Conscription versus penal servitude: army reform's influence on the Brazilian State's management of social control, 1870-1930.

The Brazilian army's size, its share of national budgets, and its preeminent role in the management of government-legitimated violence made it the primary institutional bridge between the state and the "criminal" underworld in the late 1800s. The State did not fashion the army to perform this function, but it proved convenient given circumstances, precedent, and the government's limited institutional capacities. Police and judicial officials, responding to perceived threats to civic order, summarily (without a trial) remanded thousands of Brazilian-born free men to serve minimum stints of six years in the barracks. Army orphanages inducted hundreds of street children. Authorities confined many civilian convicts, considered too dangerous for peacetime military service, in army-administered prisons, penal colonies, and stockades. Unlike Mexico's Rurales, Brazil did not have a national police force, but army soldiers commonly performed police duties across the country.(1) The army's largely unrecognized centrality to Brazil's criminal justice system meant that reforming its recruitment, training, and function altered important aspects of the State's bureaucratic management of public discipline.

The sequence and depth of institutional reforms constitutes a little explored approach for comparative history. Theda Skocpol and others, for instance, have initiated significant comparative analysis of social welfare programs in mostly North Atlantic nations highlighting (in some instances) the understudied importance of military mobilizations in their development.(2) This article takes a similar point of departure in a different direction by using military mobilization as a lens to analyze the development of public institutions of social control. Patterns in Brazil may indicate differences in the timing and emphasis of State building strategies in other developing nations. The point is that institutional reforms need to be situated more firmly in the context of a web of institutions in State building projects. In tightening and strengthening one institutional thread, the State might unintentionally open new gaps in its organizational web and weaken other strands.

This analysis works from the supposition that the rapid proliferation of European military practices proved the most efficient conduit for the global dissemination of more modem modes of social control. For emerging nationalist States, defense of the "respectable" citizenry from internal and external threats became an important justification for increasing State institutional capacity.(3) Wars, internal insurrections, and crime represented credible threats to the central State's claims to a monopoly on the management of violence. Such claims became justifications for more widely studied global arms races and the less analyzed martial organizational races of the 1800s and 1900s. In the latter case, military conscription and the maintenance of a reserve of trained men capable of rapid mobilization became touchstones of martial modernity. The sweeping adoption of largely Prussian-inspired recruitment methods and service models altered the internal structure and composition of armed forces throughout much of the world and shifted their role in civilian societies.(4) The character, pace, and results of military reforms clearly differed from nation to nation. Some States eschewed them proving that they were not preordained. The legacy of this martial reformation, however, has certainly proven more durable, uniform, and ubiquitous than attempts to emulate or transplant other European institutions.

The impetus to reform the Brazilian army's role began in earnest with the mobilization crisis of the Paraguayan War (1864-1870) discussed in more detail below. By 1874, Parliament approved a limited conscription system to eliminate coercive impressment, but the State, thwarted by popular resistance and bureaucratic intransigence, only succeeded in implementing the law in a modified form in 1916 amid the tense international environment of World War I.

To illuminate the difficulties that impeded conscription's adoption, this article examines the army's role as the central institution in Brazil's fledgling penal justice system. A number of important studies have illuminated the evolution of Brazil's criminal justice system in the 1800s and early 1900s. Most bring a regional or single institutional focus to the interplay between individuals, groups and developing police, prison, and judicial institutions. From 1850 to 1900, as slavery declined, migration and immigration increased, and urban populations mushroomed, police forces grew and arrest rates rose - specially for vagrancy and crimes against public order. Scholars generally suggest that the State's role as a disciplining agent over the free wage labor force expanded.(5) This essay suggests some reasons why expanding arrest rates signified little in terms of establishing more modern forms of disciplinary practices. The analysis makes an important distinction between traditional disciplining methods (e.g., beatings, intimidation, simple incarceration, military impressment) and more modern ones (conscription, surveillance, education, social segregation, more extensive institutional socialization, the use of uniforms, prison at hard labor, coordinated exercises, etc). Authorities did try to apply (with limited success) more modern disciplining methods to men serving in police corps, but cops walking a beat could rarely rely on auxiliary institutions to "socialize," "reform," or even detain their local nemeses. In part, this institutional weakness frustrated the police's attempts to practice ideals espoused in professional journals.(6) The analysis here narrows its focus to this weak spot, or to the State's bureaucratic capacity to "socialize" citizens within more modern disciplinary regimes.

I argue that army enlisted service slowly changed from a punitive institution of social reform to a preventative one as Brazil's national draft lottery replaced impressment in 1916. Reformers charged the conscripted military to instill lessons of nationalism, hygiene, and discipline to youths from more "respectable" poor families, whereas impressment had targeted "criminals." By "criminals," I refer to poor men authorities apprehended in the act of committing crimes or whom they merely suspected of specific offences. Officials often remitted such "criminals" to the army without due process. "Unprotected" men (desprotegidos), however, often fell under the same "criminal" rubric even without being suspected of specific offenses. The unprotected by choice or inability did not conform to norms of behavior authorities expected of "honorable" poor men. These "criminals" did not maintain a fixed residence or seek steady employment and "protection" from "respected" patrons. Nor did they marry and responsibly support a family: a traditional legal exemption from military service before 1916. Authorities also pressed sons who did not provide sustenance and protection to dependents such as an elderly widowed mother or orphaned younger siblings. The unprotected lacked the capital to pay for an exemption, nor did they practice a skilled trade legally exonerated from military service. Officials viewed unprotected men as dangerous "loners" and "criminals" in potential whose behavior threatened the sanctity of the family.(7) Most officials considered impressment an exemplary punishment for a visible, if small, portion of the unprotected poor. Only through an awareness of these factors can the disciplining strategy of impressment be understood as distinct from that of conscription.

Civil Prisons and Penitentiaries

To gauge the dimensions of the army's role in police work and criminal justice, it is necessary to briefly examine the civil penal system. In the colonial era, summary induction to the army was a common punishment for "less serious" public offenders, and army fortresses and stockades formed an essential part of Brazil's civil prison system. In part, the lack of intensive wars to fight led Brazil's colonial armed forces and militias to perform other functions, as Caio Prado Junior notes.(8) The army's function as a jailer and for policing constituted a more consistent part of its duties (based on a review of records) as opposed to what it claimed to do: guard Brazil from enemy invasions. After Independence, liberal reformers hoped to change Brazil's criminal justice system.

Brazil's 1830 Penal Code called for the construction of civil prisons suitable to apply Walter Crofton's Irish System for reforming criminals) It introduced "prison with hard labor," a sentence to be carried out within prison walls, entailing the construction of new correctional prisons. The 1824 Constitution required provincial governments to plan, fund, and build their own prisons, but over the course of the 1800s, only four out of twenty provinces (Sao Paulo, Bahia, Rio Grande do Sul and Pernambuco) and the imperial capital district (Corte) constructed new prisons conforming in some respects to Jeremy Bentham's panopticon design and the requirements of the 1830 Penal Code.(10) Under ideal conditions, these prisons added capacity for approximately thirteen hundred correctional convicts.

The construction of modern prisons stalled under the early Republic with some notable exceptions like the 1902 founding of Brazil's first juvenile prison in the prosperous state of Sao Paulo and the expansion of some older facilities. Overcrowding, stingy budgets, and the lack of qualified prison administrators and personnel undermined the ability of these new facilities to apply the essence of modern rehabilitative discipline: work regimes, cellular isolation, rigorous surveillance, and segregation by gender, age, and severity of crime. Intended to serve only correctional convicts, in practice these new penitentiaries also accepted those awaiting trial or being held for disturbing the peace. Like the army, civil prisons performed a variety of functions. In the 1860s, Rio's Correctional Prison also housed a civil trade school for minors, a depository for free Africans immigrating to Brazil, and a branch of the fire department. Prison inspections almost always lamented the miserable conditions and the lack of applied reformative disciplinary practices.(11)

Atrocious conditions in new correctional prisons were topped by more traditional jails. A second tier of large and medium-sized prisons and a third tier of ad hoc jails (mostly private properties rented by municipalities) warehoused convicts along with those awaiting trial or held for misdemeanors.(12) Ceara's provincial police chief complained in 1876 that the roof of Fortaleza's prison was not completely covered, creating great discomfort for the inmates. The jail's 408 prisoners crammed into twenty-eight cells, and carried their excrement daily to a drain inside the prison walls which overflowed from time to time, bringing fecal matter back to the surface. The chief admitted that the prison was indistinguishable from those of the "old regime," meaning Portuguese colonial rule. He added that "a prisoner is not able to escape, but with certainty, will die prematurely and if not, will not be rehabilitated but become more degenerate."(13) Thus, rehabilitative disciplinary regimes were sorely lacking in both modern and traditional Brazilian prisons.

The data presented below in figure one are intended to give a ballpark sense of the limited capacity of civil prisons. While far from comprehensive, these figures, culled mostly from inconsistent data given in governors' reports, is only a preliminary attempt to measure the size of Brazil's prison population (for lack of more accessible sources). The numbers for Sao Paulo in 1886 (977) and Minas Gerais in 1910 (2,180) are particularly instructive because they include the greater part of the jail population throughout these states. The large figures for these populous and wealthy states must be balanced against smaller, less prosperous ones like Alagoas which in 1923 logged a total of 180 prisoners.(14) State and local governments resented the burdensome costs of supporting "idle" prisoners and the even higher costs of running working prisons.(15) Prison populations naturally fluctuated according to circumstances, but even though the data are only suggestive, it seems reasonable to assert that Brazil's civil jails in the late 1800s held no more than ten thousand individuals at any one time under normal circumstances: a tiny proportion of Brazil's 14.3 million inhabitants estimated in the 1890 census. The actual figures for civil prison inmates may have been lower, but even if there were 20,000, it would not substantially alter my argument.

The majority of inmates within prison walls were convicts; most had been sentenced for homicide. In fact, one was almost forced to commit a heinous crime to land an extended sentence in a civil jail. This stands in stark contrast to industrializing nations like England that exported thousands of convicts for crimes against property to Australia before constructing modern penitentiaries in the mid-1800s.(17) From 1855 to 1923, Bahia's penitentiary accepted 5,039 convicts, of those 3,012 were sentenced for homicide compared to only 800 for crimes against property (roubo and furto). Over this period, convicts entering the penitentiary averaged a little more than 74 per year, and most looked forward to long sentences for homicide. As late as 1923, 261 of Bahia's 351 penitentiary inmates were held for homicide. A comparison of homicide convicts with total prison population in other major state prisons in 1923 reveals a general trend: Sergipe 63/73; Alagoas 79/112; Maranhao 131/146; Amazonas 50/51; Para 43/74; Minas Gerais 92/111; Rio Grande do Norte 37/54; and Pernambuco 189/381.(18) Overall, murder convicts average approximately 70 percent of the prison populations [TABULAR DATA FOR TABLE 1 OMITTED] of the largest, most secure, and "modern" state prisons listed above. Unlike the penitentiaries of the antebellum American South, Brazilian prisons did not use a color bar to limit penitentiaries to white convicts, but they often required that slave convicts wear manacles. Authorities anxiously put slave convicts to work (often in navy and army arsenals) in the hopes of discouraging other bondsmen from committing heinous crimes to escape a life of arduous labor for one of incarcerated "leisure."(19)

Civil jails, clogged with hardened convicts, had limited capacity for lesser offenders. Since there was little expansion of permanent prison facilities from 1870 to 1916, rising arrest rates meant that civil prisons were little more than revolving doors that turned with increasing rapidity for those accused of less serious crimes.(20) Army enlisted men and prison inmates came from similar backgrounds. Most were illiterate lavradores (agricultural workers, share croppers, or smallholders) or jornaleiros (day laborers often employed in the casual labor market); "mesticos" predominated in the army and prison populations though sizeable numbers of whites and blacks were also present.(21)

Even though impressment could relieve pressures on crowded jails, local officials could frustrate the efforts of state and imperial officials to conduct policing business as usual. As one police delegate in the rural Termo of Barreiros, Pernambuco, emphasized in 1874: "One cannot effect anyone's imprisonment, even [those] suspected [to be] deserters or [those usually destined] for [military] impressment, because the local district attorney will act to free the prisoner which the local judge will sanction leading to fines and court costs that triple normal [recruiting] costs. Under these conditions, I do not know how one can do police work."(22) For this delegate, the disruption of cost-effective impressment and apprehension of deserters unjustly impeded every day police work.

Enlisted army service worked as a partial escape valve for overburdened civil prisons by incorporating offenders and the "criminally" idle. Authorities tended to try only those accused of serious felonies, mostly homicide. Judges did not dragoon men suspected of homicide or other crimes they deemed worthy of prosecution. Army officials routinely returned those indicted for crimes who turned up in recruitment sweeps. But, many "less" threatening offenses, even suspected horse thievery, could be punished with summary impressment.(23) It is impossible to estimate precisely how many men pressed into military service had committed criminal acts. Since there were no trials, the police delegates' assertions were the only evidence of transgressions. Conditions in the fetid bowels of Brazil's civil prison system may have given some officials pause before trying and sentencing a man for what they considered a lesser offense. With some justification, authorities could reassure themselves that summary impressment was a more satisfactory solution. It diminished court costs and eased pressures to build expensive new prisons. Local officials thus shifted the costs of social control and "rehabilitation" to the central State. In turn, the army's manpower needs obliged it to take on tough cases.

Compared to prison conditions, army service probably offered better chances of "reforming" offenders. In 1872, War Minister Rio Branco noted the general repugnance most Brazilians demonstrated for army service while lauding the army's important social mission: "the army has saved many individuals from dangerous idleness, living lives useless to society. In military institutions [the idle] encounter severe vigilance and prompt correction of wrongs, reforming their habits while instructing and preparing them to be better citizens." Even Liberal opponents of conscription such as the senior statesmen Jose Tomas Nabuco de Araujo praised impressment's social benefits, "Now, is it inconvenient to exercise influence, and even corruption, to give a means of making a living to a vagrant? ... I do not want the army to be composed of criminals and perverts.... There is nothing incorrect [however] ... in forcing a man that has no means of making a living, who could lose himself in idleness, to adopt a profession in the army."(24) Impressment's apologists reasoned that army service reformed the slothful in a Benthamite manner by watching over them and putting them to work for society. The ways discipline played out on a day to day basis in the barracks, however, often did not conform to these wishful precepts. The use of "dishonorable" men to enforce the law tended to undermine its credibility and legitimacy in the eyes of the poor and the privileged alike. The Brazilian phrase, "For my friends, anything; for my enemies, the law," captures well the partriarchal flavor of Brazilian justice in the late 1800s.

Officers had to negotiate an often uneasy discipline with reluctant troops whom much of the larger public regarded with fear and disdain.(25) Some officers took more of an interest in shaping their troops' characters than others, but the army certainly offered a more consistent disciplinary regime than other public institutions with the possible exception of police forces. Troops wore uniforms, performed group exercises, and shaved their heads, and officers required their men to perform work: an element central to a man's moral rehabilitation and health for theorists from Jeremy Bentham to Karl Marx.

The Army as a Proto-Penal Reform Institution and Labor Regime

In most cases, police happily turned over to the army a small portion of apprehended vagrants, "deflowerers," thieves, abandoned children, runaway husbands, migrants, squatters, and hapless passers-by. This practice lightened judicial dockets, freed prison space, and swiftly removed undesirables from the street. Impressment allowed authorities to police family morality, segregating some dangerous noncomformists in barracks and attempting to watch over them.(26)

Police purveyed most recruits. The provincial police of Rio de Janeiro, for example, reported in 1876 that they remitted 325 prisoners to the army as recruits in addition to 229 army deserters.(27) In the Capital District, an 1871 Justice Minister's report showed that the Police Jail sent 231 prisoners to the navy and 245 to the army.(28) Sometimes police or army troops escorted recruits alongside civilian convicts through rural villages and urban centers on their treks to deliver their wards to officials in provincial capitals, thus further confusing the status of soldiers and criminals in the public mind.(29)

Police chiefs used their discretion to send men to the barracks. In forwarding an appeal to Pernambuco's Governor from the father of one pressed soldier, the police chief defended dragooning the youth by stating, "he had sent Claudino Maria de Encarnacao to serve in the army...because he did not have a legal exemption and was trouble maker and a filcher (larapio)."(30) This was common practice in provinces across Brazil, but not every accused transgressor sent to the army became a soldier.

In 1875, Pernambuco's governor reported that from November 1872 to December 1874, only 529 of 1003 men sent to the army recruit depository were actually enlisted.(31) Of the 398 men sent to Pernambuco's recruit depository in 1874: 232 were enlisted, 108 were liberated after proving or paying for legal exemption, 29 were rejected for physical incapacity, two died, one deserted, and 26 remained in the facility. One of the recruits "placed at liberty" in 1874 was Claudino Maria de Encarnacao. His father's appeal to the Governor apparently worked despite the Police Chief's denigrating comments on his conduct. Perhaps the Chief's comments were a pro forma defense of impressment in many cases as similar accusations appeared in police correspondence for most men sent to recruit depositories.

In 1875, Pernambuco's governor lamented that once again he was unable to meet his recruit quota, "325 entered the recruit depository: of these seventy five were freed for proving legal exemption; fourteen, for physical incapacity; two, for being deserters; and one, for being a wanted criminal. Of those approved for service, fifty four were assigned to local garrisons and 158 embarked for Rio" the southern hub of the troop trade.(32) Such aggregate figures on recruit depositories are rarely noted in official reports, but if these figures were typical of other provinces, only around half to two-thirds of those remanded to the army were actually pressed into service while a quarter to a third managed to prove or pay for their exemption. Even if they managed to exempt themselves, their temporary incarceration in recruit depositories gave the incautious reason to proceed with greater trepidation in their dealings with local authorities. Families and patrons lost valuable time and money petitioning officials to secure the release of inductees with exemptions.

Analysis of War Ministry recruitment data from 1870-1882 indicates that the army dragooned more than half of its recruits for enlisted service. This pattern appears to have carried through to the end of the century.(33) Most of those dragooned were individuals who lacked a skilled occupation, a national guard post, an influential patron, a certifiable marriage, or the capital that could exempt them. For these reasons, enlisted army service connoted vulnerability to impersonal authority, poverty, and captivity; it marked an important boundary between the "respectable" or "protected" poor and the "dishonorable" or unprotected free poor, criminals, and slaves.(34) Paradoxically, the forces of order were built largely from the very elements authorities perceived as threats to public tranquility.(35)

The small percentage of Brazilians who actively served in the military might make it appear deceptively insignificant as an institution of social "discipline." In peacetime, only about 2,000 men (volunteers and dragonees) entered the army's ranks in a given year. As the war minister lamented in 1912, in Brazil only 1.1 per thousand inhabitants actively served in the armed forces, compared to 2.2 per thousand in Mexico, 4.6 in Japan, 8.7 in Germany, and 13.5 in France.(36) The Brazilian army's institutional reach through recruitment was shallow compared to Continental Europe where powerful rivals pressured one another to expend vast resources on defense. However, when compared to Brazil's civilian penal system, the importance of the army's role in criminal justice and policing becomes readily apparent. Peacetime praca ranks typically numbered between twelve and sixteen thousand men. As impressment accounted for more than half of these enlistments, then the army incorporated between at least six and eight thousand individuals who were considered "criminal," menacing, or at best unproductive.(37) If civil jails around the country held ten thousand prisoners, then, in this sense alone, the army was by comparison far and away Brazil's largest institution for "criminal" detention and "reform."

Most volunteers for enlisted service signed on to escape hunger, unemployment, homelessness, or even slavery. Paltry wages, spartan conditions, low social esteem, and sometimes brutal treatment meant that volunteers were few and generally of meager circumstances.

In spite of these conditions, some chose an army enlisted career over other possibilities, and a few advanced through the ranks to become officers. But, whether they volunteered or were dragooned, most soldiers demonstrated more similarities than differences. According to General Dermeval Peixoto: "At times there was a scarcity of volunteers in Rio's regiments. [Then] the levas [pressed men] from the North[east] arrived.... The volunteer and the leva were, at the beginning of the [twentieth] century, the same bucolic men, the rude backlander (sertanejo bronco), the unhappy fellow countryman who failed, but who attempted to remake himself in the army."(38) Except for assuming his fellow citizens' "failure," Peixoto provides a sympathetic description of common soldiers. Most were poor uneducated rustics down on their luck. Authorities could safely treat such unprotected men as if they were criminals, and wily police and recruitment agents could extort obeisance and money from the protected poor with the threat of recruitment. When Congress renewed debates on conscription in 1906, for instance, the cordel (penny handbill) songster Leandro Gomes de Barros published a parody, "O sorteio" (the military draft lottery). In it, humble folk planned to bribe officials with pigs, burros, horses, and other goods to protect themselves from the draft.(39)

Desprotegidos existed in relative abundance throughout Brazil. Some regions and provinces, however, produced a disparately large share of soldiers; others enjoyed greater autonomy from recruitment. For instance, the Northeast, Rio Grande do Sul, and the Capital District (Corte) accounted for a little more than half of the nation's 1872 population, but they produced eighty percent of enlistments from 1870 to 1882. Conversely, Sao Paulo, Minas Gerais, and the province of Rio de Janeiro were home to more than one-third of the nation's population, but they provided less than five percent of enlistments in those same years. Thus, the army had a more prominent role in social control in some places than in others. The impoverished Northeast provided more than half of all enlistments, but garrisoned only about one-fifth of the army's troop population. This manpower surplus was exported to depleted regiments in other regions, principally to the south and southeast. Minimum contracts of six years and the possible relocation of troops to garrisons distant from hearth, kith, and kin made being pressed tantamount to a sentence of temporary exile for many. The "troop trade" was in many ways a type of internal penal transportation system for strategically relocating "dangerous" men. This practice derived from the Portuguese tradition of using penal exiles (degredados) as soldiers in undermanned colonial regiments. Some regions traded the security risk of "criminal" and "idle" men to those hungry for labor. Outside of areas where large garrison populations existed such as the capital district and Rio Grande do Sul, economically dynamic provinces like Sao Paulo, Minas Gerais, and Rio de Janeiro tended to supply the army with few troops (dragoonees or volunteers). The troop trade's southward route paralleled that of Brazil's internal slave trade after the State interdicted international sales of bondspeople in 1850.(40) It is little wonder most Brazilians demonstrated repugnance for enlisted army service and for enlisted soldiers.

The Drive for Army Reform and Modernization

For four decades after Independence, traditional recruitment practices sufficed to man Brazil's small army despite numerous complaints and parliamentary debates. Auxiliaried by a provincially organized National Guard (formed in 1831), the army served Brazil's basic defensive needs. The catalyst for reforming the army's recruitment, training, and function was the unprecedented mobilization required by the Paraguayan War (1864-70). Mobilization difficulties led the government to manumit hundreds of slaves and scores of convicts to fight at the front and to conduct widespread impressment, violating the norms that had traditionally limited induction to "unprotected" free men. Unchecked impressment disrupted society on the home front while mobilization's sluggishness helped to turn what was expected to be a brief campaign into a dreary and unpopular five-year ordeal. The Brazilian army's poor performance led officers and politicians to legislate reforms. Their principal concerns included the lack of adequate training and the need for a more modern reserve system to facilitate rapid mobilization. As a war minister complained as late as 1914, conscription was necessary because impressment integrated recruits a few at a time, year-round, hampering the ability of officers to train men according to "consecrated methods."(41)

Discipline was further impaired by the policing duties army troops were routinely required to perform. One year after the Paraguayan War, war minister Rio Branco urged that army service be restricted to garrison duty, warning that this could only succeed "if [provincial] police forces stopped depending on the assistance of army troops," a practice "prejudicial to military discipline."(42) Despite the pleas of Rio Branco and others, provincial governments continued to rely on local army troops as police auxiliaries. In fact, after the 1873 reform of the National Guard largely deactivated that force, the army appears to have been required to perform more police duties than before.(43) For example, in the small town of Iguarassu, Pernambuco (where once prosperous smallholders were now squeezed by the fall in cotton prices after the American Civil War), the police delegate complained in 1874 that the Guard's deactivation had "weakened the principle of authority in a frightening manner." In this particular case, the delegate and others like him were too undermanned to respond to the 1874 Quebra-Quilos (Kilo-Breaking) Revolt: a jacquerie led by frustrated smallholders that swept the interior of Paraiba and Pernambuco. Rioters broke newly adapted standard metric scales in markets in rural towns to protest new taxes, a religious dispute between the emperor and the Vatican, and fears of the new Recruitment Law of 1874. Governors of both these northeastern states relied to army troops to restore order in this regions. Officials pressed some poor men accused of stirring up the rebellion as exemplary punishment.(44)

Even during more peaceful times, army troops performed routine police work. In 1885, a circular letter from the War Minister complained that "north[eastern] states, which lack [adequate] police forces, use army pracas for police detachments."(45) This common procedure could separate troops from their garrisons and officers for extended periods of time. In 1875, Maranhao's governor ordered a local army commander to station his battalion in the troubled town of Caxias:

the presence of a respected force ... will restore tranquility of spirit to Caxias' good citizens, so shaken by the fear of disorder ... The police force currently present in Caxias will withdraw when your force arrives, the requisition of pracas [enlisted men and NCOs] for patrol work will be directed to you ... I recommend that you help the police with its recruitment duties, exemptions will be respected with the utmost scruples, you should not remit any recruits to the capital without honoring the time allowed to each one to prove his exemption.(46)

While patrolling and establishing order, this army battalion was also encouraged to carry out another traditional police function: chasing down their future comrades. In a similar vein, Ceara's governor reported: "The 15th Infantry Battalion and the police corps ... continue to perform duty here. The 15th has for a long time guarded the prison and patrolled the capital, and the police corps is always employed in detachments [in the interior]."(47) At least this governor sought to appease the war ministry by keeping army troops concentrated in the capital, but army reformers wanted more. They demanded that troops be exempt from extraneous duties that distracted them from assiduously drilling for war under the watchful eyes of their officers.

To this end Parliament passed the 1874 Recruitment Law, enacting a limited conscription lottery. Legislators intended the law to abolish impressment, improve the quality of inductees, distribute recruitment more equally among and within provinces, establish a modern reserve system, provide a steady supply of recruits at regular annual intervals, and make army service a respectable duty. However, popular resistance, political party rivalries, the lack of infrastructure, and the army's role in penal justice undermined attempts to enroll citizens and execute the conscription lottery.

Common understandings of the draft lottery's procedures must have been distressing, perhaps even more so if clearly explained. For most of the "honorable" poor, a conscription lottery probably seemed analogous to drawing lots among law-abiding citizens (those who would cooperate with enrollment) to see whose son should serve a lengthy jail term. Though a national draft lottery would not be implemented until 1916, the 1874 Recruitment Law threatened to alter the army's traditional role as Brazil's most important penal reform institution. Meanwhile, army recruitment continued to rely on a mix of impressment and voluntarism. Public perceptions and the civil justice system's dependence on the military as a penal dumping ground could not be altered over night by decree. Other changes in the army's standing and function were essential to alter public attitudes. The low esteem associated with army enlisted service was inextricably linked to its role in penal justice.(48) This role went beyond the incorporation of an important share of the nation's "criminal" elements and routine policing duties. The army also directly administered prisons.

Army Prisons and Orphanages

As noted earlier, army fortresses and stockades played a key role in Brazil's prison system. Forts and army arsenals provided prison space for some of the nation's most dangerous military, civilian, and slave convicts.(49) Liberal reformers had hoped to reduce the civil justice system's reliance on military facilities to confine convicts by requiring provinces to build penitentiaries. However, the foundation of an army-administered penal colony on the island of Fernando de Noronha in the 1830s greatly extended the army's role in Brazil's penal system.

This penal alternative, about 290 miles off Pernambuco's coast followed Portuguese precedents of exiling criminals to colonize distant largely unsettled frontiers. For Brazilian jurists, however, much of its inspiration and legitimacy sprang from the more modern penal transportation and colonization models of England and France.(51) Though much smaller in scale than Europe's penal colonies, by the 1860s, Fernando de Noronha's civil prison population was more than four times greater than the largest of Brazil's provincial prisons and greater than the ideal capacity of all the new correctional penitentiaries combined.

Military tribunals remitted only about one-fifth of the island's inmates; the remainder came from overflowing civil jails. Authorities attributed the rapid growth of the civilian convict population to the lack of penitentiaries capable of adequately administering the sentence of "prison with hard labor." Governors regularly petitioned the War Ministry to accept their convicts to relieve overcrowded prisons. Convict slaves, freemen, and soldiers, some accompanied by their wives and children, formed the island's penal population. Most convicts lived in privately constructed huts that they bought and sold. Some worked as domestics for public employees, army guards, private vendors and their families in the colony. The army organized inmates into military companies to perform daily work and maintain discipline. A company of cobblers made shoes, but most inmates tended fields or herds or fished while better educated forgery convicts manned bureaucratic posts. The colony produced comestible crops and sea-island cotton, but production only partially offset onerous costs.(52)

Table 2

Fernando de Noronha's Convict Population Selected Years
1865-1920(50)

                                    Civilian
              Military          Men          Women          Total

1865            278            1,400           ni           1,678
1869            248              988           21           1,257
1873            231            1,154           29           1,414
1879            278            1,364           36           1,678
1881            257            1,375           36           1,688
1886            251            1,216           ni           1,436
1896            112              260           ni             372
1899              0              232           ni             232
1920              0              447           ni             447
Source: in note above, ni - no information.

During the Paraguayan War, the island's penal population fell brusquely because the government dragooned a few hundred convicts to fight at the front (some of whom petitioned to enlist in return for pardons).(53) In defending the 1874 Recruitment Law, the War Minister deplored wartime mobilization methods: "It is precisely to avoid arriving at the extreme where it becomes necessary to manumit slaves and to free convicts from Fernando de Noronha that it makes sense to adopt a law that enables the government to raise, during a war, the contingents required ... without these improper extraordinary methods."(54) Nevertheless, by the late 1870s, a continuing stream of convicts had restored this population to its antebellum levels. As in civil prisons, inmates convicted of homicide predominated. Among the convicts on the island in 1880, 1,050 civil prisoners had been sentenced for murder; 130 for larceny; twenty-four for counterfeiting; and the crimes of over eighty prisoners were unknown. More than 700 had been sentenced to perpetual imprisonment; others served temporary sentences, except for 13 who should have been executed and 24 whose sentences were unknown.(55) The War Ministry normally only relocated convicts with a sentence of six years or more to Fernando de Nornoha because of transportation expenses.

Thus, the island held the most hardened and recalcitrant of provincial inmates. Prisoners from Pernambuco were the exception to this rule. The port of Recife supplied the island with food, libations, and convicts, giving Pernambucan authorities privileged access to the colony. In 1876, 681 of 1,260 civilian prisoners originated from Pernambuco; other northeastern provinces contributed 282; the Southeast, 215; the South, sixty; the North, twenty-two; and the Central West, none.(56) Again, the weight of the army's role in managing criminals varied among regions and provinces, but even so, Fernando de Noronha wove most provincial prison systems into a national penal network. The penal colony's administration was reformed in 1877, but the noted jurist Evaristo de Morais observed "even after the Justice Minister assumed responsibility for Fernando de Noronha's maintenance the army continued to administer it."(57)

In the 1890s, the Republican government's federalist policies and the prerogatives of reformist army officers led to the closing of what had become, in the words of one prison inspector, Brazil's "central" prison. Fernando de Noronha reverted to Pernambuco's domain, and all inmates from other states and the military apparently had been removed by 1900. Pernambuco continued to rely on the island as a state penal colony for vagrants and capoeiras (practitioners of the unique Afro-Brazilian martial art, dance, and musical form and sometimes members of urban gangs). The closing of Fernando de Noronha as a national prison in 1896 represented a sizeable reduction in civil prison capacity, placing even more pressures on overcrowded provincial jails. While some civilian convicts were still held in military fortresses and stockades, the army's administration of permanent and predominantly civilian prisons drew to a temporary close.(58) Acting as a jailer for civilian criminals was just as inconsistent as was police work and the impressment of "criminals" into the army's ranks in terms of its modernizing goals. Meanwhile, a similar and nearly simultaneous process of expansion and contraction characterized the army's role in managing "dangerous" juveniles.

Army apprentice schools got their start in the early 1800s when Dom Joao VI ordered Rio's army arsenal to take on abandoned children as apprentices; Parliament regularized their status in 1831.(59) Similar schools later spread to army arsenals in Rio Grande do Sul, Bahia, Pernambuco, Mato Grosso, and Para. During the Paraguayan War, the lack of men adequately trained in the use and maintenance of artillery prompted the founding of Rio's artillery apprentice school in 1865. In 1876, infantry apprentice schools opened in Minas Gerais and Goias. At its peak, the apprentice school system handled nearly twelve hundred boys from age ten to eighteen, annually producing scores of "volunteers" for ranks that often experienced troop shortfalls.(60)

Dissatisfied with the quantity and quality of troops, the army high command hedged its bets on the 1874 Recruitment Law's success. It expanded its investment in the education of wayward boys to secure cheap labor and guarantee a future generation of more highly skilled soldiers and arsenal artisans. Poor parents or guardians who wished to enroll their boys were paid a small premium, and the sons of operarios militares (arsenal employees) and army privates and non-commissioned officers (pracas) were also given privileged access to army apprentice schools. If an artisan apprentice demonstrated aptitude he could be graduated to the status of "army laborer" at age sixteen. Regulations obligated elected apprentices to serve the arsenal for a minimum of ten years. Those who had not demonstrated ability would serve as "volunteers" in regular army units for the standard six years of service as did artillery and infantry apprentices when they reached nineteen years of age (legally lowered to seventeen in 1880).

Some artillery and infantry apprentices who demonstrated capacity and discipline were transferred to active duty with a non-commissioned officer's rank. A handful climbed from apprentice schools to the army's system of higher education, the gateway to a commissioned officers rank.(61) A small minority of apprentices achieved appreciable social mobility through this system, but the vast majority were not so fortunate.

While some parents sent their sons to serve in army apprentice schools, the police and juizes de orphaos (literally, the orphans' judge) enrolled most enlistees. For instance, the provincial police of Rio de Janeiro reported sending forty nine minors to the army's artillery apprentice school in 1875. A large number of Paraguayan orphans were remanded to Rio and incorporated into the artillery apprentice school at the end of the Paraguayan War, temporarily swelling its ranks.(62)

Table 3

Population of Army Apprentice Schools(63)

                   1871-2   1875   1880   1885   1890   1895   1899

Bahia                146     152     50     50     62     80      0
Goias                  0       0     40     34     35      0      0
Mato Grosso           ni      82     11     50     50     ni     79
Minas Gerais           0       0     25     40     43     80      0
Para                  50      96     18     49     50     50      0
Pernambuco           128     120     50     50     50     80      0
Rio (Artisans)       203     232    100    104    265    250    250
(Artillery)          504     430    319    280    198      0      0
Rio G. do Sul         58      50     50     49     50     77     79
                   1,089   1,162    663    706    753    617    408

Source: in note above. ni - no information.

Army apprentice schools shared a picaresque reputation with their sister institutions: naval apprentice schools.(64) The rough company and severe discipline of army and navy apprentice schools became popular folklore. It remains a commonplace for parents to threaten to enroll their children in a navy apprentice school when they misbehave. Confidential Naval Ministry correspondence reveals that the apprentices of Rio's army and navy arsenals included some of the city's most notorious capoeiras, who terrified the public and frustrated the police. Correspondence between Rio's police chief and the Navy Ministry recorded several cases of apprentices illegally remitted for active military service as punishment for their involvement in capoeira gangs.(65)

In September 1876, the slave mother of one naval apprentice, Pedro Nicolau Liberal, requested that her son be returned from active duty in the navy because he was only fifteen years old. Rio's police chief claimed Nicolau, armed with a razor, had been involved in a capoeira street fight. Despite his age, the chief convinced the Naval Minister to punish Nicolau by enrolling him for active duty and shipping him out of Rio. Nicolau's mother explained that her son worked "assiduously" at the arsenal in order to buy his mother's freedom and had never involved himself in capoeira. Nicolau's mother, probably with the assistance of her owner, had her son's baptismal certificate certified and notarized by local officials. After a few months at sea, the navy reluctantly returned Nicolau to police authorities for normal judicial processing.(66) Other boys who did not have family or patrons in a position to protect them were probably not so lucky as Nicolau.

Correspondence from the previous year indicates that the impressment of underage apprentices like Nicolau probably formed part of a clandestine effort to bust up gangs (maltas) composed by "army pracas and army and navy arsenal workers."(67) On 1 June 1876, a navy arsenal worker was killed by a "Portuguese" in a capoeira rumble. At the crime scene, the police apprehended three members of the "malta of navy arsenal workers."(68) In response, Rio's Police Chief sent a special confidential request to the Naval Ministry:

authorities cannot always find a means for any legal action [against capoeiras] as they ... disperse before police forces arrive, I judge it my duty to suggest to your excellency [Naval Minister] that these turbulent maltas, many of them navy arsenal workers, [be] transferred to active duty ... and sent away from Rio ... I add that the War Minister [army], at my request, agreed to transfer his arsenal workers, when they are imprisoned for practicing capoeira ... this will reduce the repetition of such events.(69)

The chiefs of the army and navy agreed to illegally transfer these troublesome arsenal workers and apprentices for regular service. These actions demonstrate the willingness of the high command to side-step the law and the rights of common citizens to reinforce public order. Only those who could activate a patronage network to file costly petitions could hope to protect their rights. The incorporation of troublesome youths in the apprentice schools was a double-edged sword for the army, navy, and police who attempted to work in concert to control these "dangerous" young men. Ironically, their military status as apprentices afforded these capoeiras some protection from arbitrary police procedures.(70) Military apprentice schools, to the dismay of officers and authorities, had been partly subverted by workers and apprentices into capoeira schools that propagated values and practices that military discipline was intended to suppress.(71) Regular service and transfer loomed as punishment for bad behavior for these underage soldiers, just as troublesome army regulars were often punished with transfer.

In the Republic's first years, army apprentice schools initially expanded, but over the course of the 1890s, the government closed them down. In 1892, the artillery apprentice school was converted into the army's Sergeant School and infantry apprentice schools in Minas Gerais and Goias were shut down.(72) A consolidation of army arsenals in 1896 closed half of the artisan apprentice schools, and by 1901 those remaining appear to have been shut down. The apprentice schools fell victim to the army's restructuring, budgetary constraints, and plans to modernize education. The army partially replaced the apprentice system with military colegios intended to prepare boys with better educational backgrounds for careers as commissioned officers. By shutting down apprentice schools the army reduced the State's institutional capacity to deal with the problems of abandoned and delinquent youths (although navy apprentice schools continued at a reduced capacity).(73) The army's modernizing ambitions and desire to improve the public image of enlisted service were inconsistent with the duties of being the nation's largest institution for formal and informal policing, orphan and juvenile delinquent care, civil prison administration, and as a punitive destination for wayward men and adolescents.

Changing Patterns

Patterns of army recruitment and service only began to change when the Republican government shortened minimum army service contracts. In 1897, minimum contracts dropped to five years and then to three years after 1900. The 1908 Obligatory Military Service Law reduced terms to two years, and established new criteria to prohibit those with criminal records from serving in the army. After 1916, regulations required infantry men to serve only one year. Concurrently, influential prussophiles, like General Hermes Rodrigues da Fonseca (invited by Kaiser Wilhem II to observe the German army in 1908 and elected as Brazil's President in 1910) insisted on the primacy of army training and troop readiness. Hermes oversaw improvements in the sanitary conditions of barracks and mess kitchens. Less police work meant soldiers had less onerous and more stable work schedules. Improved conditions, however modest, and shorter service contracts attracted more volunteers.

Reports of increasing numbers of volunteers reflected these improvements. In 1914, the War Minister complained that the 1908 Obligatory Military Service Law was flawed. While it called for the organization of a draft, the lottery could only be executed when there were not enough volunteers. Since there had been an increasing number of volunteers after 1908, Congress had consistently questioned the need for a draft lottery. The War Minister then ironically celebrated that "in the last few years the numbers of volunteers has happily diminished" making it necessary to institute the draft.(74)

The focus on training and field exercises began to make a perceptible difference in army discipline and performance. Army desertion and crime rates began to decline slightly. Improving discipline allowed officers to focus more energy on training as they were less burdened by the bureaucracy of punishment.

In 1916, the army finally guaranteed that the draft would be instituted by allowing only one month out of the year for voluntary enlistment. They further required that volunteers meet more rigorous health standards, be literate, certify their good comportment with local police, and be single between the ages of twenty one and thirty. Instead of relying on draft enlistment rolls which had failed to produce adequate lists in the past, the 1916 draft lottery used civil registry records, a bureaucratic instrument created during the early Republic and unavailable to previous generations of reformers.

As improvements occurred, officers bristled at suggestions that their troops perform duties that smacked of police work.(75) In 1913, the War Minister complained that a few scattered units in Amazonas and Acre were being used by local prefects to perform "police functions," duties "alien to those required of the army in the 1891 Constitution." The War Minister requested that all such isolated units be fused with larger battalions where soldiers would receive proper training. It would appear that elsewhere the high command had succeeded in preventing state and local officials from consistently diverting army troops for police duties.(76) Most troops now spent their active hours in and around their garrisons training. Crowds in Rio marvelled at the novelty of Hermes' battalions who could march smartly in tight formations.(77) Discipline and training, however, often remained slack, especially in frontier garrisons.(78)

In the 1910s an influential coalition of educated Brazilians reached a consensus that military conscription would help resolve a variety of threats to national strength and unity, ranging from regionalism to poor hygiene, racial "degeneration," and national defense. Pro-military intellectuals had some success in selling conscription to more prosperous Brazilians as a masculine eugenic motor for improving the health, hygiene, intelligence, discipline, and sense of national identity among a broader cross-section of Brazil's racially and ethnically diverse lower classes. They also began to attempt to market conscription to the poor with propaganda and symbolism that emphasized the "manliness" and "respectability" of defending national honor as conscripts.(79) For example, a 1916 editorial in a popular Rio newspaper praised the youth for its vigor, playing soccer instead of crowding cafes and flop houses. It noted that without "obligatory [military] service for every able-bodied man, a nation cannot confront an enemy without running the risk of an inevitable defeat."(80) The renowned poet and lobbyist for the draft lottery, Olavo Bilac, portrayed conscription as a regenerative institution with more of a preventative than a punitive role in social reform. The grim lessons of World War I and heightening concerns over the need for public sanitation to prevent the spread of epidemic disease made Brazil's upper and middle-classes more concerned with the health and hygiene of its common citizens. If Brazil hoped to mobilize an army capable of warding off powerful potential European aggressors (let alone its regional rival Argentina), it had to take stock of its human resources. Since most poor Brazilians did not attend public schools, the draft became a way of socializing a larger proportion of the lower classes and providing them in Bilac's words with "obligatory primary education" in the barracks. These changing conditions and emerging ideologies converged to make the realization of a national draft lottery possible in 1916. The public image of army service had undergone a partial facelift even though many continued to evade the draft.

The 1916 draft lottery brought a significant change in the army's function in Brazilian society. The army now actively tried to reject "criminals," and began to draw young men from more stable urban working-class families (those more likely to appear on civil registries relied upon to compile draft lists). Upper and middle-class men continued to exploit legal loop holes to dodge the draft, but the army no longer depended heavily on the "criminal" classes to fill its ranks. Henceforth, younger raw recruits entered the barracks at the same time with peers of similar age, serving in most instances closer to their home and families. These reforms helped make army service seem less of a threat to the dignity of "honorable" poor families. Garrisons now marched to a new rhythm of perennial basic training aimed at building a large reserve force.

The growth of militarized state police forces in the Old Republic has correctly been interpreted as a government response to rising crime rates, the decline of slavery, increasing migration and immigration, and dizzying urbanization. Powerful and wealthy states also built up militarized police forces as a hedge against coercive federal intervention by the army.(81) In addition to these pressures, one must add the army's withdrawal from everyday police work: a factor overlooked up to now which highlights the need to look across institutions of public discipline. Thus, army modernization and specialization itself sped the growth of militarized state police forces that rivaled the army's hegemony over government-sanctioned violence during the Old Republic. Reformist officers viewed large militarized state police forces as a threat to their modernization platform: "The same politicians that in their states organize these armies [of state police] try to weaken and retard the improvement of the federal army . . . police armies are organized to counterpoise the federal army!!!!"(82) While state police forces may have rivaled army power in wealthy states, temporary incarceration and intimidation were in most instances the best these growing police forces could muster to counter lesser crimes for lack of jail space, schools, orphanages, or other institutions capable of incorporating the unprotected poor into State institutions.

The Republican government's greater propensity simply to abandon public offenders in uninhabited and insalubrious frontier regions was likely a symptom of this transition in institutional strategies of social control. As Miguel Calmon noted, "Since the beginning of the Republic, deportations were always made to the extreme North. . . . Marshal Floriano Peixoto's government decreed the internal exile (desterro) of various persons of public notoriety to Cucuhy and Tabatinga. After the Vaccine Revolt (1904) . . . deportations to [the state of] Amazonas were made larga manu."(83) Similar deportations followed the Canudos Rebellion (1896-97), the Chibata Revolt (1910), and the Tenente Revolts (1922; 1924-27). Police took advantage of these deportations to round up their local nemeses and ship them off with rebels and rioters.(84) This type of unregulated ostracism had occurred under the Empire, but not on this scale. Earlier in the 1800s, the foot soldiers of insurgent forces had frequently been transferred to other regions and integrated into the army there. Clearly army reforms cut off an important penal destination for Brazil's "dangerous classes."(85)

In the absence of institutions capable of handling these problems, the Republic abdicated much of its role in reforming criminals. In 1923, an advocate of prison reform called on the government to reappropriate Fernando de Noronha as a national penitentiary for military convicts because of the overcrowded and inhumane conditions of army and navy jails.(86) Instead, the Federal Government developed new penal colonies on Brazil's extreme frontiers, packed military fortresses, and revived other island penal colonies. Officials even restored the practice of using ship hulks as fetid floating prisons for political and common criminals.(87) Clevelandia was originally founded in the vast wetlands of Northern Amapa to serve as a refugee colony for those fleeing drought in Ceara in the early 1920s. After the first Tenente Revolt (a coup attempt by junior army officers in Rio de Janeiro) in 1922, however, the few colonists who had arrived were overrun by exiled "criminals." Clevelandia received some summarily exiled rebel enlisted men (rebel officers were not sent), but most of those relocated were common criminals rounded up by the police. According to a resident, the exile population rose to 1,630 individuals. Disease decimated colonists and inmates alike, and many fled to Guiana. The thirty soldiers posted there could not adequately police much less guard the exile population surrounded by vast rain forests. The army was unwilling to use more resources and reassume the role of the nation's largest jailor. Clevelandia was not so much a reform institution as it was a penal dumping ground. When the Old Republic fell in 1930, the remaining prisoners in Clevelandia were pardoned and reformers could point to it as a symbol of a decadent regime.(88) Army reforms and the lack of development in Brazil's civil penal system during the Old Republic encouraged the government to resort to these desperate and inhumane penal solutions. Political instability would see the revival of these brutal penal practices under President Vargas' dictatorial rule from 1937-45, despite his increased federal spending to build modern penitentiaries in states across Brazil.(89)

Conclusions

Theorists from Max Weber to Michel Foucault, E. P. Thompson, and Benedict Anderson have emphasized in different ways the importance of bureaucratic public institutions (penitentiaries, schools, armies, navies, orphanages, police forces, asylums and administrative bodies among others) in disciplining individuals for participation in nationalist societies.(90) Other scholars, such as Eugen Weber, have more specifically focused on the significant role national militaries played in introducing citizens, particularly rural inhabitants, to nationalist ritual and doctrine.(91) Less attention, however, has been focused on how different public institutions developed in relation to one another. The question of institutional fit is particularly useful when examining how one of the largest agro-export economies in the world selected and adapted North Atlantic bureaucratic ideologies and practices. Developing nations like Brazil lacked the funds and conditions to reform the entire panoply of modern institutions in a more contemporaneous fashion. Difficult choices had to be made on competing institutional agendas supported by constituents with differing degrees of political influence. The struggles among groups and individuals within the polity and the State shaped tortuous, difficult to follow paths toward institutional reforms and their enforcement. These paths could not be reliably predicted by measuring levels of capitalist development although it certainly shaped political battles and outcomes.

This essay suggests, nonetheless, that there may be significant patterns within apparently chaotic patchworks of smoke raised by struggles to reform a range of public institutions in different international, national, and regional contexts. The evidence examined here indicates the importance of understanding how deeply interconnected institutions of social control were in the late 1800s and early 1900s in Brazil. The army's attempt to rationalize its training, recruitment, and function directly impinged on Brazil's police, prison, orphanage, and judicial systems. The weight of this transition varied among provinces and regions, but no other set of reforms, save those abolishing slavery, could compare to the scale of its impact on public institutions of social discipline and the lives of the poor. Halting efforts to modernize a variety of state institutions occurred in the late 1800s. Those who held power in Brazil could have selected to give greater emphasis to modernizing prisons, police, educational, or orphanage institutions before reforming the army. In some instances, these institutions did receive the attention of reformers, but the primacy of national defense and the political influence of army officers and their allies ultimately made military reform a higher priority of the central government.

The choice of conscription over other institutional reforms also speaks to some of the doubts that Brazil's leaders had about the abilities of their population. Lingering strains of racist social Darwinism caused many of Brazil's elite to doubt the state's ability to "civilize" the poor whose composition was predominantly of mixed African, European, and Indian heritage. By contrast, Argentina adopted conscription much earlier (1901) than Brazil and gave greater stress to public education for its largely European-based population. Brazil's racial heterogeneity and the assumed "inferiority" of African and Indigenous peoples may have made the discipline and hierarchy of the barracks more attractive than the classroom for often bigoted political decision makers.(92) Indeed, one army "anthropologist" used data collected on draftees to hint that conscription would accelerate the "whitening" of Brazil's population.(93) In any case, the uncertainties of World War I pushed the army over a threshold that transformed its relationship to civil society (particularly the poor) as well as related institutions of social discipline.

The regional focus of secondary literature on police, prisons, and social control in Brazilian historiography of the 1800s and early 1900s (referred to at the head of this article) suggests that the State stepped into the breach to discipline freedmen and immigrant workers as slavery declined. The army's retreat from its once prominent position in penal justice and everyday police work indicates that caution and qualification need to be used when generalizing about the State's increasing role in "disciplining" the free wage labor cohort. As federal armed forces began to modernize, new government institutions had to supplant the army's former duties. The Republic's federalist system shifted the brunt of police problems to state governments, and some were in a better position than others to make this transition. For example, prosperous Sao Paulo and Minas Gerais did not press many of their "dangerous" citizens for army service in the late 1800s, and because they possessed small army garrisons, they did not depend heavily on the army as an auxiliary for police forces. Conversely, Rio Grande do Sul, the Corte, and the Northeast had comparatively high rates of impressment and a greater dependence on the army to perform police duties, guard civilian convicts, and harbor abandoned children. For these areas, army reforms had a more profound impact on the web of institutions responsible for social control. The ability of state officials to discipline citizens appears to have been greatly reduced by army reformism in these cases. This important and largely overlooked factor needs to be taken into consideration when making comparisons among regions, something that has gone unmentioned in scholarly literature on social control in Brazil. In future, researchers must look more intensively beyond the State for answers as to how immigrants, former slaves, poor freeman, employers, and political bosses negotiated the transition from a slave to a wage labor market.(94)

In 1880, if one conservatively adds the number of men pressed into the army and its apprentice schools to those whom it guarded in stockades and penal colonies, the army directly handled some eight to twelve thousand men and adolescents considered "criminal," threatening, or idle. While a fuller survey of the nation's prison population awaits further research, the evidence examined here indicates that the army alone dealt with a large portion of the civil "criminal" population, perhaps equal to or greater than the population of the civil prison system. Under the Republic, military reforms slowly closed off this significant institutional destination for "criminals" and reduced the credible threat of such punishment. Most states did little to expand or modernize penal institutions as their populations grew. Army reforms shut down a crucial part of the penal justice system's institutional capacity to deal with "dangerous" men and boys. In this sense, the State's ability to discipline its wayward citizens diminished in most areas.

Unlike the North American South, Brazilian states did not establish large-scale systems of publicly organized convict-lease systems nor a good roads campaign to make chain-gang labor a viable public necessity. These changes indicate that more traditional patriarchal strategies of social control (including paramilitary forces) probably played a more important role than the "State" in the transition from slave to wage labor in Brazil, particularly outside of the more prosperous, populous, and urban states of the South and Southeast.(95) Only when the central State became a more aggressive patron of the protected poor and successfully challenged certain patriarchal precepts did this role change. This process of government centralization which undermined the influence of regional political bosses began under the Old Republic, and conscription played a fundamental part in this State building.(96) Conscription laid an important foundation that made possible Getulio Vargas's dictatorial Estado Novo (1937-45) and his more profound centralization of State power and direct bureaucratic rule.

If one equates the army's modernization with progress, as turn-of-the-century Brazilian leaders often did, then one has to question exactly how this progress enhanced the state's ability to "discipline" its citizens. Under the Republic, the results of reforms were very uneven. The 1916 conscription lottery made the army responsible for educating and disciplining a much larger cross-section of Brazilians drawn from documented "honorable" poor families. Conversely, army reforms and the lack of prison development diminished the state's ability to reform and discipline "criminals" and the unprotected poor more generally. The importance of understanding the historic connections between the army and penal justice becomes even more apparent when one considers the periodic role expansion of the military in this arena since 1930 in Brazil.

Thus, 1 suggest that, at least in the Brazilian case, rather than relying on rather blanket assertions that the state began to increase its role in disciplining the work force as slavery declined, one must distinguish among strata of the poor. Also, analysis should consider the interconnectedness and reach of public disciplining institutions in different regions and the strategies they pursued. When taking these factors into consideration, I argue that army reform significantly reduced the state's institutional capacity to discipline common "criminals" and the "unprotected" poor during the Old Republic (1889-1930): precisely the moment in which a post-abolitionist order was being negotiated. The approach emphasizes the importance of national trends to an examination of issues often obscured when examined through a regional framework. The analysis also looks across several key institutions to trace patterns that do not become apparent when case studies limit themselves to a single public organ (as most archive-based studies do). Because of a lack of more solid statistics on those incorporated into public "disciplining" institutions, however, the arguments in some instances are necessarily more suggestive than definitive.

The army's centrality to policing and the criminal justice system in Brazil and many other nations was a crippling impediment to military reformism. Reformers hoped to make enlisted service acceptable to the sons of "respectable" families, ending its association with emasculating punishment and criminality. After the Paraguayan War (1864-1870), changes in the army's role as an agent of institutional discipline traced a zigzagging path. In the late 1800s, State disciplining strategies turned away from and then back toward the draft lottery of 1916 based on the ideal, if not the essence, of universal male conscription. Modern military mobilization and its outcomes differed around the world, but their linkages with a variety of other social and political issues remain poorly illuminated. It is probable that many of the patterns examined here were common to the experience of developing nations in Latin America and elsewhere. The primacy of defense concerns for States across the world made military reforms the most widely accepted Trojan horse of western European bureaucratic practices.

In much of Europe and the United States the development and expansion of penitentiaries, reformatories, poor houses, schools, penal colonies, and orphanages enabled their armies to make a smoother transition from impressment to conscription and modern military discipline. Particularly in continental Europe, the need to mobilize effective and efficient armies to deter the threat of invasion by powerful neighbors facilitated conscription's adoption. Conversely, in Brazil and most of Latin America, political decision makers were not so pressured to invest heavily in armed forces to protect their property from foreign incursions or internal subversion.

Conscription began to make the implementation of these bureaucratic European disciplining methods more possible in Brazil, but in the meantime, the government turned its back on most efforts to incorporate and reform the "unprotected poor" in favor of socializing a larger portion of the "honorable" poor. The advance of conscription relieved the army of one of its primary traditional functions as a proto-penal reform institution of the ancien regime. Thus, in industrializing Europe, the rise of conscription, emigration, fears of foreign invasion, and the development of modern reform institutions and penal colonies were complementary processes in the development of social "discipline" and nationalist sentiments. In Brazil, conscription reform also worked to accelerate the development of other modern institutions to replace the army's previous function as de facto jailor for unruly youths and non-homicidal "criminals." But the expansion of Brazil's prisons, orphanages, and schools did not keep pace with the demands created by the reduction in the state's capacity to deal with less serious offenders and healthy vagrants brought about by army reforms. The reach of these disciplining institutions never extended to as large a proportion of the Brazilian populace as they did in Europe, North America, and Japan. This rough comparison is drawn out here to suggest the utility of examining the sequencing and depth of institutional reform in Latin America to develop a more rigorous comparative history of state building and the culture of social discipline. Comparative studies of conscription, one of the last widely accepted forms of corvee labor, should provide rich new insights into state building because it is an institutional thread linked to the histories of coercive and free labor systems as well as open and autocratic political systems.

Department of History East Lansing, MI 48824-1036

ENDNOTES

The author would like to thank Robert M. Levine, Elizabeth Kuznesof, Hendrik Kraay, Jeff Lesser, Joan E. Meznar, Timothy Joel Coates, Mark D. Szuchman and the anonymous reviewer for penetrating comments on earlier versions of this manuscript. I claim full credit, however, for any mistakes or misinterpretations herein. Generous support from the Fulbright Foundation and the National Endowment for the Humanities helped make the research and writing of this article possible.

Abbreviations for archives in the notes include: Arquivo Nacional, Rio de Janeiro (ANR); Biblioteca Nacional de Musica, Rio de Janeiro, (BNMR); Biblioteca Nacional Obras Raras (BNOR); Arquivo Geral da Cidade do Rio de Janeiro, (AGCRJ); Arquivo Historico do Exercito, Rio de Janeiro (AHER); Arquivo Publico do Estado de Pernambuco, Recife (APEP); Arquivo Publico do Estado do Rio de Janeiro (APERJ).

1. On the Rurales, Paul J. Vanderwood, Disorder and Progress: Bandits, Police, and Mexican Development, revised ed. (Wilmington, 1992).

2. Theda Skocpol, Protecting Soldiers and Mothers: The Political Origins of Social Policy in the United States (Cambridge, 1992), 39; Also see essays in States, Social Knowledge, and the Origins of Modern Social Policies, eds. Dietrich Rueschemeyer and Theda Skocpol (Princeton, 1996); Richard M. Titmuss, Essays on the Welfare State (London., 1958), Chap. 4; Jaime A. de Araujo Oliveira and Sonia M. Fleury Teixeira, (Im)previdencia social: 60 anos de historia da Previdencia no Brasil (Petropolis, 1985), 19-22; James M. Molloy, The Politics of Social Security in Brazil (Pittsburgh, 1979).

3. Charles Tilly, "War Making and State Making as Organized Crime," 169-86, in Bringing the State Back In, eds. Peter Evans, Dietrich Rueschemeyer, and Theda Skocpol (Cambridge, 1985); Tilly built on Frederic C. Lane's classic analysis in "Force and Enterprise in the Creation of Oceanic Commerce," in The Tasks of Economic History (Supplemental issue to the Journal of Economic History) 10 (1950): 190-31; Lane, "The Economic Meaning of War and Protection," in Venice and History: The Collected Papers of Frederic C. Lane (Baltimore, 1966 [1942]); For an application of Lane's ideas in Latin America see John Markoff and Silvio R. Duncan Barreto, "Civilization and Barbarism: Cattle Frontiers in Latin America," Comparative Studies in Society and History (Oct. 1978): 587-620.

4. As Eugen Weber noted even the French who had invented modern conscription began to adopt Prussian models by 1868; in Peasants into Frenchmen: The Modernization of Rural France, 1870-1914 (Stanford, 1976), 292. Academicians seldom compare the coerced labor performed by many soldiers and sailors to other coercive labor forms like slavery, indentured servitude, serfdom, debt peonage, and so forth (although Brazilians in the 1800s often did). James Scott's theoretical analysis of resistance, for instance, often mentions in passing military impressment and conscription as a type of labor tax exacted from under classes, but in his bibliography there is not one book and only one article that specifically addresses these topics. Scott, Domination and the Arts of Resistance: Hidden Transcripts (New Haven, 1990). Interestingly, the one article he does cite is on Latin America, Carolyn Fick's "Black Peasant and Soldiers in the St. Domingue Revolution: Initial Reactions to Freedom in the South Province," In History from Below: Studies in Popular Protest and Popular Ideology in Honour of George Rude ed. by Frederick Krantz (Montreal, 1985).

5. On crime and punishment in Brazil see, e.g., Martha Huggins, From Slavery to Vagrancy (New Brunswick, 1985); Thomas Flory, Judge and Jury in Imperial Brazil, 1808-1871 (Austin, 1981); Thomas H. Holloway, Policing Rio de Janeiro: Repression and Resistance in a Nineteenth-Century City (Stanford, 1993; Holloway, "The Brazilian 'Judicial Police' in Florianopolis, Santa Catarina, 1841-1871," Journal of Social History 20:4 (Summer 1987): 733-56; Sydney Chalhoub, Trabalho, lar, e botequim: O cotidiana dos trabalhadores no Rio de Janeiro da belle epoque (Sao Paulo, 1986); Boris Fausto, Crime e cotidiano : a criminalidade em Sao Paulo (Sao Paulo, 1984); Patricia Ann Aufderheid, "Order and Violence: Social Deviance and Social Control in Brazil, 1780-1840," (Ph.D. diss., University of Minnesota, 1975; Marta Abreu Esteves, Meninas Perdidas: Os populares e o cotidiano do amor no Rio de Janeiro da belle epoque (Rio, 1989); Maria Helena P. T. Machado, Crime e escravidao: trabalho, luta, e resistencia nos lavouras paulistas 1830-1888 (Sao Paulo, 1987); Bernice Cavalcante Brandao, et. al., A policia e a forca policial no Rio de Janeiro, (Rio, 1981); Magareth Rago, Do cabare ao lar: a utopia da cidade disciplinar, Brasil, 1890-1930 (Rio, 1985); Ricardo D. Salvatore, "Penitentiaries, Visions of Class, and Export Economies: Brazil and Argentina Compared," The Birth of the Penitentiary in Latin America: Essays on Criminology, Prison Reform, and Social Control, 1830-1940 Ricardo D. Salvatore and Carlos Aguirre eds. (Austin, 1996), 194-223; Paulo Roberto Staudt Moreira, "Entre o Deboche e a Rapina: Os cenarios socials da criminalidade popular em Porto Alegre (1868-1888)," (M.A. Thesis Universidade Federal do Rio Grande do Sul, April 1993); Carlos Antonio Costa Ribeiro, Core Criminalidade: Estudo de Analise da Justica no Rio de Janeiro (1900-1930) (Rio, 1995); Carlos Eugenio Libano Soares, A Negregada instituicao: os capoeiras no Rio de Janeiro (Rio, 1994); Marcos Luiz Bretas, A guerra nos ruas: Povo e Policia na cidade do Rio de Janeiro (Rio, 1997).

6. On problematic police discipline see, e.g., Holloway, Policing; Bretas, A guerra nas ruas.

7. Exemptions and their significance for social and gender hierarchies are explored in Peter M. Beattie, "The House, the Street, and the Barracks: Reform and Honorable Masculine Social Space in Brazil 1864-1945," Hispanic American Historical Review 76:3 (Aug. 1996): 439-73.

8. Caio Prado Junior, The Colonial Background of Modern Brazil, trans. Suzette Macedo (Berkeley, 1967), 361-85.

9. Evaristo de Morals, Prisoes e instiuicoes penitenciarias (Rio, 1923), 49.

10. Morais noted that for lack of funds Rio's Casa de Correccao did not exactly follow the panopticon design of its original architectural design. Morals, Prisoes, 12-13; The American missionary of penal reform, Enoch Cobb Wines, reported that Fernando de Noronha was Brazil's "central" prison and that the only prisons with some attributes of modern reform institutions were in the Capital, Sao Paulo, Bahia, and Pernambuco. The State of Prisons and of Child-Saving Institutions in the Civilized World (Cambridge, 1880), 552-4; On Rio Grande do Sul see Moreira, "Entre o deboche," 121-58.

11. Morais, Prisoes, 15.

12. In 1876, Pernambuco had forty-two jails; twenty-two were private properties; thirteen, provincial; five, municipal; one, national (Fernando de Noronha); and one, clerical. The governor noted that "except for the Casa de Detencao, these jails do not conform to constitutional precepts." Relatorio do Presidente Joao Pedro Carvalho de Moraes (Recife, 1876), 20. A similar report on prisoners in Rio Grande do Sul in 1866 noted that the state-owned prisons in 10 major urban centers while private buildings were used in 16 less populous areas. Letter from the Police Secretary of Rio Grande do Sul to the Police Chief of the Corte, SPCOA, Porto Alegre, Dec. 29, 1866, ANR, maco IJ6 517, no folha nos. A description of a small municipal jail in the interior of Bahia can be found in E. Raulino, O sentenciado 304 (Leitura para as prisoes ([Bahia, 1902]), 8-9.

13. "Police Chief's annexo" in Relatorio do Presidente Francisco de Faria Lemos (Fortaleza, 1876), 7.

14. Jose Gabriel de Lemos Brito, Os sistemas penitentiarias do Brasil, 3 vols. (Rio, 1924-27), I:309-17.

15. Enoch Cobb Wines noted that Brazilian delegates to his international conference on prisons complained that Pernambuco spent the equivalent of US $500,000 annually on its House of Detention in Recife. Wines, State of Prisons, 554.

16. To simplify lengthy citations of Governors' reports to provincial (after 1889 "state") governments, I have listed sources found on microfilm in the periodical section of the Biblioteca Nacional by locale, date, prisons listed, and page number. Often data are found in appendixes to reports written by police chiefs as noted below. Ceara: Casa de Detencao, (July 4, 1871), 6-7; (May 31, 1876: Police Chief's annexo), 6; (April 1, 1881: Police Chief's annexo), 62-3; (September 9, 1886: Police Chief's annexo), 7; Pernambuco includes Casa de Detencao and after 1900 the state Penal Colony of Fernando de Noronha: (March 1, 1871), 9; (March 1, 1876: Police Chief's annexo), 4; (March 1, 1881), 25; Reported 32 deaths in 1885; (July 3, 1886), 15; (March 6, 1900: Police Chief's Annexo), 6; Bahia includes the Cadeia de Prisao com Trabalho, the Casa de Correccao, and the Prisao de Gales in the Navy Arsenal up to 1881: (March 1, 1871: Police Chief's annexo), 17-8; (March 1, 1876); (April 3, 1886), 43; (April 7, 1892), 8-12; Rio Grande do Sul: Casa de Correccao, (March 14, 1871), 81; Cadeia Civil da Capital, (March 7, 1886), 123; (September 22, 1926), 8; Sao Paulo: Casa de Correccao and Cadeia da Capital: (February 5, 1871), 11, 14; (February 2, 1876), (January 13, 1881), 156; In 1886 includes Cadeia de Capital (171), Limoeira (112), Pindamonhangaba (117), Mogymirim (131), Queluz (134), Campinas (312), (February 15, 1886: annexo 8), 12, 22-30; Minas Gerais: Cadeia da Capital (March 2, 1871: Police Chief's annexo); (May 4, 1881), 29; Minas: (1891) 29; Includes Minas' 110 jails in 1906 and 1911, (1906: Police Chief's annexo), 47; (1911: Police Chief's annexo), 8; Mato Gross: includes Cadeia da Capital and in 1876 Presos Correcionais: (August 20, 1871), 12; (May 3, 1876), 11, 17; (May 3, 1881: Police Chief's annexo), mapa #1; (September 7, 1921), 56; Para (March 25, 1886), 78; Rio (Capital District): Casa de Detencao, in 1886 includes Casa de Correccao and Prisao Civil, in 1896 Casa de Detencao and Casa Correccao, in 1901, 1906, and 1910 only the Casa de Correccao, in 1916 Casa de Correccao and Casa de Detencao: (1871), 29; (1877), 267; (1881), 155; (1886), 117-8, 126; (1891) 96, 100; (1896), 84, 87; (1901), 151-4; (March, 1906), 68; (April 1911), 70; (April, 1916), 81-2, 88; Paraiba: Cadeia da Capital (October 16, 1871), 8; Cadeia da Capital and de Pombal for 1881 only, (September 21, 1881), 11; Cadeia da Capital only (August 1, 1886: Police Chief's annexo), 20-1; (October 1, 1901), 15; (September 1, 1906), 6; (September 1, 1911), 6; Rio de Janeiro (province/state) (October 22, 1876), mapa 2; (June 3, 1881: Police Chief's annexo), 24; Casa de Detencao only (August 6, 1886), 52.

17. See, e.g., Robert Hughes, The Fatal Shore: The Epic of Australia's Founding (New York, 1987), chapter 6.

18. Lemos Brito, Os sistemas, II:70-85.

19. Edward L. Ayers, Vengeance and Justice: Crime and Punishment in the Nineteenth Century South (New York, 1984).

20. Martha Huggins, for instance, suggests that vagrancy laws replaced slavery as a repressive labor system in Pernambuco. She suggests that petty offenders held in Recife's House of Detention were turned over to plantation owners as a source of cheap labor. However, there is little solid evidence of the formation of a major state apparatus capable handling and distributing hundreds of annual arrestees. Huggins noted that in 1890 the governor "suggested" that an agricultural correctional camp be established for vagrants. This became a reality when Fernando de Noronha became a state penal colony later in the decade. But, Fernando de Noronha held only a few hundred prisoners [ILLUSTRATION FOR FIGURE TWO OMITTED]. This clearly could not account for most arrestees. The exceptions would be abandoned minors which juizes de orphaos regularly attempted to relocate on farms and households as apprentices and domestics. Huggins cites precedents set under the extraordinary events of the Great Drought (1877-79) when the government established work camps for drought refugees. Pernambuco's officials also attempted to contract refugee workers to planters and railroad contractors. But, Gerald Greenfield shows that Pernambucan officials used relief funds from the imperial government to subsidized this labor, making it more attractive to employers. Pernambucan officials wanted to remove refugees from Recife. In short, it was the desire to push these people out of the city rather than a demand for their labor in the countryside that dominated this process. If such a large-scale labor-intensive bureaucracy existed in the 1890s, it seems unlikely that it would escape the notice of foreign and local observers. It seems improbable that Pernambuco would have the money to pay for such an expensive enterprise considering that the federalist republican policies greatly reduced the revenues of northeastern states. The vast majority of these offenders were probably intimidated and then turned back out into the street. Huggins, Vagrancy, 105-8; Gerald M. Greenfield, "O comportamento dos migrantes e as atitudes das elites durante a Grande Seca do nordeste: 1877-1879," Cadernos de Estudos Socials 5:2, (1989): 21933; Joan E. Meznar, "Orphans and the Transition from Slave to Free Labor in Northeast Brazil: The Case of Campina Grande, 1850-1888," Journal of Social History (Spring 1994): 499-503.

21. Lemos Brito, Os sistemas, II:70-85; For a survey of the procas' racial and social composition see Peter M. Beattie, "Transforming Enlisted Army Service in Brazil 1864-1940: Penal Servitude Versus Conscription and Changing Conceptions of Honor, Race and Nation," (Ph.D. diss., University of Miami at Coral Gables, 1994), chapter 8.

22. Letter from Police Delegate Manoel Honrato de Barros to Police Chief Antonio Francisco Correia d'Araujo, Termo de Barreiros, Dec. 19, 1874, APEP, Livro PC-140, folha 438.

23. See, e.g., Letter from the Police Chief Antonio Francisco Correia d'Araujo to Governor Henrique Pereira de Lucena, Recife, Dec., 16, 1874, APEP, PC-140, folha 363; Ibid., Dec., 17, folha 368; Ibid., Oct. 17, 1874, folha 197.

24. Relatorio cia Reparticao dos Negocios do Ministerio da Guerra apresentados ao Parlamento (hereafter, RRMGP) (Rio de Janeiro: Typ. Universal de Laemert, 1872), 3; This opinion was echoed by other military and civilian authorities. The war minister in 1874 who succeeded in passing limited conscription

legislation argued, "It is said with reason, that manhunts . . . have reduced the army to a penitentiary that should be proscribed as soon as possible." "Corpo legislativo," O Diario do Rio de Janeiro, 8 June 1874, 1. Senator Leitao de Cunha linked the issue of recruitment reform to police reform. The involvement of army and police pracas in crime became an increasing embarrassment. Holloway, "A Healthy Terror," 664-9; Senator Nabuco is cited in O Diario do Rio de Janeiro, June 8, 1874, p. 2.

25. See Hendrik Kraay's discussion of this negotiation in "Soldiers, Officers, and Society: The Army in Bahia, Brazil, 1808-1889" (Ph.D. diss., University of Texas, Austin, 1995), Chap. 8.

26. Joan E. Meznar provides a number of examples from Paraiba of men sentenced to army service when they refused to marry women they had deflowered. Meznar, "The Ranks of the Poor: Military Service and Social Differentiation in Northeast Brazil, 1830-1875," Hispanic American Historical Review 72:3 (August 1992): 342-3; Beattie, "The House," 441-6.

27. During the same period, the provincial police sent 28 prisoners to serve in the navy and returned 82 naval deserters. Fifty two minors were remitted to the Naval Apprentice School. Relatorio do Presidente Dr. Luiz Hollanda Cavalcanti de Albuquerque, (Niteroi: October 22, 1876), chart #7; Correspondence from police in Rio de Janeiro reveals that of 114 recruits only four were volunteers. See, e.g., Mappa dos individuos recrutados na Provincia do Rio de Janeiro e remettidos ao Quartel Geral do Exercito no mez de Janeiro, Feb. 1, 1860, APERJ, Fundo PP, Col. 8, maco 7, no folha nos.

28. Relatorio do Ministro e Secretaria de Justica Francisco de Paula de Negreiros Sagao Labato, (Rio, 1871), 36; In 1876, the Xadres de Policia remitted 188 prisoners to the Ajudante General da Armada and 366 to the Ajudante General do Exercito. Relatorio do Ministro e Secretaria de Justica Conselheiro Diogo Velho Cavalcante de Albuquerque (Rio, 1877), 266; And in 1881, seventy five prisoners were turned over to the navy and forty seven to the army; Relatorio do Ministro e Secretaria de Justica Conselheiro Manoel Pinto de Souza Dantas (Rio, 1881), 3.

29. On recruits escorted along with convicts see, e.g., Carta do Cabo do 2nd Infantry Batallion Antonio Schalopprink to the Chief of Police Antonio Francisco d'Araujo, Recife, October 1, 1874, APEP Livro PC-140, folha 6-7; Police authorities often depended on army troops to escort civilian convicts to prisons in provincial capitals, Letter from the Police Chief Antonio Francisco Correia d'Araujo to Governor Henrique Pereira de Lucena, Recife, Nov. 5, 1874, livro PC-140, folha 197.

30. Letter from the Chief of Police Antonio Francisco d'Araujo to Governor Henrique Pereira da Lucena, Recife, Oct. 7, 1874, APEP, Livro PC-140, folha 68.

31. Relatorio do Presidente Henrique Pereira de Lucena (Recife, 1875), 35-6.

32. Ibid.; Relatorio do Presidente Joao Pedro Carvalho de Moraes (Recife, 1876), 27.

33. Beattie, "Transforming," chapter 6.

34. The social boundary marked by army service is emphasized in Antonio Edmilson Martins Rodrigues, et. al., A guarda nacional no Rio de Janeiro, 1831-1918 (Rio, 1981), 14-5; Also Meznar, "The Ranks of the Poor," 337-40.

35. Similar contradictions in the formation of Mexico's Rurales are skillfully examined in Vanderwood's Disorder and Progress.

36. Relatorio da Reparticao dos Negocios do Ministerio da Guerra apresentados ao Congresso (hereafter RRMGC) (Rio, 1913), 6.

37. On impressment levels see Beattie, "Transforming," chap. 6.

38. Dermeval Peixoto, Memorias de um velho soldado: nomes, coisas e fatos de meio seculo atras (Rio, 1960), 72-3.

39. Leandro Gomes de Barros, "O tempo de hoje. O sorteio militar," pamphlet (Guariba, Paraiba: P. Baptista, 1918 [on the cover of the pamphlet it states "first published in 1906"]), 12-5, in BNMR.

40. Beattie, "Transforming," chapters 1, 6.

41. RRMGP (1914), 3-4.

42. RRMGP (1871), 3.

43. Rodrigues et. al., A guarda nacional, 301-12.

44. Letter from Police Delegate Francisco Joaquim Cavalcante to Galvao to Police Chief Antonio Francisco Correia d'Araujo, Igaurassu, Dec., 9, 1874, Livro PC-140, in APEP, folha 336; Also a similar letter indicates that National Guard troops were not as dependable, Letter from Police Delegate Jose Alves Marinho Falcao to Police Chief, Brejo, Dec. 14, 1874, Ibid., folha 382; Letter from the District Attorney Graciliano Auguost Cesar Mandelo to the Police Chief, Villa Bella, 28 Out., 1874, ibid., folha 216.

45. Of these only one Northeastern province, Maranhao, had a battalion with less troops than called for by government budgets. All other Northeastern provinces had a troop surplus. War Ministry Correspondence with the President of Maranhao, Rio, July 15, 1885, ANR, maco IG127, no folha number.

46. Governor of Maranhao's Correspondence with the War Ministry, Sao Luis, January 22, 1875, ANR, maco I[G.sup.1]26, no folha numbers.

47. "Relatorio do Chefe da Policia," in Relatorio do Presidente da Provincia apresentado a assemblea cearense (Fortaleza, 1876), 6.

48. On the low esteem of soldiering in 19th century France see, e.g., Weber, Peasants, 294-5; Alan Forrest, Conscripts and Deserters: The Army and French Society During the Revolution and Empire (New York, 1989); For Prussia, Ute Frevert, "Das jakobinische Modell: Allgemeine Wehrphlict und Nationsbildung in Preussen-Deutshland," in Militar und Gesellschaft im 19. und 20. Jahrhundert, ed. Ute Frevert (Klett Cotta, 1997), 17-47.

49. Augusto Fausto de Souza, "Fortificaoes no Brasil: Epoca da respectiva fundacao, motivo determinado dela, sua importancia defensiva, e valor actual," Revista do Instituto Historico Geographico e Ethnographico do Brasil, Vol. XLVIII, Part II (1885): 5-140; Morais, Prisoes, 6, 10; Thomas Holloway noted that when the Regency gained the upper hand over riotous Republican crowds in July 1831 the army and navy forts around Rio were "filled with former rebels, military and civilian." Holloway, Policing Rio de Janeiro, 74; Also see, e.g., Cadeias e Prisoes Civis e Militares, Rio de Janeiro, AGCRJ, Codice (363) 40-2-60. This latter document reported that on the navy's Ilha das Cobras there existed four prisons of which two were for civilian inmates.

50. RRMGP, (1870, 1874, and 1880), annexos; On simplification of notes for Governor's reports see note 16. Pernambuco (March 1, 1881), 26; (February 19, 1900: Police Chief's annexo), 13; and (March 6, 1921), 19; for 1865 see Morals, Prisoes, 35.

51. On the British "transportation system" see Robert Hughes, The Fatal Shore; on the French, Gordon Wright, Between the Guillotine and Liberty: Two Centuries of the Crime Problem in France (New York, 1983), 82-152; On the Portuguese, Timothy Joel Coates, "Exiles and Orphans: Forced and State-Sponsored Colonizers in the Portuguese Empire, 1550-1720" (Ph.D. diss., University of Minnesota, 1993); Also, Gerald Bender, Angola Under the Portuguese: The Myth and the Reality (Berkeley, 1978), chapter 3.

52. Wines, The State of Prisons, 553; also see more detailed surveys of penal island properties in RRMGP in the 1870s; for more detail on production and disciplinary conditions see Dr. Antonio Herculano de Souza Bandeira Filho, "Informacoes sobre o Presidio de Fernando de Noronha: Relatorio apresentado ao governo," (Rio: n.p., 1880); Conselheiro Padua Fluery, "Parecer sobre o presidio de Fernando de Noronha," (Rio, April 30, 1880).

53. See, e.g., Letter from Tenente Coronel Jose Soares [illegible] to Governor Francisco Paula da Silveira Martins, Fernando de Noronha, Dec. 18, 1874, APEP, FN-12, no folha nos.

54. Annaes do Senado Brasileiro, vol. I (sessao May 29, 1874), 131.

55. Fluery, "Parecer," 5.

56. The provincial origins of Fernando de Noronha's prison population in 1876 were: Pernambuco 681; Sao Paulo 84; Ceara 76; Minas 76; Rio Grande do Sul 60; Paraiba 55; Maranhao 54; Rio de Janeiro 52; Bahia 34; Alagoas 22; Rio Grande do Norte 22; Para 22; Piaui 15; Sergipe 4; Espirito Santo 3. Amazonas, Santa Catarina, Goias, Parana, and Mato Grosso were not represented. Relatorio do Presidente Joao Pedro Carvalho de Moraes (Recife, 1876), 29.

57. Justice Ministry pressure did manage to bring about a virtual halt to the remission of additional prisoners, but an 1880 Justice order that provinces (in particular, Pernambuco) remove inmates was apparently ignored. Morais, Prisoes, 35.

58. As late as 1923, a year of relative peace and stability, the naval prison on the Ilha das Cobras held "twenty two sailors, nine navy marines, fifteen convicts from Rio's Casa de Coreccao, ten dishonorable discharged soldiers, and twelve sailors for disciplinary infractions." Brito, Os sistemas, I:306.

59. Morais, Prisoes, 76.

60. War Ministry reports harped on the problem of recruitment shortfalls, stressing repeatedly that traditional recruitment methods were "insufficient" to resolve the army's manpower needs. RRMGP (1870), 25; RRMGP (1872), 2-4; RRMGP (1874), 3-5.

61. Morais, Prisoes, 76. The artillery school's greater emphasis on education made it a narrow avenue for even more dramatic social mobility. The top three artillery school graduates were guaranteed spots in the two-year prep course for the Military Academy. A commander of the artillery apprentices bragged in 1882 that forty army officers were sons of this institution. Admittedly this was forty out of thousands of apprentices, but nonetheless these officers could be pointed to as role models. RRMGP (1882), 23-4.

62. During the same period, the provincial police sent 28 prisoners to serve in the navy and returned 82 naval deserters. Fifty two minors were remitted to the naval apprentice school. Relatorio do Presidente Dr. Luiz Hollanda Cavalcanti de Albuquerque (Niteroi: October 22, 1876), chart #7.

63. On simplification of cites from Governor's reports see note 15. Pernambuco, (March 1, 1871), 13; (March 1, 1876), 31; (March 1, 1881), 112; (March 6, 1886), 10; Rio Grande do Sul, (March 14, 1871), 36; (May 8, 1886), 15; Minas Gerais (1881), 28; (June 15, 1891), 32; Mato Grosso (May 3, 1881: Police Chief's Annexo), 19; Para (February 18, 1876), 27-8; (February 15, 1881), 90; (March 25, 1881), 20; RRMGP (1877), 28-9, 30-1; RRMGP (1880), 15-8; RRMGP (1885), 15-8; RRMGP (1886), 20, 22; RRMGC (1891), 20; RRMGC (1896), 42-8; and RRMGC (1899), 62, 64.

64. During the late Empire, the active duty force of the navy remained around three thousand pracas. The navy apprentice school system undoubtedly provided the majority of its sailors. Because of its need for skilled seamen, the navy depended much more heavily on its apprentice school system as a source of recruits. In 1882, the navy had eighteen apprentice companies around the nation had space for fifteen-hundred boys. Morais, Prisoes, 77. But in 1883, there were only 1050 boys actively enrolled. Relatorio da Repartic ao dos Negocios do Ministerio Marinha ao Parlamento (Rio, 1883), 17.

65. Naval Ministry Correspondence on Enlistment and Reenlistment, Rio de Janeiro, March 6, 1872, ANR, maco XM 268, no folha numbers. The governor of Minas Gerais reported in 1881 that "since the installation of the infantry apprentice school 51 boys were enrolled and of these 21 were sent to the army because they had completed 14 years." Relatorio do Senador Joao Florentino Meira de Vasconcellos Netto (Ouro Preto: May 4, 1881), 28; On police repression and capoeiras in Rio, see Holloway, Police; Holloway, "A Healthy Terror;" Soares, A negregada instituicao.

66. Naval Ministry Correspondence on Recruitment, Rio de Janeiro, September 21, 1876, ANR, maco XM 874, no folha numbers.

67. See copies of police report in Naval Ministry Correspondence on Recruitment, Rio de Janeiro, June 13, 1876, ANR, maco XM 874, no folha numbers.

68. One police report noted that a group of "uniformed navy arsenal apprentices were armed and practicing capoeira in front of a tavern near the arsenal, and when a policemen saw them and went to the station to call more men [to the scene], then [the apprentices] ran into the arsenal." Navy Ministry Correspondence on Recruitment, Rio de Janeiro, June 9, 1876, ANR, maco XM 874, no folha numbers.

69. Navy Ministry Correspondence on Recruitment, Rio de Janeiro, June 9, 1876, ANR, maco XM 874, no folha numbers. Unfortunately, army correspondence from the Corte does not include confidential letters on this initiative.

70. Thomas Holloway noted that police authorities complained that their was little they could do about capoeiras who were veterans, national guardsmen, military arsenal employees, or army enlisted men. Holloway, " 'A Healthy Terror,' " 668; Hendrik Kraay, "The Protection of the Uniform: The Brazilian Army and Runaway Slaves, 1899-1888," Journal of Social History 29:3 (March 1996): 637-57; Also Kraay, "Slavery, Citizenship and Military Service in Brazil's Mobilization for the Paraguayan War," Slavery and Abolition 18:3 (Dec. 1997): 228-56.

71. Scott, Domination and the Arts of Resistance.

72. RRMGC (1892), 15-17.

73. Relatorio apresentado ao Presidente da Republica dos Estados Unidos do Brasil pelo Ministro de Estado dos Negocios da Marinha Vice Almirante Antonio Coutinho Gomes Pereira (Rio, 1919).

74. RRMGC (1914), 3-4.

75. In their professional journals, the Prussophile junior officers, known as the Young Turks, also insisted that the "action of the army should not have the character of police operations." This was reported in the context of the Contestado campaign where the army was suppressing Brazilian millenarian insurgents. A Defesa Nacional (Rio: October 10, 1914): 2; Leila Maria Correa Capella, "As malhas de aco no tecido social: a revista," A Defesa Nacional e o servico militar obrigatorio (Dissertacao de Mestrado, Universidade Federal Fluminense, 1985).

76. RRMGC (1913), 9-11.

77. Edgar Carone, A Republica Velha: Evolucao politica (1889-1930) (4th ed., Sao Paulo: DIFEL, 1983), II:242.

78. For example see Todd Diacon, "Searching for a Lost Army: Recovering the History of the Federal Army's Pursuit of the Prestes Column in Brazil, 1924-1927," Americas (forthcoming).

79. Beattie, "The House, 460-71.

80. "Os voluntarios," Correio da Manha, Rio de Janeiro, October 25, 1916, p. 1.

81. See the state by state comparison of army and police forces and analysis of its significance in Jose Murilho de Carvalho, "As Forcas armadas na Primeira Republica: o poder desestabilizador," Cadernos do Departamento de Ciencia Politica, No. 1 (Belo Horizonte, 1974), 181-183; June Hahner demonstrated how Sao Paulo built up its state police forces to guarantee greater autonomy from the central government in Civil-Military Relations in Brazil, 1889-1898 (New York, 1969), chapter VII.

82. Lieutenant Castro Ayres, "O Orcamento da Guerra," A Defesa Nacional, n. 5, ano I (Rio: February 10, 1914):138.

83. Miguel Calmon, Discursos pronunciados nos sessoes do Senado Federal de 29 a 30 de Dezembro, 1927 (Rio, 1928), 13.

84. Jose Murilho de Carvalho observed that the police used the Vaccine Revolt as an excuse to clean up the streets. He argues that they tended to deport known criminals with little regard for the extent of their involvement in the revolt, Os bestializados: Rio de Janeiro e a republica que nao foi (Sao Paulo, 1991), 113-26; On the Chibata Revolt see Edgar Morel, A Revolta da Chibata, 4th ed. (Rio, 1986), 161-77; On desterros after the Canudos Rebellion see Calmon, Discursos, 19; For a damning report on political prisons during the Tenente Revolts see, Everardo Dias, Bastilhas Modernas 1924-26 (Sao Paulo, n.d).

85. Marcus Joaquim Maciel de Carvalho, "Hegemony and Rebellion in Pernambuco, 1821-1835" (Ph.D. diss., University of Illinois, Urbana), 78; Izabel Andrade Marson noted the reluctance of armed factions of the Praieira revolt to give up the fight, knowing that army recruitment awaited them. Marson, O imperio do progresso: A revolucao praieira (1842-1855) (Sao Paulo, 1987), 108-19; Hendrik Kraay cautions, however, that many such rebel recruits sent out in punitive impressment sweeps after the 1837 Sabinada Revolt in Salvador, Bahia were discharged soon after their induction. But this action may have been due to the Liberal reforms that slashed the army size in the 1830s cutting demand for troops. Kraay, "Soldiers, Officers, and Society," 231-2.

86. Brito, Os sistems, I:292.

87. Dias, Bastilhas.

88. Padre Rogerio Alcinio, Clevelandia do Nortet (Rio, 1971), 94-8.

89. See novelist Graciliano Ramos harrowing personal account of his experience as a political prisoner, Memorias do Carcere, 2 vols., 27 ed., (Sao Paulo, 1994).

90. Michel Foucault, Discipline and Punish: The Birth of the Prison, trans. Alan Sheridan (New York, 1979); Max Weber, Economy and Society, Ed. Guenther Roth and Claus Wittich (New York, 1958); Weber, The Theory of Social and Economic Organization (New York, 1947); Benedict Anderson, Imagined Communities: Reflections on the Origins and Spread of Nationalism (New York, 1991); E. P. Thompson, "Work, Time-Discipline, and Industrial Capitalism," Past and Present 38 (Dec. 1967): 58-97.

91. Michael Howard, The Causes of War (London, 1983), 182; Weber, Peasants into Frenchmen, Chap. 17; Charles Tilly, "The Emergence of Citizenship in France and Elsewhere," 223-36, in Citizenship, Identity and Social History, Ed. Charles Tilly (Cambridge, 1996); Mark Von Hagen, Soldiers in the Proletarian Dictatorship: The Red Army and the Soviet Socialist State, 1917-1930 (Ithaca NY, 1990); Frederick S. Harrod, Manning the New Navy: The Development of a Modern Naval Enlisted Force 1889-1940 (Westport, CT, 1978); Frevert, "Das jakobinische."

92. On Brazil's limited public education system and its low rates of retention even in the immediate Post world War II period see, e.g., Manfredo Berger, Educacao e dependencia (Sao Paulo, 1976). On Argentina's relatively more precocious development of primary and secondary public education see, e.g., Mark D. Szuchman, Order, Family, and Community in Buenos Aires 1810-1860 (Stanford, 1988), Chap. 5. In a related vein, Ricardo D. Salvatorre argues that ideas about race impeded the implementation of modern penitentiary methods in the Northeast of Brazil, distinguishing it from reforms undertaken earlier in Buenos Aires. Salvatorre, "Penitentiaries, Visions of Class, and Export Economies."

93. Coronel Dr. Arthur Lobo da Silva, "Accao eugenica dos exercitos," Memoria Apresentado ao Primeiro Congresso Brasileiro da Eugenia, reprinted in A Folha Medica (Rio: August 25, 1929), 2-14.

94. Rebecca J. Scott's work promises to reveal much in this regard in a comparative perspective; see "Defining the Boundaries of Freedom in the World of Cane: Cuba, Brazil, and Louisiana after Emancipation," American Historical Review (Feb. 1994): 70-102; Also Hebe Maria Mattos de Castro, Dos Cores do silencio: os significados da liberdade no sadeste escravista, Brasil seculo XIX (Rio, 1995).

95. On the penitentiary system in Georgia in comparative perspective with other southern states, Ayers, Vengeance.

96. On the importance of conscription in developing European ideas of citizenship see Tilly, "The Emergence of Citizenship," 223-34.