Printer Friendly

Ancillary evidence for the decline of medieval slavery.

The demise of slavery is generally understood to be one of the most distinctive features marking off the medieval West from the ancient world; and this remains true even though as early as the 1850s Wilhelm Heyd noted the persistence of a lingering east-to-west traffic in Mediterranean slaves in medieval times.(1) Charles Verlinden's archival researches have now established that fact beyond doubt.(2) That may not be significant, for after all the demise argument is concerned not with trade, but with medieval agricultural work; in Marc Bloch's words, "[f]rom the 11th century onwards, it was freely admitted that whereas in former times the tenant of servile birth had been under no law but the arbitrary caprice of his master, his obligations henceforward, like those of 'free' dependants, were regulated by the custom of the group".(3) Regional studies have challenged the pace of this transformation and the means by which it was achieved, but there has been no dispute that in changing his status the servus, whose very name lost precise meaning, altered the terms of agricultural production and with that the economy and society of the Middle Ages.

Shifting the lens from the male slave to the female, or ancilla, who traditionally provided the eponymous ancillary aid to great households, changes the picture. Both Heyd and Verlinden also recorded the disquieting fact that, in however diminished a form it survived, the medieval traffic in slaves was overwhelmingly a traffic in women. Historians might then ask: while great households survived in the countryside and later thrived in towns, can the altered status of servi alone account for slavery's demise? Or does slavery persist, albeit in diminished numbers, because of ancillae? Bloch, for whom the demise argument was of great importance, found a "profound essence" of slavery that was not easily erased.(4) He argued that medieval people had long memories about slave origins, and believed that the servile condition was passed on through inheritance. In the words of the late thirteenth-century English Mirror of Justices, there existed "a subjection issuing from so high an antiquity that no free stock can be found within human memory".(5) Shifting focus to ancillae or women slaves may help explain how such thinking persisted even after slavery itself ceased to be an essential institution for agricultural production. After briefly considering the surviving philological evidence about ancillae, in order to make comparisons with the largely philological arguments about the changing status of servi that lie at the heart of the demise argument, this study will move on to a discussion of the role of ancillae in both countryside and town.

One or two caveats ought to be sounded at the outset. First, it may well be that demise arguments, wherever applied, are perverse even at their best: like the rising middle class, which took centuries to make its ascent, the purported demise of slavery in Europe took too many centuries to reach completion. This is certainly the tenor of recent criticism directed at Guy Bois and Pierre Bonnassie, who see a swift change away from slavery in Europe in the eleventh century.(6) If Marc Bloch identified slavery's demise in feudal France in the eleventh century, Ruth Mazo Karras has found that it did not disappear in Sweden until the beginning of the fourteenth.(7) Rodney Hilton and Leo Verriest note vestiges of slavery in the heartland of Europe, while the class-conflict arguments advanced by Robert Brenner and Pierre Dockes stress that for centuries slaves rebelled to gain their freedom.(8) By the thirteenth century slavery had revived in a brisk Mediterranean trade with the Levant. The fifteenth century saw the introduction of slaves from Africa into Europe, and this was almost immediately followed by the overseas expansion of the sixteenth century, itself characterized by a massive sea-borne migration of slaves to New World plantations and mines. In the interim, Christian Europe tolerated slavery. Traders sold pagans, infidels and schismatics, and set up a lively traffic in souls (animae),(9) that is to say, baptized Roman Christians.

A second complication arises when we examine the processes which historians believe contributed to the demise of slavery: informal processes like the choice of marriage partners, formal processes like manumissions of individuals or groups, and lastly, revolt or rebellion. All three yield alternative readings if evidence of gender is taken into consideration. In ninth-century villages on the estates of Saint-Germain-des-Pres, a male slave's marriage to a lida or colona gave his offspring half-free or colonus status.(10) Not in terms of production, but of reproduction within recognized households, ancillae - that is, female slaves who were equal in status to servi - were shunned as marriage partners. They might work, they might well produce some offspring, but they were unlikely to marry and be accepted among householders. This meant that some persons with chattel slave status continued to live generation after generation in the countryside, occupying positions of the lowest status.

Generally speaking, in formal processes of manumission, freedom was granted to a male slave as an individual, or as the head of a household. In 1937 Pierre Petot spoke of the serf fortunate enough to gain the status of colon after the eleventh century, mentioning only once in passing that a wife and children might also profit from that gain.(11) Michael Postan's eponymous famulus is also presented as male.(12) Indeed, a good half-century of research presents the demise argument in terms of men gaining rights or freedom in the rural countryside.

The prospect of retracing the paths of scholars through regional archives all over Europe to see if their conclusions accurately reflected the surviving evidence on the issue of gender is daunting indeed. On certain occasions, it is true, both women and men are mentioned in manumissions already published. Petot himself quoted the abbess and sisters of Saint-Memmie, near Chalons, granting cessation of servile dues to both homines and mulieres.(13) More recently, William Jordan has cited thirteenth-century grants of manumission from Senon which explicitly mention both men and women: "Likewise the aforesaid manumitted men and women and their heirs as long as they dwell and reside in our lands . . .".(14) When in 1050 the Emperor Henry III struck a coin from the hand of Sigena, the "female serf" of a nobleman named Richolf, a woman escaped from chattel servitude. The charter specified that her status was now that of serf (serva), so that she no longer passed chattel status to her children, but "assumed the same liberty and legal status as all the other female serfs who have been freed in the same way by kings and emperors".(15) The question remains, how many women had access to the emperor and his grant of manumission? Equal numbers to men? Regional archival studies suggest that was not the case. Either some scholars have mistakenly cast their arguments into the masculine mode, or in truth the grants they have found were made overwhelmingly in men's favour.(16)

So if Sigena's grant of freedom was an exceptional instance then men were overwhelmingly the manumitted persons of the Middle Ages. If so, manumissions would join the two other routes to individual freedom - running away and revolt - as measures more readily adopted by unfree men than by unfree women. While freeing a female slave on equal terms with a male slave had the capacity to bring about the demise of the institution itself, were it repeated until it became policy, a policy of freeing male slaves alone while women slaves remained unfree produced no such effect, suggesting one way in which slavery might have survived.(17)

Three gender-related issues offer insight into the persistence of slavery for women. First, the Latin term for a female chattel slave, ancilla, showed remarkable stability of meaning, whereas the corresponding term for a male chattel slave, servus, underwent permutations over the medieval centuries. Secondly, medieval custom never jettisoned the Roman notion that women passed on their servile condition to the heirs of their body. Both ancillae and their offspring could be bought and sold, and they required manumission to revoke their status of legal servitude. Thirdly, in the central Middle Ages, when slavery no longer suited the conditions of agricultural production, it may still have accommodated the labour needs of great households. Because wealthy rural and urban households, whether of ecclesiastical, noble, royal, or commercial composition, all played their own roles in the evolution of the medieval economy, slavery for women might well have continued to hold value in the eyes of those wealthy enough to possess slaves.

The term ancilla possessed a powerful resonance, and clung to the newly enslaved and to those who inherited their status. In medieval practice, it retained much of the precise meaning of "female chattel slave" assigned to it in Roman law - namely, lifelong legal servitude that passed to the heirs of a slave's body.(18) Sigena's manumission in 1050 from unfree to serva marked her transition to a more privileged state. Rodney Hilton found both "ancille" and mulieres serviles in the earliest of the three twelfth-century Caen surveys.(19) Serva may have denoted a looser or higher category of status; in other words, the word may have evolved in meaning in much the same way as did servus.(20) The ancilla, however, continued to have lowest status.

Debate over the meaning of the term servus and the vernacular "serf" continues unabated because, time and again, their meaning has proved unstable. Derivatives of the Latin term, such as servientes, generally meant merely "servant(s)".(21) The term servi, on the other hand, may denote different states or conditions even within the same document. Variants such as the Celtic gwas introduce other challenges to philology, and when analysed reveal the vast complication of medieval social evolution. Associated terms and synonyms that were generally expressed in the masculine - lidus, "villein", "boor", cliens, mancipius, famulus - ran the social gamut from chattel sieve to preferential status in a culture that relied upon ties of dependence for protection and social cohesion. Scholars often rely on evidence about residence to sort out rural status: for example, whether the famulus was "hutted" and thus afforded a degree of autonomy in the rural countryside.(22)

The famulus of the Mediterranean region, however, was generally not domiciled separately but lived and served in the household of his master.(23) Mediterranean households appear to have been staffed by persons answering to a bewildering number of terms: homines (in such expressions as Bogdan, homo Mergnani), servientes, (those who) dedi me ad serviendum or ad standum (roughly, "wait on me"), as well as famuli and famulae. Pueri and puellae were also used in the traditional sense of persons in dependence rather than as a designation of age.(24) These terms obscure rather than reveal legal status, perhaps intentionally so, since they hint at the deep dependence of the served upon the servant, who was expected to be constantly at beck and call - a totally familiar presence. Almost any understanding of the change from unfree to free is problematical, because of the tendency of old terms to remain in use even after they had lost old meanings and gained new ones.

If attention refocuses on unfree women, famula or comparable terms replaced ancilla in day-to-day usage, but they failed, as did the term ancilla itself, to suggest a new and freer status.(25) From the barbarian law-codes and the eleventh-century glossary of AElfric of Eynsham to the slave sale contracts of late medieval Mediterranean trade, ancilla retained the same meaning.(26) The twelfth-century revival of Roman law strengthened this tendency. In canon law, unfree serf fathers gained some rights over their marrying offspring by the thirteenth century and thus modified their status; however, ancillae, whether of unrecognized paternity or newly enslaved, remained chattel slaves in canon law, especially those deemed res ecclesiae, that is, the property of the church.(27)

Since neither medieval law nor custom treated slavery as the same institution for women as for men, the laws governing the sexual unions of slaves and free persons indicate that being a slave had different consequences for women than for men. A union of a free woman with a male slave was treated as analogous to bestiality from the era of Roman law to that of the barbarian law-codes; not until the ninth century, when servile status was complicated by the creation of half-free conditions, did non-slave women marry servi, and then only in rural areas where custom condoned the practice.(28) But while such unions often incurred penalties, neither severe punishment nor opprobrium were visited on the sexual union of a free man and a slave woman. In such cases no ambiguity about the status of possible offspring existed. Medieval law tended to punish a free man only when he copulated with a female slave who did not belong to him, and even then only in cases of mistreatment or if the union occurred in the house of the master of the ancilla - for the woman was seen as another owner's possession.

So when Marino Ranina, a noble citizen of Ragusa, brought charges against a fellow nobleman for assaulting his two famulae, who were working in his vineyard outside the town walls, he was protecting his property interests. The defendant, Marino de Getaldi, needed only to restrict his attentions to his own slaves to avoid prosecution.(29) Michel Balard noted that a free man in medieval Genoa who fathered the child of another man's slave owed the woman's master twenty-five lire - fifty, if she died in childbirth.(30) Having interfered with another man's property, most likely diminishing the slave's market value, such a man faced a stiff fine.

The dangers of illicit association between clerics and ancillae form a major sub-topic in many of the eighty-one canon law texts (pre- and post-Gratian) to survive from the central Middle Ages.(31) Long after the church claimed victory against clerical marriage, the cohabitation of priests and serving-women (carefully termed ancillae to distinguish them from free women) remained common enough to be the subject of legislation in provincial synodal councils and church decrees. Causa 15 of the second part of the Decretum provides a case in point. Any ecclesiastic, from bishop to subdeacon, who produced children with an ancilla met with the church's censure. More importantly, the offspring, being of "polluted" birth, were not counted as heirs; indeed, they were themselves the property of the church.(32) This clear distinction separated the cleric, who administered the church's property by right, from the offspring he had fathered, who became the church's chattel, creating an almost unbridgeable social distance between the generations. This concerted effort to establish beyond a doubt the offspring's unfree status would probably have in turn required legal action to gain freedom for that offspring. As John Gilchrist has noted, churchmen hesitated before alienating the church's goods, inanimate or human.(33) Once identified as such, slaves who were by birth church property confronted a legal apparatus that enforced their unfree condition.

Throughout the medieval era the notion persisted that an ancilla's offspring were by inheritance unfree, so that the union of a free man with a slave became one that required a legal initiative to undo the effects of the servitude - that is, if the free partner desired such a change.(34) The authority of the law on the Continent, whether secular or ecclesiastic, tended to enforce this construction of the inheritance principle, so that the statutes of the church, of royal realms and of city-states often followed Roman precedent and enforced the principle of maternal inheritance of unfree status.(35)

As a result there existed a wide range of legal initiatives that a free man, who was both master and father of a slave's child, might adopt in the day-to-day practice of slavery. A woman's slave status need not stand in the way of her master freeing her, marrying her and legitimating her offspring (unless he was a priest). Short of that her child might be acknowledged, and any number of provisions in wills attest to the widespread acceptance of slave offspring into the household. It was a pious act to free slaves and make them bequests in last wills and testaments. Pasque de Goce had an illegitimate child from a union with a household slave, and the father's noble parent - rather than the father himself - left that child a testamentary bequest.(36) In great households which assumed responsibility for the offspring of their inmates, free or unfree, such a solution ameliorated the bleak terms of unfree birth.

Such alternatives created freedom of action for owners of female slaves, and, as a result, masters held vastly greater power over female slaves and their offspring than a man might wield over his offspring in an out-of-wedlock union with a free woman. Options ranged from turning the ancilla out, punishing her, selling her or her offspring, acknowledging the issue, freeing mother and child immediately or at a later time, freeing them and finding another man to marry the former slave, settling mother and children in the household or far away, to marrying the former slave and thus legitimating the offspring. In sum, an extraordinary range of choices existed, all of which may be found somewhere or other in the "documents of practice" (as in the example of Orenetta discussed below).

Intentionally or otherwise, great medieval households employing domestic slaves produced a renewable substratum of persons labouring at the will of their masters. Such persons faced a wide diversity of future prospects. In a world of rights and obligations shaded hierarchically from the most to the least privileged, domiciled slaves were useful precisely because they stood below all others; being virtually rightless, their owners could use them in almost any way they pleased, a fact well illustrated by references that are found to the ancilla's child-bearing capacities.(37) Furthermore, this condition was inherited and of long standing, so that householders did not need to impose new forms of servitude, as was found necessary in the late medieval countryside in such regions as Catalonia. New types of servitude tended to be difficult to enforce, often required co-operation with higher authority and demanded careful attention and surveillance.(38) By contrast, affluent householders equipped with traditional female slave labour had at their disposal a time-sanctioned supply of unfree workers who could be deployed at will. This ancient system of labour may have survived because it worked well.

Nevertheless, the question of whether female slaves survived in the countryside rests ultimately on their utility in work performed there. Decades of painstaking scholarship indicate that Europe never fell into such "primitive" conditions that undifferentiated agricultural labour provided mere subsistence in rural areas. Complex work and manufacturing processes continued to be important features of economic life at monastic centres and on other large rural estates. These made demands for skilled as well as unskilled labour. For example, on great estates gynaecea, with multiple-skilled textile operations organized under one roof and staffed by female slaves, survived and prospered.(39) Monasteries and cathedral chapters, like royal and noble households, consumed cloth made on the premises. The gynaecea of Charlemagne's realm were renowned, and the emperor's gift of woollen cloaks for Harun-al-Rashid was produced at his own workshop in Aachen.(40) In the tenth century St Liutbirg began life as an unfree member of such a workshop; so did a number of young women whom she instructed in the art of fine textiles. Eventually those skills gained Liutbirg and her students freedom, evidence offered in her Vita as proof of her sanctity.(41) If the romance Yvain may be trusted to reflect twelfth-century expectations, young women taken as hostages were also thought appropriate staff for the gynaeceum.(42) Apparently, women worked together at intricate textile production in workshops under one roof well into the central Middle Ages.

If ancillae persisted into this era it was largely as a response to the demand for textiles. In cloth-making, repeated labour-intensive tasks predominate. Slaves, unmarried, without their own households protected by "the custom of the group", could be ordered to live and work together under one roof and to carry the burden of repetitive tasks that produced some economy of scale and possibly some quality control. David Herlihy has argued that in time labour shortage destroyed the workshop, leading to the domestication of cloth production in Europe (when, presumably, free women and men of the household took over the entire tedious production but lost whatever economy of scale the workshops had provided). This change, he contended, occurred only with the re-urbanization of Europe. It involved centuries-long processes that began in the twelfth and thirteenth centuries and were completed by the fifteenth. By then the revival of trade had begun to reintroduce a supply of slaves into some towns and cities of Europe, where they were marketed to great households.(43)

Perhaps ancillae persisted so long in the economy because they lacked the means to change their condition that male slaves exercised when they ran away to another manor, or to a town willing to turn a blind eye to their escape. It cannot be overemphasized that workshops could supervise personnel in much more efficient fashion than manors. Women with inherited servile status were then reliant upon formal measures of manumission far more than were men of their station. In an age with a high ratio of unskilled to skilled labour, great households needed many women to perform menial tasks. Domiciled slaves of the household, like the domiciled slaves of the workshop, afforded an opportunity for surveillance impossible where hutted workers were concerned. Workshops and great households were effective isolating institutions, where either fleeing, with all the associated risks, or remaining with the hope of winning future manumission from one's master were the only avenues open to a female slave.

In addition, the frontiers over which slaves were transported to augment the supply of those of inherited servile status never completely closed. On Europe's frontiers - in Scandinavia, the Balkans and Castile - slave women were a reasonably easily obtained resource.(44) Heath Dillard noted in Daughters of the Reconquest that slave-traders worked on both sides of the Castilian frontier, supplying Christian ancillae to Muslim households, and Muslim ancillae to Christians. Fueros or statutes were passed aimed at curtailing such traffic - at least when it preyed upon the Christian community - but slavery of Muslims in Christian homes was a tolerated institution; presumably enslaved Christian women were similarly tolerated in affluent Muslim households.(45) More problematically, in Scandinavia, as in England and France, the domiciled slave remained a feature of rural life even after the disappearance of the unfree agricultural worker hutted separately.(46)

As for the Balkans and the Adriatic region, there appears never to have been a time when the slave-trade died down or slavery fell into disuse, so a full apparatus was in place when the demand for slaves intensified in the thirteenth century. The west Balkans was a region where towns had survived from ancient times, maintaining some continuity of practice with late antiquity. The traders at the mouth of the Narenta (the present-day Neretva) and in Dalmatian cities continued to buy and sell at least some slaves from the countryside in coastal towns. In 985 Prince Crnomir of Bosnia had deplored the conditions that prevailed in nearby rural areas of Bosnia, where people had no defence against foreign traders capturing and enslaving them.(47) But slavery persisted, even at the court of the ban of Bosnia itself, well into the thirteenth century, while traders from the coast bought or captured slaves from Bosnian villages.(48)

The law-code of the fourteenth-century Serbian king, Dusan, prohibited the sale of Christians to unbelievers or heretics on pain of the vendor having his hands cut off and his tongue cut out. The true faith was Orthodoxy, and the slave-vendor envisaged in the code was a Roman Christian from the coast.(49)

Nevertheless slavery was practised in Old Serbia (Rascia). In Catar (now Kotor), the kingdom's urban window on the Adriatic, Dobroslava and two small slaves, almost certainly her infant sons, fetched thirty hyperperi (about fifteen ducats) in 1326, while the slave Calia gained freedom through manumission there in 1332.(50) Both rural Serbs and Greeks were enslaved and reached Venice in the later Middle Ages.(51) Domestic slaves never truly disappeared; they remained close at hand, for localities practising slavery were never very distant from those lands where slavery survived on only a diminished scale.

Furthermore the slave of the household or the workshop was a single woman, not a wife in a servile family. Single and unfree women appear to have fallen into a different category from married and hutted labourers, whether female or male. They continued to be referred to as ancillae, they did not marry (although they may very well have borne children),(52) and they were domiciled workers set at labour-intensive and tedious chores on their masters' orders. In this regard, great households, whether rural or urban, did not differ substantially from each other in their demand for skilled and unskilled labour; both sought and used female slaves and their offspring.

With the growth of towns the number and wealth of great households increased, reviving the demand for slaves at least in the Mediterranean south, where there was some hope of reviving supply through Levantine trade. By the late thirteenth century increased demand had led to active efforts to import slaves into the cities. Imported female slaves served in affluent commercial households from Dalmatia to Barcelona, and often beyond. A market niche existed for imported slaves destined to work alternately in domestic tasks, commerce and production. Francesco Datini instructed his partner Andrea de Bonnano, resident of Genoa:

Pray buy me a little slave-girl, young and rustic, between eight and ten years old, and she must be of good stock and strong enough to bear much hard work, and of good health and temper, so that I may bring her up in my own way. I would have her only to wash the dishes and carry wood and bread to the oven, and work of that sort . . . for I have another here who is a good slave, and can cook and sew well.(53)

With typical clarity Datini defined the niche occupied by a newly enslaved child in a complex commercial household. Iris Origo has noted that a Florentine register of 1363 mentions 34 girls under twelve and 85 under eighteen out of a total of 357 slaves.(54) Study of the Adriatic trade in slaves reveals a similar market preference for mere children;(55) Verlinden's archival researches stress the extreme youth of slaves imported from the Crimea, as well as from the Balkans, with Moorish or African slaves later fitting the same narrowly conceived market niche in the fifteenth century. When James II of Aragon and the citizens of Barcelona expressed scruples about selling enslaved Greeks in the market in 1314, on the grounds that they were found to be Christian, the trade may have been rendered doubly repugnant to them because many of those Christian slaves were children.(56)

Nevertheless there were great advantages for commercial households in purchasing very young female slaves. Foreign merchants, travelling agents, factors and porters were not always fed or housed in the town but rather in the commercial household.

Goods arrived seasonally or when the fleet came in, lending commercial establishments a rhythm of intense work interrupted by periods of inactivity that did not lend itself to hiring wage-labour. Datini's eight- to ten-year-old child would thus supplant hired male servants in a reallocation of tasks - a reassigned sexual division of labour. The new ancilla would indeed tend the ovens, haul the wood and wash up the pots and cauldrons. (Whether she would be trusted with the good plate is another question.) She would also haul loads, sew commercial goods into bales, sort, fetch and carry.(57)

Female workers were considered more tractable than males, and housed as they were with other servants, subordinate to the "good" slave Datini mentioned, they found scant opportunity to act otherwise. Extreme youth, separation from kin and friends and a wholly foreign environment cut off most paths of rebellion. Also, the mistress of the house might prefer to take on the challenge of directing female rather than male slaves.(58) This medieval system of household slavery was enforced largely by lack of alternatives, in contrast to the beatings that had "motivated" male slaves in former times. Affluent householders seldom judged such domestic arrangements inhumane. When Ragusans outlawed the slave-trade in 1416 it was the export of slaves overseas that was deemed harsh and illegal. What they did in their own households escaped censure.(59)

To return to Prato: Genoa did not supply Datini with a suitable slave, so he turned instead to the market at Venice, where he bought a child named Orenetta. She was very likely a Bosnian, and therefore a Christian child imported from across the Adriatic Sea.(60) And she pleased the household. On New Year's Day 1395, Francesco noted in his private account-book: "2 soldi to Orenetta, the little slave who comes from Venice".(61) This "little slave" had begun her ascent through the complex layers of servitude in a great commercial household.

It is valuable to compare the Datini household in Prato with similar wealthy households in port cities. In Ragusa (now Dubrovnik), merchant citizens' homes were scattered all over the city, with a concentration in the oldest sexteria on the high perch of Castellum. Commercial goods, firewood and foodstuffs had to be transported up steps from the two city gates and the harbour in order to reach these homes. While men, a few of them also enslaved, might be used to haul, cart, or ship outside the city walls, within the town loads were often packed on the backs of women or girls for the trek up the steps.(62) One clear advantage of this arrangement was the fact that a squad of unfree women, unlike a specialized labour-force of men, could be redirected to work inside the household when the commercial needs of the establishment had been satisfied. In this community, even the fifteenth-century Renaissance palaces newly built outside the town continued to double as commercial premises (on the ground floor) and family homes (on the first floor and above). Domestic architecture in most port cities favoured similar solutions. Venice, whose trunk, arterial and neighbourhood canals most closely approximated to our modern highway systems (down to the attached boat-house equivalent to our garages) gives the clearest visual evidence of this efficient urban deployment of labour. Gondolas were (literally) manned, while women often hefted commercial goods once they were delivered to the household.

A division of labour into public and private components assigned to men and women respectively suited urban commercial households. Where cost-efficiency lay in keeping down overheads, a multi-purpose labour force of tractable individuals was a sufficiently powerful incentive for private householders to alter the sexual division of labour and employ numbers of enslaved women and girls. Italian merchants may, in fact, have observed the use of slave labour when trading in the eastern Adriatic and Black Sea regions and brought the solution back home. In the markets of Ragusa, where slaves had traditionally served households, seasonal variation in the price of slaves emerged by the late thirteenth century, at just the time when urban household slavery had begun to spread more widely in Italy. When Venetian ships lay in port in the late summer sales increased quite dramatically. Venetians would buy virtually any slave they could find, including the personal or "dowry" slaves of noble households. It was entirely probable that Italian merchants used the stop at Ragusa, on the Levantine fleet's return voyage, to secure themselves slaves for their own households.(63)

For the child Orenetta in the Datini household, provided she pleased her employers an incentive system would in time lead to more skilled and less exhausting tasks, including those traditionally identified as the preserve of women. None the less, Orenetta's assimilation took place during the years of puberty, and it was not out of the question that a child would be born to her in slavery. In the childless Datini household Mona Marguerita Datini had already agreed to raise as her own daughter Lucia, the child of a household slave who had been fathered by her husband. This child was remarkably fortunate, as was the Florentine story-teller Franco Sacchetti, the child of a female slave, when his father's Florentine bride accepted him into her household.(64)

A snide scrap of doggerel from Florence noted how householders "married off" their locally recruited rural servants with dowries once they became pregnant by a man of the casa. On balance, though, this seems a far kinder fate than that of the child-bearing female slave.(65) Rural girls serving great households for a contract had the legal recourse of marriage with dowry, but the pregnant slave-girl produced a child who belonged to a household. Her masters could recognize, pamper, free or disregard the child at will. Even the happy outcomes for Lucia Datini and Franco Sacchetti may conceal stories of violated motherhood: neither slave mother called her child her own.(66) A system of assimilation that left the definition of a "good" slave entirely to the whim of the household exerted enormous power over a female slave. Harriet Beecher Stowe, who placed the abuse of domestic relations at the top of the list of slavery's outrages in nineteenth-century America, might have commented similarly on late medieval domestic slavery, that took a child from its mother even if the mother qualified as a "good" slave.(67)

Over time a compliant slave learned skills. Datini's older "good" slave could both cook and sew, and was expected to exercise both skills at the command of her owners. In 1400 a ship sailing from Venice to Majorca listed "nine Turkish heads" worth 360 lire on its ship's lading. One of these slaves was a woman who could "sew and do everything". The agent found her too good for the Majorcans, so she travelled on with the ship to be sold in Valencia.(68) Any woman with skills, particularly cooking, spinning, weaving and sewing, would command a high price. And slaves were expensive. In Italy a ten-year-old might cost as little as 20 florins in the fourteenth century; a skilled woman in her late teens cost 80 florins or even more.(69) In fact, an earlier thirteenth-century slave-girl's market price rose from about 5 ducats in Ragusa to 30, 40 or 50 ducats once she was transported to Venice, an increase which demonstrates the strong demand for female slave labour in Italian port cities. And on top of this, Italian markets also taxed slave sales.

Because slaves were less likely to leave a household than were rural girls who worked under contract, slaves' learnt skills took on long-term consequence. Did these great households ever supplant the gynaecea or female workshops of an earlier age? It is certainly suggestive that both institutions were supplied by female slave labour. Market forces, and the competitive quest for status that had heightened the role of fashion and displays of conspicuous consumption in Italian cities, brought new and heavy responsibilities to urban households. Economic conditions might well produce a "squeeze" which in turn affected the use of labour in households. Householders, beset by varied and mounting pressures without new labour-saving technologies to ease the strain, might increase their demands on unfree domestic servants. Indeed the swift redeployment of labour which was possible when directing slaves may have appealed to households facing numerous demands with only limited financial resources.

Frank Walbank's discussion of the decentralizing of glass production in the late antique economy prompted by reliance upon slave labour may be relevant here.(70) Great Roman estates in all corners of the empire began to employ semi-skilled slaves to produce glassware for local consumption, leading to the gradual disappearance of the higher-calibre glass, produced by waged workers, that had previously been exported from the city of Rome to rural consumers. No direct parallel to the Renaissance household economy is intended here, but it is worth exploring whether the factors that led to the long-drawn-out demise of women's workshops from the central Middle Ages until the fifteenth century, and the introduction of a domestic setting as a primary system for the delivery of women's production of goods (most significantly textiles), may rest in some part on the persistence of slavery as a domestic institution.(71) Resolving this issue might help explain how producers cut costs to meet market competition.

Among those factors which might be analysed to answer this question is whether, under the duress of the depressed economy of the fourteenth and early fifteenth centuries, some well-to-do households assumed production tasks as well as the commercial and domestic work performed routinely within their confines. With an increased emphasis upon display, urban families may have found that some costs could not be absorbed easily into the household - for instance, one could not afford one's own aurifici who made plate or painted portraits unless one were extraordinarily wealthy - but the moderately wealthy urban household might absorb some of the endless, time-consuming tasks associated with the production of fashion apparel. Even if households purchased fine cloths from urban workshops they could be finished and embellished to create fashion at home - if the proper skills were found there. And the production of fashionable clothes was highly labour-intensive, repetitive and tedious, as well as demanding skilled work.

Added to this, by their very design banquets and revels relied on labour-intensive service. With slave labour at a household's disposal, both ostentations might be better brought within cost range, if for no other reason than that the same slaves could be employed alternately in the work at the owner's command. No labour supply system except slavery offered this advantage, because hierarchy among servants in urban households depended on an accepted ordering of rank based on a complex grading of tasks among the free. Even in its long-drawn-out demise, the traditional gynaeceum may have exerted its influence through suggesting to wealthy householders that slavery represented a cost-effective solution to the competitive endeavours of display, entertainments and wearing rich clothes.

If so, this adds a new twist to a commercial economy, though one in line with Richard Goldthwaite's consumption model of the Italian urban economy of the fifteenth century. Goldthwaite asserts that by this period urban Italians imported few products because they satisfied consumption preference with goods they produced locally.(72) While Florentines actually imported slaves for only a brief period, returning by the end of the fifteenth century to the traditional cheap source of rural labour supplied by contract, they did make thrifty use of the slaves they owned, renting them out to others and generally treating them as valuable resources.(73) It has been suggested that the presence of imported slaves served as a silent warning to local servants that they could be dispensed with if they pressed too hard for better rewards and working conditions.(74) The lifelong use of those slaves whom householders sought out and imported may have prompted owners to seek carefully for trained slaves, or alternatively to see that their own purchased slaves acquired skills.(75)

What provokes this line of reasoning is the persistent mention of sewing as a valued skill among slaves. Annarosa Garzelli has noted that numbers of foreign embroiderers entered Italy in the century after the Black Plague.(76) While production was controlled at the highest level by guild statutes, the fashionable woman desiring the almost prohibitively expensive skills of an embroiderer gained a great advantage if her slave had the necessary expertise. In her study of Italian sumptuary law, Diane Owen Hughes mentions women who registered gowns and mantles above the value laid down by law.(77) Did they employ slaves in their households? Lace-making, generally assumed to be in its modern guise an Italian invention of the fifteenth century, may well have owed a debt to the nimble fingers of slaves serving for a lifetime in domestic households.

It is less conjectural that Europeans kept before their eyes the example of slave labour's efficacy throughout the Middle Ages. Bloch's argument about the "profound essence of slavery" may provide insight into how the institution coloured thought about the free and the unfree, and perhaps even help account for some change in the sexual division of labour. Slavery for women may have influenced thinking about the work appropriate to women in the household. Where the domestic establishment assimilated its own workers, introducing slaves to less taxing, more skilled and more gendered roles as they won favour with the household, benefits to owners appear to have outweighed slaves' high purchase-prices. This was so because slaves did all kinds of work over a lifetime. Cateruccia, the garrulous, angry old slave of Alessandra Strozzi (whose voluminous correspondence gives a precious insight into Florentine domestic life), seems to have been the only mortal whom Alessandra ever feared, for were she let go her gossip about the household would "ruin" the family name. Whatever domestic austerities Cateruccia fretted under, one thing is clear: Alessandra could be left to live in her own home in her old age because her old slave was still with her. Even in old age Cateruccia had value to the Strozzi family.(78)

The demise argument about slavery is challenged in the rural countryside by the survival of ancillae, and later by the revival of trade in slaves and their use in numerous European towns. Throughout the medieval centuries the term ancilla evoked deeply moving images to medieval people; indeed some saintly women (even a few of privileged birth) found the epithet so compelling that they adopted it to express their sense of themselves as the most humble, the lowest in status: ancilla Dei, the handmaiden of the Lord.(79)

Domiciled slaves, although much reduced in numbers by the central Middle Ages, had survived in Europe. They kept the idea of slave labour alive, and within the domestic sphere in southern towns slavery saw a revival in the later medieval period. Were these enslaved women as essential to the economy as slaves employed in farming endeavours? Clearly not, because nothing was as essential to the medieval economy as agricultural production. Nevertheless, textile workshops and great households employed slaves through the central Middle Ages, and thereafter the diversification of the economy created conditions where domestic slavery flourished. A corps of enslaved women was found to be a useful adaptation to market forces. This helped assure the institution's survival.

While domestic slavery's methods of control were less physically abusive than was the case with slavery in its other guises, it is well to remember that slavery was and is an essentially violent institution. Aside from a few Tartar, Greek and Slavic names and a few verses in linguaggio di schiave, slaves working in Italy's households have left almost no trace.(80) The same is true of those in ports like Barcelona and Valencia, and in the islands of the Mediterranean; indeed, historians' neglect of the survival of the institution throughout the central Middle Ages may be owing to the very scarcity and elusiveness of the evidence for it. The institution itself was renamed, memorializing the women from the Balkans transported around the Mediterranean, but this was surely a notoriety that no Slav sought.(81) The true violence done to slaves lay in their almost complete deracination, for their culture was erased and replaced, and even running away allowed almost no chance to return to homes far across the sea.

This sea-borne migration of unfree female labour anticipated the larger sea-borne migration across the Atlantic Ocean which began in the sixteenth century. In fact Mediterranean slaves, men and women alike, were employed on the plantations of Crete and other islands - a dress rehearsal, so to speak, for their introduction on to the plantations of the New World.(82) And domestic slavery in medieval households played its own instructive role in the early modern diffusion of slavery. Slaves in medieval households were rendered compliant through effective controls: they served cheaply and well. Thus this slave corps composed chiefly of women provided a compelling model for others who later used the plantation household as the locus of both commerce and production.

1 Wilhelm Heyd, Histoire du commerce du Levant au Moyen Age, 2 vols. (Leipzig, 1923), i, p. 67. The term "slave" is used here in its modern sense, and refers to both women and men bought and sold as chattels. For one of the early medieval uses of slavi, see n. 81 below.

2 Charles Verlinden, L'esclavage dans l'Europe medievale, 2 vols. (Bruges and Ghent, 1955-77).

3 Marc Bloch, "Mediaeval 'Inventions' ", in his Land and Work in Mediaeval Europe, trans. J. E. Anderson (London, 1967), p. 178. See also M. Bloch, Slavery and Serfdom in the Middle Ages, trans. William R. Beer (Berkeley, 1975).

4 Marc Bloch, "Personal Liberty and Servitude", in Slavery and Serfdom in the Middle Ages, pp. 33-92, discussed by Pierre Bonnassie, From Slavery to Feudalism in South-Western Europe (Cambridge, 1991), pp. 331-4 (quotation from p. 331).

5 [Andrew Horne], The Mirror of Justices, ed. William Joseph Whittaker (Selden Soc., vii, London, 1895), p. 77.

6 Adriaan Verhulst, "The Decline of Slavery and the Economic Expansion of the Early Middle Ages", Past and Present, no. 133 (Nov. 1991), pp. 195-203; Dominique Barthelemy, "Qu'est-ce que le servage en France au X[I.sup.e] siecle?" Revue historique, clxxxvii (1992), pp. 233-84; William Jordan, rev. of Bonnassie, From Slavery to Feudalism, Slavery and Abolition, xiii (1992), pp. 97-102.

7 Ruth Mazo Karras, Slavery and Society in Medieval Scandinavia (New Haven, 1988), pp. 138-40.

8 Robert Brenner, "Agrarian Class Structure and Economic Development in Pre-Industrial Europe", Past and Present, no. 70 (Feb. 1976), pp. 30-75; Pierre Dockes, Medieval Slavery and Liberation, trans. Arthur Goldhammer (Chicago, 1982). For the Russian literature on medieval servitude, see Yuri L. Bessmertny, "August 1991 as Seen by a Moscow Historian, or the Fate of Medieval Studies in the Soviet Era", Amer. Hist. Rev., xcvii (1992), pp. 803-16. For the deviation of Russian historians from a strict Marxist approach to the issue of slavery, see Z. V. Udal'tsova et al. (eds.), Istoriia krest'ianstva v Europe: epokha feodalizma [History of the Peasantry in Europe: Feudal Epoch], 3 vols. (Moscow, 1985-6). For an English translation of the views of one participant in this scholarly venture, see Aron Gurevich, Categories of Medieval Culture, trans. G. L. Campbell (Chicago, 1985).

9 Iris Origo, "The Domestic Enemy: The Eastern Slaves in Tuscany in the Fourteenth and Fifteenth Centuries", Speculum, xxx (1955), p. 332.

10 Polyptique de l'Abbaye de Saint-Germain-des-Pres, ed. A. Longnon (Paris, 1886), pp. 58-61, cited in David Herlihy, Medieval Culture and Society (New York, 1968), pp. 52-5. The problem of Carolingian terminology has recently been reopened: Hans-Werner Goetz, "Serfdom and the Beginnings of a Seigneurial System in the Carolingian Empire", Early Medieval Europe, ii (1993), pp. 29-51.

11 Pierre Petot, "L'evolution numerique de la class servile en France du I[X.sup.e] au [XIV.sup.e] siecle", in Le servage, 2nd edn (Recueils de la Societe Jean Bodin, ii, Brussels, 1959; first pubd Brussels, 1937), pp. 159-68.

12 Michael Postan, The Famulus: The Estate Labouter in the XIIth and XIIIth Centuries (Econ. Hist. Rev. Suppl. ii, Cambridge, 1954), pp. 2-14.

13 Archives Nationales, Paris, X A 14, fos. [502.sup.r]-[508.sup.r], cited in Petot, "Evolution numerique de la classe servile en France", p. 163.

14 Archives Departementales de l'Yonne, Auxerre, H169, "Inventaire par ordre chronologique", 1703, p. 3, cited in William Jordan, From Servitude to Freedom (Philadelphia, 1986), p. 56: "Item predicti homine et femine manumissi et eorum heredes quamdiu erunt manentes et commorantes in terris".

15 Monumenta Germaniae Historica, Diplomatum Regum et Imperatorum Germaniae, v, ed. H. Bresslau and P. Kehr (Berlin, 1931), doc. 253, pp. 336-7 (16 July 1050), cited (from an earlier edition) in A Source Book for Medieval History, ed. O. J. Thatcher and E. H. McNeal (New York, 1905), pp. 547-8.

16 For example, on royal manumissions in France, see Bonnassie, From Slavery to Feudalism, pp. 56-7, especially the table on p. 57, based on Veronique Sablayrolles, "De l'esclavage au servage de Louis le Pieux a Philippe Ier. d'apres les actes royaux" (Univ. of Toulouse-Le Mirail, memoire de maitrise, 1982). The issue is cast in the masculine by Marc Bloch, "How and Why Ancient Slavery Came to an End", in Slavery and Serfdom in the Middle Ages, pp. 1-31, and indeed by Bonnassie and the authors of many regional studies. For canon law favouring the manumission of unfree persons marrying outside a lord's holdings, see John Gilchrist, "The Medieval Canon Law on Unfree Persons: Gratian and the Decretist Doctrines, c. 1141-1234", in Stephan Kuttner and Alfons M. Stickler (eds.), Melanges G. Fransen, 2 vols (Studia Gratiana, xix-xx, Rome, 1976), i, pp. 285, 290-3.

17 Compare the evidence offered by Paul Freedman, The Origins of Peasant Servitude in Medieval Catalonia (Cambridge, 1991), app. 4, pp. 227-30. This series of redemptions from serfdom in the diocese of Girona between 1200 and 1282 favours women over men. However, a number of these redemptions are nominal sums paid to enable the serf to marry outside the lord's jurisdiction: ibid., pp. 132-3. Presumably a lord gained as well as lost in this kind of transaction.

18 For a useful study showing the variety of practices relating to slaves under the Roman Empire, not all of which squared with Roman law, see Keith Hopkins, Conquerors and Slaves (Cambridge, 1978). See also Moses Finley, Ancient Slavery and Modern Ideology (New York, 1980); David Davis, Slavery in Western Thought (Ithaca, 1966).

19 R. H. Hilton, "Freedom and Villeinage in England", Past and Present, no. 31 (July 1965), p. 8.

20 During the canonization process of St Margaret of Hungary (d. 1270), when female witnesses were asked if they were free or slaves ("Si est libera, vel ancilla?"), some are said to have replied, "Serra et libera": Michael Goodich, "Ancilla Dei: The Servant as Saint in the Late Middle Ages", in Suzanne Wemple and Julius Kirschner (eds.), Women in the Medieval World (Oxford, 1985), p. 120 n. 5. However, I have not found this phrase in any servant's testimony reported in the record of the process: Monumenta romana episcopatus Vesprimiensis, 4 vols. (Budapest, 1896-1907), ii, pp. 306-58.

21 Benjamin Arnold, German Knighthood, 1050-1300 (Oxford, 1985), pp. 29-37. For a different view of the unfree status of Salzburg knightly families, see John Freed, "Nobles, Ministerials and Knights in the Archdiocese of Salzburg", Speculum, lxii (1987), pp. 575-611. On Carolingian nomenclature, see also Goetz, "Serfdom and the Beginnings of a Seigneurial System in the Carolingian Empire", pp. 29-51.

22 For a variety of opinions, see M. M. Postan, Famulus, pp. 1-14; [Horne], Mirror of Justices, ed. Whittaker, pp. 77-9; Paul R. Hyams, Kings, Lords and Peasants in Medieval England: The Common Law of Villeinage in the Twelfth and Thirteenth Centuries (Oxford, 1980), p. 126; John Hatcher, "English Serfdom and Villeinage: Towards a Reassessment", Past and Present, no. 90 (Feb. 1981), pp. 3-39.

23 See Josip Lucic, Obrti i usluge u Dubrovniku do podetka xiv stoljeca [Trades and Services in Dubrovnik at the Beginning of the Fourteenth Century] (Zagreb, 1979), pp. 136-60; Susan Mosher Stuard, "Urban Domestic Slavery in Medieval Ragusa", Jl Medieval Hist., ix (1983), pp. 162, 170.

24 Stuard, "Urban Domestic Slavery in Medieval Ragusa", p. 162; Lucic, Obrti i usluge u Dubrovniku, pp. 136-40.

25 Stuard, "Urban Domestic Slavery in Medieval Ragusa", p. 162. This blurring of distinctions might have helped the smooth conduct of household labour, because co-operative work among free and unfree servants may have been more easily obtained by incentives rather than by punishments. However, the language of slave contracts preserved the precise Latin usage.

26 On the condition of slaves in ancient society, see Finley, ncient Slavery and Modern Ideology, pp. 93-122. On the evolution of the late Roman law of slavery, see C. R. Whittaker, "Circe's Pigs: From Slavery to Serfdom in the Later Roman World", Slavery and Abolition, viii (1987), pp. 88-112. On Lombard law and the purchase of a female slave, see Die Gesetze der Langobarden, trans. Franz Beyerle, 3 vols. (Weimar, 1947), iii, no. 231, p. 94 ("Rothair's Edict"); The Lombard Laws, trans. Katherine Fischer Drew (Philadelphia, 1973), p. 98; AElfric, "Glossarium", ed. William Somner, in Dictionarium Saxonico-Latino-Anglicum (repr. Menston, 1970), p. 101. For a discussion of sales contracts for slaves, see Stuard, "Urban Domestic Slavery in Medieval Ragusa", pp. 155-71. Some Roman law re-emerged intact in medieval city-state statute law, especially in the region of the Adriatic Sea: The Rhodian Sea-Law, ed. Walter Ashburner (Oxford, 1909), pp. clxviii-clxix. This was particularly true in the case of the laws governing slavery.

27 Gilchrist, "Medieval Canon Law on Unfree Persons", pp. 276-92. An even stronger case for the amelioration of the servile condition of men as heads of families has been made: Michael M. Sheehan, "Theory and Practice: Marriage of the Unfree and the Poor in Medieval Society", Mediaeval Studies, 1 (1988), pp. 457-87; James Brundage, Law, Sex and Christian Society in Medieval Europe (Chicago, 1987), p. 196.

28 See, for example, Gesetze der Langobarden, trans. Benerle, iii, no. 221, pp. 88-90, cf. Lombard Laws, trans. Drew, no. 221, p. 95; for examples from the central Middle Ages and later, see Gilchrist, "Medieval Canon Law on Unfree Persons", pp. 276-92.

29 Historijskog Arhiva u Dubrovniku (hereafter H.A.D.), Lamenta de Intus, 1404-7, fo. [181.sup.r] (15 Apr. 1406).

30 Michel Balard, "La femme-esclave a Genes a la fin du Moyen Age", in Michel Rouche and Jean Heuclin (eds.), La femme au Moyen-Age (Maubeuge, 1990), p. 306.

31 Gilchrist, "Medieval Canon Law on Unfree Persons", p. 277, citing the following: Gratian, Decretum, pt i, dist. 34, c. 15, in Corpus iuris canonici, ed. Emil Friedberg, 2 vols. (Leipzig, 1911), i, col. 129; dist. 81, c. 30 (ed. Friedberg, i, col. 288); pt ii, causa 15, questio 8, c. 3 (ed. Friedberg, i, col. 760).

32 "Ideoque quilibet ab episcopo usque ad subdiaconum deinceps vel ex ancillae vel ex ingenuae detestando conubio in honore constituti filios procreaverint, illi quidem, ex quibus geniti probabuntur, canonica censura dampnentur; proles autem, aliena pollutione nata, non solum hereditatem numquam accipiet, sed etiam in servitute eius ecclesiae, de cuius sacerdotis vel ministri ignominia nati sunt, lure perhenni manebunt": Gratian, Decretum, pt ii, causa 15, questio 8, c. 3 (ed. Friedberg, i, col. 760). It is open to interpretation whether all female servants cohabiting with clerics were to be treated as ancillae by canon law. If this was so, canon law had gone a considerable way to enforce clerical celibacy.

33 Gilchrist, "Medieval Canon Law on Unfree Persons", pp. 290-2. David Herlihy finds an economic motive for gifts to the church of ancillae (mature women as well as girls), believing that these women staffed textile workshops for the continued production of essential textiles: D. Herlihy, Opera muliebria: Women and Work in Medieval Europe (Philadelphia, 1990), pp. 83-94.

34 "I will not examine here in detail the solutions offered to the problem of status when the two parents were of different conditions. Two principles have been opposed: at times the opinion was that 'the worse won'; at times servitude was considered transmitted matrilineally. The latter thesis was Beaumanoir's and seems to have generally triumphed in juridical literature. In earlier eras the difficulty did not often present itself, since one rarely allowed a free individual to marry a servile one without agreeing to be a serf himself": Bloch, Slavery and Serfdom in the Middle Ages, p. 205 n. 23.

35 Is the common law the exception? Paul Hyams believes that in medieval England unfree status was inherited from either parent, while in The Mirror of Justices it is said to descend only from the father: Hyams, Kings, Lords and Peasants in Medieval England, p. 126; [Horne], Mirror of Justices, ed. Whittaker, p. 7.

36 H.A.D., Testamenta, V, fo. [41.sup.v].

37 Some evidence of property owning exists where slaves "purchased" their own freedom: Stuard, "Urban Domestic Slavery in Medieval Ragusa", p. 168; Balard, "Femme-esclave a Genes", pp. 306-8; Origo, "Domestic Enemy", p. 351.

38 See, for example, Freedman, Origins of Peasant Servitude in Medieval Catalonia, pp. 119-78.

39 Herlihy, Opera muliebria, p. 83.

40 Notker Balbulus ("the Stammerer"), Gesta Karoli Magni, ii.9, in Einhard and Notker the Stammerer: Two Lives of Charlemagne, trans. Lewis Thorpe (New York, 1969), p. 147.

41 For this example, and for many others of slave-staffed textile workshops, see Herlihy, Opera muliebria, pp. 83-94; see also Julia M. H. Smith, "The Problem of Female Sanctity in Carolingian Europe, c. 780-920", Past and Present, no. 146 (Feb. 1995), p. 19, p. 20 n. 58.

42 Chretien de Troyes, Yvain, the Knight with the Lion, trans. Ruth Harwood Cline (Athens, Ga, 1975), lines 4957-70, p. 147.

43 Herlihy, Opera muliebria, pp. 185-91.

44 See Richard C. Hoffmann, "Outsiders by Birth and Blood: Racist Ideologies and Realities around the Periphery of Medieval Culture", Studies in Medieval and Renaissance Hist., new ser., vi (1983), pp. 14-20. It is perhaps more than a coincidence that these same "frontiers" practised slavery and understood that inheritance conveyed unfree status.

45 Heath Dillard, Daughters of the Reconquest (Cambridge, 1984), pp. 98, 154, 207-8.

46 Karras, Slavery and Society in Medieval Scandinavia, pp. 21, 26, 31-2, 70, 91, 94, 129, 146-7. The latest references in England I can find occur in late fourteenth-and fifteenth-century wills: P. J.P. Goldberg, Women, Work and Life Cycle in a Medieval Economy: Women in York and Yorkshire, c.1300-1520 (Oxford, 1992), p. 182.

47 H.A.D., fr. 985, "Pismo Kneza Crnomira knezu i opcini Dubrovackoj" [Letter of Prince Crnomir to the Ruler and Community of Dubrovnik], in Stare srpske povelje i pisma [Serbian Charters and Letters], ed. L. Stojanovic, 2 pts (Belgrade, 1929-34), i, doc. 25, pp. 23-4.

48 Stuard, "Urban Domestic Slavery in Medieval Ragusa", p. 157.

49 Malcolm Burr, "The Code of Stefan Dusan", Slavonic and East European Rev., xxviii (1949-50), pp. 380-1. Serbs as well as Greeks were enslaved, however: Charles Verlinden, "Orthodoxie et esclavage au bas Moyen Age", in Melanges Eugene Tisserant, 7 vols. (Studi e testi, ccxxxi-ccxxxvii, Vatican City, 1964), v, pp. 427-56; C. Verlinden, "Le recrutement des esclaves a Venise aux [XIV.sup.e] et [XV.sup.e] siecles", Bulletin de l'Institut Historique Belge de Rome, xxxix (1968), pp. 83-202.

50 Kotorski spomenici, god. 1326-35 [Commemoration Volume for Kotor, 1326-35], ed. Antun Meyer (Zagreb, 1951), doc. 171, p. 72; doc. 1019, p. 343.

51 A. Tenenti, "Gli schiavi di Venezia alla fine de cinquecento", Rivista storica italiana, lxvii (1955), pp. 52-69.

52 David Herlihy emphasizes the reputation for licentiousness that surrounded gynaecea, believing it originated in the fact that in the early and central Middle Ages they housed the only concentration of adult unmarried women outside convents: Herlihy, Opera muliebria, pp. 80-94. Georges Duby strikes the same chord when discussing humble women-servants in great feudal households who provided a sexual outlet for "youth" (that is, nobly born bachelors): G. Duby, Love and Marriage in the Middle Ages, trans. Jane Dunnett (Chicago, 1994), pp. 22-35, 59-60. See also Goodich, "Ancilla Dei", pp. 122-3.

53 It was not unusual for slaves to be so young: Iris Origo, The Merchant of Prato (New York, 1957), p. 206. Origo directed attention to the issue of domestic slavery well before others recognized its importance, and was pleased to see it "addressed again" before her death in 1988 (personal communication, April 1984). The following illuminating example from her research in the Datini archives is included here to honour her insight.

54 Archivio di Stato, Florence, classe XI, distr. 8, no. 81, cited in Origo, Merchant of Prato, p. 206 n. 4.

55 Stuard, "Urban Domestic Slavery in Medieval Ragusa", pp. 155-71. For important studies of slavery in the Adriatic, see Gregor Cremosnik, "Pravni polozal naseg robilja u sredjem veku" [Legal Position of our Labourers in the Middle Ages], Zemaljski Muzej u Bosni i Hercegovini: Glasnik, new ser., ii (1947), pp. 69-73; G. Cremosnik, "Izvori za istoriju robilia i servijalnih odnosa u nasim zemljama sr. vijeka" [Evidence for the History of Labourers and Servants in the Medieval World], Istorijski pravni zbornik, i (1949), pp. 146-52. See also A. Teja, "La schiavitu domestica ed il traffico degli schiavi", Rivista dalmatica, xxii (1941), pp. 33-44; A. Tenenti, "Gli schiavi di Venezia alla fine del cinquecento", Rivista storica italiana, lxxvii (1955), pp. 52-69.

56 Verlinden, "Orthodoxie et esclavage au bas Moyen Age", p. 428.

57 This argument follows the line established in studies of great noble households where retinues of men served at table and prepared food. For such tasks to be assigned to women constitutes a shift which may be dated to the later Middle Ages, although of course regional differences in sexual division of labour exist: see Joseph and Frances Gies, Life in a Medieval Castle (New York, 1974), pp. 95-124; Pierre Riche, Daily Life in the World of Charlemagne, trans. Jo Ann McNamara (Philadelphia, 1978), p. 100.

58 Compare the experience of Lizzie Neblett, when she was left to manage a farm and male slaves for two years during the American Civil War: Drew Gilpin Faust, "'Trying to Do a Man's Business': Slavery, Violence and Gender in the American Civil War", Gender and History, iv (1992), pp. 197-214.

59 Stuard, "Urban Domestic Slavery in Medieval Ragusa", p. 170. But punishment of slaves could be harsh, and include beating, maiming and assignment to harder labour.

60 At Genoa, Tartar or Black Sea slaves were often baptized before they were allowed to serve in Christian households. In Adriatic towns this practice was far less common, suggesting that slaves imported from the west Balkans were understood to be at least nominally Christian.

61 Origo, Merchant of Prato, p. 206.

62 While almost 90 per cent of local slave sales specified women, there were always some males within the urban slave population. Servi ran away more often than ancillae. Over half the fugitive slaves advertised for capture in early fourteenth-century Ragusa were male, whereas the great preponderance of the enslaved were female: Stuard, "Urban Domestic Slavery in Medieval Ragusa", pp. 155-8.

63 Ibid., pp. 165-6.

64 Origo, Merchant of Prato, pp. 205-16; Ignacij Voye, "Bencius del Buono", Istorijski casopis, xix (1971), pp. 189-99.

65 Christiane Klapisch-Zuber, "Childhood in Tuscany", in her Women, Family and Ritual in Renaissance Italy, trans. Lydia Cochrane (Chicago, 1985) pp. 107, 165-77; C. Klapisch-Zuber, "Women Servants in Florence during the Fourteenth and Fifteenth Centuries", in Barbara Hanawalt (ed.), Women and Work in Preindustrial Europe (Bloomington, 1986), pp. 56-80.

66 Gilchrist, "Medieval Canon Law on Unfree Persons", pp. 289-94. A serf (originarius) had rights over a daughter, in contrast to a slave (famulus), who did not. Here, as elsewhere, the terminology is not consistent.

67 Cited by Elizabeth Fox-Genovese, "Ghosts and Memories: The Legacy of Slavery in Women's Imaginings", unpubd paper, Anna Howard Shaw Symposium, Bryn Mawr Coll., 16 Feb. 1991.

68 Archivio Datini, Prato, file 1142, cited by Origo, "Domestic Enemy", pp. 330-1. The date (September 1400) is supplied by another letter about the same goods.

69 Origo, "Domestic Enemy", pp. 336-40.

70 F. W. Walbank, The Awful Revolution (Toronto, 1969), pp. 85-6, 122.

71 Herlihy, Opera muliebria, esp. pp. 185-92 ("Conclusion"); see also n. 15 above. Ragusa, which practised domestic slavery, also represents an exceptional urban environment in which the women's textile workshop was restored. In the fifteenth century, when cheap cloth could not be easily obtained from overseas, Ragusans initiated large-scale production of cheap cloth for export inland through a certain Petar Pantella, brought from Italy for his expertise. He employed numerous country women in a factory-like environment for cloth manufacturing at very low cost. As soon as imported cloth became available the workshop was broken up. See Dusanka Dinic-Knezevic, "Petar Pantella, trgovac i suknar u Dubrovniku" [Petar Pantella, Merchant and Cloth-Maker in Dubrovnik], Godisnjak filozofskog fakulteta u Novom Sadu, xiii (1970), pp. 87-114.

72 Richard A. Goldthwaite, "The Renaissance Economy: The Preconditions for Luxury Consumption", in Atti del Convegno di Studi nel X anniversario della morte di Federigo Melis (Florence, 1985), pp. 659-71; R. A. Goldthwaite, Wealth and the Demand for Art in Italy, 1300-1600 (Baltimore, 1993), ch. 1.

73 Klapisch-Zuber, "Women Servants in Florence", pp. 68-9.

74 Piero Guarducci and Valeria Ottanelli, I servitori domestici della casa Borghese toscana nel basso Medioevo (Florence, 1982), pp. 79-80.

75 On the issue of surveillance, the presence of slaves may have acted to reduce the status of the nominally free servant in towns. In Ragusa, during the terrifying months of the Black Plague when disorder as well as disease haunted the city, the authorities strengthened legal sanctions over servants of all kinds; a new law stated flatly that all the conditions of surveillance pertaining to enslaved persons were now extended to all other servants of the household: Liber statutorum civitatis Ragusii, compositus anno 1272, ed. V. Bogisic and C. Jirecek (Monumenta historico-juridico Slavorum meridionalium, ix, Zagreb, 1904), VIII, c. 93, p. 224; VI, c. 33 (4), p. 136.

Origo has found that laws governing slaves in Florence and Venice similarly increased in severity in times of disaster: Origo, "Domestic Enemy", pp. 340-1. Balard finds the same true in Genoa: Balard, "Femme-esclave a Genes", pp. 299-310. Strengthening the householder's authority was intended to restore urban order, but for nominally free servants the legislation had the effect of levelling them to the lesser status of the unfree.

76 Annarosa Garzelli, Il ricamo nella attivita artistica di Pollaiolo, Botticelli, Bartolomeo di Giovanni (Florence, 1973), p. 43.

77 Diane Owen Hughes, "Sumptuary Law and Social Relations in Renaissance Italy", in John Bossy (ed.), Disputes and Settlements (Cambridge, 1984), pp. 87-8. The registered mantles of the Florentine sisters-in-law Nera and Piera de Albizi were made of the finest lined white wool, embroidered with vines and red grapes. Their mother-in-law registered an equally rich purple mantle with a reverse trim of green vines and white grapes. The "at-home" production of elegant co-ordinated mantles might well come within reach of an affluent household, if it contained a skilled needlewoman.

78 Alessandra Macinghi Strozzi, Lettere di una gentildonna fiorentina del secoli XV ai figliuoli esuli, ed. Cesare Guasti (Florence, 1877), pp. 103-4 (6 Dec. 1450), 474 (13 Sept. 1465), cited by Origo, "Domestic Enemy", p. 343.

79 For example, Gunthild of Suffersheim, Zita of Lucca, Margaret of Louvain, Margaret of Citta di Castello, Sibillina Biscossi of Pavia, Veridiana Attavanti of Castelfiorentino, Jane of Orvieto and Margaret of Hungary: Goodich, "Ancilla Dei", pp. 119-28. Though the subtitle of his paper suggests a bias to the late Middle Ages, Goodich's list of ancillae Dei in fact begins with examples from north of the Alps in the central medieval period. The phrase refers back to the Vulgate translation of Luke 1:38: "ecce ancilla Domini".

80 Mario Ferrara, "Linguaggio di schiave nel quattrocento", in Studi di filologia italiana, viii (1950), pp. 320-8. Tartar female names used in Italy include Cotlu, Jamanzach, Tholon, Charactas, Aycholu and Bollaza.

81 Liber statutorum civitatis Ragusii, ed. Bogisic and Jirecek, I, c. 14, p. 11. It is sometimes claimed that the earliest occurrences of slavi, in the sense of "slave", are to be found in German sources. However, the use of the term in the thirteenth-century Adriatic example cited here is particularly unambiguous, as it is used in statute law defining slave status.

82 Charles Verlinden, "La Crete, debouche et plaque tournante de la traite des esclaves au XIVe et XVe siecles", in Studi in onore di Amintore Fanfani, 6 vols. (Milan, 1962), iii, pp. 591-669.

Susan Mosher Stuard Haverford College
COPYRIGHT 1995 Oxford University Press
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1995 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Stuard, Susan Mosher
Publication:Past & Present
Date:Nov 1, 1995
Previous Article:Historians and nation-building in Germany after reunification.
Next Article:Blasphemy and the play of anger in sixteenth-century Spain.

Terms of use | Privacy policy | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters