The Term Mamluk and Slave Status during the Mamluk Sultanate El término mamluk y la condición de esclavo durante el sultanato mameluco

Scholars of the Mamluk Sultanate generally maintain that the status of all the mamlūk was that of an elite, and that the mamlūk were proud of their slave origin even after manumission. It is here argued that these assertions are based on a misconception of the term mamlūk as used in Mamluk sources. The term mamlūk has a double meaning: slave and servant, and it frequently expresses subordination, obedience and servitude. It is never used to express pride in slave status or slave origin. There is no evidence that manumitted mamlūk were proud of their slave status. On the contrary, manumitted slaves with aspirations made great efforts to repress their servile past by claiming an exalted origin or by creating marital ties with established families. Mamlūks were considered property and they lacked a legal identity of their own. They were often manumitted only upon their master’s death. They perceived themselves as slaves for lacking family ties. Only an outstanding few succeeded in completely freeing themselves of their slave status and become members of a ruling elite with family ties. It seems that starting from al-Nāṣir Muḥammad b. Qalāwūn’s third reign the enslavement of Turkish mamlūks who had been sold by their families became more of a formality. On the other hand, non-Turkish mamlūks , who were generally Christian war captives, were subject to discrimination. They were disdained, manumitted at a later age and prevented from establishing marital ties with the Qalawunids and creating their own families at a young age. They were perceived by their contemporaries as being “more slaves” than the Turkish mamlūks .

Scholars of the Mamluk Sultanate generally maintain that the status of all the mamluks was that of an elite, and that the mamluks were proud of their slave origin even after manumission.It is here argued that these assertions are based on a misconception of the term mamluk as used in Mamluk sources.The term mamluk has a double meaning: slave and servant, and it frequently expresses subordination, obedience and servitude.It is never used to express pride in slave status or slave origin.There is no evidence that manumitted mamluks were proud of their slave status.On the contrary, manumitted slaves with aspirations made great efforts to repress their servile past by claiming an exalted origin or by creating marital ties with established families.Mamluks were considered property and they lacked a legal identity of their own.They were often manumitted only upon their master's death.They perceived themselves as slaves for lacking family ties.Only an outstanding few succeeded in completely freeing themselves of their slave status and become members of a ruling elite with family ties.It seems that starting from al-Nasir Muhammad b.Qalawun's third reign the enslavement of Turkish mamluks who had been sold by their families became more of a formality.On the other hand, non-Turkish mamluks, who were generally Christian war captives, were subject to discrimination.They were disdained, manumitted at a later age and prevented from establi- Al-Qantara XXXIV 1, 2013, pp.7-34 ISSN 0211-3589 doi: 10.3989/alqantara.2013.001Mamluk authors almost always refer to the political regime that ruled Egypt, Syria and adjacent areas for two-and-a-half centuries (648/1250-923/1517) as "the state of the Turks" (dawlat alatrak/dawlat al-turk/al-dawlah al-turkiyah).They seem to be aware of the fact that the reign of the "Turks" is divided into two periods -the rule of the Turks (dawlat al-atrak) and that of the Circassians (dawlat al-jarakisah), and they clearly emphasize the ethnic origin or language of the ruling elite.Only rarely, and only in the Circassian period of the Sultanate (784/1382-923/1517), do they explicitly refer to the Sultanate as being ruled by slaves.Despite this fact, modern scholars almost without exception use the term "the Mamluk Sultanate" (dawlat al-mamalik), that is, an appellation that emphasizes the elite's and rulers' slave status or slave origin. 1 This appellation distorts Mamluk writers' perceptions of their ruling elite and its defining characteristic, and reflects a view propagated by David Ayalon and still held among modern scholars, that in that period the right to rule and hold key positions in the Sultanate was reserved almost exclusively for mamluks; that the ruling elite's main characteristic was mamluk descent; that the status of all the mamluks was that of an elite; and that the mamluks were proud of their slave origin even after manumission.2formalidad.Por otro lado, los mamluks que no eran turcos, generalmente cautivos de guerra cristianos, eran discriminados y despreciados; solo se les liberaba cuando eran ancianos y se les impedía establecer lazos matrimoniales con los Qalawuníes así como crear sus propias familias siendo jóvenes.Eran percibidos por sus contemporáneos como «más esclavos» que los turcos mamluks.Palabras clave: mamelucos; Sultanato mameluco; condición de esclavo; esclavitud militar; élites de esclavos.
shing marital ties with the Qalawunids and creating their own families at a young age.They were perceived by their contemporaries as being "more slaves" than the Turkish mamluks.
Although D. S. Richards maintained that "Mamluks entered no charmed circle, no special caste" and that it is "absurd to think that any mamluk, merely by virtue of that legal status, had a real expectation of power, wealth and influence",3 and Robert Irwin argued that most mamluks cannot be regarded an elite,4 generally, scholars of the Mamluk Sultanate are still of the above-mentioned opinion.For example, Linda Northrup maintains that "having been a slave was a condition for eligibility to the highest ranks of military society", and that even after manumission "the recruit, now free, retained his mamluk and, therefore, elite status".5In a similar manner, Reuven Amitai holds that "even after official manumission at around the age of twenty or younger, at the ceremony known as kharj, where the trainees received a certificate of release, the soldiers still proudly regarded themselves as mamluks, jealously guarding their status…", or that "officially free Mamluks still overtly referred to themselves as mamalik, proud of their special slave origins". 6It will be argued below that these assertions are based on a misconception of the term mamluk as used in Mamluk sources.Servile status was not considered a source of pride; on the contrary, it seems to have been considered degrading and manumitted slaves with aspirations made great efforts to repress the servile phase of their life.

The meaning of the term mamluk
We often come across declarations of manumitted slaves to the effect that they are "the mamluks of the sultan" (nahnu mamalik al-sultan).However, when checking the context in which these expressions appear, we find that they are meant to express obedience and subordination to the ruler and not slave status or pride in slave (mamluk) descent.Often, when a Mamluk amir rebels or is suspected of being disloyal, he expresses his (real or dissimulated) subordination to the ruler with the words "I am a mamluk of the sultan and obey him" (ana mamluk al-sultan wa-tahta ta,atihi). 7The connection between the ex-  Al-Qantara XXXIV 1, 2013, pp.7-34 ISSN 0211-3589 doi: 10.3989/alqantara.2013.001pression mamluk/mamalik al-sultan and obedience (ta,ah) is clear in many other cases. 8A Mamluk amir might also designate himself the mamluk of a fellow khushdash in order to express subordination and obedience to him.For example, after al-,Adil Kitbugha (d.702/1302) was deposed by al-Mansur Lajin (d.698/1299), he is quoted as saying of Lajin that "he is my comrade and I am his mamluk who obeys him" (huwa khushdashi wa-ana mamlukuhu wa-tahta amrihi). 9In other cases it is clear that mamluks expressing obedience and subordination with the expression "we are your mamluks who obey you" (nahnu mamalikuka wa-,alà ta,atika) do not address their master, but rather a patron whom they serve.10Moreover, even free persons might express obedience by using the expression "we are mamluks" (nahnu mamalik). 11Al-Maqrizi (d.845/1441) provides a very clear example of the link between the expression nahnu mamalik and total subordination.When the rebel Jukam min ,Awad (d.809/1406) claimed the title of sultan in the year 803/1400 he tried to reassure al-Nasir Faraj (d.815/1412) that he and his followers were not opposing him but the amir Nawruz al-Hafizi (d.817/1414), saying: "We are the mamluks of the sultan… had he wanted to kill us we would not have opposed his command" (nahnu mamalik al-sultan… wa-law arada qatlana ma khalafnahu). 12While such expressions are clearly hyperbole (or even plain lies), the phrase "we are mamluks" (nahnu mamalik) is almost always meant to express subordination and obedience.It often appears in the course of revolt or when an amir is suspected of planning one.It never expresses pride in mamluk status or origin.
Many times the term mamluk is used to convey the fact that a person is a ruler's or a patron's servant, and not his slave.In the same manner, the term ustadh is used to denote a patron and not a master. 13or example, the Khawarizmiyah, a group of free mercenaries, ad-dressed the Ayyubid sultan al-Salih Ayyub (d.647/1249) as his mamluks.Qaratay al-,Izzi al-Khazindari (d. after 708/1308) mentions that al-Salih Ayyub "enslaved them with money" (ista,badahum bi-lamwal), 14 and the context makes it clear that al-Salih Ayyub bought their services and that Qaratay al-,Izzi al-Khazindari is referring to patron-client ties based on exchange relationships: favors of the patron (ni,mah/ihsan) in return for service (khidmah). 15From the above-mentioned example it is clear that this usage of the term mamluk was not unique to the Mamluk period. 16The same Khawarizmiyah make it clear that the usage of the term mamluk is strongly related to service (khidmah) and obedience (ta,ah), when they say to al-Salih Ayyub: "We came to serve you and we are your obedient mamluks" (nahnu qad hadarna ilà khidmatika wa-nahnu mamalikuka wa-tahta ta,atika). 17They highlight the metaphorical meaning of the term mamluk in this case, denoting servitude and not slavery, by adding that they are "slaves of the Ayyubids" (,abid li-Bani Ayyub). 18In a similar manner, the free Turkmen amir, Qarajah bin Dhu l-Ghadir (d.754/1353), who aspired to the post of governor of al-Abulustayn in the year 738/1337, expressed his request by saying that he "wishes to be the 11 THE TERM MAMLUK AND SLAVE STATUS DURING THE MAMLUK SULTANATE Al-Qantara XXXIV 1, 2013, pp.7-34 ISSN 0211-3589 doi: 10.3989/alqantara.2013.001   Qaratay al-,Izzi al-Khazindari, Ta,rikh Majmu,p. 65. 15 We can find another example for the usage of the term mamluk related to service (khidmah) given in return to favors (ihsan) in the words of the Bahriyah to the ruler of Anatolia (sahib al-rum): "If you will be pleased of us and give us from your favors we will be your mamluks" (fa-in ahsanta ilayna wa-radita bina fa-nahnu mamalikuka), Qaratay al-,Izzi al-Khazindari, Ta,rikh Majmu, al-Nawadir, p. 145; and see also Ibn Qadi  Shuhbah, Ta,rikh Ibn Qadi Shuhbah, vol.4, p. 302; for the connection between the term mamluk, favors and total subordination, see Mufaddal b. Ab al-Fada 'il, al-Nahj al-Sadid,  p. 379; for the importance of patron-client relationships, ihsan and ni,mah in the Mamluk Sultanate, see Van Steenbergen, Order out of Chaos, pp.57-75; for their importance prior to the Mamluk period, see Mottahedeh,Loyalty and Leadership, For another example of a usage of the term mamluk expressing obedience and service concerning the Ayyubid period, see Qaratay al-,Izzi al-Khazindari, Ta,rikh Majmu, al-Nawadir, p. 78.In a similar manner, during the ,Abbasid period, servile terminology (mawla, ghulam, and more rarely ,abd) was used to express servitude and loyalty, see for example Golden, "Khazar Turkic Ghulams", pp.285-287 (especially page 286); Beckwith, "Aspects of the Early History", p. 38; Pipes, "Mawlas", p. 224; Crone, "Mawla", p. 881; and see also Golden, "The Terminology of Slavery".
sultan's servant (mamluk) in that land" (yas,alu an yakuna mamluk alsultan fi tilka al-ard). 19e also come across instances in which mamluks, after being manumitted, offer their service to persons other than their masters, in words similar to those of Qarajah bin Dhu al-Ghadir. 20In both the Turkish and the Circassian periods, it was common enough that sons of mamluks, Mongol immigrants (wafidiyah) or other free persons, were listed among the members of the sultan's mamluks (mamalik al-sultan) or that of an amir. 21We also find references to Christian bureaucrats who converted to Islam as servants (mamluks) of the sultan. 22Sometimes the sultan's khushdashiyah refer to themselves as his mamluks (mamalik al-sultan), and in these instances it is clear that the term mamluk denotes servitude rather than servile status. 23In addition, the term mamluk was frequently used as part of the protocol of the civilian and military elite, in order to express subordination and low rank.We have evidence that at times civilians found this term degrading and refused to use it. 24According to Nasser Rabbat the meaning of the term mamluk was transformed in the beginning of the Mamluk period from a warrior-slave who was subjugated all his life to his master, to one destined to be manumitted and promoted in the ranks of the military. 25It turns out that the term, even prior to the period of the Mamluk Sultanate, also simply denoted a servant.
By now it should be clear that the term mamluk denotes obedience, servitude and subordination at least as much as it denotes slave origin and slave status.It is never used to express pride in slave origin.A misinterpretation of the term mamalik al-sultan led scholars to argue that free immigrants (wafidiyah) were refused promotion to high ranking offices because they were not slaves.Al-Maqrizi has Baybars al-Jashankir (d.709/1310) address an immigrant amir (wafid) by the name of Sanjar al-Barawani (d.731/1330) as follows: You are a person who was expelled, an immigrant.How can you think that your status and that of mamalik al-sultan is equal?"(anta wahid manfi, wafidi, taj,alu nafsaka mithla mamalik al-sultan?). 26vid Ayalon saw in this text evidence that the wafidiyah were discriminated against for not being slaves; however, as already mentioned, the term mamalik al-sultan denotes servitude, subordination and obedience, and not only slave origin.Al-Maqrizi emphasizes that the immigrant is an expelled person, that is, one who abandoned his previous master, and so betrayed him.In this context, the meaning of the expression mamalik al-sultan is "the obedient servants of the sultan".
The wafidiyah were perceived as a treacherous, disloyal and untrustworthy element in the Mamluk society.They were accused of cons piring against Kipchak Sultans, of collaborating with the Mongol Ilkhans, or of trying to escape to the Ilkhan's territories. 27They were denied promotion because they had betrayed their previous masters, by immigrating to the Sultanate, and not because they were not slaves.In a similar manner, al-Maqrizi mentions that when al-Nasir Muhammad b.Qalawun (d.741/1341) decided to promote Oirat immigrants, who had served the amirs before abandoning them and becoming his servants, al-Nasir Muhammad's mamalik al-sultan made him change his mind after protesting: "They harshly criticized and condoned them for betraying their masters, and said that they are no good" (aktharu 13 THE TERM MAMLUK AND SLAVE STATUS DURING THE MAMLUK SULTANATE Al-Qantara XXXIV 1, 2013, pp.7-34 ISSN 0211-3589 doi: 10.3989/alqantara.2013.001 26 Al-Maqrizi, al-Suluk, vol.2, p. 22; and see also David Ayalon, "The Wafidiya in the Mamluk Kingdom", p. 93; Ayalon maintains that Sanjar al-Barawani was not an immigrant, but there is evidence that he might have been a wafid from Anatolia, al-,Ayni, ,Iqd al-Juman, vol.2, p. 166. 27See for example al-Dhahabi, al-Mukhtar min Ta,rikh Ibn al-Jazari, p. 305; Ibn  Kathir, al-Bidayah wa-l-Nihayah, vol.13, p. 268; Ibn Shaddad, Ta,rikh al-Malik al-Zahir,  pp.104-105; al-Maqrizi, al-Suluk, vol.2, pp.67, 87; al-Safadi, al-Wafi bi-l-Wafayat, vol.24, pp.178-179.min dhammihim wa-l-,ayb ,alayhim bi-kawnihim khamaru ,alà asatidhahum wa-annahum la khayr fihim). 28Also in this case, the wafidiyah were accused of being disloyal and untrustworthy (this time in the territories of the Sultanate), and hence unworthy of promotion.On another occasion, when al-,Adil Kitbugha decided to promote the Oirat immigrants, he was accused of promoting them "not according to the norms [of promotion], for he promoted them over the senior amirs of the state" (,alà ghayr al-qiyas, fa-qaddamahum ,alà akabir al-dawlah). 29In this case the complaint was that the immigrants served too short a period for being promoted.Nowhere it is mentioned that their not being slaves was problematic, an assertion that could only arise through a misinterpretation of the term mamalik al-sultan.

Was slave status an elite status and a source of pride?
That said it should still come as no surprise that none of the mamluk sultans saw fit to boast of his slave origin, or to claim that his legitimacy to rule was based on such an origin.To the contrary, mamluk sultans were in great pains to rid themselves of the negative connotations attached to their servile past, by associating themselves with established dynasties or by claiming an exalted origin.The mamluks were criticized for their servile origin by the local population and by their external enemies.The Ilkhans treated the mamluk sultans as inferiors for the latter's humble descent (nasab). 30In response, the mamluks tried to highlight their relative advantage over the Ilkhans -their being Muslims and defenders of Islam (jihad). 31In general, the local population was also not pleased with mamluk rule ("they generally did not want that one of the mamluks will be the ruler" -kana ,adatuhum an la yuriduna an yaliya al-mulk ahad min al-mamalik), and there is evidence that they supported (sometimes physically) the Qalawunids against mamluk amirs trying to take the reins of power from them. 32The mamluks could not ignore such perceptions. 33l-Mu,izz Aybek (d.655/1257), the first mamluk sultan, based his legitimacy on his marital ties with Shajar al-Durr, al-Salih Ayyub's widow.Shajar al-Durr even claimed that she was the one who gave Aybek the reins of power. 34In early Mamluk sources, al-Mu,izz Aybek and his son al-Mansur ,Ali are considered to be Ayyubid kings. 35In later sources, it is mentioned that Aybek had to spent great sums of money in order to convince the local population, that said to him "we want only a sultan from an established house, born as a leader" (la nuridu illa sultanan ra,isan mawludan ,alà fitrah), to comply with the rule of a mamluk sultan (man massahu al-riqq). 36l-Muzaffar Qutuz (d.658/1260), the second mamluk sultan, claimed that he was a relative of the Khawarizmian king Jalal al-Din Khawarizm Shah.In a biographical anecdote it is related that one day Qutuz' master beat him and cursed his fore-fathers.Qutuz, who burst into tears, explained that he was not crying because of the beating, but because his fore-fathers were cursed.When told that he had no reason to cry on account of his fore-fathers, for he was just "a Turkish mamluk, infidel son of infidels" (mamluk turki kafir b. kafirin), he replied that he was the relative of Khawarizm Shah. 37This anecdote illustrates that mamluk origin was far from being a source of pride, and that mamluks tried to repress such an origin and replace it with a more exalted one. 38l-Zahir Baybars (d.676/1277), who was criticized for being a slave by the local population and by the Sultanate's external enemies, 39 tried to associate himself to established dynasties in several ways: he asked the Qadi Ibn Khallikan (d.681/1282) to forge a Genghisid genealogy for him; 40 he married a Khawarizmian princess whose family was related by marriage to al-Salih Ayyub; 41 his relative, Baysari al-Shamsi (d.698/1298), also established marital ties with the Ayubbids; 42 Baybars also related himself to al-Salih Ayyub by adopting his nisbah; 43 he established marital ties with families of senior Mongol immigrants; 44 and he connected himself symbolically to the Saljuqs. 45In the popular treatise Sirat Baybars, Baybars is said to be the son of the king of Khurasan, born as a Muslim by the name of Mahmud, sold into slavery, adopted by al-Salih Ayyub and designated his heir.According to Thomas Herzog, the purpose of this treatise was to legitimize the rule of mamluks. 46l-Mansur Qalawun (d.689/1290) boasted of his exalted ethnic origin and his marital ties to the family of al-Zahir Baybars. 47Al-Mansur Lajin (d.698/1299) made use of the fact that he was married to al-Zahir Baybars' daughter, and therefore also indirectly related to the Qalawunid dynasty, in order to legitimize his rule.Al-Nuwayri (d.733/1333)  mentions that shortly after becoming sultan, he brought the exiled Khidr b.Baybars and his mother from Constantinople to Egypt, because he was married to the daughter of Baybars. 48He also brought to Egypt the coffin of Salamish, al-Zahir Baybars' exiled son.Al-Muzaffar Baybars al-Jashankir (d.709/1310), the relative of Al-Mansur Lajin, also emphasized his indirect tie to the family of al-Zahir Baybars.In his days, Khidr b.Baybars was allowed to leave the fortress and live in the palace of the amir Aqush al-Afram, a relative of Baybars al-Jashankir. 49l-Zahir Barquq (d.801/1399), who after al-Muzaffar Baybars al-Jashankir was the first mamluk sultan to hold the reins of power in about a hundred years, was criticized from all sides for being a slave. 50n response he attached himself to the Qalawunid dynasty 51 and boasted of his ethnic origin. 52Other Circassian mamluk sultans, such as al-Mu,ayyad Shaykh (d.824/1421) and al-Zahir Tatar (d.824/1421), boasted of an exalted ethnic origin, and Shaykh even claimed to be a descendant of Circassian kings. 53In some of the biographies of mamluk amirs in the Circassian period it is mentioned that they were of inferior origin (radi, al-asl). 54From this we learn that a mamluk's descent was a matter of importance, and that an exalted origin was a source of pride.
Like mamluk sultans, mamluk amirs did not boast of their servile origin and tried to claim for themselves an exalted descent.Qawsun al-Nasiri (d.741/1341) was proud of not being a real slave (mamluk) and for not having had to undergo the normal route of training in the barracks and slow promotion. 55Sources of the Turkish period mention the high ranking position of the fathers of Mongol war captives who became senior amirs in the Sultanate. 56For example, Salar al-Mansuri's (d.710/1310) father was in charge of the hunt (amir shikar) in the court of the ruler of Anatolia (sahib al-rum), and we know that when Qibjaq al-Mansuri (d.710/1310) fled to the Ilkhanid territories he met his father and brothers who served at the court of the Mongol khan. 57We also know that Aytamush al-Muhammadi (d.736/1336), a high ranking Mongol amir, was descended from an exalted Mongol lineage (the circumstances of his arrival into the Sultanate are not mentioned in the sources). 58It is certainly reasonable that these amirs were proud of their exalted origin, which was quite likely taken into consideration when they were promoted. 59Muslim ("free") origin was also a source of pride, and we have evidence that some mamluks (like Qutuz) claimed to be Muslim war captives. 60Some of the mamluks even claimed to be descendants of the Prophet (ashraf). 61At times the status of the mamluks in their homeland was remembered, and mamluks who came from a humble background were ridiculed. 62Even in the "more mamluk" Circassian period, service in the barracks as a mamluk was certainly not a source of pride.We hear, for example, that when the Sultan al-Zahir Jaqmaq (d.858/1453) bought the grandson of the brother of the senior amir Yashbak min Salman Shah al-Mu,ayyadi (d.878/1473), he exempted him from service in the young mamluks' barracks out of respect for his uncle ("raqqahu ,an dhalika ikraman li-,ammihi"). 63ar from being a source of pride, there is evidence that being a slave was in fact considered degrading.When al-Ashraf Khalil b.Qalawun (d.693/1293) wanted to humiliate the amir Lajin al-Mansuri, he made him a mamluk of the amir Baydara al-Mansuri (d.693/1293).According to Baybars al-Mansuri (d.725/1325), Lajin became Baydara's slave and not his servant (wahabtuka lahu haqqan li-tasira mamlukan riqqan). 64Mamluk writers usually differentiate quite clearly between rank-and-file mamluks and amirs, 65 and the expression "the amirs and the mamluks" (al-umara, wa-l-mamalik) is quite common in Mamluk sources.Whereas amirs, who were generally manumitted slaves, were respected, we come across expression of contempt towards simple mamluks.For example, when the amir Aqbay al-Hajib (d.805/1402) beat one of the amir ,Alibay al-Zahiri's (d.800/1397) mamluks, ,Alibay complained to al-Zahir Barquq, but Barquq dismissed the complaint with the words "am I supposed to beat Aqbay on account of a [simple] mamluk?" (adribu Aqbay li-ajli mamluk?). 66ccording to Shaun Marmon, "…the enslaved individual suffered a kind of legal and social metamorphosis.He left the realm of human beings and entered the realm of commodities thus losing his legal capacity to act of and for himself". 67Marmon is referring to household slaves, but it seems that the status of the military slaves (mamluks) was not much different.At times, it is implied that, unlike amirs, mamluks were not considered human beings.When the amir Al,akuz al-Nasiri (d.738/1337) cursed another amir, al-Nasir Muhammad is quoted as saying to him: "How do you allow yourself to call an amir, like you, a pimp?You were just a page in the stables until I promoted you and made you a human being" (taqulu li-amir mithlaka qawwad?wa-aysh kunta anta fi-l-istabl aushaqi, talla,tu bi-ka wa-,amaltuka b.Adam). 68ilitary slaves (mamluks) are quite often mentioned as part of a deceased amir's estate. 69When the Circassian amir Qara Sunqur al-Mansuri (d.728/1327) was pursued by al-Nasir Muhammad, he was advised to turn himself in.He refused, claiming that al-Nasir Muhammad would surely kill him, for he was originally just "a piece of Circassian slave… that was not even worth 300 Dirhams" (qit,at mamluk jarkasi… ma yaswu 300 dirham). 70This is yet another indication that a slave, not to mention a Circassian slave, was perceived as property and not as a human being.
The master's domination over his mamluks was total.Masters had the right to prevent their mamluks from marrying or to arrange a marriage as they wished. 71Cases of disobedience by mamluks were considered a severe breach of the master's honor. 72We hear of mamluks who fled from their masters out of fear. 73At times, amirs are praised for not cursing their mamluks; 74 we may thus assume that cursing, humiliation and even beating of mamluks were not that rare (and see above the story of Qutuz).It is even implied that the master had the right to take his mamluks' lives. 75A slave's manumission is occasionally compared to release from imprisonment or captivity.For example, when al-Ashraf Khalil b.Qalawun (d.693/1293) released Baysari al-Shamsi after a long period of imprisonment, Baysari took on Khalil's nisbah (al-ashrafi), like a manumitted slave, 76 and the son of the Armenian king, who was released from captivity, is called ,atiq (manumitted slave). 77Mamluks had no separate legal identity and no legal capacity to act on their own; their actions were attributed to their masters.For example, Baybars al-Mansuri refers to Kitbukha's mamluks as "his slaves, whose actions are attributed to him" (mamalikuhu, almansub sani,uhum ilayhi). 78At times a mamluk acting on behalf of his master is called ,abd ma,mur (a legal category that relates to slaves empowered by their masters to act on their behalf), 79 thus accentuating the mamluk's lack of legal capacity.
Orlando Patterson defined a slave as a powerless, violently dominated, natally alienated and generally dishonored person, who has no existence without his master. 80According to Dror Ze'evi, the near-absolute power of the master was softened by the fact that the relationship between slave and owner sometimes resembled family relations, and especially in the case of elite slavery, integration into the family of the master was a necessary phase. 81However, as Richards has already argued, only few especially favored mamluks were treated as quasi-kin by their masters and his relationship with the mass of them must have been of a more material nature. 82Elsewhere I have argued that mamluks perceived themselves as slaves because of the absence of family ties, and that only an outstanding few succeeded in completely freeing themselves of their slave status and become members of a ruling elite with family ties. 83It would thus seem that Patterson's definition fits military slaves (mamluks) quite nicely. 84

Slave status and manumission
Modern scholars have commonly argued that the servile phase in the life of a mamluk was only formal and quite limited in time.It is usually maintained that mamluks were manumitted automatically by the end of their religious and military training, at the age of twenty or less. 85This is consistent with the claim that military slaves were not slaves in the full sense of the word and that slave status was that of elite.However, at least with respect to the Turkish period, our knowledge about manumission is quite limited, and the commonly held view may well reflect only the situation in the Circassian period.Rabbat is the only scholar who elaborates on the matter (Ayalon did not explore the issue of manumission in any great depth).According to Rabbat, it seems that until the Ayyubid period mamluks remained slaves even after becoming high ranking military commanders.There is no unequivocal evidence that during the Ayyubid period mamluks were manumitted automatically at the end of their training.Rabbat assumes, but does not prove, that in the days of al-Zahir Baybars or al-Mansur Qalawun automatic manumission at the end of the training period became the norm. 86here is some evidence that at least until al-Nasir Muhammad b.Qalawun's third reign (709/1310-741/1341) mamluks were not manumitted automatically.Qalawun, who was originally the mamluk of the Ayyubid amir Qara Sunqur al-Kamili (d.647/1249), became upon the latter's death in the year 647/1249 the slave of al-Salih Ayyub.He was manumitted later in the same year, shortly before his new master died. 87alawun died in the year 689/1290, at the age of more than sixty or seventy (most sources claim that he was more than sixty years old). 88f we estimate his age as sixty-five, then he was manumitted when he 22 KOBY YOSEF Al-Qantara XXXIV 1, 2013, pp.7-34 ISSN 0211-3589 doi: 10.3989/alqantara.2013.001appear to belie somewhat the suggestion of Orlando Patterson that 'social death' was also the status of the military slave of the Islamic world", Amitai, "The Mamluk Institution", pp.67-68.However, Ibn Khaldun puts more emphasis on the benefits that the institution has for Islam rather than for the slaves themselves ("Islam rejoices in the benefit which it gains through them").Moreover, even when mamluks who became sultans make a link between their enslavement and their ascendance to power, they clearly regard themselves as being redeemed from slavery, and they consider the servile phase of their lives as a difficult one, see for example al-Nuwayri al-Iskandarani, Kitab al-Ilmam, vol.4, p. 79; for a detailed discussion, see Yosef, "Mamluks and Their Relatives", pp.67-69. 85See for example Ayalon, L'esclavage du Mamelouk, p. 9; Amitai, "The Mamluk Institution", p. 62; Rabi,, "The Training of the Mamluk Faris", p. 162. 86Rabbat, "The Changing Concept of Mamluk", pp.89-93.
87 Ibn Taghribirdi,vol. 7,Ibn Iyas,Bada,i,vol. 1,p. 189;vol. 1,p. 755;Ibn Taghribirdi,vol. 7, was about twenty-three.Significantly, he was not manumitted automatically but only upon his master's death.89 Salar al-Mansuri, who was captured in the year 675/1276, was bought by Qalawun for his son ,Ali.Upon ,Ali's death in the year 687/1288, he became Qalawun's property once again.Salar died in the year 710/1310 at the age of fifty or little less.90 If we estimate his age at death as forty-eight, then he was still a slave when he was about twenty-five.He was probably manumitted upon Qalawun's death in the year 689/1290, when he was about twenty-seven years old, for his nisbah indicates that Qalawun manumitted him.
We know that before al-Nasir Muhammad b.Qalawun's third reign some halqah members holding a military estate (iqta,) were still slaves (mamalik ariqqa,). 91These were probably low-ranking mamluks or mamluks of amirs, on whom we generally have little information.We have to take into consideration the possibility that such mamluks were manumitted at a later age. 92We also know that in the year 670/1272, al-Zahir Baybars bought two amirs; therefore, in his time amirs could still be slaves. 93 Qalawun.Perhaps he only became his servant [and not his slave], for his master, al-Muzaffar Baybars, made him an amir of ten at the end of his reign, and had he not manumitted him he would not have made him an amir" (kana asl Baktamur min mamalik al-malik al-Muzaffar  Baybars al-Jashankir, thumma intaqala ilà-l-malik al-Nasir Muhammad b.Qalawun, la,allahu bi-l-khidam fa-inna ustadhahu al-Muzaffar  Baybars kana ammarahu ,ashra fi awakhir dawlatihi wa-lawla annahu a,taqahu ma ammarahu). 94Ibn Taghribirdi's account makes it clear that in the Circassian period a slave yet not manumitted could not have become an amir.But apparently he projects this state of affairs onto the Turkish period as well.Baktamur's nisbah is al-Nasiri, and the nisbah normally goes by the manumitter.Apparently, Ibn Taghribirdi had some further indications that Baktamur was the slave of al-Nasir Muhammad, for he finds this worthy of comment.
We have some evidence that starting from al-Nasir Muhammad b.Qalawun's third reign the enslavement of some of the mamluks was a mere formality. 95This might be related to the fact that starting from this period the Turkish mamluks were generally sold by their families, who knew about the fine treatment that al-Nasir Muhammad gave his mamluks. 96Certainly, the enslavement and the conditions under which these mamluks lived were less traumatic than those of war captives, 97 and it is possible that the servile phase in their lives was considered more formal or more limited in time.
Nevertheless, there is some evidence that until Barquq's reign at least some of the mamluks were not manumitted automatically at a young age. 98Shahin al-Shaykhi (d.834/1430) was originally a mamluk of Shaykh al-Safawi (d.801/1398), but upon his master's death he was bought by Barquq.Shahin was about eighty years old when he died, and therefore he was still a slave at the age of about forty-six. 99In the year 785/1383, Aytamush al-Bujasi (d.802/1399) was bought by Barquq and immediately manumitted when the latter discovered that he was still a slave, for his original master, Jurji al-Idrisi (d.772/1370), had not manumitted him.After the latter's death the amir Bujas al-Nawruzi (d.803/1400) apparently took him from his heirs and manumitted him illegally. 100We know of other instances in which an adult mamluk (sometimes on the verge of a natural death from old age) was still legally a slave, usually due to illegal buying or selling.From such cases we cannot draw any conclusions about mamluks' manumission age. 101Nevertheless, the sources provide us with hints as to the supposed time of Bujas' manumission.Ibn Hijji (d.816/1413) comments that it is strange that Aytamush was not manumitted, for his master died in the year 772/1370 and Aytamush had been acting as a free person (yatasarrafu tasarruf al-ahrar) for a long time. 102Ibn Hajar al-,Asqalani (d.852/1449) also found it strange (min al-ghara,ib) that when he was bought by Barquq in the year 785/1383 Aytamush was still a slave, because his master "Jurji died in the year 772/1370, and Aytamush was acting as a free person for 17 (!) years, although he was still [legally] a slave" (Jurji mata sanat 772, fa-aqama Aytamush 17 sanah fi l-riqq yatasarrafu tasarruf al-ahrar). 103These reports indicate that Aytamush was a slave while his master was alive, but was supposed to be manumitted upon Jurji's death.Apparently his master did not have the chance to do this, or he did it in an illegal manner.When Aytamush died he was almost sixty years old. 104If we estimate his age as about fifty-eight, then he was still a slave at the age of twenty-eight.Note that here once again a mamluk's manumission is linked to the death of a master.It seems that, at least during the Turkish period, manumission upon the master's death (probably by a testament) was a common occurrence, and that at least some mamluks were not manumitted automatically upon the termination of their training period.
The fact that Shahin and Aytamush were originally mamluks of amirs might explain their late manumission (see page 23, footnote 92 above).Another explanation, that does not contradict but rather complements the above-mentioned explanation, is that the late age at which Shahin and Aytamush were manumitted was due to the fact that they were not Turks -Aytamush was a Circassian and Shahin was probably a Rumi. 105There is evidence that during the Turkish period non-Turkish mamluks (mainly Circasians and Rumis) did not enjoy the same treatment as their Turkish counterparts.Generally, the former were originally Christians, as were many of the Sultanate's enemies, and therefore they were described in negative terms and were subject to discrimination. 106Since there is no evidence for the sale of non-Turkish mamluks by their families during the Turkish period, we may assume that most of them were war captives and therefore their enslavement was more traumatic than that experienced by Turkish mamluks. 107There is also some evidence that non-Turkish mamluks started families at a later age than their Turkish peers, maybe due to the fact that they were manumitted at a later age.Moreover, as soon as the non-Turkish mamluks entered the Sultanate, their connection to their fami lies was severed forever.Whereas the Turkish mamluk had the option of becoming a favored mamluk, marrying into the Qalawunid family, establishing a family while still young, and of bringing his relatives into the Sultanate, this option was almost totally closed to non-Turkish mamluks.Since the creation of a family was the only way to ultimately shed one's slave status, the non-Turkish mamluks could not fully leave behind this status even after manumission, and were perceived by their contemporaries as being "more slaves" than the Turkish mamluks. 108eginning in Barquq's reign a drastic change took place in mamluk manumission practices.There is evidence that Barquq was in the habit ("ka-ma hiya ,adatuhu") of buying mamluks and immediately manumitting them, and apparently this was an innovation. 109Indeed, starting in his days, it seems that the mamluks' period of slavery was limited in time, and mamluks were manumitted automatically upon the termination of their training, or even immediately after being bought.The ex-pression "bought him and manumitted him" (ishtarahu wa-a,taqahu) is quite common in sources from the Circassian period. 110Another common expression, which, as far as I know, does not appear in texts from the Turkish period, is "manumitted him and made him one of his mamluks" (a,taqahu wa-ja,alahu min jumlat mamalikihi). 111As noted above, already in the Turkish period the term mamluk had a double meaning: slave and servant.It seems, however, that in the Circassian period the servile phase in the life of a mamluk was more formal, and he was perceived more as a servant than as a slave.Paradoxically, the expression "manumitted him and made him one of his mamluks" (a,taqahu waja,alahu min jumlat implies that until manumission, the mamluk (slave) was not considered a mamluk (servant).

Conclusion
Even though Mamluk authors emphasize the ethnic origin or language of the Sultanate's ruling elite, modern scholars emphasize its slave status or origin.The commonly held view by modern scholars is that the status of all the mamluks was that of an elite, and that the mamluks were proud of their slave origin even after manumission.I have argued that this view is in need of modification.
There is no evidence that manumitted mamluks were proud of their slave status.On the contrary, it seems to have been considered degrading and manumitted slaves with aspirations made great efforts to repress their servile past by claiming an exalted origin or by creating marital ties with established families.The term mamluk has a double meaning: slave and servant, and it frequently expresses subordination, obedience and servitude.When manumitted slaves refer to themselves as "mamluks of the sultan" they do not express pride in their slave status, but rather their subordination and obedience to the ruler.

27
THE TERM MAMLUK AND SLAVE STATUS DURING THE MAMLUK SULTANATE Al-Qantara XXXIV 1, 2013, pp. 7-34 ISSN 0211-3589 doi: 10.3989/alqantara.2013.001 110 See for example Ibn Taghribirdi,vol. 16,p. 357. 111 See for example Ibn Taghribirdi, al-Manhal al-Safi, vol.3, p. 506; Turkish-period sources usually refer to the process of the mamluk's enslavement and manumission by expressions such as "he remained his mamluk until he was manumitted" (wa-lam yazal fi jumlat mamalikihi ilà an a,taqahu), see for example al-Nuwayri, Nihayat al- Arab, vol.33,  p. 291.At least until al-Nasir Muhammad b.Qalawun's third reign mamluks were not manumitted automatically, and the servile phase of their life was not a mere formality.Until that period many of the mamluks were war captives, and we may safely assume that their enslavement was a traumatic event. Miltary slaves were considered property and they lacked a legal identity of their own.They perceived themselves as slaves for lacking family ties.The master's domination over them was total and their manumission is sometimes compared to a release from imprisonment or captivity.They were often manumitted only upon their master's death.
It may be that starting from al-Nasir Muhammad b.Qalawun's third reign the enslavement of Turkish mamluks who had been sold by their families became more of a formality or more limited in time.On the other hand, non-Turkish mamluks, who were generally Christian war captives, were subject to discrimination.They were disdained, manumitted at a later age and prevented from establishing marital ties with the Qalawunids and creating their own families at a young age.They were perceived by their contemporaries as being "more slaves" than the Turkish mamluks.
Only in the days of Barquq a norm of automatic manumission emerged, and in the Circassian period the servile phase in the life of a mamluk became more of a formality and limited in time.The mamluk was perceived more as servant rather than slave.Still, slave status never became a source of pride.

9
THE TERM MAMLUK AND SLAVE STATUS DURING THE MAMLUK SULTANATE