Scanlon - Rawls Theory of Justice (3)

8/12/2019 Scanlon - Rawls Theory of Justice (3)

1/51

Rawls' Theory of JusticeAuthor(s): Thomas M. Scanlon, Jr.Source: University of Pennsylvania Law Review, Vol. 121, No. 5 (May, 1973), pp. 1020-1069Published by: The University of Pennsylvania Law ReviewStable URL: http://www.jstor.org/stable/3311280.

Accessed: 30/12/2010 23:40

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=uplr..

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].

The University of Pennsylvania Law Reviewis collaborating with JSTOR to digitize, preserve and extend

access to University of Pennsylvania Law Review.

http://www.jstor.org
http://www.jstor.org/action/showPublisher?publisherCode=uplrhttp://www.jstor.org/stable/3311280?origin=JSTOR-pdfhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/action/showPublisher?publisherCode=uplrhttp://www.jstor.org/action/showPublisher?publisherCode=uplrhttp://www.jstor.org/page/info/about/policies/terms.jsphttp://www.jstor.org/stable/3311280?origin=JSTOR-pdfhttp://www.jstor.org/action/showPublisher?publisherCode=uplr


2/51

RAWLS'THEORYOF JUSTICETHOMAS M. SCANLON, JR.t

Rawls' book is a comprehensiveand systematic presentationof aparticular ideal of social life. The aim of the book is to analyze thisideal in a way that allowsus to see clearlyhow it differsfromprominentalternatives and on what grounds it may be preferred to them. Incarrying out this analysis Rawls presents and draws upon not only atheory of distributivejustice and a theory of political rights, but alsoa theory of value, a theory of obligation and a theory of moralpsychology.Contemporarypolitical philosophy has already been altered byRawls' book, but the leading ideas of his theory are in a number ofrespects familiar ones. They are familiar within the philosophicalcommunity since they have been set forth by Rawls in a series ofimportant and influential articles over the course of the last fifteenyears.l Rawls has altered and clarified important points in his argu-ment since these articles first appeared,and the book presents a largeamountof new material,but the main thrust of Rawls' theory can beseen in these shorter works. The idea of Rawls' book will also seemfamiliar to many who have not read his articles, for the ideal of sociallife he describesis one that lies at the heart of liberal political theory,and the principlesand policies which this ideal supportsare, in general,ones that liberals have traditionally supported. Finally, the centralanalytical device in Rawls' argumentis a variant of the familiar ideaof a social contract. Such a contract is perhaps most often thoughtof in connection with accounts of the origins of political authority.While Rawls does give an account of political obligation (one whichdoes not requireactual consent) his use of the idea of acceptinginsti-tutions in an initial contractualsituation is not confinedto this purposebut serves morebroadly as the basis for critical appraisalof economic,

t Associate Professor of Philosophy, Princeton University. A.B. Princeton University,1962; Ph.D. Harvard University, 1968.1 Rawls' basic thesis about justice was set forth in Justice as Fairness, 67 PHIL.REV. 164 (1958). This article has been widely reprinted, e.g. in PHILOSOPHY,OLITICSANDSOCIETY: ECONDERIES 32 (P. Laslett & W. Runciman eds. 1962) and in JUSTICEANDSOCIAL OLICY 0 (F. Olafson ed. 1961). His ideas on liberty were developed inConstitutional Liberty and the Concept of Justice, in NOMos VI: JUSTICE 8 (C. Fried-rich & J. Chapman eds. 1963). Distributive Justice, in PHILOSOPHY, POLITICSANDSOCIETY: THIRD SERIES58 (P. Laslett & W. Runciman eds. 1967) outlines economicinstitutions which would satisfy Rawls' conception of justice; The Justification of CivilDisobedience, in CIVL DISOBEDIENCE240 (H. Bedau ed. 1969) presents his theory ofpolitical obligation. Finally, Rawls' account of the development of moral sentiments isset out in The Sense of Justice, 72 PHI. REV. 281 (1963).

1020


3/51

RAWLS' THEORY OF JUSTICEsocial and political institutions. Rawls' work thus develops and carriesforward n a systematicway that branch of contracttheory representedby Kant and Rousseau, who saw the idea of a hypothetical initialagreement as a necessary condition for the legitimacy of politicalinstitutions.In this study I will present and discuss what I take to be themost importantargumentsof Rawls' book, emphasizingin each casethe way in which these argumentsare related to the ideal of sociallife which forms the core of his theory. I will begin with a discussionof Rawls' method and overall aims. In Section II I will consider hisarguments against what he calls perfectionism and in favor of histhesis of the priority of liberty. Section III is concerned with hisarguments against utilitarianism and Section IV with his thesis thatsocial and economic inequalities are just only if they work to theadvantageof the worst-off members of society.

I. RAWLS'PROGRAM NDMETHODPrinciples of justice, according to Rawls, are principles whichprovidea way of assigning rights and duties in the basic institutionsof society and . . . define the appropriatedistribution of the benefitsand burdens of social cooperation. 2One can think of principles ofjustice as instruments of theoretical criticism and practical choice,guiding our appraisalof differentsocial institutions, but it is essentialto think of such principlesalso as one of the crucial operativeelementsin a functioning set of institutions. This is the perspective taken byRawls and expressedin his notion of a well-orderedsociety.

Now let us say that a society is well-orderedwhen it isnot only designed to advance the good of its members butwhen it is also effectively regulated by a public conceptionof justice. That is, it is a society in which (1) everyone ac-cepts and knows that the others accept the same principlesof justice, and (2) the basic social institutions generallysatisfy and are generally known to satisfy these principles.In this case while men may put forth excessive demands onone another, they neverthelessacknowledgea commonpointof view from which their claims may be adjudicated. Ifmen's inclinationto self-interest makes their vigilance againstone another necessary, their public sense of justice makestheir secure association together possible. Among individualswith disparate aims and purposes a shared conception ofjustice establishes the bonds of civic friendship; the generaldesire for justice limits the pursuit of other ends. One may

2 J. RAWLS, A THEORY OF JUSTICE 4 (1971) [hereinafter cited as RAWLS1.

1021


4/51

1022 UNIVERSITY OF PENNSYLVANIA LAW REVIERW Vol. 121:1020think of a public conception of justice as constituting thefundamental charter of a well-ordered human association.3The idea of an agreementon principlesor procedureswhich standsbehind conflicts of interest and provides the basis for their resolutionis, in various forms, a familiar one. One may think of such an agree-ment as somethingwhich just grows up in fortunate circumstancesoras something inherited from tradition, but the content of such anagreement s clearly subject to rationalscrutiny. It is a comprehensiblequestion whether some principles of justice are rationally to be pre-

ferred to others for the crucial social role Rawls assigns to them. Thisquestion is not immediately answerable,however, without some con-sideration of what the appropriate tandardsare for such a far-reachingand headily abstract choice.Rawls' approachto this problemis analogousto what must havebeen the dream of many an arbitrator: to separate the contendingparties in a dispute and to force them somehow to come to an agree-ment on principles which they would accept in general as a basis forthe adjudicationof disputes of the kind in which they are involved.In arrangingsuch an argument-within-an-argumentn arbitratormighthave a numberof differentaims in view. First, as a practical matter,he wants the parties to come to some agreementon principles and hewants this agreementto be one that will actually hold up when theprinciples are applied to particular cases. In addition, he may havehis own views as to which principles should be chosen, views aboutparticular substantive decisions these principles should or should notdictate, and views about the kinds of considerationswhich it is relevantfor such principles to give weight to and the kinds of considerationsgood principles should ignore. The arbitrator'sproblem is to devisethe ground rules for the argument-within-an-argument nd provideincentives to the parties in such a way as to ensure that these aimswill be met.What Rawls calls the OriginalPosition can be seen as a construc-tion on this model, and the chief strategic device it employs to achieveits aims is the notion of choice behind a partial veil of ignorance.Theparties in the OriginalPosition know that they are contemporaries nsome society. While they do not know any details of the circumstancesprevailing in this society, they do know that these circumstancesareones in which cooperation is both possible and necessary, and theyhave full access to general truths of social science telling them howsocial institutions work and the difficultiesto which they are subject.

3 Id. 4-5.


5/51

RAWLS' THEORY OF JUSTICEThe parties do not know their own places in their society, their par-ticular talents and skills or even their own conceptions of the good(i.e. their tastes, goals and objectives). They are supposed to be ra-tional and to be mutually disinterestedin the sense that in their choiceof principles they are not motivated either by sympathy or by envy,but rathereach by a desire to do as well for himself as he can.Since the parties do not know their own conceptionsof the good,there is a problemof specifying the terms in which they are to under-stand judgmentsof relativewell-being.Rawls' strategy here is to focuson certain categories of goods which it is rational to want no matterwhat one's conceptionof the good may be. A good may have this prop-erty either because, like self-respect, it has a central place in any con-ception of the good or because, like income and liberty, it is necessaryas a means to most other things one may desire. Some goods whichare neutrally desirable in this sense, such as a sound constitution orgood eyes, are not things whose allotment is a function of social insti-tutions (although they are to some extent dependent on the circum-stances of one's life). Neutrally desirable goods whose allotment isdirectly dependenton social institutions are what Rawls calls primarysocial goods.4He lists these as rights and liberties, powers and oppor-tunities, income and wealth, and the bases of self-respect.Since Rawlsuses primary social goods as an index of relative well-being, he needsto claim not only that these goods are things which it is rational foranyone to want no matterwhat his conceptionof the good may be, butalso that there is at least a rough rankingof bundles of such goodswhich is similarly neutral, i.e., a ranking of combinations of primarygoods (so much income, certain liberties, etc.) as greater andsmaller such that it is rational for anyone to prefer a greaterallotment of primarysocial goods to a smaller one.5In addition to wanting for themselves the largest possible bundleof primary social goods, the parties are assumed to be motivated tochoose principleswhich it will be rationaland psychologicallypossiblefor them to adhere to, and principles which will provide the basis fora stable and lasting social orderby generatingtheir own psychologicalbase of support; i.e., it is supposed that the parties will make theirchoices only after they have determined that persons growing up ina society governedby their chosen principles will naturally acquire asense of justice motivatingthem to act in accordwith those principles.

4 Id. 62, 92.5 This problem, which Rawls calls the index problem for primary goods, is discussedat id. 93-95.

1973] 102.1


6/51

1024 UNIVERSITYOF PENNSYLVANIALAW REVIEW [Vol. 121:1020Rawls puts forward the followingTwo Principlesas the principlesof justice that would be chosen in the OriginalPosition:[First Principle]Each person is to have an equal right to the most extensivetotal system of basic libertiescompatiblewith a similarsystemof liberty for all.6[Second Principle]Social and economic inequalities are to be arrangedso thatthey are both (a) to the greatest benefit of the least advan-taged and (b) attached to offices and positions open to allunder conditions of fair equality of opportunity.7It would be a mistake to see Rawls' argumentas proceedingde-ductively, first from the descriptionof the role of a conceptionof justicein a well-orderedsociety to the idea of the OriginalPosition as a meansfor choosing the (instrumentally) best such conception,and thence tothe Two Principlesof justice as the particular conceptionwhich passesthis test. To construe Rawls' argument n this way would be a mistakefirst because, as he emphasizes,8his argumentdoes not proceed onlyin this direction.The choice of the constraintswhich define the OriginalPosition is guided initially by the idea of the role that justice is toplay in a well-orderedsociety and by strategic considerationsdesignedto insure that some agreement is reached. But these constraints arealso trimmed and shapedto insure that the OriginalPosition will yieldresults which conform with our consideredjudgments of what is just

and unjust in particularcases. What is sought is not a proof of theTwo Principles but a setting out of their relations, on the one hand,to our considered judgments of justice and, on the other, to certaingeneral ideas of social cooperation. Each of the elements in such atotal picture, and the way in which all of them fit together, elucidatesand provides support for the others. Such a method of reasoning isnot viciously circular since it is not assumed, nor is it by any meansobvious from the outset, that our consideredjudgmentsof justice andour most generalnotionsabout social cooperationcan be fitted togetherin a systematic and cohesive way.The second reasonwhy it would be a mistake to see Rawls' strat-egy as deductive is that some of the most likely alternatives to theId. 250.7 Id. 83. This principle is advanced as the favored interpretation of the more ambig-uous principle that social and economic inequalities are to be arranged so that they areboth (a) reasonably expected to be to everyone's advantage, and (b) attached to positionsand offices open to all. Id. 60. On the relation of these two formulations of clause (a),see note 68 infra.Id. 20, 21.


7/51

RAWLS'THEORYOF JUSTICETwo Principles of justice appear to be ruled out by the form of theOriginal Position itself or by the very idea of a well-orderedsocietyas one in which conflicts are regulated by a notion of justice in theway Rawls describes.Two exampleswill be helpful here: first, it mayseem to many that to take as an ideal a well-orderedsociety inwhich conflicts of interest arise and are settled with reference to prin-ciples of justice chosen on the basis of their appeal to parties whosemain motivation is each to secure as much as possible for himself isalready to build in too many features of the societies with which weare most familiar. Ruled out without serious consideration,it may bethought, is the possibility of a society in which the sources of majorconflict are largely eliminated and in which the basis of associationlies in relations of love and sympathy rather than in principles ofdistributivejustice.

Second, the possibility of a society in which the ruling principlesare given by a particular religiousideal or a particularsecular ideal ofhuman excellence appears to be ruled out in the Original Position bythe requirement hat the parties are ignorantof their own conceptionsof the good. But if judgmentsof relative value are not just matters oftaste and opinion but also of objective fact, then it appears to be ir-rational (and tendentious) to block considerationof such facts fromthe deliberations of the parties in the Original Position.It would not be unreasonable to say that these two objectionsrepresent alternative social ideals-which I will call, respectively,communitarian and perfectionist-which are not merely rivals toRawls' Two Principles of justice for consideration in the OriginalPosition but rivals to the outlook representedby the OriginalPositionitself. If Rawls is not to be construed as begging the question againstthese alternatives,then he cannotbe construed as following the deduc-tive strategy outlined above.Rawls' response to the communitarianchallenge proceeds alongtwo related lines. First he argues that it is a mistake to see love andsympathy as in themselves rivals to justice as guides for social action.These feelings may move us to sacrifice for others, but in themselvesthey provide no guide to the degree of sacrifice appropriate,no guideas to which of two loving persons should in a given case make way forthe other, and, most clearly, no guide to how one should act when theinterestsof two loved personscome into conflict.9Second,Rawls arguesthat the natural attitudes of love and trust themselves presuppose aconception of what it is to respect another as a distinct person, i.e. a

9 d. 191.

10251973]


8/51

1026 UNIVERSITYOF PENNSYLVANIALAWREVIEW [Vol. 121:1020notion of justice of the kind which the Original Position constructionis designed to capture.10 f these arguments are correct, then a com-munitarian ideal of social life of the kind I have described is not, atthe fundamental evel, an alternativeto cooperationon terms of justice.It consists rather in putting new flesh on the bones which a theory ofjustice provides.Differencesbetween the two outlooks remain,however,which aretoo deep to be called mere matters of emphasis. The communitarianmay be understoodas asking in part how far a healthy society couldrely upon justice in its distributive sense as the main counterforcetocertain tensions and instabilities. The most important question, thecommunitarian contends, is not whether the institutions maintainstrict distributivejustice; it is whetherthey providea basis for healthyhuman relations or whether, on the contrary, they foster social rela-tions which are antitheticalto the growth of natural attitudes of love,sympathy and trust. Rawls argues that the connectionbetween justiceand the values of communityis much more intimate than this criticismwould suggest. Throughthe adoptionof his Two Principlesas a publicconception of justice, Rawls argues, the membersof a society expresstheir respect for one another as moral persons; moreover, since theSecondPrincipleconstitutes anundertakingto regardthe distributionof natural abilities as a collective asset so that the more fortunate areto benefit only in ways that help those who have lost out, this prin-ciple embodies an idea of fraternity.12Thus, he maintains, a well-orderedsociety foundedon these principlesis a communityin a strongsense, a social union in which each person may pursue his own goodwithin a form of association which is itself a good for all.13I cannotconsiderthese arguments n detail here; some will be treated at greaterlength below in the discussion of the Two Principles themselves. Mypresent purpose in this example has been merely to indicate how theforce of Rawls' theory depends not only on the details of the deriva-tion of his principlesin the OriginalPosition, but also and more impor-tantly on the coherentview of social cooperationwhich the theory asa whole provides, in this case particularlyon the related accounts ofmoral psychology, of the concept of a person and of the idea of thegood.

0Id. 486-87.I Id. 179.12Id. 105.13Id. 522-29.


9/51

RAWLS' THEORY OF JUSTICEII. LIBERTY

A. The ArgumentAgainst PerfectionismThose theories which Rawls calls perfectionist direct us to ar-rangeinstitutionsand to definethe duties and obligationsof individualsso as to maximizethe achievementof humanexcellence in art, science,and culture. 14 n my previous remarksI have groupedsuch theoriestogether with theories which take as the ruling aim of social institu-tions the promotionof a particularreligious ideal. This groupingmayseem somewhat unfair since there is in theoriesof the first sort a strong

tendency toward elitism-i.e. towards placing much greater emphasison the needs and interests of some membersof society than on those ofothers-and while some religious-based heoriesmay exhibita tendencyof this kind in singlingout a small group (e.g. the elect or the clergy)for special privileges, this need not be regarded as a characteristicfeature of the type.What all of these theories, religious and secular, share is first ofall a teleological structure:15once the value of a certain end is estab-lished, social institutions are to be appraised strictly on the basis oftheir tendency to promote this end. In addition, quite apart fromtendencies to elitism, all of these theories raise serious problems con-cerning individual liberty: institutions which preserve the opportunityfor each personto adopt and pursuehis own interests and ideals and totry to persuade others to follow him will be justified on perfectionistgrounds only to the extent that they are the most effective means tothe promotionof the given end.Now it would be possible to reject theories of this kind simply onthe basis of their tendency to support institutions which conflict withour considered judgments of justice, and then to design the OriginalPosition in such a way that the offending heories are ruledout. Adoptedalone, however, this strategy is not wholly satisfying. If we can giveno independentrationale for the design of the Original Position thenthis maneuverappearssomewhat ad hoc. To provide such a rationale,based on a non-perfectionist ideal of social cooperation, would notconstitute a refutation of perfectionism; but without such a rationalewe are left with no response to the basic theoretical challenge whichthese theories raise: If there is an objective difference in the intrinsicvalue of differenttalents, goals and pursuits why should not informa-tion about these differences be used by the parties in the OriginalPosition as the basis for their choice of the principles by which social

14 Id. 325.15The notion of a teleological theory is discussed more fully in Section IV infra.

1973] 1027


10/51

1028 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 121:1020institutionswill be judged?How, in short, can we defend an egalitarianor libertarian position without embracing some form of skepticismabout values?Rawls' responseto this challenge (and his rationalefor the designof the Original Position) is groundedin the notion that social institu-tions are just only if they can be defended to each of their memberson the basis of the contribution they make to his good as assessedfrom his point of view. We must be able to say to each member thatthe arrangementshe is asked to accept provide as well for him as theypossibly can, consistent with satisfying the parallel demandsof others.In order to spell out this idea more fully it is necessaryfirst to considerRawls' analysis of the notion of an individual person's good.16Those experiences,ends and activities are components n the goodfor a particular individual, Rawls argues, which have an importantplace in a plan of life which it would be rational for him to choose.Now it may seem that a person could be said rationally to choose aplan of life (if at all) only after he has developed a conceptionof hisown good, on the basis of which he can judge and rank alternativeplans of life. But Rawls argues, persuasively I think, that this is notthe case. In real life our deliberationsabout those actual choices which,taken together, determine our plan of life proceed on the basis ofknowledge of our present tastes and capacities, knowledge of whatthings we have in the past found satisfying, and knowledgeof generalprinciples governing the ways in which our tastes and capacities aresubject to growth and change over time. This information allows usto decide on courses of action not only with the aim of satisfying ourcurrentdesiresbut also with the knowledgeand intent that our choiceswill be instrumentalin determiningwhat interests, talents and desireswe will come to have in the future. Long range choices such as thechoice of a career or a place to live give perhaps the best example ofchoices which, because they may be foreseen to have far-reachingeffects on our interests and objectives, must be made on some basiswhich goes beyond the satisfaction of our current desires and specificinterests.Rawls puts forward a negative and a positive thesis about thisprocess of deliberation. The negative thesis consists of an attack onthe idea that there must be some single overriding general goal (e.g.the maximizationof satisfaction or happiness) which underlies all ofour deliberations and explains how we can compare and choose be-tween disparatealternatives.17The positive thesis consists of a sketch

16 See generally RAWLS 60-65.17 Id. ?? 83-84.


11/51

RAWLS'THEORYOF JUSTICEof standards of rationality with reference to which our choices, par-ticularly those most general and far-reaching choices described aschoices between alternative life plans, can be criticized. This sketchconsists of two parts. First, there are general principles of rationalchoice according to which it is irrational, e.g., for anyone to preferplan of life A to plan of life B if B involves the developmentof exactlythe same interests and desires as A and provides for their satisfactionat a markedly higher level. Not all of these principles, which Rawlscalls counting principles, are as uncontroversialas this example,butall are fairly weak, and taken together they by no means can be ex-pected to determine a unique plan as the only rational choice for apersonto make. A choice from amongthe plans not ruled out by theseprinciples (the set of maximalplans) will involve such things as com-paring the relative intensity of differentdesires and the relative valuefor us of differentkinds of accomplishments.For this choice there areon Rawls' view no principles of rationality which directly require achoice of some plans over others. The only relevant standards concernthe manner n which the choice is made-whether the relevantevidencehas been duly weighed, the possible sources of uncertainty and errorproperly discounted for, etc. These criteria are grouped together byRawls under the heading deliberativerationality.Thus, to say that a certain thing is, objectively, a good for acertainpersonis, on Rawls' analysis,18o say that it would be a promi-nent feature in a plan of life which that person would hypotheticallychoose, with deliberativerationality, from among the class of maximalplans. Under any actual conditions,of course, not only the means forattainingthose things which are goods for us, but also ideal conditionsfor determiningwhat things are such goods, will in some measure belacking. What the parties in Rawls' Original Position look for in asociety is not only the means for securingthose things, whatever theymay be, which are objectively componentsof their good, but also theconditions necessary for determiningwhat these goods are.

From the fact that the parties in Rawls' Original Position sup-pose that as membersof a society they will choose their own plan oflife, and hence also determine their own conception of the good, itshould not be thought that they suppose themselves to be independentof social forces which will in large part shape and influencethe choicesthey make. It would be idle to deny that such influences exist, andirrational to object to all such influences as interfering with one'sliberty. But it is still reasonableto prefer some institutions to others

is Id. 417.

1973] 1029


12/51

1030 UNIVERSITY OF PENNSYLVANIA LAW REVIETI' [Vol. 121:1020on grounds of the conditions they provide for rationally forming aconceptionof one's good. Obviouslyone may reasonablyobject, simplyon groundsof efficiency,to institutions which place arbitraryobstaclesand difficulties in the way of individuals'attempts to get a clear viewof the alternatives open to them, of their own potentialities, and ofwhat they and others can expect from various courses of action. Amore difficultcase is presentedby the fact that some features of insti-tutions will not merely be randominferences but can be seen clearlyto favor certain choices and to discourage others, and to do this notby just enlarging people's views or by approaching ideal conditionsthereby favoring the correct answer, but ratherby skewing the evi-denceavailableor by restrictingthe alternatives ikely to be considered,or by affecting people's deliberations n other more subtle and indirectways. Systematic interference of this kind might be the result of rela-tively fixed impersonalfeatures of institutional arrangements.Alterna-tively, certain individuals may be charged with overseeing and main-taining these influences through censorship or other devices.

It is one of the features of perfectionist views which strike usintuitively as objectionablethat such views may authorize the use ofmeans of this sort in order to produce individuals conforming to aparticular ideal. Now we cannot simply reject as involving unaccept-able conditioning all social institutionswhich mold a person'schoicesand beliefs without his consent with the aim of bringinghim closer tosome ideal. Certainly Rawls cannot do this. For as he himself says,his own view involves a certain ideal of the person, and he is at somepains to show19 hat there are psychologicallaws which give us reasonto believe that persons growingup in a well-orderedsociety governedby his Two Principles of justice will naturally acquire what he callsa sense of justice-the tendency to understand and be motivated byconsiderationsof justice as specified by those principles. The actionof these psychological laws is in part dependent upon the intellectualactivity of the person on whom they are acting, but is also in largepart something which happens to a person without his knowledge orrational scrutiny.How then is one to distinguish among the various ways in whichsocial institutions may be arrangedto influence the choices and beliefsof their memberswithout each member'sconsent? Can one distinguishacceptablefrom unacceptable nfluencesof this kind on any basis otherthan an appraisalof the relative value of the particulartypes of per-sons these influencesproduce? The appropriatestandards for making

9l d. ?? 51-59.


13/51


14/51

1032 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 121:1020tion as well for susceptibility to the natural attitudes of friendship,love and trust.24These are things, Rawls argues,which almost25anyonehas reason to want.Without going fully into the argumentsfor these claims, we maycomparethemto the case which mightbe madeon perfectionistgroundsfor featuresof social institutionsdesignedto mold or restrict the choicesof their members so as to promote a particular secular or religiousideal. There is a clear sense in which such featureswill have a rationaljustification: they will be justifiableon the basis of the objective valueof the particular deal in question. A perfectionistmight thus maintainthat the interferences with a person's liberty which these featuresrepresent are ones which he should, rationally, come to accept. Butthe justification which is offered by the perfectionist will not neces-sarily be one which claims that these features promote the good ofthe person whose liberty is restricted or which claims that they areconsistent with his desire to determine his own conception of thegood; it is apt to appeal instead to some impersonalscheme of values.Moreover,this justificationneed not be based on considerationswhichwould be agreed upon by almost anyone regardlessof his conceptionof the good. Rather, it is likely to be based on one specific conceptionof the good which, even if it is objectively correct, may nonethelessbe something which is a matter of some disagreement among rationaladults in the society in question. Indeed, it is just the fact that thisconception of the good, though correct, does not compel generalagreement, which may be taken on perfectionist grounds to makenecessary the intervention in question. On Rawls' theory, however,such interventions are permitted only when there is evident failureor absence of reason and will, a phrase intended to cover cases suchas infancy, insanity or coma which involve major diminution of ra-tional capacities relative to the standard of a normal adult in fullpossession of his faculties.Thus, while Rawls' theory bases principles of justice on a hypo-thetical choice made by persons who may appear to be standing tem-porarily outside any particular society, the point of view which thetheory takes as fundamental is actually that of a person in society.The parties in the Original Position do not act from special wisdomor knowledgewhich enables them to make choices which they later,as persons under the limiting and distorting conditions of real life inan actual society, will have to take on faith. Rather, the parties' aim

24 Id. 570.25 Rawls does allow for the possibility that there may be some persons for whomthe affirmation of their sense of justice is not a good. Id. 575.


15/51

RAWLS' THEORY OF JUSTICEis to make choices which they, as real citizens, will have reason toaccept. Each party thereforeregardshis own judgmentas a real citizenas sovereign-not as infallible or immune from limitations, but as thebasis fromwhich his life will be lived, his choices madeand his work asideal contractorappraised.Rawls remarks that embedded n the principles of justice thereis an ideal of the person that provides an Archimedean Point forjudgingthe basic structure of society. 26AlthoughI have not describedthis ideal in full, the precedingargumentseems to me to illustratepartof the force of this remark. The ideal of each person as a rationalchooser of his own ends and plans provides an Archimedean Pointpartly in virtue of the fact that this conception of a person is takento be prior to any particular independently-determined onception ofhis good. One need not be a skeptic about values or truth to holdthat each of us does in fact look at himself in this way. If this is so,then the assumption that the parties in the Original Position adoptthis view of themselves should seem a natural one, and the fact thatcertain principlesof social cooperationinvolve the recognitionof eachmemberof society as in this sense a sovereign equal, while others in-volve the denial of this status to at least some members, should seema fact of some importance.The conceptionof the persondescribedby Rawls is of course notan ArchimedeanPoint in the sense of being itself a notion formedoutside of or independent of particular social and historical circum-stances. It may well be that this conception of the person and theideal of social cooperationfounded on it are typical of particularhis-torical eras and civilizations. But this is not in itself an objection toRawls' theory, particularly if, as it seems to me, the conception ofthe person in question is one that has a particularlydeep hold on usand is not a matter of great controversy or of significant variationacross the range of societies to which the theory should be expectedto apply. The question is not whether this conception of the personis in some sense absolute, but whether the particular features of thisconceptionthat are appealedto in Rawls' argumentare more contro-versial than the conclusionsthey are used to support.

Certainly this conception of the person involves a number ofimportant parameterswhich must be fixed before the notion can beappealedto in support of conclusions about the justice or injustice ofparticular institutions. The most obvious of these is the standard ofrationality: what is to count, for example, as evident failure or

26 Id. 261-65, 584.

10331973]


16/51

1034 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 121:1020absence or reason or will ? Other parametersare representedby thegeneralfacts of social science which the parties in the OriginalPositionuse in reaching their conclusions, by the notion of the primary socialgoods, and by other appeals to the idea that certain goods or circum-stances are to be desired no matter what one's conceptionof the goodmay be. The latter appeals depend upon some idea of the normalrange of variation in conceptions of the good and upon some idea ofthe means and conditions requiredfor the pursuit of these goods. Allof these may be subject to some variation over time and social circum-stances. But here again the theory need make no claim to absolutenessin these matters. It is sufficient to ask whether the appeals the theorymakes to facts about persons and the circumstancesof human life arecontroversialfor us; in particular,whether the facts appealed to aremore controversialthan the conclusions at issue; and finally, whetherthe ways in which conclusions about the justice of institutions aremade to depend on such facts strike us as plausible.Much of the preceding discussion has been internal to Rawls'particular conception of social cooperation and is thus not in anyproper sense a refutation of perfectionism.It is, rather, a descriptionof an alternative ideal of social life, one which might be called co-operationon a footing of justice. The developmentof this ideal en-ables Rawls to move beyond the observationthat perfectionismseemsto supportarrangementswhich are at variancewith our intuitive judg-ments of justice to a theory which explains why this should be so andprovides a point of view from which we can see how the perfectionistchallengecan be answered.

B. The Priority of LibertyI turn now from these theoreticalissues to considerationof Rawls'morespecific conclusionsconcerningthe place of liberty in just institu-tions. Rawls' substantiveaccountof justice is put forwardin two forms

which he calls respectively the General Conception and the SpecialConceptionof Justice as Fairness. The GeneralConceptionof Justiceas Fairness provides that [a]11 social primary goods-liberty andopportunity,income and wealth, and the bases of self-respect-are tobe distributed equally unless an unequal distributionof any or all ofthese goods is to the advantage of the least favored. 27The SpecialConceptionis expressedin the two principlesof justice stated earlier,28with the proviso that the First Principle is to be held prior to the

27 Id. 303.`8 Text accompanying note 7 supra.


17/51

RAWLS'THEORYOF JUSTICESecond in a sense to be discussed more fully below. The Second Prin-ciple allows for inequalities in the distribution of goods other thanbasic liberties on terms similar to those specified by the GeneralConception,but the First Principlelays down a more stringent require-ment of equality in basic liberties, a requirementwhich is not to beset aside for the sake of greater economic or social benefits. Thisprinciple and the rule specifying its priority receive their final state-ment in the following form.

First PrincipleEach person is to have an equal right to the most extensivetotal system of equalbasic liberties compatiblewith a similarsystem of liberty for all.Priority RuleThe principles of justice are to be ranked in lexical order29and therefore liberty can be restricted only for the sake ofliberty. There are two cases: (a) a less extensive libertymust strengthen the total system of liberty shared by all,ancd b) a less than equal liberty must be acceptableto thosecitizens with the lesser liberty.30The basic liberties with which the First Principle is concernedare specified by Rawls as follows.The basic liberties of citizens are, roughly speaking, politicalliberty (the right to vote and to be eligible for public office)together with freedom of speech and assembly; liberty ofconscience and freedom of thought; freedom of the personalongwith the right to hold (personal) property; and freedomfrom arbitrary arrest and seizure as defined by the conceptof the rule of law. These liberties are all requiredto be equalby the first principle, since citizens of a just society are tohave the same basic rights.31A liberty in the sense in which Rawls uses the term is definedby

a complex of rights along with correlative duties of others to aid ornot to interfere. Thus, by a restriction on liberty or an unequalliberty he means a restriction or inequality in what people are legallyentitled to do (or, perhaps,entitled to do by the nonlegalrules definingthe basic institutions of their society). Inequalities in people's ability29 This is an order which requires us to satisfy the first principle in the orderingbefore we can move on to the second .... A principle does not come into play untilthose previous to it are either fully met or do not apply. RAWLS43. For Rawls'

initial statement of the Second Principle, see text accompanying note 7 supra.30 RAWLS250 (footnote added).:l Id. 61.

1973] 1035


18/51

1036 UNIVERSITYOF PENNSYLVANIALAWREVIEW [Vol. 121:1020to take advantageof their rights due, e.g., to unequal economic meansdo not count as inequalities in liberty for Rawls but rather asinequalities in what he calls the worth or value of liberty. Whilethe basic liberties must be held equally, the worth of these libertiesmay vary since any significant inequality in wealth, income orauthority (allowed under the Second Principle) will represent aninequality in the ability of citizens to make use of their liberty inorder to advance their ends.32Rawls stresses at a number of points33the importance of preserving the fair value of the basic liberties,particularlypolitical liberties, but strict equality in the worth of theseliberties is not requiredby the First Principle itself.Two examples, frequentlycited by Rawls, of restrictions on basicliberties that are justified on the ground that they strengthenthe totalsystem of basic liberty are the restrictions on the scope of majorityrule imposed by a bill of rights and the restrictions on the freedomto speak imposed by a system of rules of order. In the first case arestriction of the legal powers of citizens is justified by the fact thatmore extensive powers could legally erase other basic liberties. Inthe second case what are sometimes called restrictions as to time,place and manner are imposed on the exercise of a basic liberty inorder, Rawls says, to preservethe worth of that liberty to all.34Thusit appearsthat while equal worth of the basic liberties is not requiredby the First Principle, securing the worth of these liberties is one ofthe goals which can justify restrictions on basic liberties under thePriority Rule.35

1. The Preference for Basic Liberties overOther Primary GoodsGiven the degree to which the content of the Priority Rule, andhence the claim of Rawls' theory to provide a secure basis for liberty,depends upon the distinction between the basic liberties and other

goods and opportunities, it may seem surprising that no theoreticalaccountof this distinction is offered. The list of familiarconstitutionalcategories given above is offered by Rawls not as a precise enumera-tion of the class of basic liberties but only as indicating roughly

321d. 204.33E.g., id. 224-26, 277-78.34Id. 203.35 Both of the cases considered here are examples of a lesser but still equal liberty

as provided for by clause (a) of the Priority Rule. Primary examples of unequal libertyallowed by clause (b) seem to be cases of justifiable paternalism in which a less thanequal liberty is acceptable to those whose liberty is restricted in the sense spelledout in the requirements discussed above. See text accompanying notes 19-20 supra.


19/51

RAWLS' THEORY OF JUSTICEspeaking what this class is to include. I suspect that here, as withthe class of primary goods itself, no precise theoretical demarcationcan be given. What is claimed for these liberties is just that, dueboth to the importance for anyone of the interests they safeguardand to their great instrumentalvalue for the enjoymentof other goods,they are not only things it is rational for anyone to want but alsothings it is rational for anyone to value particularly highly relativeto other primary social goods.It is not claimedthat these liberties are always to be valued morehighly than any other goods. Rawls allows that under particularlydire conditions, when bare survival or the pursuit of the means for aminimally comfortable life is the dominant concern, and when thenecessary prerequisitesfor the effective exercise of the basic libertiesare lacking, it may be rational to sacrifice basic liberties for the sakeof other goods such as increased security or economic development.It is under such conditions that the GeneralConceptionof Justice asFairness applies. Rawls argues,36however, that as conditions improveand the possibility for the effective exercise of the basic liberties be-comes real, people will set an increasingly high marginal value onbasic liberties relative to other goods. After the most urgent wantsare satisfied, people come to set greater importance on the libertyto determineand pursue their own plans of life. They will thereforeinsist on the right to pursue their own spiritual and cultural interests,seek to secure the free internal life of the various communitiesofinterests in which persons and groups seek to achieve . . . the endsand excellences to which they are drawn and, in addition, come toaspire to some control over the laws and rules that regulate theirassociation, either by directly taking part themselves in its affairs orindirectly through representativeswith whom they are affiliated byties of culture and social situation. 37Recognizingthese tendencies,theparties in the OriginalPosition will see that [b]eyond some point itbecomes and then remains irrational [for them] . . . to acknowledgea lesser liberty for the sake of greater material means and amenities.. . .,38 Thus the position of liberty underthe Special Conceptionmakesexplicit the priority that emerges under the General Conception asthe natural preference for basic liberties over increases in otherprimary social goods asserts itself.There are a number of questions one might raise concerningthisargument. First, since the appeal to an increasing preference for

36 RAWLS ? 82.37Id. 543.38 d. 542.

1973] 1037


20/51

1038 UNIVERSITY OF PENNSYLVANIA LAWTREVIEI' [Vol. 121:1020basic liberties over other primary social goods represents Rawls'most detailed claim about the way in which the parties in theOriginal Position would order bundles of primary social goods, itnaturally gives rise to questions of the sort considered above underthe heading of parameters. Rather than to consider the generalquestion of whether this preference is in some suitable sense uni-versal, however, it seems to me more profitable to ask whether anappeal to such a preference provides adequateand interesting answersto those questions about liberty (and about the particular basicliberties listed by Rawls) that one would want a philosophicaltheoryof liberty to answer.Foremost among these is the question to what extent the basicliberties have some kind of absolute status and to what extent, andwithin what limits, they are to be understood and interpreted interms of a balancing of competing interests. Rawls appears to havetwo answers to this question. The first, given by the Priority Rule,makes the limitation on acceptable balancing depend upon the dis-tinction between basic liberties and other primary social goods: basicliberties are to be limited only for the sake of the total system ofbasic liberty itself. The second answer, and the one most often usedby Rawls to indicate when a lesser but still equal liberty is just, isgiven by what he calls the Principle of the CommonInterest:

According to this principle institutions are ranked by howeffectively they guarantee the conditions necessary for allequally to further their aims, or by how efficiently theyadvance shared ends that will similarly benefit everyone.Thus reasonable regulations to maintain public order andsecurity, or efficient measures for public health and safety,promote the common interest in this sense. So do collectiveefforts for national defense in a just war.39

Rawls does not formulate this principle explicitly, but his discussion40suggests the following formulation: basic liberties may be restrictedonly when methods of reasoningacceptable to all make it clear thatunrestricted liberties will lead to consequencesgenerally agreed to beharmful for all.Rawls seems to hold41 hat these two doctrinesare consistent, i.e.that cases in which a restriction of basic liberties is justified by thePrincipleof the CommonInterest are also cases in which basic libertyis being limited for the sake of the total system of basic liberty itself.

39Id. 97.40ld. 213-14.41 Id. 246-47, 212-13,


21/51

RATWTLS'HEORY OF JUSTICEThis appears to be true in the most apocalyptic cases, e.g. cases inwhich a restriction of basic liberties is necessaryas part of the commondefense against an invasion. It may be true as well in some moremundane cases, such as Rawls' example of the restrictions imposedupon the right to speak by fair rules of order (taking into account, aswas noted above, that what is protected in this case is not, strictlyspeaking, liberty but rather the worth of liberty). But if the re-strictions on utterances imposed by such a set of rules count asrestrictionson a basic liberty, then so also must similar restrictionsonthe time, place and manner of political demonstrations, religiousfestivals, parades, the placing of posters and the use of loudspeakersand sound trucks. Regulation of these activities is normally thoughtto be acceptable, and appears to be justified by something like thePrinciple of the Common Interest, but it seems to me difficult tomaintain (without considerable stretching of the notion of a basicliberty) that in these cases basic liberties are being restrictedonly forthe sake of the same or other basic liberties. It seems to me muchmore plausible and straightforward o say that in order to arrive at apolicy in these cases we must balance the value of certain modesof exercise of a basic liberty not only against the exercise of otherbasic liberties but also against the enjoyment of other goods (unin-terruptedsleep, undefaced public buildings, etc.). Somethinglike thisis surely true in the case of the restriction of expression by lawsagainst defamation: different standardsof defamation for, on the onehand, private, artistic or culturalexpressionand, on the other, politicaldebate, seem to me obviously appropriate,and I take this to be thereflection of the differing values we place on the unfettered exerciseof these forms of expressionrelative to, among other things, the valueplaced on safeguardingthe primarygood of self-respect.One could of course maintain that what is balanced againstliberty in these cases is not liberty itself but the worth of liberty.Since almost anything, including any significant increase or decreasein material well-being, can affect the worth of liberty, the generalprinciple that basic liberties may be restricted only when this bringsan increase (or is necessary to avoid a decrease) in the worth of thetotal system of basic liberties appears to be a weaker principle thanRawls wishes to defend. I suggest that Rawls' response here wouldbe that while a great number of things can contributeto the worth ofliberty, not every restriction of basic liberty that yields gains in othergoodswill yield sufficientgains to constitute a net increase in the worthof the total system of such liberties. This is what a restriction mustdo in orderto be acceptable.

19731 1039


22/51

1040 UNIVERSITYOF PENNSYLVANIALAW REVIEW [Vol. 121:1020Conceivably, this principle can be made to fit the most obviouscases in which a restriction of basic liberty is justified. Given the

rather diffuse character of the notion of the worth of the totalsystem of basic liberties, however, it is not a principlethat is easy toapply. Under any accountthe decisionas to when a restrictionon basicliberty is justified will involve some difficult balancing, but I do notthink that a clear guideline between acceptable and unacceptablebalancingis obtained by describing everythingin terms of the worthof liberty. Such an approachmight seem inviting if one thought thatthe notion of an increasing preference for basic liberties over othergoods representedthe most important theoretical element in the casefor liberty. But I do not think that this is so. On the contrary itseems to me that the idea of an increasing preference for basicliberties leaves out or obscures the most importantfactors in the casefor certain of the basic liberties, factors which Rawls' own discussionof these particularbasic liberties brings out quite clearly.2. Freedom of the Person

The argumentfrom an increasingmarginalpreferencefor libertiesover other primary goods is most appropriateas an account of thebasis of freedom of the person. It is not completely clear fromRawis' discussionwhat this category of basic liberties is to encompassother than the protections against arrest and seizure embodied underthe rule of law, but I take it to include at least freedom of move-ment within the country and across its borders, freedom of choicein aspects of one's personal life, and perhaps also freedom fromsurveillance. The increasing preference for these liberties claimedby Rawls can be seen as deriving in part from the fact that theyrepresent important conditions for the use and enjoyment of othergoods. Beyond this, however, there is the fact that the interventionsthese liberties are intended to preclude constitute particularly deepintrusions into a person's life which anyone has strong reasons towant to avoid, both because of the real disruption they cause andbecause of their great symbolic impact.We can of course imagine people who felt quite differentlyaboutthese matters. To the extent that such differences are not merelythe object of speculative imagination but the subject of real dis-agreement and controversy, the force of Rawls' argument for thepriority of freedom of the person will be seriously weakened. Butin such an event it seems clear that the case for these liberties will begenuinely in doubt. Rawls' analysis of the case for the freedoms ofthe person as a matter of relative preference thus seems quite appro-


23/51

RAWLS' THEORY OF JUSTICEpriate; there is no obvious theoretical element in the case for theseliberties that his analysis leaves out.

3. Liberties of Expression, Thoughtand ConscienceFreedom of speech and assembly, liberty of conscience and free-dom of thought present a slightly differentcase. The argumentfor thepriority of these liberties rests upon the recognition by the parties inthe OriginalPosition that as material conditionsimprove there will be

a growing insistence upon the right to pursue our spiritual andcultural interests. 42As Rawls says in arguing for freedom of con-science, the parties mustassume that they may have moral, religious,or philosophical interests which they cannot put in jeopardy unlessthere is no alternative. 43Now this argument contains two distinguishable elements. Thefirst is the recognition by the parties in the Original Position that,for the reasons discussed in connection with the argumentagainst per-fectionism, they cannot concede to the government any authority inmattersof religious,moralor philosophicdoctrine. As Rawls says,The governmenthas no authority to render religious asso-ciations either legitimate or illegitimate any more than it hasthis authority in regardto art and science. These matters aresimply not within its competence as defined by a just con-stitution. Rather, given the principles of justice, the statemust be understood as the association consisting of equalcitizens. It does not concern itself with philosophical andreligious doctrine but regulates individuals' pursuit of theirmoral and spiritual interests in accordance with principlesto which they themselves would agree in an initial situationof equality.44

The second element is the recognition by the parties that they willcome to set a particularly high value on the pursuit of their spiritualand cultural interests.These two elements are clearly independent.To take the case ofreligion, the value that a group of people place on keeping theirreligious commitmentswill be reflected in such things as the amountof economic loss and disruptionof the pattern of life they are willingto undergo to allow everyone to observe the holidays of his religion,attend services, etc. and in the lengths to which they are prepared

42Id. 543.43Id. 206.44Id. 212.

1973] 1041


24/51

1042 UNIVERSITYOF PENNSYLVANIALAWREVIEW [Vol. 121:1020to go to recognizeand respect the religious scruplesof individualmem-bers against taking part in certain necessary tasks and activities. It iscertainly possible that the cost a society is willing to bear in orderto allow full freedom of religious observancemight vary widely whilethe principle of the lack of governmental authority to decide betweenparticular religious doctrines remainedquite fixed.This kind of variation in the value attached to religious obser-vance, while possible, may in fact be unlikely if, as Rawls says, [a]nindividual recognizing religious and moral obligations regards themas binding absolutely in the sense that he cannot qualify his fulfill-ment of them for the sake of greater means for promotinghis otherinterests. 45 This extraordinary importance attached to religiousmatters tends to overshadow the distinction I have tried to draw andmakes it inviting to rest the case for toleration entirely on the claimthat the parties in the Original Position can foresee that they willcome to set an incomparablyhigher value on religious liberty (i.e. onthe freedom to meet their religious commitments) than on otherprimary social goods. But this approachbecomes less attractive if wethink not only of religious liberty but of freedom of thought andexpression more broadly construed. A society is apt to set ratherdifferent values on the fulfillment of religious commitments, thepursuit of scientific knowledge and the pursuit and enjoyment ofexcellence in the arts, and these differences will be reflected in theprice the society is willing to bear in order to allow these activities togo forward. But in a society which recognizesfreedomof thought andexpressionthe regulationof these pursuitswill be guided by a commonprinciple that governments lack the authority to decide matters ofmoral, religious or philosophic doctrine (or of scientific truth) andhence also lack the authority to restrict certain activities on thegrounds that they promulgate false or corrupting doctrines. Let mecall this principle, which I have formulated only very crudely, thePrincipleof LimitedAuthority.Taken alone such a principle does not constitute a completedoctrineof liberty or even of freedomof thought and expression.Butit seems to me that this principle is the most important element insuch a doctrine that can be established from the point of view of theOriginalPosition. It is not possible to determine from that standpointexactly what relative values are to be assigned to these pursuits andto other interests with which they may conflict. Nor is it possible toforesee from that standpoint what will be the best way of regulating

45 Id. 207.


25/51

RAWLS' THEORY OF JUSTICEthese pursuits so that they do not conflict. These are problems thatcan be dealt with only at a later stage when the full facts about asociety and the preferencesof its members are known. (I suspect thatthis process of balancingand coordination is what Rawls has in mindwhen he speaks of restricting particular basic liberties in order tostrengthenthe total system of basic liberties.) While it may be possiblefor the parties in the OriginalPosition to foresee that in general theywill attach a high value to their spiritual and cultural interests, such ageneral preference, or a resultant general principle that in thebalancing process these liberties are to take precedence over othergoods, seems to me to be less useful as the basis for a doctrine offreedom of thought and expression than the idea that the process ofbalancing must take place within the constraints imposed by some-thing like the Principle of Limited Authority.A doctrine of freedom of expression founded on this idea issuggested by Rawls on a number of occasions, in particular in hisprincipleof the common interest, with its emphasis on the distinctionbetween what might be called neutral and non-neutral groundsfor restricting liberty. I think that some account of freedom of ex-pression of this general type must be correct, although there are anumberof difficulties n formulatingsuch a view.46While I have somemisgivings about Rawls' particular formulation (misgivings, e.g., asto whether too much may be conceded to the doctrine of clear andpresent danger by his blanket allowance that liberty of consciencemay be limited when there is a reasonable expectation that notdoing so will damage the public order which the government shouldmaintain 47), t seems to me one of the strong points of Rawls' theory(as described in the first part of this section) that it provides aphilosophicalbasis for an account of liberty of this type. It thereforeseems to me important to ask whether this strength is adequatelyrepresented n his doctrineof the priority of liberty.

While it is not explicitly stated in the Priority Rule, the Principleof Limited Authority will be implied by clause (b) of that rule if (asseems plausible on the basis of the argumentof the precedingsection)we take governmental authority over matters of religion, etc. torepresent an unequal liberty which would not be acceptable to thosewhose liberty is restricted.It is unclear,however,how this principleisrelated to the argument for the priority rule based on the increasing46 I have myself put forwarda view of this kind in Scanlon,A Theory of Freedomof Expression, 1 PHIL. & PUB. AFFAIRS 04 (1972).47RAWLS13.

1973] 1043


26/51

1044 UNIVERSITYOF PENNSYLVANIALAW REVIEW [Vol. 121:1020marginalvalue of liberty. There seem to be two possible interpreta-tions of this argument.While the parties in the Original Position might readily agreethat there are conditions under which the pursuit of spiritual andculturalinterests may be severely curtailedfor the sake of other morepressing needs, it may seem unlikely, given the close relation betweenthe Principle of Limited Authority and the conception of individualautonomy underlying the argument against perfectionism, that thepartieswould ever concedeto a governmentthe right to decide mattersof moral, religious or philosophic doctrine. This suggests an inter-pretation of Rawls' argument according to which the Principle ofLimited Authority applies under the GeneralConceptionof Justice asFairness as well as under the Special Conception.What distinguishesthe Special Conception,on this view, is just the increased importancethat is attached to spiritual and cultural interests as the opportunityto pursue these interests presents itself and the demands of meresurvival become less pressing. This interpretation is faithful toRawls' description of the transition from the General Conception tothe Special Conception as consisting of a shift in the ordering ofprimary social goods. But the Principle of Limited Authority is nota factor in this shift; it stands instead as a constant element of thetheory. Given the importanceof this principle from Rawls' point ofview, it seems somewhatsurprisingon this interpretationthat nothingresemblingthis principle is either stated or implied in Rawls' accountof the GeneralConception.An alternative, somewhat more extreme interpretation,and onewhichseems to me morelikely to representRawls' view, would identifythe Principle of Limited Authority as one of the distinguishingele-ments of the Special Conception. This means that there must becircumstancesto which the GeneralConceptionof Justice as Fairnessapplies but in which the parties in the Original Position would notonly allow the severe curtailment of expression on the groundsallowed under the Principle of the Common Interest but would alsosuspend the Principle of Limited Authority itself. I am not quitecertain what such situations would be like. Presumablythey would besituations in which cooperation on certain common tasks is notmerely mutually advantageous but essential for survival or for theamelioration of intolerable conditions. If deep disagreementswere toexist which made the basis of this cooperation fragile, and if closeand uninterruptedcooperation were required to avoid consequencesthat would be disastrous for all, then perhapsit would be rationalnotonly to accept rigid regulationon the time, place and manner of ex-


27/51

RAWLS' THEORY OF JUSTICEpression to prevent interferencewith essential work, but also to grantto the government the power to ban the expression of views likelyto give rise to dangerouscontroversyor to dissention and doubt.It seems to me most accurate to describe such situations as onesin which the circumstances of justice would be present only to alimited degree. Cooperation in certain tasks may be feasible andprofitable and in these areas of common purpose considerations ofjustice may apply, dictating,e.g., that the benefits and burdens of thiscooperation (including liberties and constraints) should be shared inaccordancewith Rawls' Second Principle of Justice. But if the basisof this cooperation s quite shaky, and if the ends at which it aims aretruly vital, then it might be rational for the parties involved to regardeach other primarily as means to these ends. This attitude would bereflected, for example, in the parties' placing the smooth functioningof their institutions ahead of the right of individual membersto raiseand discuss with each other questions about the wisdom, viability orpropriety of these institutions. I have some inclination to say thatsuch a case would not representcooperationon a footing of justice atall; collective actions to quell controversy in such circumstancesarebest seen not as the exercise of the distinctive authority of a justgovernmentin the sense defined by the Original Position, but ratheras acts which must be justified on a case-by-case basis by appeal tothe residual rights of the individuals involved to undertake thosemeasuresnecessary to their self-defense and survival.48However this may be, it is at least clear that the justification Ihave offered for limited tolerance in what might be called situationsof partial justice dependsupon the presenceof conditionsunder whichanything which undermines effective cooperationrepresentsan imme-diate threat to all. When these conditions are lacking, such justifica-tion is also lacking, and, in addition, it becomes rational for peopleto seek to establish cooperationon a footing that gives full recognitionto the status of the participants as autonomousequals, i.e. to some-thing like Rawls' Special Conception.One thing making this transition rational is the fact that underimproving conditions individuals will develop religious, moral andphilosophical interests and will want their institutions to safeguardtheir pursuit of these interests. But on the interpretationI have beendiscussing the Special Conceptionof Justice as Fairness can no longerbe seen simply as what emerges under the General Conception oncethese interests begin to develop. For the transition to the Special

48Such a view is suggested n Scanlon,supranote 46, at 224-26.

1973] 1045


28/51

1046 UNIVERSITY OF PENNSYLVANIA LAWTREVIEI [Vol. 121:1020Conceptioninvolves a fundamentalchange in the basis of cooperation,namely a move to what I called in the first part of this sectioncooperation on a footing of justice. Cooperationon this basis wouldbe less apt to be rational for people if they did not place a high valueon certain kinds of opportunity,but the definingelements of this formof cooperationgo beyond this configurationof preferences, just as thedefiningelementsof just cooperation n the economicsphere go beyondthe structure of needs and interests that make such cooperation in-viting.

III. CONTRA UTILITARIANISMA dominant place in twentieth-century Anglo-American moralphilosophy and, even more, in the normative thinking of Americanlegal and social theorists, has been occupied by the family of theoriescalled utilitarianism.Like the theories which, following Rawls, I havecalled perfectionist, these theories have a teleological structure. Thatis, they set out to define notions of right (the moralpermissibilityand

impermissibilityof actions, the justice or moral acceptability of socialinstitutions) solely in terms of tendencies to promote certain specifiedends (understood independently of considerations of right). In thecase of classical utilitarianism as espoused by Bentham49 his meansthat an action is held to be right (i.e. permissible) if and only if thereisno alternative action available to the agent at the time which wouldyield greater net balance of pleasure over pain (or, alternatively,greater total happiness). Social institutions and pieces of legislationare to be judged similarly in terms of their tendency to promote thisend. In determining the moral status of a course of action or aninstitution we are to take into account the happiness of every personaffected, giving each equal weight.It is useful to divide teleological theories into two groups on thebasis of the end which they call upon us to promote.What I will callhard teleological theories are those which define notions of right interms of the tendency to bring about certain states of affairs whichare not states of consciousness in individualsand which are held to bevaluable quite independently of their tendency to bring enjoyment,pleasureor satisfaction to individuals.Soft teleologicaltheories, on theother hand, are those which take as the end to be promoted somestate or states of consciousness of individuals, e.g. pleasurablesensa-tions of some kind.

49 See J. BENTHAM, INTRODUCTION TO THE PRINCIPLES OF MORALS AND LEGISLATIONch. 1 (1907).


29/51

RAWLS' THEORY OF JUSTICEHard teleological theories (some or all of which are what Rawlscalls perfectionist theories-I am not certain how broadly he intends

this term to apply) leave open the possibility that an overall netsacrifice in the quality of the lives of all persons now and in thefuture could be justified if this were required for promotion of thespecified goal. Thus, soft teleological theories, of which utilitarianismis the principal example, put themselves forward as a humanisticalternative.If the states of consciousness on which a utilitariantheoryis based are ones which all human beings are capable (in roughlyequal degrees) of experiencing,then that theory also has a democraticcast by comparisonwith perfectionist theories, in which all weight isgiven to the promotionof accomplishmentsof which only a few per-sons may be capable. Under utilitarianism everyone counts and iscounted equally. This does not at once mean that everyone is to betreated equally, however, for distributive considerations per se areirrelevant in the utilitarian scheme, and it is quite conceivable thattwo quite different distributions of material goods might yield thesame total of satisfactions and thus have equal claim to be promotedon utilitarian grounds. Natural empirical assumptions about the wayin which satisfaction is in fact produced suggest, however, thatequality of material circumstanceshas a specialclaim to the utilitarian'sattention.The principal empiricalassumptionhere is what is generallycalled the principle of diminishingmarginal utility: the principle thateach successive increment of a good produces a smaller increase insatisfaction than the preceding (equal) increment. If this is true forall of us and for all goods (hence generally for increments of realincome), and if differentpersons' utility functions (i.e. the curve re-lating amounts of goods to satisfaction produced) are roughly similar,then every step away from equality of real income will be disad-vantageousfrom a utilitarianpoint of view: the loss in satisfaction ofthe persons who receive a less than equal share will be greater thanthe gain in satisfaction for those whose real income is increased.

A parallel empirical argument provides a prima facie case for akind of libertarianismon utilitarian grounds. If each person is thebest judge of his own satisfactions, and if each person is naturallymotivated to choose those alternatives which promise him greatersatisfaction, then there is reason to think it good utilitarian policy toleave every personalone in those actions that concernprimarilyhimselfand affect others scarcely at all. It is worth pointing out that themotivationalassumptionmade here must be plausible if the utilitariantheory itself is to be plausible. The main intuitive argument for thetheory rests on the claim that what I have been calling satisfaction

1973] 1047


30/51

1048 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 121:1020

representsfor each person what is desirable in any event or course ofaction; the moral point of view, it is then claimed, just requiresof usthat we care equally about everyoneelse's fate in just the way we careabout our own, i.e. that we seek to maximize the total of their (andour) satisfactions, counting each for one and none for more than one.To a much greater degree than the other alternative views sofar considered,utilitarianismcan be formulatedand given a plausibledefense within Rawls' OriginalPosition construction.Although I haveso far been consideringthe classical version of utilitarianism in whichit is the total of satisfactions that is to be maximized, it will be ad-vantageous to focus on the arguments which can be offered in theOriginal Position for and against the principle of average utility, theprinciplewhich holds those social institutions to be morally preferableunderwhich the averagelevel of satisfactions is maximized.What the proponent of the principle of average utility claims,accordingto Rawls,50s that each party in the OriginalPosition shouldreason as follows:

The worth to me of a given allotment of materialgoods andsocial opportunities (i.e. the satisfaction that allotment willbring me) will depend upon my tastes, desires and abilities.All I can say without knowledge of these is that I want tohave that combinationof, on the one hand, social goods and,on the otherhand, tastes and abilities which will yield me thegreatest level of satisfaction. In the absence of any in-formation about my position in society or my tastes andabilities it is rationalfor me to assumethat if I am a memberof a particular society, the probability of my being a personin a certain social position is representedby the proportionof the total population who are in that position. Similarly,knowing nothing of my particularinterests and talents, it isrationalfor me to supposethat the probabilitythat I, if I ama person in a certain social position in a given society, willhave or develop certain interests and talents is representedby the proportionof the people in that position in the societywho have those interests and talents. It follows that the ex-pected value for me of being some person (I know not whichone) in a given society is representedby the average of thelevels of satisfaction enjoyed by people in that society.Consequently,it is rational for me to choose as the funda-mental principle by which the basic structureof my societyis to be judged the principle that material goods and oppor-tunities are to be distributedin such a way that the averagelevel of satisfaction is maximized.Rawls offers a number of argumentsin support of his claim that

50RAwLs 163-66.


31/51

RAWLS' THEORY OF JUSTICEhis own Two Principles (with the SecondPrinciple interpretedby whathe calls the Difference Principle) would be chosen over the principleof averageutility by the parties in the OriginalPosition. The argumentwhich has received widest attention is based upon an appeal to whatis known as the maximin criterion for rational choice under uncer-tainty. A choice under uncertainty in the technical sense in whichthis phrase is used in decision theory is a choice between alternativessome of which may produce a number of different outcomes depend-ing upon circumstanceswhich are beyond the chooser's control and towhich he has no positive grounds for assigning probabilities. Themaximin criterion specifies that in such a situation one shouldmaximize the minimum, i.e. seek to insure oneself the highestobtainable minimumby choosingthat alternativewhose worst outcomeis at least as good as the worst outcome of any other alternative.Applied to the problem of the Original Position this means that theparties' main concern should be to maximize the expectations of themembers of the worst-off class in their society. This leads them tochoose as a principle of distributive justice what Rawls calls theDifferencePrinciple.The maximin criterion is an extremely conservative principle ofchoice. Faced with a choice between two alternatives, the first ofwhich will yield us either one dollar or $100 depending on circum-stances the relative likelihood of which we cannot determine, andthe second of which will guaranteeus two dollars in either case, themaximincriterionbids us choose the latter alternative. Rawls argues5'that even though the maximin criterion may be too conservative tocommend itself for general adoption,nonetheless the circumstances ofthe particularchoice faced by the parties in the Original Position aresuch as to make a conservative principle of choice appropriate foranyone in that situation (and appropriateon objective grounds quiteindependent of subjective factors such as one's taste for gambling).These circumstancesare, first, the fact that the principles chosen inthe Original Position will determine the basic features of the societyin which the parties live. What is at stake here is not one choiceamong many of approximately equal import but a choice of morefundamentalimportancethan any other, one whose consequencesarenot likely to be erased or modified by subsequent choices or circum-stances. In addition, given the assumption that prevailing conditionsare ones which make social cooperationpossible (we are not talkingof two castaways fighting over one plank) it is presumably possible

51 d. 28.

1973] 1049


32/51

1050 UNIVERSITY OF PENNSYLVANIALAW REVIEW [Vol. 121:1020in the society to which the parties belong to assure everyone oftolerable conditions of life. If this is so, then the parties to theOriginal Position, while they naturally have an interest in improvingtheir condition above whatever level can be guaranteedfor all, shouldrationallybe more concernedto insure that they do not have to acceptconditions below this level, which could be entirely unbearable.Like the argumentfor the principle of average utility, this argu-ment addresses itself to the interest which the parties to the OriginalPosition have in doing as well for themselves in society as they can.Both arguments are thus examples of the reductive strategy whichRawls mentions as one of the primary strategiesof the OriginalPositionconstruction: to reduce the problem of determining principles ofjustice to a problemof individualprudentialchoice under uncertainty,a type of problemwhich has been extensively studied and one aboutwhich we may hope to have a clearer view. Rawls advances otherarguments for the choice of his Two Principles, including argumentswhich appeal to other aims of the parties, e.g. their aims to selectprinciples which it will be possible for them to uphold in the condi-tions of their actual society, to select principles which will support astable social order, and to select principles which will express theirnature as rational beings and provide a secure foundation for theirself-respect. I will discuss some of these arguments in the followingsection. In the remainder of this section I want to consider in moredetail the contrast between Rawls' principles and various forms ofutilitarianism.It is absolutely crucial to the argument presented above for theprinciple of average utility that a subjective quantity such as satis-faction rather than, say, primarysocial goods or some other objectiveindex of real income be taken as the measure of relative well-being.As an alternative to be consideredin the Original Position, a principlerequiring the maximization of the average share of primary socialgoods would totally lack plausibility; for, since there will be wildlydifferent distributions of such goods which yield the same average,the adoption of such a principle would representan extreme form ofgambling. Against the choice of this principle Rawls' argument forthe rationalityof conservatism n the choice of basic institutions seemsexactly right. But the effect of the standard empirical assumptions(i.e. diminishing marginalutility) about how satisfaction is producedby material goods and opportunitiesis greatly to narrowthe range ofpossible variation among those distributions of primary social goodswhich yield the same averagelevel of satisfaction. Since these assump-tions have quite a bit of intuitive plausibility for most people (perhaps


33/51

RAWLS' THEORY OF JUSTICEas much as the maximin criterionitself) this considerablyunderminesthe force of Rawls' argument.This reliance on the special propertiesof satisfaction, however, exposes the utilitarian to another line ofattack, one directed against the notion of satisfaction itself.In accordancewith the reductivestrategy I have mentioned,whatis appealed to in the argumentfor the principle of average utility isnot what might be called the ethical role of the notion of satisfaction,but rather its psychological role in the motivation and explanationofthe rational choices of a single self-interested individual. Here satis-faction (or in other formulations, happiness or pleasure) is supposedto represent that thing for the sake of which all other things aredesired, and, since other things are desirable in proportion to thesatisfaction etc. they yield, satisfaction serves as a common denomi-nator which makes possible rational choice between apparently dis-parate alternatives (a day at the races, a night at the opera, anafternoon in bed). Rawls argues at some length that neither satis-faction nor happiness nor pleasure nor any other psychologicalquantity is a plausible candidate for the role of dominantend, i.e.that thing for the sake of which all other things are to be desired.Wedo often speak of some things bringingus more satisfaction than others(or morepleasureor happiness) and of some courses of action turningout not to bring us satisfaction, but, Rawls argues, this should notbe taken as a reference to some psychological quantity which under-lies all our choices; rather, it is just another way of saying thatcertain things are to be preferred or that others do not, after all,seem to be preferableand that we are inclined to look for some alter-native. Rawls also arguesthat there is no necessity to posit a dominantend in order to give an account of rational choice between disparatealternatives. It is simply a fact that we can make such choiceswithout the aid of some psychologicalcommon standard. Rawls giveshis own account of how some choices of this kind are possible, anaccountwhich I have briefly sketchedin my discussion of the good fora person.52On Rawls' analysis, one method we may employ inchoosing between disparate alternatives is to ask ourselves whetherthey do in fact answer to some of the same ends and, to the extentthat this is so, which is preferable on these grounds. But this is notthe whole process of rational choice; there are many other methods.

Although I find Rawls' argumentson these points entirely com-pelling and will suppose them to be correct in the discussion whichfollows, I cannot rehearse them in detail here. It should be noted,

52 See Section II.A supra.

10511973]


34/51

1052 UNIVERSITY OF PENNSYLVANIA LAW REVIEW [Vol. 121:1020however, that Rawls' attack on the idea of a dominant end (satis-faction, pleasureor happiness) which is to be maximized n all rationalself-interestedchoices is not an attack on the notion of utility as it isused in moderndecision theory. In this theory (as to a large extent inRawls' own account of rational choice) it is the notion of preferencewhich is taken as primary.Decision theory uses the notion of a utilityfunction to describe and systematize a person's preferences, butmakes no claim to having explained those preferences by isolatingthat thing for the sake of which the preferred alternatives are to bepreferred.The idea that satisfaction represents that thing for the sake ofwhich all other things are to be chosen is invoked in a particularlystrong form in the argumentfor the principleof averageutility. In thisargumentthe particular tastes, desires, talents and abilities which aredeveloped and exercised in a life, the choices which determine thatlife and the circumstances n which it is lived are all treated as quiteincidental and secondary to the fundamental question of how muchsatisfaction the life promisesto the person who lives it. The utilitarianchooser in the Original Position takes no facts about himself intoaccount in making his choice of principles; he appraises alternativesnot from the point of view of a personwho will have some (at presentunknown) system of ends, but from the point of view of a subjectwho will enjoy one quantity or anotherof satisfaction.Now it may be said here that a person in the OriginalPosition ispreventedby the veil of ignorancefrom taking any facts about himselfinto account in making his choice of principlesand thereforethat

Scanlon - Rawls Theory of Justice (3)

Documents

Transcript of Scanlon - Rawls Theory of Justice (3)