Slaves (nubi 奴婢) in Daoxuan’s Vinaya writings

ABSTRACT Daoxuan’s view on monastic slavery is based mainly on the principle of the Vinaya, but it seems that he particularly opposes individual monastics to possess slaves and supports the releasing of monastic slaves and the ordination of slaves of a certain kind.

in China. Thus it may be said that he contributed a great deal to the foundation of later Chinese Buddhist monasticism.
As has been indicated by the fact that later masters of this Vinaya school frequently quoted and/or referred to his views as authority, Daoxuan's Vinaya books are very influential on the later generations of the school. Among them, two are specifically consulted for this study as they contain information on Buddhist monastic slaves. One is the Sifen lü shanfan buque xingshi chao 四分律刪繁補闕行事鈔 (Guiding notes for practices: simplification of and supplements to the Dharmaguptaka-vinaya. T. No. 1804. Hereafter SLSBXC), the other the Liangchu qingzhong yi 量處輕重儀 (Standards for calculating and dealing with the light and heavy [possessions]. T. No. 1895. Hereafter LQY). 4 The first is an enormous compendium of Buddhist monastic disciplinary code based on the Chinese translation of the Dharmaguptaka-vinaya with supplements of rich quotations from many other Vinayas and Vinaya-natured texts, and scriptures, and it has served as an authoritative guide to the Chinese Vinaya school. The second is a treatise starting off with a short sentence from the foresaid translation, aiming, as its title suggests, to set standards for how to deal with the belongings left behind by deceased monastics and with other related issues. 5 According to his own prologue to this piece, it is a reconstructed and enriched work of the relevant materials presented in the compendium. 6 Much useful information on slaves from these two works is scattered in the discussions on topics such as the eligibility for ordination, precepts of monastics, and the belongings of dead monastics, and it has not be made used of so far. With this information this paper presents how Daoxuan viewed monastic slavery so as to provide some fragmentary, normative information about slavery in early medieval Chinese Buddhism. It starts with a few words on terminology.

Terminology
The Oxford English Dictionary has four definitions for the word 'slave'. The first and most relevant is: 7 '(Especially in the past) a person who is the legal property of another and is forced to obey them.' This is the basic meaning of the word understood and used in this study. As for more precise and academic definitions, let us borrow from Orlando Patterson's work, Slavery and Social Death: A Comparative Study, in which both 'slave' and 'slavery' are clearly defined. According to him, 'slavery' is 'the first and foremost of "relation of domination"' 8 in which 'a person is dominated and bonded in three respects': under full control of the master, not belonging to the community in which he was inserted or born, and without honour. 9 This will be the understanding of the word 'slavery' in this study, although not all these features are applicable to the Chinese concept and practice of slavery in the time we are concerned with.
Traditional Chinese texts show that in pre-modern China there were quite a few terms for the concept of slave, and as time went by, a few other terms appeared to refer to slaves, some of them are continuously used in the texts of later dynasties. 10 Even in Daoxuan's days, outside of the legal texts (and some official documents), slaves were also referred to by many other terms. 11 But besides using tongpu 僮僕 to refer to 'slaves' and 'servants', respectively (see below), Daoxuan frequently uses three terms. 12 The most standard and frequent one is nubi (奴婢) which is a combination of nu and bi. The etymological meaning of nu is still a matter of debate, being either a slave or woman slave. 13 A social and cultural definition is given in an allegedly pre-Qin (i.e. prior to the third century BCE) text which states that a nu is a slave originally made from a male criminal and that a bi is one from a female criminal (其奴,男子入於罪隸, 女子 入於舂槁。凡有爵者，與七十者，與未齔者，皆不為奴). 14 This meaning was incorporated into the Eastern Han dynasty  dictionary, the Shuowen jiezi 説文解字 (Explanations of graphs and analysis of characters). 15 Therefore the dissyllabic nubi can mean a slave of either gender. Presumably, losing some freedom and honour of oneself and being forced to work for others with very limited material returns is the punishment for the crime committed. By Daoxuan's time, nubi had become a legal term for slaves, and socially and legally nubi belonged to the lowest class of society and had the least legal rights among all walks of life. 16 The second term Daoxuan used is shengkou (生口), which initially meant 'captives in war' and then also meant 'slaves' in the Eastern Han dynasty, 17 but commonly used in the Buddhist translations only since the fifth century. 18 Daoxuan's third term merits our close attention. It is shou sengqielan ren 守僧伽藍人 (monastery housekeeper), an umbrella term that covers varied forms of monastic lay labour, including workers of servitude and bondage. This is introduced in Daoxuan's commentary on the short sentence of the Sifen lü: '[In the possessions left behind by the deceased monastic] there are many monastery housekeepers' 多有守僧伽藍人. 19 So shou sengqielan ren was originally from the Vinaya translation and not Daoxuan's coinage. As has been noted by Jonathan Silk, this term is used to render the Sanskrit word ārāmika which was translated into Chinese in various ways in the Chinese translations of other Vinayas. 20 Daoxuan's expounding of this term seems to have been based on the reality of the Chinese Buddhist monasteries, the types of the workforce may also refer to those available there rather than to the Indian case. It covers three types of lay labour force in the monasteries in his days: shili gongji (施力供給) 'donated and supplied labours', buqu kenü (部曲客女) 'attached labours and female servants', and nubi jianli 奴婢賤隸 'slaves and the lowly menial workers'. The meaning of the first type is obvious and contains crucial information on how a Buddhist monastery obtains some of its lay workforce and will be dealt with later. The other two deserve some brief explanations. Buqu kenü (部曲客女) consist of two groups, i.e. buqu and kenü, both are low-class people. 21 Some have considered them slaves, 22 but it is simplistically incorrect. For in the legal codes and official documents of Daoxuan's time, buqu are normally, not exclusively, skilled musical or ritual professionals and other workmen attached to government offices or rich families, whereas kenü includes buqu's wives and children and released female slaves who still worked for their previous master. 23 They are low-class people but legally and socially not slaves, and most importantly, according to the law, unlike slaves they cannot be sold. But, as Daoxuan himself differentiates them from the first type, they are still bonded one way or another. 24 That is to say that in some respects they are not complete free people.
As for the phrase nubi jianli 奴婢賤隸, it could be either a combination of nubi and jianli, 'lowly menial worker', or the real conceptual term is nubi, whereas jianli 25 just functions as a descriptive phrase for the former's social status and the nature of their works, although if it is used independently it could also stand for the former, just like many other terms.
Occasionally, shou sengqielan ren is also related to jingren (淨人 'pure person') 26 which is another term important to this study as it will become clear that jingren is also relevant to slaves. Monastics are prohibited by Vinaya rules from engaging in many activities or accepting stuff, both those activities and stuff are considered as bujing 不净 'impure' for their occupation but could be needed in real life, thus they need lay people to 'purify' the activities or stuff so that they could engage in or possess them without breaking the prohibitions. The lay people employed to purify the activities and things are called jingren. 27 While in the Theravada tradition they are called kalpikāraka ('legitimizer') 28 which is sometimes believed to be a synonym of ārāmika, they are among shou senqielan ren in Chinese Buddhism.
With this brief information on the key terms, we proceed to find out whether Buddhist monastics can keep slaves or not.

Keeping slaves by monastics
In early history of Indian Buddhism the Buddha and elderly monastics did take junior monastics as personal attendants or servants, but the Vinaya rules prohibited monastics from ordaining people with the intention of making them servants, 29 and certainly from keeping slaves. The idea that Buddhist monastics should not keep slaves appears in a few places in Daoxuan's works and more intensively in his SLSBXC. 30 It is explicitly mentioned in the commentary on the rule against 'possess[ing] money and gems', the eighteenth rule in the category of naisargika-prāyacittika (confession with forfeiture) in the Dharmaguptaka-vinaya. He quotes from the Duolun 多論 31 stressing that for 'three benefits monastics are not allowed to possess money and gems: to cease slanders [of Buddhism and the monastics], to extinguish quarrelling, and to accomplish the seed of saints [by] practising frugality' 多論云：佛制此戒有三益：一為息誹謗故，二為滅鬪 諍故，三為成聖種，節儉行故. 32 What is more, the possessing of money and gems, keeping slaves and servants (tongpu 童仆), 33 possessing lands, farming, storing silk clothes and grains, keeping animals, using wool blanket and gold vessels, and using a bed decorated with gold and silver and other 'heavy' stuff make up the 'eight impure properties' 八不淨財. 'They all give rise to craving and destroy the practice' 皆長貪壞 道, he emphasizes. 34 Again in another place, he quotes from the Daji jing 大集經 (Mahāsaṃnipata-sūtra) saying that for monastics, accepting the eight properties is a sign of breaking the rules, which corresponds with the first reference just dealt with. 35 Some of the properties reappear when he refers to the Niepan jing 涅槃經 (Nirvanasūtra), saying 'as for Śrāvakas, they don't have savings and collections such as slaves and servants…' 聲聞僧者無有積聚，所謂奴婢僕使…. 36 Later in the same text he also states a straightforward ban on the possessing of these eight properties with the rationale that they are the obstacles which affect religious practice most. 37 He stresses in both the SLSBXC and the LQY that it is because slaves and other things would disturb one's faith and practice that the monastics should neither receive nor keep them. 38 Then while further explaining each of these 'eight impure properties' and in an effort to convey the idea of 'not keep[ing] and hold [ing] servants/slaves' he provides more references and quotations. First, he briefly relates a story from the Zengyi ahan jing 增壹阿含經 (Ekottarāgama) about the Buddha refusing a householder who offers him his daughter and likening craving to a Rākṣasī 羅剎女 (female ogre), indicating that accepting women for a (male) monastic brings about craving which in turn leads to pitfalls as dangerous as being eaten by a Rākṣasī. This could be understood as either that a male monastic cannot accept a woman as a donation or that any monastic cannot accept any person. His next reference confirms that the first assumption is the case. He paraphrases a long passage from the Mahasaṅghikavinaya which prevents monastics from accepting the offering of slaves or servants and female garden keepers with the exception that if the keeper was a jingren ('servant') for the reason that a jingren is there to manage stuff for monastics. 39 The jingren, of course, must be the same sex as the offering receiver. The same source is paraphrased in a short form in his LQY but with the remark that 'after receiving them [one should] pass them onto the monastery' 施僧奴婢及諸畜生，一 切別人，不得自受。為料理僧故，受已付僧, 40 which indicates that individual monastics should not hold jingren. In another place, Daoxuan quotes the instruction of a senior Vinaya master named Lingyu 靈裕 (517-605) which says that monasteries should not keep female jingren as that would destroy one's religious practice. 41 He even quotes from another scripture that proclaims that 'there are ten kinds of offering which do not generate merit, and the first of which is to donate women' 十 種施無福。一謂施女人. 42 Nevertheless, the permission to accept jingren applies to female monastics too, in which case the gender of jingren ought to be female. 43 Further and important information on the accepting of servants/slaves can be found in his references to the story of Pilindavaccha 畢陵伽 in the Pinimu jing 毘尼母經 (Vinaya-mātṛkā ?) and the Mahasaṅghika-vinaya. 44 This story has been studied by Schopen and Yamagiva in different contexts. 45 Here we present Daoxuan's paraphrased version from the Mahasanghika-vinaya. 46 One day Pilindavaccha was making his own hut in a village, and the king happened to be passing and saw him. The latter accordingly proposed to offer the monk some labour but he turned it down three times and accepted it at the fourth on the condition that those labourers agreed to observe the five precepts and the fast for life. 47 Daoxuan's passage specifies those people as 'shiren' 使人 (workman) while the original source has it as 'yuanmin' 園民 (garden keeper) which is another translation of ārāmika. This means that the people offered are not slaves, which is supported by the Chinese translation of other four Vinayas. 48 In any case, this, as Daoxuan tries to show, marks the start of individual monastics accepting lay servants or workmen. Then he quotes from the Niepan jing to support the acceptance of slaves under some circumstances: Seeing his disciples having people to offer what they need and lacking nothing, the Buddha then did not allow them to accept the offering of the eight impure properties. But if his disciples have nobody to support for a living, especially at the time of famine when food and drink are hard to come by, for the sustainment of the wonderful dharma, 'I allow my disciples to accept and store animals, slaves, gold, silver, carts, lands, houses, and grains, and to buy and sell what they need…' 涅槃 云：若有人言，如來憐愍一切眾生，善知時宜，說輕為重說重為輕。觀知我等弟 子，有人供給，所須無乏，如是之人，佛則不聽受畜一切八不淨物。若諸弟子，無 人供須，時世饑饉飲食難得，為欲護持建立正法，我聽弟子，受畜奴婢、金銀、車 乘、田宅、穀米，賣易所須。雖聽受畜如是等物，要須淨施，篤信檀越。如是四法 所應依止。我為肉眼諸眾生說是四依，終不為慧眼者說。 49 This seems to say that if a monastic is in desperate situations the ban on the use and acceptance of the 'eight impure properties' can be lifted. By the term 'disciples', we may well speculate that they could mean individual monastics as well as the whole Sangha. In either case, it is the beginning of monastics and monasteries accepting slaves.
In fact, Daoxuan's works offer abundant evidence showing monasteries taking servants/slaves. He refers to the Mahāsaṅghika-vinaya saying that it is alright to accept the offering of persons if the offer was made to the whole Sangha, 50 which is supported by the Mūlasarvāstivāda-vinaya, 51 and adds that in the Jetavana (Monastery) there were five hundred housekeepers, and that there were the same number of helpers in a monastery in Rājagṛha. 52 But not any monastery is eligible to receive or use servants. He writes that the Rizang fen (日藏分) proscribes a monastery of less than five members from receiving the offering of slaves and other property, 53 and that the Shanjian 善見 instructs that as long as the community has five members, even the offer of people can be accepted. 54 These two pieces of evidence differ in terms of the social status of the people offered; only one states that they were slaves. But Daoxuan's other piece of information strongly suggests that monasteries held slaves. In passing, he rephrases the Shanjian saying that bi 婢 (i.e. female slaves) in the monastery should not be addressed as bi, but they should be called 'sister' 大姊 (lit. 'senior sister'). 55 In addition, a further two references state even more directly that because slaves and other special objects are in principle not permitted for the individual monastics to possess and belonged only to the whole community of the four quarters (i.e. the Sangha), that stealing them, selling them, lending them to others, and using them for private purposes are classified as serious wrongdoings. 56 Finally, he refers to two passages of the Foshuo Mulian wen jielüzhong wubai qingzhongshi 佛說目連問戒律中五百輕重事 to show that both individual monastics and monasteries can accept and keep slaves. 57 So far it is clear that the way that individual monastics and monasteries obtained slaves is through donation. In fact, by Daoxuan's time, apart from the Vinaya texts, there had actually been quite a few translations encouraging people to donate slaves to the monastery, 58 as a Buddhist cataloguist, Daoxuan was bound to have read them, or at least some of them. In effect, Chinese Buddhist monasteries holding slaves seems to have been an existing fact in Daoxuan's time for in the context in which he was talking about how to handle the dead monastics' belongings, slaves are clearly mentioned among other items (see above). 59 Furthermore, besides the evidence from Daoxuan's own work (see below) the unearthed Dunhuang and Turfan manuscripts reveal that for centuries before his time Chinese Buddhist monasteries had already been using and holding slaves. 60 Some monastics were even found occasionally engaging in trading slaves. 61 This activity is to be dealt with in paragraphs that follow.

Monastics are not allowed to trade slaves
Clearly, as is reflected in his works, Daoxuan knows that in the Vinaya fully ordained monastics are in general forbidden to trade; 62 he even quotes from the Wubai wen 五百 問 ('Five hundred questions') saying that if a monastic is sent by his teacher to undertake any trade he should leave his teacher. 63 The same text also provides him with a prohibition that prevents individual monastics from renting out stuff and lay labourers belonging to the Sangha for personal gain. 64 Certainly, he seems to be fully aware of the exceptions to the rule against trading. He writes that trading is allowed if it is carried out in kind and done without being as commercial-minded as in the secular context and only among the five groups of Buddhists [i.e. male and female monks, male and female trainee monks, and female monks on probation]. 65 The actual dealing of all other trading for practical daily use needed in the monastery should be handled by jingren. 66 Another exception noted by Daoxuan in the Sarvāstivāda-vinaya is that the Buddha allows his disciples to sell one house to finance the refurbishing of another. 67 Thus the monastics can sell stuff in some circumstances.
However, in any case, for monastics trading in people is completely out of question under any normal circumstances. 68 By a random search in the electronic edition of the Chinese Buddhist canon, there appear numerous Indian Buddhist texts translated before Daoxuan's time and containing the message prohibiting monastics from trading slaves or people, 69 but Daoxuan seems to have ignored most of them. Instead, he only briefly refers to a couple of texts as supporting literature. He first rephrases a few lines of the Niepan jing saying that in the practices of the bhikṣu (monk) selling and buying people and animals are banned. 70 He then quotes from the Shanjian specifying that trading people is not allowed even for the welfare of the Triple Gem. 71 Finally, he adds that a monastic especially cannot sell slaves possessed by the monastery. 72 What is interesting is the reason for the prohibition. He says that the trading bonds the person's freedom and will eventually make the trading monastics become like lay people and nothing of religious nature. 73 In other words, trading slaves is no good for both monastics and the people traded.
Despite the injunction, however, there are a few scriptures which predicate that future generations of monastics would engaged in trading slaves but of which Daoxuan did not make use. 74 This predication seems to have turned into (or rather been based on) reality since there were some monastics engaging in buying or selling slaves. In two places at least Daoxuan lists the buying and selling of slaves as one of the appalling activities in which some monasteries were engaged in his days. In one passage he complains, 'nowadays, many monasteries keep women or sell and buy slaves. Who knows [how much] corruption there is in such [activities]?' 今諸伽藍多畜女人，或賣 買奴婢者，其中穢雜，孰可言哉 75 In another passage, he criticizes 'monasteries which set up different practices that often do not comply with the Buddha's teaching [such as] engaging in match-making for the jingren, buying and selling slaves and other properties.' 寺別立制，多不依教…… 媒嫁淨人，賣買奴婢及餘畜產… 76 Daoxuan's condemnation of monastics trading slaves and the lack of a straightforward rule against monasteries receiving and keeping lay servants and slaves no doubt confirm what is discussed in the previous section -that is that there were lay servants and slaves in Buddhist monasteries. But did the monasteries release them at some point? This is to be answered in the next section.

Manumission of monastic slaves
Despite the fact that Daoxuan appears to believe that karma is at least one of the factors which causes a living being to be reborn as a slave or in the status of being a slave as he quotes from the Baoliang jing 寶梁經, 77 he does seem to believe that fate can be changed through human intervention since he supports the releasing of slaves. According to his works, the releasing happens on two occasions: being released or being ordained as a monastic. Since ordaining slaves will be treated in a separate section below, here we only discuss his treatment of the manumission of slaves.
The releasing of monastic slaves mainly happened when their owner died, and the slaves manumitted all belonged to individual monastic masters. In the passage of his LQY referred above we find more detailed treatment about lay manpower left behind by the dead individual monastics. 78 There he informs us that in the Vinaya texts the monastery lay labourers are classified into the category of heavy possessions 79 because they (including animals, servants/slaves) can only be collectively owned by the monastery. As has been highlighted earlier, Daoxuan divides the lay labour of a monastery into three types of which only the last obviously consists of slaves. Although Daoxuan also suggests that those buqü (belonging to the second type of the lay workforce), who work for the master on a written contract, should also be treated like slaves, i.e. to invalidate the contract and let them go when their master dies. 80 As for the slaves, he suggests manumission. He writes: 81 The third [type] is slaves and lowly menial workers. [They and] all their offspring should be taken by the permanent Sangha. If they die and leave no relatives, the permanent Sangha take them. 82 The first case should be solved according the [pini] mu lun which instructs if individual monastics have slaves they should manumit them and let them go, if not, they should be made monastery jingren. 三謂奴婢賤隸所有子息資生，並入常住。

若身死無親者，常住收之。已前一條，判如母論云：若私有奴婢應放令去(如前條中) 。若不放者，作僧祇淨人。
This instruction turns slaves owned by individual monastics into communal servants. One may ask why the word '放' is taken to mean 'manumission' rather than 'releasing'. The answer lies in his reference to the same piece of information of the Pinimu lun in another of his works in which, after repeating the same instruction, he adds the following remark in brackets: 83 'releasing slaves means upgrading them to a free men status and give him/her a family name. After that [the matter] should be dealt with according to the Vinaya. 準此放去，謂賜姓入良，後終依律'. 84 Strictly speaking, the death of slaves is not a proper occasion of manumission, but it could be one for their children. Besides, since it concerns a slave's life in the monastery, it may be useful to explore this a bit here. Daoxuan quotes from the Vinayas talking about the management of the property belonging to a dead slave. The matters are solved according to the circumstances. First, when a monastery-owned slave dies, his clothing should be handed over to his relatives. In cases where there is no relative, they are given to the resident monastics. Second, if an individual-owned slave dies, after proper counting of his or her belongings in the presence of a witness, there could be two solutions according to their living arrangement. One, if they (i.e. the monastic and his/her slave) lived together, the owner should have the right to take what he likes from the belongings of the dead; if they do not live together and the owner provides food and clothing, all the slave's property ought to be given to his/her relatives. If he does not have relative, they should be considered as ownerless and taken by the monastery. 85 To the latter case, he adds that it is just like the case in which a monk is expelled by the monastery. What is important is the offspring of the slaves. They should be set free or transferred to the monastery to be jingren, as instructed by the Pinimu jing, referred to a couple of times before. 86 It is clear that the slaves to be manumitted or transferred to the Sangha to be jingren are only individually owned; slaves belonging to the monastery or owned by the Buddha are not mentioned for release. Perhaps it is because there were monastery slaves who could not be released, the government from time to time issued edicts to order the monasteries to manumit slaves. 87 Yet, the real issue from Daoxuan's paraphrased passages seems to be the options between manumission and transference. It seems that only one of these two options can be chosen: either manumitting the slaves and letting them go or transferring them to the monastery as jingren. In other words, it is unlikely that the slaves made jingren are manumitted, although the monastery could certainly do so, which means there are jingren who are still slaves. Besides, we have seen at the beginning of this study that in light of terminology, shou sengqie lan ren includes jingren and slaves. We also know that some jingren were transferred from slaves who might not be manumitted. Thus there is an overlapping of roles, i.e. jingren includes slaves. Hence the next section looks at the role of jingren in the monastic lay labour system, and thereby the role of some slaves will become clearer.

Ordination of slaves
In the paragraph discussing the releasing of slaves, Daoxuan quotes from the Chujia gongde jing 出家功德經 (Sutra on the merit of going forth from home to homeless), with more information such as analogies of the merit generated by releasing slaves to become a monastic being greater than that gained through offering for a hundred years to as many Arahants as fill up all the universe. 88 This means that slaves can be ordained as monastic, which is also a way of manumission.
On the point of admitting slaves to the Sangha, Daoxuan's comments and remarks are mainly contained in his SLSBXC, which basically means that his position accords with the rule that in principle prevents slaves from being ordained. 89 This rule is shared by all the Vinayas. Namely, the attempt of a slave to become a monastic is in vain at the ordination ceremony, 90 because as part of the ritual process for the ceremony, every candidate is asked to answer thirteen major and ten minor questions for the purpose of eliminating the unqualified. The question, 'You are not a slave, are you? 汝非奴不' is asked in both questionnaires. 91 Daoxuan seems to have accepted this rule. In the early part of his SLSBXC while discussing the importance of observing the Vinaya rules for the sustenance of the Buddha's teaching he quotes from the Moye jing 摩耶經 the examples of breaking the Vinaya rules in the future, which includes the phenomenon that slaves will become monastics. 92 In other words, slaves becoming monastics is a sign of the decline of the Dharma.
His clear position on this rule can be seen in another passage worthy of being translated in full here: 93 The Mahāsaṅghika-vinaya states, 'For those who were born, bought or forcibly taken as slaves, were not allowed to obtain ordination in this [ There are a few points worth noting in this passage. First, on the one hand it is a monastic rule that slaves are generally not allowed to enter the Sangha, on the other hand some slaves can actually be exempted from the effect of this rule, that is to say that the donated and self-made slaves are exceptions. In fact, it is because of the exception to these two types of slaves that we may better understand what the underlying reason is for the general rule. To be exact, slaves cannot be ordained because they belong to others, while those donated and self-made ones can be ordained because their ownership is with the monastics, be that individual or corporative. And as is clear above, monastic-owned slaves are recommended for release. This line of reasoning can be justified by Daoxuan's equalizing of slaves' ordination with that of children. He writes the principle of deciding whether or not a slave can enter the Sangha is the same as that used in ordaining children, which is discussed before this case in his work. A quick crosscheck shows that his sources for the exceptions of both cases are the Mahāsaṅghika-vinaya in which both cases are even described with the same wording, i.e. that children without parents' permission cannot be ordained, but run-away and monastic-adopted ones can. 98 It may not be a pure coincidence that the wording describing the exception for ordaining children is the same as in the case of slaves because they share one thing and that is custodianship or ownership. In other words, both slaves and children are subjects of some caretaker, as it were, and need permission to be taken away, otherwise the act will be qualified as 'taking without permission' which is the definition of 'stealing', an act prevented by the third cardinal monastic rule. In fact the importance of the ownership of slaves may explain why Daoxuan specifically stresses 'stealing slaves commits a serious wrongdoing' 盜奴犯重 when commenting on the cardinal rule of 'not stealing'. 99 For a justifying illustration to this interpretation of custodianship and ownership, we may refer to another set of questions and answers in the Wubai wen. It is asked whether or not a monk who is held and put up for sale should run away. The answer is that 'initially he can, but after the change of ownership he can not' 初時 得, 經主不得. 100 This shows a certain Buddhist understanding of the ownership of property at the time, which may be at work in the case under discussion. Accordingly, the principle of custody/ownership may well be the reason why slaves given to the Sangha by others or self-made are allowed to be ordained: the monastery or the individual monastic has the proprietorship of the slaves and therefore they have the right to decide whether the slaves can or cannot be released. 101 Second, Daoxuan seems somewhat in favour of ordaining slaves. He appears to be fully aware of the spirit of equality in Buddhism. 102 In one early page of his SLSBXC, he states 'in Buddhism there is no [discrimination regarding] the noble and the humble, the closely related and the unrelated, there is only the dharma of equality…' 佛法中無貴賤親疎，唯以 有法平等應同護之. 103 In the last part of the same work he seems to become less stringent on the slave-unfriendly rule. For we read in the section on sramaṇera (novice monks) a list of many sutras which exhort the marvelous karmic benefits of becoming a monastic and that those texts sending the message that stops people from becoming monastics generate tremendous suffering. 104 Indeed, he supports the common rule of all the Vinayas in objecting to slaves joining the Sangha, but he also repeatedly refers to and paraphrases the exception from the only Vinaya that supports the ordination of the given and self-made slaves. In addition, he also refers to the advice of the scriptures that encourages people to join the Sangha and exhorts the merit of helping people to be ordained, especially the Chujia gongde jing. While rephrasing a passage of this sutra, he changed the original 'person'(人) into 'slaves (奴) and other men and women', the slaves were singled out without excluding other people. Thus the original '放人', 'releasing people', in the sutra becomes releasing slaves and other people. Apart from his loyalty to obey the authority of the disciplinary code there could be more than one reason why he accepts the common rule that prevents the ordination of slaves. The first could be that he places the priority on the third cardinal rule, i.e. the rule of not stealing. The second reason could be that he supports the law of the country as well. In the last part of his SLSBXC, we read the initial screening criteria for the ordination once again quoted from the Wubai wen, that 'secretly ordaining those without the consent of their parents or the permission of the king's law is a serious breach of the vinaya rules', and his remark which says 'because that could teach and show others to disobey and abandon [their] 105 This remark may suggest that Daoxuan strictly upholds the principle that Vinaya rules should be in accordance with governmental law 106 and local customs, 107 especially in view of the fact that one imperial court before him had ordered that slaves not be ordained. 108 Notwithstanding this, in Daoxuan's time, there were monastics who were originally slaves; at least two slaves who not only became monastics but whose lives were also included in a book of hagiographies of eminent monks several decades before his time. 109 Third, in the Wubai wen only 'Buddha slaves' are referred to, whereas Daoxuan adds sengnu 僧奴, 'monastery slaves', extending the ban from Buddha slaves to monastery slaves and accidentally indicating that slaves in the Buddhist monasteries were differentiated by their symbolic or actual ownerships. This suggests that the monastic slaves of his time included at least those belonging to the Buddha, the monastery and individual monastics. Unfortunately, his writings do not offer any explanation about what 'Buddha slaves' were. Despite there being people by the names of 'So-and-So Buddha Slave' or 'So-and-So Monastery Slave' found in surviving manuscripts, 110 it is equally impossible to say anything about the difference between these two types of slaves in terms of their work and treatment in Chinese monasteries before or during Daoxuan's time. What is reasonably clear is the work of jingren, and it is to be discussed next.

Jingren
Like the use of nubi 'slaves', Daoxuan uses the term jingren equally frequently when referring to the lay workforce of the Buddhist monastery. But in none of his works does he define or annotate what jingren is, rather he borrows it from the translation of the Vinayas and uses it as a word deserving no annotation. From most cases of his use of the terms, jingren are distinct from slaves. And, as has been shown above, what makes jingren different from slaves is that individual monastics are never banned from accepting and keeping jingren. The only specification in the acceptance is on the gender. As has been seen from Daoxuan's reference to the Mahāsaṅghika-vinya twice, monks should not accept female jingren and female monastics cannot accept male jingren almost for the same reason given in the monks' case. 111 In the list of stuff and staff belonging to the monastery, he used nu and pu 僕 together, 112 which, refer to slaves and servants respectively. 'Servants' may refer to jingren, as we will see that jingren do run errands and do other trifle tasks. From all the information provided by Daoxuan, the works and duties charged upon jingren can be grouped into two major categories. The first type is performing the duties of the 'purifier' for the monastics by doing the things which the monastics are not allowed to do by the Vinaya rules. These duties include receiving money for buying medicine and robes on behalf of the monastery, 113 receiving gold or silver for the purposes of building or repairing the monastic facilities, 114 receiving the donation of lands and houses, 115 buying lands and doing other general trading, 116 carrying raw grains donated to the monastics, 117 digging dirt, 118 trying on new shoes given to the monastics, 119 cooking for the monastics, 120 picking fruit and cutting the overhanging branches of trees for the monastics, 121 handling over to monastics the food that is on other places or leftover, 122 eliminating grasses and small trees, 123 removing pests. 124 They are also there on behalf of monks to deal with women, such as passing things on to women, being witnesses for monks in situations in which misunderstanding or accusation could arise from the public, such as sitting with women. 125 One would assume that female jingren in the nunneries were expected to perform similar tasks.
The second type of jingren's work is serving tasks, which includes telling the monastics the passing of the time in the day and reminding them about the time for assembly for the prātimokṣa recital (although these two tasks can also be assigned to novice monastics), 126 serving the monastics meals and medicine, 127 building and repairing monastery facilities, 128 arranging altar offerings, 129 and administering invitations to meals for monastics. 130 Daoxuan's writings also provide information about how jingren should be treated in the monastery. He refers to the Dharmaguptaka-vinaya and Mahīśāsaka-vinaya saying that whatever donations the monastics receive, jingren should also have a portion. 131 This could only be meant for those permanent jingren. What is more, he modified some passages of other Vinayas in relation to the payment of jingren. He refers to the Shanjian and says that if the work is rotated by (two) jingren, the earlier worker is provided with food and clothing, the second comer has none. The long-term jingren is also provided with both food and clothing. Referring to the Sarvāstivāda-vinaya he advises paying the jingren according to the amount of work he has done, in the case he has to end the job due to some emergency, this is said to be in accordance with what was practised in India at the time. Namely, if he quits the work before lunchtime, he is provided with one meal but without pay. If he quits in the afternoon he is paid for a full day's work. His own idea is that the monastics have to judge the jingren's performance: whether he is lazy or hard-working; if the latter, and even if he has only wored for half a day, he should be given a full-day's pay. 132 Finally, monastics are also supposed to look after jingren when they are ill, and in that situation monastics are allowed to do many things they cannot do without jingren, especially cooking and preparing medicine. 133 In addition, Daoxuan also suggests that if the relatives of jingren are staying in the monastery and fall ill, monastics are supposed to prepare medicine for them. 134 But monastics are discouraged from being matchmakers for jingren. 135

Conclusion
Although nowhere in Daoxuan's works does he specially treat slaves or servants as a separate topic, his information has informed us about some aspects of the lay workforce of the Buddhist monasteries, among which are slaves. As has been shown above, there are different types of slaves: individually-owned, monastery-owned, and the Buddha slaves. Of these types, apart from mentioning the names of the last two in a passing way Daoxuan has not provided us with information substantial enough to work out anything concrete. So in the materials he presents, the focus is almost exclusively on the slaves owned by individual monastics. Aside from sources about these slaves, there is also some information on jingren, a type of monastic paid servant, which Daoxuan's writings seem to deliberately differentiate from slaves although his own evidence indicates some of which may still keep the status of slaves. All in all, from what we have presented above we can gain some understanding of some aspects of Buddhist monastic slaves in China of his time as well as in the Indian Vinaya texts he made use of. This understanding can be summarized into the following few points.
Originally, individual monastics were not allowed to use and accept and keep lay servants and slaves. But in some situations the prohibition was lifted: when monastics had no one to support for a living, and that if the lay servants or slaves offered were willing to observe the five Buddhist precepts and the eight-fold fast for life. Moreover, there was no clear rule preventing a monastery from accepting slaves or lay servants although some sources require that for a monastery to accept servants or slaves it must have a minimum five resident monastics. Eventually both individual monastics and monastery are found using and holding slaves.
As for the sources of slaves/servants, they were all initially offerings of Buddhist patrons, although there were other means through which slaves were obtained, such as trading. But Daoxuan has clearly presented that according to the Vinaya buying and selling slaves is prohibited because trading harms both the salves and the religious faith and practice of the monastics. This prohibition of the Vinaya did not stop the Chinese Buddhist monastics from doing that, as can be seen from Daoxuan's criticism of some monasteries which engaged in such activities. Although probably due to the fact that well before Daoxuan's time, Chinese Buddhist monastics and monasteries held slaves, Daoxuan did not show disapproval of the monastery receiving or holding slaves, but he does object to monastics engaging in trading slaves.
All slaves (including their offspring) of the dead monastics had two chances to become a free person: the first is manumission, the second being ordained as a monastic. This means that monastic slaves and their children were also likely to be manumitted by releasing or becoming a member of the Sangha. Daoxuan constantly refers to the Pinimu jing to argue that slaves should be manumitted and let go or transferred to the monastery as jingren.
As for another chance of manumission, that was to be ordained as a member of the Sangha, Daoxuan has shown us from the Vinaya point of view that only two types of slaves were eligible for these chances. They are slaves offered by others and the selfmade slaves because only their ownership is with the monastics or the monastery. He indicates that ownership/custodianship is the key factor to consider while ordaining slaves, just as admitting children to the Sangha. It is because the ownership matters most that individual monastics were prohibited from secretly ordaining slaves belonging to the monastery.
Finally, we have some information on jingren, some of who were originally from slaves. It is in this context that we have some descriptions of the work of slaves and of how they are supposed to be treated in the monasteries. They basically help the monastics doing things that are normally prohibited by the rules as well as serving them. In return they also get paid, and care when in need.

Disclosure statement
No potential conflict of interest was reported by the author.