plugin-conceptions on freedom in the political thougth of kant and rousseau

8/6/2019 Plugin-Conceptions on Freedom in the Political Thougth of Kant and Rousseau

http://slidepdf.com/reader/full/plugin-conceptions-on-freedom-in-the-political-thougth-of-kant-and-rousseau 1/29

Reason, Self-Legislation and Legitimacy: Conceptions of Freedom in the Political Thought ofRousseau and KantAuthor(s): Alexander KaufmanSource: The Review of Politics, Vol. 59, No. 1 (Winter, 1997), pp. 25-52Published by: Cambridge University Press for the University of Notre Dame du lac on behalfof Review of Politics

Stable URL:http://www.jstor.org/stable/1408116

Accessed: 24/04/2009 13:18

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR's Terms and Conditions of Use provides, in part, that unlessyou have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and youmay 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 athttp://www.jstor.org/action/showPublisher?publisherCode=cup .

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 thescholarly community to preserve their work and the materials they rely upon, and to build a common research platform thatpromotes the discovery and use of these resources. For more information about JSTOR, please contact [email protected].

University of Notre Dame du lac on behalf of Review of Politics and Cambridge University Press arecollaborating with JSTOR to digitize, preserve and extend access to The Review of Politics.

http://www.jstor.org

http://www.jstor.org/stable/1408116?origin=JSTOR-pdf

http://www.jstor.org/page/info/about/policies/terms.jsp

http://www.jstor.org/action/showPublisher?publisherCode=cup

http://www.jstor.org/action/showPublisher?publisherCode=cup

http://www.jstor.org/page/info/about/policies/terms.jsp

http://www.jstor.org/stable/1408116?origin=JSTOR-pdf



Reason, Self-Legislationand

Legitimacy: Conceptions ofFreedom in the Political Thought

of Rousseau and KantAlexander Kaufman

Rousseau and Kant both argue for contractarian heories of justice. In spiteof their common contractarianism, however, Rousseau and Kant argue forconceptions of legitimacy which differ markedly. The substantive differencesbetween their accounts of legitimacy, I suggest, illustrate the political implicationsof disagreement regarding the status of practical reason. Rousseau, in assigningreason to a merely instrumental status, anticipates both postmodern and empiricistskepticism regarding the power of reason to ground the choice of ends. Kant isthe forerunner of contemporary accounts of justice which reject such skepticalviews of practical reason. Rousseau's skepticism about practical reason ties his

criterion of legitimacy directly to the actual preferences of individuals. Kant'smore robust conception of practical reason (1) allows him to argue for a criterionof great generality and flexibility, but (2) ties the plausibility of his account oflegitimacy directly to the soundness of his conception of practical reason.

Contractarian theories of government ground politicallegitimacy n the consent, actual or hypothetical, f the governed.Legitimacy can only be plausibly grounded in a conception of

consent, however,f consent embodies he free

willingof citizens.

Thus, the conception of "freedom" r "free will" underlying acontractarian heory will necessarily determine he nature of thetheory's account of legitimacy. Both Kant and Rousseau arguefor contractarian heories of government. Moreover, oth theoriesground freedom in the agent's capacity for self-legislation:freedom s defined as obedience to laws one has prescribed oroneself.

In spite of their shared conception of freedom as autonomousself-legislation, Rousseau and Kant argue for conceptions of le-gitimacy which differ dramatically. Rousseau argues that onlylaws which are authorized by an entire people are just. A legiti-mate form of sovereignty must, therefore, provide a mediumthrough which the entire people can express ts will. Kant arguesthat laws are just if an entire people couldhave authorized hem.Thus, Kant's criterion f justice substitutes a hypothetical est for



THE REVIEW OF POLITICS

Rousseau's requirement f actual willing. This substitution hasprofound institutional mplications which sharply distinguishKant's political heory from that of Rousseau.

What can account for this dramatic contrast between thepolitical arguments of two contractarian heorists whose concep-tions of freedom appear so similar? In particular, why doesRousseau require hat the entire people must actually authorizelegislation, while Kant s content f each citizen couldhave willedthe

legislation?I will argue that this disparity derives from differences n theconception of reason, n its practical mployment, underlying heconceptions of freedom employed by Kant and Rousseau.Moreover, his disagreement nticipates central controversy ncontemporary olitical heory. Rousseau, n assigning reason o amerely instrumental status, anticipates both (1) postmodernskepticism regarding rationalizing narratives, and (2) the

empiricist ational hoice model's skepticism egarding he powerof reason to direct action. Kant, in assigning reason to afoundational and constitutive status, is the forerunner ofcontemporary accounts of justice and legitimacy which rejectsuch skeptical views of practical reason. Thus, the analysis, inKant and Rousseau, of theoretical roblems relating o the statusof practical reason addresses a controversy between skepticismand rationalism which remains entral n contemporary iscourse.

The postmodern claim that all foundational interpretiveschemes are merely historically ontingent onstructs1 rounds apervasive skepticism egarding ationalizing arratives nd thusa "positive attitude ... [to] undecidability and indeterminacy."2

1. These concerns are grounded most particularly n Heideggerian insightsregarding the givenness of the world from which being takes its possibilities "inaccordance with the way things have been interpreted by the 'they"' (Martin

Heidegger, Being and Time, rans. John Macquarrie and Edward Robinson [NewYork: Harper & Row, 1962], p. 239).2. Chantal Mouffe, "Democratic Politics and the Question of identity" in The

Identity in Question, ed. John Rachman (New York: Routledge, 1995), p. 43."Perhaps the most well-known short description of postmodernism is thatprovided by Jean-Francois Lyotard: 'incredulity towards metanarratives.' By'metanarratives'... Lyotard means those foundational interpretive schemes thathave constituted the ultimate and unquestioned sources for the justification ofscientific-technological and political projects in the moder world" (Stephen K.White, Political

Theoryand Postmodernism

[Cambridge:Cambridge

UniversityPress, 1991], pp. 4-5).

26



FREEDOM N ROUSSEAU AND KANT

This"ontology

ofdiscord,"3

nturn, engenders openness

to valuepluralism and "something like a moral-aesthetic sense ofresponsibility to otherness."4 Skepticism regarding rationalizingnarratives, however, constitutes a problematic ground for apractical political theory. If, for example, all legitimating narrativesare historically contingent language games,5 we still require anaccount of "what would count as the unity or success of suchnarratives."6 Yet determinate criteria for the unity and success of

a linguistic performance seem conceptually inconsistent with apositive attitude to undecidability: "political will and anyphilosophical attempts to evaluate it" must necessarily e entangledin the contingent and the arbitrary.7 n its purest form, postmoderntheory appears to allow for "no normative criteria, ... no realway of drawing distinctions between political ideals andmovements."8

Thus, White argues persuasively that any sustained discus-

sion of justice and collective action must imply "at least someelements of metanarrative."9 Since postmodern skepticismgrounds respect for a pluralistic diversity of values and prefer-ences, an acceptable metanarrative could, at most, mediate amongsubjectively held values and preferences in a way that respectsthe diversity of otherness. A theory of communicative action,whose criterion of legitimacy requires the "actual"10 acceptabilityof proposed norms, seems perhaps the most plausible practical

3. William E. Connolly, Political Theory and Modernity (Ithaca: CornellUniversity Press, 1993), p. 371.

4. White, Political Theory nd Postmodernism, . 20.5. "The observable social bond is composed of language 'moves"'

(Jean-Francois Lyotard, The Post-Moder Condition: A Report n Knowledge, rans.Geoff Bennington and Brian Massumi [Minneapolis: University of MinnesotaPress,

1984], p. 11).6. Robert B. Pippin, Modernism as a Philosophical Problem Oxford: BasilBlackwell, 1991), p. 159.

7. White, Political Theory nd Postmodernism, . 40, emphasis added.8. Ibid., p. 18.9. Ibid., p. 140.10. "The test is whether or not a proposed norm is acceptable in an actual

argumentation to all who are potentially affected by that norm" (Stephen K.White, The Recent Work of Jirgen Habermas [Cambridge: Cambridge UniversityPress, 1988], p. 49).

27




theory consistent with a postmoder orientation.1 Thus, as in thecase of Rousseauean political theory, a practical postmoderntheory must derive its substantive content from the "actual"subjective preferences expressed by individuals.

The empiricist model of rational choice "now orthodox amongeconomists and decision theorists"12 mbodies a similar skepti-cism regarding the power of reason to direct human action. Thismodel rejects the plausibility of the rational evaluation of ends:"the content of one's ultimate ends cannot be assessed as rationalor irrational."13 Reason plays a "strictly instrumental"14 ole inchoice: "reason must be the slave of whatever intrinsic prefer-ences one may have."'5

As in the case of postmoder theory, empiricist skepticismregarding practical reason grounds a criterion of social choice(the Pareto criterion) which mediates among given subjectivepreferences.16 While Pareto analysis constitutes perhaps the most

coherent formulation of this empiricist skepticism, the theorypresents, in their most salient form, the problems associated witha social evaluative criterion grounded in given subjective prefer-ences. First, Pareto analysis does not identify a unique preferredoutcome, since different Pareto efficient outcomes with differingdistributional impacts are feasible for a given economy.Microeconomic theory resolves this indeterminacy by specifyinga social welfare function which chooses the socially preferredoutcome from the set of feasible optima.

11. Habermas, thus, manifests attentiveness both to postmodem skepticism,and to the requirements of practical theory, in arguing that his account ofcommunicative reason "makes an orientation to validity claims possible" JiirgenHabermas, Between Facts and Norms, trans. William Rehg [Cambridge, MA: MITPress, 1996], p.5, emphasis added).

12. Allan Gibbard, Wise Choices, Apt Feelings: A Theory f Normative Judgment

(Cambridge,MA: Harvard

University Press, 1990), p.10.

13. Gibbard, Wise Choices, Apt Feelings, p. 10.14. "Instrumental rationality is all of rationality" ibid., p. 10). "The theory of

rational choice... treats practical reason as strictly nstrumental" David Gauthier,Morals By Agreement Oxford: Oxford University Press, 1987], p. 25). "Unlikemoral theory, rational choice theory offers conditional imperatives, pertaining tomeans rather than ends" (on Elster, "Introduction," n Rational Choice, ed. JonElster [New York: New York University Press, 1986], p. 1).

15. Gibbard, Wise Choices, Apt Feelings, p. 11.16. The Pareto criterion defines as optimal a state in which the subjective

preferences of individuals are satisfied such that no transfer increasing the

28



FREEDOM IN ROUSSEAU AND KANT

This solutionto the

problemof

indeterminacy, however,introduces a second problem. A social welfare unction s definedby the preference rderings of the relevant et of individuals; etpreferences an be based on a wide range of motivations, ome ofwhich will be more plausible as a basis for social choice thanothers. A social choice theory which cannot distinguish prefer-ences grounded in benevolent inclinations from preferencesgrounded n malice operates with "an impoverished onception

of individual preferences."17Rousseauean political theory addresses these problems bydefining legitimacy in terms of pure procedural ustice: lawsenacted by the entire people, and applying to each citizen inprecisely he same way, are ust. This account of legitimacy voidsindeterminacy, ince institutions are designed "so that the out-come is just whatever t happens to be."'8 In addition, Rousseauavoids the problem of malicious preferences by requiring hat

legislation must apply equally to all members of society.Thus, Rousseau's account responds o two serious objections

to political heories grounded n skepticism bout practical eason.Nevertheless, wo important bjections o a Rousseauean ccountof legitimacy remain. First, ince subjective nds are determinedentirely by inclinations, hey are temporally unstable. This notionof the temporal instability of subjective willing groundsRousseau's rejection of representation: legitimate legislativebranch must be constituted not of "representants"(representatives) ho would "usurp he functions of the people,"but of "commissaires" agents/trustees) who remain directly andcontinuously responsive to individual preferences SC 15 102).The requirement of continuous and direct responsivenessintroduces a serious practical impediment to the framing ofcoherent ong-term policy. Second, since legitimacy s grounded

satisfaction of any one person can be made without a corresponding transferdecreasing the satisfaction of another.

17. "'Utility information' can and should be seen to include informationabout why individuals want what they want" (Robert E. Goodin, "LaunderingPiefeieaces," in Foundations f Social Choice Theory, d. Jon Elster and AanundHylland [Cambridge: Cambridge University Press, 1989], p. 76).

18. John Rawls, A Theory of Justice (Cambridge, MA: Harvard UniversityPress, 1971), p. 85. Rawls's precise definition of pure procedural justice indicates

clearly why indeterminacy is not a concern for such an account of justice.

29




whollyin the

positive willingof individuals, no substantive

justification of norms or institutions will be possible. Yet if anethical tradition exists merely as a brute fact, agents may lackgrounds to identify with that tradition.19

If Rousseau anticipates modern skeptical quandaries, Kant isthe forerunner of contemporary contractarian heories groundedin a robust notion of reason's power to direct human action.20Kant's conception of practical reason allows him to retain

Rousseau's conception of freedom as autonomous self-legislationwhile arguing for an account of political right responsive to thetwo remaining objections. Kantian right avoids the problem ofthe temporal instability of preferences, since Kant's criterion oflegitimacy takes the form of a hypothetical test. Kant can, therefore,coherently ground arguments in favor of representativegovernment in a Rousseauean conception of autonomy. Inaddition, the ideal of explanatory transparency and justificatory

consistency which grounds Kantian politics is, arguably, sufficientto ground identification with an ethical tradition.Thus, while Kant and Rousseau argue from a similar

conception of freedom, their divergent accounts of legitimacyclearly illustrate the political implications of disagreementregarding the status of practical reason. Rousseau's skepticismabout practical reason ties his criterion of legitimacy directly tothe actual preferences of individuals. Kant's more robustconception of practical reason: (1) grounds a criterion of greatgenerality and flexibility, and (2) offers resources arguablyadequate to ground identification with an ethical tradition, but(3) ties the plausibility of his account of legitimacy directly to thesoundness of his conception of practical reason.

In order to underline the nature and significance of thisdisagreement, my analysis will focus narrowly on specification

19. See Taylor's helpful discussion of the ontological presuppositionsnecessary to ground a plausible account of willing identification with an ethicaltradition. Charles Taylor, Sources of the Self (Cambridge: Harvard UniversityPress, 1989), p. 165ff.

20. See, for example, Rawls, A Theory f Justice; Ronald Dworkin, "What isEquality? Part 2: equality of resources," Philosophy nd Public Affairs 10 (1981):283-345; T. M. Scanlon, "Contractualism and Utilitarianism" n Utilitarianism ndBeyond, ed. Amartya Sen and Bernard Williams (Cambridge: CambridgeUniversity Press, 1982), pp. 103-128.

30



FRFnOM IN ROUSSEAU AND KANT

of: (1) the precise conceptions of practical reason employed byRousseau and Kant; and (2) the influence exercised by theseconceptions of practical reason on the conceptions of freedom,and the related heories of justice, developed by each theorist.

Such a narrow ocus will necessarily ilter out much which isimportant n the writings of both theorists. I submit, perhapscontroversially, hat, n both cases, the underlying conception ofpractical eason played a crucial role in determining he charac-ter of the

political theoryarticulated.

Therefore,a careful

examination of the interrelations etween the notions of reasonand freedom employed can substantially enhance our under-standing of the political work of both theorists. n addition, suchan approach may help us to recognize the distinct character ftwo theorists whose political deas are often conflated.

Perhaps qually controversially, will assume hat Rousseau'spolitical hought s fundamentally onsistent. n particular, will

assume that Rousseau's arly and late political writings express acoherent and consistent body of political thought. While onehesitates o disagree with such eminent theorists as Shklar2' ndRiley,22 believe Dent23 s correct o caution against the commonassumption hat Rousseau was an unsystematic hinker.

Rousseau, himself, explicitly asserts he complete consistencyof his theoretical writings: "I wrote on diverse subjects, but al-ways with the same principles: lways the same moral, he samebelief, the same maxims, and, if you like, the same opinions" LB928). To the extent that a plausible account can be generated osupport Rousseau's laim, I shall favor such a reading. My argu-ment is not, however, dependent on this assumption. As I arguebelow, my central claims have textual support n both the earlyand late works.

I will also approach Rousseau's nstitutional proposals withless skepticism than many recent commentators.24 ousseau's

21. Judith Shklar, Men and Citizens. A Study of Rousseau's Social ContractTheory Cambridge: Cambridge University Press, 1969).

22. Patrick Riley, Will and Political Legitimacy: A Critical Exposition of SocialContract Theory in Hobbes, Locke, Rousseau, Kant and Hegel (Cambridge, MA:Harvard University Press, 1982), pp. 105-109.

23. N. J. H. Dent, Rousseau Oxford: Basil Blackwell, 1988).24. Shklar, Men and Citizens; Riley, Will and Political Legitimacy, p. 112-21;

James Miller, Rousseau, Dreamer of Democracy New Haven: Yale UniversityPress, 1984), pp. 185-210; Richard L. Velkley, Freedom and the End of Reason

31




account ofpolitical

nstitutionsrepresents pragmatic ttempt

oresolve the tension, which Riley persuasively stresses, betweenRousseau's "contractual heory of obligation and his model ofpolitical perfection."25 hus, for example, Rousseau's account ofthe role of the General Assembly guarantees hat: (1) legislationreflects the actual willing of all citizens; and (2) citizens willgenerally, rather than partially, when they legislate. WhileRousseau fails to resolve fully the tension between

contractarianism nd perfectionism, his institutional proposalsrepresent a determined effort to grapple with this issue.In Section One, I suggest that the salient similarities between

the political theories of Kant and Rousseau have obscuredconceptual distinctions with significant nstitutional mplications.In Section Two, I examine the grounding of each theorist'sconception of freedom n an underlying conception of reason. InSection Three, examine he criteria of justice grounded n these

conceptions of freedom. n Section Four, examine he theories ofsovereignty derived from these criteria o identify the influenceof the differing underlying conceptions of reason.

Two Accounts of Freedom

In The SocialContract, ousseau defines reedom as "obedienceto the law one has prescribed or oneself" (SC I/8:56).26 f anagent acted solely from his inclinations, rather than from

(Chicago: University of Chicago Press, 1989), pp. 38-39, 99. See Jean Starobinski,Transparency nd Obstruction (Chicago: University of Chicago Press, 1971); Dent,Rousseau; nd Arthur M. Melzer, TheNatural GoodnessofMan (Chicago: Universityof Chicago Press, 1990) for more sympathetic readings of Rousseau's account ofpolitical institutions.

25. Riley, Will and Political Legitimacy, . 107.26. Works by Rousseau are cited in the text. I have used the followingtranslations: Allan Bloom, Emile or On Education New York: Basic Books, 1979).Donald A. Cress, Discourse on the Origins of Inequality DOI) in Jean-JacquesRousseau: The Basic Political Writings, d. Donald A. Cress (Indianapolis: Hackett,1987). Judith R. Masters, Geneva Manuscript GM), and On the Social Contract SC)(New York: St. Martins, 1978). Frederick Watkins, Considerations n the Governmentof Poland in Jean-Jacques Rousseau: Political Writings, ed. Frederick Watkins(Madison: University of Wisconsin Press, 1986). Other citations from Rousseaurefer to the French text in the following editions: Lettre a C. Beaumont 1762),Edition de la Pleiade, vol. 4 (LB) (Paris: Gallimard, 1959); Lettres Ecrite de la

32



FRFDOM IN ROUSSEAU AND KANT

autonomous self-legislation, he would actually "have donesomething other than what [he] wanted. It is then that [he] wouldnot have been free" (SC IV/3:111). This conception, uggestingthat a free being must transcend he dictates of his inclinationsand formulate moral legislation or himself, is clearly related toKant's notion that freedom s the property of the will "that makesit effective ndependent of any determination y alien causes" G446/49). Thus, for Kant, as for Rousseau, freedom is the

independentwill

determininghe

practicalecisions of the actor.

Rousseau's conception, however, contains a qualificationwhich unequivocally distinguishes his view from the Kantiannotion of freedom: "the impulse of the appetite alone is slavery"(SC 1/8:56, emphasis added). While freedom requiresself-legislation, self-legislation does not require choiceindependent of "alien causes" such as appetites. Freedom s notnecessarily ncompatible with the inclinations. Rather, reedom

is self-legislation which is not determined by the inclinations"alone." Thus, Rousseau believes that when a free agentself-legislates, her choices are grounded in a mixture of reasonand impulse; reedom s not derived solely from the applicationof pure reason.

For Kant, such an account of freedom fails to address thecentral question of practical philosophy: What conception ofpractical reedom grounds an adequate account of the will of a

human being conceived of as a spontaneous agency (CPR A803/B831/634; CPr 64/66)?7 Rousseau rgues hat agents self-legislate

Montagne, Edition de la Pleiade, vol. 3 (LEM) Paris: Gallimard, 1959); "PoliticalFragments" (PF), Edition de la Seuil, vol. 3 (Paris, 1959).

27. Works by Kant are cited in the text, with the Academy page numberfollowed by that of the English translation used. I have used the followingtranslations: Lewis White Beck, Critique of Practical Reason (CPr) (New York:

Macmillan, 1956). James W. Ellington, Groundingfor he Metaphysics fMorals G)(Indianapolis: Hackett, 1983). Theodore M. Greene and Hoyt H. Hudson, ReligionWithin he Limits of Reason Alone (R) (New York: Harper Torchbooks, 1934/1960).Mary J. Gregor, The Metaphysics f Morals, ncluding The Doctrine of Virtue (DV),and The Metaphysical Elements of Justice MJ), Cambridge: Cambridge UniversityPress, 1991). Werner S. Pluhar, Critique of Judgment CJ) (Indianapolis: Hackett,1987). Norman Kemp Smith, Critique f Pure Reason CPR) New York: St. Martin'sPress, 1965). Citations from On the Common Saying: This May be True in Theory,But It Does Not Apply in Practice' TP) and Towards Perpetual Peace TPP) are fromthe H. B. Nisbet translation n Kant's Political Writings, d. Hans Reiss (Cambridge:Cambridge University Press, 1970).

33




freely merely by groundingheir

self-legislationn

natural, atherthan alien and corrupted, passions (Emile 212).28While suchself-legislation s not determined y the inclinations, Kant wouldconsider such self-legislation merely spontaneous, and thereforenot practically ree.29Kant argues that only practical freedom(that is, autonomy), n which the will chooses its ends throughpure reason, provides an account of the self-legislative capacityadequate o ground a conception of moral agency: only "a beingendowed with inner reedom, s regarded as a being that can beput under obligation" DV418/215). Therefore, he self-legislationwhich grounds the obligation of citizens in Kant's contractariantheory must instantiate practical, and not merely spontaneous,freedom.

The apparent similarity of the conceptions of freedom setforth by Rousseau and Kant has nevertheless encouraged theperception hat these two conceptions re substantively he same.

Thus, Ernst Cassirer has famously argued that freedom, inRousseau, constitutes "the overcoming and elimination of allarbitrariness, he submission o a strict and inviolable aw whichthe individual erects over himself."30 t follows that "man mustfind within himself the clear and established aw before he caninquire nto and search or the laws of the world."31

Thus, for Cassirer, Rousseau's heory of just rule relates o hisconception of freedom n the following fashion. First, reedom srational elf-legislation, ndependent of the urges of the inclina-tions. Second, the moral law is derived from the "necessarysolidarity"32 etween the will of the individual and the general

28. "Our passions are the principal instruments of our preservation ...Their source is natural.... But countless alien streams have swollen it ... Ournatural passions are very limited. They are the source of our freedom; they arethe instruments of our freedom; they tend to preserve us. All those which subjectus and destroy us come from elsewhere.... We appropriate them to the detrimentof nature" (Emile 212).

29. "Practical reason... [is] a causality of reason in the determination of thewill" (CPR A803/B831/634).

30. Ernst Cassirer, The Question of Jean-Jacques Rousseau, rans. Peter Gay(New Haven: Yale University Press, 1954/1989), p. 55.

31. Ibid., p. 57.32. Necessary because the individual wills rationally, and thus universally.

See Cassirer, The Question of Jean-Jacques ousseau, pp. 55, 124.

34




will which is rPali7Pd when the individual achievesrational

freedom.33 inally, he sovereign s "the orm n which the will, asethical will, really exists."3 Thus, the sovereign rpali7ps the nec-essary solidarity between the individual will and the generalwill.

Such solidarity or unanimity f willing), in Cassiver's ac-count of Rousseau, is possible only if each individual willsrationally and thus justly). Therefore, he criterion of justice in

Cassirer's description of Rousseau's ethical state must be therequirement hat laws and principles of government ould possi-blyhave been agreed to by an entire people. Rousseau's theory ofgovernment appears dentical with Kant's heory as articulatedin Theory nd Practice.3

Such a view of Rousseau's theory of governmentmisrepresents ousseau because t misunderstands is conceptionof freedom. Rousseauean reedom does not require hat the willact in complete independence rom all grounding nclinations.Rather, Rousseau simply requires hat such inclinations houldnot constitute he entire round of the will. Therefore, will arguebelow, since the determinations f individual wills are grounded,in significant part, in the particularistic nclinations of theindividual, Rousseau's criterion of justice cannot be ahypothetical est.

Freedom and Self-Legislation

Both Rousseau and Kant define positive freedom asautonomous elf-legislation. Each defines autonomous egislationdifferently, however, since each understands the relation ofrationality o autonomy differently. n this section, I will examinethe implications of these differing conceptions of autonomy

regarding: 1) rationality as a necessary ground of autonomouslegislation; 2) the role of reason as (a) an instrumental faculty or(b) an unconditioned faculty of autonomy; nd (3) the possibilityof identifying determinate moral ends of the autonomous will.

33. Ibid., p. 124.34. Ibid., p. 63.35. "If it is at least possible hat a people could agree to [a law], it is our duty

to consider the law as just" (TP 79).

35




Rationality as a Ground. For Rousseau, self-legislation is notprimarily an application of pure practical reason. Unlike Kant,Rousseau does not believe that man's freedom must be groundedin rationality alone. Rather, man "consults only his passionsbefore he acts"; reason "serves only to correct the follies" whichman would otherwise be led to by his passions (PF 554).

Reason is not intrinsic to the well-being or freedom of man inhis original state. In Rousseau's state of nature, man lived inisolation. Individuals met

rarelyand

parted company quickly.No lasting relationships existed. Man's desires were limited, and"his modest needs [were] ... easily found at hand" (DOI 46).Thus, social cooperation was unnecessary and unknown.

Man in this natural, nonsocial state, had no operational ratio-nality: "Willing and not willing, desiring, and fearing will be thefirst and nearly the only operations of his soul" (DOI 45). Reasonarises only from man's need to satisfy his desires:

the progress f the mind has been precisely roportionate o the needsreceived y peoples rom nature r to those needs which circumstanceshave subjected hem. DOI 6)

Rationality and intellectual inquiry constitute departures fromman's natural condition which produce dependence and unhap-piness:

Sinceall our errors ome rom our udgments, t is clear hat fwe neverneeded o udge, we would not need o learn... Wewould ehappier ithour ignorance han we can be with all our knowledge..." Of what impor-tance s this ome?" s the phrase most amiliar othe gnorant man andmost suitable o the wise one.... But unhappily his phrase does notwork or us anymore. Emile 04-205, mphases dded)

Although Kant, like Rousseau, asserts that autonomy isachieved

through self-legislation, Kant requires that this self-legislation must be grounded in reason. An autonomous will isdetermined "independently of sensuous impulses ... throughmotives which are represented only by reason." Man realizesthat he is free because he is capable of being motivated by pureprinciples:

we have the power to overcome he impressions n our faculty ofsensuous desire

by calling up representationsf what, n a more ndi-

rect manner, s useful or injurious. CPRA802/B830/633)

I6




An autonomous will can bedistinguished

from a determinedwill only if the autonomous will acts according to a nondeterminedprinciple of choice; therefore, the autonomous will must groundits choices in principles of practical reason. Kant specifies hisconception rational autonomy in the Formula of Autonomy, whichrequires that the agent

always choose in such a way that in the same volitions the maxims ofthe choice are at the same time

presentas universal aw. (G 440/44)

The formulation contains two requirements. First, the agent mustcreate the law which she will obey: "autonomy of the will is theproperty that the will has of being a law to itself" (G 440/44).Second, the law which the agent creates for herself must bewilled as universal law: "man ... is bound only to act in accor-dance with his own will, which is, however, a will purposed bynature to will universal laws"

(G 432/39).The principle of autonomy follows from combining elementsfrom the Formula of Universal Law and the Formula of Human-ity: "the ground of all practical legislation lies objectively ... inthe form of universality [and] ... [s]ubjectively ... in ... everyrational being as an end in himself" (G 431/38).

This combination of elements reveals a quality of the Cat-egorical Imperative which was only implicit in the first two

formulations: the compulsion to obey universal law is suppliedby the rational will of the individual. The Formula of UniversalLaw, in requiring that the agent act only on maxims which can bewilled as universal laws, recognizes the role of the rational will inoriginating legislation. The Formula of Humanity, in enjoiningrespect for "humanity," defined as "the capacity to set oneself anend" (DV 392/195), appears to view the rational will as the pointof origination for moral law.36 Only in the Formula of Autonomy,however, does Kant make explicit that the imperative quality ofthe law derives from the agent's own rational will.

Thus, Kant and Rousseau disagree upon the intrinsic charac-ter of autonomous self-legislation. Kant holds that rationality

36. In addition, he Formula f Humanity equires hat the individual reatall rational agents as ends. Since an individual cannot be compelled o adoptends, the compulsion must be supplied by the free will of the individual.

I37




must be the ground of autonomy. Rousseau argues that theexistence of a developed faculty of reason s evidence of enslave-ment: only a being burdened with needs and desires requiresreason.

Reason: nstrumental r Unconditioned aculty? ousseau viewsreason as an instrument which can serve the achievement ofcertain ends, rather han as an unconditionally aluable groundof freedom. Rousseau's nstrumental iew of reason s implicit n

his account of the original state of nature. Since man's desires inthe state of nature "do not go beyond his physical needs," manhad no need of reason until he had left this state: "it s impossibleto conceive why someone who had neither passions nor fearswould go to the bother of reasoning" DOI 46).

Departure rom this asocial, prerational tate led to the evilsassociated with civil association. Once men left isolation to jointhe company of their ellows, comparisons f talents and abilities

became natural. Individuals ound it necessary to adopt or as-sume socially desirable raits. Forced o assume a false character,man developed ostentation, eceptive cunning, "and all the vicesthat follow in their wake" (DOI 67). Man, who had been free,became subject to his fellow men "by virtue of a multitude offresh needs" (DOI 67). Finally, "consuming mbition" or socialsuperiority, arising from vanity rather han need, "inspire[d] nmen a wicked tendency o harm one another" DOI 68).

Instrumental eason thus became a tool of man's passions,continually stimulating awareness of new social needs, andincreasing his enslavement to the passions aroused by civilassociation. Reason is not, however, an inherently destructivefaculty. Reason can also "correct he follies" o which the passionslead man (PF 554). The faculty of reason assists man in pursuingor restraining is passions. In both cases, however, man "consults

onlyhis

passions"before

choosingthe

groundof his actions

(PF554). In neither case is reason essential to the function ofautonomous willing: "cold arguments an determine ur opinions,but not our actions" Emile 23).37

37. Note Velkley's persuasive argument that, for Rousseau, "reason neverloses its instrumental character" (Velkley, Freedom nd the End of Reason, p. 37).

38




For Rousseau, man's freedom s best secured by pursuit of anapproximation f the conditions of the state of nature. Only insuch a state could man be freed from the corrupting desires andpassions which were aroused by civil association. This state,however, cannot be regained by man within the civil order: "hewho in the civil order wants to preserve the primacy of thesentiments of nature does not know what he wants" (Emile 0).Since man lives within a civil order, and "will, n spite of himself,have to judge," Rousseau argues for an account of judgmentwhich best preserves he virtues of the natural tate: "let us teachhim, therefore, o judge well" (Emile 05).

The "best way to teach someone to judge well" involves"simplify[ing] ur experiences" Emile 05). Man's "original dis-positions" are the source of his "enlightenment" and his"capab[ility] f using his senses" (Emile 9). It is the corruption fthese dispositions by society which endangers man's capacity o

judgewell.8 Since

judgingwell is a matter of

"extend[ing]nd

strengthen[ing]" he natural dispositions Emile 9)and "put[ting]order and regularity into the passions" (Emile 219),39Rousseau'sacknowledgment f the necessity of judging does not involve achange in his notion of the status of reason. Rousseau's socialparadigm continues to involve an approximation f a state inwhich reason s neither needed nor exercised.

As a consequence of this conception of reason, Rousseau

argues explicitly against political analysis which "reasons n thesilence of the passions about what man can demand of his fellowman and what his fellow man has the right to demand of him"(GM I/2 161). Such reasoning, Rousseau claims, can provide noreliable guidance: "where is the man who can be so objectiveabout himself?" GM I/2 161). How could such rational ntro-spection be successful, even among those desiring to groundtheir acts in the general will:

38. "What makes man essentially wicked is to have many needs and todepend very much on opinion... the dangers of society make art and care all themore indispensable for us to forestall in the human heart the depravity bor oftheir new needs" (Emile 214). Thus, the best form of education delays exposure tosociety and opinion. "The child raised according to his age is alone" (Emile 219).The individual should be exposed to "high society" only once he is "in acondition to evaluate it himself" (Emile 222).

39. "Ourpassions

are theprincipal

instruments of ourpreservation" (Emile212).

39




When it would be necessary to consult the general will concerning aparticular act, wouldn't it often happen that a well-intentioned manwould make a mistake about the rule or its application, and follow onlyhis own inclination while thinking he is obeying the law? What will hedo, then, to avoid error? Will he listen to the inner voice? But it is saidthat this voice is formed only by the habit of judging and feeling withinsociety and according o its laws. It cannot serve, therefore, o establishthem. And then it would be necessary hat there had never arisen n hisheart any of those passions which speak louder than conscience, muffleits timid voice, and cause philosophers to assert that this voice isnonexistent. (GM I/2 161)

Thus, Rousseau denies the possibility of rational analysisindependent of the influence of (1) the passions, or (2) politicalcontext. In this passage, Rousseau rejects the fundamental as-sumptions of Kant's moral and political philosophy.

For Kant, reason constitutes the "faculty of systematicthought"40 which allows human consciousness to press beyondmere general formulations of what is given in sense data (rules ofthe understanding) in order to obtain a systematic comprehen-sion of experience (see CPR A302/B359/303).41 Since Kant assertsthat reason aims at "a priori rules" (CPR A571/B599/487), Kantexplicitly rejects a Rousseauean conception of reason as instru-mental and determined, in its objects, by the inclinations.

While theoretical reason seeks to uncover systematic order inobjects and relations in experience, practical reason groundsprinciples of choice which can determine the will. For Kant, theautonomous will acts, by definition, on principles grounded inpure practical reason. Practical reason, like theoretical reason,seeks "unconditioned conditions for all that is empiricallyconditioned."42 For practical reason, an unconditioned conditiontakes the form of a normative principle which (1) is constitutiveof the

experienceof freedom, and

(2) governsthe

adequacyof

potential grounds of free action.43 Thus, practical reason must

40. Lewis White Beck, A Commentary n Kant's Critique of Practical Reason(Chicago: University f Chicago Press, 1960),p. 23.

41. Kant defines reason as "the faculty which secures he unity of the rulesof the understanding nder principles" CPRA302/B359/303).

42. Beck, Commentary n Kant's Critique f Practical Reason, p. 50.43. "Pure eason establishes he goals of action hrough he formulation f

an intrinsically ractical nd unconditional aw. This s its real use" (Ibid., . 41).

40




specifya

principlewhich determines what

maycount as a reason

adequate to ground the choice of a maxim of action.44Korsgaard argues that, under Kant's account, the free will

must choose the principle of autonomy as the principle governingthe adequacy of grounds of action, since only the principle ofautonomy preserves the will's spontaneity.45 From the stand-point of spontaneity, no specific content is required of the principleto be chosen. The will must simply choose a law for itself. If the

will chooses the principle of autonomy (the Categorical Impera-tive), the only constraint on the will is that its maxims have theform of law.46 Thus, by choosing the Principle of Autonomy, thewill retains its spontaneity.

Kant distinguishes such autonomy of the will fromheteronomy, "where an object of the will must be laid down asthe foundation for prescribing a rule to determine the will" (G444/47). Heteronomous grounding of the will violates the re-

quirement of the Formula of Humanity to respect "the capacityto set oneself an end" (DV 392/195). In such a case, the will failsto propose its own ends. Rather, the will is determined by rela-tion of the material object to the will. If the relation of the object tothe will involves inclination, then the law governing the will isgiven by nature; the will is naturally determined.47 A determinedwill cannot propose ends to itself.

Kant's rejection of heteronomy as a ground of practically freewilling does not require that a particular choice to ground actionin an inclination must be inconsistent with autonomy. The agent's

44. See the helpful discussion in Henry E. Allison, Kant's Theory f Freedom(Cambridge: Cambridge University Press, 1990), pp. 136-45.

45. Christine Korsgaard, "Morality as Freedom" n Kant's Practical Philosophy,ed. Yirmiahu Yovel (Kluwer, 1989), pp. 23-48.

46. The principle of autonomy directs that the will "seek the law that is todetermine it ... in the fitness of its maxims for its own legislation of universallaws" (G 441/45). Since the imperative that self-legislation take the form ofuniversal law is not based on any interest, "it alone of all possible imperativescan be unconditional" (G 432/39). Thus, the determination of the autonomouswill is distinguished by "the renunciation of all interest" (G 431/38).

47. If the agent fails to prescribe laws for herself, relying on hypotheticalmaxims grounded in material objects, "the will does not give itself the law, but aforeign impulse gives the law to the will" (G 444/48).

41




rejectionof

heteronomys embodied in the choice of a

higherorder principle governing what may count as a reason groundinga maxim of action. The agent chooses a disposition [Gesinnung]embodying her free nature: "the disposition, i.e., the ultimatesubjective ground of the adoption of maxims ... appliesuniversally o the whole use of freedom" R 20).48A particularchoice to take an inclination r feeling as a reason or acting maybe consistent with a free disposition49 s long as the higher order

principle (which justifies the determination hat the inclinationshould count as a reason) s not itself grounded n the inclinations.Self-determination f the will in accordance with the rational

conception of a law is thus a precondition or autonomy. Onlyagents who are capable of acting in accordance with their con-ception of laws can transcend atural etermination y inclination.Only rational beings who transcend heir inclinations and willthe law for themselves can be said to be practically ree.

Thus, Rousseau and Kant disagree regarding he characterand function of reason. Rousseau argues that the will choosesactions based on the urgings of the inclinations. Reason merelyoperates to (1) facilitate he achievement of the will's ends; (2)restrain he will from acting on foolish inclinations; nd (3) de-velop new tastes and inclinations related to the will's existinginclinations. Kant argues hat only a will that chooses ts maximsand ends in accordance with a higher order principle groundedin practical eason can be said to be autonomous; hus rationalityis the defining characteristic f a practically ree will.

Determinate Moral Ends. Since, in Rousseau's view,self-legislation must be grounded n a combination f inclinationand reason, no specific maxims or ends can be identified asnecessary for the self-legislating agent. Rather, the choice ofmaxims for action and ends for pursuit s an empirical questionto be determined

bythe individual will alone, in the exercise of

its self-legislative unction.

48. See Allison, Kant's Theory f Freedom, p. 136-45.49. A material (rather than formal) principle "can ndeed remain" as a factor

affecting the content of free self-legislation. (CPr 34/35) "Kant clearly allows forthe possibility that morally worthy [i.e., free] actions might be prompted bynonmoral motives" (Ibid., p. 119).

42




faculty,while Kant views reason as constitutive of the

experienceof freedom. Finally, Rousseau views the ends of the will asinfinite and unspecified, while Kant argues that reason estab-lishes certain objective ends which rational beings must adopt.

Criteria of Justice

Rousseau and Kant both argue for criteria of justice groundedin the same principle: the legitimacy of law derives from theunanimous authorization of the entire people. Rousseau's theoryof justice takes the form of a practical requirement for the enact-ment of law: the principle of double generality.52 Kant's criterionis presented as a hypothetical test: a law is just "if it is at leastpossible that a [whole] people could agree to it" (TP 79). In thissection, I will examine each criterion and assess the extent towhich the character of the criterion is determined by the underly-ing conception of rational autonomy.

Double Generality. Rousseau's criterion of justice requires that"the object of the law should be general, as is the will dictating it"(GM 189). By this, Rousseau means that the law must be enactedby all citizens (generality of will) and must apply to all citizens inexactly the same way (generality of the object). This relationguarantees the just nature of the laws enacted, since "a bodycannot will to harm itself"

(LEM#6

807).Since each citizen

enacting law knows that the law enacted will apply equally tohim, the citizen will be constrained to will generally. Thus, willsthat would act out of self-love are converted into wills thatlegislate justly, willing for others what they will for themselves.

Rousseau asserts that this criterion of justice derives from"the nature of man" (SC IV/4 62). The collectivity of the citizens"cannot have any interest" contrary to the interest of all: "it is

impossible or the body ever to want to harm any of its members"(SC I/3 55, emphasis added). Moreover, each individual memberis constrained to will the good of every other member:

52. In Of The Social Contract, Rousseau introduces a second criterion ofjustice, which holds that all individuals in a certain category will be treatedsimilarly (SC 66). I will not discuss this criterion in this article.

44



FREEDOM N ROUSSEAU AND KANT

There s no one who does notapply

hisword each ohimself, nd doesnot think of himself as he votes for all ... the equality f right, and theconcept of justice t produces, are [thus] derived from each man'spreference orhimself nd consequently rom he nature f man. SC I/462)

This conception of justice is introduced, at least in part, tosolve the problem created by the requirement of "total alienationof each associate, with all his rights, to the whole" (SC I/6 53).

Since each participant in the Social Contract must alienate to thesovereign his complete rights and powers, a powerful mecha-nism is required to guarantee that laws enacted by the sovereignwill be just. Rousseau's criterion of justice is thus designed toguarantee the just quality of the laws adopted.

Rousseau's criterion requires concrete implementation. Thatis, the principle of justice can only be applied if the wills of allcitizens can be brought to bear on proposed legislation: the law

must actually be enacted by all citizens.The condition that all citizens actually enact the law is required

by Rousseau's conceptions of reason and autonomous legislation.53Since an autonomous Rousseauean person "consults only hispassions before he acts" (PF P1:554), he grounds his actions inheteronomous or contingent inclinations. While a hypotheticaltest could reconstruct the determinations of wills grounded inreason, no such test can reconstruct the determinations of willsfor which reason is merely a faculty instrumental to the realizationof ends dictated by the inclinations or interests. It is for thisreason that Rousseau describes the general will as "an admirableagreement between interest and justice" (SC 11/4:63).

The possible determinations of wills grounded in the inclina-tions alone are infinite and unspecifiable. This claim groundsRousseau's rejection of representation. The will cannot be alien-ated because the will cannot bind itself for the future: the

generalwill which directs the state must be "of the present moment"(GM 168). Since "power can ... be transferred, but not will,"therefore ndividuals cannot alienate their "sovereignty," or powerof willing, to representatives (SC II/1 59).

53. In addition, Rousseau's discontinuous onception of time led him toreject notions of representation. will not develop this point n this article.

45




Thus,Rousseau's

conceptionsof

autonomyand reason re-

quire that legislation be ratified through a concrete procedure.That is, if (1) the possible determinations f the will are infinite;and (2) the principle of double generality requires generality ofwill in enacting legislation; hen (3) double generality requiresthat all wills actually articipate n enacting egislation.

Hypothetical est. Kant's criterion of justice acknowledgesthat "the egislative authority can belong only to the united will

of the people" (MJ 313/78). Kant's development of this concep-tion reveals the tangible nfluence of Rousseau:

Because ll right and justice s supposed o emanate rom the egisla-tive] authority, t can do absolutely o injustice o anyone.... Henceonly the united and consenting ill of all-that is, a general nited willof the people by which each decides he same or all and all decide hesame or each-can legislate. MJ 13/78)

Kant's actual criterion of justice, however, utilizes a hypo-thetical est instead of requiring concrete egislative procedure.Thus, Kant transforms he principle of Rousseau's legislativemechanism nto a standard of justice: "if it is at least possible hata people could agree to [a law], it is our duty to consider he lawas just" (TP 79).

Kant is willing to accept the hypothetical test in place ofconcrete xpressions of individual wills because of his conceptionof the autonomous will. Since Kant believes that all autonomouswills ground their determinations in practical reason, allautonomous wills must reach the same conclusion whenevaluating the justice of specific legislation. Therefore, ahypothetical test of justice can serve in place of the explicitexpression of individual wills.

Conclusion. While Kant and Rousseau agree on thefundamental

principlefor

determiningthe

legitimacyof

legislation, each designs a different riterion of justice groundedin that fundamental principle. Rousseau, holding that thedeterminations f individual wills are grounded primarily n theindividual's inclinations and are thus unpredictable, requires aconcrete procedure through which all citizens can ratifylegislation. Kant, holding that all autonomous wills must reachthe same conclusion regarding the justice of a particular piece of

legislation, proposes a hypothetical est as his criterion f justice.

46



FRFEDOM IN ROUSSEAU AND KANT

Forms ofLegitimate

Rule

Kant and Rousseau argue in favor of forms of rule designedto achieve just conditions in accordance with their respectivecriteria of justice. Rousseau designs a legislative process whichwould allow all citizens o participate n the process of legislationthrough agents. Kant proposes the representative epublic as theform of government best suited for the achievement of ethical

progress.Rousseau: gents, Assemblies nd the General Will. Rousseau'sconception of legitimate sovereign power54 s premised on thenotion that freedom can be achieved only through "obedience othe law one has prescribed or oneself" (SC I/8 56). Rousseauperceives only two polarized alternatives or the sovereign. Thesovereign must consist of either: (1) the subjects hemselves, or(2) some entity entirely superior o, and thus different rom, thesubjects. The latter form of sovereignty s a danger to freedom,since the sovereign may then have interests contrary o those ofthe subjects, and thus an incentive o oppress the subjects.

Rousseau thus concludes that the sovereign power must beexercised by the entire body of subjects. The Social Contract mustthus design a form of sovereignty which

protects he person and goods of each associatewith all common orce,and by means of which each one, uniting with all, nevertheless beysonly himself and remains s free as before. SC /6:53)

Rousseau argues that these goals require he total alienationof the rights of each individual o the whole community. Such atotal alienation s necessary or two reasons. First, rights cannotbe reserved by individuals against the sovereign, because nocommon superior exists who could judge the limits of theserights against the community. Second, if individuals are givenreserved rights n some cases, the authority of the sovereign willbe damaged, and the entire system will unravel.

Although individuals alienate their full rights to the sover-eign, individuals retain certain protections against unjustoppression. These protections derive from: 1) the double gener-

54.Sovereign ower efers o the egislative ower, s distinguished romthe executive ower,whichRousseau alls hegoverment.

47




alityof the

legislative process;and

(2)the

periodicassemblies

which meet to ratify or reject (a) all legislation, (b) the form ofgovernment, and (c) the policies and personnel of the executive(SC III/13, 18:100-101, 106-107).

First, the double generality of the law, described in sectiontwo, guarantees the people, as sovereign, will enact legislationwhich they believe to be just. Second, the citizens, or their agents,possess the right not to ratify the laws as proposed. This right can

be exercised through the General Assembly.Rousseau provides two differing accounts of the GeneralAssembly. In the Social Contract, he General Assembly consistsof the people meeting in person (SC III/12 99). In a later work,Gouvernement e Pologne, however, the Assembly is made up ofthe elected agents of the people. Rousseau distinguishes sharplybetween "agents" and "representatives." Since Rousseau rejectsthe notion of representation of the will, the agents receive a

transfer only of the power to enact specific legislation on behalfof the people; such a transfer can be achieved through an impera-tive mandate, binding the agent strictly to the will of hisconstituents. Since the general will must be "of the present mo-ment" (GM 168), both (1) transfers of legislative power to theagents and (2) meetings of the Assembly must occur frequently.

Thus, Rousseau requires that the general will must be ex-pressed concretely because of his notion of wills determined bythe inclinations.55 f subjective determinations grounded in theinclinations are necessarily unpredictable, Rousseau's double gen-erality criterion can be satisfied only through concrete expressionof the subjects' wills under appropriate conditions.

Representative epublics. Kant grounds his contractarian theoryof the state in his account of practical reason. Every humanpossesses an innate right to freedom, Kant argues, "by virtue ofhis humanity" MJ 237/44). If individuals are to coexist

freely,the

exercise of freedom must be regulated by coercive laws to pre-vent the free acts of one individual from constraining the freedomof another individual (MJ 231/35-36, 232/37, 307/71, 312/76). If

55. The weak quality of rationality creates additional complications inRousseau's theory of government, since it requires the introduction of theLegislator to compensate for the flawed rationality of the general will, itself. I

will not develop the issue in this article.

48




right is to secure external liberty, Kant argues, all members ofcivil society must acknowledge the binding nature of principlesof right. No obligation to respect rights can exist unless eachmember of civil society is assured that all other members ac-knowledge and respect such rights (MJ 256/65).

Thus, external right can only obligate all members of civilsociety if it derives from a general, rather than merely unilateral,will. Since: (1) mutual freedom requires enforceable obligations

of external right; and (2) binding obligations of external right canonly exist within a civil society governed by a general will;therefore (3) mutual freedom can only exist within a civil societygoverned by a general will. Thus, the state can be understood asthe product of a hypothetical social contract under which indi-viduals unite into a general will and surrender their naturalfreedom in order to obtain the more complete freedom of amember of a civil society (MJ 315-316/80).

Finally, discreet civil societies can only maintain the conditionsnecessary for external freedom if perpetual peace among states ismaintained by a "union of states" (MJ 350/123). Perpetual peace,Kant argues, therefore constitutes the "highest political good"which constitutes "the whole of the ultimate purpose of law" (MJ355/128-129). Kant's notion of the highest political good derives,in great part, from his teleological view of history. If man obeyedthe imperatives of practical reason, a Kingdom of Ends, or ethicalcommonwealth, could be realized on earth. Since man ispathologically susceptible to the inclinations, he predictably willnot realize an ethical commonwealth. Since an ethicalcommonwealth could, theoretically, be realized, however, such acommonwealth remains an objective end which the moral lawconstrains us to pursue.

Kant asserts that the republican form of government is best

designedto

preserveindividual

freedom,since "without

[therepublican form of government], despotism and violence willresult, no matter what the kind of constitution in force" (TPP102). In addition, republican government is best suited to achievethe related goal of approximating perpetual peace, since if "theconsent of the citizens is required to decide whether or not war isto be declared, it is very natural that they will have great hesita-tion in embarking on so dangerous an enterprise" (TPP 102).

I49




Republicanism, in Kant's view, seems to constitute simplythe separation of powers: "Republicanism s that political princi-pal whereby the executive power (the government) is separatedfrom the legislative power" (TPP 101). Moreover, Kant criticizesthe democratic form of government as "in the truest sense of theword,... necessarily a despotism" TPP 101).

Although the combined weight of these statements mightseem to indicate that Kant opposed representative government,Kant, in fact,

requiresof a

legitimate governmentthat it be

"rep-resentative." Kant does not always distinguish carefully betweenhis uses of the terms "republican" and "representative." Thus, inTowards Perpetual Peace, Kant writes: "Any form of governmentwhich is not representative s essentially an anomaly, because oneand the same person cannot at the same time be both the legisla-tor and the executor of his own will" (TPP 101). Here, Kantclearly utilizes "representative" as a synonym of "republican,"

meaning the separation of powers.In the Rechtslehre, however, Kant appears to associate theidea of "representative" government with a legislative powercomposed of the elected representatives:

Every true republic is and can be nothing else than a representativesystem of the people if it is to protect he rights of its citizens in the nameof the people. Under a representative ystem, these rights are protectedby the citizens themselves, united and acting through their representa-tives. (MJ 341/113)

The representative form of government does not, on its face,appear to be required by Kant's universalizability criterion. Sinceall rational wills must reach the same conclusion when evaluat-ing the justice of a particular piece of legislation, Kant's conceptionappears compatible with a single legislative will. This possibilitymust be rejected, however, since no purely rational will can beknown to exist; all men are pathologically affected by the inclina-tions in some degree. Some procedural protection is required toprevent arbitrary egislation by the government.

Thus, Kant's argument appears to commit him to the prin-ciple of representative government. The form of the argument,however, is odd. Kant's account of justice, in the abstract, ap-pears consistent with enlightened despotism. If representation isa necessary element of Kant's political theory, it is only becauseman's nature is imperfect.

50




It is arguable, however, that a commitment to representativegovernment is implicit in the fundamental principles whichstructure Kant's moral and political theory. First, representativegovernment is consistent with the injunction of the Formula ofHumanity to treat humanity "always ... as an end and neversimply as a means" (G 429/36). Kant defines "humanity" as "thecapacity to propose an end to oneself" (DV 392/50). This rationalcapacity can be best developed through the exercise of the facultyof rational choice in matters of importance. A form of governmentwhich allowed the individual no choice in legislative decisionsmight stunt the individual's powers of rational self-determination.

Second, representative government is consistent with Kant'scriterion that a whole people could agree to the justice of aproposed law. If the assent of a whole people were sought for aform of government, it is likely that they would only consent to aform which reserved to the people certain rights and powers.56Such

powerswould be

perceivedas

necessaryto

protectthe

citizens from arbitrary or self-interested legislation.Finally, and I believe most persuasively, representative gov-

ernment is required by Kant's teleological politics. Kant notesthat the republican form of government is best suited to preserveindividual freedom, since without republican government, "des-potism and violence will result" (TPP 102).

The same argument holds true, perhaps to a lesser degree, of

representative government. Unless the interests of the citizensare represented effectively in government, any government maybe tempted to enact arbitrary legislation or to seek to achievedespotism. Thus, Kant's conviction that objective ends of reason(i.e., perpetual peace) can both exist and provide imperativemotivation to individuals serves as the basis for his theories ofpolitics and government.

Conclusion

Rousseau and Kant ground their theories of justice and legiti-macy of government in substantively differing conceptions ofreason. This disagreement leads to procedural differences in thecriteria of justice derived, and substantive differences in thetheories of government articulated.

56. As Rawls has famously argued.

51



52 THE REVIEW OF POLITICS

Rousseau's original conception of freedom as autonomy edhim to discover a principle of justice which has influenced Kant,Rawls and numerous ther heorists o the present day. Rousseau'sconception of the will as grounded primarily n the inclinations,however, led Rousseau o argue for a theory of justice requiringconcrete expression of the general will.

Kant adopted Rousseau's criterion of justice in his politicaltheory. Kant's onception f rational utonomy, however, allowed

him to transform he criterion rom a mechanism requiring heconcrete egislative action of all citizens to a standard by whichlegislation can be evaluated. The criterion, thus transformed,grounds Kant's heory of representative epublicanism. Kant, bysubstituting a notion of rational autonomy for Rousseau'sempirically grounded notion of freedom, provided a criterion ofjustice which could continuously guide the choices of thesovereign.

plugin-conceptions on freedom in the political thougth of kant and rousseau

Documents