Mind Association

Mind Association
Plato's `Parmenides' (II.) Author(s): Gilbert Ryle Source: Mind, New Series, Vol. 48, No. 191 (Jul., 1939), pp. 302-325 Published by: Oxford University Press on behalf of the Mind Association Stable URL: http://www.jstor.org/stable/2250440 . Accessed: 27/02/2014 14:52
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Oxford University Press and Mind Association are collaborating with JSTOR to digitize, preserve and extend access to Mind.
http://www.jstor.org
This content downloaded from 157.92.4.12 on Thu, 27 Feb 2014 14:52:10 PM All use subject to JSTOR Terms and Conditions
II.-PLATO'S
BY
' PARMENIDES'
GILBERT RYLE.
(II.).
I HAVEsaid that the Parmenideandialecticcontainsfourmain stages or operations whichI have labelled Al, A2, NI and N2. Each of these containstwo movements. Let us call theseMl as Al (M2) to a given movement and M2, so that we can refer or N2 (Ml), (' M' formovement). The references to themare as follQws: Al (Ml) Al (M2) A2 (Ml) A2 (M2) NI (Ml) NI (M2) N2 (Ml) N2 (M2) 137 c 4 142 b 1 157 b 6 159 b 2 160 b 5 163 b 7 164 b 5 165 e 2
withinone The generalrelationbetweenthe two movements operationis this,that while MI (say) proves that the subject underinvestigation, namelyUnity(or,in the othercases,what is predicates, otherthan Unity),possessesboth of two antithetical M2 provesthat that same subjectpossesses the othermovement neither of two antitheticalpredicates. Or rather, in each movement the label of whichis Ml, say, it is provedthat there bothofthemembers predicates are numerous pairsofantithetical the subject,whileM2 establishes of all whichpairs characterise that the subject is characterised by neitherof the members of theseseveralpairs of antithetical predicates. And in general in Ml are moreor less the same considered the predicate-couples M2. in the corresponding as the predicate-couples of the two movements in Actuallyin A2, NI and N2, the first each case provesthat the subjectpossessesboth of the members
PLATO S
PARMENIDES
303
of the pairs of antithetical predicates, while the second movement proves that it possessesneither; but in Al the orderis the otherway round,Ml provingthat it has neitherand M2 provingthat it has both.
A1 (Ml).
The firstmovementof the firstoperation,namelyAl (Ml), is (according to my interpretation) as follows: If Unity exists, it cannot be manifoldand therefore must be unitaryor single. It cannot therefore be a whole of parts. It will not therefore have outer or inner parts, and so it will lhaveno figure. It will have no location and no surroundings and so no change of position or stationarinessof position. Change and fixityof relationsare forbidden to it. It cannot be numerically different fromanythingor identicalwith anything: It cannot be identicalwith anythingelse or different from itself,for obvious reasons; and it' cannot be different *from anythingelse, because being different is different from being single,so that if it is single it cannot be that and be different fromanything. Equally it cannot be identicalwith anything, even itself. For unity is one thing and identityis another. [This seems a dubious step. Certainlyunity is not the same as eitheridentity or difference.But it does not seem to followthat it cannotenjoyidentity or difference, save on the assumptionthat unity is single and has no other properties than singleness. However, this point is now affirmed.]If Unityhas any otherattributes than that of beingunitary, then it is ipso facto shownto be several things, whichseveralness is inconsistent with its unitariness. Unity cannot be both unitary and anything else at all, even identical with itself. Since similarity and unlikenessare identityand difference of attributes, Unity cannot enjoy eithersimilarity or unlikeness, and so neitherequality nor inequalityof dimensions. So it cannot have equalityor inequalityof age withanything, and so cannot have an age at all, and is therefore not in time. Its existence therefore is existence at no date, and thisis nonexistenceat every date. It cannot,therefore, exist, and if it does not exist it cannot carry its alleged special propertyof being single,since therewould be nothingin existencefor the propertyto characterise. So Unity neither exists nor is it single. No name can be the name of it, no description the of it, and therecan be no knowledge, description opinionor per ceptionof it. It cannotbe talked or thought about (sincethere isn't any 'it '), whichis absurd.
304
GILBERT BYLE:
Comment. This, like all the other operations, smells highly artificial. There must be somethingwrong with the several deductions. We are inclinedto say that the starting-point was and to write off 'Unity exists' and 'Unity is illegitimate, unitary' as bogus sentences-the latterformakingan universal one of its own instances,the former for tackingthe verb 'to exist' on to what is supposed to be a logicallypropername. We may also suspectthat the argument presupposes that singleness is a quality,when it is nothingof the sort. Doubtless we are correct on all these scores-but how can the illegitimacy of be established? Not by primafacie unplausisuch procedures forthe Theoryof Formsdid seemplausibleand did entail bility, (1) that everyuniversalis single; (2) that everyabstractnoun is not only possibly but necessarilythe subject of a true affirmative and (3) that being single is a existence-sentence; case of having an attribute. The illegitimacy of the starting-point is establishedby the impossibility of the consequences that mustfollow if the original propositionsare taken to be both legitimateand true. We must not be superiorand appeal to sophisticated distinctions betweenformaland non-formal conceptsor to professionalised into 'categories' or 'types' of the various sorts classifications of logical terms; for the necessityof such distinctions and classifications to be shown. Plato is showing had first it, though it may well be that he could not formulate what it was that he was showing. Of necessity he lacked the languageof categories and types. That there are different formsof judgementand whattheirdifferences are couldhardlybe familiar at a timewhen theverynotionof 'judgement'had yetto receiveits introductory examination, e.g. in the Sophist. And little progresscould be made in the former of inference enquiryuntilprinciples became of specialisedresearch. the subject-matter We can say, gliblyenough, that qualitiesdo not have qualities and also that existenceand unity are not qualities. For we have been taught these lessons. But what firstmade it clear to whom that these lessons were true, unless some such ratiocinationsas these? To say that a termis of such and such a type or category is to say something about its 'logical behaviour', namely,about the entailmentsand compatibilities of the propositionsinto whichit enters. We can only show that termsare not of one type by exhibitingtheir logical misbehaviourwhen treated alike. And thisis what Plato is heredoing. To complainthat the severalconclusions are absurdis to miss
PLATO S
PARMENIDES'.
305
must the wholepoint. Plato means to provethat the premisses be illegitimatebecause the conclusionsare absurd. That is ad absurdumarguments, the sole and entireobject of reductio are. whichis what all thesearguments A1 (M2). is the longest ofthe first operation, This,the secondmovement tedious. Its object is to of them all. And it is insufferably of all the predicateprove that Unityhas both of the members couples, the lack of both of the membersof which had been in Al (MI). established ofexistence. it mustpartakein orbe an instance If Unityexists, So beingunitaryis one thing and being an existentis another. So the Unity to whichexistencebelongswill be a compoundof Unity and Existence, a compoundhaving those two parts or be unitary thesepartswillitself members. The wholecontaining willbe bothunitary and so also each ofits members and existent, and existentand thus will be anothercompoundof these two forever. So if Unity over again, and thiswill continue elements manifold. it mustbe an infinite has existence, must Next ' Unity' and ' Existence', not beingsynonymous, of difference things. So bothwillbe instances standfordifferent a thirdtermover and above whichis consequently or otherness, two. We can now speak of one couple consisting those original of Unity and of Unityand Existence,anothercouple consisting and a thirdof Existenceand Otherness. Otherness, of a couple are unitsboth of whichmust And the constituents be unitaryin orderto be instancesof unit. A couple plus the thirdunit will make threeobjects,and as couples are instances and threes of odd-ness,the Forms of Even-ness of even-ness, and Odd-nessare also now on our hands. And as multiplying consistsin, e.g., takingcouples threeat a time,or threestwice in thisway. All arithmetical at a time,we can get any number concepts are automaticallygenerated; fromthe existenceof unity the existenceof every numberfollows,i.e., an infinite of objects mustexist. Every numberyieldsan infinity nuimber with by its interlocking so Unity is fractionised of fractions, as therecould be arithmetical Existenceinto as many members number. i.e., an infinite fractions, Being a wholeof partsit must containits parts. Theremust by it. what is and what is not contained be a distinction between be finite, for all that So it must have limitsand consequently number ofpartswhichit contains. thereis an infinite 20
306
GILBERT RYLE:
and If it has limitsor boundariesit must have a beginning an end as well as a middle: and it must have a configuration jumps to the conor shape. [Parmenideshere unwarrantably Being a clusion that it must have a spatial configuration.] whole of parts,Unity cannotbe a part of any of its parts,nor be one can it be just one of its own parts. It cannottherefore it must of the thingsthat it itselfcontains. To be anywhere countable otherthanitself; yetsinceeverything be in something is among its parts,it must be containedin itself. This is supthatit must,qua self-containing, to imply invalidly, posed,I think else, be mobile. by something and, qua contained be immobile, Next, Unity, not standingto itselfas part to whole or as and not partially, fully withitself, wholeto part,mustbe identical and it must also be fullyand not partiallyotherthan wlhatever is not Unity. But the next stage seems very paradoxical. fromwhat is For it is to be arguedthat Unity is not,diflerent withitself. otherthanit and also is notidentical is not whereits contents are, sincetheyare inFor a container side it, whichit cannot be. Now Unityhas just been shownto itself and container, so it mustbe elsewhere thani be both content and so not be identicalwithitself. The oppositepoint,that Unity is identicalwith what is not Unity, is shown in this way. Othernesscannot characterise is 'itself and not anotherthing'. So foreverything anything, neitherUnity nor what is not Unity can possess otherness. And as what is not Unity cannot be eithera part of Unity or wholeofwhichUnityis a part,it is onlyleftforUnity an unitary pretends, and what is not Unityto be identical. [This argument ' is the name of a quality. for the moment,that ' otherness Of courseit isn't a quality-but whynot ?] Next, since Unity is otherthan what is not Unity,and vice otherversa,both Unityand what is not Unitymust exemplify ness. But in theirboth beinginstancesof the same attribute, theymust be similarin that respect. namelythat of otherness, For that is what similarity is, the possessionby two thingsof is the oppositeof otherness. the same character. Now identity that Unitymust But it has been shown,in an earlierargument, as be identicalwith what is not Unity [146-7]; consequently, of otherness, the possessionof identityis the non-possession between Unity there must be this respect of dissimilarity and what is not Unity. For by this argumenta suggested is not shared. It followsthat Unity is both shared property both to what is not Unityand to Unity similarand dissimilar itself.
PLATO S
PARMENIDES
307
I skip the detail of the next fewstages of the argument. It is argued that Unity must be both in and out of contact with itselfand with the 'field'; 1 that it must'be both equal and unequal to itselfand the 'field', that it must be greater and smallerthan itselfand the 'field' and also olderand younger and the 'field', and also be neither than itself of these. the end conclusion of Al (MI) it is shown Then,to controvert that Unity does exist at everytime and is thereto be named and described, knownand thought about. Finally,sincethe onlyway in whicha subjectcan be conceived is that it alters, both to have and to lack a given property having the propertyat one date and lacking it at another, it is argued that Unity changes,develops,decays, and moves as -wellas being immutable and static,and that the time of its and movingsmust be a time whichtakes no timechangings at whichtimeit is in neither of the conditions fromor to which its transitionis. (This looks like a variant of a Zenonian paradox about motion.) we feelthat most of the foregoing Comment.Naturally assertions,withthe arguments leadingto and from them,are absurd. Conceptsare being played with fast and loose. Those of one type,withone sortof logicalrole, are beingmade to understudy or deputise for othersof quite different sorts. Different concepts shouldnot be treatedas if the rulesof theirco-functioning were all similar. Precisely-but only absurdities reveal the different rules,and the reductio ad absurdum argument marshal the absurdities. A2 (Ml). Parmenidesnow enquires: From the assumption that Unity follow exists,what consequences about 'rada'a ? He will argue that this subject too must possess opposite predicates. What exactly does a'a&AAadenote? We have no reason to restrict it, for example,to the objects of sense or opinion; nor yet to the Forms other than Unity. It must be taken to cover all of whatever termswhatsoever, whichare otherthan Unity. sorts, as well as Alcibiades,the Equator as well as my So Circularity presentpang of pain,will be members of thisomnium gatherum. Let us just call it, in racingparlance,'tthefield'. The fieldis otherthan Unity,yet it embodiesit. For it has and so must be one aggregateor members, being a plurality, whole of those members. Moreover,each of those parts or
1Meaning ofall thatis distinguishable by thisthetotality from Unity.
308
GILBERT RYLE:
So sincethe fieldis not Unityit mustbe a plurality or manifold. which I skip,is developed And theargument, thatsucha plurality must be both a finiteand an infinite so each of its plurality, will be so too. members the field and its several Being both limitedand unlimited, are similarto one another,since theyall co-exemplify members limitednessand unlimitedness; yet since these are opposite what exemplifies one mustbe unlikewhat exemplifies predicates, the other,as what is black is unlikewhat is white. Similarly it could be shown,thoughit is not shown,that the fieldand its several membersmust enjoy both identityand otherness and both changeand changelessness, etc.
them is Unity. A thing is notthatofwhich it is an instance.
must be one part or member. A whole is a plurality members of units,so it is a unit-and each of themis a unit. But thoughor because theyexemplify it, it is not and noneof
A2 (M2).
Unity and the fieldare an exhaustivedisjunction; therecan be nothingwhich does not belong to the one camp or to the other. So therecan be no superior camp, to whichboth these as members. Hence Unity will have no camps are subordinate truckwith the field,eitherso as to constitute it one whole of parts,or as an assemblageof unitary parts. So the fieldcannot norwill any number be a plurality, be applicableto it, or to any ofit. So the fieldcannotpossesseither part or feature similarity or dissimilarity or both at once. For both together would be a pair and each by itselfwould be single,and theseare applications of number. For the same reason the field cannot be identicalor different, or mobile,cominginto or going stationarv out of existence, or smalleror equal. greater The conclusionof all the movements of both operationsAl and A2 is thus summedup. If Unity exists it both has every predicate and lacks every predicate,includingthat of unity. And the same holdsgood forthe fieldtoo.
Ni (Ml).
We nowturnto the consequences of the hypothesis that Unity does not exist. The propositionthat Unity does not exist in havinga different clearlydiffers subjectfrom the propositions that largenessor that smallnessdoes not exist. So we know what 'Unity' denotesand that it denotessomething otherthan what these othernouns denote,whether our judgementis that theredoes or that theredoes not exist such a thing. So Unity
PLATO S
PARMENIDES'.
309
is something whichwe apprehend, and it possessesand is known to possessthe attribute of being otherthan the termswhichwe have distinguished fromit. Consequently Unity,forall that it does not exist,is an instanceof variousthings. The word ' it' applies to it. Being distinguished, it has dissimilarities from what it is distinguished from, and as it is not so distinguishable fromitself,it must have the oppositeof dissimilarity, namely, to itself. [We may grumbleat this step. The insimilarity ' I am notunlikemyself, ference ' therefore I mustbe likemyself contains a fallacy. But what sort of fallacy? The inference is valid if I am comparedwith my father, so why does it not hold good in this case ? If we say 'because the termsto the relations oflikeness and unlikeness mustbe numerically different ', then we are asserting a very special sort of ' must'. Namely we are sayingthat ' I ' and ' like (or unlike)' are termswhich are of such formal constitutions that absurdity results from their juxtapositionin this way. And that is a discoveryabout the formal propertiesof certain sorts of terms. It shows that is nota quality. But thedistinction similarity ofquality-concepts and relation-concepts is a distinction between typesofconcepts.] Being unlikethe field, it cannotbe equal to it or its members; so it must be unequal to them. But inequalityis in respect of largeness and smallness(sincefortwo thingsto be unequal in size one mustbe relatively large and the otherrelatively small). So Unitypossesseslargeness and smallness would [the argument only prove that it must possess at least one of the two]; but as beingbig is the oppositeof beingsmall,Unitymust,by way of compromise, have what is betwixtand betweenthe two, i.e., equality with itself. [This is fallacious-but why?] Unity is an instanceof bigness, therefore smallness and equality. But if it has all these predicates,Unity must, thoughnonexistent,still enjoy being in existencein some fashion. For if the above descriptions were true,theydescribedit as being what it reallyis. Unitymust be thereforus to be able to say or thinkthat it does not exist. But also it must not be there, forits non-existence to be trulypredicatedof it. But hovering in this way betweenexistenceand non-existence is change,and is motion[this is illegitimate-butto see change or transition why it is illegitimate is to see something about the important and change.] non-existence conceptsof existence, Yet since it does not exist it cannot be anywhere or move anywhence anywhither. And the other sort of transition, fromstate to state,is also ruled out; for if unity changed in this way it would cease to be Unityand becomesomething else.
310
GILBERT RYLE:
and changeis to be stationary But to be exempt from movement and immutable. So Unity both is and is not mobile,and both is and is not mutable. And it also followsboth that it is and and annihilation. that it is not subjectto generation Comment.The interesting parts of this movementare the that that of which stages wherewe findthe famousargument in some sense,to it is true that it does not exist must be there, accept thisascription of non-existence and also to be distinguishable fromother terms,existentor non-existent.We are enlightened enoughto say (withKant) that ' exists' is not a predito verbs cate or (withlatter-day logicians)that the nominatives of existence or logically do not function as demonstratives proper names; but the penaltiesof not sayingso are here exhibited. Doubtless the rules governing the logical behaviourof verbs of existence are stillobscure to Plato; but that thereare suchrules, and that theyare different fromthose governing ordinary prefrom result dicates,is herebeingrealisedby him. For absurdities treating themalike. Plato seemsto be ahead of Meinonghere. Ni (M2). If Unity does not exist, it is lacking in all modes, departmentsor sortsof existence. It can enjoy neithercoming-to-be nor annihilation; it cannot be subject to mutationor motion, nor,beingnowhere, can it be stationary anywhere. Indeed, it can have no attributes or properties, neither largeness, smallness,nor equality, neithersimilarity nor difference. It cannot even be correlated with a field,forits having such a correlate would be a relationalproperty of it. It has no attributes, parts,relations, dates, and it is not thereto be known, thoughtor talked about, perceivedor named. There is no 'it' at all. Comment. It seems to follow from this that all negative existencepropositions must be nonsenseif they are true, since thereis nothing leftto supportthe negativepredicate. So the name ofthe subjectofpredication is the name ofnothing. From this it is a shortstep,whichPlato does not take (any morethan to verbsofexistence Meinong did), to seeingthatthe nominatives are notthe names of anything, a and 'exists' does not signify quality,relation, dimension or state,etc.
N2 (Ml).
If Unity does not exist, what predicatesattach consequentially to the field? Plainly the field must by definition be other, yetit cannotbe other thanUnity, sincethis,by hypothesis,
PLATO S
PARMENIDES'.
311
does not exist for the field to be demarcated against it. The fieldmustbe otherin the sense that its members are other than one another. Yet, since Unity does not exist, the membersof the field cannotbe unitary or be units; so the fieldcan onlybe a manifold of manifolds withoutend. Only of such manifolds can we say that theyare otherthan each other-since thereis nothing else to say it of. Each manifoldof manifolds will seem to be single,thoughnot reallybeing so. And numberswill seem to be applicable to them, though the seemingwill be illusory. Derivativelythe concepts of odd and even, greater,smaller and equal, limit and unlimitedness will appear to have application,together with those of unityand plurality, similarity and dissimilarity, etc., etc. Yet if unity does not exist, none of theseconceptscan reallyhave application to the field.
N2 (M2).
If Unitydoes not exist,the fieldcannot be single,nor can it be a plurality, else it would be one pluralityand its members would be units. Nor could the fieldseem to be eithersingle or a plurality. For since thereis no Unity,thereis nothing of the sortforthe fieldto exemplify or participate in in any respect whatsoever. So the fieldcannot be thought, even, to be single or pluralor to be an instanceof anything else, such as similarity or dissimilarity, identityor otherness,contact or separation, or anything else at all. The fieldcould not therefore be thought to exist. So ifUnitydoes not exist,nothing exists. So, whether Unityexistsor not,Unityand thefield bothhave and lack every predicateand its opposite. 'Very true' is the last word of the dialogue. What is the outcomeof all this tiresome ? chain of operations First,ad hominem it seems to have been proved,in the case of at least one extremely eminent Form,whatSocrates was reluctant to believe could be proved, that a Form does undergo hosts of incompatiblepredicates,and that these disagreeableconsequencesflownot onlyfrom the palatable hypothesis that that Form exists but also fromthe unpalatable hypothesis that it does not exist. But what does Plato thinkto be the important lesson of the wholedialogue? Here we can onlymake moreor less plausible conjectures. 1. Plato mightthink that the whole argumentproves that no universalcan be the subject of an attributive or relational
312
GILBERT
RYLE:
proposition; and he may have confusedwith this the quite different point that no universal can be the subject of an or negative existence-proposition. (For he may affirmative have thought wrongly,as Descartes and Meinong did, that ofthesamecategory, ' exists' is a predicate i.e., withthesamesort '.) Universals of logical behaviour,as ' is square' or ' is green or abstractnounsare not propernames,and are not substances, sentences in whichwe talk as if theywere are logicallyvicious. This conclusionis true, and relevantto the question of the truthof the Theoryof Forms. So it may be what Plato had in his mind. and us of a seemingly himself 2. But Plato may be apprising more parochial discovery,namely that some concepts do not behavein the same way as some others. He may, forexample,be makingthe discovery that 'exists' and ' does not exist' do not have the same sort of logical be' or ' is square '. If we haviour as 'breathes ' or ' resembles considerthe conceptswhich occur in our ordinary descriptions of things,they seem to fit reasonablywell and classifications into scales of generaand species. And we can imaginea table depictingall the ladders or pyramidsof genericand specific concept concepts, such that any descriptiveor classificatory in one and not more wouldhave its place fixedforit somewhere than one such ladder or pyramid. But thereare some concepts which can be peculiar to no one ladder or pyramidbut must somehowpervade them all. Such are the conceptsanswering like 'not', 'exists', 'same', ' other', 'is an into expressions stanceof', ' is a speciesof', ' single', ' plural' and manyothers. '. Some conceptsare ' syncategorematic which At first to take such concepts, sightwe may be tempted are obviouslyof very generalapplication,to be merelyhighly genericconcepts,perhaps actually SuimmaGenera. But if we do so take them,our enterprise collapses,forjust theseconcepts are again requiredwhen we attemptto describethe affiliations or non-affiliations between Summa Genera themselves,and also betweenthe sub-divisions, not of one but of all the sorthierarchies. Formal concepts, as we may now call them,differ from generic ones not in beinghigher than theyin the way in whichtheyare higherthan specificconcepts,but in some other way. They from differ not,forexample,as 'Even Number' generic concepts from' 2 ', but as ' + ' and ' /- ' differ differs from either. Or again, to pick up again the two analogieswhichPlato uses in the Theaetetus from and the Sophist,formalconceptsdiffer
PLATO S 'PARMENIDES
313
genericand specific conceptsnot as one letterof the alphabet fromanother or as one bunch of letters differs differs from anotherbunch of letters,but as the mode in whichlettersare arrangedinto a syllableor word differs fromthe letterswhich are so arranged: or' else as the way in which nouns, verbs, adjectives, etc., are combinedto forma significant sentence is different from those elementsor even fromthe way in which one such element,like a noun, differs from another,like a preposition. What a grammatical is to the words construction of a sentenceembodying that construction, that a formalconcept is to the terms(particulars and ordinary which universals) enterintothe proposition or judgement. Now whenwe treata formal conceptas ifit werea non-formal or proper concept, we are committing a breach of 'logical syntax'. But what shows us that we are doing this? The deductive derivationof absurditiesand contradictions shows it, and nothingelse can. Russell's proof that, in his codeb cannotbe a value ofx in the propositional symbolism, function is only another exercisein the same genreas Plato's proof Ox tbat 'Unity ' cannot go into the gap in the sentence-frame . . . exists' or ' . . . does not exist'.' I feel fairlysure that this is something like the point which Plato was trying to reveal in this dialogue. I feel this partly because the imputeddoctrineis true and important and partly because,so construed, the dialoguethen linkson directly to the later parts of the Theaetetus and to almost the whole of the Sophist. Whereas the firstinterpretation which I suggested has no echoesof importance in eitherdialogue. we know that Aristotle Moreover, was alive to the fact that therewas a special crux about,Unityand Existence; and also that these conceptswith some others(e.g., Good) did not come under any one of the Categoriesbut exhibitedthemselvesin all of the Categories: nor were they concepts of the genusspeciessort.2
l It is worthnoticing that the conceptof being-an-instance-of, about whichthe discussion in the first turned part of the dialogue,is in fact a form-concept, and not a properconcept; the contradictions and circles whichembarrassed Socratesdid arise fromhis attemptto treatit as if it was fromthe same basket with ordinary relations. However,Plato does not pointthis out. We can conjecture that the secondpart of the dialoguedoes contain(betweenthe lines) the answerto the problem of thefirst part; but we cannotsay thatPlato was awareofit. 2And cf. De Interpr., 16b, where Aristotleexplicitly says that 'is' and 'is not' onlyfunction significantly in the assertion of somesynthesis, and cannotbe thought excepttogether withwhat is combined in such a synthesis.
314
GILBERT RYLE:
And (in Met: 1003band 1053b)he uses forboth 'existence' and 'singleness' the argument whichHume and Kant used for attributes;namely 'existence', to showthattheydo not signify of a man, an existentman and a single that the descriptions of different sortsof men. man are not descriptions to the And lastly I am temptedto prefer this interpretation otheron the scorethat it does morecreditto Plato's powersof in logical questions. Thereis, indeed, the important discerning in that suggestedmessage of the an agreeable sweepingness the generalpoint dialogueaccording to whichPlato was proving that universalsare not subjects of qualities or relations. But for it would only be would only be sanitary, its sweepingness wrong the negativepoint that therewas something establishing of the theory of Forms. withthe foundations on thinkers who had effect It would have small instructive never adopted the beliefthat abstractnouns are the names of substances. It would leave open and, worse,it would leave alnost unthe profounder question,What is wrongwith those formulated ? This question requires the discovery of the foundations concepts-and this difference between formaland non-formal and not discoveryis requiredfor all sorts of logical problems, only this special historicalone of the nature of the fallacy of SubstantialForms. underlying the special doctrine of the dialogue One objectionto the foregoing interpretation is sure to be made. It is incredible, it will be said, that the centraldoctrineof Platonism,namely,that Circularity, Unity, to be Difference, etc., exist, should be shownby Plato himself logicallyvicious, even though he mitigatesthe crueltyof his exposure of his earlier childrenby showingthat there would be a precisely parallel viciousnessin the doctrinethat they do not exist. On minorpoints,doubtless,Plato's second thoughts on his firstthoughts, mightbe expected to be improvements but that he should overtly demonstratethe untenabilityof of the system the veryprinciples whichhis wholeinfluence from a supposition. derivesis too shocking upon subsequentthinking But such an objection does less than justice to a great philosopher. Kant is felicitatedfor being capable of being awoken from dogmatic slumbers; Aristotle is permittedto be fonderof truththan of Platonism; those of Russell's contributionsto logical theory are consideredimportantwhich to Kant, Bradley,and belongto the periodsafterhis affiliation the illuminaBosanquet. Why must Plato alone be forbidden tionsof self-criticism ?
PLATO 'S 'PARMENIDES
315
that in the wholeperiod it has long been recognised Moreover, the Sophist,the which includesthe writingof the Thecaetetus, if is not entirely, Politicus,and the Philebus,Plato's thinking of the Theoryof Forms. by the premisses at all, governed He attendsto the theoryon occasions,but he does so in a dispassionate and critical way. In the Sophist [246] the exponentsof the theoryof Formsare treatedin the same way the Eleatic S-tranger's can answer as are thematerialists; neither puzzles about existence and non-existence. Similarlyin the if it is true that the theoryof SubPhilebus[15]. Moreover, stantial Forms embodied radical fallacies,to praise Plato as withcrediting as we do, would be consistent a greatphilosopher, him both with the acumen to recogniseand the candour to expose them. is this fact. But more importantthan these considerations the Theory of Whatever its sublimityand inspiration-value, Forms had been fromthe start,interalia,' a doctrineintended to resolvecertainpuzzles of a purelylogical nature. How can severalthingsbe called by one name or be of one sort or character ? And how is it that only those systemsof propositions express certain knowledgewhich contain neither the names of actual instancesof sorts or characters nor the descriptions ? and philosophy -namely mathematics The Theory of Forms was intended to answer both these theory,forthe reason, questions. It fails to be a satisfactory that exactlyanalogousquestionsarise about Substantial mainly, Forms to those questionsabout the instancesof Forms which the theory had been intendedto resolve. And in so far it was the wrongsortof answer. remains. It remainstrue that every judgeBut something termor embodiesat least one non-singular ment or proposition of mathematics element. It remainstrue that the propositions are universalpropositions. And it remainstrue that in some sense, some or all philosophicalquestions are of the pattern to be so-and-so'? (where'being-so'What is it forsomething and-so' is an universal). The criticisms of the doctrineof SubstantialForms given in the dialogue have no tendencyto upset these positionseven yield an answerto the problemswhich if they do not directly they raise. But the road is cleared for an answer to them,a theory but erroneous road whichwas blockedby the fascinating whichthey dispose of. Nor could the new advances have been the stage gone through begun save by someonewho had himself of being at least very familiarwith the theoryof Substantial Forms.
316
GILBERT
RYLE:
the road is now clearedforthe I shall suggest,. In particular, advance whichwas partiallymade in the Sophist,whereforthe and the need of a tbeoryof categories time the possibility first betweengenericconcepts or types is realised.1 The distinction and logicalenquiries is hereseenor half-seen, concepts and formal are at last capable ofbeingbegun. In fine,on my theory,the Parmenidesis a discussionof a and most of the problem of logic-as part of the Theaetetus Sophistwere discussionsof problemsin logic. Not that Plato Epistemology says 'let us turnback fromEthics, Metaphysics, to theprovince somequestions belonging and Pbysicsand consider of Logic ', forthesetitlesdid not exist. in this dialogue sbould But his questionsand his arguments by us as belongingto the same sphere to which be classified belong, for example, Aristotle'stheory of Categories,Kant's concepts,Russell's theory fromnon-formal separationof formal and Carnap's theoriesof logical of types, and Wittgenstein's syntax. is a question the dialogueis interesting ifI am right, Whether, seemsto me ofradicalimportance problem oftaste. The central who to any philosopher potentially, interesting, and therefore cares to get down to the roots. But the detail of the argument must find everyone and so sustainedthat, is arid and formalistic it tedious-in the same way as the methodicaldissectionof ViciousCircleFallacies is tediousifit is thorough. I do not thinkthat the dialoguecould or shouldbe interesting anxious to know Plato's later to a studentwho is primarily or physics, viewsabout the humansoul, or God, or immortality, Monism. For, as I read it, the dialoguecontains or Parmenidean fromwhich conto such topics and no premisses no references clusionsabout thesetopicscan be deduced. The dialogue is an exercisein the grammarand not in the proseor the poetryof philosophy. of theoryabout the programme the foregoing To corroborate the Theaetetus about some remarks I the Parmenides, append and the Sophist,in which,I think,the same or kindredlines of thought are to be traced. These dialogues were certainly composed close to the date of the Parmenides. The Sophist, written was certainly whichis a sortof sequel to the Theaetetus, after the Parmenides,to which indeed it makes one or two it undoubtedallusions and of which,in an importantstretch, partly echoes and partly presupposesa part of the dialectical
1I use the word 'category' in a less misleadingly preciseway than Aristotle.
PLATO'S ' PAPRMENIDES'.
317
operations. The Theaetetus was almost certainly in part, and perhaps as a whole, composed after the Parmenides, and it to it. containswhatis probablya reference The Theaetetuss. Withthe mainproblemof this dialogueI have bereno special concern. It is an enquiryinto the nature of knowledge. It of the theory begins with a sustained expositionand criticism is to have sense-acquaintance with it that to know something of sense-acquaintance withit. It is soon shownthat or memory neitherthis theorynor a more generalisedanalogue to it can account for our knowledgeabout the future,or of the truth of theoriesabout what is rightor expedient, especially of the truthor falsehoodof this theoryof knowledge or even of itself, mathematical truths. And it is brieflyindicated that even withinthe fieldof the objects of sense-acquaintance it will not do. For to knowthat sense-given objects ecist or do not exist, single or plural is to do or experience are similaror different, more than merelyhaving sense-acquaintance. So a something new hypothesis is considered, the gap betweenwhich and the previous view is of the greatestimportancenot only for the but also for our special problem. For it theoryof knowledge, is now suggestedthat to know is to judge, or is a species of threadsfrom earlier judging. And thismeans-to bring together and later parts of the dialogue-that knowledgerequiresfor its expressionnot just a name but a sentence or statement. And what a sentenceor statementexpressesalways contains a plurality,at least a duality of distinguishable elementsor as well as trueand falsebeliefand opinion, factors. Knowledge, cannot be expressedjust by a propername or demonstrative for some simpleobject, but only by a complexof wordswhich a sentence. constitute together which at firstsight At this point Socrates does something the directpath in orderto followup a seems to be deserting side-track. For he suddenly opens up a prolonged enquiryinto the nature of false beliefsor mistakes,and is of necessityat once led to debate how we can eithertbinkor state that which is not. How can I eitherthink or describesometbing whichis not thereto be the object of a thoughtor description ? But I thinkthat this is in fact no digression at all. For, first, it is true that I can only be describedas knowing the same sort of as mistaken about. To knowis, thingsas I could be described no mistake. And, second,any descripat the least,to be urnder tion of any actual or possible mistake automaticallyreveals
318
GILBERT
RYLE:
both of what is falselyjudged and, correspondthe complexity ingly,of what would be truly judged. For to mistake is to forsomething else. insteadof forsomething take something So a ' simple' could never be the object of a mistake. I could mistakenlythink that 7 + 5 = 11, and unmistakenly judge or know that 7 + 5 = 12. But 7 could not be the total not the total object object of a mistake,and so, by implication, either. And this is,what of a piece of truebeliefor knowledge withsensewas at bottomwrong withthe equationofknowledge acquaintance. This noise or that stenchis not the sortof thing as what I mistakenly believe,and therethat could be described as what foreit is not the sortof thingwhichcould be described I correctly believe or know. There must be a complexof disas well in what I know as in what I miselements tinguishable believe. What I knoware facts,and facts takenlyor correctly always have some complexity. So 'simples' could not be facts, thoughthey would be elementsin facts.. Only a propername could directlystand for a simple, and only a sentencecould state a fact. Now, withoutraisingforthe momentthe questionwhat are the simplesor elementsof which what I know or believe are complexes,or even whetherthere are any such elements,we must be one of two things. can see that a complexof elements or it is some sort Eitherit is just a lot or assemblageof elements them. Either the required of union of or fabric embodying complexof elements A, B and C just is A and B and C, so that to knowthe complexwould just be to knowA and to knowB, be to go back on the result and to knowC, whichwould merely alreadyarrivedat and to supposethat what can be named but not stated could be what I know. Or the requiredcomplex is some sort of an organisedwhole, of which the principleof fromthe elements is distinguishable whichit comorganisation is something oforganisation bines. And in thiscase theprinciple unitaryand not to be resolvedinto a pluralityof elements; to the original that is, it is a new ' simple', somehow superadded into the singlecomplex. But if we whichit organises elements may not say that simplesare what we know,we may not say it of thisnew combining either simple This point is broughtout by means of the analogy of letters whichthemselves in syllables. A syllableis a complexofletters, a syllable is nothingbut the are not complexes. Now either in it, in whichcase to knowit is just to knoweach lot of letters if what I knowmust always of them,an illegitimate hypothesis of be a complex. Or a syllable is some orderedarrangement
PLATO S
PARMENIDES'.
319
letters. But in this case the order of arrangement is not a lot of lettersbut something unitary and irresoluble. And then it is an extra simpleelement(thoughnot, of course,one of the same typeas a letter). Finallyit is.argued,on the tacit assumptionthat by a ' complex' can onlybe meanteither a conjunction of similarelements ('letters') or a conjunction of some elements of one sort ('letters') plus at least one elementof a different sort (' order of arrangement '), that in fact such conjunctions or assemblagesare not moreknowablebut less easily knowable than whattheyare conjunctions of. If knowing was inventorying collections,certainlysimple elementscould not be known. But in fact,whateverknowing is, collections are not more accessibleto knowledge than their membersare. Moreover,inventories are just as well capable of beingthe objects of trueor falsebeliefsas of knowledge. So the differentia of knowledge is not to be foundin this direction. Now this discussionrevealsat least two extremely important points. 1. It is true that if the universecontainssimples,such that foreach therecouldbe, in principle, a proper name,the utterance just of this propername could not be the expression of true or false belief or of knowledge(in the sense of 'knowledge that . . .'). What I believe or know requiresa whole sentencefor its expression, and what a sentencestates is in some sense a complex. It is always possibleto findforany sentenceanother of which is partlysimilarand partly sentencethe signification to that of the given sentence,i.e., what a sentence dissimilar fromeach other says containsparts or factorsdistinguishable variations and capable of some independent by substitution. Now, thoughPlato does not makethisapplication, Substantial Forms were supposed to be just such simple namables. And ifwe ask ourselves:What wouldit be liketo be knowing Equality or knowingJusticeor knowingExistence?, and, still more, if about Equality we ask: What would it be like to be mistaken or Existence?, we find ourselves or Justice and bothered bothered forthe veryreasonthat Plato here gives,namelythat we know so and thatwhenwe describeourselvesas 'believingor knowing so', a propername cannot go into the place of the accusative to thoseverbs. who approvesof the refutation of the Oddly,Prof. Cornford, is havingsense-acquaintance, since knowing view that knowing stillbelievesthatPlato's realtheory is, or is a speciesof,judging, in this dialogue,was that Substantial of knowledge, unexpressed is of. Yet this would involvethat Forms are what knowledge
320
GILBERT RYLE:
' Equality ' and 'Circularity ' do express knowledge,for all that it would be nonsenseto assertthat any such abstractnoun could expresseithera mistaken or a truebelief. Socratesdrawsattention to an important affiliated pointwhen he asks how we can mistakeone thingforanothereitherwhen we know both (supposingstill that we may speak of knowing 'things '), or whenwe know neither, or when we knowone and not the other. And he asks: Who has evermistaken the number 11 forthe number12 or viceversa,forall that plentyof people have taken 7 + 5 to equal 11 ? No one has ever told himself that an ox mustbe a horseor that two mustbe one, that beauty is uglinessor justice is injustice. By analogy we might ask (thoughSocratesdoes not): Who evertold himself the infallible tidingsthat 11 is not 12 or that 11 is 11, that justice is not injustice or that uglinessis ugliness ? It is tempting to suggestthat the moial of this puzzle and of later developments of it is somethinglike this, that while a mistakenor a true judgementmust contain some pluralityof this requirement elements, as it stands is too hospitable. Not any combination of any sortsof elementsconstitutes a possible mistake,or in consequence,a possible truth. '7 + 5 - 11 ' is a possible mistake, but ' 12 is 11 ' is not. ' Theaetetus is ' is not a possible mistake,but ' Theaetetus is the Theodorus ' is. The elementsof what I know or believe son of Theodorus will not all be of the same type. But Plato does not hereallude to any such lesson. 2. But anyhow it is unquestionablethat Plato is in this dialogue alive to the following matters. WlhatI know or truly believe or falselybelieve is some sort of a complexof elements, and one the verbalstatement of whichrequires not a name only, nor even a conjunction of names, but a complex expression of which the special formof unity is that of a sentence. What constitutesa complex,like a syllable,a unityis some feature of it otherthan any one or the merelot of its elements, such as letters. That is, Plato is nowconsidering theplaces and r6lesof' terms' in truthsand falsehoods, withhis eye on the underlying question of what are the principlesof organisationwhich govern the combinationof such ' terms'. He does not say, nor are we warrantedin inferring fromthe contentsof this dialogue that he saw, that there are some concepts,namely form-concepts, whichcannot do duty forproperconceptsor ordinary 'terms', much less that he saw that 'exists', ' not ', ' one ', ' several' and othersdo expresssuch form-concepts. But it is clear that
PLATO S
PARMENIDES'.
321
the formal he is consciously developinga methodof inspecting properties of such complexesof elements as constitutetruths and falsehoods. He knows that names are not true or false, that sentencesare not names, that sentencesare not just assemblages of names or composite names resoluble without residue into several componentnames; and he knows that nothingless than sentenceswill expresswhat we know or truly or falselybelieve. A mereinventory of namable simpleswould not onlynot be all that we know,or wishto know,it would not even be any part of what we knowor wishto know. In any truth or falsehoodthere must be some multiplicity a -ofdistinguishable and of these at least some perform factors, different sortoffunction from some others-the orderof arrangementof lettersin a syllabledoes not play the same sort of role and so is not the same type of factoras the individualletters. Of course, Plato has not got a substitution-method, or, what this involves,a code-symbolism to indicate those with whlich similarities which sanction or and differences of factor-types veto particular substitutions. But that thereis a co-functioning of distinguishable factors in truths and falsehoods and that their functions are not all similaris, I suggest,a thingwhich Plato is hereclearlyrealising. The Sophist. to arriveat a cleardefiniThis dialoguebeginswithan attempt is thatofdichotoma Sophist. Its method tionofwhatconstitutes ous division. Some highlygenericconcept,which is assumed withoutproofto be the correct one, is dividedinto two species, divided into two sub-species, and one of these is then similarly so on untila pointis reachedwherethe conceptunderenquiryis of the original seen to be such and such a sub-sub-species genus. Many commentatorsregard this method of Dichotomous of Plato, and some even identify Division as a grand discovery it withthe MethodofDialectic forwhichPlato makeshis famous claims. It is clear to me that the Methodof Dialectic as this is describedin outlinein the Republicand in detail in the Parand the later parts of the Sophist,and is actuallyexermenides has almostnothing cised in the secondpart of the Parmenides, to do withtheMethodof Division. The Methodof Dialectic has links with Zeno's antinomianoperations, or it may just be an expansionof them; but this process of DichotomousDivision sort. In particular, it is not is an operationof quite a different as Aristotle a processof demonstration, pointsout.' 21
1 In Prior Analutics. 46a. PosteriorAnalutics. 91b and 96b.
322
GILBERT RYLE:
WhetherPlato did or did not believe that the Method of Divisionwas a powerful -philosophic we can be quite instrument, clear that it is not so. No philosopher, includingPlato, has it fortheresolution evertriedto employ ofany serious philosophand if theyhad done so theywould not have sucical problem, of all it can onlybe appliedto conceptsof the ceeded. For first or determinable-determinate genus-species sort, and it is not conceptsof thissortthat in general, if ever,engender philosophical problems. And, next, most genericconceptsdo not subdivide into just two polarly opposed species; usually there are numerousspecies of a genus or sub-speciesof a species.' a sort divides into two or seventeen And the questionwhether sub-sorts is, in general,a purelyempirical question. So nearly any case of a philosopher's operation by Division could be upset of sortslyingon neither by the subsequentempirical discovery side of the philosopher'sboundarylines. And, finally,there is roomforalmost any amountof arbitrariness in the selection fromthe ladders of sorts en routeforthe definition of a given concept. Except in artificial hierarchies,such as library catalogues and regimental ranks, there are few, if any, rigid scales of kinds. So there are many tolerableand no perfect mostof the sort-concepts ways of defining that we employ. Had Plato wishedto exhibitthese and kindredblemishesin of definition the programme by Dichotomous Division,he could have chosenno moreeflective than that of exhibiting procedure of one and the same concept,all achievedby severaldefinitions scales of kinds. And this is what in fact descendingdifferent definitions of 'sophist', he does. He gives six or seven different all arrivedat by different he does not say that paths. However, he is revealing defects in the method, and the subsequent dialogue,the Politicus,is anotherexercisein it; so some of his commentators may be right in believingthat Plato thought well of its potentialities. However,thereis a pair of conceptswhich are forcedupon our notice in the course of the operationswhich turn out to sortof elucidation, requirea verydifferent namelythoseof nonexistence and existence. For a Sophist is a pretenderwho or says that what is not so is so. The puzzle which either thinks arose in the Theaetetus arises again here. How can what does or thought of ? And if it cannot, not exist be named,described how can we or Sophiststalk or thinkof it, falsely, as existing ?
2 Cf. Aristotle's criticism of the programme of dichotomous division, De Part. An., 642.
PLATO'S
PARMENIDES'.
323
So the questionis squarelyput: What does it mean to assert or denyexistence ? of something What do Pluralistsor Monistsmean when they assert that thereexist a lot of thingsor that thereexists only one thing? or idealistsmean whentheyassertor deny Wlhatdo materialists that bodies or that Formsare real ? for our main question to Now, it is of the firstimportance notice certain points. (1) With referenceto Parmenidean Monismit is shownthat the conceptsof Unity and Existence interlockin an important way, withoutbeing identical. And of the Parmenidesis echoed here part of the argumentation upon just thismatter. (2) No attemptis made to elucidatethe concepts of existence and non-existence by the Method of Division. The heroic attempt of Meinongto show that they are co-ordinate species of a genericconcept is not anticipated by Plato. And we can see-so perhaps Plato saw-that the Method would not work just because these concepts are not but that thereis an important diflerence between sort-concepts, and these two which is the source of the insort-concepts applicabilityof the Method of Division to them. (3) There are some otherconcepts, identity, otherness, changeand changelessnesswhichhave to be operatedupon alongsideof existence and non-existence. (4) The procedure of investigating the interrelationsof these conceptsis called Dialectic-which, I think, is only remotelyconnectedwith the operationof tracingout, sort-hierarchies whichis called Division. Now in attempting to elucidatethe conceptsof existenceand, Plato makes use of two analogies,one of which non-existence, he had used in the Theaetetus. Namely,he comparesthe ways in whichsome conceptswill combinein only certainways with certainothers(a) to the ways in whichletterswill only admit of certainsorts of alliances so as to formsyllables,and later (b) to the ways in which wordswill admit only of certainsorts of alliancesso as to form sentences. For a syllableto be constituted vowelsmust be thereas well as consonants, and fora sentence to be constituted a noun must be conjoinedwitha verb and not a noun witha noun or a verb witha verb. If we like to build metaphors from theseanalogies we can say that some, but not all, conceptsmust be 'vowel 'concepts,or that some,but not all, conceptsmustbe 'syntax 'conceptsas opposed to 'vocabulary '-concepts. And existence and non-existence are of thesenew types. It is further indicated[253, 259, 260b]that thesetwo concepts of existence and non-existence, togetherwith certain others
324
GILBERT RYLE:
whichare associatedwiththem,namelychangeand changelessare in an important ness, otherness and identity, way pervasive -they crop up, that is, in all the Division-scales in whichwe in the same sortof way,I take it, as ' nonlocate otherconcepts, ' cropped up in one of the definitions existence of ' sophist'. We are remindedof Aristotle'sassertionthat Existence and Unity and Goodness belong to no one of the Categoriesbut pervade them all, though his Categoriesare not, of course, Summa Genera. There appears then to be quite good internalevidencein the an important Sophistforthe view that Plato was now discerning and in parbetweentypes of conceptsor universals, difference ofsorts, whichcan be scaledwithor without ticularthat concepts of genera,species and sub-species, precisionin hierarchies obey rules fromsome others,like existenceand nonvery different what existence. And the conceptsof this latter class perform I may call a logicalrole whichis analogous' to the role of vowels in sentences. They function in syllablesor that of syntax-rules not like the bricksbut like the arrangement of the bricksin a building. Now the interesting thingis that it is true that existenceand non-existence are what we should call 'formal concepts', and further that if modernlogicianswere asked to describethe way in whichformal differ concepts from proper ormaterial or contentthe role of formal concepts,theirmethodof exhibiting concepts would be similarto that adopted here by Plato. But we need than to say that Plato was becomingaware of not go further of type between concepts. There some importantdifferences is no evidenceof his anticipitating Aristotle'senquiryinto the which enquiryit is which firstrenders principlesof inference, the antithesis of formaland otherconceptsthe dominantconsideration. There is, consequently,in Plato, no essay at the formalfromthe contentualfeaturesof proposiabstracting forthe formal in abstraction tions,and so no code-symbolisation from the materialfeatures of propositions. Thereis, of course,alwaysa considerable hazard in attempting to elucidatea doctrineof an earlierphilosopher in the lightof doctrines. It is subsequent and especially of contemporary and often alwaystempting easy to read palatable lessonsbetween the lines of some respectedbut inexplicitScripture. But the to chartthe driftof some adolescent oppositepolicy of trying to the progressof any more adult theory without reference is subjectnot to the riskbut to the certainty theories of failure.
PLATO'S
PARMENIDES'.
325
We cannot even state what was a philosopher's puzzle, much less what was the directionor efficacy of his attemptto solve it, unless subsequent reflections have throwna clearer light upon the matter than he was able to do. Whethera commentatorhas found such a light or only a will-of-the-wisp is always debatableand often verywell worth debating. Thus I may be wrongin believing bethat thereare affinities tween Plato's enquiries in these dialogues and Hume's and Kant's account of assertionsof existence,Kant's account of of proposiforms of judgement and categories, Russell's doctrine tional functions and theoryof types, and, perhaps,more than Tractatus any other,nearlythe whole of Wittgenstein's Logicothese dialogues Philosophicus. I may be wrongin construing as, so to speak, forecasting most of the logical embarrassments into whichthe infinitely courageousand pertinaciousMeinong was to fall. But at least my error, if it is one, does not imply that Plato's puzzleswereso factitious that no other or ephemeral serious philosopher about has ever experienced any perplexity them.

Mind Association

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mind Association

Uploaded by

Copyright:

Available Formats

Mind Association

them is Unity. A thing is notthatofwhich it is an instance.

PLATO 'S 'PARMENIDES

PLATO'S ' PAPRMENIDES'.

You might also like