Cangaço and Cangaceiros: Rural Banditry in the Brazilian Northeast
Cangaço and Cangaceiros: Rural Banditry in the Brazilian Northeast
- Luiz Bernardo PericásLuiz Bernardo PericásDepartment of History, Universidade de São Paulo
Summary
The cangaço was a social phenomenon related to rural banditry in the backlands of the Brazilian Northeast (an area referred to as the sertão). Beginning in the nineteenth century, the cangaço reached its peak with the actions of Virgulino Ferreira, popularly known as Lampião, the most important and emblematic leader of these outlaws, during the 1920s and 1930s. Its demise came with the start of the dictatorial Estado Novo regime in 1937. The cangaço received widespread coverage in the local press and was amply depicted in the visual arts, literature, and cinema, enduring as one of the most distinctive and controversial subjects in Brazilian cultural history.
Keywords
Subjects
- History of Brazil
- Social History
Origin and Definition of the Term “Cangaço”
The term “cangaço” appears to have been used for the first time in the 1830s and 1840s, by the population of the northeastern backlands of Brazil, according to some scholars.1 Others, however, argue that the word had been employed regularly since the eighteenth century, although these early cangaceiros, as they were called, were hired by colonizers to hunt down Indians.2 This second version is less likely. In any case, the oldest dictionary of the Portuguese language, Vocabulário portuguez e latino by the priest Raphael Bluteau, published in Coimbra between 1712 and 1728, does not include the word “cangaço.”3 Luís da Câmara Cascudo asserts that the word appeared in dictionaries created by Domingos Vieira for the first time around 1872 and meant a “gathering of minor and varied objects, utensils of modest families, items of the poor or enslaved.” The definition that most closely approximates what is meant by “cangaço” today, however, is an entry in Dicionário de vocábulos brasileiros by Henrique de Beaurepaire Rohan (published in Rio de Janeiro by the National Press in 1889), which refers to the “set of arms usually wielded by ruffians [valentões].” This is probably the earliest formal definition of the term specifically related to rural banditry in the Northeast.4 Juvenal Galeno, however, had already employed the term in his 1871 book Cenas populares; Franklin Távora also used it in another literary work, O cabeleira, published in 1876; and a little later, Irenêo Joffily would again mention the cangaço in his Notas sobre a Parahyba, published in 1892.5 From the beginning of the republican period in 1889, the designations “cangaço” and “cangaceiro” were already consistently appearing in official records, such as police investigative reports, judicial sentences, and letters between law enforcement officials and judges describing acts of rural banditry.
One explanation for the origin of the term can be found in the word canga, from which derived the name for both a torture device (a version of which was used on slaves) and the cangalha, a mechanism used on the back of oxen and other draft animals.6 The former resembled the tronco (trunk) which, according to Rodrigues de Carvalho, could be described as, “A device affixed with wooden support beams, with a sucupira (a heavy dark wood native to the backlands of the Northeast) board in the shape of a half-moon, through which the prisoner placed one leg, suspended, with the other leg set at a height established by the user.”7
The Austrian doctor Johann Baptist Emanuel Pohl, who traveled throughout the country in the 1820s, observed that the canga device:
consisted of two planks of wood, each one two fathoms [4 yards] long, a foot and a half tall, and three inches thick, perfectly fitted on top of each other. Three openings are cut into the wood and used to secure the offender's feet and neck. Consequently, this tronco is a modification of the cavalete, once commonly used in Europe, and which here, as I later learned, is often employed out of simple malice on the part of the militia and entirely at their discretion.8
For his part, Jean-Baptiste Debret noted that:
It is common to find in the home of the Brazilian farmer a tronco, an old instrument of punishment, consisting of two pieces of wood six to seven feet in length, fastened on one end by iron hinges and fitted on the other with a padlock the key for which remains in possession of the maker. The purpose of this device is to overlap the two round holes, through which wrists or legs are passed, and sometimes the neck of the tortured. The instrument is usually placed in a closed shed or an attic.9
The tronco, therefore, in the words of José Alípio Goulart, would be a kind of “cousin-brother” of another device known as the canga chinesa. It was used mainly on African slaves, but also, at times, on free workers. This explains the popular association of the apparatus with subjugation and oppression. Popular expressions like “no canga is placed on this neck” [nesse pescoço não se bota canga] are thus used to suggest freedom, haughtiness, independence, and dignity.
The cangalha, or ox-canga, is a mechanism composed of two wooden forks, set about three palm-widths apart, fastened by lateral beams—called prendas—while leather strips (repuxos) secured on the inside hold the straw mats placed on the backs of the pack animals.10 Also included are the surcingle, or cilha (a buckled leather strap passed around the animal’s body to fasten on a blanket or a mat), rabichola, and breastplate—implements to ensure device’s balance and stability. The load to be transported is placed on the heads of the wooden forks.11
The oxcart has played an important role in the rural landscape of the Brazilian Northeast over several centuries. Indeed, a scholar of social relations in the backlands has argued that the oxcart was the vehicle for monoculture and economic individualism, without which the rural landlord would not have existed.12 This mode of transporting people and cargo was used in mines, sugar mills, and other forms of commerce. It was fundamental to the penetration and conquest of the vast northern territories, especially in those places where there were no navigable rivers. Its social, domestic, and economic impact is undeniable.
The oxcart also aided the construction of a particular image of the backlands, giving it a distinctive identity. It was so central to rural life that it inspired several place names, like Carro Quebrado (broken car) and Passagem do Carro (car passage), in Rio Grande do Norte, for example. Among the various practical components of the oxcart, the canga stands out for its size, shape, and function. It therefore is not surprising that such a prominent feature of the central economic engine of life in the backlands should become symbolic in the sertanejo (“of the backlands”) popular imagination. It makes sense that this apparatus would be used as a reference to illustrate or designate some region, such as Baixa da Cangalha and the Cangalha mountains (Bahia), as well as perhaps some modality or social phenomenon like the cangaço.
There are two more common allusions related specifically to the canga. The first compares the subordination or dependence of the cangaceiros to the so-called coronéis (literally, colonels, or powerful hinterland power brokers) who imposed their will on their underlings with the harsh rigidity of a torture or fastening device (like the cangalha on the back of a pack animal). The second associates the canga to the paraphernalia that cangaceiros hung on their bodies. Some, however, claim that the connection of the term cangaço to the oxcart stems from the proximity of the animals, joined to one another by the wooden device. The term, in this telling, denotes unity of purpose.13
The suggestion that the word is associated in some way with a version of the popular saying “to be under God” [estar debaixo de Deus]—that is to say, “to be under the cangaço” [estar debaixo do cangaço], which suggests a positive relationship with the bandits defined not by subjugation but by protection—seems less probable. Nevertheless, the same saying, used to express the uniform, customs, and weapons characteristic of the bandits, is a much more likely possibility. In that case, cangaço refers to the lifestyle of the rural rogues.
Still others, like the folklorists Pedro Batista and Batista Caetano, argue that the origin of the term comes, in fact, from the indigenous [abanheenga] word cang [kang], another spelling for the Tupi word acanga or akanga, which some translate as head or skull, while others define it as the remains or skeleton of domestic or wild animals.14 According to that definition, the term evokes an arrangement of bones with the weapons and equipment associated with the bandits.15 Finally, there are a few scholars who claim that the term cangaço is actually of African origin, coming from the Kikongo or the Kimbundu words konganso or nkangunsu, meaning band or group of bandits.
The Independent Cangaço
The notion of an independent cangaço (especially during the period of Lampião) refers to autonomous bandits with no direct links to specific coronéis. These gangs carried a considerable amount of equipment, armament, and ammunition, and raided the rugged outskirts of the Northeast, crossing into several states. They often formed with the pretext of exacting revenge (usually interfamilial), or as refuges for outcasts seeking protection from the police or other enemies.16 They would then come to rely on rural banditry as a means of survival, that is, obtaining material gains by robbing, looting, and extorting, although, as we will see later, their actions, depending on the individuals involved, presented differing variables and motivations, which should be analyzed on a case by case basis. In general, however, the concept of the independent cangaço relates both to the equipment of the bandit and to his way of life, mainly in the hinterlands of the Northeast.
The period from about 1890 to 1940 was the most emblematic for cangaceiro activity. This was the era in which the iconic leaders of the independent cangaço, Antônio Silvino, Sinhô Pereira, Ângelo Roque, Jararaca, Lampião, and Corisco, operated; this was also a time of increasing independence and mobility in the region. This was the interregnum between the establishment of the First Republic in 1889 and the beginning of the Estado Novo in 1937 (in the previous decades, specifically in the 1870s, the most important cangaceiro was Jesuíno Brilhante, in Rio Grande do Norte).
There are several examples of banditry in Brazil before the golden age of the cangaço. In the Northeast, regional outlaws operated in both the Zona da Mata and the Bahian Recôncavo, as well as in the backlands of Bahia, Pernambuco, and other provinces. These rogues were not called cangaceiros, however, nor did they display several of the traits associated with the phenomenon at its height, traits that will be analyzed later in this article. There are well-known cases of Dutch soldiers who became brigands, and of bandeirantes from São Paulo who abandoned their expeditionary missions in the northeastern backlands in the second half of the seventeenth century. These then formed bands of marauders who “knew neither King nor Justice” and committed various crimes. A notable case was the invasion of Porto Seguro by forty raiders from São Paulo, who killed and raped indiscriminately, and plundered the entire city.17 In the eighteenth century, the activities of José Gomes, a native of Pernambuco better known as Cabeleira, stood out. And later, in the first half of the nineteenth century, bandits like Lucas Evangelista dos Santos, the infamous Lucas da Feira (who operated around Feira de Santana in his native Bahia), might be seen as precursors of what later came to be referred to as the cangaço.18 Although these cases exhibit some of the particularities of the criminal phenomenon that followed, they lacked the culture that organically developed around this outlaw lifestyle. There was still no set image of the cangaceiro as an emblematic figure and representative of the social fabric of the northeastern backlands. The scale of criminality and its regional influence remained relatively limited.
During the aforementioned period, autonomous groups were the norm, informing the image of the cangaço still popularly held today. The cangaço, in other words, was already an identifiable phenomenon in the middle of the nineteenth century, before reaching widespread recognition in the early 1900s, with infamous figures like Antônio Silvino, Sinhô Pereira, and Lampião (the latter at one point commanded around 150 men). During the Lampião era, Brazil experienced the influential Modern Art Week in São Paulo; the founding of the Brazilian Communist Party; a radical movement for professional and national reforms led by junior army officers known as tenentismo; the Prestes Column, an outgrowth of the tenentista movement; the 1930 Revolution that brought Getúlio Vargas to power; and the beginning of the Estado Novo dictatorship. This was a time of great political change, social upheaval, and economic modernization across the country.
While there were a number of independent bandit groups active since at least the 1850s (and even earlier), it was not until the 1890s that the cangaço went from endemic to epidemic. Between 1850 and 1889, according to an analysis of reports from northeastern provincial presidents (as governors were then called) conducted by Hamilton de Mattos Monteiro, there were forty-seven notable bands of outlaws active throughout the region.19 Among the most important were the Viriatos, the Meirelles, the Quirinos, and the Calangros. But between 1919 and 1927 alone (a period of only eight years), there were roughly fifty-four groups active in the hinterland and outskirts of the Northeast.20 The number of groups and individuals involved had grown considerably, as had the size of the territory in which they operated.
The ambitions of cangaceiros in the twentieth century shifted as well. The goal became to lead a lifestyle around, and live off, the cangaço, which became increasingly professionalized. During the Lampião era, the bandits also became much more brutal than their predecessors had been. Torture and deadly acts of cruelty were more commonplace and widespread.
Between 1890 and 1940, this mode of rural banditry so particular to the backlands of Northeastern Brazil reached its peak and began to decline, although there are somewhat similar cases in other parts of the world.21 After all, violence in the countryside is not exclusive to any one nation.
On the surface, in other words, the cangaço resembles other instances of rural banditry around the world. For example, revenge as a motivator, and the idea of criminal gangs as protection against the law, can be found in different contexts across time and geography. But these similarities are superficial, as there are structural elements that give the phenomenon of the cangaço very particular cultural characteristics, including their clothing, language, guerrilla tactics, and relationships with women, sertanejos (residents of the backlands), farmers, and the police. Even with possible similarities to other cases, these specificities can only be fully understood within the historical evolutionary process of the Brazilian Northeast, and the backlands in particular. The cangaço, then, was more than just a manifestation of marginality; over time, it imbued itself with a diversity of distinctive cultural elements that lent it a singular, socially constructed aesthetic.
In order to understand the complexity of the social dynamics of the Northeastern hinterland, the emergence and the end of the independent cangaço, and the cangaço’s impact on the local populations, it is necessary to address the interplay between the region’s apparent immutability and the endurance of cultural remnants, along with the ruptures and structural changes that took place at a crucial moment. The distinctive features of the local way of life, marked by mysticism, fanaticism, superstition, religiosity, coronelismo (rule of the coronels), family disputes, political structures, and the culture of armed mercenaries, as well as lax attitudes toward criminal behavior within the so-called “Leather Civilization,” are key.22 So too is careful analysis of the emergence and expansion of railroads, roads, labor movements in major cities (contrasted against the supposed isolation of populations in remote areas), the state and federal legal superstructure, the inflow of capital and investments in the different northeastern states, the federal and state governments’ policies regarding rural banditry, the labor market, drought cycles, the physical environment, migration, industrialization, and national economic change and modernization, among other factors. Simplistic readings fail to satisfactorily explain the cangaço. The roots of modern northeastern banditry are deep and complex.
Nevertheless, many authors have tried to understand life in the backlands by overemphasizing the physical, emotional, or psychological characteristics of local residents or the influence of climate and geography. This tendency produces incomplete interpretations that reinforce clichés about a society which, like any other, has always produced a wide range of cultural experiences, patterns, and values. The massive territory of the Brazilian Northeast, in other words, cannot be seen as a single, homogenized landscape. Rather, it should be interpreted as a dynamic environment with a series of characteristics ranging from a latent cultural conservatism to renewals, adaptations, and incorporations of material conditions and forms of social coexistence, which are sometimes underappreciated by scholars.
In a single group of cangaceiros, for example, it was possible to find individuals from various states of the federation with different life experiences, distinctive ways of speaking and acting, idiosyncratic religious practices, and unique personal relationships, all of which would not necessarily align with those of their partners-in-crime. Even if, through coexistence and necessity, some personal psychological traits were diluted by or blended in with those of the group, other characteristics would have remained unspoken and closely held. The ranks of Northeastern bandits during the modern period were so diverse as to include newly freed slaves (or their children); farmers; tradesmen; mule drivers; fugitives; deserters; blacks; whites; mixed-race people of indigenous, African, and European heritage; and natives of Paraíba, Bahia, Rio Grande do Norte, Alagoas, Sergipe, Ceará, and Pernambuco. There were also some who had spent time São Paulo, Rio Grande do Sul, Rio de Janeiro, as well as other states of the Northeast. This illustrates the heterogeneity of the cangaço phenomenon.
These gangs were often formed by a father and his sons, a group of siblings, cousins, relatives, or members of extended social networks, as well as any others with direct or indirect kinship ties. They would then gradually and sporadically incorporate other local residents.23 Beginning in the 1930s, when Maria Bonita joined Lampião’s gang, women also started organically joining such groups.
The Class Backgrounds and Motivations of Cangaceiro Leaders
Contrary to popular assumptions, many cangaço leaders came from relatively wealthy and traditional families. That is to say, to a large extent, they were part of what could be considered the local elite: landowners, descendants of members of the National Guard and slaveholders (in the case of the nineteenth century, before the Republic), or their children, could be found among leaders of the cangaço. At various points, these individuals enjoyed good relations with local power brokers, the police, and conservative politicians, from whom they received protection, arms, and ammunition. Their minions (jointly referred to as arraia miúda, or the rank and file of the cangaço) considered these leaders, who often shared no class identity with the poorer people of the backlands, to be their bosses. Cangaço leaders generally preferred the company of coronéis over that of humble farmers or peasants. While the vast majority of the population in the Northeast was illiterate, all the leaders of the cangaço knew how to read and write, even if crudely, which further set them apart from most local residents.
Such leaders were relatively few, but they were the ones who set the tone for the gangs, leaving their mark and imposing their will. They were the commanders who ultimately decided on the course of action for individual groups.
Many of these individuals managed to amass considerable gains from extortion, kidnapping, and robbery (targeting the rich and poor alike). These stolen goods were not used to improve the life of the dispossessed and or the needy population. Rather, the bandits kept everything for themselves, thought they occasionally distributed negligible amounts as handouts in village squares. Reports of families, and even entire populations, fleeing a particular village, terrified upon learning of the imminent arrival of cangaceiros, are plentiful. Cruelty and torture were commonplace, with victims ranging from workers on anti-drought infrastructure projects, to women branded with the initials of one bandit or another (José Baiano was most infamous in that regard). There was thus widespread fear of encountering these rogues throughout the backlands, although the same was often true regarding the government forces intended to combat the cangaço, suggesting that the forces of both law and disorder embodied the most violent and aggressive aspects of the hinterlands at that time. In the end, it is clear that the cangaceiros were not rebelling against the traditional order and the rural establishment.
The Area of the Cangaço
The northeastern backlands are a varied landscape, both physically (topography, rainfall, soil types, vegetation, geology, and fauna) and in terms of social composition. An important feature of life in the region is the vast area known as the Drought Polygon, composed of wide arid areas distinct from one another. It includes the sertão, Seridó, and the agreste, with mountains and areas of desert vegetation called caatinga.24 The terrain is mostly dry and rocky, and dotted with short, spiky vegetation.
The caatinga, which extends through the subregions of the Seridó, sertão, Cariris Velhos, carrasco, agreste, Curimataú, and cerrado, runs from the north of Minas Gerais to Piauí, across approximately 800 thousand kilometers of the Northeastern territory. Excluding Bahia, the caatinga covers an area of 27,497,171 hectares and, including that state, it reaches 60,246,021 hectares. It is the most common type of landscape vegetation in the arid Northeast.
With the expansion of agricultural frontiers and cattle ranching, human settlement significantly altered the natural landscape of the agreste. In the first half of the twentieth century, the golden age of the cangaço, the region was characterized by small vegetation, mostly velames and cacti, the most common varieties being the mandacarus, xiquexiques, alastrados, palmatórias, and cardeiros. Also interspersed in the area were some species native to the Atlantic Forest and caatinga. Once covered in jungle, this region was completely eroded, transforming into something very similar to that of the sertão itself, that is, inhabited by only those plants and shrubs able to survive the invasive new agricultural processes that destroyed the original ecosystem. The primitive formations of dense caatingas with taller trees (such as braunas, Lithraea molleoides [aroeiras], and Anadenanthera colubrina [angicos]) were greatly reduced.
The northeastern hinterland, therefore, was not at all homogenous. There were salient local differences in terms of dress, habits, cuisine, beliefs, folklore, and cultural sophistication. In some areas, like Conquista, cattle drivers did not wear leather clothing, whereas in others, the thickets and thorny vegetation made full leather suits (almost from head to toe) a requirement.
The clothes and riding gear used also displayed regional variations, from the cut, decorations, and even the design of the heads of saddles and hats, to the footwear, which could be espadrilles or so-called Russian boots. The shape of leather hats in Camisão, for example, was quite different than that of those produced and used in Riachão or Alto São Francisco. Indeed, the way one bent the brim of one’s hat indicated, at that time, where one came from.25 On the arid coast of Rio Grande do Norte, for example, cowboys wore hats with shorter brims, claiming that they were more practical, given the closer vegetation and prolonged rainy seasons. In the upper part of the state and in Seridó, on the other hand, where soil and climate conditions were different, long brims, four to five fingers long, were imperative, considering that they offered greater eye protection in the intense sunlight.26 The hats of the tucanistas in the state of Bahia did not resemble those of other cowboys. And although the leather hat was the most common type in the region, simpler hats made of straw were also widely used, especially by peasants and muleteers.
Cangaceiros, coronéis, preachers, priests, mystics, cowboys, gunslingers, and mercenaries are the most common archetypes of the northeastern backlands, but there were other sorts of people in the region as well. Muleteers, peddlers, and travelers were scattered across the vast expanse of territory, imbuing the sertão, with a sense of transience, both of information and products, between the coast and interior. For that reason, their influence on life in the backlands was crucial.
It was in this vast expanse of land and settlement, spanning several northeastern states, with all its geographical differences, social specificities, and ongoing physical, human, and cultural changes, that modern rural bandits roamed. In the words of Gustavo Barroso, a writer from Ceará:
The large area between the São Francisco River and the Cariri Valley, stretching from the Quicuncá to Martins mountain ranges, from there to the fault of Borborema, the foothills of the Baixa Verde and the Dois Irmãos, is the habitat of banditry. That is where they meet and traverse the borders of seven states, with shards of territory jammed into one another like wedges. The environment, the complicity of residents, and the ease of fleeing from one state to the other offers safe haven to all criminals.27
The End of the Cangaço
Several factors contributed to the end of the cangaço. From 1935 (a year of intense action by the National Liberating Alliance and Communist insurgents), several cangaceiro leaders were assassinated, among them Manuel Torquato in Rio Grande do Norte; the Ingrácia brothers, Medalha, Suspeita, Limoeiro, and Fortaleza, in Alagoas; José Baiano, Mariano, Pai Velho, and Zepelim in Sergipe; and Paizinho Baio in Pernambuco. After the installation of the Estado Novo dictatorship in 1937, government repression increased considerably: the state government of Alagoas created the Second Military Police Regiment, whose headquarters were in Santana do Ipanema; Bahia, for its part, already had a contingent of 1,200 men tasked with chasing Lampião in the backlands (Jeremoabo was the center of operations); finally, in Pernambuco, the interventor (the chief executive of the state appointed by Getúlio Vargas), Agamenon Magalhães, sought to disarm the population of the hinterland by increasing the number of police officers and modernizing their equipment.
In July 1938, Lampião was killed in Grota do Angico, Sergipe (on the border with Alagoas), along with his wife Maria Bonita and nine other companions. They were beheaded and their heads were photographed; those photographs were then showed to the public in villages across the region. It was notable that Lampião had come to be seen as a sort of ally by the administration of former president Artur Bernardes (1922–1926) after he agreed to participate in the Patriotic Battalion of Juazeiro to fight against the Prestes Column, although that did not end up happening. For their part, the National Liberating Alliance and the communists sought to exploit the image of the cangaceiros as social rebels, so as to co-opt them in the struggle against the Vargas regime. For one reason or another, cangaceiros were not well regarded by the government. In the mid-1930s, a film by Benjamin Abrahão depicted Lampião and his comrades favorably, which greatly displeased the Vargas regime, who saw in the work a provocation to the authorities and an exaltation of backland bandits who purportedly represented the nation’s most archaic and backward elements (the film was banned for decades, with the few minutes of images that could be recovered included in the documentary Memória do Cangaço, directed by Paulo Gil Soares in the mid-1960s). With the dawn of the Estado Novo, many cangaceiro leaders turned themselves in to the authorities. Government pressure was mounting.
The use of modern weaponry such as Hotchkiss and Thompson machine guns; the overwhelming manpower advantage (with a large increase in police forces); a spike in federal funding to combat rural banditry; the political will of the Vargas government to end the phenomenon once and for all; the weakened position of coronéis (many of whom had lost their stature and stopped supporting the gangs); a crackdown on those who aided and sheltered criminals; and the fact that many corrupt police officers were no longer able to supply arms and ammunition to the outlaws all contributed to the decimation of the cangaço as a criminal enterprise and as part of the region’s social fabric. The only important leader who remained active after the death of Lampião was Corisco, who died in 1940. The cangaço, for all intents and purposes, died with him.
Discussion of the Literature
A great many authors have written about the cangaço. Much of the work on the subject, however, is narrative in nature. Written in almost literary (and sometimes prejudiced) language, or to justify certain political positions, these works cannot always be rightfully characterized as studies or sophisticated investigations. That is, some of these texts—many of which are biographies or biographical sketches—are undoubtedly of historical interest, but lack source citations, and describe the cangaceiros in racist language or inaccurate terms that are unacceptable today.
The first biography of Virgulino Ferreira, the real name of the so-called “King of cangaceiros,” is an emblematic case. Written in 1926 by Érico de Almeida, a journalist from Paraíba, Lampeão, sua história contains an entire chapter of excessive praise for the “honorable,” “principled,” “righteous,” “generous,” “loyal,” “firm,” “brave,” and “energetic” João Suassuna, the state governor who spared no effort in combatting rural banditry. In Almeida’s estimation, Suassuna was a “titan,” defined by intelligence, character, and “incomparably brilliant” efficiency. In the book, apparently commissioned by Suassuna and financed by Congressman José Pereira Lima, the author argues that the “outstanding statesman” evoked the purity of purpose of Plutarch, immune to the moral dissolution of his time. The people’s trust in such a politician was thus supposedly limitless.28 Virgulino Ferreira, in turn, was characterized as a “supercriminal,” born and bred to commit the most heinous crimes. He killed for sport, stole obsessively, humiliated enemies, and set everything ablaze for fun.29
Another example of this genre of writing is Lampião by Ranulfo Prata, a doctor from Sergipe, completed in 1933 and published the following year. According to Prata, Lampião, the so-called “governor of the sertão,” was a “satyr, governed by hypersexuality, of evident somatic imbalance,” and a cruel, vain, and sanctimonious man.30 Prata’s portrait of the infamous cangaceiro was derived from descriptions and testimonies of third parties who supposedly had interacted with him. Prata, the son of a coronel (colonel), never met Virgulino Ferreira. His writing was informed by a partisan point of view committed to defending law and order. Other critics included renowned authors like Gustavo Barroso (especially in Heróis e bandidos, from 1917), Xavier de Oliveira with his Beatos e cangaceiros (released in 1920), and Abelardo Parreira in the book Sertanejo e cangaceiros, published in 1934—not to mention works by the “sociologist” Rodrigues de Carvalho—Lampião e a sociologia do cangaço—and the police officer Optato Gueiros—Lampeão.
Some authors, on the other hand, made unreasonable apologies in defense of the noted outlaw. These include Frederico Bezerra Maciel, who wrote a six-volume biography of Virgulino Ferreira, and Eduardo Barbosa, who wrote Lampião, rei do cangaço, a biased and romanticized narrative closer to fable than reality.31 These authors, unlike the aforementioned writers, relied on hearsay, popular stories, supposed conversations with bandits, andsecondary sources to lavish praise on the bandit and insist upon his essential decency. Barbosa notes that:
We all have our good side and our bad side. In the King of the Cangaço, the good side triumphed after meeting Maria Bonita. Thus, we can affirm that Lampião was our Robin Hood or Dick Turpin of the sertão. Against all statements to the contrary, there are the countless songs, poems, and sertanejo legends that immortalize Lampião as the protector of the lowly and the destitute.32
In another typical passage, this time regarding Lampião’s arrival in a town in the interior, Barbosa affirms that the people—the common folk who loved the outlaw, and who saw in his imposing figure the defender of their rights against the injustices of the coronéis and powerful politicians—ran into the streets, cheering loudly for their idol.33
Books like these, written in specific contexts and with specific goals in mind, are historical artifacts that must be read with careful detachment.
The theme of the cangaço also found fertile ground in the literary field. Novelists like Franklin Távora, Ulysses Lins de Albuquerque, Graciliano Ramos, José Lins do Rego, Maximiano Campos, Rodolfo Teófilo, Carlos Dias Fernandes, and even the noted painter Emiliano Di Cavalcanti, among others, depicted the phenomenon or illuminated some aspect of it through fictional or biographical narratives (sometimes only in single chapters in greater works, or in isolated excerpts from their works). Even if they did not necessarily produce accurate analyses of northeastern banditry, they nevertheless underscored the social importance of the cangaço and its place in the popular imagination of the time. Dozens of films from the 1920s to the present day—like O cangaceiro, directed by Lima Barreto; O primo do cangaceiro, directed by Mário Brasini; Lampião, o rei do cangaço, directed by Carlos Coimbra; and Deus e o diabo na terra do sol, directed by Glauber Rocha—did the same.
Later, interpretations of the cangaço by left-wing scholars began to emerge, focusing primarily on the context of social injustice in the backlands to explain and even justify the development of rural banditry. Authors like Rui Facó and Christina Matta Machado, from partisan, journalistic, or academic backgrounds, in some cases seemed to support the outlaws, seeing them as embryonic popular uprisings in the region.34 Even the noted British historian Eric Hobsbawm reinforced the image of cangaceiros as social rebels (especially in his book Bandits, originally published in 1969), although he lacked any field research or empirical analysis, relying exclusively instead on an extremely small and limited bibliography. The Communist Party of Brazil (PCB) endorsed a similar view starting in the 1930s, classifying the cangaço as an unconscious social revolt carried out by the impoverished, dispossessed, and exploited peasant masses—that is, a rebellion against the injustices of the backlands. While these pioneering analyses offered more sophisticated interpretations than their predecessors, emphasizing social issues in the countryside, they still did not present a complete picture of the motives and actions of most rural bandits.
Of course, there are exceptions. Several scholars, Brazilian and foreign alike, have produced fine works on the subject. Frederico Pernambucano de Mello, in Guerreiros do sol (2004), presents one of the most enlightening assessments of the cangaço.35 The author carried out sociological analyses based upon a classification of individual outlaws according to their criminal intentions and personal backgrounds. He also coined original concepts like the “ethical shield,” contextualizing cangaceiros within the irredentist tradition of violence in the sertão. In 2010, the same author published Estrelas de couro, an extremely intriguing study of cangaceiro attire. Works by Brazilianist Billy Jaynes Chandler also stand out, as do those by anthropologist Jorge Villela, who analyzed a vast set of documents in his O povo em armas (2004) to produce a seminal study of rural banditry in the Northeast during the First Republic.36 Relying on oral testimonies and interviews with old cangaceiros, Antônio Amaury Correia de Araújo was able to construct a linear and factual account of the cangaço, while giving voice to some of the protagonists themselves. Journalists have also produced key works, for example, Melchíades Rocha’s Bandoleiros das catingas (1938), and Moacir Assunção’s Os homens que mataram o facínora (2007).
Primary Sources
Several letters written by cangaceiros were published contemporaneously in the press, especially in the 1920s and 1930s, in magazines and newspapers such as Correio do Ceará, Noite Ilustrada, Jornal Pequeno, Jornal do Recife, and A Tarde. Judicial and police records, as well as declarations put out by the bandits themselves, can be found in the archives of different public institutions and private collections, such as the Museu do Sertão (Petrolina, Pernambuco), Museu Xucurus (Palmeira dos Índios, Alagoas), Melchíades da Rocha collection (Rio de Janeiro), Luís Torres collection, Frederico Pernambucano de Mello collection (Pernambuco), Instituto Histórico e Geográfico de Alagoas (Maceió), Arquivo Público Estadual (Recife), Biblioteca Pública Municipal (João Pessoa), Arquivo PMPE/Comissão de História, Arquivo Público Jordão Emereciano, Arquivo do Fórum de Floresta (Pernambuco), Arquivo do Fórum de Triunfo (Pernambuco), Arquivo do Fórum de Serra Talhada, and Fundação Joaquim Nabuco (Recife). Some documents have also been reproduced in scholarly works, like Frederico Pernambucano de Mello’s, Guerreiros do sol: violência e banditismo no Nordeste do Brasil (2004), and Luiz Bernardo Pericás’s Os cangaceiros: ensaio de interpretação histórica (2010), among others.
Further Reading
- Assunção, Moacir. Os homens que mataram o facínora. Rio de Janeiro: Record, 2007.
- Chandler, Billy Jaynes. The Bandit King: Lampião of Brazil. College Station: Texas A&M University Press, 1978.
- Freixinho, Nilton. O sertão arcaico no Nordeste do Brasil. Rio de Janeiro: Imago, 2003.
- Gueiros, Optato. Lampeão. São Paulo: Linográfica, 1953.
- Hobsbawm, Eric. Bandits. New York: New Press, 2000.
- Hobsbawm, Eric. Bandidos. São Paulo: Paz e Terra, 2010.
- Lewin, Linda. “The Oligarchical Limitations of Social Banditry: The Case of the ‘Good’ Thief Antonio Silvino.” In Bandidos: The Varieties of Latin American Banditry. Edited by Richard W. Slatta, 67–96. Westport, CT: Greenwood Press, 1987.
- Lima, Estácio de. O mundo estranho dos cangaceiros. Salvador, Brazil: Itapoã, 1965.
- Mello, Frederico Pernambucano de. Benjamin Abraão: entre anjos e cangaceiros. São Paulo: Escrituras Editora, 2012.
- Mello, Frederico Pernambucano de. Estrelas de couro: a estética do cangaço. São Paulo: Escrituras Editora, 2010.
- Montenegro, Abelardo. Fanáticos e cangaceiros. Fortaleza, Brazil: Henriqueta Galeno, 1973.
- Mota, Leonardo. No tempo de Lampião. Fortaleza, Brazil: Imprensa Universitária do Ceará, 1967.
- Oliveira, Antonio Xavier de. Beatos e cangaceiros. Rio de Janeiro: Revista dos Tribunais, 1920.
- Paiva, Melquíades Pinto. Cangaço: uma ampla bibliografia comentada. Fortaleza, Brazil: Editora IMEPH, 2012.
- Pang, Eul-Soo. “Banditry and Messianism in Brazil (1870–1940): An Agrarian Crisis Hypothesis.” In 8th Proceedings of the Pacific Coast Council on Latin American Studies, 1981–1982. San Diego, CA: San Diego State University Press, 1982, pp. 1–24.
- Pericás, Luiz Bernardo. Os cangaceiros: ensaio de interpretação histórica. São Paulo: Boitempo, 2010.
- Singlemann, Peter. “Political Structure and Social Banditry in Northeast Brazil.” Journal of Latin American Studies 7, no. 1 (1975): 59–83.
- Souza, Amaury de. “The Cangaço and the Politics of Violence in Northeast Brazil.” In Protest and Resistance in Angola and Brazil. Edited by Ronald L. Chilcote, 109–131. Berkeley: University of California Press, 1972.
- Young, Augustus. Lampion and his Bandits: The Literature of Cordel in Brazil. London: Menard Press, 1994.
Notes
1. See Abelardo Parreira, Sertanejos e cangaceiros (São Paulo: Paulista, 1934), 34.
2. See Maria Isaura Pereira de Queiroz, Os cangaceiros (São Paulo: Duas Cidades, 1977), 44.
3. See Raphael Bluteau, Vocabulário portuguez e latino.
4. See Luís da Câmara Cascudo, Flor dos romances trágicos (Natal, Brazil: EDUFM, 1999), 211.
5. See Juvenal Galeno, Cenas populares (Fortaleza, Brazil: Tip. do Comércio, 1871); See Franklin Távora, O cabeleira (Rio de Janeiro: Edições de Ouro, 1966); See Irenêo Joffily, Notas sobre a Parahyba (Brasília: Thesaurus, 1977).
6. According to Dicionário Houaiss, the term “canga,” of controversial origin, probably arose in the fourteenth century, and has several meanings. It could refer to a wooden frame on thatched roofs, a piece of wood to attach a shaft to a cart or a plow, a beam rested on the shoulders of two loaders and used to carry heavy objects, or an instrument of Chinese torture, bargain, dominion, or oppression, among other possible definitions. See Instituto Antônio Houaiss, Dicionário Houaiss da língua portuguesa, 597. Regarding the word, for his part, Nei Lopes writes: “controversial etymon. Springs, separating definitions into two entries, assign, respectively, Celtic and Chinese origin. In our view, the origin may be in the Kikongo term kanga, tying, holding, capturing, squeezing, from nkanga, the act of binding; that which is tied (Laman). Among the Cuban Congos, kanga is the binding, the magic connection.” See Nei Lopes, Novo dicionário Banto do Brasil (Rio de Janeiro: Pallas, 2003), 64. On the other hand, Rodolfo Garcia suggests the term comes from the Tupi acanga. See Bernardino José de Souza, Dicionário da terra e da gente do Brasil (São Paulo: Companhia Editora Nacional, 1961), 80.
7. See Rodrigues de Carvalho, cited in José Alípio Goulart, Da palmatória ao patíbulo, castigos de escravos no Brasil (Rio de Janeiro: Conquista, 1971), 64.
8. See João Emanuel Pohl, cited in José Alípio Goulart, Da palmatória ao patíbulo, castigos de escravos no Brasil, 64.
9. See Jean-Baptiste Debret, cited in José Alípio Goulart, Da palmatória ao patíbulo, castigos de escravos no Brasil, 64.
10. According to Dicionário Houaiss, “cangalha,” a term used since 1518, refers to a device made of wood or iron, usually padded, which attaches to the backs of horses to hang cargo on both sides, a wooden triangle placed on the neck of pigs to prevent fires, a cart pulled by only one animal, or a frame of wood and iron in which cannons or boxes of ammunition are placed when driven on the backs of animals. See Instituto Antônio Houaiss, Dicionário Houaiss da língua portuguesa, 598. According to Nei Lopes, this is a controversial etymon which some scholars believe shares the same origin as canga, while others argue the word stems from the Kikongo kangala, meaning to obstruct or impede. See Nei Lopes, Novo dicionário Banto do Brasil, 64.
11. See José Alípio Goulart, Meios e instrumentos de transporte no interior do Brasil (Rio de Janeiro: Ministério da Educação e Cultura, 1959), 205–206.
12. See Manoel Rodrigues de Melo, Patriarcas e carreiros (Natal, Brazil: Universidade Federal do Rio Grande do Norte/Universitária, 1985), 167.
13. See Aglae Lima de Oliveira, Lampião, cangaço e Nordeste (Rio de Janeiro: O Cruzeiro, 1970), 51.
14. See Protásio Pinheiro de Melo, Contribuição indígena à fala norte-rio-grandense (Natal, Brazil: Imprensa Universitária, 1971), 15.
15. See Maria Christina Russi da Matta Machado, As táticas de guerra dos cangaceiros (São Paulo: Brasiliense, 1978), 24. See also Maria Gouveia Seitz, Trilhando com a imaginação: uma visão romântica do banditismo na literatura do Nordeste brasileiro (PhD diss., Indiana University at Bloomington, January 2004), 59. While the term “canga” appears in Nei Lopes’s Novo dicionário banto do Brasil (Rio de Janeiro: Pallas, 2003), the word “cangaço” does not. The author, therefore, does not suggest the word is of African origin. Bernardino José de Souza, on the other hand, based on Rodolfo Garcia, argues that the term actually originates from the Tupi “acanga,” as previously mentioned. See Bernardino José de Souza, Dicionário da terra e da gente do Brasil, 80–81.
16. There are exceptions, of course. The case of the cangaceiro Gato is an interesting example. When he was already a member of Lampião’s gang he went away on one occasion to visit relatives. He then massacred his entire family. Since his parents were no longer alive, he killed his grandmother, two aunts, four brothers, and two cousins. This case, however, was highly abnormal. Another example is that of Adolfo Meia-Noite, from Afogados do Ingazeira, Pernambuco. Beaten by his uncle, who wanted to dissuade him from courting his daughter, Adolfo took revenge and murdered his uncle. See Gregg Narber, Entre a cruz e a espada: violência e misticismo no Brasil rural (São Paulo: Editora Terceiro Nome, 2003), 118.
17. See Renato Castelo Branco, Domingos Jorge Velho e a presença paulista no Nordeste (São Paulo: TAQ, 1990), 7.
18. Lucas da Feira’s area of activity was not restricted to the Feira de Santana, although that city was his main source of support. His actions took place mostly between the Agreste and the Recôncavo Baiano.
19. See Hamilton de Mattos Monteiro, Crise agrária e luta de classes (Brasília: Horizonte, 1980), 74.
20. See Frederico Pernambucano de Mello, Guerreiros do sol: violência e banditismo no Nordeste do Brasil (São Paulo: A Girafa, 2004), 28.
21. For more on the cangaço in the agreste, see Antonio Vilela de Souza, O incrível mundo do cangaço (Garanhuns, Brazil: Bagaço, 2006). The most famous cangaceiro of the agreste was Paizinho Baio, also known as the Lampião of the agreste. He operated in Garanhuns, Bom Conselho, Correntes, Águas Belas, and Buíque, in Pernambuco. He remained a bandit for twenty years. Corisco, on the other hand, after mid-1934, settled in the backlands of Alagoas. See Gregg Narber, Entre a cruz e a espada: violência e misticismo no Brasil rural, 145.
22. For more on the concept of the “Leather Civilization,” see João Capistrano de Abreu, Caminhos antigos e povoamento do Brasil (Rio de Janeiro: Sociedade Capistrano de Abreu/Livraria Briguiet, 1960); Capítulos de história colonial, 1500–1800 (Rio de Janeiro: Sociedade Capistrano de Abreu/Livraria Briguiet, 1954); and R. P. Castelo Branco, A civilização do couro (Teresina: Departamento Estadual de Imprensa e Propaganda, 1942). Capistrano de Abreu observed that, “the door of the huts was leather, the rough bed on the hard floor, and later the bed for childbirths; all the ropes were leather, the receptacle to carry water, the mocó or saddlebag to take food, the sack to store clothes, the knapsack to ride on horseback, the reigns to tie it on a long journey, the knife sheaths, the brocades and burlap, the clothes to enter the brush, the hide for curtains or to set salt; for the dams, the material was carried on leather straps pulled by oxen that tread the earth with their weight; on leather, tobacco was stepped on for the nose.” See João Capistrano de Abreu, Capítulos de história colonial, 1500–1800, 217–218.
23. See Jorge Villela, O povo em armas (Rio de Janeiro: Relume Dumará, 2004), 86.
24. According to José Guimarães Duque, the sertão “is the hot, inland region, with a height of no more than 300 meters with an area of 6,982,000 hectares, more rainy than Seridó and Carrasco, with yellow, compact, and shallow ground, partially covered by rolled pebbles, where a blanket of wild grass and weeds in winter are interspersed with trees and sporadic shrubs; the sudden downpours, not finding in the soil the permeability and the depth for rapid imbibition, drag in the floods the slopes of clay into the streams. In the summer, overgrazing and wind cleanse the land to receive in other uncertain rains a new contingent of water […] The average evaporation of the area is from 200 m to 2,200 m; with 9,900 to 3,400 [hours] of sunlight per year; the average temperature ranges from 14 to 37 degrees centigrade; there is no dew. The dryness index is 4.5 in the most arid years and 5.6 in the wetter years. The soil is mostly Archaean. In the winter, the vegetables that appear in the soils are herbs, grasses, legumes, malváceas, and subterranean convolvulaceae; shrubs of low canopies, short and twisted branches interspersed with cacti; trees and shrubs are distant enough to allow for the growth of pasture, presenting more pungent vegetation and soils with more water than Seridó.” See José Guimarães Duque, cited in Carlos Bastos Tigre, “Regeneração natural das formações arbóreas da caatinga,” in As regiões naturais do Nordeste, o meio e a civilização, ed. J. Vasconcelos Sobrinho, 182 (Recife, Brazil: Conselho de Desenvolvimento de Pernambuco, 1970), 182; According to Guimarães Duque, the Seridó “is a region of about 3.400.000 hectares characterized by low vegetation of cactaceous and aggressive clinging to the ground, of shrubs spaced with grasses and bare stretches in Archaean soil, highly eroded and rough, with rolled pebbles everywhere, and the masses of round granite standing out here and there, demonstrating how the slow erosion, through the centuries, leaves cyclopic traces. In Rio Grande do Norte and Paraíba, rainfall from January to May ranges from 127 mm to 916 mm per year, with an annual average of 497 mm. The average rate of sunshine is 2,988 hours per year; the maximum average temperature is 33 degrees centigrade and the minimum is 22 degrees; the aridity index is 3.3; in Ceará, the annual rainfall is 750 mm with an aridity index of 4.4. There is no dew.” See José Guimarães Duque, cited in Tigre, “Regeneração natural das formações arbóreas da caatinga,” in As regiões naturais do Nordeste, o meio e a civilização, 181–182.; The “agreste” is an intermediate region located between a humid area and a semi-arid area, or between the sea and the caatinga. It is subhumid, with milder temperatures at night; sometimes this region is affected by the rains of the Mata or the surplus rainfall of the caatinga; this allows the crops of foodstuffs, cereals, cassava and even tomatoes, such as the great industrial crop in Pesqueira (PE), to flourish. The rains are little later than in the Sertão and less irregular. The soils may be shallow, of Archean origin as in Pernambuco, or silica and deep sandstone as there are in the Agreste of Piauí. The Agreste of Piauí is marked by a sedimentary formation with acidic sandstone, humidity, and abundant of groundwater; the topography is very flat, and the soil is not fertile enough to sustain for food crops. The vegetation is of trees spaced with wild grass underneath. Grasses for pastures haveprevented the growth of shrubs. The cashew (Anacardium occidentale Linn), the faveiro, the piqui (Caryocar glabrum Pers), the carnauba (Copernicia cerifera Mart), and the tucum (Pyrenoglyphis maraja burret), are the most frequent. The region is susceptible to the effects of the rains of Maranhão (isoietas of 800 to 900 mm). The Agreste of Ceará is of little importance here; it is a track in the Serra do Araripe, after the Mata, in the Ceará coast of the mountains, found in the interior of the plateau. Between the Mata, a rainy area, and the interior caatinga, is the Agreste, marked by very sandy and weak soil, with few agricultural possibilities, because the water is at a great depth. The Agreste of Rio Grande do Norte, or Potiguar, includes eleven municipalities from Torres, according to the isoieta of rains of 1,000 mm to the border of Paraíba, near New Cross. The region is susceptible to the rains of the Brejo da Paraíba, diverted by the winds that bump into the foothills of Borborema; also, the proximity of the sea influences the atmospheric humidity. The soil is sandy, yellow, and full of underground water, which is sometimes of good quality, but sometimes brackish or full of limestone. The topography is flat and undulating, lending itself to large-scale cultivation. The valleys are damp or soaked with water. The Agreste in Paraíba spans the municipalities of Remígio and Esperança between the Brejo (Mata), the Cariris Velhos and the Curimataú. The climate is that of the mountain range of Borborema, with pleasant temperature and humid air coming from the city of Areia. The soil is siliceous, wavy, and eroded. The primitive vegetation was devastated and in its soils was planted small potatoes and beans. Currently, the agave crop is predominant. The agreste of Pernambuco spans 27 municipalities. The soil formed by the decomposition of granite and gneiss is very shallow, already eroded and depleted, and native vegetation is greatly altered in its initial composition. The Agreste in the Northeast occupies 6,276,000 hectares, excluding the Agreste of Bahia.” See J. Vasconcelos Sobrinho, As regiões naturais do Nordeste, o meio e a civilização.
25. See Oswaldo Lamartine de Faria, Encouramento e arreios do vaqueiro no Seridó (Natal, Brazil: Fundação José Augusto, 1969), 26.
26. De Faria, Encouramento e arreios do vaqueiro no Seridó, 27.
27. See Gustavo Barroso, Heroes e bandidos (Rio de Janeiro: Livraria Francisco Alves, 1917), 11.
28. Érico de Almeida, Lampeão, sua história (João Pessoa, Brazil: Universitária, 1998), 7–12.
29. De Almeida, Lampeão, sua história, 87.
30. See Ranulfo Prata, Lampião (São Paulo: Traço, [s. d.]), 30.
31. See Eduardo Barbosa, Lampião, rei do cangaço (Rio de Janeiro: Edições de Ouro, 1968).
32. Barbosa, Lampião, rei do cangaço, 7.
33. Barbosa, Lampião, rei do cangaço, 32.
34. See Rui Facó, Cangaceiros e fanáticos (Rio de Janeiro: Civilização Brasileira, 1963). Although published in 1963, the book collects pieces Facó had published in Revista Brasiliense since 1958. See Carlos Alberto Dória, “O Nordeste ‘problema nacional’ para a esquerda,” in História do marxismo no Brasil, Visões do Brasil, vol. 4, eds. João Quartim de Moraes and Marcos del Roio, 283 (Campinas, Brazil: Editora da Unicamp, 2007); See Christina Matta Machado, As táticas de guerra dos cangaceiros (Rio de Janeiro: Laemmert, 1969). See also Maria Christina Russi da Matta Machado, “Aspectos do fenômeno do cangaço no Nordeste brasileiro, I,” Revista de História 93 (1973): 139–175; Maria Christina Russi da Matta Machado, “Aspecto do fenômeno do cangaço no Nordeste brasileiro, II,” Revista de História 95 (1973): 177–212; Maria Christina Russi da Matta Machado, “Aspectos do fenômeno do cangaço no Nordeste brasileiro, III,” Revista de História 96 (1973): 473–489; Maria Christina Russi da Matta Machado, “Aspectos do fenômeno do cangaço no Nordeste brasileiro, IV,” Revista de História 97 (1974): 161–200; and Maria Christina Russi da Matta Machado, “Aspectos do fenômeno do cangaço no Nordeste brasileiro, V,” Revista de História 99 (1974): 145–174.
35. See Frederico Pernambucano de Mello, Guerreiros do sol, violência e banditismo no Nordeste do Brasil (Recife, Brazil: Fundação Joaquim Nabuco/Massangana, 1985).
36. See Billy Jaynes Chandler, “Brazilian Cangaceiros as Social Bandits: A Critical Appraisal,” in Bandidos: The Varieties of Latin American Banditry, ed. Richard W. Slatta, 97–112 (Westport, CT: Greenwood Press, 1987), and Billy Jaynes Chandler, Lampião, rei dos cangaceiros (Rio de Janeiro: Paz e Terra, 1981); See Jorge Mattar Villela, O povo em armas, violência e política no sertão de Pernambuco (Rio de Janeiro: Relume Dumará, 2004).