Catherine hezser introduction
Although mass (chattel) slavery was a specifically Roman phenomenon in antiquity, Jews do not seem to have been less affected by this common and all-pervasive institution than Romans, Greeks and other provincials in whose immediate environment they lived. Ancient Jewish literary sources suggest that Jews were both slaves and slaveholders in antiquity, and that slavery had a major impact on almost all areas of Jewish daily life, whether in Hellenistic and Roman Palestine or in the Mediterranean Diaspora. The phenomenon of slavery shows how and to what extent Jews were part of Greco-Roman society while at the same time maintaining their biblical roots. It is therefore of particular interest to examine similarities as well as differences between the Jewish and Greco-Roman discourse on slavery to determine whether there was a particularly Jewish perspective and approach.
This chapter will focus on the Jewish experience and rhetoric of slavery from post-exilic to Roman imperial times and late antiquity (fifth century bc to fifth century ad). Previous scholars have usually assumed that slavery was insignificant in ancient Judaism of the post-biblical period. Since the Hebrew Bible refers to Jewish slaves as temporary bondsmen who are to be released in the seventh or Sabbatical Year (cf. Exod. 21:2—3; Lev. 25:40—41; Deut. 15:12), it was commonly assumed that Jews could not be proper slaves of Jews and that Jewish slaveholders treated their gentile slaves as slaves only. Some scholars argued that after the Babylonian Exile the institution of Hebrew bondage came to an end, and that Jews advocated a particularly humane treatment of their non-Jewish slaves. Scholars such as Farbstein (1896) who propagated such theories usually wrote in the second half of the nineteenth or the beginning of the twentieth century and were sympathisers of the Reform movement in Judaism. They wrote in a German-Jewish context in which Jews were eager to assimilate into a predominantly Christian society. In this context, they were eager to show that the moral values represented by Jewish traditions were equal or even superior to those of their Christian fellow-citizens. The
438
Claim that ancient Jewish sources reveal a proto-abolitionist stance towards slavery can therefore be considered apologetic rather than historically persuasive.
Although the two studies which followed in the 1960s (Zeitlin 1962—3; Urbach 1964) acknowledge the significance of slavery in Judaism throughout antiquity, they did not yet share contemporary scholars’ concerns for a historical-critical examination of the sources, especially as far as the rabbinic material is concerned (Mishnah, Tosefta, Palestinian and Babylonian Talmud, Midrashim). It has been recognised since then that references to slaves and slavery in ancient Jewish sources must primarily be understood as literary discourses revealing their authors’ and editors’ worldviews and ideologies, rather than being taken as historically reliable depictions of everyday life. What actually happened can only be assumed hypothetically, on the basis of certain recognisable patterns in the sources. Only those literary works which originated in the Greco-Roman cultural environment can be properly interpreted in the Greco-Roman context. This means that slavery in Babylonian Jewish society, for which the proper context would be ancient Persia, is excluded from the discussion here.612
With the exception of a few inscriptions from Rome and elsewhere, most sources on ancient Jewish slavery are literary in nature. They consist of Greek Jewish writings from the Hellenistic and early Roman period, including the works of Philo and Josephus, and Palestinian rabbinic literature from the Mishnah (edited around ad 200) to the Talmud Yerushalmi and amoraic Midrashim (fourth-fifth centuries ad). None of these texts can be considered to provide historically reliable information on the actual use and treatment of slaves by Jews in antiquity or allow us to determine the exact percentage of slaves within ancient Jewish society. What the literary sources do allow us to determine is the ancient Jewish discourse on slavery and the moral values and ideology on which it is based. Recurrent patterns of behaviour in different literary forms are especially relevant for social historical inquiries.
Jews as slaves and slaveholders in antiquity
Throughout the period under discussion, Jews appear as both slaves and slaveholders. They were the slaves of Jews and non-Jews and owned both Jewish and non-Jewish slaves.
In Jewish as in Greco-Roman society, slavery was one ofthe consequences of imperialist politics. According to Josephus, Jewish enslavement of non-Jews seems to have occurred in the Hasmonaean and Herodian periods in connection with various rulers’ attempts to expand their territories. The inhabitants of the conquered lands were partly enslaved. For example, after the capture of Sebaste, John Hyrcanus and his sons allegedly ‘reduced the inhabitants to slavery’ (Josephus, Jewish War 1.2.6—7, 63—5). This pattern is mentioned several times in connection with other places and regions. Alexander Jannaeus seems especially to have either killed or enslaved war captives in conquered territories (cf. BJ 1.4.2—3, 87—88). However, due to the nature of our sources, the exact number of people who were enslaved by the Hasmonaeans and later Herodians cannot be determined.
As far as the enslavement ofJews is concerned, Josephus states that Jews were already enslaved by Hellenistic rulers. For example, both Antiochus III and Antiochus IV are said to have captured and enslaved some of the inhabitants of Jerusalem (Josephus, Jewish Antiquities 12.3.3,144). The main periods of Jewish enslavement, however, were during and immediately after Pompey’s conquest of Jerusalem (63 bc), and during the first and second Jewish wars against Rome (ad 66—70, 132—135). Josephus (e. g. BJ 3.10.10, 539—542.) repeatedly mentions the murder of the old and weak and the enslavement of women and children. Men were either killed, if they were rebellious, or enslaved and sold or sent to work on large-scale projects. Josephus states of the first war (BJ 6.9.3, 420) that ‘the total number of prisoners taken throughout the entire war amounted to ninety-seven thousand’. Although the historical trustworthiness of this number is uncertain, we can nevertheless assume that tens of thousands of Jews were enslaved by the Romans during the revolt.
About the enslavement of Jews during the Bar Kochba revolt, few literary sources exist. From Jerome we know, however, that enslavement took place. He writes (Ad Jer. 31.15.6) that numerous Jewish captives of different ages were sold at slave markets in Hebron, Gaza and Egypt, and that the price of Jewish slaves was very low at that time due to their wide availability (Chronicon Paschale i: 474, ed. Dindorf).
Jerome’s remarks suggest that not all Jews enslaved by Romans during the first and second century were sent to Rome. A certain percentage was sold at local slave markets, where they could be bought by both Jews and non-Jews. In all likelihood, some of these slaves would eventually become the slaves of Jewish owners. Wealthy Jewish landowners who had remained loyal to the Romans or went over to the Roman side towards the end of the war, such as Josephus, were allowed to maintain their estates after ad 70. The Romans may have given land confiscated from rebels to their loyal supporters. These Jewish landholders would have considered the use of cheap Jewish slaves economically advantageous. And they may have considered the enslavement a just fate imposed on those who were rebellious.
Once they were enslaved, slaves could hardly maintain their original religious orientation. Slaves who worked in Roman households would necessarily come into contact with pagan worship. Gentile slaves bought by Jews would almost always be circumcised or immersed for purity reasons.613 Similarly, Christian slaveholders would baptise their pagan (and Jewish?) slaves. These slaves would be forced to live together with slaves of other ethnic and religious backgrounds and have sexual relations with them. Accordingly, ethnic and religious identity could hardly be preserved or at least not be exposed during enslavement. The slave familia was naturally of mixed origin and therefore denationalised, and the view of the slave as a blank slate was in the owner’s interest. The very small number of identifiably Jewish (freed) slaves in the few slave inscriptions from Rome and elsewhere is therefore understandable.
Circumcision and immersion did not make the gentile slaves of Jewish households Jewish. These rituals merely qualified them to work in Jewish houses. The circumcised and immersed slaves became avde Yisrael rather than converts (cf. y. Yeb. 8:1, 8d). Yet it seems that once manumitted, former male slaves could become members of the Jewish community. Freed women would forever be stigmatised as sexually promiscuous and rabbis would warn against marrying them (T. Hor. 2:11).
Forms of enslavement and manumission
In Jewish society slaves could be enslaved and manumitted in a number of ways, some of which were similar to and others different from Greco-Roman practices. The main form of enslavement, as in Greco-Roman society, was the enslavement of war captives seized during conquests. The number of Jews enslaved in this way will have been highest from the first century bc to the second century ad, that is, after the Roman conquest of Palestine and the first and second revolts. Such slaves would officially be imperial property, but at least some of them would be given to Roman officials and loyal clients as gifts.
The revolts against Rome will also have increased the occurrence of other forms of enslavement: self-sale, the sale of children, and debt-slavery. After the revolts large amounts of farm land were confiscated by the Romans, a phenomenon which left part of the peasant population impoverished. At least some of them would have had no other choice than to sell themselves or their children in order to survive. While Roman law prohibited debt-slavery, the New Testament and rabbinic literature suggest that it was widely practised in Roman Palestine. Although rabbis shared Roman jurists’ resentments against self-sale, they recognised the difficult situation of the rural population and allowed it under certain circumstances. According to T. Ar. 5:8, only the poor person is permitted to sell himself (cf. Lev. 25:39).
Similarly, a father’s right to sell his children (or at least his daughters) as slaves is already mentioned in the Hebrew Bible (cf. Ex. 21:7—11) and confirmed by the rabbis (M. Sot. 3:8). Some texts try to limit the right to sell a girl to the time before her puberty (M. Ket. 3:8; M. Qid. 1:2), probably to avoid selling daughters directly into prostitution. Therefore the Mekhilta also prohibits an adult Jewish woman from selling herself (Mekh. Mishpatim/Neziqin 3). In contrast to the unlimited power of the paterfamilias in Roman law, the Mekhilta explicitly states that a father may not sell his son. But whether such rabbinic advice was always followed by the impoverished Jewish families of the first centuries ad is doubtful. Yet for economic reasons, one may assume that families were more inclined to sell daughters than sons.
Other forms of enslavement seem to have been less prevalent but were nevertheless important in that they occurred empire-wide: the use of abandoned children as slaves and the capture and enslavement of travellers by bandits on the road. The existence of foundlings is already taken for granted by the Mishnah (M. Qid. 4:1—2; M. Makh. 2:7), and threptoi are also mentioned in a few Jewish inscriptions, the most famous being the Severus-inscription from the synagogue of Hammat Tiberias (Severus, ‘the threptos of the very illustrious patriarchs’). It must be assumed that abandoned children (threptoi) were usually raised as slaves, even if the status of Severus — (manumitted) slave or adopted son of the patriarchal family? — remains uncertain.
Although the sale of free people into slavery was prohibited, bandits who seized travellers and tried to sell them as slaves abroad, where their real identity was unknown, were a common phenomenon in antiquity known to Jewish writers. Alluding to the biblical Joseph story, the Testament of Joseph (12:2—3,13:6) discusses the possibility that Joseph had been stolen by the slave-trader who sold him to Pharaoh. The fact that rabbinic literature discusses the theft and sale of fellow human beings indicates that rabbis were also aware of this practice (e. g. T. B. Q. 8:1).
Besides conquest, debt, poverty and banditry, the slave population was also increased through slave-breeding and natural procreation. Masters were obviously interested in encouraging their female slaves to have children in order to increase the number of slaves born within their household (vernae), and these slaves were considered particularly reliable. The practice ofslave-breeding is reflected in a rabbinic story about a matron’s encounter with R. Yose b. Halafta (Lev. R. 8:1 par. PRK2:4). The matron rejects the rabbi’s suggestion that God spends a lot of time joining couples. In order to prove to him that she can do this work in seconds, she randomly couples her slaves, an action which results in chaos and broken bones. Although the story’s impact is theological rather than sociological, it indicates in a humorous way that intentional slave-breeding was not always easy and successful.
Slaves were sold at fairs, slave markets and individually, from one owner to the next. It is evident from several rabbinic traditions that Jews attended fairs in order to purchase Jewish and non-Jewish slaves (T. A. Z. i:8 and y. A. Z. i:i, 39b). The specific ways in which the ownership of slaves was transferred from one person to another were regulated by both Roman property law and rabbinic halakhah. Whereas earlier Jewish literature does not provide any guidelines about the sale of slaves, rabbis were as scrupulous as Roman jurists with regard to regulating the sales transaction. According to M. Qid. 1:2—3, slaves were acquired though the transfer of money and the writing of a document. It is not clear whether the monetary transaction was sufficient by itself or whether it had to be accompanied by a document. The first possibility seems more likely in view of the fact that gentile slaves could also be acquired by usucapion (M. Qid. 1:3), which is clearly presented as an alternative form of gaining ownership. According to M. B. B. 3:1, usucapion is effected by occupation, that is, employment or usage of the slave for three full years, whereas T. Qid. 1:5 provides examples of domestic services without specifying a time.
According to some biblical passages, Hebrew slaves were to be released by their masters in the seventh year of their service or in the Jubilee year (Ex. 21:2; Deut. 15:12; Lev. 25:40). The special treatment of Hebrew slaves is explained by reference to God’s liberation of Israelites from Egypt (Lev. 25:42; Deut. 15:13). Only if the Hebrew slave freely wants to stay with his master will he be allowed to do so (Ex. 21:6; Deut. 15:16—17). For Canaanite slaves, on the other hand, no such recommendations concerning manumission are given. In the book of Leviticus, non-Israelite slaves are considered the Israelite owner’s permanent property.
It is obvious that the biblical manumission laws concerning Hebrew slaves are diverse and contradictory. They must be understood as social ideals which offered differing solutions to the moral problems involved in Israelite ownership of fellow-Israelites as slaves. Whether these ideals were ever put into practice cannot be determined. On the two occasions where a general release of slaves is mentioned in the Hebrew Bible, the seventh and Jubilee year rules are not mentioned (Jer. 34:8—11; Neh. 5:1—13). In both texts the release is presented as a one-time event, not an institution to be repeated at fixed intervals. These texts therefore suggest that the seventh and Jubilee year rules were not commonly practised. This is also evident from Philo and Josephus: besides mentioning the biblical rules in exegetical contexts, they refer to actual releases of slaves on particular occasions by individual political leaders only (e. g. Josephus, BJ4.508).
In M. Qid. 1:2 the contradictory biblical rules concerning the release of slaves in the seventh and Jubilee years seem to be harmonised: those slaves who had not been released in the seventh year are to be manumitted in the Jubilee year. The Jubilee year manumission is applied to ‘permanent’ slaves (slaves who have their ears pierced) here as well. A third way of manumission, specified in M. Qid. 1:2, is financial redemption, which could take place at any time. Purchasing freedom was also customary in Roman society.
Several rabbinic texts suggest that the release of slaves in the seventh and Jubilee year could not be enforced by a court but depended on the individual master’s own conscience (e. g. Sifra Behar, pereq 2:4—5, 74a—b: ‘For it is possible to have a Jubilee year without the sending forth of slaves’). It seems that in rabbinic times the manumission of Jewish slaves in the seventh or Jubilee year continued to be seen as an ideal which rabbis knew was not always (or rarely?) practised.
If the seventh and Jubilee year rules were not generally observed, however, there would have been little difference between the manumission of (originally) Jewish and non-Jewish slaves. This conclusion stands in harmony with rabbinic literature’s major interest in the distinction between slaves and freeborn people rather than in the ethnic distinction between Israelite and Canaanite slaves put forth by the Hebrew Bible.614 Since the slave population of the Roman Empire must be considered a denationalised mass that was hardly able to maintain its religious and ethnic affiliations, it is understandable that differences between Hebrew and gentile slaves would have been negligible as far as manumission and other aspects of slavery were concerned.
Domestic and agricultural slavery
Most Jewish sources on slavery are rabbinic documents from Roman Palestine of the second to fifth centuries ad. Their chief concern is domestic slavery. This corresponds with the situation known in late antiquity from other Roman provinces.615
Slaves were part of the ancient Jewish household (bayit). Alongside wives and children, they were the householder’s dependants who stood under his authority and had no property on their own.616 This power situation was the basis of their role within the family and their relationship to other family members. Since the status ofwomen, slaves and minors within the household was so similar, they are often mentioned together in rabbinic sources, and often the same regulations concerning religious practice and observance of holidays are applied to them.6
It can be assumed that the presence of slaves within the household had a large impact on family relationships.7 Wives who knew that their husbands slept with their female slaves are likely to have become jealous or apathetic. Wealthy women whose female slaves took over their household work could dedicate their energies to other pursuits, a phenomenon which worried rabbis and caused them to warn such women against laziness and adultery (M. Ket. 5:5). If slave nurses and pedagogues were employed to care for the family’s children, mothers’ relationships with their children would have been more distant than if they had had more contact with their children themselves. Nurses and pedagogues are repeatedly mentioned in rabbinic king parables, where the ambiguity involved in their intimate relationship with the sons and heirs of the family is revealed: they can be dangerous as well as protective, distance the son from the father or reunite them (e. g. Gen. R. 28:6; Lev. R. 10:3; 11:7). Such texts reflect the ambiguities that characterised master-slave relationships: the slave was potentially dangerous but could, at the same time, become the saviour of the family.
It is well known that slaves could be sexually exploited by their masters, and the use of slaves as prostitutes within or outside the family was common in antiquity. The Hebrew Bible already mentions slave concubines. The so-called ‘Concubine Law’ of Exodus (21:7—11) suggests that Israelite daughters sold by their fathers could have sexual relationships with their owners without being their owners’ wives. The passage rules that the slave concubine will remain in her master’s household throughout her life. Ifhe does not like her any more, he may not simply sell her to someone else; she should rather be redeemed and freed.
In addition, there are many biblical stories about Israelite patriarchs’ relationships with their foreign-born slave women. The sons whom the biblical patriarchs are said to have had with their slave women could be accepted as proper children and heirs to the householder’s property. Hagar, Zilpah and Bilhah, the female slaves of Sarah, Leah and Rachel, could function as their mistresses’ substitutes as far as the production of children was concerned (Gen. 16: 29). The children born to slave women are considered the patriarchs’ proper children, regardless of their mothers’ foreign and servile status. In a tribal social structure with its emphasis on large families, the increase of one’s offspring was presumably considered most important, and slave women and their offspring became an integral part of the household. A later repercussion of this notion is found in the Elephantine papyri, where a free Israelite is said to have been married to an Egyptian slave woman with whom he had a child while she still remained the slave of someone else (Kraeling, Aramaic Papyri, no. 2).
This liberal attitude towards slave women seems to have changed in Hellenistic and especially Roman times, when we find polemics against sexual relationships with slave women in the Jewish sources. Such relationships were increasingly seen as illegitimate, and clear boundaries between the freeborn members of the family and the slave familia were established. The new position on marriages with slaves is explicitly stated by Josephus: ‘Let your young men, on reaching the age of wedlock, marry virgins, freeborn and of honest parents.. . Female slaves must not be taken in marriage by free men, however strongly some may be constrained thereto by love: such passion must be mastered by regard for decorum and the proprieties of rank’ (Ant. 4.244). Rabbis were also opposed to marriages with slaves. They explicitly stated that they did not consider marriages between slaves and freeborn Israelites valid, and they declared all offspring of slave mothers slaves. M. Qid. 3:14 rules that in legally valid marriages the offspring obtains the status of the father. In cases where a free Israelite had sexual intercourse with a slave woman, however, the union was not considered legally valid and the offspring followed the status of the mother, that is, the children are considered slaves. This ruling would prevent the householder from considering the children he had with his female slaves proper sons (and daughters) and heirs. Sexual relations between masters and their slaves were still tolerated (although not by all; cf. Lev. R. 9:5), but they did not have any repercussions as far as the family and inheritance structure was concerned.
The reason for this change may be the greater significance of the nuclear family in Roman society at large.617 The creation of large families was no longer the ideal. On the contrary, families were now interested in limiting the number of heirs to the householder’s property. Roman law similarly ruled that the child of a slave mother was a slave (cf. Gaius, Institutes 1.81). Rabbinic and Roman law arrived at amazingly similar conclusions on this issue. Besides inheritance issues, the maintenance of family purity may also have motivated rabbis to exclude illegitimate children: only children born to a Jewish mother were considered proper Jews. Thus the matrilineal principle and the ruling that the children ofslave women were slaves served one and the same goal: to preserve the purity of the nuclear Jewish family at a time when the Jewish family was thought to be threatened by a lack of political autonomy.
Although slaves seem to have fulfilled a great variety of functions in the ancient Jewish family and differed among themselves in status, rabbis — like Roman writers — tend to talk about them as a homogeneous and faceless collectivity. The attempt to distinguish a free male adult Israelite — or a free Roman citizen — from the slave as ‘Other’ is clearly recognisable in this regard. Only traces of the actual diversity among slaves have survived.
Education, skills and beauty distinguished one slave from the next and determined slave prices. Defects and proven misbehaviour would, on the other hand, lower the prices. Rabbis were as concerned about these issues as Roman jurists (T. B. B. 4:5). They also knew that some slave-owners had slaves to whom they were very close and whom they treated favourably. The stories about Rabban Gamliel and his slave Tabi portray an ideal master-slave relationship (e. g. M. Ber. 2:7; M. Suk. 2:1) in which Tabi represents the main value of sages, namely Torah learning. Although slaves, as slaves, could not officially become disciples of sages, they may sometimes have been able to overhear discussions and memorise teachings of their masters.
Besides education and training, another major difference among slaves was whether they were agricultural or domestic slaves. The living and working conditions of domestic slaves can be assumed generally to have been better than those of agricultural slaves, especially those who had to perform menial tasks such as planting, harvesting and ploughing.
Although domestic slavery seems to have prevailed in Jewish society in late antiquity, agricultural slavery will have continued to exist. But the number of Jewish sources that speak to this issue is limited. In Roman Palestine of the first centuries ad, as elsewhere in the Roman empire, slave labour will have coexisted with other forms of agricultural work. In the gospels of the New Testament and in rabbinic literature, slaves are mentioned alongside free labourers, tenants and smallholders. But we cannot determine the relative percentage of these types of labour in Roman Palestine any more.
Their respective significance can only be considered hypothetically, on the basis of what we know about the ancient Jewish economy and the economy of the Roman Empire at large. Slave labour was economical only on large estates where slaves could be occupied throughout the year. Since slaves were part of a property that needed to be preserved, day labourers were generally preferred for arduous and exhausting agricultural tasks, as Cato and Columella noted. Tenants would guarantee a certain degree of continuity, but they would transfer to the landowners part of the proceeds only. But the relative advantages of these forms of agricultural labour are never discussed in ancient Jewish literary sources.
In the Hebrew Bible, slaves appear as the servants of wealthy, cattlebreeding and land-owning masters for whom they fulfilled a variety of functions, including shepherding and farming. In the post-exilic, Hellenistic and Roman periods, land-owning upper-class Jewish families will have continued to employ slaves both on their estates and in their households, as evidenced by the book of Tobit, Judith, Philo and Josephus. But there is a recognisable shift to the predominance of domestic tasks and the role of the slave as an intermediary in business transactions.
Slaves could not own property but were sometimes put in charge of (part of) their master’s property for business reasons. In Roman society slaves were given apeculium, or working capital, which theoretically belonged to the master but was practically used by the slave - to increase his master’s property through the profits he was able to achieve (cf. Digest 41.2.49.1).
If the profits were high, and also to motivate the slave to achieve high profits, the master would reward the slave with gifts which, once accumulated, could enable him to purchase his freedom. If the master decided to manumit the slave, the slave could take part of his peculium with him into freedom (Codex lustinianus 4.14.2).
Cohen (1951) maintained that the Hebrew term segullah, which already appears in the Hebrew Bible, was equivalent to the Latin peculium. But this is unlikely. The term rarely appears in either tannaitic (first and second centuries) or amoraic (third to fifth centuries) sources. When it does, it refers to a fund set aside for minor sons, with money that usually comes from outsiders. Slaves are not mentioned in connection with it.
What is true, however, is that in both rabbinic and Roman law the status of the slave resembled that of the son with regard to his dependence on the householder and his inability to own property independently. In Roman society, children inpotestate could not have their own property. According to a ruling in the Tosefta, ‘the son who does business with what belongs to his father, and likewise the slave who does business with what belongs to his master, behold, they [the proceeds] belong to the father, behold, they [the proceeds] belong to the master’ (T. B. Q. 11:2). This text indicates that in rabbinic as in Roman law the income gained through the slave’s and the son’s work is the property of the paterfamilias under whose authority they stand.
Although Roman and rabbinic law allowed slaves to make use of their masters’ property for business purposes, they were not allowed to make gifts to a third party from the peculium or their master’s property (cf. Dig. 20.3.1.1; M. Shen. 4:4, 55a). This indicates that masters wanted to maintain control over the use of their property and curtail the monetary independence of slaves. Independence was granted only if they made use of the property for the master’s benefit.
In certain circumstances it seems to have been beneficial for slave-owners to use their slaves as intermediaries rather than conduct business transactions themselves, not only as far as time and effort are concerned. Both rabbinic and Roman law discuss the many and complex issues involved in business transactions with slaves as intermediaries between owners and third parties. According to a tannaitic ruling attributed to R. Meir, which appears in various contexts, ‘the hand of a slave is like the hand of his master’ (cf. y. Peah 4:6, i8b). On the one hand, slaves were given a certain amount of freedom in conducting business and entering contracts; on the other hand, they acted as extensions of their masters in all regards.
The most important incentive to use slaves in business was probably the master’s limited liability in case anything went wrong. Since slaves were not considered legal persons and had no legal rights, they could not be sued in cases of fraud. Since slaves were nevertheless considered responsible for the damages they incurred (M. Yad. 4:7), masters enjoyed a certain immunity. Although slaves were generally seen as chattel, they were attributed reason in as far as they were considered responsible for any damages they caused.
Can the evidence from the Hellenistic and Roman periods that locates slaves in domestic and business contexts more than in agricultural settings be considered representative for Jewish society? Or should this emphasis perhaps be attributed to the mostly urban environment in which the writers of the surviving sources lived? Some rabbis will have been landowners themselves, and at least in the first two centuries many lived in villages. The over-representation of domestic slavery in the sources may have had some basis in reality. In as much as both authors of Greek Jewish writings such as Philo and Josephus and some later rabbis participated in the lifestyle of the Roman urban elite, they would represent themselves as owners of slaves who performed various household tasks. Although landholders may have had a few slaves to assist them in maintaining their rural landholdings, alongside day labourers and tenants, slaves will have been more prevalent in wealthy urban households that did not depend on them for their economic survival. The ownership of slaves endowed their masters with prestige and was probably considered a status symbol in Jewish as much as in Roman society.
As members of the intellectual rather than the political and economic elite, rabbis would have had fewer slaves than wealthy urban grandees, but they may have followed the model of Libanius, Augustine and other intellectuals in considering the possession of a few slaves desirable, especially from the third century onwards, when they seem to have resided in cities more often. Very few rabbis are presented as slaveholders in tannaitic texts of the first two centuries. But the number of rabbis associated with slaves increases in amoraic texts from the third and fourth centuries.618 This phenomenon may partly be due to the fact that the Talmud is much more voluminous than the Mishnah and Tosefta and gives more space to narrative traditions. In addition, we may assume that the rabbinic movement had increased over the centuries, so that more traditions about rabbis circulated. Nevertheless, the development is striking and may be related to the greater urbanisation of rabbis from the third century onwards.
Slavery metaphors
Since slavery played such a large role in almost all aspects of ancient Jewish everyday life, it was also used theologically to express human beings’ relationship with God, politically to express Jewish subjugation to Rome, and psychologically to express personal ‘enslavement’ to bad habits and emotions.
Theological usage of the slave metaphor appears already in the Hebrew Bible. Patriarchs, kings, prophets, but also less important figures and ordinary Israelites are presented as ‘slaves’ of God. The terminology appears, for example, in prayers to express humility before God. Especially in Deutero-Isaiah, Israel as a collectivity is called ‘slave of God’ (Isa. 41—2). The slave metaphor is used to denote the close and intimate relationship between individual Israelites or the entire people of Israel and God. It also indicates the status difference between the two partners, the difference between the Divine sphere and humanity. Given the Hebrew Bible’s emphasis on obedience to Divine laws and regulations revealed at Sinai (cf. Exod. 19—23), the slave metaphor stresses the need for Israelites’ compliance to God and God’s option to punish misdeeds.
The theological usage of the slavery metaphor is expanded in the many slave parables found in rabbinic texts, especially in Midrashic contexts.619 Here a king generally stands for God, whereas human beings are represented by slaves, pedagogues and sons. The parables concentrate attention on aspects of the relationship between God and human beings, such as obedience and disobedience, praise and punishment, attraction and escape. Sometimes the king’s son and slave are compared with regard to their relation to the father and master (e. g. Lev. R. 1:15). Interestingly, both the son and the slave metaphors are used side by side and interchangeably to denote two different aspects of God’s relationship to human beings: that of the caring father and that of the strict and just master.
In political discourse, slavery metaphors were commonly used in antiquity to denote subjugation to foreign rulers. Subjected people would lack the independence and freedom they enjoyed at the time of political autonomy. In ancient Jewish literature this usage of the slave metaphor appears most frequently in Josephus’ writings. The term servitude (douleia) is continuously used in the Jewish War and the Jewish Antiquities to express Jewish subjugation to Roman rule, both of the rebels and of those who opposed rebellion. For example, it appears in the speech of Agrippa, who allegedly tried to dissuade his fellow Jews from fighting against the Romans. Unlike the rebels, he is said to have advocated submission to the Romans, since he considered them too powerful to defeat: ‘For servitude is a painful experience and a struggle to avoid it once and for all is just; but the man who having once accepted the yoke then tries to cast it off is a contumacious slave, not a lover of liberty’ (BJ 2.355—6).
The argument that Jews had a long experience in enduring slavery qua subjugation seems to have been advanced by Roman politicians to justify their policies. According to Titus, as reported by Josephus (BJ 6.42), Jews who have lived under foreign dominion for so many centuries should be experienced in slavery and be aware of the uselessness of revolt. Josephus is aware of the anti-Jewish argument implied in Titus’ speech, that the lack of political independence would render Jews a ‘nation of slaves’, as explicitly propagated by Apion: A clear proof, according to him, that our laws are unjust and our religious ceremonies erroneous is that we are not masters of an empire, but rather the slaves, first of one nation, then of another, and that calamity has more than once befallen our city’ (Contra Apionem 2.125). Josephus responds to such claims that due to political misfortune almost all people have at times been the victims of political subjugation, so that Jews are not exceptional in this regard (C. Ap. 2.126—7).
The psychological usage of the slavery metaphor is most frequently used by Philo of Alexandria and has analogies in Stoic and other Hellenistic philosophical writing. Among others, Seneca, Plutarch, Epicurus and Zeno wrote philosophical works on restraining personal passions. It was emphasised, for example, that the Stoic wise man had to be in control of his senses and emotions and showataraxia (psychic tranquillity). He could neither be seduced by food, wine, or women and other vices, nor by anger, fear, jealousy or other negative emotions that held the mind in their grip. He was rather free of all such constraints on his mind and body and could devote himself to philosophy. Similarly Philo argues (De cherubim 107; Legum allegoriae 3.198—99, 221, 240) that the purified soul, which has only God as its master, must be free from enslavement to passions and emotions; those who are inclined to give in to their passions will always remain slaves and never achieve spiritual freedom. This argument is further developed in the tractate Every Good Man Is Free (Quod omnis probus liber sit), where Philo distinguishes between slavery of the body and slavery of the soul (17). True liberty can be achieved through liberation of the spirit only. Therefore a person can be physically free and at the same time possess an enslaved soul, or be physically enslaved and spiritually free: ‘Those in whom anger or desire or any other passion, or again any insidious vice holds sway, are entirely enslaved, while all whose life is regulated by law are free’ (45).
The notion of spiritual slavery and freedom was later also adopted by early Christians and elaborated in the theory of original sin.620 It probably helped to justify the status quo of real slavery, while at the same time claiming that spiritual freedom was what mattered most. The Stoic focus on spiritual freedom and slavery may also, at least ideally, have improved relations between slaves and free people, since it transposed the boundaries between them into the moral-spiritual sphere. A master would have been more likely to behave in a friendly manner towards a slave who shared his moral and religious concerns, if those concerns were most important to him. The transfer of the slave-free distinction to the moral-spiritual sphere probably takes account of the fragility of freedom in the ancient world and the constant danger of enslavement through political subjugation, banditry and poverty.
In ancient Judaism slavery was closely associated with the Exodus experience, which was commemorated during Passover. After ad 70 the sacrificial rite performed on Passover was transformed into a ritual family meal in the course of which each participant was supposed to identify with the Exodus generation and its liberation from Egyptian slavery (M. Pes. ch. io). The Hebrew Bible already stresses that slaves, in contrast to hired labourers, may eat from the Passover sacrifice as soon as they are circumcised (Exod. 12:44). Slaves were considered part of the household and could therefore participate in the ritual. Arguably the Torah states that slaves may eat from it, but that they are not obliged to do so.
The biblical inclusion of slaves reappears in the Mishnah and Tosefta, where aspects not mentioned in the Torah (half-slaves, female slaves) are also addressed. M. Pes. 8:7 rules that women, slaves and minors should not eat from the sacrifice in separate congregations. They should rather partake of the meal together with the male Israelite members of the household or community (cf. T. Pes. 8:6). As a banquet in which women, slaves and minors participated, the Passover meal would differ from ordinary banquets held in Greek society, which would usually be all-male affairs.
A possible analogy to the Passover meal was a meal at the Saturnalia, an annual festival in honour of Saturn, where according to tradition Roman slaves were served at table by their masters. The reversal of roles was what mattered. Passover lacked the element of role reversal: masters did not assume the role of slaves because it would have stood in opposition to the liberational aspect of the holiday. Nevertheless, Passover and the Saturnalia shared the extraordinary aspect of slaves’ dining at their masters’ table, an exception that perhaps made normal status differences all too clear. The shared meal of the Passover seder, even if more ideal than real, would celebrate liberation and autonomy and point to human beings’ essential equality before God. In this sense it seems to have been unique in antiquity.
Jewish slavery in the greco-roman context
How can the relationship between the Jewish and Greco-Roman discourse on slavery be assessed? Martin (1993: 113) has suggested that Jewish slavery was entirely identical with Greco-Roman practice: ‘Jewishness itself had little if any relevance for the structure of slavery among Jews.. .. Slavery among Jews of the Greco-Roman period did not differ from the slave structures of those people among whom Jews lived.’ Martin reached this conclusion from epigraphic and papyrological material only, however, without having analysed the literary sources. Although the socio-economic basis ofslavery may have been similar in Roman Palestine, Roman Italy and other Roman provinces, it is likely that the discourse on slavery and the actual treatment of slaves would have differed in each society due to its particular political situation, ancestral traditions and moral values.
One major difference between Roman and Jewish society of the first centuries ad was that Roman society was an imperialist power, whereas Jews were subdued by the Romans and victims of Roman colonialism. We already noted that the discourse on slavery was closely interlinked with colonialist discourse in antiquity: the conquest by foreign rulers was considered ‘enslavement’, and the actual enslavement of at least apart of the native population was a common aspect of military defeat. Since Jews were conquered rather than conquerors for most of the period under discussion, did their attitude towards slavery differ from that of the imperialists? Were they more prone to sympathise with slaves because they were sometimes identified as such by Greeks and Romans?
Another issue to consider is the biblical tradition of the Exodus experience. The liberation of the Israelites from Egyptian slavery is to be relived by each generation during the annual Passover celebration. Would the Exodus experience have caused Jews to be more friendly towards their slaves? Could the ritual remembrance of the enslavement of ancestors have changed contemporary Jewish views and practices of slavery?
The sources clearly indicate that Jewish writers, like their Greco-Roman counterparts, took slavery for granted as an integral part of everyday life in the Land of Israel and the Diaspora. Neither Philo and Josephus nor the rabbis were opposed to slavery. They never questioned the institution of slavery as such, nor did they consider the possibility of its abolition. They favoured returning runaway slaves to their owners like any other kind of lost property. Whether the slaves were of Jewish or non-Jewish origin does not seem to have made much difference. What mattered most was free or servile status, and free male adult Israelites distinguished themselves from slaves just as male Roman citizens did. The slave was the quintessential ‘Other’ in ancient Jewish as in Greco-Roman society.
Such similarities were probably due to the basic socio-economic structures of wealth and poverty, authority and dependence in which slavery developed, as outlined above. Ancient Jews lived in a context in which certain power structures were taken for granted or were maintained ‘for the good order of the world’. We can assume that the wealthy slave-owning strata of Jewish society were quite assimilated to the Greco-Roman lifestyle. Even if rabbis did not generally belong to the upper strata of society, as free Jewish men they seem to have shared the rhetoric of free Roman citizens, who distinguished themselves from women, slaves and minors.
Yet despite the similarities we can also recognise certain nuances which give this discourse its particularly Jewish flavour and seem to be partly based on biblical regulations. Ancient Jewish intellectuals such as Philo and the rabbis sometimes advocated a mild treatment of slaves and tried to protect slaves from a too cruel treatment. They opposed the ‘natural slave’ theory which suggested that certain people were born to be slaves.621 While advocating a master’s punishment of his disobedient slaves, Palestinian rabbis did not share Roman jurists’ notion of the master’s power of life and death over his slaves (T. B. Q. 9:10, 21). Since rabbis were interested in guarding the purity of the Jewish family, they criticised men’s sexual relationships with their slave women (M. Ket. 2:2—4). Although rabbis focus on the distinction between slaves and free persons, some traces of the biblical distinctions between Hebrew and Canaanite slaves survive in rabbinic literature and reappear in exegetical and halakhic contexts. For example, M. B. M. 1:5 associates the objects found by Canaanite slaves with the finds of minor children and wives, whereas the finds of adult children and divorced wives are compared with those of Hebrew slaves. But the traces of these distinctions are sporadic, disconnected and not further elaborated.
Although the biblical manumission laws advocated the manumission of Hebrew slaves in the seventh or Jubilee year, these rules seem to have been more ideal than real already in biblical times. We do not know whether ancient Jewish slave-owners actually practised manumission more than their Greek and Roman counterparts. In all likelihood manumission was practised whenever it was economically advantageous in both Jewish and Greco-Roman society. The status of the freedman was not as well defined in Jewish as in Roman society. The freedmen of Jewish owners did not form a special order within society, but rabbinic taxonomy assigned them a special status between slaves and free people. No fixed obligations that the freedman would have to render to his master are specified, but Jewish masters seem to have expected their slaves to remain loyal to them after liberation. As in Roman society, a certain stigma was attached to former slaves. Rabbis tried to limit their choice of marriage partners and to assign them the very lowest status in the Jewish community. To what extent they succeeded remains unknown.
Bibliographic essay
Books and articles on Jewish slavery are sparse. Besides the works which were written approximately a hundred years ago (e. g. Farbstein 1896) and therefore do not meet the standards of contemporary critical methodology, two relevant articles were written in the 1960s (Zeitlin 1962—3; Urbach 1964). Hezser (2005) is the first comprehensive and methodologically sophisticated study of Jewish slavery in antiquity. Goldenberg (2003) treats Jewish slavery in the broader context of racism and Christian and Islamic attitudes and practices.
World History
