Hariti Domesticated: Re-evaluating Structures of Patronage in Gandharan Art
Strategically located on pan-Asian trade networks in modern-day northwestern Pakistan and Afghanistan, the region named Gandhara was of tremendous economic importance in the early centuries of the Common Era, a locus of exchange between cultures, both in terms of trade and ideas. Over the centuries, it was successively occupied by Achaemenid, Mauryan, Greco-Bactrian, Scythian, Parthian and finally, Kushan rulers, with each of these foreign dynasties leaving imprints of their visual culture and belief systems. The art produced in Gandhara has been a focus of Western scholarly attention since the first excavations began in the region in the 19th century. Early discourse on the subject is infused with imperialist ideology, with scholars such as James Fergusson (1808–86) and Alfred Foucher (1865–1952) tracing the region’s innovations—particularly, the vexed question of the appearance of an anthropomorphic image of the Buddha—to Hellenistic or Roman influences, revealing biases against the possibility of Indian originality.
In this discussion, I want to move away from the notion of ‘influence’, which presupposes a local population passively receiving Greco-Roman iconography, and consider instead the ways in which Buddhist sculptures produced in Gandhara around the 2nd and 3rd centuries CE can be seen as an active response from the reigning dynasty, the Kushans (c. 78–240), to their socio-political circumstances. To illustrate this larger process of dynastic patronage, as well as the local contexts in which its visual culture was implanted and adopted, this essay will concentrate on the figure of Hariti, a yakshi (spirit-deity) later incorporated into the Buddhist pantheon. As a new import to Buddhism, the figure of Hariti is unusually ubiquitous in the Gandharan visual repertoire, and even appears to be the single most prevalent female deity in the archaeological evidence from the region.
A particularly noteworthy, large-scale sculpture of Hariti is permanently exhibited in the Joseph E. Hotung Gallery of Oriental Antiquities at the British Museum, and depicts Hariti seated with a child on her lap, wearing a heavy shawl slipping from her right shoulder to reveal the detailed carving of her tunic’s fabric (Fig. 1). The child is grabbing her long, beaded necklace with one hand, while seven other children or minor deities stand at her feet. Before discussing this sculpture and the art of Gandhara in greater length, it will be useful, one the one hand, to situate the figure of Hariti in both Buddhist texts and ritual practices, as these provide crucial insights into the process of the adoption of popular cults within organized religion, and on the other, to see how this particular process facilitates political outreach to the deepest layers of society.
Fig. 1 Hariti
Gandhara, Kushan period, 2nd–3rd century
Grey schist; height 77 cm, width 42 cm, diameter 14 cm
British Museum (1886,0611.1)
(Image © The Trustees of the British Museum)
Hariti is referred to in a variety of textual sources, a good number of which have been translated into French by Noël Peri (Peri, 1917). Peri himself remarks on the relatively high volume of texts or passages that are dedicated to this secondary deity, which bears witness to her popularity. Although he takes his source material from the Chinese versions of Indian scriptures, whose approximate dating and aggregative nature are problematic, it is nonetheless possible to draw a fairly good picture of the story of Hariti. In this regard, the Samyuktavastu, closely related to a section of a Chinese scripture titled Guizimu shizi yuan (Hariti’s Fate of Losing a Child), dated to 471 CE, is the most detailed text, developing the story of Hariti’s conversion to Buddhism in considerable length (Hsieh, 2009). Although a mother herself—she has no less than 500 children, 10,000 in the Chinese source—Hariti is described as an insatiable ogress who devours the children of the city of Rajagrha (present-day Rajgir in north India). Distressed, the townspeople have tried to propitiate her with offerings and sacrifices, but to no avail. As a last resort, they decide to beseech the Buddha to pacify the deity, and the next day, he hides the youngest and most beloved of Hariti’s children, Priyankara, underneath his begging bowl. Priyankara’s absence drives Hariti mad with worry and despair as she traipses to the four corners of the earth in search of her lost son. When the Buddha finally approaches her, he makes her realize that her suffering for the loss of one child among her 500 offspring is nothing compared to what she has inflicted upon the mothers of single children. Hariti immediately converts to Buddhism and her child is returned, unharmed. She promises to uphold the precepts of Buddhism and protect the inhabitants of Rajagrha and monastic communities. In exchange, monks and nuns are to offer her food regularly in substitution for her cannibalistic diet.
This story is revealing of the early cult of Hariti, which probably mandated offerings and sacrifices intended to win her clemency. Spirit-deities require ritual action both to appease their anger and secure their blessings. Interestingly, Buddhism gradually lodged itself in this reciprocal relationship between the deity and her devotees. The main problem encountered when discussing the adoption of spirit cults by Buddhism is the general scholarly dismissal of the process as a ‘concession to popular superstitions’, as Foucher puts it, which is largely informed by a narrow definition of this religious tradition (Foucher, 1918, p. 134). Modern scholarship has tended to overplay the role of philosophy over ritual, or texts over the archaeological record. In an attempt to counter these antinomies, the adoption of spirit-deities such as Hariti can here be examined as a continuous process of negotiation between sangha (monastic community) and lay practitioners. The basic reciprocal relationship between laity and deity is one where the ritual offering of food will secure blessings. When the Buddhist sangha enters the equation, we find a triangular system of exchange: members of the laity support the monastic community, who, on their behalf, make ritual offerings to the deity, who in return protects the local community. This is precisely the pattern we find towards the end of the Samyuktavastu story, where offerings by the monastic community take over from those of ordinary men and women, and prove to be more effective in appeasing Hariti.
The process of converting local deities to Buddhism was beneficial to the sangha because it allowed it to gain social relevance in local communities, thereby attracting more followers and more donations. But as Robert DeCaroli has recently demonstrated, members of the sangha probably did not coldly see the co-opting of local deities as an opportunity for material enrichment: ‘We see a group of people who are still very respectful and wary of the beings with whom they share their homes. The monks and nuns, it seems, stepped carefully for fear of having to answer to an irate spirit-deity’ (DeCaroli, 2004, pp. 132–33). In fact, because monks and nuns came from the same communities as the laity, they most likely shared the same belief systems. Although the transformation of spirit-deities into Buddhist devotees was an intentional process conducted through textual, ritual and artistic activity, the relationship between laity and monastic community was probably far more fluid than traditionally thought by scholars. This process of domestication and adoption helps explain how Hariti came to enter the Buddhist pantheon, but does not account for her unusual ubiquity in the sculptural and architectural remains of Gandhara, a land located some distance away from her north Indian origins.
Fig. 2 Stupa drum panel with a framing element containing a female figure, presumably a yakshi
Gandhara, Kushan period, 2nd–3rd century
Grey schist; height 23.5 cm, width 14 cm, diameter 6.7 cm
British Museum (1899,0609.3)
(Image © The Trustees of the British Museum)
Contrary to many cults and religious traditions, which are tied to their locale, Buddhism travelled extensively. In a few cases, it also carried local elements along with it. Figures such as Kubera or naga deities, for example, were uprooted and adopted by foreign Buddhist communities. But in Gandhara under the Kushan dynasty, we notice the presence of Hariti in sculpted forms on an unparalleled scale. Granted, the sheer quantity of available evidence is shaped by the collecting practices of early archaeologists, who were fascinated by the art of Gandhara because of its resemblance to Greco-Roman styles. As a result, we find Gandharan art in nearly every Western museum that collects Asian art, and we have access to a much greater number and range of sculptures from this region than we have from Mathura (the second Kushan capital) for the same period. It is therefore important to bear in mind that an assessment of the popularity of Hariti in Gandhara is informed by the availability of the evidence.
Several scholars have noticed the prevalence of Hariti in Gandharan art, but few have provided convincing arguments to explain the popularity of her cult in this region. Alfred Foucher and John M. Rosenfield put forth similar arguments around the idea that secondary deities govern sub-religious functions such as wealth, luxury, protection and control over the forces of nature, which reflect the secular interests of the lay people who support the sangha (Foucher, 1918, pp. 153 and 174; Rosenfield, 1967, p. 248). Reiko Ohnuma applies this argument to the figure of Hariti, who, in her dual role as a fecund laywoman and protectoress of children, would have appealed primarily to an audience of women (Ohnuma, 2007, p. 114). She points out that a common feature of most Buddhist tales of women’s conversion is the imperative of renouncing family attachments and taking monastic vows. In contrast, Hariti remains a lay mother after her conversion. This ecclesiastical sanction of motherhood provided a model for a less rigorous form of religious involvement than taking monastic vows, thus avoiding the loss of female devotees to smaller and more localized forms of religion. Ohnuma’s argument partly explains why Hariti became a popular deity among lay people despite her status as an outsider to the standard Buddhist pantheon, but a closer look at the religious, social and political life of Gandhara will shed more light on her successful dissemination in that particular region.
Fig. 3 Relief depicting two deities with a staff, purse and cornucopia, possibly Pharro and Ardochsho
Gandhara, Kushan period, 2nd–3rd century
Grey schist; height 18 cm, width 17 cm, depth 4.5 cm
Ashmolean Museum of Art and Archaeology, University of Oxford
Presented by Major P. C. Hailey, 1962 (EA1962.42)
First, there are few indications of other female goddesses competing with Hariti in Gandhara. Foucher’s study of the material evidence did not yield a single Tara or Lakshmi type, and the only other female figures found in sculpture are either small-scale, anonymous yakshis fulfilling ornamental functions (Fig. 2), or occasional so-called Athena or Skarah Dheri types which lack sufficient attributes to identify them unequivocally (see von Drachenfels and Luczanits, 2008, cat. nos 43, 65, 107). In fact, the only other identifiable female deity that appears in Gandharan Buddhist complexes is Ardochsho, the Iranian goddess of fortune, commonly represented holding a cornucopia symbolizing plenty and prosperity (Fig. 3). There is only a small leap between the symbolic associations of a goddess of fortune with those of a bestower of children, and both would certainly have appealed to a lay audience of both men and women. I was not able to locate any large-scale sculptures of Ardochsho; she always appears on smaller reliefs, paired with Pharro, Persian god of wealth, much as Hariti is often paired with her husband Panchika, a yaksha also associated with wealth and prosperity. When comparing examples of these ‘tutelary couple’ sculptures of Ardochsho, such as that shown in Figure 3, with those of Hariti (Fig. 4), one finds a striking resemblance in the pose and treatment of the figures, and a few particular reliefs even show the goddess with both cornucopia and child, which suggests a gradual coalescence of Ardochsho and Hariti into a single figure (Fig. 5). It is tempting to presume that the worship of Hariti in Gandhara grew from that of Ardochsho, and the precedent of a cult of a goddess of fortune could well explain why Hariti, as a Buddhist import, took root in the region. Local predispositions certainly play a role when a foreign religion brings in new sets of beliefs and imagery, and a pre-existing local template could have facilitated the adoption of this foreign deity.
The vitality of the cult of Hariti in Gandhara may also have been bolstered by a devastating pestilence that ravaged the Roman Empire and spread across Asia in the 2nd century CE. Adrian D. H. Bivar has documented textual echoes of this epidemic as far as southern Arabia and China, and the symptoms described have been ascribed alternatively to smallpox, plague or typhus (Bivar, 1970, pp. 19–21). Bivar further notes: ‘The focus of smallpox infection during the 2nd century AD was no doubt in South Asia, as at the present day. If our dating … is correct, the epidemic was already growing in the reign of Kaniska [Kushan emperor, r. 126–151 CE].’ It is therefore possible that the proliferation of Hariti images paralleled the growing desperation of populations who lost children and family members to an unknown epidemic disease, which provided a compelling reason to pay homage to this new protective deity.
But perhaps the strongest impetus to the diffusion of the cult of Hariti was the intense Kushan patronage of Buddhist art in Gandhara. The Kushans, originally a nomadic people from western China called the Yuezhi, arrived in the north of the subcontinent and established their dynasty in the 1st century. They are known for their political acumen and pragmatic approach to ruling foreign territories with different cultures and beliefs. One of the ways in which the early Kushan rulers adapted to local circumstances was by creating a sense of historical continuity through the coinage system, minting images of past kings or local gods opposite their own portraits (see Bopearachchi, 1997, pp. 189–213). Importantly, the Kushans were strong supporters of Buddhism, and they appear to have carried out a general agenda of conversion.
Fig. 4 Panchika and Hariti
Gandhara, Kushan period, 2nd–3rd century
Grey schist; height 30.48 cm, width 29.84 cm, depth 8.25 cm
Los Angeles County Museum of Art
Gift of Tom and Nancy Juda (M.83.66)
(Image © 2014 Museum Associates/LACMA)
What is striking about the Gandharan sculptures that survive today, beside their sheer quantity, is their regularity. Susan Huntington has remarked that the sculptural programs which decorated stupa and monastery complexes were rather clear and complete, illustrating jataka tales and scenes of the Buddha’s life in a standardized and organized manner (Huntington, 1985, p. 140). This apparent planning and adherence to iconographical prescriptions suggests a top-down patronage of the arts rather than an organic development of local art forms. Furthermore, in a lecture at the British Museum, Indian art curator Michael Willis highlighted the fact that art from Gandhara is indiscriminately quoting from a wide variety of sources in terms of style; motifs from Greece, Rome and Persia are used, sometimes in the same reliefs. It appears that Kushan patrons were producing art using a visual vocabulary that would have been familiar to the diverse and hybrid local population (‘Object Study Session: Gandhara’, Postgraduate Diploma in Asian Art, School of Oriental and African Studies, British Museum, London, 19 October 2010). We know very little about the so-called ‘Indo-Greeks’ that inhabited these regions since the incursion of Alexander of Macedon in 327 BCE, but their mere geographical situation suggests highly syncretic social and religious systems. In order to communicate their Buddhist message effectively, the Kushans needed to introduce a visual language that would be recognizable and accessible to this composite audience. This might explain why the sculpture of Hariti in the British Museum (see Fig. 1) is modelled with a Hellenistic attention to naturalism, dressed in heavy Greco-Roman tunic and scarf, and seated on a throne, in Iranian fashion.
In the Kushan strategy of westward expansion and state building, Buddhism functioned as a social binder. In an insightful discussion, Cibele Aldrovandi and Elaine Hirata lay the groundwork for a new approach to the Gandharan archaeological evidence. Using quantitative methods to survey a large number of pieces and narrative reliefs in South Asian, European and American museum collections, they have discovered that identifiable Greco-Roman motifs featured on only 33.4 per cent of Gandharan stone reliefs (‘Buddhism, Pax Kushana and Greco-Roman Motifs: Pattern and Purpose in Gandharan Iconography’, Antiquity 79, June 2005, p. 311). Crucially, the three most recurrent elements, Corinthian capitals, Eros-Amorini and Acanthus leaves, never stand on their own, but rather appear as architectural or decorative motifs that frame Buddhist scenes. For the authors, this selective use of Greco-Roman elements suggests a conscious borrowing of external form rather than substance on the part of Kushan patrons. Being established in this region shaped by trade routes and cross-cultural exchange for centuries, local populations were likely to be familiar with Greco-Roman styles already. For Kushan rulers, the use of a visual vocabulary that would be recognizable in Gandhara proved to be an effective way to reach a highly diverse population and propagate a Buddhist discourse. Gandharan sculptors were dressing Buddhist deities in Greco-Roman clothing not because they could not imagine any other, but because of a conscious decision of Kushan rulers to dress foreign ideas up in familiar clothes. Recognizable forms facilitated the receptivity of local audiences to the religious content of the reliefs. As Aldrovandi and Hirata conclude, ‘Buddhism would in this way reinforce the integration of society using common recognisable patterns and avoiding inner ethnical conflicts … [It] would develop an iconographic repertoire that could be understood by other cultures, assuring that its ideals would reach further Asian and Western regions’ (p. 314). It appears that the combination of local visual vocabulary and a Buddhist message eased the tensions inherent in a multicultural empire.
I have attempted to retrace the processes by which a local spirit-deity became relevant to a wider audience, and was successfully adopted by a foreign population. Through both scriptural and ritual apparatuses, the Buddhist sangha effectively domesticated and assimilated Hariti within its pantheon, a process that could either be seen as a gradual penetration of local culture into organized religion, or a rather more deliberate absorption of competing forms of lay worship. Particularly, the figure of Hariti was a trope that could be carried to any land and still have a strong resonance with the daily concerns of men and women. The conversion and widespread reproduction of this figure in stone serves to illustrate a larger trend in the program of Kushan patronage of Buddhist art. In the massive architectural and sculptural projects undertaken in Gandhara, both form and subject-matter appear to have been purposefully mobilized to appeal to a local audience. Popular deities served as potent tools to reach and convert the deepest layers of society, providing a sense of unity to the disparate lands under Kushan tutelage and thereby avoiding conflicts that could stifle economic activity. A study into the prevalence of Hariti therefore suggests that Buddhist material culture in Gandhara could advantageously be understood as a body of work closely enmeshed in a pragmatic project of empire building.
Fig. 5 Tutelary couple, possibly Panchika and Hariti
Gandhara, Kushan period, 2nd–3rd century
Grey schist; height 27 cm, width 24.7 cm, depth 10.3 cm
British Museum (1950,0726.2)
(Image © The Trustees of the British Museum)
Julie Bellemare is an independent scholar of Chinese art and graduate of the MSt in History of Art and Visual Culture, University of Oxford.
Selected bibliography
Cibele Aldrovandi and Elaine Hirata. ‘Buddhism, Pax Kushana and Greco-Roman Motifs: Pattern and Purpose in Gandharan Iconography’, Antiquity 79 (June 2005): 306–15.
A. D. H. Bivar, ‘Hariti and the Chronology of the Kusanas’, Bulletin of the School of Oriental and African Studies, University of London 33, no. 1 (1970): 10–21.
Osmund Bopearachchi, ‘The Posthumous Coinage of Hermaios and the Conquest of Gandhara by the Kushans’, in Gandharan Art in Context: East-West Exchanges at the Crossroads of Asia, Cambridge, 1997.
Robert DeCaroli, Haunting the Buddha: Indian Popular Religions and the Formation of Buddhism, New York, 2004.
Dorothee von Drachenfels and Christian Luczanits, Gandhara, The Buddhist Heritage of Pakistan: Legends, Monasteries, and Paradise, Mainz, 2008.
James Fergusson, History of Indian and Eastern Architecture, James Burgess and R. Phené Spiers, eds, 2 vols, London, 1876; reprint, 1889.
Alfred Foucher, L’Art Gréco-Bouddhique du Gandhâra: Étude sur les Origines de l’Influence Classique dans l’Art Bouddhique de l’Inde et de l’Extrême-Orient, Vol 2, Paris, 1918.
Hsieh Ming-liang, ‘Guizimu zai Zhongguo: cong kaogu ziliao tansuo qi tuxiang de qiyuan yu bianqian’ (‘A Study of the Origin and Development of the Representation of Hariti in the Chinese Tradition’), Guoli Taiwan daxue meishushi yanjiu jikan (Taida Journal of Art History) 27 (2009): 107–140.
Susan L. Huntington, ‘The Northwest and Northern Regions under the Kusanas’, in The Art of Ancient India: Buddhist, Hindu, Jain, New York and Tokyo, 1985, pp. 125–62.
Reiko Ohnuma, ‘Mother-Love and Mother-Grief: South Asian Buddhist Variations on a Theme’, Journal of Feminist Studies in Religion 23, no. 1 (2007): 95–116.
Noël Peri, ‘Hariti, La Mère-De-Démons’, Bulletin de l’École Française d’Extrême-Orient 17 (1917): 1–102.
John M. Rosenfield, The Dynastic Arts of the Kushans, Berkeley and Los Angeles, 1967.
This article first featured in our October 2014 print issue. To read more, purchase the full issue here.
To read more of our online content, return to our Home page.