Edinburgh University Press

Chapter Title: ARISTOTLE AND THE STRANGENESS OF SLAVES

Book Title: The Politics of Slavery

Book Author(s): Laura Brace
Published by: Edinburgh University Press. (2018)
Stable URL: https://www.jstor.org/stable/10.3366/j.ctt1tqxvk1.5

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Edinburgh University Press is collaborating with JSTOR to digitize, preserve and extend
access to The Politics of Slavery

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms
 Chapter 2

ARISTOTLE AND THE STRANGENESS OF SLAVES

Aristotle’s theory of natural slavery, set out in the Politics and in the
Nicomachean Ethics is a useful starting point for thinking about slavery
in the history of political thought, and for introducing and developing
some of the key themes of this book. The idea of conceiving of the
slave as an animate tool raises a whole set of questions about the sup-
posed subhumanity of the slave and how that status is understood in
the history and politics of slavery. I am particularly interested in the
‘incompleteness’ of the slave, and the ways in which his or her soul
was understood to be lacking in spirit, in the constituent elements
required to build a free citizen. Aristotle is an important place to start
because his arguments bring together political slavery and what Mary
Nyquist calls psycho-ethical slavery, and because his explicitly political
approach to slavery draws attention to the question of how we should
theorise the relationship between slavery as metaphor and slavery as
lived experience. Did his theory of natural slavery have anything to
say about the lives of actually existing slaves in ancient Greece?
Nyquist discusses the differences between figurative, political slav-
ery and chattel slavery, where political slavery is about the threat to
the democratic polis and not about the condition of chattel slaves. Her
analysis of the ‘polyvalent metaphor of slavery’ (Nyquist 2013, 5) draws
attention to the ‘entangled interrelations’ (Nyquist 2013, 2) between
political servitude and chattel slavery, and these entanglements are
particularly gnarly in Aristotle’s theory. Within the polis, political slav-
ery is represented as the illegitimate domination of free, male citizens
who expected to participate as equals in the political process, exercising
their freedom as political agents, none of them ruling over others. Polit-
ical slavery comes about when ‘a leader fails to protect the citizenry’s

16

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 16 12/01/18 6:03 PM

 Aristotle and the Strangeness of Slaves 17

freedom, instead attempting to become its master’ (Nyquist 2013, 22).

As Nyquist points out, the injustice of this political enslavement lies ‘in
the attempt to enslave those who patently ought not to be enslaved’
(Nyquist 2013, 23). Participants met in the political arena as equals, but
they were masters within their own households. Aristotle’s opposi-
tion to political slavery was not an attack on slavery as wrong in itself.
Democratic citizens were the masters of slaves within their households,
and the boundary between the household and the polis was crucial in
guaranteeing the freedom of the citizen. Someone who failed to dis-
tinguish the polis from his own private household and presumed to
treat citizens as if they were slaves became a tyrant by falling victim to
his ‘grandiose desire for power’ and failing to maintain the boundary
between public and private (Nyquist 2013, 38). Slavery was entangled
in the structure of politics.
Nyquist argues that focusing on the structure of slavery within
the household ‘has the effect of naturalizing it’ by bringing it into the
same space as marital and parental relations, which are understood to
be determined by nature (Nyquist 2013, 25), but at the same time to
be social relations. The slave–master relationship within the house-
hold was ambiguous because the slave was classed as chattel and so
‘ostensibly belongs to the same category as nonhuman animals and
other possessions’ (Nyquist 2013, 25). As long as the master’s power
was directed towards the fulfilment of purely private needs, despotic
power was justified and it was accepted that the household master
was not accountable to others for how he used his power. Free citi-
zens were the masters of natural slaves. Nyquist identifies an opposi-
tion between those for whom slavery ‘would represent a demeaning,
traumatic loss and those for whom it was supposed to be natural’
(Nyquist 2013, 26). Aristotle ‘argues at one and the same time for
the categorical naturalness of household slavery and the unnaturalness
for Greeks of despotism in the political sphere, associating enslaved
barbaroi with both’ (Nyquist 2013, 49).
The big, behind-the-scenes question is about the where race fits in
to these ancient conceptions of slavery. This is, of course, an anachro-
nistic question, but the construction of the barbarism of non-Greeks,
and the ways in which they were represented as fitted for slavery are
fascinating and clearly have parallels with seventeenth-century nar-
ratives of the state of nature and with the eighteenth and nineteenth

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 17 12/01/18 6:03 PM

 18 The Politics of Slavery

centuries’ apologies for slavery. This brings us to the vexed and diffi-
cult question of how to make comparisons and draw parallels between
ancient Greece and, for example, nineteenth-century America. In one
sense, of course, this is impossible to do, but the temptation is almost
irresistible when you read and engage with the arguments of George
Fitzhugh, or William Harper and his defence of slave-owning in 1838,
where he makes self-conscious use of the ancients’ theory of natural
slavery and argues that slaveholders in Carolina should derive inspira-
tion and wisdom from the slavery model of Greece and Rome. Like
Aristotle, Harper sees the structure of slavery as resting on the cat-
egorical naturalness of household slavery and the unnaturalness of
attempting to enslave people who ought not to be enslaved. Harper
uses Aristotelian arguments about natural slavery to argue that ‘society
must exclude from civil and political privileges those who are unfitted
to exercise them, by infirmity, unsuitableness of character, or defect
of discretion’ (Harper 1838, 7). In his view, the civilised and cultivated
man had a right over ‘the savage and ignorant’: ‘It is as much in the
order of nature, that men should enslave each other, as that other ani-
mals should prey upon each other’ (Harper 1838, 11). Aristotle made
the same analogy between slavery and hunting. Harper’s question
about slavery was one that had been answered by Aristotle: ‘If there
are sordid, servile and laborious offices to be performed, is it not better
that there should be sordid, servile, and laborious beings to perform
them?’ (Harper 1838, 33).
As Page DuBois argues, slavery appears in classical historiographi-
cal work as a closed and static system, and that stasis needs to be
challenged through the recognition that slavery is not monolithic,
but has its own histories and variations (DuBois 2008, 25). In Greek
slavery, the distinction between helotry, as the collective enslavement
of conquered peoples who remained in their communities, and chat-
tel slavery, which brought slaves into individual households through
traumatic displacement, was central to the complex meanings of
slavery (DuBois 2008, 25). Then, there was a hierarchy among the
slaves. Some were trusted members of the household, others were
regarded as dangerous and hostile prisoners of war, and still others
worked down the silver mines or on grand public building projects.
There were slaves everywhere; public slaves worked in the police
force and picked up the bodies of the dead. State slaves were used

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 18 12/01/18 6:03 PM

 Aristotle and the Strangeness of Slaves 19

as the police force to restrain, arrest and detain citizens because ‘[a]n
important part of what being a citizen meant was not being manhan-
dled by other citizens’ (Fisher 1993, 56). Public slaves also worked as
managers of coin, weights and measures, as keepers of the archives
and as clerks and assistants to the council. They were paid regular
salaries and were able to accumulate some wealth. They are there in
the archaeological record, in literary, historical and theoretical texts.
Once we start to think about the ancient context, what does it mean
to conceptualise slavery as a place of no return, an order of nature,
especially in the context of manumission?
This means thinking about the meanings of the contested, in-
between statuses of freedom and the mobile borders between human-
ity and personhood. This is particularly interesting in the context of
ancient Greece and the processes of manumission. Rachel Zelnick-
Abramowitz asks what it is to be a manumitted slave. That is a ques-
tion that resonates throughout this book, but is posed most sharply
in this chapter. The question of what it means to be a slave is linked
to the question of what it means to stop being one. What lies on
the other side of the slavery/freedom binary that ancient Greece and
Rome did so much to construct and defend? (Zelnick-Ambramowitz
2005). The interesting thing is how many common themes emerge
here, and how blurry the line is between slavery and freedom even
as it is being drawn. In Aristotle’s theory of ‘natural’ slavery, we can
already trace the elements of freedom, belonging and labour that
intersect with each other to define what it means not to be a slave.

THE POLIS

The first and most important point to make is that Aristotle’s conception
of nature was ‘thoroughly teleological’ (Aristotle 1995, xi). The identity of
the polis lay in its organisation and structure, and this organisation was
the constitution. Human beings were polis-creating and polis-inhabiting
animals, and the city or the polis existed for the good life. The positive
moral purpose of the city was to enable citizens to live a life of virtue
or excellence, to make possible a life of Aristotelian moral virtue. From
the start, as R. F. Stalley points out in his introduction to the Politics, it is
implicit in Aristotle’s conception of the good life that not everyone can
achieve it, and that the institution of slavery is essential in guaranteeing

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 19 12/01/18 6:03 PM

 20 The Politics of Slavery

the possibility of a good life of virtue for some. When William Harper
looked back to the great republics of antiquity, the lesson he took from
them was that ‘slavery is compatible with the freedom, stability and long
duration of civil government, with denseness of population, great power,
and the highest civilization’ (Harper 1838, 45). The first natural form of
association was the family, the next was the village and the final or perfect
form of association was the polis, which ‘while it comes into existence for
the sake of mere life, . . . exists for the sake of the good life’ (Aristotle 1995,
1252b7). Because in Aristotle’s teleological theory the whole was neces-
sarily prior to the part, the city existed by nature and was prior to the indi-
vidual. For Aristotle, there was a natural impulse in all men towards an
association of this sort because man is a political animal. He is, by nature,
what Millett calls ‘a polis-creature’ (Millett 2007, 181), and the polis has
priority in nature over the household and the individual. This inspiring
vision of the city-state as the final or perfect form of association contin-
ues to define what many political theories mean by ‘politics’, and so what
it means to be constructed as outside the scope of the political.
The polis as an inclusive system of social ethics underpinned by
laws and unwritten rules was a shared, dynamic enterprise geared
to the highest goal: of living a good life. The polis was ‘a community
of persons who associate because of their need to make a living, but
who have as their goal the good life, i.e. a life of fulfilment exempli-
fying the characteristically human virtues’ (Schofield 1999, 103). The
exemplary humans who were members of the polis were assumed to
be free and equal and, as Malcolm Schofield puts it, capable of deter-
mining their own strategies for living (Schofield 1999, 103). Freedom
is rational self-direction. The free man could see for himself, through
the exercise of his reason, the ends he ought to pursue for the sake
of his own well-being (Walsh 1997, 499). This rational self-direction
is connected to democratic freedom in particular ways through the
shared capacity for deliberation. Such free men had the capacity to
live as they wished, to share equally in public responsibility and office,
and to be treated equally before the law. They were true citizens, not
mere subjects, meaning that they were allowed to share in delibera-
tion and decision, and that the good of the polis included their own
good (Walsh 1997, 501). In a democratic polis, all the free inhabitants
are ‘full citizens, sharing in decision and office’, and pursuing a com-
mon good that includes their own good. The flourishing of a properly

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 20 12/01/18 6:03 PM

 Aristotle and the Strangeness of Slaves 21

ordered city is the highest good, and perfect freedom is to contribute

to its flourishing, and to flourish within it (Walsh 1997, 503).
As Schofield argues, these egalitarian principles opened up the
potential for hierarchy, and some people who were not slaves, such
as farmers, manual workers and people engaged in trade, lived lives
that were devoted to ignoble purposes, or left them no leisure for
noble pursuits. Such men were only able to think in terms of wealth
and freedom, rather than of virtue and excellence, and so it was clear
that they should be excluded from citizenship (Schofield 1999, 106).
The status of manual labour was demeaning in Ancient Greece.
Greek cities developed as communities of farmers, and the ideal citi-
zen was a soldier and an independent, nearly self-sufficient farmer.
As Athens developed its craft and manufacturing, farming remained
the most respectable basis for wealth. Working the land was regarded
as more gentlemanly and masculine than manufacture, and as a bet-
ter preparation for military action. These equal, democratic citizens
were comfortable and materially secure, they listened to reason and
they did not envy others (Patterson 1991b). The slave owner who was
also a free citizen was capable of self-direction and of directing oth-
ers, and the free citizen was also a slaveholder.
In Book 1 of the Politics, the basic unit of the polis is the house-
hold, which includes husband and wife, father and children and mas-
ter and slave. Women, children, slaves and animals are all members
of the polis ‘constructed as a geographic, agricultural, social, religious,
productive and reproductive community, though not as a political
community’ (Nyquist 2013, 49). In Book 3, the polis is less inclusive,
and the basic unit of politics is the male citizen. The polis becomes
a political community, whose end, the good life, excludes slaves and
animals. Against this background, slavery appears as an institution
of benefit to the master, and its ‘individual, private character is sud-
denly thrown into relief’ (Nyquist 2013, 49).The equality of politi-
cal rule envisaged by Aristotle creates the polis as a community of
those who are capable of virtue, where ‘equality is secured only at the
price of hierarchy: the subordination of those who perform the lower
functions’ (Schofield 1999, 112). Women and chattel slaves were rel-
egated to the household, and placed ‘below the threshold of political
discourse’ (Schofield 1999, 110). Women were incorporated into the
oikos under the guardianship of their male kin. They were barred from

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 21 12/01/18 6:03 PM

 22 The Politics of Slavery

legal proceedings and from making contracts to dispose of significant

amounts of property, and they were not polis-creatures. Women and
slaves were understood to be naturally subordinate beings in relation
to free men, and free men were intended by nature to exercise per-
manent leadership over them. Women must obey because of men’s
innate inability to command, and because the good life of the polis
concerns only free men ‘whom women must serve as a functional
contribution towards the males attaining perfection’ (Femenias 1994,
170). Women played a vital role in transmitting citizenship rights and
in contributing to the survival of the oikos. The polis depended on the
presence of slaves and foreigners who were not members to sustain
its own vision of itself as an inclusive system.

THE NATURE OF RULE AND THE NATURE OF THE SLAVE

Schofield argues that the dominant question throughout the Politics

is about how many forms of rule there are, and Aristotle’s answer is
that there are several. His interest in slavery arises only in the con-
text of ‘his preoccupation with the different forms of rule’ (Schofield
1999, 132). These different forms of rule are grounded in the nature
of human beings, and slavery works at the limit, representing ‘the
extreme case in a range of cases in natural rule’ (Schofield 1999, 132).
Aristotle argued for a natural basis for the difference between slavery
and political rule, and made a sharp distinction between rule over
slaves and the rule of equals in the political process.
For Aristotle, there must necessarily be a pairing of those who
cannot exist without one another. He was aiming for the union of the
naturally ruling element with the element that was naturally ruled for
the preservation of both: ‘The element which is able, by virtue of its
intelligence, to exercise forethought, is naturally a ruling and master
element; the element which is able, by virtue of its bodily power, to
do the physical work, is a ruled element, which is naturally in a state
of slavery; and master and slave have accordingly a common interest’
(Aristotle 1995, 1252a24). Masters and slaves had a different kind of
knowledge. Slaves were instructed in the nature of their duties, and
could learn to be skilled and proficient in cookery and domestic ser-
vice, but the character of their knowledge remained servile. Masters
of slaves must simply know how to command what the slave must

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 22 12/01/18 6:03 PM

 Aristotle and the Strangeness of Slaves 23

know how to do. This knowledge, Aristotle points out, is not great or
majestic, and many free men delegated the management of slaves to
a steward, and spent the time they saved on politics or philosophy.
The question then arose of whether the slave had any ‘goodness’
or virtue beyond that of discharging his or her function as an instru-
ment and performing his or her menial service. Did they possess
goodness of a higher value, such as temperance, fortitude or justice,
or did they have no virtue beyond the bodily services they provided?
Either alternative presented difficulties for Aristotle’s account, and
for the theory of natural slavery. If they did possess such virtues, how
were they different from free men? If they did not, how could they be
characterised as human and as possessing reason? For Aristotle, the
difference between those who were naturally ruled and those who
naturally ruled was a difference in kind. The subject could only be
properly ruled if he was temperate and just, and so those who were
naturally slaves shared in goodness, but their goodness was of a dif-
ferent kind. Free men, women and slaves shared in moral goodness,
‘but not in the same way – each sharing only to the extent required
for the discharge of his or her function’ (Aristotle 1995, 1260a4). The
goodness of the slave was all about his relation to his master. Slaves
were useful for the necessary purposes of life, and to discharge that
function they needed just a little goodness: ‘only so much, in fact, as
will prevent them from falling short of their duties through intem-
perance or cowardice’ (Aristotle 1995, 1260a33). Slaves lacked the
prohairesis that enabled moral choice in advance of action (Millett
2007, 185). Prohairetic activity combined desire and intelligence and
disclosed the character of the one who acted and allowed him to
live a ‘life based on choice’ (Frank 2004, 96). Natural slaves could not
engage in this kind of activity, and instead had to have their choices
made for them by someone with foresight. The slave ‘is a partner in
his master’s life’ (Aristotle 1995, 1260a33), and once he was placed
in that relation, he could mirror or approximate prohairetic activity
(Frank 2004, 96).
It was better for masters that natural slaves existed because it
made it possible for them to live better lives. It was an incidental
benefit that being enslaved was good for natural slaves, making them
‘capable of participating in and contributing to an intrinsically worth-
while life’ (Heath 2008, 266). Dobbs argues that once we put the slave

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 23 12/01/18 6:03 PM

 24 The Politics of Slavery

relationship into its proper teleological context, it becomes clear that

it ‘in no way involves the dehumanization of the slave’ (Dobbs 1994,
87). For Aristotle, despotic rule exercised in accord with nature was
not exploitative, and the slave was both ‘property and partner of his
master’. Through his relationship to the master, the slave shared in
a distinctively human way of life and, according to Dobbs, ‘property
in a natural slave derives from the fact that all human beings belong
in a life partaking in the distinctively human telos’ (Dobbs 1994, 87).
Dobbs argues that it was the master’s responsibility to ‘bring out such
virtue as the slave can achieve’ (Dobbs 1994, 87), and he should do
so out of concern for the excellence of his property, not just in order
to maximise output. The rightfulness of the master’s dominion was
‘conditioned upon the subordination of his own conduct to the natu-
ral order’ (Dobbs 1994, 88). Dobbs reads a kind of ‘humanisation’ of
the slave into the partnership that resurfaces in Harper’s nineteenth-
century defence of slavery when he talks about the ‘virtues of slaves’,
which include fidelity, submission to authority and the disposition
to be attached to superiors. The slave, he declared, had no need for
heroic virtues or elegant accomplishments. It was for the master ‘to
compensate for this, by his own more assiduous cultivation, of the
more generous virtues, and liberal attainments’ (Harper 1838, 30).
This kind of despotic rule was only justified as right by nature as
long as ‘the master’s proprietorship in the slave derives from and duly
respects the nature of the slave qua human being’ (Dobbs 1994, 86).
Those who were fitted by nature to be slaves possessed only bodily
powers and the faculty of understanding ‘the directions given by
another’s reason’. The natural slave lacked nous, and ‘has no insight
into what is fundamental’ (Dobbs 1994, 86). In Aristotle’s schema,
the soul ruled the body with the authority of a master, reason ruled
the appetite with the authority of a statesman. The body should
be ruled by the soul, the affective part of the soul should be ruled
by the rational part, and in the same way, animals should be ruled by
man and the female ruled by the male. For Plato, men and women all
had the same virtues, exercised in different realms, and to be a ruler
required the exercise of these same virtues. Anyone could (in principle
for Plato) become a ruler, and anyone could be ruled. For Aristotle,
virtue was different in different kinds of people, and ruling was differ-
ent in different contexts, so that ‘what it is to rule well cannot simply
be a matter of science’ (Deslauriers 2006, 59). The master was a master

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 24 12/01/18 6:03 PM

 Aristotle and the Strangeness of Slaves 25

because he was a certain kind of person, who possessed phronesis, a

‘practical intellectual virtue that informs moral character’ (Deslauriers
2006, 61). The desires of a free man were informed by the dictates of
his reason. The naturalness of slavery came from these differences in
intellectual and moral virtues, the distribution of these ‘psychic facul-
ties’ that rendered people capable of ruling well and with authority
(Deslauriers 2006, 62). All men who differed from others as much as
the body differs from the soul, or an animal from a man, ‘all such are
by nature slaves’ (Aristotle 1995, 1254b16). For these people who were
slaves by nature, it was better to be ruled by a master. For those whose
function was bodily service, they produced their best when they sup-
plied such service to their masters; ‘Someone is thus a slave by nature
if he is capable of becoming the property of another’ (Aristotle 1995,
1254b16). This definition of what it means to be a slave by nature
resonates through the centuries, putting the focus clearly on the defi-
ciencies of the slave, gendering the slave as male, describing what it
means to be ‘slavish’, not what it means to try to force someone else
to become your property. The slave apprehended reason in another,
but had none of his own. The deficiency of the natural slave was
‘his failure to actualize the first-level capacity for logos he possesses’
(Frank 2004, 96).The part and the whole, like the body and the soul,
had an identical interest, and the slave was part of the master, ‘in the
sense of being a living but separate part of his body’ (Aristotle 1995,
1255b4). Moira Walsh gives the example of her hand. It has no inher-
ent purpose and cannot direct itself, but receives its purpose ‘from
me or my intellect, which commands it to move in a particular way’
(Walsh 1997, 498).

THE SLAVE AS AN ANIMATE ARTICLE OF PROPERTY

Aristotle located the slave as property within the household. Prop-

erty was part of the household, and the art of acquiring property was
part of household management. The household had to be furnished
with appropriate instruments if its function was to be fulfilled, and
those instruments were partly inanimate and partly animate: ‘Each
article of property is thus an instrument of the purpose of life, prop-
erty in general is a quantity of such instruments, the slave is an ani-
mate article of property’ (Aristotle 1995, 1253b23). This is Aristotle’s
most famous formulation of what it means to be a slave, the slave as

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 25 12/01/18 6:03 PM

 26 The Politics of Slavery

a living tool. The description sums up what is understood to be the

‘problem’ of slavery: the (im)possibility of reducing a human being
to the status of an object. As Schofield argues, the ‘ensouled tool’ is
not the name for a distinct species of animal, but a way of describing
‘a perfectly recognisable sort of human being’ (Schofield 1999, 128),
just not the sort of exemplary and virtuous human being who can
inhabit the polis. In belonging to someone else, slaves were, accord-
ing to Vincent Rosivach, ‘something less than fully human, and prob-
ably closer to animal than to free man’ (Rosivach 1999, 146). Malcolm
Bull argues that the slave’s incomplete soul was not enslaved by the
master’s soul, but subject to it, so that the slave was still another soul.
Aristotle constructed ‘a justification of the enslavement of those whose
humanity is somehow incomplete while retaining an awareness that
slavery does not obliterate such humanity as they possess’ (Bull 1998,
4). Someone who belonged to another person worked primarily for an
external end, with no innate purpose or reason for being, but a ‘being-
for-other’ remained a human being (Femenias 1994, 169).
Aristotle made a distinction between the slave as a slave and
the slave as a man. He made some room for justice in every rela-
tion between humans who were capable of participating in law and
agreement (Millett 2007, 186). Slaves needed to have a small amount
of appropriate virtue, developed by their masters, and to be given
reasons for their instructions. In Bull’s analysis of the multiple self in
slavery, this means that Aristotle captured something of the duality
inherent in the condition of the slave. Fisher argues that the idea of
friendship with a slave as a man ‘brings out the fundamental con-
tradiction very clearly’ (Fisher 1993, 97), and is compounded by the
‘single most glaring contradiction’, the use of manumission as an
incentive for all slaves. The slave was, potentially at least, fully human.
The relationship between master and slave created a community of
interest, a relationship of friendship between the master and slave
‘when both of them naturally merit the position in which they stand’
(Aristotle 1995, 1255b4). Zelnick-Abramowitz’s analysis of friendship
draws attention to the bonds of service and loyalty, and the wish to
benefit the other. She argues that these relations of benefactors and
beneficiaries could exist between equals and between unequals, and
philia could exist in vertical and asymmetrical relationships. Slaves
as human beings in relations with their masters depended on each

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 26 12/01/18 6:03 PM

 Aristotle and the Strangeness of Slaves 27

other, and on social connections based on suspicion, fear, co-oper-

ation, expectations and obligations. Faithful and resourceful slaves
expected the freedom they had been promised by their masters (Zel-
nick-Abramowitz 2005). In return for their promised manumission,
slaves were encouraged to be loyal and diligent, and then often found
themselves still bound by their debts and by the conditions of their
manumission long after they were supposed to have gained their
independence. Manumitted slaves remained ‘essential, yet outsiders’
(Zelnick-Abramowitz 2005, 60).
The slave as an animate tool was an instrument of action. Aristo-
tle recognised different kinds of rationality, practical and technical.
Natural slaves could not achieve eudaimonia, the best kind of human
life, because it consisted in virtuous activity, beyond the scope of tech-
nical rationality. Natural slaves did not share in the practical, architec-
tonic wisdom that provided overall guidance for life (Heath 2008, 247).
They were incapable of prohairesis and eudaimonia, of virtuous action
or deliberated choice, because they were characterised as not able to
reason back from a goal to the action required to implement that goal.
The slaves’ deliberative capacity was permanently impaired. They lived
without the guidance of a stable conception of the overall good (Heath
2008, 251). Aristotle’s theory of chattel slavery within the private realm
of the household made clear that slavery is about ownership: ‘While the
master is merely the master of the slave, and does not belong to him,
the slave is not only the slave of his master; he also belongs entirely to
him’ (Aristotle 1995, 1254a8). It was a slave’s nature to belong to another,
and his nature was fulfilled only when he actually belonged to another
person (Rosivach 1999, 146). As Zelnick-Abramowitz argues, it is crucial
that we understand slavery as a complex set of social relations, and not
just as a straightforward property relationship. At the same time, it is
important that we take the property element seriously, and recognise
that it is the asymmetry of property and ownership that Aristotle uses to
construct his vision of natural slavery:

Anybody who by his nature is not his own man, but another’s,
is by his nature a slave; anybody who, being a man, is an article
of property is another’s man; an article of property is an instru-
ment intended for the purpose of action and separable from its
possessor. (Aristotle 1995, 1254a13)

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 27 12/01/18 6:03 PM

 28 The Politics of Slavery

The slave, as DuBois argues, is ‘a sort of uncanny object, standing at

the blind spot of modernity where the place of the subject and that of
object intersect’ (DuBois 2008, 31).
The slave in this account is alive but socially dead, ‘a personality
without personhood’ (Bull 1998, 17). The slave is at the same time both
a tool and a human being, and Aristotle ‘inscribed the social contra-
dictions of slavery within the soul(s) of slaves’ (Bull 1998, 17). This is
a controversial and disputed reading of Aristotle which links him to
Hegel and to W. E. B. DuBois, but it makes an interesting point about
social death and about what Bull calls ‘the duplicated selves formed
through slavery’ who do not fit the model of unified, authentic indi-
viduals (Bull 1998, 18). Aristotle’s theory of natural slavery lays the
ground for later theories based on the premise that selves that do not
fit or are somehow incomplete will be excluded ‘from consideration
as moral subjects’ (Bull 1998, 18). ‘From Aristotle onwards, having a
unified moral self has been seen as a privilege confined to a social
elite’ (Bull 1998, 20). Bull is making a further point about the rela-
tionship between interpersonal and intrapersonal models of slavery.
In both Aristotle and Hegel, he argues, the master–slave dyad is an
interpersonal relation constructed as an intrapersonal one. Freedom
in Aristotle is the freedom of rational self-direction both within the
polis and within the soul. Individual deficiency, for both slaves and
women, entails political exclusion. Not having a unified moral self
will involve not being a citizen, so that the privilege is confined to an
explicitly political as well as social elite. The distinction between citi-
zens and inhabitants turns out to make the number of beneficiaries
of Aristotle’s inspiring vision of the polis as an inclusive system much
smaller than we might at first imagine, and opens up the question
of the humanity of the slave, and of the process of dehumanisation.

SLAVERY AND LABOUR

The great majority of slaves in Athens were household slaves, in the

sense that they were formally the property of an individual oikos.
Large numbers of them worked in the fields, especially at harvest
time, and the enlarged household in Athens included slave-craftsmen
and artisans. Outside the confines of the household, there were hired-
out slaves and those employed in the mines (Millett 2007, 203). They
were all part of a differentiated and hierarchical system of slavery, with

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 28 12/01/18 6:03 PM

 Aristotle and the Strangeness of Slaves 29

domestic slaves at one end of the continuum and slave miners at the
other. As Millett and Nyquist point out, Aristotle’s main concern was
with the potentially close relationships and the complex psychological
interactions of the household.
In his discussion of slavery in Roman history, Keith Bradley
describes the dehumanising process of buying and selling slaves in
the marketplace. Slaves stood on display on a raised platform, and
potential buyers could insist that they jump up and down or undress.
They were subject to intense scrutiny, and had to disclose any illness,
wounds, scars or sores, deformities, pregnancy or menstrual difficul-
ties. It seems, Bradley argues, a ‘reasonable inference that the physical
examination on the castata reduced the slave to the level of an object
– an object that was generally mute, passive, and devoid of any human
dignity’ (Bradley 1992, 129). Bradley concludes that ‘it was as though
the slave were in fact an animal’ (Bradley 1992, 129). Sales of oxen, cat-
tle and mules, he points out, required similar disclosure of diseases and
defects, and jurists made no distinction between animal and human.
This chapter and the rest of the book will return to the ‘as though’, the
‘as if’ construction of conflating slaves and non-human animals, but
Bradley’s wider point about the brutality of slavery in the ancient world
is important.
The wealth of the rich in Athens was traditionally based on small-
scale, scattered landed estates, and there is debate over how much
slave labour was employed in agriculture. Fisher’s survey of the evi-
dence in the sources concludes that ‘all prosperous and rich Athenians
owned some slaves, as domestics and status-symbols’ and would have
used them for agricultural work at peak periods (Fisher 1993, 41).
He paints a fairly fluid picture of slave ownership, with some poorer
households investing in slaves when they were prospering and sell-
ing them again when times got hard. Slaves seem to have cost about
twice as much as a cow or an ox, though they were cheaper than a
mule, and poor families would have required a sizeable loan to make
the purchase. It may well have been worth it for them to employ a few
all-purpose slaves much of the time, both for the labour provided and
the status it gave them in terms of freedom and citizenship.
Large numbers of ‘the most expendable’ slaves were employed
in the state-owned silver mines (Fisher 1993, 49) doing the most
unpleasant, dangerous and unhealthy work. The less dangerous work
at the surface was done by less expendable slaves who were bought

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 29 12/01/18 6:03 PM

 30 The Politics of Slavery

or hired by the people who leased the land. Slaves were also used in
manufacturing weapons, pots, statues, knives, lamps and clothes, and
manufacture of such items was an important source of ‘new wealth’
(Fisher 1993, 50). It is interesting that slaves, even in ancient Greece,
were associated with new as well as landed wealth. This seems to me
to undermine some of the assumptions behind the old/new slavery
split as it operates now. Slavery appears from the start as a flexible,
adaptable form of labour and source of wealth. Occupations such as
artisanal crafts, smithing or potting were considered to be demean-
ing, and living by buying and selling encouraged double-dealing
and misrepresentation (Fisher 1993, 100). Manufacturing warped the
spirit of the free Greek, rendering it slavish ‘since the work was done
inside, in the dark and crouching over, not outside in the open air
with the upright posture of a free man’ (Cartledge 1993, 173). It was
shameful to be poor and compelled to perform such slavish labour.
This brings us back to Harper’s description of sordid, servile and
laborious beings, fitted by nature for sordid, servile and hard work.
Like a domesticated animal, the slave-by-nature helped his or her
master by means of their body (Rosivach 1999, 147).
The identification of the slave with the body was central to what it
meant to be enslaved in Athens, and it extended beyond their sale and
their labour. The testimony of slaves in an Athenian court was admissible
only if it had been extracted under physical torture. As Cartledge points
out, this was because, by definition, only free men could tell the truth:
‘Against the pristine purity of the self-controlled democratic citizen’s
body there was counterposed ideologically the aboriginal impurity of
the un-self-controllable servile body – a body which was controllable,
moreover, only by the master’s whip’ (Cartledge 1993, 175). Corporal
punishment was a central and defining feature of what it meant to be
a slave. Slaves were understood to lack the moral capacity to lie or to
tell the truth without being compelled to do so, ‘and so the truth must
be sought from their bodies’ (Rosivach 1999, 152). Giving evidence in
court was a signal of free status, of having witnesses who were prepared
to vouch for your story. Open access to the courts, as Fisher argues,
was part of what it meant to be free, and so had to be denied to slaves
(Fisher 1993, 60).
Slaves were often treated legally as the property of their masters
and bought, sold, bequeathed or confiscated ‘like land or beds’, as

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 30 12/01/18 6:03 PM

 Aristotle and the Strangeness of Slaves 31

Fisher puts it (Fisher 1993, 62). They could not form sexual relation-
ships without their masters’ approval, their partners or children might
be legally sold, and they were precluded by law from exercising in
wrestling grounds or being the lovers of free-born boys. Slaves, Fisher
concludes (with an implicit assumption that the slaves are male), ‘had
to accept that their bodies might be penetrated by free males, but
they might not themselves penetrate the bodies of the free, male or
female’ (Fisher 1993, 62). At the same time, the law assumed that
slaves were persons, and had some minimal honour that deserved
to be protected. The death of a slave was treated legally as the kill-
ing of an inferior person. If the slave were killed by someone other
than the master, the master could prosecute the killer for homicide
(Fisher 1993, 62). It is important to remember that the process of
dehumanisation was never complete, even where slaves were most
closely identified with property, limited to their bodies and described
as animate tools. It was impossible for them not to be involved in
entangled interrelations, not only of chattel and political slavery, but
also of social and property relations.

NATURE AND SLAVERY

For Aristotle’s structure of slavery to work, to function smoothly, nature

needed to erect a physical, tangible difference between the bodies of
freemen and the bodies of slaves: ‘But nature, though she intends, does
not always succeed in achieving a clear distinction between men born to
be masters and men born to be slaves’ (Aristotle 1995, 1254a13). Nature
was able to distinguish slaves from non-slaves, but Frank argues that
nature ‘secures no absolute boundaries and offers no permanent foun-
dations’ (Frank 2004, 96) because nature continued to be guided and
determined by activity. Natural beings could always be otherwise, and
claims about their identities were claims about their activities, and our
activities are changeable and in constant interaction with our circum-
stances (Frank 2004, 98). The contrary of nature’s intention often hap-
pened, and so Aristotle argued that slavery by nature was a matter of the
differences between souls, so that he could conclude that ‘just as some
are by nature free, so others are by nature slaves, and for these latter the
condition of slavery is both beneficial and just’ (Aristotle 1995, 1254b39).
Aristotle made a distinction between natural slavery and conventional

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 31 12/01/18 6:03 PM

 32 The Politics of Slavery

slavery, and rejected the idea that just slavery could be based on force.
For Aristotle, it was a question of enslavability, of who naturally deserved
to be in a condition of slavery. If slavery became about force rather than
desert, then it would risk enslaving men of higher rank if they happened
to be captured and sold. The possibility of slavery as a result of war con-
tradicted the idea of natural slavery, and the categorical claim that ‘there
are some who are everywhere slaves, and others who are everywhere
free’ (Aristotle 1995, 1255a21). Those who were ‘everywhere slaves’ were
barbarians. Aristotle pointed out that among barbarians, the female and
the slave occupied the same position because there was no naturally rul-
ing element among barbarians.
The assumption was that barbarian and slave were by nature one
and the same. For Aristotle, barbarians were naturally slavish, and
since most of the slaves in Athens were barbarians, Aristotle can be
characterised as accepting that most slaves in his own society were
natural slaves (Schofield 1999, 133). Athenian chattel slaves were
barbaroi, either born outside Greece or born in Greece of forbears
who were born outside Greece. Usually, they had names that marked
them as barbaroi, ethnic labels that were used to denote a ‘typical
slave’ (Rosivach 1999, 129). The Greeks’ stereotypes of different peo-
ples were a product of their interactions with them (E. Hall 1989,
108). As the barbaroi lived for longer amongst the Greeks, the domi-
nant myths of slavery changed, and instead of being characterised
as speaking different languages, they were coded as ‘language-less’,
not speaking any intelligible language at all, but making the sound
of twittering birds (Rosivach 1999, 153). Not speaking Greek became
a sign of their intellectual deficiency, their inability to command
themselves and their lack of interior logos. The natural slave was
described as always having his neck bent and slanting, and as dis-
proportionately small, ugly or tattooed. The archetypal slave was
Aesop, a Thracian who was imagined as ‘pot-bellied, weasel-armed,
hunchbacked, a squalid, squinty, swarthy midget with crooked legs’
(Millett 2007, 196). The Scythian hinterland was seen as remote,
intractable, desolate and untamed (E. Hall 1989, 114), and its inhab-
itants were coded both as nomads, armed with powerful bows, and
as unspoiled innocents living in a well-governed utopia. The barbar-
ians were portrayed as impudent, rash and unsophisticated, relying
on the ‘outdated machismo’ of their monarchs that was no match for

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 32 12/01/18 6:03 PM

 Aristotle and the Strangeness of Slaves 33

the ‘covert and cunning activities of the Greeks’ (E. Hall 1989, 122–3).
Many of the stereotypes surrounding the invention of the barbar-
ians coalesced around their inability to restrain their passions. The
failure to control his sexual desire was the standard trait of the
barbarian male in Greek writers. He was a tyrant who let loose the
savage appetites in his soul (E. Hall 1989, 125).
Edith Hall makes the important point that as well as being con-
structed as savage, violent and highly masculinised, at other times
barbarians appear as excessively refined and effeminate, luxuriating
in exotic clothing and sumptuous funerals. Their concern for comfort
and their excessive displays of grief were part of the ‘vocabularies of
barbarism’ (E. Hall 1989, 128) which emphasised the Greek ‘appropri-
ation of moderation’ against different kinds of extremes (E. Hall 1989,
127) and the ‘systematic feminization of Asia’ as emotional and sub-
servient (E. Hall 1989, 157). Hall argues that the Greeks’ view of the
barbarians was inherently contradictory, incorporating ‘the idea not
only of primitive chaos, but of a more virtuous era, when men were
nearer to the gods’ (E. Hall 1989, 149). The Greeks’ idea of the past
overlaps with their conception of ‘the elsewhere’, and they lived with
and produced a barbarian world of their own that was the home both
of savages and tyrants and of idealised, innocent peoples with a har-
monious relation to heaven (E. Hall 1989, 149). In terms of thinking
about the parallels between ancient Greek slavery and the slavery of
the eighteenth and nineteenth centuries, this vision of ‘the elsewhere’
is strikingly similar to the ways in which Africa was imagined as an
uncorrupted state of nature and as a space of savagery and degenera-
tion. For both imperial Greece and colonial Europe and America, the
people they enslaved came from a place of no return. Societies outside
Europe are positioned in what Nyquist terms ‘a privative age’, a pre-
political space (Nyquist 2013, 16), without ordered social relations, ‘a
state where slavery is at home’ (Nyquist 2013, 17). For Aristotle, man
achieved completion as an exemplary human as part of the city, as a
citizen (S. D. Collins 2006, 172). He ‘is worst of all when apart from
law and justice’. In Aristotle, slaves ‘and other creatures’ were explic-
itly denied the ability to constitute a polis (Millett 2007, 182). Without
the polis, men were rendered unscrupulous and savage. For Jill Frank,
the distinction between Greeks and non-Greeks rested not on their
immutable natures, but on Aristotle’s observations of the behaviours

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 33 12/01/18 6:03 PM

 34 The Politics of Slavery

of those foreigners. The proper determinant of slavery, as we have

seen, was not foreignness, but worthiness or character and it was
these internal characteristics and activities that could justify enslave-
ment (Frank 2004, 101–2). Asians, because they lived in hot climates
and under hereditary tyrannies, tended not to act on their own initia-
tive. Living under tyrannies rendered them naturally slavish, and this
brings us back to the entangled interrelations between chattel and
political slavery that define non-Greeks as natural slaves.
Rachel Zelnick-Abramowitz’s study of manumitted slaves in ancient
Greece draws attention to the complex statuses in between freedom
and slavery that unsettle the freedom/slavery boundary. Freed slaves
found themselves subjected to a range of conditions, often including
being obliged to remain with their ex-masters for a fixed period. On
leaving slavery, they were not wholly free, but in a state of semi-slavery,
‘a twilight zone between the completely free and the completely non-
free’ (Zelnick-Abramowitz 2005, 6). To understand this twilight zone,
Zelnick-Abramowitz argues that we need to define slavery in social
terms rather than as a property relationship. Manumission was a social
transaction involving exchange and reciprocity, and social bonds that
protracted dependence. Manumitted slaves did not become citizens,
they were barred from taking part in the political process and from land
ownership, and they were required to pay a poll tax that marked them
as inferior to citizens and to metics and continued to bind them to their
former masters. It is important to recognise here the complex social
realities behind the binary of slavery and freedom. The metic was less
free than the citizen; the foreigner was less free than the metic; the helot
was regarded as less slavish than the chattel slave. Zelnick-Abramowitz
extends Moses Finley’s spectrum of statuses at the free end, insisting
that ‘freedom itself has different shades’ (Zelnick-Abramowitz 2005,
38). Freedom and slavery are better conceived of as dependence and
independence, as economic, moral, internal and external, rather than
just as property relations.

CONCLUSION

Jill Frank draws attention to the complex relation between nature and
politics in Aristotle’s theory, and her arguments are important for con-
structing a historical and dynamic politics of slavery. In her view, Aristotle
treats nature as a ‘question for politics’ (Frank 2004, 92). Human nature

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 34 12/01/18 6:03 PM

 Aristotle and the Strangeness of Slaves 35

is, at least in part, constituted politically and nature and politics are both
changeable. This has several important effects. First, it means that citizens
are involved in constructing their own identities through making and
doing, ‘where doing is a kind of self-making’ (Frank 2004, 94). Citizens
are made citizens by the collective activity of sharing a constitution and
making institutions. The polis is an active, dynamic space, full of individ-
uals acting in concert. The fundamental difference between citizens and
slaves is that slaves ‘are the product of citizen activity alone’ (Frank 2004,
95). What starts off looking like a defence of a static model of slavery as
based on unchanging constituent elements turns out to be a theory that
it is politics that produces the institutions that help make citizens and
slaves. Human beings and the polity itself emerge ‘as both natural and
made’ (Frank 2004, 99).
On this reading, human nature is vulnerable to and shaped by both
politics and by self-determining activity (Frank 2004, 102). Aristotle’s
defence of natural slavery ‘at the same time serves as a warning about the
dangers slavery poses to politics’ (Frank 2004, 102). In the end, accord-
ing to Frank’s interpretation of Aristotle, Greeks can become slaves if
they act like slaves. Even the hierarchies of the polity and the household
are insecure and reversible. Aristotle is offering not immutability, but
perpetual ‘boundary-setting and keeping’ (Frank 2004, 102). This is an
important corrective to a static conception of slavery as fixed by nature,
and opens up for discussion what DuBois (2005, 109) calls ‘the strange-
ness of slaves’, their unsettled status at the boundary of subject of object,
their instability as both property and partner. It shows us how problem-
atic the idea of ‘dehumanisation’ can be when it gets divorced from these
intensely political questions of boundary-setting and keeping. It takes us
back to Aristotle’s own emphasis on the participatory and deliberative
character of politics and challenges us to think again about our capacity
to see which means will lead to the given ends. In thinking about the
spaces in between slavery and freedom, and in between the household
and the polis, women, slaves and barbarians emerge from their privative
state of nature into‘a complex, cross-institutional, cross-discursive’world
of politics (Nyquist 2013, 27). At the same time, I am not convinced that
the boundaries of this space are so porous that Greeks could become
slaves if they acted like slaves, or that all the relations between master
and slave were social rather than property relations. There is something
about Aristotle, about the nature of political rule and the structure of
the soul, and about the private world of the individual set against the

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 35 12/01/18 6:03 PM

 36 The Politics of Slavery

public world of the citizen, that fixes that difference between those who
are considered as full moral subjects and those who are lacking, dupli-
cated and incomplete. In the next chapter, the focus is on Locke and
his understanding of the space in between slavery and freedom, and in
particular the borderland between the state of nature and civil society.
The social relations of slavery in the seventeenth century continued to
be intensely political, and deeply concerned with boundary-setting and
keeping, even as the ground shifted over the nature of slavery and the
intersections between subject and object.

This content downloaded from 128.122.149.96 on Thu, 14 May 2020 18:25:02 UTC
All use subject to https://about.jstor.org/terms

5606_Brace.indd 36 12/01/18 6:03 PM