The Surprising Survival of Constitutionalism in the Caudillo Republic of Rio Grande do Sul, Brazil, 1836–1845

Abstract

The caudillo strongman remains emblematic of Latin American authoritarianism, but scholarship has seldom reflected on the semantic shifts that this concept suffered over time and its implications for the history of political thought. Numerous political experiments have been marginalized from historical and state-building narratives as the irrelevant work of caudillos, such as the short-lived Rio Grandense Republic in southern Brazil (1836–45). By explaining the Rio Grandense caudillos’ engagement with constitutionalism, this article argues that ‘caudillo’ can be a useful category of analysis if historically contextualized. The article thus reconsiders the history of political thought and state-building in Latin America and beyond in the age of revolutions, suggesting the serious need to scrutinize ‘failed’ states and revolutions. This argument is pursued in three steps. First, the article describes shifting understandings and usages of ‘caudillo’ in nineteenth-century Brazil and neighbouring River Plate states. Second, it analyses the Rio Grandense Republic’s 1842–3 constituent assembly and the novel electoral procedures it employed. Third, it examines the never-promulgated constitutional draft produced by its assemblymen. This constitutional draft is then compared to contemporary River Plate and Brazilian constitutions and its rejection is explained through the assemblymen’s divergent understandings of constitutionalism and democracy.

Of the influence exercised by Bento Gonçalves over the people, we can get an idea from what [they] said of him: ‘If he did not agree to something, who could force him?’ These words translated the conviction of those people. For the inhabitants of the hinterland, the colonel was the king of the countryside; no one had the right to give him orders.

José de Alencar, O gaúcho (1870)

This was the classic image of the caudillo: a powerful proprietor adorned with military commissions and unfettered by political loyalties. The caudillo utilized charisma and physical feats to cause ‘men to adore him; women to admire him’. And he was ultimately moved by territorial conquest, whether for self-gratification or for rewarding clients and cronies.¹ Bento Gonçalves da Silva, an estancieiro (estate-owner) from Rio Grande do Sul, the Brazilian Empire’s southernmost province, perfectly fitted this bill. He held lands and offices in both Brazil and Uruguay, and he alternately fought for monarchists and republicans.² His pièce de résistance was the backdrop to Alencar’s historical novel O gaúcho: a fiscal-turned-secessionist revolt in Rio Grande do Sul that remains Brazil’s longest armed conflict, the so-called Farroupilha Revolution (1835–45).³

When the farroupilhas declared a tentative Rio Grandense Republic in 1836, they elected Gonçalves their president. Combining civil and military authority, he ruled by decree and consolidated the separatists’ hold over the provincial hinterland. But when his own ministers undercut his rule by convening a constituent assembly in 1842, he put up no resistance – why? And why, despite their similarities, did Alencar call foreign revolutionaries caudillos, but not Gonçalves and other farroupilhas? These questions’ intertwined answers reveal the flexibility of constitutionalism in post-colonial Latin America and the nationalist polemics that have often informed commentary on caudillos.

Caudillos remain emblematic of Latin American authoritarianism. John Lynch theorized the ‘caudillo state’ as an early stage of Latin American state-building, wherein caudillos’ ‘personal sovereignty subverted constitutions’.⁴ Thomas C. Wright recently reiterated this view: caudillos ‘controlled or ignored their country’s legislature and courts…[lacking] ideologies or political principles’.⁵ A caudillo-led constituent assembly may therefore sound oxymoronic, but the farroupilha constituent assembly deeply disturbs this understanding and points to an older, flexible reading of caudillo. The farroupilhas were promptly recognized by contemporaries as caudillos, but they also sought to protect their interests through constitution-writing and a language of rights. They even came to blows over how ‘democratic’ their republic should be, and it arguably was their failure to agree on a constitution that doomed the Rio Grandense Republic. Their assembly’s internal disagreements and novel electoral proceedings make the historiographical dearth on this episode notable, although the very idea of the caudillo may be to blame.⁶

The past thirty years have witnessed increasingly nuanced studies of caudillos, but these contributions have been slow to change perceptions of Latin American politics.⁷ Eduardo Posada-Carbó noted that the conceptual shadow of caudillismo or caudillaje (caudillo-rule) still discourages historical research on Latin American representative institutions, so perhaps we ought to move away from studying caudillos.⁸ José Carlos Chiaramonte similarly argued that scholars have too eagerly dismissed deviations from liberal constitutionalism as caudillismo.⁹ Reducing caudillos to popular leaders in post-colonial class struggles, Gabriel Di Meglio recently advised historians to avoid ‘such a loaded term’.¹⁰ More productively, Jorge Myers proposed that caudillismo should not be understood ‘as an “active” category of analysis’ but as a historically unstable concept whose deployments must be individually examined.¹¹

Constitutionalism has conversely enjoyed a brighter place in Latin American historiography.¹² François-Xavier Guerra, in particular, inspired transnational and comparative studies by asserting that Latin America’s first constitutions and relatively broad suffrages signalled a Hispanic ‘precocious modernity’.¹³ José Antonio Aguilar Rivera subsequently placed Latin America within a transatlantic ‘constitutional moment’ (1787–1830), and Joshua Simon similarly compared North and South American constitution-writers. They and other historians saw in early constitutional experiments across the Atlantic world attempts to protect individual liberties while establishing socially exclusionary representative governments.¹⁴ Collaborative projects, moreover, recently spotlighted how local expressions of constitutionalism underscored popular mobilization, territorial reconfigurations, and reconceptualizations of sovereignty across the continent.¹⁵ Global histories of the age of revolutions, meanwhile, have evidenced similar insights. Linda Colley stressed that empires and revolutionaries everywhere effected regime-change through constitution-writing, and Maurizio Isabella analysed southern European revolutions and constitutional compromises bearing similarities and connections to Latin America.¹⁶ Analysing Latin America’s constitutional experiments, especially those with transnational links like the farroupilha, can therefore explain the global allure of constitutionalism and its relationship to state-building and independence movements in the age of revolutions.

This article seeks to contribute to the history of Latin American political thought by transnationally reconsidering the dichotomy caudillo–constitution. It recovers nineteenth-century meanings of caudillo in Brazil and neighbouring Argentina and Uruguay (whose politics influenced the farroupilhas) before analysing the farroupilha constituent assembly and its constitutional draft. The article concludes that ‘caudillo’ was then undergoing a major semantic shift, which turned it into a concept used by historical actors to interpret, shape, or challenge political institutions and authorities. The farroupilha experiment, meanwhile, proved that constitution-writing was not the preserve of a single party or ideology, attesting the importance of scrutinizing ‘failed’ states and revolutions to recover the multiple state-building pathways explored by Latin Americans after independence.

I

Early nineteenth-century writers knew that caudillo was a medieval Castilian term for ‘leader’.¹⁷ Moses was therefore described in 1806 as ‘caudillo and legislator of the people of God’.¹⁸ But the Spanish imperial crisis prompted by the Napoleonic Wars encouraged an identification of caudillos with informal or demagogic military authority: none was worse than Napoleon Bonaparte, ‘caudillo of the French’.¹⁹ The image of the caudillo as a demagogue eventually prevailed, but this was a gradual shift.

In Spanish America, metropolitan weakness caused the rise of republican and monarchist ‘caudillos’ who created makeshift armies through clientelism and charisma.²⁰ In 1810, for example, Montevideo’s monarchist governor Gaspar de Vigodet rallied his troops by declaring ‘I will be your caudillo, a friend, a companion, and not a superior.’²¹ Vigodet was no democrat, but his self-fashioning suggested the caudillo as a popular leader – a recurring image that Argentinian President Bartolomé Mitre (1862–8) associated with the emergence of democracy in America.²²

Despite Vigodet’s claims, contemporaries recognized José Artigas as the Uruguayan caudillo par excellence. Artigas was a radical republican who championed independence and agrarian reform to foster a smallholding citizenry.²³ Republicans and monarchists alike criticized Artigas’s democratic tendencies and charismatic sway over the ‘vandals’ who ‘blindly followed his ideas’, meaning his army of gauchos (peripatetic horsemen), runaway slaves, and Amerindians.²⁴ Even some young landowners joined Artigas, if only to earn their spurs alongside a feared warlord.²⁵ Among them was Bento Gonçalves, then a junior magistrate, who volunteered under Artigas in 1811–12.²⁶

At the municipal level, historians have also identified ‘aspiring caudillos’.²⁷ One was Pedro José Vieira, a Rio Grandense foreman in Uruguay and later a farroupilha colonel. He raised 400 militiamen in 1811 and offered his services to Buenos Aires’s revolutionary junta in exchange for the title of caudillo of the village of Mercedes.²⁸ It is unclear how his request was answered, but it showed that self-styled caudillos attempted to legitimize their authority by collaborating with emergent representative institutions.²⁹ For Vieira, there was no contradiction between being a caudillo and supporting a junta that championed popular sovereignty and constitutionalism.³⁰ His rise also illustrated how ‘new men’ could enter politics amidst revolutionary wars.³¹ Artigas embraced similar ideas, although he had enough influence to dominate regional politics and to advance his own constitutional experiments.³² Within a context of extreme political uncertainty, caudillos thus sought the protection or the leadership of new regimes.

Argentina observed similar developments. After independence in 1816, powerful, landowning provincial governors (often dubbed caudillos) jealously guarded their autonomy and opposed centralizing ‘national’ constitutions.³³ The 1820 Treaty of Pilar, for instance, was an inter-provincial agreement that allowed each province to promulgate their own constitution in lieu of a stricter union. A decade of instability followed, but the caudillos who battled over this period did so at the head of ‘province-states’ – increasingly complex polities equipped with constitutions, banks, semi-professional armies, and legislatures directly elected by broad franchises.³⁴ Buenos Aires notoriously allowed all ‘free men’ to vote from 1821, leading to unprecedented levels of political mobilization and campaigning.³⁵ Caudillos certainly rigged elections and strong-armed assemblymen, but provincial assemblies successfully imposed themselves as consultative and legitimizing bodies.³⁶After civil wars between centralizing unitarios and decentralizing federales, Buenos Aires’s federal dictator Juan Manuel de Rosas inaugurated a new arrangement. Under his ‘Argentinian Confederation’ (1831–52), a mixture of bilateral treaties and coercion held the nation’s provinces together, while Rosas manipulated elections and legislators to legitimize his ‘extraordinary faculties’.³⁷ Rosas and other caudillos arguably contributed to a tradition of authoritarian personalism in countries like Argentina.³⁸ But their success depended on meeting the expectations of heterogeneous social groups, not on their embodiment of peculiarly ‘Hispanic’ or ‘pre-modern’ politics.³⁹ Even Rosas’s brutal, constitution-less dictatorship relied on negotiations with Buenos Aires’s merchant community, and some critics blamed his rise as a demagogue on the city’s ‘democratic’ tradition.⁴⁰

The federales’ victory inspired the earliest debates on caudillismo, beginning with Domingo Faustino Sarmiento’s Facundo: civilización y barbarie (1845). For Sarmiento, violent caudillos like Rosas emerged in Argentina because ‘civilized’ urban society had failed to check ‘barbarian’ rural impulses after independence.⁴¹ Sarmiento thought little of provincial constitutions and contrasted ‘barbarous, arbitrary, American’ caudillos to ‘civilized, constitutional, European’ unitarios.⁴² Facundo proved influential beyond Argentina, beginning with an 1847 Uruguayan polemic amidst a civil war between Montevideo’s liberal colorados and Rosas-backed rural blancos.⁴³ Following Sarmiento, the colorado Manuel Herrera y Obes argued that caudillos were a colonial hangover who exploited uneducated gauchos and opposed ‘the empire of intelligence and law’.⁴⁴ To remedy this situation, Americans should ‘[import] European civilization [through] its men, its books, its industry’, as the colorados had done by recruiting adventurers like Giuseppe Garibaldi to their cause.⁴⁵ In response, the blanco Bernardo Prudencio Berro dismissed Sarmiento’s civilization–barbarism dichotomy. The real struggle was between European oligarchy, represented by Montevideo’s merchant elite, and American democracy, located in the virtuous rural masses whose ‘unanimous’ will was upheld by Rosas and his allies.⁴⁶ Berro agreed that ‘caudillaje [respected] no law other than [the caudillo’s] arbitrary caprices’, but he denied that Rosas was a caudillo.⁴⁷ These three authors wrote within fractious domestic contexts, but they anticipated subsequent scholarship (including anglophone debates) by theorizing caudillaje as a hemispheric ailment.⁴⁸

In Brazil, the farroupilhas were accused of caudillismo because they attempted to ally neighbouring republics and because of River Plate–Rio Grandense societal connections.⁴⁹ Brazilian republicans so rebuked farroupilha separatism, while a Rio Grandense monarchist senator lamented that ‘the likes of Bento Gonçalves…abused the people like the caudillos of Spanish America’.⁵⁰ In 1881, the first historian of the Farroupilha Revolution, Tristão de Alencar Araripe, described the Rio Grandense Republic as a Spanish American ‘republic of caudillos’.⁵¹ Besides condemning a rebellion, these assertions were part of a Brazilian nation-building discourse that contrasted ‘fragmented’ Spanish America to ‘united’ Brazil.⁵² In 1857, the prominent historian Francisco Adolfo de Varnhagen triumphantly wrote that ‘Sensible Brazilians’ were monarchists because they ‘[feared] the anarchy [of] neighbouring republican states’.⁵³ This explained Alencar’s avoidance of the term caudillo to describe Bento Gonçalves. He wished to emphasize the Brazilian-ness of Rio Grandense borderlanders by downplaying their past separatism and foreign connections.⁵⁴

The farroupilhas were well aware of contemporary accusations of caudillismo and comparisons with Spanish America, and their responses trod a fine line around these issues. They agreed that ‘caudillo’ was an insult associated with dictators like Rosas, but they wished to rebuke caricatures of neighbouring republics. In 1839, they criticized Brazilian writers for resorting to a foreign word, caudillo, to describe them (even though they occasionally labelled their enemies so).⁵⁵ Bento Gonçalves also condemned the ‘despot’ Rosas and praised the ‘enlightened’ unitarios, but only after Rosas refused to secretly aid the farroupilhas.⁵⁶ In 1840, Gonçalves dismissed Brazilian critiques of ‘our republican neighbours [as] an often cited and successfully refuted commonplace’. He argued that Brazilian monarchists cherry-picked Spanish American problems (and ignored their own) to tarnish republicans everywhere.⁵⁷

The term caudillo had thus acquired its negative connotation by 1840. This was the result of decades of polemics: from Vigodet’s self-fashioning as a royalist caudillo, to Sarmiento’s pessimistic Facundo. This label, however, did not nullify Artigas’s republican experiment or stop Argentinian provincial state-building – examples of alternative territorial and institutional configurations explored by Latin Americans. In Brazil, accusations of caudillismo were inexorably tied to a nationalist monarchical discourse, although farroupilha refutations highlighted their duplicity towards Rosas. As the following sections explain, the farroupilhas were engaged in their own state-building enterprise, which culminated in their ill-fated constituent assembly.

II

On 20 September 1835, tax-weary estancieiros led by Bento Gonçalves overthrew Rio Grande do Sul’s provincial governor and demanded greater provincial autonomy. The rebels’ federalist demands were not unusual for 1830s Brazil, but the ensuing civil war pushed them to declare their independence on 11 September 1836.⁵⁸ Gonçalves was elected interim president of a new ‘Rio Grandense Republic’ and was extolled ‘to arrange a date for the election of deputies for the Constituent Assembly’.⁵⁹ Right from their separatist turn, the farroupilhas envisioned a constituent assembly to shape their tentative republic. Why, then, did this assembly only convene in December 1842?

War was a key factor, bitterly dividing the province. While ideological fissures should not be overstated (federalism, for instance, had adherents across the province), republicanism more readily appealed to estancieiros with River Plate connections like Gonçalves.⁶⁰ The farroupilhas thus became concentrated in the rural hinterland, whose sparsely populated municipalities were dominated by estancieiros. But the more populous coastal towns, closely linked to Rio de Janeiro and whose powerful beef jerky manufacturers (charqueadores) enjoyed fiscal protections, remained monarchist.⁶¹

Military setbacks, moreover, kept the farroupilha leadership on the run and reliant on draconian expropriations and drafts to pursue the war effort.⁶² In 1837, they attempted to introduce ‘military chiefs of police’ to republican municipalities, armed with impressment powers outside the jurisdiction of elected local authorities. Municipal protests prevented this measure, but violent repression met anti-draft demonstrations.⁶³ This highlighted the absence of a forum to debate ‘national’ questions, as Gonçalves created but never met a Council of Municipal Attorneys in 1838.⁶⁴ His actions were justified by O Povo, the republican government’s newspaper, whose Italian editor Luigi Rossetti defended Gonçalves’s undisputed ‘Power [to direct] the revolution’.⁶⁵ But the patience of many farroupilhas was wearing thin, as evidenced by municipal complaints and Rossetti’s dismissal in 1839.⁶⁶

That same year, two of Gonçalves’s ministers, José Marianno de Mattos and Domingos José de Almeida, took upon themselves to schedule elections for a constituent assembly for March 1840. They did so while Gonçalves was on campaign, during the municipal attorneys’ only ever meeting, in December 1839.⁶⁷ They agreed the assembly’s deputies would be indirectly elected after Brazilian electoral law and that their powers would be ‘constituent and Legislative’, ending Gonçalves’s rule by decree.⁶⁸

Electoral returns, however, showed that the 1840 proceedings broke with tradition by allowing thousands (rather than hundreds) to participate in direct elections. This followed a major discrepancy between electoral instructions issued by O Povo in December 1839 and February 1840. Whereas in December it announced that elections would be indirect, the February instructions stated that deputies would be elected ‘directly by the People in primary Assemblies’, like municipal councilmen and justices of the peace.⁶⁹ Over 3,600 ballots were cast, against the less than 200 of the 1834 provincial elections (which included the more populous coastline).⁷⁰

This novelty has surprisingly attracted little scholarly attention. The only attempt to explain it unsatisfactorily suggested that this was an anticipation of the 1843 constitutional draft, and that direct elections had always been one of the farroupilhas’ goals because they ‘[intended to] bring Rio Grandense citizens closer to central political decisions’.⁷¹ This argument disregarded the fact that the 1843 draft was divisive to the point it was never promulgated, and it ignored the strictly hierarchized society envisioned by leading farroupilhas. Almeida made this belief clear in his plans for public education: ‘universal, but not uniform [because] the Labourer [must be] educated to be a Citizen Labourer, and not to be a Magistrate or a General’.⁷² Mattos arguably was more radical, but also less influential.⁷³ His egalitarianism was described as ‘bizarre’ and his proposal to fully enfranchise formerly enslaved soldiers was quickly rejected.⁷⁴ We can more reasonably compare the 1840 elections to River Plate practices. Some Argentinian provinces had enfranchised rural smallholders through direct elections, and even Rosas found them an expedient way to galvanize his rule.⁷⁵ As explained below, the farroupilhas clearly saw direct elections as tools of political legitimation and reflected on Argentinian provincial constitutions.

Fraud was also an unlikely cause of this electoral oddity, since farroupilhas discussed and even criticized aspects of the election, but not the number of ballots and voters. The February 1840 electoral instructions did not explain how one could stand to be a candidate in the first place – similarly to existing Brazilian electoral laws. Well-connected republicans nevertheless circulated slates of preferred candidates, as in prior elections.⁷⁶ Almeida wrote one such slate in February 1840. It sadly has not survived, but his justification for writing it does: ‘so our co-citizens will have something to go by’, adding that ‘no one [should be able to] say that the Government restricted the freedom to vote’.⁷⁷ Making no mention of an unusually high franchise, Almeida was more concerned with the embattled republican government’s image. The farroupilha diplomat Antônio Manuel Correia da Câmara expressed similar concerns. Many citizens, he noted, feared police chiefs’ oversight over electoral proceedings in the eventual absence of an elected justice of the peace (traditional election overseers), which could facilitate the intimidation of voters. Câmara pleaded with ministers to adopt a ‘more constitutional and less military’ policy, reserving electoral oversight to ‘popular magistrates’.⁷⁸

Across Brazil, to be sure, armed threats and intimidation were common occurrences during elections. But contrary to Câmara’s assessment, elected authorities were just as likely to condone or partake in fraud and coercion as police officers.⁷⁹ Studies of Brazilian municipal politics have pointed to the 1840s as the moment from which electoral violence and corruption truly became ‘endemic’ thanks to better organized political parties, and Rio Grande do Sul’s ongoing civil war made its situation all the more delicate.⁸⁰ The farroupilhas, moreover, were struggling to build and legitimize a state, having so far failed to gain the recognition of neighbouring republics.⁸¹ Câmara and Almeida might therefore have been more concerned with the smooth running of proceedings than with the actual integrity of elections – what Richard Graham dubbed ‘the appearance of fairness’.⁸² Similar considerations may have also swayed Gonçalves against blocking his ministers’ initiative to convene elections, as he and associates like Rossetti were already under heavy criticism.

The elections saw leading farroupilhas chosen for the constituent assembly. O Americano (O Povo’s successor) published a list of deputies-elect and their occupations, attesting the election of ministers like Mattos and Almeida alongside a plurality of military officers.⁸³ These results were announced in October 1840, yet the assembly was only convened on 1 December 1842. This delay was probably due to the farroupilha capital moving from Caçapava to Alegrete in May 1840, as they were driven further inland by monarchist forces.⁸⁴ Opening the assembly, President Gonçalves reiterated Rio Grandenses’s right to self-determination and downplayed chances of reconciliation with the Brazilian ‘monarchy’ though not with the Brazilian ‘nation’. ‘Royalty’, he imagined, would one day be ‘banished’ from Brazil, and so Rio Grande do Sul would again ‘unite [itself] by strong ties of federation to the magnanimous Brazilian Nation’. Until then, ‘the first necessity of the State is a political Constitution [to guarantee] internal political stability’.⁸⁵ While the farroupilhas had not all been avowed republicans since 1835, the possibility of a federal reconciliation with Brazil had always been contemplated as a solution to the conflict. Gonçalves and other estancieiros imagined this would protect their interests from metropolitan fiscal impositions.⁸⁶

The assembled deputies, however, seemed more preoccupied with routine government tasks than with constitutional debate. General João Antônio da Silveira, for instance, complained that he was given the ‘frivolous’ task of creating an ‘Army Archive’ to record pensions and promotions. He was also concerned that the assembly was moving to form an unaccountable Council of State (similar to Brazil’s) to oversee Executive matters, which could discredit the assembly in the eyes of ‘the sensible mass of the Rio Grandense People’.⁸⁷ Despite these issues, what most preoccupied Silveira was the emergence of a ‘minority’ party within the assembly, which held ‘fierce opinions’ and was just large enough to paralyse legislative proceedings. While he did not specify what these ‘fierce opinions’ were, he denounced their preferred modus operandi: to abstain from sessions while blocking the admittance of elected substitutes, preventing the chamber from reaching quorum. Silveira advised Gonçalves to dismiss naysayers from the assembly ‘[to deny] the minority the right to complain of the majority’.⁸⁸ While Gonçalves refused to do so, he shared the general’s dissatisfaction that the assembly had failed to promulgate a constitution after two months. He believed that this was done out of spite for himself, and he lamented that personal ambitions ‘[made] us look incapable of independence’.⁸⁹ But Gonçalves’s and Silveira’s disagreements were equally revealing of farroupilha internal divisions. The Council of State berated by Silveira as an imitation of the Brazilian monarchy, for instance, was in fact suggested by Gonçalves’s supporters, as explained below. Gonçalves, conversely, had long avoided as much as summoning elections for the assembly, angering many republicans.

Gonçalves’s exchange with Silveira reflected, too, that global preoccupation with constitutions identified by Colley and other historians: between the late eighteenth and early nineteenth centuries, state-building and regime-change required constitution-writing. Rosas’s constitution-less Confederation was an anomaly. Far more common was Brazil, where a constitution had been a popular argument for independence in 1822: the authoritarian ‘Charter’ promulgated by Dom Pedro I in 1824 indeed endured until 1889, even though it followed his violent dissolution of a liberal-leaning constituent assembly.⁹⁰ His constitution protected royal prerogatives (exercised through the Council of State), but by enshrining individual freedoms and parliamentary representation it tied his regime to what Andréa Slemian called a ‘new constitutional legitimacy’. This meant the contemporary transformation of ‘constitution’ from a description of ‘dominium’ into an ‘objective-concept’: an expression of expectations and rules for government and society.⁹¹ Isidoro Vanegas added that Latin American revolutionaries especially clung to written constitutions as challenges to the colonial order. That is, a written constitution proved a society’s capacity to exercise ‘free will’ and explained how the absence of colonial law would be addressed.⁹²

The Rio Grandense Republic’s leaders were similarly concerned with differentiating themselves from the Brazilian Empire, but they lacked the legitimacy to be recognized abroad and had failed to protect individual rights through their wartime policies. The direct elections of 1840 perhaps were a first step towards assuaging fears of despotism. But as Silveira’s complaints demonstrated, a constitution was required to convince their citizenry and neighbouring republics that their state would last.

III

Bento Gonçalves’s son, Joaquim Gonçalves da Silva, served as a substitute in the farroupilha constituent assembly. He reminisced that the assembly ‘was dissolved just as it discussed the Constitution project’ because of an approaching monarchist army, and he claimed that ‘the behaviour of the oppositionist minority’ was far less disruptive than Silveira had argued.⁹³ He also defended his father from accusations of despotism, an allegation that gained traction after December 1842, when the constituent assembly discussed suspending individual guarantees to help the government requisition military supplies and arrest ‘dissidents’.⁹⁴ He claimed ignorance of who defended the suspension, although O Americano identified the lawyer Serafim dos Anjos França as its author, supported by Almeida and Mattos (then reconciled with Gonçalves), among others.⁹⁵ This was the cornerstone of the Council of State described by Silveira, which was meant to step in once Gonçalves’s rule by decree officially ended.⁹⁶ The suspension was approved by the assembly on 8 February 1843, although how many deputies voted for or against it was not reported.⁹⁷

Another episode that solidified Gonçalves’s image as a tyrant was the murder of Vice-President Antônio Paulo da Fontoura, whose recurring absences from the constituent assembly suggested that he was a member of the abstaining opposition.⁹⁸ Due to a lack of sources, Fontoura’s murder has not been further examined by scholars, but rumours implicated him in a plot to overthrow Gonçalves.⁹⁹ Fatefully, Gonçalves himself believed and was vocal about those rumours, which, for many, implicated him in the vice-president’s death.¹⁰⁰

Against this divided backdrop, the assembly nevertheless named a commission to draft a constitution – although their draft proved a non-starter. Among its authors were the main advocates for the suspension of individual guarantees: in signature order, José Pinheiro de Ulhôa Cintra, Francisco de Sá Brito, José Marianno de Mattos, Serafim dos Anjos França, and Domingos José de Almeida. Sá Brito and França were the only ones who had never been government ministers under Gonçalves. Their draft mixed Brazilian and River Plate constitutions, both in terms of its prescribed institutions and language. It particularly paraphrased the constitution of Corrientes, an Argentinian province with which the farroupilhas had attempted an alliance.¹⁰¹ Its preamble, for instance, closely followed Corrientes’s 1838 constitution by promising ‘[To protect] with all efficacy life, honour, liberty, individual security, property and equality, [the] essential bases of the rights of man…[and to] secure justice, promote public happiness and safeguard the enjoyment of all these goods for us and our posterity.’¹⁰²

The draft defined Rio Grande do Sul as a ‘free and independent nation [with a] republican, constitutional, and representative [government]’.¹⁰³ And it further departed from the Brazilian monarchical constitution by embracing ‘popular sovereignty’ instead of ‘national sovereignty’.¹⁰⁴ Rather than a rejection of the idea of ‘nation’, this word choice could again refer to the Correntino document, which mapped sovereignty onto ‘the universality of Correntino Citizens’.¹⁰⁵

O Americano clarified the farroupilha understanding of popular sovereignty. It lambasted followers of Jean-Jacques Rousseau’s interpretation of popular sovereignty because, although they had used his ideas to ‘victoriously oppose the pretend legitimacy of kings, proving that sovereignty resided in the peoples, [and] that all legislative power emanated from the general will’, they had committed a ‘grave error’ when they ‘put [the exercise of] sovereignty in the hands of the peoples’. O Americano preferred Benjamin Constant’s and Montesquieu’s emphasis on the rule of law and individual liberties, embodied in a representative government wherein ‘the power of a small number was sanctioned by the consent of all, [this] is the general will’.¹⁰⁶ This argumentation resembled Rousseau’s reception across the River Plate. In post-1810 Buenos Aires, as explained by Gabriel Entin, there were attempts to employ Rousseau to simultaneously ‘legitimize revolution’ (challenging royal authority) and to establish ‘social order’ (through a new social contract). These ‘contradictory’ objectives were meant to be reconciled through a constitution that acknowledged ‘the sovereign people’ as its ‘constituent power’.¹⁰⁷ Artigas had likewise stated that Uruguayans could only protect their ‘primitive freedoms’ by exercising their ‘inalienable sovereignty’ to adopt a new ‘social constitution’.¹⁰⁸ In Brazil, debates on national sovereignty had evolved differently, highlighting the crown’s unifying power over a large and diverse empire.¹⁰⁹

Defining the three branches of government, the farroupilha draft gave most power to the Legislative, composed of a bicameral General Assembly (with an upper Senate and a lower Chamber of Deputies). The General Assembly’s authority encompassed military affairs, commerce, and foreign relations, including ‘treaties [of] federation’.¹¹⁰ Its bicameral structure contrasted with Corrientes’s unicameral system, but while it superficially resembled Brazil’s, it was most likely influenced by Uruguay’s Legislative.¹¹¹ Deputies were to be directly elected every four years but, as in Uruguay, wealthy ‘electors’ voted for senators every twelve years. Senators were then divided into three ‘classes’: one serving for twelve years, another for eight years, and the third for four years. Until the twelve-year limit was reached, vacant seats were filled by the president, picking candidates from a list drafted by the deputies.¹¹² As in both Brazil and neighbouring republics, the draft stipulated high property requirements for deputies and electors, and higher still for senators and presidents, benefitting estancieiros and the few charqueadores who turned republican (like Almeida).¹¹³ The draft also barred the illiterate from voting, a restriction then only present in Uruguay, where the 1830 constitution stipulated literacy as a requirement for those ‘[entering] the exercise of citizenship from the year 1840’.¹¹⁴ This system aimed at making the Senate not only an elite-controlled institution, but also a ‘permanent’ one, which could fulfil Legislative functions year-round (unlike the deputies).¹¹⁵ The Rio Grandense Senate, like its Uruguayan counterpart, would thus constantly check the Executive, summoning the deputies if the president failed to regularly do so, and would assume the attributions of both deputies and the Supreme Court when required.¹¹⁶ This arrangement resembled Corrientes’s ‘permanent commission’ of legislators, which oversaw the everyday governance of the province during parliamentary recesses.¹¹⁷

The Executive was thoroughly dependent on the Legislative. The draft indeed suggested a Council of State, but this was a consultative body whose members were indirectly elected by voters.¹¹⁸ The president, following the Uruguayan precedent, was elected by the General Assembly every four years, and while he could nominate ministers, they were held individually responsible before the assembly.¹¹⁹ He did not hold veto powers and was only able ‘to make observations on legislative drafts’.¹²⁰ The president also presented nominees for the Supreme Court and other tribunals to be approved by the Senate (a power his Uruguayan counterpart lacked).¹²¹ Most of the Judiciary thus depended on the other two branches, but popularly elected justices of the peace were maintained in accordance with Brazil’s outgoing liberal model (Brazil moved in 1842 to disempower elected justices and to centralize law enforcement).¹²² On municipal administration, however, the draft dithered on its liberal commitments. It established municipal ‘directors’ and ‘intendants’ to preside over municipal assemblies as ‘immediate [agents] of the Executive Power’, somewhat resembling the controversial military chiefs of police of 1837.¹²³ Uruguay again possessed analogous offices: jefes politicos (‘political bosses’), centrally appointed police chiefs who enforced electoral regulations.¹²⁴

Joaquim Gonçalves’s claim that the constituent assembly never officially debated the constitutional project appears to have been true. On 11 February 1843, O Americano announced its publication would be put on hold so the government press might be used for printing the ‘Constitution Project’, but once it resumed publication on 1 March (its last ever issue) there was no reference to the constitution.¹²⁵ There is no evidence that the draft circulated publicly or privately, although at least one copy was printed in Alegrete in 1843 and survives at the Brazilian National Library.¹²⁶ The draft’s circulation notwithstanding, there was visible confusion among farroupilhas on what happened during the assembly’s final days. On 7 March, a friend of Almeida enquired of him what had become of the assembly and if the constitution had been ‘completed’.¹²⁷ At the same time, Gonçalves wrote to Silveira complaining that he could not find enough deputies to keep the assembly open.¹²⁸ Moacyr Flores conjectured that many deputies simply left Alegrete in opposition to Gonçalves’s continuing rule. The president had become associated with dictatorial measures such as the suspension of individual guarantees, and the constitution drafted by his allies offered no assurance that Gonçalves would not find his way back to the presidency after its promulgation.¹²⁹

Few historians have commented on the constituent assembly or the 1843 constitutional draft. Writing from a monarchist perspective, Araripe considered the farroupilha constitutional experiment a performance, since mounting military defeats meant its drafters ‘held the conviction that [the constitution] would never become a reality’.¹³⁰ As a statement of intent, Araripe saw in the farroupilha draft an attempt to ‘[combine] the dispositions of [the Brazilian] constitutional code with the republican constitutions of America’, the result being ‘a kind of Roman Senate’ fated to ‘degrade’ into despotism or oligarchy.¹³¹ More recently, Setti Reckziegel and Cittolin Abal attempted a comparative study of the Rio Grandense and Uruguayan constitutions. They argued that both texts upheld ‘caudillo liberalism’, by which they meant an ‘elitist’ understanding of individual rights and representative government characteristic of the ‘Uruguayan-Rio Grandense frontier’.¹³² While they contrasted this liberal tradition to Brazil’s ‘caboclo [Euro-Amerindian] liberalism’ (distinct for the emperor’s powers), it was unclear what made caudillo liberalism uniquely ‘caudillo’, especially as they conflated other constitutional experiences into a broad ‘European and North American’ tradition.¹³³ Tanger Jardim conversely noted the similarities between the Rio Grandense and Brazilian constitutions: the 1843 text was a ‘possible’ constitution because it was based on legal practices already observed by Rio Grandenses.¹³⁴ But he inexplicably believed that the farroupilha draft was an abolitionist text.¹³⁵ As Reckziegel and Abal correctly noted, the document was completely silent on slavery, diverging from its Uruguayan counterpart that stated ‘in the territory of the State, no one shall be born a slave’.¹³⁶ Most farroupilhas indeed supported slavery; Almeida even protested Uruguay’s 1842 abolition of slavery because his captives held there were freed.¹³⁷ The draft was similarly silent on indigenous rights, although this was unsurprising: in neighbouring Corrientes, liberal constitutionalism had justified communal expropriations, but most Rio Grandense indigenous communities had been violently destroyed in 1754–9.¹³⁸

A more perceptive interpretation of the draft emphasizing its Brazilian links was Gabriel Paquette’s. He described the farroupilhas’ suggested system as a ‘meretricious democracy’, a representative government based on popular sovereignty but upholding existing hierarchies through a restricted suffrage.¹³⁹ A similar conclusion may be reached following Bernard Manin’s argument that contemporary ‘representative governments’ operated through a ‘principle of distinction’: the belief that elected officials should possess superior ‘wealth, talent, and virtue’ to voters.¹⁴⁰ Consequently, by focusing on property requirements for office-holding rather than for voting, the farroupilhas’ 1840 direct elections emerge as a less pressing enigma. Their priority was to create a representative government that, headed by a senatorial elite, followed models observed from American republics to European parliamentary monarchies. In one of its few mentions of North American constitutionalism, O Americano indeed paid little attention to elections but thoroughly described how Massachusetts’s constitution reserved public office for wealthy proprietors.¹⁴¹ Quoting Benjamin Constant, O Americano also praised direct elections for their capacity to keep ‘voters of the inferior classes…docile and industrious…Satisfied with having exercised their rights.'¹⁴² The farroupilhas believed that direct elections, alongside constitutional guarantees, legitimized political authority, but government should be kept to the ‘superior’ classes.

It should nevertheless be noted that the farroupilha defence of ‘representative democracy’ was articulated not by the constitutional drafters but by their opponents in the constituent assembly. On 18 February 1843, as the constituent assembly stopped holding meetings, six deputies headed by Antônio Vicente da Fontoura (brother to the murdered vice-president) and Onofre Pires penned a ‘Manifesto of the Minority Deputies’.¹⁴³ They argued that President Gonçalves and his supporters had lost touch with the ‘general will’ by suspending individual guarantees and conspiring to create a Council of State, which could be used to reward Gonçalves’s underlings and to indefinitely prolong his ‘absolute power’.¹⁴⁴ Only a constitution that upheld ‘the system of representative Democracy’ could prevent Gonçalves from perpetuating his presidency, but the opposition deputies feared that their only option was to flee Alegrete.¹⁴⁵ Should they stay, they believed that they would either be killed or compelled to join the assembly so that it would reach quorum. If the latter happened, ‘without liberty to discuss, without the liberty of the press’, they would be powerless to prevent Gonçalves from promulgating a constitution ‘of his own moulding’.¹⁴⁶ It was unclear if these deputies had read the 1843 draft. They also did not detail what they meant by ‘representative Democracy’, only stating that it was characterized by a limited Executive, individual liberties, and was followed by the United States.¹⁴⁷ Their opposition nevertheless was successful because a lack of quorum meant that the 1843 draft was never promulgated and the assembly indefinitely paused its works. Military pressures also meant that by early March the assembly was effectively disbanded, a blow which utterly disorganized and dispersed the republic’s government.¹⁴⁸

Following the constituent assembly fiasco and further military defeats, Bento Gonçalves resigned the presidency in August 1843.¹⁴⁹ He was succeeded by José Gomes de Vasconcelos Jardim, who was elected behind closed doors by the remaining farroupilha commanders and who soon sued for peace with Brazil.¹⁵⁰ The resulting Treaty of Ponche Verde dissolved the Rio Grandense Republic in February 1845, although rebel estancieiros benefitted from concessions like military commissions.¹⁵¹ The treaty also outlined the incorporation of farroupilha soldiers into the Brazilian army, including enslaved soldiers who were promised manumissions in exchange of military service. But predicting this outcome and unwilling to free enslaved soldiers, republican and monarchist commanders engineered the Porongos Massacre of November 1844, when 400 black farroupilhas (mainly enslaved or freedmen) were ambushed just as peace talks were about to start.¹⁵²

IV

The hopes the farroupilhas placed on their much-anticipated assembly and constitution raised fundamental questions about the history of constitutionalism and caudillismo in post-colonial Latin America. While first animated by landowners’ fiscal interests, the Rio Grandense rebels quickly adopted a republican discourse and representative institutions. As they did so, they were accused of caudillismo for following their River Plate neighbours, which they sought to rebuke while defending republicanism. The farroupilhas’ own understanding of ‘caudillo’ certainly was pejorative, but this was the result of an ongoing semantic shift and did not necessarily imply a disinterest in constitutionalism. Since the 1810s, after all, self-styled caudillos had defended republican and monarchical understandings of popular sovereignty and constitutionalism. It ultimately was the ramification of debates around Rosas’s dictatorship, in addition to Brazilian nationalist narratives, that definitely opposed caudillos and constitutions.

This is not to say that the farroupilhas, as constitutionalists, were democratic social reformers – quite the contrary, as episodes like the Porongos Massacre proved. But this article has shown that many of Latin America’s early and often forgotten state-building experiments were understood by contemporaries both through the prisms of caudillismo and constitutionalism. The caudillo, therefore, still has an important role in studies of Latin American politics, but it should not be taken as a historically static concept. ‘Failed states’ like the Rio Grandense Republic likewise did not illustrate the inexorable fate of a ‘republic of caudillos’, just as ‘provincial’ constitutions like Corrientes’s should not be seen as destined to be replaced by ‘national’ ones. Distinct state-building experiments indeed shared key assumptions regarding constitutionalism, elections, and representative government. The conflicts surrounding these ideas and experiments highlighted the deep uncertainties that characterized post-colonial Latin America and elsewhere in the age of revolutions, and the various means by which contemporaries sought to understand or influence their situation.