A eunuch at the threshold: mediating access and intimacy in the Mughal world

Abstract Across the early modern Islamicate world, the phenomenon of eunuch slavery constitutes a significant aspect of courtly contexts and royal households. Although Mughal historiography has focused on the eunuch primarily in relation to the harem, this article analyses the function of such figures in regulating elite male space, in order to explore how these practices shaped both the representation of courtly life as well as the dynamics animating the Mughal court and the inner palace. As is shown in both textual and visual materials, enslaved, castrated men appear as figures both marking and mediating the perimeters of such spaces. In the process they played an important part in the spatial formation of access, intimacy, and hierarchical relations. However, their formative role in mediating elite social interactions at times entangled eunuchs in political conflict. The article concludes with an examination of a particularly dense archive of evidence from the reign of Aurangzeb dealing with royal princes. This material underlines the sometimes-precarious situation of eunuchs in moments of intrafamilial struggle, a fact which suggests the complicated reality of these kinds of intimate roles not only in Mughal princely households but wherever they took on such proximate positions.

palace or imperial camp as a whole can help overcome the sometimes-artificial division between the study of royal men and royal women, effectively segregated into the respective histories of harem and court. Approached from this broader perspective, the article will demonstrate how eunuchs emerge as human border markers, who not only served as liminal figures able to move between certain restricted spaces but who furthermore were instrumental in forming mobile, calibrated boundaries around and between elite men, both within built environments such as the Mughal palace-forts as well as on the road.
While this article will not deal with the many other kinds of posts taken on by eunuchsfrom trusted messenger, to harem guard, to high-ranking noble and adviser-it will provide a key towards understanding the larger significance of these enslaved, castrated figures in the Mughal world from the reign of Akbar (1556-1605) to that of Aurangzeb (1658-1707). 2 Over this period, eunuchs served concrete purposes as attendants and guards to the emperor in private as well as relatively public spaces such as the royal court. It is within this capacity that they emerged as a significant element of imperial iconography in paintings of audience scenes. Yet in contrast to the idealised representation of eunuch roles proximate to the emperor, this article's examination of the evidence on eunuchs within princely households also suggests how such intimate posts and access could embroil such figures in intrafamilial political conflict, a potentiality which becomes especially evident during the long reign of Aurangzeb. Examining this sometimes-uneasy intimacy clarifies both how such mediating positions could shape interactions among the elite, as well as the consequences-both positive and negative-for eunuchs themselves.
Who were these eunuchs who took on such crucial roles? In the South Asian context, the word 'eunuch' itself is used in several ways, a legacy of the colonial-era translation of a broad range of distinct social categories and practices under the umbrella of a single term. 3 On the one hand, the term 'eunuch' can refer to enslaved, castrated men (khwāja sarās) who historically formed a part of elite households, a practice with a long history prior to the Mughals both within and outside of South Asia. 4 It has also been used to refer to members of the still-extant hijra community, who have been characterised variously as trans, third gender, and/or gender-liminal individuals who may or may not undergo castration. 5 It is also encountered as a translation for an incredibly wide range 2 As has been widely discussed in the secondary literature on the Mughal period, Akbar's reign saw the establishment of enduring imperial institutions such as the imperial harem and the mansabdārī system. The article concludes with the reign of Aurangzeb due to the significant shifts in the eighteenth century alongside the dramatic contraction of Mughal domains. 3 Shane Gannon, 'Exclusion as language and the language of exclusion: tracing regimes of gender through linguistic representations of the "eunuch"', Journal of the History of Sexuality 20.1 (2011), pp. 1-27. 4 Slavery of course existed in South Asia from ancient times, and while the evidence on the existence of castration is unclear in this earlier period, there is a much longer history of this practice globally speaking. For an introduction to the long and complicated history of slavery in South Asia, see Indrani Chatterjee and Richard Eaton (eds), Slavery and South Asian History (Bloomington, 2006). For an overview of the longer history of eunuchs, see, for instance, Vern L. Bullough, 'Eunuchs in history and society', in Eunuchs in Antiquity and Beyond, (ed.) Shaun Tougher (Swansea, 2002), pp. 1-18. There are occasional cases in the Mughal context of free adult men being castrated as adults, often for sex crimes. However, the reasons for their castration, the advanced age at which it occurred, as well as their non-slave status set them apart from the castrated, enslaved individuals that form the focus of this article.  5 Descriptions based solely on gender identity or physiological status are inadequate to represent the complexity and variation within this group, but are helpful for understanding the broad distinctions between the primary usages of the term 'eunuch'. For more on the hijra community and its history, see Gayatri Reddy, of terms 'denoting individuals who are not normative in their sexual or gender-role behavior' in Sanskrit and Pali classical texts. 6 This article addresses the first category of eunuchs, who were a highly visible aspect of Mughal society throughout this period and are thus much attested to in contemporary sources. 7 Even as recent work on early colonial Awadh has pointed to conflicting narratives around eunuch masculinity-and in particular occasional elite assertions of 'eunuch effeminacy'-the preponderance of the evidence in the Mughal context, including naming practices, dress, and positions held, tend to suggest that eunuchs were broadly categorised as men, albeit men whose manhood might be viewed as different or inferior. There is also no evidence to suggest that eunuchs viewed themselves as anything else, such as third gender or feminine. 8 Therefore throughout this article I will refer to them as male.
Eunuchs were enslaved as children and castrated before being given or sold into the service of the Mughal elite. Given Mughal bans on this practice with regard to all subject populations within the imperial territories, while some eunuchs were castrated and sold illegally-as is demonstrated by mention of the capture and punishment of such 'eunuchmakers' (khwāja-gars)-many, if not most, eunuchs would have originated from the borderlands of the empire or else further afield. 9 Eunuchs may have shared this starting point, entering elite households as castrated, enslaved boys, but their life trajectories could vary quite significantly, with some eunuchs ultimately entering the nobility and achieving high rank, while others (likely the majority) remained in lower-status posts With Respect to Sex: Negotiating Hijra Identity in South India (Chicago, 2005); Jessica Hinchy, Governing Gender and Sexuality in Colonial India: The Hijra, c. 1850-1900 (Cambridge, 2019); and Jessica Hinchy, 'Hijras and South Asian Historiography', History Compass 20.1 (2022). 6 Michael J. Sweet and Leonard Zwilling, 'The first medicalization: the taxonomy and etiology of queerness in classical Indian medicine', Journal of the History of Sexuality 3.4 (1993), pp. 595-596; see also Gannon, 'Exclusion as language', pp. 8-14. 7 It is important to note that, while a full history of the hijra community prior to the nineteenth century is still to be written, the masculine-presenting eunuchs discussed in this article are understood as ancestors by members of the hijra community today. Vanja Hamzić, Sexual and Gender Diversity in the Muslim World: History, Law and Vernacular Knowledge (London, 2016), pp. 159-160. In fact, the term khwāja sarā has recently been adopted by this community in Pakistan, as having fewer pejorative connotations. For more on this, see Faris A. Khan, 'Translucent citizenship: khwaja sira activism and alternatives to dissent in Pakistan', South Asia Multidisciplinary Academic Journal 20 (2019), pp. 7-8. 8 Emma Kalb, 'Framing gender in Mughal South Asia', History Compass 19.11 (2021), pp. 5-6. See also Emma Kalb, 'Slaves at the Center of Power: Eunuchs in the Service of the Mughal Elite, 1556-1707', (unpublished PhD thesis, University of Chicago, 2020), pp. 21-29. For early colonial Awadh, see Nicholas Abbott, '"In that one the Ālif is missing": eunuchs and the politics of masculinity in early colonial North India', Journal of the Economic and Social History of the Orient 63.1-2 (2019), pp. 73-116. 9 It is worth noting that the regional or ethnic origins of individual eunuchs is generally unspecified in Mughal sources of this period. This silence may be a form of embarrassed avoidance, given the abovementioned prohibitions on the practice of forced castration within Mughal territories, most notably by Jahāngīr, who ordered that the castration of young boys and their sale (kharīd o farūkht) come to an end, and that any violators be punished (tanbīh). Tūzuk-i Jahāngīrī (Lucknow, 1863), p. 73; (trans.) Alexander Rogers (London, 1909(London, -1914 such as guards or pages. 10 Whatever their position, eunuchs were identifiable as such through the physical consequences of the castration procedure, undertaken prior to puberty. The fact that eunuchs lacked facial hair, spoke with unbroken, high-pitched voices, and seem to have experienced distinct forms of bodily fat distribution would have marked them visually and audibly as different-regardless of how that difference was understood and interpreted by contemporaries. 11 As has been widely argued in the larger comparative literature, eunuchs' relative social marginality, alongside their inability to reproduce, served as the basis for the belief in their particular reliability and trustworthiness. 12 This ostensible reliability possibly underwrote the deployment of eunuchs not only within the female space of the harem, but also within male spaces of the palace complex as well. Despite the rich literature on the Mughal state, formal male spaces such as the court have often remained peripheral to historical studies of this period. This article takes its cue from recent scholarship on Islamicate South Asia which has begun working to fill the gaps created by the tendency in prior work to focus on individual emperors or on administrative institutions. 13 While the institution of the harem and its attendant 10 For more on the career trajectories of elite eunuchs during this period, see Bano, 'Eunuchs', pp. 419-422; Kalb, 'Slaves at the Center', Chapters 3 and 4. In the context of eighteenth-century Awadh, Hinchy has also mentioned that while some eunuch children underwent extensive education and training, others remained uneducated and served more menial functions. Jessica Hinchy, 'Enslaved childhoods in eighteenth-century Awadh', South Asian History and Culture 6. questions of space have been more fully explored, 14 scholars are only beginning to examine how the regulation of male space infused politics within the court itself as well as the more restricted areas of the palace. 15 This article will especially build on prior studies of Mughal architecture in considering the significance of the spatial with respect to men, and how eunuchs played a role in forming and regulating spaces such as the audience hall. This scholarship has emphasised how palaces and camps, rather than being haphazard agglomerations of structures, were mindfully designed not only to serve specific concrete functions but also to create calculated impressions upon the mind of the viewer. As Necipoğlu has argued, the Mughals, along with fellow early modern Islamicate empires such as the Safavids and Ottomans, used architecture and ceremony to construct 'distinctive images of absolute kingship', by controlling the gaze of contemporaries through 'framing and staging'. She refers to this as a 'theatrical "display culture"' which used the tool of the gaze to articulateand naturalise-social hierarchies and dynastic legitimacy. 16 This understanding of the palace as theatre, shaping the attention and experience of the viewer, is echoed in Andrews' work on the imperial camp. 17 In this staging of power, the organisation of the palace itself was centred on the person of the emperor, conceived of as the embodied political centre of the empire. 18 This did not express itself in the same way throughout this period, as both the precise arrangement of buildings and the ritual practices within them shifted over the course of the Mughal period. But the secondary literature on the subject emphasises a general principle of spatial differentiation and managed access to the emperor. 19 This formation of social space can be seen to play out in the layout of Mughal palaces and imperial camps as a whole. For example, during the reign of Shāh Jahān (1627-1658) extant records describe a carefully formulated schedule, in which the degree of access to the emperor was calibrated differently according to the particular location. 20 While all of the nobility would be allowed to attend the emperor at the public audience hall (Dīwān-i 'Āmm), only the select few would be permitted to enter the private audience hall (Dīwān-i Khāss). Beyond the private audience hall, the King's Tower (Shāh Burj) constituted a zone where even more private meetings could take place with the emperor's closest confidantes. The last, least accessible, area consisted of the sleeping quarters of the emperor and the harem. 21 Here, then, boundary-building between 'outer' and 'inner' was created through a gradual, layered process. While the specific structures and associated terminologies vary across the period under review, throughout a distinction was made between spaces of public audience, private consultation, and the inner palace. 22 As has been pointed out in comparative contexts, this management of proximate access to the emperor suggests that when it comes to the inner spaces of the palace, the qualities of being forbidden, secluded, and sacred apply not only to the female harem but also to male spaces as well. 23 The restriction of access to elite men such as this can be read as a mark of status and the expression of hierarchical superiority and of control, in addition to the more basic need to ensure the physical safety and security of the ruler. 24 This differentiation encompassed the entire palace complex, from the explicitly 'private' spaces such as the harem or emperor's sleeping quarters, to the public audience hall. This process of space-making can be seen both at the scale of the entire complex as well as at the scale of the individual structure, where space was arranged through the use of platforms or balconies, railings, and other material objects and ritual practices. 25 Eunuchs served as another element in this formation of space, as embodied boundaries and mediators. 26 While Mughal sources tend to be reticent on the particulars of how such spatial divisions were enforced, the evidence regarding the harem gives us some clues. In an oft-quoted normative description from the Ā'īn-i Akbarī, the historian Abū al-Fazl poses the harem as an impeccably administered zone, guarded by four layers of staff: female guards at the innermost post, eunuchs at the threshold, Rajput guards stationed a little further away, and beyond them, porters. 27  the administrative process by which individuals entered or exited, points precisely to the kind of questions of mediation and, indeed, policing discussed in this article. When it comes to the male spaces of the palace, it is known that individuals were scrupulously ordered within the court, and that access to more-private areas such as the privy chamber (referred to as the ghuslkhāna, lit. 'bath-house') were carefully regulated. 28 Yet, unlike Abū al-Fazl's account of the institution of the harem, there is no known guide regarding how access would have been regulated. Nor is there detailed information about the organisation of lower-ranking eunuchs or their titles. 29 Through emphasising the complexity of this spatial regime, this article explores the potential arenas for both negotiation and conflict between the varied elite and non-elite individuals who were present. 30 While the above-described literature on the Mughal court and palace suggests the uses of architecture and ritual to create calculated impressions and normalise certain hierarchies-an understanding which can suggest a unidirectional, one-sided process-this article will emphasise how such spaces contain the possibility of navigation and negotiation for all individuals inside them, albeit within unequal power dynamics. Although the sources dealing with eunuchs in close proximity to the emperor tend to gloss over the personal aspect of such relations, it is likely that such posts led to the advancement of certain eunuchs into the nobility. 31 It is these prominent, highranking eunuchs that the Mughal archive highlights, and whose positions of influence leave significant archival traces. 32 In addition to high-ranking eunuchs, the evidence also suggests hundreds, if not thousands, of eunuchs operating in and around royal and elite households at any given time, the majority of whom would have occupied subordinate posts such as guard or page. 33 28 See Kinra on Chandar Bhān Brahman's account of standing arrangements within the court, where all were required to keep silent and none allowed to move without permission; see also the description of access to the ghuslkhāna: Kinra, Writing Self, Writing Empire, pp. 111 and 115. 29 There is information in the Ā'īn-i Akbarī about the various figures known to be stationed in and around the palace more generally-just without details as to precise placement and the flow of individuals in and out. See, for instance, Ā'īn 6, which discusses the thousand guards appointed to guard the court, and Ā'īn 9, regarding the mounting of the guard (chaukī) around the palace ( perāman-i daulatkhāna). Ā'īn-i Akbarī, pp. 2.188 and 2.192; (trans.) 1.252 and 1.257. 30 The recent work of Flatt, The Courts of the Deccan Sultanates and Kollatz, 'Where is "the audience"?' point to an emerging scholarship on the social dynamics within early modern South Asian courtly milieus. For medieval South Asia, see Daud Ali, Courtly Culture and Political Life in Early Medieval India (Cambridge, 2004). 31 This is implied in passing by Bano, 'Eunuchs', p. 420. Also relevant is the comparative literature on how proximity and intimacy, and specifically access to the ruler's person, allowed for the development of relations of trust between sovereign and servant, while also opening new routes to influence and power during the early modern period. 32 This can be seen in mentions of eunuchs within textual sources, which tend to centre on high-ranking eunuch mansabdārs such as the eunuch Iʻtibār Khān discussed below; it is also evidenced in the limited archive of writings by eunuchs, most notably Bakhtāwar Khān, among whose many works several are still extant. With this in mind, the focus of this article will be on the intimate, mediatory roles taken on by both high-ranked as well as lower-ranked, often unnamed, eunuchs. In examining how eunuchs were critical in the differentiation of space within palaces and camps and in managing this kind of spatial access to the emperor and other elite men, through working as guards, attendants, jailers, and informants, this article will also be attentive to how such roles may have been experienced in contradictory ways. Even as some eunuchs may have advanced into the elite, leveraging their relations with members of the royal family, such intimate service could also entail danger, as they often placed eunuchs directly in the middle of conflicts between elite figures, and in particular of intrafamilial power struggles. 34 The first part of the article, through taking together visual and textual evidence, will trace both the symbolic and functional embodied roles of eunuchs in demarcating and forming the rarefied space around the emperor, which was often represented in a seamless, idealised manner. The second part of the article will examine the more complicated archive of eunuchs in princely households, considering how they were embedded in the political dynamics, and sometimes embroiled in the political conflicts, of the Mughal elite. Taken together, these two strands demonstrate the ongoing significance of these seemingly marginal, lesser-known figures as mediators between elite men and makers of elite space, as well as the ramifications of such roles within the social and political contexts of the palace and court.

Eunuchs and the imperial presence
In 1626, while the imperial camp was being moved across a river, a high-ranking noble named Mahābat Khān staged a coup which would see the emperor held hostage for over six months. 35 At the moment when most of the camp had been shifted over, leaving only the emperor Jahāngīr (r. 1605-1627), a few officials, and a reduced number of guards, Mahābat Khān made his move. On his approach, as eyewitness and historian Muʻtamad Khān relates, Mahābat Khān passed the entrance to the harem, and instead entered through the great gate to dismount in the public audience hall, where there were posted a number of guards. His men passed by these guards and approached three or four eunuchs standing watch in front of the threshold of the privy chamber (ghuslkhāna). However, although they came to the edge of the ghuslkhāna, they did not enter this highly restricted zone, instead simply tearing down the boards that had been put up for safety and privacy and throwing them into the courtyard. The emperor, hearing of the insolence that had been committed, came out of his tent and seated himself in a palanquin that had been prepared for him. Even as the rebel soldiers prepared to take full control of the camp, Mahābat Khān only approached the emperor's palanquin after first performing the ritual salutations and bringing his head to the ground in prostration (kornish and zamīn-bos). Muʻtamad Khān's narrative demonstrates an attentiveness to the differentiated space of the imperial camp which puts the scope of transgression into context. The drama of Mahābat Khān's coup unfolds in real time, as he first enters the main audience hall with his troops and then goes so far as to interfere with the ghuslkhāna's defences. Yet even in this moment of dramatic departure from courtly norms, there is a limit: he does not enter the ghuslkhāna nor any other part of the innermost palace. Rather, the emperor is allowed to come out and take his seat in his palanquin, and Mahābat Khān performs the requisite ritual greetings. In other words, the regime of space is not completely overturned or inverted; it remains tenacious (if not insurmountable) even in this moment of crisis. This event thus reveals the tension between, on the one hand, the rules and practices observed around certain spaces and the body of the emperor, and on the other, the fact of the coup itself, which led ultimately to a full occupation of the premises.
Here, the location of eunuchs is of particular interest. Muʻtamad Khān notes that Mahābat Khān did not turn to enter the harem, but his account attests to a comparable spatial organisation, with a similar layering of guards, progressing from armed men in the main audience hall to eunuchs guarding the ghuslkhāna. This suggests that eunuchs are used here to mark these points of access to the emperor, similar to their function as part of a layered deployment of guards around the harem, with eunuchs placed closer to the women's quarters, at the threshold, and non-castrated male guards at a further distance away. In this section, after briefly examining the evidence for the mundane roles of eunuchs proximate to the emperor, I will consider their appearance in both court audiences as well as imperial paintings as embodied elements serving alongside other visual elements to produce imperial space.
The appearance of eunuchs within the narration of Mahābat Khān's coup is just one example of how eunuchs were stationed in proximate positions close to the emperor and around the more restricted parts of the palace. In textual accounts, eunuchs are described as serving the emperor during mealtimes (perhaps alongside female servants), with non-castrated male servants only allowed to carry food as far as the threshold of the dining hall. 36 In less-exclusive spaces such as the public audience hall, eunuchs are remarked upon as normatively close to the emperor, presenting jewels to him in court, standing behind him holding a fan, and accompanying him as he entered and exited the external viewing balcony (known as the jharokhā balcony) via a small door behind the throne. 37 Sources particular to Aurangzeb's reign relate that he ordered petitions to be received directly from the oppressed (mazlūmān) during his appearance on the jharokhā balcony. 38 Those gathered outside the palace walls were carefully cordoned off, with an official located between them and the palace walls to collect petitions and deposit them in a bag attached to a cord. A eunuch on the balcony, alongside the emperor, would haul these upwards and present them directly to him. 39  other servitors placed at a further distance, in the process embodying the separateness and inviolability of the space surrounding the sovereign. These contemporary accounts are supported by a later sketch of the arrangement of individuals within the court dating from the mid-to late-eighteenth century, claiming to represent courtly norms through the reign of Muhammad Shāh (r. 1718-1748). It notes personal attendants (khawāsān) 40 and palace eunuchs (mahallīyān) present behind the emperor, with the nāzir (eunuch superintendent of the household) also flanking the emperor on stage left. In contrast, other officials and nobles are seen to the left, right, and in front of the throne; the master of the ceremonies (mīr tūzuk) stands in front of the emperor, behind the most powerful Mughal state officials such as the wazīr al-mamālik, with mace-bearers (gurz-bardārs) standing behind them, beside the entrance. 41 In this arrangement, this diagram echoes examples from the broader Islamicate world where eunuchs are similarly represented flanking the ruler in formal contexts. 42 As will be visible in several of the images discussed below, eunuchs would not have been the sole figures given access to such areas, but they were a regular, marked presence within them.
Given this cumulative textual evidence, it is clear that eunuchs, in their posting proximate to the emperor, played a fundamental role in producing space through embodied practice, both in the inner palace as well as in formal public arenas. Eunuchs are described in the texts as taking on practical functions, such as holding fans, passing on petitions, or standing guard. They are not alone in performing these services, as texts mention both female as well as non-eunuch male guards policing the various layers of access within the imperial palace or camp. However, the regular reference to eunuchs specifically-such as in the account of the coup, as well as of Aurangzeb's acceptance of petitions during public appearances in the jharokhā balcony-suggest the particular prominence of such figures.
This information is crucial context for understanding the presence of eunuchs within imperial paintings of court audience (darbār) scenes. While most of the paintings examined here accompanied imperial chronicles and refer to historical personages, they should not be understood as simply visual renderings of historical events, but rather as imperial documents, if not propaganda in their own right, working in parallel with the textual sources consulted above. Even as such manuscript illustrations would not have had the breadth of circulation of written histories, they nonetheless speak to how the world of the court was represented among the most elite. 43 Crafted for this select audience, such works employed forms of imperial iconography to formulate assertions of imperial authority, which inevitably shifted over time; accordingly, 'the imperial image was relentlessly recast'. 44 Within these imperial images, eunuchs play a role in defining the space 40  around the emperor himself as distinct from the larger space of the audience hall occupied by Mughal nobles. As such, eunuchs formed a part of the imperial iconography, as a tool to distinguish the figure of the emperor through both extending the imperial aura and buffering imperial space. This remains the case even as there is significant variation in the style of court scenes over this period, 45 as well as changes in the architectural structures and courtly practices themselves. 46 Thus this section will consider the presence of beardless figures within distinct contexts, ranging from the audience scenes produced during Akbar's reign, which were 'asymmetrical and full of movement', 47 to the highly canonised, symmetrical spatial organisation of darbār scenes from the reign of Shāh Jahān. 48 Given the limitations of space, this article will confine itself to a small number of illustrative examples, beginning with an image of a court audience during the reign of Jahāngīr. This painting was produced during the reign of his successor Shāh Jahān as a part of the Windsor Pādshāhnāma, a lavish imperial copy of Lahorī's history of Shāh Jahān's reign (Figure 1). Unlike most court illustrations, the images included in this work include text that identifies particular figures by name. 49 The subject of this painting is then-Emperor Jahāngīr receiving Prince Khurram, the future Shāh Jahān, as he returns from a military campaign in Mewār. Jahāngīr sits in the upper centre on a raised platform, with Khurram bending down to touch his feet. At a significant distance below, courtiers witness the events taking place above them, arranged according to rank and divided by a gold railing. 50 Behind Jahāngīr on the platform stand three beardless figures, labelled in the painting itself as Iʻtibār Khān, 51 Khidmat Khān, 52 and Fīroz Khān. 53 These labels confirm to us what their hairless faces suggest: that they are eunuchs, identified as such in the textual sources. 54 Some aspects of this painting suggest the need for a highly situated reading, focusing on the particular circumstances of its production. Prior work has argued that the corpus of historical paintings commissioned by Shāh Jahān to commemorate events in his father Jahāngīr's reign can be read to subtly critique his predecessor. 55 In that regard, the prominence and attention given to these eunuchs, and especially the presence of Iʻtibār Khān at the very edge of the throne-almost looming over the emperor-could be interpreted as intending to imply that Jahāngīr's reign, which featured a famously powerful queen in Nūr Jahān, was characterised by poor rulership and the elevation of 'inappropriate' individuals. While these eunuchs do not seem to be subjected to the same kinds of negative treatment as Nūr Jahān in Shāh Jahān-era histories, 56 elsewhere there is evidence for the belief in this period of a link between the power of both women and eunuchs to poor rule. 57 Yet this interpretation of the painting, as engaging in implicit critique, is made complicated by the contextual imagery of this darbār scene: as Beach and Koch note, the portrait of Akbar above Jahāngīr's head, as he embraces his son below him, emphasises dynastic continuity and stability, and the poem inscribed below the throne serves as a blessing of the emperor and his son. 58 This interpretation is further called into question by the longer tradition of such audience paintings, including other darbār scenes within the Windsor Pādshāhnāma itself, which underline the representation of eunuchs in such locations as part of an already-extant imperial iconography. One representative example is the well-known painting of Abū al-Fazl presenting the Akbarnāma to Akbar, appearing within the Chester Beatty copy of this same text (Figure 2). There are many striking differences between this scene and the symmetrical, axial arrangement of Shāh Jahānī court scenes, but one can nevertheless perceive a careful delineation of imperial and other space. Here Akbar himself is at the centre, seated on the throne beneath a canopy. These visual cues serve to identify the emperor, distinguished by these architectural elements and his elevation above the others present, in a distinct yet analogous way to the balcony and railings used in the previous image. Abū al-Fazl kneels in front of him to present the manuscript, and a large number of other standing figures are present, most (but not all) paying attention to these events.
Within this scene, attention paid to facial hair reveals a striking division of space. Standing behind Akbar, on the right side of the painting, are four men without facial hair; two moustachioed men are also seen as a part of this group. Other than these four beardless men, there is not a single figure within the crowded confines of the court not sporting some form of facial hair. The fact that this is evidently not a haphazard distribution, potentially linked merely to personal preference, allows us to identify these four men as probable eunuchs. Eunuchs and khawāsān would have been an expected presence behind the emperor, as was discussed above. While it is possible that this latter category may have included beardless youths, extant evidence points to such attendants possessing facial hair. 59 This would be all the more likely 56 For instance, see Lāhorī, Pādshāhnāma, p. 70; (trans.) p. 8. Muʻtamid Khān also refers to Nūr Jahān's fitna sāzī, or mischief-making, in his Iqbālnāma, p. 199, even as elsewhere (p. 56) he praises her abilities. 57   given that most adult men at this time sported facial hair of some kind. 60 Through specifically depicting likely eunuchs in close attendance to the emperor, this imperial image suggests both their intimate proximity as well as their symbolic valence as embodied designators of imperial space.
In other words, the 'soft' architecture of the eunuchs themselves sets off imperial space in a way analogous to other movable elements such as canopies, railings, and thrones. 61 This interpretation is underlined by the unfailing appearance of beardless figures in attendance on Akbar in audience scenes throughout versions of the Akbarnāma in both the Victoria and Albert Museum (circa 1590-1595) and the Chester Beatty Library (circa 1603-1605). 62 The regular appearance of such attendants even outside of formal contexts-for instance, in a painting illustrating Akbar's oversight of the building of Fatehpur Sikri, or another where he inspects dead animals after a hunt-underlines their usefulness as movable embodied elements framing the emperor within paintings of this period. 63 In light of this evidence, the eunuchs in the first painting from the Windsor Pādshāhnāma should be viewed as continuing the tradition of prior imperial iconography by including eunuchs as embodied extensions of imperial space. In addition to historical paintings of the prior ruler, this manuscript also represents probable eunuchs proximate to Shāh Jahān himself in audience scenes, particularly positioned slightly below the emperor bearing fans and fly whisks, reinforcing the architectural distinction between imperial space and the space occupied by the nobles. 64 In doing so, they serve as part of the move within these paintings to create for the emperor 'a separate pictorial space, positioned above a crowd of assembled courtiers'. 65 Yet not all of the darbār paintings in this manuscript include eunuch attendants; some of them only feature bearded men or no attendants at all. 66 This suggests that with the development of a more fully and formally articulated vertical spatial separation of emperor and court in Shāh Jahān's reign, the beardless attendants became less essential to differentiate space.
There are fewer, and less consistent, examples of audience scenes from Aurangzeb's reign, as more intimate scenes seem to have come into favour. 67 Thus it is not possible to track with as much clarity a specific modality to how eunuchs are made use of in such images. Nevertheless, those that exist often attest to the continued representation of eunuchs operating in close proximity to the emperor. One striking instance is a painting dating from an early period of his reign, which depicts Aurangzeb on a throne under a rich canopy with a large crowd of men lined up in the background, those lines converging on an arched gate in the distance (Figure 3). 68 Just behind the emperor, among a row of men bearing swords and fans, is one beardless man holding one of the fans, while another beardless man or youth stands to the left of this small group with his arms folded. Thus far, this scene is similar to prior examples. But there is something more: among the throng past the end of the carpet's edge stand at least eight eunuchs bearing maces, four on each side. Their distinctive lack of facial hair sets them off from the figures who succeed them, all of whom clearly sport moustaches and/or beards. The size, heavier-set bodies, and wider faces of these figures suggest that they are not youths, and the fact that they are grouped so particularly in contrast to other men with facial hair allows confident identification of them as eunuchs. In the foreground, there is an additional group of eight mace-bearing eunuchs, lined up beside the canopy's poles. The placement of eunuchs here behind the tent poles, embodied 'soft' architecture echoing and reinforcing other ephemeral elements within the space, is especially interesting to note as the prior two paintings examined similarly align beardless figures with pillars and poles. This image has a radically different arrangement from the other paintings examined here, which all position the emperor as occupying the upper third of the page and the nobles occupying the lower portion, closer to the viewer. These other works place the viewer among the nobles, or behind them, looking inwards and upwards towards the emperor. Here, in contrast, the emperor is in the foreground, and although some figures stand closer to our viewpoint-including, notably, eight of the mace-bearing eunuchs-the majority of the audience is beyond the emperor, barely discernible as they recede towards the vanishing point. While executed in a different way, this image suggests an analogous understanding of eunuchs, with the eunuch mace-bearer taking on a comparable role in terms of demarcating space. It is also notable that the eunuchs just beyond the platform are scaled to a similar or even larger size as several figures closer to the viewer under the canopy, emphasising not only their importance or relative rank, but also possibly their function as physical barriers between the space around the emperor and the crowds lined up behind them. 69 This reworking of the audience scene underscores the separation of the emperor and his immediate environs from the crowds beyond the canopy through its use of these bulky, mace-bearing eunuchs.
Taken as a whole, the evidence demonstrates that throughout the period under review eunuchs were given concrete posts in close proximity to the emperor, in privileged locations largely inaccessible to other men. Both in the inner palace, beyond the sight of the average person, as well as in the context of the court, eunuchs served a particular kind of 67 Parodi notes, from Aurangzeb's reign onwards, a move towards less 'exalted' representations of the emperor, as well as a more general 'disappearance of the public audience from Mughal painting'. Parodi, 'Darbars in transition', pp. 89-93. 68  purpose. Practically speaking, they may simply have been present to guard boundaries and mark the threshold (as in the case of Mahābat Khān's coup), perform specific ceremonial or menial tasks (waving a fan, bearing jewels, passing on petitions), or else, as essentially expensive luxury items, utilised as a way of demonstrating wealth and power. But their presence was more than pragmatic. Within both the texts and paintings surveyed, a symbolic, rhetorical role can also be perceived. In these idealised representations, eunuchs were incorporated as one of a range of elements setting off and defining the space around the emperor in court, marking it as fundamentally different and distinct. As will be evident in the next section, the seemingly seamless setting-off of elite (and, here, imperial) space described above is in stark contrast to the complex, strained dynamics often evidenced in textual documentation regarding princely establishments. Reflecting on this can help illuminate what is left out of representations of the imperial palace and royal court, as well as how eunuchs may have operated differently within princely households.

Eunuchs and Mughal princes
Up to this point, this article has demonstrated the significant boundary-marking function of eunuchs with respect to the emperor. In reviewing the sources, both textual and visual, it is clear that eunuchs were intentionally deployed as a kind of embodied 'soft' architecture within the imperial court and palace, for both practical and symbolic purposes. In the section that follows, accounts of eunuchs operating in relation to princes present a different picture, pointing to how the human, embodied nature of these boundaries comes into play in shaping social and political relationships. This section will therefore seek to understand how both royal men and eunuchs might have experienced these forms of intimacy, particularly in times of conflict. While discussion of eunuchs close to the emperor in courtly sources tends to emphasise the smooth maintenance of the order of the court and the distance of the emperor, evidence for the more complicated scenarios eunuchs confronted when operating proximate to princes demonstrates the possible consequences of eunuch mediation of space. While the textual sources considered here should be understood as just as invested in imperial discourses as the materials discussed in the previous section, the level of detail they provide on tensions around princely households nevertheless speaks to mundane concerns that are less apparent with respect to imperial households. While the specific meaning of the increased inclusion of such information within Aurangzeb-era texts will be returned to in concluding this section, this material nevertheless establishes how eunuchs could be used as tools of surveillance and control, as well as become personally involved in interpersonal conflicts. Perhaps in large part because of political anxieties around Mughal princes, there is a rich body of evidence regarding the administration and management of their establishments. Generally speaking, eunuchs formed a standard part of princely and other elite households, serving in intimate capacities similar to the imperial context. 70 As will be seen in the coming pages, within Mughal princely households, eunuchs are found serving as guards, masseurs, harem superintendents (nāzirs), and go-betweens. While throughout this period, princely rebellions were a regular feature of political life, Aurangzeb's extended reign provided an especially lengthy era of tension between the ageing monarch and his adult male sons, including both outright princely rebellion and imprisonment as well as more general distrust and surveillance. 71 It is against the background of these intrafamilial struggles that eunuchs emerge as significant figures.
In the process, eunuchs were instrumental in shaping the social dynamics of these households, at times in unexpected ways. For example, when Prince Muhammad Akbar attempted to rebel against Aurangzeb in 1681, eunuchs played a key, if inadvertent, role in the prince's failure. Aurangzeb, in order to create dissension among the would-be rebels, sent a letter to Muhammad Akbar praising him for his help in entrapping Durgādās, a prominent Rathor leader and co-conspirator. As planned, the message was 70  intercepted by Durgādās himself. However, when he went to speak with the prince, the tent's entrance was guarded by eunuchs who refused to disturb the prince while he was sleeping. Taking this as a pretext to avoid him, Durgādās and his companions made a hasty departure that very night. But the eunuchs were no pretext: when Muhammad Akbar awoke the next morning, and was informed by those same eunuchs of what had transpired, he understood the 'requirements of the moment' and fled to join Durgādās. 72 In this way, in their control of access to the royal prince, eunuchs could create levels of ambiguity and distance, even between such co-conspirators. Even as this example suggests how the additional layer of mediation via eunuchs influenced communication in curtailing the accessibility of such royals, an impact which would have been true for eunuchs in the imperial household as well, other information points to how this could also be explicitly and intentionally used against the royal male occupants of such spaces. Much as eunuchs functioned in relation to women, sometimes as allies and sometimes as wardens, the evidence suggests that for men, too, this mediation could be a double-edged sword. 73 Sometimes this could be counter to expectations, as in the case of the arrest of Prince Murād Bakhsh during the war of succession from 1657 to 1659. Seeking to pursue an alliance, Murād Bakhsh entered the camp of his brother, then-Prince Aurangzeb; meanwhile his companions remained outside. Having spoken and feasted, Aurangzeb left Murād Bakhsh to rest; he lay down, and Aurangzeb's eunuch Bashārat busied himself with massaging the prince's feet until he fell asleep. At this juncture, a female servant entered and disarmed him, and immediately thereafter soldiers were sent in to arrest him. 74 In this moment, the intimacy and trust existing between Murād Bakhsh and the subordinates present, and their close proximity during times of vulnerability, quickly became the instruments of his manipulation and downfall.
While the above two examples point to how the access of eunuchs to elite men within normally functioning households could come into play in moments of political tension and conflict, eunuchs were also explicitly deployed by emperors as jailers and prison guards. This is not a phenomenon unique to Aurangzeb's period. Prince Khusrau (son of Jahāngīr), for instance, was imprisoned and placed in the charge of Iʻtibār Khān in 1607. 75 But it is during the reign of Aurangzeb that the management of troublesome family members by eunuchs, and potentially recalcitrant princes in particular, emerges as an important recurrent theme.
The degree to which sources describe the careful oversight and control exerted over the objects around a jailed prince (or even former emperor) underlines the degree of intimacy, and intimate control, this position implied in managing the mundane aspects of an elite prisoner's life. A news digest from 1688 reports that eunuchs had been providing food to the imprisoned Prince Muʻazzam by way of a hook (miʻlāq), allowing them to hoist his meals into his cell. 76 Eunuchs also proliferate around the imprisonment of Shāh Jahān following his removal from power, responsible for monitoring him and controlling his access to various things. To take one example, a eunuch was sent to Shāh Jahān to confiscate the large cache of jewels in his possession, leading to an altercation between the eunuch and the deposed monarch about a rosary of pearls which the eunuch demanded and which Shāh Jahān refused-ultimately successfully-to hand over. 77 This case shows how eunuchs could come to serve as the points of interaction-and therefore contestation-by virtue of the roles with which they were entrusted, even if those posts seem superficially routine.
The question of what objects jailed royals had possession of could be a complex arena not only for self-assertion but also for communication and negotiation. For example, a eunuch was used by Aurangzeb to test his imprisoned son Prince Muʻazzam in 1691, as the emperor considered whether his son was trustworthy enough to be forgiven. One day the emperor sent him a pencase, including a pen-knife, through a trusted eunuch. This was a test of the prince's honesty: the eunuch was instructed that if the prince said something he was to inform him that the pen-knife was sent on purpose; if he kept the pen-knife without saying anything, that omission was to be reported to the emperor. 78 In this case, the prince passed this test and, in doing so, implicitly accepted the emperor's authority. This instance demonstrates the utility of a eunuch as a particular kind of loyal go-between, with personal access to the imprisoned prince as well as enjoying the level of trust needed to be given such a sensitive mission. 79 Not all forms of such oversight and control took the form of literal imprisonment. Even for princes who had not been detained, eunuchs could take on the tasks of spying or informing. Such roles are not absent in the earlier period, but recent scholarship has argued that in the later years of Aurangzeb's rule, and especially from the 1680s onwards, financial difficulties led to the emperor increasingly providing 'personnel' to princely households under economic duress. This shift in turn meant royal princes had less independence and came under increasing imperial surveillance, making princely rebellions more difficult, if not impossible. 80 These practices can thus be viewed as part of a larger body of work around the controlling of potential successors in the seventeenth century across the Islamicate world, albeit without utilising the 'cage' system of house-arrest seen in the Safavid and Ottoman contexts. 81 While the Mughals did not institute a similar system until the following century, 82 here eunuchs serve to enhance imperial control and curtail princely powers in a more mobile environment.
The responsibility for monitoring princes could make life complicated for such eunuchs, as they operated within the complex power dynamics between emperors and their often-ambitious sons. There are accounts of unnamed nāzirs (harem superintendents) who are criticised or removed from their posts for failing to report violations of protocol, such as when Prince Muʻazzam erected qanāt (screens) in the jāmiʻ mosque or held an elephant fight, both prerogatives of kings. In the latter instance, the emperor interpreted the nāzir's report as wilfully concealing the depth of his own knowledge and possibly his own complicity, implying that the eunuch in question was not acting out of true loyalty but rather betraying the sovereign. 83 One can only imagine the subjectivity of the eunuch at this moment: caught between the emperor who had placed him in this position, and the royal prince who might one day be emperor, and who in any case was the man of status closest at hand. Although these moments can be seen to reflect the choices made by individual eunuchs to make certain alliances or bets on the future, these choices would have been drastically constrained by the larger context, and one can also read a different story, of vulnerability and even bodily risk. Mention of eunuchs being tortured for information on their elite masters-as in the case of the eunuch confidante Nawayd, who was almost tortured to death in 1686 for information on Muʻazzam-underlines how tenuous such positions could be and the difficult decisions eunuchs sometimes had to make. 84 The material described above testifies to the varying ways in which eunuchs were operating in-and often policing-intimate spaces not only in reference to women, but also to men. Eunuchs emerge repeatedly as guards and spies, sometimes in the service of princes, yet they were also sometimes utilised by the emperor to police them. However, these examples come overwhelmingly from the reign of Aurangzeb, and the evidence related to imperial-appointed nāzirs comes more specifically from the 1680s onwards. This material does not speak so much to continued, equally distributed practices across the entirety of the Mughal period, but rather to the forms eunuch service could take at distinct historical moments. An empire under financial strain, or an ageing emperor fearing impatient sons, might see a shift in the deployment of eunuchs, both in numerical terms as well as in the exact post assigned. Of course, such economic or political anxieties could also potentially lead to a shift in representations of eunuchs; the granular depictions seen here could reflect concerns about imperial stability, thus revealing structures that have otherwise been mostly hidden from sight. Most likely, the relative density of archival information surveyed here is the product of both of these factors.
This body of evidence falls within a much longer history in which eunuchs occupied intimate spaces, and so were available to be deployed in particular ways. In this respect, the emerging picture of eunuchs in and around princely households clarifies how, more generally speaking, eunuchs could both gain access to certain opportunities as well as become personally vulnerable due to their mediatory roles. While this is less documented outside of the context of Aurangzeb-era princely households, the ongoing practice of spatial control and boundary-marking throughout the period under review indicates that the specific posts held by eunuchs would have made them liable to such entanglements, even as it also suggests that such eunuchs could use the political valences of their positions strategically to their own advantage.

Conclusion
In this article, I have approached the question of elite Mughal life from the perspective of limitations around access to elite men. As embodied boundaries and human mediators, eunuchs took on representational and material functions in the making of these rarefied spaces, both in the stone palaces of Agra and Delhi as well as in princely and imperial camps. Within the gradated layers of space around the emperor and royal princes, eunuchs were embodied markers and boundaries, alongside other figures such as noncastrated attendants, mace-bearers (gurz-bardārs), and officials of the court such as the master of ceremonies (mīr tūzuk). In representations of imperial audiences, eunuchs are particularly visible due to their beardless faces and their placement, as they are regularly represented close to the emperor, either behind or below him, marking the space he occupies within the larger, carefully ordered, and formally managed arena of the court. These paintings point to the value of representations of ordered space to project imperial 84 Muntakhab al-Lubāb, pp. 331-332; (trans.) p. 334. authority and to naturalise relations between emperor and his subjects. Textual sources similarly attest to a careful and layered definition of space within the palace complex or camp as a whole, with eunuch guards and attendants playing analogous boundarymarking roles.
At certain moments, however, this spatial order is negotiated or even violated. The coup of Mahābat Khān demonstrates that, at a moment of crisis, the embodied spatial order created by eunuchs and other subordinate figures can exert a restraining force and shape the unfolding of events. While sources from this period do not speak to the complicated relations that are likely to have existed between eunuchs and other subordinate figures-which are better attested to by eighteenth-century archives-they do demonstrate how the particular posts occupied by eunuchs could come into play in shaping their own lives as well as those of the elite men they served. 85 These dynamics appear especially complicated when it comes to eunuchs proximate to ambitious princes during the reign of Aurangzeb. In addition to reflecting developments specific to Aurangzeb's reign, this material also points to the difficulty of reading eunuch activities in a straightforward manner as direct expressions of their master's power. Rather, it becomes clear that the inner imperial palace or camp, as well as the camps of the royal princes, were complex locations of the assertion of power and authority, informed by tensions within royal families and eunuchs' own strategic choices. On the one hand, such functions could allow eunuchs to achieve positions of intimacy, knowledge, and influence with the emperor and members of the royal family, a fact which provides an at least partial explanation of the rise to high status of numerous eunuchs over the course of the Mughal period. Yet the evidence also underlines how even unnamed, low-ranked eunuchs could become entangled in moments of political conflict, intrafamilial and otherwise, a situation that provided greater opportunities but also heightened risk.