haṭhayoga philosophy

18
Haṭhayoga’s Philosophy: A Fortuitous Union of Non-Dualities James Mallinson Oriental Institute University of Oxford [email protected] 0044 1672852294 0044 7973615591 Unpublished draft submitted to the Journal of Indian Philosophy Please do not cite without the author’s permission Abstract: In its classical formulation as found in Svātmārāma’s Haṭhapradīpikā, haṭhayoga is a Śaiva appropriation of an older extra-Vedic soteriological method. But this appropriation was not accompanied by an imposition of Śaiva philosophy. In general, the texts of haṭhayoga reveal, if not a disdain for, at least an insouciance towards metaphysics. Yoga is a soteriology that works regardless of the yogin’s philosophy. But the various texts that were used to compile the Haṭhapradīpikā (a table identifying these borrowings is given at the end of the article) were not composed in metaphysical vacua. Analysis of their allusions to doctrine shows that the texts from which Svātmārāma borrowed most were products of a Vedantic milieu - bearing testament to Vedānta’s newfound interest in yoga as a complement to jñāna - but that many others were Śaiva non-dual works. Because of the lack of importance given to the niceties of philosophy in haṭhayogic works, these two non-dualities were able to combine happily and thus the Śaiva tenets incorporated within haṭhayoga survived the demise of Śaivism as part of what was to become in the medieval period the dominant soteriological method in scholarly religious discourse in India. Keywords: yoga, haṭha, Haṭhapradīpikā, Śaivism, Vedānta

Upload: jacob-barger

Post on 23-Jan-2016

11 views

Category:

Documents


1 download

DESCRIPTION

Haṭhayoga Philosophy

TRANSCRIPT

Page 1: Haṭhayoga Philosophy

Haṭhayoga’s Philosophy: A Fortuitous Union of Non-Dualities

James MallinsonOriental Institute

University of Oxford

[email protected] 16728522940044 7973615591

Unpublished draft submitted to the Journal of Indian PhilosophyPlease do not cite without the author’s permission

Abstract: In its classical formulation as found in Svātmārāma’s Haṭhapradīpikā, haṭhayoga is a Śaiva appropriation of an older extra-Vedic soteriological method. But this appropriation was not accompanied by an imposition of Śaiva

philosophy. In general, the texts of haṭhayoga reveal, if not a disdain for, at least an insouciance towards metaphysics. Yoga is a soteriology that works regardless of the yogin’s philosophy. But the various texts that were used to compile

the Haṭhapradīpikā (a table identifying these borrowings is given at the end of the article) were not composed in metaphysical vacua. Analysis of their allusions to doctrine shows that the texts from which Svātmārāma borrowed most

were products of a Vedantic milieu - bearing testament to Vedānta’s newfound interest in yoga as a complement to jñāna - but that many others were Śaiva non-dual works. Because of the lack of importance given to the niceties of

philosophy in haṭhayogic works, these two non-dualities were able to combine happily and thus the Śaiva tenets incorporated within haṭhayoga survived the demise of Śaivism as part of what was to become in the medieval period

the dominant soteriological method in scholarly religious discourse in India.

Keywords: yoga, haṭha, Haṭhapradīpikā, Śaivism, Vedānta

Page 2: Haṭhayoga Philosophy

Rigorous philological study of haṭhayoga is in its infancy despite the global popularity of its practices and derivatives thereof. One reason for this may be that the root texts of haṭhayoga and their exegesis make little room for discussions of philosophy and so philologists of the usual bent have not been attracted to their study. Until recently Śaivism had suffered the same fate. For a long time its notoriety deterred scholars from its study, but the richness of its root texts and their philosophically underpinned exegesis has proved irresistible to a new generation of indologists, who have discovered that the study of Śaivism can not only satisfy their desire to understand more of India’s pre-modern religious culture through philology but also stimulate their philosophically enquiring minds.

In this paper I wish to draw on recent text-critical study of works on haṭhayoga in order to make preliminary assessments of the place of philosophy therein. In doing so I hope to encourage scholars better qualified than I am to examine the minutiae of philosophical pronouncements within the haṭhayogic corpus. While such study may not unearth the treasures of the Kashmiri Śaiva exegesis, it will, I believe, contribute to our understanding of the development of Indian religion in the medieval period by revealing more details of how haṭhayoga, elements of which can be traced to all three of the first-millennium Brahmanic religions of India, i.e. Vedism, Vaiṣṇavism and Śaivism, enabled some of the tenets of the last to remain part of widely accepted soteriological methods even as Vedānta-influenced manifestations of the first became the dominant paradigm of scholarly religious thought.1

Śaivism’s Appropriation of Haṭhayoga

Before we turn to the philosophical teachings found in haṭhayogic texts, in order to delineate the relationship between Śaivism and haṭhayoga, attention will be drawn to the way in which the latter was appropriated by the former.

Classical haṭhayoga as formulated in its locus classicus, the fifteenth-century Haṭhapradīpikā, combines elements from a wide range of yogic teachings, but in essence comprises the gross physical techniques of an ancient extra-Vedic ascetic tradition overlaid with subtle visualisation-based Śaiva yoga. The purpose of Svātmārāma, the Haṭhapradīpikā’s compiler, was to lay claim to this new synthesis for a broad tradition of Śaiva siddha schools. In this he was entirely successful, although the parameters of the tradition he invoked were subsequently narrowed, with haṭhayoga’s originators coming to be identified with the first gurus of the then fledgling Nāth saṃpradāya.2

The distinguishing characteristic of haṭhayoga in its pre-Haṭhapradīpikā formulations is the practice of physical techniques known collectively in the Haṭhapradīpikā as mudrās, which are used to make the breath enter the central channel and to raise bindu, semen, up to its source in the head and keep it there. We find forerunners of two of these techniques, the relatively simple constrictions known as mūlabandha and jālandharabandha,in Śaiva works, but the quintessentially haṭhayogic techniques of khecarīmudrā and vajrolimudrā are first taught in a Vaiṣṇava text, the Dattātreyayogaśāstra, in which they are used for the preservation of bindu, unlike in subsequent formulations in which they are overlaid with Śaiva features such as the raising of Kuṇḍalinī, the flooding of the body with amṛta, and the absorption of commingled sexual fluids (Mallinson, “Śāktism and Haṭhayoga”, forthcoming).

It is in the Haṭhapradīpikā that the practice of non-seated āsanas, with which haṭhayoga has come to be identified, is first taught as one of its key components.3 Of the fifteen āsanas taught therein, eight are relatively simple seated

1 Doubtless some elements of haṭhayogic practice originated outside of Brahmanic religion, in particular within the śramaṇa ascetic traditions (see Mallinson forthcoming), but when taught in the Sanskrit texts on haṭhayoga under consideration here they are of course couched in the language of Brahmanic religion.

2 The mahāsiddhas listed by Svātmārāma at Haṭhapradīpikā 1.5-8 include the Vīraśaiva Allāma Prabhu, who in contemporaneous hagiography is portrayed as a rival of Gorakṣa, the supposed founder of the Nāth saṃpradāya (Śūnyasaṃpādane upadeśa 21).

3 haṭhasya prathamāṅgatvād āsanaṃ pūrvam ucyate |kuryāt tad āsanaṃ sthairyam ārogyaṃ cāṅgalāghavam || 1.17 || ...āsanaṃ kumbhakaṃ citraṃ mudrākhyaṃ karaṇaṃ tathā |

Page 3: Haṭhayoga Philosophy

postures for meditation, such as those found in a wide variety of earlier texts including Vyāsa’s Bhāṣya on the Yogasūtras (ad 2.46), several Śaiva Tantras (from the earliest we have, the Niśvāsatattvasaṃhitā, onwards),4 and various Purāṇas.5 Of the seven non-seated āsanas taught in the Haṭhapradīpikā, one is śavāsana, the corpse pose, which is a reworking of a posture taught not as an āsana but as a technique of layayoga in the Dattātreyayogaśāstra.6 Another, paścimottānāsana, is taught in the Śivasaṃhitā, which was probably composed not long before the Haṭhapradīpikā.7 I have failed to identify source passages for three of the non-seated āsanas taught in the Haṭhapradīpikā, namely uttānakūrmaka, dhanur and matsyendra āsanas. The verses describing the remaining two, kukkuṭa and mayūra āsana, are taken from the Vasiṣṭhasaṃhitā, a c.13th-14th-century Vaiṣṇava work.8 The Vasiṣṭhasaṃhitā’s teachings on mayūrāsana can be traced back through a variety of earlier Vaiṣṇava works to the c.9th-century Vimānārcanākalpa,9 a Vaikhānasa Saṃhitā, which teaches mayūrāsana as one of the lowest of three grades of āsana. Its prose description is found reworked into ślokas in the Pādmasaṃhitā,10 a Pāñcarātrika Saṃhitā. Those ślokas are found with some changes in another Pāñcarātrika Saṃhitā, the Ahirbudhnyāsaṃhitā,11 and again in the Vasiṣṭhasaṃhitā.12 The Vasiṣṭhasaṃhitā’s verses are then found in the Haṭhapradīpikā13 with the metre changed to upajāti. Kukkuṭāsana is not taught in the Vimānārcanākalpa, Pādmasaṃhitā or Yogayājñavalkya, but the verses describing it in the Haṭhapradīpikā are found in the Ahirbudhnyāsaṃhitā and Vasiṣṭhasaṃhitā in very similar forms.14

On its own this evidence would not point to much; all we could say is that the oldest known textual evidence for non-seated āsanas is found in the Vaiṣṇava tradition. What makes it of particular interest is that the Matsyendrasaṃhitā, a

atha nādānusandhānam abhyāsānukramo haṭhe ||1.56.4 Niśvāsatattvasaṃhitā Nayasūtra 4.14c-15d. See Goodall 2004:348-351 nn. 728-732 and Vasudeva 2004:397-402 for

references to āsana in other Śaiva sources.5 E.g. Skandapurāṇa 179.27, Mārkaṇḍeyapurāṇa 36.28.6 Compare Dattātreyayogaśāstra 24c-25b with Haṭhapradīpikā 1.32.7 Compare Śivasaṃhitā 3.108-109 with Haṭhapradīpikā 1.28-29. The Śivasaṃhitā borrows and paraphrases several

verses from both the Amṛtasiddhi and Dattātreyayogaśāstra: Śivasaṃhitā 2.1b, 2.1cd, 2.2, 2.3, 2.4ab = Amṛtasiddhi 1.15b, 1.16ab, 1.17ab+1.16cd, 1.17c-1.18b, 1.19ab; Śivasaṃhitā 2.6c-9d, 2.11-12, 3.31, 4.27ab, 4.27dc, 4.28ab, 4.34cb, 5.13, 5.17c-20b = Amṛtasiddhi 3.1-4, 4.3-4, 19.2, 11.3cd, 11.4bc, 11.5ab, 11.7cd, 15.1, 16.1-3; Śivasaṃhitā 4.31, 4.38, 5.14-15 ≈ Amṛtasiddhi 11.6, 12.6, 15.3. Several other verses in the Śivasaṃhitā’s descriptions of mahāmudrā, mahābandha and mahāvedha are derivative of verses in the Amṛtasiddhi. Śivasaṃhitā 3.44a-3.45e, 3.48ab, 3.62ab, 3.63ab, 3.102-105, 4.88ab, 5.71 = Dattātreyayogaśāstra 143-146a, 155, 177, 178, 68-75, 313, 43-44. Dattātreyayogaśāstra 143-162, 195-198, 221-241 are paraphrased at Śivasaṃhitā 3.42-48, 3.60-61, 3.72-75. 8.5 of the 13.5 verses that the Śivasaṃhitā shares with the Haṭhapradīpikā are not to be found in other texts so it is likely that the Śivasaṃhitā was composed at some time between the Dattātreyayogaśāstra and Haṭhapradīpikā.

8 The Vaiṣṇavism of the Vasiṣṭhasaṃhitā is not stated explicitly but Vasiṣṭha bows to Jagannātha before beginning his exposition of Brahmā’s yoga (1.12) and the saguṇa dhyāna that he teaches is of Viṣṇu Nārāyaṇa (4.26-31, 49-53).9 Vimānārcanākalpa paṭala 96: karatale bhūmau saṃsthāpya kūrparau nābhipārśvayor nyasya na(unnata)taśirāḥpādau

daṇḍavad vyomni saṃsthito mayūrāsanam iti |Colas (1988:279) remarks that the yogas of the Vimānārcanākalpa, Ahirbudhnyāsaṃhitā and Vasiṣṭhasaṃhitā are

similar but does not note the textual parallels in their descriptions of mayūrāsana. He dates the Vaikhānasa Saṃhitās to the 9th to 13th/14th centuries and tentatively identifies the Vimānārcanākalpa as the earliest (2010:158).

10 Pādmasaṃhitā yogapāda 1.21-22:avaṣṭabhya dharāṃ samyak talābhyāṃ hastayor dvayoḥ ||kūrparau nābhipārśve ca sthāpayitvā mayūravat |samunnamya śiraḥpādau mayūrāsanam iṣyate ||11 Ahirbudhnyāsaṃhitā 31.36-37:niveśya kūrparau samyaṅ nābhimaṇḍalapārśvayoḥ |avaṣṭabhya bhuvaṃ pāṇitalābhyāṃ vyomni daṇḍavat ||samonnataśiraḥpādo māyūrāsanam iṣyate |etat sarvaviṣaghnaṃ ca sarvavyādhinivāraṇam ||12 Vasiṣṭhasaṃhitā Yogakāṇḍa 1.76-77 (which is also found at, and is likely to be the source of, Yogayājñavalkya 3.15a-17b):avaṣṭabhya dharāṃ samyak talābhyāṃ ca karadvayam |hastayoḥ kūrparau cāpi sthāpayan nābhipārśvayoḥ ||samunnataśiraḥpādo daṇḍavad vyomni saṃsthitaḥ |mayūrāsanam etad dhi sarvapāpavināśanam ||76c kūrparau] em.; kharpare ed.13 Haṭhapradīpikā 1.30:dharām avaṣṭabhya karadvayenatatkūrparasthāpitanābhipārśvaḥ |uccāsano daṇḍavad utthitaḥ khemāyūram etat pravadanti pīṭham ||14 Ahirbudhnyāsaṃhitā 31.38, Vasiṣṭhasaṃhitā Yogakāṇḍa 1.78, Haṭhapradīpikā 1.23.

Page 4: Haṭhayoga Philosophy

13th-century Śaiva work attributed to Matsyendra, one of the gurus of the Nātha saṃpradāya, also includes mayūra and kukkuṭa āsanas among the fourteen āsanas it teaches.15 The descriptions are rather obscure (and the text a little corrupt), but it is clear that they are different āsanas from those taught in the Vaiṣṇava works mentioned above and that like all the other āsanas taught in the Matsyendrasaṃhitā they are seated postures.16 We cannot be sure that the Matsyendrasaṃhitā was known to Svātmārāma, but there were certainly overlaps in their milieux: the Matsyendrasaṃhitā incorporates most of the Khecarīvidyā, from which four verses were borrowed to compile the Haṭhapradīpikā, and one other half-verse in the Matsyendrasaṃhitā is also found in the Haṭhapradīpikā.17 We can thus say with some confidence that the practice of non-seated āsanas originated in a Vaiṣṇava rather than Śaiva milieu.

Haṭhayoga’s Universalism

In the light of this Śaiva appropriation of the techniques of haṭhayoga one might expect Svātmārāma also to declare haṭhayoga to be grounded in one or other formulation of Śaiva metaphysics. Yet, as much as any other work that teaches haṭhayoga, the Haṭhapradīpikā is devoid of detailed philosophical teachings of any kind. Furthermore, its Śaiva orientation, which is to be inferred by its maṅgala verses (1.1 and 4.1) and invocations of Śaiva siddhas (1.5-8), is not corroborated by descriptions of sect-markers such as mantras or maṇḍalas, nor even of the cakras that came to be synonymous with the practice of haṭhayoga. What we see here is the apotheosis of a process identified by Csaba Kiss in his analysis of the Matsyendrasaṃhitā, a text which he suggests is indicative of

“a phase in the history of yoga when yogic teachings tried to become detached, perhaps not for the first time, from the mainstream religion, in this case tantric Śaivism, by eliminating sectarian boundaries through the concealment of sectarian marks such as easily decodable deity names, mantras, and iconography and to prepare for a formative period of pan-Indian yoga, which can again become an alternative for the official/conservative religion.” (Kiss 2011:162).

Yet Svātmārāma, despite going further down the line towards universalism than the compiler of the Matsyendrasaṃhitā, did not go as far as he might have done. The Haṭhapradīpikā, hamstrung by its purpose of claiming haṭhayoga for the Śaiva siddha tradition, makes only a half-hearted claim to universality when it states that through the practice of haṭhayoga success can be attained by “the young, old, very old, ill or weak”.18 This verse is taken from one of the fifteen or so texts that Svātmārāma used to compile the Haṭhapradīpikā, namely the Dattātreyayogaśāstra, which, as noted above, is a product of a Vaiṣṇava school in the tradition of extra-Vedic asceticism in which some of the physical practices of haṭhayoga first developed. The Dattātreyayogaśāstra, perhaps because its compiler, unlike Svātmārāma, had no sectarian axe to grind, is more generous in its universalism than the Haṭhapradīpikā, following the verse borrowed by Svātmārāma with the following two:

brāhmaṇaḥ śramaṇo vāpi bauddho vāpy ārhato ’thavā|

15 Matsyendrasaṃhitā 3.8a-13b:vāmagulpham adhaḥ kṛtvā kṛtvopari ca dakṣiṇam |gudenāpīḍya vai jānuyugalaṃ paripālya ca ||8||vāmahastāṅgulīmūle-m-avaṣṭabhya yathābalam |dakṣahastāṅgulīmūlaiḥ parivartya ca tat talau ||9||ūrumūladvaye pīḍya nāsāgre sthāpayed dṛśau |kukkuṭāsanam etad vai manasaḥ sthirakāraṇam ||10||ūrumūladvayādhastāt pādayugmaṃ nidhāyā ca |tayor madhye +sthira guda pṛchan+ udyamya niścalaḥ ||11||tathā karataladvandvaṃ samāśliṣya parasparam |ūrvor upari vinyasya nāsikām avalokayet ||12||mayūrāsanam etad vai sarvavyādhivināśanam |Cf. Kubjikāmata 23.115-117 which also teaches a seated kukkuṭāsana.16 Kiss (2009:52) qualifies his statement that the āsanas in the Matsyendrasaṃhitā are all seated positions with

“probably”, noting that the text is corrupt.17 On the inclusion of the Khecarīvidyā in the Matsyendrasaṃhitā, see Mallinson 2007:5-9. Matsyendrasaṃhitā

4.44ab = Haṭhapradīpikā 4.17cd. Matsyendrasaṃhitā 4.23cd is the same as Haṭhapradīpikā 2.17ab, but it is also shared with Vivekamārtaṇḍa 100ab, which, since Haṭhapradīpikā 2.16-17 corresponds to Vivekamārtaṇḍa 99-100, is almost certainly the Haṭhapradīpikā’s source. At Haṭhapradīpikā 1.18 Svātmārāma says that he will teach āsanas that were accepted by “munis such as Vasiṣṭha and yogis such as Matsyendra”. None of the Haṭhapradīpikā’s āsana teachings has parallels in the Matsyendrasaṃhitā. Variants of siddhāsana and padmāsana are said at 1.35 and 1.48 to be matsyendramata, but the verses teaching them are from the Vivekamārtaṇḍa (7, 35B), which is attributed to Gorakṣa in its earliest manuscript and subsequent citations. Matsyendrāsana is taught (as śrīmatysanāthodita āsana and matsyendrapīṭha) at Haṭhapradīpikā 1.26-27 in verses which I have not found in earlier works.

18 Haṭhapradīpikā 1.64: yuvā vṛddho 'tivṛddho vā vyādhito durbalo 'pi vā |abhyāsāt siddhim āpnoti sarvayogeṣv atandritaḥ ||

Page 5: Haṭhayoga Philosophy

kāpāliko vā cārvākaḥ śraddhayā sahitaḥ sudhīḥ ||41||yogābhyāsarato nityaṃ sarvasiddhim avāpnuyāt |kriyāyuktasya siddhiḥ syād akriyasya kathaṃ bhavet ||42||

“Whether a Brahmin, an ascetic, a Buddhist, a Jain, a Skull-Bearer or a materialist, the wise man who is endowed with faith and constantly devoted to the practice of [haṭha] yoga will attain complete success.”

Such universalism is at least implicit in most works on haṭhayoga. Teachings on yoga in earlier Śaiva works, in contrast, are for initiates into the traditions of which their texts are products,19 and often involve a progression of meditative conquests of the elements particular to their tradition’s ontology. A tattvajaya of this kind is found in the Niśvāsatattvasaṃhitā’s Nayasūtra (paṭala 3) and similar techniques feature in a wide range of subsequent Śaiva works. Haṭhayogic texts, on the other hand, restrict such practices to dhāraṇās on the five elements accepted by all Indic metaphysical systems or laya, dissolution, into those five elements, often in the course of Kuṇḍalinī’s rise up the central channel.

A relatively late Śaiva work, the Śāradātilaka,20 shares verses and practices with some haṭhayogic texts and also echoes their universalist understanding of yoga. At the beginning of its 25th and final paṭala we read:

atha yogaṃ pravakṣyāmi sāṅgaṃ saṃvitpradāyakam|aikyaṃ jīvātmanor āhur yogaṃ yogaviśāradāḥ||1||śivātmanor abhedena pratipattiḥ pare viduḥ|śivaśaktyātmakaṃ jñānaṃ jagur āgamavedinaḥ||2||purāṇapuruṣasyānye jñānam āhur vīśāradāḥ|

“Now I shall teach yoga, with its ancillaries, which bestows understanding [or “mokṣa in the form of the experience of eternal bliss” - Rāghavabhaṭṭa]. The experts in yoga say that yoga is union of the jīva and the ātman. Others say that it is knowledge of Śiva and the ātman as not being different. Those who know the āgamas say that it is knowledge of the nature of Śiva and Śakti. Other wise ones say that it is knowledge of the ancient puruṣa.”

Rāghavabhaṭṭa, the fifteenth-century commentator on the Śāradātilaka, understands these four different opinions to be those of Vedāntins, Śaivas, Śāktas and Bhedavādins (whom he identifies as Sāṃkhyas, Vaiṣṇavas and Naiyāyikas, the latter two identifying puruṣa with Nārāyaṇa and Īśvara respectively).Then, with no further ado, the text goes on to teach an eight-fold system of yoga, the implication being that the practice of yoga will get the yogi the reward he or she wants, regardless of the yogi’s philosophical standpoint.

Concomitant with this understanding of yoga is the widespread notion of yogipratyakṣa,“yogic perception”, with which the founders of a variety of traditions are said to have discovered the ultimate truth, whether that truth is conceived of as, for example, the Vedas for Naiyāyikas or śūnyatā for Mādhyamika Buddhists. The founder of the tradition did not need to know these truths to succeed in yoga; they were revealed as a result of his success in yoga. Later followers of such yogis have no need to rediscover these truths; they use meditation to deepen their understanding of them (Franco 2009a:9-10).21

Of course the yogic universalism espoused in varying degrees by the Śāradātilaka, Dattātreyayogaśāstra and Haṭhapradīpikā does not mean that those texts were created in a doctrinal vacuum, nor is that same universalism found in all of the texts used by Svātmārāma to compile the Haṭhapradīpikā. In what follows I shall examine those texts in order to make a preliminary assessment of their philosophical standpoints. More detailed analysis by specialists of the various traditions will no doubt cast a brighter light on the texts’ relationship with specific philosophical schools.

The Philosophical Orientation of the Haṭhapradīpikā’s Source Texts22

The Vivekamārtaṇḍa, Gorakṣaśataka and Dattātreyayogaśāstra are three of the four texts which contribute twenty or more verses to the Haṭhapradīpikā. These three texts are all products of schools influenced by Vedanta. The opening

19 E.g. Mālinīvijayottaratantra Yogapāda 4.6cd: na cādhikāritā dīkṣāṃ vinā yogo ’sti śāṅkare ||20 Sanderson (2007:230-233) gives the 13th century as the terminus ante quem of the Śāradātilaka, adding that it is likely to have been composed in Orissa.21 Franco (2009:6 n.22) distinguishes between yoga as “a technique of gaining control over the body, senses and mind

in order to attain a liberating insight” and Yoga, the philosophical school. He adds that yoga “is a technique or a method and as such is not connected to any philosophy or religion in particular”; this is the understanding of yoga found in haṭhayogic texts.

22At the end of this paper is an appendix in which all the texts from which Svātmārāma borrowed verses are listed, together with the locations of the shared verses in the Haṭhapradīpikā and the source texts.

Page 6: Haṭhayoga Philosophy

verse of the Dattātreyayogaśāstra23 nails its Vaiṣṇava and vedantic colours to the mast:

nṛsiṃharūpiṇe cidātmane sukhasvarūpiṇe|padais tribhis tadādibhir nirūpitāya vai namaḥ||1|

“To him who has the form of Narsiṃha, whose self is consciousness, whose true form is bliss, and who is defined by the three words beginning with tat, homage!”

Elsewhere in the Dattātreyayogaśāstra, samādhi is said to be the union of ātman and paramātman (126ab), and when the yogin wants to cast off his body he is to dissolve it into parabrahman (127ab). Yet jñāna, the key to vedantic liberation, has almost no place in the Dattātreyayogaśāstra. It is mentioned only once - and then somewhat disparagingly - in the context of mantrayoga, the lowest of the three yogas taught in the text (the others being laya and haṭha). Through twelve years of mantra-repetition, the sādhaka will, usually (prāyeṇa), attain jñāna and the siddhis of aṇimā and so forth. This yoga is for the lowest type of sādhaka, he of little wisdom (alpabuddhiḥ).24

Both the Vivekamārtaṇḍa and Gorakṣaśataka combine Śaiva yoga25 with vedantic metaphysics.26 There is no internal evidence to identify the place of composition of either text, but parallels with the more or less contemporaneous 27 Bhāvārthadīpikā, Jñāndev’s Marathi commentary on the Bhagavadgītā (popularly known as the Jñāneśvarī), together with other external evidence, suggest origins in the Deccan. Like the Vivekamārtaṇḍa and Gorakṣaśataka, the Jñāneśvarī teaches physical Śaiva yoga and the vedantic identification of ātman and brahman. The yoga of the Gorakṣaśataka in particular is very close to that of the Jñāneśvarī.28 The teachings in both the Vivekamārtaṇḍa and Gorakṣaśataka are attributed to Gorakṣa, and Jñāndev names Gorakṣa among his teachers at Jñāneśvarī 18.1756. Gorakṣa is said to have come from several different parts of India, but the majority of the early references to him are from the south or the Deccan.29 The title of the Vivekamārtaṇḍa, as well as being suggestive of a vedantic, or at least

23 nṛsiṃharūpiṇe cidātmane sukhasvarūpiṇe|padais tribhis tadādibhir nirūpitāya vai namaḥ||24 Dattātreyayogaśāstra 12-14:aṅgeṣu mātṛkānyāsapūrvaṃ mantraṃ japet sudhīḥ |yena kenāpi sādhyaḥ syān mantrayogaḥ sa kathyate ||12||mṛdus tasyādhikārī syād dvādaśābdais tu sādhanāt |prāyeṇa labhate jñānaṃ siddhīś caivāṇimādikāḥ||13||alpabuddhir imaṃ yogaṃ sevate sādhakādhamaḥ |mantrayogo hy ayaṃ prokto yogānām adhamas smṛtaḥ ||14||12c yena kenāpi sādhyaḥ syān ] em; yena kenāpi siddhaḥ syāt Haṭharatnāvalī 1.9c, yaṃ kañcanābhisiddhyai B, yaṃ kaṃ

canābhisiddhyai W1, ekena cāpi siddhiḥ M, ekaṃcanābhisidhyai J1, ekaṃ ca tābhiḥ sidhyai W225 The earliest Śaiva works to teach the central practices of what came to be known as haṭhayoga, i.e. the Amṛtasiddhi,

Vivekamārtaṇḍa and Gorakṣaśataka, do not call their yoga haṭha.26 See Vivekamārtaṇḍa 88 and 151-152 for evidence of its Śaiva orientation. The Gorakṣaśataka is not so explicit, but

vv. 13, 80 indicate that Śiva is the highest god. For the texts’ vedantic teachings, see Vivekamārtaṇḍa 106-110, 153-60, 164, 170-171 and Gorakṣaśataka 88-100.

27 Jñāneśvarī 18.1711 says that the text was written in 1290 CE and this date is widely accepted, but see Kiehnle (1997:5) on the likelihood of this particular line having been interpolated at a later date. Our earliest definite evidence for the existence of the Vivekamārtaṇḍa is its being named in the c. 1400 CE Khecarīvidyā (1.14), although five of its verses are found in the 1363 CE Śārṅgadharapaddhati (Śārṅgadharapaddhati 4308a-4309b, 4374, 4407abc, 4418 = Vivekamārtaṇḍa 27a-28b, 7, 57abc, 59). The terminus ante quem of the Gorakṣaśataka is also 1400 CE (Mallinson 2011b:263).

28 Their yogas use the three bandhas, jālandhara° (Jñāneśvarī 6.207-208; Gorakṣaśataka 61c-63b), uḍḍīyāna° (Jñāneśvarī 6.209-210; Gorakṣaśataka 57c-61b) and mūla° (Jñāneśvarī 6.192-199; Gorakṣaśataka 52c-57b), to raise Kuṇḍalinī.

29 Our earliest references to Gorakṣa are from the 13th century. Despite frequent claims in secondary literature that he came from Punjab (e.g. Briggs 1989:229), none of these early references is from the northwest of the subcontinent. In addition to that from the Jñāneśvarī, other 13th-century references to Gorakṣa from the Deccan or south India include the c.1200-1220 CE Kannada Ragales of Harihara (Revaṇasiddheśvara Ragale, Sthala 3, ll.25-65; I thank Professor Prithvi Datta Chandra Shobhi for confirming Harihara’s dates and assisting me with the text - email communication June 2009); the Matsyendrasaṃhitā (paṭalas 1 and 55, the frame story; on the text’s date and provenance, see Kiss 2009:26-28); the 1279 CE Kalleśvara temple inscription from the Chitaldroog district in Karnataka (Saletore 1937); the Marathi Līḷācaritra (although its surviving recension may not be as old as the 13th century (Tulpule 1979:319)). 13th-century references to Gorakṣa from places other than the Deccan or south India include the 1287 CE Cintra Praśasti inscription from Somnath (verse 42 - Bühler 1892:284); the Amṛtakaṇikodyotanibandha which is probably from the Bengal region (for bibliographic reference see Āryamañjuśrīnāmasaṃgīti; on the life of the text’s author, Vibhūticandra, see Stearns 1996); Abhayadattaśrī's c.13th-century Tibetan account of the lives of the 84 Siddhas (I thank Professor Harunaga Isaacson for alerting me

Page 7: Haṭhayoga Philosophy

jñāna-oriented, worldview, is paralleled by two early Marathi works, the Vivekadarpaṇ and Vivekasindhu, of which the latter is said to be the first Marathi vedantic work (Vaudeville 1987:218). This evidence suggests that the Vivekamārtaṇḍa and Gorakṣaśataka represent the Sanskrit textual underpinnings of the integration of Śaiva physical yoga and Vedanta in early medieval Maharashtra.30

The Yogabīja, a dialogue between the goddess and Śiva, teaches a similar combination of Śaiva yoga and vedantic philosophy and so may well be a product of the same milieu as the Vivekamārtaṇḍa and Gorakṣaśataka. Yoga is the union of, among other pairs, ātman and paramātman (89cd). It brings about jīvanmukti (59ab, 170ab, 181cd), in which the body becomes one with brahman (186ab). The Yogabīja is the source of 13 verses in the Haṭhapradīpikā, but of these 8 are also found in the Gorakṣaśataka or Dattātreyayogaśāstra, indicating that the Yogabīja postdates those works. Its yoga is identical to that of the Gorakṣaśataka: śakticālana is used, along with the three bandhas, to raise Kuṇḍalinī, and it teaches the Gorakṣaśataka’s four sahita kumbhakas.

to the Amṛtakaṇikodyotanibandha reference and for sharing with me his opinion of the date and original language of Abhayadattaśrī’s text (email communication April 2009)). Texts associated with Gorakṣa show direct or indirect connections with the Kaula Paścimāmnāya tradition (the Gorakṣasaṃhitā is an inflation of the Kubjikāmatatantra (Heilijgers-Seelen 1994); other works attributed to Gorakṣa, such as the Vivekamārtaṇḍa, teach the Paścimāmnāya system of six cakras). Sanderson notes that the sites of 24 yoginīpīṭhas listed in the Kubjikāmata and other Kaula works “are located in various parts of the subcontinent with eastern India and the Deccan strongly represented. The Far South, the upper Ganges valley, the Panjab, Kashmir and the North-West are absent” (2011:77) and identifies the Maharashtran origins of the Kubjikāmata and another important Paścimāmnāya text, the Manthānabhairava. The latter laments that “the north is full of impediments and devoid of the Siddha lineages”. From Maharashtra manuscripts of these works spread to “Kerala, Tamilnadu, Kashmir, East India and the Kathmandu valley” (ibid.:44-45), a spread matched by that of the Siddha Gorakṣa’s renown. The earliest reference to Gorakṣa from Nepal is the 1382 CE Itum Bahal rock inscription, which describes Madanarāma Varddhana, a senior minister of King Jayasthiti Malla, as gorakhātmajaśiṣya, “a pupil of a descendant of Gorakh” (Vajracarya 1975:34). Narharinath (1953:5) reported that a copperplate inscription from the Maru Sattal temple in the Kāṣṭhamaṇḍapa in Kathmandu’s Durbar Square dated 1379 CE mentions Gorakṣa in its opening line and this assertion has been repeated in other reports of the contents of the inscription and in secondary literature (Slusser & Vajrācārya 2005:451, Locke 1980:435). A new transcription by Kashinath Tamot reveals that Narharināth misread sambata 499 vēśāṣaśukla as samvat 499 devo go(gva)rakṣo (Tamot 2009; I thank Kashinath Tamot and Jason Birch for for helping me with this and other inscriptions from the Kathmandu valley). On links between south India and Nepal in the 11th to 13th centuries, see Michaels 1985 and Lévi 1905:364-365. Gorakṣa is not mentioned in the Śaiva exegesis of Kashmir. Perhaps the earliest text from that region with which he is connected is the Amaraughaśāsana, which is attributed to Gorakṣa in its manuscripts, one of which is dated 1525 (Amaraughaśāsana ed. intro. p.1). The five manuscripts of the text of which I am aware are all in Śāradā script (Ōtani Collection Unidentified Fragment No. 628 (Hori 2005:93-94); Benares Hindu University Acc. No. C4250 and C4723; Oriental Research Library Srinagar No. 2344-1; and the manuscript on which Shastri’s 1918 edition was based, which may be the same as the last). Supporting claims of its Kashmiri origin, the Amaraughaśāsana shares its system of three śaktis with the Netratantra (Amaraughaśāsana 35, 42; Netratantra 7.1-2) and has verses in common with the Siddhasiddhāntapaddhati (Amaraughaśāsana 12-16 = Siddhasiddhāntapaddhati 1.37-41; the rivers listed at Siddhasiddhāntapaddhati 3.11 suggest that the text is from the northwest of the subcontinent) - but not with any earlier works on haṭhayoga. The earliest reference to Gorakṣa from the northwest of the subcontinent is in the Śārṅgadharapaddhati (vv. 4372 and 4373), which was compiled near Jaipur in 1363 CE. Despite numerous reports of interactions between Sufis and Jogis from the thirteenth century onwards (see Digby 1970 and 2000:221-233, 288-291; Rizvi 1971:vii; Ernst 2007), no northwestern vernacular text mentions Gorakh until the Alakhbānī of Gangohi (p.31 l.4, p.40 l.1, p. 46 l.3, p. 57 l.2) which probably dates to 1480 CE (Weightman 1992: 171-172).

30 Jñāndev names the teachers of his tradition as Śrī Tripurāri (Śiva), Macchaprasava (Matsyendra), Cauraṅgī, Gorakṣarāya, Gahinīnāth, and Nivṛttināth (Jñāneśvarī 18.1751-1760). These are all understood in hagiography and secondary literature to have been members of the Nāth saṃpradāya. Similarly, Kiehnle (1997:8-9) notes that in the contemporaneous Līḷācaritra Cakradhar mentions “many Nāths” including “Udhaḷīnāth, Adaṃḍināth, Viśvanāth, Indrāī Luipāī and Jāḷandhar”. What is striking about both these lists is that only those names which are not suffixed with -nāth are recognisable from later lists of “Nāths”. Meanwhile, the guru-paramparā of Mukundarāja, who was approximately contemporaneous with Jñāndev, shows that the tradition of teachers whose names ended in -nāth in thirteenth-century Maharashtra was of a different kind from that associated with today’s Nāth saṃpradāya: Mukundarāja’s two works, the Paramāmṛta and Vivekasindhu are “exposition[s] of the Vedānta written in the peculiar intellectual style of the Vedāntin” and he traces the lineage of his gurus to Ādinātha as Śiva, through his guru Raghunātha and Harinātha (= Cakradhara) (Tulpule 1979:325-326). Because of his identifying Jñāndev’s teachers as Nāths of the same stripe as the later Nāth order, Tulpule has to hypothesise that the order curbed its literary output for nearly 500 years: “A more solid contribution to the Marāṭhī literature of the 18th century came from those poets who belonged to the Nātha cult. They were somewhat dormant after the age of Jñānadeva.” (ibid.:420). Many scholars have fallen upon early instances of the word nāth in a variety of languages as evidence of

Page 8: Haṭhayoga Philosophy

Unlike the aforementioned texts, the Yogabīja sets out the relationship between jñāna and yoga. The first 89 of its 190 verses are an exposition of why yoga must be practised: jñāna is a prerequisite for liberation but yoga is needed too, in order to perfect the body. This is the first articulation within a text that teaches haṭhayoga of a doctrine that was starting to find currency in contemporaneous vedantic works, namely that to achieve jīvanmukti, yoga is needed in addition to jñāna. For Vedāntins, yoga was necessary not to perfect the body but to eradicate the vāsanās and annihilate the mind, as expounded by Vidyāraṇya in the 1380 CE Jīvanmuktiviveka (prakaraṇas 2 and 3).31

Svātmārāma used more than a quarter of the Amaraughaprabodha’s 74 verses when compiling the Haṭhapradīpikā. A Śaiva work32 attributed to Gorakṣa and invoking gurus of the Nāth saṃpradāya in its opening verse, it contains little in the way of metaphysical teachings, or even allusions to metaphysical principles, and in the main describes physical haṭha techniques.

The Vasiṣṭhasaṃhitā, 15.5 verses of which, describing āsanas and prāṇāyāma, were used by Svātmārāma, was composed before 1300 CE (Bouy 1994:82 n.343) and incorporates Kuṇḍalinī yoga within a Vedic and Vaiṣṇava vedantic framework.33 Its philosophical standpoint is bhedābhedavāda, not pure advaita (Vasiṣṭhasaṃhitā introduction p.23). Vasiṣṭha states that both Vedic ritual action and jñāna, which consists of yoga, are needed for liberation (1.29, 31), which is attainable while alive (4.47).

The Amanaska, of which 11 verses are found in the Haṭhapradīpikā, is, at first glance, an odd choice of source text for Svātmārāma. It pours scorn on a variety of soteriological methods including archetypal haṭha practices (2.31-32), espousing instead the practice of the no-mind state of its title. Accordingly, all of the verses which Svātmārāma borrows from it are found in the Haṭhapradīpikā’s fourth upadeśa, which teaches samādhi rather than physical yoga practices. Perhaps it was with a view to catering to the vedantic interest in yoga as a means of eradicating the vāsanās and destroying the mind that he included the Amanaska’s teachings in the Haṭhapradīpikā. At 4.60, in a verse for which I have not identified a source text but which precedes a verse from the Amanaska, Svātmārāma equates the mind with jñāna and says that both must be destroyed.

The Amanaska incorporates both Śaiva and vedantic non-dual terminology.34 Thus the supreme tattva is described in terms redolent of Śaiva non-dualism such as niṣkala (2.41, 77, 91), while the sleep and waking states described at 2.59-64 are derived from the four states of consciousness taught in Gauḍapāda’s Kārikās.

The last text that contributes a significant number of verses to the Haṭhapradīpikā is the Śivasaṃhitā. 8.5 of the 14 verses that they have in common are not found in any other text which predates the Haṭhapradīpikā. The Śivasaṃhitā teaches a yoga grounded in the Vedanta-inflected southern Śrīvidyā tradition, a manifestation of Dakṣiṇāmnāya Kaula Śaivism associated with the Śaṅkarācāryas of Shringeri and Kanchi.35 Its first paṭala sets forth its advaita world-view from the first verse, and dismisses a variety of other doctrines. Elsewhere its Śaiva orientation is made clear. 36 But none of its verses which are found in the Haṭhapradīpikā teaches doctrine; they are exclusively concerned with the physical techniques of haṭhayoga.

the existence of a Nāth saṃpradāya, but in all such instances nāth either refers to a semi-divine being, is being used as an honorific term of address, or, as is the case in the majority of instances, is referring to a god, usually the supreme being. An example from the tradition under examination here is given by Kiehnle, who (2005:484) translates Jñāneśvarī 6.291 (piṃḍeṃ piṃḍācā grāsu | to hā nāthasaṃketicā ḍaṃsu | pari dāunu gelā uddeśu | mahāviṣṇu ||) with “the swallowing of the body by the body, this is the secret in the context of the Nāths, but [here] Mahāviṣṇu gave the explanation”. It is from this verse that Kiehnle takes the title of her article, “The Secret of the Nāths”, but nātha here refers to Śiva (cf. Dattātreyayogaśāstra, 15-26 which teaches Śiva’s saṃketas, i.e. the secret doctrines of layayoga, and Yogabīja 136). The context confirms this, since Śiva, the original teacher of the technique, is being contrasted with Viṣṇu, who (as Kṛṣṇa) is now passing on the teaching. See Mallinson 2011a:409 on how the use of the word “Nāth” to denote a saṃpradāya of human ascetics is not found until the eighteenth-century.

31 Cf. Aparokṣānubhūti 143. Haṭha texts typically propose a simpler method than Vidyāraṇya’s: Gorakṣaśataka 9 (which is found at Haṭhapradīpikā 4.22) says that there are two hetus of the mind, vāsanās and breath. On one being destroyed, so is the other (and so too the mind): hetudvayañ ca cittasya vāsanā ca samīraṇaḥ | tayor vinaṣṭa ekasmiṃs tau dvāv api vinaśyataḥ|| On Vidyāraṇya’s preferral of gentle (mṛdu) yoga over haṭhayoga, see Jīvanmuktiviveka 1.3.27.32 See verses 16, 25, 27 and 64.33 The Vasiṣṭhasaṃhitā’s yoga does not include the mudrās which distinguish early haṭhayoga as set out in the earliest text to teach it, the Dattātreyayogaśāstra.34 I am grateful to Jason Birch for this and many other insights into the Amanaska that he has passed on to me in

conversations and email correspondence.35 On the evidence for the Śivasaṃhitā’s being a product of this Śrīvidyā tradition, see Mallinson forthcoming, Śāktism

and Haṭhayoga.36 Thus the final reward of the practices taught in the text, which culminate in one crore repetitions of the Śrīvidyā

mantrarāja, is the attainment of the place of Śiva (5.252).

Page 9: Haṭhayoga Philosophy

Of the remaining six texts borrowed from by Svātmārāma (none of which contributes more than four verses), five are from Kaula Śaiva milieux: the Khecarīvidyā, Matsyendrasaṃhitā, Candrāvalokana, Yogaviṣaya and Kaulajñānanirṇaya. Again, none of the verses taken from these texts teaches metaphysics, but these works are, where explicit, grounded in Śaiva non-dualism. The remaining text, the Uttaragītā, is a short Bhagavadgītā-style dialogue between Kṛṣṇa and Arjuna, said in some of its manuscript colophons to be part of the Mahābhārata, but certainly composed considerably later than the main body of the epic. Kṛṣṇa teaches Arjuna a yogic method of realising the truth of advaita Vedānta. Indicators of the text’s relatively late date are its inclusion of the doctrine of 72,000 nāḍīs with the Suṣumnā forming the central channel in between the Iḍā and Piṅgalā (Uttaragītā 2.15, 20) and of the Śaiva non-dual practice of meditating on the absolute as space,37 which is found in the first of its two verses borrowed by Svātmārāma (Uttaragītā 1.9c-10b = Haṭhapradīpikā 4.55):

khamadhye kuru cātmānam ātmamadhye ca khaṃ kuru |sarvaṃ ca khamayaṃ kṛtvā na kiṃcid api cintayet ||

“Put the self in space and space in the self. Make everything space and think of nothing.”

Thus Svātmārāma, while rarely borrowing verses that teach metaphysical doctrine and being somewhat indiscriminate in his choice of those,38 continued and contributed to a process that was already underway, in which vedantic and Śaiva non-dualism were synthesised, albeit with the vedantic brahman ultimately winning out as the accepted understanding of the absolute. Vedantic early haṭha texts already echoed the Uttaragītā’s instruction to meditate on the void in order to obtain mokṣa, in contrast to Śaivism, in which meditating on the void is a means of attaining Śivahood.39

This synthesis is in the main harmonious and unchallenged by critical exegesis such as that found in Abhinavagupta’s commentaries on Utpaladeva’s Īśvarapratyabhijñākārikā (II.4), in which passive vedantic ontology is rebutted in favour of the dynamic consciousness of the Pratyabhijñā (Ratié 2011:668-712). And it is assisted by fortuitous ambiguities and yogic śleṣas. In Śaiva haṭha sources, the stem form brahma refers to Brahmā. When reinterpreted in a vedantic light, however, it can mean brahman. Thus the brahmarandhra, for example, is no longer the aperture of Brahmā, but that of brahman (Mallinson 2007:205 n.240). The polyvalence of certain elements of Śaiva terminology, which in Śaiva texts refer to visualised elements in the subtle body, allows them to be reformulated in haṭhayogic texts with referents all along the ontological spectrum, from the grossest to the most sublime. In khecarīmudrā, when the tongue enters the hollow (kha) above the palate, the mind enters the void (kha).40 The bhastrikā kumbhaka, the bellows method of breathing, removes the phlegm and so forth that constitute the brahmārgala, the bolt of Brahmā/brahman at the top of the central channel, allowing Kuṇḍalinī to pierce the granthis and continue on her upward path (Gorakṣaśataka 47-48, Yogabīja 111-112, Haṭhapradīpikā 2.66-67).41

The Śaiva subtle body, as a microcosm of the universe, is predicated on Śaiva ontology, but this was soon forgotten when it was adopted as the template of the haṭhayogic body. Even the relatively early (pre-1300) Vasiṣṭhasaṃhitā had no trouble accommodating Kuṇḍalinī yoga within its orthodox brahmanic teachings. Where we do see some awkwardness is in the synthesis of the two maps of the subtle body that underly haṭhayogic theory, that of Kuṇḍalinī's ascent through the cakras and that of the keeping of bindu, semen, in its lunar store in the head. The former originated within Paścimāmnāya works such as the Kubjikāmatatantra (Heilijgers-Seelen 1994); the latter is first found in the circa 11th-century Amṛtasiddhi.42 Thus the all-inclusive Haṭhapradīpikā teaches two khecarīmudrās. One (3.31-53) seals bindu in the head, the other (4.43-4.55) floods the body with amṛta.

Conclusion

Haṭhayoga is a practical soteriology independent of metaphysical speculation. Metaphysics underpins gnostic and ritual

37 Early Vedāntic works do compare jñāna and brahman with the void (Watson 2010:99,108) but the practice of meditating on the void is not taught therein.

38 Svātmārāma includes verses from texts from different vedantic schools, such as the bhedābhedavādin Vasiṣṭhasaṃhitā and kevalādvaita Uttaragītā, and, despite attributing the creation of haṭha to a group of Śaiva mahāsiddhas, includes verses with Vaiṣṇava overtones (e.g. 4.100, or 4.58, which is taken from the Laghuyogavāsiṣṭha and preserves the vocative rāma found in its source text).

39 E.g. Dattātreyayogaśāstra 124, Vivekamārtaṇḍa 153. On meditation on the void in Śaivism, see Vasudeva 2004:263-271.40 Vivekamārtaṇḍa 50ab ( = Haṭhapradīpikā 3.40ab): cittaṃ carati khe yasmāj jihvā carati khe gatā|41 On the haṭhayogic “corporealisation” of elements of the subtle body, see Mallinson 2007:27-28).42 Although it shares two and a half verses with the Haṭhapradīpikā, the Amṛtasiddhi appears not to have been used directly by Svātmārāma, since those verses are also found in the Amaraughaprabodha, which shares a large number of verses from the Amṛtasiddhi (as does the Śivasaṃhitā). On the details of these shared verses, see Mallinson forthcoming notes 31 and 34.

Page 10: Haṭhayoga Philosophy

soteriologies, providing the object of knowledge for the former and encoding the latter with meaning, but neither gnosis nor ritual alone can bring about liberation according to haṭhayogic texts. The Haṭhapradīpikā declares that there is no jñāna until the breath is led into the central channel and bindu is held firm. He who claims to have jñāna without doing so is a liar (4.114 cf. 4.15). Kriyā is essential for haṭhayoga, but it no longer means ritual action, having become yogic practice itself (Dattātreyayogaśāstra 42, 45-46).

Yet haṭhayogic works do not do away with metaphysics altogether and include a variety of references to non-duality, in particular in the context of samādhi, the yogic summum bonum.43 This is of course in keeping with an understanding of yoga as a unificatory liberating experience, its ineffability concomitant with pronouncements of the nirguṇatva of the absolute. The period of composition of the haṭhayogic corpus came at a time when the Śaiva Age was reaching its end and Vedānta was becoming the dominant paradigm of scholarly religious thought. Haṭhayoga’s partiality for Vedāntic non-duality would by then have raised few sectarian hackles.44

Eliade (1973:143-161) declared the period from the 4th century BCE to the 4th century CE to be the “Triumph of Yoga”, yet yoga’s true triumph came during the first half of the second millennium CE. It is then that, thanks to the composition of the haṭhayogic corpus, yoga’s practices ceased to be the preserve of ascetics or initiates into tantric cults; that mainstream formulations of yoga - in which haṭha and Pātañjala yoga were not distinguished - first teach it to be an essential counterpart to jñāna in the pursuit of liberation, wedding it forever with Vedāntic soteriology; and that yoga first appears as one of the six darśanas in a Sanskrit doxography, the twelfth-century Sarvasiddhāntasaṃgraha (Halbfass 1988:352-353).45

The texts of the early haṭhayogic corpus quickly floated free of any sectarian moorings and became common property, allowing them to be used not only to compile the Haṭhapradīpikā, but, in the 17th century, to create a corpus of yoga Upaniṣads (Bouy 1994). Indeed, haṭhayoga’s lack of sectarianism and metaphysical dogma was such that its practice was readily adopted by Muslims in India in the 15th to 17th centuries (Ernst 2003, Sakaki 2005) and many of its tenets, if not its grosser practices, were espoused and recycled by Hindi nirguṇī poets such as Kabīr.

What of the place of Śaivism in this new yoga? Although haṭhayoga provided a home for a variety of Śaiva practices and concepts, its philosophical basis came to be dominated by advaita Vedānta. The lack of importance given to metaphysics and the concomitant antisectarianism found in the texts of haṭhayoga composed in its formative period were mirrored by the homogeneity of the ascetic milieu of that time, amongst which yoga’s innovators were usually to be found. It was not until the 16th to 17th centuries that concrete sectarian identities began to take shape among these groups and it was then that the Nāth saṃpradāya first established its corporate identity (Mallinson 2011a). Part of this process was the composition and writing down of a Nāth sectarian textual corpus including the Siddhasiddhāntapaddhati and vernacular works such as the Macchīṃdra Gorakh Bodh and Prāṇ Sāṃkalī.46 These works display an ambivalent attitude towards haṭhayoga - teachings on the subject are trumped by the all-encompassing scorn of the avadhūta - but they do make a place for expositions of Śaiva metaphysics. With the Nāth saṃpradāya coming to be understood erroneously as the sole originators of haṭhayoga, scholars seeking to write about its philosophy, on failing to find a thorough metaphysical grounding in its source texts, have turned to the Siddhasiddhāntapaddhati, thus continuing the Śaiva appropriation of haṭhayoga by placing excessive emphasis on its Śaiva origins and neglecting its Vedāntic heritage (e.g. Banerjea 1961).

Appendix

Borrowings in the Haṭhapradīpikā47

Amanaska (AY) 11 Amaraughaprabodha (AP) 20.5

43 E.g. Vivekamārtaṇḍa 164 ( = Haṭhapradīpikā 4.7):tatsamaṃ ca dvayor aikyaṃ jīvātmaparamātmanoḥ |pranaṣṭasarvasaṅkalpaḥ samādhiḥ so 'bhidhīyate ||44 But haṭhayoga’s (for some) incompatibility with bhakti did make it a target for poets of the later “bhakti movement”

such as Tulsīdās (e.g. Kavitāvalī, Uttarakāṇḍa 7.84; I thank Patton Burchett for pointing out this reference to me).45 The Sarvasiddhāntasaṃgraha explicitly subordinates all darśanas to Vedānta (Halbfass:356); its inclusion of yoga among the darśanas points to the Vedāntins’ new interest in yoga.46 Both these vernacular texts are included in the Gorakhbāṇī, Pītāṃbaradatta Baḍathvāl’s compilation of Hindi Nāth

works.47 As can be seen in the following table, many verses are found in more than one source text. When counting the

number of verses contributed by a text, I count only those for which I consider it the primary source. Thus the Haṭhapradīpikā shares 12.5 verses with the Yogabīja, but only 4 of those are found in the Yogabīja and not elsewhere.

Page 11: Haṭhayoga Philosophy

Uttaragītā (UG) 2 Kaulajñānanirṇaya (KJN) 1 Khecarīvidyā (KhV) 4 Gorakṣaśataka (GŚ ) 31.5 Candrāvalokana (CA) 4 Dattātreyayogaśāstra (DYŚ) 20.5Matsyendrasaṃhitā (MaSaṃ) 0.5Laghuyogavāsiṣṭha (LYV) 2Yogabīja (YB) 4Yogaviṣaya (YV) 0.5Vivekamārtaṇḍa (VM) 47.5 versesVasiṣṭhasaṃhitā (VS) 15.5 versesŚivasaṃhitā (ŚS) 8.5

The parallel that I consider to be Svātmārāma's source is given first, with additional parallels in parentheses. If there are two likely sources, the text I consider oldest is given first. Some sources are found only in parentheses (the Amṛtasiddhi (AS), Ahirbudhnyasaṃhitā (ABS), Yogayājñavalkya (YY) and Śāradātilaka (ŚT)); they are likely to be the original source texts of those verses, but Svātmārāma used texts which borrowed from them.

1.11 ŚS 5.254

1.12cd GŚ 32cd

1.19 VS 1.68 ( = ŚT 25.12, ABS 31.40, YY 3.3, ŚS 3.113)1.20 VS 1.70, ( = YY 3.5)1.21 VS 1.72, ( = YY 3.8)1.22 VS 1.80, (ab = ABS 31.35ab)1.23 VS 1.78 (ABS 31.38)

1.28 ŚS 3.108 (change of metre)1.29 ŚS 3.109 (change of metre)1.30 VS 1.76-77 (change of metre)

1.32ab DYŚ 24cd

1.35 VM 71.36 VS 1.81

1.38 ≈ DYŚ 32c-34b

1.43 ŚS 5.471.44 VM 81.45-46 DYŚ 35-36 (ŚS 3.102-103)

1.48 VM 35B48

1.50a-1.52b VS 1.73-75b (YY 3.9a-11b)

1.53ab VS 1.79ab (YY 3.12ab)

1.54 VS 1.79c-f (YY 3.12c-f)

1.57 VM 371.58 GŚ 12c-13b

1.61ab AP 44cd

48 The Baroda Vivekamārtaṇḍa manuscript has two verses numbered 35, which I distinguish with “A” and “B”.

Page 12: Haṭhayoga Philosophy

1.64 DYŚ 401.65 DYŚ 42c-43b1.66 DYŚ 46

2.2-3 VM 71-72

2.5 VM 76

2.7 VM 772.8 VM 792.9ab DYŚ 63ab2.9cd DYŚ 61cd2.10 VM 812.11 cf. DYŚ 63a-65b2.12 cf. DYŚ 75a-76b (VS 3.22)

2.14 ŚS 3.432.15 VM 1012.16-17 VM 99-100 (2.17ab = MaSaṃ 4.23cd)2.18 VM 1022.19ab DYŚ 67cd

2.20 VM 82

2.40 VM 73

2.45ab GŚ 61cd (YB 121cd)2.45cd GŚ 57cd (cf. YB 118ab)2.46 YB 123c-124b

2.48cd GŚ 34ab

2.50ab cf. GŚ 352.50cd GŚ 36ab (YB 103cd)

2.51-53 GŚ 36c-39b

2.57-58 GŚ 39c-41b

2.59 GŚ 14 (ŚT 25.10cd+11cd, ABS 31.34)

2.60-67 GŚ 41c-49b (2.65-67 = YB 110c-112d)

2.71cd VS 3.28cd (YY 6.32ab)2.71ef VS 3.29ab (YY 6.32cd, GŚ 30ab)2.72ab VS 3.28ab (YY 6.31cd)2.72c-72f VS 3.27 (YY 6.30c-31b)2.73 VS 3.30a-30d (DYŚ 146-147, YY 6.33)

2.76 ŚS 5.222a-d

3.2 ŚS 4.21

3.5 ŚS 4.22

3.6ab ŚS 4.23ab

3.9 AP 29 (AS 11.3; cd also = AP 37ab) 3.10-11b AP 30-31

3.13ab AP 32ab3.13cd AP 32cd (AS 11.9cd)

Page 13: Haṭhayoga Philosophy

3.14-17 VM 60-63

3.18ab DYŚ 132cd

3.19 AP 34

3.22ab AP 35ab

3.23cd AP 35cd3.24 AP 36 (≈ AS 13.3; 3.24cd = ŚS 4.47ab)3.25ab cf. SS 4.43ab3.26ab cf. SS 4.43cd3.26cd AP 40cd3.27 AP 41

3.29ab AP 42ab3.29cd AP 43ab3.30ab AP 42cd3.30c-f AP 43c-44b3.31 VM 47

3.33-35 KhV 1.44-46

3.38-40 VM 48-503.41 VM 503.42 VM 533.43 VM 1253.44 VM 1303.45 VM 131

3.49 VM 1283.50 VM 118

3.53ab Kulacūḍāmaṇitantra (cited in Śivasūtravimarśinī ad II.5)

3.54 GŚ 58 (YB 118c-119b)3.55-56 VM 43-443.57-8 DYŚ 141c-143b (YB 119c-121b)

3.60 VM 423.61 GŚ 52c-53b ( = GŚ 75)3.62-63 DYŚ 144-145 (YB 116-117)3.64 VM 413.65-68 GŚ 53c-57b3.69 DYŚ 1383.70-71 VM 45-463.72ab YV 19ab

3.78 VM 1353.78c-79f DYŚ 146-1473.80-81 DYŚ 148c-150b3.82a-83b DYŚ 152a-153b (3.82ab = ŚS 4.79ab)

3.86c-87d DYŚ 156c-157d (3.87cd = ŚS 4.88ab)

3.88cd VM 52ab

3.101 VM 35A3.102 VM 333.103 VM 39

3.109 YB 923.110 GŚ 59

Page 14: Haṭhayoga Philosophy

3.111 YB 1253.112 YB 943.113-114 GŚ 22c-24b3.115 GŚ 26c-27b

4.2 GŚ 63c-64b

4.5-6 VM 162-1634.7 VM 1644.8 AY 2.5

4.16 CA 30

4.17cd MaSaṃ 4.44ab

4.19 DYŚ 108

4.22 GŚ 9

4.24-25 AY 2.27-28

4.31-32 AY 2.21-224.33 KJN 3.2c-3b

4.35 AY 2.94.36 AY 2.10 (CA 1)4.37 CA 2

4.39 AY 2.84.40 AY 2.114.41 CA 34.42 KhV 3.19

4.54 CA 254.55 UG 1.9c-10b4.56 LYV 6.15.79

4.58 LYV 3.7.27

4.61 AY 2.79

4.69 AP 45 (AS 19.2, ŚS 3.31)4.70-77 AP 46-53

4.100 UG 1.42

4.108 VM 1684.109 VM 1664.110 GŚ 7

4.112 AY 2.594.113 VM 169

I am very grateful to Jason Birch for giving me general advice on how I might go about this article and specific comments on a draft of it. I thank also Alex Watson, who gave me very useful feedback on a draft, and Isabelle Ratié, who helped me with references. The table of borrowings in the Haṭhapradīpikā owes a great deal to the work of the late Christian Bouy, whose pioneering 1994 monograph paved the way for philological study of the haṭhayogic corpus and who kindly provided me with copies of several manuscript sources.

Aparokṣānubhūti, Vidyāraṇyakṛtayā Aparokṣadīpikākhyaṭīkayā saṃvalitā, ed. Kamla Devi. Akṣayavaṭa Prakāśana,

Page 15: Haṭhayoga Philosophy

1988.

Amanaskayoga, ed. Jason Birch. Critical edition submitted for DPhil. University of Oxford, 2012.

Amaraughaprabodha, of Gorakṣanātha, ed. K.Mallik in Mallik 1954.

Amaraughaśāsana, of Gorakṣanātha, ed. Pt. Mukund Rām Śāstrī. Kasmir Series of Texts and Studies 20. Srinagar, 1918.

Amṛtasiddhi of Virūpākṣanātha. Maharaja Mansingh Pustak Prakash, Jodhpur, Acc. No. 1242.

Alakhbānī or Rushdnāmā of Shaikh Abd-ul-Quddus Gangohi, ed. Saiyid Athar Abbas Rizvi and Shailesh Zaidi. Aligarh: Bharat Prakashan Mandir, 1971.

Ahirbudhnyāsaṃhitā, ed. M.D.Ramanujacharya under the supervision of F.Otto Schrader; revised by V.Krishnamacharya. Madras: Adyar Library and Research Centre, 1966.

Āryamañjuśrīnāmasaṃgīti with Amṛtakaṇikā-ṭippanī by Bhikṣu Raviśrījñāna and Amṛtakaṇikodyota-nibandha of Vibhūticandra, ed.B.Lal. Sarnath: Central Institute of Higher Tibetan Studies, 1994.

Uttaragītā, ed. S.V. Oka. Bhandarkar Oriental Research Institute Post-Graduate and Research Department Series No. 3. Poona: Bhandarkar Institute Press, 1957.

Kavitāvali of Tulsīdās, in Śukla, Rāmcandra; Bhagvān Dīn; and Brajratna Dās, eds. 1973 (2030 VS)–1974 (2031 VS). Tulsī Granthāvalī, vols. 1–2. Benares: Nāgrī Pracāriṇī Sabhā.

Kubjikāmatatantra, Kulālikāmnāya version, ed. T. Goudriaan and J.A. Schoterman. Leiden: E.J.Brill, 1988.

Kaulajñānanirṇaya of Matsyendranātha, ed. Prabodh Candra Bagchi in Kaulajñānanirṇaya and Some Minor Texts of the School of Matsyendranātha. Calcutta Sanskrit Series, No. 3. Calcutta: Metropolitan, 1934.

Khecarīvidyā, ed. James Mallinson. The Khecarīvidyā of Ādinātha. A critical edition and annotated translation of an early text of haṭhayoga. London: Routledge, 2007.

Gorakṣaśataka. Government Oriental Manuscripts Library, Madras, MS No. R 7874. See also Mallinson 2011b.

Gorakṣasaṃhitā, ed. Janārdana Śāstrī Pāṇḍeya. Sarasvatībhavanagranthamālā Vol. 111.Varanasi: Sampūrṇānandasaṃskṛtaviśvavidyālaye, 2006.

Gorakhbāṇī, ed. P.D. Baḍathvāl. Prayāg: Hindī Sāhity Sammelan, 1960.

Candrāvalokana. Government Oriental Manuscripts Library, Madras, MS No. D 4345.

Jīvanmuktiviveka. See Goodding 2002.

(Śrī)-Jñāneśvarī of Jñānadeva, ed. G.S. Naṇadīkar. 5 vols. Mumbai: Prakāś Gopāl Naṇadīkar, 2001.

Dattātreyayogaśāstra. Unpublished edition by James Mallinson based on Dattātreyayogaśāstra, ed. Brahmamitra Avasthī, Svāmī Keśavānanda Yoga Saṃsthāna 1982 (B); Man Singh Pustak Prakash Nos.1936 (J1); Wai Prajñā Pāṭhaśālā 6/4-399 (W1), 6163 (W2); Baroda Oriental Institute 4107 (V); Mysore Government Oriental Manuscripts Library 4369 (M); Thanjavur Palace Library B6390 (T).49

Niśvāsatattvasaṃhitā

Netratantra with commentary (Uddyota) by Kṣemarāja, ed. Madhusūdan Kaul Śāstrī. KSTS 46. Srinagar, 1926.

Pādmasaṃhitā, ed. S.Padmanabhan and R.N.Sampath. Madras: Pancaratra Parisodhana Parisad, 1974.

Bhāgavatapurāṇam, ed. Vasudeva Śarman. Bombay : Nirnaya Sagar, 1905.

49 This edition was read with Professor Alexis Sanderson, Jason Birch, Péter Szantó and Andrea Acri in Oxford in early 2012, all of whom I thank for their valuable emendations and suggestions.

Page 16: Haṭhayoga Philosophy

Matsyendrasaṃhitā. See Kiss 2009.

Mārkaṇḍeyapurāṇa, ed. K.M. Banerjea. Calcutta: Bishop’sCollege Press, 1862.

Laghuyogavāsiṣṭha, ed. Vasudeva Sharma Panasikara. Delhi: Motilal Banarsidass, 1985.

Līḷācaritra of Mhāibhaṭa, ed. V.B. Kolte. Muṃbaī: Mahārāṣṭra Rājya Sāhitya Saṃskṛti Maṃḍala, 1978.

Ragales of Harihara. Mahākavi Hampeya Hariharadevakṛta Nūtana Purātanara Ragalegalu, ed. M.S.Sunkapura. Dharwar: Kannada Adhyayana Pitha, Karnataka University, 1976.

Yogabīja, ed. Rām Lāl Śrīvāstav. Gorakhpur: Śrī Gorakhnāth Mandir, 1982.

Yogayājñavalkya, ed. P.C. Divanji. Bombay: Royal Asiatic Society, 1954.

Vasiṣṭhasaṃhitā (Yogakāṇḍa), eds. Swami Digambarji, Pitambar Jha, Gyan Shankar Sahay (first edition); Swami Maheshananda, B.R. Sharma, G.S. Sahay, R.K. Bodhe (revised edition). Lonāvalā: Kaivalyadhām Śrīmanmādhav Yogamandir Samiti, 2005.

Vimānārcanākalpa, ed. Śrīsvāmīhāthīrāmjī. Madras: Venkateshwar Press, 1926.

Vivekamārtaṇḍa of Gorakṣadeva. Oriental Institute of Baroda Library. Acc. No. 4110.

Śāradātilakam of Lakṣmaṇadeśikendra, paṭala 25, ed. G.Bühnemann. See Bühnemann 2001.

Śārṅgadharapaddhati, ed. Peter Peterson. Bombay: Government Central Book Depot, 1888.

Śivasaṃhitā, ed. and tr. J. Mallinson. New York: YogaVidya.com, 2007.

Śivasūtra with the commentary (vimarśinī) of Rājānaka Kṣemarāja, ed. J.C.Chatterji. Kashmir Series of Texts and Studies 1. Shrinagar, 1911.

Śūnyasaṃpādane, eds. S.C. Nandimath, L.M.A. Menezes, R.C. Hiremath, M.S. Sunkapur. 5 Vols. Dharwar: Karnatak University, 1965-1972.

Siddhasiddhāntapaddhati of Gorakṣanātha, ed. M.L. Gharote and G.K. Pai. Lonavla: Lonavla Yoga Institute, 2005.

Skandapurāṇa, ed. Kṛṣṇaprasāda Bhaṭṭarāī (Skandapurāṇasya Ambikākhaṇḍaḥ).Kathmandu: Mahendrasaṃskrṭaviśvavidyālayaḥ, 1988.

Haṭhapradīpikā of Svātmārāma, eds. Svāmī Digambarjī and Dr Pītambar Jhā. Lonavla: Kaivalyadhām S.M.Y.M.Samiti, 1970.

Banerjea, Akshaya Kumar. (1983). Philosophy of Gorakhnath with Goraksha-Vacana-Sangraha. England: Combe Springs Press.

Bouy, Christian. (1994). Les Nātha-Yogin et les Upaniṣads. Paris: Diffusion de Boccard.

Briggs, George Weston. (1989 [1938]). Gorakhnāth and the Kānphaṭa Yogīs. Delhi: Motilal Banarsidass.

Bühler, G. (1892). The Cintra Praśasti of the reign of Sarangadeva. Epigraphia Indica Vol. 1, 271-287. Calcutta: Archeological Survey of India.

Bühnemann, Gudrun. (2001). The Iconography of Hindu Tantric Deities volume II. The Pantheons of the Prapañcasāra and the Śāradātilaka. Groningen: Egbert Forsten.

Colas, Gérard. (1988). Le yoga de l'officiant Vaikhanasa. Journal Asiatique 276 (3-4), 245-283.

Page 17: Haṭhayoga Philosophy

----------------------. (2010) Vaiṣṇava Saṃhitās. Brill Encyclopedia of Hinduism Vol.2, 153-167. Leiden: Brill.

Digby, Simon. (1970) Encounters with Jogīs in Indian Ṣūfī hagiography. Unpublished paper presented at a seminar on Aspects of Religion in South Asia at the School of Oriental and African Studies, University of London.

Eliade, Mircea. (1973 [1954]). Yoga: Immortality and Freedom. Princeton: PrincetonUniversity Press.

Ernst, Carl W. (2003). The Islamization of Yoga in the Amrtakunda Translations. Journal of the Royal Asiatic Society, Series 3, 13:2, 1-23.

----------------------. Accounts of yogis in Arabic and Persian historical and travel texts. Jerusalem Studies in Arabic and Islam 33, 409-426. The Hebrew University of Jerusalem.

Franco, Eli. (2009) Introduction. Yogic Perception, Meditation and Altered States of Consciousness, ed. Eli Franco in collaboration with Dagmar Einer, 1-51. Vienna: Verlag der Österreichischen Akademie der Wissenschaften.

Goodall, Dominic. (2004) The Parākhyatantra. A Scripture of the Śaiva Siddhānta. A critical edition and annotated translation. Pondicherry: Publications de l’Institut français d’Indologie No. 98.

Goodding, Robert. (2002) The Treatise on Liberation-in-Life, Critical Edition and Annotated Translation of theJīvanmuktiviveka of Vidyāraṇya. Unpublished PhD Thesis, University of Texas.

Halbfass, Wilhelm. (1988) India and Europe. Albany: State University of New York Press.

Heilijgers-Seelen, Dory. (1994) The System of Five Cakras in Kubjikāmatatantra 14-16. Groningen: Egbert Forsten.

Hori, Shin’ichirō. (2005) Additional Notes on the Unidentified Sanskrit Fragments in the Ōtani Collection at Ryūkoku University Library. Journal of the International College for Postgraduate Buddhist Studies, 91-97.

Kiehnle, Catharina. (1997) Songs on Yoga. Texts and Teachings of the Mahārāṣṭrian Nāths. Stuttgart: Franz Steiner Verlag.

—————–. (2005) The Secret of the Nāths: The Ascent of Kuṇḍalinī according to Jñāneśvarī 6.151-328. Bulletin des Études Indiennes 22-23, 447-494.

Kiss, Csaba. (2009). Matsyendranātha's Compendium (Matsyendrasaṃhitā). A critical edition and annotated translation of Matsyendrasaṃhitā 1-13 and 55 with analysis. Unpublished DPhil. thesis submitted to Oxford University.

Kiss, Csaba. (2011) The Matsyendrasaṃhitā: A Yoginī-centred Thirteenth-century Yoga Text of the South Indian Śāmbhava Cult. Yogi Heroes and Poets: Histories and Legends of the Nāths, eds. David N.Lorenzen and Adrián Muñoz, 143-162. New York: SUNY.

Lévi, Sylvain. (1905) Le Népal: Étude Historique d'un Royaume Hindou. Vol. 1. Paris: Ernest Leroux.

Locke, John K. (1980) Karunamaya. The Cult of Avalokitesvara-Matsyendranath In the Valley of Nepal. Kathmandu: Sahayogi Prakashan.

Mallik, Kalyani. (1954) The Siddha Siddhānta Paddhati and Other Works of Nath Yogis. Poona: Poona Oriental Book House.

Mallinson, James. (2007) The Khecarīvidyā of Ādinātha. A critical edition and annotated translation of an early text ofhaṭhayoga. London: Routledge.

----------------------. (2011a) Nāth Saṃpradāya. Brill Encyclopedia of Hinduism, Vol. 3, ed. Knut A. Jacobsen, 407-428. Leiden: Brill.

----------------------. (2011b) The Original Gorakṣaśataka. Yoga in Practice, ed. D.G.White, 257–272. Princeton: Princeton University Press.

Michaels, Axel. (1985) On 12th-13th Century relations between Nepal and South India. Journal of the Nepal Research Centre Vol. VII, 69-73. Wiesbaden: Franz Steiner Verlag GmbH.

Page 18: Haṭhayoga Philosophy

Narharināth, Yogī. (1953) Kāṣṭhamaṇḍapa. Saṃskṛta-sandeśa Vol. 1, No. 6, 4-10.

Ratié, Isabelle. (2011) Le Soi et L’Autre. Leiden: Brill.

Rizvi, S.A.A. (1971) See Alakhbānī.

Sakaki, Kazuyo. (2005) Yogico-tantric Traditions in the Ḥawd al-Ḥayāt. Journal of the Japanese Association for South Asian Studies, 7, 135-156.

Saletore, B.A. (1937) The Kānaphāṭa Jogis in Southern History. The Poona Orientalist 1, 16-22.

Sanderson, Alexis. (2007) Atharvavedins in Tantric Territory: The Aṅgirasakalpa Texts of the Oriya Paippalādins and their Connection with the Trika and the Kālīkula, with critical editions of the Parājapavidhi, the Parāmantravidhi, and the Bhadrakālāmantravidhiprakaraṇa. The Atharvaveda and its Paippalāda Śakhā: Historical and Philological Papers on a Vedic Tradition, eds. Arlo Griffiths and Annette Schmiedchen, 195-311.

----------------------. (2011) Śaivism, Society and the State. Unpublished paper.

Slusser, Mary & Vajracarya, Dhanavajra. (2005) Two Medieval Nepalese Buildings: An Architectural and Cultural Study. Art and Culture of Nepal: Selected Papers, by Mary Shepherd Slusser with contributions by Gautam V. Vajracharya and Manuela Fuller, 429-503. Kathmandu: Mandala Publications. [This article first appeared in Artibus Asiae, vol. 36 No. 3 (1974) pp. 169-218.]

Stearns, Cyrus. (1996) “The Life and Tibetan Legacy of the Indian Mahāpaṇḍita Vibhūticandra. TheJournal of the International Association of Buddhist Studies Vol.19 No.1, 127-171.

Tamot, Kashinath. (2009) ‘Devo Gorakṣo’ is not there. Blog post at http://www.nepalmandal.org/archive/200901.

Tulpule, Shankar Gopal. (1979) Classical Marāṭhī Literature. A History of Indian Literature Vol. IX, fasc. 4. Wiesbaden: Otto Harrowitz.

Vajrācārya, Dhanavajra. (1975) Śaktiśālī Bhadra Rāmavarddhanaru rā tatkālika Nepāl. 12-36 in Pūrṇimā 7 (vikram samvat 2022 Kārttik), 12-36.

Vasudeva, Somadeva. (2004) The Yoga of the Mālinīvijayottaratantra. Pondicherry: Publications de l’Institut français d’Indologie No. 97.

Vaudeville, Charlotte. (1987) The Shaiva-Vaishnava Synthesis in Maharashtrian Santism. The Sants: Studies in a Devotional Tradition of India, ed. Karine Schomer and W.H.McLeod, 215-228. Delhi: Motilal Banarsidass.

Watson, Alex. (2010) Rāmakaṇṭha’s Concept of Unchanging Cognition (nityajñāna): Influence from Buddhism, Sāṃkhya and Vedānta. From Vasubandhu to Caitanya: Studies in Indian Philosophy and Its Textual History, ed. Johannes Bronkhorst and Karin Preisendanz, 79-120. Delhi: Motilal Banarsidass.

Weightman, S.C.R. (1992) The text of Alakh Bānī. Devotional Literature in South Asia: Current Research 1985-1988, ed. R.S.McGregor., 172-178. Cambridge: Cambridge University Press.