Liberalism and Its Discontents

Author(s): Raymond Geuss

Source: Political Theory, Vol. 30, No. 3 (Jun., 2002), pp. 320-338
Published by: Sage Publications, Inc.
Stable URL: http://www.jstor.org/stable/3072588
Accessed: 23/02/2010 12:04

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=sage.

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].

Sage Publications, Inc. is collaborating with JSTOR to digitize, preserve and extend access to Political Theory.

http://www.jstor.org
 LIBERALISM AND ITS DISCONTENTS

RAYMONDGEUSS
Universityof Cambridge

Agents in contemporaryWesternsocieties find themselvesin an odd situ-

ation. On the one hand,we seem to have no realisticalternativeto liberalism;
thatis, we know of no otherapproachto humansociety andpolitics thatis at
the same time as theoreticallyrich andcomprehensiveas liberalismand also
even remotely as morally acceptableto wide sections of the populationin
Westernsocieties, as they arenow in fact constituted.'Liberalideaspermeate
oursocial worldandoureverydayexpectationsabouthow people andinstitu-
tions will and oughtto act; they constitutethe final frameworkwithin which
ourpoliticalthinkingmoves. Primafacie nonliberalforms of habitualbelief,
such as those associatedwith certainreligions,formsof nationalism,residual
class enmities, and so on, still, of course, exist, but they seem to be, at best,
isolated and localised foreign bodies in a universe,the overall structureof
which is essentially liberal;in societies that are or are aspiringto be 'West-
ern', even these nonliberalideologicalfragmentssometimesadoptprotective
colourationin the form of the best veneer of compatibilitywith liberalism
they can muster.
On the otherhand,there are signs of a significanttheoretical,moral,and
political disaffection with some aspects of liberalism.Liberalismhas for a
long time seemedto lack muchinspirationalpotential;it is good at dissolving

AUTHOR'SNOTE:Thistextis a revisedversionofan article I wrotein Germanandpublishedin

December2001 in the DeutscheZeitschriftfir Philosophieunderthe title 'Das Unbehagenam
Liberalismus'.The original Germanarticle in turn was the content of a series of three talks I
gave at the Universityof Saarbriickenin December 2000. My thanks to Professor Wilfried
Hinschof Saarbriickenforthe kindinvitationto speak there,and also to the colleagues in Cam-
bridge with whomI have discussed this topic mostfrequently,John Dunn, Zeev Emmerich,and
QuentinSkinnerI also owe a greatdebtof gratitudeto Hilary Gaskinand to the two anonymous
readersfor thisjournal who helped me to correctseveral mistakesand significantlyimprovethe
original Germanversion of the essav.
POLITICALTHEORY,Vol. 30 No. 3, June 2002 320-338
? 2002 Sage Publications
320
 Geuss / LIBERALISMAND ITS DISCONTENTS 321

traditionalmodes of life andtheirassociatedvalues, but less obviously good

at replacingthem with anythingparticularlydistinctiveor admirable.2It fits
all too comfortablywith some of the more ignoble aspects of commercial
society. What contributioncould liberalism conceivably make to thinking
about the general degradationof the planetaryenvironment?Liberalideals
like individualism,toleration,or limitationof statepower,seem eithershort-
sightedly confused or mere covers for hegemonic designs. As the Harvard
political scientistSamuelHuntingdonnotoriouslywrote, 'Whatis universal-
ism to the Westis imperialismto the rest'.3Oldercriticismsof liberalismhave
also lost none of theirpower and plausibility:thatit has no clear remedyfor
poverty,for reprehensibleforms of inequalityof power,of conditionsof life,
and so on. To the extentto which liberalismis committedto the principlesof
individualinitiativeandthe defence of privateproperty,it is hardto avoidthe
suspicionthatit is ratherpartof the problemthanpartof the solution.Politi-
cal theories, however, which, like liberalism, are deeply anchored in the
social institutions,the mentality,and the form of life of large and wealthy
populations cannot easily be shifted by even the most vigorous forms of
intentionalhumanaction. This inertiaeven in the face of massive andtelling
criticismis not merely the disreputableresultof the brutepower of the past;
rather,in an uncertain,dangerous,and unpredictableworld there are good
general reasons not to embarkon radicalchanges in one's social formation
unless one is forced to it by demonstrableoverwhelmingnecessity.
The title of this essay is modelled on that of a late essay by Freud.4For
Freud we modems are condemnedto suffer from culturalimperativesand
regulationsthat do not allow us to lead a biologically fulfilling life, but that
we are also not able simply to throwoff. 'Discontent'with civilisation is an
unavoidablefate, given the incompatibilitybetweenourbiology andthe nec-
essary demandsof any form of specifically humansociety, and Freudthinks
thatit is strictlyimpossibleto do awaywith it altogether;the best we can do is
try to mitigatesome of its worsteffects. In contrastto this, the discontentwe
feel with liberalismis of a differenttype, if only because we can be surethat
changes in the world aroundus, in our politics, our social arrangements,our
economic circumstances,or perhapssimply an improvementin our powers
of theoreticalimagination,will sooneror laterdissolve liberalismandrender
it as irrelevantto us as feudalismor theoriesof moralitybased on honour.In
the meantime,though,we are stuck with a political and social regime and a
set of associateddoctrineswhose deficiencies are palpable.
Historically,liberalismis aninventionof the nineteenthcentury.5'Liberal'
was originallya wordused to designatea politicalparty;it seems to havebeen
used for the first time in about 1810-11 to refer to a group in Spain whose
membersadvocateda limitationof the privilegesof the king andthe introduc-
322 POLITICALTHEORY/ June 2002

tion of a constitutionalmonarchyon the Britishmodel.Expost, a legitimising

prehistory of liberalism is constructed in which Spinoza, Locke,
Montesquieu,Adam Smith, and others are made to featureprominentlyas
theoreticalprecursors.Since at the latestthe middleof the nineteenthcentury,
then, 'liberalism'refersboth to a relativelyabstracttheoreticalstructure-a
collection of characteristicarguments,ideals, values, concepts-and to a
social reality,a politicalmovementthatis at leastpartiallyinstitutionalisedin
organisedparties.Janus-facedhistoricalphenomenaof this kindthatencom-
pass both conceptualor theoreticalelements and real social forces pose spe-
cial difficultiesfor traditionalforms of philosophy.Since its beginning,phi-
losophy has oriented itself primarily on the analysis and evaluation of
relativelywell-definedarguments,butthe strugglebetweencompetingpolit-
ical groupsis not a seminardiscussion.Questionsof definitionandof purely
theoreticalconsistencyareoftennotthe mostrelevantones to ask in politics.6

II

Classical liberalismis best understoodas a negativephenomenon,a reac-

tion againstcertainevents, theories,andsocial andpoliticaltendenciesin the
late eighteenthand early nineteenthcenturiesthatearly liberalsidentifiedas
especially dangerous.7In addition,this was a twofold reaction,a kind of war
on two fronts. In one direction, as it were, vis-a-vis the past, liberalism
opposes absolutismand also the cameralistidea that the state had the duty
andthe rightto carefor the positivewell-being of its membersin an extensive
sense. In the other direction,facing the future,classical liberalismstrongly
rejectsthe exaggeratedmoralisationof politics thatit sees as propagatedby
the FrenchRevolutionaries.The ideological precursorsof liberalismin the
eighteenthcenturywere staunchopponentsof the subordinationof politics to
theology,andto the extentto which an absolutistethics simply steps into the
place thatnow discreditedtheology once occupied in the political and social
spherewithoutchangingthe existing structures,it too becomes an appropri-
ate object of liberal criticism. Rousseau's theory of the republic as the
embodimentof a unitarygeneralwill opens up a highly insalubrious,specifi-
cally modernpossibility,thatof clothing political decisions with the mantle
of an unlimited,secularmoralauthority.Kant'sattemptto groundpolitics on
a nonnaturalistcategoricalethics is understoodby liberalsas a parallelphe-
nomenon and correspondingly condemned. Thus, for the early liberal
Benjamin Constant, Robespierre's 'republique de la vertu et de la terreur' is
a naturaloutcomeof takingRousseau'scentralconceptionsat face value,and
 Geuss / LIBERALISMAND ITS DISCONTENTS 323

absolutistethics of the Kantiantype is just another,slightly etiolatedversion

of the same basic position.8
There are four chief componentsof the classical liberalismof Constant,
Mill, andde Tocqueville.First,liberalsassign a high positive value to tolera-
tion, as the cardinalvirtueof humansocieties. This is the oldest layer in the
liberalsynthesis.Second, liberalsattributespecial normativeimportanceto a
particularkind of humanfreedom. Society should consist as much as possi-
ble of voluntaryrelationsbetweenpeople, andin particular,the free assentof
the membersis the only sourceof political authority.Third,liberalsarecom-
mittedto individualism:a society is good only to the extentto which the indi-
viduals in it are well off. Fourth,liberalismis characterisedby a particular
kind of anxiety,the fearof unlimited,concentrated,or arbitrarypower.Limi-
tation of such power is thus always a goal of liberalpolitics. These four ele-
ments constitute the political substanceof the traditionalliberalism of the
nineteenthcentury.
Since my intentionis to startfrom liberalismas a historicalphenomenon,
it is importantto try as far as possible to avoid anachronism,thatis, to avoid
narratingthe historyof liberalismfroman end-pointin the presentthatis pos-
itively valuedand assumedteleologically as the naturalgoal of the historical
process. Precisely this kind of anachronisticview seems to me to have
become increasingly common in late twentieth-centuryliberalism, espe-
cially underthe impactof the workof JohnRawls. Startingin the later 1950s,
Rawls's work gave impetus to a revival of political philosophy,a discipline
thathadbeen pronouncedmoribundby some of its most distinguishedpracti-
tioners a few years before, and his early achievement,especially as docu-
mentedby Theoryof Justice9led to a correspondingattemptto reinterpretthe
historyof liberalismretrospectivelyin the light of his position.This had some
peculiarresults, given that Theoryof Justice (and the associatedearly writ-
ings) representeda significantdeparturefromwhathadbeen the mainline of
liberalthinkingin a numberof importantrespects.
Firstof all, as the title of Rawls's majorearlyworkindicates,he placedthe
conceptof justice at the centreof attention.Since 'justice'for him is the chief
virtueof a humansociety, it is understandablethathe organiseshis political
philosophyarounda 'theoryof justice'. This, however,is a rathersurprising
development.To be sure, justice was of great importanceto a numberof
pagan thinkersin the ancient world-the qualification'pagan'is important
herebecausethe Paulinestrandof primitiveChristianityonce againdemoted
justice (andthe 'law') in favourof 'grace''0-but I thinkit is fairto say thatno
particularsaliency had been attributedto 'justice'in the political philosophy
of the modem period. The two originatorsof modem political philosophy,
324 POLITICALTHEORY/ June 2002

Machiavelli and Hobbes, set the tone. For Hobbes, security and self-
preservationare the basic political virtues and the highest goals of politics.
'Justice'is a mereword,the contentof which is given by the law laid downby
the sovereign;it is thusa highly derivativeandnot very significantphenome-
non. Machiavellirecognises the varietyof disparategoals thathumanspur-
sue and a correspondingvarietyof differentconceptionsof the good and of
the good life-there is the life of piety, of wealth accumulation,of politics.
Even withinthe realmof politics, a politicalcommunityis the objectof praise
on accountof its 'greatness',not its justice (in the Discorsi), and an individ-
ual is 'virtuoso'by virtueof being able to attainfame, honour,glory, praise,
andso forth,ratherthanfor being 'just'in mattersof the distributionof goods
or the administrationof given laws. The theoreticalupshot of the work of
these two theoristsis thatjustice is a minorpropertyof subordinateadminis-
trativesystems ratherthan the chief virtueof a society as a whole, and that
'being just' is the appropriatedefining charactertrait of the administrator,
functionary,orbureaucratratherthanof the politicianor citizen. Tojumpfor-
ward by several centuries from Hobbes, Marx, too, treats justice as an
epiphenomenon.Eachsocioeconomicformationgeneratesthe conceptionof
justice it 'needs'to allow productionto proceedas smoothlyas possible, and
this conceptionremainsdependenton and has no standingoutside the mode
of productionin question.Thisrelativetheoreticalinsoucianceaboutthe con-
cept of justice is not merely a generalfeatureof muchof the most interesting
modem political philosophy,but it seems especially characteristicof classi-
cal liberalism.After all, for Humboldt,Constant,Mill, and de Tocqueville,
toleration, freedom, and individualismwere focal issues, but justice was
either completely invisible (Constant),or at best a minor side-issue (J. S.
Mill), or finally an object of some suspicion because it could be thoughtto
presupposea unitary,centralisingview of society thatwas a dangerto indi-
vidualism(Humboldt).Primafacie, it seems highly unlikelythatthe analysis
of a conceptlike 'justice', which is so highly dependenton shiftingforms of
economic activity and on historicallyextremelyvariableconceptionsof the
good life, could give one any real graspon the centralphenomenonof poli-
tics. If this is correct,the Rawlsianprojectwas headedin the wrongdirection
from the start, but even if Rawls's reorientationof political philosophy
aroundthe concept of justice was on its own termsa philosophicallyfruitful
move, it representsa singularlyunfortunateposition from which to try to
rewritethe history of liberalism,a movementwhose membersoverwhelm-
ingly had very differentconcerns.
The second main element in Rawls's early programmewas a remoral-
isation of political philosophy. For him, in contrast to most nineteenth-
centuryliberals,political philosophywas 'appliedethics', andthe 'ethics' in
 Geuss / LIBERALISMAND ITS DISCONTENTS 325

questionis a complex andoriginalconstruction.Rawls is not in any interest-

ing sense a Kantianbecause he has no room in his theory for such central
Kantiandoctrinesas thatof the 'a priori',buthe is also at painsto emphasisea
certaincontinuitybetween his position and Kantianethics, particularlyon
two issues: the centralityof individual 'autonomy'and the priorityof the
rightto the good." The characteristicearlierliberalview, however,was one
of great suspicion towardthe intrusionof specifically moralcategoriesinto
politics, and in particularof principled rejection of the Kantian ethics.
Rawls's work had the curious effect of advancingKantto the position of a
kind of patronsaint of liberalism.This is mildly paradoxical,because Kant
hadbeen seen for most of the nineteenthandearlytwentiethcenturiesby the
main philosophic proponentsof liberalism (Constant,J. S. Mill, I. Berlin;
also Benthamand Dewey) as an archanti-liberal.
Pre-Rawlsianliberals had two main objections to Kant. First of all, the
concept of the a priori,which is structurallyindispensablefor all forms of
Kantianism,is not acceptableto liberals. Kant'sabstractconception of rea-
son (whichcan in some sense be seen as the sourceof his doctrineof the a pri-
ori) constitutesan attemptto absolutiseaccidentalformsof thinkingthathap-
pen to be socially importantat some particulartime, andthusto freeze human
developmentat some given level. Because people at a certaintime andplace
all thinkthatmurderersshouldbe executed,thatall formsof telling anuntruth
are intolerable,or thatthe rightsof propertyare incompatiblewith taxation,
and cannot perhapseven coherentlyimagine any alternatives,these beliefs
will be stylised as universalprinciplesandcircumflexedwith the hyperbolic
radianceof the a priori.An a prioriphilosophyis for liberalsa fetteron human
progress.A Kantianethics of unvaryinga prioriprinciplesis incompatible
with the openness,flexibility,and willingness to revise one's view andadapt
to the realitiesof the situationdemandedof liberalpolitics. Second, although
both Kant and classical liberalismare committedto the value of freedom,
theirrespectiveconceptionsof freedomareradicallydifferent.Most liberals
are highly suspicious of Kantian freedom-based-on-reasonand, in fact,
stronglysuspectthatthis 'positive'conceptionof freedomcan be used to jus-
tify forms of totalitarianism.12
To avoid any possible misunderstandingon this point, I am certainlynot
claimingthata Kantianstyle philosophyis absolutelyincompatiblewith any
formof liberalism.To makean assertionlike thatwould be to makeprecisely
one of the mistakesI am suggestingthat(some) moder liberalsmake,thatis,
to assumethatthereis an essence of liberalismandan essence of Kantianism
andthatthe two can be compatibleor incompatible.If 'liberalism'and 'Kant-
ianism' are open concepts, it is not excluded that after a sufficiently long
periodof time, it might be possible thatthe two could be made to converge.
326 POLITICALTHEORY/ June 2002

Similarly,nothingpreventsus fromusing ourpresentconcepts anachronisti-

cally if we wish to do that,especially if we can give some plausiblereasonfor
wantingto do it. WhatI do wish to assert,though,is thatas a matterof fact the
majorityof liberaltheoreticiansin the nineteenthcentury,anda not insignifi-
cant numberin the early twentiethcentury,saw Kantas an opponentof their
basic project and that this is a fact that liberals who wish to be Kantians
should recognise and take some kind of position on ratherthan ignoring. If
they were wrong, why exactly were Constant,J. S. Mill, Dewey, and Isaiah
Berlin wrong aboutthe compatibilityof Kantianismand liberalism?What-
ever the best way forwardfor liberals in the twenty-firstcenturymight be,
neitherKantnorRawls providesan illuminatingmode of cognitive access to
the historical phenomenonof liberalism.

III

To pass now fromthe historyof liberalismto its presentstateandpossible

future,one sometimeshearsthe claim thatliberalismdiffersfromotherpolit-
ical philosophiesthroughits recognitionof the pluralityof potentiallyvalu-
able modes of life. This is a highly misleadingassertion.Firstof all, liberal-
ism has no monopoly on the praise of pluralism.After all, Marx, too, was
convinced that the capitalisteconomic formationmade it possible for indi-
viduals to develop and participatein a wide varietyof diverseforms of life.
Second, the multiple forms of life which liberalismrecognises are always
assumedto be embeddedin an overridingconsensus thathas a latentmoral
significance.Whatis distinctiveaboutliberalismisn't, therefore,so muchits
openness to pluralismas its view thatall societies shouldbe seen as capable
of attainingconsensus,despitea lack of homogeneityin the manners,beliefs,
andhabitsof theirmembers.Can one give any reasonsfor adoptingthis atti-
tude towardconsensus? It is not completely clear what 'consensus' means.
The term vacillates between descriptiveand normativeuses in a way that is
confusing.One can distinguishfourkindsof case. The firstis the case of sim-
ple empiricalagreement.We areboth standingin the rain,and undernormal
circumstancesI will assumethatyou too know it is raining.The second kind
of case is that of adaptivebehaviour,conformism,acquiescence, or modus
vivendi.People do as othersdo in some particularareaof life withoutgiving it
much thought,or because they thinkthey must bow toforce majeure.Thus,
certain Islamic groups in the United Kingdom no longer circumcise their
young women becausethey don't wantproblemswith the Britishpolice and
courts,despite the fact thatthey by no means agree thatthey should give up
this practicethat they take, to use the now fashionablejargon, to be partly
 Geuss / LIBERALISMAND ITS DISCONTENTS 327

constitutiveof their 'identity'. They just think they have no choice. A third
group of cases concerns formal agreements,as in the paradigmaticcase of
contracting.In a contractall partiesexplicitly affirmthatthey will behavein a
certainway, usually by transferringcertainresourcesor performingcertain
services. However,all partiesto a contractneed not have equally good rea-
sons to enterinto it, and they certainlyneed not have the same reasons.Two
people can agree on state-enforcedvegetarianism,the one for religious, the
otherfor medical or sociopolitical reasons.The fourthpossible case of con-
sensus is one in which the participantshave the same reasonsfor agreement.
Even if the agentshavethe samereasonsfor agreeing,it does not follow from
thatfact alone thatthe agreementhas anyparticularnormativevalueor stand-
ing. Two thievescan have the samereasonsfor wantingto cooperatein a bur-
glary.If one agrees thatincreasingthe numberof personsinvolved does not
changethe standingof any agreement,it isn't clearthateven the existence of
universalconsensus need be anythingmore than one fact among others.
One standardliberalline of argumenttendsto runthe notions of 'consen-
sus' that are prominentin these differentcases together.Effective coordina-
tion of actionis highly desirableif humansareto surviveandlive a life anyof
them will find worth living, but coordinationof action requiresthat some
kind of at least minimalandtacit agreementin values andnormativeconcep-
tions exist betweenthe cooperatingparties.If the partiesdid not sharea large
number of such values, cooperation would break down. Therefore, it is
claimed, thereexists in every society a basic consensus thatcan serve as the
basis on which furtheragreementscould be reached,therebyexpandingeven
furtherthe humansocial spherein which freedomandnormativitypeacefully
intertwine.Fromthis the furtherconclusionis drawnthatit is alwayspossible
and rationalfor humansto try reach consensus with their fellows, or at any
rate with those with whom they must regularlydeal.13
To be more precise, there are three variantsof the liberalthesis. First an
empiricalversion: in fact, in every functioningsociety there is, one way or
another,a basic consensus. Second, the politicalthesis thatit is alwayspossi-
ble 'in principle'to elaboratethe basic consensuson which social life rests so
thatpeaceful resolutionof conflicts is possible. The thirdmoralisingvariant
has a strongerand a weakerversion.
The strongerassertsthatwe areall in some sense obliged to reachconsen-
sus or thatit is alwaysrationalfor us to tryto reachconsensus;the weakerthat
it is always a good idea to try to reachconsensus.
Againstthese liberalpositions,MarxistsandNietzscheanscan makecom-
mon cause. Nietzsche sees humansociety as a field of potentialand actual
conflict, althoughthe 'conflict'in questionmay not alwaysbe a matterof fist-
icuffs but may involve only the exchangeof argumentsandwitticisms.In the
328 POLITICALTHEORY/ June 2002

real world,Nietzsche argues,any existing 'consensus'can be no morethana

momentarytruceenteredinto for pragmaticreasonsandwith no moralimpli-
cations, and to expect anythingmore is a utopianhope.
Marxistsin any case have always been of the opinion that irreconcilable
conflict, continuingdisagreement,andsocial divisionarethe normalstatesof
all formsof society thathaveexisted up to now.Apparentpublicconsensusis
merely the false (and thin) ideological cover thathides a chasm of division
thatis as deep andunbridgeableas anythingin the humanworldcan be. In its
classic form,Marxismteachesthatevery class society is dividedinto groups
that not only have no common good but have diametricallyopposed basic
interests.Whatis good for the capitalistsis bad for the proletariat,and vice
versa.Only a classless society could lack socially entrenchedinsolublecon-
flicts of interest.In capitalistsocieties, politics-as-usualis a pointlessactivity
for membersof the proletariat,andthe only sensibleway to act in the long run
is active engagementin the class struggle.
For a varietyof reasons, the above analysis nowadaysseems out of date.
The thesis thatthe economically and politically relevantstructureof a mod-
em society can be exhaustivelydescribedby the contrastbetween capitalists
and proletariatis no longer plausible.This should not, however,be takento
imply thatliberalconceptionsof social harmonyand the unlimitedpossibil-
ity of peaceful consensus have become any more convincing, because the
main problemof the Marxistanalysis is thatit oversimplifiesthe sources of
conflict anddivision in the modem world.Insteadof one maincontradiction
between workersand capitalists,there is an almost unsurveyablevarietyof
groupsthatarepotentiallyor actuallyin conflict with each other,groupsthat
in some cases have very sharplydefined, completely incompatibleinterests
and control over considerablepowers and resources.
In a given case, it may sometimesbe possible to attainagreementabout
some points of disputein real or hypotheticaldiscussion. Sometimesthereis
neutralgroundor a groundconstitutedby sharedbeliefs on to which one can
withdrawto find compromises-sometimes, not always. In every society
therearebothareasof consensusandareasof conflict. Both shouldbe under-
stood naturalistically,and individuals,social groups, and institutionsmust
learnto deal with both. Naturallywe often-but not always-have perfectly
good reasonsfor takingpartin discussion, especially when the alternativeis
physical violence with opponentswho are strongerthanwe are,but whether
the reasonswe do (or do not) have in a particularcase aregood, less good, or
ridiculouslybad is an empiricalmatter.'4
None of the threeliberaltheses aboutconsensus seems to me at all plausi-
ble. Firstof all, it seems obvious thatmanysocieties areperfectlywell able to
maintainthemselves althoughtheirmembersdo not takepartin a consensus
 Geuss / LIBERALISMAND ITS DISCONTENTS 329

thatis in anyway normativelybinding;manypeople in manysocieties simply

putup with existing arrangementsthatthey mustendureas best theycan. Sec-
ond, the claim that it is 'in principle'always possible to attainconsensus is
completely uninformativeuntilone knows, in more detailthanhas ever been
providedby liberals,whatexactly 'in principle'means.Furthermore,even if
the claim were true,why shoulda statementabouta consensusthat 'couldbe
attained'under some fictive or hypotheticalcircumstanceshave any direct
relevanceto a given realpoliticalsituation?Finally,it is alwaysan open ques-
tion whetheror not it is a good idea to enterinto discussionor attemptto reach
consensus. If I am dealing with a small group of armedfanatics,it is by no
meansclearthatI oughtto arguewith themratherthanimmediatelyanduni-
laterallydisarmingthem.To be sure,I will probablyhave variousreasonsfor
tryingto do this with as little use of force myself as possible, buteven if I use
minimalforce I won't be discussing anythingwith them, and a prioriI can't
know that any particularlevel of applicationof force will be sufficient.

IV

Which parts,then, of classical liberalismdeserve to be furtherdeveloped

and cultivated?In the first place the criticism of theocraticconceptions of
society or, what is anotherform of the same thing, of absolutist (that is,
explicitly or implicitly theocentric)forms of ethics. The Kantianphilosophy
is no more thanat best a half-secularisedversionof such a theocraticethics,
with 'Reason'in the place of God. This does not amountto muchmorethana
change of names.15The purenormativestandpointthat Kant'sethics tries to
occupy,a standpointin which we consideronly the normativelyrelevantfea-
tures of a possible world, abstractingstrictly from the real world and the
empiricalaccidents of concrete situations,is an expression of what Dewey
called 'thequestfor certainty'.16In an insecureworld,weak humansstruggle
convulsivelyto reachsome kind of stability;the a prioriis an overcompensa-
tion in thoughtfor experiencedhumanweakness.17This is one of the origins
of Kant'snotoriousrigidity,his authoritarian devotionto 'principles',andhis
to
tendency promote local habits of thoughtto constituentsof the absolute
frameworkwithin which alone (purportedly),any coherentexperience was
possible; thus, Euclidean geometry is declared the a priori condition of
humanexperience,and sadisticremnantsof puritanismbecome demandsof
purepracticalreason.18ClassicalliberalismrejectedKant'spracticalphiloso-
phy, but perhapsthis is not enough. Perhapsone should also reject the very
idea of a pure normativestandpoint.
330 POLITICALTHEORY/ June 2002

This might be thoughtto be a ratherextreme suggestion. Kantianshave

some humanfailings like everyoneelse; these need not be thoughtto reflect
negativelyon the purenormativestandpoint.Is thereany reasonto thinkthat
the very idea of a purenormativestandpointimplies the attemptto absolutise
accidentalexisting habits of thought?
Ratherthantryingto give a directanswerto this question,I would like to
approachit by discussing two examples. Both are drawnfrom the work of
JohnRawls. As I said, Rawls was never a strictKantian,and as his thought
developed,he movedfurtherandfurtherawayfromcommitmentto anyform
of purenormativity.This is a furtherreasonto use him as an example:if some
of the deficiencies inherentin adoptinga purenormativestandpointarevisi-
ble even in a philosopherwho has moved as far beyond Kantas Rawls has,
this seems to me to give furtherweight to suspicions about the normative
standpointas a whole.
To startwith the first example, in Theoryof Justice, Rawls claims to be
describingthe 'reflectiveequilibrium'thatwould be attainedby certainfully
rationalagentswho engagedin discussionundercertainidealisedconditions.
This state of reflective equilibriumis best understoodas a kind of successor
to the purenormativeperspective.After all, the pointof one of the maincon-
structions-the introductionof the 'veil of ignorance'- is precisely to
exclude from considerationempiricalinformationthat might prejudicethe
overridingnormativeforce of the outcome.It is, then,extremelystriking,not
to say astounding,to the lay readerthatthe complex theoreticalapparatusof
Theoryof Justice, operatingthroughover 500 pages of densely arguedtext,
eventuatesin a constitutionalstructurethat is a virtualreplica (with some
extremely minor deviations) of the arrangementsthat exist in the United
States.19It strainscredulity to the breakingpoint to believe that 'free and
rationalagents'(with no furtherqualifications),even if they were discussing
behindan artificialveil of ignorance,andassumingthatthey were to agreeon
anything at all under those circumstances,would light on precisely these
arrangements.Some criticsmightfastenon this as an indicationof the essen-
tially conservativebias of Rawls's discussion:the theoreticalimaginationis
employed not to think about alternativesto the status quo, but in orderto
reproduceit schematicallyin thought,presentingit as the outcome of full,
free, rationaldiscussion.20This mightseem grossly unfair,given Rawls'sevi-
dent intention to produce a work that would have some powerful
redistributiveimplications. If, however, one thinks it at all reasonable to
judge what is afterall presentedas a politicalphilosophyby its actualpoliti-
cal effects, it is hard to see how Rawls's perfectly genuine redistributive
hopes could have any chance of being realised-and not merely because
Rawls has no theoryof political action or agency, althoughthat is also true.
 Geuss / LIBERALISMAND ITS DISCONTENTS 331

The actualeffect of Rawls's theoryis to undercuttheoreticallyany straight-

forwardappealto egalitarianism.Egalitarianismhas the advantagethatgross
failure to comply with its basic principlesis not difficult to monitor.There
are, to be sure, well-known and unsettled issues about comparabilityof
resourcesandaboutwhetherresourcesarereallythe properobjectsfor egali-
tariansto be concernedwith, buttherecan be little doubtthatif personA in a
fully monetarisedsociety has ten thousandtimes the monetaryresourcesof
person B, then undernormalcircumstancesthe two are not for most politi-
cally relevantpurposes 'equal'.
Rawls's theoryeffectively shifts discussion away from the utilitariandis-
cussion of the consequences of a certaindistributionof resources,and also
away from an evaluation of distributionsfrom the point of view of strict
equality;instead,he focuses attentionon a complexcounterfactualjudgment.
The question is not, 'Does A have grossly more than B?'-a judgment to
which within limits it might not be impossible to get a straightforward
answer-but ratherthe virtuallyunanswerable:'WouldB have even less if A
had less?' One cannot even begin to think about assessing any such claim
withoutmakingan enormousnumberof assumptionsaboutscarcityof vari-
ous resources,the form the particulareconomy in question had, the prefer-
ences, and in particularthe incentivestructure,of the people who lived in it,
andunless one had a ratherrobustanddetailedeconomic theoryof a kindthat
few people will believe anyeconomisttodayhas. In a situationof uncertainty
like this, the actualpolitical onus probandiin fact tacitly shifts to the have-
nots;21the 'haves'lack an obvious systematicmotivationto arguefor redistri-
butionof the excess wealththey own, or indeedto find argumentsto thatcon-
clusion plausible.Theydon't in the sameway needto proveanything;they,ex
hypothesi, 'have' the resourcesin question: 'Beatipossedentes'.
How, however, are the have-nots-or intellectuals speaking in their
name-supposed to makean argumentthatdependsboth on convincingoth-
ers of the generalplausibilityof Rawls's approachand in additionon what
cannot be more than a highly speculative evaluation of a complex
counterfactual claim? That Rawls's early views have had no real
redistributiveeffect is not merely a result of the usual difficulty of imple-
mentingpoliticaltheoriesin the realworld.The second examplecomes from
Rawls's late work On the Law of Peoples. In this work in which Rawls dis-
cusses certainaspectsof internationalrelations,he introducesthe categoryof
an 'outlawstate', a 'regimethatrefuses to comply with a reasonableLaw of
Peoples' (p. 90), and writes that 'France, Spain, the Hapsburgs-or, more
recently Germany'were instances of 'outlaw states' (pp. 105f.). 'Outlaw
state' is a slightly more refined variantof the term 'rogue state', which has
come to fashionableuse in the contextof the attemptby the Bush administra-
332 POLITICALTHEORY/ June 2002

tion to justify its missile defence programme,22and Rawls's claims about

'outlaw states' are the philosophicalpendentof formerU.S. PresidentRea-
gan's characterisationof the Soviet Union as an 'evil empire'. At this late
point in his career,Rawls has moved very farindeedaway from Kantianism,
but this is still the sort of easy-going, but narrow-minded,moralisationthat
some of the most interestingpolitical theorists of the nineteenthand early
twentiethcenturies-Hegel, Marx,Nietzsche, Freud,Dewey-wished to put
an end to and replacewith morehighly differentiatedmodes of dealing with
history and politics. For Rawls, it seems a truthtoo self-evident to require
mentioning that Spanish hegemony over Latin America in the eighteenth
century was something utterly differentfrom and much worse than North
Americanhegemonyoverthe sameregionin the earlytwentiethcentury.The
BritishEmpiredid not always use kid gloves in dealingwith competitorsand
subjects,butfor Rawls it was, in contrastto France,apparentlyneveran 'out-
law state'. It also does not seem to occurto him even as an abstractpossibility
thatthe UnitedStatesmightbe consideredby some an 'outlawstate', despite
a history of annihilationof indigenous populations, slavery, and repeated
militaryinterventionin CentralAmerica(andelsewhere).It is hardeven for
those of us who belong to the privileged, inherently nonoutlaw, Anglo-
Americanworldto resistthe conclusionthatthispartof Rawls'stheoryis sig-
nificantly influenced by ethnocentrism.Naturallythere are massive differ-
ences betweenthe SpanishEmpireof the seventeenthcenturyandthe British
Empireof the nineteenthcentury-who would deny that?It is also truethat
politicians have a strong interest in distinguishing as sharply as possible
betweentheirown policies (andthe actualeffects of these) andthose of their
analoguesin otherstates-what is firmnessof purposewith us is repression
in them.23Nowadays most moder governmentswill have huge staffs of
experts,lawyers, and researcherswho arepaid to seek out groundsfor mak-
ing the appropriatedistinctionsas vividly and convincinglyas possible. The
factthatoccasionallyin some particularextremecases one can'tfindanycon-
vincing differencesis not really an argumentfor the politicalrelevanceof the
strictlynormativestandpoint.In those extremecases in which adoptingthis
standpointdoes deliver a practicallyuseful answer, we usually have suffi-
cient reasonsto come to a decision of a varietyof kinds, and in most run-of-
the-millcases normativitygives us a cleardecision thatseems plausibleonly
because the analysis that must precede the normativityjudgment rendersa
complex situationartificiallysimple and perspicuous.This analysis, which
eventuatesin thejudgment 'this is murder','this is fraud',and so on, is what
is actuallydoing the work.
The historicalstruggleagainsttheocracy,absolutism,and dogmatismhas
left behindin liberalisma thick deposit of scepticism not only vis-a-vis all-
 Geuss / LIBERALISMAND ITS DISCONTENTS 333

encompassingworldviews,butalso vis-a-vis universalistpoliticaltheoriesof

any kind. On this point Constant,Berlin, Popper,and Rorty (and also, of
course, Burke) are of one accord.Classical liberalismdid not wish to be an
all-encompassing,universal worldview but merely a political programme
aimed at eliminatingspecific social and political evils.
In its origin, liberalismhad no ambitionto be universaleitherin the sense
of claiming to be valid for everyoneand every humansociety or in the sense
of purportingto give an answerto all the importantquestionsof humanlife.
There is no clearly developed single epistemology for classical liberalism,
but it would seem that a liberal would have to believe that liberal views are
easily accessible to humanswho have no special expertiseor epistemically
privilegedposition. The ideal of liberalismis a practicallyengagedpolitical
philosophythatis bothepistemicallyandmorallyhighly abstemious.Thatis,
at best, a very difficultandpossibly a completelyhopeless project.It is there-
fore not surprisingthatliberalssuccumbagainandagainto the temptationto
go beyondthe limits they would ideally set themselvesandtryto makeof lib-
eralisma complete philosophyof life. Forcomplicatedhistoricalreasons,in
the middle of the twentiethcentury,Kantianismpresenteditself as a 'philo-
sophicalfoundation'for a versionof liberalism,andliberalsat thattime were
sufficiently weak and self-deceived (or strong and opportunistic)to accept
the offer.Even with the infusionof a significantdose of the Kantianphiloso-
phy, however, liberalismhas not succeeded in producinga position that is
'universal'in any relevantsense. Neither has it demonstratedan ability to
remainfaithfulto its originaltheoreticaland moral abstemiousnesswithout
losing political effectiveness.
Rorty has made the extremely astute and importantobservationthat the
a priori,theocentrism(even in its attenuatedform as a 'philosophy of rea-
son'), the purely normativestandpoint,and a specific form of the 'spiritof
heaviness'24all naturallygo together.A consistentliberalismwould have to
turnits back on all of them.
Unfortunately,Rorty strongly suggests an interpretationof this observa-
tion thatdoes not dojustice to it, andhe seems to drawfromthis interpretation
two false consequences.Firstof all, Rortyis obviouslykeen to promoteirony
as the most appropriateattitudefor a contemporaryliberal.While, however,
it is truethatthe rejectionof a theocentricview of the world will most likely
bring with it a discreditingof a certainnumberof humanattitudesthatwere
closely associated with it-automatic deference to authority,attractionto
certainkinds of solemnity,unctiousness,andobscurantism-irony is not the
only alternativeto piety. Anotheralternativeis to adopt an extremelybusi-
nesslike attitude,to identify oneself fully with variousprojectsin the world,
334 POLITICALTHEORY/ June 2002

and so forth.Indeed,looking at the matterhistorically,therewould not seem

to be any particularnaturalaffinitybetween liberalismandirony.If one con-
sidersthe most significantironistsof the moder period-Pascal (in the 'Let-
tresProvinciales'),Swift, Voltaire,Kierkegaard-only Voltaireseems in any
importantway a precursorof liberalism,andI thinkone would searchin vain
in the writingsof the majorfiguresof liberalism(Humboldt,Constant,Mill)
for tracesof irony.In fact, the only obvious 'ironist'amongthe political phi-
losophersof the nineteenthcenturyis de Maistre,who was anythingbuta lib-
eral.Whende Maistrein a famouspassage25reportsthatthe executionerafter
discharginghis functionreturnshome in a self-satisfiedmood, sayingto him-
self, 'No one can breaka man on the wheel as well as I can', this is an arche-
typicalinstanceof whatwe usuallycall irony.de Maistreis invitingus hereto
look at this situation simultaneouslyfrom three distinct perspectives that
conflict. Thereis the point of view of the executionerhimself ('the only man
who wearsgloves in church');the pointof view of (post-Beccarian)common
sense, which finds the executioner (and his family) repellent and his self-
satisfactionnauseous;andfinally the pointof view of God, who sees the exe-
cutioneras the necessarycentralpoint of any society, holding it all together
and making civilised life possible. 'Ironically', the executioner is (from
God's point of view and, therefore,also de Maistre'sto some extent) right,
not perhapsright to be self-satisfied, but right to glory in the efficient dis-
charge of a dignified and commodious office. The examples of de Maistre
and Kierkegaardalso show thatirony is not in itself inherentlyincompatible
with a theocentricview of the world.
Rorty to be sure would be unmovedby all this, because he is not using
'irony'in the normalsense in which we use thatterm-which is admittedly
hardto grasp-but ratherhe is engaged in the projectof 'using old wordsin
new senses'26 so as to breakdown existing vocabularies.Rather,for him an
ironistis someonewho has doubtsaboutthe existing 'final'vocabularyin use
in society and 'does not believe that her vocabularyis closer to reality than
others, that it is in touch with a power not herself'.27 On this use of the term,
neitherPlato, de Maistre,nor Kierkegaardare ironists,a consequencethatI
thinkRortywouldwelcome. Nor,althoughI thinkRortywoulddisagreewith
me, is Hegel.28Most oddly of all, Heideggerturnsout on this readingto be an
ironist,a claim thatwill not, I think,immediatelyrecommenditself to anyone
who has followed the earth-heavyfootfall of the Sage of Messkirchthrough
any of his worksandwho retainsa graspon any of the senses 'irony'has had
in Europeanlife since antiquity.29 None of this, again,would botherRorty-
of course, breakingdown the old vocabularywill generate paradoxeslike
this, and my pointingthem out is just partof my strategyof being, in Rorty's
eyes, conventionaland boring, or of rejigging the meaningof 'irony'to suit
 Geuss / LIBERALISMAND ITS DISCONTENTS 335

my own purposes.If, however,'irony',in the traditionalorthe Rortyansense,

is not the only possible attitudewe can adoptif we wantto avoidpiety, andif
'irony'in Rorty'ssense is ratherdifferentfrom whatthe traditionwould have
called 'irony',it is hardto avoidaskingwhy we shouldaccede to Rorty'ssug-
gestion. And to answer this in turn,it might be useful to think about what
motivatesRortyto makethis suggestionandwhatthe consequencesof adopt-
ing it wouldbe. This bringsme to the secondplace in which he seems to me to
pointus in the wrongdirection.I thinkhis motivationis to detachus as much
as possible from trying to approachpolitics theoreticallyand to denigrate
political action in a very subtle and sophisticatedway. Ironystandsorthogo-
nal to any form of active, practicalengagementwith the world. It is a luxury
of people who do not pressingly have to act, the kind of people Rorty calls
with admiration'bookishintellectuals',andwhom he wishes to encourageto
find self-realisationin privatelife, not politics.30This is why it is particularly
impressive that Socrates and Kierkegaard,who thought they did in some
sense have to act, also allowed themselves to indulge in irony. An 'ironic'
executionis eitherno executionat all (buta literaryor theatricalevent),or it is
a form of attemptedadditionalmockery of the victim, or both at the same
time: 'Thisis Jesus,King of the Jews'. The Christianthinksthatthis is doubly
ironic, and that the joke is finally on Pilate.
The liberal who gives up the sanctimoniousnessof the purely normative
standpointwill perhaps,as Rortycorrectlyrecognises, standat a certainkind
of distanceto some kinds of beliefs, butthe attitudeinvolvedin this does not
seem appropriatelycapturedby calling it either 'irony'or (anotheroldercon-
tender) 'scepticism'. With this, one is returnedto the issue from which I
started,namely,whatattitudewe shouldadopttowardliberalism.This cannot
be completely separatedfrom questions about the intentions of liberalism
andhow to describethe kinds of attitudesthatare,havebeen, or mighteasily
be or have been associatedwith existing forms of liberalism.Ironywill not
allow the rightkind of theoreticallyreflective, engaged political practice.
Some will (correctly) object that the demand that I made earlier that
anachronismbe avoidedis an ideal impossible fully to attain.Of course, the
account of 'liberalism'I have given is a selective one-an ideal type-that
arises from emphasising certain features and downplayingothers, and the
choice of what to emphasise to some extent depends on what I judge to be
philosophicallyfruitfuland morallyand politically valuable.We all have no
alternativebutto constructthe past in the light of what we taketo be a viable
future, but it does not follow from this that all constructionsare equally
enlighteningor thatthe usualempiricalanddocumentarystandardsforjudg-
ing historical accounts are irrelevant.The anti-Kantianand anti-Rawlsian
perspectivehas, in my view, a twofold advantage.It is a betterguide to liber-
336 POLITICALTHEORY/ June 2002

alism as a historicalphenomenon,thatis, one thatallow us to attaina fuller,

moredetailed,andmorecorrectunderstandingof its history,andit also at the
same time provides a more promising orientationfor thinking and acting
politically in the future.
As long as the real social, economic, andpoliticalinstitutionsandcircum-
stances of our life do not change,31we cannot expect to rid ourselves com-
pletely of our discontent with liberalism. This might, however, even be
thoughtto be a vindicationof one strandin the liberaltradition,the strandthat
is action-orientedbut reflexively anti-utopianand asserts that no system
eitherof actionor thoughtis perfect.This shouldhold as muchfor liberalism
as for anythingelse. This kindof discontent,then,mightnot necessarilybe an
objectionbut a sign of the continuingvitality of this tradition.

NOTES
1. John Dunn, WesternPolitical Thoughtin the Face of the Future, 2d ed. (Cambridge:
CambridgeUniversityPress, 1993).
2. Alastair MacIntyre,After Virtue(London: Duckworth, 1981); FriedrichNietzsche,
Jenseits von Gut und Bise, in Kritische Studien-Ausgabe,ed. Giorgio Colli and Mazzino
Montanari(Berlin:de Gruyter,1980), vol. 5, ?260 and Zur Genealogie der Moral in Kritische
Studien-Ausgabe,ed. Giorgio Colli and Mazzino Montanari(Berlin:de Gruyter,1980), vol. 5,
Essay II, ?? 11-12.
3. Samuel Huntington,The Clash of Civilizationsand the Remakingof the WorldOrder
(London:Simon & Schuster,1977), 184.
4. Sigmund Freud,Das Unbehagenin der Kulturin Studienausgabe,vol. IX (Frankfurt:
Fischer, 1974).
5. See also chapter2 of my Historyand Illusionin Politics (Cambridge:CambridgeUniver-
sity Press, 2001).
6. See also my History and Illusion in Politics (Cambridge:CambridgeUniversityPress,
2001), 1-13 and 69-73.
7. I'm particularlyinterestedin four theorists as representativesof classical liberalism:
Wilhelm von Humboldt(especially his Ideen zu einem versuch, die Grenzendes Staates zu
bestimmen[Stuttgart:Reclam, 1967]), BenjaminConstant(De la libertechez les moderes, ed.
M. Gauchet [Paris:Hachette, 1980]), Alexis de Tocqueville(L'ancien rdgimeet la revolution
[Paris:Gallimard1967]), andJ. S. Mill ('On Liberty',in 'OnLiberty'andOtherWritings,ed. S.
Collini [Cambridge:CambridgeUniversityPress, 1989]).
8. See below, footnote 16.
9. Cambridge,MA: HarvardUniversityPress, 1971.
10. See, for instance,Alan Badiou,Saint Paul: Lafondationde l'universalisme(Paris:PUF,
1997), esp. chapter VII. This element becomes even more prominent in the work of the
'Ultrapauliner'(GiinterBornkamm,Paulus [Stuttgart:Kohlhammer,1969], 24) Marcion,about
whom the best work is still Adolf von Harnack,Marcion:Das EvangeliumvomfremdenGott
(Leipzig, 1924, reprintedDarmstadt1985).
11. See Theoryof Justice, ? 40, for the first of these and Theoryof Justice, p. 31 n. 16 and
throughoutfor the second.
 Geuss / LIBERALISMAND ITS DISCONTENTS 337

12. See Isaiah Berlin, FourEssays on Liberty(Oxford).

13. To be sure,one mustperhapsdistinguishmerelyapparentconsensus,pseudo-consensus,
from real voluntaryagreement,but this is a minorcorrectionthatchanges nothing in the basic
structureof the argument.
14. See also my Public Goods, Private Goods (Princeton,NJ: PrincetonUniversityPress,
2001), 96-104.
15. It was a commonlyheld objectionto Kantin the late eighteenthcenturythathis criticism
of traditionaltheology was substantivelyradicalin name only. The whole contentof traditional
theology could be reintroducedsimply by renamingit 'Postulatesof Pure PracticalReason'.
This was not exactly fair because not all of traditionaltheology would surviveKant'sattack.A
later version of basically the same line of thought occurs in Stirner'scriticism of Feuerbach
(Stirner,Der Einzige undsein Eigentum[Stuttgart:Reclam, 1967]). Feuerbach,Stirnerclaims,
doesn't, as he pretends, radically detheologise religion, he simply uses the words 'human
essence' in place of the word 'God'. The structureof Feuerbach'stheoryand the contentof the
moral obligations it imposes on individuals, though, remain the same as that of traditional
theology.
16. John Dewey, The Questfor Certaintyin John Dewey: TheLater Works1925-1953, ed.
John Boydston (SouthernIllinois Press, 1988), vol. 4; similarthoughtsin TheodorAdornoand
Max Horkheimer,Die Dialektikder Aufklarung(Frankfurt:Fischer,1969).
17. In a complex industrialisedworld, there is a furtherreason for adherenceto fixed and
rigidgeneralprinciples:efficiency andsimplicityof administration.See my Historyand Illusion
in Politics (Cambridge:CambridgeUniversityPress, 2001), 150-52.
18. For Kant as 'gallows-philosopher',see JacquesLacan, Seminaire VII:L'ethiquede la
psychanalyse(Paris:Seuil, 1986), andalso AdornoandHorkheimer,DialektikderAufklarung.
19. Of course it is not at all difficult to see how Americansmight find it plausible that any
rational agents discussing politics under favourable conditions would agree on these
arrangements.
20. See Theodor Adorno, Der Positivismusstreitin der deutschen Soziologie (Berlin:
Luchterhand,1972) 'Einleitung';see also my Morality,Culture,and History(Cambridge:Cam-
bridge UniversityPress, 1999), 69-76.
21. This is not a logical point.
22. See Noam Chomsky,Rogue States (London:Pluto Press, 2000).
23. Sometimes, of course, politicians have the reverse interest, one in presenting their
programmesas being as like as possible those of some favouredmodel.
24. Nietzsche, Also sprach Zarathustra.
25. Joseph de Maistre, Les Soirees de Petersbourg (Paris: Edition du vieux colombier,
1960), 40.
26. RichardRorty,Contingency,Irony,and Solidarity(Cambridge:CambridgeUniversity
Press 1989), 78.
27. Ibid., 73.
28. Hegel did change his mind on a numberof things duringhis philosophicallyactive life,
but the one thing he never gave up was the commitment to a form of absolute knowledge
(couched in a final vocabulary)and also a rejectionof what he called 'irony'(which he saw as
instantiatedin the work of FriedrichSchlegel).
29. It is, of course,also the case thatHeideggeris not a liberal,andis a paradigmof those atti-
tudes of willful obscurantism,authoritarianism,and sanctimoniousnessthat liberalismshould
terminate.
30. Rorty,Contingency,Irony,and Solidarity,65.
338 POLITICALTHEORY/ June 2002

31. 'Le liberalisme n'est evidemmentpas une ideologie ni un iddal. C'est une forme de
gouvernementet de "rationalite"gouverementale fort complexe.'Michel Foucault,Dits et
ecrits, vol. IV (Paris:Gallimard,1994), 36.

RaymondGeuss is a Reader in Philosophy at the Universityof Cambridge;his most

recent books are History and Illusion in Politics (CambridgeUniversityPress, 2001);
PublicGoods,PrivateGoods (PrincetonUniversityPress, 2001); and At CrossPurposes
(London:Hearing Eye, 2001).