A Precariedade de Liberdade

IRSH 56 (2011), pp. 405–439 doi:10.
1017/S002085901100040X
r 2011 Internationaal Instituut voor Sociale Geschiedenis
The Precariousness of Freedom in a Slave Society

(Brazil in the Nineteenth Century)*
SIDNEY CHALHOUB
Department of History, Universidade Estadual de Campinas
(UNICAMP), Brazil
E-mail: [email protected]
SUMMARY : One of the main features of slavery in Brazil was that slaves had a better
chance of achieving freedom than was the case in other slave societies. However
difficult freedom may have been to obtain, significant rates of manumission resulted
in a high percentage of free and freed people of color in the population of the country
throughout the nineteenth century. This article analyzes facets of the structural
precariousness of freedom in nineteenth-century Brazil. It deals with such themes as
the constitutional restrictions on the political rights of freed persons; the masters’
interdiction of their slaves’ learning how to read and write; the practice of granting
conditional manumissions; the masters’ right to revoke liberties; the illegal enslave-
ment of free people of color; and police profiling of free and freed blacks under the
allegation that they were suspected of being slaves. The idea is to highlight situations
which often blurred the distinction between slavery and freedom, therefore rendering
insecure the condition of free and freed people of African descent.
INTRODUCTION
One of the main features of slavery in Brazil was that slaves had a better
chance of achieving freedom than was the case in other slave societies.
However difficult freedom may have been to obtain, significant rates of
manumission resulted in a high percentage of free and freed people of
color in the population of the country throughout the nineteenth century.
Nevertheless, much remains to be known regarding the experience of
freedom for freed people and their descendants while slavery still existed.
* This paper was written while I was a visiting scholar at Stanford University in the fall of 2010,
associated with the Spatial History Project and the Center for Latin American Studies. During
this period, I held a fellowship from the Fundação de Amparo à Pesquisa do Estado de São
Paulo (FAPESP). Research in Brazil was funded by the Conselho Nacional de Desenvolvi-
mento Cientı́fico e Tecnológico (CNPq) and FAPESP. My thanks to Michael Hall for his
corrections of my English. João José Reis and Rebecca J. Scott provided detailed comments that
were very helpful in preparing the final version of this article.
Downloaded from https:/www.cambridge.org/core. IP address: 189.6.16.57, on 09 Jun 2017 at 18:24:34, subject to the Cambridge Core
terms of use, available at https:/www.cambridge.org/core/terms. https://doi.org/10.1017/S002085901100040X
406 Sidney Chalhoub
This article seeks to analyze facets of what I will denominate the struc-
tural precariousness of freedom in nineteenth-century Brazil. Thus, I will
deal with such themes as the constitutional restrictions on the political
rights of freed persons; the masters’ interdiction of their slaves’ learning
how to read and write; free and freed blacks’ very limited access to ele-
mentary schooling; the practice of granting conditional manumissions; the
masters’ right to revoke liberties; the illegal enslavement of free people
of color; and police profiling and arrest of free and freed blacks under
the allegation that they were suspected of being runaway slaves. In
approaching all these complex themes very briefly, the idea is to highlight
situations which often blurred the distinction between slavery and freedom
in nineteenth-century Brazil, therefore rendering insecure the condition of
free and freed people of African descent.1
Comparing data on the free population of color in Brazil, Cuba, and the
United States, the three main slave societies in the Americas in the nine-
teenth century, helps to bring our subject into relief. According to the
Brazilian census of 1872, the only national census carried out before the
abolition of slavery in 1888, the country had 9,930,478 inhabitants, of
whom 8,419,672 were free (84.7 per cent) and 1,510,806 remained slaves
(15.2 per cent). With regard to racial composition, 38.1 per cent were white,
19.7 per cent black, 38.3 per cent pardos (mixed race) and 3.9 per cent
Indian. Blacks and pardos together, including all social conditions (free,
freed, and slave) comprised 58 per cent of the total population (5,756,234
people). Slaves excluded, the free population of 4,245,428 people of African
descent represented 42.7 per cent of the inhabitants of the country, just
about the same number of people deemed white in the 1872 census.2
In Cuba, according to historian Rebecca Scott, the total population
reached 1,389,880 in the early 1860s, of whom 57.1 per cent were white,
26.6 per cent slaves, and only 16.2 per cent free blacks – a sharp contrast
with the 42 per cent in this category in Brazil at about the same time.
1. For a previous presentation of the concept of the ‘‘structural precariousness of freedom’’, see
Sidney Chalhoub, ‘‘Precariedade estrutural: o problema da liberdade no Brasil escravista (século
XIX)’’, História Social: Revista dos Pós-graduandos em História da UNICAMP, 19 (2010), pp. 33–62.
The present article is a considerably revised and enlarged version of that earlier piece.
2. Nelson de Castro Senra, História das estatı´sticas brasileiras (Rio de Janeiro, 2006), I,
pp. 418–419, 423. Concerning the status and reliability of the 1872 census, it is worth noting
that census tables yield incorrect figures quite often. In addition, masters provided the infor-
mation regarding their slaves, and it seems that they did not care to give accurate information,
for example, about their captives’ ages and places of birth. With respect to the slave population,
the 1872 slave registration or matrıćula appears more accurate. In general, however, demo-
graphers deem the 1872 census reasonably reliable, especially in comparison with the ones that
followed it in 1890 and 1900. My thanks to Robert Slenes for these clarifications; for further
details, see Robert W. Slenes, ‘‘The Demography and Economics of Brazilian Slavery:
1850–1888’’ (unpublished Ph.D. thesis, Stanford University, 1976), passim.
Precariousness of Freedom in Nineteenth-Century Brazil 407
BRAZIL
Recife
PERNAMBUCO
BAHIA
Salvador
MINAS GERAIS
Ouro Preto
SÃO PAULO RIO DE JANEIRO
Valença Cantagalo
Campinas
Rio de Janeiro
Sao Paulo
Figure 1. Nineteenth-century Brazil: provinces and cities cited in the text.
408 Sidney Chalhoub
According to the same author, in the state of Louisiana, in 1860, just

before the beginning of the American Civil War, the population consisted
of 708,002 inhabitants, of whom 50.5 per cent were white, 46.8 per cent
slaves, and just 2.6 per cent free people of color, and those were very
concentrated in the city of New Orleans.3
Another way of pondering this sort of data is to focus on the relative
presence of free people of color in the total population of African descent.
Thus, in 1872, 73.7 per cent of blacks and pardos living in Brazil were
free.4 In the United States, in 1860, Ira Berlin found that no more than
11 per cent of the population of African descent were free; this figure
included data from the northern states, where slavery had virtually dis-
appeared and which registered 99 per cent of their colored population as
free. Regional differences remained important in southern slavery – in the
Upper South (Maryland, Virginia, North Carolina, etc.), 13 per cent of the
people of African descent were free; in the Lower South (South Carolina,
Georgia, Florida), 2 per cent; in the Deep South (Louisiana, Alabama,
Mississippi, Arkansas, Texas), 1 per cent.5
These brief comparisons of population data for Brazil, Cuba, and the US
in the nineteenth century lead to two observations. First, slave societies in
the Americas seemed extremely diverse with respect to opportunities for
obtaining manumission. In the US South, the overall tendency was for masters
to lose the right to grant freedom as slavery expanded and the national split
over it deepened.6 In some states, as in Alabama, for example, masters who
wished to manumit slaves had to petition the state legislature; as a con-
sequence, in the period from 1829 to 1839, the local assembly authorized the
granting of freedom to approximately 200 slaves, while Alabama had 120,000
people held in bondage in 1830. In the 1850s, several states prohibited masters
from freeing slaves altogether.7 In Dutch Surinam, from the 1730s to final
emancipation in 1863, slave owners needed to submit requests to concede
freedom to a manumission court. According to the historian Brana-Shute, in
any given year, from the 1760s to the 1820s, the number of slaves manumitted
never exceeded the proportion of 0.8 per cent of the total slave population.8
In Brazil, although achieving freedom remained difficult for as long
as slavery existed, the state never took the initiative of forbidding owners to
manumit their slaves. On the contrary, the masters’ exclusive prerogative
3. Rebecca J. Scott, Degrees of Freedom: Louisiana and Cuba after Slavery (Cambridge, MA
[etc.], 2005), pp. 273–274.
4. Senra, História das estatı´sticas brasileiras, p. 423.
5. Ira Berlin, Generations of Captivity: A History of African-American Slaves (Cambridge, MA
[etc.], 2003), pp. 278–279.
6. Idem, Slaves without Masters: The Free Negro in the Antebellum South (Oxford, 1974), p. 138.
7. Ibid., pp. 138, 140–141.
8. Rosemary Brana-Shute, ‘‘Slave Manumission in Suriname, 1760–1828’’, Slavery and Abolition,
10 (1989), pp. 40–63, 44.
to decide whether, when, or how to free their slaves constituted a major

aspect of social control within Brazilian slavery; as a consequence, the
seigneurial class resisted and resented every government attempt to create
ways for slaves to obtain freedom irrespective of the masters’ will.9 In 1871,
a law that declared free the children born to slave women after that date
and gave slaves the right to freedom by self-purchase, regardless of their
masters’ consent, infuriated coffee planters but won approval in parliament
after months of political turmoil.10
In any case, the data gathered by historian Robert Slenes, based on the
two national slave registrations, in 1872–1873 and 1886–1887, show that
manumission rates varied enormously in different regions of the country. In
the city of Rio de Janeiro, for example, access to freedom was significant:
36.1 per cent of slaves listed in the 1872–1873 slave registration had
obtained freedom by 1886–1887. The situation in the capital contrasted
sharply with areas dominated by coffee cultivation. Hence, 11 per cent of
captives held in 1872–1873 had been manumitted in the province of
São Paulo by 1886–1887; 7.8 per cent in the province of Rio de Janeiro; just
5.6 per cent in Minas Gerais.11
The second observation derived from these numbers is that in Brazil, to
a greater extent than in other slave societies in the Americas, slave
emancipation occurred concomitantly with the continuity of slavery
itself, leading to the figure, just mentioned, of about 74 per cent of people
of African descent being free or freed in 1872. In this context, historians’
traditional focus on ways of obtaining freedom in nineteenth-century
Brazil must be balanced by further attention to the experience of freedom
for this ever-growing contingent of ex-slaves and their descendants. In
addition to the widespread practice of illegal enslavement, there existed
several legally sanctioned situations – such as conditional manumissions
and revocation of freedoms – that often made the boundaries between
slavery and freedom uncertain, thus constituting a structural feature of
that society, conducive to strategies for the control of workers, slave and
free, based on personal dependence and paternalist ideology.
9. Manoela Carneiro da Cunha, Antropologia do Brasil: mito, história, etnicidade (São Paulo,
1986), pp. 123–144.
10. Sidney Chalhoub, Machado de Assis, historiador (São Paulo, 2003), ch. 4.
11. Slenes, ‘‘Demography and Economics of Brazilian Slavery’’, pp. 495, 501, 504, 542. The data
on manumission presented by Slenes refer to a period of decline of slavery in Brazil. Because of
the lack of a national census and a national slave registration prior to 1872, it is not possible to
compare the rates found by Slenes with those pertaining to earlier periods. Nonetheless, for the
sake of my argument here, it suffices to say that recent studies continue to confirm that
manumission rates in Brazil have consistently guaranteed the presence of a relatively high
population of free and freed people of African descent in the country during the nineteenth
century. For a survey of the literature, see Herbert S. Klein and Francisco Vidal Luna, Slavery in
Brazil (Cambridge, 2010), ch. 9, especially pp. 267–269.
410 Sidney Chalhoub
Figure 2. Portrait of a couple; carte de visite; photographer: Militão Augusto de Azevedo,

São Paulo, c.1879.
Museu Paulista, São Paulo. Used with permission.
S L AV E RY A N D F R E E D O M I N B R A Z I L I A N
HISTORIOGRAPHY
Until recently, historians seemed to share a paradigm, so to speak, for the
study of slavery in Brazil and perhaps in the Atlantic world in general. In
the mid-1980s, Rebecca Scott described the frame of mind prevailing in
the field, stressing that studies on the history of slavery had overcome
the common association between subordination and passivity. In other
words, historians had learned to consider slave initiatives within the
system that oppressed them, and to explore the creation of alternative
values and beliefs despite the burdens of economic exploitation and
social and ideological domination.12 It appeared possible to bring slave
communities to life in historical narratives, largely because there was
agreement regarding what constituted the politics of dominion in slave
societies, and to approach the question of what slaves did with what was
done to them – just to retread the battle cry of social historians in those
glorious times.
The politics of dominion in slave societies centered upon three aspects:
the slave market, with its transformations over time and impact upon
slave communities; ideologies and practices with regard to manumission
and the constitution of slave families;13 and, last but not least, physical
punishment, then seen not just as a reaffirmation of violence as the cor-
nerstone of slave discipline, but as happening in the context of class
struggle, somewhat limited by slaves’ pressures and concepts about fair
treatment and tolerable punishment.14 More recently, the effort to pursue
slaves’ views of slavery in Brazil, where African-born captives comprised
the majority of the slave labor force in most regions until past the
mid-century, entailed seeking the forms of re-appropriation and recom-
bination of cultural legacies in the context of New World slavery. People
who arrived in Brazil through the slave trade ‘‘discovered’’ Africa in New
World slavery, meaning that their common fate made them aware of the
connections and shared assumptions pertaining to their otherwise unique
ethnic legacies.15
12. Rebecca J. Scott, ‘‘Exploring the Meaning of Freedom: Post-Emancipation Societies in

Comparative Perspective’’, Hispanic American Historical Review, 68 (1988), pp. 407–428.
13. Robert W. Slenes, Na senzala, uma flor: esperanças e recordações na formação da famı´lia
escrava (Brasil sudeste, século XIX) (Rio de Janeiro, 1999); Sidney Chalhoub, Visões da liber-
dade: uma história das últimas décadas da escravidão na Corte (São Paulo, 1990).
14. Silvia H. Lara, Campos da violência: escravos e senhores na capitania do Rio de Janeiro,
1750–1808 (Rio de Janeiro, 1988).
15. Robert W. Slenes, ‘‘‘Malungu, ngoma vem!’: África coberta e descoberta no Brasil’’, in idem,
Mostra do redescobrimento: negro de corpo e alma – Black in body and soul (São Paulo, 2000),
pp. 212–220; also by Robert W. Slenes, ‘‘The Great Porpoise-Skull Strike: Central African Water
Spirits and Slave Identity in Early Nineteenth-Century Rio de Janeiro’’, in L. Heywood (ed.),
412 Sidney Chalhoub
In a critical essay still to appear, Robert Slenes remarks that divergent

perspectives now prevail among historians regarding the slave family
and manumission.16 On the one hand, authors such as Manolo Florentino
and José Roberto Góes consider slavery in Brazil to have been char-
acterized by significant social mobility: a considerable number of slaves
achieved freedom consistently over time; freed people often had access to
land, either as tenants or owners of a few acres, also becoming masters of
slaves.17 These openings for social mobility weakened horizontal solida-
rities among captives, making rivalries and competition among them
rather common, therefore strengthening seigneurial domination. In other
words, owners seemed more able to impose their rule on slave quarters
divided ethnically and appreciative of the opportunities for the formation
of family and kinship ties – resistance ‘‘pacified’’, as Florentino and Góes
would have it. Slenes deems such interpretation static and culturalizing,
failing to account for historical change, although recognizing the
oppressive nature of the slave system to a certain degree.
On the other hand, there are historians who emphasize social trans-
formation and patterns of everyday conflicts as determining factors for
the political views and strategies of both masters and slaves. Hebe Mattos,
for example, argues that slaves’ and freed people’s expectations concern-
ing social mobility changed considerably over the course of the nineteenth
century.18 In the decades following Independence (the 1820s to 1850s), the
expansion of coffee cultivation intensified the competition for land,
making its acquisition increasingly difficult for free and freed people of
color. Furthermore, because of the rise in prices after the cessation of the
slave trade in 1850, poor people in general, including freed persons, had
their access to slave property, a prerequisite for capital accumulation and
social mobility in that society, significantly reduced.19 As the century
wore on, the concentration of slave property in fewer hands and the
Central Africans and Cultural Transformations in the American Diaspora (Cambridge, 2001),
pp. 183–208, and Robert W. Slenes, ‘‘Saint Anthony in the Crossroads in Kongo and Brazil:
‘Creolization’ and Identity Politics in the Black South Atlantic, ca. 1700–1850’’, in L. Sansone,
E. Soumonni, and B. Barry (eds), Africa, Brazil and the Construction of Trans-Atlantic Black
Identities (Trenton, NJ, 2008), pp. 209–254.
16. Robert W. Slenes, ‘‘A ‘Great Arch’ Descending: Manumission Rates, Subaltern Social
Mobility and Slave and Free(d) Black Identities in Southeastern Brazil, 1791–1888’’, in John
Gledhill and Patience Schell (eds), Rethinking Histories of Resistance in Brazil and Mexico
(Durham, NC, forthcoming).
17. Manolo Florentino and José Roberto Góes, A paz das senzalas: famı´lias escravas e tráfico
atlântico, Rio de Janeiro, c.1790–c.1850 (Rio de Janeiro, 1997).
18. Hebe M. Mattos de Castro, Das cores do silêncio: os significados da liberdade no sudeste
escravista – Brasil, século XIX (Rio de Janeiro, 1995).
19. See also Zephyr Frank, Dutra’s World: Wealth and Family in Nineteenth-Century Rio de
Janeiro (Albuquerque, NM, 2004).
massive transfer of captives from the northern to the southern coffee-

growing provinces deprived the institution of slavery of the broad social
backing it had enjoyed in the first half of the nineteenth century.
In making social mobility the crux of their interpretations, these his-
torical perspectives, however different, assume the existence of areas of
uncertainty or blurring between slavery and freedom in nineteenth-century
Brazil. Nonetheless, these indefinite boundaries are conceived of mainly as
continuous or interrupted opportunities for upward social mobility. In
other words, there is a shared lack of emphasis – more so for Florentino
and Góes than for Mattos – on the conditions that rendered freedom
precarious, and also on how such precariousness changed over time. To put
it bluntly, the pressing insecurity of their condition shaped free and freed
blacks’ expectations and daily strategies at least as much as their often
protracted hopes for upward social mobility.
CITIZENSHIP RIGHTS
The Constitution of 1824 considered freedmen born in Brazil to be citizens.
It did not mention African-born freed people, who were numerous because
of the high percentage of Africans among slaves for as long as the slave trade
continued. It was clear, however, that an African captive, when freed,
remained a foreigner, thus eligible for citizenship only through naturaliza-
tion. Perdigão Malheiro, in his influential study of the legal aspects of
Brazilian slavery published in the 1860s, said that there existed no legislation
to impede the naturalization of Africans, although the underlying assump-
tion of the comment was that such a possibility would be unexpected or
unwelcome.20 Police documents pertaining to the city of Rio de Janeiro,
from the 1830s to the 1860s, show that there was great hostility toward freed
Africans, often suspected of luring slaves to run away, planning rebellions,
and having an inclination toward vagrancy and crime.21 The police often
had them summarily deported to Africa, especially in the aftermath of slave
rebellions such as the one in Bahia in 1835.22
Regarding the political and civil rights of Brazilian freedmen, Perdigão
Malheiro thought that their condition ‘‘in our society is highly restricted’’
because of ‘‘a generalized racial prejudice against the African race’’.23 In
the electoral system established by the Constitution, with two principal
stages of citizenship rights mainly on the basis of wealth, freedmen could
20. Perdigão Malheiro, A escravidão no Brasil: ensaio histórico, jurı´dico, social, 2 vols (Petrópolis,
1976), I, p. 141. The first edition appeared in 1866 and 1867.
21. See, for example, Arquivo Nacional, Rio de Janeiro [hereafter AN], Secretaria de Polı́cia da
Corte, Ofı́cios com Anexos (1837), maço IJ6–174.
22. João José Reis, Rebelião escrava no Brasil: a história do levante dos malês em 1835
(São Paulo, 2003).
23. Malheiro, A escravidão no Brasil, I, p. 141.
414 Sidney Chalhoub
vote in primary or local elections. They could vote for and be elected to
the municipal council, but they were not allowed to participate in the
following stage, the election of provincial deputies, general deputies, and
senators. Since citizenship rights also served as criteria for the occupation
of jobs in the state bureaucracy, a freedman could not be appointed a
justice of the peace, ranked police officer, attorney, judge, diplomat, or
bishop – they were even barred from serving as jurors in jury trials. They
could be admitted to the National Guard, but not as officers.24 As for civil
rights, still following Perdigão Malheiro, manumission gave freed people
the condition of persons, ‘‘who could exercise freely their rights and
activities, according to the law, as could other citizens, form a family,
acquire property’’ etc. (emphasis in the original). He compared the
situation of freed persons to that of minors coming of age.25
The 1832 code of criminal procedure gave freed Africans and slaves the
same treatment concerning travel within the country: they had to have a
passport, even when accompanied by their masters or protectors.26 At local
level, restrictions on freed Africans’ economic and social activities were
commonplace. In Salvador, in the province of Bahia, a city ordinance issued
in 1859 established a fine for slaves caught on the streets at night who did
not carry the written permission of their masters; likewise, freed Africans
would be fined in the amount of 3 mil-réis (worth at least three days work),
or eight days in prison, if found on the streets at night without a note ‘‘from
any Brazilian citizen’’.27 In other words, an African freed person could not
move about the city freely at certain times without the protection of a
citizen, someone willing to vouch for his or her conduct.
The example from Salvador comes from historian João Reis’s recent book
on the life of Domingos Sodré, an African slave, later freed and becoming
the owner of slaves, who lived in the city in the years following the
1835 rebellion. Restrictions on the economic activities of freed Africans
increased in the period. It seemed that local authorities intended to keep
freed Africans under constant pressure, perhaps to convince at least some of
them to cease living in Bahia, preferably assuming for themselves the cost
24. See also Keila Grinberg, O fiador dos brasileiros: cidadania, escravidão e direito civil no
tempo de Antonio Pereira Rebouças (Rio de Janeiro, 2002).
25. Malheiro, A escravidão no Brasil, I, pp. 141–143.
26. Actually, the law just said that citizens were exempted from having passports for travel
within the country; afterwards the government ‘‘interpreted’’ the law, to make it clear that non
citizens in general – freed Africans included – needed passports; Collecção das leis do Imperio do
Brasil, law of 29 November 1832, articles 118–120; for the interpretation, AN, Secretaria de
Polı́cia da Corte, Ofı́cios com Anexos (1835), Minister of Justice to the Chief of Police,
18 March 1835, maço IJ6–170.
27. In Portuguese, ‘‘que não levarem bilhetes de qualquer cidadão brasileiro’’; João J. Reis,
Domingos Sodré, um sacerdote africano: escravidão, liberdade e candomblé na Bahia do século
XIX (São Paulo, 2008), p. 88.
of a return to their native continent. A law enacted in May 1835 almost

went so far as asking freed Africans to leave, for it created an annual tax of
10 mil-réis that applied only to them if they wished to continue residing in
Salvador. The same law prohibited them from acquiring any real estate,
thus leaving slave property as their only option for investing personal
savings. João Reis sums up the situation saying that, ‘‘respecting Africans, a
tenuous line separated the condition of freed from that of slave’’.28
As for Brazilian freed people’s political rights, a serious setback occurred
in the 1880s. A new electoral law enacted in January 1881 abolished the
different stages of political rights established in the Constitution of 1824. In
theory, Brazilian freedmen could now vote and be elected to the same
offices open to other citizens, without exception; in addition, all jobs in the
state bureaucracy became a possibility for them. However, the new law
raised income requirements significantly; even worse, it made mandatory
that voters be able to read and write, a novelty well attuned to the Western
liberal tenets of the times.29 But such a law was enacted in a country where
blacks attending school seemed a matter for the police to deal with. On
27 March 1835, the chief of police of the city of Rio sent a ‘‘reserved’’ letter
to the justice of the peace of the parish of Santana asking him to investigate
meetings taking place at a given house in the neighborhood, ‘‘allegedly a
school for Mina blacks [people from West Africa] to learn how to read and
write’’. The police chief added that the Africans gathered everyday in the
afternoon, therefore the justice of the peace had to prepare a ‘‘detailed’’
report on their activities ‘‘urgently’’.30
Given the year of this occurrence, 1835, the chief of police may have
taken action on this gathering of Africans not because of the alleged
purpose of their meetings – learning how to read and write – but because
he suspected that they might be plotting an insurrection or something else
surely more threatening than elementary school. Furthermore, since the
Africans involved were identified as Mina blacks, it is probable that
the police feared that they were engaged in Quranic lessons, an activity
seen as dangerous because the authorities thought that the 1835 rebels in
Salvador communicated in Arabic amongst themselves. There was no law
prohibiting teaching slaves to read and write in Brazil; nonetheless, there
can be no doubt of the total lack of interest of Brazilian masters in the
literacy of their slaves, for security concerns or whatever other reason.
The census of 1872 registered no slaves at all attending school in the
whole country. However, the census found that 1,401 captives knew how
to read and write, not an impressive number given a slave population of
28. Ibid, p. 92.

29. Chalhoub, Machado de Assis, historiador, pp. 281–288.
30. AN, Registro de Correspondência Reservada Expedida pela Polı́cia (1835–1844), códice
335, I, fos 2–3.
416 Sidney Chalhoub
more than one and a half million – that is, 0.08 per cent of slaves were
literate.31 In the city of Rio, where the situation regarding literacy seemed
much better than in the rest of the country, 329 captives were able to
read and write in a total slave population of more than 48,000 – that is,
0.7 per cent of captives were literate in the capital. Data about literacy in
the 1872 census appeared appalling in general, for it became known that
for every 100 inhabitants, 81.4 were illiterate; by excluding children under 5,
too young to go to school, one comes to a figure of 77.4 illiterate per 100.
The city of Rio was the only place where more than half the population
could read and write – and just barely, 50.1 per cent.
Apparently shocked by the census results, politicians devoted much
time and talk to the question of elementary schooling in the following
years, but no meaningful action ensued. Instead, the literacy requirements
established by the electoral law of 1881 curtailed decisively the citizenship
rights of generations of people of African descent that had become or
were about to become free at the time. The perverse combination of
literacy requirements for voting rights and lack of elementary school for
the poor continued for decades to come.32
CONDITIONAL MANUMISSIONS,
R E V O C AT I O N O F F R E E D O M S
There were diverse ways to obtain freedom in Brazilian slavery: letters
of manumission; freedoms conceded in last will and testaments and post-
mortem inventories;33 liberties granted at the baptismal font; and manu-
missions achieved by judicial means. Manumissions could be gratuitous or
unconditional, in which case the freed person had no further obligations
toward the ex-proprietor (other than gratitude and due respect); and
onerous or conditional, which lumped together a variety of situations,
31. As I said before, masters provided information about their slaves to the 1872 census takers;
therefore, it is possible that they did not care to report correctly the literacy of their slaves.
Regardless of that, there is no reason to believe that the rate of literacy was in any way
significant among the slave population in Brazil. It is perhaps worth mentioning that in almost
three decades of research in the archives of Brazilian slavery, especially in police papers and
criminal and civil trial records, I have found the signature of one single slave – his name was
Luiz.
32. For the data presented here, see Diretoria Geral de Estatı́stica, Recenseamento Geral do
Brazil de 1872, and Senra, História das estatı´sticas brasileiras, pp. 428–431.
33. This clear-cut rendering of paths to freedom in Brazilian slavery is not unproblematic. For
example, bequests of freedom in wills often made their way to the notary books as manu-
mission letters. In addition, post-mortem inventories normally carried out what was established
in wills and testaments, and thus it may make sense to comprise all of those in the category of
liberties given by bequests. A further complication is that at times a slave bought his or her
freedom for the price of his or her evaluation during the post-mortem inventory proceedings,
an obvious instance of self-purchase rather than manumission by bequest.
ranging from the establishment of conditions such as continued service

for a number of years – often seven – or until the death of the owner, to
self-purchase, which could be done in instalments paid over the years.
Self-purchase was different from other types of conditional manumission
because the initiative belonged to the slave, or to a third party acting on
her or his behalf.34 A vast area of untapped research concerns the relation
between the type of manumission obtained by a freed person and his or
her degree of autonomy vis-à-vis the ex-master thereafter.
It seems that conditional freedoms associated with continued service
constituted a sizable number of all manumissions in nineteenth-century
Brazil. In his classic piece about the subject in Campinas, Peter Eisenberg
computed letters of manumission for the period 1798 to 1888, and found
that, despite variations over time, liberties granted upon the requirement
of continued service for a time consistently represented about 50 per cent
of cases. Self-purchase appeared in another 30 per cent of the letters.35 In a
more recent study, also about Campinas, Lizandra Ferraz analyzed
manumissions in last will and testaments and post-mortem inventories, to
discover that conditional freedoms comprised 30 per cent of the sample in
the period 1836–1845, then reached 34 per cent in 1860–1871.36 These
studies suggest that there may be variations in the results obtained
depending on the historical sources used: for example, in Campinas,
conditional manumissions requiring continued service were more com-
mon in letters of freedom than in last will and testaments and post-
mortem inventories. More important, though, for my purposes here, is
that manumissions requiring a condition to be fulfilled by freed persons
represented a considerable percentage of freedoms granted whatever the
historical source analyzed, thus launching freed people on to murky
terrain between slavery and freedom, to be shaped by a process of
negotiation and conflict with their masters or patrons.
Klein and Luna collected the results of more than twenty studies on
manumission in Brazil, about different periods and regions, organized
them in tables that are very useful for comparative purposes, and there we
find that, despite variations to be expected given the diversity of situa-
tions, conditional manumission was significant everywhere, at any time,
ranging from about 20 per cent to 70 per cent of cases in the samples, but
more typically representing around 30 or 40 per cent of manumissions.37
34. Manumission studies have been a vast sub-field in slavery studies in Brazil for decades; for a
recent review of the literature, see Klein and Vidal Luna, Slavery in Brazil, ch. 9.
35. Peter L. Eisenberg, Homens esquecidos: escravos e trabalhadores livres no Brasil, séculos
XVIII e XIX (Campinas, 1989), pp. 255–314, 282.
36. Lizandra Meyer Ferraz, ‘‘Entradas para a liberdade: formas e freqüência da alforria em
Campinas no século XIX’’ (M.A. thesis, UNICAMP, 2010), pp. 122–123.
37. Klein and Vidal Luna, Slavery in Brazil, p. 261.
418 Sidney Chalhoub
As to what conditional manumission meant for freed people, whether it

entailed a situation markedly different from slavery or not, little is yet
known, with historical sources seemingly open to the most contradictory
interpretations, perhaps because the situation remained very indetermi-
nate for masters and slaves, patrons, and the freed persons themselves. For
instance, legal controversies abounded on the condition of the child of a
freed woman who had been granted a letter of manumission but had yet
to fulfill the requirements there established. Did she remain a captive until
serving fully the period of further labor demanded of her, in which case
her child had been born in bondage? Or was it otherwise, with freed
women in the same position as minors, awaiting full emancipation, but
then giving birth to free children?38
Manolo Florentino’s study on manumission in the city of Rio from 1789
to 1871 shows that important changes occurred over time concerning the
relative importance of each type of manumission. In periods in which
freedom by self-purchase appeared to be more accessible, meaning that
captives were having more economic opportunities and therefore saving
more to gain liberty, the tendency was to see more people in general
becoming freed, increasing likewise the presence of freed people in the total
population of the capital. In other words, only freedom by self-purchase
guaranteed periods of significant increase in the rate of manumissions,
with other types of liberty merely sustaining the proportion of freed
people in relation to the total population of the city.39 It is also possible
that freed persons granted conditional manumissions remained more
vulnerable to re-enslavement, since slaves who succeeded in buying
themselves out of bondage probably strengthened their position vis-à-vis
former masters – maybe they enjoyed a higher degree of autonomy than
people with lingering obligations. Also, freedom by self-purchase may
have guaranteed more autonomy than voluntary, unconditional manu-
mission, with the recipients of the latter having to show eternal, unabated
gratitude to their benefactors.
Revoking freedoms by alleging ingratitude on the part of an ex-slave
remained a master’s right until the enactment of the Law of the Free
Womb, in 1871. Until recently, students of Brazilian slavery thought that
revocations of freedom were rare. Not any more. Keila Grinberg has
been suggesting that historians, myself included, have misread, at least in
part, the many slaves’ appeals for freedom existent in Brazilian archives.
38. Malheiro, A escravidão no Brasil, I, pp. 117–121; Chalhoub, Visões da liberdade, pp. 122–131;
Eduardo Spiller Pena, Pajens da casa imperial: jurisconsultos, escravidão e a lei de 1871 (Campinas,
2001), pp. 79–88.
39. Manolo Florentino, ‘‘Sobre minas, crioulos e a liberdade costumeira no Rio de Janeiro,
1789–1871’’, in idem (ed.), Tráfico, cativeiro e liberdade. Rio de Janeiro, séculos XVII–XIX
(Rio de Janeiro, 2005), pp. 331–359.
Studies using judicial records have focused upon slaves’ initiatives to file
for freedom in court, their ability to find allies for their causes, the
arguments mustered to forward their legal cases.40 However correct this
approach, it has missed evidence that often times in these lawsuits an
ex-slave was seeking to defend a condition of freedom that he or she had
obtained in the past, and had been enjoying when the legal case started –
in other words, the freed person was actually struggling against an
attempt to re-enslave him or her. According to Grinberg, in a sample of
402 lawsuits to decide the bondage or freedom of one or more persons
pertaining to the Higher Court of Rio (Tribunal da Relação do Rio de
Janeiro), 27 per cent consisted of appeals to re-enslave someone, or for
someone to maintain freedom. As is obvious, in appeals for re-enslave-
ment, the ex-master or his/her heir initiated the case, hoping to regain
property rights over a given person; in appeals to maintain freedom, freed
people took action to resist an ongoing attempt to return them to bondage.
Grinberg’s study hints that captives who had received conditional manu-
mission were more likely to face the threat of revocation on the supposed
grounds of ingratitude.41
Grinberg’s work details legal arguments put forward in appeals to
maintain freedom and in attempts to revoke manumissions, and computes
their results, seeking to evaluate freed people’s chances of confirming their
condition through these means, and comparing what she finds in the
judicial cases with discussions conducted in law books and periodicals.
Although the point is not developed in her work, it suggests the need to
relate the different forms of manumission to the frequency and content of
judicial cases motivated by attempts to re-enslave freed persons.
The following extract, dated 28 July 1870, tells the story of Augusta, a
woman who sought legal protection for her liberty:
The black Augusta, a Brazilian, says that about eight years ago, the Lieutenant
Colonel José Rodrigues Gonçalves Valle freed her by letter, with the condition
to accompany him during his life.
Upon the Lieutenant Colonel’s death on 8 November 1864 (see attached
document), the Plaintiff left the house and separated her savings, as a free
person, that she is, and in this state has remained ever since, without con-
testation from anyone. Recently, however, João Carlos Álvares Valle, who
alleges to be in charge of the post-mortem inventory of the property of the
40. For examples, see Chalhoub, Visões da liberdade; Keila Grinberg, O fiador dos brasileiros,
and Liberata, a lei da ambiguidade: as ações de liberdade da Corte de Apelação do Rio de
Janeiro no século XIX (Rio de Janeiro, 1994).
41. Keila Grinberg, ‘‘Senhores sem escravos: a propósito das ações de escravidão no Brasil
imperial’’, in José Murilo de Carvalho and Lúcia Maria Bastos Pereira das Neves (eds),
Repensando o Brasil do Oitocentos (Rio de Janeiro, 2009), pp. 415–435. Actually, historian Hebe
Mattos had remarked the importance of appeals to maintain freedom in Das cores do silêncio,
ch. 9, p. 200.
420 Sidney Chalhoub
deceased Lieutenant Colonel, has intended to consider the Plaintiff a slave

pertaining to the inheritance, seeking to force her to pay him a daily sum, which
is a violation against the freedom that the Plaintiff has legitimately acquired, and
has possessed for more than five years.
And because the Plaintiff’s letter of liberty, given by the deceased Lieutenant
Colonel, has disappeared, the Plaintiff wants to justify her allegations with the
depositions of persons who saw the mentioned letter, or know of its existence,
in order to obtain a ruling in her favor.42
Augusta’s petition suggests that she might be having trouble with the
heirs because she was not able to present written proof of the arrange-
ments she had made with her ex-master. The fact that the freedom granted
her had been pending on a condition may have weakened her position as
well. But the judge deemed reliable the witnesses presented by Augusta
and decided to confirm her freedom. She succeeded because she was able
to enroll people willing to confirm her story, thus showing that the
condition had been fulfilled, that the letter of manumission existed, and
that the heirs had not questioned her condition for years, therefore tacitly
recognizing that she had a right to freedom. Furthermore, the judge’s
ruling ‘‘maintaining her in possession of her freedom’’ (‘‘manutenida na
posse de sua liberdade’’) provided her with a document, a valuable piece of
paper to use thereafter in defense of her freedom.
Actually, it seems that often times freed people filed petitions to
‘‘maintain the possession’’ of their freedom to create a paper trail of their
condition, as was emphatically asserted by the ‘‘pardo Matias’’, who once
avoided an arrest demanded by his alleged master waving a certified copy
of a judge’s ruling of this sort to police officials about to detain him.43
I L L E G A L E N S L AV E M E N T
On 7 November 1831, as a consequence of a treaty signed with Great
Britain associated with the formal recognition of independence, the
Brazilian government enacted a law prohibiting the slave trade. Such law
was systematically disrespected for the following two decades, a period
during which about 750,000 Africans were smuggled into the country,44
therefore making illegal enslavement a major source of slave labor for the
rapid expansion of coffee in the south-eastern provinces. Notwithstand-
ing its failure to abolish the slave trade, recent research has been exploring
precisely the political and social consequences of such law, how non-
compliance with it impacted relations between masters and slaves, and
public policies regarding slave emancipation.
42. AN, processo cı́vel (1870), Segunda Vara Cı́vel da Corte, no. 3395, maço 849, galeria A.
43. AN, processo cı́vel (1848), Tribunal da Relação do Rio de Janeiro, no. 4, caixa 42.
44. Data from www.slavevoyages.org; accessed 15 November 2010.
Figure 3. Portrait of Glyceria da Conceição Ferreira; carte de visite; Carneiro & Gaspar Studio,
Rio de Janeiro, c.1872.
Arquivo Nacional, Rio de Janeiro. Used with permission.
422 Sidney Chalhoub
There were moments in which it appeared possible that the 1831 law
could become decisive in slaves’ strategies for achieving freedom, threa-
tening to bring to court the question of the illegality of much of the slave
property existent in the country. According to historian Elciene Azevedo,
the black abolitionist Luiz Gama attempted just that in the 1860s,
a decade during which the British also kept the imperial government
under pressure to resolve the question of the africanos livres – liberated
Africans – who were those Africans apprehended in ships involved in the
contraband trade, declared free and put under the protection of the
Brazilian government, which in its turn exploited their services in public
works and hired them out to private citizens. Luiz Gama rightly argued in
judicial petitions – although to no avail – that the concept of africanos
livres should encompass all Africans imported after the enactment of the
law of 1831, and not just those recovered by British warships and given to
the imperial government.45 Beatriz Mamigonian concluded her detailed
study of africanos livres reflecting on the question of how the experience
of these few thousand Africans may have influenced the strategies and
visions of slavery and freedom of others, hundreds of thousands, introduced
by contraband and held illegally under bondage.46
Mamigonian returned to the theme more recently, to ask the following
question: ‘‘Considering that the illegality of the slave property originated in
importations occurred after 1831 was uncontroversial, how did slave owners
sustain their right to such property?’’47 The author examined resolutions by
the Council of State, annals of proceedings and debates of the Senate,
ministerial reports, and other official documents to surmise that politicians
and public officials upheld the enslavement of Africans arrived after 1831
because such conduct was legitimated by custom – that is, planters had
traditionally had their access to slave labor guaranteed, a seigneurial right so
much recognized that the illegal enslavement of Africans seemed to count on
the connivance of public authorities at all levels and to enjoy the active
support of several sectors of the population. Nevertheless, there remained
the problem of British pressure to broaden the meaning of africanos livres,
with Her Majesty’s diplomats in Rio eventually taking initiatives to
denounce the open sale, sometimes in highly publicized auctions, of Africans
whose ages made obvious their illegal importation.
45. Elciene Azevedo, Orfeu de carapinha: a trajetória de Luiz Gama na imperial cidade de São
Paulo (Campinas, 1999).
46. Beatriz Galotti Mamigonian, ‘‘To Be a Liberated African in Brazil: Labour and Citizenship
in the Nineteenth Century’’ (unpublished Ph.D. dissertation, University of Waterloo, Canada,
2002). Beatriz Mamigonian and Keila Grinberg have recently edited a collection of articles on
the social and political aspects of the law of 1831; Estudos Afro-Asiáticos, 29 (2007), pp. 87–340.
47. Beatriz Galotti Mamigonian, ‘‘O direito de ser africano livre: os escravos e as interpretações
da lei de 1831’’, in Silvia H. Lara and Joseli M. N. Mendonça (eds), Direitos e justiças no Brasil.
Ensaios de história social (Campinas, 2006), pp. 129–160, 132.
In 1854, for example, British diplomats repeatedly complained of African

slaves advertised for sale in local papers with the ages of nineteen or twenty,
sometimes younger, rendering impossible their legal arrival before 1831.
When the Minister of Justice asked the police chief for explanations, the
answer he got was that the ages were incorrect because sellers had the habit
of making their captives seem younger to obtain better prices for them; the
chief explained that he had brought the Africans to his presence, had talked
to them, had seen that they spoke Portuguese as ladinos (expected of
Africans who had arrived years before), and ascertaining that they looked
older than the ages advertised for them. But the police chief proceeded to
remark that if he were supposed to
[y] inquire into these cases, of slaves announced for sale with 24, 25 years, just
because they may have been imported after the law of 1831, then the same
supposition could be made about other Africans, however their ages, 5, 10, 15
and older, because they too could have been imported after the aforementioned
law, and it does not seem to me that our present circumstances make such
investigations convenient.48
As can be seen, there were serious risks involved in non-compliance
with the law of 1831, risks that sometimes materialized in British initia-
tives and judicial appeals, forcing authorities to be vigilant to silence such
attempts and deny them legal due process.
Concerning the slave owners approach to the law of 1831, it was surely
cautious at the beginning, with importations dropping dramatically from
almost 73,000 in 1829, when the trade was still legal, to little more than
6,000 in 183149 – according to the 1826 treaty between Brazil and Great
Britain, the trade was supposed to have stopped by March 1830. However,
the expansion of coffee, class solidarity, and the strengthening of the hold
of conservative politicians in central government emboldened planters,50
with the trade back to its full strength by 1837, when the daring was such
that a proposal reached the Senate floor to revoke the law of 1831 and
legalize all slave property acquired against its determinations.51 The
proposal to revoke the law met strident parliamentary opposition in both
48. AN, Secretaria de Polı́cia da Corte, Ofı́cios com Anexos (August–December 1854), Chief
of Police to Minister of Justice, 5 August 1854, maço IJ6–218. There are other examples in
IJ6–217 (January–July 1854). For more on British initiatives, see W. D. Christie, Notes on
Brazilian Questions (London [etc.], 1865), pp. 82–85.
49. www.slavevoyages.org; accessed 21 October 2010.
50. The classic work on the conservatives, or saquaremas, is Ilmar Rohloff de Mattos, O tempo
saquarema (São Paulo, 1987); for a study that relates the resumption of the illegal trade with the
particiapation of the conservatives in the central government, see Tâmis Peixoto Parron,
‘‘Polı́tica do tráfico negreiro: o Parlamento imperial e a reabertura do comércio de escravos na
década de 1830’’, Estudos Afro-Asiáticos, 29 (2007), pp. 91–121.
51. Anais do Senado, 30 June 1837, pp. 175–181.
424 Sidney Chalhoub
chambers and irked the British; although it passed in the Senate, the
deputies buried it.52 The illegal trade continued unabated in the following
years, and there is a fascinating glimpse of what may have gone on in the
minds of planters at the time in the excerpt below, taken from the last will
and testament of Antônio Machado de Campos, from Campinas, province
of São Paulo, 1837:
Due to ignorance and because other people told me that I could do it, I bought
two Africans after the law that forbade these transactions, and because my only
wish is to save my soul and in matters of conscience one has to proceed very
cautiously, I ask the person in charge of my last will and testament to take the
aforementioned Africans to the judge of the orphans, requiring that they be
deposited with my heirs until they are educated, and they will be baptized, but
not as slaves. Therefore I finish this will, and I want fulfilled what I establish
here, except that if there is a law determining that the Africans now existent
should be slaves, then the two about whom I made the declaration above must
belong to my heirs as their slaves.53
The passage bears witness to the class solidarity prevailing among slave
owners, with planters encouraging each other to disrespect the law, all
of them patriotically sharing the ideal of increasing their wealth through
contraband and the illegal enslavement of Africans – incidentally, ‘‘to
reduce a free person to slavery’’ was a crime specified in the 1830 criminal
code. On the subject of class conscience, Antônio Machado performed
equally well, for he showed that he was aware of parliamentary debates
about the revocation of the law of 1831 and the legalization of slave
property held illegally, thus the conditional form used in his will – ‘‘if there
is a law determining that the Africans now existent should be slaves’’.
Although seigneurial expectations for the repealing of the 1831 law did not
prevail, at least one deputy argued convincingly that the mere discussion of
the subject in the parliament seemed to have helped to legitimate the
contraband trade, contributing to increase the pace of importations.54
The appalling rate of illegal enslavement that occurred after 1831
impacted the daily experience of freedom for people of African descent in
general, since it caused insecurity and rendered freedom precarious. The
connection between illegal enslavement and the precariousness of freedom
52. For opposition in the Senate, see Anais do Senado, July 1837, in particular the interventions
of Senator Teixeira de Gouvêa; for opposition in the Chamber of Deputies, see, for example,
Annaes do Parlamento Brasileiro, Camara dos Srs Deputados, 3 October 1837, p. 615. Also, for
general accounts, see Robert E. Conrad, World of Sorrow: The African Slave Trade to Brazil
(Baton Rouge, LA [etc.], 1986), ch. 4, and Leslie Bethell, The Abolition of the Brazilian Slave
Trade: Britain, Brazil and the Slave Trade Question, 1807–1869 (Cambridge, 1970), ch. 3.
53. Cited in Meyer Ferraz, ‘‘Entradas para a liberdade’’, pp.75–76.
54. Annaes do Parlamento Brasileiro, Camara dos Srs Deputados, 2 September 1837, p. 453;
see also Conrad, World of Sorrow, p. 97.
is crucial, both to understanding the logic permeating public policies and to

observing the strategies used by blacks and pardos, slave, free, and freed, to
deal with it.55
Slave owners’ interest not to abide by the law of 1831 necessarily meant
increasing slackness in property requirements thereafter. For example, in
the early 1830s, when attempts to curb the illegal trade were still in order,
the police authorities in the city of Rio quickly realized that they needed
to prevent the transportation of Africans to where they were in demand,
that is, to the interior of the provinces of Rio de Janeiro, São Paulo, and
Minas Gerais. Despite the fact that conductors of slaves for sale in the
interior needed to have the captives’ passports ready for presentation to
the authorities in controlling posts along the road, the issuing of these
documents remained confusing. They could be obtained from different
authorities, such as justices of the peace, police officials, and even from
employees in ministerial offices, making it difficult to establish equal
requisites for all cases; furthermore, the alleged owners of the slaves to be
transported did not have to present documentation of the original
transaction that brought them into the possession of the enslaved persons
in the first place.56
The absence of this requirement is telling, since it alone made it rather
easy to claim ownership of recently arrived Africans, who after 1831
could no longer be introduced through the alfândega (customs), with its
ensuing certificates and receipts associated with the collection of due
taxes. With government revenue on slave property suddenly downsized to
Lilliputian proportions, laxity regarding primary documents of owner-
ship became a must, with authorities in Rio and elsewhere eager to have
proprietors pay taxes on their slaves.57 Hence, a law enacted 11 April
1842, establishing rules for slave registration and for the payment of
annual and sale taxes on slave property, reassured owners saying that, ‘‘On
the occasion of the first registration, nobody can be required to present
the title through which [he or she] came into the possession of a slave.’’58
Rules and procedures to make it possible not to recognize illegally
enslaved Africans and to give the appearance of legality to property ori-
ginated in contraband had two consequences. First, it encouraged slave
stealing, an activity that seems to have acquired epidemic proportions in
55. Sidney Chalhoub, ‘‘Illegal Enslavement and the Precariousness of Freedom in Nineteenth-
Century Brazil’’, in John D. Garrigus and Christopher Morris (eds), Assumed Identities: The
Meanings of Race in the Atlantic World (College Station, TX, 2010), pp. 88–115.
56. AN, Secretaria de Polı́cia da Corte, Ofı́cios com Anexos (1831–1832), maço IJ6–165.
57. Wilma Peres Costa, ‘‘Estratégias ladinas: o imposto sobre o comércio de escravos e a
‘legalização’ do tráfico (1831–1850)’’, Novos Estudos CEBRAP, 67 (2003), pp. 57–75.
58. Collecção das leis do Imperio do Brasil, decree n. 151, 11 April 1842, article 6: ‘‘No acto da
primeira matricula a ninguem se exigirá o titulo por que possue o escravo’’.
426 Sidney Chalhoub
the 1830s and 1840s, at least judging from the amount of time and effort
the police dedicated to prevent it during the period.59 Slave stealing
necessarily involved establishing networks with the participation of a
variety of individuals, beginning with the captives themselves, who often
agreed, and sometimes asked, to be taken away.
In May 1837, the justice of the peace of the first district of Sacramento
parish received the information that a tavern in the neighborhood served
as a gathering place for the enticement of slaves to run away.60 As was
usually the case in such matters, a slave made the first contact with his or
her fellows and attracted them to the tavern. In this case, his name was
Matheus, an African from Congo and a mason. Interrogated by the judge,
one of the blacks to be carried away, Manoel Monjolo, said that he wanted
‘‘to embark to get rid of the bad master he had’’; in addition, he explained
that Matheus had approached him saying that Joaquim José Ribeiro, a
Portuguese man who worked in the tavern, could arrange for him to be
sent away. The negotiations continued for a while, with Manoel Monjolo
perhaps seeking to be assured that he would be moving to a more
advantageous situation within bondage; he said that he had had meals
with Matheus in the tavern ‘‘several times’’, without ever paying for them.
The police caught Manoel Monjolo and Joaquim José Ribeiro as they
walked to Prainha, a beachfront location, supposedly the departure point
for the stolen slave.
Had the journey proceeded, Manoel Monjolo would have joined other
slaves and been conducted, maybe by another Portuguese man, to the
interior of the province of Minas Gerais or São Paulo, for example. There
he would have been sold to a coffee planter, probably under a new name,
João Monjolo, for instance; the master would have included him on
the list of slaves pertaining to his household, registered him, and paid
the annual slave fee. More importantly, the alleged master would have
counted on never being asked to document how he had originally
obtained the captive.61 As this case shows, the business of acquiring
smuggled Africans became inseparable from having a constant influx of
stolen ladino slaves as well, therefore occasionally giving captives the
opportunity to take the initiative to change masters, hoping to improve
their lot within slavery.
59. Marcus J. M. de Carvalho, ‘‘‘Quem furta mais e esconde’: o roubo de escravos em Pernambuco,
1832–1855’’, Revista do Instituto Histórico e Geográfico Brasileiro, 150 (1989), pp. 317–344.
60. AN, processo criminal (1837), Tribunal da Relação do Rio de Janeiro, Joaquim José Ribeiro
(apelante), furto de escravos, no. 1281, maço 139, galeria C.
61. This synthesis of how Manoel Monjolo’s story would have evolved is based on evidence
collected from police papers, series IJ6, several packets for the 1830s and 1840s; for a detailed
account on how stolen slaves were sold in the interior, see AN, Secretaria de Polı́cia da Corte,
Ofı́cios com Anexos (1842–1843), maço IJ6–199.
The second consequence of looseness respecting proof of slave property

was that illegal enslavement became a greater threat to free and freed
people of color in general – both African- and Brazilian-born. Although it
is not possible to know the frequency of such events, they pop up suffi-
ciently often in police correspondence, prison books, and trial records to
suggest that potential victims had to deal with the problem, calculating
their moves and remaining vigilant. Judy Bieber published a path-breaking
study on the subject some years ago, relating cases of the kidnapping and
sale of crioulos (Brazilian-born blacks), free and freed, quite often children,
in the province of Minas Gerais.62 On 1 September 1854, the Chamber of
Deputies discussed a proposed law to control the internal slave trade, which
had been increasing since the abolition of the African trade in 1850. João
Maurı́cio Wanderley, a deputy from Bahia, spoke extensively on the con-
sequences of the internal trade for the northern provinces; slave owners
there, especially those of modest means, lured by the high prices paid for
slaves in the coffee-growing provinces, had been selling their labor force,
condemning themselves to poverty. Furthermore, demand for labor was so
high that a new ‘‘speculation’’ had developed – namely, reducing free people
to slavery. The kidnapping and sale of free and freed black and pardo
children to slave dealers had become commonplace; sometimes guardians
of children of color sold them to slavery.63
Criminal and civil records render vivid tales of the illegal enslavement of
Brazilian-born slaves. In April, 1865, a black man, José, supposedly a slave
belonging to Fuão Goulart, gave a detailed account of his plight when
arrested for the homicide of Silva Pereira, his alleged master’s son-in-law.64
Although one can never be sure that José presented events as they really
happened, or even as they seemed to have happened from his point of view,
his narration was not contradicted by other witnesses in any important
aspect, thus allowing us to take the tale as plausible – that is, truthful or not
in its particular details, it provides valuable information about the experience
of free and freed people of African descent in the period.65
62. Judy Bieber Freitas, ‘‘Slavery and Social Life: Attempts to Reduce Free People to Slavery
in the Sertão Mineiro, Brazil, 1850–1871’’, Journal of Latin American Studies, 26 (1994),
pp. 597–619; see also Marcus J.M. de Carvalho, Liberdade: rotinas e rupturas do escravismo no
Recife, 1822–1850 (Recife, 1998), pp. 242–253.
63. Annaes do Parlamento Brasileiro, Camara dos Srs Deputados, 1 September 1854,
pp. 349–350.
64. Arquivo Edgard Leuenroth, UNICAMP, processo criminal (1865–1866), Tribunal da
Relação do Rio de Janeiro, reel 84.0.ACR.163 (microfilm copy of original belonging to the
Arquivo Nacional).
65. For an insightful discussion of how to read depositions in police and judicial records, see
Arlette Farge, Le gout de l’archive (Paris, 1989). Social historians in Brazil have been sys-
tematically using criminal and civil trial records for the study of slavery and working-class
culture since the 1980s: see, for early examples, Boris Fausto, Crime e cotidiano: a criminalidade
428 Sidney Chalhoub
José, who was a shoemaker, had been born in Pau d’Alho, in the north-
eastern province of Pernambuco. He was thirty-seven years old, single,
then living in the city of Rio; he said that he resided on Catumby Street,
but did not know the number. He confessed to having stabbed the victim
out of despair caused by the harsh physical punishment imposed by his
master. Answering questions asked by a police officer, José declared that:
He, the respondent, is free by birth; his mother is alive in Pau d’Alho, and her
name is Joanna Maria da Conceição; besides her mother, he also has an aunt in
Recife, Silvéria Maria da Conceição, who lives on Direita Street [y], and the
Lieutenant Colonel of the National Guard Antônio Laurianno Lopes Coutinho,
who also lives in Recife, on Queimados Street, served as godfather at his bap-
tism. [He said] [t]hat he came to this capital engaged by the Portuguese Manoel
Teixeira de Araújo, as his servant; but this man had to go to a hospital and left
him with Bernardo José Pinto, who lives in this city on Rosário Street, from
where he the respondent was taken to Cantagalo [a county in the province of
Rio], lured by promises to earn more money there working at his profession;
but there arriving he learned that he had been enslaved, and not being able to
suffer the punishments inflicted upon him, he tried to commit suicide by cutting
his own throat with a razor. From Cantagalo he was sold to Fuão Goulart, who
now lives on Catumby Street, there receiving the same punishments he had been
suffering in Cantagalo; yesterday, as he left the shop where he worked to take a
knife to sharpen, without the intention of killing anybody, it happened that he
met Pereira on Cano Street; Pereira then grabbed him [y] to conduct him to
Goulart’s house, and it was on this occasion that the respondent, having received
several blows given by Pereira with his umbrella, stabbed him, which he did not
intend to do, and Pereira contributed for this outcome because he inspired
terror in the respondent, threatening that he would be put in shackles and
receive beatings; that he regrets what happened to Pereira, and wished he had
done it to himself [y].
José probably belonged to a family whose members had recently
obtained freedom – perhaps the very first generation of a family of freed
people. His mother carried the names of ‘‘Maria’’ and ‘‘Conceição’’,
alluding to the Blessed Virgin Mary, and very common among freed
women;66 also, she managed to have her son baptized by a person of a
em São Paulo (1880–1924) (São Paulo, 1984); Maria Odila Leite da Silva Dias, Quotidiano e
poder em São Paulo no século XIX – Ana Gertrudes de Jesus (São Paulo, 1984); Sidney Chalhoub,
Trabalho, lar e botequim: o cotidiano dos trabalhadores no Rio de Janeiro da Belle Époque (São
Paulo, 1986); Hebe Maria Mattos de Castro, Ao sul da história: lavradores pobres na crise do
trabalho escravo (São Paulo, 1987); Maria Helena P.T. Machado, Crime e escravidão: trabalho,
luta e resistência nas lavouras paulistas, 1830–1888 (São Paulo, 1987), and many others. For a
recent survey of Brazilian social history since the 1980s, see Sidney Chalhoub and Fernando
Teixeira da Silva, ‘‘Sujeitos no imaginário acadêmico: escravos e trabalhadores na historiografia
brasileira desde os anos 1980’’, Cadernos AEL, 14 (2009), pp. 11–50.
66. Jean Hébrard, ‘‘Esclavage et dénomination: imposition et appropriation d’un nom chez les
esclaves de la Bahia au XIXe siècle’’, Cahiers du Brésil Contemporain, 53/54 (2003), pp. 31–92.
higher social status, a ‘‘Lieutenant Colonel of the National Guard’’, a

strategy to bring recognition to her child’s status as free or freed.
Nonetheless, poverty and destitution, and perhaps the hope of achieving
a better economic situation, made José take his chances in the nation’s
capital, therefore falling prey to criminal gangs willing to enslave free
people and sell them to coffee planters. The detailed information provided
about relatives in Pernambuco suggests that he had a safety net in his
native province, with his mother having devised spiritual and social
strategies to keep him safe from birth.
It is curious that José was a native of the parish of Pau d’Alho and
said that his mother lived there. The parish of Pau d’Alho was at the
center of a major popular revolt against a civil registration law in 1852.
According to contemporary accounts of the revolt, people who partici-
pated in it alleged that mandatory civil registration was actually a strategy
to enslave free people of African descent because ‘‘the English no longer
allowed the coming of new Africans’’.67 We do not know if José lived with
his mother in Pau d’Alho during the 1852 events. In any case, however
truthful his narration in its details, it seems clear that he could articulate
meaningful aspects of the precariousness of freedom in nineteenth-
century Brazil.
I will indulge in one last story on the same subject, an almost rocam-
bolesque one in the way it brings together successive episodes of illegal
enslavement.68 Joaquim de Azevedo Ramos initiated a petition on behalf
of the pardo Odorico in 1862, alleging that he was his father and seeking
to free him from a bondage he deemed illegal. Ramos said that by virtue
of ‘‘the weakness of the flesh’’ he had had relations with the black woman
Felismina, ‘‘who was free, but had been sold to the deceased father of the
notary public Perdigão as a captive with the name of Thereza, and she
died in 1842’’.
Three witnesses offered detailed depositions to support the plaintiff’s
allegations. As stated by Leocádia Maria da Conceição, a seamstress,
forty-five years old, and a freed woman who had worked side by side
with Felismina in bondage, Perdigão and his family recognized Felismi-
na’s condition as free; nonetheless, they had her son, Odorico, baptized as
a slave. Ana Rangel de Macedo, a school principal, seventy-two years old,
and daughter of the notary public Perdigão, gave her account of the illegal
67. Guillermo Palacios, ‘‘Revoltas camponesas no Brasil escravista: a ‘Guerra dos Maribondos’
(Pernambuco, 1851–1852)’’, Almanack Braziliense, 3 (2006), pp. 9–39, 22 (quotation from
correspondence sent by a local judge to the president of the province of Pernambuco); see also
Mara Loveman, ‘‘Blinded Like a State: The Revolt against Civil Registration in Nineteenth-
Century Brazil’’, Comparative Studies in Society and History, 49 (2007), pp. 5–39.
68. AN, processo cı́vel (1862), Juı́zo Municipal da Segunda Vara do Rio de Janeiro, no. 1, caixa
523, galeria C.
430 Sidney Chalhoub
enslavement of both mother and son, Felismina and Odorico. When she
was seven, Felismina had been sold to her father by a man named José
Bento, from Valença, a coffee county in the province of Rio, and given the
name of Thereza. However, one day, as the supposed Thereza walked on
the street with another slave woman pertaining to the household, a black
woman who passed by recognized her as Felismina, born in Valença as a
free girl and her godchild. The Perdigão family believed in the story and
considered Felismina free thereafter. Later her sister, Dona Mariana,
brought Felismina to live with her. Felismina gave birth to Odorico
during this period; thus it seems that Dona Mariana had him baptized as
slave, although her sister does not go so far as to say that. The third
witness, Miguel Geraldo da Silva, a typesetter, thirty-four years old,
confirms the whole story, and suggests that the family considered
Odorico free because the young man, unlike slaves, went about the city
wearing shoes.
In spite of the witnesses’ declarations that the Perdigão family deemed
Odorico a free person, he was sold to Leopoldino dos Santos Pereira, who
sold him to José Antônio Teixeira, who died, but whose widow intended
to sell him again, finally prompting legal action by his father, Joaquim de
Azevedo Ramos. Asked to present documentary evidence that Felismina
had been a freed person, therefore proving that Odorico had been born
free, Ramos argued that he took the trouble of filing the petition just
because he did not have a paper trail to sustain his allegations. Since
Felismina had been born free, and not freed after birth, he did not have a
letter of manumission or another type of document to support his story.
Here, again, as in the case of the black Augusta, seen above, part of the
motivation to go to court consisted of creating a paper trail to safeguard
the right to freedom. Paradoxically, Felismina’s freedom seemed harder to
prove than Augusta’s because she had never been a slave. Augusta gath-
ered witnesses to show that a letter of manumission had existed and
disappeared. Felismina’s ex-lover had to resort to witnesses to reconstruct
her story from scratch, a mission made more difficult by the sheer fact
that illegal enslavement entailed erasing or muddling the evidence of the
original act of reducing someone to bondage.69
69. For a discussion of illegal enslavement in a different context, see Rebecca J. Scott, ‘‘Paper
Thin: Freedom, Re-Enslavement, and Determinations of Status in the Diaspora of the Haitian
Revolution’’, Law and History Review, (November 2011). Scott discusses the re-enslavement of
refugees from Haiti – first in Cuba, then in Louisiana – in the early nineteenth century, to
conclude that ‘‘The same codes and statutes that determined that some persons could be held as
property required no sound or consistent account of how the property right in a given indi-
vidual had lawfully come into being in the first place. They required no proof that the person in
question was subject to such ownership; property in his or her person was, like property in
other ‘immovables,’ presumed from the fact of possession’’ (cited with permission).
S U S P E C T E D O F B E I N G F U G I T I V E S L AV E S
During the long period during which he held the job of chief of police
in Rio, from 1833 to 1844, Eusébio de Queiróz organized the routine
procedures of the institution according to the tenet that ‘‘since it is not
easy to obtain proof of slavery, when a black insists that he is free’’, it
seemed ‘‘more reasonable, in the case of blacks, to presume their bondage,
until they present a certificate of baptism or a letter of liberty to prove
otherwise’’.70 It is noteworthy that the policy of deeming a slave any
person suspected of being one took root while the contraband trade
brought more than 750,000 Africans to the country in the 1830s and
1840s. Eusébio de Queiróz’s doctrine provided a rationale, so to speak, for
overlooking procedures that would have made it difficult not to recognize
the illegal enslavement of Africans. Moreover, it transferred to Africans
and their descendents the burden of proving their liberty, at a time when
masters would have problems if pressed hard on the question of how they
had originally acquired much of their slave property.
The consequences of such police conduct were durable for people of
African descent, since their detention on the allegation that they were
suspected of being fugitive slaves remained common for as long as slavery
existed in Brazil. Not surprisingly, the police appeared particularly prone
to making mistakes in this matter, not least because, as mentioned before,
three out of every four blacks and pardos were free or freed according to
the census of 1872. In the city of Rio, the 1872 census computed a
population of 274,972, of which 226,033 were free people (82.2 per cent),
and 48,939 slaves (17.7 per cent). As regards the racial composition of the
free population in relation to the total population, 55.2 per cent were
considered white, a high percentage due, in part, to the strong presence of
Portuguese immigrants in the capital (more than 55,000), 10.3 per cent
black, 16.3 per cent pardos, 0.33 per cent caboclos (mixed with Indians);
among the remaining 17.7 per cent slave population, 77.3 per cent were black,
22.6 per cent pardos.71 On the aggregate – that is, slave and free combined –
the population of African descent in Rio represented 44.4 per cent of the
total. Considering only the population of African descent, 59.9 per cent was
free, 40 per cent slave; in other words, for every five people of color residing
in the capital in 1872, three were free, two slaves.
Let us turn briefly to the Livros da Casa de Detenção da Corte, that is,
the books that registered the entry to the House of Detention of people
70. AN, Secretaria de Polı́cia da Corte, Ofı́cios com Anexos (1835), Chief of Police to Minister
of Justice, 7 December 1835, maço IJ6–171.
71. Luiz Carlos Soares, O ‘povo de Cam’ na capital do Brasil: a escravidão urbana no Rio de
Janeiro do século XIX (Rio de Janeiro, 2007), pp. 376–380, and Senra, História das estatı´sticas
brasileiras, I, p. 442.
432 Sidney Chalhoub
arrested daily in the city of Rio in the 1860s and 1870s.72 Despite several
gaps in the available data, the House of Detention books convey a wealth
of information on each person detained – that is, name, age, nationality,
place of birth, parents’ names, reason for arrest, marital status, address,
occupation, height, color, clothes worn on occasion of detention, date of
entry, date of release, name of master (for slaves), social condition (slave,
free, freed), authority who ordered the arrest. There were different books
for slave and free people. The database contains 8,445 entries: 2,697 in
books pertaining to slaves (31.9 per cent), 5,748 in books for free detainees
(68 per cent).
Focusing on slave entries reveals that throughout the period the main
function of the police with regard to the institution of slavery consisted of
assisting masters in disciplining their captives. Thus 61 per cent of slave
arrests may be described as actions to help masters maintain control of
their bonded labor force – for instance, detentions made on a master’s
request, capture of supposed runaways, arrests associated with slaves
being out on the streets after curfew without a master’s permission; 29.8
per cent of the detentions resulted from slaves being accused of com-
mitting a variety of felonies and misdemeanors; in 6.1 per cent of the cases
the slaves took the initiative of seeking the police – for example, to
complain of mistreatment or to file a petition for freedom; 3.1 per cent of
the entries do not specify a motive for the detention.
Searching the database for entries associated with being a ‘‘fugitive’’
slave, or ‘‘suspected of being a fugitive’’ slave, we come to a total of 1,097
such detentions: 447 for ‘‘fugitive’’, 650 for ‘‘suspected of being a fugitive’’.
Although 70 per cent of entries for these motives appeared in books
dedicated to slaves, it is perhaps surprising to find nearly 30 per cent of
them annotated in books pertaining to free people. However difficult to
be sure about these matters, it seems that there were situations in which
the police tended to take information as presented to them; therefore, if
those detained for allegedly being runaway slaves declared themselves to
be slaves and to have run away, they would enter a book for the registry
of slaves as ‘‘fugitives’’. In other words, the police appeared inclined to
believe a person who admitted to being a slave, despite the fact that the
police records carry several examples of free people who alleged themselves
72. The Livros da Casa de Detenção da Corte are kept in the Arquivo Público do Estado do
Rio de Janeiro; there are copies of them in microfilm in the Arquivo Edgard Leuenroth at
UNICAMP. The contents of the books pertaining to the 1860s and 1870s are being loaded on to
a database belonging to the Centro de Pesquisa em História Social da Cultura (CECULT/
UNICAMP). The following paragraphs are based on what the database held as of July 2010,
which represented nearly 90 per cent of cases existent in the books for the 1860s and 1870s.
I owe a special thanks to Anita Lazzarin, a student research assistant at CECULT, who helped
me to consolidate the data used in this article.
to be captives to avoid an immediate danger of forced recruitment for the

army or the navy.73 In addition, there were several cases in which the
detainee would admit to being a slave, but deny the accusation that he or
she had run away. Finally, there appeared entries in which the police
doubted it all: a person was seized on suspicion of being a slave and of
having run away; not giving credence to the suspect, but having to deal
with his or her assertion of freedom, the police could register the case in a
book pertaining to free people, creating a logical contradiction, since a free
person could not be seized for being a ‘‘fugitive’’ slave. Alternatively, from a
social logic specific to slave society, a slave could not be detained for being a
‘‘vagrant’’, since the condition of slave linked someone immediately to labor
for a given master.74
Thus the 650 detentions given the justification of ‘‘suspected of being a
fugitive’’ are of special interest to us because police conduct in this regard
had an impact on the lives of free and freed people of African descent. The
search yields 344 results in books for the registration of slaves (53 per cent),
306 in books for free people (an appalling 47 per cent). Nonetheless, results
for the 1860s differ significantly from those of the 1870s. When 11 entries
with illegible or unreliable dates are excluded, there appear 366 results for
‘‘suspected of being a fugitive’’ in the 1860s, distributed thus: 77.8 per cent in
books for slaves, 22.1 per cent in those for free people. For the 1870s, there
are 273 entries: 20.1 per cent in books for slaves, nearly 80 per cent in those
for free people. In other words, the registration of arrests for ‘‘suspected of
being a fugitive’’ largely moved from books dedicated to slaves in the 1860s
to ones pertaining to free people in the 1870s.
Why did this change happen in the pattern of registry? Until the 1860s,
the notion that a person taken to be a fugitive slave remained a slave
unless proven otherwise continued to prevail. In the 1870s, no doubt as a
consequence of the gradual emancipation law of 28 September 1871, the
tendency became to consider free a person whose bondage could not be
ascertained. After the 1871 law, only the certificate of slave registration
(matrıćula) served as the legal means for claiming slave property, or
proving someone’s condition as a slave. Hence, despite their profiling on
the streets for ‘‘suspicion of being a fugitive’’, it is probable that the police
began to enter in books for free people the cases of detainees of African
descent who asserted their freedom. In addition, these inconsistencies
in the House of Detention books suggest again the frequency of social
73. Chalhoub, ‘‘Illegal Enslavement and the Precariousness of Freedom in Nineteenth-Century

Brazil’’, passim.
74. Coherently, the database carries just a dozen examples of slaves detained for being
‘‘vagrant’’, while this was perhaps the most commonly alleged cause for the detention of free
people (together with drunkenness, which also applied to slaves).
434 Sidney Chalhoub
situations in which the frontiers between slavery and freedom remained

blurred in nineteenth-century Brazil.
On the one hand, the doctrine of chief of police Eusébio de Queiróz
helped to strengthen slavery and incorporate into it a massive number of
Africans smuggled into the country in the 1830s and 1840s, as well as their
increasing number of descendants, also illegally held in bondage for
decades to come. As a consequence, generations of free and freed people
of color had their experience of freedom largely structured by these cir-
cumstances, having to learn to deal with situations which could bring
them under the threat of being reduced to slavery. On the other hand, in a
society in which more than 70 per cent of blacks and pardos were free in
the early 1870s, the 1871 law allowed many slaves to turn upside down the
crux of Eusébio de Queiróz’s doctrine: if the uncertain frontiers between
slavery and freedom reduced opportunities for free and freed people of
African descent, they also gave slaves the chance of seeking to pass as free,
mingling and hiding among thousands of free and freed blacks and pardos
living in the capital and elsewhere, challenging alleged owners, police
officers, and judicial authorities to find their matrıćula.75
STRUCTURAL PRECARIOUSNESS
AND HISTORICAL PROCESS
Seen from a distance, the abolition of the slave trade and of slavery proper
were perhaps the most impressive nineteenth-century historical achieve-
ments of western European countries and their former colonies in the
Americas. By the end of the century, nationalist deliriums, scientific
delusions, and imperialist predation, on the part of several European
countries, rested on representations about the progress of freedom and
civilization that the suppression of slavery and other forms of compulsory
labor epitomized. The end of slavery appeared as a prerequisite for
industrial and technological progress, the expansion of markets, the
improvement of financial institutions, the voluntary migration of workers,
the acquisition of civil and political rights, and urbanization.
Nonetheless, such a panoramic perspective elides the indeterminate and
fragmentary character of much of the history of slavery in the nineteenth
century, particularly with regard to the strengthening of the institution
in ways that rendered even more tragic the experience of the African
diaspora. At the end of the eighteenth century, slavery remained strong in
the British and French sugar-producing colonies. The Haitian Revolution
interrupted the prosperity of the main French sugar colony; successive
abolitionist waves and political turmoil ended British involvement in the
slave trade and, subsequently, in the 1830s, brought slavery to an end in
75. Chalhoub, Visões da liberdade, ch. 3.
Figure 4. Portrait of a man; carte de visite; photographer: Militão Augusto de Azevedo, 1870s.
Museu Paulista, São Paulo. Used with permission.
436 Sidney Chalhoub
its colonies. The same historical process that consolidated the pre-
eminence of Great Britain in world markets and international relations
caused both the abolition of slavery in British colonies and the expansion
of it in other areas of the Americas. In the following decades, as the cotton
industry grew, the US South deepened its commitment to slavery; in 1830,
Cuba had become the leading world producer of sugar, while its slave
population of 85,900 people, in 1792, reached 286,900 in 1827, and
436,500 in 1841.76
Due to low birth rates among the slave population, the expansion of
coffee in Brazil in the second quarter of the nineteenth century depended
heavily on the importation of enslaved Africans. Actually, demand for
bonded Africans had been increasing since the 1790s as a consequence of
markets made available after the collapse of sugar production in Haiti. In
the 1820s, coffee remained the third item in the country’s exports, after
sugar and cotton.77 In other words, the restructuring of slavery in Brazil,
following the demise of gold mining, preceded the growth of coffee in the
provinces of Rio de Janeiro and São Paulo.
Current estimates for the number of Africans taken to Brazil during the
whole period of the slave trade, from the mid-1500s to the early 1850s,
stand at more than 4.8 million. In the first quarter of the nineteenth
century (1801–1825), 1,012,762 Africans disembarked in the country; in
the second quarter (1826–1850), there were 1,041,964, plus 6,800 after the
1850 law that effectively abolished the trade. These figures show that
42 per cent of the total number of Africans transplanted to Brazil during
three centuries of the slave trade arrived there in the first half of the
nineteenth century.78 Most Africans introduced in the second quarter of
the century ended up in coffee-growing south-eastern provinces; further-
more, a vast majority of them, approximately 750,000, came in contempt of
the international treaties and national legislation that had prohibited the
African trade.
As I have argued in this article, the core of the concept of the structural
precariousness of freedom in nineteenth-century Brazil rests upon the
idea that there prevailed, in the longue durée, modes and imperatives of
social interaction that often made the boundaries of slavery and freedom
76. Dale W. Tomich, ‘‘The ‘Second Slavery’: Bonded Labor and the Transformation of the
Nineteenth-Century World Economy’’, in idem, Through the Prism of Slavery: Labor, Capital,
and World Economy (Lanham, MD, 2004), pp. 56–71; for data on the Cuban slave population,
see p. 64. See also idem, ‘‘World Slavery and Caribbean Capitalism: The Cuban Sugar Industry,
1760–1868’’, Theory and Society, 20 (1991), pp. 297–319, and in the same issue, by Sidney Mintz,
‘‘Comment on Articles by Tomich, McMichael, and Roseberry’’, pp. 383–392.
77. Tomich, ‘‘The ‘Second Slavery’’’, p. 67.
78. www.slavevoyages.org; accessed 6 September 2010. Robert Slenes has called my attention to
the magnitude of these numbers.
uncertain. However important this feature of Brazilian slavery may have

been in previous times, it acquired a new weight after 1831, when dealing,
and eventually condoning, massive illegal enslavement gradually became
an organizing principle of state formation and institutional building in
the country. By the end of the 1830s, political groups that would soon
coalesce in what became known as the conservative party, or saquaremas,
held a firm grip on ideological views and political power,79 henceforth
directing public policies to defend the monarchical regime, the expansion
of latifundia, and the institution of slavery – which increasingly meant
procrastinating about the abolition of the slave trade and protecting the
interests of those who held slave property illegally. With respect to the
experience of freedom of people of African descent, they had to learn to
cope with the consequences of the politics of precariousness in public
policies and institutions, which shaped, for example, their routine inter-
actions with the police, county officials, local judges and, last but not
least, slave dealers and slave owners, for whom illegal enslavement seemed
to have become a customary right.
The abolition of the slave trade after the enactment of a new law, in
1850, led to a period of political stability, with liberal and conservative
political currents in agreement as regards basic ideological tenets and
temporarily relieved from British pressure respecting the problem of
slavery. In addition, the cessation of African arrivals quelled fears that a
growing demographic imbalance might render more dangerous the
rebellious disposition of slaves, whose actions by the end of the 1840s
reminded masters and authorities of the situation they had confronted
in the mid-1830s.80 Nevertheless, as the abundance of labor since the
large-scale importation of Africans in the late 1840s wore out, especially
after a devastating cholera epidemic in 1855–1856, the experience of
precariousness within slavery increased by virtue of a growing internal
slave trade.81 As a massive number of slaves began to be forcibly moved
from the north-eastern provinces – where slave communities with a
majority of Brazilian-born captives were more common – to coffee
plantations in the provinces of Rio and São Paulo, concerns about slave
79. Rohloff de Mattos, O tempo saquarema; see also Jeffrey D. Needell, The Party of Order:
The Conservatives, the State, and Slavery in the Brazilian Monarchy, 1831–1871 (Stanford, CA,
2006).
80. Marcos Ferreira de Andrade, ‘‘Rebelião escrava na comarca do Rio das Mortes, Minas
Gerais: o caso Carrancas’’, Afro-Ásia, 21:2 (1998–1999), pp. 45–82; Flávio dos Santos Gomes,
Histórias de quilombolas: mocambos e comunidades de senzalas no Rio de Janeiro, século XIX,
2nd edn (São Paulo, 2006); Reis, Rebelião escrava no Brasil.
81. Robert W. Slenes, ‘‘The Brazilian Internal Slave Trade, 1850–88’’, in Walter Johnson (ed.),
The Chattel Principle: Internal Slave Trades in the Americas (New Haven, CT [etc.], 2004),
pp. 325–370.
438 Sidney Chalhoub
resistance resurfaced, this time in the form of crimes committed by

bonded people against masters, overseers, and their families.82
The 1860s brought with it the recrudescence of British pressure,
now focused upon the question of africanos livres – ‘‘liberated Africans’’.
W.D. Christie, Her Majesty’s Minister in Brazil, argued vigorously that
the definition of africanos livres should encompass all Africans introduced
into the country in disrespect of the 1831 law, not just those Africans
actually recovered during actions to repress the clandestine trade.83 If
such a thesis took hold, the recognition of the illegality of much of the
existent slave property would follow. Not incidentally, perhaps, there
appeared at the time judicial petitions for freedom defending the right to
liberty of Africans imported after 1831.84 The struggle for freedom that
began to surface, in the context of mounting international isolation in the
aftermath of the American Civil War, led to the first governmental and
legislative moves toward gradual emancipation, resulting in the 1871 law.
The 1871 law seemed ambivalent. On the one hand, legislators sought
to settle the question of slave property rights by establishing mandatory
slave registration, thus creating an official document to legitimate slave
ownership. On the other hand, government and judicial intervention on
the question of manumission – with the law granting freedom to children
of slave mothers born thereafter and opening paths to liberty irrespective
of the masters’ consent – empowered slaves and their allies to struggle for
freedom by resorting to the legal apparatus made available by the imperial
government itself.
During the 1870s, patterns of slave arrests in the city of Rio show that
the main rationale for police intervention continued to be that of assisting
masters in disciplining their captives. In spite of that, slaves went more
often to the police to confront the injustices of life in bondage, and
appeals for freedom became more common in judicial courts.85 Gradually,
82. As early as 1854, the Provincial Assembly of São Paulo sent a letter to the imperial govern-
ment to complain of slave crime and to protest against the non-application of the death sentence,
due to the intervention of the Emperor, to several slaves who had murdered masters and their
relatives; see ‘‘Consulta de 30 de Outubro de 1854. Sobre a representação da Assembléia Legislativa
da Provı́ncia de São Paulo acerca da pena de galés que se impõe aos escravos’’, in José Prospero
Jehovah da Silva Caroatá, Imperiais resoluções tomadas sobre consultas da seção de Justiça do
Conselho de Estado. Desde o anno de 1842, em que começou a funccionar o mesmo Conselho, até
hoje (Rio de Janeiro, 1884), part 1, pp. 507–509. For a detailed study of the subject in the 1870s and
1880s, see Célia Azevedo, Onda negra, medo branco. O negro no imaginário das elites – século
XIX (São Paulo, 1987).
83. Christie, Notes on Brazilian Questions, passim.
84. Elciene Azevedo, O direito dos escravos: lutas jurı´dicas e abolicionismo na provıńcia de São
Paulo (Campinas, 2010).
85. In the House of Detention database that I have been using in this paper (see n. 72, above),
entries that originated with slaves taking the initiative of going to the police went from 2.6 per cent
the customary situations of uncertainty between slavery and freedom

began to offer opportunities for slaves to challenge masters’ control; for
instance, their attempts to pass for free or freed people after 1871 required
owners to present a certificate of slave registration to undo.
Although the 1880s started with growing expectations concerning
slave emancipation, with the abolitionist movement spinning out of the
parliament on to the streets of the capital, planters and their representa-
tives resisted the demise of slavery with incredible tenacity. Planters
would achieve a new legal compromise in 1885, which would not last
long, but proved sufficient to guarantee to Brazil the dubious honor of
having been the last country in the western hemisphere to abolish African
slavery, in 1888.86
The slaves who achieved manumission in growing numbers faced social
and political redefinitions of the precariousness of freedom. They had
been excluded from any aspiration to political rights as a consequence of
the literacy requirements established in the 1881 electoral law, remained
without access to elementary education for decades to come, and were not
allowed to create mutual aid societies based on racial or ethnic identities.87
In addition to these hurdles, people of African descent faced prejudices
associated with scientific racism, which emboldened European imperial-
ism and influenced ideologies and public policies in Brazil. The emergence
of new labor ideologies changed the scope of the concept of vagrancy,
tending to value wage labor, rejecting other options for earning a living
and subjecting the urban poor to renewed forms of police profiling, with
blacks and pardos being specially targeted.
of the cases in the 1860s to 9.6 per cent in the 1870s; on slaves going to the police and filing appeals
for freedom, see Chalhoub, Visões da liberdade, passim.
86. On the 1885 law, see Joseli Mendonça, Entre a mão e os anéis: a lei dos sexagenários e os
caminhos da abolição no Brasil (Campinas, 1999).
87. Sidney Chalhoub, ‘‘The Politics of Silence: Race and Citizenship in Nineteenth-Century
Brazil’’, Slavery and Abolition, 27 (2006), pp. 73–87, 73–76; idem, ‘‘Solidarité et liberté. Les
sociétés de secours mutuel pour gens de couleur à Rio de Janeiro dans la seconde moitié du
XIXe siècle’’, Cahiers du Brésil Contemporain, 67/68 (2007), pp. 363–392.

A Precariedade de Liberdade

Uploaded by

Copyright:

Available Formats

A Precariedade de Liberdade

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A Precariedade de Liberdade

Uploaded by

Copyright:

Available Formats

IRSH 56 (2011), pp. 405–439 doi:10.

The Precariousness of Freedom in a Slave Society

Figure 1. Nineteenth-century Brazil: provinces and cities cited in the text.

According to the same author, in the state of Louisiana, in 1860, just

to decide whether, when, or how to free their slaves constituted a major

Figure 2. Portrait of a couple; carte de visite; photographer: Militão Augusto de Azevedo,

12. Rebecca J. Scott, ‘‘Exploring the Meaning of Freedom: Post-Emancipation Societies in

In a critical essay still to appear, Robert Slenes remarks that divergent

massive transfer of captives from the northern to the southern coffee-

of a return to their native continent. A law enacted in May 1835 almost

28. Ibid, p. 92.

ranging from the establishment of conditions such as continued service

As to what conditional manumission meant for freed people, whether it

deceased Lieutenant Colonel, has intended to consider the Plaintiff a slave

In 1854, for example, British diplomats repeatedly complained of African

is crucial, both to understanding the logic permeating public policies and to

The second consequence of looseness respecting proof of slave property

higher social status, a ‘‘Lieutenant Colonel of the National Guard’’, a

to be captives to avoid an immediate danger of forced recruitment for the

73. Chalhoub, ‘‘Illegal Enslavement and the Precariousness of Freedom in Nineteenth-Century

situations in which the frontiers between slavery and freedom remained

75. Chalhoub, Visões da liberdade, ch. 3.

uncertain. However important this feature of Brazilian slavery may have

resistance resurfaced, this time in the form of crimes committed by

the customary situations of uncertainty between slavery and freedom

You might also like