Essays in Ancient Philosophy

8/12/2019 Essays in Ancient Philosophy

1/410


2/410

Essays n ncient Philosophy


3/410

The University of Minnesota ressgratefully acknowledges publication assistanceprovided for this book by Princeton University


4/410

Essaysin ncient hilosophyMichael Frede

niversityofMinnesota Press inneapolis


5/410

Copyright 1987 by theUniversity ofMinnesota. All rightsreserved.Nopart ofthis publica-tion may be reproduced, stored in a retrieval system, or transmitted, in any form or by anymeans, electronic, mechanical, photocopying, recording, or otherwise, withoutthe prior writtenpermission of the publisher.Published by the Universityof MinnesotaPress, 2037University Avenue Southeast, Min-neapolis, MN55414. Published simultaneouslyinCanadaby Fitzhenry Whiteside Limited,Markham. Printedin the United StatesofAmerica.Library o Congress Cataloging in Publication DataFrede, Michael.

Essays in ancient philosophy.Bibliography: p.Includesindex.1 Philosophy, Ancient. I Title.

B171.F69 1987 180 86-6974ISBN 0-8166-1274-9ISBN 0-8166-1275-7 pbk.)See p vi for further copyright information.

The UniversityofMinnesotais anequal-opportunityeducator an demployer.


6/410

ContentsPreface vii

Introduction: The Study of Ancient Philosophy ixPlato1 Observations on Perception in Plato s Later Dialogues 3Aristotle

2 The Title, Unity, and Authenticity of the Aristotelian 11 ategories

3 Categories in Aristotle 294 Individuals in Aristotle 495 Substance in Aristotle s Metaphysics 726 The Unity of General and Special Metaphysics:Aristotle s Conception of M etaphy sics 81

Stoics7 Stoic vs Aristotelian Syllogistic 998 The Original Notion of Cause 1259 Stoics andSkepticsonClear and Distinct Impressions 151

Skeptics The Skeptic s Beliefs 179 TheSkeptic s Two Kinds of Assent and theQuestion of the Possibility of Know ledge 201

Medicine2 Philosophy and M edicine in An tiquity 225 The Ancient Empiricists 2434 The Method of the So-Called Methodical School of Medicine 2615 On G alen s Epistemology 279

Grammar6 Principles of Stoic Grammar 3017 The Origins of Traditional Gram m ar 338

Notes 363Index of Ancient Authors 375Index of Subjects 379


7/410

The following chapters were originally published as indicated and are reprinted with permission ofthe copyright holder:Chapter3, CategoriesinAristotle, StudiesinAristotle, editedby DominicJ. O'Meara (StudiesinPhilosophyandHistoryof Philosophy, Volume9), 1981,The Catholic UniversityofAm ericaPress. Chapter 5, Substance in Aristotle's Metaphysics, Aristotle on Natureand Living Things,editedbyAllan Gotthelf, 1985Mathesis P ublications, PittsburghPA. Chapter 8 The OriginalNotion of Cause, Doubt and Dogmatism: Studies in Hellenistic Epistemology, edited by M.Schofield, M. Burnyeat,and J. Barnes, 1980Oxfo rd University Press. Chapter 9 StoicsandSkeptics on Clear an d Distinct Impressions, The Sceptical Tradition, edited by M. Burnyeat, 1983Un iversity of California Press. Chapter 10 TheSkeptic's Beliefs, Des SkeptikersMeinungen, Neue Hefte fr Philosophie,AktualitderAntike, Heft15/16,pp. 102-129, 1979Vandenhoeck & Ruprecht, Gttingen. Chapter 11 TheSkeptic's Two Kindsof Assent and theQuestion of the Possibility of Knowledge, Philosophyof History,edited by R. Roty, J. Schneewind,an d Q. Skinner, 1984Cambridge University Press. Chapter 12 Philosophy and MedicineinAntiquity, Human Nature and Natural Knowledge. Essays Presented to M arjorie Grene on the Oc-casion of Her Seventy-Fifth Birthday, edited by Alan Donagan, Anthony N. Perovich, Jr., an dMichaelV. Wedin,pp.211-32, 1986by D. Reidel Publishing Company, Dordrecht, Holland.Chapter 14 TheMethodof the So-Called Methodical Schoolof Medicine, Scienceand Specula-tion: Studies in Hellinistic Theory and Practice, edited by Jonathan Barnes, Jacques Brunschwig,Myles Burnyeat, and Malcolm Scholfield, pp. 1-23, 1983by Cambridge University Press.Chapter15, On Galen's Epistemology, Galen: ProblemsandProspects, editedby V. Nutton,1987 The Wellcome Institute for the History of Medicine, 183 Euston Rd.,London NW12BP.Chapter 16 Principlesof Stoic Grammar, TheStoics, edited byJ.M.Rist, pp.27-75,1978Universityo f Califo rnia Press. Chapter 17 TheOrigin o f Traditional G rammar, HistoricalandPhilosophicalDimensions ofLogic Methodology, an d Philosophy of Science, edited byR.E.Buttsan dJ. Hintikka,pp.51-79, 1977by D. Reidel Publishing Company, Dordrecht, Holland.Chapters 2, 4, and 7 were previously published as fol lows:Chapter 2, TheTitle, Un ity,and Authenticityof theAristotelian Categories, appeared as Titel,Einheit und Echtheit der Kategorien, in Zweifelhaftes im Corpus Aristotelicum, 1983, Walter deGruyter& Co. Chapter 4 Individuals in Aristo tle, appeared as Individuenbei Aristoteles, inAntike and Abendland, 1978,W alter de Gruyter & Co. Chapter 7, Stoic vs. AristotelianSyl-logistic, appeared in Archiv fr Geschichte der Philoso phie, Band 56, Heft 1, 1956(Walter deGruyter &Co.).


8/410

Preface

This volumecontains seventeen papers whichIhave writtenoverthecourseofthe last twelve years and an introduction written for this volume in whichwithsome hesitation I try toexplainhow Iconceiveof mystudyofancient phi-losophy.Mostof thepapershavebeenpublished before butsomeofthemwerepublished involumes which are not readily accessible and three of them hadappeared only in German. Hence I amgratefulto the University of MinnesotaPress for this opportunity to present these papers in a formwhich makes themmore easily available and toPrinceton University which by agrant made thispublicationpossible. But in particular I would like to thank the editorial staffof University of Minnesota Press who went through the manuscript with ex-traordinary care and tact andWolfgang Mann who did an excellent piece ofwork in translating the rather stubborn German of Chapters 2 4 and 10. I amalso indebted to the various publishing houses which allowed me to reprintpapers in this collection. Finally I am glad to have this opportunity to expressmygratitude toPearl Cavanaugh AnnGetson andBunnyRomano whosepatience I have tried too often in the course of the years.


9/410

his page intentionally left blank


10/410

Introduction:TheStudyofAncientPhilosophy

Ancient philosophy can be studied in many ways.1 The thoughts of ancientphilosophers are of great interest not just as philosophical thoughts. Many ofthem in one way or another are also of great historical importance. They helpto explain a great many historical facts not just in the history of philosophybut in many otherhistories e.g. th e history oftheology the history ofpoliticaltheory even the history of literature. Or they are reflections of some historicaldevelopm ent we may be interested in; again this m ay be a developm ent in thehistory ofphilosophy or in some other history even one that at first may seemto have very little to do with philosophy e.g . the rise of literacy . In historicalaccounts of ancient life there are few aspects of that life which do not involvesome reference to the fact that some philosopher had a certain view and manyaspects of thatlife into wh ich philosophy enters quite substantially e.g. Rom anlaw.Equa lly thereis hardly a facet ofancientlifetha t does not find its reflectionin ancient philosoph y and there are m any aspects of that lifew hich seem to havea substantial influence on the thought ofphilosophers. Thusthereare many ap-proaches to the thou ght of ancient philosophe rs all of w hich contribute to a bet-ter understanding of it. One can pursue each of the many histories in which an-cient philosoph y either as a w hole or in part plays a role and try to determ inewhat this role is in am anne r appropriate for thehistory in question. One reasonwhy the study of ancient philosophy is so attractive and so lively is that it allow sfor soma ny interests and approaches. Clearly itwouldbe am istaketo thinkthatthere is only one way to study ancient philosophy.

It would be asgreata m istake to think thatone could fruitfully study the sub-ject in any way one cared to. The different approaches have to be carefully dis-tinguished and kept distinct. Different approaches are appropriate for differentinterests and the results one obtains are relative to this interest and to the ap-proach chosen. Th us one m ight well im agine that one could explain the thoughtof a political philosopher on the d istribution ofgoods in terms of the history of

ix


11/410

x INTRODUCTIONthat philosopher s society and social status in it, if this is the kind of history oneis interested in and if one chooses this approach to the thought of the philosopherin question. But it would be a mistake to think that th e explanation one foundwas the only possible e xplana tion. For the philosopher may have had very goodreasons for his v iews on the distribution of goods, reasons that we find so con-vincing that we feel the need to explain why not everybod y in his society adoptedthem. M oreover, we m ay have no reason to doubt that it was for these reasonsthat he adopted the view in question. Thus, depending on the way w e approachhis thought, we account for his thought in two quite different ways. This doesnot mean there is anything wrong with either of these explanations or that wehave to declare one of them illegitimate. All this does show is that the fact thatsomeone had a philosophical view is an extraordinarily complex fact, and that,if w e want to capture some of its complexity, we have to allow for a wide varietyof approaches to it and resist the temptation to declare one of these approachesth e only legitimate one.

In principle one can look at a philosophical view that someone has held intw o different ways. One can look at it primarily as a philosophical view thatsomeone might entertain; one may wonder whether or not it is t rue, fo r whatkinds of reasons one m ight w an t to take this view , what its implications are, an dentirely disregard, as irrelevant to one s purposes, th e fact that it is a view thathas actually been taken by a certain person u nder certain circum stances. To con-sider a view in this way is to consider it p hilosophically. But one can also lookatthis view prim arily as one that w as actually held, be interested in the fact thatit was the view of a certain person und er certain circumstances, and try to under-stand it as such. Now, presumably, one is not interested in understanding thefact that someone had a certain philosophical view quite independently of whohadth e view a nd what the v iew w as. W e are interested in und erstanding the factthatsomeone had a certain philosophical view only if we th ink this fact has somesignificance, is in some way revealing. The fact that someone held a certainphilosophicalview has some significance, is somehow revealing, if the view in-trinsically is of philosophical interest or if it has considerable historical in -fluence, either in the history of philosophy or in some other history; or it mightbe revealing in a number of other ways, e.g., because it shows how considera-tions or events that form part of some other history influence the thought ofphilosophers, or how the influence of certain events and changes was so perva-sive that it was reflected even in the thought of philosophers. I will call factsaboutthe past that ha ve this significance, that are revea ling, historical facts. Onemight, of course callall facts about the past historical facts. B ut it seems impor-tant to emphasize that history in the sense in which the historian is concernedwith it is not the whole of the past, but some abstractions from it into which onlysome facts about the past enter, namely those we find interesting or importantor those that we have to refer to account fo r those facts that w e deem interesting


12/410

INTRODUCTION xior important. To do justice to this it seems preferable to restrict the notion ofahistorical fact to those facts aboutth e past that enter into a history. To under-stand the historical fact that someone took a certain philosophical view is to beable to explain it in the way in which one explainshistorical facts.Now if the historical fact is the fact that a certain agent performed a certainaction we try to explain it in the way in w hichw e normally try to explain whysomeonedid som ething.We first askourselves w hether theagenthad good rea-son to do whathe did and if we see thathe did we think we have understoodhis action. By good reason I mean here and throughout what weourselveswould regard as good reason. It is of course quite true that the agent in ques-tion may have a different view ofwhat constitutes a good reason and may acton whathe considers to be good reasons. But in this case his behavior wouldnot be readily and imm ediately intelligible to us precisely because w e wouldfirst have to realize that he acted on a different conception of what constitutesa good reason and then we would have to understand why he had this differentconception. In the end we have no alternative but to understand w hat othersdidor thought in terms of our notion of what constitutes a good reason though intrying to und erstand others we may come to realize thatit is our notion ofwhatcounts as a good reason that needs to be changed and that stands in the way ofunderstanding them. But we may also come to the conclusion that the personeven given his ownconception, did not have agood reason to do whathe did.And in this case we have to try to find a more complicated explanation that w illexplain why the agent did what he did though he had no good reason for it.Now w ha t is true of action seems also to be true of taking a philosophical v iew .Ifw e thin k thataphilosopherhad a good reason to adopta certain view wethinkweunderstand why he held this view. It may take us some time to find out thathe had a good reason. It may be that the reason we do not readily understandthethoughtof a philosopher is thatat first we fail to see thathe in fact did havea good reason to adopt his view; it may take us some time to change our ownviews and possibly even our notion of what constitutes a good reason before wecanrealizethathe had agoodreasonforholding his views. One reason we studythe though t of great philosophers w ith such care w ould seem to be precisely thisthat wetrust that in many cases they had good reason to say w hat they did al-though because of limitations in our understanding we do not readily under-stand it. These limitations are one of the things we hope to remove by studyingthe greatphilosophersof thepast. W e may ofcourse, in some cases come tothe conclusion that the philosopher after all had no good reason to adopt theview in question.It is perhaps w orth pointingoutthatit often is noteasy to come to this con-clusion. For to claim that someone did not have a good reason to think w hat hedid is to claim that it is not owing to our lack of understanding that we find itdifficult to understand why the person held thisview-a claim not easily made


13/410

xii INTRODUCTION

in the case of philosophers whose peer power of intellect and depth of insightgenerally fa r exceedsour own. Still, we may feel sufficiently confident that thephilosopher had no good reason to think what he did. In this case wethink wehave to look for a more complex explanation of why he took the view, in spiteof the fact that he had no good reason to do so. If this is correct it is apparentthat a full historical understanding of the fact that somebody held a certain viewwillalway s involve a philosophical understand ing of the view itself. For how isone to judge whether someone had a good reason to hold a view, unless one hasa philosophical understanding of that view by virtue of which one knows whatit is to have a good reason for holding it? Even if the philosopher did not havegood reason fo r holding the view, the explanation of why he held it will have tomakesomereference to the fact that it was not for a good reason that he adoptedit. It is an ex planation of why the philosopher held the view, in spite of the factthathe had nogood reason fordoing so. For instance, merely tocite the bad rea-son he had will not satisfactorily explain the fact that he held the belief, thoughitwas, in fact, fo r this reason that h e held i t. We wo uld still not understand whyhe held the belief for this reason, unless something were added that made usunderstand why he held the belief, though his reasons for doing so were bad.

But even in the casewhere we have come to the conclusion that the philoso-pher heldhis view for no good reason, there are two quite different kinds of ex-planation that m ight account for the fact that he held it. In one kind of case wecan explain why the philosopher held the view he did by providing him with aset of assumptions and a line of reasoning such that we can understand howsomeone who made these assumptions and argued in this way could think thatthe inadequate reasons he offered for adopting his view did constitute a good rea-son to do so. W e wo uld not share these assum ptions or we w ould find fault withthe argument, or both, but we might be able to understand how even one of usmight make these assumptions or use an argumen t of this kind. We m ight, e.g.,decide that the author had fallen victim to a simple fallacy, the kind of fallacywe can see ourselves committing, and this might explain why he thought whatin fact are bad reasons to constitute a good reason to adopt the view. Hence wehave an explan ation for why he adopted the view , though he had no good reasonfor doing so .

In another kind of case, though, no ex planation of this kind m ay be available.However hard we try, there is no set of assump tions and no line of philosophicalargument that we could easily see ourselves adopting and that would explainwhy the philosopher thought his bad reasons good reasons. It is in these casesthat we think we have to appeal to some historical context from which we canexplain why the philosopher held the view. Thus, we might discover that all ofthe philosopher s contemporaries made certain assum ptions, w hich , althoughnone of us would make them, readily explain why the philosopher in questiontook hisreasonsto be good reasons to adopt his view.


14/410

INTRODUCTION xiii

Now all the explanations we have considered so far are explanations of ahistorical fact, and in that sense one might call all these explanations historicalexplanations. But these explanations are of two radically different kinds inthatonlythelastkindofexplanation triestoexplainthehistoricalfactfromitshistor-ical context. It may be useful, then, to make a distinction between these twokinds of explanation by reservin g the term historical explanation for the kindof explanation that must appeal to a particular historical context to explain thefact that someone held a certain philosophical view.

How, then, do we explain historically the historical fact that someone helda certain philosophical view, if he had no good fo r reason fo r doing so, and ifwecannotfindsome lineo freasoning andcertain assumptions thatwe caneasilyimagine ourselves using?Weconsiderthehistorical contextof thethoughtto seewhether there is some history that will help explain why someone, given hishistorical situation, would come to hold this view.

But at this point it is, perhaps, worthwhile to note the fact that it does notfollow from thefact thatsomeoneheldaphilosophicalview whichhas to be ex-plained historically that it has to be explained in terms of the history of philoso-phy,by thehistoriano f philosophy. Perhaps we canavoid some confusion if wedistinguishbetween ancient philosophy, or quite generally the philosophy of thepast, on the one hand, and the study of this philosophy, on the other. Thereisanobject, ancientphilosophy and this object allowsfor acertainkindof study.Often one uses the expression the history of ancien t philosophy to refer to theobjectas awhole,bu t toavoidconfusion we mayprefertoreservetheterm his-tory of philosophy for a certain kind of studyof this object and for the aspectof the object that is studied this way, namely the kind of study that tries to dophilosophical justice to ancient philosophy.The reason I think it isuseful to make th is distinction is this: it is not the taskof the historian of philosophy to explain whatever philosophical view someonemay have had, even if it is a historical fact, i.e., a fact of some significance,thata certain person held this view . Nor is it the task of the historian of philoso-phyto findsomeexplanation orother forsuchahistoricalfact.It is rather histask to find a certain kind of explanation for the view in question, namely thekindofexplanation thatisappropriate for thehistoryofphilosophy , rather than,say, the history of morals. Thus it may be a historical fact of great significancethat a certain politician held certain philosophical view s, and thisfactm ay admitonlyof ahistoricalexplanation.Butthismay be afactof nosignificance for thehistory of philosophy. The thought may not be remarkable as a philosophicalthought,it mayshed nolighton thethoughtof earlier philosophers, and it maybe of no help in understanding the thought of later philosophers. It may evenbe that it is an important historical fact that a philosopher held certainphilosophical views,but this in itself doesnotguaranteehim aplace in thehis-tory of philosophy, since the only reason his view s were so important may have


15/410

xiv INTRODUCTIONbeen that he was the friend of an important politican whose politics were verymuch influenced by his philosophical views.It is easy to see that some philosophical thoughts do not enter the history ofphilosop hy because they lack historical s ignificance . It is also easy enough to seethat some philosophical thoughts d o notenter the history of philosophy becausethey are of no significance for this history. It is not so easy to say positively thata philosophical thought is to be considered a part of the history of philosophy.Ultimately this will depend on the conception one has of the history of philoso-phy. But it does seem safe to say that we want those philosophical thoughts tobe part of the history of philosophy that had a considerable philosophical in -fluence on later philosophical thought. A thought m ay have philosophical in -fluence on later thought in any number of ways : it may make the philosophicalproblem at issue appear different it may suggest other views one could take onthisproblem it may open up new ways to argue for a given view it may revealthe limitations of a line of a rgum ent that had been accepted thus fa r. If a gooddeal of later philosophical thought can be seen to depend on some earlierphilosophical thoughtin this wa y the earlier thought no d oubt forms part of thehistory of philosophy.A nd the more th e thou ghts that are influenced by earlierthoughtin turn are ph ilosophically influential the clearer it w ill be that the origi-nal thought should be part of a history of philosophy.Now to say that a philosophical thought has been philosophically influentialis to say that there are philosophical thoughts that somehow depend on it that insome way have to be explained in terms of it. But a thought may depend on anearlier thought in several w ays . The simplest case wo uld seem to be one in w hicha later philosopher ad opts a view for a good reason but the view and the reasonare sufficiently complex so that one assumes that his taking this view for this rea-son wasfa cilitated or even made possible by the fact that an earlier philosopherhad taken this view for this reason. M ore complex cases are those in w hich a laterphilosopher adopts a view for reasons that do no t constitute good reasons becausehe has convinced him self that some earlier philosopher who adopted the view forthese reasons had good reasons to adopt it or mo re generally cases in which alater philosopher adopts a view for reasons that do not constitute good reasonsbecause he has been persuaded by the thought of some earlier philosopher thatwhathe regards as reasonsto adopt the view a re good reasons. Almost all philo-sophical thoug ht depends on earlier thoug ht in this w ay . Wh at this reflectsis sim-ply the fact that we always do philosophy against the background of the philo-sophical views and the philosophical reasoning of at least our immediatepredecessors that we canno t at least to begin w ith see the problem s except intermsof the views and the reasons of our pred ecessors and that however m uchwe free ourselves from their views and reasons there w ill alw ays be some depen-denceon them. And in general even in the case o f highly original p hilosophers


16/410

INTRODUCTION xv

this dependence seems to be overwhelming. If early modern philosophy seemsor even at times pretends to stand on its own feet, it can do soonly as long asweknow very little aboutthehistoryofH ellenisticandlate Medieval philosophy.So what the history of philosophy in the narrow er sense seems to be made up ofare those philosophical thoughts which are influential in this way.Nor is it the task of the historian of philosophy to find some explanation orother for the philosophical thoughts that enter the history of philosophy. Thehistorian of philosophy will, rather, go on the assumption that philosophicalviews are usually set forth for philosophical reasons. He recognizes that some-times philosophical viewsare put forth byphilosophersw ho arequite aware thattheydo not have agood reason tohold them,but the historian ofphilosophynevertheless, and often rightly, thinks that it wouldbe worthwhile to considerthese views. More may be gained by this thanby considering uninteresting orboring views for which excellent reasons have been offered. But the paradigmis that of a philosopher who adopts a v iew because he think s he has a good rea-son to do so. The historian of philosophy will try to identify the reasons forwhichheadopts theviewandwillseewhether they constituteagood reason fordoing so. Failing this, he will see whether he can reconstruct some line ofreasoning that would makei tintelligblewhy thephilosopher thoughthis reasonsconstituted good reasons and hence adopted the view, a philosophical line ofreasoning that evenone of us might still avail himself of. Only if thisalsofailswill thehistorianofphilosophy resort to ahistorical explanation intermsof thehistory of philosophy. But he will still insist that it is because the philosopherhad reasons forholdingacertain viewandthat there mu stbe some philosophicalconsiderations that will explain why thephilosopher inquestion took these rea-sons to be adequatereasons exceptthatnowthese philosophical considerationsare dated; only someone in thehistorical situationof thephilosopher inquestioncould avail himselfof such considerations. They are thekindsof considerationswe would expect someone who is dependent on the thoughts of those predeces-sors to take seriously. W e ourselves can imag ine that if we were in those circum-stancestherewouldbenothing remarkable noteworthy, surprising, orastonish-ing, if we examined these considerations and concluded that the reasons we hadfor the view in question constituted good reasons to adopt it. It is at this pointin particular thatthehistorianofphilosophy will havetodisplayall hishistoricallearning and his philosophical in gen uity . For he will have i) to try to reconstructsome philosophical line of reasoning that would explain why the au thor in ques-tion thoug ht his reasons for holding the belief adequate, and ii) to make a casefor saying thatitwas, indeed, because ofsuchalineofreasoning thattheauthorthoughthis reasons adequate. To do the first often requires much philosophicalresoucefulness; to do thesecond requires a firmgrasponw hat kindofreasoning,which kinds of philosophical considerations were available at the time.


17/410

xvi INTRODUCTION

Nevertheless how ever succe ssful we m ay be in reconstructing a line ofreasoning that we can imag ine ourselves espou sing in this historical context andthat we have reason to think the philosopher adopted or at least might haveadopted it will still be a flawed line of reasoning. It must rely on assumptionsthat not only are un w arran ted but that one can plausibly make only in such ahistorical context. Or it will rely on a mode of reasoning that is inconclusiveand that could be found acceptable only in such a historical context. And wemust be able to identify these flaws or mistakes. For we do want to say that theauthor came to hold his view because he made these mistakes and that it wasbecause of these mistakes understandable as they may be that he thought thathis reasons for holding his view were adequate.

Often tho ug h not even this kind of explanation is available to us. For how -ever hard we t ry we are not able to find a set of philosophical considerationswe ourselves m ight have used in this historical situation on purely philosophicalgrounds. Even given th e thought of the relevant predecessors we cannot seeourselves making these assumptions or finding these arguments acceptable. Inpurely philosophical terms and in terms of the history of philosophy in the nar-row sense there is something remarkable notew orthy surp rising astonishingabout the flaws and the mistakes that led the philosopher to take his reasons tobe good reasons for his vie w . It is at this point that we have to look for a histori-cal explanation outside the history of philoso phy an explanation in terms ofsome other historical context some other history. Thus we m ight conclude thatthe only way to und erstand w hy the philosopher came to avail himself of a cer-tain line of reasoning is by assuming that he found it difficult to avail himselfof certain lines of reasoning that would have been preferable on philosophicalgrounds because of his religious convictions the religious convictions of thetime and because of the way in w hich such convictions were encouraged andconflicting views were discouraged.One may note first that in actual practice it is quite difficult to determinein a particular case how far one should go in tryin g to provide a philosopher w itha line of reasoning that is intelligble at least in the light of the history of philoso-ph y and when one should just give up and look at an explanation in terms ofsome other history. Na turally eno ugh historians of philosophy try to take thephilosophers of the past seriously as philosophers and hence go as far as theypossibly can to explain their thought in terms of purely philosophical consider-ations.Second ly we may assum e that the selectivity with w hich the historian of phi-losophy deals with the philosophy of the past results in much philosophicalthought that stands in need of a historical explanation in terms of some historyother than the history of philosophy being dropped from consideration.Philosophers who adopt philosophical views for reasons that could not makemuch philosophic l sense even to their contempor ries tend to have little


18/410

INTRODUCTION xvi iphilosophical influence and hence to disappear from th e history of philosophy.It seem s that th e philosophers who play a crucial role in the history of philosophyare in general those whose thought we can explain without having to re fer tosomeother h istory. But howeve r narrowly we conceive of the history of philoso-phy, it will still be the case that some of the thought it deals with will have tobe understood in terms of some other history.

So though the historian of philosophy usually explains those philosophicalviews of the past that enter into the history of philosophy in term s of philosophi-calconsiderations, it is obvious for the reasons given above that this will oftennot suffice to unde rstand th e fact that a philosopher took a certain view, becauseit will not suffice to explain the mistakes he made. And un less these m istakesare trivial because they are the kinds of mistakes any of us occasionally make,they need an explanation in terms of some other history. We might, e.g., thinkthat the fact that a philosopher availed himself of a certain line of reasoningcould b e und erstood on ly in term s of som ething in the history of his life that sug-gested this line of reasoning to him, which made it tempting for him to thinkof a particular matter in a certain wa y, which m ade it difficult for him to thinkof it otherw ise. W e m ight com e to the conclusion that the fact that a philosopheravailed himself of a certain line of reasoning had to be understood in t erm s ofth e history of the social structure of his society, which made it very difficult forhim to think of certain matters other than as he did. W e m a y suspect that th ereason he was inclined toward a certain line of philosophical reasoning hassomething to do with the history of religion an d that this will also explain whyit was rather difficult to adopt certain lines of philosophical reasoning, thoughon purely philosophical grounds they may have seemed preferable even then.Ne ither last nor least, it m ight occur to one that the puru sit of philosophy is alsoa social institution, with its history in t erms of which we can explain that stu-dents have views resembling the views of their teachers, and that at t imes itwould have been quite difficult to have views different from th e v iews of one steachersor one s school. There are any number of ways in which some historyother than the history of philosophy m ay in terfere with th e thought o f a philoso-pher in such a way that it no longer is intelligible just on philosophical grou nd s,not even on the philosophical grounds available at that point in the history ofphilosophy.

Now, though I think that one should conceive of the history of philosophyin this wa y, I also thin k that thinkin g of it in this way involves an enorm ous ab-straction and idealization. One goes on the assumption that, in gen eral, philoso-phers adopted certain views because they had certain philosophical reasons fordoing so. But , in fact, it seem s that philosophical views grow on one in a highlycomplex manner, of which our philosophical reasons and our philosophical con-siderations form only a part. We have seen that even in the case in which a phi-losopher has a good reason for adopting the view he does and no doubt holds


19/410

x v i i i INTRODUCTION

it for this reason w e m ay nevertheless think thathe depends for his view onsome earlierphilosopher from whom he has learned to see the matter correctlyand wi thout whom we m ight think hewould never come to hold th e right viewfor the right reasons; and this in tu rn is perfectly compatible with the furtherassumption that our philosopher given h is nonphilosophical e.g. m oral con-cern s could under these historica l circum stanc es e.g . these social conditionshardly fail to avail himself of this lineof reasoning and adopt the v iew in ques-tion. W e w ill never un derstand the origins of Greek philosophy by looking onlyat the philosophical considerations that led Thales A nax im ande r and An axi-menes to their philosophical views unless w e understand enough about thehistory of Greek society to understand why at this point this society neededsomething like philosophy and how this influenced the thoughts of the firstphilosophers. That philosophers hold their views for philosophical reasons isperfectly com patible with the assumption that there are ma ny other histories in-fluencing their thought. This is most apparent w hen their thoug ht gets derailedin such a way thatwe can no longer u nderstand it in terms ofpurely philosophi-calconsiderations. But thesame kindsof influences that reveal themselves in thiscase are also operative even when the philosopher adopts a view for purelyphilosophical reasons.

In fact one may become quite impressed by how firmly embedded thethought of philosophers is in the life of their societies and even in their ownl ives. I have been struck for a long time by how autobiographical as it werethe thought ofphilosophers is. It does not take much reflection to see that it isnot surprising that the topics philosophers concentrate on the general approachthey take to their topics the way they argue the way they set forth their viewsand often even the questions they consider are very much a reflection of theirlife and their personality. And it is no less surprising that the thought ofphilosophers should closely reflect the life the histo ry and the character of thesocieties they live in. One cannot understand why friendship plays such an im-portant role in ancient moral philosophy that A ristotle devotes tw o books of histhi s to it unless one understand s the enormo us role friendship played as a so-cial institution in classicalGreece One cannot understand w hyPlato and Aris-totle subordinate ethics to politics unlesso ne recognizes that the relation betweenthe individual and the political comm unity was v ery different in classical Greecefrom whatit is now and correspondingly thati t was conceived of rather differ-ently. It is difficult to understand on purely philosophical grounds w hy almostthe who le philosophy of late antiquity should be some form of Platonism; obvi-ously there is a connection between the dominance of Platonism and the newreligions that conquered th e Roman Empire. But itwould be a m istake to be soimpressed by this thought to think that the reasons philosophers offer for theirv iews or the philosophical considerations th e historian of philosophy attributesto them are m ere rationalizations of views that they in fact held fo r other rea-


20/410

INTRODUCTION xix

sons.To think this is to underrate the intellectual pow er ingenu ity resourceful-ness and honesty of certain philosophers who w ould have been ready to trans-form modify or ifnecessary giveup any of their views to arrive at a set ofbeliefs for which they could have produced satisfactory reasons, even thoughtheym ight have started out by trying to justify a view they w ere inclined tow ardon other ground s.And it is interm s o fthese reasons tha tw e have to try to under-stand their views unless we want to think that there is something misguidedabout the whole enterprise of philosophy that allows us to discount thephilosophers claim to hold philosophical viewsfo r philosophicalreasons. More-over we have to keep in mind that even if we came to believe that thephilosophical reasons given frequentlyare mere rationalizations they nonethe-less are reasons that have to be considered as such and that they m ight turn outto be perfectly good reasons in spite of the fact that they m ay have been es-poused fo r other reasons. What is more the way they influence this history ofphilosophy is not as rationalizations but as reasons as good or bad plausibleor implausible reasons. It is because of this that the history ofphilosophy triesto explain the viewso f philosophers as far as this is possible on purely philo-sophical grounds.But even if we think of the history of philosophy in this way wem ay fo rthe reasons given also w ant to insist that the thoughto fphilosophers is tied tovarious histories, several ofwhich m ay help to explain why a certain philoso-pher held a certain philosophical view even if it was for philosophical reasonsor even good philosophical reaso ns that he held it. W hat is m ore these historiesoften help to shape philosophical thought namely when its precise form andcontent is no longer determined by purely philosophical considerations. More-over, we have to keep in mind that philosophical thought itself helps to shapemany other histories.

Thus if we regard ancient philosophy an an object this object either as awhole or in part enters into m any histories. It is because of this that it can bepursued in many different ways all ofwhich have something to contribute toa fuller understanding of this object. To consider the philosophical thoughts ofancient philosophers only as such will provide one with a very partial under-standing ofancient philosoph y. The history ofphilosophy goes further than this.But it too does notprovide us w ith more than a n abstract general understand-ingofancient philosophy .To und erstand it as much as possible in its concretecomplex detail one has also to look at all the other histories to which it is tiedby an intricate web of causal connections which run both ways.Hence if Iw ere asked w hetherm y interest in an cient philosophy was prim ar-ily an interest in philosophy or an interest in the history ofphilosophy Iwouldsay neithe r since I am primarily interested in ancient philosophy itself as itturns up in the various histories into which itenters and in the way it actuallyenters these various histories.


21/410

xx INTRODUCTION

It is because I conceive of my interest in ancient philosophy in this w ay thatI have taken an interest in the whole history of ancient philosophy. For if, asI believe, a good deal of ancient philosophical thought cannot be und erstood interms of reasonswe migh t avail ourselves of, and if even w hat can be understoodin this way is more fully understood if we also understand it in term s of the his-tory of ancient philosophy, then an understanding of the history of ancient phi-losophy is crucial. But one does not arrive at a full understanding o f a historyby looking at just a few parts of it, especially if these parts are no t selected witha view to what is important in terms of this history, but, rather, in terms, e.g.,ofour current philosophical interests and tastes. One can not hope to understandth e history of ancient philosophy by looking at just it s beginnings, fo r obviouslyhow the history is to be constructed depends crucially on how it continues andhow it ends. This is just another way of saying that we can understand aphilosophical view only in terms of the history of philosophy, that is , only if wesee how it fitsinto this history as a whole, that is, if we understand not just whatleads up to it, but also how it leads up to w hat follows. If we try to understandAristotle s Ethics we are not only greatly helped by seeing it against the back-ground of Plato s moral philosophy, but also by considering w hat became of itas it was passed down in the Peripatetic school and by considering how Stoics,Epicureans, and Skeptics reacted to it and transformed it.Hence, a large part of my work has been devoted to Hellenistic philosophy,in particular to the Stoics and the Skeptics, because until fairly recently we hadvery little understanding o f this part of ancient philosophy. One reason for thiswas that Hellenistic philosophers were regarded as second- or third-rate philoso-phers oflittleor no philosophicalinterest.A s we cometo have a better under-standing of them, we increasingly realize three things: (i) Hellenistic philoso-phers are extrem ely interesting philosophically, once w e do the tedious w ork ofthe historian to restore and reconstruct their actual view s, instead of jus t believ-ing what philosophers have been telling us about them since the beginning ofmodern times; (ii)W e will understand early modern philosophy from Descartesto K ant m uch better once w e fully realize how enormous th e debt of early mod-ern philosophy to Hellenistic philosophy is; (iii) Pre-Hellenistic ancient philoso-phy begins to appear in a different and better perspective. Hence, it is not sur-prising that the last ten years have seen an enormous increase in the interest inHellenistic philosophy.

I very much hope, though, that one will soon be able to say the same aboutthe philosophy o f late a ntiquity. The objection is that it is philosophically boring ,if not repellent. Again, the judgment is not based on careful study of the evi-dence, but on what has been commonly said about th e philosophy of late antiq-uity. It seems clear to me (i) that Plotinus is extremely interesting philosophi-cally, (ii) that we will never be able to understand medieval philosophy in itsvarious traditions (the traditions of Byzantine philosophy, Islamic philosophy,


22/410

INTRODU TION xx iand the Latin West) unless we understand th e philosophy of late antiquity, and(iii) that the philosophy of late antiquity sheds a greal deal of light on the historyof Hellenistic philosophy and classical philosophy. One can learn much morefrom Plotinus about Aristotle than from most modern accounts of the Stagirite.Thus, I am confident that the near future will bring us a renaissance of studiesof late ancient philosophy.

Because it seemed to me that one has a chance to understand the fact thatsomeone took a certain view o nly if one fully understands that view, I chose tostudy the ancient history of one subject-matter, namely logic. Since we nowseem to have a particularly clear understanding of the subject-matter, it is rela-tively easy for us to attain an un usually high level of understanding of v iews oflogic held in the past. And, indeed, the enormous advances logic has made inth e course of the last century have had the effect that we now h ave vastly betteraccounts of ancient logic. But this case also shows that a mere understanding ofth e subject-matter is not sufficient. To explain th e reasons for wh ich views w ereheld,we also have to know which lines of reasoning wereava ilable and whichnot. Modern accounts of ancient logic almost invariably suffer from anachro-nism, and often grossly so.Moreover, it seemed to me a good th ing to take particular view s or complexesofview s and to follow them through history, to see how they were interpretedand reinterpreted, what was made of them in which context. A treatise likeAristotle s ategories offers a unique opportunity to do this, since it is one ofthe two or three philosophical texts that have been studied con tinuously through-out the history of philosophy, there are commentaries on it from all periods, ithas had an enormous influence on the history of philosophy, and its contentswere diffused at all levels of learning through compendia. Hence, a good dealof m y w ork has centered on this treatise, in particular its metaphysics with itsdoctrine of substance.Given that I am interested in the way in which ancient philosophy fits intoth e life of an tiquity in general, I have not only tried to come to some und erstand -ing of ancient philosophy in terms of the history of ancient philosophy as awhole, but I have also taken an interest in some of the other histories in w hichparts of anc ient philosophy play an importan t role. In pa rticular I have been in-terested in the connections between philosophy and other branches of learning,e.g. , grammar, medicine, and rhetoric.

Grammar seemed to me to be a particularly interesting casefor the followingreason. In school I had great difficulty understanding traditional grammar,whether Greek, Latin, or German. Later I learned from modern linguists thattraditional grammar is utterly confused. Part of the reason for this confusion,though , seems to me not to ha ve been properly und erstood. Trad itional gramm arwas heavily influenced first by Stoic philosophy and later by Peripatetic philoso-phy . But, of course, the quite substantial philosophical assum ptions tha t explain


23/410

xxii INTRODUCTIONmany of the features of traditional grammar were no longer accepted under-stood or even acknowledged when the subjecthad gained a lifeof i ts own andwaspursued by scholarswho knew little m oreo fStoic philosophy tha n itsname.Hence crucial featuresof the theory were no longer understood by those whowere supposed to teach revise and expand thetheory. In thison e can see howfa r and in what disguise philosophical ideas can travel and what damage theycan do if they go unrecognized.An cient medicine is of particular interest because here w e have a case wh erethere is a close connection on different levels which goes in both directions.There is notjust a close connection between philosophical theory and medicaltheory ow ing to the fact that both philosophers and doctors are interested inphysiology and even pathology. There is also a connection between philosophi-ca lview s concerning thena tureo fhum an knowledge thesciencesand arts andthe way doctors conceive of their art. In fact doctors develop quite elaboratephilosophical theories concerning their expertise and expert knowledge ingeneral which in turn influence philosophers. Moreover one might a t leastthinkthat these philosophical view s thatare ofgreat concern fo r ancient doctorsmight significantly affect their medical practice. One can observe that they didbut it ism ore interesting to notice how principles based on philosophical con-siderations whichonem ight im agine wo uld lead to w ide divergencesin practicein the end are supplemented by further principles so thatd ifferences in medicalpractice between adherents of different schools were greatly reduced if notabolished.The case of m edicine is also of interest in this context becau se an cient doctorshad their own tradition of philosophical thou ght they as it w ere insisted ontheiro wn philosophy w hichis rich en ough to haveits own history closely inter-woven with the history of philosophy of the philosophers but not part of itrather parallel to it. To m ake things m ore complicated some ancient doctorslike Asclepiades of Bithy m a M enodotus Sextus Em piricus and Galen w erealso philosophers of sufficient stature to secure them selvesa place in the historyof philosophy. But it is the history of philosophy within medicine as it werewhichhasbeen m y particular concern. For na turally en oug h historians o fphi-losophy have not taken much interest in it and historians of m edicine equallynaturally have been reluctant to deal with philosophical matters.

So I have tried to study ancient philosophy in these v arious w ay s in the hopeof getting as complex an understanding as possible of its complex reality. Itseems to me that all these are perfectly good w ay s to study ancient philosophyto shed light on the subject. Som etim es though philosophers talk as if therewere only one way to study ancient philosophy and the philosophy of the pastin general. And sometimes they talk as if itwere notreally w orthwh iletostudyancientphilosophy and the philosophy of thepast in general obviously assum ingthat there is thisone way to study the philosophy of thepast but thatnotm uch


24/410

INTRODUCTION xxiiiprofit is to be gained by studying it this w ay. I am sure that no one really meansthis, but some comments on the matter may at least clarify my v iew.To start with, it is merely an institutional fact that ancient philosophy ismainly studied in p hilosophy departments by philosophers. There is no separateprofession of students of ancient philosophy. And this is all for the better, sinceto understand ancient philosophical thought one first of all has to understand itphilosophically. It does cause a noticeable problem, though, for classicists, an-cient historians, Roman lawyers, historians of medicine, historians of science,historians of theology, and a great many others, who may get the feeling thatthey are supposed to approach the subject in the way philosophers tend to ap-proach it , when they, in fact, have their own legitimate approaches to it. Forphilosophers naturally want to study ancient philosophy in such a way as to un-derstand it philosophically and to benefit philosophically f rom this understand-ing. It goes without saying, or rather apparently it does not, that this is noteveryone else s ultima te aim in studying ancient philosop hical tho ugh t. The onegreat history of ancient philosophy was written by a theologian, E. Zellerwhose primary interests in w rit ing this history were in theology and the historyof theology. But the philosophers are encouraged in their attitude toward thestudy of the philosophy of the past by a historical accident, namely the accidentthat ancient philosophy and the philosophy of the past in general came to be asubject o f research and teaching by philosophers for a certain reason. It seemsthat the philosophy of the past came to be studied and taught by philosophersat the end of the eighteenth and in the course of the nineteenth century to comp le-ment or supplement th e systematic study o f philosophy. It seems that it wasthoughtthat the great philosop hers could serve as models o f wha t it is to do phi-losop hy, that they h ad raised certain questions in an exemplary way and that theyhad formulated classical answers to them, from the s tudy o f which one couldgreatly benefit, even if one disagreed with their view s, because they were exem-plary even in their mistakes.This at t i tude toward the great philosophers of the past had, o f course, a longtradition. The tradition o f studying great philosophers as philosophical classicsgoes back to antiquity. At the end of the second and in the first century B.C.certain figures in the history of philosophy, primarily Plato and Aristotle, weresingled out as classical p hilosop hers, jus t as one singled out classical histo rians,classicalorators classicaldramatists, authors who were supposed to serveas amodel for , and in a way to define, a genre. Within another two centuries thestudy of philosophy was reduced to the study o f these classical ph ilosophers.Philosophy was taught by commenting on the texts o f these historical authors.Much historical and philological learning went into their study: reliable editionsfo r these authors had to be prepared, authentic w ritings had to be distinguishedfrom inauthentic ones numerous historical allusions in the text had to beclarified. To understand what Plato and Aristotle say, one often has to know that


25/410

xxiv INTRODUCTIONthey are addressing certain long-forgotten philosophical views. In short, m enlike A lexander ofA phrodisias and Porphyry were men of vast historical learn-ing, but there is no reason to suppose they had an interest in the history of philos-ophyassuch. They just learned wh atittook todetermine Plato'sand Aristotle'sthought and to understand it philosophically. For their purposes it was an acci-dent that Plato and Aristotle were figures of the past, historical figures.Obviously, this way of teaching and studying philosophy by studying classi-caltextswas no longer acceptable inmodern times.It hadbeen astrain alreadyinan tiquity, and in the M iddle A ges literary forms were developed that allowedone to formally comment on a text when, in fact, one was systematically ex-pounding one'sow n views, e .g. , the questiones commentary. Once one startedto study and to teach philosophy by settingforth systematicallyone 'sown viewsor byteachingfrom acontemporary textbook, another problemarose.Theviewsofearlierphilosophers m ayhave been outdated,but onecould notfail torealizethat there was a noticeable difference between Descartes, Leibniz, Locke,Hume, andoneself or the author of the textbook one used. There issom ethingstrange in thenotion that Baumgarten, Reimar, Cruse, Knutzen should take theplace of these earlier philosophers, and there is something incongruous in theidea that K ant should teach philosophy by com menting on these a uthors. In anycase, it is easy enough to see why it could be thought that the systematic studyof philosophy should be supplemented by a study of the great philosophers ofth e past as philosophical models, to be understood and appreciated asphilosophers. And this kind of study came to be called th e studyof the historyof philosophy, since it was after all a study of the philosophy of the past andsince it could involve some, or even a great deal of, historical learning, as wesaw above in the cases of Alexander and of Porphyry. Hence, it is only naturalthatphilosophers to the present day should thinkof the study of the history ofphilosophy in this way and decide that it should be abandoned if it no longerbenefited usphilosophically. After all, it was introduced for this purpose, andif it no longer serves it, it has lost its rationale.

But ,clearly, there is an equivocationhere.Thestudyof the history of philos-ophy as asubject as Idescri ed it , i.e., as asystematic historical discipline isquite adifferent enterprise from the study of the philosophy of the past as it hasbeen practiced by philosophers from the nineteenth century onward and con-tinues to bepracticed by them to the present day, though both come under th et it le of history of philosoph y. The historian of philosophy wan ts to unde rstandth ehistoryofphilosophy,and hewantstoexplain philosophical view sof thepastin terms of this history . H e is not, at least as such , concerned tofully appreciatehow past philosophers have managed to thinkor fail to think the way w ethinkor the way one ought to th ink .

This switch in approach to past thought and the resulting equivocation mayhavebeen obscuredfor along timeby acertainconceptionof thehistoryofphi-


26/410

INTRODU TION xxv

losophy in the historian s sense of history. If one conceives of the history of phi-losophy as essentially a process in which certain questions that define th ephilosophical enterprise are seen and understood ever more clearly and in wh ichthe answers to these questions become moreand more apparent if it is perhapseven assumed that there is some mechanism or force that guarantees this kindof progress and in terms of which the history of philosophy, therefore, has tobe understood, the two appraoches to the philosophical past might easily seemto coincide. For now the philosophical classics w ill serve to sho w a certain un -derstanding of the philosophical problems, but also the limitations of this pastunderstanding and the necessity to overcome these limitations by the progresslater thinkers have made. This seems to be the spirit in which the first detailedhistories of philosophy were written at the end of the eighteenth century. But,surely,i t was a m istake to think that th e proper way to understand and to explainAristotle s thou ght was to see it as a crucial step f orw ard in the d irection of K ant-ianism, or some other philosophical view. The nineteenth century abounded inviews that explained why philosophy, along with the culture o f which it is a part,was set on a steady path of progress, in which the steps could be understood,almostideologically in terms of the position they led up to. But if the historyof philosophy is as much a history of failure where success was possible, as ofachievement where failure was possible or almost guaranteed, what reason dowehave to think that there is something that guarantees philosophical progresssuch that we have to understand the history of philosophy in terms of it?Thus, it seems to me that there is no reason to suppose that the study of thegreat philosophers of the past as models of philosophical thought and the studyof the history of philosophy in the historian s sense will somehow amount to thesame thing. And, hence, I think that the question whether the study of the greatphilosophers of the past as philosophical models is philosophically profitable isquite different from th e question w hether th e s tudy of the history of philosophyis philosophically profitable. In both cases I find it difficult to believe that th eanswer should not be positive. It is difficult not to see, even without any histori-cal learning, that K ant is a much better philosopher than the fam ous Kruse, notto mention any of our contemporaries, and that much is to be learned from thecomplexityof his thought. It is equally difficult to see how one would not benefitphilosophically when, in doing the history of philosophy, one tries to find asgood a philosophical reason as possible to take the most d iverse, if not perverse,philosophical views. W hat better w ay could there be to expand one s repertoireof philosophical lines of reasoning than to find one for almost any conceivablephilosophical position? What better way is there to learn to see things in fun-damentally different ways and to appreciate th e merits and the defects of thedifferent positions one could take?

In all this it should not be forgotten, either, that the philosophical views o fcontemporary philosophers are as much a part of the history of philosophy in


27/410

xxvi INTRODUCTIONthe wider sense as the philosophical views of thepast. If then, by chance, weshould be interested notjust in some contemporary philosophical view, but alsoin the question why a certain philosopher holds it, we will try to get the kindofanswer we are looking for when we do the history ofphilosophy, or moregenerally, studythephilosophical thought of thepast. Onewould expect that theansweronegets is the kind ofanswer onegets throughout the history ofphiloso-phy, an answer that will show how much the view dependson earlier views,atleastaslikely as not there is no good reason toholdthe view, butthere arecon-siderations that allowone to understand why thephilosopher does think that hehas a good reason tohold thebelief. Onething, though, that will almost neverhappen isthat wecome to think that the philosophical considerations weattributeto theauthor aredated, are the kinds of considerations we would no longer availourselvesof,and, hence, have to be explained in terms of the history ofphiloso-phy.This must be part of the reason why some philosophers seem to think thatcontemporary philosophy does notdepend on its history. For one can, indeed,understand contemporary philosophical thought, at least in general, withoutreference to the history of philosophy, because thekinds of philosophical con-siderationscontemporary philosophers avail themselves of are the kinds of con-siderations in terms of which we can understand any philosophical view,whether presentorpast, without having tohave recourse to the history of philos-ophy. But,ofcourse, it does not follow from the factthat one can explain some-one shaving a philosophical view without recourse to thehistory of philosophythat itdoes notdepend on thehistory ofphilosophy. In fact it might depend soheavilyon it that in the future onewill nolonger beable tounderstand itexceptin terms of thehistory ofphilosophy. This is just obscured from us by the factthat wehave little idea which contemporary considerations in thefuturewillap-pear dated.

Now if one does not take theview that thehistory ofphilosophy by itsverynature is ahistory of increasingly rational and philosophically satisfactory an-swersto a set of pernennial problems, but, rather, a history of achievements andfailures where thefailures often had more influencethan theachievements, andif onebelieves that philosophical thought does heavily depend on the history ofphilosophy, there mightbesomething to be learned philosophically from thehis-tory of philosophy as described above. If we were able to get a good enoughgraspon the actual history of philosophy, we should be able to see ever moreclearlyhow our own philosophicalthought depends on the philosophicalfailuresof the past. As long as the history ofphilosophy is seen primarily as a seriesofachievements that did not go farenough and, hence, naturally invited furtherachievements that would take thematter astep further it seems that notmuchis to be learned philosophically from the realization that one s thought is indebtedtoone s predecessors.But it is exactly because thehistorian ofphilosophy triestotake thephilosophers of the past seriously asphilosophers that hemight come


28/410

INTRODUCTION xxviito theconclusion that thehistoryofphilosophy atcrucial junctureshasgone inth e wrong direction.

Ifthis were a fact itwould be difficult for us tosee. For wewould have tomake such judgments interms ofwhatwethinkof asgood reasons or at leastasconsiderationsin thelightofwhich someone might take somethingto begoodreasons. Given thattheseare matters conditionedby thehistory ofphilosophy,they arelikelyto beconditionedby thevery failureswewanttodiagnoseintheirterms. Obviously, this willbe a d ifficult task. For to theextent thatou rnotionsandassumption ofwhatisrationalandreasonableareconditionedby the historyofphilosophy, they will make that history appear rationalandreasonable, ahis-tory ofachievements rather than offailures.

Fortunately, the historian of philosophy has more to rely on than contem-porary philosophical views. His work, ideally, would have taught him newviews thatonecould take, new reasons for oragainst old views; he mayhavediscovered therewasgood reason forviews whichat firstseemed unreasonable.All this work mayhave substantially changed his notions and hisassumptionsof what constitutes good reason and ofwhatat least is reasonable. Hence, thehistorianofphilosophy might very wellbe in aposition todiagnose adevelop-mentin thehistoryofphilosophyas anaberration when, from thepointofviewof contemporary philosophy, this development seems entirely reasonable. Thedifficulty ofcourse, is that the historian ofphilosophy shouldbe able to per-suade philosophers that this is so onpurely philosophical grounds.

Butif onestudiesthephilosophyof thepast notjust as ahistorianofphiloso-phy but in all itsaspects, one has further resouces to fall back on. It may bethat atsome of thejunctures in thehistory ofphilosophy where thehistorianofphilosophy believes he has todiagnose afailure, thefailure may be the resultofthoughts which themselves are to beexplained ingood part intermsofsomeother history. One mayeven beable toshow that this other history interferedwiththe natural development of philosophical thoughtatthis point, howeverphilosophically reasonable this development may now seem to us.

Once one asked questions such as What isphilosophy? A way toanswerthis questionis to lookat the thoughtof thepast tostudyancient philosophye.g., in the way Ipropose to do notjust bystudying ancient philosophers asparadigms, nor byjust trying to fitthem into thehistory ofphilosophy, but bylookingat all thehistories inwhich they occur, to see bytheir example, ascon-cretely as possible, what itactually meansand amountstowhenonedoes philos-ophy. Onethingone canlearn from this isthat to be aphilosopher in antiquitywassomething rather different from what it istoday. There is nodoubt that the ives and iews of thePhilosophers of a Diogenes Laertius are bad history ofphilosophy, bu tperhaps theydocapture anaspectofancient philosophy thatthescholarly history ofphilosophy, given itsaims, passes over, bu tthat, nonethe-less is realand of interest.


29/410



30/410

l to


31/410



32/410

Observationson Perception inPlato sLaterDialogues

Ast, in hisLexicon Plantonicum gives the followingas thegeneral meaningofthe verb aisthanesthai in Plato: to sense, to perceive by a sense, and hencegenerally toperceive by the senses. This not only seems to me to bewrong,it also seemsto be seriously misleadingif onew antstoarriveat anunderstand-ing of what Plato has to say about perception. For it suggests that in generalwhen Plato uses the verb aisthanesthai, he is relying on a common notionofsense-perception, a notion wh ich Plato ju st tries to clarify. This suggestionseems natural enough. Surely, onewill say, the Greeks even before Plato musthave had a notion of sense-perception, and aisthanesthai m ust hav e been theverb they comm only used wh en they w antedtotalk aboutsense-perception.A ndyet itseems to methatone fails tounderstandwhatPlato istryingto do, inpar-ticular in the Theaetetus unless one understands that it is only Plato who in-troduces a clear notion of sense-perception, because he needs it for certainphilosophical purposes. What he has to say about perception has to be under-stood againstthebackgroundof theordinaryuse of theve rb aisthanesthai andagainstthebackgroundof thephilosophical intentionswithwhichPlato narrowsdown this common use so that itdoes come to have the meaning toperceiveby the senses.Though aisthanesthai presumably is formed from a root which signifies hearing, itsordinary use isqu ite general. It can be used in any case inwhichone perceives something by the senses andeven more generally in any case inwhich one becomes aware of something, notices something, realizes or evencomes to understand something, however this may come about. There will, ofcourse, be atendency to use the word incases inwhich it isparticularly clearthat somebody is becoming aware of something or noticing something, as op-posed tojust venturing aguess, makinga conjecture, learning ofsomething byhearsay. These will be cases of seeing, but then also cases of sense-perceptionquite generally. But the use of theverbis notrestricted tothese cases. It isused


33/410

4 PERCEPTION INPLATO'S LATER DIALOGUESwhenever someone becomes aware of something. And up to Plato's time, andoften far beyond it, there is no clear recognition that there are two radicallydifferent ways inwhich webecomeaware of something, one by way ofsense-perception and the other in some other w ay , e. g. , by a grasp of the m ind. Thu s,there is no reason to suppose that the verb aisthanesthai, strictly speaking,refers only to sense-perception, but is also used metaphorically in other cases.It, rather, seems that allcases of becoming awareofsomething are understoodand construed along the lines of the paradigm of seeing, exactly because onedoes not see a radicaldifference between the w ay the m ind grasps something andthe way the eyes see something. B oth are supposed to involve some contact w ithth eobject byvirtueofw hich, througham echanism unknow nto us, we becomeaware of it.But in addition to this very general use of the verb aisthanesthai, we findin Plato a second, narrower use of the term, e.g., in the Phaedo and in theRepublic. In this use the term is restricted to cases of awareness that somehowinvolve the body and that constitute an awareness of something corporeal. Buteven now it wou ld be rash to assum e that the verb mean s sense-perception.For inthesecases it is used almost interchangeabely with dokein and dox-azein, toseem and tobelieve. The realmofbelief, asopposed to the realmof knowledge, is the bodily w orld w ith which we are in bodily contact as a resultofw hich this world appears to us in a certain way,as aresult ofwhichw ehavecertain beliefs about it. There is no doxa, no belief about the ideas, becauseideas are not the kinds of things with w hich one could hav e the kind of contactthat gives rise to a belief or a perception. Bu t, jus t as it w ould be a mistake toinfer from this that doxa m eans sense-perception, so there also is no needto assume that aisthesis m eans sense-perception, thoug h standard cases of aisthesis will be cases of sense-perception.

It is also in the later dialogues that we clearly have an even narrower use ofaisthanesthai, inwhichit,indeed, doesmean toperceiveby thesenses. Andit is this third sense of aisthesis whose introd uctio n I w ant to discuss.

Unfortunately, our main evidence fo r this very narrow notionof aisthesisis contained in a passage of the Theaetetus 184-187, whose interpretation hasbecome highly controversial, sinceitinvolves basic claims aboutPlato's philoso-phy and his philosophical development.

In this passage Plato tries to show not only that perception is not identicalwith knowledge, but that no case of perception as such is a case of knowledge.The argument assumes that if we perceive something, a bodily sense-organ isaffected, and that through this change in the sense-organ a change is broughtabout in themind (186 Cff . ; 186 D). Wh a t theargument ,as Iwantto interpretit, mainly turns on is that if we have aclear andprecise notion ofperception,we seethat perception is apurely passive affection of themindandthatfo r that


34/410

PERCEPTION IN PLATO'S LATER DIALOGUES 5very reason itcannot constitute knowledge, since knowledge minimally involvestrue belief and since any belief involves an activity of the mind.

Ifthis is correct, then it would seem that Plato's point in introduc ing this verynarrow notion of perception is to un tang le the conflation of perception, appear-ance, belief, and knowledge with which the main discussion of the dialogue be-ginsin 151D f f There perception is firstidentified w ith knowledge in Theaete-tus' first definition of knowledge as perception, and perception gets quicklyidentified withappearance(152C 11),which then throughout this section of thedialogue istreated as if it were the same asbelief (cf.,e.g., 158 A 1with158A2 and185B 2).But,obviously,it isusefultodistinguishbetween thesecogni-tive states: to perceive is not the same as to believe (though in the middle dia-logues we had not paid much attention to the distinction); neither is the sameas to beappeared to, and toknowis yet a fourth thing.But it is notonly usefulto make these distinctions, as Plato tries to make them in the Theaetetusand theSophist (264 A-B).It isnecessary tomake these distinctionsif wewanttocom-bat a certain philosophical view that we first encounter in Protagoras, but that,inoneversionoranother, willlaterbeespousedbysome rhetoricians, Skeptics,and the so-called Empiricists, namely the view that the beliefs which we havearejustamatterof howthingsappearto us, howthey strikeus, ofwhatimpres-sion, given the contact we have with them, they leave on us. Plato and thephilosophical tradition that depends on him, on the other hand, think that weshould not rest con tent with how thin gs strike us, that we have to go beyond thatto find out how they really are, quite independently of how they appear to us.Theoppon ents, like Protagoras, questionordenythepossibility thatweevergetbeyond appearance, seeming, belief. And, hence, they doubto rdeny that thereis anypoint inreserving theterm knowledge for something thatgoes beyondbelief. It is in this context that I wan t to see the argu m ent of the Theaetetus andin particular the section from 184 to 187. Plato thinks thatour beliefs and ourknowledge abo ut the physical world involve a passiveaffection of the m ind, buthe also thinks that they go much beyond this passive affection. And he wantsto reserve the term aisthanesthai, or to perceive, for this passive element inour beliefs, which he was willing to grant the opponents. It is in this way thatthe term came to have the meaning ofsense-perception.

With this as a background let us turn to the details of the argument. The con-clusion thatperceptionandknowledge are twod ifferent things isdrawnin 186E9-10 on the basis of the argument in the preceding lines, 186 E 4ff. It is as-sumed that to know is to grasp the truth and that to grasp the truth is to graspbeing. But in perception we do not grasp being, hence we do not grasp truth.Therefore, to perceive is not to know. This argument has two crucial assump-tions: (i) tograsp the truthis to grasp being, and (ii)toperceive is not to graspbeing. It isdifficult to understand and to evaluate these assumptions, since we


35/410

6 PERCEPTION IN PLATO'S LATER DIALOGUESdo notknow what ismeantby tograspbeing. Thereis noargument for thefirstassumption thatcanshed lighton themeaningof thephrase.But thesecondpremise is supposed tohavebeenestablished by theargument that extends to186C 6. Hence,we canlookatthis argumentto seewhetheritgivesus acluetowhatismeant by to grasp being.

Now, if welook at theargument, it seems that the reason given for the as-sumption that inperception we do notgrasp being is that the mind considersquestions concerning thebeing ofsomethingby itself, rather thanbymeansofone of thesenses.This would suggest thatthemind graspsorgets holdofbeingintherelevant sense whenitmanagestosettlethequestion concerning thebeingofsomething whichit hasbeen consideringbyitself.This seemsto beconfirmedby the finalcomments on theargument in 187 A Iff. There Plato says thatwehave learned from theargumentatleast thatwehavetolook forknowledgenotinperception, but inwhattheminddoeswhenitconsiders questions concerningbeingby itself (187 A5-6), when itforms beliefs (187 A7-8). It isbecausewe are supposed to draw this moral from theargument that the dialogue pro-ceedstodiscussthesuggestion that knowledgeistrue belief(187 B4-6).It isinbelief thatw egrasp truth,if thebelief istrue, though,as thefurtherargumentwillshow, thisis not yet a sufficient condition forknowledge,since knowledgerequires that this truthbegrasped in aparticular way.But if it is in true belief that wegrasp truth, it isalso intrue belief thatwegrasp being. This suggests thatby grasping being Plato here meansnomorethanthatthemind informingatrue belief manages tosettle thequestionof thebeing of something correctly. And it iseasy to see howPlato could think this,givenhisviewsonbeing.For heassumes thatanybelief, explicitlyorimplicitly,is of theform A ifF, and hethinks that inassuming thatA is F oneattributesbeing bothto A and toF-ness. Toassume that Socrates isjust is, onthis view,to attribute being to Socrates and to justice. Hence, any true belief will pre-suppose thatone has correctly settled questions concerningthebeingof some-thing.

Onemay,ofcourse, think thatby grasping being Plato here means some-thingmuch stronger than settlingthequestion whether being shouldbeattributedto something in this way. One may think that Plato wants to distinguish twokinds ofgrasps or intutions, a perceptual graspor intuitionand an intellectualgrasp or intuition. Thus, one may think that Plato, having distinguished twokindsoffeatures, perceptualfeaturesandnonperceptualorintelligible features,wantstoclaim that knowledge involvestheintellectual graspofintelligible fea-tures andhence that perception will never giveusknowledge. Buteven ifthisshouldbePlato'sview, this is not the way heargues inthis passage. Instead ofdistinguishing tw okindsoffeaturesandcorrespondingly tw okindsofgraspsorintuitions, hedistinguishestwokindsoffeaturesandcorrespondingly tw okindsofquestionsthemindconsidersandtriestosettle (cf. 185E 6fF.). IfF-nessis


36/410

PERCEPTION INPLATO'S LATER DIALOGUES 7

aperceptual feature, then, whenthemindconsidersthequestion whethersome-thingis F, itdrawson thetestimonyof thesenses (cf. 185 B10-12).IfF-nessis a nonperceptual feature like being, then the mind considers the questionwhethersomethingis F by itself.What little Platohas to sayabouthow themindgoes about doing this makes noreference to some intellectual grasp. Plato isreferring to comparisonsand to reasoningsthemind goes throughtocometoajudgment (186 A lOff . ; 186 B8ff.; 186 C 2ff . , thekindsof thingsthe minddoes whenittriestodecideamatter.And thefactthat Platois 1 87 A5ff. charac-terizes whatthemind does whenitconsiders questonsby itself as doxazein,i.e.,ascomingtoform abelief, certainly should warnusagainst assuming thatsome special powerof themindtograsp intelligible entitiesisappealedtohere.Allthat seemsto beappealed to is whatthe mindhas to beableto do to formbeliefs. Andthisis agreat deal, though Plato here doesnotcaretospellit outin anydetail. To be able to form the belief that A is F, the mind has tohavearrived atsome idea ofwhat it is to be for A andwhat it is to be for F-ness,orwhatit is to be for an F and it has to find outwhetherA is suchas to be anF. What Plato here wantstoemphasizeis themere fact thattheperception isa purely passiveaffection (cf. 186C 2 and186D 2), whereas the simplest beliefeven if itconcerns aperceptual feature, requires andpresupposes agreat dealof mental activity. And he infers from this thatsinceall this activity is neededtoarriveattruth, perception itself does notgiveus truthand, hence, cannotbeknowledge.

Now one maywanttointerpret theargumentof 184-187differently and ar-guethus:Plato distinguishestwokindsof questions, those themindsettlesbyitselfandthosethemind settlesbyrelyingon asense. Since therearequestionsthe mindhas tosettlebyitself, andsince,presumably, theanswertotheseques-tions can beknown,w ehere haveanargument which shows that knowledgeisnotto be identified with perception. But we do nothaveanargument,nor doesPlato intendtoargue, that perception never givesusknowledge.Afterall, therearequestionsforwhose solutionthemind relieson asense.Theanswertothesequestionsseemsto beprovidedbyperception.Itseemsto methat this interpreta-tion iswrong. Plato isquitecareful never to saythat some questionsaresettledbyperception or by a sense. Allquestions are settled by the mind, thoughfo rsome itdoes rely on perception. Thus,Itakeit that Plato wantsto argue thateventhequestion whetherA is red is notsettledbyperception.We may bepas-sively affected by the color red, but to form the belief that something is redpresupposes and takes a great deal of activity on the part of the mind. Hence,weperceive the color red, but we donot, strictly speaking, perceive thatA isred. Hence, knowledge, since it always involves belief, never isjust amatterofperception.

The only textual evidence thatseemsto stand in the way of this interpretationis the following. In 186 B 11-C 5 we aretold that whereas animalsand we as


37/410

8 PERCEPTION IN PLATO'S LATER DIALOGUESchildren perceive many things right from birth, there are other things that ittakesus along time, much trouble, and some educationto grasp. Surely, onewill say, to see that something is red does not take much troubleand a lot ofeducation.It issomethinganyinfantcando.But,it may beworth rememberingthat even the Stoics later will deny that children, properly speaking, perceivethatsomething isred. Forperception inthis wider sense presupposes astateofthe development of reason that allows us to articulate a visual impression intermsofconceptsandthatallowsus toacceptsuchanimpression astrue. Thus,eventhesimple judgment that somethingis redpresupposes some notionofwhatit is to be andsome notionofwhatit is to bered. Andthiswe do nothave rightfrom birth. Nor is itgiven to us byperception, butonly byreflection onwhatweperceive.Whatweperceive,strictlyspeaking, arejusttheproperobjectsofthe different senses,e.g.,colors in thecase ofsight (184 E 7ff . ) . Thus, strictlyspeaking, we do noteven perceive the object ofwhichwecome tobelieve thatit isred. And ifthis isso, it iseven more difficult to see how wecould besaidtoperceive that something isred, given this very narrow notion of perception.

N o w , Plato, in restricting perception to apassive affection of themindan dinemphasizing theactivity of the mind in forming beliefs, thinks ofbeliefs assomething we deliberately arrive at after a good deal of consideration andratiocination. AsPlato puts itlater in thedialogue (189E-190 A), belief is theresult of asilent discussiononeleads with oneself. In the ophist (263 Eff.) an din the hilebus 38C-E), we get a similar view of belief. Thus, belief iscon-ceivedof assomething thatisactively espousedon thebasisofsomeconscious,deliberate activity. This, nodoubt, is an idealization of how wecome tohavebeliefs. Formany beliefs wejustfindourselveswith,and intheir casethereisnoreason to suppose that weever went through aprocess of deliberation as aresult of which we espoused the belief. The Protagorean view, on the otherhand,an d theother views alluded to in thebeginning, whichare like it,assumethat beliefs normally are something wejust find ourselves with, which havegrown onus,whichw ehave just comeby bybeing struckbythingsin acertainway. Andthey try toassimilate allbeliefs towhat they take to be the normalcase. Hence, they emphasizethepassive elementinbelief-formation. Thus,onecan see whyPlato shouldbeinterestedinemphasizinghowsmallthepassive ele-mentinbelief-formationis. To do so, herestrictsthegeneral notionof percep-tiontosense-perception insuchanarrow sense and, moreover, tosuchanarrownotion of sense-perception that wecannot even an y longer be said to perceivethatsomethingisred. It isthis philosophical motivation that underlies Plato's in -troduction of anarrow use of aisthanesthai in thesense of sense-perception,a sense which the word did nothave ordinarily and which it did not have inPlato's earlier writings.


38/410

ristotl


39/410



40/410

2

TheTitle Unity andAuthenticityoftheAristotelian Categories

I IntroductionThe Ca tegor i e s ascribed to Aristotle, has played a unique rolein our tradition.It is the only philosophical treatise that has been the object of scholarly andphilosophical attention continuously since the first century B .C. , when peoplefirst began writing commentaries on classical philosophical texts. From earlylate antiquity until the early m odern period, one w ould begin the study of Aris-totle and the study of philosophy quite generally with the Ca t e go r i e s and Por-phyry sIsagog e . For several centuries, these tw o treaties, and the D e Interpreta-t ione formed the core of the philosophical corpus which w as still beingseriously studied. Thus, it is hardly surprising that our received view of Aris-totle whetherw e are aw are of this in all its details or no t was colored substan-tially by the Ca tegor i e s .Already in late antiq uity , how ever, doubts were raised about its authen ticity, 1though w e know of no ancient scholar w ho , on the basis of such doubts, declaredthe treatise to be spurious. On the contrary, Ammonius claims that everyoneagreed that it was authentic.2T he w riters of the Middle Ages and the scholasticsof the early m odern period seem to have had no doubt about the authen ticity ofthe treatise;3 presumably, they were relying mainly on the authority ofBoethius.4 It is tempting to suppose that this acceptance of the

Essays in Ancient Philosophy

Documents