C’étoit des deux côtés la même vertu guidée par différentes maximes.
Rousseau, Parallel between Sparta and RomeHow should institutions be designed so that the votes of the people reflect the general will and not the corporate will of the politically powerful? Rousseau's Social Contract provides us with two mutually exclusive solutions: eliminate pluralism through educational institutions, and manage it through electoral rules. Given that Rousseau ambitiously describes these as “the only precautions that will ensure that the general will is always enlightened, and that the people make no mistakes,” exploring these two solutions can shed light on the vexed question of voting the general will, particularly in political societies characterized by pluralism (SC, 60; 3:372).Footnote 1
To investigate these two solutions, this article explores the two primary examples Rousseau provides: Lycurgus's elimination of “partial societies” in Sparta and Servius's management of pluralism through electoral reforms in Rome.Footnote 2 By highlighting these two models, it aims to correct two tendencies in the literature concerning the Social Contract. The first is a general disregard for parts of book 4 and the Roman model presented therein. David Lay Williams says of SC 4.4–7 that “the dominant approach of most scholars to understanding these chapters … is simply to ignore them.”Footnote 3 Christopher Kelly concurs and suggests the following explanation: “Most commentators look at the practical details covered in Book IV as standing outside Rousseau's theoretical account of the first three books.”Footnote 4 If Kelly's assessment is accurate, then connecting the Roman institutions of book 4 with the problems of discerning and voting the general will articulated in books 2 and 3 will provide scholars with reasons to more carefully examine the Roman model.
The second is the tendency to view Sparta—a small, homogeneous, and illiberal city-state—as the primary, if not the only, model held up in the Social Contract. Interpreters often identify a single Sparta-Rome model that primarily embodies the unique features of the Spartan context. Judith Shklar is the best-known exponent of this view, explicitly reading the two cities together as a single “Spartan utopia.”Footnote 5 Subsequent accounts treat Sparta and Rome as exemplifying the same civic virtue,Footnote 6 “militant patriotism,”Footnote 7 and “unity and cohesiveness.”Footnote 8 To the extent that differences are briefly noted, they convey the message that Sparta was the model and Rome its less perfect approximation.Footnote 9 Kelly is an exception to this, but though he notes a number of differences between the Spartan and the Roman models, he offers no systematic comparison of their political institutions.Footnote 10
These interpretive tendencies have three primary implications for our understanding of the Social Contract. First, Rousseau's institutional proposals appear utopian. Political societies even in his time were much larger than Sparta and more pluralistic along religious, ethnic, and economic lines. From this, some conclude that Rousseau intended his utopia simply as a critique of existing societies.Footnote 11 In recent decades, scholars have turned to Rousseau's constitutional projects for Poland and Corsica as examples of his realism and pragmatic concern with institutional design.Footnote 12 However, this realist turn has largely avoided engagement with Rousseau's most famous political work, the Social Contract. A second consequence of the prominence of the Spartan model is that Rousseau's institutional proposals appear incompatible with liberalism. If Rousseau's principles of political right necessarily require Spartan institutions, then his critics would be right to fear the illiberal implications of his political theory.Footnote 13 Third and finally, the focus on Sparta has distorted the image of Rousseau's (in)famous legislator. This role has largely been associated with Lycurgus and his institutional work on eliminating pluralism. There are no scholarly works attending primarily to Rousseau's second category of lawgivers on the model of the Roman Servius who takes the second path of multiplying and equalizing the number of “partial societies” in the state.
This article comprises three parts. Part 1 explains Rousseau's concerns about partial societies and outlines the Spartan solution of eliminating pluralism. Part 2 briefly provides textual evidence that Rousseau saw Sparta and Rome as diverging on a number of aspects salient to his political theory. Part 3 turns to the Roman model of managing pluralism through political institutions. Although imperfectly successful, the Roman solution resembles liberal constitutional proposals for controlling the effects of factions through institutional design. I conclude by noting the presence of the two models in Rousseau's later constitutional writings for Corsica and Poland, and compare Rousseau's discussion of factions to Madison's. By focusing on the Roman model of legislation and the Roman political institutions described in book 4, I argue that we can identify a more liberal and realist model in Rousseau's Social Contract, one based on the Roman republic that Rousseau held up as “that model of all free Peoples” (SD 115; 3:112). I also challenge the utopian reading of the Social Contract, in part by identifying the continuities between the two institutional models of dealing with pluralism in the Social Contract and Rousseau's later political writings concerned with the practice of constitutional design.
1. The Problem of Partial Societies
My investigation of the two mutually exclusive institutional paths takes its bearings from Rousseau's discussion of “partial societies” in SC 2.3. This is the canonical chapter where Rousseau introduces the distinction between the general will (la volonté générale) and the will of all (la volonté de tous) and therefore between the general will and electoral procedures. There are a number of debates about the nature of the general will. Patrick Riley identifies a tension between its “voluntarist” and its “rationalist” elements,Footnote 14 while David Lay Williams distinguishes between “formal” and “substantive” dimensions.Footnote 15 The problem of partial societies described in this section appears across these conflicting interpretations of the general will. Take, for example, what Bertram calls the “democratic” conception, which “identifies the general will with the decisions of the sovereign people as they legislate together.”Footnote 16 The key political problem is distinguishing between procedures conducive to the general will and procedures reflecting the will of only a few. Alternatively, take what Bertram calls the “transcendent” conception, in which the general will is “a transcendent fact about the society which may or may not be reflected in actual legislative decisions.”Footnote 17 The question then is how this fact can be discovered through collective decision-making.Footnote 18 Because the problem of partial societies for identifying the general will does not depend on the particular definition adopted, this article does not directly intervene in the ontological debate. Instead, its focus is on the political question of how to design institutions and electoral rules in order to avoid the problems caused by factions.
Rousseau argues that, under the right conditions, the aggregation of individual votes can produce a reliable approximation of the general will through a (mysterious) procedure of canceling out pluses and minuses (SC, 60; 3:372).Footnote 19 The main problems emerge with the appearance of factions or “small associations at the expense of the large association” (SC, 60; 3:371). Rousseau argues that when such factions emerge, “there can then no longer be said to be as many voters as there are men, but only as many as there are associations” (SC, 60; 3:371–72). Within a faction, members look to the interest of their particular group instead of the interest of the entire political community. Their votes therefore no longer reflect their opinions concerning the general will. The smaller the number of factions, Rousseau notes, the harder it is to approximate the general will. In the extreme, where one single faction dominates the majority decision, “the opinion that prevails is nothing but a private opinion” (SC, 60; 3:372). Once the general will no longer corresponds to the will of the majority, the freedom of the political community is compromised regardless of which side obtains a majority of votes.Footnote 20
There are multiple ways to make sense of these passages about the problem of “partial societies.” One approach, pioneered by Grofman and Feld, is to read these passages through the lens of the Condorcet Jury Theorem (henceforth, CJT).Footnote 21 Under certain conditions, the CJT establishes that (a) the majority within a group has a higher likelihood of reaching a correct decision than any individual member; and (b) that the likelihood of a correct decision by the majority approaches certainty as the size of the group approaches infinity.Footnote 22 Not only that, but the convergence is rapid for even moderately sized groups. For the majority vote to correctly approximate the general will, the votes must be (statistically) independent and the average competence of the voters at discerning the general will sufficiently high. As Grofman and Feld have argued, Rousseau's description of the conditions under which the will of all can approximate the general will resemble the conditions for the CJT. The requirement that voters be informed indicates a concern for citizen competence, while the lack-of-communication requirement can be read as a concern with voter independence.Footnote 23 If individual voters primarily follow their faction or interest group, this could violate both key assumptions. Voters following the consensus of their partial associations are less likely to express their individual opinion about the common good, violating the independence condition. Similarly, voters primarily considering the interest of their faction or corporate group rather than the interest of the community as a whole would be answering the wrong question and therefore be less competent at identifying the general will.
However, one can understand the problem of partial societies without appealing to the formal mechanisms of the CJT. As Melissa Schwartzberg argues, Rousseau was influenced by natural law authors considering the moral implications of decision rules.Footnote 24 His discussion of factions makes clear that voters belonging to a partial society change the meaning of their vote: “So that instead of saying with his vote, it is advantageous to the State, he says, it is advantageous to this man or to this party that this or that opinion pass” (SC, 122; 3:438; emphasis in original). This change in orientation can be described as a form of voter corruption, where private or factional interests override the concern for the common good. In SC 4.1, Rousseau provides an example of electoral corruption in which the general will is no longer in the majority decision because selfish interests everywhere overwhelm the concern for the common good. Of this nonpropitious set of circumstances he asks: “Does this mean that the general will is annihilated or corrupt?” and answers “No: it remains constant, unalterable, and pure; but it is subordinated to others which have vanquished it” (SC, 122; 3:438). Although each citizen continues to have a general will that could be consulted, moral corruption leads to prioritizing the corporate will over the general one and the subsequent decline of the political community.
Regardless of whether one interprets the problem through the CJT or through the moral lens proposed by Schwartzberg, the concern with factions looms large in Rousseau's treatment of the general will. The dominance by a single faction spells the end of political freedom and the collapse of the social bond. It is in response to this concern that Rousseau proposes the two mutually exclusive institutional solutions. He writes:
It is important, then, that in order to have the general will expressed well, there be no partial society in the state, and every citizen state only his own opinion; this was the unique and sublime institution of the great Lycurgus. That if there are partial societies, their number must be multiplied and inequality must be prevented, as Solon, Numa and Servius did. These precautions are the only good ways to ensure that the general will always be enlightened and that the people are not mistaken. (SC, 60; 3:372)Footnote 25
In what follows, I briefly summarize the Spartan solution of eliminating all pluralism, then turn to the underexplored Roman solution of managing pluralism.
The Spartan Model: Preventing Pluralism
The Spartan solution is the most radical response to the problems potentially caused by “partial societies”: preventing the emergence of pluralism. As Shklar vividly describes it, the Spartan model sacrifices not only private associations, but any private aspect of life in competition with the civic identity: “The Spartan city excludes all private affections and associations, not only the family. It precludes contemplative and universal religiosity, as all inclinations are bent before xenophobia, communal isolation and pride.”Footnote 26 This thoroughgoing elimination of pluralism requires a demanding program of social engineering. This occurs through two channels. The first is “the pressure of public opinion,” active throughout the life of the citizen.Footnote 27 The second is “a rigorous public education that teaches and breeds patriotism.”Footnote 28 Kelly describes the Spartan model as producing “an unprecedented devotion to the community” by keeping citizens constantly occupied with public festivals.Footnote 29 He cites Rousseau on Lycurgus: “He imposed on them a yoke of iron, the like of which no other people has ever borne; but he attached them to it, identified them so to speak, with the yoke by always occupying them with it.”Footnote 30
Lycurgus designed the public institutions of Sparta in order to exert homogenizing pressures over the individual from birth to death.Footnote 31 Upon entering the world, children were examined by a governing body. Those determined insufficiently “robust” were dropped off a cliff (SD, 135; 3:135). The rest were subjected to a uniform public education program. In Life of Lycurgus, Plutarch reports that fathers in Sparta had no legal authority to raise their children as they saw fit. This resembles the model Rousseau proposed in the Political Economy, where he argued that fathers would only be able to make collective decisions about children's education (PE, 21; 3:260). This communal education project is key to Lycurgus's solution. Rousseau's attitude towards it is one of awe, oscillating between its positive and negative valences. In Political Economy, he describes Sparta as one of the few known ancient societies to practice public education and among which “it achieved wonders” (PE, 22; 3:261). In the First Discourse, however, he calls the policy of Lycurgus concerning education “in truth monstrous in its perfection” (FD, 22; 3:24).
The life of Spartan adults was no less constrained than that of children. The primary object of Spartan institutions was “war” (SC, 79; 3:393). Rousseau describes the all-encompassing nature of the mobilization efforts introduced by Lycurgus: “He constantly showed it the fatherland, in its laws, in its games, in its home, in its loves, in its feasts. He did not leave it a moment's respite to be by itself” (GP, 181; 3:957) The development of “partial societies” would be impossible to mask during the daily gatherings for communal meals and military training. The comically large iron currency made its secret accumulation similarly impossible. Public opinion and public censorship operated constantly on the mores of citizens, directing their attention exclusively towards their collective civic goals (SC 4.7).
The institutional model sketched above should be familiar to any reader of secondary literature on Rousseau. Its main advantage is that Spartans so educated would avoid the problems caused by factions. However, the Spartan model delivers these advantages with significant, if not unbearable, costs. It is not uncommon to notice the incompatibility of the Sparta model with the size and political economy of the states prevalent in Rousseau's time. Not only is the model inapplicable, but its near-complete sacrifice of the private sphere in favor of an all-encompassing civic life is incompatible with the normative aspirations of a liberal regime. According to Crocker, who reads Rousseau's Social Contract as a thoroughly Spartan project, the elimination of dissent and pluralism make it “scarcely necessary to underscore the difference between Rousseau's concept of law and liberty and that which obtains in all liberal and ‘open’ societies.”Footnote 32
2. Sparta and Rome: Similar, but Different
There are three key features associated with Sparta that do not apply in the Roman context: (1) its very small size; (2) its political economy in which even agricultural labor was held in low regard and performed by slaves; (3) the homogeneity of citizens along religious, ethnic, and economic dimensions. These aspects are of significant theoretical and practical importance for Rousseau's political theory. In this section, I briefly review these differences as a prima facie case for expecting that institutional solutions applicable in one context would not be applicable in another.
Sparta was a small polity even among ancient Greek city-states. Herodotus's History gives an estimate of five thousand Spartiate soldiers during the Persian War, while Plutarch's Life of Lycurgus estimates between four thousand five hundred and nine thousand Spartiates at its founding.Footnote 33 There is little doubt that Rousseau thought of Sparta as a small city-state. In a fragment labeled “History of Lacedaemonia,” Rousseau begins by saying, “It is the inhabitants of a country of such little extent whose history I am undertaking to write” (PF, 64; 3:544). In SC 3.12, Rousseau explicitly contrasts the small Greek city-states with Rome. He calls the Roman Republic “a large state” and the city of Rome “large” (SC, 110; 3:425). He supports this claim by citing ancient statistics concerning the Roman census, including approximately four hundred thousand armed citizens for the Roman Republic and four million soldiers for the Empire.Footnote 34 The emphasis on the large size of Rome is the entire focus of that chapter. The relevance of its large size is further reflected in the summary of the Social Contract presented in his simultaneously published book Emile. There, Rousseau lists among topics for the young man to discuss “whether the Roman populace was not a large populace” (E, 463; 4:843). Not only was Rousseau aware of the important size differences between Sparta and Rome, but he made it a focal point of the comparison between the two ancient polities. In an unpublished fragment comparing the two, he writes: “As for the size of the State, there is no comparison between these two Republics. Sparta, almost limited to its walls, could not even succeed in subjecting Greece, which was so to speak only a dot in the Roman empire” (PF, 63; 3:542).
A second salient difference concerns their political economy. Sparta is peculiar in the low regard for labor among its citizens. Not only did Spartans avoid arts associated with luxury, but they also avoided agricultural labor, which was done by the extensive population of Helot slaves. Plutarch tells of an Athenian who was punished for idleness. Upon hearing of this punishment, a Spartan sojourning in Athens is reported to have “desired his friend to show him the man who was condemned for living like a freeman.”Footnote 35 Rousseau agrees with Plutarch that the ability of Spartans to dedicate themselves so completely to their community depended on the complete enslavement of the Helot population. Of the small Greek city-states, he claims: “It [the people] lived in a mild climate, it was not greedy, slaves did its work, its chief business was its freedom” (SC, 115; 3:430–31). The case of Sparta was even here exceptional. Its “unfortunate circumstances” meant that for the Spartans “the Citizen can be perfectly free only if the slave is utterly enslaved” (SC, 115; 3:430). Without the extensive leisure obtained through Helot slave labor, the Spartan social engineering project would be impossible.
The Roman model illustrates the opposite approach to labor, especially agricultural labor. In the Social Contract, Rousseau argues that “since all of Rome's illustrious men lived in the country and cultivated the land, it became customary to look there for the mainstays of the Republic” (SC, 129; 3:446). He draws on both Varro's De re rustica and Pliny's Natural History for evidence that the Romans respected agricultural labor and preferred the rural farmers to the urban proto-bourgeoisie. He cites Varro as claiming that the village was “the nursery of those robust and valiant men who defended them in time of war and fed them in time of peace” (SC, 129; 3:446). He cites Pliny as confirming that “the rural tribes were honored because of the men who composed them” (SC, 129; 3:446). Rousseau attributes this Roman predilection to the work of an unnamed founder who “coupled freedom with rural and military labors” (SC, 128–29; 3:445–46). In the First Discourse, he claims that Rome was “founded by a shepherd and rendered illustrious by Tillers of the soil” (FD, 10; 3:10). Emile describes the Roman model of combining military service with labor as a standard for the pupil to emulate: “Remember that the Romans went from the plow to the consulate” (E, 474; 4:860).Footnote 36 One could write a separate paper drawing out the implications for property rights and prevailing social norms of these different attitudes to labor. For our purposes, it suffices to note that the Romans had neither the collectivist Spartan political economy nor the same amount of leisure as the Spartans.
Perhaps the largest distinction between Sparta and Rome involves the degree of pluralism. For the Spartans, the primary division was between the homogeneous group of Spartan citizens and the racially distinct population of Helot slaves. Because the Helots had no membership within the Spartan polity, this key division did not affect electoral politics.Footnote 37 In contrast with Spartan homogeneity, Rome was characterized by pluralism from its earliest period. Early Rome grew by conquest and alliances and incorporated the conquered people regardless of their religious worship or ethnicity. In the chapter “Of Civil Religion,” Rousseau attributes the spread of paganism to the Roman practice of incorporating the religion of conquered peoples (“They left the vanquished their Gods as they left them their laws”) or adopting the divinities while incorporating the people into Rome (“often having themselves adopted those [Gods] of the vanquished by granting them as well as their Gods freedom of the City”) (SC, 144; 3:461–62). The original division of the city of Rome into thirty curiae reflected this original religious pluralism: “all of the Roman people which was enclosed within the city walls at the time was composed of thirty Curiae, each with its own Temples, its Gods, its officers, its priests, and its festivals called compitalia” (SC, 130; 3:447).
Another of the famous initial internal divisions in Rome was the one between patricians and plebeians—a division that Machiavelli made the core of his analysis of Roman institutions. For Rousseau, this division was the most pernicious. He calls it “a vice inherent to the body politic” which resulted in “so to speak, two States in one” (SC, 123; 3:439). It is also the reason he cites for the difficulty Romans had in achieving institutional stability in the early Republic: “it did not assume a stable form from the first, because the failure to abolish the patriciate left the work only half done” (SC, 107; 3:421). In considering the electoral divisions of the Roman model, the preexisting pluralism along religious, racial, and economic lines serves as both a constraint and an affordance utilized by Servius in his reforms.
3. The Roman Model: Managing Pluralism
With a large and already diverse population, the legislators of the expansionist Roman Republic could not rely on a Spartan model of social engineering. In this section, I introduce the key features of Rousseau's Roman model that address the problem of factions without sacrificing pluralism. The discussion of Servius's reforms appears soon after Rousseau recapitulates his concerns about voting the general will and the problems caused by factions: “When earlier I showed how particular wills were substituted for the general will in public deliberations, I indicated clearly enough the practicable ways to prevent this abuse; I shall have more to say on this subject later” (SC, 124; 3:441). I take this passage to indicate that Rousseau's extensive discussion of the Roman voting assemblies represents his fullest statement on the question of voting the general will within a pluralist society. The mechanisms for tempering and checking the power of factions, particularly factions based on rank or wealth, primarily rely on Servius's role as legislator and the subsequent work of the Roman censors who acted as electoral officers in charge of maintaining and adjusting electoral districts.
Servius as the Model Roman Legislator
Although Servius is not mentioned by name in the famous chapter on the legislator (SC 2.7), he is the only legislator whose work is carefully documented in the Social Contract. Rousseau attends to every detail of Servius's reorganization of the Roman people into voting blocs based on residential, economic, and military criteria. These three divisions are introduced in part to correct the “dangerous abuse” of the original organization by Romulus, which allowed one segment of the population to become larger and more powerful than the rest (SC, 128; 3:445). However, once introduced, the divisions turn out to have wide-ranging electoral effects, some of which Rousseau judges more favorably than others.
The first of Servius's divisions provides a corrective to Romulus's division into three racially distinct tribes: Ramneses, Tatienses, and Luceres (SC, 128; 3:445). Early on, the tribe of Luceres grew rapidly by the influx of new members while membership in the other two tribes grew only at the natural rate of reproduction (SC, 128; 3:445). In order to prevent the Luceres from becoming a majority faction, Servius abolished the racial divisions and replaced them with a new division based on place of residence. He first organized the inhabitants of the city into four tribes, each one corresponding to a different hill, then added fifteen rural tribes based on as many rural cantons. This division was later supplemented until the total of thirty-one rural and four urban tribes was reached. Rousseau praises this new division on multiple counts. First, it provided an effective and immediate remedy to the dominance of the Luceres tribe (SC, 128; 3:445). Second, it prevented a similar inequality between tribes from occurring in the future since the residential division evenly divided the population into different districts.Footnote 38 Third, by forbidding movement between tribes Servius ensured that tribal identities would be stable and that inequality would not emerge as a result of strategic relocations into different areas (SC, 128; 3:445).
Not only did these divisions organize the people into coherent and flexible groups, but Rousseau admires their effect on the balance of power between country and city. He claims that “Rome owed to it [the division by tribes] both the preservation of its morals, and the growth of its empire” (SC, 128; 3:445). Instead of allowing the capital and the urban tribes to “arrogate to themselves the power and honors, and to lose no time in debasing the rural tribes,” Rousseau claims that the countryside remained dominant and dictated the course of both moral and political life in Rome (SC, 128; 3:445). Rousseau had noted the threat to both freedom and expediency created by powerful capital cities: “For each palace I see rise in the capital, I seem to see an entire countryside reduced to hovels” (SC, 112; 3:427). His suggestion for avoiding this dominance by the city resonates with the division into tribes adopted by Servius: “People the territory evenly, extend the same rights throughout it, spread abundance and life throughout it, that is how the state will at once have as much force and be as well governed as possible” (SC, 112; 3:427).
In addition to the new division by tribes and the old division into thirty curiae which he left basically unaltered, Servius introduced an entirely new division based on wealth. The division included six classes and 193 centuries. The first class, composed of the wealthiest citizens, was divided into ninety-eight centuries. What Rousseau labels the moderately wealthy classes controlled another ninety-four centuries, while the poorest class had a single century. If this division was fully in line with the advice to multiply partial societies, it certainly did not follow the second prescription of preventing inequality among them. Rousseau is well aware that the centuries were of highly uneven size, which resulted in uneven electoral power for their members and disproportionate power for the wealthy: “Thus it came about that the class with the smallest number of men had the largest number of centuries, and that the entire last class counted as only one subdivision although it alone contained more than half the inhabitants of Rome” (SC, 130; 3:447). Throughout his analysis, Rousseau refuses to pass judgment on “whether this third count was in itself good or bad” (SC, 131; 3:448).Footnote 39 Instead, he notes that it became “the most important division of all” as a result of its effects (SC, 130; 3:447).
The importance of these divisions is revealed through Rousseau's discussion of their electoral consequences. These different rules for collecting and counting votes can strongly influence the character of the assembly: “these various divisions [of the population] were not simply forms indifferent in themselves, but … in addition to determining the order in which the votes cast by such a large people were counted, each one of them had effects relative to the opinions that led to its being preferred” (SC, 134; 3:452). The division by tribes was most conducive to popular or democratic government.Footnote 40 Rousseau describes the assembly by tribes as “properly the council of the Roman people” (SC, 134; 3:451). This assembly was convened and presided over by the tribunes of the people. He disapproves of the formal exclusion of senators and patricians, but notes that even with their formal inclusion, the method of one person one vote would have led to their opinions counting for as little as that of “the least proletarian” (SC, 134; 3:452). If the assembly by tribes was most conducive to popular government, the assembly by centuries was most conducive to aristocratic government. In the following sections, I note Rousseau's careful discussion of institutional ways to temper both the aristocratic and the democratic assemblies in order to avoid the dominance of factions and check potential abuses.
Factions based on Rank and Wealth
For Rousseau, aristocratic government could lead to either the best or the worst results. The best form of aristocratic government is elective, particularly when “probity, enlightenment, experience, and all the other reasons for public preferment and esteem are so many further guarantees of being well governed” (SC, 93; 3:407). However, even under elective governments there are significant concerns about the transformation of the government into a corporate body whose will no longer reflects and implements the general will (SC, 93; 3:407).Footnote 41 Not only is even the best aristocratic government at risk of partiality and factionalism, aristocracy also carries with it the risk of decline into the worst form of government, namely, a hereditary aristocracy, where the ruling elites are guaranteed to pursue their factional interests at the expense of the common good.
In Rousseau's discussion of the assembly by centuries, his first concern is precisely the power of the hereditary aristocracy (i.e., the patricians in the senate).Footnote 42 As he puts it, “one is left to wonder why the senate did not always prevail in that comitia which bore that name and which elected the consuls, censors, and other curule magistrates” (SC, 133; 3:450). Had the patriciate been successful in imposing its will in the assembly by centuries, this would have translated into control over other key offices in the state and therefore almost unchecked power. Fortunately, Servius's institutional design included checks that prevented the dominance of the patriciate in the assembly. Although the patricians were a powerful group, they were not necessarily the majority in each of the centuries of the first class. Rousseau claims that “the tribunes ordinarily, and a large number of the plebeians always, were in the class of the rich and thus balanced the influence of the patricians in the first class” (SC, 133; 3:451). By preventing a single patrician faction from dominating the centuries controlling a majority of the votes, this balance of power between patricians and plebeians could also prevent the same faction from imposing its private will on the whole electorate.
In addition to the concern about the power of the patricians, Rousseau was concerned about the power of the wealthy, whether patrician or plebeian. He rejects Aristotle's analysis that the rich would always be preferred in aristocratic regimes and argues that “an opposite choice should occasionally teach the people that men's merit offers more important reasons for preference than do riches” (SC, 94; 3:408). The division of the population on the basis of wealth exposed the assembly by centuries to the dominance of the rich. The primary check on the wealthy was the censorship.Footnote 43 Previous discussions of censors in the Roman model have emphasized their moral function, connecting their role to censorship of the arts. Rousseau's discussion, however, emphasizes their role as electoral commissioners in charge of redistricting. The censors had the power to reassign citizens to different centuries on the basis of their behavior. According to Rousseau, “some men found themselves relegated to the class of the poor for having made an excessive display of their riches” (SC, 131; 3:448). This relegation would serve a dual purpose. First, it would diversify the economic composition of the different centuries, operating against the possibility of shared interests within groups. Second, it would eliminate precisely the leaders of potential factions from the group, limiting their influence on citizens of similar wealth and status.
By noting the way in which censors could limit the development of factions in the assembly by centuries, one can get clearer about Rousseau's attitude towards the military cast of this division. Rousseau notes Servius's attempt to disguise his electoral reforms as a military operation without taking a clearly critical stance.Footnote 44 One of the reasons might be that the military conceit allowed censors to more frequently reconsider the electoral districts and their composition. Rousseau explicitly draws this connection in his discussion: “In each Class, except the last, he [Servius] distinguished the young from the old, that is to say those who were obliged to bear arms from those whose age by law exempted them from bearing them; a distinction which more than the distinction by wealth made it necessary frequently to take a new census or count” (SC, 130–31; 3:448). Because of the frequent transitions from youth to adulthood and from adulthood to old age, the censors had the option of frequently reconsidering the electoral divisions and adjusting them in cases of concern over factional interests. Of course, the success of this measure depends on censors being willing to make the necessary adjustments, which in turn requires a general agreement on the inappropriateness of economic factions. Otherwise, the censors could gerrymander the voting districts according to their own private economic interests. The moral aspects cannot fully be removed from the equation.Footnote 45 In this, Rousseau agrees with Madison that no electoral checks can make up for a complete absence of civic virtue.Footnote 46
Factions of the Poor
As a democratic assembly, the assembly by tribes could provide a better approximation of the general will in its decisions. As Rousseau notes, “the more numerous the body of the magistrates, the more closely does the corporate will approach the general will; whereas under a single magistrate, the corporate will is, as I have said, merely a particular will” (SC, 89; 3:402). However, this form of government also carries potential disadvantages.Footnote 47 With respect to factions, one concern is the susceptibility of the poorest citizens to bribery.Footnote 48
The selling of votes was one of the key issues that brought down the Roman Republic. When Rousseau expounded on the dangers of extreme wealth and extreme poverty, he had noted that “it is always between these two that there is trafficking in public freedom; one buys it, the other sells it” (SC, 78; 3:392). In accepting bribes, citizens would be evading the general will and changing the meaning of their vote. This form of corruption appears most clearly in Rousseau's discussion of the assembly by curiae. After Servius's organization of the population into tribes and centuries, the old division by curiae remained as a separate organization of the urban population. Rousseau found the assembly by curiae to be the least successful of the Roman voting assemblies. He claims that the urban tribes “sold the state to those who deigned to buy the votes of the rabble who composed them” (SC, 130; 3:447).Footnote 49 Given the widespread bribery and corruption, this assembly eventually fell into discredit in the republic as even “seditious parties” became unwilling to show their designs through such an openly corrupt venue (SC, 135; 3:452).
In the assembly by tribes, the citizens most likely to accept bribes were confined to the four urban tribes out of a total of thirty-five. Maintaining this pattern once again required the active involvement of the censors. As Rousseau explains it, the censors initially followed an “excellent” maxim of maintaining the virtue of the rural tribes and reassigning citizens whose dedication to the common good was in doubt (SC, 129; 3:446). The most noble families, such as that of Appius Claudius, were automatically enrolled in the rural tribes. Those who showed courage and honorable behavior remained in these rural tribes while “cowards whom they wanted to degrade were transferred to the urban tribes as a disgrace” (SC, 129; 3:446).Footnote 50 However, this control over the shape of electoral groups eventually turned into a form of abuse. Rousseau attributes the decline of the tribal assembly to the capriciousness of censors who “permitted most of them [citizens] to enroll in whichever one [Tribe] they pleased; a permission which was certainly not good for anything” (SC, 129; 3:446). The result of this change was that “the tribes generally no longer had a district or territory” and “the idea of the word tribe thus shifted from the residential to the personal, or rather it became almost a chimera” (SC, 129; 3:446).Footnote 51 If membership in tribes was determined by individual preferences, one of the consequences would be a selection effect with individuals from the same faction strategically assigning themselves to tribes.Footnote 52
Conclusion: Rousseau's Roman Model beyond the Social Contract
Many interpreters of the Social Contract have focused on Sparta as the primary model for Rousseau's political theory. Spartan institutions certainly feature prominently across Rousseau's political writings. Its small size, homogeneous population, and extensive leisure based on slavery made Sparta uniquely able to resist the potential problems characteristic of pluralistic societies. Given Rousseau's unquestionable pessimism about modernity, the “monstrous perfection” of the Spartan model carries a powerful appeal. At the same time, however, Rousseau's interest in the science of politics extended beyond criticism and utopian theorizing. In the famous opening lines of the Social Contract, he claimed to be searching for “a legitimate and reliable rule of administration in the civil order, taking men as the are, and laws as they might be” (SC 51; 3:351).Footnote 53 A key part of this project involves close attention to the particulars of institutional design, particularly as they pertain to discerning and approximating the general will under conditions of pluralism. As Kelly suggests, “Book IV is not a mere appendage to Rousseau's theory. Institutions and practices that aim at maintaining the constitution are a part of the theory.”Footnote 54 This article's recovery of the Roman model suggests that scholars have been mistaken to avoid the Roman institutions as either peripheral to or inconsistent with the rest of the Social Contract. Instead, these chapters offer a new way to read key aspects of his political theory: the role of the legislator, the acceptable limits of pluralism in a democratic society, and the role of checks and balances in a liberal theory of government. As I argue in this conclusion, the mutually exclusive approaches to pluralism characteristic of Rousseau's two model republics extend to his later constitutional writings for the small and fortuitously situated island of Corsica and the larger and pluralistic republic of Poland. Attending to these institutional details throughout Rousseau's political writings reveals a more complex, more liberal, and more realistic approach to politics than interpreters have generally acknowledged.
If Sparta is a model for the priority of patriotism among a small group of equals, Rome is a model for the priority of liberty in a larger and more pluralistic polity.Footnote 55 The Roman model and its emphasis on liberty are particularly prominent in the Social Contract. The epigraph to the work comes from Virgil's Aeneid, a story about the mythical origins of Rome's ancestors. The frontispiece shows a woman surrounded by Roman symbolism.Footnote 56 Not only does the frontispiece speak of the Roman rather than the Spartan model, it also specifically emphasizes features of Roman political life that were distinct from the Spartan one: manumission, liberty, individualism. Finally, the majority of book 4 is dedicated to a discussion of the minutiae of Roman institutions and the legislative work of Servius. Although this article aims to preserve both the Spartan and the Roman models as independent solutions to the problem of pluralism, both of which Rousseau expressed admiration for under different circumstances, there is strong evidence that Rousseau considered the institutions of Rome more relevant as a model to describe to his contemporaries.
Take the example of Rousseau's (in)famous legislator. Reconsidering the role of Servius in contrast to the more commonly discussed Lycurgus challenges the monolithically illiberal interpretation of this key figure. Lycurgus is the ultimate example of denaturing. The perfection of his system of social engineering encapsulates the priority of patriotism over liberty characteristic of the Spartan model. It is not surprising that Shklar, reading Lycurgus as Rousseau's primary legislator, concluded that the legislator is the “the least genuine, the most wooden, one-dimensional figure.”Footnote 57 On the other hand, Servius's work in the Roman model can be seen as an alternative way through which the legislator can “rally without violence and persuade without convincing” (SC, 71; 3:383). Instead of eliminating pluralism and dissent, Servius worked to multiply the number of potential partial societies and prevent political dominance by a single interest group. Unlike Lycurgus, Servius is not described as godlike and beyond reproach. His solutions require the willing cooperation of a number of censors acting as electoral officers that continue the work of reassessing electoral districts. Furthermore, not all of his solutions constitute unmitigated successes, showing us a more human and plausible model of a legislator designing institutions in light of local constraints and with what social scientists might call bounded rationality.
Taking the two models of the Social Contract seriously also reveals a larger degree of coherence between this best-known work and Rousseau's later political writings, particularly his projects for Corsica and Poland. Each of these constitutional projects is, of course, adapted to the relevant circumstances of time and place. In neither of these two cases is Rousseau performing a simple transplant of either the Spartan or the Roman models into a modern context. However, Rousseau's constitution for the small island of Corsica prioritizes the elimination of pluralism, while his constitutional treatise for Poland introduces numerous checks and balances analogous to the Roman model and continuous with the prescriptions in book 4 of the Social Contract. Rousseau's proposals for Corsica and Poland speak positively to the plausibility of these two different ways of approaching pluralism and to Rousseau's preference for a liberal model in the case of a large and pluralistic republic such as Poland.
Rousseau's Constitutional Project for Corsica has often been read as a Spartan project of social engineering.Footnote 58 As Schaeffer and Putterman both observe, Rousseau's model for Corsica emphasizes unity, patriotism, and a systematic rejection of commerce, money, and industrial pursuits in favor of agriculture and rustic simplicity.Footnote 59 When it comes to the two possible approaches to pluralism, the focus in Corsica is on preventing divisions among the population. In his assessment, “the divisions among the Corsicans have always been an artifice of their masters for making them weak and dependent” (CC, 125; 3:903). The primary concern should be to forestall the rebirth of factions in the absence of an external enemy (CC, 125; 3:903).Footnote 60 Owing to the “fortunate condition that makes a good foundation possible,” the circumstances of Corsica permit a focus on shaping people to prevent divisions instead of managing existing sources of pluralism (CC, 123; 3:902). This assessment of Corsica in the later writings fits with Rousseau's assessment of Corsica in the Social Contract. As the extensive discussion of propitious circumstances indicates, Corsica represents a unique confluence of Spartan proportions: a small, egalitarian society with an already strong communitarian identity and capable of supporting an elaborate social engineering project: “It is true that it is difficult to find all of these conditions together…. There is one country left in Europe capable of receiving legislation; it is the island of Corsica” (SC 78; 3:391).
In comparison to the small and egalitarian republic of Corsica, Poland represents a large and pluralistic community. Poland is mentioned in the Social Contract as an example of a divided and less fortunately situated community where “there is no unity in the government” and “the state lacks cohesion” (SC, 99; 3:413). The turmoil of the Polish situation certainly requires civic virtue and dedication to the community, each citizen saying every day of his life “in the bottom of his heart what a virtuous Palatine said in the Diet of Poland: ‘I prefer a perilous freedom to quiet servitude’” (SC, 92; 3:405). While civic virtue is important for Rousseau's constitutional project for Poland, its internal divisions on the basis of rank, wealth, and residence require careful institutional design to manage pluralism. As Williams notes, Rousseau proposes a multitude of checks and balances in Poland in order to prevent a single faction from tyrannically imposing its will on a community. The institutional solutions include, but are not limited to, division of executive power, legislative checks on the executive, frequent legislative assemblies, reforms of the decision procedures in the diets, and the introduction of electoral mechanisms in the selection of the king.Footnote 61 Rousseau makes explicit mention of Roman institutions throughout Poland. In assessing the conditions, he claims, “I seem to see Rome beleaguered, tranquilly ruling the territories where its enemies had just pitched camp” (GP, 178; 3:954). Rousseau even recommends the Roman practice of clientelage to the Polish nobles as a way to partially address the inevitable political challenges of a republic stratified by class and birth (GP, 188; 3:965).Footnote 62 The clearest statement of Rousseau's Roman model, however, appears in his discussion of voting within the Polish diet (GP, 206; 3:984).Footnote 63 Scholars like Williams who use Rousseau's discussion of Poland to reconstruct a more liberal version of Rousseau's political thought can now trace this Polish model back all the way to the Roman model in the Social Contract.
The final advantage of taking the Roman model seriously is the potential for a comparative political theory of eighteenth-century republicanism across the Atlantic. The emphasis on Sparta has led scholars to see Madison's vision for the American republic as profoundly at odds with Rousseau's civic republicanism suited to small, virtuous communities. Rousseau's Roman model, however, brings him much closer to the Madisonian vision, opening up a potential debate about institutional design between the two thinkers. Madison shared Rousseau's concern about the risk that factions would substitute their will for the will of the community. His solution for the United States included selecting representatives with more enlightened views, multiplying the number of potential parties and interests, and establishing obstacles for different factions to act in concert. Although there are important differences concerning the role of representative government in the solutions proposed by Madison, there are important continuities with Rousseau's discussion of the Servian solutions: dividing the electorate into multiple voting groups, equalizing the power of the different voting groups, and avoiding the development of cross-group factions. Rousseau's explicit construal of the problem of factions offers new reasons to appreciate the multiplication of electoral divisions proposed by Madison as a positive solution to the problem of voting the general will in addition to his more commonly noted emphasis on protections against tyranny. Both Rousseau and Madison were concerned about the possibility of majority factions of the poorest citizens, either corrupted through bribery or led astray by populist leaders. At the same time, Rousseau's specific concerns about the prominence of rank and wealth-based factions in legislative assemblies can suggest new mechanisms for controlling the influence of the rich and powerful that occupied the authors of the Federalist Papers to a lesser degree. Furthermore, acknowledging the importance of officers charged with designing electoral districts and the significance of gerrymandering for the health of the republic provides a timely push to reconsider the institutional role of modern-day censors (i.e., electoral officers). Although there is much to be said on this topic, I leave a fuller comparative discussion of the two approaches to the problems of pluralism within a large democratic society for further scholarly investigation.
