3

The khāṣṣa and the ‘āmma: Intermediaries in the Samanid Polity

Louise Marlow

The recurrence of the paired categories of khāṣṣa and ‘āmma in descriptions of societies far apart from one another in time and distance attests to the remarkable flexibility of this binary terminology.¹ Although applications of the terms might be precise or imprecise, the former invariably implied ‘special’ and ‘distinctive’ status, the latter ‘common’ status and lack of distinction. The broad scope of the terms, their abstract quality, as well as their sheer persistence, tend to obscure specific meanings and implications that they may have carried in the contexts in which they are invoked.² This paper addresses the designations khāṣṣa and ‘āmma, the ‘special people’ and the ‘common people’, in an early tenth-century Samanid context. It relates the testimony of an Arabic mirror for princes to the information furnished in various contemporary and near-contemporary sources, and considers the role of specifically enumerated categories among the khāṣṣa and the ‘āmma in the functioning of governance under the Samanids in the first half of the tenth century.

It is assumed in the present article that the Arabic mirror, known as Naṣīḥat al-Mulūk, despite its attribution to al-Māwardī, originated in the Samanid domains, probably in Balkh but possibly in Samarqand, towards the middle of the tenth century. The author’s conclusion of his chronology of praiseworthy rulers with ‘al-Shahīd’, that is, al-Amīr al-Shahīd Aḥmad II b. Ismā‘īl (r. 295–301/907– 14), establishes the book’s composition after 301/914, and suggests a dating during the reign of Aḥmad’s successor, Naṣr II b. Aḥmad (r. 301–31/914–43), an inference supported by many other aspects of the text. A combination of indicators, including a set of references to the Buddha and to Buddhist iconography, suggests Balkh as a likely location for the text’s first composition.³

Intended for an unnamed ruler, the book consists of ten thematic chapters. At the centre of his mirror, the author has placed a sequence of three chapters dedicated to the three ‘governances’ (siyāsāt): governance of the self, the khāṣṣa and the ‘āmma. As the later tenth-century writer al-Khwārazmī (d. 387/997) observed in his encyclopaedic Mafātīḥ al-‘ulūm, this three-fold cluster of siyāsāt corresponded to the (Aristotelian and late Alexandrian) division of practical philosophy into ethics, economics and politics, or, in al-Khwārazmī’s language, ‘ilm al-akhlāq, the science of moral dispositions; tadbīr al-manzil, the management of the household; and siyāsat al-madīna wa-l-umma wa-l-mulk, governance of the city, the community and kingdom.⁴ The author of the Naṣīḥat al-Mulūk, who appears to have participated in the development of the Kindian tradition,⁵ was certainly familiar with the branches of practical philosophy, but did not invoke them in this presentation for a courtly audience.

This essay explores Pseudo-Māwardī’s sixth chapter, devoted to the governance of the khāṣṣa, and parts of his seventh chapter, devoted to the governance of the ‘āmma. It seeks to demonstrate that, notwithstanding its prescriptive nature, the author’s unusually detailed treatment of the king’s khāṣṣa, read in conjunction with passages concerning the ‘āmma, illuminates the particular character of governance in his early tenth-century environment. The Naṣīḥat al-Mulūk, it is proposed, offers a local perspective on the relationship of the Samanid court to the society in which it was situated, and accordingly complements the impression, frequently emphasised among modern scholars, that Samanid governmental practices constituted replications of Abbasid models.⁶ For Pseudo-Māwardī, individual caliphs, like individual kings, exhibited exemplary conduct, but the institution of the caliphate held little relevance or interest. Instead, Pseudo-Māwardī’s presentation evokes a distinctive political culture that draws its support from, above all, the sacred sources of the Qurʾan and the Prophetic model, and next from human wisdom, exemplified in the ‘Testaments’ of Ardashīr, Shāpūr and the Indian king Sāb.t.r.m, Kalīla wa-Dimna, and the epistolary cycle of Aristotle and Alexander – in accordance with mā jāʾat bihi al-sunna wa-waṣafahu ahl al-ḥikma, ‘that which the Prophet’s example has established and that which persons of wisdom have described’, as he writes in one instance.⁷ In its general principles and in several matters of detail, the author’s presentation anticipates later Persian mirrors, such as the Pandnāma ascribed to Sebüktegin (r. 366–87/977–97), the Qābūsnāma or Andarznāma (475/1082–3) of Kaykāʾūs b. Iskandar (r. from 441/1049), and the Siyar al-Mulūk of the vizier Niẓām al-Mulk (410–85/1019–92).

Within a hierarchical symbolic framework, Pseudo-Māwardī portrays governance as a co-operative affair in which the ruler relies of necessity on the co-operation of numbers of intermediaries. Addressing the king and beginning his exposition with the figure of the monarch, Pseudo-Māwardī posits a ruler-centred political culture; yet as he develops his presentation, he accentuates the indispensable nature of the king’s engagement with a diverse set of mediating and often local holders of authority.⁸ At the opening of his chapter devoted to the governance of his khāṣṣa, the author presents a sacralised conception of kingship: he likens the king’s position in the social order to God’s position in relation to the world of His creation.⁹ Just as God has singled out amongst His creation angels, whom He has placed closest to Himself, and then prophets, whom He appointed as stewards over His creatures, the king, ‘according to his capacity and the extent of his strength’, should train and govern his khāṣṣa, so that they become the quickest of people to show obedience to him and the most determined in coming to his assistance (ḥattā … lā yakūna fī ahl mamlakatihi … man huwa asra‘ ilā ṭā‘atihi … wa-aqwā ‘azman fī nuṣratihi).¹⁰ This portrayal, which resumes the author’s consistent urging of the king to emulate the divine through the cultivation of the divine attributes, conforms to Roy Mottahedeh’s observation of the king’s perceived position as above and apart from society.¹¹ From this detached position, the king possessed the potential capacity to wield his khāṣṣa, who constituted an instrument (āla), in the formative management of his ‘āmma, who constituted the raw material (mādda) of the kingdom.¹² The author’s depictions of the king, parallel to the divine on the one hand and the artisan (al-ṣāni‘) on the other, convey both the unique isolation associated with the royal office and the mediated nature of the ruler’s relationship with his kingdom. The khāṣṣa who, in Pseudo-Māwardī’s presentation, constituted the king’s ‘instrument’ correspond to the ‘intermediaries’, to whose vital activities in the maintenance of the societies of eastern Iran and Transoxiana Jürgen Paul, in particular, has devoted illuminating attention.¹³ As Pseudo-Māwardī explains, it is not the king but the king’s khāṣṣa who operate in contact with the ‘āmma, and their conduct reflects on him:

(He should devote his highest attention to his khāṣṣa) because the larger part of his affairs are entrusted to them and tied to them, while they (in turn) are related to him and compared to him. From their modes of conduct (ādāb), his conduct (adab) (is inferred); from their moral dispositions (akhlāq), his disposition (khulq), and from their religion, his religion; he will be judged, favourably or unfavourably, according to that which is witnessed from them.¹⁴

Somewhat later in his chapter, the author, in an unusually precise and detailed fashion, supplies a definition of the khāṣṣa:

It is incumbent on the virtuous king to follow the command of God and imitate (the example of) His Prophet in the governance of his khāṣṣa, his family (ahl), his retinue (ḥāshiya), his armies (junūd) and his notables (a‘yān). The king’s khāṣṣa whom we mean in this place consist of categories (ṭabaqāt) that are constructed so that some of them are more special than others (buniyat ba‘ḍahum akhaṣṣa min ba‘ḍin). The most special among them to him are his children, and his khadam among his relatives and his family members; then his personal and military slaves (‘abīduhu wa-mamālīkuhu), and the élite among his personal military staff and his personal body-guards (khāṣṣ fityānihi wa-ghilmānihi); then his viziers and his secretaries, and those who discharge the affairs of his palace (kufāt ashghāl ḥaḍratihi); then his army, his senior commanders, his cavalrymen and soldiers (junduhu wa-quwwāduhu wa-asāwiratuhu wa-muqātilūhu);¹⁵ then his officials (‘ummāl), to whose assistance he has recourse in the improvement (ensuring the welfare) of his kingdom beyond his palace and his court, and outside his central and established place of residence (‘ummāluhu alladhīna yasta‘īnu bihim fī iṣlāḥ mamlakatihi al-nāʾiya ‘an bābihi wa-dārihi wa-l-khārija ‘an markazihi wa-qarārihi).¹⁶

In this evocative depiction, the author portrays the khāṣṣa as comprising internal divisions and rankings, correlated to degrees of closeness, physical or symbolic, to the polity’s centre, represented in the king’s person. The theme of proximity recurs in the author’s several references to ‘the khāṣṣ and the ‘āmm, the near and the far’ among them (for example, an yuẓhira lahum khāṣṣahum wa-‘āmmahum wa-aqṣāhum wa-adnāhum).¹⁷ The quality of ‘specialness’ dissipates in proportion to the degree of involvement beyond the confines of the Amir’s dār and qarār – the pair of terms employed, incidentally, in Ibn Ḥawqal’s near-contemporary account of Amir Abū Ibrāhīm Ismā‘īl b. Aḥmad (r. 279–95/892–907)’s selection of Bukhara as his capital: ‘The first member of the Samanid dynastic family to adopt it [Bukhara] as a (royal) residence (dār) and to make it a permanent settlement (qarār) was Abū Ibrāhīm…’.¹⁸ The royal court remained mobile, however, and the Amir moved periodically from one location to another.¹⁹ Naṣr II b. Aḥmad (r. 301–31/914–43), who reportedly developed a fondness for the city of Herat, famously engaged his troops in an extended period of settlement there; the soldiers’ desire to return to their families in Bukhara prompted Rūdakī (d. c. 329/940–1) to compose a celebrated poem, which effectively moved the Amir to return.²⁰

Pseudo-Māwardī thus depicts the king at the centre of a set of five ranked groupings. The first group consists of his immediate family and his relatives, including his khadam, literally ‘servants’ but in this case, since they are members of the dynastic family, evidently a designation for individuals who were servile in a purely metaphorical sense. (As David Ayalon established several decades ago, the term khādim also connoted ‘eunuch’, a meaning that did not apply in this context.)²¹ In this instance, the author’s reference to the king’s khadam is likely to have connoted persons who, whether or not they were related to him by consanguinity or marriage, constituted, by reciprocal bonds of loyalty and obligation, extensions of his family. As Paul has demonstrated, the khādim, in this sense, entered into a formalised relationship with his makhdūm, the king, who, in exchange for the pledge of obedience and loyalty, incurred the obligation to provide ‘benefits’ (ni‘am).²² The second grouping consists of linked categories, tied to the king’s person, and employed in various capacities at the court, in the royal bodyguard, or in the Amir’s military activities; the varied vocabulary of ‘abīd, mamālīk, fityān and ghilmān evokes the diverse functions undertaken by these groupings.²³ The third category among the khāṣṣa consisted of the admin-istrators, who discharged the ruler’s affairs in situ. The fourth group (junduhu wa-quwwāduhu wa-asāwiratuhu wa-muqātilūhu) comprised the various territorial levies that, when summoned, still constituted large portions of the Samanid armies.²⁴ These categories, separate from the Amir’s personal troops, characteristi-cally resided on their estates, which they possessed either as familial properties or as gifts, marks of ni‘ma, ‘favour’, dispensed by the ruler.²⁵ The fifth group consisted of the king’s official representatives, who, dispersed across his dominions, were particularly responsible for the collection of revenue.

The equation of distinguished status and physical proximity to the royal person found expression in the architectural planning of palaces and their precincts. In a celebrated example, Narshakhī, who composed his Tārīkh Bukhārā in 332/943, relates that Naṣr II b. Aḥmad had constructed in Bukhara, in the ancient vicinity of the Rīgistān, where kings had built palaces since antiquity, a new palace, at great expense and of great loveliness, for himself, and ‘at its door’ a palace for his officials (sarā-yi ‘ummāl), each of whom had his own dīvān bar dar-i sarā-yi sulṭān.²⁶ The design suggests in physical terms the close relationships of the groups accom-modated within these dīvāns with the ruler, and their status within his khāṣṣa.

After this explanation of the groups that comprised the ruler’s khāṣṣa, the author turns his attention to the king’s governance of these groupings. He begins with the king’s immediate family. Addressing the physical, intellectual and moral formation of princes, he first advises the king regarding the choice of a mother, and, in the event that the mother is unable to breastfeed her child, of a wet-nurse; and the criteria for the selection of a pleasing and suitable name. After discussing the prince’s infancy and early childhood, he turns to the prince’s education. The choice of a teacher, he avers, is at least as important as the choice of the child’s mother and wet-nurse. The prince should learn the Qurʾan with Arabic, and the author expatiates on the importance of acquiring a working proficiency in that language for religious, cultural and intellectual reasons. He offers advice on methods of instruction in Arabic and recommends appropriate reading matter. He stipulates the importance of physical education, and, in an echo of ancient practices of royal training, includes polo among the athletic pursuits appropriate for princes.²⁷

After a brief treatment of the king’s relatives, whom, regardless of their behaviour, the ruler is obliged to treat with generosity and kindness, the author turns to the king’s khadam and ḥasham. The latter term, usually rendered as ‘retinue’, appears to have functioned as a synonym for ḥāshiya (pl. ḥawāshī).²⁸ As the extract reproduced below reveals, Pseudo-Māwardī links these two groups with the previously mentioned categories of ghilmān and fityān. This grouping, as well as the author’s recommendation of a common course of instruction and a common mode of treatment for all of these categories, would seem to suggest that they shared a common status; it is possible, however, that the author grouped the khadam and ghilmān, and the ḥasham and fityān, in this order primarily for reasons of rhyme. The groups constituted sub-categories of the khāṣṣa, and were bound to the ruler through relationships of khidma, ‘service’, and iṣṭināʾ, ‘nurturing’:

As for the needs of the ‘servants’ (khadam) and ‘retinue’ (ḥasham): it is essential that there be in the kings’ palace instructors and teachers who will teach the ghilmān, khadam, fityān and ḥasham that which each one’s condition can bear of the Qurʾan and religion; remind them time after time of God; inform them of the principles of the religion, the religious laws (sharāʾi‘), and the performance of the prayers with the full ritual purity (ṭuhūr), bowing (rukū‘) and prostration (sujūd); explicate to them the religion’s supereroga-tory acts (nawāfil) and its branches; teach them fine moral dispositions and praiseworthy acts; admonish them, remind them and instill in them fear of the Fire; call them to Paradise and the Abode of Permanence; and urge them to jihād. He should restrain them from corruption and bad habits, and committing that which God has forbidden of the major sins (kabāʾir), such as fornication and false accusation. (If) anyone among them commits a sin out of inadvertence and negligence, then the way to proceed in this case is to turn aside and feign inattention. (In the case of) anyone who commits a minor sin (ṣaghīra) – to the exclusion of that which requires one of the statutory penalties, or brings about corruption against the kingdom and the religion (wa-man irtakaba ṣaghīratan dūna mā yajibu fīhi lillāh ḥadd aw ya‘ūdu ‘alā l-mamlaka wa-l-dīn bi-fasād) – then (the path to adopt is that of) exhortation, reproach, arousing distaste for it and deterrence. If he repents of (the misdemeanour), abandons it and turns remorsefully to God, then (he should receive) pardon and forgiveness for it. If, on the other hand, he continues (in it), then (he merits) punishment and exemplary chastisement, in accordance with his obstinacy and persistence and in proportion to the degree of the crime and sin.²⁹

This passage reflects a courtly environment in which the ruler’s ghilmān, evidently acquired from culturally distant locations, had attained such numbers and prominence that Pseudo-Māwardī largely equated them with the ruler’s khadam and ḥasham. The Samanids’ location at the frontier facilitated their predominance in the capture, training and sale of slaves, whose ethnicities and native languages pass without mention in Pseudo-Māwardī’s account, but who, he indicates, required rudimentary instruction in prayer and fitting behaviour, and initiation into the established culture of the court and society.³⁰ Unlike the author’s earlier invocation of ‘abīd and mamālīk, the terms ghilmān, fityān and ḥasham need not connote the status of slavery, and even in the case of terms, such as mamālīk and khadam, that evoked servile (or formerly servile) status, their usage appears to have encompassed a set of figurative meanings.³¹ The individuals who composed the ruler’s personal khadam and ḥasham were assimilated in the likeness of foster-children into the royal household.

The author’s diversified vocabulary of ghilmān, khadam, fityān and ḥasham is likely to correspond to a variety of posts and activities, as well as a range of levels of status and prestige; it is likely that, following Central Asian custom, a number of the ghilmān served in the Amir’s personal bodyguard.³² In a later passage, Pseudo-Māwardī discusses the moral status of the purchase of slaves; his discouragement of the acquisition of slaves for purposes of pleasure and his endorsement of their purchase for purposes of defence indicate the prevalence of their employment in both domestic and military settings in the early tenth-century Samanid court.³³ In the present context, Pseudo-Māwardī’s allusion to the preparation of ghilmān for participation in jihād indicates the military purposes for which some, perhaps many, of the slaves were intended. According to Niẓām al-Mulk (410–85/1019– 92), Ismā‘īl had sent an army of ghulāmān against Abū Bilāl al-Qarmaṭī as early as 295/907,³⁴ and it is possible that over the course of the tenth century, the Amirs sought to reduce their dependence on the essentially voluntary services of military and landowning leaders whose wealth, property and status enabled them to dispose of large numbers of recruits, animals and weapons, and to replace these contingents with a greater proportion of soldiers under the command of their personal retainers.³⁵ Military ghilmān owed their training, equipment and social positions to the ruler, their patron or ṣāni‘, to whom they were taught to display obedience and gratitude.³⁶

Pseudo-Māwardī’s account emphasises acculturation rather than professional training of the kind ascribed to the Samanids in the Siyar al-Mulūk, yet the Samanid author anticipates Niẓām al-Mulk’s encouragement of tolerance and readiness to pardon minor and inadvertent indiscretions, especially, in Niẓām al-Mulk’s presentation, amongst persons promoted to high rank (kasānī-rā kih bar kashand wa-buzurg gardānand).³⁷ Pseudo-Māwardī’s exclusion of such lenience in cases of ‘corruption against the kingdom’ perhaps responds to the involvement of ghilmān in the murder of Aḥmad II b. Ismā‘īl in 301/914 and their subsequent attempt to occupy Bukhara and control the succession.³⁸

Pseudo-Māwardī continues:

If the king is a person who takes the matter of religion into account especially, then it is necessary in (his) governance that there should be with him, in his presence and in his palace some persons from among the people of tawḥīd and religious understanding (ahl al-tawḥīd wa-l-fiqh fī l-dīn) who will teach them its roots and acquaint them with its general principles, and increase (the level of his teaching) for anyone in whose nature (ṭab‘) he perceives a capacity to accept such increase, and from whom he expects aptitude for learning.³⁹

The king, then, should engage qualified instructors to educate the ghilmān, and respond to the potential of those in whom they detect a capacity for further education. The passage suggests the potential and encouragement for particularly able ghilmān, initiated into the new cultural milieu, to advance to prominent positions.

The author next enumerates ten principles according to which he should treat all members of his khāṣṣa, regardless of their positions within that category. Some of these recommendations anticipate the advice set forth in the Siyar al-Mulūk: the timely payment of salaries, predictable patterns of advancement in accordance with individual merit, the prevention of injustice towards the subjects, the forgiveness of minor infractions, the impartial application of punishment in cases where it is required, the avoidance of inactivity among the khāṣṣa, care in the selection of trustworthy advisers from among them, the prevention of dissolute behaviour, accessibility, care for their sick, aged, orphans and heirs, and replacement of their animals, weapons and possessions when these have been expended in his battles.⁴⁰

After identifying these general principles, the author introduces another aspect of ‘special’ or ‘distinguished’ status, namely the specialised nature of the offices and functions subsumed within the ruler’s khāṣṣa:

There is no alternative for the king other than to avail himself of the most distinguished (specialised) persons (al-akhaṣṣ), and (particularly) those persons among his khadam who are most distinguished (al-akhaṣṣ min khadamihi) in the important matters of his districts, (such as) the collection of wealth in the kingdom, and the disbursement of it among the armies (juyūsh) in accordance with persons’ rightful claims (fī sabīl al-ḥuqūq).⁴¹

These individuals, in the author’s depiction, participate in the king’s khadam, a differentiated group within which some persons were more distinguished than others. The author proceeds to detail these particular participants in the maintenance of the kingdom:

In the maintenance of the kingdom and the great provinces (fī iqāmat al-mamlaka wa-l-wilāyāt al-‘aẓīma), it is impossible to dispense with viziers, deputies, secretaries, commanders of armies, overseers of military affairs, directors of police, overseers (of the conveyance of information) or leaders (of communities), officers of the guard, gatherers of information, (fiscal) agents and judges (lā budda … min wuzarāʾ wa-khulafāʾ wa-kuttāb wa-aṣḥāb juyūsh wa-‘āriḍīn wa-aṣḥāb shuraṭ wa-nuqabāʾ wa-aṣḥāb ḥaras wa-aṣḥāb akhbār wa-wukalāʾ wa-quḍāt).⁴²

The author’s list of these eleven positions, operative in ‘the kingdom’ and ‘the great provinces’, augments the accounts of Narshakhī and al-Khwārazmī, who, writing during the first and second halves of the tenth century respectively, also provided information pertinent to the offices and specialised functions involved in Samanid governance. In his account of Naṣr b. Aḥmad’s construction of the royal precincts in Bukhara, Narshakhī enumerated ten dīwāns located in the ‘palace of the officials’ (sarā-yi ‘ummāl) that the Amir had had constructed next to the royal palace. The Amir had the complex built in order that ‘every official should have a separate dīwān’ (chunān-kih har ‘āmilī-rā … dīwānī būdī).⁴³ Narshakhī’s list consisted of, in this order, the dīwān-i wazīr; the dīwān-i mustawfī, the department of the treasury; the dīwān-i ‘amīd al-mulk, the office of correspondence; the dīwān-i ṣāḥib shuraṭ, the office of the chief of police;⁴⁴ the dīwān-i ṣāḥib-i muʾayyid;⁴⁵ the dīwān-i sharaf (or dīwān-i ishrāf), department of inspection;⁴⁶ the dīwān-i mamlaka-yi khāṣṣ, the office for the private domains of the ruler; the dīwān-i muḥtasib, the office of the inspector of markets and public space; the dīwān-i awqāf, the department responsible for the oversight of religious endowments; the dīwān-i qaḍāʾ, the office for the administration of justice.⁴⁷ Al-Khwārazmī, who possessed expert knowledge of Samanid administrative affairs and procedures, supplies supple-mentary information in his list of seven divisions among the Amir’s secretarial staff: the kuttāb dīwān al-kharāj; the kuttāb dīwān al-khazn; the dīwān al-barīd; the kuttāb dīwān al-jaysh; the dīwān al-ḍiyā‘ wa-l-nafaqāt; the dīwān al-māʾ; and the kuttāb al-rasāʾil.⁴⁸

These three lists of administrative components are complementary rather than equivalent. The author of the Naṣīḥat al-Mulūk presents a list of the specific functions necessary to rulers in their governance of the kingdom and the great provinces subsumed within or affiliated to it; Narshakhī provides a list of the administrative offices represented in physical form in the environs of the Amir’s palace; al-Khwārazmī supplies a more restricted list of the divisions among the administrative and secretarial staff. Similarities among the lists are nevertheless apparent. Narshakhī’s first and last dīwāns correspond to the first and last posts mentioned in the Naṣīḥat al-Mulūk. In many depictions, the vizier was expected to remain in close proximity to the monarch; Pseudo-Māwardī’s contemporary Abū Zayd al-Balkhī (d. 322/934), however, wrote that a vizier who upheld high standards of competence and trustworthiness might be permitted to reside in the more distant provinces and dispatched to the remote regions in order to promote the prosperity of the lands and to govern the subjects.⁴⁹ The office of ṣāḥib shuraṭ appears in both lists. All three authors refer to the ruler’s secretarial staff; where Narshakhī mentions the dīwān of the ‘amīd al-mulk, Pseudo-Māwardī lists the kuttāb, and al-Khwārazmī the kuttāb al-rasāʾil.⁵⁰ These offices, of course, existed in every administration of the period, and the authors’ accounts reflect the frequently noted parallels between Samanid and Abbasid institutions. Consideration of the presentation of these categories among the khāṣṣa in the light of Pseudo-Māwardī’s later presentation of the ‘āmma, however, yields a more complex and particular impression of the workings of Samanid governance.

Pseudo-Māwardī, Narshakhī and al-Khwārazmī include in their listings, under various designations, the gatherers of information. Pseudo-Māwardī refers to the aṣḥāb akhbār, ‘conveyers of information’, and al-Khwārazmī to the department of information, under the rubric of dīwān al-barīd. The dīwān al-barīd appointed gatherers and conveyers of information (aṣḥāb al-barīd) to the provinces, including the regions under the governance of vassals to the Samanids, who appointed officers in turn. The ṣāḥib al-barīd was responsible for informing the administration of the political circumstances in his district; in the employ of the state, he was not subject to the authority of the provincial governor under whom he served, and might inform against him.⁵¹ The author’s aṣḥāb akhbār performed the functions of the mukhbir, an agent of the barīd.⁵² Pseudo-Māwardī accentuates the importance of the collection and communication of intelligence in a later passage, in which he advises the king to appoint over all his officials ‘spies, overseers and supervisors’ (an yaj‘ala ‘alā kullin minhum ‘uyūnan wa-mushrifīn wa-azimma).⁵³ Pseudo-Māwardī’s use of the terms aṣḥāb akhbār and mushrif[ūn], as opposed to aṣḥāb al-barīd, is likely to reflect the shift, underway during his lifetime in Iraq and elsewhere, in the responsibility for gathering information to these groups from the barīd.⁵⁴ Mushrifs operated in the domains of the Ghaznavids and Seljuqs in particular, and the Naṣīḥat al-Mulūk suggests that they operated in the Samanid period as well.⁵⁵ Narshakhī’s list includes two offices possibly related to the gathering and conveyance of intelligence.⁵⁶ If the emendation of Narshakhī’s dīwān-i sharaf to dīwān-i ishrāf is adopted, it probably suggests, as in Pseudo-Māwardī’s vocabulary, the function of inspection or oversight;⁵⁷ and if the reading of dīwān-i barīd is substituted for dīwān-i muʾayyid, it connotes the office mentioned by al-Khwārazmī. Later advisory writings in Persian, notably the Pandnāma of Sebüktigin and the Siyar al-Mulūk, affirm the importance of the gathering and reliable communication of information.⁵⁸

The apparent omission from Narshakhī’s list of departments dedicated to military affairs and administration has occasioned considerable scholarly reflection.⁵⁹ Pseudo-Māwardī, by contrast, includes in his inventory of essential categories the aṣḥāb juyūsh, the ‘āriḍūn and the aṣḥāb shuraṭ, mentioned in sequence and apparently all distinct from one another. The term jaysh (pl. juyūsh) sometimes, though by no means invariably, appears to imply a permanent military body, and Pseudo-Māwardī’s aṣḥāb juyūsh perhaps represents the ruler’s standing military personnel, including commanding officers and senior staff.⁶⁰ His category of ‘āriḍūn, officials in charge of military affairs, chiefly the mustering and inspection of troops and the administration of the troops’ allowances and salaries, corresponds to al-Khwārazmī’s kuttāb dīwān al-jaysh. The aṣḥāb shuraṭ, police commandants, possessed deeds of appointment from the ruler for the performance of police duties in regional locations, where they discharged these duties in conjunction with local notables.⁶¹ The ṣāḥib shuraṭ appeared in Narshakhī’s list, whereas the aṣḥāb ḥaras, officers of the guard, appear only in Pseudo-Māwardī’s text, where they are listed immediately after the police commandants. The Round City of Baghdad, the quintessential example of the architectural embodiment of political relationships, included quarters for both the ḥaras and the shurṭa, in a design perhaps replicated at Samarra.⁶² Under the Ghaznavids, the amīr-i ḥaras or commander of the guard maintained discipline at court and in the vicinity of the palace.⁶³

The representations of Narshakhī in particular and al-Khwārazmī to a somewhat lesser degree presuppose a Bukharan perspective; the extent to which the offices detailed in the Tārīkh-i Bukhārā functioned beyond the confines of the city and region of Bukhara is questionable.⁶⁴ By contrast, the author of the Naṣīḥat al-Mulūk, who lived, it is proposed, probably in Balkh but possibly in Samarqand, wrote at remove from the capital. In his list of specialised functional categories within the king’s khāṣṣa, Pseudo-Māwardī employs only indefinite plural forms; this unvarying usage, together with his references to ‘the kingdom’ (al-mamlaka) and ‘the great provinces’ (al-wilāyāt al-‘aẓīma), evoke a network, its elements linked by means of a system of intermediaries, replicated at differing levels across the vast terrains of the Samanid realm. The near-contemporary author of the Ḥudūd al-‘ālam, completed in 372/982 and similarly the product of a regional, possibly Chaghani perspective, explains, ‘The mīr of Khorāsān resides at Bukhārā …; he is from the Sāmān family (āl-i Sāmān) …. These (princes) are called Maliks of the East and have lieutenants (‘ummāl) in all Khorāsān, while on the frontiers (ḥadd-hā) of Khorāsān there are kings (pādhshāhān) called “margraves” (mulūk-i aṭrāf)’.⁶⁵ Pseudo-Māwardī provides an elaboration of this general observation of the relationship of Bukhara to the provinces and extremities of the Samanid domains.

In an indication of his regional point of view, Pseudo-Māwardī, unlike Narshakhī and al-Khwārazmī, includes in his account the Amir’s deputies (khulafāʾ), to whom he assigns a prominent position.⁶⁶ The amirs appointed deputies to represent them during periods of absence, usually for purposes of military campaigns, from their principal bases of authority. Reports of the appointment of deputies appear frequently in relation to the reigns of the earlier amirs, who, unlike their successors, participated actively in ghazw-related warfare,⁶⁷ and whose example Pseudo-Māwardī invoked for the benefit of his recipient.⁶⁸ In the early Samanid period, the amirs often appear to have chosen as their deputies members of collat-eral branches of the dynastic family. When Ismā‘īl, at the beginning of his conflict with his brother Naṣr (I) (r. 250–79/864–92), left Bukhara for Samarqand, he left as his deputy (khalīfa) in Bukhara his nephew Abū Zakariyyā Yaḥyā b. Aḥmad b. Asad, and when Naṣr II in [3]13 left Bukhara for Nishapur, he appointed Abū l-‘Abbās Aḥmad b. Yaḥyā b. Asad Sāmānī in the capital as his deputy.⁶⁹

The ruler’s khulafāʾ represented him in situations that involved direct contact with the population, and for this reason, like his tax collectors (‘ummāl), they provided and disseminated an impression of the character of his rule. In a later passage, Pseudo-Māwardī alludes to this representational role when he writes, ‘The king … should know that he has no more implacable enemy … than his (financial) official (‘āmil) if he is unjust and oppressive, and his deputy (khalīfa) if he is aggressive and tyrannical’.⁷⁰ Ibn Ḥawqal, who embarked on his travels in 331/943, visited the Samanid domains in the reign of Manṣūr I b. Nūḥ b. Naṣr (r. 350–65/961–76) and is likely to have been a younger contemporary of the author, demonstrates the force of the impression created by the king’s representatives when he praises the efficiency and fairness of the Samanid provincial administration. Each district (nāḥiya) had a qadi, a ṣāḥib khabar wa-barīd, a ṣāḥib ma‘ūna and a kātib salla, also known as a bundār, responsible for the collection of the kharāj and other sorts of funds owed to the authorities, most of them due to the governor of Khurasan (ṣāḥib Khurāsān).⁷¹ In this descriptive passage, Ibn Ḥawqal mentions two offices missing from the lists of Pseudo-Māwardī, Narshakhī and al-Khwārazmī. The ṣāḥib (‘āmil) al-ma‘ūna , as Patricia Crone has indicated, from the ninth century onwards assumed responsibility for policing and the execution of justice; in some locations his office perhaps came to resemble that of the military governor.⁷² The kātib salla or bundār, also absent from the lists of Pseudo-Māwardī, Narshakhī and al-Khwārazmī, appears to have exercised the functions of the ‘āmil; the titles perhaps represented regional designations for the widely used term.⁷³ The mustawfī, mentioned only in Narshakhī’s report, was responsible for supervising the work of the tax collectors who levied the kharāj and other types of revenue in provincial locations.⁷⁴

In a further indication of his regional rather than Bukharan orientation, Pseudo-Māwardī omits from his list of essential categories the personnel associated with the department of royal lands (dīwān-i mamlaka-yi khāṣṣ), and the department of religious endowments (dīwān-i awqāf). These two departments, included in Narshakhī’s portrayal, were principally concerned with lands situated in the environs of Sughd.⁷⁵

The department of correspondence or chancery, the dīwān al-rasāʾil, which appears in al-Khwārazmī’s account, corresponds to Narshakhī’s dīwān-i ‘amīd al-mulk. Many of the poets whose biographies appear in the Yatīmat al-dahr were employed in this department as secretaries (kuttāb).⁷⁶ Al-Khwārazmī mentions specialised groupings within the administrative staff, most of whom were considered, in a general sense, kuttāb. He alone mentions the dīwān al-māʾ, which, he states, was situated in Marw and had responsibility for matters of distribution and taxation related to water;⁷⁷ as Paul has indicated, the maintenance of irrigation systems constituted a major area in which effective mediation between the state and the population was an absolute requirement. Pseudo-Māwardī refers to the kuttāb, secretaries, without further differentiation, and it is likely that his reference applies primarily to the secretaries involved in the king’s correspondence, al-Khwārazmī’s kuttāb al-rasāʾil. This interpretation finds support in Pseudo-Māwardī’s later reference to the qualities required in the kātib: ‘an excellent knowledge of literary culture (adab) and language, a fine hand and clarity of expression, facility of presentation and excellence of talent’. At the same time, ‘an excellent knowledge of accounting and matters of income and expenditure’ constituted the necessary qualification for ‘the agents and collectors of taxes among the secretaries’ (al-wukalāʾ wa-jubāt al-amwāl min al-kuttāb).⁷⁸

This last reference permits the inference that Pseudo-Māwardī’s wukalāʾ, listed independently as an indispensable group among the king’s khāṣṣa, constituted a specialised grouping within the ranks of the kuttāb. Like the jubāt al-amwāl, they were perhaps involved in the collection or administration of revenue. Under the Ghaznavids, the designation wakīl or wakīl-i khāṣṣ would denote an agent involved in the management of the court, and especially the financial administration of the court’s expenditures and provisions.⁷⁹ Whether Pseudo-Māwardī’s wukalāʾ performed these specifically court-related functions in the Samanid context is uncertain. Figures closely associated with the court, such as the chamberlain (ḥājib), do not appear in his enumeration of essential functions; nor do such individuals as the boon companion (nadīm), the pen-case holder (davātdār, dawātī), cupbearer (sharābdār) and keeper of the wardrobe (ṣāḥib al-kiswa).⁸⁰ By his own testimony, the author had no intention of addressing the trappings of courtly life, which, he averred, had been thoroughly addressed already in books ancient and modern.⁸¹

Though Pseudo-Māwardī displayed little interest in the ceremonial offices attached to the court, his treatment of the khāṣṣa reflects his perceptions of governance in the context of the provincial cities. In his list of essential personnel, he includes, like Narshakhī, the judicial function. Since the leading judges in the provinces held their offices by central appointment, the dīwān-i qaḍāʾ constituted a link between the state and the populations of regional locations. The qadi was, of course, responsible for the administration of justice in accordance with the sharī‘a; in some cases, he also presided over courts of maẓālim, the redress of grievances, which, in the early Samanid period, the amirs and members of the dynastic family often dispensed in person.⁸²

Unlike Narshakhī and al-Khwārazmī, Pseudo-Māwardī also mentions the nuqabāʾ, by which term he perhaps meant individuals involved in the relay of information from the provinces to the court,⁸³ or the nuqabāʾ al-ashrāf, the ‘marshals of the nobility’, namely the descendants of the Prophet and his family, the ahl al-bayt, especially the ‘Alids, descendants of ‘Alī and Fāṭima. The nuqabāʾ al-ashrāf oversaw genealogical, material and moral matters pertaining to the ahl al-bayt; their duties included the proper administration of the awqāf established for the ashrāf. By the end of the ninth century, all of the larger towns had naqībs, who fell under the supervision of a chief naqīb, the naqīb al-nuqabāʾ.⁸⁴ The author’s reference to nuqabāʾ perhaps denotes the leaders of communities, and especially the ‘Alids, whose status and level of activity in public affairs had risen markedly over the preceding half-century, as the Tārīkh-i Bayhaq (completed in 563/1167), a later work informed extensively by earlier sources, illustrates with particular clarity.⁸⁵ In Nishapur, Hamadan, Qum, Rayy and Samarqand, ‘Alid families frequently combined their Prophetic genealogy with other kinds of social capital, such as marriage relations, wealth and scholarship.⁸⁶

Among the things that must be known is that the king’s adornment lies in the wellbeing of the subjects. The more affluent and distinguished the subjects, the nobler their condition in (terms of) religion and the world, and the more prosperous and extensive his kingdom, then the greater in authority (a‘ẓam sulṭānan) and the more illustrious in repute (ajall shaʾnan) the king; (whereas) the more base in state and dejected in mind (the subjects), the more negligible in sovereignty, the more insignificant in income and demeaned in reputation is the king. For it is not suitable for the governing king to seek the prosperity of his station (manzila) by ruining the stations

In accordance with his discursive purposes, Pseudo-Māwardī’s discussion of the essential personnel within the king’s khāṣṣa presents the role of intermediaries from the ruler’s point of view. The author’s social description reflects and accentuates the relativity of the ruler’s power and his dependence on the co-operation of parallel networks of individuals and groups that linked disparate regions and populations to the court. In a later passage, situated in his next (seventh) chapter, devoted to ‘the governance of the common people’, Pseudo-Māwardī offers a description in which he treats the role of intermediaries from a complementary perspective. In this context, he describes the ruler’s reputation and effectiveness (his ‘adornment’) in the following terms: (manāzil) of the subjects, nor to amplify (the contents of) his treasuries and the repositories of his wealth by depleting and diminishing the houses of the common people.⁸⁷

His adornment and magnificence lie as much in the prosperity of the kingdom, the plenty of its revenue, the abundance of its wealthy (aghniyāʾ), its elders (mashāyikh), its dihqans (dahāqīn), its scholars (‘ulamāʾ), jurists (fuqahāʾ), exponents of (political) opinions (dhawū l-ārāʾ),⁸⁸ its notables (sarawāt), judges (ḥukkām), ascetics (nussāk), philosophers (ḥukamāʾ), and the categories of those possessed of ranks and meritorious status (office-holders and members of respected families) (aṣnāf dhawī l-marātib wa-l-manāqib) among them, as in the prosperity of his castles (quṣūr), the lavishness of his palaces (dūr), the abundance of his horses, his troops (junūd), his khadam and his furnishings (athāth).⁸⁹ The might that he possesses against his enemies, derived from his listening and obedient subjects, bound in love to him and defending him, is no weaker than the strength (that he acquires) through his assistants (a‘wān) and troops (junūd). Moreover, the fear that he harbours against his enemies, those who are external to his kingdom and those who oppose him in his religious community (milla) and his community (umma), is no more severe than his fear of the opposition of his subjects’ hearts – rather, that which comes to him from all of these quarters concerning his subjects is more extensive, more far reaching, more consequential and more decisive.⁹⁰

In this passage, Pseudo-Māwardī provides a further description of the categories that constituted indispensable intermediaries in the political culture of the Samanid domains. In two successive articulations, he distinguishes between two groups: categories of persons who were directly dependent on the ruler and subject to his command, and categories whose service was primarily voluntary and accordingly negotiable. Pseudo-Māwardī asserts that the king’s enduring sovereignty and the stability of the polity owed as much to the second set of categories as to the first. He refers to the first set of categories in terms that liken them to possessions; the king’s ‘assistants’ (a‘wān), troops (junūd) and khadam, here almost certainly connoting persons who have entered into relationships of khidma with the ruler, are in metaphorical terms ‘his’, and assimilated in the author’s description to the king’s castles, palaces, horses and furnishings. The second set of categories, in contrast, are associated not with the king’s person but with the kingdom. This set consists of groups largely independent of the king’s direct control, on whose co-operation (in the author’s language, ‘love’) and goodwill he nevertheless depends. The social stature of these persons stemmed variously from wealth, inherited status and lineage, merit acquired through learning, the cultivation of the intellect and personal conduct, esteemed ways of life, and professional rank.

It is quite striking that the author begins his list of categories whose support for the ruler was of a voluntary nature with ‘the wealthy’. According to al-Iṣṭakhrī and Ibn Ḥawqal, the wealthy people (ahl al-tharwa, ahl al-amwāl) of Transoxiana were distinctive in their expenditure not for the sake of their own amusements but for the numerous ribāṭāt and public works associated with jihād.⁹¹ Pseudo-Māwardī is likely to have invoked the wealthy, as well as the dihqans, on account of their capacity to supply material support and muster numbers of recruits for the ruler’s military undertakings. The author’s early mention of the dahāqīn reflects the enduring distinctiveness of this category in the eastern regions during the first half of the tenth century. Al-Iṣṭakhrī and Ibn Ḥawqal remarked on the dahāqīn’s importance as commanders of troops and praised their loyal service.⁹² In the eastern territories, unlike in Iraq and western Iran, the dihqans remained a significant social force, and contributed in a range of political, military and cultural functions; the poetry of Rūdakī provides a contemporary testimony to their continuing prominence in the Samanid polity, and the figure of Aḥmad b. Sahl (d. 307/920) provides a particularly telling example.⁹³

While the wealthy and the dahāqīn provided troops and material support for the ruler’s campaigns, religious scholars and ascetics also featured in Pseudo-Māwardī’s listing of categories whose co-operation was necessary to the ruler, as they possessed the authority to confer legitimacy on his undertakings, military and otherwise. In addition, individual scholars and renunciants sometimes participated in and led ghazi campaigns.⁹⁴

Pseudo-Māwardī’s distinction of the mashāyikh among his second set of social groupings corresponds to the capacity of this group, usually associated with a particular urban setting, to intervene in affairs by means of mediation and endorsement. In one example, the mashāyikh of Samarqand, having learnt of the death of Aḥmad II in 301/914, pledged allegiance to his uncle Isḥāq b. Aḥmad, the governing amir of the city; in another case, ‘Amr b. al-Layth (r. 265–88/879–901) dispatched to Ismā‘īl b. Aḥmad a group of the mashāyikh of Nishapur, apparently in an effort to avert hostilities.⁹⁵ The author’s ‘categories of those possessed of ranks and meritorious status’ (aṣnāf dhawī l-marātib wa-l-manāqib) is likely to have referred to office-holders, including, perhaps, local figures whose involvement was required in the collection of revenue. The second element in the formulation perhaps designated above all the ‘Alids, who, in the cities of Khurasan by the late ninth century, constituted, in Richard Bulliet’s phrase, ‘a blood aristocracy without peer’.⁹⁶

Pseudo-Māwardī’s exposition of the categories whose satisfaction and co-operation were essential to the ruler’s governance articulates a fundamental characteristic of the political culture of the Samanid kingdom, namely that the dynasty’s capacity to endure depended on the periodic participation and active support of the region’s multiple communities and constituencies. The governors of cities and their subordinates; the leading members of local landowning families, including the distinctive category of the dahāqīn; prominent members of the urban elites, such as religious scholars, judges, members of eminent ‘Alid families and renunciants: these categories, like Pseudo-Māwardī’s earlier listing of the essential personnel among the king’s khāṣṣa, mediated between the changing rulers and the local populations, and their co-operation was essential to the functioning of the Samanid system. The ‘āmma, often portrayed as an undifferentiated collectivity, similarly consisted of agents whose co-operation, or at least acquiescence, was required.⁹⁷ Paul has analysed the varied arrangements by which the intermediaries functioned; some individuals undertook certain responsibilities, such as the collection of taxes, on a contractual basis; some held official appointments, but required the consensus of the local population in order to discharge their duties; and some held no official appointment, but possessed authority through the loyalty of the local population.⁹⁸ The two passages discussed in the present article illustrate, from differing perspectives, the modes of mediation within the Samanid polity. Various groupings among the khāṣṣa represented the king and acted on his behalf; their ability to act among the population, however, required the co-operation of further levels of intermediaries. This system of interdependent networks suggests that the distinction between khāṣṣ and ‘āmm, so pervasive in expressions of political culture and seemingly absolute, was often relative and indeterminate. Large categories of individuals occupied positions between khāṣṣ and ‘āmm, or, perhaps more accurately, their standing was khāṣṣ in relation to persons subordinate to them, and ‘āmm in relation to persons endowed with superior authority. In his following chapter, devoted to the governance of the common people, the author makes this point explicitly:

Among them [ten qualities that the author proceeds to enumerate] are some in which is no difference between the khāṣṣa and the ‘āmma, because the king’s khāṣṣa, according to the amount of (their) acquaintance (with the king) (‘alā miqdār al-ta‘āruf), are, in relation to those apart from them, commoners (‘āmma), since it is agreed that in no one of the countries is there a greater abundance among the sum of the kings than his slaves (‘abīd) and his khadam, nor the like of their numbers. Also among them (the ten qualities) are those that are unique to the ‘āmma to the exclusion of the khāṣṣa.⁹⁹

Nor did proximity to the king’s person, and a higher position in the hierarchy of the khāṣṣ, necessarily correspond to greater power; in some instances, greater distance coincided with greater independence.

To conclude, the Naṣīḥat al-Mulūk, an advisory text composed in the Samanid domains probably towards the middle of the tenth century, attests on the one hand to the semi-sacralised representation of kingship and the exclusive place of kings in the divine order, and on the other to the indirect and contingent nature of governance in the Samanid polity. It illustrates the author’s perception of a political culture that, while intensely hierarchical in its representation, entailed the amirs’ dependence on networks that stretched across their domains, and in which a variety of local figures performed essential mediating functions. Pseudo-Māwardī distinguishes between groupings composed of individuals who were bound to the ruler through ties of personal loyalty and obligation, and categories whose voluntary cooperation and support the ruler was obliged to cultivate. He writes at a moment when it was still possible to imagine and promote human mechanisms by which to bridge the chasm between the ruler and the population; the personal involvement of rulers in matters of governance, such as the convening of courts of maẓālim and leadership of military campaigns, remained within living memory.¹⁰⁰ Finally, it might be noted that, in articulating this perception, the author refers not to Abbasid models but to the Prophet’s practice and the customs of the sagacious kings of the past. He emerges as a strong advocate of the mode of governance associated with the earlier generation of Samanid amirs, exemplified by the three sons of Aḥmad, Naṣr I, Ismā‘īl and Isḥāq, whose ‘horizontal’ style of leadership, as Treadwell has suggested, intentionally blurred the boundaries between the amir’s household and the leaders of the community.¹⁰¹ His presentation attests to a distinctive and confident regional culture with deep roots in Islam and universal wisdom.

Notes

1On these categories, see Roy P. Mottahedeh, Loyalty and Leadership in an Early Islamic Society (Princeton NJ, 1980), pp. 115–16, 121–2 – or the second edition (London, 2001), pp. 115–16, 121–2; Bernard Lewis, The Political Language of Islam (Chicago and London, 1988), p. 67; M. A. J. Beg, ‘al-Khāṣṣa waʾl-‘āmma’, EI², iv, pp. 1098–90. The present discussion does not address the usage of the term khāṣṣa to denote the king’s private property.

2Cf. Axel Havemann, Riʾāsa und qaḍāʾ: Institutionen als Ausdruck wechselnder Kräfteverhältnisse in syrischen Städten vom 10. bis 12. Jahrhundert (Freiburg, 1975), pp. 18–19.

3That the Naṣīḥat al-Mulūk does not belong to the oeuvre of al-Māwardī has been established by Fuʾad ‘Abd al-Mun‘im Ahmad, in the essay that precedes his critical edition of the text: ‘Muqaddimat al-taḥqīq wa-l-dirāsa’, Naṣīḥat al-Mulūk al-mansūb ilā Abī l-Ḥasan al-Māwardī (Alexandria, 1988), pp. 5–33; and also in a separate monograph, Abū l-Ḥasan al-Māwardī wa-kitāb Naṣīḥat al-Mulūk, (Alexandria, [n. d.]). On the likely location and dating for the work’s composition, see, in addition to Ahmad, ‘Taḥqīq’, Louise Marlow, ‘A Samanid work of counsel and commentary: The Naṣīḥat al-Mulūk of Pseudo-Māwardī’, Iran xliv (2007), pp. 181–92; and at greater length, Wisdom and Politics in Tenth-Century Iran (Edinburgh University Press, forthcoming), Vol. 1, chapters one and two. See also Hassan Ansari, ‘Yak andīsha-nāma-yi siyāsī-yi arzishmand-i mu‘tazilī az Khurāsān dawrān-i Sāmānīyān’, Bar-rasī-hā-yi tārīkhī, http://ansari.kateban.com/entryprint1951.html (accessed 3 October, 2013); Ansari’s conclusions coincide to a large extent with those of the present author.

4Al-Khwārazmī, Mafātīḥ al-‘ulūm, ed. G. van Vloten, Liber Mafâtîh al-olûm (Leiden, 1968), p. 132.

5On the ‘Kindian tradition’, see Peter Adamson, ‘The Kindian tradition: The structure of philosophy in Arabic Neoplatonism’, in Cristina D’Ancona (ed.), The Libraries of the Neoplatonists (Leiden, 2007), pp. 351–70, especially p. 352 and n. 3.

6Perhaps first explored in W. Barthold, Turkestan Down to the Mongol Invasion, fourth edition, (London, 1977), pp. 226–32. See further W. L. Treadwell, ‘The Political History of the Sāmānid State’, Unpublished D.Phil thesis, Oxford, 1991, pp. 121–30.

7Fuʾad ‘Abd al-Mun‘im Ahmad, Naṣīḥat al-Mulūk al-mansūb ilā Abī l-Ḥasan al-Māwardī (Alexandria, 1988), p. 211.

8In their coinage and in their posthumous regnal titles, the Samanids adopted the title of Amīr. The term malik does not appear in their titulature until the reign of Nūḥ b. Naṣr (r. 331–43/943–54), who assumed the title al-Malik al-Muʾayyad. Many contemporary sources refer to the Samanid rulers as mulūk, however, and Pseudo-Māwardī states that malik was the term in most common use in his milieu (Naṣīḥat al-Mulūk, p. 66).

9Cf. Patricia Crone, God’s Rule: Government and Islam: Six Centuries of Medieval Islamic Political Thought (New York, 2004), pp. 153–4, 252–4.

10Naṣīḥat al-Mulūk, p. 203.

11Mottahedeh, Loyalty and Leadership, pp. 178–9.

12Naṣīḥat al-Mulūk, p. 205. Probably in an indication of his acquaintance with the Kindian tradition, Pseudo-Māwardī likens the king’s khāṣṣa to the artisan’s tool and the ‘āmma to raw matter in more than one location; see further Naṣīḥat al-Mulūk, pp. 66–7, 255.

13Jürgen Paul, Herrscher, Gemeinwesen, Vermittler: Ostiran und Transoxanien in vormongolischer Zeit (Beirut, 1996). Paul has especially indicated the areas of irrigation, taxation and armed action (related to military and police matters) in which mediation and co-operation between the structures of power and the population were essential (pp. 31–66, 66–92, 93–139).

14Naṣīḥat al-Mulūk, p. 205.

15Pseudo-Māwardī, like al-Khwārazmī, employs the term jund, ‘more characteristic of early Islamic Syria’, in a Samanid context (cf. C. E. Bosworth, ‘Abū ‘Abdallāh al-Khwārazmī on the technical terms of the secretary’s art: A contribution to the administrative history of mediaeval Islam’, Journal of the Economic and Social History of the Orient xii [1969], pp. 113–64, 145). On the term asāwira in the Samanid context, see also al-Khwārazmī, Mafātīḥ al-‘ulūm, p. 115 (jam‘ al-uswār wa-huwa al-fāris).

16Naṣīḥat al-Mulūk, pp. 208–9.

17Ibid, pp. 226, 237.

18Ibn Ḥawqal, Kitāb Ṣūrat al-Arḍ, ed. J. H. Kramers, (Leiden, 1938–9), p. 491.

19C. E. Bosworth, The Ghaznavids: Their Empire in Afghanistan and Eastern Iran 994–1040 (Edinburgh, 1963), p. 34.

20Niẓāmī ‘Arūḍī Samarqandī, Chahār Maqāla, ed. Muhammad Mu‘in (Tehran, 1336 [1957]), pp. 22–4; cf. E. G. Browne, A Literary History of Persia, Vol. II: From Firdawsí to Sa‘dí (Cambridge, 1928), pp. 15–17.

21David Ayalon, ‘On the eunuchs in Islam’, Jerusalem Studies in Arabic and Islam i (1979), pp. 67–124, esp. pp. 74–89; Ayalon, Outsiders in the Lands of Islam: Mamluks, Mongols and Eunuchs (London, 1988), section III.

22On the institution of khidma, see Herrscher, Gemeinwesen, Vermittler, pp. 166–7, n. 18, 173, and Jürgen Paul, ‘Khidma in the social history of pre-Mongol Iran’, Journal of the Economic and Social History of the Orient lvii/3 (2014), pp. 392-422. I am grateful to the author for providing me with a copy of this article before its publication. See also Mottahedeh, Loyalty and Leadership, pp. 82–96.

23In other contemporary contexts, the term fityān designates free soldiers; cf. D.G. Tor, Violent Order: Religious Warfare, Chivalry, and the ‘Ayyār Phenomenon in the Medieval Islamic World, (Würzburg, 2007), esp. pp. 231–51. In the present context, as a later passage indicates, the linkage of the fityān with ghilmān appears to suggest that the categories shared a common status.

24See D.G. Tor, ‘Privatized jihad and public order in the pre-Seljuq period: The role of the Mutatawwi‘a’, Iranian Studies xxxviii (2005), pp. 555–73.

25Cf. Mottahedeh, Loyalty and Leadership, pp. 72–82 and passim.

26Narshakhī, Tārīkh-i Bukhārā, tr. Abū Naṣr Aḥmad b. Muḥammad b. Naṣr al-Qubāwī, ed. Mudarris Raḍawī, (Tehran, 1351 [1984)], p. 36.

27Naṣīḥat al-Mulūk, pp. 209–21.

28Mottahedeh, Loyalty and Leadership, p. 115.

29Naṣīḥat al-Mulūk, p. 225.

30The narrative of a Chinese embassy to the Samanid territories during the reign of Naṣr b. Aḥmad refers explicitly to ‘Turkish’ ghilmān (al-Qāḍī al-Rashīd Ibn al-Zubayr, Kitāb al-Dhakhāʾir wa-l-tuḥaf, ed. M. Hamidullah, (Kuwait, 1984), p. 145 [alf ghulām turkī], p. 146 [miʾa ghulām turkī]), perhaps as if to distinguish them from other groups of ghilmān.

31The Kitāb al-Dhakhāʾir also refers to kull ghulām mamlūk ṣaghīr wa-kabīr, possibly in order to specify mamlūk status of the ghilmān in the case in question (al-Dhakhāʾir, p. 141); Bosworth translates the phrase as ‘every military slave … young and old’, ‘An alleged embassy from the Emperor of China to the Amir Naṣr b. Aḥmad: a contribution to Sâmânid military history’, in M. Minuvi and Iraj Afshar (eds), Yadnameh-yi Iranu-yi Minurski (Tehran, 1969), pp. 17–29, 19.

32Cf. C. I. Beckwith, ‘The early history of the Central Asian guard corps in Islam’, Archivum Eurasiae Medii Aevi iv (1984), pp. 29–43; Étienne de la Vaissière, Samarcande et Samarra: Élites d’Asie Centrale dans l’empire abbasside, (Paris, 2007), pp. 59–88, esp. pp. 68–77.

33Naṣīḥat al-Mulūk, p. 377.

34Siyar al-Mulūk (Siyāsatnāma), ed. Hubert Darke, (Tehran, 1347/1962), p. 297 = Hubert Darke, The Book of Government or Rules for Kings, second edition, (London, 1978), p. 220.

35The extent of this replacement and the factors involved in the transformation are by no means self-evident, however; see Jürgen Paul, The State and the Military: The Samanid Case (Bloomington IN, 1994), esp. pp. 20–30; Treadwell, ‘Political History’, pp. 131–2; D.G. Tor, ‘The Mamluks in the military of the pre-Seljuq Persianate dynasties’, Iran xlvii (2008), pp. 213–25, esp. pp. 214–15.

36Mottahedeh, Loyalty and Leadership, pp. 84–9. On the military training of the ghilmān under the Samanids, see de la Vaissière, Samarcande et Samarra, pp. 262–5.

37Siyar al-Mulūk, 174 = 127, 141 = 103–4, cf. 166–8 = 121–3. Bosworth has observed that Niẓām al-Mulk’s elaborate account was likely to represent ‘an ideal rather than an actuality’ (Bosworth, Ghaznavids, p. 102); see further ‘The Pand-nāmah of Subuktigīn’, ed. M. Nazim, Journal of the Royal Asiatic Society (1933), pp. 605–28, 613, 623, and Bosworth, ‘An alleged embassy’, pp. 17–29.

38See W. L. Treadwell, ‘Ibn Ẓāfir al-Azdī’s account of the murder of Aḥmad b. Ismā‘īl and the succession of his son Naṣr’, in C. Hillenbrand (ed.), Studies in Honour of Clifford Edmund Bosworth, Vol. II: The Sultan’s Turret: Studies in Persian and Turkish Culture (Leiden, 2000), pp. 397–419, and Luke Treadwell, ‘Urban militias in the eastern Islamic world (third-fourth centuries AH/ninth-tenth centuries CE)’, in Teresa Bernheimer and Adam Silverstein (eds), Late Antiquity: Eastern Perspectives (Exeter, 2012), pp. 128–44, 133–5.

39Naṣīḥat al-Mulūk, p. 225.

40Ibid, pp. 225–38.

41Ibid, pp. 238–9.

42Ibid, p. 239.

43On the topography and urban design of Bukhara, including the Rigistan, see Ibn Ḥawqal, Kitāb Ṣūrat al-Arḍ, pp. 482–5, 491.

44Barthold rendered the phrase as ‘captain of the guard’ (Turkestan, pp. 229, 230), Frye as ‘chief of the guards’ – see Richard N. Frye, Al-Narshakhi’s The History of Bukhara (Princeton, 2007), p. 31. Cf. Havemann, Riʾāsa und qaḍāʾ, pp. 51–7.

45The phrase remains unidentified; Frye, apparently following Barthold (Turkestan, p. 229), suggests the reading dīwān-i barīd and renders the office as ‘bureau of the postmaster’. See R. N. Frye, ‘The Sāmānids’, in R. N. Frye (ed.), The Cambridge History of Iran, Vol. IV: From the Arab Invasion to the Saljuqs (Cambridge, 1975), pp. 136–61, 144; Frye, History of Bukhara, p. 31.

46The text reads dīwān-i sharaf; the proposed emendation follows Bosworth, ‘Technical terms of the secretary’s art’, p. 142, Ghaznavids, p. 29, and Frye, ‘Sāmānids’, p. 144. Cf. Barthold, Turkestan, pp. 229, 231 (‘Dīwān of the Mushrifs’).

47Narshakhī, Tārīkh-i Bukhārā, p. 36.

48Mafātīḥ al-‘ulūm, p. 53.

49Cited in (Pseudo)-Tha‘ālibī, Tuḥfat al-Wuzarāʾ al-mansūb ilā Abī Manṣūr ‘Abd al-Malik b. Muḥammad b. Ismā‘īl al-Tha‘ālibī, eds Ḥ. ‘Alī al-Rāwī and I. M. al-Ṣaffār, (Baghdad, 1977), pp. 81–2.

50Cf. Bosworth, ‘Technical terms of the secretary’s art’, p. 118.

51Adam J. Silverstein, Postal Systems in the Pre-modern Islamic World (Cambridge, 2007), pp. 125–31; Barthold, Turkestan, pp. 230–1; Treadwell, ‘Political History’, pp. 126–7.

52Cf. Bosworth, ‘Technical terms of the secretary’s art’, p. 127.

53Naṣīḥat al-Mulūk, p. 244.

54Silverstein, Postal Systems, pp. 114–15.

55Bosworth, Ghaznavids, pp. 93–6.

56See above, nn. 45 and 46.

57Barthold surmised that the mushrif, ‘observer’, in charge of ishrāf, oversaw the sums allotted for the maintenance of the court, and Frye translates the term as ‘chief of protocol’ (Turkestan, p. 231; Frye, History of Bukhara, p. 31).

58Nazim, ‘Pand-nāmah’, p. 620, 627; Siyar al-Mulūk, pp. 63–78, 85–96, 101–16 = 47–59, 63–71, 74–87.

59Barthold suggested that the ‘āriḍ operated within the dīwān-i ṣāḥib-i shuraṭ, the office of the ‘captain of the guard’, to whom he was subordinate (Turkestan, p. 230); Bosworth suggested that ‘unlike those departments with fixed quarters, the Department of the Army was peripatetic and followed the troops on their campaigns’, and accordingly lacked permanent premises in the Rigistan (Bosworth, ‘Technical terms of the secretary’s art’, pp. 117–18); Mudarris Raḍawī suggested that the office of the ‘āriḍ had not yet come into existence (Tārīkh-i Bukhārā, p. 213); Treadwell, observing the absence from Narshakhī’s list of two departments, the dīwān-i lashkar (jaysh) and the dīwān-i ṣāḥib al-shuraṭ, responsible for army and police respectively, considered Barthold’s suggestion that the functions of both were confined to the dīwān-i ṣāḥib al-shuraṭ ‘quite reasonable’ (Treadwell, ‘Political History’, p. 128, n. 115).

60Frye notes the use in Arabic of the term ṣāḥib juyūsh for sipah-sālār (Frye, ‘The Sāmānids’, p. 143). In his narrative of the Chinese embassy, Ibn al-Zubayr lists, and apparently distinguishes between, the juyūsh of Farghana and the ‘volunteers’ (muṭṭawwi‘a), although he refers to the ‘gathering’ of both categories (al-Dhakhāʾir, p. 141).

61Paul, Herrscher, Gemeinwesen, Vermittler, p. 94.

62Alistair Northedge, ‘An interpretation of the palace of the Caliph at Samarra (Dar al-Khilafa or Jawsaq al-Khaqani)’, Ars Orientalis 23, Pre-Modern Islamic Palaces (1993), pp. 143–70, 147 (with reference to al-Ya‘qūbī, Kitāb al-Buldān , [Leiden, 1891], p. 240).

63Siyar al-Mulūk, pp. 181–6; Bosworth, Ghaznavids, p. 138.

64Paul, Herrscher, Gemeinwesen, Vermittler , p. 70, n. 11.

65Ḥudūd al-‘ālam. ‘The Regions of the World’: A Persian Geography 372 A.H.–982 A.D., tr. V. Minorsky, second edition, (Cambridge, 1970), p. 102.

66The term khalīfa also, of course, means ‘successor’; when Nūḥ b. Asad died, he made his brother Aḥmad b. Asad his khalīfa, and when Aḥmad died, he made his son Naṣr khalifa, all of them governing in Samarqand (Tārīkh-i Bukhārā, pp. 105–6).

67See D.G. Tor, ‘The Islamization of Central Asia in the Sāmānid era and the reshaping of the Islamic world’, Bulletin of the School of Oriental Studies, lxxii (2009), pp. 279–99.

68Naṣīḥat al-Mulūk, pp. 106–8; cf. Marlow, ‘Counsel and commentary’.

69Tārīkh-i Bukhārā, pp. 110, 130. For further examples, including ninth-century cases in which a khalīfa presided over the funeral prayers of a qadi and a ghazi, see J. Paul, ‘The histories of Samarqand’, Studia Iranica xxii (1993), pp. 69–92, 88, citing the Istanbul MS of the Arabic Qandiyya.

70Naṣīḥat al-Mulūk, p. 243. The formula perhaps evokes the combined governmental, military and fiscal responsibilities of the early ‘ummāl (Havemann, Riʾāsa und qaḍāʾ, pp. 38–45).

71Ibn Ḥawqal, Kitab Ṣūrat al-Arḍ, p. 424 (kātib salla yu‘raf bi-l-bundār yuṭālib bi-l-kharāj wa-wujūh al-amwāl al-wājiba lil-sulṭān) = La configuration de la terre (Kitab surat al-ard), tr. J. H. Kramers and G. Wiet, (Paris, 1964), p. 411. Cf. Treadwell, ‘Political History’, p. 125, n. 104; Barthold, Turkestan, pp. 229–30.

72Ibn Ḥawqal, Kitab Ṣūrat al-Arḍ, pp. 424, 469–70. See P. Crone, ‘Ma‘ūna’, EI², vi, p. 848; Silverstein, Postal Systems, pp. 101–2.

73Cf. Paul, Herrscher, Gemeinwesen, Vermittler, pp. 86, 73–4, n. 19. On the practices and personnel involved in the collection of revenue, see further pp. 66–92.

74Treadwell, ‘Political History’, p. 125, n. 104; cf. Barthold, Turkestan, pp. 229–30.

75Treadwell, ‘Political History’, p. 129.

76Al-Tha‘ālibī, Yatīmat al-Dahr fi maḥāsin ahl al-‘aṣr, ed. Mufīd Muḥammad Qumayḥa (Beirut, 1983), iv, p. 73 (Abū Aḥmad b. Abī Bakr al-Kātib), p. 87 (Abū Manṣūr al-‘Abdūnī), p. 90 (Abū l-Ṭayyib al-Muṣ‘abī), p. 96 (Abū Bakr Muḥammad b. ‘Uthmān al-Naysābūrī al-Khāzin), p. 103 (Abū Muḥammad al-Sulamī), p. 108 (Abū ‘Alī al-Sallāmī), pp. 108, 112 (Abū l-Qāsim ‘Alī b. Muḥammad al-Iskāfī al-Naysūbūrī). In certain cases, as attested elsewhere for the early Samanid period, the term kātib overlaps with or is even synonymous with the term wazīr.

77Mafātīḥ al-‘ulūm, p. 68; see further Paul, Herrscher, Gemeinwesen, Vermittler , pp. 31–66, esp. pp. 55–65.

78Naṣīḥat al-Mulūk, pp. 240–1.

79Siyar al-Mulūk, p. 119 (Wakīl-i khāṣṣ) = p. 88 (‘the steward of the household’); Bosworth, Ghaznavids, pp. 68–9, 138; Barthold, Turkestan, p. 229; Frye, ‘The Sāmānids’, p. 144.

80Treadwell, ‘Political History’, pp. 129–30 and notes; cf. Barthold, Turkestan, pp. 227–8.

81Naṣīḥat al-Mulūk, p. 143.

82Examples among the Samanid dynastic family include the brothers Isḥāq b. Aḥmad (al-Sam‘ānī, al-Ansāb (Hyderabad, 1382–1402/1962–82, vii [1396/1976]), p. 25), Ismā‘īl b. Aḥmad (Siyar al-Mulūk, pp. 28–9 = 21–2), and both brothers together (al-Khaṭīb al-Baghdādī, Taʾrīkh Baghdād, iii, [Cairo, 1931] p. 318). Other figures who presided over courts for the redress of grievances included, in Nishapur, the governor Muḥammad b. ‘Abd al-Razzāq (Tārīkh-i Gardīzī, ed. ‘Abd al-Ḥayy Ḥabībī, [Tehran, 1363/1984–5], p. 353), and in Balkh an individual named Abū ‘Alī al-Misbakhī (al-Tha‘ālibī, Yatīmat al-dahr Yatimat al-Dahr, iv, pp. 167–8).

83Siyar al-Mulūk, p. 117 (naqībān, specifically identified with ‘the custom of the past’ [‘ādat-i guzashteh]) = p. 87 (‘sergeants’); cf. Silverstein, Postal Systems, p. 130.

84A. Havemann, ‘Naḳīb al-ashrāf ’, EI², vii, pp. 926–7. See also Richard W. Bulliet, The Patricians of Nishapur: A Study in Medieval Islamic Social History (Cambridge MA, 1972), esp. pp. 234–40; Bosworth, Ghaznavids, pp. 196–8.

85The evidence of the Tārīkh-i Bayhaq speaks especially eloquently of the prominence of ‘Alid families; see Ibn Funduq, Tārīkh-i Bayhaq, ed. Ahmad Bahmanyar, (Tehran, 1960), pp. 54–65; 168, 169, 170, 179, 180, 186, 190, 221, 231, 232, 246, 250, 253–5, 284–6 and passim.

86Paul, Herrscher, Gemeinwesen, Vermittler, p. 123; Teresa Bernheimer, ‘The rise of sayyids and sādāt: The Āl Zubāra and other ‘Alids in ninth- and tenth-century Nishapur’, Studia Islamica c/ci (2005), pp. 43–69.

87It is possible that, in a parallel to the khazāʾin and buyūt of the following clause, the author’s manzila and manāzil refer to physical residences rather than social stations.

88I am grateful to Christopher Melchert, who advises that the phrase dhawūʾl-ārāʾ is likely to signify ‘political advisers’; cf. Christopher Melchert, ‘Traditionist-Jurisprudents and the framing of Islamic law’, Islamic Law and Society viii (2001), pp. 383–406, 386–7.

89Niẓām al-Mulk would write that senior officers should prepare appurtenances, weapons and the tools of war and buy slaves, for their stature lies in these possessions rather than in the appurtenances, equipment and adornment of the household (ma‘rūfān-rā … bibāyad guft tā tajammul wa-silāḥ wa-ālat-i jang sāzand wa-ghulām kharand kih jamāl wa-shikūh-i īshān andar īn chīz-hā būd na andar tajammul wa-ālat wa-zīnat-i khāna) (Siyar al-Mulūk, p. 165 = 121).

90Naṣīḥat al-Mulūk, pp. 253–4. See also Naṣīḥat al-Mulūk, pp. 264–5.

91Al-Iṣṭakhrī, Masālik al-Mamālik, ed. M. J. de Goeje, (Leiden, 1870), p. 290; Ibn Ḥawqal, Kitāb Ṣūrat al-Arḍ, p. 466.

92Al-Iṣṭakhrī, Masālik al-Mamālik, pp. 291–2; Ibn Ḥawqal, Kitāb Ṣūrat al-arḍ, p. 468.

93Tārīkh-i Sīstān, ed. Malik al-shu‘arāʾ Bahār, (Tehran, 1314 [1935]), p. 319; cf. Ahmad Tafażżoli, ‘Dehqān’, EIr, vii, pp. 223–5. Cf. Jürgen Paul, ‘Nachrichten arabischer Geographen aus Mittelasien’, Bamberger Mittelasienstudien; Konferenzakten, Bamberg 15–16. Juni 1990 (Berlin, 1994), pp. 179–91, 184–5. In ‘Where did the Dihqāns go?’ Eurasian Studies xi (2013), pp. 1–34, Jürgen Paul has argued that even after the eclipse of the term dihqan, the persons to whom it had applied persisted and continued to exercise authority; I am grateful to the author for making this article available to me prior to its publication. On Aḥmad b. Sahl, see Tārīkh-i Gardīzī, pp. 332–4; Ibn al-Athīr, al-Kāmil fī l-Taʾrīkh, ed. C. Tornberg, (Beirut, 1965–7), viii, pp. 118–19; Jürgen Paul, ‘Aḥmad b. Sahl’, EI³, i, pp. 60–1; Treadwell, ‘Political History’, p. 111 and n. 40, and Appendix One, ‘Aḥmad ibn Sahl ibn Hāshim al-Kāmkārī al-Marwazī’, pp. 313–14; Bosworth, ‘Aḥmad b. Sahl b. Hāšem’, EIr, i (1985), pp. 643–4.

94Paul, ‘Histories of Samarqand’, pp. 82–7; Paul, The State and the Military, esp. pp. 20–30; Tor, ‘Privatized Jihad’, pp. 555–73.

95Treadwell, ‘Ibn Ẓāfir’, pp. 141 = 155, 139 = 154; Narshakhī, Tārīkh-i Bukhārā, p. 103.

96Bulliet, Patricians of Nishapur, p. 234.

97Urban populations, described as ‘āmma or ahl, not only participated in revolts, but also rose to the defence of their cities and intervened in political contests; for examples, see Paul, Herrscher, Gemeinwesen, Vermittler, pp. 201–4; 94, n. 2; 96–7; 119; cf. 210.

98Ibid, pp. 6–7, 237–51.

99Naṣīḥat al-Mulūk, p. 257.

100On the theme of direct access to the ruler, with particular reference to its prominent treatment in the Siyar al-Mulūk, see Paul, Herrscher, Gemeinwesen, Vermittler, pp. 148–56.

101Luke Treadwell, ‘The Samanids: The first Islamic dynasty of Central Asia,’ in Edmund Herzig and Sarah Stewart (eds), The Idea of Iran, Vol. V: Early Islamic Iran (London, 2012), pp. 3–15, 8.

Bibliography

Aḥmad, Fuʾād ‘Abd al-Mun‘im, ‘Muqaddimat al-taḥqīq wa-l-dirāsa’, Naṣīḥat al-Mulūk al-mansūb ilā Abī l-Ḥasan al-Māwardī (Alexandria, 1988), pp. 5–33.

Adamson, Peter, ‘The Kindian tradition: The structure of philosophy in Arabic Neoplatonism’, in Cristina D’Ancona (ed.), The Libraries of the Neoplatonists (Leiden, 2007), pp. 351–70.

Ansari, Hassan, ‘Yak andīsha-nāma-yi siyāsī-yi arzishmand-i mu‘tazilī az Khurāsān dawrān-i Sāmānīyān’, Bar-rasī-hā-yi tārīkhī, http://ansari.kateban.com/entryprint1951.html (accessed 3 October, 2013).

Ayalon, David, ‘On the eunuchs in Islam’, Jerusalem Studies in Arabic and Islam i (1979), pp. 67–124 = Outsiders in the Lands of Islam: Mamluks, Mongols and Eunuchs, London, Variorum, 1988, III.

al-Baghdādī, al-Khaṭīb, Taʾrīkh Baghdād aw Madīnat al-salām, 14 Vols, (Cairo, 1931).

Barthold, W., Turkestan Down to the Mongol Invasion, fourth edition, (London, 1977).

Beckwith, C. I., ‘The early history of the Central Asian guard corps in Islam’, Archivum Eurasiae medii aevi iv (1984), pp. 29–43.

Beg, M. A. J., ‘al-Khāṣṣa waʾl-‘āmma’, EI2, iv, pp. 1098–90.

Bernheimer, Teresa, ‘The rise of sayyids and sādāt: The Āl Zubāra and other ‘Alids in ninth- and tenth-century Nishapur’, Studia Islamica c/ci (2005), pp. 43–69.

Bosworth, C. E., The Ghaznavids: Their Empire in Afghanistan and Eastern Iran 994–1040 (Edinburgh, 1963).

Bosworth, C. E., ‘Abū ‘Abdallāh al-Khwārazmī on the technical terms of the secretary’s art: A contribution to the administrative history of mediaeval Islam’, Journal of the Economic and Social History of the Orient xii (1969), pp. 113–64.

Bosworth, C. E., ‘An alleged embassy from the Emperor of China to the Amir Naṣr b. Aḥmad: a contribution to Sâmânid military history’, in M. Minuvi and Iraj Afshar (eds), Yādnāmeh-yi Īrānī-yi Mīnūrskī, (Tehran, 1969), pp. 17–29.

Bosworth, C. E., ‘Aḥmad b. Sahl b. Hāšem’, EIr, i, pp. 643–4.

Browne, E. G., A Literary History of Persia, 4 vols (Cambridge, 1906–1928).

Bulliet, Richard W., The Patricians of Nishapur: A Study in Medieval Islamic Social History (Cambridge, MA, 1972).

Crone, Patricia., ‘Ma‘ūna’, EI², vi, p.848.

Crone, Patricia, God’s Rule: Government and Islam: Six Centuries of Medieval Islamic Political Thought (New York, 2004).

Frye, R. N., ‘The Sāmānids’, in R.N. Frye (ed.) The Cambridge History of Iran, Vol. IV: From the Arab Invasion to the Saljuqs (Cambridge, 1975), pp. 136–61.

Gardīzī, Tārīkh-i Gardīzī, ed. ‘Abd al-Hayy Habibi, (Tehran, 1363 [1984–5]).

Havemann, Axel, Riʾāsa und qaḍāʾ: Institutionen als Ausdruck wechselnder Kräfteverhältnisse in syrischen Städten vom 10. bis 12. Jahrhundert, (Freiburg, 1975).

Havemann, Axel, ‘Naḳīb al-ashrāf ’, EI², vii, pp. 926–7.

Ḥudūd al-‘ālam: ‘The Regions of the World’: A Persian Geography 372 A.H.–982 A.D., trans. V. Minorsky, second edition, (Cambridge, 1970).

Ibn al-Athīr, al-Kāmil fī l-Taʾrīkh, ed. C. J. Tornberg, (Leiden, 1862); reprinted Beirut, 1965–7.

Ibn Funduq, Tārīkh-i Bayhaq, ed. Ahmad Bahmanyar, (Tehran, 1960).

Ibn Ḥawqal, Kitāb Ṣūrat al-arḍ, ed. J. H. Kramers, (Leiden, 1938–9) = La configuration de la terre (Kitab Surat al-Ard), tr. J. H. Kramers and G. Wiet, (Paris, 1964).

al-Qāḍī al-Rashīd Ibn al-Zubayr, Kitāb al-Dhakhāʾir wa-l-tuḥaf, ed. M. Hamidullah, (Kuwait, 1984).

al-Iṣṭakhrī, Masālik al-Mamālik, ed. M. J. de Goeje, (Leiden, 1870).

al-Khwārazmī, Mafātīḥ al-‘ulūm, ed. G. van Vloten, Liber Mafâtîh al-olûm, (Leiden, 1968).

Lewis, Bernard, The Political Language of Islam (Chicago & London, 1988).

Marlow, Louise, ‘A Samanid work of counsel and commentary: The Naṣīḥat al-Mulūk of Pseudo-Māwardī’, Iran, xliv (2007), pp.181–92.

Marlow, Louise, Wisdom and Politics in Tenth-Century Iran, Vol. I: The Naṣīḥat al-Mulūk of Pseudo-Māwardī: Contexts and Themes; Vol. II: The Naṣīḥat al-Mulūk of Pseudo-Māwardī: Texts, Sources and Authorities (Edinburgh, [forthcoming]).

Melchert, Christopher, ‘Traditionist-Jurisprudents and the framing of Islamic law’, Islamic Law and Society, viii, (2001), pp. 383–406.

Mottahedeh, Roy P., Loyalty and Leadership in an Early Islamic Society (Princeton, 1980) = Roy P. Mottahedeh, Loyalty and Leadership in an Early Islamic Society, second edition, (London, 2001).

Narshakhī, Tārīkh-i Bukhārā, tr. Abū Naṣr Aḥmad b. Muḥammad b. Naṣr al-Qubāwī, ed. Mudarris Razavi, (Tehran, 1351 [1984]) = Frye, Richard N., Al-Narshakhi’s The History of Bukhara (Princeton, 2007).

Naṣīḥat al-Mulūk al-mansūb ilā Abī l-Ḥasan al-Māwardī, ed. Fuʾad ‘Abd al-Mun‘im Ahmad, (Alexandria, 1988).

Nazim, M., ‘The Pand-nāmah of Subuktigīn’, Journal of the Royal Asiatic Society (1933), pp. 605–28.

Niẓām al-Mulk, Siyar al-Mulūk (Siyāsatnāma), ed. Hubert Darke, (Tehran, 1347/1962) = The Book of Government or Rules for Kings, trans. Hubert Darke, second edition, (London, 1978).

Niẓāmi ‘Arūḍī Samarqandī, Chahār Maqāla, ed. Muḥammad Mu‘in, (Tehran 1336/1952).

Northedge, Alistair, ‘An interpretation of the palace of the Caliph at Samarra (Dar al-Khilafa or Jawsaq al-Khaqani)’, Ars Orientalis, xxiii, Pre-Modern Islamic Palaces (1993), pp. 143–70.

Paul, Jürgen, ‘The histories of Samarqand’, Studia Iranica, xxii (1993), pp. 69–92.

Paul, Jürgen, ‘Nachrichten arabischer Geographen aus Mittelasien’, Bamberger Mittelasienstudien; Konferenzakten, Bamberg 15–16. Juni 1990 (Berlin, 1994), pp. 179–91.

Paul, Jürgen, The State and the Military: The Samanid Case (Bloomington IN, 1994).

Paul, Jürgen, Herrscher, Gemeinwesen, Vermittler: Ostiran und Transoxanien in vormongolischer Zeit, (Beirut, 1996).

Paul, Jürgen, ‘Aḥmad b. Sahl’, EI³, i, pp. 60–1.

Paul, Jürgen, ‘Where did the dihqāns go?’ Eurasian Studies, xi (2013), pp. 1–34.

Paul, Jürgen, ‘Khidma in the social history of pre-Mongol Iran’, Journal of the Economic and Social History of the Orient lvii/3 (2014), pp. 392–422.

al-Sam‘ānī, al-Ansāb, (Hyderabad, 1382–1402/1962–82).

Silverstein, Adam J., Postal Systems in the Pre-modern Islamic World (Cambridge, 2007).

Tafażżoli, Ahmad, ‘Dehqān’, EIr, vii, pp. 223–5.

Tārīkh-i Sīstān, ed. Malik al-Shu‘arāʾ Bahār, (Tehran, 1314 [1935]).

al-Tha‘ālibī, Yatīmat al-Dahr fī maḥāsin ahl al-‘aṣr, ed. Mufīd Muḥammad Qumayḥa, (Beirut, 1983).

Tor, D. G., ‘Privatized jihad and public order in the pre-Seljuq period: The role of the Mutatawwi‘a’, Iranian Studies, xxxviii (2005), pp. 555–73.

Tor, D. G., ‘The Mamluks in the military of the pre-Seljuq Persianate dynasties’, Iran, xlvii (2008), pp. 213–25.

Tor, D. G., ‘The Islamization of Central Asia in the Sāmānid era and the reshaping of the Islamic World’, Bulletin of the School of Oriental Studies, lxxii (2009), pp. 279–99.

Tor, D. G., Violent Order: Religious Warfare, Chivalry, and the ‘Ayyār Phenomenon in the Medieval Islamic World (Würzburg, 2007).

Treadwell, W. L., ‘The Political History of the Sāmānid State’, Unpublished DPhil Thesis, Oxford, 1991.

Treadwell, W. L., ‘Ibn Ẓāfir al-Azdī’s account of the murder of Aḥmad b. Ismā‘īl and the succession of his son Naṣr’, in C. Hillenbrand (ed). Studies in Honour of Cliford Edmund Bosworth, Vol II: The Sultan’s Turret: Studies in Persian and Turkish Culture, (Leiden, 2000), pp. 397–419.

Treadwell, Luke, ‘Urban militias in the eastern Islamic world (third-fourth centuries AH/ ninth-tenth centuries CE)’, in Theresa Bernheimer and Adam Silverstein (eds), Late Antiquity: Eastern Perspectives (Exeter, 2012), pp. 128–44.

Treadwell, Luke, ‘The Samanids: The first Islamic dynasty of Central Asia’, in Edmund Herzig and Sarah Stewart (eds), The Idea of Iran, Vol V: Early Islamic Iran (London, 2012), pp. 3–15.

de la Vaissière, Étienne, Samarcande et Samarra: Élites d’Asie Centrale dans l’empire Abbasside (Paris, 2007).

al-Ya‘qūbī, Kitāb al-Buldān (Leiden, 1891).