‘Harem’: moving beyond the cliche

‘Harem’: a word that conjures up the popular image of a closely guarded pleasure palace filled with scantily clad nubile courtesans idling away their days in languid preparation for nights of sexual adventure in a sultan’s bed. It is a world of scatter cushions, jewels in the bellybutton, and fluttering eyelashes set above gauzy yashmaks. These cliches find their most vivid expression in nineteenth-century Orientalist paintings and literature, and in popular Hollywood movies of the last 100 years (see Llewellyn-Jones 2009b). Unsurprisingly, this vision of Oriental sensual excess has often led scholarship to dismiss the notion of the harem as a western fabrication, an open sesame to an Arabian Nights fantasy world, and little more than that.

If we want to utilise the word ‘harem’ in its correct context and use it to consolidate some facts about royal women in the Persian Empire, we must dispense with the Orientalist cliches entirely. Let us start by expanding our awareness of what a harem really is.

While a harem can be a physical space, an identifiable area of a palace or house which is used by women - and by children, eunuchs, and privileged men for that matter - a harem can also simply refer to women and their blood kin grouped together; a harem does not necessarily need a defining space. ‘Harem’ has at its core the Arabic ha’ram, meaning ‘forbidden’ or ‘taboo’. By implication it means a space into which general access is prohibited (or limited) and in which the presence of certain individuals or certain types of behaviour are forbidden (see Peirce 1993: 3; Marmon 1995; Schick 2010). The fact that the private quarters in a domestic residence, and by extension its female occupants, are also referred to as a ‘harem’ comes from the Islamic practice of restricting access to these quarters, especially to males unrelated by blood kinship to the resident females. The word ‘harem’ is therefore a term of respect, evoking religious purity and personal honour and, as Hugh Kennedy has stressed, in Middle Eastern royal practice a ruler would use ‘harem’ to refer to his women and to all other individuals under his immediate protection - children, siblings, courtiers, and slaves - in other words, the personages who made up his inner court (Kennedy 2004: 160-99). Is this the way to think about using the term in its ancient Persian context? The lack of documentary and archaeological evidence makes it difficult to establish who made up the Great King’s harem, let alone speculate on how and where its members were housed and hierarchically structured, but perhaps Kennedy’s observation might work here also. But there can be no denying that the study of any royal ‘harem’ must include (perhaps even privilege) women.

It is difficult to know how the ancient Persians actually referred to a harem - either in its physical or in its ideological form - although it has been suggested that the Old Persian *xsapd. stdna, meaning ‘place where one spends the night’, might have been employed (Shahbazi 2003) but it is hard to substantiate this. As we have seen, the Old Persian term vid as used by Darius I in his inscriptions seems to carry with it the triple sense of ‘dynasty’, ‘house’ (‘palace’), and ‘household’, so vid might have been used to describe the harem in its double meaning of a (flexible) space and a group of people, but it is impossible to say so with any certainty. Another candidate for ‘harem’ is the Old Persian word tagara, ‘suite of rooms’, but this, while attractive, is far from certain and the word does not have a double meaning to incorporate the people who might inhabit those rooms.

We have already seen how the Achaemenids cultivated a separation between the public and private spheres, between visibility and invisibility, in terms of palace structure, court ceremonial, and monarchic ideology, yet this was not a system in which seclusion was endorsed but one in which separation was desired. Separation was not exclusive to women, given that the Great King himself consciously played with the notion of his separation from his subjects, nor (as we shall see) did the ideology of royal separation exclude royal women from active participation in the affairs of the dynasty, or from economic transactions, or from independent travel, or even from the owning and maintenance of personal estates of land.

Separation is the central issue of the spatial and representational divide of traditional palace and elite house structures in the Middle Eastern world and indeed the modern Farsi word andarunl, a term used by Iranians for the private family quarters and for the people who inhabit them, literally means ‘the inside’. It is used in opposition to birun - the public space and sphere of a household used for welcoming and entertaining guests of both sexes. In contemporary Iran the andarunl consists of all the males of a family and their wives, mothers, and grandmothers, and a whole array of male and female offspring ranging from babies to adolescents. Like andarunl, the Arabic-root ‘harem’ also refers to a distinct group of people who inhabit a permeable but hierarchically bound space which is separated but not secluded from the wider social space.

The ideology of the harem is a hallmark of almost all ancient Near Eastern monarchies and it makes little sense that in the long history

Table 2. The known ancient words used for ‘harem’

Of royal harems this important institution should be absent from the Achaemenid royal court. Moreover, spatial polarity is often highlighted in the known ancient words used for ‘harem’ (Table 2), stressing over and again its removal (in the physical and abstract forms) from the outer court.

Despite the interesting and illuminating work recently undertaken on ancient Near Eastern royal harems (Marsman 2003; Solvang 2003), studies of the ancient Persian court by and large either underplay the place and role of the harem or totally deny its presence. Kuhrt, for instance, generally questions the practice of royal polygamy in Near Eastern civilisations and is reluctant to acknowledge the institution of the harem in any Near Eastern society, arguing that historians rely too heavily on its existence to explain or concoct a ranking system for royal women (Kuhrt 1995: 149, 526), although a hierarchical structure among court women is of fundamental importance to the maintenance of dynastic order (see below). Briant (2002: 283) likewise speaks of ‘the myth of the harem’ and in this he follows Heleen Sancisi-Weerdenburg, who argued that generally the pernicious roles attributed by the Greeks to harems and queens were nothing more than a widespread literary cliche (Sancisi-Weerdenburg 1987a: 43, 38). Certainly, Plato’s representation of the imperial harem as the route of royal degeneracy and the inevitable decline of empire (D1) or the problematic epilogue of Xenophon’s Cyropaedia with its diatribe against Persian effeminacy (and probably not by Xenophon at all) are representative of the wider Greek paranoia about, and misunderstanding of, the part played by Persian women - but it would be hard to maintain that this pejorative view of Persian royal women pervades every Greek source (Llewellyn-Jones and Robson 2010: 66-8, 82-7; see further pertinent comments by Harrison 2011: 64-8). Nonetheless, for Briant the word ‘harem’ conjures up so effectively the misguided stereotypes promulgated by Orientalist art, literature, and cinema that he seems unable to move beyond the fantasy. He reluctantly, and obliquely, concedes that ‘although the term harem must be retained for convenience, the usual meaning cannot be applied to any women other than the royal concubines’ (2002: 285). This gets us nowhere.

More puzzling still is Maria Brosius’ methodical exclusion of the harem from her important study Women in Ancient Persia (1996). While she notes the wide array of royal females found at the Achaemenid court, and correctly observes that there was a distinct hierarchy of women, ranging in importance from the king’s mother and the king’s wives (of Persian stock) to the non-Persian concubines and, ultimately, slaves, she does not attempt to describe these women as a specific unit within the court. This does not make sense, given that in any developed court system the presence of women of specific social status would have called for a codified hierarchical structure which must have been reflected in such issues as court protocol and even designated (if not permanent) social and living spaces. Brosius is correct to refute the idea of female seclusion, noting that ‘It is clear. . . that there is no truth in suggestions that women lived in seclusion and were confined to the palace’ (Brosius 1996: 188) but her belief in a ‘Greek notion that women lived in the seclusion of the palace, hidden away from the outside world’ (Brosius 2006: 43) needs dissecting. It is not satisfactory to accept at face value that all Greeks advanced an image of ‘Oriental seclusion’ onto their construction of court women; in fact, the Greek texts rarely say as much. Persian women’s confinement is not an issue ever envisaged by Classical Greek authors of the fifth or fourth centuries bce and, in fact, key writers contemporaneous with the Achaemenids like Herodotus, Xenophon, Plato, Isocrates, Aristotle, and even Ctesias (the most maligned of Greeks when it comes to ‘women’s matters’) show royal women operating in a wide array of public spheres: travelling the country, having economic autonomy and political agency, and even hunting in the open countryside. There is no Greek text of the Achaemenid period which specifically talks about female seclusion or the hidden or carefully guarded lives of women. On the contrary, Plato states that the Persian king had no need to keep his queen in seclusion or to have her guarded because her own sense of social superiority kept her self-vigilance in operation (D2).

It is later Greek authors, like Plutarch, who over-dramatise the Persian fixation on the rigorous policing of their wives and concubines (D3). It must be recognised, though, that Plutarch was writing during the first century CE, when the romantic stereotype of the secluded Persian woman had become a stock image in Greek fictional literature.

It was the Greek-speaking authors of the newly emerged genre of the novel who first gave rise to a vogue for romantic adventure stories set within the palaces of Achaemenid kings and their romantic tales deliberately played with the tensions associated with viewing women in the harems of eastern monarchs. The Greek novelists, working from the fourth-century bce Persica of Ctesias, Deinon and Heraclides of Cumae, concocted their stories long after the fall of the Persian Empire. While they recognised the historical truth that royal Achaemenid women had been part of a regulated court society in which the harem played a key role, the stories which they composed were intended to arouse the passions of (male) readers through an erotic voyeurism. Chariton’s novel Callirhoe, written at some time between 25 bce and CE 50, is generally regarded as the earliest extant piece of Greek prose fiction and tells the story of a beautiful Greek girl, Callirhoe, who is forced into concubinage in the harem of Artaxerxes II. By locating his story in old romantic Persia, and within the harem of his imagination, Chariton allows a distinct form of Orientalism to permeate his narrative. In fact Callirhoe can be seen as a formative contributor to a long line of beautiful, if deeply misunderstood and precarious, Orientalist cliches that permeate later Greek works of literature. Edward Said in his seminal study of 1978 regarded Aeschylus as the first proponent of Orientalism, while Sancisi-Weerdenburg (1987a) regards Ctesias as the culprit; in fact it is Chariton, working some 200 years after Aeschylus, who is responsible for a particularly passe visualisation of the east, which might be termed the ‘the jewel in the bellybutton’ school of Orientalism (see further Llewellyn-Jones 2009b).

By the first century CE the romantic ‘harem motif’ had embedded itself so firmly within the popular imagination of Greek and Roman readers that historians like Plutarch and, later, Aelian were using the stereotypical image of the secluded Oriental harem as factual content in the construction of their eastern biographies and histories (for a full discussion of this process see Llewellyn-Jones forthcoming a). It is also vital to recall that in writing his Greek Lives Plutarch had a particularly virulent anti-Persian prejudice (see Llewellyn-Jones and Robson 2010: 40-3). Plutarch’s Life of Themistocles has a definite agenda and his Boys’ Own-style adventure story of the Greek statesman’s flight from Persia in a curtained carriage, travelling disguised as a woman, necessitates Plutarch’s exaggeration and gives him a vehicle to express his opinion that the despotic Persians exercised extreme control over their women. However, in his Life of Artaxerxes, which revolves around the workings of the inner court in some detail, and is derived in large part from Ctesias’ and Deinon’s observations of court life, nothing of this strict barbaric ‘Oriental seclusion’ is suggested (for the unreliability of Plutarch on the Themistocles matter see Nashat 2003: 21, 23).

Moreover, misunderstandings regarding the ‘Oriental seclusion’ of Persian women arise in part from cavalier translations of the Greek spatial term gynaikaion or gynaikonitis as ‘women’s quarters’. Janett Morgan (2007, 2010) has revealed that this translation is widely off the mark and that a better rendering of these Greek terms would simply be ‘the place where the women are’, suggesting any temporary space utilised by women and by family. In the ancient Greek understanding, gynaikaion is never a fixed, let alone secluded, female-only space, so that to read Persian court structures through an imperfect understanding of the Greek terminology is futile and to think of ‘harem’ in terms of a secluded female-only space or as a form of oppressive purdah is a crucial misconception of the nature of the term and the institution.

Brosius (2007: 25) asserts that the strong economic status of royal women ‘provides a clue to palace organisation’ but unfortunately she offers no further exploration of this potentially important statement. Despite recognising the Great King’s immediate family as the most important group of the inner court and acknowledging the presence of several groups of women (‘female royalty and noblewomen’, including mother, wives, heir, other children, and royal siblings, ‘attendants’, ‘royal concubines, and administrative personnel’) she cannot find a ‘place’ for them at court, and her reading of ‘harem’ (which might be rooted still in a Saidian reading of Orientalism, although she does not explicitly state this) inhibits her from drawing them into a collective institution (Brosius 2007: 31-3). In contrast, Jack Balcer (1993: 273-317) and Tony Spawforth (2007a: 93, 97, 100) both see the logic of a female court hierarchy and thus employ ‘harem’ as the simplest and most effective way to talk about the women and personnel of the Persian inner court without any pejorative associations.

There is no reason to abandon using ‘harem’. We can use the term safely, without an Orientalist gloss and free of misconceptions or preconceptions. Scholarship needs to rise above and beyond the harem cliche and recognise that, in the light of not having an Old Persian term which survives, ‘harem’ is the most appropriate term to use to describe the domestic make-up and the gender ideology of the Persian inner court.