The Ocean of Yoga: An Unpublished Compendium Called the Yogārṇava

The Yogārṇava (‘the ocean of yoga’) is a Sanskrit compendium on yoga that has not been published, translated or even mentioned in secondary literature on yoga. Citations attributed to it occur in several premodern commentaries and compendiums on yoga, and a few published library catalogues report manuscripts of a work on yoga called the Yogārṇava. This article presents the results of the first academic study of the text. It has attempted to answer basic questions, such as the work’s provenance and textual sources. The authors then discuss the importance of the Yogārṇava within the broader history of yoga based on their identification of citations and parallel verses in other Sanskrit texts and a detailed analysis of the Yogārṇava’s content.


Introduction
The Yogārṇava ('the ocean of yoga') is a compendium with an interesting combination of yogic theory and praxis. It contains extensive discourse on the yogic body, including the vital winds (vāyu), points (marman) and five sheaths (pañcakośa), and more general topics, such as nasal dominance and the astrological signs in the body, prognostication and cheating of death, and the importance of retaining the body to know Brahman. The author combines these topics with a yoga of eight auxiliaries (aṣṭāṅgayoga), the particulars of which are very similar to those of two related texts: the Vasiṣṭhasaṃhitā and Yogayājñavalkya. A close comparative analysis of these texts is presented in this article to reveal that the author of the Yogārṇava borrowed from both works. We combine this analysis with other evidence to propose a tentative date for the Yogārṇava's composition. This article will also provide a brief overview of the text's available manuscripts, catalogue references and content that may assist further research, and perhaps lay the foundations for a critical edition and translation of the Yogārṇava. We conclude that the Yogārṇava is an early example of a yogic compendium that anticipates larger compilatory works that foreground yoga within a vedāntic framework, such as Śivānanda's Yogacintāmaṇi and Bhavadevamiśra's Yuktabhavadeva.

Manuscripts
The research for this article is based on two transcripts of manuscripts of the Yogārṇava. The first is a Devanagari transcript at the Government Oriental Manuscript Library, Chennai (GOML), 1 and the second is a Malayalam transcript at the Oriental Research Institute, Trivandrum (ORI). 2 Only the latter has been reported in the New Catalogus Catalogorum of Madras (NCC). 3 There is also a Jyotiṣa work by the name Yogārṇava, of which the NCC (vol. 22: 146) lists many manuscripts. It is likely that this work is mistaken for the 'yogic' Yogārṇava in some catalogues. There is at least one instance of this. In volume ten of the Mysore Oriental Research Institute's catalogue of Sanskrit works, a Jyotiṣa Yogārṇava has been included in the section on yoga texts. 4 The editors appear to have made this mistake because of the title 'Rājayoga' at the beginning of the text. However, the opening verses make it clear that this Rājayoga is not the type of yoga concerned with samādhi, but with the constellations relevant to kings.
In The Descriptive Catalogue of Yoga Manuscripts compiled by Kaivalyadhama Yoga Institute (2005: 386-387), only three manuscripts of the Yogārṇava are reported. Two of these are held at the Sanskrit University (Sampurnananda) Library, Varanasi. 5 Both are in Bengali script. The third is at the Palace Granthappura Library, Trivandrum. 6 This manuscript is in Malayalam script, and we are yet to determine whether it was the exemplar of the Malayalam transcript at the ORI.
The transcript from the GOML is in Devanagari script and on paper. It was created in the early twentieth century and is probably a copy of a south-Indian manuscript. It is complete but some lines of the text are missing. The transcript from the ORI is in Malayalam script and also on paper. It is complete and contains all the verses. We have been able to reconstruct much of the text with these witnesses because the verses which were missing in the GOML transcript can be found in the one from the ORI. Also, the ORI transcript has fewer scribal errors than the GOML one. The quotations in this paper are based on this reconstruction.

Possible Source Texts of the Yogārn . ava
There is a complex relationship between the Yogārṇava, Vasiṣṭhasaṃhitā (12th c.) and Yogayājñavalkya (13th-14th c.). 7 The editors of the Kaivalyadhama Yoga Institute's critical edition of the Vasiṣṭhasaṃhitā (2005: 31-32) argue that the Yogayājñavalkya borrowed much material from it. This hypothesis is supported by a comparison of parallel passages in both works that was published in Birch (2018, pp. 21-22), which demonstrated that the redactor of the Yogayājñavalkya borrowed a lengthy discussion on the yogic body from the Vasiṣṭhasaṃhitā and supplemented it with material from elsewhere. Therefore, the Vasiṣṭhasaṃhitā is probably the earlier work, which the editors of the Kaivalyadhama Yoga Institute's edition date to the twelfth century.
As seen in Table 1 of the Appendix, approximately two hundred and thirty-three verses of the Yogārṇava are found in the Vasiṣṭhasaṃhitā, and one hundred and seventy-eight in the Yogayājñavalkya. Some of these verses occur in both the Vasiṣṭhasaṃhitā and Yogayājñavalkya, whereas others are peculiar to only one. Therefore, it appears that the author of the Yogārṇava used both the Vasiṣṭhasaṃhitā and Yogayājñavalkya to compile sections on the yogic body and aṣṭāṅgayoga, which is the main topic of chapters four to eight in the Yogārṇava. As Mallinson (2014, pp. 227-228) has observed, this type of aṣṭāṅgayoga can be found in the early Vaiṣṇava samḥitās, including the Vimānārcanākalpa, Sūtasaṃhitā and Ahirbudhnyasaṃhitā. In fact, some of the verses that the Yogārṇava shares with the Vasiṣṭhasaṃhitā and Yogayājñavalkya are also in the Ahirbudhnyasaṃhitā and Sūtasaṃhitā (see Table 1). 8 The Yogārṇava does not add much new material to the discussion of the first five auxiliaries of aṣṭāṅgayoga found in these sources. However, it contains more extensive and detailed sections on the last three. 9 The Yogārṇava is a larger compilation than either the Vasiṣṭhasaṃhitā or Yogayājñavalkya. It covers topics that are not found in these earlier works, such as the five sheaths (pañcakośa), the development of a foetus (garbha), gross elements (mahābhūta) and bodily constituents (dhātu), regions of the body (maṇḍala), aspected (sakala) and aspectless (niṣkala) meditation (dhyāna), meditation on the sun (sauradhyāna), the four states of consciousness (avasthā), four levels of speech (vāṇī), visualising the alphabet in the navel and a ritual for oneself (ātmayāga). Furthermore, there are more elaborate discussions of the eight letters (aṣṭavarga) of the central channel (suṣumnā), and the bodily channels (nāḍī) and winds (vāyu), as well as dreams (svapna), stages of life (āśramakrama), length of life (āyuḥpramāṇa), conquering death (mṛtyuñjaya), immortality (amaratva), the self (ātman), om (praṇava) and meditative absorption (samādhi). 10 Citations and Provenance of the Yogārn . ava The name of the author and the region in which the Yogārṇava was composed remain unknown to us. The work is not mentioned at all, let alone discussed, in secondary sources on yoga. Also, the exact date of the text is unknown. We are yet to find a dated manuscript of the Yogārṇava and, as far as we are aware, no such manuscript has been reported in a published catalogue. However, as mentioned above, the Yogārṇava's terminus a quo is the Yogayājñavalkya, which means it was composed sometime after the thirteenth or fourteenth century.
There are citations of the Yogārṇava in various texts that date from the fifteenth century or later. The most important of these for establishing a terminus ad quem is Rāghavabhaṭṭa's commentary on the Śāradātilakatantra called the Padārthādarśa. According to Sanderson (2007, p. 230), Rāghavabhaṭṭa was a Maharashtrian scholar who completed this commentary in Varanasi in 1494 CE. Rāghavabhaṭṭa cites the Yogārṇava by name five times on the topics of the formation of the foetus, the nāḍīs and the ten vāyus. 11 Rāghavabhaṭṭa also cites a passage on the process of digestion and attributes it to the Yogārṇava, 12 but this passage is not in the transcripts of the Yogārṇava that we have consulted, which suggests that he was using a slightly different, perhaps longer, version than is currently available. Owing to the content shared between the Yogārṇava and Yogayājñavalkya and the relevant citations in Rāghavabhaṭṭa's commentary, we can conclude that the Yogārṇava was probably composed in the late fourteenth or early fifteenth century.
The Yogārṇava is also cited by name in the Upāsanāsārasaṅgraha and Yogasārasaṅgraha. Both of these works are compilations on yoga that cite other texts profusely. The Upāsanāsārasaṅgraha was composed in South India, possibly before 10 See the Table of Contents that we have created for the Yogārṇava (Table 2 of the Appendix) for the chapter and verse numbers of these topics. 11 Rāghavabhaṭṭa on Śāradātilakatantra 1. 32, 1.40, 1.42, 1.45 (1996: 31, 38, 40-41) cites Yogārṇava (with attribution) 1. 32-1.36, 2.17, 2.21-2.25, 2.37-2.57. Rāghavabhaṭṭa on Śāradātilakatantra 1.32 also cites the untraced passage in the next footnote. Yogārṇava 1.32-1.36 is on the formation of the foetus; 2.17 and 2.21-25 are on the nāḍīs, and 2.37-2.57 is on the vāyus. 12 Rāghavabhaṭṭa on Śāradātilakatantra 1.32 (1996: 31-32): atra prakāro yogārṇave -āviśya bhuktam āhāraṃ sa vāyuḥ kurute dvidhā | saṃpraviśyānnamadhyasthaṃ pṛthak kiṭṭaṃ pṛthag jalam || agner ūrdhvaṃ jalaṃ sthāpya tadannañ ca jalopari | jalasyādhaḥ svayaṃ prāṇaḥ sthitvāgniṃ dhamate śanaiḥ || vāyunā vyūhyamāno'gnir atyuṣṇaṃ kurute jalam | annaṃ taduṣṇatoyena samantāt pacyate punaḥ || dvidhā bhavati tat pakvaṃ pṛthak kiṭṭaṃ pṛthag rasam | rasena tena tā nāḍīḥ prāṇaḥ pūrayate punaḥ || pratarpayanti sampūrṇās tāś ca dehaṃ samantataḥ | mātū rasavahā nāḍī manuviddhā parābhidhā || nābhisthanāḍīgarbhasya mātrāhṛtarasāvahā | iti | the seventeenth century (Bouy, 1994, pp. 89-92). It contains citations of three passages in the Yogārṇava's sections on meditation (dhyāna) and absorption (samādhi). 13 The Yogasārasaṅgraha may post-date the seventeenth-century Haṭharatnāvalī (Birch, 2020, p. 464 n. 43). It cites a verse from the Yogārṇava's section on dhyāna. 14 Verses of the Yogārn . ava in Other Works As seen in Table 1 of the Appendix, the Yogārṇava has verses in common with some earlier śruti and smṛti texts, such as the Bhagavadgītā and various Upaniṣads. We have also found verses of the Yogārṇava in works that probably post-date it, including yoga compendiums, like the Yogasārasaṅgraha and Yogacintāmaṇi; a commentary called the Haṭhapradīpikājyotsnā; and various yoga Upaniṣads, such as the Śāṇḍilyopaniṣad, Dhyānabindūpaniṣad, Vārāhopaniṣad, Yogacūḍāmaṇyopaniṣad, Yogatattvopaniṣad and so on. Much of this borrowed material probably derives from the Vasiṣṭhasaṃhitā and Yogayājñavalkya, which (as noted above) were sources for the Yogārṇava and, more importantly in this regard, for the Haṭhapradīpikā and many other subsequent works. 15 However, the encyclopaedic compendium called the Prāṇatoṣinī cites the Yogārṇava by name. Also, other so-called Yoga Upaniṣads, such as the Varāhopaniṣad, Amṛtanādopaniṣad and Dhyānabindūpaniṣad, contain verses in the Yogārṇava that are not in the Vasiṣṭhasaṃhitā and Yogayājñavalkya. 16 The Prāṇatoṣinī was composed in Bengal and completed in 1820 (Goudriaan and Gupta 1981: 147) and the so-called Yoga Upaniṣads were created in South India for a corpus of one hundred and eight Upaniṣads in the mid-eighteenth century (Bouy, 1994). Therefore, the Yogārṇava appears to have remained a valued source of material on yoga until the nineteenth century.
Interestingly, verses in the Yogārṇava are cited in the Yogasārasaṅgraha with attribution to the Śivayoga, Praṇavacintāmaṇi, Yogasāramañjarī and Kāśīkhaṇḍa, as well as Ādinātha, which suggests a strong association with Śaiva works. At the very least, it is clear that many teachings of the Yogārṇava were reproduced in compilations on yoga composed after the sixteenth century, particularly those that were orientated towards Advaitavedānta.

The Yogārn . ava's Content
As seen in Table 2 of the Appendix, the first three chapters of the Yogārṇava discuss the yogic body; prognostication by observing the breath and seeing the signs of death; the paths of rebirth and liberation; caste, stages of life and duty (varṇāśramadharma); and the importance of cheating death (kālavañcana) in order to live long enough to know Brahman. Much of this content derives from the Vasiṣṭhasaṃhitā and Yogayājñavalkya, the main exceptions being discussions of the five sheaths (pañcakośa), the development of a foetus and the eight letters of the central channel (suṣumṇā). 17 Most of the additional content was probably inspired by, or perhaps even borrowed from, vedāntic and tantric works. 18 The last five chapters discuss aṣṭāṅgayoga and each of its auxiliaries. The first five auxiliaries are dealt with in the last one hundred and four verses of chapter four and the first six verses of chapter five. Most of these verses derive from the Vasiṣṭhasaṃhitā and Yogayājñavalkya. However, the discussion of the last three auxiliaries (i.e., dhāraṇā, dhyāna and samādhi) comprises the last three chapters of the work (5-8), which amount to three hundred and sixteen verses. Much of the content of the last three chapters goes beyond that of the Vasiṣṭhasaṃhitā and Yogayājñavalkya, and we are yet to trace it to earlier sources. 19 The discussion of dhāraṇā has additional material on concentration methods that cure great diseases (mahārogahara), cheat (vañcana) and conquer death (mṛtyuñjaya), and bring about immortality (amaratva).
The section on dhyāna begins with an overview of various twofold schemes of meditation, such as aspected (sakala) and aspectless (niṣkala), internal (ābhyantara) and external (bāhya), all-pervading (sarvaga) and on a particular place (pradeśaviśiṣṭa), or on what is differentiated (bhinna) and undifferentiated (abhinna) from oneself. 20 The syncretic nature of this introductory passage reveals the author's familiarity with different systems of meditation in earlier traditions and his intention to integrate them. The discussion of meditation on bodily supports (ādhāra) which follows it continues in the same vein. The author first notes that some yogins meditate on eighty-one supports whereas others know sixty-four, thirty-six, thirty-two and so on. He concludes this discussion with those who meditate on only one support and states they are the foremost (pradhāna). 21 The rest of the chapter contains passages on different visualisations ranging from the sun to the goddess and various worlds (loka).
The topic of dhyāna appears to continue into the seventh chapter, which weaves together various meditations and theoretical discussions on praṇava, the Self and the supreme deity, known as Īśvara or Brahman. A significant portion of the chapter is devoted to explaining four levels of speech (sūkṣmā, paśyantī, madhyamā and vaikharī) and their relation to the yogic body and the Self. The content of this chapter largely derives from vedāntic and tantric sources, and the author seems to have taken advantage of differences in terminology, expression and metaphors to augment the compilation. The section on samādhi, the eighth auxiliary, begins with the twelfth verse of the eighth chapter. In the same style as earlier chapters, the author compiles various meditations that result in samādhi, ranging from the contemplation of the letters of praṇava to realize 17 Please see Table 2 of the Appendix for the verse numbers of these topics. 18 We are yet to trace the Yogārṇava's verses on these additional topics to an earlier source. However, seeing that the Yogārṇava is mainly a compilation and these topics are peculiar to earlier vedāntic and tantric traditions, it is likely that at least some of the verses were borrowed by the Yogārṇava's author. 19 The few exceptions occur mainly in the Yogārṇava's seventh and eighth chapter where some verses may derive from the Mahābhārata (including the Bhagavadgītā) and some Purāṇas. See Table 1 of the Appendix for the references. 20 Yogārṇava 6.2-6.3 21 Yogārṇava 6.4-6.8. that 'I am only Brahman' (brahmaivāham) to meditations on the Self, the void, the three phases of the breath, raising Kuṇḍalinī and so on. The discussions of dhyāna and samādhi are similar in style and content but are somewhat distinguished by the fact that the section on dhyāna has greater emphasis on visualizations of things with attributes, whereas the section on samādhi emphasizes meditations on what is free of attributes.

Historical Significance of the Yogārn . ava
Although the Yogārṇava's content derives largely from earlier traditions of Vedānta and Tantra, its style of composition anticipates several compilations on yoga that were composed in the sixteenth and seventeenth century. Examples include the Yogacintāmaṇi of Godāvaramiśra, Yogacintāmaṇi of Śivānandasarasvatī, Yuktabhavadeva of Bhavadevamiśra, Upāsanāsārasaṅgraha and Yogasārasaṅgraha. 22 Like the Yogārṇava, these compilations foreground yoga in a vedāntic framework. They present yoga with eight auxiliaries (i.e., yama, niyama, āsana, prāṇāyāma, pratyāhāra, dhāraṇā, dhyāna and samādhi) as the means to attaining gnosis of Brahman, the supreme Self. In particular, the structure of the Yogārṇava is similar to the Yogacintāmaṇi of Śivānandasarasvatī, the first half of which consists of general topics on yoga whereas the second half is structured on the eight auxiliaries of yoga. Both compilations integrate doctrine and theory from vedāntic and tantric traditions, while emphasizing gnosis of the Self as the goal of yoga. Likewise, the first three chapters of the Yuktabhavadeva address diverse topics, some of which are integral to yoga, such as the obstacles to achieving yoga and the yogic body, and others more tangential, such as elixirs (kalpa). The remaining seven chapters of the Yuktabhavadeva (i.e., 4-11) are structured on the eight auxiliaries of yoga.
A significant difference between the Yogārṇava and the afore-mentioned compilations is that the author of the former did not reveal the textual sources from which verses were borrowed whereas the authors of the latter do. In this regard, the Yogārṇava's style of composition is closer to the Vasiṣṭhasaṃhitā and Yogayājñavalkya, both of which can also be seen as syncretic works structured on the eight auxiliaries of yoga. However, the scope of topics outside aṣṭāṅgayoga and the extent of vedāntic and tantric doctrine is far greater in the Yogārṇava than the Vasiṣṭhasaṃhitā and Yogayājñavalkya, which seem almost rudimentary by comparison. The Yogārṇava's broader range of content and diversity of sources is a salient feature of the yogic compilations that followed it, yet the authors of the subsequent works accentuated this syncretic style by explicitly citing their source material.
The vedāntic orientation of the Yogārṇava is most clearly seen in the work's emphasis on the realization of the Self and Brahman, as well as the incorporation of the upaniṣadic sheaths (kośa). In fact, as far as we are aware, the Yogārṇava and Yuktabhavadeva are the only premodern yogic works that incorporate the five sheaths. 23 As Bouy (1994) and Birch (2020) have noted, the foregrounding of yoga in vedāntic compendiums and Upaniṣads represents a burgeoning interest in yoga within vedāntic milieus that flourished in the early modern period. However, the Yogārṇava pushes the epoch for such yogic compilations back to the fifteenth century, and one wonders whether the success of the Yogārṇava, as evinced by the citations in Rāghavabhaṭṭa's commentary, the Upāsanāsārasaṅgraha and Yogasārasaṅgraha, inspired subsequent authors to write more comprehensive compilations on yoga for a learned audience who were primarily interested in the role of yoga within vedāntic soteriology.
Finally, it should also be noted that, unlike subsequent compilations, the Yogārṇava does not mention Haṭhayoga or any of its distinct techniques, such as the mudrās and bandhas that feature in the third chapter of the Haṭhapradīpikā. This somewhat supports our hypothesis that the Yogārṇava was composed before the late fifteenth century, for this relatively early dating of such a compendium suggests that it arose before Haṭhayoga become too significant for Vedāntins to ignore. 24

Conclusion
The Yogārṇava appears to have been an important work in the history of yoga because it was cited in several prominent works, such as Rāghavabhaṭṭa's Padārthādarśa and the Upāsanāsārasaṅgraha, and was a likely source of many yoga compendiums and Upaniṣads that were written after the sixteenth century. If we are correct in dating the Yogārṇava to the late fourteenth or early fifteenth century, then it is an early and sophisticated attempt to weave yogic, vedāntic and tantric teachings into a wideranging compendium that posits the eight generic auxiliaries of yoga as the chief means to realizing gnosis of Brahman.
Statements and Declarations S.V.B.K.V Gupta and Jason Birch do not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and have no relevant affiliations beyond their academic appointments.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.   Installing the alphabet on the body and visualising Praṇava 2 Visualising the Self as Īśvara

3-12
Knowing what the Self is not, and achieving liberation through gnosis of the Self (jñāna)

13-18
The identity of the Individuated Self and the Self as Īśvara

28
Four levels of speech (sūkṣmā, paśyantī, madhyamā and vaikharī). The highest is equated with the Self 29-32ab Sūkṣmā in the form of the Self pervades the body 32cd-44 Paśyantī is the letters of the alphabet on a twelve-spoke cakra above the navel.
The alphabet is in Praṇava and the alphabet is the source of all mantras. Therefore, one should recite Praṇava, the ātmamantra

45-51ab
Madhyamā is the various forms of Vedic metres in a fourteen spoke cakra below the heart 51cd-52 Vaikharī is in the heart as 64 arts in 64 triangles 53-54 How all things emerge and resorb into the great Self in the heart 55-65 The great Self is the fourfold Praṇava, the basis of everything and in all creatures. It is identical with Brahman and the Self.

66-69
Brahman pervades all things 70-74 Installing the Tattvas in Praṇava, the three letters of which are creation, sustenance and destruction 75-79 Praṇava, Brahman and the Self are in all things and vice-versa 8 1-11 Equivalences between knowledge of the Self (ātma) and Vedic ritual and knowledge 12-13 Definition of Samādhi

14-16
Knowing that the letters of Praṇava (i.e., a, u, m) represent the Individuated Self, the Supreme Self and their union, one should contemplate 'I am only Brahman' and unite the Individuated and Supreme Selves

17-20ab
Meditating on the Self until one thinks of nothing else

20cd-22ab
Meditating on what is free of all qualities (nirguṇa), after dissolving the material elements, sense objects and senses in their own causes Using the bodily winds to stoke the fire and the fire to burn Kuṇḍalinī, all three go into the central channel (suṣumṇā) and ascend upwards through the heart, throat and middle of the brows 39cd-42ab One should unite the Self in the void, supreme bliss 42cd-43ab One sees the orb of the sun, the supreme, untainted light, like a mass of lightning, emitted from the aperture in the head.

43cd-44
Making the fire go to the aperture of Brahman (brahmarandhra), churning it, burning the whole body up, the Self becomes Brahman and one is not reborn