International Phenomenological Society Hegel's Critique of Kant's Theoretical Philosophy Author(s): Karl Ameriks Source: Philosophy and Phenomenological Research, Vol. 46, No. 1 (Sep., 1985), pp. 1-35 Published by: International Phenomenological Society Stable URL: http://www.jstor.org/stable/2107654 . Accessed: 08/02/2011 20:28 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=ips . . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent 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 [email protected]. International Phenomenological Society is collaborating with JSTOR to digitize, preserve and extend access to Philosophy and Phenomenological Research. http://www.jstor.org
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Hegel's Critique of Kant's Theoretical PhilosophyAuthor(s): Karl AmeriksSource: Philosophy and Phenomenological Research, Vol. 46, No. 1 (Sep., 1985), pp. 1-35Published by: International Phenomenological SocietyStable URL: http://www.jstor.org/stable/2107654 .
Accessed: 08/02/2011 20:28
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .http://www.jstor.org/action/showPublisher?publisherCode=ips. .
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
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 [email protected].
International Phenomenological Society is collaborating with JSTOR to digitize, preserve and extend access to
Philosophyand PhenomenologicalResearchVol.XLVI, No. i, Septemberi985
Hegel'sCritiqueofKant'sTheoreticalPhilosophy
KARL AMERIKS
UniversityofNotre Dame
Ivan Soll has remarkedthat "Hegel's entireprogramand conceptionofphilosophydependedupon refutingKant's limitationof reason."' ButwhileSolldiscussesHegel'sattemptsinthisregard,he admitshe has not"attemptedto corroborateor criticizeHegel's interpretationofKant."'Soll is not alonehere,forevenwiththegreatrenewalof interestinHegel
today,3therehas been surprisinglylittlecriticaldiscussionof Hegel'streatmentofKant,especiallywithregardto thedifficultcoreofthattreat-ment,namely,therejectionofthetwo centralcomponentsofKant'stheo-reticalphilosophy:thetranscendentaldeductionof thecategoriesandthedoctrineof transcendentalidealism.
Therehave beena fewhelpfuldiscussionsof thissubjectrecently,4butnone has givena systematicaccountof thedistinctionbetweenthesetwocrucialcomponentsin Kant'sown view and of thegeneralnatureandgroundof Hegel'streatmentofthedistinction.Inthefollowingsections,Ifirstoffersuch an account(insectionI) and thendistinguishand evaluateHegel's threetypesofobjectionsto Kant's deduction(insectionsII-IV)and his threetypesof objectionstoKant'sidealism(insectionsV-VIII). Iarguetheseobjectionsall failbecauseof a closelyrelatedsetoferrors,errorswhichareunderstandablebecausetheyconcernsomeof themost
Ivan Soll, An Introductionto Hegel's Metaphysics(Chicago: Universityof ChicagoPress,i969), pp. 48-49.
2 Ibid., p. xiv.
See, e.g., CharlesTaylor,Hegel (Cambridge:CambridgeUniversityPress,1975); Rich-ard Bernstein,"WhyHegel Now?" Review of Metaphysics31 (1977): z9-60; WilliamMaker, "UnderstandingHegel Today," Journal of the History of Philosophy19
(1981): 343-75.4 See especiallyJohn Smith, "Hegel's Critique of Kant," Review of MetaphysicsZ6
difficultissuesinKant'sphilosophy.I also contendthattheseerrorsillus-tratecertaingeneralpatternsofapproachingtranscendentalphilosophywhichare stillveryinfluential,andhencean especiallyappropriateroute"backto Kant" inour owntimeisvia the re-examinationofHegel'scri-
tique ofthatphilosophy.5I. Hegel'streatmentofKant'stranscendentaldeductionturnsout to be
so closelyconnectedto his objectionsto transcendentalidealism thatbeforeanyassessmentcan bemade ofhisspecificattacks,itisnecessarytoreviewtheessentialsof thegeneralrelationof Kant's deductionto hisidealism.6The pointofKant'sdeductionis basicallythecentralclaim oftheTranscendentalAnalyticof the Critiqueof Pure Reason, namely,
that (i) thereare determinablea priori principlesforspatiotemporalexperience,principlesinvolvingcategoriessuch as substanceand causal-ity.The pointof Kant's idealismis basicallythecentralclaimof theTran-scendentalDialectic oftheCritique,namely,that(z) metaphysicallythespatiotemporalrealmhas a nonultimatestatus,so thatwhateverwe orotherbeingssuchas God maybe inthemselves,suchthingsinthemselvescannothave intrinsicmaterialproperties.Ifwecombinethesepointsandadd that(3) ourobjectivetheoreticalknowledge7cannot transcendthe
realmgovernedbytheprinciplesofAnalytic,wethengetthelessonoftheCritiqueas a whole,a doctrinewhichhas beencalledKant's "restrictionthesis,"namely,that(4) althoughourknowledgehas a prioristructure,itis all only phenomenal.
By "knowledge"I will meangenerally(withoutalwaysmakingit explicithereafter),asKant does, "objective"or determinateas opposed to merelyanalyticor formalknowl-
Antinomiesthat Kant shows transcendentalidealism is inescapablebecausewithoutitwe are leftin contradiction.This is Kant's ultimate
groundfor(z), and it is onlyvia thisclaimand (3)10 thathe concludeswithhisrestrictionthesis.HenceforKanttherestrictionthesiscanbenei-therconflatedwithnorwhollycut offfromthededuction:thededuction
alonedoes not provethespatiotemporaldatawe dealwithare onlyphe-nomenal,andtheAntinomiesalone doesnotprovethattherealmofthese
data exhaustswhat we can know.Hegelwas not veryclearabout these relationsbetweenKant'sbasic
argument.For example, he says repeatedlythat in using the term
"transcendental"in theproofofthecategories,Kantmeansto be express-ingtheirideality."This issimplytooverlookthetwostagesin thededuc-
8 Seee.g.,B 41, B 146. "A" and/or"B" willrefer,as is customary,to thefirstand/orsecond
editionsof the Critique of Pure Reason. Cf. K. Ameriks,"Kant's Transcendental
Deductionas a RegressiveArgument,"Kant-Studien69 (1978): z73-87.
9 B xix.
0 The basis for(3) is Kant's doctrineof judgment,i.e., thatour objectiveknowledge
Thus,tosaya unityofapperceptionis transcendentalis notto meanthatitmustattachto what is beyondexperienceand intrinsicallynonspatio-temporal,but is justto saythatitisa kindofapperceptionnecessaryforexperience.
Some historicalconsiderationsmayhelpaccountforHegel'sconfusionhere.'3We know thatinhis crucialformativeyearsHegelneglectedclosestudyof the centraltheoreticalcomponentsof Kant'sphilosophy,'4andwhenhe did approachthemit was throughtheperspectiveofsomeone
interpretationswhichgivea verypeculiarslant to Kant's work.ThusFichte'srouteto therestrictionthesis'5restson cruderepresentationalistworries'6and ignoresthespecificargumentsof the deductionand theAntinomies.Hegel followedFichte's lead in skippingover theseargu-ments,and evenin hislatestand mostsystematicaccountsofKantthereisnopatientanalysisof thesetextsand henceno clearstatementofthetwo
stagesin Kant's argument.Fortunately,though,Hegel did not acceptFichte'srepresentationaliststartingpoint. He appears to have beenspurredto idealismratherbySchelling'snotionthatour freedomimmedi-atelydemonstratesthat our ordinaryknowledgeis merelyphenome-nal.'7ThusHegel arguesthattheitemswithintheworldaremerelyphe-
nomenalsimplyinthesensethattheydo nothavetheirgroundin them-selves(as an absolutelyfreebeingsupposedlydoes).'8 In thiswayHegelcan manageto hold on to the metaphysicaldowngradingofspace andtime'9thatis attheheartoftranscendentalidealism,whileavoidingcom-
mitmentto Kant's restrictionthesisoranyofthespecificargumentstoit.As a consequence,however,itbecomesall the easierforHegel to misswhatis reallyinvolvedin Kant's deductionand idealism."
The firstobjectiondemonstratesFichte'sinfluence,forHegel repeat-edlyremarksthatFichteis to begivencreditforbeingmoreconsistentandrigorous,forat leasttryingto deduceall of thecategoriesfromthe "I"."
17 F. W. Schelling,Werke,(Stuttgart:J.G. Cotta,1856), Vol. I, p. 340. Cf. Hegel,I: z34.
Thisobjectionalso revealshowuncertainHegelcouldbe about Kanthim-self,forwhenhe firstexpresseshisidealconceptionofthededuction,hehesitatessayingKant didnot seethisideal. In "principleorspirit,"ifnotin"form,"Kantisallowed tohaveunderstoodthefundamentaltruththat
all realityis relatedto and foundedin an I, an absolute mind." Or, tonotetwo otherways Hegelhas ofexpressingthisfundamentaltruth,it issaid thatthe"principleofthededuction"is justtheFichteannotionof"genuineidealism"(whichis supposedto rejectanydistinct"thinginitself"),"3and thatthisis the sameas the"generalprincipleofspecula-tion," theprincipleof theabsolute (and notmerelyformal)identityofsubjectand object.'4Verysoon,however,HegelchangedhisestimateofKant,and althoughhe continuedto hold to thefallacynotedearlier,the
idea that a transcendentaldeductionmustbyitselfbe an argumentforidealism,hecame to recognizethateveninthevery"spirit"of hisworkKant himselfwas not directedtoward the formofidealismwhichdoeswithoutthingsin themselves."5Hegel thenchargedKant withinternalinconsistency,forheheldthattheprincipleof Kant'sown deduction(the"I") demandedsuchidealism,despitewhateverKant himselfsaid.
Beforewe evaluateHegelon thispointit should be madeclear thatin
callingfora deductiontieddirectlyto absoluteidealismand basedentirelyon a simple,mentalstartingpoint (thepure representationof the"I"),Hegeldidnotmeanto insiston a Cartesianoregoisticbasis forphiloso-phy.On thecontrary,hestressedthatitispreciselya virtueof hisview,asopposed to what he now understands as the "subjective" or"psychological"idealismofKant,that thereis no suggestionthattheformofrealityarisesfromourimposition,as ifitwereliterallyuptous,asfinite,particularminds,to determinethecategoriesthatobtain."6Unfor-
tunately,whilethisclaimshowsHegel'sownviewis atleastnotas radicalas somehavefeared,itrevealsbutanotherinjusticetoKant,whowouldhavebeenonlytoohappytoagreethatsuch a subjectiveidealismis inap-
critique,namely,therejectionof thethingin itself- to theinfluenceof ReinholdandG.E. Schulze.See above n. zo and belown. 87, and D. Breazeale,"Fichte'sAesnesidemusReview and the Transformationof German Idealism," Review of Metaphysics34
(I981): 545-68.2 Hegel, II: IO, Surber,p. i.
23 Hegel,II: 9, Surber,p. i. Hegel goeson to arguethatFichtehimselfdid not hold truetothisidealism.
propriate.It ispreciselyforthisreasonthatKant stresses(as Hegelmaysee elsewhere"7)thattheformofthe individualempiricalego is just aslimitedto theconditionofmerephenomenalityas is therestof thespatio-temporalrealm,and thatthephenomenalityof thisrealmrestsultimately
on thedistinctivequantitativefeaturesof spaceand time,featuresthatarea prioriand notimposedbyhumansinanyimaginablesense."8Thusifitis wrongto beginwithto connectKant'sdeductiondirectlywithidealism,itisevenmoreunfairtoconnectitat all witha subjectiveversionof ideal-ism.
Thisall stillleavesitunclearjustwhyHegelbelievedone shoulddevotethe deductionto an argumentforabsoluteidealism.In generalterms,Ibelievetheanswerhererestson theidea thatifthecategoriesofthingsin
themselveswerecompletelydeterminablefromwhatisinvolvedina purerepresentationof the I, thenthiswould demonstratethe basic subject-objectidentityto whichHegelis committed.Thentheworldwouldnotbethemereimageofa particularself,as inegoisticidealism(solipsism),norwould its ultimatefeaturesbe beyondthereachofall suchselves,as inKantian idealism,'9butitwould ratherbe distinctfromus and yet (inessence)whollyaccessibleto our mind.Expressedsimplyintheseterms
Hegel'spositionis not all thatimplausible,althoughforitto be distin-guishablefromordinaryrealism,as Hegel surelytakesitto be,he alsoneedsto show how the"pure" representationwithwhichhe wants thedeductionto begincan havesomekindof mentalor subjectivemeaningwithoutbeingidenticalwiththe notionof a mereempiricalego. Hereeverythingrestson preciselywhatHegelunderstandsbythepurerepre-sentationoftheI.
In so faras HegelcriticizesKantfornot havingmovedproperlyfrom
thisnotion,itwould seemthatone couldfindout whatthenotionis sim-plybylookingat what Kantsays. Yet,thefactthatKantreachesconclu-sionsverydifferentfromwhatHegelexpectsalreadysuggeststhatKant'sownstartingpointmaybe unlikeHegel's,and whathasalreadybeenseenof Hegel's interpretationwould supportsuch a suggestion.Furtherconfirmationforit can be foundinthe factthatat onepointHegeltakesthenotionoftheI to standforanabsoluteidentitywhichisexpressednot
inwhatis calledthe "shallowness"3"of theargumentof thedeductionitself,but ratherin Kant's referenceto the "productiveimagination,"
whichis saidto be thegroundofboththe"subjectiveI" and the"objectiveworld."3'ElsewhereHegel identifiesthisfacultywitha unitythatis the"absoluteidentityofself-consciousness,whichpositsjudgment[and so
outof itself."3Hegel thus movesfarbeyondthetext,forKanthimself(unlikeFichte)does notspeakof sucha "positing"absoluteconscious-ness,nordoes he hypostatizetheproductiveimagination.Althoughit
soundsquiteforeboding,"productiveimagination"in theCritiquecan beunderstoodas a termthatstandsforlittlemorethanthefactthatthereare
a priori(and hence"productive"ratherthan"reproductive"or merelycontingent)ruleswhich governparticularformsof sensibility(hence"imagination"ratherthanabstractthought)."
Hegel sensesthis restraintin Kant'sown talk,and he chargesthatitleaves Kant with the absurd picture of an empty (not literally"productive")I confrontingan absolutelydistinctmanifoldof data,
which as such would retainan unsurpassableprimacyvis-A-visthe
mind.34The pictureis calledabsurdbecauseHegelthinksthattheI and
have sucha startingpointand thatthisis at leastin somewaystronglyhintedat bywhatKanthas to say.
The mainclueforhandlingthiscomplexquestionhas todowithsomebasic ambiguitiesin thenotionofapperceptionin thepurerepresentation
oftheI. For Hegel,as forKant,thenotionofa purerepresentationoftheIis mostcommonlyexpressedas thenotionofa necessarysyntheticunityof apperception.However,bysayingthereisnecessarysyntheticunityofapperception,one could meaneither:
(i) (a) all representations35mustbe able to belongto one self-con-sciousness;
(b) sincethereis a manifoldofsuchrepresentations,theunity
here can be called synthetic,evenwhen it involvesanalyticrelationsamong-thecontentsoftherepresentations;
(c) and sinceit is a unitywithin(andpotentiallyfor)a self-con-sciousbeing,itcan be calledapperceptive,evenifthereis atno pointan actualall-encompassingthoughtof theform,"Ithink . .
or(ii) (a) all representationsin ourcognitionare componentsof judg-ment and expressan objectiveunityof terms,whichcanalwaysbe formulated,"I think . . . is .
(b) sincethereis a manifold . representations(as above);
(c) and sinceitis a unity. . . "I think"(as above).
Notethat(i) stressesa conditiononthoughtsas representationssimplyhad byus, whereas(ii) stressesa conditiononthemas cognitiveelements,i.e., stateswhichat leastmake a claim thatcan be trueor false. They
perception,suchas "x seemsto be y," whichnecessarilycan be embeddedin the cognition,"it is the case thatI thinkx is y.,,36
These distinctionshave greatconsequencesfortheissueof theapplica-bilityofthecategories.Anargumentwhichmovesfirstfromonlytheepis-
temicsenseofapperceptionwould show thatthecategoriesarenecessaryinso faras theyarenecessaryforjudgment(as inKant's "metaphysicaldeduction").Ifone thenholds,as Kantdoes,thatour(determining)judg-mentsdependongivenintuitions,andso onparticularformsof sensibil-ity,a beliefintheidealityofthose formswill block usingthe categorieswhich are deduced (via theepistemicrelation)to determinethingsinthemselves.On the otherhand,ifone believescategoriesare deduciblefromthemerepossessivesenseofapperception,thentherewill benosuchoriginalrestrictionofthecategoriesderivedandnotevena clearneed forconsiderationofwhatis involvedintheconditionsofjudgment.Inotherwords,onewillbeina naturalpositionto be an absoluteidealistand to beperplexed,as Hegel was, by any "restrictionthesis."(Note also thatthiskindofunderstandingofapperceptioncanmeettheneeddiscussedearlierforsomekindof mentalelementwithoutimmediatesubjectivisticconse-quences.)
I contendthislineofthoughtrepresentsnot onlya path Hegel couldhave takenbut the one he actuallytook.37He explicitlycommitshimselfto the crucial conflationnecessaryfor this linewhenhe definespureapperceptionas the"actbywhichthe'I' makesthematerials[ofrepresen-tation] 'mine'."38Elsewhere he contendssimilarlythat the "specificgroundof thecategoriesis declaredbythecriticalsystemto lieinthepri-maryidentityofthe'I' inthought- what Kant calls thetranscendentalunityofself-consciousness."39
This conflationis notunprovokedby Kant's language.In additiontothecomplicatedcloseconnectionultimatelyexistingbetweenwhatiscov-eredbythetwo sensesofapperception,40thereis thefactthatKant some-
That thisis Kant'spositionis arguedinAmeriks,"RecentWork,"pp. 14-i8.
3 Here I am passingover the contrastbetweenHegel's radical (theone thatemphasizes
productiveimagination)and tame(theone thatspeaks merelyof the possessivesense)
modelsof theI, and I take myselftobe doinghima favorby concentratingon thetame
model.Thereis of coursea close relationbetweenthesemodels,sinceit appearsthatfor
Hegel the primordialsubject-object,particular-universalrelationis instantiatedwhen
onegeneratesone's thoughtand considersit to beone'sownandsubjectto theprinciple
of apperception.Cf. VI: z5 5,JohnstonandStruthers,vol. 2, p. 219. See also below,atn.
45-38 Hegel, Encyclopedia,?42, Zusatz i, VIII: i i 8, Wallace, p. 69. HereHegel speaksof a
remarkable"Tdtigkeitdes Vermeinigens."
39 Hegel, Encyclopedia,42, VIII: ii6, Wallace,p. 68. My emphasis.40
timesspeaksas ifhis principleof apperceptionhas an ontologicalmean-ing,as ifthepointis how all ourpossiblerepresentationsare necessarilyrelatedto ourbeingone and the samethinkerhavingthem.Fortunately,Kant also saysenoughelsewhereto indicatethedifficultieswithresting
thedeductionon thepossessivesense of apperception.4"In theParalo-gismshe indicatesthat the argumentwould go througheven on thehypothesisthatthereis an ultimatepluralityof substancesunderlyingtheactinwhichthoughtis syntheticallyunited.42And withinthe deductionitself,hestressesthat'itis theepistemicconditionofthe"objectiveunityof
subjectiveunityofassociativerelations.A furtherdifficultywithstressingthe possessivesense is thatit is
unclearjustwhatcan be meantbysayingitisan actofoursthatmakesourrepresentationsourown.As Hegelhimselfindicates,inmockingtalk ofrepresentationsbeingmerely"accompanied"bya mind,"we have norelevantnotionof a representationthatis nota representationhad byamind. The dependentnatureof representationsmakes themreadytobelongto us quiteindependentofwhateverwe do. Moreover,ifthere
weresomeactbywhichrepresentationsweremadecapableofbeingours,itis remarkablethattheveryprincipleofapperception(on thisreading)impliesthatthereisno specialeffectorlimitationthatthisactionwouldinvolve- anyandall representationsaresubjectto it,and by beingputintorelationwithoneselftheythemselvesarenotsupposedto change.(Isuspectthisfactis relatedto theearlierpointthatHegelstressedabout themanifoldhavingtobewhollyamenableto us.Thisisobviouslytrueinthesense thatthereis nothingdata would have thatwould block thema
mustbecounteredthatthisis to invokean empiricalnotionofappercep-tionwhichis preciselywhat was to be excludedbythenotionofa purerepresentationoftheI. Suchapperceptionmightberesponsiblefora quite
intelligiblerelationto representations,but it would have no chanceof
4' See K. Ameriks,Kant'sTheoryofMind (Oxford:ClarendonPress,i982), pp. 137-42.4 A 364.4 B i39-40.
claiminga prioritobeapplicableto allrepresentationsandwouldfailtherequirementsof both transcendentalphilosophyand absoluteidealism.
So far,Hegel'scriticismof Kantappearsatbestto tradeon a conflationofthe substantivebut notabsolutelyuniversalapplicabilityof theepis-
temicformoftheprincipleofapperceptionwiththegenuinelyuniversal,but only apparentlysubstantivecharacterof thepossessiveformof theprinciple.This is not the whole story,forin his last treatmentof Kant,Hegel givesa moredetailedattempttospellout thekindofactionthathetakes to be involvedin apperception,admittingthatit is "not exactly"explainedthisway byKant himself:
I am theentirelyuniversal,thecompletelyindeterminable,abstract.Insofaras I setanempir-
contentto be able to enterinto theOne (das Eine), thesimplicityofthe I, it itselfmustbe
made simpleand infectedby simplicity.A contentin consciousnessthus becomesOne,
becomesmycontent.I am I, am One. Thus thethoughtis put intoa unityand so becomes
one . . . What thought produces is unity, and thus it produces itself,for it is
one . . . WhateverI touchmustbe able to allow itselftobeforcedintotheseforms[ofthe
synthesisofapperception]ofunity.45
In thisargumentHegel expressesa combinationof his earlierradical
interpretationof
theI in
termsofa
productivepower,and hislaterposses-
sive conceptionof apperception,a combinationwhich he may havealwayshad in mindbut nevermade so clear before.The I isnowgivenaveryspecificpower,onewhichis togo beyondthemeredesignatingofarepresentationas one's own. Apperceptionis saidto involvetheimposingofa formontherepresentation,theformofsimplicity,whichiswhatthenallowstherepresentationto become one's own.Unfortunately,themerenotionofsimplifyinga thoughtremainsas mysteriousas thatofdirectly
makingit mine. Once again,it is not explainedhow anythingis reallygivento thethoughtthatit doesnotalreadyhave.On theotherhand, ifsomeordinaryactionisproposed,suchas somehowmakingthethoughtsimplebygivingita specialvividorpersonaltone,thenthisagainconflictswiththenecessaryuniversalscopethatisto be attachedtoapperception.One wayout of thisimpassewould be to switchfromdiscussingwhatmustbe doneto a thoughtorrepresentationtomakeitsimpletodiscuss-ingtheconditionsformakinga manifoldor"congeries"of datasimpleinthe sense of beingunited.46But then anotherdilemmareturns,for itseemsthateithertheseconditionscan be spelledout inminimaltermswhich,as Humeindicated,requirenoreferenceto an I,letalone absoluteidealism,or one focusseson a richerkindofunity,suchas thatofjudg-
ment,therebybringingin the epistemicsense of apperceptionwhichHegel avoidsand whichis whatmakesKant'sapproachdistinctfromhisown.47
AlthoughHegel's firstobjectionthusfailsbecause it restson an inter-
pretationwhichis inaccurateas well as intrinsicallyinadequate(giventhelimitedconsequencesoftheHegeliannotionofapperception),it shouldbe notedthatitat leastfitsintothemainstreamofrecentinterpretationsoftransendentalarguments,and especiallyofthemanyquasi-Strawsonianreconstructionsof the deductionwhich abstractfromthefeaturesthateventuallyled Kant to an espousal of transendentalidealism.48In thisway Hegel's attack expressesa veryinfluentialand understandableresponseto Kant thatinprinciple,ifnotinorigin,is largelydetachablefromany peculiarmetaphysics.Moreover,evenifHegel mayhave gonetoo farincallingfora dialecticalargumentfromone categoryto all theothers,andevenifhe is unfairinchargingKantwitha merely"historical"and "psychological"methodin themetaphysicaldeduction,49one can-nothelpbut feelthatHegelhas at least a strongersenseoftheneed foracompleteandfullypersuasivemetaphysicaldeduction,andto thisextenthis call fora morerigorousproofofthecategoriesis well taken.
III. The considerationsadvancedso farmakeitrelativelyeasyto deal withHegel's secondobjection,namely,thatwithKantthe"universalityand
ity.Here again Hegel is conflatingtheargumentofthedeductionproperwith(hisunderstandingof)Kant's doctrineofidealism,a conflationmadeunderstandablebythejustnotedfactthatHegelmissedKant'sown argu-menttothecategories.Thisconflationmayalso havebeenencouragedbythefactthatinKant'sProlegomenathe"necessityincognition"isineffecttakenas a presupposition."We know,however,thatthe methodofthe
Hegel mayappear to approachKant herewhenhe linkspossessingwithconceiving,but
unfortunatelyhe uses the formerto understandthe latter,ratherthan vice-versa:"Inordertorecognizewhata notionis,oneisto considerthenotionoftheI." (VI: 25 5,John-ston and Struthers,vol. z, p. zi9). My translation.
48 See Ameriks,"RecentWork," p. 13. Cf. below,at n. 54.49 See, e.g.,VI: 289, Johnstonand Struthers,vol.2, p. z247;and XX: 346. Hegel is well chal-
lengedhere by Smith,"Hegel's Critique,"pp. 445-48.50 Hegel, Encyclopedia,540, VIII: 113, Wallace, p. 65; cf.XX: 336.51 KantsgesammelteSchriften(Berlin:KoniglichePreussischeAkademieder Wissenschaf-
ten,i9ozf),
IV: 279.
HEGEL S CRITIQUE OF KANT'S THEORETICAL PHILOSOPHY I3
Prolegomenais notthatof theCritiqueitself,5"and in anycase a mereglanceat argumentssuchas the SecondAnalogyshowsKant is out toderivea principleof necessityratherthan simplyto assumeone."
Hegeldoesnot attendto thesetextualpoints,and inanyeventhe may
havebeenunwillingtogivethemmuchweight.Givenour analysisof thefirstobjection,it seemslikelythatforanyargumentotherthanonewhichdefeatsscepticismandpartialidealism(bydeducingthe structureoftheworldfromwhatis givenin thepossessivesenseof therepresentationoftheI),hewouldinclineto theviewthatwe areat thelevelofmerepresup-positionand "explanation"ratherthanofthedemonstrationof genuinenecessity.The ascriptionof such a view seems evenmore properifoneconsidershow highingeneralHegel's requirementsareforwhatcountsasa proof,and ifone recallshow evenmanycontemporaryinterpretersoftranscendentalargumentshave-tendedto assumethatapartfroma refuta-tionofscepticismKant's deductionloses epistemologicalforceand canhave only an odd metaphysicalmeaning.54Hegel explicitlysharesthisassumptionwhen he statesthatwiththepresuppositionof"a thingthatsensesand a thingthatis sensed,all [true]philosophyis drivenfromthefield."55 Thereis no appreciationhere forthe Kantianidea of an argu-
mentwhichbeginswithcertaincommonsensepresuppositionsandthenmoves to unearththeirdependenceon various controversialprinciplesthatare to constitutea necessaryconceptualframework.
In sum,whereasKantpresentsa relativelymodestargumentfrom(a),thenatureofempiricaljudgment,to (b),a listofvalidcategories,and thento (c),a systemoftranscendentalidealism,Hegeldesiresa radicaldeduc-tionwhichwouldbeginwithsomethingpriorto (a), such as (aa), a purepossessiveor "simplemaking"powerofapperception,andwhichwould
gobeyond(c) to (d),a systemofabsoluteidealism.Notfindingsucha rad-icalargumentworkedoutintheCritique,HegelscoldsKant fornotmov-
ingdirectlyfrom(aa) to (b), and thenforsupposedlynotmovingto (b) atall butonlymovingfromit to (c). Inmissingthebasicstructureof Kant's
S SeeA 176/ B zi8, A i89/ B z3z. HeremyinterpretationcontrastswithSmith,"Hegel'sCritique,"p. 459. My defenseofKant hereis also meantto meetat leastthe firsttwo
ancillaryconsiderationsthatareto makepossiblethetransitionsfrom(a)to (b) and from(b) to (c). The omissioninregardto thelatterinferencewill be stressedlater;themainoversightsinregardto theformerconcernnot onlythe epistemicsenseofapperceptionbutalso the useof thepure
formsofintuition,whichis crucialtoKant's conclusionthatthecatego-rieshave a fullyuniversalapplicabilitywithinourexperience."
IV.Hegel's thirdobjectionto Kantdoes notappearto be so closelyrelatedtodetailsofthededuction,forinHegel'soriginalpresentationit isstatedas a very generalproblemthat applies just as well to Locke.57Theobjectionis directedagainstall philosopherswho tryto set down thescope and limitsofknowledgebyfirstexaminingourcognitivefaculties
and forgettherebythatsucha preliminaryinquiryis itselfa partoftheprocessofknowledge.In thePhenomenologyKant's nameis notmen-tionedwhenthisobjectionis madeinitsmostfamousform,butitisobvi-ous thatit istheCritiquethatHegelmeanstoreferto whenhe attackstheprojectof a preliminaryinquiryas motivatedbya conceptionofknowl-edgeas a "medium"or "instrument"to be examinedbeforehandforitsreliability."8The problemsthatsuch a criticalphilosophyis takentoemphasize,thatthemediumof ourcognitionmay,as passive,be obscure
Hegel's explanationof thedifficultywiththisconceptionisveryindi-rect;hisfavoritewayofexpressingit is thatthecriticalphilosophercom-mitsthe errorof "refusingto enterthewateruntilyou have learnttoswim.,,6, If thatis thedifficultyinvolved,thenit appears it could be
know priorto knowingthegeneralcriterionwhichmakes themknowl-edge,and it is throughreflectingon thesegivenparticularcases thatwethencandevelopa formulationofwhat thegeneralcriterionis.6zIn con-trast,a criticalphilosopherwould appear to be a methodistwho urgeswithholdingon particularknowledgeclaims - perhapsbecause he isawareof thedifficultiesjustnoted withour cognitive"instruments"untilhe has ascertainedcertaingeneralprinciples.On thebasisof theseprinciples,and an examinationof our capacityforsatisfyingthem,hispreliminaryinquirycan thenbedevelopedintoa fullaccountofthestruc-tureand scope of our knowledge.
admittingsomeknowledgeclaims,i.e., byaskinghowKant intendedtolearnhow to swimwithoutbeingalreadywet.6' However,ifoneworksinsteadwiththe construalofKant'sprojectdevelopedin oppositionto
Hegel's in the firstpartofouranalysis,thenthereisno suchproblematthispoint.For,on thatconstrualthedeductionisnotout to givea meth-odistresolutionofthegeneralproblemof a criterionforknowledge,foritratherassumesthattherearesome warrantedstatements6,(e.g.,"thereis
somethinghappeningat t") and asks onlyabout thea prioriprinciples
requiredbysuch statements.Hence ifthereis no more to Hegel's thirdobjectionthanwhathas justbeendeveloped,it can be dismissed(as a
directcritiqueofKant)inthesamewaythatthepreviousobjectionswere.It turnsout that thereis somethingmore to the thirdobjection,for
Hegel stressesthatcloselyintertwinedwith(whatwe have called) the
problemofthecriteriononefindsthebasicerroroftreatingknowledge,orour cognitivefaculty,as an instrument.The difficulty(or deep water)
Kant supposedlygetsintohereis mostfrequentlyexplainedby Hegel bysayingthatin assumingsuch a "tool" conceptionof knowledgeKantoverlookedthegreatdifferencebetweenit and realtools." Whereas anordinarytool can be testedindependentlyto see ifit is proper(e.g.,to
checka hammerwedo nothave touseit,letalonehitit witha hammer),our cognitionis suchthatit seemsthereis no testforit whichdoes notitselfappeal to theuse of ourcognition.Thusthereis somethingfunda-mentallyquestionableaboutassumingcognitionisa meretool, and oncewehavedroppedthisassumption,we havedroppedwhatHegel givesasthemotivationforthe criticalphilosophy(namely,the idea ofcheckingbeforehandhow thistoolworks).
Unfortunately,thisnewwayofformulatingHegel's objectioninvolvesthemisleadingsuggestionthatthetoolconceptionofknowledge(as justexplained)isa presupposition-ofKant'sphilosophy.Perhapsitdoesfunc-tionthiswayforsomephilosophers,butinKant'scase it mustbereiter-ated thatthe talkofformsof intuitionand understandingis meantpre-ciselyas the conclusionratherthanthestartingpointofhis argument.Kantno doubtbelievesitwouldbewrongto assumethatourempiricalrepresentationsdirectlyrevealthingsin themselves,"forthe groundof
suchan assumptionishardlyevident,and he also surelybelievesthatthehistoryofmetaphysicalcontroversieswarrantssomeinitialwithholdingabout thereliabilityofreason whenitgoes beyondthe empiricalrealm,but all thisdoes not mean thatfromthestarthe is sayingthatwhat wehave to workwithare onlyour formsof accessto objects.On the con-trary,oncetheseformsarediscoveredand elaborateditbecomesanextrapointto show thattheyare onlyour forms,onlyour instrumentsfortryingto getat somethingin itselfquitedistinctfromus.
Here one mightinsistthatHegel's objectionis independentofwhenoneholds thatthereis sucha thingas a tool ofcognitiondistinctfromthingsinthemselves(whatHegelcalls"theAbsolute").For as longas itistruethatknowledgecannotbelikeanordinarytool,thenthiswould seemto be enoughto makeimproperanytalkofknowledgeas a mediumorinstrument,and so evenifcriticalphilosophyisnotoffbaseat thebegin-ning,itisultimatelyoffbase.TheKantian'sreplytothispointis thatall he
isevercommittedto isananalogybetweencognitionsandtools,an anal-ogywitha particularmeaningresultingfroman argumentto a particularkindofindependenceofthingsfromour mind.More specifically,he canreplythatwhilethe reflexivecapacityofourfacultyofcognitioningen-eralandtheabsolutelyuniversalscopeof a certainpartof thatfacultydo
65 See the passagescited above at n. 6o.66
See especially,Kants gesammelteSchriften,IV: z8z.
HEGEL S CRITIQUE OF KANT S THEORETICAL PHILOSOPHY I7
make it quite unlikean ordinarytool, it is also truethatthispartof ourcognitivefacultycanuse theprincipleofcontradictionto testotherparts,and the resultsof thattestcan implythatthosepartshavelimitedscopeand to thatextentare like mediaormeretools. So, if(as intheAntino-
mies)thereare contradictionsarisingwhen certainpropositionsof com-mon senseare combinedwitha particularmetaphysicalthesissuchas thethesisoftranscendentalrealism,thenthecriticalphilosopher(who,onceagain,does not have to be committedtoan absolutelypresuppositionlessstartingpoint)can assertthenegationofthemetaphysicalthesis.And ifithappens that a consequenceof that negationis the assertionof a gapbetweenwhatwecanempiricallydetermineaboutitemsandwhatcan betrueabouttheminthemselves,thenit is appropriatetosaythatin a sense
our empiricalcognitivefacultyis likea tool,one whichreachesso farandonly so far.
One can,ofcourse,challengethespecificargumentsmadeinthe pro-cess ofdemonstratingtheallegedcontradictions,butanysuchobjectionisno longerone tothemereidea of cognitionas a tool butis ratheronlyanobjectionto a particularkind of justificationforone formofthisidea.One can also seeksomegeneralwayto oppose Kant's entireattemptto
make a meaningfuldistinctionbetweenwhatourcognitivefacultiescandetermineandwhatthingsareinthemselves,and infactthisis justwhatHegeldoes in hissecondmajorgroupofobjectionsto Kant's theoreticalphilosophy,thecritiqueof Kantianidealism.Thereisa strikingargumentalongthislinewhichcould also be consideredanothervariantofthethirdobjectionto thededuction,namely,thatifoneweretotryto asserton itsbasisthatall we have areprinciplesthatapplyonlytophenomena,thenthisbareassertionitselfwouldconflictwiththerestrictionthesis,forthe
assertionwould not be merelyabouthowthingsappearto our tools ofcognitionbutwouldratherbea claimaboutwhatisabsolutelythecase."The Kantian'sreplyto thischargemustbe thatit was neverclaimedthatthereisnothingwe can know thatis absolutelytrue.Rather,thepointofKant's thesisis onlythatthe determinationofthingsrequiresintuitionandthatthepredications(andall thatrestsonthem)oftheintuitivequali-tiesgivenvia our formsofsensibilitycannotcharacterizethingsinthem-
distinctionbetweenanabsolutetruthand a truthof someothersort. .69
Herethebasicobjectionis thateithertheCritiqueleaves uswitha "baredirection,"an "emptyplace,"or itshouldadmitto a knowledgewhichis"itselfsomethingreal."Note, however,thatforall thathas beenexplicitlyargued,theknowledgewhichsupposedlymustbeadmitted,andadmittedas anembarrassmentto Kant,is notsaidto bethegeneralclaim"weknowonlytheformof our'tool,'the'law ofrefraction,'not thethinginitself,"butis ratherthesetof claimsmadethroughthistool. Inotherwords,itisbeingurgedthat theparticularcontentof our sensibleknowledge(the"rayitself,"or "knowledgebyitselfand- supposedly- cutoffformtheAbsolute")mustbe admittedas a partof theAbsolute,i.e., as a partof
whatis absolutelytrue.Byitselfsuchurgingishardlypersuasive,fortheaim oftheAntinomiesispreciselytoargueindetailagainstit.It is mostlikely,though,thatHegeldid not meanhis claimto be persuasivehere,70foritoccursinwhat isonlya briefIntroduction,and that theissue is to be settledonlywhenHegel considerstheAntinomiesitselfmoredirectlyand systematically.We will focuson theseconsiderationsin the next section,but firstitshouldbe madeclearhow muchrestson them.The factisthatalthough
Hegelrepeatedhisgeneralattackon Kantmanytimesand thereforehadampleopportunitytofleshout hisgroundsforrejectingthecriticalphilos-ophytoutcourt,he neverofferedanythingmorerigorousat this levelthan the passage just examined. The talk of "tools," "rays," and"swimming"is aboutall thatcanbe foundapartfromwhatis inthesec-tionsdealingspecificallywiththeAntinomies.
new line of argument,but it is not strongenoughto bear verymuchweight.Theargumentis thatthemereadmission(whichKantiscertainlywillingtomake)thatthereareempiricaltruthscountsagainsttherestric-tionthesisand thephenomenal/noumenaldistinctiontiedupwithit.Sup-posedly,thedistinctionbetween"empiricaltruth"and "truthofanotherkind"isonlyso much"randomtalk,""wordswhichpresupposea mean-ingthathas to begotat."7' This isnota pettyissue,forsurelytheKantianmustsaysomethingaboutthedifferentkindsof truthswhichhe consid-
noumenalmaterialtruthsaboutthecharacterofthingsin themselves.ButI see noproblemhereas longas Kantgives,as he surelydoes,someindica-tionabouthow thesetruthsare to be distinguished,suchas bydifferentproceduresof verification(respectively:transcendentalarguments,for-
mal logic,science,intellectualintuition),and as long as it is clearthatthesedistinctionsneednotbe meanttoimplythattruthitselfis ofdifferentkinds(ratherthanjust that thereare differentkinds of itemsthatare
opposedto thatofknowledgeor itsobjects,72butthereis no reasonthisshould amountto a weakness specificto criticalphilosophy.At most,Hegelcan chideKant fornot beingquiteopen enoughaboutthelimita-
tionsof"empiricaltruth,"sinceforKantitis afterall thecasethatwhatisempiricallytrueofx,e.g.,thathisexistenceistemporallylimited,doesnotentail,as onemightsuppose,a correspondentorbasing (ina Leibnizianway)facttrueofx in itself.On thecontrary,thenoumenaltruthaboutx,e.g.,thathisexistenceiswithouttemporallimits,can be quitetheoppositeofwhat we wouldempiricallyassertand of whatanyonecouldanalyzeout of suchassertions.ButthisjustmeansthatforKant our empirical
truthis truthonlyin a particularirreduciblebut limitedcontext,73andthisshouldnotbe a thoughtwhich,inour linguisticage,mustbeheld toputKant intospecialdifficulty.74Thus,evenifinsomedeepsense"truthis one," thereis, forall Hegel says,stillan understandablepointandmeaningtothedistinctionsbetweenthedifferentkindsof truthsthattheKantianwantsto introduce;it is hardlyall "randomtalk."75
V. As weturnnowtoassessHegel'ssecondgroupofobjections,hisrejec-
tionofKant's transcendentalidealism,it mustbe admittedthatthishasalreadybeeninviewforsometimebecauseof the way Hegel tendedtoconnectitwithhis firstgroupofobjectionstoKant'stheoreticalphiloso-phy.Nonetheless,we haveyetto look at Hegel'smoredirectattackson
72 See Gerold Prauss, "Zum Wahrheitsproblembei Kant," Kant-Studien60 (1969):
theKantiannotionof thethinginitselfandon thespecificargumentsoftheTranscendentalDialectic,whereinKant triestocompletehis case forthe restrictionthesis.
The generalperspectivefromwhichall these attacksstartis Hegel's
ownabsoluteidealism,whichwas discussedearlieras beingremarkablylikea highlyconfidentrealism,a beliefinthein principletransparencyofall realityto ourrationalfaculty.76Ofcourse,Hegelalso buildsintothisidealisma setofteleologicaldetailswhicharenotso easytoaccept,fortheworldis saidnecessarilyto be notonlyopentoreason,butalso so opentoit thatitis justas ifa supernaturalIdea, inthefullestprovidentialsense,generatedrealityandall itsbasicforms(a levelofcontingencyis allowedbyHegel,butthatthereis sucha levelissomethingthatis itselfduetothe
Idea). Importantas themoreexuberantaspectsofidealismwereto Hegel,theywillnotconcernus,forourquestionis simplywhetherthereis anyminimalway in whichHegel demonstratesthat the restrictionsKantimposeson reason mustbe rejected.
Unfortunately,it isnot alwayseasytokeep aparttheminimaland theexuberantsidesofHegel's critique.Forexample,intryingto embarrassthe Kantian aboutthestatusofempiricaltruths,whatHegel eventually
wantstoemphasizeisthathe(Hegel)has thebestexplanationofhow"therayitself,"thedomainofappearances,has a kindoftruthwhichispartoftheAbsolute.Thatis,forHegelthetruthofappearances(and,ina sense,their"untruth")is theirverynatureas mereappearances,thefactthat,asmerelysensibleand finite,77theyrequirean ultimateexplanationin a
76 To thisextentHegelisremarkablyclose to whatMichael DummettcallsFrege'sobjectiv-ism: "In sayingthatwhat is objectiveis not independentof reason,Fregedoes not mean
thatitsexistencedependson ourthinking. . . he meansthatitcannotbeapprehendedsave by,or by referenceto, rationalthought"("Objectivityand Realityin Lotze andFrege,"Inquiry25 [Ig82], p. i i i). Frege'sown languageis quiteHegelian: "By objectiv-
ityI understandan independencefromoursensation,intuition,ideation . . . but notindependencefromthereason,forto answerthequestionwhatthingsareindependentof
reasonwould be as muchas to judgewithoutjudging,to washthefurwithoutgettingit
wet," cited byDummett,ibid., p. i IO, fromDie GrundlagenderArithmetik(Breslau,
I884), ?z267 Note.thata mathematicallyinfinitemagnitude,such as thedomainofevennumbers,is
stillmetaphysicallyfiniteforHegel,sinceithas an essentialreferencetoanother,inthis
case thedomainof odd numbers.For Hegel, to be trulyinfiniteis to be whollyself-
grounded.Thus with Hegel liststheinfiniteitemsthathe insists(contraKant)can be
termsatEncyclopedia,?45,VIII: i z i, Wallace,p. 72, and V: 5z, JohnstonandStruthers,vol, I, p. 67. For a finetreatmentofHegel'sviewoftheinfinite,seePaulGuyer,"Hegel,
Leibniz, and the Contradictionin the Finite," Philosophyand Phenomenological
self-groundednotionofreason."8Appearanceshave truthinsofaras theyaregroundedinreasonin a waythatcanbe revealedeitherfromthebot-tomup,as inthePhenomenology,wherelimitedconceptionsof limitedthingsleadus toan allembracingexplanation,or fromthetopdown,as in
theLogic,wherethesequenceof basiclogicalcategoriesunfoldsin suchawayas todisclosethestructureoftheempiricalworld.For theKantian,ontheotherhand,thegapbetweenappearancesandthingsinthemselvesis absolute,forinneitherdirectionarewe able to derivethe(particularpositive)qualitiesoftheone fromthoseoftheother.Similarly,thetruthoftheempiricallies notin itsbeinggroundedin an ultimatelyfinalexplana-tion,a knownIdea,but ratherin an internalconsistencywiththelimited(thougha prioristructured)proceduresof our empiricalinquires.
WhatthismeansisthatapartfromthegeneraloppositiontoKant'sdis-tinctionoftruthlevels,reviewedintheprevioussection,Hegel'sultimateobjectiontoKantrestson a veryelaborateclaimabout howappearancesandrealityentaileach other.Thereis no wayto testthisclaimnow,and soI willadmitto be challengingHegelonlywithrespectto thoseaspectsofhiscritiquethatare detachablefromthedetailsofhisownpositivemeta-physicalsystem.Sucha challengewillno doubtstrikean orthodoxHegel-
ian as incompleteandunfair,butI willassume thereareunderstandablereasonsforit,and thatifa Hegelianis reallytohope somehowto breakinto contemporarydiscussionsin theoreticalphilosophyit must be bysomeargumentsthatdo notalreadyrequireappealtothe(moreelaboratepartsoftheHegelian)system.Moreover,I presumeHegelhimselfwouldnothavetoregardmytacticaswhollyunfair,for,as ourmanyquotationsfromtheearlyJenapublicationshaveshown,HegelclearlyformulatedhisbasiccritiqueofKantbeforedevelopinghisownsystem,and(as theparal-
lel citationsfrommuch later texts show) what he formulatedthenremainedto theend the coreofhis directcritique.
Thereremainto be examinedthreebasicaspectsofthiscritique:(A) the
generalattackon theconceptionofa thinginitself;(B) thegeneraltreat-mentofKant's basicargumentforidealismintheAntinomies;and (C) the
specifictreatmentofindividualdoctrinesintheTranscendentalDialecticthataredevelopedin line withtranscendentalidealism.
Thefirst-kindofobjectioncomesinvariousforms,buttheseall appearto reston thepresumptionthatKant'scategoriesareto be equatedwiththeirsensiblemeaning.79Thus therestrictionthesisis readin sucha way
78 See Hegel,Encyclopedia?13i, Zusatz,VIII: z6zf,Wallace, p. i87.
79 It is notablethatanotherwayinwhichHegelexpressestheprincipleoftrueidealismisin
termsof theidea thatrelateditemshaveno senseinisolation,and hetakesas an expres-
sionof thisviewKant's doctrinethatconceptsand intuitionsalone are eachinsufficient
as to implythatsincewe cannotapplythecategoriesbeyondthesensibleand phenomenalconditionsspelledoutintheSchematismandPrinciples,thedomainof thingsin themselvesiswholly,and absurdly,blockedfromour thought:"realityis absolutelybeyondtheNotion,"" "theobject,as
a thingin itself,simplyremainsa somethingbeyondthought."8iTimeand again Hegel assertsthat the essence of Kant's idealism is its"subjectivity,"its wholly limitingthecategoriesto our mind,so thatwhereasforJacobithecategoriesarelimitedbecause theyapplyto a con-ditionedand finitedomain,forKant the limitationcomesdirectlyfromtheirsourcein a merelypsychologicalconceptoftheself.8,
The traditionalKantianwouldreplythatsuchchargeswhollymisstheessentialdistinctionbetweenthepureandtheschematizedmeaningofthe
categories.On thebasis ofthisdistinctiontheKantiancan saythat(contraJacobi)whatgoesbeyondthesensibleisnota whollyamorphousdomainbutrathersomethingwhichcanbeallowedsomesortofconceptualorder.This orderis one thatholdsforallthinkers,anditcanevenbemadedeter-minateby us as longas wehaveanothertypeofdatathanthespatiotem-poralto makeuseof,as in factoccurswithourmoralfaculty.83So evenifschematizedcategories,suchas thosefoundin our ruleoftemporalcaus-
ality,arenotto beappliedtothingsinthemselves,suchthingsneednotbewhollybeyondthought,and in factKantbelieveswe mustthinktheminaccord witha kind ofpure,moral causality.
Hegel'sobjectioncan also be counteredby notingagainthathemisses
mentwhichdoesnotimmediatelyhavesuchlimits,extraconsiderationsareneededtogettotherestrictionthesis.Werethisnotso,therewouldbeno needforhisfocuson thefactthatour use ofthecategoriesinvolvespar-ticularformsofsensibility,orfortheargumentoftheAntinomiesaboutthespecificlimitationsoftheseforms.84
seemsto havemissedthe factthatforKantconceptscan have somemeaningevenif,
withoutintuition,theyare "empty."See J.Nolan, "Kant on Meaning:Two Studies,"
clopedia,56z, VIII: 149, Wallace, p. 96; XVIII: 318; XX: 322, 350.83 See above, n. 79, and Kants gesammelteSchriften,V: 484-85, VIII: 136.84 It is truethatKantnotes his deductionassumes the passivenatureof our intuition(B
I45), and so isnot claimedtoapplyforknowerssuchas God,whomightcognizethings
Hegelis notunawareofthisKantianresponse,forhe triestoblockitbyremarking,"it is no escape to turnaroundand explainthatreasonfallsinto contradictionsonly byapplyingthecategories."8"This suggestsanacceptancethattheKantainneednot believethatthingsinthemselvesare
directlyunreasonable,and thereforefully"beyondthought."Nonethe-less,Hegeldoes not allow thispointto makeanydifference,forhe addsimmediately,"applicationofthecategoriesismaintainedtobenecessary,and reasonisnotsupposedto be equippedwithanyotherformsbutcat-egoriesforthe purposeof cognition.But cognitionis determininganddeterminatethinking:so thatifreasonbe merelyemptyindeterminatethought,it thinksnothing."86This chargeinvolvesseriousconfusions.Hegelappearsto conflatethefactthat(forKant)theapplicabilityofthecategoriesis necessaryinthatit is indispensableforourempiricalcogni-tion(seequoteabove),withtheidea thatsomehowwithoutsuchanappli-cationthecategoriesthemselveswouldbe senseless.Thereisno basisforconcludingthat withoutapplyingthe categoriesreason is absolutelyempty,it "thinksnothing."WhenI considerGod as the groundoftheworld,I do notcognizehimin accord withtheprincipleof theSecondAnalogy,I do notapplythecategoryof causalityinitsschematizedmean-
ing,andyetI stillhavesomethingotherthana "merelyempty"represen-tation.In particular,I can be thinking- perhapswithoutjustification,but stillwithsense- thatsomehowGod standsin at leastthatgeneralrelationto the worldinwhichanynecessaryreal conditionofa thing'sbeingstandsto thatbeing.
But this charge,just like Hegel's own,restson a refusalto distinguishbetweenpureandempiricalmeaningsofthecategories.Aslongas thelat-ter (the temporalmeaningofcausality)is notapplied to the noumenal,and as longas the formeris not used to makeanydeterminateclaims
about particularobjects,thereisno directcontradictionof Kant's princi-ples.88
In a second textHegel makesa claimthatalso leads intothesecondofhisobjectionsto Kant's idealism,theobjectionsconcerningthegeneralargumentof theAntinomies.On Hegel's readingthe argumentis that"insofaras itis determinedbythecategories,theinfiniteentanglesitselfincontradictions."89On thisreading,therestrictionsKant is layingdownforpurereasonarerequiredbyreasonall by itself,and the specificfea-turesofspaceand timebecomewhollysecondary.Ifoneacceptsthisread-ing,thenit is naturalto say (as inHegel's objectionto Kant's idealism)thatforKantobjectsmustbe absolutelybeyondthought,thatthingsinthemselvesmustbe above thecontradictoryfacultiesofthemind.It thenbecomesnaturaltoobject,asHegeldoes repeatedly,thatKantis beingtoohardonreason,90foritdoes seemtoo muchto thinkthatreasonitselfisacontradictoryfaculty.
This objectionis unfairbecauseKant's Antinomiesis not about thecontradictionsof reasonas such,theentanglementsthat arise fromthemereconceptoftheinfinite.On thecontrary,thechapteris about how theinfiniteor unconditional(supposedly)cannotbe determinedto be presentintheempiricalrealm.Wheresuchdeterminationisnotconsideredneces-sary,wherethereis nothingdemonstrablyinconsistentabout an endless(and supposedlyindeterminate)seriesofcomplex,ordependent,orcon-tingentitems,Kantproposesit is enoughto say thattheremay be an
unconditionedfoundina realmofthingsinthemselves.Thus theremaybe bothsimple,free,andnecessarilyexistentnoumenalbeings,as wellascomplex,determinedandcontingentsensiblecharacters.9'None of these
Theory,pp. z84-86, and M. Westphal,"In Defenseof theThingin Itself,"Kant-Studien
59 (i968), pp. 135-41.89
Hegel,XX: 3 53. Cf.V: 39, Johnstonand Struthers,vol, I, p. 56, whereKantis attackedfortheerrorofsupposingit is reasonthatis in contradictionwithitself,and fornotreal-
topicsare taken to forceone to transcendentalidealism:theysimplyencourageit as an optionto be keptin mind.On theotherhand,whenKant discussesspaceand time,thereis no roomevento entertaina sepa-ratenoumenalcharacter.WhereastheThirdAntinomy,forexample,can
introducetheidea ofnoumenalfreedom,thereis no idea of a separatenoumenalfinitudeorinfinitudeof thespatiotemporaldimensionsof theworldfortheFirstAntinomyto encourage.
In his basic argumentKantpresumesthatifspatiotemporalrelationsweretranscendentallyreal,thenthe worldwould haveto have eitheradeterminateand infinitelyor a determinateand finitespatiotemporalmagnitude.93The particularargumentsof the First Antinomyareadvanced to show that neitherof these disjunctsare possible, andpreciselyratherthanassertthatreason itselfis contradictory,Kant con-cludes onlythatthepremiseoftranscendentalrealismmustbe givenup(unfortunatelyhe does notconsidergivingupinsteadwhathepresumestofollowfromthispremise,i.e.,thesupposedlyexhaustivedisjuncts).Noneofthestepsto thisconclusion,questionableas theymay be, invoketheclaimthatthe notionofthe infiniteitselfentanglesus incontradictions,andKant evenseemswillingto allowan infinitefuture.Thereis no asser-
tionofa singleandbasic difficultywithreason;rather,thereisa sequenceofverydifferentconsiderationsabout space and timewhichsupposedlylead to conclusionsaboutthe world's dimensionsthatin turninvolveacertainlimitedrestrictionof reason.
Hegelneveracknowledgedtheroleand structureofthebasicargumentof theFirstAntinomy.In a letterof I 8I 2, heexplainsthatas faras he isconcernedthe contentofthe Antinomiesshouldbe explicatedlogicallywithoutanyreferenceto cosmology:"indeed,all furthercontentabout
theworld,matter,etc.,isa uselessballast,a confusingimage (Nebelbild)of theimagination."94The sameviewis expressedin theLogic, whereHegel makesno basic distinctionbetweenthe various antinomiesandassertsthat all conceptsof the sensibleworldinvolvecontradictionsbecausetheyare aboutmereparticulars,notbecausetheyareconceptsofreason.95His casual attitudeto the textmayagainexplain whyHegel'ssecondobjectiontoKant'sidealism,likehisfirst,appliesonlyagainstan
thatofan internallyincoherentreason- whichis not at all a partof thatidealism.
VI. Similarproblemsariseas we move fromHegel'sgeneraltreatmentoftheAntinomiesto histhirdand lastgroupof objectionstoKant's ideal-ism,thosewhichconcerntheindividualtreatmentof thespecifictopicsoftheDialectic.In each case Hegel urgesthatKant has failedto be idealisticin the genuinesensein thathe has remainedfixedon a choicebetweenabstractlyopposed alternativesand has overlooked furtheroptionswithintheknowablerealm.Expressedinthesegeneraltermshisobjectionpointsto whatis no doubta propercriticismofKant,andyetas onelooksmorecloselyatexactlyhowHegelworksout hisown critique,itbecomes
more difficultto approveofit.Hegel's tendenciesare alreadyrevealedin his earlyessay, FaithandKnowledge,which givesan overviewof the entireKantianphilosophyandemphasizeshowKanthas misseda "positive"mediatingsolutiontothe Antinomies.6 The FirstAntinomyis describedas arisingsimplybecause"being-otherispositedas wellas being,i.e.,thecontradictioninitsabsoluteinsuperability.Henceonesideoftheantinomymustconsistinpositingthedeterminatepoint,andtherefutationinpositingtheopposite,
thebeing-other. . ."97This is an ellipticalwayofsayingthatwe seemforcedtoposita firstpointofspaceandtime,andthenalso a pointbeyondthat.Nothingis said hereto challengethespecificsof theroutetakentothisdilemma.Instead,Hegel again expresseshisgeneraldisappointmentwithKant'sgoingon to insiston transcendentalidealism,andhestressesthatKant'sargumentsare difficulttoreconcilewithhislatertalkaboutan"endlessprogress"in ourmorallife.98Correctas he is on thelatterpoint,
Hegel's discussion adds nothingto solve the particulartheoreticalproblemat stake.Afewyearslater,whileteachinginNuremberg,Hegelprovidedhisstu-
dentswith a moredetailedtreatmentoftheAntinomies,whichat leastincludesa synopsisofthevariousargumentsas wellas theclaimthattheproblemcan be resolvedbecause theconceptsofa limitedandan endlesstime(and space)eachinvolvetheother,forthelimitedis whatcanbe tran-scended,and theendlessiswhat transcendslimits.99In contrastto these
poralas suchbutratherreason,whose reflexivity"penetratestheconceptand essenceof the world.""'0This is, of course,one ofHegel's favoriteideas- thatreflectionhas thepeculiarkindofendlessnesssuggestedbytheformofa circle,everturningupon itself,whereasthenon-spiritualhas
a merelylinearform(and "bad infinity").But whateverthevalueofthisidea, it too simplybypassestheoriginalproblem,the factualquestionoftheworld'sphysicaldimensions,whichremainsevenifthereis thekindofnecessaryrelationHegel stressesbetweentheconceptsofthelimitedandthe endless.
Similarlyfrustratingis the treatmentin theEncyclopedia,wheretheFirstAntinomyis describedsimplyas implying"recognitionof thedoc-trinethatspace and timepresenta discreteas well as a continuousaspect . . . we can go beyond everydefinitespace and beyond everydefinitetime,butitis nolesscorrectthatspaceandtimearerealandactualonlywhentheyaredefinedorspecializedinto 'here' and 'now'.""'0 Herethesubjecthaschangedagain- and in a thirdnewway- forthereis noreasonto presumethatanypartyto theAntinomywouldhave to disputethe "discreteness"oftimeandspaceinthewaythatit is introducedhere.The questionatissue,though,ishow largetheworldis,andto thisHegel
is givingeithera strangedogmaticanswer,namely,thatwe reallycangobeyond "everydefinite"point,or no answerat all. In his finallecturesHegel takes the same ambivalentline.He repeatsthepoints made inNurembergandconcludesthat"theworld,as theuniverse,is thewhole;itis thusa universalIdea,andthereforecan bedeterminedeitheras limitedor unlimited.̀02
VII. The distinctiverole of the FirstAntinomymakesHegel's indirect
approachtoitespeciallydisturbing,and all themoreso becausehemightwell havemadespecificobjections(e.g.,to theexhaustivenessofthebasicdisjunctionKant poses) that would have blocked Kant's argumenttotranscendentalidealism.Hegel's treatmentof theotherAntinomiesisevenbriefer,and the evaluationof the treatmentis even moredifficult.SinceHegelbelievesthattheargumentof the SecondAntinomyis basi-callythe sameas thatoftheFirst,'03and thatthatof theFourthis basi-callythesame as that of theThird,'04we shall examineonlyhis treat-
mentof the Third Antinomy.Hegel's basicpositionis that"freedomandnecessity,as understoodby
abstractthinkers,are notindependentlyreal,as thesethinkerssuppose,butmerelyideal factors(moments)in the truefreedom."'"5Thismight
be readas anexpressionofthetraditionalcompatibilistposition,andonemightimagineHegel pressingagainstKantthechargethatthewholeneedand motivationfortheprojectoftryingto cleara wayfortranscendentalidealism disappearsif therecan be an account of humanaction andmoralitywhichallowsmetaphysicalnecessity.Thus,onemightimagineHegelchallengingKantby pointingout how littleKanthad saiddirectlyagainstsucha position,andhowmuchHegelandothershad saidintheiraccounts of actionand moralitywithoutinvokingtranscendentalfree-dom.106In fact,however,Hegeldoes not take thisroute,and insteadhedeclaresagainthatboththesisandantithesiscanbetrue,thatboth(abso-lute) freedomand necessityhold, thoughonlyas part of a largertruth.
PreciselywhatHegel has inmindhereis difficultto determine,forhedoes notoptfortheKantiansolution,thatwhatcan hold isbothtranscen-dentalfreedomandphenomenalnecessity,nor,I believe,forthecompat-ibilistpositionthata lackof transcendentalfreedomstillleaves a merely
phenomenalbutadequatekindoffreedom(viz.,rationalself-determina-tion)thatis consistentwithabsolutenecessity.It istruethatsometimesitappearsthatallHegelhas in mindbyour freedomistheordinaryself-de-terminationofthecompatibilist,or ourparticipationinthebroaderself-determinationof the rationalcourseofthe worldas a whole.'07In fact,however,he also implies(as a partofhisgeneralstrategyofadoptingasynthesisthatincludesbothof theopposingtheses)thatwe do havetheabsolutefreedomwhich concernstheKantian,the "abilityto abstract
fromeverythingwhatever,"thenegativefreedomtobecompletelyunde-termined(in some contexts)by anythingoutsideone."08What Hegelargues is not that such freedomis unrequitedbut ratherthat it isinsufficient,that,as Kantindicated,wewantnotonlytheabilitytochoose
Wallace p. 55f.i06 See Ameriks,Kant's Theory,pp. i85-233.107 See thesketchfor"a proofthatthewillis free"inHegel's Philosophyof Right,?S4-5,
VII: 48-50, trans.T. M. Knox (Oxford:ClarendonPress,1952), pp. zi-zz.Io8 Ibid.I believethispointisnotstressedenoughinR. Schacht'sfineessay,"HegelonFree-
dom," in Hegel: A Collection of CriticalEssays, ed. A. MacIntyre (Garden City:Doubleday Anchor,1972), pp. 289-3z8. For example,Schachtsays,"The experienceoffreedomis takenbyHegel . . . to reflectthefactthatwhatarecommonlyreferredto as
the laws of naturedo not governthewhole of phenomenalreality"(ibid.,p. 301).
Schachtdoesnotpointouthowimportantthisclaim is and how nonethelessHegelfails
as anuncausedcausebut also theabilityto chooseinthiswayto followarationallaw.His disputewithKant arisesonlywiththeissueof howsucha law is to be understood,forHegel believesthatit can and must beexplainedin a muchlessformalwaythan Kantwould accept.
WhatisdisturbingaboutHegel's positionisthat,whilehe doesappearto allow thatabsoluteor "negativefreedom"is at leastpartof whatmanhas,hedoesnotputanyeffortintoarguingforthepossibilityofsuchfree-dom (he appearsto thinkitis obvious on introspection).'09Kant's doc-trineof transcendentalidealismis at leastaddressedtothisissue,and itatleasttakesseriouslythe deterministicimplicationsof scienceand sociallife.ThusKantusestheThirdAntinomyto introducetheideathat,despiteuniversalnaturalnecessity,theremaybe some uncausedcausality.Ofcoursehis argumentmaybefaulty,and itstillprovidesnotheoreticalrea-son to affirmwe have suchcausality,but at least it makesa starttowardtheargumentneeded. Hegel,on the otherhand,simplytriesto havethebestof bothworlds,thatis,he buildson themetaphysicaldoctrineof ourabsolutenegativefreedom,whileimpugningtraditionalmetaphysicsanddismissingall particularargumentsto thedoctrine.
ofstrategicsilenceabout thebasicissueoftheexistenceof absolutefree-dom.InthefirstJenawritings,forexample,hepresentsno detailsof thediscussionsin the Third Antinomyand simplydeclares,"Freedomandnecessityare idealfactors,and thusare notinreal opposition . . . Free-dom is the characterof the AbsolutewhentheAbsoluteis positedassomethinginner . . .Necessityis thecharacteroftheAbsoluteinsofarasit is observedas somethingexternal. . . Freewill (Willkfir)and acci-dent,whichhave a place onlyin subordinatestandpoints,are banished
fromtheconceptof thescienceoftheAbsolute.""0.The firstpartof thisstatementsidestepsthefundamentalpointthatin one sensefreedomandnecessitymustbe in "real opposition,"forto hold to transcendentalornegativefreedomisto hold to theexistenceofuncausedcausality,andtoholdto transcendentalorabsolutenecessityis to assertthedirectdenialofthis.The latterpartofHegel'sstatement,theremarksabout an arbitrary
lo9 Cf. above, n. 107, andHegel'sPhilosophyofMind(intheEncyclopedia),??473-48z,
X: 295-302, trans.W. Wallace and A. V. Miller (Oxford:ClarendonPress,I97I), pp.
234-40."1 II: io8, Surber,p. 83. InFaithandKnowledgeHegel does goon to reviewKant'sdiscus-
sionbriefly,but he reducesit to theidea that"whentheyarethoughtwithoutanycom-
munionat all freedomand necessitydo not conflict"(II: 320, Cerfand Harris,p. 84).
This isa veryinaccurateexpressionof Kant's position,sinceforKant thefreedomofthe
intelligiblecharacterhas responsibilityfor,andso is "in communionwith"thenecessity
of theempiricalcharacter.See A 55i/ B 579.
HEGEL S CRITIQUE OF KANT S THEORETICAL PHILOSOPHY 3 I
conceptofan absolutesciencehardlymeanstheycannot exist at all. ItseemsratherthatoncemoreHegel's pointisthatevenifthereissuchfree-dom,itbyitselfis not thehigheststate.
InlaterworkHegel expressesa slightlydifferentbutequallyundecisiveposition.InNuremberghe introducestheidea thatthe "truesolution"ofthe Antinomyis in therelationofcommunityor interaction(Wechsel-wirkung).Eachiteminsucha relationis a causeandinthatsenseissaidtobe free,"' and yeteach is also an effect,and in that sense is subjecttonecessity.But clearly,whethersuch a relationexistsor can be provedwouldnotbyitselfsettletheissue of humanfreedom.The fundamentalquestionremainsof whetherand on whatbasiswe can believein someuncausedcausalityon ourpart.IntheLogic,"' as intheLectureson theHistoryofPhilosophy,"3 no newideas areintroducedon thisissue,andHegel impliesthatthematterhas alreadybeendealtwithsatisfactorily.The argumentsof theAntinomyare dismissedas "proving"theirown
claimsonly by findingopposed views to be contraryto theseoriginalclaims.Thustheargumentsareheld to be entirelyquestionbegging:thedeterministdisprovesfreedomonlyby findingit inconsistentwithdeter-minism,andviceversa. HereHegel entirelymissestheunderlyingaimofKant's discussion,theexposureofthe dogmaticassumptionthatforthelaw ofcausalityto hold at all it musthold in a metaphysicallyabsolutesenseas a temporalrule.Oncethissuppositionisdropped,we areleftnotwithquestionbeggingclaims,butratherwithconsistentdemonstrations
(givenKant's Analytic)thatitwould be improperto asserttherecan be atemporalfreecausality,and improperto asserttherecannot be a non-temporalfreecausality.
Hegeldoes notconsider,letaloneuse,thissolution,nordoes he evershowpreciselyhowanythingshortofthissolutionwould besatisfactory.The mostnaturalformulationofhisownpositionis thathewantsto con-tinueto assertourfreedominthetraditionalsensebutsimplyhasno inter-
estin or meansforbackingthatassertion.A lesscharitablereadingwouldhavehimassertingthatthecontradictionoffreedomand necessityis notamereappearancearisingfromtheintroductionofdogmaticpremisesand
Hegel,IV: i90, Hegelspeakshereofan "absolutecausalityoffreedom."Cf.VI: z37-40,Johnstonand Struthers,vol. z, pp. 203-5.
themisapplicationof our reasonbutis rathera truthabout ournatureasfinitethings,i.e., justwhat one should expectfromitemswhichare onlyfinite."I4I willpresume,however,thatwhenHegel seemsto be talkinginthiswaywe aretosaynotthathe isascribinga literallycontradictoryrela-
tionto things,but ratherthathe has inmindsomekindoflesserand atleastcomprehensiblecontrastlikethatbetweenwhathe calls thecontinu-ityand discretenessofspaceand time.Butofcoursethisis onlytosaythatatbestHegel'streatmentof theThirdAntinomyis no betterthanthatoftheFirst;itleavestheoriginalquestion,as well as Kant's treatmentof it,untouched.
VIII. A similarand finalexampleofHegel's treatmentof Kantcan be
foundinhis discussionoftheParalogisms.ThemainissuethatHegel dis-cusseshere,thesoul's simplicity,is treatedjustliketheissuesoftheAntin-omies (notinappropriately,forI suspectKantoriginallymeanttotreatitundertheSecondAntinomy).We areto see how bothsides of thedebatecan be affirmed,forthe soul is bothsimpleand complex: "Thus, forexample,whilethesoulmaybe admittedto be simpleself-sameness,it isat the sametimeactiveand institutesdistinctionsin itsown nature.""15
Here againwe finda shiftfromtheoriginalissue,forneitherpartyto the
traditional question need dispute that the soul involves some"distinctions."The questionis ratherwhetherit is a simplebeingor acomplexofbeings;whicheveritis,it can involve"distinction"and neednotbe what Hegel calls "a meredead thing."To talkaboutthe soul assimpleandsomethingelsebesides,is to ignorediscussinghow we aretoresolvethe originalquestionofwhetherthesoul is simpleat all.
In the same paragraph in which he appears to encourage the
assumptionofthesoul's simplicity,Hegelalso seemsto saythatsuchanassumptionisfalse,that"predicateslikesimplicity,permanence,etc.,areinapplicableto the soul.""6 Taken literally,thisimplies Hegel thinksthatwhatissimpleandcomplexis not evensimple,whichisofcourseanabsurdthought.Taken morecharitably,hispositionmustbethatwhatissimpleand complexin a sense,is notmerelysimple,i.e., is not "a deadthing,"andin thatsensealoneisnottobe saidtobe simple.Butagain,itwould seemthereis noneedto makethispoint,fornopartyto theParalo-
gismholdssucha "dead" view.In fact,however,Hegel ascribedsome-thingverylikethispositionto Kant.Hegel givesthename"dogmaticsub-jectivism"to the Kantianpositionwhichsupposedlyholds thatthe soul(qua simple)is onlya form,in contrastto the rationalistpositionof
"dogmaticobjectivism,"which holdsthat the soul can be describedintermsoftranscendentobjective(materialas opposed to merelyformal)predicates."I7LaterHegel impliedthatKant's viewis thatiftheselfis nota meresensiblething,thenit is notsomethingactual atall."8ThusHegel
explainsthatwhereasKantthoughtthe terms"being,thing,substance"are "too high"fortheself,on his own viewtheyare "too low.""9
Hegel's critiqueofKantgivessomesenseto hiscommentsabout thesoul,but thecritiqueitselfis veryquestionable.Inparticular,(a) Hegelimpliesthat Kant denies the existenceand substantialityofthe soul,"'2somethingwhich Kantdoes notdo (whatKantdoes is criticizecertainwaysofarguingaboutthesoul); and (b) Hegel failsto explain on whatbasishe means to defendsuchassertions,giventhosewho argue (partlyundertheinfluenceofKant) that thesoul mightbe a merecomplex ofmaterialpartswithrationalfunctions.Once again,inthecontextofurg-ingus to go beyondtwo opposed and limitedpositions- thatthesoul is amerething,justlikeanyother,and thatitisa mereform,nota thingatall
to a reconcilinghigherview,Hegelwhollyfailsto focuson the argu-mentsbehindeitheroftheoriginalpositions(thatthesoulissimple,oritisnot) and fallsintoa dogmaticacceptanceofone ofthem(simplicity).
In one importantrespectHegel's handlingof the ParalogismsdiffersfromhistreatmentoftherestoftheDialectic,forthischapterhas a differ-entrelationto thedoctrineoftranscendentalidealism.IntheAntinomies,Hegel'sfailuretoread Kantcloselyhad as itmostimportantconsequencethatitlefthimwithouta directcountertotranscendentalidealism.IntheParalogisms,Hegel'sfailurehas onlytheimmediateconsequencethathemisperceivesthecomplexityofKant's position,and so fails to see thedepthof Kant'scritiqueof thetradition(a critiquethatwouldalso apply
to Hegel) as well as the tolerancewithinKant's positiveview (whichallowsthe selfmayhaveall thecomplexityHegel asserts).Buta furtherconsequenceofHegel's approachhereis thathe missestheepistemologi-cal complementto Kant's critiqueof the metaphysicaldoctrinesofrationalpsychology.Kant'sParalogismsimplynotonlya questioningofcertainkindsofspecificclaimsaboutthesoul,butalso a challengeto the
17 Hegel, II: 3I9, Cerfand Harris,p. 83.
xx8Hegel, XX: 355.
91Hegel, XX: 356.
? See Hegel, Encyclopedia,547, VIII: Iz5, Wallace, p. 75: "That the soul cannot be
basic methodonwhichsuchclaimswerebased,namely,via an appeal toan allegedlyprivilegedinnerperception.The mainlocusofthischallengeis in thedoctrineof apperceptionand the Refutationof Idealism,whereKant arguesthatour self-knowledgenot onlyrequiresjudgmentand gen-
eral rules but also is parasiticupon spatial data and externalobjects.Hegelpracticallyignoresthisargument,and whenhedoes cometo ithe.seriouslymisrepresentsit,sayingthatitspointis thattemporaldetermina-tion requiresreferenceto somethingpermanentin oneself."' ThusHegelmissesKant's mostsignificantchallengeto themetaphysicaltradi-tion,thechallengetoexplainhowmereself-reflectioncan giveus determi-nate knowledgewithoutreferenceto a publicrealm.Hegeldid challengetheCartesiantraditioninmanyways,buthe oftenremainedboundtoits
presumptionthatself-perception(as in themetaphysicalarticulationoftheformsofGeist)has a specialprivilege.In thisrespecthis treatmentoftheParalogismsinvolveshismostseriousunderestimateofKant's work,and inthiswayhiscritiqueof the Dialecticsuffersfromthe sameweak-nessas thatof hiscritiqueof theAnalytic:an overconfidenceaboutwhatis obviousfromthemererepresentationoftheI.
IX. In describingand evaluatingHegel's attackon Kant,I have triedto
presentthe simplestand moststraightforwardKantianresponses.Theseresponsesshould be familiar,butrecentdiscussionsindicatethatit hasnotbeenappreciatedhowfartheygoto meetthepointsinHegel's explicitcritique.Ifthispaperdoes no morethanprovokea Hegeliancounterat-tackthatattemptsto meettheseresponsesin a comparablystraightfor-wardway,I wouldconsiderthisworkquite justified.In themeantime,aKantiancannothelpbut draw attentionto the obviousimplicationsfor