James Mallinson

The earliest textual descriptions of yogic techniques date to the last few centuries BCE and show their practitioners to have been ascetics who had turned their backs on ordinary society.1 These renouncers have been considered practitioners of yoga par excellence throughout Indian history. While ascetics, including some seated in meditative yoga postures,2 have been represented in Indian statuary3 since that early period, the first detailed depictions of Indian ascetics are not found until circa 1560 in paintings produced under the patronage of Mughal Emperor Akbar (reigned 1556–1605) and his successors.4 These wonderfully naturalistic and precise images illuminate not only Mughal manuscripts5 and albums but also our understanding of the history of yogis6 and their sects. Scholars have argued for these paintings’ value as historical documents;7 their usefulness in establishing the history of Indian ascetic orders bears this out. The consistency of their depictions and the astonishing detail they reveal allow us to flesh out—and, sometimes, rewrite—the incomplete and partisan history that can be surmised from Sanskrit and vernacular texts, travelers’ reports, hagiography, and ethnography.8

The eleventh to the fifteenth centuries saw the composition of a corpus of Sanskrit works that teach the haṭha method of yoga, which places the greatest emphasis on physical practices.9 The techniques of haṭha yoga—some of which were probably part of ascetic practice for more than a thousand years before they were taught in texts—became integral to subsequent formulations of yoga, including orthodox ones such as those found in the later “Yoga Upaniṣads.”10 They form the basis of much of the yoga practiced around the world today.

Within the texts of the haṭha yoga corpus, we can identify two yogic paradigms. One, the older, is the tradition of the yogis described in our earliest sources and is linked to the physical practices of tapas—asceticism. It uses a variety of physical methods to control the breath and to arrest the downward flow and loss of semen,11 which is said to be the essence of life. Control of breath and semen leads to control of the mind, as well as perfect health and longevity. In classical formulations of haṭhayoga—such as that found in the most influential text on the subject, the fifteenth-century Haṭhapradīpikā—a second paradigm, that of Tantric yoga, is superimposed onto this ancient ascetic method. As taught in its root texts, which were composed between the fifth and tenth centuries CE, Tantric yoga consists for the most part of meditations on a series of progressively more subtle elements, a progression represented in some Kaula Tantric texts from the tenth century onward by the visualization of the ascent of the serpent goddess Kuṇḍalinī through a series of wheels (cakras) or lotuses (padmas) located along the body’s central column.

The ultimate goal of both of these yogic paradigms is liberation (mokṣa), which can be achieved while alive. Along the way various supernatural abilities or siddhis are said to arise, ranging from mundane benefits such as overcoming hunger and thirst through the power of flight to the attainment of an immortal body. In the ancient ascetic tradition, these siddhis are ultimately impediments to the final goal; in the Tantric tradition, they may be ends in themselves.12

This mixing of yogic traditions suggests an ascetic milieu in which techniques were exchanged freely, a suggestion corroborated by the lack of emphasis on sectarianism in the texts of the early haṭhayogacorpus. The earliest text to teach a yoga explicitly called haṭha declares: “Whether a Brahmin, an ascetic, a Buddhist, a Jain, a Skull-Bearer or a materialist, the wise one who is endowed with faith and constantly devoted to the practice of [haṭha] yoga will attain complete success.”13

Early Mughal paintings bear witness to an ascetic archetype. Yogis have long, matted hair and beards, are naked or nearly so—what cloth they do wear is ochre-colored—and smear their bodies with ashes. In addition to these long-attested ascetic attributes, Mughal-era yogis display some more recent traits: they wear hooped earrings,14 sit around smoldering fires,15 and drink suspensions of cannabis.16 See, for example, some of the finest early Mughal depictions of Indian yogis—a single folio from the St. Petersburg Muraqqa‘ (Album), which shows a camp of ascetics (fig. 1) or two folios from a manuscript of the Akbarnāma showing a battle between two Saṃnyāsī suborders (figs. 2 and 3).

Figure 1. Mughals Visit an Encampment of “Sadhus,” from the St. Petersburg Album.
India, Mughal dynasty, ca. 1635. Opaque watercolor and gold on paper, 46 x 29.5 cm. St. Petersburg Institute of Oriental Manuscripts, f.47r
Figure 2. Akbar Watches a Battle between Two Rival Groups of Saṃnyāsīs at Thaneshwar. (left folio) By Basawan and Asi. India, Mughal dynasty, possibly Pakistan, 1590–95. Opaque watercolor and gold on paper, 38.1x 22.4 cm. Victoria and Albert Museum, IS.2:62-1896
Figure 3. Akbar Watches a Battle between Two Rival Groups of Saṃnyāsīs at Thaneshwar. (right folio)
By Basawan Asi. India, Mughal dynasty, possibly Pakistan, 1590–95. Opaque watercolor and gold on paper, 38.1x 22.4 cm. Victoria and Albert Museum, IS.2:62-1896











But although the two yogi traditions clearly interacted, sharing both theory and practice, their lineages remained distinct.17 They were represented, in the case of the ancient tradition of celibate asceticism, by groups that today constitute sections of the Daśanāmī Saṃnyāsī and Rāmānandī ascetic orders, and, in the case of the tradition of Tantric adepts such as Matsyendra and Gorakṣa,18 by groups that today constitute sections of an ascetic order now known as the Nāths.19 These orders were only starting to be formalized in the early Mughal period.20 Today they remain, together with the Sikh-affiliated Udāsins, the biggest ascetic orders in North India.

Figure 4. Rāmānandī sādhus putting on ashes after bathing at the Ardh Kumbh Melā, Allahabad, February 1995; photo by James Mallinson
Figure 5. Daśanāmī Nāgā Saṃnyāsīs processing to bathe at the Ardh Kumbh Melā, 1995; photo by James Mallinson
Figure 6. Young Nāth sādhus at Jwalamukhi, November 2012, photo by James Mallinson

We know from external evidence that the ascetics depicted fighting in two folios (figs. 2, 3) from the Akbarnāma (1590–95) and those depicted in two folios (figs. 7, 8) from the Bāburnāma are from lineages belonging to the two separate yogi traditions.

Figures 2 and 3 depict a battle, witnessed by Emperor Akbar, that took place in 1567 on the banks of the bathing tank at Kurukshetra. The combatants belonged to two rival yogi suborders, and they were fighting over who should occupy the best place to collect alms at a festival. In his description of the battle, Akbarnāma author Abu’l Fazl called the combatants Purīs and Giris, which remain to this day two of the “ten names” of the Daśanāmī or “Ten-Named” Saṃnyāsīs.21

Fig. 7 The Yogis at Gurkhattri, from Vaki’at-i Baburi (The Memoirs of Babur). By Gobind. India, Mughal dynasty, ca. 1590-3. ©The British Library Board, Or. 3714, f.197r
Fig. 8 Babur’s 1519 Visit to Gurkhattri, from Vaki’at-i Baburi (The Memoirs of Babur) By Kesu Khurd. India, Mughal dynasty, ca. 1590-93. ©The British Library Board, Or. 3714, f.320v









Figures 7 and 8 are illustrations from a circa 1590 manuscript of the Bāburnāma and depict a visit Emperor Bābur made in 1519 to a monastery at Gurkhattri in modern-day Peshawar, Pakistan. The manuscript and its illustrations were made under the patronage of Akbar, who himself visited Gurkhattri twice in 1581,22 so the illustrations are likely to depict the monastery and its inhabitants at that time.23Until the partition of India, Gurkhattri was an important center of the Nāth ascetic order,24 and there is still a temple to Gorakṣa, its founder, at the site today.25 This does not confirm that Gurkhattri was in the possession of Nāths at the time of either Bābur’s or Akbar’s visit—many such shrines have changed hands over time—and the inhabitants of Gurkhattri are not identified in the Bāburnāma as Nāths, but rather as jogī(s),26 a vernacular form of the Sanskrit yogī, which can refer to ascetics of a variety of traditions. However, we can infer that they were Nāths27 from three attributes that they do not share with the Saṃnyāsīs shown fighting at Kurukshetra in the Akbarnāma.

The first is the wearing of horns on threads around their necks. Today, the single most reliable indicator of Nāth membership is the wearing of such horns (see fig. 11).28 Nāths now call their horns nāds, but they were formerly known as siṅgīs, and this appears to have been the case in the medieval period. In medieval Hindi literature siṅgīs are frequently mentioned among the accoutrements of yogis, and siṅgī-wearing yogis are sometimes identified as followers of Gorakṣa.29 In keeping with their lack of sectarianism, Sanskrit texts on haṭha yoga, even those associated with Gorakṣa, make few mentions of sect-specific insignia, and none of siṅgīs, but other Sanskrit sources associate yogi followers of Gorakṣa with the wearing of horns. Thus an early sixteenth-century South Indian Sanskrit drama describes a Kāpālika ascetic as uttering “Gorakṣa, Gorakṣa” and blowing a horn,30 and the tenth chapter of a Sanskrit narrative from Bengal dated to the second half of the sixteenth century or earlier31 tells of the yogi Candranātha being awoken from his meditation by other yogis blowing their horns.32 From the fourteenth to the sixteenth century travelers to the regions in which the earliest references to Gorakṣa are found33 reported the use of horns by yogis.34 The identification of ascetics who wear horns as Nāths is supported by a painting of the annual Urs festival of Mu’inuddin Chishti at Ajmer completed in the 1650s35 and now in the collection of the Victoria and Albert Museum, London.36 At the bottom is a group of Hindu ascetics. The fourth and fifth figures from the right, who both sport siṅgīs, are identified on the painting itself as Matsyendra and Gorakṣa, the first human Nāth gurus.

The other two specifically Nāth attributes are the necklace and fillet worn by three of the ascetics in figure 8. At the end of the sixteenth century the Jesuit traveler Monserrate visited Bālnāth Ṭillā, a famous Nāth shrine in the Jhelum district of Pakistani Punjab, which was the headquarters of the order until the partition of India.37 Describing the monastic inhabitants of the Ṭillā, Monserrate wrote, “The mark of [the] leader’s rank is a fillet; round this are loosely wrapped bands of silk, which hang down and move to and fro. There are three or four of these bands.”38 This description seems to conflate two items of apparel often depicted in Mughal paintings of yogis: a simple fillet and a necklace, hanging from which are colored strips of cloth (Monserrate’s silk bands).39 Neither of these is worn today,40 but they serve to identify their wearers in Mughal paintings as Nāth yogis.41

These indicators of membership of the Nāth order—the horns, fillets, and necklaces—enable us to identify ascetics in a large number of early Mughal paintings, including those depicted in this beautiful seventeenth-century painting of yogis (fig. 9), as Nāths.42

Figure 9. A Party of Kanphat Yogis Resting around a Fire. By Mas’ud. India, Mughal dynasty, 1630-40.
Tinted drawing with gold; on an album leaf with inner border of marbled paper and an outer border of leaf-motifs in blue and gold; 22.4 x 13 cm (folio), 36.1 x 24 cm (page).
© The British Library Board, Johnson Album, 22,15.

Once members of the Nāth saṃpradāya have been identified, it is possible to note other attributes that Nāths do not share with the Saṃnyāsīs depicted in contemporaneous illustrations. These include the wearing of cloaks and hats, the accompaniment of dogs, and the use of small shovels for moving ash. The Saṃnyāsīs, meanwhile, in keeping with the renunciation implied by their name, do relatively little to embellish their archetypal ascetic attributes and are thus best distinguished by the absence of the specifically Nāth features noted above.43 Indeed, in some cases, their renunciation is such that they are naked, which the Nāths never are. Figure 1, then, shows a Saṃnyāsī encampment.

There are fewer Mughal pictures of Saṃnyāsīs than of Nāths.44 The north Indian ascetic Nāth traditions encountered by the Mughals were closely linked to the Sant tradition of holy men and, like them, believed in a formless, unconditioned god. This theological openness—which manifested in, among other things, a disdain for the purity laws adhered to by more orthodox Hindu ascetics—allowed them to mix freely with those such as the Muslim Mughals, who more caste-bound Hindu traditions would consider mlecchas (barbarians).45 Furthermore the Nāths were not militarized, unlike the Saṃnyāsīs, whose belligerence would have proved an impediment to interaction with the Mughals.46 The Nāths’ greater influence on the Mughal court is further borne out by the preponderance of their doctrines in Persian yoga texts produced during the Mughal period.47

The criteria used above to identify the Nāths and Saṃnyāsīs in early Mughal paintings have been taken exclusively from sources contemporaneous with or older than the paintings themselves. This is because using modern ethnographic data to interpret these images has its pitfalls. By now the reader acquainted with the Nāths may have wondered why little mention has been made of earrings. Today, Nāths are renowned for wearing hooped earrings through the cartilages of their ears, which are cut open with a dagger at the time of initiation.48

Fig. 10 Antelope horn kānphaṭā earring, Jvalamukhi, November 8, 2012; photo by James Mallinson

For this reason, they are sometimes referred to as kānphaṭā (split-eared), a pejorative term that they themselves eschew. Very few other ascetics today wear earrings of any sort and, to my knowledge, none wears them kānphaṭā-style.49 The current exclusive association of Nāths wearing hooped earrings has led many scholars to take textual mentions or artistic depictions of such insignia as indications that the wearers are Nāths, but this is not always the case. In India, earrings have long been emblematic of both divinity50 and rank.51 Thus many representations of the Buddha show him with earlobes that are distended and pierced but empty, signifying his renunciation: he had abandoned the heavy jeweled earrings he wore as a royal prince.52 In contrast, Mahāyāna bodhisattvas and Tantric adepts (siddhas) were conceived of as sovereigns of their realms and are often described and depicted as wearing earrings (and other regal accoutrements).53 These Hindu and Buddhist siddhas may have been the first ascetics to wear earrings; a related type of ascetic, the Kāpālika (Skull bearer), is often said to wear them.54

In medieval vernacular texts contemporaneous with early Mughal paintings, earrings are almost always included (usually as mudrā) in lists of yogi insignia.55 Often they are associated with yogis who follow Gorakṣa. If we look at the ears in figures 1–3 and 7–9, however, we see two surprising features. First, almost all, whether they belong to Nāths or Saṃnyāsīs, sport earrings. Second, no earring goes through cartilage. Depictions of Saṃnyāsīs up to the eighteenth century often show them wearing earrings, and it is not until the late eighteenth or even early nineteenth century that we come across the first depictions of Nāths wearing earrings kānphaṭā-style. A fine example is a painting of two ascetics that illustrates a manuscript of the Tashrīḥ al-aḳvām, an account of various Indian sects, castes, and tribes commissioned by Colonel James Skinner and completed in 1825 (fig. 11). The ascetic on the left is identified in an expanded version of the picture from the same period as an Aughaṛ, i.e., a Nāth who is yet to take full initiation; the one on the right, who wears a siṅgī around his neck and kānphaṭāearrings, is a full initiate by the name of Śambhu Nāth.56

Travelers from the sixteenth century onward commented on the wearing of earrings by yogis,57 but there are no outsider reports of them being worn kānphaṭā-style until circa 1800.58 The seventeenth-century poet Sundardās, whose earliest manuscript is dated 1684,59 contrasts earring-wearing jogīs with jaṭā-growing Saṃnyāsīs (pad 135) and elsewhere derides splitting the ears (kān pharāi) as a means of attaining yoga (sākhī 16.23).60 Since no paintings of yogis from the Mughal heyday (up to 1640) show split-eared yogis, it thus seems likely that the practice developed in the second half of the seventeenth century. The use of the pejorative name kānphaṭā, however, is not found until the second half of the eighteenth century, suggesting that the practice did not become widespread until then. The Nāths’ adoption of this extreme kānphaṭā style led to earrings in general being closely associated with the Nāth order, with the result that other ascetic orders eschewed the practice.61

Figure 11. Aughar and Kanphata Yogi, from Tashrih al-aqvam, p. 399. India, Hissar, Hansi Cantonment, 1825. Manuscript, watercolor; 31.5 x 22 cm (folio). The British Library Board, Add.27255, f.399b

The received history of the Nāths is based on hagiography and has the twelfth-century Gorakṣa founding the order, complete with its twelve subdivisions, by putting earrings through the cartilages of his disciples’ ears. The order is said to have flourished until the eighteenth century or thereabouts and to have been in steady decline ever since. But close examination of the historical sources shows that the opposite is more likely.62 The first organization to claim authority over all Nāth lineages was founded in 1906.63 The Nāth saṃpradāya (Nāth order) often referred to in histories of yoga and yogis was in fact a variety of disparate orders that traced their lineages to one or another Tantric siddha. Thus Jālandharnāth was the tutelary deity of Maharaja Man Singh’s Jodhpur in the early nineteenth century, and Gorakṣa played a subsidiary role in the texts and paintings produced at Man Singh’s court64 until late in his reign (1803–43).65 The adoption of kānphaṭā-style earrings appears to have been part of the process of Gorakṣa’s becoming the titular head of the order and is always associated with Gorakṣa in legend.66 The earliest image of Jālandharnāth from Man Singh’s reign, a painting of him at his seat in Jalore, shows him and his attendants wearing earrings in their earlobes (fig. 12).67 In subsequent depictions of Nāths from the region, such as another of Jālandharnāth in a folio from the Nāth Carit (fig. 13), they sport kānphaṭā-style earrings.68 The Nāth Carit identifies the previously preeminent Jālandharnāth with Gorakṣa.69 Jālandharnāth was also identified with the Bālnāth of Bālnāth Ṭillā, which, as noted above, was known as Gorakh Ṭillā by the second half of the eighteenth century.70

Fig. 12 Jālandharnāth at Jalore.
By Amardas Bhatti. India, Rajasthan, Marwar, Jodhpur, ca. 1805–10. Opaque watercolor and gold on paper; 39 x 29 cm. Mehrangarh Museum Trust, RJS 4126
Fig. 13 Three Aspects of the Absolute, folio 1 from the Nath Charit. By Bulaki. India, Rajasthan, Jodhpur, 1823 (Samvat 1880). Opaque watercolor, gold, and tin alloy on paper; 47 x 123 cm. Mehrangarh Museum Trust, RJS 2399










Just as the Nāths’ earrings changed as the result of changes in the Nāth saṃpradāya, so too did their horns. The siṅgī worn by Nāths today is a more complex affair than that depicted in Mughal painting, which appears to have been an antelope horn eight to ten centimeters long, worn on a short thread around the neck so that it rested on the upper part of the chest. Today’s siṅgī ensemble consists of a stylized miniature horn—more of a whistle—about three centimeters long and one centimeter in diameter, which is made from a variety of different materials, ranging from gold to plastic. It is worn around the neck with a ring (pāvitrī) and a rudrākṣa (Elaeocarpus ganitrus Roxb.) seed on a long thread of spun black wool that hangs almost to the waist.

Fig. 14 Nāth janeo, Bābā Bālaknāth Temple, April 2009; photo by James Mallinson

The Nāths call this ensemble either a selī or a janeo (fig. 14). The latter is a Hindi word for the yajñopavīta or “sacred thread” worn by twice-born Hindus, and suggests a clue to the changes in Nāth neckwear.

The watershed in the Nāths’ siṅgī configuration can be seen in paintings from Man Singh’s reign in Jodhpur. Figure 12 has Jālandharnāth and his companions wearing their stylized siṅgīs on short threads around their necks, without a ring or rudrākṣa seed, in the manner of those shown in figures 7, 8, and 9. Once the “mature archetype” of Jālandharnāth was established,71 he and his companions were always shown wearing their siṅgīs (without a ring or rudrākṣa seed) on waist-length black threads, usually around their necks (in the same manner as the yogi in fig. 11) but sometimes over one shoulder and under the other in the manner of a brahmin’s sacred thread.72 It seems that the newer, longer ensemble came about in imitation of the brahmanical janeo. During their heyday, the Jodhpur Nāth householders “began to adopt high-caste Hindu ways,”73 and we see in texts commissioned by Man Singh an alignment between the previously unorthodox Nāth tradition and classical Hinduism.74

The most significant fault line in Hindu theology is the division between Śaivas, who hold that the supreme being is Śiva or his consort, Devī, and Vaiṣṇavas, who hold that it is Viṣṇu or one of his incarnations (avatāras), usually Rāma or Kṛṣṇa. This division was at its most violent in the eighteenth century, when battles between the military wings of two yogi orders, the Śaiva Daśanāmī Saṃnyāsīs and Vaiṣṇava Vairāgīs (whose largest suborder is that of the Rāmānandīs), resulted in the deaths of thousands of ascetics. To this day, the sādhu camps at the triennial Kumbh Melā festivals are divided into the army of Śiva and the army of Rām (fig. 15). Mughal-era paintings of ascetics, however, show that the situation was somewhat different in the sixteenth and seventeenth centuries, as we shall see below.

Fig. 15 Rāmānandī Nāgā at the Kumbh Mela, Haridwar, April 2010; photo by James Mallinson

Nowadays the Nāths, like the Saṃnyāsīs, are overtly Śaiva, but the pictorial record indicates that this has not always been the case: Nāths are not shown sporting Śaiva insignia, such as rudrākṣa seeds or tripuṇḍras (horizontal forehead markings made with ash) until the late eighteenth century.75 The current Nāth janeo configuration, in which a ring and a rudrākṣa seed have been added to the long black thread and siṅgī, appears to be an innovation of the nineteenth century at the earliest.76 The Nāths’ roots in Śaiva Tantric traditions make the absence of Śaiva insignia in Mughal depictions of them surprising; perhaps it is symptomatic of their devotion to a formless absolute, an attitude prevalent in North Indian ascetic orders in late medieval India.77

But it is not only the Nāths who are free from Śaiva insignia in Mughal paintings; to my knowledge, no ascetic of any stripe wears the horizontal tripuṇḍra forehead marking or necklaces of rudrākṣa seeds. The unmistakable Śaiva denomination of today’s Daśanāmī Saṃnyāsīs makes the absence of Śaiva insignia in their Mughal depictions particularly surprising. In myths, Śiva is often portrayed as the yogi par excellence, with the result that asceticism and yoga have come to be thought of as originally Śaiva, and their non-Śaiva manifestations as adaptations of Śaiva traditions. But in our earliest sources, the association of asceticism and yoga with Śiva is by no means exclusive,78 and Śaivism did not dominate subsequent teachings on yoga.79 It is perhaps the association of asceticism with Śiva and the Śaiva affiliation of today’s Daśanāmī Saṃnyāsīs that have led scholars to assume that the ascetics in Mughal paintings are Śaivas.80 Yet, as I have remarked, there are no Śaiva insignia in any Mughal pictures of ascetics.81 On the contrary, many of the Saṃnyāsīs depicted therein sport on their foreheads the distinctive ūrdhvapuṇḍra V-shaped Vaiṣṇava marking. A large number of the Saṃnyāsīs fighting in figures 2 and 3 clearly have these markings (see details in 16b, 16c, 16d), as does the leader of the Saṃnyāsī troop (figs. 1, 16a). Other Mughal paintings of Saṃnyāsīs from the same period also show them wearing ūrdhvapuṇḍras (e.g. figures 18 and 19).82

Fig. 16 Leader of the Saṃnyāsī troop, detail from Mughals Visit an Encampment of “Sadhus,” from the St. Petersburg Album, ca. 1635. St. Petersburg Institute of Oriental Manuscripts, f.47r (see fig. 1)
Figure 16b. Detail from right folio of Akbar Watches a Battle between Two Rival Groups of Saṃnyāsīs at Thaneshwar.
By Basawan Asi. India, Mughal dynasty, possibly Pakistan, 1590–95. Opaque watercolor and gold on paper, 38.1x 22.4 cm. Victoria and Albert Museum, IS.2:62-1896










Fig. 16c. Detail from left folio of Akbar Watches a Battle between Two Rival Groups of Saṃnyāsīs at Thaneshwar.
By Basawan Asi. India, Mughal dynasty, possibly Pakistan, 1590–95. Opaque watercolor and gold on paper, 38.1x 22.4 cm. Victoria and Albert Museum, IS.2:62-1896
Fig. 16d. Detail from left folio of Akbar Watches a Battle between Two Rival Groups of Saṃnyāsīs at Thaneshwar.
By Basawan Asi. India, Mughal dynasty, possibly Pakistan, 1590–95. Opaque watercolor and gold on paper, 38.1x 22.4 cm. Victoria and Albert Museum, IS.2:62-1896











Vaiṣṇava features of Daśanāmī Saṃnyāsī identity are in fact legion. To this day, all Daśanāmī ascetics greet one another with the ancient Vaiṣṇava aṣṭākṣara (“eight-syllabled” mantra): oṃ namo nārāyaṇāya. Śaṅkarācārya, who was retroactively claimed to have founded their order, was Vaiṣṇava.83 Three of their four pīṭhas or sacred centers—Dwarka, Puri, and Badrinath—are Vaiṣṇava places of pilgrimage.84 Prior to the sixteenth century, the Daśanāmī nominal suffix Purī is found only on the names of Vaiṣṇava ascetics.85 The tutelary deities of the two biggest akhāṛās (regiments) of the Daśanāmīs today are Dattātreya and Kapila, both of whom are included in early lists of the manifestations of Viṣṇu.86

It is the ūrdhvapuṇḍras in these Mughal miniatures, however, and the absence of Śaiva insignia that provide us with the most compelling evidence that at least some of the groups that came to form the Daśanāmī Saṃnyāsī order were originally Vaiṣṇava. It is not clear how, when, or why the Daśanāmīs acquired an overarching Śaiva orientation, but it is likely to have been a result of the formalization of the order, in particular its affiliation with the southern Sringeri monastery and the concomitant attribution of its founding to Śaṅkarācārya, who by the seventeenth century had been rebranded a Śaiva.87 During the seventeenth century, the three main ascetic orders of North India—the Daśanāmīs, Rāmānandīs, and Nāths—forged links with southern institutions as they staked claims to dominion over all of India. The Daśanāmīs joined forces with the Sringeri maṭha, whose teachings, a blend of Advaita and the sanitized form of Śaivism known as Śrīvidyā, they adopted.88 As part of this process, both the Sringeri maṭha and the Daśanāmīs claimed Śaṅkarācārya as their founding guru. Together with Śaivism, the Daśanāmīs would have taken northward the antipathy between Śaivas and Vaiṣṇavas that had afflicted South India for at least five hundred years. It persisted in debates between different Brahmin and Saṃnyāsī factions, some of which were connected with the Sringeri maṭha, in Vijayanagar until its downfall in 1565 and, latterly, in Varanasi.89

The rapid hardening of the Daśanāmīs’ Śaiva orientation over the course of the seventeenth and eighteenth centuries was likely to have been in reaction to the formation of their archrivals, the Rāmānandīs, ascetic worshipers of Viṣṇu’s Rām incarnation. Today, Rāmānandīs wear Vaiṣṇava ūrdhvapuṇḍra forehead markings like those depicted in the early Mughal portrayals of Saṃnyāsīs (figs. 17, 18, 19).

Fig. 17 Rāmānandī sādhus putting on ashes after bathing at the Ardh Kumbh Melā, Allahabad, February 1995; photo by James Mallinson
Fig. 18 Detail, Saṃnyāsī with attendants. By Dhanrāj, 1595-1600. From the collection of Ludwig Habighorst
Fig. 19 Detail, An ascetic in a landscape. By Govardhan, 1620-30. ©The British Library Board, Add.Or.3129, f.11v









Indeed, one might contend that figure 1—whose subjects, unlike those in figures 2 and 3, are not identified in contemporaneous sources as Daśanāmī Saṃnyāsīs—portrays Rāmānandīs (or rather their forerunners, since the order was yet to be formalized or refer to itself as Rāmānandī).90 But three features of the ascetics in figure 1 set them apart from today’s Rāmānandīs.

First, there is the ancient ūrdhvabāhu penance of permanently holding one or two arms in the air undertaken by the ascetic in the bottom left of the picture. Today this is the preserve of Daśanāmīs (fig. 20).

Fig. 20 Daśanāmī Saṃnyāsī Amar Bhāratī, who has held his arm in the air for forty years, Kumbh Mela, Allahabad, February 2013; photo © Cambridge Jones


Rāmānandīs will not practice it because it is likely to permanently disfigure the body, rendering it unsuitable for the orthodox Vedic ritual acts that they, unlike the Saṃnyāsīs, perform (fig. 21).

Fig. 21 Rāmānandī Tyāgīs performing a Vedic fire sacrifice, Ardh Kumbh Mela, Allahabad, 2007; photo by James Mallinson

Rāmānandīs prefer austerities such as dhūni-tap, sitting in the summer sun surrounded by smoldering cow-dung fires (fig. 22), or khaṛeśvarī, standing up for years on end (fig. 23).

Fig. 22 Rāmānandī Tyāgī performing dhūni-tap at the Kumbh Mela, Haridwar, April 2010; photo by James Mallinson
Fig. 23 Rāmānandī Tyāgī khaṛeśvarī, Kota, November 2001; photo by James Mallinson








Second, two of the ascetics, including the figure who has undertaken the ūrdhvabāhu penance, are naked. Rāmānandīs today are scornful of the Daśanāmīs’ nakedness, saying that it offends Lord Rām.91Third, the remaining ascetics wear ochre-colored cloth, unlike the Rāmānandīs, who wear white cloth, saying that the Daśanāmīs’ ochre robes are the color of the menstrual fluid of Pārvatī, Śiva’s consort.92

Other features differentiate the Rāmānandīs from the Daśanāmīs, such as the former’s insistence on “pure” (i.e., lacking onion and garlic) vegetarian food, their taking of the nominal suffix -dāsa at initiation, their practice of orthodox rituals, and the associated preservation of the topknot when they have their heads shaved at initiatory and other ceremonies. These differences are all emblematic of the Rāmānandīs’ ultra-Vaiṣṇavism, a trait shared with other members of the “four traditions” (cār saṃpradāya) of Vaiṣṇavas, which were formalized in the seventeenth century and sought to unite North Indian devotional traditions with more established South Indian lineages.93

If one puts these ultra-Vaiṣṇava traits aside, however, the Daśanāmīs and Rāmānandīs are remarkably similar, and not just because they both embody a shared ascetic archetype and lead almost identical lives. Their organization and initiation procedures are very close.94 They both worship Hanumān and gods and sages associated with the ancient ascetic yoga tradition, such as Dattātreya and Kapila.95 They share a secret vocabulary.96 The nominal suffix –ānanda found in the names of early Rāmānandī gurus prior to the adoption of the suffix –dāsa is still used by certain subdivisions of the Daśanāmīs.97 Both have a military unit (akhāṛā) called (Mahā) nirvāṇi.

Today, the Rāmānandīs are the largest ascetic order in India, and ascetics who worship Rāma have been part of the North Indian religious landscape since at least the twelfth century.98 But our Mughal miniatures have shown us only Daśanāmī Saṃnyāsīs and Nāths. Where were the ascetic worshippers of Rāma hiding? A close inspection of Akbar Watches a Battle between Two Rival Groups of Saṃnyāsīs at Thaneshwar and a folio from Jahangir’s 1618 Gulshan Album tells us that they are right before our eyes: the forerunners of the Rāmānandīs were Saṃnyāsīs.99 Some of the yogi warriors in the Akbarnāma depiction of the battle at Thaneshwar have, in addition to Vaiṣṇava insignia, words written on their bodies. Only one word—ramā—is discernible, on the chest of a Saṃnyāsī in the bottom right (figs. 2, 24). And we can see similar markings on the body of a Vaiṣṇava in a beautiful collage of paintings from the Gulshan Album, which depicts a Nāth yogi encountering a Vaiṣṇava ascetic very similar to the Thaneshwar Saṃnyāsīs (fig. 25). The words are not clearly written—one wonders how good the Devanāgarī orthography of the Mughal court painters was—but rāma is the most likely reading.

Fig. 24 Detail, Akbar Watches a Battle between Two Rival Groups of Saṃnyāsīs at Thaneshwar, 1590–95. Victoria and Albert Museum, IS.2:62-1896 (see fig. 2)
Fig. 25 Folio from the Gulshan Album.
India, Mughal dynasty, first quarter of the 17th century. Opaque watercolor and gold on paper, 53.5 x 40 cm.
Staatsbibliothek zu Berlin, Preussischer Kulturbesitz, Orientabteilung, Libri pict. A 117, f.6b
Fig. 26 Rāmānandī Tyāgī with “Rām Rām” written on his forehead, Kumbh Mela, Allahabad, February 2013; photo © Cambridge Jones










In matters of doctrine, the Saṃnyāsī tradition is most closely associated with the rigorous philosophies of Vedānta. Bhakti (devotion), however, has held an important, if overlooked, place in their teachings,100 and some medieval North Indian Saṃnyāsī ācāryas were renowned for their devotion to Rām.101 The formalization of the Saṃnyāsī order involved the incorporation of a broad variety of different renouncer traditions, whose followers considered themselves part of the ancient tradition of renunciation (saṃnyāsa). In the course of the seventeenth and eighteenth centuries, the generic name for a renouncer, Saṃnyāsī, became associated with this formalized order. When the Rāmānandīs seceded from it in the course of their adoption of ultra-Vaiṣṇavism, their ascetics differentiated themselves from the Saṃnyāsīs by giving themselves the name Tyāgī, which is an exact Sanskrit synonym of Saṃnyāsī (fig. 26). In a similar fashion, as Nāth corporate identity solidified in the eighteenth century, the name Yogī came to be associated exclusively with the Nāths and was shunned by the Saṃnyāsīs and Rāmānandīs.

The Śaivism of the Daśanāmī Saṃnyāsīs and Vaiṣṇavism of the Rāmānandīs, while ostensibly responsible for a lengthy, and sometimes lethal, antipathy, should be taken with a pinch of salt. Doctrinal differences are highlighted in texts composed by the learned of both traditions but, as noted above, the rank-and-file yogis were (and remain) very similar, and their shared Sant heritage of anti-scholastic nirguṇabhakti is still prevalent today. The Śaiva and Vaiṣṇava denominations were adopted in the course of the consolidation of the two orders and provided a convenient ideological justification for what was in fact competition over resources rather than a dispute over doctrine.102 Not only do the ascetics of both orders lead very similar lives, but many features of the two orders fly in the face of their supposed incompatibility. An important Saṃnyāsī commander of the late eighteenth century, when battles between the two orders were at their fiercest, was called Rāmānand Gosāīṃ.103 At the 2010 Haridwar Kumbh Melā, I met a Saṃnyāsī called Rāmānand Giri in the Saṃnyāsīs’ Jūnā Akhāṛā. Recently, when making inquiries in Himachal Pradesh about historical religious affiliations, my informants were confused by my attempts to categorize local rulers or religious institutions as exclusively Vaiṣṇava or Śaiva. Taruṇ Dās Mahant, a householder Rāmānandī from Kullu, told me that “here the devotees of Rām all worship Śiva and the devotees of Śiva all worship Rām.”104

There has long been confusion over the identity of the yogis depicted in Mughal and later paintings. This has resulted from a lack of understanding of the complex and constantly changing makeup of yogi sects in the early modern period, and the concomitant absence of terminological rigor in both Indian and foreign descriptions of yogis from the Mughal period to the present day. Yet a close reading of these pictures and other historical sources allows us to identify the sectarian affiliations of the depicted yogis and thereby to cast new light on their history and the nature of the yoga that they practiced. The pictures’ naturalism and the associated consistency of their depictions mean that seemingly insignificant details, such as the position of an earring, are of great significance.

Mughal-era and later paintings provide evidence for, and have inspired, many of the new ways of looking at Indian yogis and their history outlined in this essay. Doubtless some of the theories proposed will be rejected or refined in the light of further research—whether textual, ethnographic, or art historical—but the details shown in these beautiful images, which have hitherto been overlooked in histories of yoga and yogis, need to be addressed by historians. They bear testament to the fluidity of India’s religious landscape and the transformations undergone by her yogis as they adapted to the changes around them.

I am grateful to Debra Diamond, Jane Lusaka, Bruce Wannell, Monika Horstmann, Arik Moran, Susan Stronge, Patton Burchett, Lubomír Ondračka, Anand Venkatkrishnan, Dominic Goodall, Jason Birch, Jerry Losty, Sunil Sharma, Péter-Dániel Szántó, Véronique Bouillier and Holly Shaffer for their useful comments on earlier drafts of this essay. Many of the arguments rehearsed had their first airing in a Mellon Foundation lecture I gave at Columbia University on September 29, 2011, at the kind invitation of Sheldon Pollock. I thank him and the audience there for their constructive criticism. I also received useful feedback from the members of the panel on “Yogis, sufis, devotees: religious/literary encounters in pre-modern and modern South Asia” at the European Conference on South Asian Studies in Lisbon, July 27, 2012. Many people have provided me with scans of images of yogis that I refer to in this essay. I would like to thank in particular Debra Diamond, who has sent hundreds of such scans my way. Ludwig Habighorst very kindly allowed me to use scans of pictures from his collection. Thanks too to Malini Roy, who has helped with my repeated requests to see images in the collection of the British Library.

James Mallinson, PhD, is a Sanskritist from Oxford University whose work focuses on the history of yoga and yogis. His publications include The Ocean of the Rivers of Story by Somadeva (2007) and The Khecarīvidyā of Ādinātha (2007).