of things. how he could avoid supposing it. Or how we could, for that matter. The notion of intentional similitude, then, turns out to be basically the same for The idea, in short, is that the intellect sometimes is in substance"mode. When it substance concepts as we found it to be for qualitative concepts, but its semantical is, it assimilates itself to some external singular substances via the perceptual role is now quite different. To say that a substance concept is a similitude of what it scheme which typically corresponds to this specific kind of substances. This represents is merely to say that it is a typical posture the intellect becomes able to amounts to say that the intellect then forms an abstract similitude of a given take on in virtue of having been in touch - through the senses ~ with individuals of individual substance which it has encountered and of all the other ones which a certain sOrt. And the relevant equivalence relation between the represented things present the same typical perceptual features. Does this mechanism yield the right is, just as in the case of qualitative concepts, maximal similarity (or cospecificity). result? Namely the formationof simple substance concepts corresponding to natural The great difference is this: the cospecificity in question is now between substances kinds, such as the concept of 'horse' or 'nuthatch'? The answer is 'yes', if we rather than qualities. And since substances are not cognized in se but only through suppose, as Ockham did, that each natural species distinctively displays certain their accidents (at least in this world), the intentional similitude relation cannot typical perceptual features. This is not as naive as it might seem. Ockham takes it to determine alone the extension of substance concepts as it does for qualitative be a commonly accepted principle that if two individuals are of the same species, concepts. Intentional similitude still comes out as an indispensable feature of simple they will always produce highly similar effects on equivalently disposed patients.36 substance concepts, and it should provide in normal situations a quick and reliable It follows from this that two individuals which belong to the same species recognitional device for categorizing substances. But since the causal powers of a specialissima should produce, in normal circumstances, highly similar perceptual substance are not all immediately apparent in perception, a discrepancy remains effects upon human observers. And we can work here, moreover, with a very wide possible in principle, given the basic function of substance concepts, between the notion of perceptual feature, including, for example, not only shape and colour, but individuals they resemble and those they signify. also the typical sounds emitted by the object, its perfume, its ways of moving, and so on. Perception, in this context, can even be extended to empathy, that is, the non­ 3.3 Simple connotative concepts inferential apprehension of the perceived object as being itself a subject for sensations, emotions, and the like. The concept of a nuthatch, for example, will be Although there might be a few exceptions, simple connotative concepts in this regard the intellectual act that results from the typical posture the mind takes on when it are more likely to be like qualitative concepts than like substance concepts: in most apprehends, in optimal conditions, a moving and tweeting nuthatch, and apprehends cases, intentional similitude will decisively determine their correctness of it as a living being. application. Let me explain. What characterizes simple connotative concepts, as was It could still happen, of course, that the correlation fails in special cases between seen in chapter' 6, is that they are naturally acquired general representations of essence and outward appearances. We can imagine a perfect imitation of a nuthatch ordered n-tuples. They can be considered,in other terms, as mental models for with respect to all its perceptible features, which would really be a teleguided robot, ordered groups of entities out there in the world. Thus understood, they are can't we? Ockham does not discuss. such cases, but his theory of substance concepts similitudes of these ordered groups, in the by now familiar sense that they cannot rule them out as impossible and should therefore take them into account. correspondto the typical postures the intellect becomes able to adopt as a result of What he has to say, obviously, is that the nuthatch-robot does not belong to the having been in intuitive contact with such ordered n-tuples in the way described in extension of our simple concept of 'nuthatch', the one we are supposed to have chapter 6. Different sorts of cases can occur according to whether the primary and naturally acquired on the basis of our previous encounter with a real nuthatch. And secondary significates of these primitive connotative concepts are substances or this has far-reaching consequences. It means that although intentional similitude is qualities, but their correct application, most of the time, will solely depend upon the what opens the way for generality in representation, it does not adequately determine perceptual features of the relevant situations, as we can check by successively by itself the extension of our simple substance concepts. The extension of such reviewing the main possibilities. concepts in Ockham's framework should contain only individuals which really belong to the same species as one another, individuals, that is, which have basically Case 1: Both the primary and secondary sign(ficates are qualities equivalent causal powers. This is how substance concepts can play their cognitive role in our mental apparatus: the falling of an individual x under the concept of This is exemplified, for instance, by the situation Ockham mentions in Quodliht'fa 'nuthatch' is supposed in etlect to warrant the conclusion that x has all the causal IV. 17. of someone who simultaneously grasps two whitenesses and forms as a result 130 OCKHAM ON CONCEPTS CONCEPTS AS SIMILITUDES 131 a certain concept of similarity.38 The relevant connotative concept of similarity' in spontaneously applied to both Socrates and Plato, as Ockham claims,40 the mind, this example does not connect the substances that have these whitenesses (as the surely, can form from this the relational concept of 'being essentially similar', just standard Aristotelian notion of similarity would do), but the whitenesses themselves, as it can form the concept of 'being qualitatively similar' from having met with two and it clearly must be ascertainable in direct experience, just as the basic qualitative white things (or two whitenesses).1f so, such a concept will be, according to the tack absolute concept of whiteness is. It will correctly apply to every couple of singular we are presently following, this typical posture of the intellect that it can adopt as a colours which are spontaneously recognized as similar in optimal conditions of result of having met with two different individuals to which it spontaneously applies observation. Unless surrounding conditions - of light, for example - are the same absolute substance concept. But since the latter recognitional application unfavourable, intentional similitude in such a case leaves no room for error. can in principle be mistaken in special cases - remember the nuthatch-robot - so can the judgement that x and yare essentially similar to each other. And for the same Case 2: The primary significates are substances while the secondary significates reason: the internal essences of things, although they are intuited, are not cognized are qualities (or vice versa) in se, but only along with their accidents. We thus have, finally, a category of primitive connotative concepts which, in this regard, are more like simple substance A paradigmatic example of this is the connotative concept 'white', which primarily concepts than like simple qualitative concepts. The presumption, however, is that signifies'certain substances - those material bodies that have a whiteness - and this category has very few members. Apart from 'being essentially similar' - or secondarily signifies, or connotes, certain qualities, the whitenesses themselves. 'essentially different' - I can think of no other connotative concept which could Although substances are involved here, the ascertainability of such concepts seems as plausibly be naturally acquired on the basis·of intuition, while connecting only strong as it was in the previous case if the conditions of observations are optimal. For SUbstances independently of their accidents. the recognition of a certain substance as being white in no way depends upon a Be that as it may, the net result of this review is that we have in principle two sorts correct identification of what this substance intrinsically - or essentially - is. It is of simple connotative concepts (even if unequally represented), just as we had sufficient that our mind be able to intuitively grasp a certain quality - a whiteness in previously found two sorts of simple absolute concepts: those that signify substances this example - as being inherent to a substance, which is a capacity Ockham explicitly only, and those that are true or false of certain n-tuples (whether substances or acknowledges.39As was explained in the previous chapter, an absolute substance qualities) in virtue of their connection with empirically ascertainable qualities. concept will normally be formed on the same occasion (for example, the concept of Intentional similitude - understood as the typical posture the intellect adopts when a man or of a horse), but the empirical application of the connotative concept 'white' it meets with certain objects - accounts in both cases for the recognitional capacities will not in general depend upon the adequacy of this substance concept. In so far as of intellectual agents. And it depends in both cases upon the perceptual features of intentional similitude accounts for the recognitional abilities of the agent, it should be the things represented. Because of this, precisely, there can in principle be a just as reliable in the case of 'white' as it is in that of 'whiteness'. discrepancy in the former case (merely substance-related connotative concepts) between what the conceptresembles and what it signifies, as there can be in the case of absolute substance concepts, while this is excluded in principle for quality-related Case 3: Both the primary and secondary significates are substances connotative concepts, where intentional similitude directly guarantees adequacy, as it does for absolute quality concepts. Two situations must be distinguished here, and they yield different results with respect to the role of intentional similitude. 3.4 Simple generic concepts 3a) The first one occurs when a simple connotative concept applies to certain couples - or n-tuples - of substances in virtue of certain observable qualities that The intellect, according to Ockham, can also acquire more general concepts of they have. 'Being taller' would be a good example: both its primary and secondary various kinds on the sole basis of experience, without engaging into compositional significates are substances (respectively: the taller bodies, and the shorter ones), but activity: absolute substance concepts, for example, such as 'bird' or 'animal', or they both are so signified by it in virtue of their external observable features. The absolute qualitative concepts such as 'colour', or again general connotative concepts empirical application of such a concept to a given couple of substances, then, does such as 'coloured'. What is required for their implementation in the mind is that the not depend, any more than in cases 1 and 2, upon the corn~ctidentification of their agent should have been in intuitive contact with representatives of more than one essences. species of the same genus}1 When I meet with a nuthatch, the first general concept 3b) The other situation is when the connotative concept applies to certain I acquire is a specific concept, that'of 'nuthatch' in this case: if I then meet with a couples - or n-tuples - of substances in virtue of what they intrinsically are. An chickadee - which is somewhat similar to a nuthatch, but not maximally so -I example here would be the relational concept 'being essentially similar' in so far as naturally acquire a more general concept, which applies both to nuthatches and it applies to substances. Such a concept, presumably, must be among those simple chickadees,42 and so on. intellectual representations the human mind is naturally able to form, according to We have been wondering with Crathorn, at the beginning of the present chapter, Ockham. If a concept like 'man', for instance, can naturally be abstracted and how such gcncril: concepts could still plausibly be seen as similitudes of what they 132 OCKHAM ON CONCEPTS CONCEPTS AS SIMILITUDES 133 represent, especially as they get more and more general. How, for example, can my Such discrepancies are rendered possible in both generic and specific concepts of concept of 'animal' equally resemble a nuthatch and an elephant, but not a pine tree substances by the fact that concepts, as we may now conclude, have two main or a chariot? The notion of intentional similitude reached in section 2 above no\\( functions in Ockham's epistemology. They account, on the one hand, for the allows us to provide a satisfactory answer to these Crathornian worries. A generic recognitional capacities of the intellect,45 and they determine, on the other hand, via concept simply is this characteristic posture that the intellect becomes able to adopt their signification and supposition, the truth-conditions of the mental propositions in after having met with several individuals of different species. The Ockhamistic which they suitably occur.46 Being based on the outward appearance of things, conjecture, here, is that as a result of having seen a nuthatch and a chickadee, I must intentional similitude has primarily to do with the recognitional function. It is to be have acquired a new mental capacity, that of adopting a certain determinate supposed, within the Ockhamistic framework, that when the intellect readopts a intellectual posture that can be reactivated by certain individuals (small birds, let's determinate posture previously rendered familiar by some perceptual experience, it say), but not others. Such a recognizable posture will be an intentional similitude of has some awareness of this. And this awareness, precisely,. is what prompts it to everything it pertains to in this way, exactly in the same sense that the original classify a newly met object within the same species or genus as some previously specific concepts dealt with in the previous subsections are intentional similitudes. encountered ones. Intentional similitude, in other words, accounts for our grouping My concept of 'animal' will thus be said to be a similitude of nuthatches and dispositions. But grouping things together typically determines our expectations elephants (among other things), but not of pine trees and cars, because it is an towards them: identifying something as a trout on the basis of its external appearance intellectual posture I have acquired the capacity of adopting by meeting with many 'raises expectations as to its behaviour, its taste, and so on. Whether these expectations different animals, and that can typically be reactivated, in normal circumstances, by are fulfilled or not depends upon the objective truth-values of our mental propositions nuthatches and elephants among others, but not by pine trees and cars. (of our thinking, for example, that this thing in the water is a trout). In so far as some In this conception of intentional similitude, there is no limit in principle to the of our naturally acquired concepts are supposed to help us classify things according possible generality of intellectual representation. The more general concept, that of to their essential causal powers - substance concepts, namely - a discrepancy is 'being' (ens), is a similitude, Ockham says, of an infinite number of possible objects, bound to be possible in principle between the grouping dispositions which are 43 every thing indeed that can exist. There is no way to understand this if we stic~to associated with these concepts and what their contribution is supposed to be to the either the Aristotelian or the picture conceptions of resemblance. But the statement truth-conditions of our beliefs and expectations, between, in other words, similitude loses its halo of mystery when we realize that an intellectual state can legitimately and signification. Fortunately, however, the outward appearance of things normally is, be called a similitude, in Ockham's vocabulary, of whatever could have produced its under good conditions of observation, a reliable indicator of what they intrinsically original implementation in the mind and can reactivate it in normal favourable are and of what causal powers they have. A nuthatch normally looks like a nuthatch, circumstances. The concept of 'being', admittedly, constitutes a special case for and what looks like a nuthatch normally is a nuthatch. Ockham, on the whole, is quite Ockham, since, contrary to' all other generic concepts, which require several justified in thinking that our simple concept of 'man' normally allows us to correctly encounters, it is supposed to be immediately impressed upon the mind by any single judge whether something is a man or not. If true of specific concepts such as encountered thing, simultaneously with the corresponding specific concept.44 Yet the 'nuthatch' or 'man', this probably holds too - although to a lesser extent, admittedly sense in which it is an intentional similitude is exactly the same as in the other cases - of our naturally acquired generic concepts of substances such as 'bird' or 'animal'. we have reviewed: it is the typical posture the mind is able to adopt in virtue of having met with any existing object. Since there are absolute substance concepts, absolute qualitative concepts, and 4. Two problems about absolute concepts connotative concepts among our generic mental terms, just as there are among the specific ones, the connections will be the same between signification and intentional Our proposed elucidation of intentional similitude now puts us in a favourable similitude in these new cases as they were in the previous corresponding ones. In position to discuss two subtle difficulties that have been recently raised against qualitative generic concepts, in particular - whether absolute or connotative ­ Ockham's nominalistic theory of absolute concepts, one by Gyula Klima (section signification should automatically coincide with similitude: the generic qualitative 4.1), and the other by Deborah Brown (section 4.2).47They don't seem to have much concept of 'colour' signifies whatever qualities it resembles (in thti relevant sense), to do with each other at first sight: Klima's puzzle pertains to the simplicity of and the connotative concept of 'coloured [thing]' signifies whatever couples of absolute concepts, while Brown's concerns the purported absence of synonymy in substance and quality that it resembles (in the relevant sense). The absolute generic Ockham's mental language. Yet a correct understanding of intentional similitude and substance concept of 'bird', on the other hand, signifies every bird and normally its role turns out to be decisive in both cases for their solution. resembles (in the relevant sense) every bird as well, but since the intentional resemblance in such cases is based on outward appearances, a discrepancy between 4./ Klima's ohjcctiol1 signification and similitude is possible in principle, just as it was for the more specific concept of 'nuthatch' or for that of 'being essentially similar' (as applied to Oyula Klima is currently one of the most penelrating analysts of the philosophical substances). import of late medieval semantics and onlology:*H In a reccnl disl'ussion on his 134 OCKHAM ON CONCEPTS CONCEPTS AS SIMILITUDES 135 website, he has claimed that Ockham's special brand of nominalism must fail, relation, both symmetrical and transitive. It is, more precisely, what we have called because, precisely, of how it copes with absolute concepts.49 There is a crucial 'essential similarity': maximal essential similarity in the case of specific concepts requirement, Klima holds, that should be satisfied by any sound theory of conceptual such as 'nuthatch', and less stringent essential similarities for generic ones. There is representation, but that Ockham's semantics for absolute concepts fails to satisfy. nothing in respect of which all nuthatches have to be similar in order to fall under The requirement in question is the following: 'our concepts should represent the absolute concept of 'nuthatch'; they simply have to be essentially similar to each particulars belonging to the same kind in that respect in which these particulars other - or 'cospecific', if one prefers. should be similar in order to belong to the same kind. Otherwise a concept could not For this approach to yield the right result in the case of specific concepts, it has represent all individuals of the same kind on the basis of acquaintance only with a to be assumed - as Ockham certainly does - that each individual belongs to one, and limited number of them.' only one, species specialissima. Once a certain sample s is present to the mind, the Once this requirement is made plain, it is obvious that Ockham's account of group of all individuals which are of the same specific kind as s is thus uniquely absolute concepts cannotmeet it since absolute concepts, for him, as Klima points determined, and so is, consequently, the extension of the simple specific concept out, do not represent their objects 'in respect of something'; they simply have them naturally formed by the intellect on this occasion. in their extension, period. This does not mean that the agent will automatically be able to recognize as such Now, Klima is certainly right on this last point, as our own discussion of any new instance of this particular kind that she might come across in the future, intentional similitude confirms. Absolute concepts are indeed similitudes of their without any possibility of mistake. Two quite different sorts of similarities are objects for Ockham, but they cannot be said to signify these objects in respect of the involved in Ockham's theory of absolute concepts: essential similarity among their properties according to which they are similitudes of them. This is especially clear significates on the one hand, which determines the extension of the concept, and in the case of simple substance concepts such as 'nuthatch' or 'man'. In so far as intentional similitude on the other hand, which accounts, I take it, for the these intellectual representations are similitudes of nuthatches or men, this must be, recognitional dispositions of the agent. The latter,admittedly, must depend upon as we have explained, on the basis of the perceptible or phenomenal features of the certain observable features of the objects, and it is true that Ockham takes it to be represented objects. Yet a substance concept of this sort is not a mental sign of these normally reliable, but a discrepancy always remains possible in principle, as we have features, but only of the cospecific substances that display them. It is not analysable, insisted, between what the agent is disposed to recognize as falling under a given as a complex sequence of simpler - qualitative or substantial ~concepts, and it does absolute substance concept of his, and what does in fact fall under this concept. not connote in any way the perceptible qualities in virtue of which it resembles its Falling or not under a given absolute specific concept is a matter of truly being objects. Being an absolute concept in Ockham's sense, its proper semantical import cospecific or not with certain salient instances of this concept, and not a matter of is restricted to its extension. And this extension contains only singular substances, having such and such recognizable features. each one of which being a member of it to the same degree as any other. 50 Intentional A non-trivial picture of the working of the mind is involved here, no doubt. But similitude opens up the way to generality in representation, but the internal structure not an implausible one. It seems reasonable to suppose that our intellect is so of the representation, in so far as intentional similitude requires such a structure, constructed -either by God or by evolution - as to form spontaneously, upon any must not be ideptified with its semantical content. Seen as a posture of the mind, an single encounter, a general mental sign which is true of all individuals of the same absolute concept might well be a complex psychological state - how, otherwise, kind as the encountered one, whether we are easily able to recognize them or not. could it be a similitude of its objects with respect to their perceptual features? - yet Since being of the same kind involves having highly equivalent causal powers, such it is to be counted as a simple concept in so far as none of its parts is itself a concept. a mental device should tum out to be most useful in the long run for guiding This, indeed, is the very reason why a discrepancy becomes possible in principle in expectations, and one especially interesting brand of them in particular: conditional the case of substance concepts between.what they resemble and what they signify. expectations. Once I have acquired the specific concept of a nuthatch, for example, I Although Ockham's absolute concepts might resemble their objects in respect of know that all the significates of this concept will essentially have the same basic their outward appearance, Klima is right that they do not signify them in respect of causal powers as the original samples do, even if these causal powers are not yet anything. known to me. What I know, in other terms, is that if one of my salient samples of It is true, then, that Ockham's theory of absolute concepts does not satisfy nuthatch turns out to essentially have a given causal power, all the other significates Klima's requirement. But why should it? What Klima presupposes here is that since of this mental concept will have it as well, precisely because'all agents of the same all the significates of an absolute concept are supposed to resemble each other, they specific kind [species specialissima] are productive ofeffects of the same sort' ,51 This must do so in some respect. Which is why, in his view, the absolute concept in is how Ockham can admit that true generalizations can sometimes be reached on the question must signify them in this very respect. But this is a notion Ockham does not basis of a single experiment. 52 Since absolute concepts are supposed, in addition, to share, and does not need to share. What determines the extension of an absolute come along with certain recognitional capacities (in virtue of intentional similitude), substance concept, for Ockham, is not any connection that its members have with the ability to produce such true generalizations, and to act on them, is undoubtedly an another domain of objects (respects, features, qualities, or the like), but the relation invaluable advantage. even if, as we have acknowledged, the recognitional aptitude thl'.v have with each other. And this relation, as we have seen,must he an equivalence should fail in exceptional cases (think of the nuthatch-rohot again!). 136 OCKHAM ON CONCEPTS CONCEPTS AS SIMILITUDES 137

The situation, admittedly, is somewhat more complicated for generic concepts, are individuated by their causal history as well as by their semantical properties. It and Ockham does not give much indication as to what concepts exactly we are follows that if an agent should unknowingly originate on different occasions two expected to acquire from having met with a nuthatch and a chickadee, let's say, or a independent causal chains of absolute conceptual acts with the same extension, those nuthatch and a duck, or a nuthatch, a duck and a trout, and so on. But the principle, would be counted as two different ooncepts in the Ockhamistic doctrine. These basically, should be the same. What we are to suppose is that at every level of remarks, however, do not suffice to solve Brown's puzzle. They merely help to locate generality, the significates of an absolute concept are uniquely determined by some it more precisely. The problem arises orily if it should be possible for a human agent degree or other of essential similarity among them, according to what function to form two distinct, but coextensive, general absolute concepts. And this, if exactly the concept in question is naturally designed to fulfil within mental possible, would not conflict with two Ockhamistic theses, but only with one, namely propositions, given the way our minds are functionally organized. The main point is that there are no synonyms among the simple concepts of a given agent. The that whether they are specific or generic, absolute concepts do not have to signify question, then, is the following: should it be admitted as possible, given Ockham's their objects in respect of anything. Klima's requirement, then, is simply bypassed. causal theory of concept acquisition, that an agent unknowingly acquires two distinct absolute concepts of exactly the same objects? 4.2 Brown's puzzle If what we have said earlier about a possible discrepancy between what a substance concept signifies and what it resembles, is correct, then the answer to this In a stimulating paper published in 1996, Deborah Brown has claimed to identify a question is: yes, it is possible in principle for an agent to form two distinct and tension within Ockham's theory of absolute concepts which, she says, threatens the coextensive absolute substance concepts, but this, from Ockham's point of view, internal coherence of his whole philosophical project.53The problem is that absolute must be quite exceptional. The normal situation, Ockham thought, is that in good concepts, in Ockham's view, are acquired in a purely passive way, through a process conditions of observation, a member of a given species can be recognizeQ as such on described in strictly causal terms. If so, there seems to be no reason why someone the basis of its outward appearance. The case of general substance concepts in this might not independently - and unknowingly - acquire two distinct absolute concepts regard is very different from the example Brown started with. Even if for the same individual: 'One would think that a person could easily acquire the Brown's agent had not recognized Cicero on her second visit to the Forum (because, concept "Marcus" in an encounter with Cicero on one day's visit to the Forum and let's say, he had grown a beard in the meanwhile, and lost all his hair), she would the concept "Tullius" from another encounter with Cicero without realizing that she normally have recognized him as a human being without any problem if she was was encountering one and the same individual.'54 Such a possibility, Brown says, can close enough and the light was good. It might be very' difficult to say whether this be generalized to any' absolute term whatsoever, and it bluntly'conflicts with two nuthatch in front of me is the same individual I saw yesterday, but it would normally principles of Ockham's theory ofmental language: (i) that concepts are individuated be much easier, if the bird is close enough and the light is good, to identify it as being only by their signification, and (ii) that there is no synonymy in mentalese. What is of the same species. questioned here, ultimately, is the compatibility of Ockham's causal account of the The point, here, is that what the agent acquires when she first meets with a implementation of absolute concepts with the fundamental semantical role he nuthatch, let's say, is' a capacity to adopt a certain intellectual posture. This posture, attributes them. which is said by Ockham to be a similitude of nuthatches, will be reactualized from Before getting to the heart of the matter about this difficulty, we need to clear up then on in two sorts of circumstances: it will occur, even in the absence of any a couple of preliminary points. First,Brown's Cicero example is ill-chosen. nuthatch, as a component of certain complex propositional acts of the agent, the Ockham, as we have seen, is explicit that the human mind cannot form simple proper truth-conditions of which, then, will have to do with nuthatches; it will be singular concepts, precisely because concepts are intentional similitudes.55 The reactivated, on the other hand, on any new encounter with a nuthatch (under singular 'Marcus' concept formed by the agent upon her first encounter with Cicero, favourable conditions of observation), thus bringing about in the agent's mind the consequently, will not be an absolute concept (no complex concept can be absolute recognitional judgement that this is a nuthatch. Absolute substance concepts can be in Ockham's theory), but a complex connotative one, and so will be the singular identified with such causal chains of intellectual acts and habitus. If everything is 'Tullius' concept formed on the second visit to the Forum. The resulting normal, then, once an agent has acquired a certain absolute concept in the guise of a coextensivity of two distinct complex connotative concepts is no problem for determinate causal chain of this sort, she will not independently ~cquirea new one Ockham: since these concepts differ as to what they connote - different with the same extension. A simple concept being an intentional similitude normally circumstances of encountering, for example - they will not wholly signify the same associated with a reliable recognitional capacity, the formation of two distinct but things under the same modes, and will not, therefore, be synonymous. synonymous absolute concepts must be unusual according to this doctrine. Secondly, it is not the case that Ockham's concepts are to be individuated solely Even though unusual, however, it is true that it cannot be ruled out. Since we have by their semantical properties, as Brown assumes. As I have argued in chapter 3, two admitted the possibility in principle of a discrepancy between signification and distinct intellectual acts within a given agent can be counted as tokens of the same similitude in substance concepts, we should also admit the possibility that the concept, in Ockham's doctrine, only if they both belong to a single causal chain recognitional aptitude fails. even under favourahle conditions of ohservation. if the originally anchored in one determinate intuitive encounter.56 Concepts. in this view. world docs not collahorate. Individuals of a given species might sometimes 138 OCKHAM ON CONCEPTS CONCEPTS AS SIMILITUDES 139 misleadingly look different,5? and. individuals of two different species might signified by one of them under a certain mode is also signified by the other under the misleadingly look alike (like water and XYZ in Hilary Putnam's famous tale of the same mode. 66 It follows, of course, that if two conventional terms are subordinated Twin Earth).58 Brown's consequence, then, reappears, albeitin a milder form: even to the same concept, they become ipso facto synonymous. But such a common if the case is exceptional, a human agent might still end up with two distinct but subordination is by no means analytically required by Gckham's definition of synonymous simple concepts in his mental apparatus. synonymy. For one thing, two different speakers might end up with synonymous This does not constitute, however, such a deep threat to the consistency of spoken terms if each one of them has subordinated its chosen word to one of his own Ockham's doctrine as Brown thinks it does. The non-synonymy thesis concerning concepts, and if the two concepts involved - although numerically distinct, being, ex simple concepts, although characteristic, is not nearly as central to Ockham's system hypothesis, qualities of different minds - do happen to signify the same things under as it has been taken to be. It is true that the Venerabilis Inceptor reaffirms the the same modes. These two concepts, moreover, will clearly be synonymous to each principle on several occasions: 'there is no multitude of concepts corresponding to other in this (quite common) case, even though none of them is subordinated to the multitude of [spoken] names'.59 But one has to look carefully at what use he anything. That a similar situation accidentally occurs within the mind of a single really puts it to. His goal in the passages where it is invoked (Summa Logicae I, 3, speaker can certainly not be analytically ruled out in Ockham's vocabulary. and Quodlibeta V, questions 8 and 9) is to identify the grammatical categories that The second consideration that brought many people to think of the non­ have to be postulated in mental language. The important principle at work in this synonymy thesis as being of importance in Gckham's system, is somewhat deeper. context is the positive one:, any distinction among grammatical categories which is It is that mental language, they think, must be a logically ideal language if it is to needed for the sake of signification must be present in mental as well as in spoken fulfil the role Ockham expects from it of accounting for the semantics of spoken or written languages. The negative converse principle, according to which the languages. 6? 'What we want from a proponent of mental synonymy', Richard Gaskin grammatical distinctions which are not semantically relevant are not to be found in writes, 'is a semantical (i.e. meaning-theoretical) role for the alleged existence of mentalese, is much less strongly asserted, and obviously considered less important. synonyms.'68 That there is no such role, however, is no problem for Gckham, because About the distinction between participles and verbs, for example, Gckham is content his theory of mental language, as should by now be clear to the reader of this book, to conclude that 'there does not seem to be any .great necessity to postulate such a is not primarily meant as the construction of a logically ideal language, but as a distinction among mental terms'.60And the corresponding negative general rule is correct description of what really goes on in the human mind. Reaching a sound presented merely as a convenient guideline: 'for we can conveniently eliminate from epistemology with no undesirable ontological commitment is, in Gckham's view, the mental names all of those grammatical features with respect to which spoken names predominant constraint on the theory of mental language. This .theory is also can differ, while remaining synonymous'.61 What we have here is a methodological expected to provide in the process an appropriate mentalistic foundation for. the principle of parsimony, the import of which, when it does apply, is the weak semantics of conventional languages, but it is not a flaw that some of its theses conclusion that it is not necessary to postulate a certain distinction in· mental should have no semantical role to play, if they are called for by a satisfactory account language,62 and not the strong conclusion that it is imperative not to postulate it. of concept acquisition. Gckham's application of these principles, moreover, has to do in effect only with Gn the matter of synonymy, what the doctrine of concepts as similitudes leads to the admission or rejection of whole grammatical categories· in mentalese. The is the double conclusion that (a) the human mind is functionally adapted to nature in corresponding thesis about the rejection of simple synonymous concepts, while such a way that there is normally no strict synonymy among the simple concepts of clearly asserted, occurs in virtually all cases as a mere comparison, usually a given agent; but (b) su~hsynonymy can exceptionally happen in special situations introduced by 'sicut':63 what Gckham means to say in all these cases is that just as if nature does not collaborate, particularly in the case of absolute substance concepts. there is no need to postulate a multiplicity of synonymous concepts in mental That (b) should be called for by considerations about the recognitional role of language, there is no reason either to postulate the presence in it of those general intentional similitude rather than by purely semantical reasons is quite all right given grammatical distinctions that have no semantical relevance. I do not see any great the intent of the theory. There is no serious threat here to its internal coherence. problem, consequently, in adopting a relaxed attitude towards Ockham's non­ synonymy thesis about simple concepts, and seeing it merely as what he took to be the normal situation. That exceptions should be admitted as possible in exceptional Notes cases does not seem to be deeply troublesome. There are two reasons, it seems to me, why some commentators - Brown among 1. The idea comes from chapter 1of Aristotle's Perihermeneias, 16a8. 'Similitudo' was the others - have thought otherwise, and have taken the non-synonymy thesis to be a Latin word used by Boethius in the translation of this passage. strong and central claim in Gckham's thought.64 Both of which, I contend, are 2. Adams 1987, pp. 121-33. wrongheaded. First, it has been thought that the very notion of synonymy in Gckham 3. Ibid., p. 132. 4. Altcri 19X9,passim. See in particular pp. 248-9. I have discussed Alferi on this in analytically requires that synonymous terms be subordinated to the same concept.65 Panacdo Ili9()b. esp. pp. 293-6. But this is not so. Gckham's preferred sense of synonymy is what he calls the hroad .5. Biard llillli. pp. 61-2. sense, according to which two terms arc synonymous if and only if whatever is 6. Biard IlJIJ7.pp. 21i-.ll. 140 OCKHAM ON CONCEPTS CONCEPTS AS SIMILITUDES 141

7. See Michon 1994, pp. 144-55; Panaccio 1992a, pp. 120-30. large forte accipiendo simile vel dissimile, potest aliquid secundum a1iquid aliud dlel 8. Ord. I, dist. 2, q. 8, OTh II, p. 272: ' ... intellectus videns aliquam rem extra animam simile vel dissimi1e'. fingit consimilem rem in mente, ita quod si haberet virtutern productivam sicut habet 23. See for example SL 1II-3, 26, OPh I, p. 690, where the definition Ockham gives for virtutem fictivam, talem .rem in esse subiectivo - numero distinctam a priori ­ 'simile' is: 'simile est quale, correspondens alteri quali, habenti qualitatem eiusdem produceret extra .... Et illud [fictum] potest vocari universale, quia est exemplar et speciei specialissimae'. See also SL 1,55, OPh I, p. 182: ' ... secundum qualitatem indifferenter respiciens omnia singularia extra, et propter istam similitudinem in esse aliquid dicitur simile vel dissimile'. obiectivo potest supponere pro rebus extra ...' (italics are mine). The same point is 24. See Panaccio 1992a, pp. 260 ss. made, for example, in Ord. I, dist. 2, q. 8, OTh II, p. 279, Ord. I, dist. 3, q. 3, OTh II, p. 25. This is illustrated, for example, in the passages from the Quodlibetal Questions quoted 428, and Ord. I, dist. 13, OTh III, pp. 419-20. above in nn. 17 and J 8. 9. Karger 1994, p. 439. 26. See, for example, Ord. I, dist. 2, q. 6, OTh II, p. 211: ' ... Sortes secundum substantiam 10. Quodl. IV, q. 35, OTh IX, p. 474: 'Ideo dico quod tam intentio prima quam secunda est est simillimus Platoni.' vere actus intelligendi, quia per actum potest salvari quidquid salvatur per fictum, eo 27. See the extract from Ord. I, dist. 2, q. 8 quoted above in n. 8, as well as the other quod actus est similitudo obiecti, potest significare et supponere pro rebus extra, potest passages mentioned in the same footnote. These, admittedly, were written from the point esse subiectum et praedicatum in propositione, potest esse genus, species etc., sicut of view of the jictum-theory, but in the last one mentioned - Ord. I, dist. 13, OTh III, p. fictum'.(trans!. Freddoso and Kelley 1991, p. 390; italics are mine). 420 -Ockham later added a short sentence saying that the same holds within the aclUJ­ 11. See above chap. 2, sect. 2for a detailed discussion of this argument. theory: ' ... esset proportionaliter dicendum de ipsa intellectione ... ' 12. Quaest. in Phys., q. 2, OPh VI, p. 399: 'Conceptus est similitudo rei extra.' 28. See John Buridan, In Metaph. VII, 8, Paris, 1588, fo. 46r: ' ... accipiendo similitudinem 13. See Michon 1994, p. 147. large vel improprie non pro eadem qualitate: immo solum per quadam appropriata 14. Quaest. in Phys., q.1, OPh VI, p. 398: 'Septimo, quia tale idolum plus differret are convenientia agentis ad effectum'. quacumquequam quaecumque res differt ab alia, quia ens reale et ens rationis plus 29. SeeJohn Buridan, Quaest. de anima I, 5, from ms. Munich 761, as editedin an appendix differunt quam quaecumque duo entia realia, et ideo tale idolum minus assimilatur rei, of Patar 1995, p. 50l: '... quilibet effectus habet et gerit quodammodo in se et per consequens minus potest supponere pro re quam intellectio quae plus assimilatur similitudinem suae causae ... ' rei; et ita minus erit communis rei extra et minus habebit rationem universalis quam 30. See Quodl. IV, q. 17, OTh IX, p. 385: ' ... conceptus speciei potest abstrahi ab uno intellectio. Sed tale idolum non ponitur propter aliud nisi ut supponat pro re vel ex eo individuo'; and Quodl. I, q. 13, OTh IX, p. 77: ' ... dico quod conceptus generis componatur propositio vel ut sit communis ad res etc., et ista verius competunt numquam abstrahitur ab uno individuo'. intellectioni quam tali ido10. SuperfIue igitur ponitur tale idolum vel fictum' (italics are 31. Ockham admits indeed that the intuitive act is already a similitude of its object. See mine). above chap. 1, n. 46. 15, See Michon 1994, pp. 147-8. 32. Exp. in Porph. 2, 15: ' ... et per ipsa accidentia innotescunt nobis individua substantiae, 16. Exp. in Perih., Prooemium, 10, OPh II, pp. 370-71: 'Ad aliud potest dici quod tale quia accidentia directe et in se cognoscuntur ab intellectu nostro, non sic autem fictum seu idolum plus distinguitur a re extra quam quaecumque res ab una alia; tamen substantiae'. See also Ord. I, dist. 3, q. 2, OTh II, p. 412: ' dico quod nulla substantia in esse intentionali magis sibi assimilantur, in tantum quod si posset produci realiter corporea exterior potest a nobis in se naturaliter cognosci '; Quodl. III, q. 6, OTh IX, sicut potest fingi, esset vere consimile realiter rei extra. Et propter istam rationem magis p. 227: 'Cum igitur de substantiis non habeamus experientiam nisi per accidentia ... ' potest supponere pro re et esse communis et esse illud in quo res intelligitur quam 33.' See Ord. I, Pro!., q. 1, OTh I, p. 23: 'Sed certum est quod intellectus potest habere intellectio vel alia qualitas.' See also Ord. I, dist. 2, q. 8, OTh II, p. 283, ad 2. notitiam incomplexam tam de Sorte quam de albedine.' Similarly, the cameo description 17. Quodl. V, q. 7, OTh IX, p. 506: 'Quarto dico quod intellectus noster de nulla creatura of the cognitive process given in SL 1II-2, 29, OPh I, p. 557, is based on the privileged potest habere aliquem talem conceptum simp1icem proprium sine visione creaturae nee example of tlie simple intuitive grasping of a man, followed by the formation of a simple cum visione; et hoc quia quaelibet talis cognitio sive conceptus aequaliter est similitudo abstractive concept of 'man'. A few pages further in the same chapter, Ockham contrastll et repraesentat omnia individua simillima, et ita non plus est conceptus proprius unius his own case, as somebody who has never seen a lion, with that of a person 'who sees quam alterius' (trans!. Freddoso 1991, pp. 422-3). or has seen intuitively the substance of a lion' (OPh I, p. 560: ' ... ille qui videt vel vidit 18. Quodl. I, q. 13, OTh IX, p. 74: 'Tertio dico quod cognitio prima abstractiva primitate intuitive substantiam leonis'). generationis et simplex non est cognitio propria singulari sed est cognitio communis 34. See in particular John Buridan, Quaest. de Anima (tertia lectura) I, 6, ed. Patar 1991, pp. aliquando, immo semper ... assumptum patet, quia nulla cognitio abstractiva simplex est 786-91; and Quaest. sup. Phys.I, 4, Paris, 1509. On this theory, see Reina 2002, pp. plus similitudo unius rei singularis quam alterius sibi simillimae, nee causatur a re nec 101-36. nata est causari; igitur nulla talis est propria singulari sed quaelibet est universalis' 35. John Buridan, Quaest. de Anima (tertia lect.) I, 6, as edited in an appendix to Patar 1991, (trans!. Freddoso and Kelley 1991, p. 65). p. 790: ' ... directa representatione substantia cum accidentibus confuse representat se 19. Biard 1997, p. 29. sensui, et sensu mediante, intellectui qui illam potest extrahere ... ' 20. Crathom, In Sent. I, q. 2, ed. Hoffmann 1988, p. 168. See Ockham, Quodl. VII, q. 13, OTh 36. See for ex. Ord. I, Pro!., q. 2, art. 1, OTh I, p. 87: ' ... omnia individua eiusdem rationis IX, p. 749: ' ... cognitio quae est conceptus entis est similitudo infinitorum obiectorum ... ' habent effectus eiusdem rationis in passo aequaliter disposito'. The same principle is 21. Crathorn, In Sent. I, q. 2, ed. Hoffmann 1988, p. 169. also found in Onl. I. Prol.. q. I, OTh I. pp. 22-3. Belonging to the same species, actually, 22. Exp. in Praedic. 15, 3, OPh II, p. 292: 'Notandum quod stricte sumendo simile et involves. from (kkham's point of view. having equivalent causal powers. See On/. I, dissimile, nihil dicitur simile vel dissimile alteri nisi secundum qualitatem .... A wider Prol.. q. I. OTh I. p. 42: .... causae eiusdem rationis habent etfeclus eiusdem rationis', _. and more indeterminatc- sense is mentioned hy Ockham immediately after: 'tamen a principle he considers as commonly IICl·cptcd. 142 OCKHAM ON CONCEPTS CONCEPTS AS SIMILITUDES 143

37. Ockham is explicit, for instance, that our concept of 'man', in so far as it is equally a 57. Note, however, that the often striking differences between males and females within a similitude of all men, is that 'according to which we can judge about any thing whether given species of bird, for example, or between youngsters and adults, would not it is a man or not' (' ... secundum quod nos possumus iudicare de quolibet si est homo constitute good examples of this. Ockham could easily admit that these correspond to vel non', Ord. I, dist. 2, q. 8, OTh II, p. 278). This is said within the framework of the different - and non-synonymous - simple absolute concepts. fictum-theory, but it must hold equally well in the actus-theory. 58. The standard presentation of the Twin Earth thought experiment is found in Putnam 38. Quod!. IV, q. 17, OTh IX, p. 386. The text was quoted above in chap. 6, n. 15. 1975a. 39. See Ord. I, Prol., q. 1, OTh I, p. 31: 'Similiter, notitia intuitiva est talis quod quando 59. Quodl. V, q. 9, OTh IX, p. 518: ' ... multitudini nominum synonymorum non aliquae res cognoscuntur quarum una inhaeret alteri ... statim virtute illius notitiae correspondet multitudo conceptuum' (transl. Freddoso 1991, p. 432). See also SL I, 3, incomplexae illarum rerum scitur si res inhaeret vel non inhaeret ... ' OPh I, pp. 11-12, and Quodl. V, q. 8, OTh IX, pp. 510-13. 40. See Ord. I, dist. 2, q. 6, OTh II, p. 211: ' ... ex hoc ipso quod Sortes et Plato se ipsis 60. SL I, 3, OPh I, p. 11: ' ... non videtur magna necessitas talem pluralitatem ponere in differunt solo numero, et Sortes secundum substantiam est simillimus Platoni, omni alio mentalibus terminis'. circumscripto, potest intellectus abstrahere aliquid commune Sorti et Platoni ... ' (italics 61. Ibid., p. 12: 'Unde quaecumque pluralitas et varietas talium accidentium, quae potest are mine). competere nominibus synonymis, potest convenienter a mentalibus amoveri '(italics are 41. See Quod!. I, q. 13, OTh IX, p. 77: ' ... dico quod conceptus generis numquam mine; transl. Loux 1974, p. 53, with a slight amendment, Loux having simply neglected abstrahitur ab uno individuo'; and SL III-2, 29, OPh I, p. 557, where after having to translate the adverb 'convenienter'!). described the formation of the concept 'man' upon a single encounter with a man, 62. About common and deponent voices among verbs, for example, Ockham's conclusion Ockham adds: 'Deinde apprehenso alio animali ab homine vel aliis animalibus, elicitur is the following: 'cum verba communia aequivaleant activis et passivis et deponentia una notitia generalis omni anima1i, et illa notitia generalis omni animali vocatur passio neutris vel activis, et ideo non oportet talem pluralitatem in verbis mentalibus ponere' seu intentio animae sive conceptus communis omni animali.' (SL I, 3, OPh I, p. 13; italics are mine). 42. Ockham does not give any precise clue as to which concept exactly that would be in such 63. See for example Quod!. V, q. 8, OTh IX, p. 510-11: 'Et ideo multitudinem talium a case. Presumably, it would not yet be the general concept of a bird, but something less accidentium quae competunt nominibus synonymis non oportet tribuere naturalibus general. signis cuiusmodi sunt conceptus mentales, sicut nec nominibus synonymis correspondet 43. Quod!. VII, q. 13, OTh IX, p. 749: 'Sicut cognitio quae est conceptus entis est similitudo pluralitas conceptuum' (italics are mine). The same 'sicut' structure is also found in infinitorum obiectorum.' See also Quod!. IV, q. 5, OTh IX, p. 319: ' ... cognitioentis est similar arguments in SL 1,3, OPh I, p. 11, and Quod!. V, q. 8, OTh IX, p. 513. cognitio generalis infinitorum ... ' 64. See for example Brown 1996, p. 79: 'In his writings on semantics and logic, William of 44. Cf. Quod!. I, q. 13, OTh IX, p. 78: 'Semper tamen imprimitur conceptus entis, quia Ockham combines two very stronge/aims about mental language', the second one being quando obiectum est debito modo approximatum, simul causatur a re singu1ari extra 'that mental language contains neither synonymous nor equivocal terms' (italics are conceptus specificus et conceptus entis.' It can even happen that the concept of 'being' mine). be impressed upon the mind without any specific absolute concept whatsoever: if the 65. This was originally suggested by Spade 1975, p. 66. It is very clearly assumed, in intuitedobject is far away, I can realize that something is there in the far distance without particular, by Gaskin 2001, p. 232: 'We can characterize spoken or written synonymy in having the slightest idea of what it is. Says Ockham: 'Manifestum est quod in isto casu terms of subordination to a single mental item, but what would synonymy of mental cognitio abstractiva quam habeo primo primitate generationis est cognitio entis et nullius items be?' Gaskin, consequently, takes 'the absence of synonymy as being constitutive inferioris' (ibid., p. 74). of mental language, rather than as being merely a regulative principle' (p. 231). 45. See above n. 37. 66. See SL I, 6, OPh I, p. 19: 'Large dicuntur illa synonyma quae simpliciter idem 46. As explained in detail in chap. 3 above. significant omnibus modis, ita quod nihil aliquo modo significatur per unum quin eodem 47. See Klima 2002, Brown 1996. modo significetur per reliquum ... ' The same definition is also found in Quod!. V, q. 10, 48. See in particular Klima 1988, 1993, 1999,2001. OTh IX, p. 519. 49. See Klima 2002. 67. The seminal paper, here, was that of John Trentman (1970). Paul Spade, of course, has 50. See SL I, 10, OPh I, p. 35: 'Nomina mere absoluta sunt illa quae non significant aliquid been the pre-eminent defender of this approach in the last decades. principaliter et a1iud vel idem secundario, sed quidquid significatur per illud nomen, 68. Gaskin 2001, p. 232. aeque primo significatur ... ' 51. Ord. I, Pro1, q. 2, OTh I, p. 92: ' ... omnia agentia eiusdem speciei specialissimae sunt effectiva effectuum eiusdem rationis'. 52. See ibid. ' ... dico quod aliquando sufficit unum experimentum ad habendum principium artis et scientiae ... Si enim in principio primo artis et scientiae subiciatur species specialissima, sufficit unum experimentum. Sicut ad sciendum istam propositionem "omnis calor est ca1efactivus", quae est principium primum, sufficit evidenter cognoscere quod iste calor calefacit vel calefecit ... ' 53. Brown 1996, p. 79. 54. Ibid., p. 88. 55. See the passages from Quodl. V, q. 7, and Quodl. I, q. U quoted above in nn. 17-18.

50. See above ,-,hap. ~,sect. 5. Chapter 8 Logical Concepts

In addition to absolute and connotative concepts, the vocabulary of the language of thought, according to Ockham, also contains syncategorematic terms.! These correspond for the most part to what we call today 'logical constants': quantifiers such as 'all', 'every', 'no', 'some', 'neither', etc.; connectors such as 'if', 'and', 'or', 'because', 'when', 'where', 'unless', and so on; prepositions such as 'except'or 'as'; some adverbs such as 'only', 'per se', etc.; negations ofcourse, and even the copula.2 Ockham's most famous thesis about this group of terms is that, in the strict sense of 'to signify', they signify nothing. There are no special objects in the world to which 'all', 'and', 'no' or 'is' could be said to refer:

None of these expressions has a definite and detenninate signification, nor does any of them signify anything distinct from what is signified by categorematic terms. The number system provides a parallel here. 'Zero', taken by itself, does not signify anything, but when combined with some other numeral it makes that numeral signify something new. Likewise, a syncategorematic tenn does not, properly speaking, signify anything; however, when it is combined with a categorematic expression it makes that categorematic expression signify something or supposit for something in a detenninate manner, or it perfonns some other function with regard to the relevant categorematic tenn.3

Ockham's position on syncategoremata thus comes down to what Wittgenstein calls his 'fundamental idea' in the Tractatus: 'the "logical constants" are not representatives' .4 This thesis, however, raises a problem for Ockham's theory of mental language. While simple categorematic terms, whether absolute or connotative, are supposed to be implemented in the mind as a result of an abstraction process rooted in direct encounters with concrete individuals, no such explanation is available for syncategorematic concepts: since they do not signify or represent anything, they cannot be acquired on the basis of particular encounters; there simply is no syncategorematic object in the world for the abstraction process to get started with. Ockham saw the problem very early and gave it a first answer in the Ordinatia from the point of view of the fictum-theory of concepts. Very briefly expounded, this answer has puzzled recent commentators. Yet it is quite interesting philosophically, and I will first reconstruct it in some detail here (section 1). It will then be argued that, even though he is not explicit about it, logical concepts can only be innate in Ockham's later actus-theory (section 2). The status of some apparently non-logical syncategorematic prepositions and verbs such as 'in', 'of', 'before'. 'to have', 'to inhere in', and so on, will finally be briefly discussed (section 3).

14~ LOGICAL CONCEPTS 147 146 OCKHAM ON CONCEPTS about the treatment of connotative concepts in it,? I will now concentrate on the 1. The earlier theory: logical words internalized doctrine it proposes for syncategoremata. What has surprised the commentators about this doctrine is the idea that mental The question of the origin of syncategorematic concepts in the mind is raised in the syncategorematic terms could be derived from the corresponding spoken units. Ordinatio along with that of connotative and negative terms as an objection to the Calvin Normore, for one, sees this as an 'astonishing view' for Ockham to have fictum-theory: held:8 'since spoken language is arbitrary (ad piacitum)', he writes, 'it would make hash of the idea that mental language is natural'.9 And Marilyn Adams expresses a The fourth problem has to do with syncategorematic, connotative, and negative concepts: where can they be taken or abstracted from? For if it is precisely from things, then it is not similar worry: 'This proposalmakes the syncategorematic force that binds terms into seen how they can be distinguished from other concepts. But it is clear that there are such the subject and predicate position of a proposition derivative from that of concepts, since to every spoken proposition, there can correspond a similar one in the conventional language.' 10 To put it in a more general way: all logical operators and mind, and therefore to this proposition 'every man is an animal' and this one 'some man connectors - including the copula - would, apparently, be derived from external language, without which, consequently, thought would be devoid all logical or is an animal', there correspond distinct propositions in the mind; somethingstherefore of corresponds to the in one proposition which is not found in the other. grammatical structure. Logical structure would be essentially conventional. Which is indeed an 'astonishing' consequence for a fourteenth-century Aristotelian thinkerI The two propositions used as examples 'every man is an animal' and 'some man is And Ockham, in addition, would not evenhave felt the need to explain himself about an animal' differ only in their syncategorematic components, but obviously they are this. There is obviously something wrong here. We must look at the text more not equivalent: distinct thoughts correspond to these two sentences. The closely. presumption, then, is that mental language contains distinct syncategorematic The theory presented there comes down to two theses: concepts corresponding to the quantifiers 'every' and 'some'. Where can they come from? (1) There are no natural syncategorematic concepts. Ockham's reply holds in a single paragraph, which is well worth a detailed (2) There are syncategorematic concepts, nevertheless, but by 'institution' only. examination. Here it is, almost in its entirety (the letters between brackets are adqed by me to facilitate further references): Let us consider these two points in turn. Thesis (1), to begin with, cannot mean, as was sometimes believed, that the [a] To the fourth problem, I say that syncategorematic, connotative and negative concepts human mind is naturally incapable of logical operations. This would be radically are not concepts abstracted from things and capable by their very nature to supposit for inconsistent with Ockham's acceptance of both the Aristotelian and the Augustinian things or to signify them in some special way with respect to other concepts. [b] And for frameworks. And it would be incompatible with Ockham's earlier philosophy as well this reason I say that there are no syncategorematic, or connotative, or negative concepts, as with his later theory of concepts. Human thought, in his view, can certainly except by mere institution ... [c] Such concepts, however, can be imposed or they can be combine a subject and a predicate without the help of spoken language; it can form abstracted from words, and this is what happens in fact either always or generally. Strictly speaking, to the spoken word 'homo' there applies such grammatical modes as the singular, universal propositions as well as particular ones, negative propositions as well as the nominative case, the masculine gender, and so on, while to the spoken word 'hominis', affirmative ones, conjunctions as well as conditionals, and so on. What thesis (I) there applies other grammatical modes. Similarly, the spoken word 'homo' signifies a thing asserts, instead, is this: such logical operations do not generate special intelligible determinately when taken by itself, while this does not apply to the spoken word'omnis', objects as categorematic·concepts are supposed to be in the fictum-theory. How which signifies only when taken with some other term. And the same holds for the spoken could they, since syncategoremata represent nothing, while concepts, according to word 'non' and for words such as 'per se' and 'in quantum', 'si' and other such the fictum-theory, have existence within the mind in so far only as they represent syncategorematic terms. [d] From words which thus signify, then, the intellect abstracts something? The key to a sound understanding of Ockham's position here is to common concepts which can be predicated of them, and it imposes these concepts at remember that he is exclusively adopting in this passage the standpoint of the fictum­ signifying the same thing as these external spoken words signify. [e] And in the same way, theory. Saying thatthere are no natural syncategorematic concepts, in this context. it forms with such concepts propositions which are similar to the spoken propositions and amounts to saying that there are no such intelligibleficta. It does not follow from this have similar properties. [f] And just as it can institute such concepts to signify in this way, it can institute the concepts that are abstracted from things to signify in the same that the intellect is incapable of logical operations without the help of conventional grammatical modes as spoken words do. [g] In order to avoid equivocities, however, this is language, but only that such operations, when they occur, are not intelligible objects more convenient with the concepts that are abstracted from the spoken words, for these before the mind, as concepts are supposed to be. They are the intellectual handling concepts are distinct among themselves just as spoken words are, which is not the case for of ideal objects, rather than special objects themselves. They are what the mind does other concepts. [h] And in this way any proposition can be distinguished, for example the rather than what it contemplates. proposition that corresponds to 'homo est homines', or to 'homo est hominis', and so on.1> An important consequence of thesis (I) thus understood is that it ruins the parallelism that is often expected from a theory of mental language hetween mental This is a complex passage, undouhtedly. which, I helieve, has not yet heen and spoken or written propositions. While the latter are linear sequences of discrete satisfactorily explained in the literature. Having said a word in a previous chapter 148 OCKHAM ON CONCEPTS LOGICAL CONCEPTS 149 units, some of which are categorematic and some syncategorematic, mental beings are n:;tturally capable of. The non-linear structure of mental propositions _ propositions, in the view we are presently considering, are operations accomplished composed of both logical acts and categorematic conceptualficta - is, cOllsequently, by the mind upon the sole categorematic concepts. The written sentence 'every horse to be accepted in this doctrine. is mortal', for example, has four successive components, while only two concepts What thesis (2) tells us in this regard is that the mind can nevertheless form, if it would be involved in the corresponding thought: 'horse' and 'mortal', which the so decides, closer mental analogs for spoken and written sentences, and thus institute mind would be handling in a determinate way, performing on them a quantification in the mentalorder a much stricter parallelism with the linguistic order, presumably operation on the one hand, and a on the other hand. Is this a picture facilitating.in this way abstract thought and mental reasoning..How the mind does it Ockham could have self-consciously subscribed to? My contention is that it is. is what Ockham mostly insists upon in the relevant passage. Adam Wodeham, indeed, who was Ockham's pupil at one point and who usually His starting point is that the intellect can form certain concepts by abstracting shows exceptional familiarity with his thought, does attribute such a position to him, them from spoken words, just as it can from anything else in the world. And it can at least with respect to the copula(which he calls a 'comparative act').ll do so in particular on the basis oLthe conventional features of spoken and written Doubts about this attribution, however, have recently been raised by Elizabeth words. This is whatthe development on singular, masculine and nominative case in Karger, on the basis of two main arguments.J2 One of them rests on a passage from sentence [c] is all about: 'Such concepts', Ockham says, 'can be imposed or they can the Reportatio where Ockham clearly treats the natural concept corresponding to the be abstracted from words, and this is what happens in fact either always or 13 copula as afictum endowed with mere objective being. He cannot, therefore, have generally.' This is how we acquire grammatical concepts, such as the concept of identified the copula with an act, Karger concludes. But the passage in question 'nominative' or that of 'singular', and semantical concepts too, such as the concept dates from a very early period in Ockham's career, when he was still accepting the of 'signifying something in a certain mode'. Even more precise metalinguistic existence of special relational entities of inherence; it corresponds to an even earlier concepts can be. formed in the same way, such as the concept of 'the genitive of stage in his thought than what is illustrated by the (favourable) discussion of tb,e "homo"', for example, or the concept of 'the syncategorematic term "if"'. Those are fictum-theory in distinction 2 of the Ordinatio, and one which he soon moved away concepts, indeed, in the sense of the fictum-theory: they are ideal intelligible objects from. It should not, therefore, be considered as relevant in the present context. representing something in the mind. They are metalinguistic concepts, more Karger's second argument is that the identification of the natural copula with an act precisely, since what they represent are linguistic signs. Remember sentence [d] of within the fic'tum-theory would bring about an unacceptable consequence for the quoted passage: 'From words which thus signify, then, the intellect abstracts Ockham, namely that: 'every act of assenting to (or dissenting from) a mental common concepts which can be predicated of them .... ' (italics are mine). The sentence would be an act which, though "direct" with respect to the terms of the concept 'genitive of "homo''', for example; can be abstracted from singular tokens sentence, which are not mental acts, would be "reflexive" with respect to its copula, of' hominis', and it can be predicated of them in the sense that of each one of these which is a mental act' .14 tokens, the following thought is true: 'this is the genitive of "homo'''. We have thus Wodeham saw that consequence and accepted it without qualms,15 as Karger reached in this case a precise grammatical metalinguistic concept.· An even more acknowledges, but Ockham could not have, she contends, since it would entail that directly relevant example for our present discussion is the concept of 'the 'every act of knowledge is partly reflexive', while Ockham admits that 'there are syncategorematic term "if"': I can have the thought that this written word I am unreflective people who are unaware of their mental acts and who are nonetheless looking at is the syncategorematic term 'if'. The mental unit which occurs as capable of knowledge' .16Karger's reference, here, is to Quodlibeta III, 8, where predicate in that thought is a concept representing certain things out there in the Ockham indeed distinguishes two varieties of assent: a reflective one and an world, namely the various tokens of the word 'if'. This concept is not itself, at this unreflective one. 17 The latter, however, which is the one that counts for Karger's stage, a mental syncategorematic term, but a categorematic concept, albeit a argument, is not considered by Ockham as knowledge in the strong sense, as I metalinguistic one. indicated earlier. IS It is.a simple judicative act in the mind, which is equivalent in But now comes the crucial move. Let us carefully read the rest of the sentence: many respects to a complex propositional sequence, but which does not have in itself ' ... and it [the intellect] imposes these concepts at signifying the same thing as these any component parts: 19 strictly speaking, it has no mental subject, no predicate, and external words signify'. Having formed the metalinguistic categorematic concept of a fortiori no copula. Only in the case of reflective assent is a mental proposition with 'the syncategorematic term "if''', the mind can, by decision - by 'institution', a copula directly involved. Contrary to what Karger suggests, then, Ockham would Ockham says - endow this very concept with the same semantical value as is have no difficulty in conceding that whenever a mental proposition does have a conventionally possessed by the word it represents. This is a distinct operation. The copula, the act of assenting to it is a reflective act. metalinguistic concept now receives a new function, not that of representing the It would certainly be, at any rate, a much more unacceptable consequence for him tokens of 'if' anymore (in the sense that the concept of 'horse', for example, if all logical operations should tum out to be conventional! When he holds in represents horses), but that of occurring as a syncategorematic term itse(l within Ordinatio I, distinction 2, question 8, that there are no natural syncategorematicfieta mental sequences. The mind can then form mental propositions which are strictly (as stated by what I called thesis (1), he cannot mean to exclude logical acts such parallel with the corresponding spoken or written sentences. This is precisely the as predication, quantification, negation, and so on, from the range of what human point of part lei in our passage: 'And in the same way. it Ithe intellectI forms with LOGICAL CONCEPTS 150 OCKHAM ON CONCEPTS 151 such concepts propositions which are similar to the spoken propositions and have 5. The metalinguistic concepts thus introduced can then be endowed, by the similar properties.' agent's decision, with a new semantical function, which reduplicates within the What do we find, now, in the curious sentence [f]: 'And just as it can institute such mind that of the spoken or written words which these concepts naturally concepts to signify in this way, it can institute the concepts that are abstracted from represent. A kind of reverse subordination occurs, in virtue of which a naturally things to signify in the same grammatical modes as spoken words do'? It is that the implemented mental unit inherits, by a voluntary institution, the. semantical mind could, if it so wanted, do the same with non-metalinguistic concepts ('the properties of certain spoken or written expressions.21 concepts that are abstracted from things')! I could decide to use my concept of 6. The mind, finally, can use these reassigned concepts to construct mental 'horse' as a genitive in some mental sequences, or maybe even as a syncategorematic sequences which are strictly parallel with spoken or written sentences, and in term, why not? In Ockham's eyes, the advantage of preferring metalinguistic which logical operations - like predication, negation, or quantification - are concepts for this kind of stipulative roles is that they 'are distinct among themselves objectivated into ad hoc units. Although Ockham does not say it in so many just as spoken words are, which is not the case for other concepts' (sentence [g]). The words, it is to be presumed that such mental objectivations facilitate reflection mind can .thus form a special mental sequence for each spoken proposition, and reasoning by displaying, so to say, the logical processes before the mind; including those that make use of both the nominative and the genitive of the same word, for example, such as 'homo est hominis', or both the singular and the plural of Even independently of the fletum-theory with which Ockham associated it, this the same word, such as 'homo est homines' (sentence [h]). approach is quite remarkable. What it draws attention to is that the process of This, then, is the theory proposed to us in this somewhat surprising page of the 'thinking with words' -often taken for granted in recent philosophy - requires that Ordinatia. Let us recapitulate, for the sake of clarity, the various stages of the some of our naturally acquired mental representations - our representations of spoken cognitive process it hypothesizes: or written items, more precisely -,--should be semantically reassigned according to linguistic conventions. If so much is admitted, then how such a process works exactly, 1. There is, first, the normal formation of simple categorematic concepts through and how it bears upon the nature and scope of human thought, is certainly something abstraction, as described in the Prologue of the Ordinatio. According to the text that today's cognitive scientists should be interested in investigating. we are now discussing, no syncategorematic concepts, or connotative or negative concepts are formed in this way.20 2. The mind can, from there, accomplish various operations uponthe concepts so 2. Logical constants in the actus-theory formed. It can combine them, in particular, through predication, negation, conjunction, and so on, and it can quantify over them. These operations are However that may be, Ockham eventually renounced this approach to mental intellectual handlings of previously acquired categorematic concepts, but they syncategoremata as he moved away from the fletum-theory. It is true, as Beatrice are not concepts themselves. Since it is not acquired on the basis of experience, Beretta remarks, that there seems to be a later allusive reappearance of it in the the aptitude to accomplish such operations must be seen as innate within the Quodlibeta,22 but one should probably suppose that the relevant passage - in which framework of Ockhamism (more on this in the next section). Ockham introduces a wide sense for 'second intention' - had originally been written 3. These intellectual operations can be conventionally expressed in spoken or at a time when he was still favouring the fictum-theory and that he failed to adjust written languages by special words, the function of which is to indicate to the this particular sentence as he integrated the whole surrounding development into the hearer (or the reader) which mental operations are accomplished by the final version of the Quodlibeta.23 It is clear at any rate that the actus-theory of speaker. These special words are the syncategorematic terms of spoken and concepts does allow, in Ockham's own eyes, for natural mental syncategoremata. written languages. Strictly speaking, they must not be said to signify the This is explicit in the very presentation of the actus-theory that he inserted in the operations they express, since they are not used to refer to these operations or revised version of distinction 2, question 8 of the Ordinatio, just a couple of pages to connote them, but their presence can modify the signification, supposition or after the discussion of syncategorematic terms that we have been scrutinizing in the truth-conditions of the other linguistic units they are combined with, according previous section. In the new account, he says, to what logical operations they correspond to. 4. Like any other things in the world, linguistic items - and in particular the ... just as among spoken words and conventional signs ... there are Some that do not syncategorematic words introduced at the previous step - can be intuitively signify things, but merely consignify along with others, such as syncategorematic .terms grasped by the intellect and thus bring about the formation of categorematic ..., similarly there are some qualities subjectively existing in the mind which ­ concepts capable of suppositing for them within mental propositions. Some of proportionally - have by their very nature such features as these spoken words conventionally have. 24 these metalinguistic concepts will be quite general, such as the concept of 'genitive' or 'masculine', and some will be much more narrowly Concepts. in general, in this approach, are no more seen as ideal objects produced circumscribed, such as the concept of 'the genitive of "homo'" or that of 'the by mental acts and liable to logical operations. They are mental acts themselves - feminine of "albus'''. 152 OCKHAM ON CONCEPTS LOGICAL CONCEPTS. 153 operations, that is, or states of the mind -'-which do not require special intelligible distinct units just as the spoken or written sentence is, and one of them corresponds correlates. The parallelism between mental and spoken or written propositions, to the copula. The same thing, moreover, must hold mutatis mutandis for the other therefore,is now more easily secured: mental propositions are concatenations of acts syncategorematic logical acts such as negation, conjunction, quantification, and so just as spoken and written sentences are concatenations of words; it becomes quite on. Even though a mental act of negation, for example, is not normally accomplished natural to postulate a distinct conceptual act in the mind for each semantically alone in the natural order, it suffices, when it is added to a bundle of categorematic relevant word in spoken or written discourse, whether categorematic or concepts plus copula, for turning the whole group into a negative mentalproposition, syncategorematic.25The overall picture can thus be significantly simplified. of which it is itself a distinct simple component.33 This is not to say that syncategorematic concepts. are now treated on a par with While clearly realizing that such was Ockham's position -at least with respect to categorematic ones. They are still not granted a determinate signification of their thecopula,34 Marilyn Adams raised interesting objections against it in the guise of own and their semantical role continues to be considered as subordinate. In the two counterexamples, and it turns out to be quite instructive to see just how the actus- as well as in the fictum-theory a syncategorematic conceptual act has no doctrine presented here can handle them. The key to it is to take seriously the semantical relevance if it is not accompanied by some categorematic concepts. Take requirement that the relevant mental units should be concatenated into wholes for the paradigmatic case of the copula. The corresponding mental term, Ockham says propositional acts to occur. Adams's first counterexample is a situation in which 'I in Quodlibeta VI, question 29 (written from the point of view of the actus-theory) is think the proposition "an oak is a tree" and think the concepts· of man and animal, 'the union of the extremes of a mental proposition'. It is not an intrinsically without thinking the proposition "a man is an animal"..'35 In this case, the relational unit for all that, but 'a quality of the mind, viz., an act of understanding'. categorematic concepts 'man' and 'animal' are both present in the mind along with 'And this concept', Ockham adds, 'is really distinct from the subject and the a copula (the 'is'of 'an oak is a tree'), but the mental proposition 'man is an animal' predicate, which are also diverse acts of understanding.'26 Yet being an act of 'union\ is not formed for all that; therefore, Adams correctly concludes, the merecopresence it seems clear that contrary to the categorematic terms, it cannotnormally be posited of the copula with two categorematic concepts is not sufficient for the corresponding alone. And the same must be true of all other syncategoremata. One can, of course, mental proposition to be formed. What is relevant in the example is that the copula think of a syncategorematic term without forming a proposition; the mind, then, has in question is not mentally concatenated with the categorematic concepts of 'man' a second-order categorematic concept representing the syncategorematic term ­ and 'animal' into a single mental unit. As we saw in chapter 2, the mereological which is what Ockham calls a 'second intention' (intentio secunda);27 but it is not grouping of mental items into distinctive wholes must be a cruCial process in actually producing in this circumstance a truly syncategorematic mental act as such. Ockham~sa.ccbunt of how complex mental propositions are structured.36 'Omnis What is the difference, Paul Spade asks at one point, 'between merely having the homo est animal' and'omnis animal est homo', for example, differ from each other three concepts "Socrates", "is", and "mortal", and having the mental proposition or in having different parts: 'omhis homo' occurs in the former and 'omnis animal' in judgment "Socrates is mortal"'?28 Ockham's answer to this, in the framework of the the latter. Adams is right that the mere copresence of certain units within the mind is actusctheory, must be that there is no difference there, as long, at least, as the three not sufficient for the corresponding complex item to be there as well. Some sort of mental items are concatenated into a single whole. As we just saw, the binding act is required. But the mental groqping of parts into wholes is enough to do syncategorematic copulative act is described in Quodlibeta VI as the 'union' of the the job in Ockham's view. The principle is that whenever two or more simple acts subject and the predicate. This means that whenever such a copula accompanies a are thus assembled, they form a new complex act, the semantical properties of which subject-term and a predicate-term within a given mental complex act, a mental depend only on those of the components, and which can in tum oCcur as a part in proposition is ipso facto constituted. No additional binding act is required, except the some further grouping.37 mereological concatenation itself.29 This does not distort the usual Ockhamistic Adams's second counterexample, however, can be seen as a challenge to parallelism between mental and spoken or written propositions, as Spade thinks it precisely that idea.38'A man is an animal' and 'an ahimalis a man', she points out. does,30 since the mental subject term, predicate term and copula are still seen as three have exactly the same components, but 'man' is the subject in the former and the distinct 'absolute' units, just as the corresponding spoken or written words are.31 predicate in the latter;mereological grouping, therefore, is not sufficient for Spade's argument is that if the copula is what binds the extremes together, then it is determining which of the two propositions we end up with when these components not a 'separable ingredient', and, consequently, not a distinct component either. But are mentally assembled. One way out, apparently, would be to concede that these this is an inference Ockham would reject,since he both holds that the copula is a two propositions, which are semantically equivalent, are in fact indistinguishable in distinct component and that its addition to the other two suffices to bind them the mind.39But Ockham, I suspect, would not be too happy with such an approach together into a proposition, just likethe addition of the written word 'est' to 'homo' since it entails that in mental particular affirmative propositions (such as 'homo est and' animal' within a single sequence suffices for turning the resulting whole, 'homo animal'), it is indeterminate which term is the subject-term and which the predicate­ animal est', into a written sentence. There is an interesting difference between the term. A more plausible solution within the Ockhamistic framework would be to two cases, admittedly, since the order is relevant in the written sequence while it is postulate a special mental syncategorematic term which turns the categorematic not supposed to be in the mental corresponding concatenation.32But this does not concept with which it is immediately grouped. into the subject-term of the substantially affect the present point: the mental proposition is composed out of proposition in which they both occur (and maybe another special predicate forming 154 OCKHAM ON CONCEPTS LOGICAL CONCEPTS 155 syncategorematic functor as well). Ockham nowhere considers such a possibility, but 3. Prepositions and non-standard copulas it is mentioned in one of his anonymous successors as an integral part of the doctrine that mental propositions are complex units.40And it does seem to be a rather natural Our focus in this chapter up to now has been on logical constants: quantifiers, elaboration on Ockham's own general approach. It saves at a low cost both the connectors, the copula 'est', the negation, and so on. These correspond indeed to subject-predicate structure for all atomic propositions, and the mereological Ockham's standard examples of syncategorematic terms.51But the category in his principle of compositionality identified above. Ockham certainly could have view also includes a bunch of terms which cannot easily be classified as logical accepted it without any great drawback. constants in the modern sense. Think of prepositions such as 'in', 'for', 'from', and The actus-theory thus provides a rather elegant account of logical concepts: they the like; or of verbs like 'to have' or 'to inhere in', which Ockham regularly treats as are functional mental acts of a special sort, that normally do not occur alone in the copulas. Those might well be the sort of terms Calvin Normore had in mind in the mind, but that can be combined with categorematic conceptual acts to form complex argument just discussed at the end of the previous section; and it must be conceded units endowed with precise semantical roles, that of being a quantified subject or that the account offered so far does not smoothly extend to such cases. How are we predicate term, for example, or a predicative proposition, or a negative one, or a to deal with them, then? Ockhamdidn't say much on the matter. The important thing complex proposition such as a conjunction, and so on. fOThim was to acknowledge the existence of non-referring auxiliary terms (whether As to the original problem raised in the Ordinatio - where do such logical spoken, written, or mental), of which he took logical constants to be the more salient syncategorematic concepts come from? - there is only one plausible answer to it in instances.52 Where exactly the limits of that category are to be drawn is not the context of Ockham's later theory as well as in the former: these logical acts something he cared to investigate in any detail. This does not mean, however, that correspond to innate capacities of the mind. Although he is not explicit on the matter, his system is without resources for facing the question of what concepts in the mind as Beretta stresses,42 Ockham can be shown to be committed to in this case correspond to prepositions and verbs. Two complementary approaches could be by the following straightforward considerations. Just as about every medieval explored. Since Ockham doesn't do it himself, I will be content to simply sketch philosopher does, he often repeats the Aristotelian motto that the specific difference them here, and to outline one important aspect which they share. of human beings with respect to other animals is that they are endowed with The first possibility is to consider some spoken or written prepositions such as . This 'potentia' or aptitude - which he squarely identifies with the 'in', 'with' or 'after', and verbs such as 'to have' and 'to inhere', as expressing in intellectual soul43- certainly has to be innate in his eyes, since it constitutes the main fact categorematic relational concepts rather than genuine mental syncategoremata. distinctive feature by which a member of the human species essentially differs from It is true that all prepositions are cursorily said by Ockham to lack determinate and a horse or a donkey.44Now, a rational being is defined by Ockham, unsurprisingly, as finite signification.53 But maybe this can be seen as a loose formulation. What a 'being or substance that can reason [ratiocinari]';45 and the capacity for reasoning, matters for him in such cases is that no special 'prepositional' objects be admitted in in his view, does require the capacity for assembling propositions, making negations, the ontology as denotations for 'in', 'from', and the like, and this constraint would drawing inferences, and so on.46The consequence follows: the basiclogical aptitudes be satisfied if the corresponding concepts were analysed as having a primary and a must be innate in rational beings. But syncategorematic logical concepts are nothing secondary signification each pertaining to nothing but singular substances and but actualizations of such aptitudes in Ockham's theory. The possession of logical qualities. The relational concept corresponding to 'in' in the spatial sense, for concepts, therefore, has to be seen as an innate feature of human beings. instance, could be said to primarily signify every substance or quality which is inside While considering this as a 'plausible' option, Calvin Normore thinks that other something, and to secondarily signify (or connote) every substance or quality which possibilities are also open to Ockham for explaining the origin of mental is such that something is inside it, just as 'taller' is supposed to primarily signify syncategoremata.47The only one he mentions, however, corresponds to what we every thing which is taller than something else, and to connote every thing which is have identified in chapter 6 as Ockham's account for the implementation of simple such that something else is taller than it. connotative concepts,48 and it could hardly apply to logical constants: 'If the objects This comes even more naturally for a verb such as 'to inhere' ('inhaerere' or of intuition are sufficiently complex', Normore says, 'if; for example, one intuits not 'inesse'), which is used to connect qualities with substances (as in 'albedo inest just Socrates's paleness or Socrates himself but, say, pale Socrates sitting on a fence, Sorti': 'a whiteness inheres in Socrates'), and which Spade, for one, has suggested then it may be that the syncategorematic elements of the complex are presented in at one point to treat as a special syncategorematic copula within Ockhamism.54 intuition.'49 But since syncategorematic terms do not signify any special external Ockham himself, in an important passage, puts the cognition corresponding to this objects, according to Ockham, there simply is no such syncategorematic element in particular verb on a par with any other intuitively ascertainable relational concept.55 the external intuited complex in such cases. Relational and other connotative terms, Why shouldn't he? It is true, as Adams points out, that he had previously rejected by contrast, do have external significates: those are not relations, of course - which such a connotative analysis of 'to inhere',56but this was when he still accepted have no independent reality according to Ockham - but n-tuples of singular inherences as special relational objects in the world, a position he abandoned very substances or qualities, as we have repeatedly insisted upon in chapter 6. When soon in his career. 57 As he reduced ontology to substances and qualities, he came to complex groups of objects are intuited, then, what is implemented in the mind are see relational tenns in general as connotative, and there is no obvious reason why connotative concepts, not syncafl'gol'l'mafa. 50 this should not hold good for the concept of inherence. That Ockham sometimes 156 OCKHAM ON CONCEPTS LOGICAL CONCEPTS 157

treats 'inest' as a copula58.is no objection here since he often uses a very general As Spade remarks, 'there is no evidence that Ockham himself would have accepted sense of the technical term 'copula', according to which it applies to any relational [this] development ofhis theory of connotation'.67 But something like this does seem verb whatsoever.5?Just as 'to see' can be analysed in mentalese into 'to be seeing', to be called for, as I said earlier, if a variety of modes of connotation is to be posited 'to inhere in' in this approach would be analysed into a regular mental copula, 'to in the language of thought, as Ockham is committed to. be', and a connotative concept'inhering in'. His treatment of oblique cases (genitive, accusative, and the like) turns ouno be Being~onnotative,this concept 'inhering in' would be liable to a nominal directly relevant here. For one thing, he clearly admits a variety of them in the definition, like any other connotative term. Contrary to what Spade and Gaskin language of thought,68 and this, as we have seen, simply amounts to accepting believe,60 this would not bring about an infinite regress. 'Inhering in' is an different sorts of grammatical complements.69It seems natural to suppose -although asymmetrical relational term in this' approach, and this means, according to Ockham does not say so - that each such distinctive sort of complement is associated Ockham's view of suc):l terms, that any language where it appears,. including in mental sentences with a distinctive syncategorematic indicator (corresponding, for meritalese, contains a converse relational term ~ 'inhered in', let's .say - which instance, to 'in', 'with', 'on', or maybe 'having', and so on). This grammatical should occur in obliquo in its correct complete nominal definition, and in the variety, on the other hand, is connected with the plurality of possible modes of definition of which,in turn, it should occur itself in obliquO.61The complete connotation by the theory ofnominal definitions: what I have called condition C3 for Ockhamistic nominal definition of 'inhering in' would be something like 'a quality such definitions in chapter 5 requires in effect that the precise grammatical functions with respect to which a substance is inheredin', and the definition of 'inhered in' of the terms in obliquo within a nominal definition (as revealed by their grammatical would be something like 'a substance with respect to which a quality is inhering in'; cases and/or by the prepositions or the copulas which govern them) should indicate just as the definition of 'taller', as we have seen, should be something like 'a body the modes under which the secondary significates of the defined term are connoted.70 with respect to which a body is shorter', and the definition of 'shorter': 'a body with Now, ifthere are simple connotative concepts, as Ockham admits, and if-as he also respect to which a body is taller'.62As I repeatedly stressed, circularity in such caSes holds ~everyconnotative term has one, and only one, correct nominal definition,7\ is not a problem, given what nominal definitions are intended for in Ockham's it follows that there must be in the language of thought one primitive and distinctive system:. What matters is that the two definitions, taken jointly, make it perspicuous way of being a grammatical complement for each primitive mode of connotation. that only substances and qualities are needed as primary and secondary significates And if each sort of grammatical complement is associated with a distinctive of these two correlative terms. Circularity, on the other hand, efficiently prevents syncategorematic indicator, there should be at least one distinctive syncategorematic infinite regress 'in this case. Spade and Gaskin think otherwise because they concept in the mind for each distinctive mode of connotation. Since terms in obliquo mistakenly take the Ockhamistic programme to aim at the ultimate elimination of all can occur as predicates or even subjects according to Ockham,72 these should also connotative terms from the language of thought through nominal definitions. correspond to different modes of predication. Which brings us very close to the A' second fruitful approach for the treatment of such prepositions and verbs Buridanian view mentioned by Spade. within Ockhamism would be to follow an intriguing suggestion, from Paul Spade, Letus take an example. Ockham refers at one point in the Summa Logicae to a and to postulate primitive mental syncategoremata corresponding to certain special grammatical situation where 'the oblique case is governed by the force of terms like 'inhering', 'having', 'doing', and a few others.63The idea is found in possession',73as in the Latin sentence 'aliquis asinus est Sortis' (literally: 'some Buridan, for example, who rightly noticed that the nominalistic theory of donkey is Socrates's'), where the genitive 'Sortis' indicates a possession connection. connotation requires the acceptance of several distinct modes of connotation. Since Ockham speaks there of the oblique case being 'governed' [regitur] by the 'Father', presumably, does not connote the children in the same way 'horseman' 'force of possession' [ex vi possessionisJ, it is permitted to extrapolate that the term connotes the horses or 'white' connotes the whitenesses. If it did, the nominal 'Socrates' in the corresponding mental sentence should indeed be grammatically definitions of these connotative terms could not distinctively spell out how their governed by a special copula (such as 'is had by') or a special possession preposition , primarysignificates should be connected in reality with theirconnotata, as required (such as the postposited 's' in English or the preposition 'd' in the French by what I called condition C3 for a good nominal definition.64Buridan spoke of corresponding sentence 'un fine est d Socrate'). Of course, it is not clear in this distinct 'modes of adjacence' in this respect (modi adiacentiae);65 he conjectured particular case that the possession connection in question does correspond to a that to every such mode there corresponds a distinct mode of predication in the primitive mode of mental connotation for Ockham (although it might), but the point language of thought, and a distinct' Aristotelian category of simple accidental here is that such a distinctive copula or preposition should be innately available for categorematic terms. Spade appropriately quotes on the matter the following passage each variety of connotation that the human mind is spontaneously capable of. from Buridan's : whatever these varieties are. These two approaches to prepositions and copulas - the relational one and the Third, note that in accordance with the different positive ways of joining connoted things grammatical one - could indeed be combined within Ockhamism. It is possible to lathe things to which the terms refer, there arise the different modes of predkating. For suppose hath that the human mind is equipped with a limited numher of innate example, how, how much, when, where, how this is related to that, and so on. From these capacities for different sorts of complementation and predication (as in Buridan's different modes of predicating, the different categories arc taken."" suggestion), and that. nevertheless. all other spoken or written prepositions and LOGICAL CONCEPTS 159 158 OCKHAM ON CONCEPTS copulas are represented in the human mind as categorematic relational concepts. As down the same chapter, he makes it clear that all conjunctions and prepositions are to be to which is which, given the complexity of the problem and its ultimately empirical treated in the same way (p. 16). Elsewhere, the negative sign 'non' is also given as an example of syncategorematic term (see SL III-3, 7, OPh I, p. 613). As to the copula, see character, it is not surprising that no solution to it is even sketched in Ockham. Quodl. VI, 29, OTh IX, p. 695 (the passage is quoted above in chap. 5, n. 28, where other Seriously trying to devise a list of the basic modes ofmental connotation would most references are also given). The case of adverbs is somewhat special, since some of them probably have led to an in-depth revision of the Aristotelian theory of categories, a are considered as partially categorematical (see SL I, 4, OPh I, p. 16: 'De quibusdam revolutionary move in the context of the early fourteenth century, which even autem adverbiis aliter est, quia quaedam eorum determinate significant illa quae Ockham might not have felt ready to make, had he been tempted to. 74 significant nomina categorematica, quamvis alio modo significandi important'); Whichever way we choose for elaborating it, however, it is clear that the 'courageously' would be a good example of this: it obviously includes some sort of Ockhamistic approach to prepositions and copulas should ultimately involve some categorematical reference to courage. further innate mental apparatus in addition to purely logical capacities. Whether they 3. SL I, 4, OPh I, p. 15: 'Termini autem syncategorematici ... non habent finitam are syncategorernaticor not, not all the required concepts could plausibly be significationem et certam, nec significant aliquas res distinctas a rebus significatis per acquired on the basis of perceptual experience without some sort of antecedently categoremata, immo sicut in algorismo cifra per seposita nihil significat, sed addita alteri figurae facit earn significare, ita syncategorema proprie loquendo nihil significat, available aptitudes. As we have previously seen, the human mind, in Ockham's sed magis additum alteri facit ipsum aliquid significare sive facit ipsum pro aliquo vel picture, must be able to judge that certain things are ordered in certain ways with aliquibus modo determinato supponere vel aliud officium circa categorema exercet' respect to others,75 and while some of these orderings might be perceptually (transl. Loux 1974, p. 55). recognizable, this cannot be generalized. The tendency to associate perceptible 4. Wittgenstein 1921,4.0312 (transl. Pears and McGuinness 1961, p. 22). qualities with underlying substances through inherence-judgements is a salient case 5. Ord. I, dist. 2, q. 8, OTh II, p.. 282: 'Quartum dubium est de conceptibus in point: in order to intuitively recognize an inherence connection in the first place, syncategorematicis et connotativis et negativis: unde possunt sumi vel abstrahi? Quia si one must, apparently, prepossess, in one form or another, the concept of 'inhering praecise a rebus, non videtur quomodo possunt distingui ab aliis conceptibus.Quod in'. And special dispositions must also be supposed for causal judgements, autem sint tales conceptus patet, quia omni propositioni in voce potest correspondere mereological judgements, and maybe a number of others as well. The theory, consimilis in mente, igitur isti propositioni "omnis homo est animal" et isti "aliquis consequently, requires the postulation, in one way or another, of an innate array of homo est animal" correspondent distinctae propositiones in mente; igitur aliquid correspondet signo in una propositione quod non correspondet in alia.' categories, which Ockham - unsurprisingly - never fully explored. This is not to say 6. Ibid., pp. 285-6: '[a] Ad quartum dubium dicerent [or rather dico according to ms. A that he was unconsciously anticipating some sort of Kantian with an a priori and F] quod conceptus syncategorematici et connotativi et negativi non suntconceptus structured mind on the one side and an ultimately uncognizable world in itself on the abstracti a rebus ex sua natura supponentes pro rebus vel ipsas modo distincto ab aliis other side. Ockham clearly admits that although orders are not absolute things in conceptibus significantes. [b] Et ideo dicerent [or rather dico according to ms. A, B, C themselves, the ordering capacities of the mind adequately correspond in principle and F] quod nullus [est according to ms. C, D, E and F] conceptus syncategorematicus to the organization of external reality. This is not as nai've as it may sOllnd: how, nec connotativus nec negativus, nisi tantum ex institutione ... [c] Possunt autem tales otherwise, could human beings survive in this world? conceptus imponi vel conceptus abstrahi a vocibus, et ita fit de facto vel semper vel Whether in.the fictum- or in the actus-theory of concepts, the human mind, in communiter. Verbi gratia, isti voci "homo" competit talis modus grammaticalis quod est conclusion, is endowed with a rich innate set of aptitudes. Some are called for by its singularis numeri, nominativicasus, masculini generis, et sic de aHis; et isti voci very rationality, in order to account for logical operations such as predication, "hominis" competunt alii modi grammaticales. Similiter, isti voci "homo" competit quod significet determinate rem per se; isti voci "omnis" non sic competit, sed quod quantification, negation, conjunction, etc. When actualized in mental discourse, they significet tantum cum alio. Similiter est de ista voce "non" et de istis "per se" et "in correspond to syncategorematic conceptual acts. Others, in addition, are required to quantum", "si" et huiusmodi syncategorematicis. [d] Tunc ab istis vocibus sic account for the various sorts of basic relational judgements that the mind is capable significantibus abstrahit intellectus conceptus communes praedicabiles de eis, et imponit of. It is left open in Ockham's writings whether these should be thought of as istos conceptus ad significandum illa eadem quae significant ipsae voces extra. [e] Et syncategorematic copulas and prepositions or as specialinnate connotative concepts, eodem modo et de talibus format propositiones consimiles et habentes consimiles the more promising approach probably being a combination of these two proprietates quales habent propositiones prolatae. [f] Et sicut potest instituere tales possibilities. conceptus ad sic significandum; ita potest instituere ipsos conceptus abstractos a rebus ad significandum sub eisdem modis grammaticalibus sub quibus significant ipsae voces. [g] Hoc tamen fit convenientius per conceptus abstractos a vocibus propter aequivocationem vitandam, quia illi conceptus sunt distincti sicut ipsae voces, quamvis Notes non omnes sint distinctae; conceptus autem alii non sunt distincti. [hi Et ita quaelibet talis propositio esset distinguenda, puta propositio correspondens tali propositioni 1. See SL I, 4, OPh I, p. 15: 'Adhuc aliter dividitur terminus, tam vocalis quam mentalis, "homo est homines", "homo est homini", et sic de aliis.' In sentences [a] and lbl, it quia terminorum quidam sunt categorematici, quidam syncategorematici' (the italics are seems prohahle that Ockham originally wrote 'dico' at a time when he was still mine). suhsrrihing tOlhe.!icllllII-theory, and laler changed to 'dicerell/' as he moved towards lhe 2. Ockham's paradigmatic examples sy"cale!-iOl'/,lIIallJ in SL I. 4, OF" I, p. 15 arc of lU·III.\·-lheory. Thai even then, he uses 'dic(,/,/'II/' ('Ihey would say') ralher Ihan 'dinll//' 'omllis'. '11111111.1",'alil/llis'. '10111.1", '!,ral'lel", 'Iallllllll' and 'illl/IIlII/IIII11'. A lillie further 160 OCKHAM ON CONCEPTS LOGICAL CONCEPTS 161

('they say') shows that he did not find this doctrine in other writers, but devised it intelligendi. Et iste conceptus distinguitur realiter a subiecto et praedicato, quae sunt himself. In sentence [b], on the other hand, the insertion of 'est', as found in some good etiam diversi actus intelligendi' (transI. Freddoso 1991, pp. 584-5). manuscripts, makes more sense. 27. See SL I, 12, OPh I, pp. 43-4. Second intentions are defined there as those concepts that 7. See above chap. 6, sect. 2. signify first intentions; and. first intentions (in the wide sense) are said to include mental 8. Normore 1990, p. 59. syncategoremata (p. 43, lines 55-7). 9. Ibid., p. 60. 28. Spade 1996, p. 134. 10. Adams 1987, p. 301. Adams provides a presentation of this particular theory on pp. 29. Spade 1996, p. 134, singles out a similar answer as being 'perhaps' Buridan's (his 298~302.Fot other short discussions of it, see Gelber 1984, pp. 148-53, Karger 1996, reference is to Buridan's Sophismata I, ed. Scott 1977, p. 32: ' ... haec dictio "est", prout pp. 226-7, and Maieru 2002, pp. 5ss. praecise est copula, nihil significat ad extra ultra significationem terminorum 11. See Adam de Wodeham; Lectura Secunda in I Sent., ed. Wood and Gal 1990, vol. I, pp. cathegorematicorum, sed significat solum illum conceptum complexivum quo 147-9. intellectus format propositiones ex. istis terminis ... '): but whether it is Buridan's or not, 12. See Karger 1996, pp. 221-6. a version of it is clearly found in Ockham. See Quodl. VI, q. 29, OTh IX, p. 696: 'Potest 13. Rep. II, q. 1, OTh V, pp. 18-23. See,for instance, p.. 20: ' ... per copulam importatur enim aliquis absolute cognitione incomplexa intelligere hominem etanimll1, et tamen conceptus absolutus qui habet tantum esse obiective in anima ...' nec homo erit subiectum nec animal praedicatum. Et hoc, quia deficit iste conceptus 14. Karger 1996, p. 222. syncategorematicus "est", quo pO$ito, sine omni alia respectu statim homo erit 15. See A. de Wodeham, Lect. Secunda, ed.Wood and Gal 1990, vol. I, p. 150. subiectum et animal praedicatum, et habetur tota propositio' (italics are mine). The 16. Karger 1996, p. 223 n. 96. italicized part of this quotation plainly states that the addition of the mental copula to a 17. See Quodl. III, q. 8, OTh IX, pp. 233-4; the relevant passages are quoted above iqchap. pair of categorematic concepts suffices to bring about a mental proposition. 2, nn. 99-103. 30. See Spade 1996, pp. 134-5. 18. See above chap. 2, sect. 4.3. 31. See Quodl. VI, q. 29, OT,h IX, p. 695: 'Sicut in ista propositione prolata "homo est 19. See Quodl. III, q. 8, OTh IX, p. 234: ' ... et iste actus aequivalet quantum ad multa alicui animal" unio extremorum ponitur istud verbum ~'est",quod est quaedam qualitas complexo quo aliquid scitur'(italics are mine). Ockham's. use of 'aequivalere' in this absoluta, ita proportionaliter est de unione istorum extremorum in mente, quod est context refers us back to his hypothesis that some mental propositions are actually conceptus' copulae, quae est qualitas quaedam mentis, puta actus intelligendi. Et iste simple acts that are equivalent in some respects with complex sequences. See the conceptus distinguitur realiter a $ubiecto et praedicato, quae suntetiam diversi actus passages from Exp. in Perih., Prooemium, and Quaest. in Phys., q. 6 quoted above in intelligendi' (italics are mine). chap. 2, nn. 83-6, as well as the discussion I propose of these in sect. 4.2 of the same 32. Seen. 25 above, as well as chap. 2, sect. 4.2. chapter. 33. Note that since syncategorematic acts such as the copula or the negation are 'absolute' 20. As argued earlier, however, Ockham admits, even at this stage, that some relational qualities, according to Ockham, he ~scommitted to accept thatGod at leastcould make concepts cart be acquired by abstraction. It must be concluded that he probably uses the these occur alone .if He so wished. But this is no more surprising than the possible word 'connotative' in this passage in its narrow sense, which does not include relational supernatural occurrence of a whiteness without a substratum or, for that matter, of a terms (see above chap. 6, nn. 10-11). categorematic conceptual act without an underlying thinking mind. 21. More on reverse subordination in chap. 9, sect. 2 below. 34. See Adams 1987, pp. 259-62. Adams bases her interpretation on a passage from the 22. See Beretta 1999, p. 172. The author refers to Quodl. IV, q. 35, OTh IX, p. 471, where Reportatio where Ockham firmly states that the copula is a distinct absolute unit rather Ockham does indeed speak of 'signa mentalia ad placitum signijicantia ... puta than a relational one (Rep. II,I, OTh V, pp. 18-19), but this is, as we saw, a thesis he syn(:ategoremata mentalia'. still maintains in Quodl. VI, q. 29, while adapting it to the actus-theory. 23. This conjecture tends to be supported by the fact that there is no such reference to 35. Adams 1987, p. 262. conventional mental syncategoremata in the otherwise parallel passages of the Summa 36. See above chap. 2, sect. 4.2. Logicae, where Ockham defines the wide senses of 'nomen secundae intentionis' (SL I, 37. It follows that contrary to what we find in 's (see, for 11, OPh I, p. 40, lines 57-60) and 'intentio secunda' (SL I, 12, OPh I, p. 44, lines 75-7). example, Goodman 1956), the identity of the ultimate parts are not sufficient here for 24. Ord. I,'dist. 2, quest. 8, OTh II, pp. 289-90: ' ... sicut voces et signa voluntarie instituta determining that of the resulting whole, since 'omnis homo est animal' and 'omnis .., quaedam sunt quae non· significant sed tantum consigrtificant cum aliis, cuiusmodi animal est homo' do have the same four ultimate components. Mereological grouping, sunt syncategoremata ..., ita sunt quaedam qualitates exsistentes in mente subiective, then, has to be granted an arborescent structure in the Ockhamistic approach. quibus ex natura competunt talia proportionabiliter - qualia competunt vocibus per 38. Adams 1987, p. 262. voluntariam institutionem' (italics are mine). The addition to the Ordinatio from which 39. In modern logic, both would be expressed by the same formula: '(::Ii:) x is a man /\ .r is these lines are drawn is placed by the editors under the suitably descriptive heading: an animal'. And Ockham admits, as a matter of fact, that in 'homo est animal' hoth 'Opinio posterior auctoris' (p. 289). extremes have the same mode of supposition (see SL I, 70, OPh I, p. 210: ' ... in ista 25. Remember, however, that since mental propositions, for Ockham, are not extended in propositione "homo est animal" utrumque extremum habet suppositionem space or in time, they cannot be thought of as displaying a simple linear sequential order determinatam '). as spoken and written words do. See my remarks on this in chap. 2, sect. 4.2 above. 40. The text is quoted hy Spade 1996, PI'. 12R-9. His reference is to Vienna, 26. Quodl. VI, q. 29, OTh IX, p. 695: ' ... dico quod unio extremorum propositionis in mente Nationalhihliothek, Pal. Lal. MS 4X5.1, 1'01.Imr. est conceptus syncategorematicus verhi copulativi sive copulantis suhiectum cum 41. Sec ahovc chap. 2, n. XX. praedicato ... quod est cOI1l:eptuscopulae, quac esl qualitas quaedam mcntis, puta actus 42. Berctta IW'J, pp, 172-.1. 162 OCKHAM ON CONCEPTS LOGICAL CONCEPTS 163

43. See, for example, Exp. in Porph. 1,5, OPh II, p. 23: 'Unde si "rationale" est differentia dicitur appellativus; et appellat illud quod connotat per modum adiacentis ei pro quo hominis, "rationalitas" importabit idem quod "anima intellectiva", et rationalitas erit supponit' (the italics are mine). anima intellectiva ...' 66. John Buridan, Sophismata 4, ed. Scott 1977, p. 62: 'Tertio notandum est quod secundum 44. See Ord. I, dist. 8, q. 4, OTh III, p. 233: 'Haec enim est prius vera "omrris homo per diversus modos positivos adiacentiae rerum appellatarum ad res pro quibus termini rationalitatem vel per animam intellectivam differt specie ab asino" ... ' Or Exp. in supponunt, proveniunt diversi modi praedicandi, ut in quale, in quantum, in quando, in Porph. 3, 1, OPh II, p. 55: 'Sed differentia magis proprie dicta est illa qua aliquid differt ubi, in quomodo hoc se habet hoc ad illud, etc. Ex quibus diversis modis praedicandi, ab alio differentia specifica, sicut homo differt ab equo per rationale.' sumuntur diversa praedicamenta, prout debet videri supra librum Praedicamentorum.' 45. Exp. in Praedic. 9, 3, OPh II, p. 187: 'Sicut si definiatur rationale, quod est ens vel The passage is quoted by Spade 1990, p. 606; the English translation I use is his. substantia quae potest ratiocinari ...' 67. Spade 1990, p. 606. 46. Even , in Ockham's view, in so far as they are intelligent beings, can form complex 68. See Quodl. V, q. 8, OTh IX, pp. 509-510: 'Prima [conclusio] est quod accidentia propositions and draw inferences; see Rep. II, q. 14, OTh V, pp. 311-37, esp. 316-21. communia nominum vocalium conveniunt nominibus mentalibus ... Nam primum 47. Normore 1990, pp. 60-61. accidens commune est casus ... Et manifestum est quod sicut istae propositiones in voce 48. See above chap. 6, sect. 2. "homo est homo", "homo est hominis" habent distincta praedicata variata solum per 49. Normore 1990, p. 60. casus ..., sic istae propositiones in mente propositionibus vocalibus correspondentes 50. Normore ultimately closes this development by raISIng a doubt about Ockham's "Sortes est homo", "Sortes est hominis" habent distincta praedicata variata solum per acceptance of such complex intuitive cognitions (Normore 1990, pp. 60-61), but the casus.' See also SL I, 3, OPh I, pp. 11-13. texts q:uoted above in chap. 6, nn. 14~15,are quite conclusive in this respect. 69. See above chap. 5, sect. 1. 51. See above n. 2: . 70. See chap. 5, sect. 2. 52. Taking logical constants as paradigms for syncategoremata was the standard attitude of 71. This is what was labelled Thesis 3 about nominal definitions in chap. 5, sect. 1. See in medieval logicians. Kretzmann (1982a, p. 213, n. 10) identifies the following logical particular the passage from Quodl. V, q. 19, OTh IX, p. 555 quoted there in n. 15. features as constituting the domain of syncategorematic terms in the thirteenth and 72. See for example SL II, 8, OPh I, p. 272, about propositions with an oblique subject or fourteenth centuries: 'Distribution (or quantification), e.g., "omnis", "totus"; Nega~ion, predicate. e.g., "non", "nihil"; Exclusion, e.g., "solus", "tantum"; Exception, e.g., "praeter", 73. Ibid.:' ... quando casus obliquus regitur ex vi possessionis'. "nisi"; Composition (or predication), e.g., "est", "incipit"; Modality, e.g., "necessario", 74. Note that Ockham, contrary to many other medieval authors, never tried to provide a full "contingenter";Conditionality, e.g., "si", "quin"; Copulation (or conjunction), e.g., "et"; justification for the Aristotelian list of categories. See Exp. in Praedic. 7, 1, OPh II, p. Disjunction, e.g., "vel", "utrum"; Comparison, e.g., "qua"; Reduplication, e.g., 161: 'Quinto, circa sufficientiam et numerum praedicamentorum est sciendum quod "inquantum", "secundum quod".' difficile est probare quod sint tantum decem praedicamenta.' In SL I, 41, OPh I, pp. 53. See SL I, 4, OPh I, p. 16: 'Et sicut hoc nomen "omnis" nihil determinate et finite 116-17, and Quodl. V, q. 22, OTh IX, pp. 567-9, he links the diversity of the categories significat ..., sic est de omnibus syncategorematibus et universaliter de coniunctionibus with the number of distinct questions which can be answered with one single term; but et praepositionibus.' he gives no argument to prove the correctness of the Aristotelian list. Buridan, on the 54. Spade 1990, p. 605. See also Panaccio 1992a, pp. 34-5 for a related idea. The suggestion other hand, goes so far as to say that no such argument is found in Aristotle himself; see is critically discussed in Michon 1994, pp. 373-9, and Gaskin 2001, pp. 246-50. his Summulae in Praedicamenta 3, ed. Bos 1994, p. 18: 'Et sciendum est quod 55. See Ord. I, Prol., q. 1, OTh I, p. 31. The passage is quoted above in chap. 6, n. 14. numquam Aristoteles posuit rationem ad ostendendum quod non essent alia 56. See Adams 1987, pp. 271-2, n. 119. The passage she refers to is found in Rep. II, q. 1, praedicamenta praeter ista decem. Nec esset inconveniens ponere alia ... ' OTh V, pp. 16-17. 75. See above chap. 5, sect. 3.3. 57. For Ockham's final position concerning the ontological status of relations in the natural world, see in particular Ord. I, dist. 30, q. 2-4, OTh IV, pp. 32~74,and Quodl. VI, q. 8-30, OTh IX, pp. 611-701. 58. See for example SL II, OPh I, p. 281. 59. In SL I, 31, OPh I, p. 94, Ockham provides the following definition: 'Copula autem vocatur verbum copulans praedicatum cum subiecto.' A 'predicate' in this way of speaking can be any grammatical complement, and the copula, consequently can be just about any verb. See for example SL II, 21, OPh I, p. 321, where 'asinum' is treated as the predicate of 'nullus homo videt asinum', and 'videt' as the copula. 60. See Spade 1990,p. 606, and Gaskin 2001, p. 250. 61. This is what I called Thesis 1 concerning nominal definitions in chap. 5 above. See, in particular, the passages quoted in chap. 4, nn. 28, 30, and 31. 62. See above chap. 6, sect. 3. 63. Spade 1990, pp. 604-6. 64. See above chap. 5, sect. 2. 65. See for example John Buridan, Summulae de Suppositiollihu,I' 5, cd. Van der Lecq )l)l)8. pp. 80-83; in partie. p. 80: 'Sed omnis lerminus connolans aliud atl co pro quo supponit Chapter 9 The Meaning of Words

Everybody admitted in the Middle Ages that spoken and written words receive their meaning by being associated with concepts somehow. But how exactly is this dependency to be theorized came to be the subject of a lively debate.:.., a. 'magna altercatio', Scotus says -towards the end of the thirteenth century.! Ockhatn in this discussion resolutely sided with those who held that words are not properly said to signify concepts, but things. Yet he did not believe, any more than anybody else, that words. signify things independently of concepts: words, he would S;lY, are 'subordinated' to concepts. How he thinks of this relation turns out, as we shall see, to be of far-reaching significance for his whole philosophY of mind and language, but still, I ;lm afraid, widely misunderstood. My intent in this final chapter is to elucidate Ockham's notion of subordination. It will be shown, in particular, to require a resolutely externalistic conception of the meanings of words, of the sort Hilary Putnam's name has been associated with in contemporary philosophy (section 1). How subordination thus understood fits with Ockham's nominalism will then be considered (section 2). And it will be argued, finally, that Ockham's best theory on this matter opens the way for the acceptance of what I will call 'reverse subordination': a direct semantlcal dependency of certain mental units upon their linguistic counterparts. The case of singular terms, at this point, will come out as especiallyrevealing (section 3).

1. Subordination

The basic scheme is the following. First, simple categorematic concepts are acquired as natural signs of external things. And then comes subordination: certain spoken sounds are conventionally associated with certain concepts, in such a way that the spoken sounds in question inherit the signification of the concepts they are associated with. The point is made in the very first chapter of the Summa Lo!?icae:

I say that spoken words are signs subordinated to concepts or intentions of the soul not because in the strict sense of 'signify' they always signify the concepts of the soul primarily and properly. The point is rather that spoken words are i!Dposed to signify the very things that are signified by concepts of the mind, so that a concept primarily and naturally signifies something and a spoken word signifies the same thing secondarily ... The same sort of relation I have claimed to hold between spoken words and impressions or intentions or concepts holds between written words and spoken sounds.2

The concept of 'horse', for example, naturally signifies horses. When the spoken sound 'horse' - or 'CqlillS' or '('/i('\'al' - is conventionally suoordinated to thaI

1M 166 OCKHAM ON CONCEPTS THE MEANING OF WORDS 167 concept, it ipso facto starts signifying horses too, albeit conventionally. And when to that of words. The term 'subordination' turns out to be judiciously chosen after the corresponding written word is conventionally subordinated in turn to that spoken all. sound, it also ipso facto starts to conventionally signify horses. So far so good. This is generally well understood. But there is more to Ockham's Ockham's source for this way of speaking is probably John in the idea of subordination, and the really interesting part of it seems to have been missed very passage where he refers to the magna altercatio about whether words signify in much of the recent secondary literature. The picture many commentators give is concepts or things: the following: whenever I speak intelligently, I have some concepts present to my mind and the spoken words I utter to communicate what I mean are subordinated to Let me concede, for short, that what is properly signified by a spoken word is a thing. those concepts. If I utter, for example, the English sentence 'some horses are black', Nevertheless, written words, spoken words, and concepts, are ordered signs [signa my token utterance is supposed to be subordinated to a corresponding mental ordinata] of the same significate, just as there are many ordered effects of a same cause, proposition which is actually present to my mind. none of which being the cause of the other, as is clear in the case of the sun illuminating This picture, however, is not Ockham's. Two preliminary considerations will several parts of the environment} begin to shake it. First, subordination, as we have seen, holds not only between spoken words and concepts, but also - and in the same sense - between written Ockham expresses himself in a closely related way in his own Ordinatio: words and spoken words. How can the received parallelistic picture be extended to this other case? Should we say that whenever I write down something, I should also ...a certain spoken word primarily signifies several things equally, because it has been utter the. corresponding spoken words, or at least have a mental representation of imposed by a single imposition to everything which a determinate concept of the impositor these corresponding spoken words? so, I would need three simultaneous levels of is common to, so that the word and the concept are to each other like ordered signs [signa If quasi ordinata]; not because the word primarily signifies the concept, but because it is mental representations for each word I actually write down: a conceptual imposed at primarily and precisely signifying every single thing the concept is true of ...4 representation of horses, let's say, a mental representation of the spoken word 'horse', and a mental representation of the written marks I am about to draw. Is this What we have here is a transitional phase, with respect to vocabulary, between a theory we want to attribute to Ockham? Well, maybe, if we have to. But the fact is Scotus's phrase 'signa ordinata' and Ockham's later use of 'subordinata' in the he doesn't say anything of the sort. And, as we shall see shortly, he is not committed Summa Logicae. to it either. But what is the advantage of saying that words are subordinated to concepts, My second preliminary consideration has to do with equivocity, which is indeed rather than signify them? For Ockham at least, this is not just an innocuous variation a subject where Ockham saliently resorts to the terminology of 'subordination'. In in the way of speaking. Having decided to count concepts as signs, Ockham clearly the Summa Logicae, for example, the very definition of what an equivocal word is, realized that it was preferable for him to use some other verb than 'to signify' for the is coined in terms of 'subordination': 'A word is equivocal if, in signifying different connection of words to concepts. For this relation, from a semantical point of view, things, it is a sign subordinated to several rather than one concept or intention of the is very different from the relation that concepts have with the things they signify. In soul.'6 This definition should strike us as not easily compatible with the parallelistic Ockham's framework, to say that a concept signifies certain things in the world is to picture. If subordination was the relation between a speaker's token utterance and say that this concept will normally be used by the agent to refer to those very things: what that speaker has in mind at the moment of utterance, the strange consequence when taken in personal supposition (which is the standard case), the concept is would ensue that in order to utter an equivocal word, I would need to have the two expected to supposit for its sign~ficata.It is exactly the same sense of 'to signify' that corresponding concepts in my mind at the moment of utterance. This is normally not Ockham wants to transpose to spoken and written words: just like conceptS, they will what happens. If I say 'I cashed money at the bank yesterday', I might very well have be said to signify whatever it is that they are normally usedto refer to when taken in in my mind, in uttering the word 'bank', only the concept of a financial institution, personal supposition. This corresponds, in fact, to the technical notion of 'signum' and not that of a shore. Is it to say that 'bank' is not equivocal in such occasions? If Ockham endorses in the first chapter of the Summa (and which I lengthily so, we could easily avoid equivocity in our external languages simply by mentally commented upon in chapter 3 above).5 Ockham's distinctive move here is to reserve sticking to one particular interpretation of some potentially equivocal word. This the technical appellation of 'signification', and the consequent use of 'to signify', for sounds strange. The very idea of 'equivocity' is typically used for cases where the the semantical connection he wants to posit between signs - whether conceptual, meanings of the words do differ from what the speaker wants to convey. spoken or written - and the individual things out there in the world that they What these two preliminary considerations jointly lead to, is that the received normally refer to when they occur within propositions. This allows for fruitful parallelistic picture of subordination as the relation holding between a given parallelisms between the semantics of mental language and that of spoken and utterance of a speaker and what that speaker has in mind at the moment of utterance, written languages. And it characteristically brings into focus the referential aspects would force us to attrihute to Ockham either strange theses, or imprecisions of of meaning which Ockham persistently wanted to stress. Yet a hierarchy is still speech, or hoth. This is not yet a refutation of the picture, of course, hut it should assumed, of course: concepts are taken to he more fundamental, and their arouse our suspicion. signitication, heing natural, is seen as usually prior and determinative with resped Let us corne down now 10 the main poin!. In the crucial passage of Summa THE MEANING OF WORDS 169 168 OCKHAM ON CONCEPTS

Logicae I, I quoted above, whereOckham introduces the technical phrase 'signa the following about Ockham: 'Insofar as a translation is supposed to "express the subordinata', he explicitly - and thoughtfully -links subordination with the same thought" as the original, we can say that a statement in one language is • imposition of the term: 'I say that spoken words are signs subordinated to concepts correct translation of a statement in another language if the two statements are ... because ... spoken words are imposed [imponuntur] to signify the very things that subordinated to the same mental proposition.'J3 are signified by concepts of the mind ... '7 This is to be taken seriously. '!mponere' Strictly speaking, this cannot be Ockham'sown notion of subordination. Spoken is a technical term in medieval philosophy of language. The 'impositio', in sentences, in Ockham's theory, are not imposed to signify anything andthey are not, Ockham's vocabulary, is the original attribution ofa conventional signification to a therefore, subordinated to anything in the relevant sense. Ockham; remember, given sound or a written mark. The subordination of a word to a concept, then, is subscribes to some form of semantical atomism. What is attributed a conventional what happens at the original moment of the imposition, not at the moment of signification at the moment of imposition normally is a simple tetm, and the semantical properties of complex phrases, such as a complete sentenGe, are supposed utterance. Look again at the passage from distinction 2 of the Ordinatio quoted above: a to be systematically derived from those of their simple components, without any new word and a concept, there, are said to be like 'ordered signs' (signa quasi ordinata) subordination being needed in the process..We could set out, of course, to introduce because the word is imposed, by means of a single act of imposition, at signifying a derivative notion of 'subordination' which would be applicable .10 complete everything that the concept signifies.8 Ockham's example is the univocal term sentences. But for one thing, it is not quite obvious how to do it. And we should be 'homo'; the reason he gives why 'homo' is indeed a univocal term overtly has to do aware, moreover, that this would not· be subordination in Ockham:s own sense with the original intention of whoever imposed that word: 'Such is the word "homo" anymore: the relation between a spoken and a mental sentence, in his view, does not and this is why it simply is a univocal term; for the impositor [imponens] intended normally depend upon a special act of imposition. this word "homo" to signify every single thing which a certain determinate mental Another .interesting example of a similar discrepancy with respect toOckham's Goncept is true of ... '9 The use of 'imponens' along with the past tense 'intendebat' use is found in Calvin Normore's oft-quoted paper on Ockham's mental language: in these lines clearly refers the reader to some past of imposition. 'Each spoken connotfltive term and its defining expression', Normore writes, 'will This is even more explicit in the Summa Logicae as Ockham discusses the same be subordinated to the same complex expression of the mental.'14 Normore supposes here that every spoken connotative term is strictly synonymous with its own example: definition for Ockham and that both are represented in the mind ofa competent Whoever it was that first instituted the use of the term' homo' saw some particular man and speaker by a single complex conceptual sequence, which is, of course, the very cQined the term to signify that man and every substance like him ... But even though it reading of Ockham we found reasons to reject in chapters 4 to 6 above. What I am signifies indifferently many men, 'man' is not equivocal, for in signifying indifferently concerned with now is .the place attributed to subordination in this interpretative many men it is a sign subordinated [signum subordinatum] to just one concept and not scheme: it does not fit very well, I contend, with Ockham'sown way of many. 10 characterizing the relation. Even admitting that a simple spoken word could in certain cases be subordinated to a complex conceptual sequence,15 the spoken Subordination, once more,is based upon the original imposition. of the term by definition of that word would not in such cases, properly be said to be subordinated, whoever it was that first coined it and associated it with a given concepr in Ockham~stechnical sense, to the same mental expression. Ockham's well-considered idea of subordination can thus be captured in the Suppose, for the sake of discussion, that 'father' is adequately defined as 'male following formula: animal having a child'. And suppose, in addition, as seems probable, that we do have the mental capacity for assembling a conceptual complex sequence of the form: 'male' A given sign S2 is subordinated to a previously signifying term Sl if and only if S2 has + 'animal' + 'having' + 'child'. It might be that the person or group who originally been imposed to signify whatever S1 signifies. II struck the word 'father' decided to subordinate it· to precisely this complex mental phrase (or some token of it). Butwhatabout the spoken definition? The English phrase This is a very general notion. And a useful one too. In the light of it, we can readily 'male animal having a child', for example, will not properly be said to be subordinated, understand, for example, that a written word can be said to be subordinated to a inOckham's sense, to the corresponding complex mental phrase. What has been spoken word in exactly the same sen.se in which a spoken word is subordinated to a independently subordinated in this case, as in that of complete sentences, is each concept: in both cases the original impositor of the newly coined sign simply component of the English expression. The English word 'male' must have been attributed it the signification of some other previously existing sign, whether natural subordinated to the concept 'male', the English word 'animal' to the concept 'animal', or conventional. and so on. But the complex English phrase as a whole is not subordinated to anything, And there is no problem either, in this approach, with Ockham's characterization since it is not (normally) the object of a new imposition. It will turn out, of course, to of equivocity in the Summa: 12 a given sign typically gets equivocal by being have some very strong semantical equivalence with the corresponding mental sequence. successively the object of several different impositions. Subordination, again, is seen But, however strong, this relation will not be subordination in Ockharn's sense. to hang on to what happened at the original mornent(s) of imposition. . This is not mere quihhling. A whole view of language is at stake here. Spade and The point seems to have heen missed hy many. Paul Spade. for example. writes 170 OCKHAM ON CONCEPTS THE MEANING OF WORDS 171

Nonnore's use of 'subordination' rests, as far as I can see, upon their common ... words, in their primary or immediate signification, stand for nothing but the ideas in assumption that what a spoken sound is subordinated to is the intellectual item the the mind of him that uses them ... That then which words are the marks of are the ideas of speaker has in mind when uttering the sound. If this was the case, it would be the speaker: nor can anyone apply them as marks, immediately, to anything else but the innocuous, indeed, to say that a complex spoken phrase or sentence is subordinated ideas that he himself hath ... 18 to the mental sequence the speaker has in mind at the moment of utterance, just as each one of its components is supposed to be subordinated to some conceptual Locke's radical intemalism in these lines is at complete variance with Ockham's component of this very mental sequence. But this is not Ockham's approach. The view. on the signification of spoken words as I read it. Let us return again to a important point, let me repeat, is the following: the meanings of words, for Ockham, passage from Summa Logicae I, 43, quoted earlier: 'Whoever it was that first depend not upon what is going on in the head of the speaker at the moment of instituted the use of the term "homo" saw some particular man and coined the tenn utterance, but upon whatthey were originally subordinated to by the impositor. Put to signify that man and every substance like him ... '19 To a modern reader, this must in modem tenns: Ockham is a proponent of externalism rather than intemalism with inevitably recall Hilary Putnam's canonical presentation of in respect to the signification of spoken and written words. his famous 1975 paper 'The meaning of "meaning"';20 and it is as remote as can be A passage that strikingly supports this reading is found in the Third Part of the from Locke's solipsistic view. Summa, when Ockham discusses the use of absolute - that is, non-connotative ­ Ockham goes as far as to say that in some cases even the impositor might not have spoken names such as 'man', 'horse', 'animal', 'flower', 'water', and the like: how in his own mind the concept which he chooses to subordinate a given spoken sound does a spoken sound get to be an absolute name? The answer, as one expects, is !bat to. For subordination to succeed, it is sufficient, Ockham thinks, that the impositor this happens when the spoken sound in question is conventionally subordinated to a be able to identify the relevant concept in some distinctive way. This does not concept which is already an absolute name, albeit a natural one. What I want to draw automatically require that he should possess it himself: attention to in this passage is Ockham's explicit acceptance that some of the spoken words I use might depend for their signification upon concepts that other people have: Moreover, it is possible for somebody to impose this name a to signify whatever animal ,." since spoken words are conventional, absolute spoken names can be imposed upon will appear to him tomorrow. This being done, the word a distinctly signifies this animal, the very same things of which we have- or others have - such absolute concepts' .16 and it will signify it for all those who are willing to use the word as it was imposed, even The example he gives is the tenn 'lion'. Someone like myself, says Ockham, who though the impositor does not have a distinct intellection of this animal, and maybe will 21 has never met with a lion, cannot have acquired the simple absolute concept of 'lion'. not have one when it appears to him. What such a person has in mind when uttering the spoken word 'lion' (in a sentence like 'a lion is an animal', for example) is actually a 'complex mental expression In such a case, the concept which the impositor intends to subordinate the word a to, composed of several simple cognitions, none of which is both simple and proper to the is uniquely identified: it is the absolute specific concept that a well-placed observer would spontaneously fonn if he or she was to have a nonnal intuitive cognition of lion'. I? Yet this does not prevent the speaker in question from using the tenn 'lion' as an absolute term in her own spoken speech. Ockham's clear suggestion in this passage the first animal the impositor will meet with tomorrow. It might happen, as Ockham is that a speaker can legitimately utter a sentence with an absolute spoken name as explicitly acknowledges,. that the impositor, at the moment of coming across the subject while not having in her own mind the corresponding absolute simple concept. animal in question, should not himself be in a position to form that concept (he might This is possible because, subordination being what it is, the meanings of our own not have a good view of the animal, for example, or his attention might be drawn words can depend upon the concepts that other people have. The spoken word 'lion' is elsewhere, or whatever ... ). Yet the relevant absolute concept is entirely detenninate. an absolute tenn in English, according to Ockham's story, because it was originally And so is its signification: it signifies a certain individual animal (the first one that subordinated to the absolute concept of 'lion' by people who had that concept. And it the impositor will come across tomorrow) plus 'every other substance like it';22 and does not lose this instituted signification when it is uttered by a speaker like Ockham nothing else. The subordinated spoken word, consequently, will inherit this same himself who does not happen to have the relevant simple concept. What detennines the signification from the very moment of its imposition; this is the signification it will semantical status of a given spoken word within a given linguistic community is the have, from then on, whenever it is uttered by a speaker who accepts to use this concept which the original impositor (or group of impositors) associated it with, not particular spoken sound in accordance with this imposition, even if this particular the concept that any particular speaker happens to have at the moment of utterance. speaker - and everybody else, for that matter - should lack the concept in question at the moment of utterance. The consequences of this are far-reaching. There will inevitably be, according to this view, frequent discrepancies between the semantical status of the spoken words uttered by a particular speaker, and the concepts this speaker actually has present to 2. Types and tokens again his mind at the moment of utterance, even if the speaker is speaking intelligently and literally. Contrast this, for example, with 's well-known account of linguistic meaning in his Essay Concernin[? Human Understandin[?: Our description of the subordination process has been given so far in terms of linguistic and conceptual types rather than tokens. This is easier to do for a start. and 172 OCKHAM ON CONCEPTS THE MEANING OF WORDS 173 it does correspond,to Ockham's usual practice. When he writes, for example, that 'a imposed' results in her uttering a token with more than one signification, even if she certain spoken word [aliquis vox] signifies several things equally, because it has been has only one determinate concept in mind at the moment of utterance, and even if imposed [etc.]',23 he speaks as if the spoken .sound that now signifies something was she is entirely unaware of the of her own speech. the very same one that was originally associated with a 'determinate concept' by the Saying, in short, that a certain linguistic type is subordinated to one Or more impositor. Yet the relevant tokens, of co~rse,can be numerically distinct. Sameness concepts amounts, ultimately, to say something about a plurality of phonologically in such cases must be generic or specific identity, not numerical identity.24 As was similar tokens, uttered under the authority of some determinate past impositions; explained in chapter3, the type idiom is acceptable to Ockham ill semiotical matters in so far as it is reducible in principle to talk about tokens.25How such a reduction canbe achieved in the particular case of subordination-theory, is not - unsurprisingly 3. Reverse. subordination? The instructive case of proper names - developed by Ockham,but it turns out to be rather straightforward. Types can enjoy no independent existence in Ockham's ontology. What must have An interesting feature of Ockham's idea of subordination, as I understand it, is that happened, then, ina typical situation of semantical subordination, is that the original it is neutral with respect to the sorts of signs it applies to. I have insisted so far on impositor associated at the moment of imposition a characteristic token of a certain the subordination of spoken sounds to concepts because it is the basic case in spoken sound with a certain conceptual token.26The latter naturally signified, prior Ockham's view, but a newly coined sign could in principle be subordinated, in the to the impositional act, certain individuals in the world, in virtue of belonging to a Ockhamistic sense, to any well-identified previously existing sign, whethernatural mental chain of intellectual acts and habitus causally grounded in the right. way. or conventional, and inherit in this way the semanticalfeatures of this previous sign, When the signification of this conceptual token is conventionally transferred by the whether they are or not entirely clear to the impositor herself. One could legitimately impositor to the spoken token, the convention put forward is that it should ipso facto propose, for example, to create a new general term in English by subordinating an be transmitted to an indefinite number of future spoken tokens as well. Which ones acceptable string of English sounds to a previously existing Latin word, even if the exactly? Two conditions, as far as I can see, must be met in the Ockhamistic context proposer himself has no clear understanding of the Latin term. Ockham's own for a newly uttered token to inherit the signification attributed to one of its ancestors example of the impositor who subordinates a newly coined word to a concept he by the impositor: first, the new token must be phonologically similar to the one that doesn't yet possess (and maybe never will) shows that he is committed to go quite was associated with a certain concept at the time of imposition; and second, the far in this direction. The question, of course, is: how far? Should he admit, in utterer of the new token must agree to use the sign as it was imposed. particular, the possibility of what I will call 'reverse subordination',. defined as the The latter condition is especially interesting. Ockham clearly points to it in a conventional subordination of a mental concept to a spoken sound or a written mark? passage of the Ordinatio I have just.quoted, where he writes that a certain word a If so, the signification of many of ourown concepts could turn out to be quite opaque will signify a:certain animal 'for all those who are willing to use the word as it was to us. imposed'.27 If a speaker is to use certain spoken sounds as linguistic signs, he or she Prima facie, the case ~eemsto be excluded by Ockham's insistence that concepts must accept the authority of some past impositional acts. When and how these .acts are natural signs. But this means only that they naturally signify certain things in the occurred, and who the impositors were, is something, of course, that most speakers world. It does not by itself prevent them from stipulatively signifying some other don't know anything about.Ockham himself does not provide any precise things. Ockham himself had clearly acknowledged this possibility in his first theory description of how such conventions are originally supposed to be implemented, and of how connotative, negative, and syncategorematic concepts receive their his general approach is compatible with a wide. variety of hypotheses on this subject. signification. As we saw in some detail in chapter 8, he thought it acceptable at one There is no need, in particular, to suppose that there was only one single impositor point to say that a ment(ll sign already endowed with a natural signification can, in in each case, and that the event of imposition always occurred instantaneously. addition, be imposed by the intellect 'at signifying the same thing as [some] external Collective and indirect processes must be at work lIlost of the time, obviously, and spoken words signify'.28The concept, admittedly, would then become ambiguous for this is something Ockham would have no problems with; The willingness to use a the agent, since it would have two different significations: a natural one and a certain token 'as it was imposed' cannot depend, for a particular speaker, upon her stipulative one. But Ockham, apparently, did not see this as fatal. The case he wanted ability to correctly locate or describe the corresponding original act of imposition. to stress was when the concepts that arestipulatively reimposed by the agent are What must suffice is an agreement from the speaker to use spoken sounds and those that naturally signify spoken words: my mental concept of the word 'horse', written marks in conformity with the impositions that are currently in force within let's say, can be reimposed, by a mere stipulation of my intellect, to signify whatever the linguistic community. it is that the word 'horse' signifies. Such reimpositions, at the time of the Ordinatio. This accounts for equivocity, in particular. A given utterance of a certain speaker were considered by Ockham as legitimate and, presumably, fruitful. He must have might be equally similar to two past tokens that were independently subordinated to taken the standard intellectual agent to be able to disentangle somehow the resulting lJ,on-equivalent concepts; given the accidents of collective linguistic history, both when assembling mental propositions. impositions might simultaneously be in force within a given linguistic community. As we have seen, Ockham later abandoned this approach as an account of how all In such a case, the general willingness of the speaker to use the signs 'as they were syncategorematic and connotative concepts are acquired. He came to accept some 174 OCKHAM ON CONCEPTS THE MEANING OF WORDS 175 natural syncategoremata in mental discourse,29 as well as some naturally signifies in virtue of its original imposition, she will be in a position to use proper implemented simple connotative concepts.30Yet, nothing in his later position forces names in her own mental reasoning with the same semantical and logical features him to renounce the very possibility of those stipulative reimpositions of concepts that spoken proper names have. Invalid idiosyncratical inferences would ipso facto that he had hypothesized in the Ordinatio. It is true that he never developed the point lose much of their attractiveness. again in his later works, but his best theory, as I will now argue, readily An ambiguity, admittedly, would thus be introduced in the language of thought. accommodates it. The same mental tokens will naturally signify the English word 'Aristotle' and It will be instructive in this regard to ponder a bit over the paradigmatic case of stipulatively signify at the same time Aristotle himself. But this should be tolerable proper names. As we saw in chapter 1, intuitive acts themselves can serve as singular to Ockham. It is very similar indeed to one specific kind of ambiguity that he does terms in the language of thought, according to Ockham.31My intuitive grasping of explicitly admit into mental language: ambiguities of supposition namely.3? In a this cat in front of me is a simple mental sign, naturally signifying the individual that mental sentence such as 'man is a species', the concept of 'man', Ockham contends, caused it, and nothing else.32 It is perfectly possible, then, on the Ockhamistic can have simple supposition since the predicate is a term of second intention conception, that a certain spoken sound be subordinated to such an intuitive act by ('species'), and it can have personal supposition as well since personal supposition the impositor. This is, presumably, how proper names are supposed to enter spoken is always possible according to him.38There is a sense, then, in which such amental language in the first place. I can subordinate the spoken sound a, let's say, to my sentence can be said to be ambiguous: it could mean that a certain real man is a present intuitive grasping of this ~atin front of me, and the word a from then on will concept (which is false, of course) or that the concept of 'man' is a concept (which signify this particular cat for all those who are willing to use the word as I imposed is true).39 This is very close to what reverse subordination results in: the mental it. Whether they ever saw the cat in question or not, the word a in their mouth will sentence 'ARISTafLE is a philosopher' (where 'ARISTafLE' is the natural sign for the normally supposit for this very cat.33 English word 'Aristotle') might mean, on this view, that a real man, Aristotle Now, Ockham also holds, as I insisted in chapter 1, that no abstractive concept himself, is a philosopher (if we favour the stipulative signification) or that a certain can naturally be a simple singular term. While the representational import of an English word is a philosopher (if we favour the natural signification). Itis true that intuitive act depends upon what causes it, that of an abstractive act, or concept, is the ambiguity now rests on signification rather than supposition as in the case based on some sort of similitude; and no similitude relation, however we understand explicitly admitted by Ockham; but considered at the level of complete mental it exactly, can discriminate between maximally similar objects.34All abstractive acts sentences, the parallelism between the two situations seems very strong, and the tum out to be general. If this is so, a speaker who uses a spoken proper name in the same disambiguating processes, presumably, could work in both. Ockham is absence of the referent will not normally have in her mind at the moment of committed to attributing a disambiguating role to some contextual factors or other utterance a simple concept naturally signifying this very individual and nothing else. with respect to suppositional ambiguities.40The same factors, whatever they are, In uttering the spoken name, she might, instead, have in mind a complex cognition should do just as well for a mental sentence like' ARISTOTLE is a philosopher'. 'composed of several simple ones';35 in other words: a description. A semantical In short, nothing prevents Ockham's former hypothesis about reverse discrepancy transpires once more between what the speaker has in mind at the subordination to be maintained within his mature theory. And there is a job for it moment of utterance and the meaning of her words. This is the very same sort of there: it would nicely account for our capacity to use mental proper names in our we acknowledged earlier with respect to general terms like 'lion'.36 reasoning. If this is so, however, the idea can - and should - be generalized to other What we now realize is that such discrepancies between thought and language are categories of terms. If reverse subordination is possible and fruitful in the case of bound to be widespread in the case of proper names on the Ockhamistic view. proper names, sO will it be for general terms, whether absolute or connotative. An This is not innocuous. Different users of the same proper name will inevitably intelligent agent who has never seen a real lion would be able, in this picture, to associate it with different mental descriptions, and will be attracted, consequently, to subordinate her mental representation of the English word 'lion' to a token of that various inferences, nOlle of which would be valid in the spoken language. This is a very spoken word, and use it, in her subsequent thought processes, as a simple potential source for confusion and misunderstanding. Suppose I associate the absolute concept of something she has never seen, lions namely. Even simple English name 'Aristotle' with a given mental description such as 'the author of the connotative concepts could be instituted in the mind in the same way, by being Categories'. I might be tempted to infer from a spoken statement of the form stipulatively subordinated to simple spoken words such as 'father' for example. 'Aristotle is F' that the author of the Categories is F, which would be both Let me insist again that this is not explicitly developed in Ockham's later works. idiosyncratical and logically unwarranted. But what he says about subordination - for example, in the passages quoted in It is to be supposed, from Ockham's point of view, that an intelligent speaker section 1 of the present chapter - does commit him to a robust form of semantic should normally be capable of avoiding such traps. But how is that to be done? One externalism with respect to the signification of spoken and written words. It is but a nice way of achieving it, at least, would be to resort to this reverse subordination small step from there to accepting at least some degree of externalism with respect process that Ockham had hypothesized himself in the Ordinatio. If a speaker can to the signification of concepts. Ockham might not have thought very deeply about stipulatively reimpose her conceptual representation of the English word 'Aristotle' it. but the hypothesis about reverse subordination that he had put forward in the so that it signifies for her, from then on, whatever individual the spoken word Ordinatio smoothly allows us to take that step without contradicting or threatening 176 OCKHAM ON CONCEPTS THE MEANING OF WORDS 177 any of the other main tenets of his later theory. Some of our mental signs would thus 3. John Duns Scotus, Ord. I, dist. 27, q. 3,in Op. Omnia VI, p. 97: ' .. , breviter concedo come to depend for their (stipulative) signification upon the concepts that other quod iIIud quod signatur per vocem proprie, est res. Sunt tamen signaordinata eiusdem signati .littera, vox et' conceptus, sicut, sunt multi effectus ordinati eiusdem' causae, people have naturally acquired. quorum nullus est causa alterius, ut patet de sole iIIuminante plures partes medii ...' Hilary lists three assumptions as In Representation and Reality, Putnam (italics are mine). constituting the basis for the internalistic picture of meaning he wants to break with 4. Ord. I, dist. 2, q. 4, OTh II, pp. 139--40: ' ... aliquam vocem aeque primo' plura (and which he traces back to Aristotle): significare, quia una impositione imp0!1itur omnibus quibus conceptus deteJ;l11inatus 'habitus ab imponente est communis, ita quodsint signa quasi ordinata; non quod vox 1. Every word he uses is associated in the mind of the speaker with a certain primo significet ilium conceptum, sed quia imponitur ad significandum primo et mental representation. praecise omne illud de quo conceptus praedicatur ... ' (italics are mine) 2. Two words are synonymous (have the same meaning) just in case they are 5. SLI,l, OPh I, p. 9; text quoted above in chap 3, n. 6. associated with the same mental representation by the speakers who use those 6. SL I, 13 OPh I, p. 45: 'Est autem vox illa. aequivoca quae significans plura non est words. signum subordinatum uni conceptui, sed est signum unum pluribus conceptibus seu 3. The mental representation determines what the word refers to, if anything.4\ intentionibus animae subordinatum' (trans!. Loux 1994, p. 75). 7. SL I, 1, OPh I, pp. 7-8; text quoted above in n. 2. Loux's translation here unfortunately renders 'imponuntur' by 'are used to' (Loux 1974, p. 50), thus losing the precise Ockham, we may now conclude, is committed to rejecting this entire budget of technical import of the term. theses. In so far, at least, as linguistic meaning is concerned, the picture he favours 8. See the text quoted in n. 4. comes out as strikingly externalistic: 9. Ord. I, dist. 2, q. 4, OTh II, p. 140: 'Talis est haec vox "homo", et ideo est simpliciter univoca, quia imponens hanc vocem intendebat quod significaret omnem rem de qua l' . Spoken and written words are subordinated to certain previously existing signs conceptus mentis determinatus praedicatur ...' (italics are mine). (whether natural or conventional), not by each particular speaker, but by the 10. SL I, 43, OPh I, p. 124: 'Illeenim qui primo instituit hanc vocem "homo", videns original impositors of the language. aliquem hominem particularem, instituit hanc vocem ad significandum illum hominem 2'. Two words are synonymous if they were originally subordinated to the same et quamlibet talem substantiam qualis est ille homo ... Nontamen esthaec vox "homo" concept (or to equivalent conceptual tokens).42 Particular speakers can fail to aequivoca, quamvis significet multa aeque primo, quia .est signum, subordinatum uni conceptui et non pluribus in significando illos plures homines aeque primo' (transl. Loux rec,ognize such synonymies, even they correctly use the words in question in if 1974, p. 136, slightly amended; the italics are mine). most situations~43 11. Literally taken, this formula applies only to categorematic terms, since only these have 3'. What a word refers to is determined by the signification ofthe sign(s) which it a determinate signification according to Ockham's way of speaking (SL I, 4, OPh I, p. was originally subordinated to. The meaning of our linguistic signs, then, can 15). It can, however, be easily extended to cover syncategoremataas well: a spoken be opaque to us up to a point. sound can certainly be associated by imposition with a given syncategorematic concept in such a way that it inherits the semantical features of this concept whatever they are And if, on top of it, reverse subordination is added to the picture, as Ockham had (its role in determining the supposition of other terms, for example). once proposed, externalism extends to thought itself, at least partially: in Ockham's 12.' See the text from SL I, 13 quoted above in n. 6. best theory, as it seems, some of our mental representations could be stipulatively 13. Spade 1996, p. 95. endowed with determinate significations, even if we do not happen to possess the 14. Normore 1990; pp. 58-9. 15. See above chap. 6, sect 5. corresponding natural concepts. 16. SL III-2, 29, OPh I, p. 558: ' ... ex quo voces sunt ad placitum, voces mere absolutae possunt imponi eisdem de quibus habemus, vel aliihabent, tales conceptus' (italics are mine). Notes 17. Ibid., p. 559 (about the sentence 'leo potest esse animal'): 'Et habeo unam propositionem mentalem cuius subiectum estcompositum ex multis notitiis incomplexis 1. See John Duns Scotus, Ordinatio I, dist. 27, q. 3, in Opera Omnia VI, Vatican edition, quarum nulla est simplex et propria leoni;. sed propositionem mentalem cuius subiectum 1963, p. 97. A survey of this discussion can be found in Panaccio 1999a, chap. 7. sit aliquod mere absolutum proprium leonibus non habeo ... ' (italics are mine). 2. SL I, 1 OPh I, pp. 7-8: 'Dico autem voces esse signa subordinata conceptibus seu 18. John Locke, An Essay Concerning Human Understanding III, 2; 2. intentionibus animae, non quia proprie accipiendo hoc vocabulum "signa" ipsae voces 19. SL I, 43, OPh I, p. 124 (text quoted above in n. 10). semper significant ipsos conceptus animae primo et proprie, sed quia voces imponuntur 20. See for example Putnam 1975b, p. 245: 'We have now seen that the extension of a term ad significandum illa eadem quae per conceptus mentis significantur, ita quod conceptus is not fixed by a concept that the individual speaker has in his head, and this is true both primo naturaliter significat aliquid et secundario vox significat illud idem ... Et sicut because extension is, in general, determined socially - there is division of linguistic dictum est de vocibus respectu passionum seu intentionum seu conceptuum, eodem labor as much as of "real" labor - and because extension is, in part, determined modo proportionaliter, quantum ad hoc, tenendum est de his quae sunt in scripto indt·xically. The extension of our terms depends upon the actual nature of the particular respectu vocum' (transl. Loux 1974, p. 50, slightly amended). things that serve as paradigms.' 178 OCKHAM ON CONCEPTS THE MEANING OF WORDS 179

21. Ord. I, dist. 22, q. unica, OTh IV, p. 56: 'Praeterea, potest aliquis imponere hoc nomen 39. Ibid., p. 198 (about 'homo est species'): 'Et est propositio distinguenda penes tertium a ad significandum quodcumque animal quod occuret sibi cras. Hoc facto, distincte modum aequivocationis, eo quod subiectum potest habere suppositionem simplicem vel significat ilIud animal, et significabit apud omnes vo1entes uti voce sicut imposita est, personalem.' quantumcumque ilIud imponens non distincte intelligat, nec forte distincte intelliget 40. In SL I, 65, in particular, Ockham mentions the 'intention of the users' (voluntas quando sibi occurret.' utentium) as a possible disambiguating factor for a sentence like' homo est species' (see 22. See the passage from SL I, 43 quoted above in n. 10. the text quoted above in n. 38). Other contextual elements might also playa role. See on 23. Ord. I, dist. 2, q. 4, OTh II, pp. 139-40; text quoted above in n. 4. this Panaccio and Perini Santos, forthcoming, esp. sect. 5: 'Le role du contexte.' 24. For Ockham's use of this distinction, see, for example, the passages referred to above in 41. Putnam 1988, p. 19 (the italics are Putnam's). chap. 3, nn. 35-9, and the corresponding discussion in the core of the text. 42. Note that I do not say 'only if' (or 'just in case', as in Putnam's formulation): being 25. See above chap. 3, sect. 5. subordinated to the same concept is a sufficient condition for synonymy, in Ockham's 26. As we saw, it remains possible for Ockham that the impositor does not himself possess view, but not a necessary one. See on this chap. 7, sect. 4.2 above, esp. pp. 240-42. the relevant concept at the moment of imposition (see the passage from Ord. I, dist. 22 43. See SL I, 6, OPh I, p. 19: 'Large dicunturilla synonyma quae simpliciter idem quoted above in n. 21), but he must have, at least, some descriptive way of identifying significant omnibus modis ..., quamvis non omnes utentes credant ipsa idem significare, it, which should not presuppose, of course, the irreducible existence of linguistic or sed decepti aestiment aliquid significari per unum quod non significatur per reliquum.' conceptual types. 27. Ord. I, dist. 22, OTh IV, p. 56: 'Hoc facto, [hoc nomen a] distincte significat ilIud animal, et significabit apud omnes volentes uti voce sicut imposita est ... ' (italics are mine; a longer portion of the passage is quoted above in n. 21). 28. Ord. I, dist. 2, q. 8, OTh II, p. 286: ' ... et [intellectus] imponit istos conceptus ad significandum ilia eadem quae significant ipsae voces extra' (italics are mine; the surrounding passage is quoted above in chap. 8, n. 6 - see sentence d). 29. See above chap. 8, sect. 2. 30. See above chap. 4 and 6. 31. See above chap. 1, sect. 4. 32. See in particular the passages from Quaest. in Phys., q. 7, OPh VI, p. 411, and Rep. II, q. 12-13, OTh V, p. 288 quoted above in chap. 1, nn. 41 and 47. 33. We thus arrive at something like the Kripkean picture of how proper names are introduced in the language. See Kripke 1972, p. 302: 'A rough statement of a theory might be the following: An initial baptism takes place. Here the object may be named by ostension, or the reference of the same [sic] may be fixed by a description. When the name is "passed from link to link", the receiver of the name must, I think, intend when he learns it to use it with the same reference as the man from whom he heard it.' 34. See in particular the passages from Quodl. I, q. 13 and V, q. 7 quoted above in chap. 1, nn. 27, 28, and 31, as well as the corresponding discussion in the text. For more on the relevant notion of similitude, see chap. 7, esp. sect. 2 and 3. 35. See Quodl. I, q. 13, OTh IX, p. 77: ' ... habeo aliquam cognitionem abstractivam propriam, sed ilia non erit simplex sed composita ex simplicibus'. 36. See the passages from SL III-2, 29, OPh I, pp. 558-9 quoted above in nn. 16-17, and the corresponding discussion in the text. 37. See SL III-4, 4, OPh I, p. 763 (about suppositional equivocations): 'Et est notandum quod iste tertius modus aequivocationis potest reperiri in propositione pure mentali ...' That simple and material suppositions are possible in mental language in addition to personal supposition is explicitly acknowledged by Ockham in SL I, 64, OPh I, p. 197: 'Sicut autem talis diversitas suppositionis potest competere termino vocali et scripto, ita etiam potest competere termino mentali, quia intentio potest supponere pro ilIo quod significat et pro se ipsa et pro voce et pro scripto' (italics are mine). 38. See SL I, 65, OPh I, pp. 197-8: 'Notandum est etiam quod semper terminus, in quacumque propositione ponatur, potest habere suppositionem personalem, nisi ex voluntate utentium arctetur ad aliam ... Sed terminus non in omni propositione potest habere suppositionem simplicem vel materialem, sed tunc tantum quando terminus talis comparatur alteri extremo quod respicit intentionem animae vel VOCemvel scriptum.' Conclusion

Concept theory has come to be cruci,al in recent philosophical work on cognitive sciences. Stephen Laurence and Eric Margolis open their well-known 1999 introduction to the state of the discussion ,on the topic by squarely asserting: 'Concepts are the most fundamental constructs in theory of the mind.' I The current situation, however, is quite perplexing. After having reviewed and discussed the main, approaches pre~entlydebated, the two authors candidly conclude thaf 'no theory stands out as providing the best comprehensive account of concepts'.2 Jerry Fodor, whose recent book Concepts is significantly subtitled Where Cognitive Science Went Wrong; thinks that important confusions about the very idea of a concept are at the root of what he takes to be, a deep theoretical crisis in cognitive sciences: 3 'most of what contemporary cognitive science aboutconcepts', he writes, 'is radicallY,and practically demonstrably, untrue',4 'Who could deny', Eric Margolis adds in a contribution of his own, 'that, we,are in need of some new theoretical options?'5 Well, if 'new' options are needed, maybe old ,- and long­ forgotten -,-ones could help too. The history of philosophy, here as on so many other questions, can be a precious source for intriguing suggestions and potential breakthroughs. And William of Ockham in tpis regard stands out as an exceptionally good prospect for concept theorists: he has an elaborate and sophistiCllted network of thes~sand argull: to offer about ontology, mind, thought and language, ,with the idea of 'conceptus' at the very centre of it. The access to Ockham's doctrine, admittedly, is not immediate fortoday's readers., Not only are the texts in Latin with few translations available, but most importantly the theoreticalbackground is very dissimilar from ours. With most of his colleagues, Ockham accepted the general adequ~cyof the Aristotelian tradition in philosophy, as well'as the predominant authority of the Christian religious beliefs. His writings often responded to, those of the scholastic heroes or the previous decades such as Thoma~Aquinas or John,Duns Scotus~and, of course, he had never read a line of Frege or Russell. The logic he had learnt at school was Aristotle's, . augmented b~lthe terministdevelopments of the thirteenth century, the theory of suppositioespecially. Asa consequence of all this, his theoretical vocabulary was different from ours all the way through. Yet there is no fatal incommensurability in this case. Ockham's approach can still be of greatrelevance with respect to certain deep concernS that are at the forefront of today's interest in cognitive sciences and philosophy of mind. Some of the main puzzles we now face in these fields have to do with how to reconcile a thoroughly causal picture of the natural world with the intentionality and intelligence of the human mind, and this, as far as I can see, was Ockham's problem as well. In his small penetrating book, The Elm and the Expert, Jerry Fodor has called attention to the difficulty there is for a modern representationalist view of the mind of harmonizing three basic. and prin/{) facie quite plausible. tenets:"

IMI 182 OCKHAM ON CONCEPTS CONCLUSION 183

• that a sound psychological theory of thought cannot do without intentional with a corresponding disposition of the agent to apply this concept to new instances terms; (through judgements ofthe form 'this is an F' in particular). Secondly, andOckham • that the content of a thought depends upon some external - especially causal is much more loquacious about this one, the concept is expected to contribute - connections it has with the world; something to the truth-conditions of the propositional sequences it occurs in as • that human intellectual processes typically have a syntactical character. subject or predicate. The former role is fulfilled by the concept being a 'similitude' of some externalthings, while the latter is what we can properly call the semantical It so happens that Ockham's theory saliently incorporates a close analog for each one role of the concept: this is where its signification comes in. If concepts are to of these: concepts are taken to be mental signs for things, their signification is made efficiently play both roles at the same time, there had better be a coincidence in most to depend upon causal connections, and they are insistently described as combinable cases between what the agent is disposed to apply the concept to, and what the with each other into syntactically structured units comparable to the sentences of a concept signifies. Yet the two relations must be distinguished in principle in language. How he strives to bring it all together in his mature actus-theory, while Ockham's theory, for the latter is the of the former. The disposition ofahuman consistently subscribing to a nominalistic metaphysics, was the object of the present agent to apply a previously acquired concept to a new instance must be decisively book. sensitive, as we have argued, to the perceptual aspects of the newly met individual. Roughly summarized, what emerges is the following. The basic apparatus Ockham Perceptual aspects, however,can sometimes be misleading. In so far as the postulates is a complex causal network of mental states. Each encounter with real judgement that this newly met thing here is an F is supposed to guide the agent's individuals is supposed to trigger in the mind one or more chains of intellectual acts future expectations and actions, what makes it true or false ultimately should be, in and habitus, ontologically identified with mental qualities. How semantics enters the many cases, the essential causal powers of the thing rather than its superficial picture depends on the roles these intellectual acts are supposed to play in the mental appearance. It should not come as a surprise, then, that what a concept is said to life of human beings. They can be described as signs in the logical sense, Ockham naturally signify, in the logical sense, should be those individuals that are essentially thinks, because they can be combined into more complex units, some of which can - and not only apparently -like the ones that originally triggered its own formation. legitimately be seen as propositions and attributed a truth-value. Be that as it may, Ockham ends up with three basic categories of simple Whether a complex propositional act is true or false is made to depend upon its concepts, which he respectively calls 'absolute' terms, 'connotative' terms, and syntactical structure on the one hand, and the signification of (some of) its syncategoremata: conceptual components on the other hand. And how a simple concept naturally gets a signification is basically presented as a causal story. The human mind is so built, • The syncategorematic concepts, when they occur in actual thought, are the according to the theory, that when the intelligent agent comes in perceptual contact very logical operations of the mind, such as predication, quantification, and so with one or more individuals of the world, a special intellectual act is immediately on. The capacity to accomplish such operations must be supposed to be produced in him, the 'intuitive cognition', which then launches various series of innate, in the. Ockhamistic view, since it essentially belongs to every rational mental effects, giving rise, among other things, to a new mental act, the 'abstractive' soul. But the actual occurrence of these syncategorematic acts normally one, which will be, in turn, the starting point of a temporally spread out chain of requires the accompanying presence of some independently SIgnifying mental future similar acts, with the corresponding habitus in between. Simple concepts in categorematic terms, whether absolute or connotative. this network are those abstractive acts of which no other abstractive act is a part. • Absolute concepts are those that equally signify all their significates without What each one of them signifies will be all the individuals (or groups of individuals) obliquely referring to anything else. The sole connection they require among that relevantly resemble, to some specified degree, the individuals (or groups of their significates is some degree or other ofessential similarity. They are, in pther individuals) that directly caused the occurrence of the original triggering intuitive words, natural kind concepts (with various degrees of generality available). act. The 'first abstractive cognition' in particular - the prima abstractiva .:...signifies • Simple connotative concepts, by contrast, are endowed with a hierarchized all the individuals that are essentially equivalent - simillimi, Ockham says - with the semantical structure: they primarily signify some things, and secondarily - or original Ones. More and more general abstractive cognitions are then supposed to be obliquely - signify some other things. Their significates, in other words, are progressively induced in the mind as the combined effects of previously acquired arranged in ordered n-tuples. These concepts are normally produced within the ones, with less and less stringently similar individuals as their significates. mind by encounters with groups of individuals, rather than with isolated Why exactly should such concepts signify individuals that are essentially similar instances, and are thus made to signify, directly or obliquely, all the individual to each other rather than any other arbitrarily chosen bunch, is something Ockham participants to similarly ordered n-tuples. Despite their special hierarchized does not explain in any detail. What we are left to suppose is that this must depend structure, they can legitimately be said to be simple acts of conceptual cognition somehow upon the functions of the conceptual units in the mental life of the agent. in so far as no part of them is itself an independently signifying concept. As seen in chapter 7, Ockham postulates at least two such functions. First, concepts are thought to play a recognitional role. Without insisting much on the point, In the end, all human reasoning must amount, in Ockham's theory, to combinations Ockham standardly presupposes (hat each simple concept is associated in the mind of such simple conceptual units, in accordance with the rigid constraints of an 184 OCKHAM ON CONCEPTS CONCLUSION 185

innately implemented grammar.. The representational value of thought is thus mind at the moment of utterance, but upon what happened at the original ultimately made to rest on the natural signification of our simple categorematic moment imposition. Some of Ockham's developments even suggest, as we concepts, whether absolute or connotative. saw in chapter 9, that a related form of externalism might extend to certain With respect to the.current discussions in philosophy of mind, this theory displays conceptual units, in so far as they could receive a stipulative signification in several interesting features. Let me brieflyrecall some of them: addition to their natural one. (xi) The adoption of a strong nominalistic constraint. All the relata. of the (i) Epistemological realism. Although universals are not countenanced as real cognitional process - agents, concepts, significates, and referents - are held things in the world, cognition by means of general concepts is considered as to be singular entities. ' adequate in principle to reality. The theory admits, in particular, that things are really ordered out there in the world in various ways (temporal, spatial, The final picture we arrive at is quite remote, on the whole, from what has been the causal, etc.), and that they can, in many cases, be known to be so ordered. standard account of Ockham's approach to concepts in the last decades; much ofthis (ii) Representationalism in philosophy ofmind. It is assumed that human thought book has been devoted to a critical discussion of the predominant 'reductivistic' is composed, in a fundamental way, of units or processes which are view, chiefly put forward by the American scholar Paul Vincent Spade. Ockham's semantically evaluable (as referring to something else, for example) and language of thought, in this interpretation, was supposed to be a logically ideal syntactically structured. language,and should, consequently, be devoid of ambiguities and redundancies; it (iii) A causal approach to cognition. The semantical values of the mental should not simultaneously'contain, in particular, a simple term and its complex representational units ultimately depend upon their (often indirect) causal nominal definition: the only simple terms it can have in the end (in addition to connections with the world. These units, in turn, play a causal role in the syncategoremata) are the indefinable - and non-relational - 'absolute' concepts. psychological life ofthe agent. What we are now led to conclude is that this approach cannot legitimately be (iv) The idea that concept$ are real mental states, or acts, rather than purely ideal attributed to Ockham himself. Some ambiguities are indeed accepted by him in the objects grasped, or produced, by the mind. This allows for an appreciable language of thought,7 and Some redundancies as well. Some connotative terms ­ simplification of the ontology of the mental, by avoiding the position of a relational ones, in particular - are admitted among naturally acquired simple special mode of being for intentional objects. concepts, and they are authorized to coexist in human minds with their complex (v) The postulation of an innate mechanism of abstraction, in virtue of which, in definitions. Even the rejection of strict synonymy among simple concepts, which is particular, a single encounter with a given individual suffices to produce in explicit enough in some of Ockham's texts, cannot be counted as a central tenet in the human mind the formation of a general concept of the' most basic level. his doctrine. 8 This makes the postulation of innate categorematic concepts superfluous. Our whole understanding of Ockham's nominalistic programme has turned out to (vi) Semantical atomism and the compositionality of thought. The basic units of be at stake in these technical points. Spade reasoned that since Ockham's ontology human thought are simple concepts, considered as signs. They naturally admitted of only two categories of individuals in the world - substances and qualities acquire .their signification before occurring .as the components of - only absolute substance concepts and absolute quality concepts could be accepted propositional or inferential sequences. Once significant, they are combinable by him as simple representational units in the mind; everything else in intellectual with each other in determinate ways, and the semantical properties of the thought should be accounted for on this narrow basis. The success of Ockham's complex units thus produced are usually derivable from the signification of nominalism was thought to crucially rest on the reduction of the connotative terms, their components, given the structure of the whole. including all the relational ones, by means of nominal definitions. Which, in the end, (vii) The distinction between the signification of concepts and the referential was unfeasible! The result of our inquiry, fortunately, is that Ockham's real aim was function (or suppositio) they happen to have in the context of a given elsewhere. What the programme he puts forward is supposed to accomplish is to propositional· sequence. This is'a crucial distinction for Ockham, and a very make it clear that none of the concepts and propositions that are needed for human useful one indeed, which is unfortunately neglected in contemporary knoWledge carries with it an ontological commitment to any entity but singular semantics (it does not correspond, in particular, to the famous Fregean couple substances and qualities. This is done not by reducing other concepts to absolute of ). substance and quality concepts, but by showing that all the relata of the relevant (viii) The idea of connotation and its application to simple concepts. Some of the semantical connections - including connotation - are singular substances and basic conceptual units in our mental equipment are attributeda hierarchized qualities. semantical structure. Among them, in particular, are some relational terms. In the last analysis, what stood behind the reductivistic interpretation of Ockham, (ix) An innatist position with respect to syncategorematic concepts. The human it seems to me, was the misleading assumption that the language of thought, as he mind is assumed to be innately capable of logical operations. conceives of it, should be entirely transparent to the thinking agent. It is this (x) An externalistic conception of the meaning ()f' words. The conventional presupposition. mainly, that made it difficult for the proponents of this interpretation signification of a spoken sound depends not upon what the speaker has in to understand how a connotative concept could bl' kept distinct from its definition in 186 OCKHAM ON CONCEPTS

Ockham's approach. Yet the Venerabilis Inceptor, as we have found, did not take the conceptual content of our mental propositions to be thoroughly manifest to us in all Bibliography cases. Not only did he accept ambiguities as possible in mentalese, but he is also committed to admit (as we haye seen in chapter 5) that one can have a concept without knowing what its definition is, and without being clear as to what entities exactly this concept signifies, and what their ontological status is. The upshot is that the externalistic trend in Ockham's theory of concepts should not be minimized: the 1. Ockham's works signification of simple concepts is made to depend upon external relations that the Ockham's philosophical and theological writings have been edited by the Franciscan Institute agent might not be entirely aware of. And this ipso facto introduces some degree of of St. Bonaventure University under the direction of Fr. Gedeon Gal in two series: opacity right at the heart of the mental language! This is not a problem for the Guillelmi de Ockham Opera Philosophica [abbr. : OPh], 7 vols., St. Bonaventure, NY: The Venerabilis Inceptor's philosophical programme, because his goal was not to Franciscan Institute, 1974-1988. construe human thought as a logically ideal and semantically translucent system. It Guillelmi de Ockham Opera Theologica [abbr. : OTh], 10 vols., St. Bonaventure, NY : The was to clean up the ontology. Franciscan Institute, 1967-1986. Lac des Erables Quoted in the present book are the following : December 2003 Expositio in Libros Physicorum Aristotelis [abbr. : Exp. in Phys.], V. Richter et al. eds., OPh IV"V, 1985. Expositio in Librum Perihermenias Aristotelis [abbr. : Exp. in Perih.], A. Gambatese and S. Notes Brown eds., in OPh II, 1978, pp. 341-504. Expositio in Librum Porphyrii de Praedicabilibus [abbr. : Exp. in Porph.], E. A. Moody ed., in OPh II, 1978, pp. 1-131 (Eng!. trans!. in Kluge 1973-74). 1. Laurence and Margolis 1999, p. 3. Expositio in Librum Praedicamentorum Aristotelis [abbr. : Exp. in Praedic.], G. Gal ed., in 2. Ibid., p. 71. OPh II, 1978, pp. 133-339; 3. See Fodor 1998, p. 161. Expositio super Libros Elenchorum [abbr. : Exp. sup. Elench.], F. Del Punta ed., OPh III, 4. Ibid., p. viii (with Fodor's italics). 1979. . 5. Margolis 1998, p. 550. Ordinatio. Scriptum in Librum Primum Sententiarum [abbr. : Ord.], G. Gal et al. eds., OTh 1­ 6. See Fodor 1995, esp. pp. 1-26. IV, 1967-,-1979. 7. Spade 1974 had already noticed the point, but he suggested that this was an Quaestiones in Libros Physicorum Aristotelis [abbr. : Quaest. in Phys.], S. Brown ed., in OPh inconsistency on Ockham's part. VI, 1984,pp. 395-813. 8. As argued above in chapter 7, section 4.2. Quaestiones Variae [abbr. : Quaest. Var.], G. I. Etzkorn et al. eds., OTh VIII, 1984. Quodlibeta Septem [abbr. : Quodl.], J. C. Wey ed., OTh IX, 1980 (Engl. trans!. in Freddo80 and Kelley 1991, and Freddoso 1991). Reportatio. Quaestiones in Libros II, Ill, IV Sententiarum [abbr. : Rep.], G. Gal et al. eds., OTh V-VII, 1981-1984. Summa Logicae [abbr. : SL], P. Boehner, G. Gal and S. Brown eds, OPh I, 1974 (partial Engl. trans!. in Loux 1974, and Freddoso and Schuurman 1980). Summula Philosophiae Naturalis [abbr. : Summ. Phil. Nat.], S. Brown ed., OPh VI, 1984, pp. 135-394. Tractatus de Corpore Christi [abbr. : Tract. de Corp. Chr.], C. A. Grassi ed., in OTh X, 1986, pp.87-234. Tractatus de Quantitate [abbr. : Tract. de Quant.], C. A. Grassi ed., in OTh X, 1986, pp. 1-85.

2. Other references

N.B.- It has not seemed useful to include such classical authors as Aristotle, Thomas Aquinas or John Locke in the present bibliography. References to their works when they occurred were given in the standard ways in the relevant footnotes. Adam de Wodeham. See Wood and Gal 1990. Adams. Marilyn (1970), '(n!uitive cognition. certainty. and scepticism in William Ockham', '1im/ifill. 20..lH9--9H.

1M? 188 OCKHAM ON CONCEPTS BIBLIOGRAPHY 189

~- (1987), William Ockham, 2 vols., Notre Dame, IN: University of Notre Dame Press. Eco, Umberto and Costantino Marmo,eds. (1989) On the Medieval Theory of Signs, Alferi, Pierre (1989), Guillaume d'Ockham. Le singulier, Paris: Minuit. Amsterdam: John Benjamins. Almog, Joseph, , and Howard Wettstein, eds (1989), Themesfrom Kaplan, Oxford: Fodor, Jerry A. (1995), The Elm and the Expert, Cambridge, MA: MIT Press. . ~- (1998), Concepts. Where Cognitive Science Went Wrong, Oxford: Clarendon Press. Ashworth, Jennifer (1973), 'The doctrine of Exponibilia in the fifteenth and sixteenth Frank, W.and Girard J. Etzkorn, eds (1985), Essays Honoring Allan B. Wolter, St. centuries', Vivarium, 11, 137-67. Bonaventure, NY: The Franciscan Institute. ~- (1981), 'Mental language and the unity of propositions', Franciscan Studies, 41, 61-96 Fredborg, K. M., Lauge Nielsen, and Jan Pinborg (1978), 'An unedited part of Roger Bacon's (reprinted in Ashworth 1985). Opus Maius: De Signis', Traditio, 34, 75-136; ~- (1985), Studies in Post-Medieval Semantics, London: Variorum Reprints. Freddoso, Alfred J. (1991), William of Ockham. Quodlibetal Questions, vol. II: Quodlibets Augustine. See Martin 1962. 5-7, New Haven, CT: Yale University Press. Bakker, Paul J. J. M., ed. (2002), Chemins de la pensee medievale. Etudes offertes aZenon Freddoso, Alfred J. and Francis E. Kelley (1991), William of Ockham. Quodlibetal Questions, Kaluza, Tumhout: Brepols. vol. I: Quodlibets 1-4, New Haven, CT: Yale University Press. Beretta, Beatrice (1999), Ad aliquid. La relation chez Guillaume d' Occam, Fribourg, Suisse: Freddoso, Alfred J. and Henry Schuurman (1980), Ockham's Theory of Propositions. Part II Editions Universitaires. of the Summa Logicae, Notre Dame, IN: University of Notre Dame Press: Berube, Camille (1964), La connaissance de l' individuel au Moyen Age, Montreal: Presses de Gal, Gedeon (1967), 'Gualteri de Chatton et Guillelmi de Ockham controversiade natura l'Universite. de Montreal, and Paris: P.U.F. concept4s universalis', Franciscan Studies, 27, 191~212. Biard, Joel (1981), 'La redefinition ockhamiste de la signification', Miscellanea Mediaevalia, Gaskin, Richard (2001), 'Ockham's mental language, connotation, and the inherence regress', 13, 451~8. in Perler 2001, 227-63. ~- (1989), Logique et tMorie du signe au XlVe siecle, Paris: Vrin. Gelber, Hester (1984), "'I cannot tell a lie": Hugh Lawton's critique of Ockham on mental ~- (1997), Guillaume d' Ockham. Logique et philosophie, Paris: P.U.F. language', Franciscan Studies, 44, 141-79. Boehner, Philotheus (1943), 'The notitia intuitiva of non-existents according to William of Goddu, Andre (1984), The Physics of William ofOckham, Leiden: Brill. ~- Ockhain', Traditio, 1,223-75 (also in Boehner 1958, 268-300). (1993), 'Connotative concepts and mathematics in Ockham's natural philosophy', ~- (1945), 'In propria causa', Franciscan Studies, 5, 37-54 (also in Boehner 1958, Vivarium, 31, 106-139. 300-319). Goodman, Nelson (1956), 'A world of individuals', in The Problem of Universals (with ~- (1951), 'The relative date of Ockham's Commentary on the Sentences', Franciscan Alonzo Church and I. M. Bochenski), Notre Dame, IN: University of Notre Dame Press (also in Goodman 1972, 155-72). Studies, 11,305-316 (also in Boehner 1958, 96-110). ~- ~- (1958), Collected Articles on Ockham, St. Bonaventure, NY: The Franciscan Institute. (1972), Problems and Projects, Indianapolis, IN: Bobbs-Merrill. Gregory of Rimini. See Trapp et al. 1978. Boler, John (1985), 'Connotative terms in Ockham', History of Philosophy Quarterly, 2, 21-37. Hamesse, Jacqueline (1974), Les Auctoritates Aristotelis. Un jlorilege medieval, Louvain: Publications Universitaires, and Paris: Beatrice-Nauwelaerts. Bos, Egbert P. (1994), Johannes Buridanus. Summulae in Praedicameilta, Nijmegen: Hamlyn, D. W. (1993), Aristotle. De Anima. Books II and Ill, Oxford:Clarelldon Press. Ingenium. Henninger, Mark G. (1989), Relations. Medieval Theories 1250-1325, Oxford: Clarendon Braakhuis, H. A. G and C. H. Kneepkens, eds (2003), 'sPeri Hermeneias in the Press. Latin Middle Ages. Essays on the Commentary Tradition, Groningen: Ingenium. Rochart, Patrick (1972), 'G!Iillaume d'Occam : Le signe et sa duplicite', in Chatelet 1972, Brown, Deborah (1996), 'The puzzle of names in Ockham's theory of mental language', The 183-203. Review of Metaphysics, 50, 79-99. Hoffmann, Fritz (1988), Crathorn. Quiistionen zum ersten Sentenzenbuch, MUnsler: Brown, Stephen (1985), 'Walter Chatton's Lectura and William of Ockham's Quaestiones in Aschendorff. Libros Physicorum Aristotelis', in Frank artd Etzkorn 1985, 81-115. Hornsby, Jennifer (1998), 'Action', in Routledge Encyclopedia of Philosophy, I, London: Cesalli, Laprent (2001), 'Le realisme propositionnel de Walter Burley', Archives d' Histoire Routledge, 37-41. Doctrinale et Litterairedu Moyen Age, 68, 155-221. John Buridan. See Bos 1994, King 1985, Patar 1991, Scott 1977, and Van der Lecq 1998. Chalmers, David (1999), 'Is there synonymy in Ockham's mental language?', in Spade 1999, Kaplan, David (1978), 'On the logic of demonstratives', Journal of , 8. 76-99. 81-98. Chatelet, Fran\(ois, ed. (1972), La philosophie medievdle, Paris: Hachette. ~- (1989), 'Demonstratives', inAlinog et al. 1989,481-563. Conti, Alessandro (2000), 'Significato e verita in Walter Burley', Documenti e studi sulla Karger, Elizabeth (1978), 'Consequences et inconsequences de la supposition vide dans la tradizionefilosofica medievale, 11,317-350. logique d'Ockham', Vivarium, 16,46-55. Crathorn. See Hoffmann 1988. ~- (1994), 'Theories de la. pensee, de ses objets et de son discours chez Guillaume Davidson, Donald and , eds (1972), Semantics of Natural Language, d'Occam', Dialogue, 33,437-56. Dordrecht: Reidel. ~- (1996), 'Mental sentences according to Burley and to the early Ockham', Vivarium. 34. Day, Sebastian (1947), Intuitive Cognition. A Key to the Significance (!f'Later Scholastics, St. 192-230. Bonaventure, NY: The Franciscan Institute. ~- (1999a), 'Ockham's misunderstood theory of intuitive and ahstractive cognition', in De Andres, Teodoro (1969), EI nominalismo de Guillermo de Ockham como tilosolla del Spade 1999.204-226. lenguaje. Madrid: Clredos. -_.- (191.)90), 'Waller Burley's ~ealism',Vi\'(/rillm, .n, 24-4(), 190 OCKHAM ON CONCEPTS BIBLIOGRAPHY 191

Kaufmann, Matthias (2003), 'The discussion of the nature of the concept in Ockham's -- (1990), Being and Knowing. Studi~,ill Thomas Aquinas and Later Medieval Perihermeneias commentary', in Braakhuis and Kneepkens 2003,119-33. Philosophers, Toronto: Pontifical Institute of MedileYalStudies. Kelley, Francis E. (1981), 'Walter Chatton vs. Aureoli and Ockham regarding the universal -- (1999), The Philosophy of William ofO~ III,,.. Light of Its Principles, Toronto: concept', Franciscan Studies, 41, 222-49. Pontifical Institute of Mediaeval Studies, King, Peter (1985), 's Logic. The Treatise on Supposition. The Treatise on McGrade, Stephen (1986), 'Seeing things: Ockhun IlIclr.preaentationalism', in Wenin 1986, Consequences, Dordrecht: Reidel. 591-7. -_. (2002), 'The failure of Ockham's nominalism', http://www.fordham.edu/gsas/phil/ Meier-Oeser, Stephan (1997), Die Spur des UiCMIII." Z.kMn und seine Funktion in der klima/pking.htm Philosophie des Mittelalters und der fruhen Neuz,',. Berlin: Walter de Gruyter. Klima, Gyula (1988), ArsArtium. Essays in Philosophical Semantics, Mediaeval and Modern, Mertz, D. W. (1997), and Its Logic, NIw Haven,CI': Yale University Press. Budapest: Hungarian Academy of Sciences. Michon, Cyrille (1994), Nominalisme. La tMorie de ItI,i,IIUfctllion d'Occam, Paris: Vrin. -_. (1993), 'The changing role of entia rationis in mediaeval semantics and ontology: a -- (2000), 'Le nominalisme et les relations', Le~"",.,ltWoi", I, 121-51. comparative study with a reconstruction', Synthese, 96, 25-58. Minio-Paluello, Laurentius (1966), Aristoteles LatirllU I, ~7:Porphyrii et Anonymi -- (1999),'Ockham's semantics and ontology of the categories', in Spade 1999, 118-42. Fragmentum vulgo locatum 'Liber Sex PrincipiarlUll', Bruge.: Descl6e de Brouwer. -- (2001), 'Buridah's theory ofdefinitions in his scientific practice', in Thijssen and Zupko Normore, Calvin (1990), 'Ockham on mental languaae', In Smith 1990,53-70. 2001,29-47. O'Callaghan, Jeremiah (1955), 'The second question of the Prologue to Walter Catton's -- (2002), 'Comments on Peter King "The failure of Ockham's nominalism''', Commentary on the Sentences on intuitive and abstractive knowledge', in O'Donnell http://www.fordham.edu/gsas/phil/klima/king.htm 1955,233-69. Kluge, Eike-Henner W. (1973-74), 'William of Ockham's Commentary on Porphyry. O'Donnell, J. Reginald, ed. (1955), Nine Mediaeval Thinkers. A Collection of Hitherto Introduction and English translation', Franciscan Studies, 33, 171-254, and 34, 306-382. Unedited Texts, Toronto: Pontifical Institute of Mediaeval Studies. Kneale, William and Martha Kneale (1962), The Development of Logic, Oxford: Clarendon Panaccio, Claude (1980), 'Occam et les d6monstratifs', Historiographia Linguistica, 7, Press. 189-200. Kretzmann, Norman (1982a), 'Syncategoremata, exponibilia, sophismata', in Kretzmann, et '-- (1984), 'Propositionalism and atomism in Ockham's semantics', Franciscan Studies, 44, al. 1982, 211-45. 61-70. (1990a), 'Connotative terms in Ockham's mental language', Cahiers d'epistemologie --, ed. (1982b), Infinity and Continuity in Ancient and Medieval Thought, Ithaca, NY: -- no. 9016, Montreal: Universite du Quebec 11Montreal. Cornell University Press. (1990b), 'Reference et representation' (about Alferi 1989), L'Age de la science, 3, Kretzmann, Norman, and JanPinborg, eds. (1982), TheCambridge History -- 287-99. of Later Medieval Philosophy, Cambridge: Cambridge University Press. (1992a), Les mots, les concepts et les choses. La semantique de Guillaume d' Occam et Kripke, Saul (1972), 'Naming and necessity', in Davidson and Harman 1972,253-355. -- Ie nominalisme d' aujourd' hui, Montreal: Bellarmin, and Paris: Vrin. Laurence, Stephen and Eric Margolis (1999), 'Concepts and cognitive science', in Margolis (1992b), 'From mental word to mental language', Philosophical Topics, 20,125-47. and Laurence 1999,3-81. -- -- (1999a), Le discours interieur. De Platon a Guillaume d'Ockham, Paris: Seuil. Leff, Gordon (1975), William of Ockham. The Metamorphosis of Scholastic Discourse, -- (1999b), 'Semantics and mental language', in Spade 1999,53-75. Manchester: Manchester University Press. -- (2001), 'Aquinas on intellectual representation', in Perler 2001, 185-201. Lindbergh, David C (1976), Theories of Vision from. Al-Kindi to Kepler, Chicago, IL: The -- (forthcoming), 'Ockham and Buridan on epistemic sentences: Appellation of the form University of Chicago Press. and appellation of reason', in the acts of the conference 'Logica Modernorum' held in Loux, Michael (1974), Ockham's Theory ofTerms. Part I of the Summa Logicae, Notre Dame, Amsterdam in 1998, Leiden: Brill. IN: University of Notre Dame Press. Panaccio, Claude and Ernesto Perini-Santos (forthcoming), 'Guillaume d'Ockham et la -- (1979), 'Significatio and suppositio. Reflections on Ockham's semantics', The New suppositio materialis', Vivarium, 42. , 53, 407-27. Parrett, Hermann, ed. (1976), History of Linguistic Thought and Contemporary , Maieril, Alfonso (2002), 'Linguaggio mentale e sincategoremi nel secolo XIV', in Bakker Berlin: Walter de Gruyter. 2002,3-25. Pasnau, Robert (1993), 'Petri Iohannis Olivi Tractatus de Verbo', Franciscan Studies, 53, Margolis, Eric (1998), 'How to acquire a concept', Mind and Language, 13 (also in Margolis 121-53. and Laurence 1999, 549-67). -- (1997), Theories of Cognition in the Later Middle Ages, Cambridge: Cambridge Margolis, Eric and StephenLaurence, eds (1999), Concepts. Core Readings, Cambridge, MA: University Press. MIT Press. Patar, Benoit (1991), Le Traite de l'ame de Jean Buridan, Louvain-Ia-Neuve: Editions de Martin, I. (1962), Augustine. De Doctrina Christiana, Tumhout: Brepols. l'Institut Superieur de Philosophie. Maurer, Armand (1978), 'Method in Ockham's nominalism', The Monist, 61,426-43 (also in -- (1995), Nicolai Oresme Expositio et Quaestiones in Aristotelis De Anima, Louvain-Ia­ Maurer 1990,403-21). Neuve: Editions de l'Institut Superieur de Philosophie, and Louvain: Peeters. -- (1981), 'William of Ockham on language and reality', Miscellanea Mediaevalia, 13, Perler, Dominik (I 996a), 'Things in the mind. Fourteenth-century controversies over 795-R02 (also in Maurer 1990, 423-30). "intelligible species'''. Vivarium, 34, 231-53. -~ (19l\4), 'Ockham's Razor and Chalton's Anti-Razor', Medilll'\'al Studies, 46, 463-75 -- (1996b), Review of Spruit 1994-95, Vivarium, 34, 2l\O-83. (also in Maurer 1990,431--4.1). -~~,cd. (200 I), Ancient ami M,·tI//'I'al 1'11,'01'/".\'of IlItl'lItiollality, Leiden: Brill. 192 OCKHAM ON CONCEPTS BIBLIOGRAPHY 193

Peter John Olivi. See Pasnau 1993. Tachau, Katherine H. (1988), Vision and Certitude in the Age of Ockham, Leiden: Brill. Pinborg, Jan (1976), 'Some problems of semantic representations in medieval logic', in Thijssen, J. M. M. H. and Jack Zubko, eds (2001), The Metaphysics and Natural Philosophy Parrett 1976, 254-78 (reprinted in Pinborg 1984). of John Buridan, Leiden: Brill. -- (1984), Medieval Semantics, London: Variorum Reprints. Trapp, Damasius et al. (1978- ), Gregorii Ariminensis Lectura super Primum et Secundum Porphyry. See Minio-Palluelo 1966. Sententiarum, Berlin: Walter de Gruyter. Putnam, Hilary (1975a), 'The meaning of "meaning''', in Putnam 1975b, 215-71. Trentman, John (1970), 'Ockham on mental', Mind, 79, 586-90. -- (1975b), Mind, Language .and Reality [=Philosophical Papers, vol. 2], Cambridge: Tweedale, Martin (1990), 'Mental representations in later medieval scholasticism', in Smith Cambridge University Press. 1990,35-51. -.- (1988), Representation and Reality, Cambridge, MA: MIT Press. -.- (1992), 'Ockham's supposed elimination of connotative terms and his ontological Reina, Maria Elena (2002), HocRic etNunc. Buridano, Marsilio di Inghen e la conoscenza parsimony', Dialogue, 31, 431-44. del singolare, Florence: Leo S. Olschki. Van der Lecq, Ria (1998), Johannes Buridanus. Summulae de Suppositionibus, Nijmegen: Richard Campsall (Pseudo). See Synan 1982. Ingenium (Engl. transl. in King 1985). Rijk, Lambert Marie de (1967), Logica Modernorum. A Contribution to the History of Early Walter Chatton. See G::il1967,Kelley 1981, O'Callaghan 1955, Wey 1989, Weyand Etzkorn Terminist Logic, vol. II: The Origin and Early Development of the Theory of Supposition, 2002. Assen: Van Gorcum. Wenin, Christian, ed. (1986), L'homme et son univers au Moyen Age, Louvain-la-Neuve: Roger Bacon. See Fredborg et al. 1978. Editions de l'Institut Superieur de Philosophie. Rosier, Irene (1994), La parole comme acte. Sur la grammaire et la semantique au XlIle siecle, Wey, Joseph C. (1989), Walter Chatton. Reportatio et Lectura super Sententias: Collatio ad Paris: Vrin. Librum Primum et Prologus, Toronto: Pontifical Institute of Mediaeval Studies. Russell, Bertrand (1918), 'The philosophy of ', in Russell1956, 175-281. Wey, Joseph and Girard 1. Etzkorn (2002), Walter Chatton. Reportatio super Sententias Liber -- (1956), Logic and Knowledge. Essays 1901-1950, ed. by R. C. Marsh, London: George I, Distinctions 1-9, Toronto: Pontifical Institute of Mediaeval Studies. Allen and Unwin. Wittgenstein, Ludwig (1921), Tractatus Logico-Philosophicus, transl. by D. F. Pears and B. F. -- (1959), My Philosophical Development, London: George Allen and Unwin. McGuinness, London: Routledge and Kegan Paul, 1961. Scott, Theodore K. (1977), Johannes Buridanus. Sophismata, Stuttgart: Frommann­ Wood, Rega (1997), Ockham on the Virtues, West Lafayette, IN: Purdue University Press. Holzboog. Wood, Rega and Gedeon Gal (1990), Adam de Wodeham. Lectura Secunda in Librum Primum Smith, 1. C., ed. (1990), Historical Foundations of Cognitive Science, Dordrecht: Kluwer. Sententiarum, St. Bonaventure, NY: St. Bonaventure University, 3vols. Spade, Paul Vincent (1974), 'Ockham's rule of supposition: Two conflicts in his theory', Yrjonsuuri, Mikko (1997), 'Supposition and truth in Ockham's mental language', Topoi, 16, . Vivarium, 12,63-7 (reprinted in Spade 1988). 15-25. -- (1975), 'Ockham's distinctions between absolute and connotative terms', Vivarium, 13, Zavalloni, Roberto (1951), Richard de Mediavilla et la controverse sur la pluralite des formes, 55-76 (reprinted in Spade 1988). Louvain: Publications Universitaires. -.-. (1980), 'Synonymy and equivocation in Ockham's mental language', Journal of the Zheng, Yiwei (1998), 'Metaphysical simplicity and semantical complexity of connotative History of Philosophy, 18,9-22 (reprinted in Spade 1988). terms in Ockham's mental language', The Modern Schoolman, 75, 253-64. -. - (1988), Lies, Language and Logic in the Late Middle Ages, London: VariorumReprints. -- (1990), 'Ockham, Adams and connotation. A critical notice of Marilyn Adams, William Ockham', The Philosophical Review, 99,593--612. -.- (1996), Thoughts, Words and Things: An Introduction to Late Mediaeval Logic and Semantic Theory, http://www.pvspade.com/Logic/index.html -- (1998), 'Three versions of Ockham's reductionist program', Franciscan Studies, 56, 347-58. --, ed. (1999), The Cambridge Companion to Ockham, Cambridge: Cambridge University Press. Spruit, Leen (1994-95), 'Species Intelligibilis'. From Perception to Knowledge, 2 vols, Leiden: Brill. Streveler, Paul (1975), 'Ockham and his critics on intuitive 'cognition', Franciscan Studies, 35, 223-36. Stump, Eleonore (1982), 'Theology and physics in De sacramento altaris: Ockham's theory of indivisibles', in Kretzmann 1982b, 207-230. . -- (1999), 'The mechanisms of cognition: Ockham on mediating species', in Spade 1999, 168-203.

Synan, Edward A. (1982), The Works o.t"Richard (~t"Campsall, Toronto: Pontifical Institute of Mediaeval Studies. Tabarroni, Andrea (1989), 'Mental signs and the theory of representation in (kkham', in Ero and Marmo 1989, 195-224. Index

abbreviations 93--4, 114-15 cospecificity see maximal similarity absolute terms (or concepts) 63-6, 69, 75, Crathom see William Crathorn 85, 89-90, 92, 103, 105, 108-9, 112, 126-9,131-7,170-71,175,183 Day, Sebastian 17 abstractive cognition 6-7, 9, 113, 121-2, De Andres, Teodoro 60 124,127,174,182 definition see nominal definitions actus-theory (of concepts) 8-9, 23-7, 47, Definitional View (of concepts) 85, 113-16 119-22, 151--4, 158 Donatus 8 Adam de Wodeham 38,148,160 Adams, Marilyn 1, 11-13, 16-20,25,37-8, equivocity 167-8, 172-3 69,74,77,98,102,116,119,139,147, Etzkorn, Jerry 38-9 153, 155, 160-62 exponib1e propositions 73--4 Alferi, Pierre 17,20,119,139 externalism 165,170-71,175-6,184-6 Alhazen 27 Aquinas see Thomas Aquinas jictum-theory (of concepts) 8, 23-7, 35, Aristotle 8, 21-2,28, 30, 34, 42, 51, 55, 60, 106, 119-21, 146-51, 158 . 67,70,78-9,81, 121-2, 139, 163, 181 Fodor, Jerry 3, 60, 97, 181-2, 186 Ashworth, Jennifer 74, 82 Frege, Gottlob 2, 65, 89, 181 assent see judgement Augustine 45, 58 Gal, Gedeon 37-9, 98 Gaskin, Richard 74-7,79,82-3,97, 139, Bacon see Roger Bacon 143, 156, 162 Beretta, Beatrice 77, 97, 102, 151, 160-61 Gelber, Hester 160 Berube, Camille 17 Goddu, Andre 77 Biard, Joel 39, 60, 119, 122-3, 139--40 Goodman, Nelson 161 Boehner, Philotheus 17-18 Gregory of Rimini 32-3, 41 Boethius 8, 139 Boler, John 71 habitus 7,21-3,28-30,36,56-8,137,172, Brown, Deborah 101, 133, 136-8, 142-3 182 . Brown, Stephen 37, 98 Henninger, Mark G. 102 Buridan see John Buridan . Hochart, Patrick 60 Hornsby, Jennifer 36 Campsall see Pseudo-Richard Campsall categories (Aristotelian theory of) 23,67, imposition 166, 168, 170-76' 75~6,156-8 in recto I in obliquo 71-2,87-93, 110-12, Cesalli, Laurent 37 156-7 Chalmers, David 101 innatism 96, 150, 154, 157-8, 183--4 Chatton see Walter Chatton intellectual acts see mental acts Classical View (of concepts) see intelligible species 27-31 Definitional View intentions see second intentions connotative terms (or concepts) 63-77, intuitive cognition 6-7, 11-14,21-2,29, 85-'97, 103-16, 129-31, 154-8, 169, 103, 107, 110, 112-13, 124, 126-7, 171, 173-5, 183-5 174, 182 Conti, Alessandro 37 copula XI), 1)1, 96, 145. 14X, 152-3, ISS-X John Buridan 72,82, 125, 127, 141, 156-7, correlative terms 6X I), X6, III 12. 156 161-3

IlJ~ 196 OCKHAM ON CONCEPTS INDEX 197

John Duns Scotus 5, 25, 28, 31, 38,105, Panaccio, Claude 18-20, 37-8,41, 58-61, Streve1er, Paul 17 Trentman, John 143 166, 176-7, 181 77-8, 139-41, 162, 176, 179 Stump, Eleonor 17, 30, 39, 101 Tweedale, Martin 39, 77, 79-80, 83 27 Pasnau, Robert 29, 39-40 subordination 114, 151, 165-76 type / token 55--'-8,136, 171-3 Judgements 31, 35-6,148 Perini-Santos, Ernesto 179 substance concepts 126-9, 131-4, 137, 139 Perler, Dominik 30, 39 supposition (suppositio) 5, 8-9, 11-12, Walter Burley 12,23,37-8 Kaplan, David 13, 20 Peter Aureoli 38 14--15,26,49-54,63,94,120,150,166, Walter Chatton 23-7, 37-8, 41 Karger, ElizabethJ7-20, 23, 37,42,120, PeterJohn Olivi 38 174--5, 181 Wey, Joseph 38-9 140, 148, 160 Peter of Ailly 74 syncategoremata 32, 34, 49-50, 52, 54, 66, William Crathorn 122, 131, 140 Kelley, Francis E. 37 Peter of Spain 8 69, 89, 91, 112, 145-58, 173-4, 183-4 William of Sherwood 8 King, Peter 82 Pinborg, Jan 82 synonymy 66, 69-74, 76-7, 85-7, 89,92, William of Ware 38 Klima, Gyula 133--,6, 142 Plato 28 114, 136-9, 176, 185 Witelo 27 Kneale, William and Martha 19 Porphyry 97 Wittgenstein, Ludwig 145,159 Kretzmann, Norman 59, 162 Priscian 8 Tabarroni, Andrea 18 Kripke, Saul 3, 178 proper names see singular terms Tachau, Katherine 17,20,29,39-40 Yrjonsuuri, Mikko 82 Pseudo-Richard Campsall 117 terms see absolute tenns, connotative tenns, language of thought see menta11anguage Putnam, Hilary 3, 138,143, 165, 171, correlative terms, in recto/in obliquo, Zavalloni, Roberto 17 Laurence, Stephen 97, 181,186 176-7, 179 mental signs, relational tenns, simple Zheng, Yiwei 115-16, 118 Leff, Gordon 18 signs, type/token Lindbergh, David C. 39 Quine, Willard Van Orman 65 Thomas Aquinas 21, 23, 28~9,31, 37-40, Locke, John 170-71, 177 51,59, 181 Loux, Michael 59-60, 143, 177 Reina, Maria Elena 141 relational tenns (or concepts) 63-9, 93-6, Maieru, Alfonso 160 106-7,109,111-12,115,130-31,155-6, Margolis, Eric 2-3, 97, 181, 186 158,184--5 Maurer, Armand 18, 39 representationalism 16, 29-30, 184 maximal similarity 10, 123, 126, 129, 135 Rijk, Lambert Marie de 18 McGrade, Stephen 20 Roger Bacon 27, 45, 58-9 Meier-Oeser, Stephan 58 Roger Marston 40 mental acts 5-7, 15-16,21-36,55-7, Rosier, Irene 59 122-4, 151-4, 182-4 Russell, Bertrand 2, 13-14,20,63,69,89, mental language 8-15, 51, 55, 63, 66-9, 181 73-5,86-7,91-2,136-9,145-7,156-7, 166, 174-5, 185-6 Scotus see John Duns Scotus mental popositions 11-12,31-4,53-5, second intentions 151-2, 175 136-7, 147-53, 169, 182~ sign see mental sign, signification mental signs 8, 11-12,26'-7,45-58, 166, signification (of concepts) 8, 12,26,48, 173-4, 182; see also signification 51-4, 132-4, 166, 171-5, 182-4; see also Mertz, D. W. 77 mental signs Michon, Cyrille 17,45-50,58-9,76,78, similitude (or similarity) 10-11,25,29, 81-2,97,102,119-21,140,162 119-39, 174, 183; see also maximal modes of signification 70-72, 91 similarity simple terms 31-2, 67-8,75, 103-5, 109, nominal definitions 65-77,85-97, 103, 134, 182-4 109-16, 156-7, 169, 185-6 Simplicius 96, 102 nominalism 1,3, 10,45, 55-8, 63-5, 74, 90, singular terms 11-15, ]36, 174--5 96-7, 115-16, 134, 156,182, 185 Spade, Paul Vincent 4]-2,63-70,72,74--8, Nonnore, Calvin 18,69,74,77, 116, 147, 80-82,85,89,91,97,99-100,103-6, 154-5, 160, 162, 169-70, 177 109,115-10,118,143,152,150-7. 101-3,109--70,177,185-6 Ockham's Razor 9,20,29-30, 120 Spruit, Lccll 28. 39