As condições de possibilidade das ocupações de terra

Lygia Sigaud

ABSTRACT

Keywords: Land occupations, Social movements, Agrarian reform, Northeast Brazil, South Africa.

RESUMO

Ocupar terras e nelas montar acampamentos é, em nossos dias, a forma apropriada para reivindicar a reforma agrária no Brasil e dela se valem as organizações do mundo rural, como o MST e o movimento sindical. O Estado tem conferido legitimidade à pretensão dos movimentos ao desapropriar as terras ocupadas e redistribuí-las. Esse fato recente na história nacional é examinado a partir de pesquisa realizada em Pernambuco, estado com o maior número de ocupações. O foco é a zona das plantações canavieiras, onde há grande concentração de acampamentos. O artigo inscreve as ocupações na história recente da região, mostra o que contribuiu para que se multiplicassem e analisa suas implicações. Ao final é feita uma digressão sobre a África do Sul, onde as ocupações, tidas pelos movimentos sociais como o procedimento adequado para obter do governo a distribuição de terras, não possuem a mesma magnitude. A comparação permite identificar as condições sociais que no caso brasileiro têm favorecido a institucionalização das ocupações e no caso sul-africano as têm obstaculizado.

Palavras-chave: Ocupações de terra, Movimentos sociais, Questão agrária, Nordeste,África do Sul.

Over the last twenty years, invading private lands and implanting squatter camps has evolved into the appropriate form of pushing for agrarian reform in Brazil. This strategy has been used by the Landless Rural Workers Movement (MST), workers unions and many other rural organizations. The Brazilian State has legitimized the aim of these movements (as these organizations are called by themselves and others) by disappropriating farms and redistributing the lands to those occupying the encampments.

This type of occupation is a new event in Brazilian history. The occupations that took place before 1964, such as those in Rio Grande do Sul and the state of Rio de Janeiro,¹ lacked the features or the amplitude of those that have become widespread over the last two decades, nor did they become the appropriate form of demanding land disappropriations. There were other ways of pursuing this objective, notably the campaigns to change the Brazilian Constitution.² Following the 1964 military coup, land invasions effectively became impossible. The lands obtained in this way were returned to their legal owners and activists from the rural workers associations were suppressed by the police and military.

From the end of the 1970s onwards, occupations were resumed in Rio Grande do Sul, invariably associated with the installation of encampments with dozens, if not hundreds, of families. The first invasions, organized by colonos,³ received strong support from the  Comissão Pastoral da Terra (CPT), an entity linked to the Catholic Church. In 1984, this nucleus created the Landless Rural Workers Movement (Movimento dos Trabalhadores Rurais Sem Terra), or MST for short (cf. Stedile & Mançano 1999; Mançano 2000). In the middle of the 1980s, thanks to the organization's policy of expansion, invasions were registered across a number of Brazilian states. In 1993, the National Congress established that non-productive lands constituted a failure to fulfil the social function of the property in question, a case stipulated by the 1988 Constitution as a basis for disappropriation. The invasions spread across the country as a whole, promoted not just by MST but also by the rural workers unions and by dozens of other organizations created with the primary aim of occupying lands. During this period, the National Institute of Agrarian Reform (Incra), which until then had played a low-key role, began to disappropriate occupied lands and redistribute them to those living in the camps, turning them into parceleiros, that is, owners of a small portion (parcela) of land. Hence, the invasions, camps and disappropriations associated with the occupations indicate a change in the way in which various rural organizations and the State acted. This allows us to speak of a new event.

In conclusion, I turn my attention to South Africa⁶ where occupations are contemplated by the Landless People's Movement – an organization that in turn claims to be heavily inspired by the actions of MST – as the most effective procedure for pressurizing the government to restitute the lands confiscated by the British in 1913, redistribute lands concentrated in the hands of whites, and guarantee social rights and land ownership to those who work on it, all measures established in the post-apartheid legislation. The land invasions registered in various provinces diverge from those taking place in Brazil, particularly insofar as they aim to gain lands to live on rather than use for production: a case in point was the occupation of Bredell on the outskirts of Johannesburg in 2001. In contrast to Brazil, rather than a spiral of invasions, there is a spiral of attacks on farms and murders of black workers and white farm owners (cf. ICG 2004; Steinberg 2002). By comparing the two countries, I shall look to identify the social conditions that have favoured the institutionalization of land invasions in the Brazilian case, and their suppression in the South African case.

The encampments

The first occupation recorded in Rio Formoso was organized by MST militantes, as its members are called, and union activists from the municipality.⁷ In April 1992, around 1,200 people (men, women and children) entered the Camaçari plantation, set up a camp and demanded the disappropriation of the land. The press reported the event as a rally for jobs and food supplements: the journalists at this time lacked the perceptual categories enabling them to recognize they were face-to-face with a new event.⁸

Camaçari was taken to be the property of the Federal Railway. The owners of the Cucaú Sugar Mill succeeded in proving that the plantation belonged to them and the district judge ordered the eviction of the occupants by a police force combining hundreds of officers. Many squatters returned to their homes. Around eight hundred, however, rebuilt the camp in Vermelho, a small landholding in Rio Formoso, using this as a base for a series of invasions of plantations that, according to Incra's criteria, could be considered non-productive and therefore subject to disappropriation. These occupations were led in conjunction by the MST activists and the union leaders from Rio Formoso until 1996, when the latter were left to organize them alone.

By reconstructing the various encampments, their recurring features came to the surface. After invading the plantation during the night or at dawn, the participants searched for high and visible locations near to forest and running water. There they set up the tents with timber taken from the forest, covered them with leaves and, finally, thick plastic sheeting which they called lona, or 'canvas.' The tents were aligned to form streets. The setting up of the campsite also included the installation of a tall mast on which the flag of the organization coordinating the occupation was hoisted.⁹ At the start, only the MST flag was used, since the unions only started to have their own flags when the Pernambuco Farm Workers Federation, Fetape, which collectively represents the unions, began to include land invasions in its program of actions.

The number of participants varied, ranging from over one hundred, as in the case of the São João plantation, invaded in 1996, to small groups, such as the nine which occupied the Brejo site in 1997. After invading the site and setting up the tents, the number of occupants could rise with the arrival of more people – again, as in the example of the Brejo occupation, which at one point had more than sixty people camped – and also fall, as occurred in São João, where just nineteen people remained, and Cipó (occupied in 1993), which went from eighty to 35 occupants. These reductions arose from people's own decision to leave or from the expulsion of those whose behaviour was deemed unacceptable by the group as a whole.¹⁰

In general, adult men were at the forefront in occupying the lands, with women and children arriving later. Setting up the tent signalled taking part in the encampment. Most individuals did not stay there the whole time, since they still needed to sustain their families, leaving to work in the sugarcane fields, undertaking casual construction work, acting as watchmen or travelling salesmen, searching for crabs in the mangrove swamps, and so on, while their families stayed to look after the tents. Others left for long periods, leaving the tent closed and either unattended or with a family member or acquaintance keeping an eye on it. They would periodically return to reaffirm their ties with the other occupants.

All the encampments organized ongoing tasks through a series of 'commissions,' such as the ones assigned with camp security, particularly at night, and the commission responsible for organizing the food supplies to the participants. The movements pressed the government bodies, especially Incra, but also the local mayors' offices, councils and churches for food for the camp occupants, as well as the imposition of tolls on the roads to raise funds and to appeal for food items from commercial establishments.

The duration of the encampments also varied, sometimes lasting months, when they were disbanded following the disappropriation of the lands, or years, as in the case of Mamucaba, an encampment started in 1998 and which was still running in 2004. Almost all the camps in the area were disbanded after judicial order when the landowners' demand for repossession of their properties was granted by the district judge. The evictions were generally followed by the reassembly of the encampment at the same location or nearby, next to the roads.¹¹ Many were subject to attacks by the private militias set up by farm owners, who took their own steps to evict the occupants, as occurred in Mascatinho, Jundiá de Cima and Mato Grosso.

After the first occupation, the plantation became the object of a claim for disappropriation, and the participants were transformed into claimants for the redistribution of the lands. Even when the camp was removed from the claimed lands, it remained associated with them and was known by the name of the plantation in question. So, for example, the Cipó encampment, disbanded soon after occupation, was re-established for more than a year at the nearby location of Vermelho; the Mato Grosso camp, emptied after an attack by militias, was set up on a tract of land at Minguito, already disappropriated; and the Jundiá camp, attacked on the day of the invasion by a force of around one hundred men assembled by the farm owner, was relocated next to the road.

A specific vocabulary was associated with the invasions and encampments. Preference was given to the term occupy rather than invade, the latter verb being used by the media, the landowners and the public in general. To describe individual occupation, the workers used the verb enter. When they arrived with the intention of entering, they asked the site coordinator whether there was a vacancy, as though asking about a job. The objective of the entry was to take land, while life in the encampment was frequently described as living under the black canvas, indicating meagre living conditions and exposure to the elements (rain, stifling heat during the day and cold at night).

The encampments were therefore much more than the mere gathering of people to demand the disappropriation of a plantation. They included ritualized techniques for carrying out the occupation, a spatial organization, an etiquette for entering the camp and setting up the tent, day-to-day rules, a specific vocabulary and elements invested with a potent symbolism, such as the flag and the black plastic awnings, which became the distinctive icons of the camps. This combination of model aspects comprised a form, the encampment form (cf. Sigaud 2000). The model was engendered in the South of Brazil during the process leading to the formation of MST. Its activists, relocated to the Northeast, implanting the model in Pernambuco's Zona da Mata where the model was gradually adjusted to local conditions.

Between 1987 and 2003, Incra disappropriated 194 landholdings in Pernambuco, including sixteen plantations,¹² in the area covered by the municipalities of Rio Formoso and Tamandaré (a former district created as a separate municipality in 1996), and handed over areas of land to those who had previously lived and worked on the plantations, as stipulated in the legislation, as well as those found in the encampments. Fourteen of these had been occupied with the setting of camps, which reveals the close relation between the encampment form and the disappropriations undertaken by the state.

The belief in the black canvas

The implementation of the encampment form in Pernambuco's Zona da Mata is not at all self-evident. There are no elements in the region's recent history to suggest that the territory of the engenhos, traditionally under the close control of the bosses, would become occupied with encampments; or that MST would work alongside the mass of workers in a region where the unions held sway virtually unchallenged; or that the union activists would start to occupy lands; or that workers would campaign to occupy someone else's property. Understanding how these developments became possible demands that we examine the occupations and encampments in terms of broader social and historical contexts.

At the end of the 1980s, the Brazilian government altered its directives concerning the sugar agroindustry in the wake of a more generalized withdrawal of the State from the economy: it ceased the subsidies that for decades had guaranteed the cane and sugar prices; it privatized the exports that until then had been made by the Sugar and Alcohol Institute; and it allowed a rise in interest rate. These measures, as well as a severe drought that hit the region during this period, provoked a crisis in the sector. Many bosses, whether the sugar mill owners or sugarcane suppliers, were unable to adapt to the lack of State protection and went bankrupt. Many others underwent restructuring. Thousands of workers lost their jobs, whether due to the bankruptcy of their bosses or the downsizing implemented by the restructuring companies (cf. Correa de Andrade 2001).

It was within this setting and the context of its own process of expansion across Brazil (cf. Mançano 2000; Stedile & Mançano 1999) that MST arrived in Pernambuco's Zona da Mata. On the southern coast, the MST activists allied with the unions and started to occupy the plantations, bringing the appropriate technology for invading lands and setting up and running the encampments. The unions collaborated with their members, the contacts between workers and their infrastructure, especially the union buildings (cf. Rosa 2004, p. 77). The occupation of Camaçari, in 1992, was a product of this cooperation and is seen even today as an inaugural landmark: “Everything began in Camaçari,” as the leaders and workers who took part in this and later occupations frequently say – indeed, an idea explicitly echoed by MST in its official history.¹⁴ From this point on, MST was able to recruit young people and in a short time assembled a network of activists who started to assist the movement and its occupations.

The alliance between union activists from Rio Formoso and MST is worth highlighting. Although the demand for agrarian reform had always been on the agenda of the union movement, it had not contemplated invading lands as a way of acquiring ownership. As Rosa shows (2004), both the desire of the younger union activists to gain promotion within the union hierarchy and the wish of the older activists to build a local political career contributed to the consolidation of this alliance on the southern coast during the crisis affecting the sugar agroindustry. From 1996 onwards, the union activists began to set up the camps in the region's plantations on their own.

At state level, Fetape was being increasingly pressurized by some union activists already involved in the encampments to include the occupations in its program of actions. At that time, the unions no longer held a monopoly in terms of representing workers – something acquired gradually from the beginnings of rural unionism in 1962¹⁵ – and feared losing the strength and prestige they had once enjoyed in Pernambuco. In 1997, Fetape was already occupying as many lands as MST.¹⁶ The change in the union movement's strategy gave a spectacular impulse to the occupations in Pernambuco, which increased exponentially. Between 1990 and 1994, the state was placed sixth in terms of the number of occupations, with 28 from a national total of 421, and fourth in terms of  families involved, with almost 5,000 from a total of around 75,000. Between 1995 and 1999, the period in which Fetape also carried out invasions, it became the first placed state both in terms of occupations, 308 from a total of 1,855, and in terms of families, 35,000 from a total of around 256,000 (cf. Mançano 2000, pp. 270-272).

For the rural workers of the Pernambuco Zona da Mata, occupying a plantation without the owner's permission in order to obtain a portion of the lands disappropriated from the bosses was not part of their horizon of possibilities. They only entered a plantation after being accepted to work there and then settled in locations designated by the boss and his managers. There was the utopia of the engenho liberto or 'free plantation' (cf. Sigaud 1979, pp. 205-222),where workers could cultivate their own smallholdings (sítios) and swiddens (roçados),¹⁷ breed as many livestock as they wished and work for the boss only when they needed money. All this presupposed the owner's presence and did not imply the idea of ownership on the worker's part. Given the evidence at hand, therefore, the occupation of plantations cannot be explained as the product of a prior desire to possess land.

The unemployment resulting from the crisis in the sugar agroindustry might be an alternative and attractive explanation for the land occupations, and indeed this is the reason given by union and MST activists to justify the influx of workers to the squatter camps: after losing their jobs, workers answered their calls and went to the camps. The problem is that unemployment was always rife during the periods between sugarcane harvests from March to August when what was called the tranca de inverno took place (cf. Sigaud 1979, pp. 167-204). Although this situation worsened with the crisis – and the occupations tended to occur precisely at the end of the milling –, being unemployed does not seem to have been sufficient reason for being in the camps. Thousands of unemployed workers preferred to continue undertaking odd jobs than set up camp under the black tarpaulins, turning down the invites with the argument that they did not want land. On the other hand, other workers with a current work contract took part in the occupations and set up their tents, striving to reconcile their presence in the camp with the formal work for the boss, such as those of Pedra de Amolar, a plantation ran by the Cucaú Sugar Mill, who formed the nucleus of the occupants of the Mato Grosso plantation in Rio Formoso, in 1999.

The workers living in the encampments between 1997 and 2000 claimed that they were there to get land. They had distinct social origins and careers, as has already been pointed out. The analysis of their histories reveals a wide range of situations preceding their arrival at the camps: some had lost their jobs; others were left homeless after a large flood struck Rio Formoso in 1997; still others wanted to restart their lives after a family crisis (separation, illness or death), or were lured by the presence at the camps of relatives and people they knew and by the proximity of the encampment to where they lived; finally, there were those who accepted the invite as they had close relations with activists and union members and trusted them.

Amaro Santino was living at the Brejo camp in September 1997. He had arrived at the encampment on the 27th of May, a month and a half after the initial occupation. He was 48 years old and had eighteen children, fifteen of whom lived with him. Born in Sirinhaém, he had resided for 29 years on a plantation ran by the Trapiche Sugar Mill, but fell out with the plantation manager and preferred to leave: he handed in his notice and went to Tamandaré, where he lived with a brother. He heard about the Brejo camp on the radio:

I said: the Brejo plantation has got a problem with Incra. Calling a lot of people [the reference is to a radio program broadcast by MST]. And then there was also this problem with me there [in Trapiche]. And so I said: I'm leaving for Incra [Brejo].

Nazareno, living at the Brejo camp, had resided in Tamandaré and made his living from selling fruit, fish and crabs prepared by his wife.

So I was there, wandering about, and I went past here [Brejo] all the time. So I saw this business with the movement. The person who told me about it was Zezinho. Zezinho is an activist, isn't he? A coordinator. So I listened carefully [to what he had to say]. And one day I came here. I spoke to the lads and they told me [to stay]... So I sent up that little tent over there.¹⁸

Palhaço lived in Rio Formoso and went to the camp in Mamucaba in 1999:

So, my little boy got sick, you know? The smell of the tide, the smell of putrid mud – yes, from the sea when the tide is out –  so I left, I said: I'm going to sell this here hut, and I sold it and used the proceeds to buy on goods for my wife, and she went to her mother's house, she spent fifteen days there, and I came here to the sem-terra.

Dalvino, who came from the Agreste, also set up camp in Mamucaba in 1999:

It was when I came here again, after a job. I stayed here, knocking and knocking on all the doors, I couldn't find any job in the companies, in the sugar mills, so I went away to the sertão again. But then some time a later a guy, a colleague said to me: “go join the sem-terra!” So I came here [to the Mamucaba encampment].¹⁹

For example, Amaro Santino (Brejo) remarked that “I wanted to get a terreno [piece of land] to work with my children more so I wouldn't have to be indebted to these bosses any longer.” Daniel Pedro, living at the Brejo camp, said:

Because I'm 44 years old. Yes, 44 years of suffering in the company, understand? And I didn't acquire anything. I worked all this time for others and acquired nothing. And me working. Now I'm going to try. Because I didn't acquire anything for myself by working. I've got nothing against trying my luck either. Because in terms of losing, I've already lost, right? [...] I think my best future is there [at the encampment]. And anyway, I've nothing to lose [...]. I'm going to wager my life, try my luck [...].

Likewise Edmilson (Brejo) remarked:

What I'd like… I want to work for myself… [to work] for others, the movement collapsed. Working so I won't fall by the wayside. I'm moving on. Because I've got a daughter, plus a wife; [working] for others, I'll be pulled down. It doesn't work anymore.

The belief that a better future could be obtained 'living under the black canvas' was, therefore, a decisive element in terms of explaining and comprehending the willingness of workers to plant themselves on the lands of the bosses. How this belief emerged is difficult, if not impossible, to reconstruct. We can only point out its existence and formulate the hypothesis that it had gradually taken shape since Camaçari. Some participants in the latter occupation recount that at the outset few went to the preparatory meetings, since they distrusted what they were told and were fearful of what could happen to them. Slowly the group increased. The presence of union activists from Rio Formoso in these meetings certainly lent the meetings more credibility and helped diminish resistance to the idea of occupation. When, from 1993 onwards, Incra began to disappropriate occupied lands, the belief acquired consistency – people began to feel less afraid and harbour fewer doubts over the potential outcome of invading the lands.

The new event at this moment was the belief that by 'living under the black canvas,' they could achieve a better future. In the 1990s, along with the migration to the south of Brazil²⁰ or to Recife and the change in jobs and bosses, the black canvas became part of a repertoire of possibilities for 'improving one's life.' It was a new alternative, though still just one alternative among others.  Interpreting it in this way, enables a better explanation of facts that otherwise remain obscured if we see the decision to join the encampments as a conversion to the 'struggle for land,' as the more misty-eyed analyses of social movements suggest. One of these facts is abandoning the camps. When an individual joins an occupation, he or she believes in – and bets on – the possibilities of the 'black canvas.' The evictions, attacks by private militias, and the slowness of the process of disappropriating the plantation (“this land isn't going to be won,” people frequently say) work to discourage participants, shaking their belief and the conviction that they have betted on the right outcome. If, under these circumstances, another possibility emerges that seems more attractive to the worker, he will not hesitate in abandoning the camp. His departure does not mean a loss of belief, though. Many workers returned after a while to the same camp or entered another.²¹

The belief in the possibilities enabled by the squatter camps is not all-pervasive within the Pernambuco Zona da Mata. But not believing today does not mean not believing tomorrow. Since 1997 I have remained in contact with many workers who in principle did not wish even to hear about taking land and who I later found on a camp site. On the other hand, the belief does not produce automatic effects. Very often those who believe in this possibility prefer to wait for a better opportunity. From a sociological point of view, what matters is that this belief came to figure in the horizon of possibilities.

A founding and legitimizing act

The invasion of the plantations and the setting up of the encampments did not occur within a pre-existing context of land conflicts. An examination of the setting in the period preceding the occupations fails to reveal any signs of a marked breakdown in social relations. There were interruptions to wage payments, as in Amaragi, Sauezinho, Saué Grande and Coqueiro; the owner's death, as in Cipó; the transfer of plantations to the Banco do Brasil to pay off the debts accrued by the Central Barreiros Sugar Mill, as in the cases of Brejo, Serra d'Água, Minguito, Mascatinho and Jundiá de Cima. All these situations could have been confronted in customary form through legal action in the Labour Courts or by waiting for the arrival of new owners,²² rather than necessarily evolving into the disappropriation of lands.

It was the movements that, by promoting the occupation of plantations and the encampments, produced a shift in the course of events: they created a land conflict where previously none had existed and then requested disappropriation by Incra. Here is not the place to speculate on why the occupation of these plantations took place; it is enough to  know that this was the objective of the movements at that moment in time and that these plantations, except Serra d'Água and Minguito,²³ fitted Incra's new criteria. The aim is only to highlight that the encampment form allowed problems that would have usually been solved in other ways to be transformed into a conflict over land.

Incra, for its part, accepted the legitimacy of the procedures: it recognized the movements as representatives authorized to request disappropriations, respecting their demands, and those taking part in the occupations as legitimate claimants to the land, allocating them portions. It also conferred an official existence to the movements and the encampment participants in its records. Thus, the tables referring to occupations, named as 'areas of conflict' (conflicts that, as we have seen, were created by the movements), contained, next to the columns with information relating to the location of the conflict, the size of the landholding and the number of resident families, one column with the number of families in the camp and another with the name of the movement behind the occupation. The forms used to register the future smallholders included acampado, 'camper,' a category without any legal status, alongside other legally recognized categories, such as rural worker, tenant, and so forth.

As more than 90% of the disappropriations made by Incra involved the so-called conflict areas, occupying plantations and setting up encampments – or, put otherwise, resorting to the encampment form – became a strategy impossible to ignore. This is the act that creates the conflict for land and sets off the process that may eventually result in disappropriation.²⁴ Here the Tentúgal plantation, owned by Central Barreiros Sugar Mill, in the municipality of São José da Coroa Grande (the far south of the coast), provides an exemplary case. When the mill company entered a state of bankruptcy, the workers who lived and worked on the plantation saw a solution to their problems in the squatter camps. They knew that they would achieve nothing without a movement taking the lead in the process. They therefore went to the union leaders from the municipality and asked them to organize an encampment on the plantation, which was soon abandoned, however, due to a lack of follow-up support from the union activists (according to the workers' version). After discovering that there were MST activists working in the area, they appealed to them to set up a new encampment and request disappropriation. In 1999, the camp was re-opened, not with the entry of workers onto the lands, since the acampados already lived on the plantation, but with the setting up of the encampment with the MST flag and the black plastic tarpaulins, symbols indicating the land issue. The owners demanded repossession of their lands and the encampment was disbanded, allowing the participants to continue living there but with the destruction of the tents and the confiscation of the flag. Over the following years, the camp was re-assembled several times and in 2002 Incra disappropriated Tentúgal. This case, which does not match the norm followed by most occupations,²⁵ has the virtue of showing how resorting the encampment form had already acquired an imperious character: it was not enough to issue a request for disappropriation, it had to be done in the proper way. The form is the way.

The occupation of lands and the setting up of encampments constitutes a symbolic language, a way of making statements through actions, a founding act in terms of acquiring legitimacy. By fomenting and leading an occupation and an encampment, the movement tells Incra that it desires disappropriation of the lands, the owner that it wants his property, and the other movements that the occupation in question has an owner. This language is well understood by everyone:  Incra accepts that there is a request for disappropriation and starts the process; the landowner perceives the possibility of losing his lands and acts in defence of his interests, requesting repossession, and the other movements respect the rival's flag and stay away from the land in question. With the act of occupation, the movements legitimate their claims for disappropriation and recognition that the occupation is theirs. By setting up his tent, the worker says that he wants the land. This claim is directed at Incra, which in selecting the future smallholders (parceleiros) will take into account those 'living under the black canvas;' at the movement, which will include him in its lists to be presented to Incra; and to the others living at the encampment, who will recognize him as someone who wants the land. The tent legitimates the desire to acquire land; it provides material proof of the interest to be taken into account when the lands are redistributed. 'Living under the black canvas' is portrayed as a form of hardship which makes those braving the ordeal deserving of land compensation. In some cases, when the occupation continued over a long period, as in Mamucaba, a hierarchical structure of legitimacy was formed, based on criteria such as the length of time spent at the camp, involvement in activities, continual presence, and the courage shown during evictions or confrontations with the private militias sent by the landowners to attack the camps.²⁶ This had no impact in terms of Incra's selection; however, it served to classify individuals into those who were more and less deserving.

The disappropriations taking place in the Pernambuco Zona da Mata have been the result, therefore, of a process fostered by the occupations and encampments, which characterized a situation of conflict over land in the terms recognized by Incra. Thanks to the legitimacy conferred by this body, the encampment form became the appropriate form of making demands. The State's sanction meant that occupying lands and setting up tents became constituted as acts aimed at legitimizing the claims of movements and individuals. Those interested in vitalizing a movement or in acquiring land found themselves obliged to adopt the form.

The public declarations of both the government authorities – especially the Ministry of Agrarian Development and Incra – and the movement representatives tend to be belligerent in tone, as though their relations involved permanent confrontation. Hence, over the last ten years, the authorities have issued frequent declarations via the media affirming that agrarian reform will be undertaken within the terms of the law and that constitutional violations (invasions of private properties) will not be accepted.²⁷ The movements, for their part, habitually accuse the government of failing to carry out agrarian reform and threaten new waves of land occupations. The tone of hostilities increased during the eight years of the Fernando Henrique Cardoso government and has become generally calmer since the beginning of the Lula government. Although this rhetoric reflects tensions between the parties involved, it obscures the relations of close cooperation and dependency between the State and the movements.

So far the Brazilian State has yet to implement a policy for the disappropriation of non-productive lands, whether at local or national level, despite the fact that both the Constitution and a 1993 regulation authorize such action. In the absence of a specific policy for conducting disappropriations, the government has relied on the occupations and encampments of the movements to indicate the farms where State intervention is viable. In this sense, the movements can be said to have provided the directives for the Brazilian state's policy in relation to the land issue: the disappropriated farms are those that have been occupied. Comparing the lists supplied by Incra of the disappropriations made under the last three governments (Itamar Franco, Fernando Henrique Cardoso and Lula) with the lists of occupations and encampments supplied by the movements clearly shows the close relationship between disappropriations and occupations. The State employees justify the disappropriations alleging that they involve conflict areas. This language is undoubtedly a survival from a period when areas with pre-existing conflicts ended up in disappropriations, as tended to occur in Amazonia in the 1970s and 1980s where bloody clashes were registered between settlers (posseiros) and land-grabbers (grileiros). As we have already seen, it is the movements that create the conflicts. The invasion and encampment characterize a situation of conflict and render it visible. The language of conflict areas involves euphemisms that conceal the arbitrary nature of the conflict.

The State also relies on the movements to identify the eventual recipients during the redistribution of disappropriated lands, chosen from among those that take part in the occupations and brought together by the movements. Contrary to public opinion in Brazil, there is not a mass of sem-terra (landless workers) eager to acquire land; the movements create the demand for land by inviting workers to occupy the farms. The movements are responsible for opening up the possibility of acquiring access to an area of land they had never even dreamt about, as it was possible to verify on a large scale in the research undertaken among the newly settled smallholders after the disappropriation of the plantations in the Pernambuco Zona da Mata.²⁸ By accepting the invitation and setting up in the encampments, the individuals became sem-terra, since they were now claiming the land for themselves. They therefore started to identify themselves in this form, taken as the appropriate modality for representing oneself in the space of the encampments, and began to be seen by the others from the rural world and the city as sem-terra. Neither those workers living in the urban outskirts, surviving off casual work and odd jobs, nor those hired by the plantations, are usually considered sem-terra: they are not involved in occupations and encampments, an indispensable condition for them to be identified in this form. The movements not only create the demand, therefore, they also generate the conditions of possibility for becoming a sem-terra and later receiving land through the process of agrarian reform.

For their part, the movements depended strongly on the State's intervention to lend impetus to their program of land occupations, since the benefits to be obtained comprised a powerful argument for calling people to join the occupations. In their accounts concerning how they were invited, the workers frequently stated that they were told that Incra was giving away lands; that the lands where they were going were non-productive and, therefore, would be disappropriated; that if they went to the camp they would receive the basic food supplies also given by Incra;²⁹ and that when the disappropriation took place, they would have access to loans to build a house and keep them going for a while until they began to produce, while they would still have funds left to invest in production. Each disappropriation of an occupied plantation and each release of loans for new settlement areas confirms the veracity of these pronouncements and encourage the acceptance of new invitations for future occupations. Hence, the dynamic of the occupations is a direct result of the State policy. Without the latter, the movements would have little grounds for the expectations they arouse in their target public and would encounter difficulties in rallying people for the occupations. They would have neither strengthened nor multiplied, as occurred in the Pernambuco Zona da Mata, where there are now in fact nine such movements (cf. Rosa 2004, pp. 172-173).

Incra and the movements are therefore connected by relations of mutual dependency and tacit cooperation. As these relations make up part of a figuration, to use Norbert Elias's term, containing individuals linked to other public powers, such as the Judiciary, as well as other movements and other social actors, such as the landowners, they tend to be complex and tense. Thus, most of the disappropriations in the Pernambuco Zona da Mata were made after an intensification in the pressure exerted by the movements on Incra, as for example those of Sauezinho, Saué Grande, Coqueiro, Cocal and Cocalzinho (plantations belonging to the Santo André Sugar Mill located in Tamandaré), which were only carried out at the end of 1999 after an encampment of more than 45 days involving around one hundred workers from these plantations in front of the Incra headquarters in Recife. These disappropriations were opposed by the owners of the Santo André Sugar Mill, with the support of leading politicians at national level. The attention given by the media seems to suggest that the tension is pervasive. However, I have tried to emphasize here the hidden dimension of dependency and cooperation that have contributed enormously to enabling the encampment form to take root.

Finally, it is worth stressing that the generalization of the encampment form has also resulted from the relations of dependency that link each movement with the people which it mobilized and successfully enabled to acquire land, as well as the competitive relations between the movements. The individuals who obtained land and access to loans by means of occupations feel indebted to the movement that made this possible. The debt implies obligations, such as loyalty and cooperation, and is described as a pledge (compromisso). The movements therefore count on the former acampados who today own a portion of the land to take part in marches and demonstrations, and especially when new occupations are involved. They go to help make up the numbers, teach the technique of occupation, give moral support to the newcomers and, by their own example, show that the hope invested in the squatter camps is grounded in reality. In all the occupations, there was a nucleus composed of successful former squatters.³⁰ The symbolic capital (prestige) and relative power (position in the correlation of forces) of the movements are constituted by their perceived achievements and victories: namely, the occupations and the disappropriations. The movements compete to accumulate ever more capital, a decisive element in terms of understanding the spiral of occupations (cf. Smircic 2000; Sigaud 2000; Sigaud et al. 2001; Rosa 2004).

South African digression

Due to legislation introduced by the British colonial administration and by the governments that implanted apartheid, most of South Africa's lands are in the hands of whites, meaning that 11% of the population controls 69% of the lands. The government established following the end of apartheid in 1994 instituted agrarian reform as one of its priorities, seeking to democratize the access to land, remedy 'injustices' committed in relation to the black population and work to ensure 'sustainable' rural development. Three programs were therefore created on the basis of laws voted through by Congress: the restitution of lands confiscated from blacks in 1913, to be made by lodging an appeal to the Land Claims Commission; the redistribution of lands through the transaction between buyers and sellers, with funds guaranteed by the government (something similar to what Fernando Henrique Cardoso tried to implement in Brazil with the Banco da Terra, a creation of the World Bank); and the regularization of land ownership for those living on communal land (in the 'homelands,' the lands occupied by tribes and controlled by tribal chiefs) or on the lands of white farmers (it is calculated that 1 million tenants³¹ live on white-owned lands).

In South Africa, non-governmental organizations represented by the National Land Committee (NLC) have been at the forefront of demands relating to the land issue. Controlled by human rights activists and lawyers, they have focused their action on the defence of rights (cf. James 2002). With the support of the NLC, the Landless People's Movement was created in 2001, becoming the first organization made up of people of rural origin and claiming the status of a social movement. There are indications that it depends on the financial support of NGOs, which limits its margin of action, especially in terms of its confrontations with the government, with whom many of the NGOs are allied and fairly unwilling to break ranks. These characteristics of the organization in the South African context contrast strongly with those of the Brazilian context. In the latter case, the rural workers organization dates from the 1950s when NGOs were yet to exist. The organizations that subsequently emerged, such as the Farm Workers Association, the Peasant Leagues, the Landless Farmers Movement (Master), the Brazilian Union of Farm and Agricultural Workers (Ultab), the rural workers unions and MST, were supported by political parties such as the Brazilian Communist Party (PCB), the Brazilian Labour Party of Rio Grande do Sul, left-wing militants (such as the Trotskyites and the Catholic left), the federal government (during the short period of João Goulart), the Catholic Church and, more recently, the latter's rural offshoot, the Pastoral Land Commission (CPT). It would be no exaggeration to say that all the organizations succeeded in gaining their independence from those sponsoring them. The union movement in the Pernambuco Zona da Mata is an eloquent example of this, in terms of its relations with the Catholic Church and the third government of Miguel Arraes. On the other hand, the existence of workers organizations at crucial moments in the history of the social relations in the rural world helped alter the course of the social transformations under way: here again, the example is the Pernambuco Zona da Mata. Thanks to their strong base, the unions succeeded in using legal action to prevent the expulsion en masse of workers from the large sugarcane plantations prompted by the approval of the Rural Workers Statute (ETR) in 1963 (cf. Sigaud 1994). In the South African case, on the other hand, the absence of a rural workers organization seems to be making it extremely easy to carry out the expulsion of tenants from farms, a process likewise enabled by a change in the legal framework: the Tenants Acts of 1996 and 1997. In both countries, the intention of farmers appears to be identical: to empty their lands of people with potential rights to them. Neither the Brazilian or the South African government has done much to guarantee the rights they themselves invented. The distinguishing feature is the existence of a workers organization in Brazil capable of opposing the force of the farmers. We can perhaps formulate the hypothesis that, in the South African case, the attacks on farms, the crimes perpetrated by whites against blacks and blacks against whites – or, in other words, this uncivilized form, by Western standards, of resolving conflicts – are related to the absence of a structure for representing interests and organizing and making demands, allied to a history of racial hatred. Clearly, other social conditions also contribute to conflicts being regulated through force, such as the control exercised by white farmers over the police and justice system, which ensures their impunity.

Following the recent events in Zimbabwe, with the seizure of the lands of white farmers sponsored by the Mugabe government, the South African white elite, the international investors and the government itself seem to have become gripped by fear, not only because of the geographical and historical proximity of the two countries, but also because of the way in which the seizure of lands has been as welcome among the South African poor. In order to tranquilize the elites and the market, the government has reiterated that it will not tolerate land occupations and that agrarian reform will be pursued strictly in terms of the law. The strong repression of the Bredell invasion forms part of this logic. For Brazil, the role played by Zimbabwe belongs to Cuba at the end of the 1950s and the start of the 1960s. The 1964 military coup in Brazil was aimed, among other things, at preventing the 'Cubanization' of the country. As the coup was made in the name of democracy, the military had to keep the courts running and maintain the existing laws, including the Rural Workers Statute. The pre-1964 land occupations, strikes and demonstrations were suppressed, but the unions remained open and were able to channel the demands to the legal institutions. The land occupations at the end of the 1970s, which proliferated in the 1980s and became widespread in the following decade, form part of the long history of confrontations between workers and their organizations and the large landowners.³² Although made outside the bounds of legality, the occupations were nonetheless accepted as legitimate by the Brazilian state, which from the outset responded by disappropriating and redistributing lands. It is easier to comprehend the repression imposed by the South African government than the acceptance of the Brazilian government, but this is a question that exceeds the limits of this work. The important point here is to highlight the fact that the severity of the South African repression has strongly inhibited the occupations. In the Pietermoritzburg area (a province in KwaZulu-Natal), I interviewed tenants in conflict with the farmers, who had been preventing them from cultivating land or using the pastures as they had traditionally, while simultaneously threatening them with expulsion. They were aware of the occupations in Brazil and admired MST. Although they nourished the idea of responding to the farmers through occupations, they were worried about carrying this project through because of the repression that they knew would come from the government and the private security forces employed by the farmers.

Thanks to ethnographic studies (cf. James 2004), we can comprehend that when the tenants express their desire to remain on the farms where they work, what drives their ambition is only the sense of security involved in continuing to have access to the land for crop growing and pasture. They have no wish to become landowners, nor small rural entrepreneurs, an idea that dominates the government's projects. In a country with a high unemployment rate and an acute housing problem, having a place to 'lay one's head' (as an interlocutor cited by James put it) is still a privilege. In Brazil, the rhetoric of using agrarian reform for the development of the rural economy also prevails, and indeed this may take place, yet it was not with the objective of becoming small entrepreneurs that individuals decided to 'live under the black canvas' in the Pernambuco Zona da Mata. What animated them was the idea of being able to own something for themselves, freeing them from the domination of the boss and benefiting from the State policies.

Conclusion

The land occupations in Brazil are recognized as a notable event both inside and outside the nation's frontiers. They are frequently seen in a positive light as a new manifestation of the 'fight for land,' or in a negative light as the devilish product of the manipulation of the masses by agitators. I have looked to explain and comprehend how they became possible, without falling into the trap of seeing them as the awakening of consciousness among the sem-terra masses. This was only possible thanks to particular methodological options. The analysis was grounded on an ethnography of the encampments and the comparison between them, which allowed the existence of a specific form to be identified. Subsequently, the conditions of possibility of implanting this form were questioned. In the process, the article looked to situate the encampments within the history of the social relations in which they occurred and, above all, attempted to comprehend the motives of those involved in the occupations. In discovering that the claim for land involved a rich symbolism of black canvas tents and flags raised on masts, it asked what was at stake in the assembly of the encampment for the movements and individuals involved: these were acts that legitimized their aims. I then sought to explain the dynamic and the institutionalization of the encampment form through the relations of mutual dependency and competition that link the State, the movements and the individuals. The comparison with the South African case allowed us to perceive the extent to which, in Brazil, the actions of the movements, their relations with the State and the actual policies of the State have contributed decisively to the institutionalization of land occupations in the country.

The occupations are not an effect of the 'struggle for land.' By this, I do not mean that this struggle does not exist: I simply wish to avoid an idealizing viewpoint that tends to obscure the actual causes. The demand for land is not pre-existent: it is produced by the movements and fed by the practices of the State. But as long as there are individuals willing to respond to the movements' invitations and to believe in the possibility of acquiring land, their actions help ensure that the 'struggle for land' comes into existence. This 'struggle' has numerous effects, including the condition of possibility of the policy of disappropriation of the Brazilian State over the last twenty years, the creation and strengthening of movements and, above all, the fact that, thanks to this policy, hundreds of thousands of people succeeded in gaining the attention of the Brazilian state, benefiting from the access to land and loan policies. Were it not for this 'struggle,' many would remain ignored, like the majority of the population, or only served as the recipients of short-term and emergency programs. Of course, the basis of the analysis is the Pernambuco region, but a well studied case illuminates the study of others, providing leads for new research projects and a model for analysis.

Notes

1. On the land occupations taking place during this period, see Eckert (1984) and Rosa & Palmeira (2004) on  Rio Grande do Sul; on the state of Rio de Janeiro, see Grynzpan (1987) and Ernandez et al. (2004).

2. On the pressures for Agrarian reform, see Camargo (1981).

3. The italicized words correspond to native categories. Colono is a term designating the descendents of German, Italian and Polish immigrants who settled in the South of the country from 1824 onwards as small producers.

4. The Pernambuco sugarcane region is one of Brazil's oldest zones of colonization. The Portuguese implanted sugarcane farming and the fabrication of sugar there in the 16^th century. Since the end of the 19^th century, sugarcane has been farmed on large estates by suppliers and mill owners with the employment of a large workforce. For a history of the social relations in the world of the engenhos, see Correa de Andrade (1964; 1989), Eisenberg (1977), Mello (1975), Palmeira (1971; 1976), Sigaud (1979), Garcia Jr. (1983) and Heredia (1979).

5. Undertaken by researchers from the Museu Nacional of Rio de Janeiro Federal University and the École Normale Supérieure (ENS), of Paris, with funding from the Ford Foundation, the José Bonifácio University Research Foundation (FUJB), the Rio de Janeiro State Research Support Foundation (Faperj), the National Council of Scientific and Technological Development (CNPq) and the ENS. This text makes use of previous analyses (Sigaud 2000; Sigaud et al. 2001), the material collected for the exhibition Lonas e bandeiras em terras pernambucanas (www.lonasebandeiras.com.br) and studies undertaken in the Pernambuco Zona da Mata. The analyzed corpus includes interviews and life histories collected from people who took part in the occupations, as well as observations made in the encampments and documents supplied by Incra.

6. The South African digression is based on the examination of the academic literature and the available documentation, and on interviews with rural leaders and workers carried out during two visits to the province of KwaZulu-Natal (2003 and 2004).

7. The Rural Workers Unions are organized on a municipal basis. In the Pernambuco Zona da Mata, the large majority of associates are workers hired on the plantations.

8. “The invasion of the area is due to the consequences of the period between harvests, when the sugarcane cutting ceases and the rural workers have no means of survival.” See Jornal do Commércio (1992).

9. For the reconstruction of an occupation in the Pernambuco Zona da Mata in 1999, see Smircic (2000, pp. 29-55).

11. The judicial order for repossession allows a single process of eviction only. When the land is reoccupied, the landowner must request a new repossession order, which many prefer not to do.

12. These were: Amaragi, Serra d'Água, Minguito and Mato Grosso in Rio Formoso, and Cipó, São João, Saué Grande, Sauezinho, Cocal, Cocalzinho, Coqueiro, Jundiá de Cima, Laranjeiras, Mascatinho, Brejo and Ilhetas.

13. See Boletins de Safra (1989-1999), published by the Pernambuco Sugar Industry Syndicate, Recife, 1999.

14. See www.mst.org.br/mstpe, 2000.

15. In 1962, the federal government regulated the law that allowed the creation of rural unions. See Camargo (1983), Bezerra (1979) and Wilkie (1964).

16. See Diário de Pernambuco, 11^th June 1997.

17. The sítio was an area of land with fruit trees conceded by the boss to the moradores ('residents') in whom they trusted (cf. Palmeira 1976); the roçado was a rotating area of land also ceded by the boss in which the moradores could plant short-cycle crops.

18. The interviews relating to the Brejo plantation were collected in September 1997 by the author.

19. The interviews relating to the Mamucaba plantation were collected by David Fajolles in September 1999.

21. Studies undertaken in encampments in Rio de Janeiro state by Ernandez (2003) and in São Paulo state by Loera (2004) indicate that the reasons for abandoning the camps are similar to those found in the Pernambuco Zona da Mata.

22. After labour rights were extended to the rural world, in 1963, plantation workers began to file lawsuits in the local justice system against the bosses. Following the military coup, the practice became a favoured strategy among the unions and turned into the form par excellence of confronting the bosses (cf. Sigaud 1999). On the other hand, changes in ownership were widespread. These involved a degree of tension, but relations tended to normalize after a period of adaptation to the new owner's style.

23. These plantations were productive, according to Incra's criteria, and were leased out. They were disappropriated because the Central Barreiros Sugar Mill, the owning company, had transferred them to the Banco do Brasil.

24. After the request for disappropriation, Incra sends a team of technicians to the plantation to carry out the inspection. If the lands are found to be non-productive, the administrative procedure continues to the next stage: the documentation is sent to Brasília and Incra's national head office, which submits the request to the President of the Republic, who has the final responsibility of signing the decree establishing that the property is failing to perform its social function. Disappropriation is then carried out, in which the landowner receives compensation for the land in the form of agrarian debt bonds (TDA) with the market value and compensation in cash for installations.

25. Most of the encampments in the Pernambuco Zona da Mata were constituted by the group occupying the plantation. However, there are cases of encampments set up by moradores to demand disappropriation. This took place in Amaragi, Sauezinho, Saué Grande and Coqueiro.

26. This hierarchy was identified by Fajolles (2000) in the Mamucaba encampment (cf. Sigaud et al. 2001, pp. 65-69).

27. A good example was the episode of the National Landless March. On this topic, see the extracts of reports published by the press compiled by Chaves (2000, pp. 265-341).

28. It was a research study on the Rio Formoso and Tamandaré settlements that enabled me to discover that the workers who had taken part in the encampments had never thought of the possibility of having their own house and portion of land on the bosses' plantations.

29. The studied encampments received basket food supplies (cestas básicas) on and off. Although the distribution was intermittent, the very possibility of having access to these staple foods was an important attraction for the workers who would not receive them were they living outside the encampments.

31. Tenants are those who live on the properties. They can be likened to the moradores (residents) of the large plantations of the Northeast and the colonos (colonists) of the coffee plantations. They work on the property, and have the right to grow their own crops and breed livestock.

32. Research into the sociogenesis of the land occupations being carried out at the Museu Nacional, coordinated by myself, shows that the first occupations at the end of the 1970s were related to the pre-194 occupations, both those in Rio Grande do Sul, which ended up with the creation of  MST, and those in Rio de Janeiro state (cf. Rosa & Palmeira 2004; Ernandez et al. 2004).

Bibliography

Association for Rural Advancement (AFRA). (2002), The emergence of the Landless People's Movement in SA. Pietermoritzburg: ZA, AFRA.

Bezerra, Gregório. (1979), Memórias. Segunda parte: 1946-1964. Rio de Janeiro: Civilização Brasileira.

Brenneisen, Eliane. (2003), “Assentamento Sepé Tiaraju: persistências do passado, fragmentos do presente.” In: Martins, José de Sousa (ed.). Travessias: a vivência da reforma agrária nos assentamentos. Porto Alegre: Editora da UFRGS.

Camargo, Aspásia. (1973), Brésil Nord'Est: mouvements paysans et crise populiste. Paris. Doctoral Thesis. University of Paris, Third Cycle (mimeo. ).

______. (1981), “A questão agrária: crise de poder e reformas de base (1930-1964).” In: Fausto, Boris (ed.). História geral da civilização brasileira. O Brasil Republicano. São Paulo: Difel, tome III.

Chaves, Christine de Alencar. (2000), A marcha nacional dos sem-terra: um estudo sobre a fabricação do social. Rio de Janeiro: Relume Dumará.

Correa de Andrade, Manuel. (1964), A terra e o homem no Nordeste. São Paulo: Brasiliense.

______. (1989), História das usinas de açúcar de Pernambuco. Recife: Massangana.

______ & Andrade, Sonia Correia de. (2001), A cana-de-açúcar na região da mata pernambucana. Recife: CNPq/FJN/UFPE.

Eckert, Córdula. (1984), Movimentos dos agricultores sem terra no Rio Grande do Sul: 1960-1964. Rio de Janeiro. Master's Dissertation. CPDA/UFRRJ.

Eisenberg, Peter L. (1977), Modernização sem mudança: a indústria açucareira em Pernambuco, 1840/1910. Rio de Janeiro/Campinas: Paz e Terra/Unicamp.

Elias, Norbert. (1991), Qu'est-ce que la sociologie? Marseilles: Editions de l'Aube.

Ernandez, Marcelo. (2003), 'Zé Pureza:' etnografia de um acampamento no norte fluminense. Rio de Janeiro. Doctoral Thesis. Postgraduate Program in Social Sciences, Rio de Janeiro State University.

______; Guedes, Mauro & Figueira, Renata. (2004), “Sementes em trincheiras: a luta pela posse da terra no Rio de Janeiro do pré-64.” Paper presented at the seminar “Workshop on Research on Social Transformation in the Rural World.” Rio de Janeiro, Museu Nacional, December.

Garcia Jr., Afrânio. (1983), Terra de trabalho. Rio de Janeiro: Paz e Terra.

______. (1990), O Sul: caminho do roçado. São Paulo: Marco Zero.

Grynszpan, Mário. (1987), Mobilização camponesa e competição política no estado do Rio de Janeiro (1950-1964). Rio de Janeiro. Master's Dissertation. PPGAS/Museu Nacional/UFRJ.

Heredia, Beatriz. (1979), A morada da vida. Rio de Janeiro: Paz e Terra.

International Crisis Group (ICG). (2004), Blood and soil: land, politics and conflict prevention in Zimbabwe and South Africa. Brussels: International Crisis Group.

James, Deborah. (2002), “Tenure reformed: policy and practices in the case of South Africa's landless people.” Paper presented at the Max Planck Institute.

Jornal do Commércio. (1992), Caderno Cidades, Recife, 29^th April, p. 2.

Loera, Nashieli Cecília. (2004), A busca do território: uma aproximação à diversidade do seu significado entre os sem-terra. Campinas. Master's Dissertation. Department of Social Anthropology, Campinas State University.

Mançano, Bernardo. (2000), A formação do MST no Brasil. Rio de Janeiro: Vozes.

Mello, Mário Lacerda de. (1975), O açúcar e o homem. Recife: MEC/Instituto Joaquim Nabuco.

Palmeira, Moacir. (1971), Feira e mudança econômica. Research Symposium, Museu Nacional/Latin American  Social Sciences Research Centre (mimeo. ).

______. (1976), “Casa e trabalho: notas sobre as relações sociais na plantation tradicional.” In: Actes du XLII Congrès des Américanistes, Paris.

Rosa, Marcelo Carvalho. (2004), O engenho dos movimentos: reforma agrária e significação social na zona canavieira de Pernambuco. Rio de Janeiro. Doctoral Thesis. IUPERJ.

______ & Palmeira, Isabel. (2004), “Fazenda Sarandi: considerações preliminares sobre a construção da legitimidade das ocupações de terra no Rio Grande do Sul.” Paper presented at the seminar “Workshop on Research on Social Transformation in the Rural World.” Rio de Janeiro, Museu Nacional, December.

Sigaud, Lygia. (1979), Os clandestinos e os direitos. São Paulo, Duas Cidades.

______. (1994), “Des Plantations aux villes: les ambigüités d'un choix.” Études Rurales, 131-132.

______. (1999), “Les Paysans et le droit: le mode juridique de règlement de conflits.” Informations sur les Sciences Sociales, 38 (1).

______. (2000), “A forma acampamento: notas a partir da versão pernambucana.” Novos Estudos Cebrap, pp. 73-92.

______. et al. (2001), “Histoires de campements.” Cahiers du Brésil Contemporain, 43/44: 31-70.

Smircic, Sergio Chamorro. (2000), Com a cara e a coragem: uma etnografia de uma ocupação de terras. Rio de Janeiro. Master's Dissertation. Museu Nacional/PPGAS/UFRJ.

Stedile, João Pedro & Mançano, Bernardo. (1999), Brava gente: a trajetória do MST e a luta pela terra no Brasil. São Paulo: Editora Fundação Perseu Abramo.

Steinberg, Johnny. (2002), Midlands. Johannesburg/Cape Town: Jonathan Ball Publishers.

Wilkie, Mary. (1964), “A report on rural syndicates in Pernambuco.” Rio de Janeiro, Latin American  Social Sciences Research Centre (mimeo. ).

Text received on 25/11/2004 and approved on 20/12/2004.

Lygia Sigaud is a senior lecturer at the Department of Anthropology of the Museu Nacional/UFRJ, a researcher for CNPq and a Faperj award holder. She has published books and articles on the world of the large sugarcane plantations of the Northeast and co-edited Antropologia, impérios e Estados nacionais with Benoît de L'Estoile and Federico Neiburg. E-mail: lsigaud@alternex.com.br