You are on page 1of 28

Justice as Fairness: Political or Metaphysical? Author(s): Patrick Neal Source: Political Theory, Vol. 18, No. 1 (Feb.

, 1990), pp. 24-50 Published by: Sage Publications, Inc. Stable URL: Accessed: 18/09/2008 03:57
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at 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 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 organization founded in 1995 to build trusted digital archives for scholarship. We work with the scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that promotes the discovery and use of these resources. For more information about JSTOR, please contact

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

JUSTICE AS FAIRNESS Political or Metaphysical?

PA TRICK NEAL Universityof Vermont


It is not uncommonthese days to hear John Rawls's post-A Theoryof Justice writingsreferredto with sentimentsof regret,even disappointment. A As has often been remarked, Theory of Justice had an impact on the intellectualworld far beyond thatachievedby most academic Anglo-Saxon books. One (certainlynot the only) reasonfor this was a widespreadperception that Rawls was thereinattemptinga projectof heroic proportionand classical scope: articulating comprehensiveand universaltheoryof justice a the foundedon firstpnnciples.It is impossibleto understand praisebestowed on Rawls by friend and foe alike for "reviving"the practiceof normative political theorywithouttakinginto accountthis perception.Of course, such as praisewas usuallythepreludeto a not so faintdamnation, crtic aftercritic to the gaps, holes and presuppositionsin the (apparentlynot so) pointed of architectonicstructure Rawls's theory,which I shall follow him in referring to as "justiceas fairness"(JAF). And thereis morethana little irony in the the fact that some who devoted their criticism to demonstrating failure of JAF to satisfy the (perceived) aim of universality now bemoan the from the heroic ambitionsof A Theory (perceived)fact thatRawls retreated of Justice and settled for "merely"systematizingand giving expression to the dominantopinions of modem liberal democracy.One cannot help but wonderwhetherRawlswould have achieved the fame which is now his had the he originallyspokenas modestlyas he now does regarding ambitionsof his theory.' Makinguse of the conceptualpoles within which he situatesJAF in his views most recentworks,I aim hereto takeissue with boththecontemporary of Rawls and of those (perhapsincludinghimself) who once perceived his
Vol. 18 No. 1, February1990 24-50 POLITICAL THEORY, ? 1990 Sage Publications,Inc. 24



work along the "heroic"lines just mentioned;in short, JAF, I shall argue, while not metaphysical, is not political either. If not metaphysical in its in original incarnation A TheoryofJustice, neitherhasJAFbecome "merely" political in its latest. It was and remains,I suggest, a theoreticalformulation which has continuouslyrested uneasily between these poles, due pnmarily to its Inabilityto reflect fully on and incorporatesuccessfully into itself the problemconstitutedby this polardichotomy.The doctrineof JAF is, I submit, an attemptto articulatea conception of justice which would overcome and transcendalternativearticulationswhich either (a), in seeking universality, rest on "metaphysical" propositionswhich do not admit, ex hypothesi, of simple verification or universal public agreement,or (b), in avoiding (a), reduce to relative and nonuniversal propositions that either (b-1) fail to achieve a critical standpointwith regardto existing opinions, norms, and institutionalpractices,or (b-2) turnout to presupposeimplicitlythe sorts of propositionsunder(a) thatone set out to avoid. The problemwith JAF then is not that it amounts to an articulationof type (a) or (b), but that it lies uneasilybetween (a) and (b), unableto move beyonda positionof oscillating between the two. To call that a "problem"is not, of course, to define a solution, but is an attemptto specify the essential natureof JAF as a theory or by defining the deeperstructure context withinwhich it moves. This essay propoundsno solution,for the simple reasonthatI knowof none to propound. It does, however,attemptto reflect on the questionof what mightand might not constitutesuch a solution by means of buttressingin detailand elaboratof ing the interpretation JAFjust sketched. Before setting on thattask, three preliminaryconsiderations in order. are First, the decision to speak of JAF rather than of "Rawls's theory" is a necessity generatedby Rawls himself. Forwhile he is the authorof the works consideredhere, he has from the startfollowed the practiceof distinguishing variouspossible interpretations JAF,often on particular of pointselaborating one withoutnecessarilycriticizingor denyingpossible others.In lightof this, it is difficultat points to clearly distinguishRawls the proponent JAF from of Rawls the interpreter JAF I have chosen, for purposesof discussion, to of of separatethe doctrineof JAF fromits author.Second,by the"doctrine JAF," I do not mean to refer to the substanceof Rawls's two pnnciples of justice nor to the applicationof those prnciples to existing institutions.Rather,I focus, as Rawls has done in his post-A Theoryof Justice writings, on the "generalconcept of justice" from which these more specific aspects of his works are derived. JAF, then, I take to refer to the idea that pnnciples of justice morally binding on citizens are those to which they would unanimously agree from a point of view which representedand gave expression



to theircapacitiesas free andequal moralpersons.Third,my aim here is not to criticize, in the sense of rejecting,Rawls's Ideaof justice as fairness.It is ratherto bringto light problemsand questionswhich seem to me to remain unresolved in his most recent elaborationof that idea. For whatever it is worth, I am for the most partsympatheticto Rawls's basic views regarding justice; the analysis undertakenhere is thus at least intended to be an
expression of immanent and constructive criticism.


In his most recentlypublishedarticles, Rawls offers an accountof "how I now understandthe conception of justice that I have called justice as is This understanding designatedas "political."There are three fairness."2 key senses in which JAF is said to constitutea political accountof justice. in First, it is interpreted such a way that it is made independent,as far as possible, of any "controversialphilosophical and religious doctrines"or "claims to universaltruth,or claims about the essential natureand identity of It of persons."3 is thenindependent any particular conceptionof the good, or as Rawls puts it, any "comprehensivemoral ideal."Second, as a consequence,a politicalaccountofjustice is said to be subjectto differentstandards than other accounts. A metaphysicalaccount of justice of "justification" would be one which definedjustice in termseitherof the conceptionof the good or the metaphysicaltheory of the self it held to be ultimately true. or Justificationof such an accountwould be an "epistemological metaphysit would ultimatelyturn on the questionof whetherthe ical problem,"for moralideal fromwhich it was derivedwas indeed the truerationalgood for humanbeingsor whetherthe metaphysicaltheoryof the self was Indeedtrue. On the other hand, Rawls says that for a political conception of justice, That "task"is specified as follows: justificationis "a practicalsocial task."4
Thus the aim of justice as fairness as a political conception is practical, and not metaphysicalor epistemological.That is, it presentsitself not as a conceptionof justice thatis true,butone thatcan serve as a basis for informedandwilling politicalagreement between citizens viewed as free and equal persons. To secure this agreementwe try, so faras we can, to avoiddisputedphilosophical,as well as disputedmoralandreligious questions.5

The third sense in which JAF is defined as political is that it is selfto historicalcontext,andis understood arise consciouslyrootedin a particular



as a responseto a practicalproblemexisting within thatcontext. In the case of JAF, the context is "modernconstitutionaldemocracy," the problem and is that"asa practicalpoliticalmatterno generalmoralconceptioncan provide a publiclyrecognizedbasis fora conceptionofjustice in a moderndemocratic state."6 JAF is not a "generalmoralconception,"but a conceptionof justice designed to secure fair termsof agreementregulatingthe basic structureof a democratic society consisting of adherentsto alternativegeneral moral conceptions, or conceptionsof the good. Now, in order to evaluate Rawls's claim that JAF is political and not metaphysical,we need a clear graspof what distinguishesthese two categories of conceptionsof justice. Yet a perplexingfeatureof Rawls's accountis thatwhile he devotes much attentionto explainingthe sense in which JAF is a political conception of justice, he devotes relatively little to explicitly elaboratingthe distinctionbetween political and metaphysicalconceptions generally. In this article, I attemptto constructthat distinction,reasoningfrom the threesenses of "political" outlinedearlier.Before elaborating substantive the differencesbetween political and metaphysicalconceptionsof justice, however, we need to take note of two formal aspects of thatdistinction.First,it is unclear from Rawls's account whether these categories are taken to be exhaustive of possible conceptions of justice or not. Rawls says nothing explicitly about this, and I shall proceed on the assumptionthat they are exhaustivecategories,on the furtherassumptionthatdid he understand there to be some third(or fourth,or so on) category,he would have discussedthem. we Second, and more important, note that as categories of conceptions of one would suppose that there would, or could, be a number of justice, conceptions within each category.This mundaneconsiderationturnsout to be quite importantin terms of assessing JAF, which is after all a political and conceptionof justice on Rawls's understanding, presumablynot the only one. One level of assessmentof JAFwould involve an evaluationof it relative to alternativepolitical conceptions in terms of how well each satisfied the alms of such conceptions.Rawls does not undertake such a task, andwhat is at Issue here is morethansimply a matterof unfinishedbusiness.As we shall see, Rawls defines the generalcategoryof political conceptionsof justice in such a way that JAF comes to be treatedas a uniquelyprivileged political of conception of justice, for crucialattributes JAF,a particular conceptionof are (unwittingly, I think) read into the criteria defining the more justice, This makes it difficultto evaluateJAF generalcategory in an ad hoc manner. at the level just specified, for it is given an inherent advantageover its competitorsas a political accountof justice.7



So, in order to render the idea of "political conceptions of justice" in a logically meaningful way, we need to reconstruct the criteria of this category in such a way that one can make sense of it as a general category containing JAF as one of its members. I hope to show, however, that once we do this, it becomes unclear as to whether JAF is political or metaphysical. Let us first turn to the problem of distinguishing political from metaphysical conceptions of justice. Taking the three aspects in which Rawls speaks of JAF as a political conception of justice, we can set out to denve the distinguishing features of a metaphysical conception. Let us label these three aspects of political conceptions as (a) independence of controversial philosophical claims, or noncontroversial foundations; (b) justification in practical terms, or practical justification; and (c) historical rootedness. If we define metaphysical conceptions by contrast, we arrive at the following: Table 1 Conceptions of Justice Distinguishing Criteria Philosophical Dependence Metaphysical of Dependentclaimsto truth controversial philosophical claims;hence philosophically controversial Political of Independent such claims; hence philosophically uncontroverslal

Standardsof Justification Scope of Application

in or justified termsof Epistemological metaphysical; Practical; justified referenceto truth by securingfairagreementamong of basic premisesor axioms free and equal persons to Universal; applicable all historical societies, not relative to one arises from limited; Historically a particular historical situation, applicableonlyto it

It is important that we understand Rawls's use of the concept of "metaphysical" in the works under consideration here, and not confuse it with meanings that terms take on in different contexts. Rawls uses the term in a straightforward, though perhaps unconventional, way. It refers most generally to controversial philosophical claims purporting to be true. We are liable to be misled if we confuse this meaning with the idea of metaphysics as, say, that "branch of philosophy that systematically investigates the nature of first including the study of being principles and problems of ultimate reality. the study of the structure of the universe (cosmol(ontology) and, often, ogy)."8 For while "metaphysics" thus understood certainly comprises con-



troverslalphilosophicalclaims, Rawls is using it in a wider and looser sense to refer to any doctrine entailing controversialphilosophical claims, not simply ontological or cosmological ones. The significant point is that doctnnes expressingmoralideals or conceptionsof the good can fall within this category.JAF is said not to rest on any such doctrine,and hence assessment of it is said not to be a matterof determiningthe truthvalue of its basic premises or axioms, thatis, its "firstprinciples." Now, given the distinctionsoutlined in Table 1, there are two levels of assessmentin which we mightengage. Onewouldproceedfromasking:How successful is JAF as a politicalconceptionof justice? The litmus test would ultimately be practical:Would a society in which citizens approachedand settled questionsof justice along the lines outlinedby JAF be a peaceful and harmoniousone, relative to possible alternatives?The second level would involve raising a logically priorquestion;ratherthan takingthe categoryof "political"accounts of justice as the given startingpoint for analysis, we would assess the rationalefor adoptingthat startingpoint in the first place. In short, it would proceedfrom asking:What are the reasons for preferring political accounts of justice to metaphysicalones, and are they sound ones? Takingup this latterquestion,we can identifytwo sortsof considerations which might lead one to undertaketo replace metaphysical accounts of justice with a political one. The commonrootof such an enterpriseis the fact of long-standing,intractable between alternativemetaphysical disagreement accounts of justice and the comprehensivemoralsystems of which they are part.Following Rawls, let us speakof thisconditionas "thefactof pluralism." But responsesto this fact can be describedas either"theoretical" "practior The is cal," dependingon the sources to which this intractability attributed. "theoretical" responsewould be that metaphysicalaccounts of justice (and the comprehensivemoral systems of which they are part)cannot meaningThis response fully be said to be trueor false, andhence shouldbe abandoned. would take the form of philosophicalskepticism. The "practical" response would abandonmetaphysicalaccountswithoutdenying thatsuch an account might ultimately and actually be true. Such a response would hold that whatever the correct answer to metaphysicalquestions relating to justice might be, the fact remains that for all practicalpurposes one should treat metaphysical accounts as if they are Intractable,for the fact of pluralism shows thatpeople do disagreeaboutthem and there is no practicallyknown method, short of the autocraticuse of state power, of publicly adjudicating these disputes, hence the need for a political accountof justice. A great deal hinges on which of these considerations motivates the preference for a political account of justice over a metaphysicalone. One



considerationwould in effect be a metaproceedingfrom the "theoretical" againstthe viabilityof metaphysics,insofaras its skeptiphysical argument cism regardingtruth failed to extend to itself. One proceeding from the would be a less skeptical,overtlypolitical,argument, practicalconsideration out" the metaphysicalquestions in the name of the apparently"bracketing It out" needfor agreement. wouldseem, however,thatthis"bracketing would presupposethe value of agreementitself (or the value of freedomunderlying It),andmustat least rateit as of highervalue thantruth.Once this evaluation would be pushedrightback is called intoquestion,the"practical" respondent to the sort of metaphysicalissue that he or she sought to avoid, for if agreementis of greatervalue thantruth,it is by no meansself-evidentlyso. of It is clear thatRawls restshis case for a politicalunderstanding JAF on He the "practical" considerations. is careful to make it clear that he wishes to leave metaphysicallevel questionsopen. As he modestlyputs it, "Justice as fairness deliberatelystays on the surface, philosophically speaking."10 This "stayingon the surface"is whatRawls calls "themethodof avoidance," wherein"we try,so far as we can, neitherto assertnorto deny any religions, philosophicalor moralviews, or their associatedphilosophicalaccountsof truthand the status of values."" He stresses the point that JAF, as he now conceives it, is not premisedon philosophicalskepticism,includingskepticism about what is truly good for humanbeings. It is meant to allow the possibility of an objectivelytruemetaphysics,withoutaffirmingor denying any particularaccount. Aware that the "method of avoidance"might be as understood restingon philosophicalskepticism,he argues,to the contrary, that "it would be fatal to the point of a political conception to see it as skepticalabout,or indifferentto, truth,much less as In conflict with it."'2It is the overridingvalue of, and need for, agreementthatserves to justify this To conscious superficiality. go beneaththe surfacewould be to begin inquirinto the truthvalue of the doctnne at hand,and this would threatenthe ing possibilities of agreement,not to say consensus. Rawls writes:
Thus, the aim of justice as fairness as a political conception is practical, and not metaphysicalor epistemological.Thatis, It presentsitself not as a conceptionof justice thatis true,but one thatcan serve as a basis for informedandwilling politicalagreement between citizens viewed as free and equal persons. This agreementwhen securely founded in public political and social attitudessustains the good of all persons and associationswithin a just democraticregime. Tosecure this agreementwe try, sofar as we can, to avoid disputed philosophical, as well as disputed moral and religious, or questions. We do this not because these questions are unimportant regardedwith and but indifference, because we thinkthemtoo important recognizethatthereis no way to resolve them politically. The only alternative to a principle of toleration is the autocraticuse of statepower Thus,justice as fairnessdeliberatelystays on the surface,

Neal /JUSTICE AS FAIRNESS philosophicallyspeaking. Given the profounddifferences in belief and conceptionsof the good at least since the Reformation, must recognize that,just as on questionsof we moral and religious doctrine, public agreementon the basic questions of philosophy cannotbe obtainedwithoutthe state's infringement basic liberties.Philosophyas the of search for truthabout an independentmetaphysicaland moral ordercannot, I believe, providea workableandsharedbasis for a political conceptionof justice in a democratic society. We try,then, to leave aside philosophicalcontroversieswheneverpossible, and look for ways to avoid philosophy'slongstandingproblems.13 (emphasisadded)


Now, it is doubtlesstrue thatpublic agreementon disputedmoral issues is the last thing to be expected in contemporary liberaldemocraticsocieties. critic of Rawls, indeed any observer of such societies, would surely Any agree with that. What is open to question is not this sociological fact of pluralism, but ratherthe precise philosophical significance which Rawls attributesto it, and correspondingly, philosophicalweight and argumenthe tative burdenit carriesin his accountof justice. The fact of pluralismalone cannotjustify adherenceto a politicalconception of justice ratherthanto a metaphysicalone. Forwhile we can agreewith Rawls that in contemporaryliberal democraticsocieties no conception of metaphysicaljustice would be universallyagreed to by "citizensviewed as free and equal persons,"a resultingpreferencefor political accountspresupposes the value of viewing one anotherin this way, thatis, in respectingeach citizen's right (as a free and equal moral person) to, as it were, veto any with his or her own metaphysicalconception which would be Incompatible with that"basicvalue"butsimply conception.My point hereis not to quarrel to point out thatit isjust that.When Rawls remarksthat"theonly alternative to a principleof tolerationis the autocraticuse of state power,"this point is obscured.State enforcementof a comprehensivemoral ideal is "autocratic" insofar as we accept the primaryvalue of equal individual liberty at issue here. The point is that the mere fact of pluralismis not sufficientto ground the acceptanceof thatvalue. Hence the preferencefor avoidingcontroversial philosophical issues commits one to affirminga position on a controversial philosophicalissue, in this case an affirmation Rawls of the priorityof the by value of liberty.However muchthe "methodof avoidance"may allow us to avoid affirmingor denying particular moraljudgments,what is unavoidable is the affirmationof the value that would justify the choice to employ it in the first place. It is this necessity which generateswhat I see as the as yet unresolved of problemof oscillation In Rawls'saccount.To arguefor the affirmation the value of the individual'slibertyto define andpursuehis or her own conception of the good is to nsk moving to a "metaphysical" level position;yet, to



avoid thatrisk in the nameof avoidingcontroversyis to leave the methodof avoidancewithoutjustificationitself. To be sure,Rawls has a good deal more to say than we have thus far considered, and we need now to take up his in arguments greaterdetail.In doing so, I attemptto show thateven as Rawls deepens his accounts of JAF as a political conception of justice in ways designed to resolve the "problemof oscillation,"thatproblemsimply reappearsat anotherlevel.

III. "Justice as fairness deliberately stays on the surface, philosophically This notion bears thinkingthrough.A deliberatesuperficiality speaking."14 is something more than a naive one; it presupposesa ground or rationale which would serve to explain it, andwe are thus led to seek the natureof the deliberationbehind it. At first glance, we might suppose that the threatof civil conflict, along with thecorresponding desirabilityof civil peace, arethe operativevalues here.On this reading,a politicalconceptionof justice would constitutean antidoteto the dangersof civil strife constitutedby the spectre of a politics of competitionamongcomprehensivemoraldoctrines.Forease of reference,let us referto such a readingas the "Hobbesianversion"of a political accountof justice. Initially,it appearsthat Rawls is moving in the directionof interpreting JAF in such a manner,given the emphasis placed on avoiding controversy over substantivemoraldisagreement.This seems even more evident when JAF to constitutea he goes on to explain the sense in which he understands JAF as a liberalaccountof justice. Forwhile Rawls explicitly characterizes liberal account of political justice, he is careful to distinguish it from the "liberalisms of Kant and J. S. Mill," which are said to stem from the "comprehensivemoral ideals"of autonomyand individuality,respectively. Just because of this commitment to comprehensive moral ideals, these are "liberalisms" said to be "unsuitedfor a political conception of justice," is for each thusunderstood said to express"butanothersectariandoctrine."15 to the contrary,is committedto no such comprehensivemoral ideal; JAF, it rather, "seeks to identifythe kerel of an overlappingconsensus"'6 among comprehensivemoral ideals.'7It would appear,then, that a further(fourth) as key sense in which JAF is to be interpreted "political"is thatof eschewing substantivemoralcommitmentaltogether,aiming instead at specifying the procedureswhich make possible a prudentiallybased peace treaty among adherents alternativemoralviews. of



on However,this appearanceis mistaken.Immediately distinguishingthe of political interpretation JAF from any conceptionof justice foundedon a comprehensivemoral ideal, Rawls turnsto distinguishit, on the otherhand, from a "merely prudential"account of justice, and it is therefore worth looking carefully at Rawls's statementsin this regard.He arguesthatwhile JAF is "political"in the sense that it presupposesno commitmentto any as comprehensivemoral ideal, it is not to be interpreted "political"in the sense that it is an amoralconception of procedural justice. He remarksthat while the idea of an overlappingconsensus among comprehensive moral ideals "mayseem essentially Hobbesian,"'8 is not to be understoodin this it way. So, while it may appearthat"thepublicacceptanceof justice as fairness is no more thanprudential; thatis, a modus vivendiwhich allows the groups in the overlappingconsensus to pursue their own good subject to certain constraintswhich each thinks to be to its fair advantage given existing Rawls arguesthatthis is mistakenand that in fact JAF "is circumstances,"19 a moral conception."2'Now in purely formal terms, it is easy enough to identify the boundanes of the idea of a "political"conceptionof justice that Rawls is pursuinghere; "political"is not defined in opposition to "moral," but at the same time it is to be understood as something other than a conceptionof justice foundedon a comprehensivemoralideal. In "The Idea of an OverlappingConsensus,"Rawls describesthis balancingaim of JAF as follows: "This view steers a course between the Hobbesian strand in liberalism-liberalism as a modusvivendisecuredby a convergenceof selfand group-interests coordinatedand balancedby well-designed constituas tional arrangements-and a liberalismfounded on a comprehensivemoral doctnne such as that of Kantand Mill."21 But, while it is clear that this is what Rawls wants of a "political" conception of justice, it remainsunclearto me how one could achieve this desideratumof a moralconception of justice which presupposesno controversial philosophical argumentsestablishingits particularmoral character. As a meansof tryingto explain how thisachievementis possible, Rawls gives two argumentsdesigned to show thatthe political interpretation JAF does of not make of it a merely prudential,or Hobbesian,account. First,Rawlsremindsus thatJAFis a moralconception:"ithas conceptions of personandsociety, and conceptsof rightandfairness,as well as prnciples which those principles ofjustice with theircomplementof thevirtuesthrough are embodied in humancharacterand regulatesocial and political life."22 Second, Rawls claims that
in such a consensus each of the comprehensive philosophical, religious, and moral doctrinesacceptsjustice as fairnessIn its own way; thatis, each comprehensivedoctrine,


POLITICAL THEORY/ FEBRUARY1990 fromwithin its own pointof view, is led to acceptthe publicreasonsof justice specified by justice as fairness. We might say that they recognize its concepts, principlesand virtuesas theorems,as it were, at which theirseveral views coincide. But this does not make these points of coincidence any less moralor reducethem to mere means.For,in general, these concepts, principlesand virtuesare accepted by each as belonging to a morecomprehensivephilosophical,religious,or moraldoctrine.23

We will take up the latterargumentlater in this essay. In regardto the of former,it seems to me thatthe preciselymoral character the overlapping consensus aimed at by JAF remainsobscure,and thatRawls's remarkshere constitutemore a (re)statement what a politicalconception must amount of to than an explanation of how such an account can be given. The polar oppositionswhich Rawls seeks to avoid are clear enough;JAF must not be for moralidealof individualcharacter, it would premisedon a comprehensive then fail to qualify as a noncontroversial conceptionof a procedurecapable of adjudicatingconflict among moral ideals. Rawls's insistence that it not, on the other hand, be conceptualizedas a modus vivendi proceedingfrom Given the stress thathe prudentself-interestis moredifficultto understand. thana metaphysical on the practicalfunctionsof JAFas a politicalrather lays account of justice, one wonders why he is so uneasy with the Hobbesian renderingof the idea of overlappingconsensus. Whatcould accountfor the insistenceon Rawls's partthatJAF not be understoodin that manner9 One supposes that the insistence proceeds from an underlyingKantian notion of the separationbetween moralityand prudence,and the resulting of fear that a Hobbesianreadingof the "political"character JAF would fail to providea sufficiently moral accountof the citizen's obligationto accept and obey the institutionalpracticeswhich would result from the original position.On a Hobbesianreading,the accountof obligationwould takeon a decidedly prudentialand utilitarianflavor; adherentsof opposing conceptions of the good would accept the terms of the cease-fire yielded by JAF it insofaras they reckonedthat,outrightvictory being unobtainable, would be preferableto peacefully coexist with theiropponentsratherthanengage them. Withinthe terms in a (presently9)futile effort to removeor "convert" of Rawls's project, then, one can identify a first rationalefor resistingthe understood of "Hobbesian along interpretation" JAF WereRawls'sargument these lines, it could not plausiblybe said to fulfill the avowed Rawlsianaim of of of providing"an alternativeto the dominantutilitarianism our tradition versionof it. Instead,it would have become a prudential politicalthought."24 calls Yet this rejectionof the "Hobbesian interpretation" into questionthe "political"characterof JAF, even as it saves it from being renderedas a prudentialdoctrine. If we reject the Hobbesianreading,it would seem that



we are forced into the position of supposing some kind of "metaphysicallevel" foundationfor JAF sufficient to give an account and defense of the moral value of tolerance itself. Now, liberalismhas offered such accounts, Yet notablyin the theoriesof KantandJ. S. Mill.25 the very featurethatwould make them attractivein this respect is thatwhich leads Rawls to place them in the category of "comprehensive moral Ideals" expressing "sectarian doctrines."How can one have it both ways? Now, Rawls does anticipatethis problemand provides a response to it. But I think it can again be shown thatthe response,ratherthanresolvingthe dilemma, simply gives expression to it at another,deeper level of analysis. To see this, we need first to look at Rawls's conceptionof moralpersonality, for It is throughthe explication of this idea that Rawls provides a response to the manifestationof the problemof oscillationjust sketched. JAF, understoodas a model procedurecapable of accommodatingan overlappingconsensus among alternativeconceptionsof the good, is premised on a particular in conception of the person. Most notably,participants the originalposition are defined as "freeand equalpersons."Rawls explains each attribute follows: as
The basic intuitive idea is that in virtue of what may call their moral powers, and the powers of reason, thought, and judgement connected with these powers, we say that personsare free. And in virtueof having these powers to the requisitedegree to be fully cooperatingmembersof a society, we say thatpersonsare equal.26

He continues by elaboratingthis conception of the person: "Since persons can be full participantsin a fair system of cooperation,we ascribe to them the two moral powers connected with the elements In the idea of social cooperation noted above: namely, a capacity for a sense of justice and a As capacity for a conception of the good."27 well as being premisedon this of the person,the procedureadvancedby JAF Is designed not to conception yield a basis for social cooperationsimply, but is designed for fair social The cooperation.28 originalpositionis theelementof JAFwhereinthese ideas receive their proceduralexpression:
The fair terms of social cooperationare conceived as agreed to by those engaged in it, that is, by free and equal personsas citizens who are borninto the society in which they lead theirlives. But theiragreement,like any othervalid agreement,mustbe enteredinto underappropriate conditions. In particular, these conditions must situatefree and equal persons fairly and must not allow some persons greater bargainingadvantages than others. Further,threatsof force and coercion, deception and fraud,and so on, must be excluded.29



Hencewe arriveat whatRawlscalls the"fundamental questionof political what is the most appropriateconception of justice for justice: namely, specifying the terms of (fair) social cooperationbetween citizens regarded as free and equal persons,and as normaland fully cooperatingmembersof Now, In light of the above, we see thatJAF society over a complete life."30 does qualify as a moraltheoryof justice, at least in the descriptivesense that it ratherobviously comprisesmoralpremises:the conceptionof persons as that free and equal, and the furtherrequirement termsof social cooperation must satisfy a condition of fairness,which they will do just insofaras they resultfrom a procedurewhich respectsand embodies this conceptionof the the person.Itwould seem however,thatthiswouldundermine claim thatJAF basis Kantian as a political accountof justice, given the apparently qualifies which is admittedto be its very heart. of this conceptionof moralpersonality In the earlierDewey lectures,Rawls seemed to accept as much.As William Galston aptly summarizedthe argumenttherein:"The Dewey lecturescan be seen as resolving the tension, which runs throughA Theoryof Justice, account of humanagency, between the Kantianand the rational-prudential a thoroughgoing reconstructionof justice as fairness along by offering franklyKantianlines."3 In two more recentarticles,however,Rawls distanceshimself from such a readingof JAF To be sure, from a purely descriptivepoint of view, the notion of moral personalityremainsat the heartof JAF To quasi-Kantian for it in these terms, however,by appealingto the "Kantianinterpreargue and doctrine" undermine interpreany tation,"would makeJAF a "sectarian tation of it as a political account of justice capable of accommodatingthe rangeof sectariandoctrineswhich receive expressionin moderndemocratic culture.Awareof this tension,Rawlsoffersno directphilosophicalargument in support thisconceptionof moralpersonality; of instead,he arguesthat"the conceptionof personsas havingthe two moralpowers,and thereforeas free and equal, is also a basic intuitiveidea assumedto be implicitin thepublic
culture of a democratic society"32 (emphasis added).

With this move, Rawls avoids the chargethatJAF is a metaphysicalacelementsof JAF is now arcountof justice. The justificationfor the "moral" aboutthe self-understanding ticulatedin the formof an empiricalproposition of personsin democraticsocieties. JAF is thus said to derive its conception of the person not from any Kantianor quasi-Kantian philosophicalfoundation, but ratherfrom the (alleged) fact that this conception of the person is accepted as more or less a given in the "public culture of a democratic succeeds In detachingJAF society."But, if we allow thatthis interpretation



from any controversialmetaphysicalfoundation,it does not follow that the new, that is, "empirical," foundation,is any less controversial. To begin, one might question the validity of the empiricalproposition itself. Is it true that, whatever the ultimate metaphysical merits of the Rawlsian conception of moral personality,this conception is the (at least of liberalcitizens? Grantingthat implicit) self-understanding contemporary it is the predominantone, Rawls recognizes that it is something less than This recognitioncan be glimpsed by looking at Rawls's explauniversal.33 nation of those conceptions of the good which are impermissiblewithin the terms of JAF He openly acknowledgesthat in JAF, "the concept of justice is independentfrom andpriorto the concept of goodness in the sense thatits Echoprincipleslimit the conceptionsof the good which arepermlssable."34 of perfectionistdoctrineswhich extends back to the incaring the rejection nation of JAF in A Theoryof Justice, Rawls excludes from consideration those conceptions which "hold that there is but one conceptionof the good which is to be recognized by all persons, so far as they are fully rational."35 This turnsout to be, to say the least, a ratherexpansive category,and Rawls He describes it as includingthe "dominant traditionsince classical times."36 mentions Plato, Aristotle, and the "Christiantraditionas repreexplicitly sented by Augustine and Aquinas,"as theoristsor theories falling underit. Liberalism,on the otherhand,of which JAF is an example, is said to be
a political doctrine [which] holds that the question the dominanttraditionhas tried to answerhas no practicalanswer;thatis, it hasno answersuitablefora politicalconception of justice in a democraticsociety. In such a society, a teleological politicalconceptionis out of the question:public agreementon the requisiteconception of the good cannotbe obtained.

By my lights, this seems a remarkableargument. But in the present context, I leave aside the questionof whetherRawls has done justice to the "dominanttradition"by so characterizingit as what amounts to narrowmindedness.Instead,I note only thatthis argument circumscribes sharplythe of conceptions of the good which might find their place within the range consensus"providedby JAF Thatconsensus,we now see, will "overlapping comprise only those doctrinesthat can accept the Rawlsian conception of moral personality. Those doctrines which do not accept or embody that conception of moral personalityare, as Rawls bluntly puts it, "out of the question." Hence we see that it is, strctly speaking, false to say that JAF serves to adjudicatefairlyamongcompetingconceptionsof the good simply. By Rawls own account, JAF serves to adjudicateamong those conceptions



of the good which, whatever their other differences, share an underlying consensualacceptanceof the Rawlsianidea of moralpersonality. Any other doctnnes are ruled out primafacie. Adherentsof those doctrineswho also happento be citizensof liberalsocieties will surelywish to pointout to Rawls that, however predominanthis conception of moral personalityis In such societies, it is something less than universal.Were Rawls more willing to embracethe HobbesianInterpretation JAF,he would have a ready reply of to any complaintsthat such adherentsmightvoice, roughlyalong the lines of "grinandbearit."In the absenceof that,it is hardto see whatsortof moral, as opposedto prudential, replycould be given, especiallywhen we recallthat Rawls'sprojectherepreventshim fromassertingandarguingthe philosophical inadequacyof any such doctrines;thatwould be to pursuea strategyof not One could, of course, claim thatno "metaphysical," political, argument. such doctrinesareactuallyheld by citizens of contemporary liberalsocieties, but the manifest absurdityof that propositionwould be evident to no one more than a professorof political theorywhose book elicited public howls of complaintfromjust such fellow citizens.38 Barringan ad hoc refusal to addressthe issue at all, anotheralternative would be to pursuethe strategy that Rawls (unwittingly9)slips into when he remarksat one point that the basic ideas of JAF "are likely to be affirmed by each of the opposing comprehensivemoral doctrinesinfluentialin a reasonablyjust democratic would society"39 (emphasisadded).But hereagain,Hobbesian prudentialism seem to have rearedits ugly head;adherents of"nonlnfluential," comprehensive moral doctrnes will surely wish an account of the moral basis for imposing the legal order sanctionedby JAF on them simply because they lack thepower to make theirviews influential. In "The Idea of an OverlappingConsensus,"Rawls revises this strategy by speakingof the boundsof the consensusas excludingunreasonablemoral views ratherthannontnfluential ones. He writes:
Certaintruths,it may be said, concern things so important that that differences about them have to be fought out, even should this mean civil war.To this we say, first that questionsarenot removedfrom the politicalagenda,so to speak,solely becausethey are a sourceof conflict. Rather,we appeal to a political conceptionof justice to distinguish between those questionsthatcan be reasonablyremovedfrom the political agendaand those that cannot, all the while aiming for an overlappingconsensus. Some questions still on the agenda will be controversial,at least to some degree; this is normalwith political issues.40

Now, if our prmary worry was that an overlappingconsensus provided by a political conception of justice would be "too tight"in the sense that it



dissolved too many, or all, sources of political disagreement,this would (so perhapssatisfy us. But if our worry is the priorone of understanding as to be able to assess) the justification for requiring adherents of moral conceptions which do not fit within the overlappingconsensus providedby JAF to neverthelessobey the legal orderconsistentwith thatconsensus,this does not help. It merely labels such conceptions as unreasonable,without of explainingwhy we should acceptthe appropriateness the label. Thatsuch conceptionsare illiberal is, of course,clearbeyond a doubt,but thatis surely WereI an adherentof one of these no argumentagainsttheirreasonableness. illiberalconceptions living underthe (state-enforced!)legal orderconsistent with JAF,I would suspectthat,althoughthe languageof Rawls's accounthas shifted here from noninfluentialto nonreasonable,the operative Idea has with remainedthe same, thatis, therearenot enoughof us illiberalsectarians enough power to prevent the more powerful liberals from erecting a legal orderreflectingtheir values and requiringus to live by it as well. A further of strategyof responseby Rawlsto adherents moralidealsfalling outside the overlappingconsensus is containedin the second of his explanations as to why the political interpretation JAF is not merely a Hobbesian of modus vivendi. To recall, Rawls arguesthat,
in such a consensus each of the comprehensive philosophical, religious, and moral doctrinesacceptsjustice as fairnessin its own way; thatis, each comprehensivedoctrine, from within its own point of view, is led to accept the public reasonsof justice specified by justice as fairness. We might say that they recognize its concepts, principlesand virtues as theorems,as it were, at which theirseveral views coincide. But this does not make these points of coincidence any less moralor reducethem to mere means. For,In general, these concepts, principlesand virtues are accepted by each as belonging to a more comprehensivephilosophical,religious,or moraldoctrine.4t

This idea, first articulated "Justiceas Fairness:Political not Metaphysin Consensus."Afterbriefly ical," is elaboratedIn "TheIdeaof an Overlapping explicating that elaboration, I want to examine and assess it at two levels: (a) as a responseto the chargethata political conceptionof justice is merely prudential,and (b) as an explanationof the moral grounds for enforcing a liberallegal orderon adherents illiberalmoralideals. As we shall see, the of two levels are closely linked. A key element of the idea is thata politicalconceptionof justice is limited in scope to the political frameworkof the state; it allows for, and is indeed built on, the recognition and acceptance of a diversity of comprehensive moralideals andconceptionsof thegood operativewithinsociety Following Rawls, let us define communityas "an association or society whose unity



It restson a comprehensive conceptionof the good."42 follows thatwhile JAF as a political conceptionof justice rules out the possibility of the state as a communityin this sense, it does allow for communitiesto exist at the level of society, as Rawls says, "in the various associationswhich carryon their life within the framework the basic structure second in those associof and ations that extend across the boundares of nation-states,such as churches and scientific societies."43 JAF as a political conceptionof justice is said to the make no judgmentregarding truthof any of "thecomprehensive conceptions of the meaning,value and purposeof humanlife, or conceptionsof the "Theyare good," operativelysharedin any of these privatecommunities.44 all equally permissable,provided they respect the limits imposed by the primcples ofpoliticaljustice"45(emphasisadded). Would, then, an overlapping consensus among such communities to accept JAF as regulatingthe basic structureof political society be a moral and not merely a prudentialagreement?The answer to this question is in indeterminable the abstract.Recall that Rawls's claim is that "in such a consensus each of the comprehensivephilosophical, religious, and moral doctnnes acceptsjustice as fairness in its own way. They recognize its andvirtuesas theorems,as it were, at whichtheirseveral concepts,principles
views coincide."46 This would be accurate in regard to those doctrines, or

communities built on shared adherenceto such doctrines,which contain somewhere within themselves a commitment to the value of individual liberty as understoodin JAF Doctrines or communities without such a commitment,though,would by definition be unable to be a partof such a consensus, at least on moral grounds. Presumably,they would either be simply excluded from it or accept it (andbe acceptedwithin it) on prudential grounds only. So, the degree to which the overlappingconsensus of JAF sowould be moral ratherthan prudentialwould depend on the particular in questionandon the substantivenatureof the "prvate"communities ciety
within It.

The "model case" of an overlappingconsensus that Rawls utilizes for illustrativepurposesis not likely to answerour questionsin this respect.He describes it thus:
It contains three view: one view affirmsthe political conceptionbecause its religious the doctrineand accountof faith lead to a principleof tolerationand underwrite values of a constitutionalregime;the second view affirmsthe politicalconceptionon the basis of a comprehensiveliberalmoraldoctrinesuch as thoseof KantandMill;while the third supportsthe political conceptionnot as foundedon any wider doctrinebut ratheras in conditions favorable itself sufficientto expresspoliticalvalues that,underthe reasonably that make a more or less just constitutionaldemocracy possible, normallyoutweigh whateverothervalues may oppose them.47



The disappointingfeatureof this examplestems fromthe fact thatall three exemplars of comprehensive views are "easy cases" for JAF One is not surprised, after all, to find that those who find the political conception acceptable in Itself as a moral view (the third case) are able to affirm it morally'And there is no difficulty in acceptingthatthe othertwo exemplars can morallyaffirmJAF,as each containswithin itself a deep commitmentto the value of equal individual liberty. One wishes that the religious group imagined by Rawls (the first case) would have been drawn more from the model of Amish, Mennonites,or Jehovah'sWitnessesthan from, as appears to have been the case, Unitarians. I conclude that as a reply to the suspicion thatthe overlappingconsensus of JAF is merely prudentialratherthan moral, Rawls's idea of alternative doctrines affirming JAF in its own way is, if not unsuccessful, at least ambiguousand incomplete. How shall we assess it as an explanation of the moral grounds for enforcing the law on adherentsof illiberalconceptionsof the good? I think the answermust be the same. Such citizens (and theirdoctnnes) eitherareor are not to be treatedas partof the overlappingconsensus.If not, then a moral justificationfor requinngthem to obey the legal orderof a liberalstate must ultimately rest on the adequacyof the philosophicaldefense which can be given in supportof liberalism(and hence against their illiberalism).But to engage such a task would be to abandona political conceptionof justice for a metaphysical one. On the other hand, if we treat them as part of the overlappingconsensus, it can be so only on prudential,not moral,grounds. The first strategy would allow Rawls to maintainthe thesis of the moral character of the overlapping consensus provided by JAF at the cost of sacrificing a degree of inclusiveness, as well as the cost of entailingcontroversial moraljudgments,or "metaphysical-level" arguments.But if to avoid these costs we take the latterstrategy,we underminethe thesis of the moral instead. characterof JAF, and its overlappingconsensus becomes prudential Rawls appears At one point in "The Idea of an OverlappingConsensus," with to endorseconsciously the prudential interpretation regardto "illiberal" groups. He writes:
Nevertheless in affirminga political conception of justice we may eventually have to assert at least certainaspects of our own comprehensive(by no means necessarilyfully comprehensive) religious or philosophicaldoctrine. This happenswhenever someone that to ensuretheirbeing insists, for example, thatcertainquestionsare so fundamental rightly settledjustifies civil strife. The religious salvationof those holding a particular religion, or indeed the salvation of a whole people, may be said to dependon it. At this point we may have no alternativebut to deny this, and to assertthe kind of thingwe had


THEORY/ FEBRUARY1990 POLITICAL hoped to avoid. But the aspectsof our view thatwe assertshould not go beyondwhat is necessaryfor the political aim of consensus.48

This would seem to let the (Hobbesian)cat out of the bag. If this is taken as Rawls'sconsideredpositionon the question,thenthe moralinterpretation of JAF would have been unequivocallyabandoned,for no attemptis made to justify the exercise of power envisioned in this example, as "we"liberals our "deny"the views of those outside our consensusand "assert" own over them. It is true that, in discussingthis case, Rawls reminds"us"to assertas little as possible. It is noteworthy, however,thatthis constraintis advocated not out of concernwith the questionof the moralstatusof those outside the consensus,butout of concernto maintainthe consensusamongthosealready within. But that they share a set of moralbeliefs is not in question;what is left unanswered the questionof whatwouldmorallyjustifytheirimposition is of these beliefs on those who do not sharethem.In the absenceof an attempt to answerthatquestion, it is difficult to see how the overlappingconsensus of JAF can be called a moral, ratherthan a prudential,one. In short, the Rawls's accounthere again. problemof oscillation characterizes


we Suppose,for the sake of argument, ignorethe problemof theunsettled moral status of those whose moral commitments leave them outside the consensus offered by JAF It seems to me there is yet a furtherand deeper, and final, sense in which JAF oscillates between its metaphysical and poles. I will now try to show this. political interpretive liberal society accept the Supposingthat all citizens of a contemporary and Rawlsianidea of moralpersonality, thereforeacceptJAF as a procedural of framework designedto yield pnnciplesgoverningthe basic structure their of society,the moralcharacter the overlappingconsensusremainsanobscure notion. By making JAF a metaphysically innocent construction, Rawls leaves us to wonderwhy we should accordmoralsignificanceto the beliefs about moral personalityheld by citizens in a democraticculture. It is not enough to say that this is simply in keeping with the political characterof JAF Thatmight be enough, if Rawls were to pursuethe Hobbesianline of interpretation. As we have seen, however, Rawls does not wish to renderJAF in this manner.There is more than squeamishness at work here. A genuinely Hobbesian,or prudential, readingof JAF would ultimatelyeviscerateit, for



JAF, through the veil of ignorance, requires that parties to the original position be situated with equal amounts of power to determine the final outcome. On a Hobbesian reading,however, partieswould have reason to of acceptthismorally-basedconstraint equalpoweronly insofaras it actually obtained in social reality,a ratherdoubtfulstate of affairs to say the least. Thatis, on the Hobbesianreading,partiesto the originalpositionwould have on no reasonnot to drivea hardbargain thetermsof social cooperationinsofar as they had the power to do so. This is simply to say thatthey would refuse to have the veil of ignoranceplaced over them. (Relatively weaker parties would, of course,happilyaccepttheveil of ignorance.)We see, then,a further rationalebehind Rawls's insistence thatJAF not be conceived as a political doctrne of justice along Hobbesianlines. To do so would be to remove a centralcomponentof the theory-the requirement no partyto the original that positionhas a bargaining advantageover any other.But thatis just to say that a centralmoralpremise of JAF is the value of equal individualliberty. However, we are then left to wonder again what argumentRawls could give to support this aspect of the theory, rememberingthat any argument the appealingto "metaphysical" groundsis ruledout. Presumably, argument would be that doing so Is a means of working out a procedurewhereby can competingcomprehensivemoralideals (or theiradherents) establishfair terms of social cooperation. But this argument,depending on how it is understood,establishes either too much or too little. It is too much if "fair" is read as equivalent to "termsacceptableto free and equal moral persons deliberatingunderthe conditions of equal power establishedby the veil of ignorance."This argumentis either circular,in that fairness is thus defined in terms presupposingJAF, or dependentfor Its plausibility on one of the more controversial argumentsadvanced by Rawls, which maintains that and inequalitiesof social power are morallyarbitrary unrelatedto personal On desert.49 the otherhand,the argument establishestoo little if, leaving aside the questionof how to interpret the "fairness," claim is the minimalone that JAF is a meansof establishingtermsof social cooperationsimply in a morally pluralisticsociety. Grantingthatit is a meansof doing so, it does not follow that it is the only means. A theory specifying a procedurewhich allowed social inequalitiesof power to be replicatedwithin (say, an originalposition withoutthe veil of ignorance)could surelyresultInan agreementestablishing termsof social cooperation. Now, to be sure, such an agreementwould legitimate social inequalities which are unacceptableto Rawls, but it could achieve the minimaldesideratum of peace in a morallypluralisticsociety. What argumentcould Rawls thengive for excluding any such procedurein favorof thatspecified by JAF9



Itwould have to be one which purported show thatsuch a procedure to would fail to yield genuinelyfair termsof social cooperationinsofaras it allowed factorsto be reflectedin the choice situation.But here we morallyarbitrary would be back to the circularargumentwhich establishedtoo much. How could one break the circularity9One could, as Rawls seemed to do in the foundation a groundfor hisparticular as Dewey lectures,appealto a Kantian of interpretation "fairness"as requiringa veil of ignorancein the original position. I thinkthere is much to be said for this strategy,but it would again makeJAF a metaphysical,not a political,accountof justice. But if one tries to breakthe circularity movingin a politicaldirection,one againrunsinto by the problemthatJAF is not the only procedurecapableof generatingterms of social cooperation, and that there is no purely political rationale for the preferring procedurespecified by JAF to any otherthatcould do thejob. with the Rawls closes "Justiceas Fairness:Political not Metaphysical" "In by following remark: a society marked deep divisions betweenopposing and incommensurable conceptionsof the good,justice as fairnessenablesus But at least to conceive how social unity can be both possible and stable."50 model of social cooperation that it is not the only procedural the problemis which enabling us to conceive these things. In the absence of an argument back to (admittedly)controversial"metaphysical" groundsfor its appeals moral commitments,we have no reason to prefer JAF to any other such model. procedural models At this point, Rawls could arguethatwhile alternative procedural JAFis uniquely can accountfor thepossibilityof termsof social cooperation, suited to account for stable terms of social cooperation.Rawls does make such an argument for JAF on grounds of stability in "The Idea of an OverlappingConsensus,"suggesting that because adherentsof alternative consensusgenerated JAF moralidealswhich do fit withinthe overlapping by will have affirmedit for moralreasons,it is likely to be more stable than a consensus resting merely on a balance of power. Referrng to the "model consensus"describedpreviously,he writes:
Those who affirm the variousviews supportingthe political conceptionwill not withdrawtheirsupportof it should the relativestrengthof theirview in society increaseand eventuallybecome dominant.So long as the threeviews are affirmedand not revised, of the political conceptionwill still be supported regardlessof shifts in the distribution political power.

This argumentseems highly doubtfulin supposing,in effect, thatpower will not corrupt.Rawls begs the question in tautologicallysaying that "so long as the threeviews are affirmedandnot revised,the politicalconception



will still be supportedregardlessof shifts in the distributionof political power." Surely, the question is whetherone of those "views," or groups, might not begin to "revise"its sense of moralobligation once it no longer found itself in the position of not having the power to advantageitself by doing so. One need not thinkmankinda race of devils to fear that it might, and to fear thatRawls is overly optimistichere in supposingthata dominant "majorityfaction"will not dominate.And as the allusion to Madison suggests, the expression of that fear is certainly within the liberal tradition. Indeed,if we accept Rawls's suppositionabouthumanbehavior,we are led to wonder why he includes the veil of ignorance,which ensures equalityof power,as a componentof JAF To be sure,therearethose genuinelyvirtuous liberalcharacters,who, as it were, do not need to have the veil of ignorance placedon themto ensurethattheytreatothersas equalto themselves;whether thereareenoughof themto lendcredenceto Rawls's argument quotedearlier is anotherquestion. We can recall that Hobbes, too, recognized that there were those who would keep their word regardlessof the advantagesto be that this "is a generosity too gained from breakingit, but he also remarked rarely found to be presumed on, especially in the pursuers of Wealth, Command,or sensuall Pleasure;which are the greatestpartof mankind.The Withoutclaiming that a balance of passion to be reckonedupon Is fear."52 roughly equal power is an absolutely necessarycondition for underwriting the stabilityof an egalitarianliberalregime, it seems at the very least fair to say thatRawls's suppositionthatit is not is open to seriousdoubtfromwithin the frameworkof liberal thought.The attemptto ground JAF throughan appealto stability is thus open to the same doubt.

I have triedto show throughout articlethatJAFmaybe said to oscillate this between its metaphysicaland political interpretations. it Interpreting along political lines. Rawls must be careful to keep it from it being renderedtoo politically, lest It become Hobbesian. Yet the remedy for the spectre of Hobbesianismis a dose of Kantianism,and this serves only to take JAF beneaththe surface,philosophicallyspeaking,raisingthe spectreof controversial metaphysicalarguments,or of turningJAF into a "sectarianmoral ideal." Is this oscillation a necessary featureof Rawls's project?It seems to me that,at this point in the developmentof thatproject,it Is. Failing a returnto the explicitly "Kantianinterpretation" JAF, I cannot see how Rawls's of



presentargumentsprovidea systematicand convincing basis for rendering a politicalinterpretation JAFwhich does notcollapse into the "Hobbesian of Yet which interpretation." one can see an alternativepathof argumentation might be pursued in an attemptto work out something like the political of interpretation JAF thatRawls is after. Grantingthat the Rawlsianidea of moralpersonalityis the predominant held by contemporary citizens of (thoughnot universal)self-understanding liberal-democratic societies, and grantingthatany practicallyviable principles of justice will have to accept and proceedfrom this fact, then what is of designed necessaryto the Rawlsianprojectis the articulation an argument to explain why we should accord moralsignificanceto this empiricalstate of affairs.Only then would we have an accountwhich "settled"the moral statusof those whose moralideals exclude this conceptionof moralpersonthat ality. In this article,I have triedto demonstrate Rawls has yet to provide such an account.Thus far,he has shown only that,insofaras (most) contemthemselves this way, any porarycitizens do (at least implicitly)understand whichcouldgovernandordertheirrelations of social coordination principles will in a manner agreeablewith theirmoralintuitions haveto acceptandbuild on such intuitions.But one would have reasonto accept such principlesof carred moral as coordination moralprinciplesonly insofaras thoseIntuitions force. What is needed, then, is an argumentto show why contemporary beliefs aboutmoralpersonality(andhencejustice) are somethingmorethan simply beliefs; we need an explanationof their rationality,that is, of their genuinelymoral,and not merelyempirical,force. In short, if JAF is to remain a moral theory, as Rawls wishes, the citizens must be shown to be someconventionalbeliefs of contemporary morethanmereconventions.Thusone comes to imaginethe possibility thing that of of an "Hegelianinterpretation" JAF,an interpretation would makeof Hobbesand Kant,butone it not a doctnnewhich "steersa coursebetween"53 that transcendsboth even as it retainseach within itself. JAF needs to be or interpreted renderedin such a way that its aim is not that of balancing Kant and Hobbes, but of synthesizingthem. Hegel may be said to have just attempted such a project. the Hegel, like Rawls,transposed Kantianideaof the free andequalmoral reasonto the realmof from the timeless realmof transcendental personality But unlike Rawls, Hegel continuedon to offer an account of the history.54 to rationalityof history,purporting explain how the empiricalrealityof the modernconstitutionalstate embodyingthis conceptionof moralpersonality ("subJective freedom") is partof the larger,rationaland morally developmentalprocessof Spiritcoming to self-consciousnessthroughthe historical



action of the human species ("objective freedom"). Hegel, through his philosophy of history,offered us groundsfor accordingrationaland moral significance to the subjectivebeliefs of the modem liberalmind. Now, to be and sure, Hegel's philosophyof historyis, to say the least, bothcontroversial demandingof a ratherlarge degree of metaphysicalcommitment.My point here is not thatits truthis somehow evident, for obviously it is not. Rather, I am suggesting thatin the absenceof some sortof philosophyof historyalong Hegelianlines, thereis no obvious reasonto accordmoralsignificanceto the subjective beliefs about moral personality held by contemporaryliberal citizens. And without that, Rawls's political interpretation JAF is left of hanging between instrumentalHobbesian prudence and austere Kantian reason.My claim, then,is thatRawls'sprojectof developingan interpretation of JAF which avoids both Hobbesianheteronomyand Kantianautonomy points, strangeas it may seem, directly towardHegel. To put it bluntly,one needs to moralizehistoryif one is going to develop a moraltheoryof justice out of historicalcircumstance.Otherwise,the moralcharacter the political of of interpretation JAF is left unexplainedand ungrounded.55 The ironyof this conclusion, if indeedit is a valid one, is evident. Rawls's attemptto groundJAF in an empiricalfoundation,motivatedby the desire to avoid any controversialphilosophicalcommitments,turnsout in the end to presupposea level of philosophicalcommitment(that is, to a philosophy of history)which, if It is noncontroversial today, is so only because it is so dismissedas absurd.Reflectingon thatIronyperhaps widely gives one reason to wonderwhether the "cunningof reason"is as absurdan idea as contemwould no doubtperceive it to be. poraryliberal"intuitions"

1. The threemost important Rawls's post-ATheory of ofJustice writings, in the sense that they elaborate,extend, and in some ways modify the argumentadvancedin the book, I take to be "Kantian Constructivismin MoralTheory:The Dewey Lectures1980,"Journal of Philosophy, vol. 77, no. 9 (September,1980), pp. 515-572; "Justiceas Fairness:Politicalnot Metaphysical,"Philosophyand PublicAffairs,vol. 14, no. 3 (Summer1985),pp. 223-251; and"TheIdeaof an Overlapping Consensus," OxfordJournal LegalStudies,vol. 7, no. 1 (Spnng 1987),pp. 1-27. of This essay focuses on the lattertwo articles;the formeris discussed in WilliamGalston,"Moral and Political Theory,vol. 10, no. 4 (November 1982), pp. 492-519. Personality LiberalTheory," 2. Rawls, "Justiceas Fairness,"p. 233. 3. Ibid., "The Idea of an OverlappingConsensus,"pp. 6-7. 4. Rawls, "Justiceas Fairness,"p. 224. 5. Ibid., p. 230; see also "The Idea of an OverlappingConsensus,"p. 8.



6. Rawls, "Justiceas Fairness," 224-225. pp. 7. In brief, the problemis this. Rawls speaksof the aim of politicalconceptionsas thatof as betweencitizens regarded free andequalpersons" "specifyingthe termsof social cooperation citizens as free and equal personsis the fundamental ("Justiceas Fairness," 234). Regarding p. constitutivecomponentof JAF,presumably only one of many possible political conceptionsof of justice.Buthere,thatcomponentis readintothecriteria politicalconceptionsgenerally,giving JAF a uniquelyprivilegedstatus. 8. Entryfor "metaphysics," AmericanHeritage Dictionaryof the English Language The (New York:HoughtonMifflin, 1971), p. 825. 9. Rawls, "The Ideaof an OverlappingConsensus," 1. p. 10. Rawls,"Justiceas Fairness," 230. p. 11. Rawls, "TheIdea of an OverlappingConsensus," 13. p. 12. Ibid.,p. 12. As the succeeding analysis will reveal, I am doubtfulthat Rawls's account of JAF as a politicalconceptionof justice satisfies these self-imposedconditions.Grantingthat it is not skepticalabout truth,I am not so sure that it is not indifferentto or in conflict with it. or Everythinghinges, of course, on what constitutes "indifference" "conflict." Insofar as a politicalconceptionof justice weighs the value of agreementto be greaterthanthatof truth,it would seem plausiblethat from the point of view of those who would weigh truththe greater value, a view such as Rawls's could be seen as being indifferentto truthor in conflict with it in the sense thatit does not comprehend accept(whatthis latterview sees as) the full andproper or weight of the value of truth.Hence one would have to decide which of these two views is true itself beforeone could say that Rawls's view is not indifferentto nor in conflict with the value of truth. 13. Rawls, "Justiceas Fairness,"p. 230. 14. Ibid. 15. Ibid.,p. 246. 16. Ibid. 17. See also "TheIdea of an Overlapping Consensus,"pp. 4-5. 18. Rawls,"Justiceas Fairness," 279; see also "TheIdeaof an OverlappingConsensus," p. pp. 9-12. 19. Rawls, "Justiceas Fairness,"p. 279. 20. Ibid., p. 247. 21. Rawls, "The Ideaof an OverlappingConsensus,"pp. 23-24. 22. Rawls, "Justiceas Fairness," 247 p. 23. Ibid.,p. 247. 24. Ibid., p. 226. 25. Locke might also be mentioned, but there is a question as to whether Locke's welldoes not stem fromconsiderations mA knowndefenseof toleration LetterConcerningToleration sketched earlier; in this regard, see the interesting similar to the Hobbesian interpretation Political in "JohnLocke:FromAbsolutismto Toleration,"American argument RobertKraynak, Science Review,vol. 74, no. 1 (March1980), pp. 53-69. 26. Rawls, "Justiceas Fairness,"p. 233. 27. Ibid. 28. Ibid.,pp. 234-235. 29. Ibid.,p. 235. 30. Ibid., p. 234; note that the criteriaof JAF are here read into the criteriaof a political accountof justice. 31. William Galston, "MoralPersonalityand LiberalTheory,"Political Theory,vol. 10, no. 4 (November1982), p. 497.



32. Rawls, "Justiceas Fairness," 234. p. of on 33. Forthe purposesof this article,I concentrate the foundations this conceptof moral I personality,not its interpretation. thus assume (as does Rawls) thatwhen we say thatcitizens thinkof themselvesas "freeandequal moralpersons,"we havea moreor less clear ideaof what this means,sufficientto enablesuch citizens to reachagreementon at least constitutional-level issues. It is, however, highly doubtfulthat this assumptionis true.Consider,for example, that Nozick too appealsto thisconceptof moralpersonality, worksout a very differentconception but of it, one thatgroundsa libertarian regime very differentthanthe regime imaginedby Rawls. It may well be that even if we ignore the foundationalproblemstreatedin this essay, JAF as a political conceptionof justice entails a particular conceptionof the concept of "free and equal persons"which is rejected by many contemporaryliberals. To the extent that this is so, the overlappingconsensus of JAF would be even more circumscribedthan I assume it to be for purposesof this essay. See, in this regard,David Paris,"TheTheoreticalMystique:Neutrality, AmericanJournal of Political Science, vol. 31, no. 4 Pluralityand the Defense of Liberalism," (November 1987), pp. 909-939. 34. Rawls, "Justiceas Fairness,"p. 249. 35. Ibid., p. 248. 36. Ibid. 37. Ibid., p. 249. 38. One such example, of many,is Allan Bloom, "Justice:JohnRawls vs. The Traditionof Political Philosophy," AmericanPolitical Science Review,vol. 69, no. 2 (1975), pp. 648-662. 39. Rawls, "Justiceas Fairness,"p. 247. 40. Rawls, "The Ideaof an OverlappingConsensus,"p. 13. 41. Rawls, "Justiceas Fairness," 247 p. 42. Rawls, "TheIdea of an OverlappingConsensus,"p. 10. 43. Ibid., p. 9. 44. Ibid. 45. Rawls, "The Idea of an OverlappingConsensus,"p. 9. 46. Rawls, "Justiceas Fairness,"p. 247. 47. Rawls, "The Idea of an OverlappingConsensus,"p. 9. 48. Ibid.,p. 14. 49. Rawls's argumentcan be found in A Theoryof Justice, pp. 65-82; notable criticisms include those of RobertNozick, Anarchy,State and Utopia (New York:Basic Books, 1974), pp. 213-231; and MichaelSandel,Liberalismand the LimitsofJustice (Cambridge: Cambridge University Press, 1982), pp. 77-103. 50. Rawls, "Justiceas Fairness,"p. 251. 51. Rawls, "TheIdea of an OverlappingConsensus,"p. 11. 52. ThomasHobbes,Leviathan,ed. by C. B. Macpherson (Harmondsworth: PenguinBooks, 1968), Pt. 1, ch. 14, p. 200. 53. Rawls, "TheIdea of an OverlappingConsensus,"p. 23. 54. The sketch of Hegel's political philosophygiven here is admittedlynothing more than that. The sketch is drawn primarilyfrom my understanding, limited as it no doubt is, of The Philosophy of Rightand the Phenomenologyof Mind. 55. At times, Rawls's languageis stronglysuggestive of Hegel. Considerthe following: We collect such settled convictions as the belief in religious tolerationand the rejection of slavery and tryto organizethe basic ideas and principlesimplicitin these convictions into a coherentconception of justice. We can regardthese convictions as provisional


THEORY/ FEBRUARY1990 POLITICAL fixed pointswhich any conceptionof justice mustaccountfor if it is to be reasonablefor and us. We look, then,to our publicpoliticalcultureitself, includingits maininstitutions as of the historicaltraditions theirinterpretation, the sharedfundof implicitlyrecognized basic ideas andprinciples.The hope is thatthese ideasandprinciplescan be formulated clearly enough to be combined into a conception of political justice congenial to our most firmlyheld convictions.("Justiceas Fairness," 228) p.

The differenceis that Hegel attempts,as Rawls does not, to respondto the question:"But how do we know that what we believe to be reasonablefor us is genuinely reasonable?" to Or, putthatquestionin the form in which it most often manifestsitself in the classroom:"Butdidn't the Nazis thinktheirinstitutionswere congenial to theirmost firmly held convictions."

PatrickNeal is Assistant Professor of Political Science at the Universityof Vermont vhere he teaches courses in political theory.His primaryinterestis In modernliberal theoryand its critics, and he has publisheda numberof articles on this topic.