Who is the Īśvara of the Yogasūtra/Pātañjalayogaśāstra?

Is Īśvara in the Yogasūtra/Pātañjalayogaśāstra just the model of a puruṣa who has realised its being separated from nature? Or is He an omnipotent (and perhaps compassionate) God?

Some months ago, Andrew Nicholson posted here a link to the talks of a conference on the Yogasūtra held at the Loyola University. The talk by Edwin Bryant (a video is available here) made me for the first time aware of the possibility of interpreting the Īśvara of the Yogasūtra as an omnipotent God. Bryant points in particular to parallels in the Bhagavadgītā and in other religious texts coeval with the Yogasūtra.

Possible questions here:

  • Is the Yogasūtra really a text predating the Yogabhāṣya? Or should not we rather speak of a unitary Pātañjalayogaśāstra? If the latter, do the arguments above still apply?
  • Does it make sense to look for parallels in religious texts (I know, Amod has repeatedly shown how slippey this definition may be, let me then just say that I am speaking about epics and Purāṇas)?

I am by no means an expert, but until now I have thought of the YS/PYŚ’s īśvara as a puruṣaviśeṣa ‘special puruṣa‘ and not as a creator God. In this sense, I have been reading the YS/PYŚ as a philosophical text, to be compared with the Classical Sāṅkhya (rather than with the less philosophically developed trends of Sāṅkhya which can be detected in, e.g., the Epics). This harmonises with my general observation that religious elements tend to be read in a rationalised way in philosophical texts.
However, Jonathan Dickstein recently wrote to me soliciting me to rethink about it. His questions are:

questions are:

Around this early period of thought (at most 5th cent. BC – 5th cent. CE), is not the term “īśvara” (w/ maheśvara, parameśvara) predominantly used in the grand sense of an all-encompassing “God”? I am aware that the term is also used to denote a “master”or “lord”, even in the Gītā, but in metaphysical and/or devotional contexts, is there any strong evidence that the term ever depicted an imaginary archetypal guru or an uninvolved supreme being (as some modern Yogasūtra interpretations offer)?

A few months ago there was a protracted conversation about this topic on RISA-L, specifically concerning the symmetry of the YŚ’s īśvara with that of the Gītā. While this aspect is of central concern to my own work, currently I am concerned with isolating the dominant use of the term “īśvara” in contemporaneous (not beyond 5th cent. CE) and content-similar (e.g. mokśa śāstra) literature. I have read Bronkhorst and others and how they come to the conclusion that the Yogasūtra‘s īśvara is this type of archetypal guru or uninvolved being. However, these discussions often invoke input from later figures such as Haribhadra, Kumārila, Mādhava, etc. To my knowledge only M.D. Shastri’s article and Lloyd Plueger’s unpublished dissertation haven taken direct aim at the question from within this earlier period.

Presently I side more with Bryant’s position that īśvara, as found in the Yogasūtra and entire Pātañjala Yogaśāstra, is more similar to the Gītā‘s understanding – an īśvara who is most likely creator, efficient cause, intervener, and object of devotion (Īśvara in the YS may be a “puruṣaviśeṣa“, but even Kṛṣṇa is called a “puruṣottama“). Maintaining this position in light of all the obvious textual and philosophical difficulties is the objective of my forthcoming thesis.

What do the experts on Yoga in this blog think? And what do you think?

About elisa freschi

My long-term program is to make "Indian Philosophy" part of "Philosophy". You can follow me also on my personal blog: elisafreschi.com, on Academia, on Amazon, etc.

22 thoughts on “Who is the Īśvara of the Yogasūtra/Pātañjalayogaśāstra?

  1. Thank you for this post. I am certainly no Yoga expert, but would like to note that one late commentator on the YS, Narayana Tirtha (17th C.), does very deliberately take īśvara to mean the God of the Gita and the Bhagavata Purana. I have a short section on this in one of my dissertation chapters. For those who may be attending the AAR in Atlanta in November, I am on a panel at DANAM that uses precisely this question as a starting point to explore issues of commentarial creativity.

  2. Being not an expert in YS I still wonder, whether it makes sence to choose a priory to which branch of Western knowledge does some Indian text belong, to philosophy or religion? The more so, as Premodern India was evidently unaware of this very opposition that historically derives from the fact, that European philosophy was initially Greek and religion – Hebrew.
    YS are composed in ‘philosophical manner’, but the same terms and topics can be also found in some other texts, composed in different styles. So why not compare with epics, Upanishads etc, if it helps us to learn more about the general context of the ideas of YS?

    • Thanks, Evgenija. I am not sure I agree completely with you, since the categories we are referring to are not just etic ones. Sanskrit authors speak of kāvya, itihāsa (like the Mahābhārata), śruti (like the Upaniṣads) and śāstra (like the various schools of philosophy). I am all in favour of reading X while having in the background the coeval texts Y and Z, but I would not say that no distinction was felt.

      • But it is not a rear case, that such Sanskrit classifications are based on the idea to classify everything, not on the real contents of the texts. And sometimes they were formulated after the texts were composed. E.g. it is ok for a modern Hindu guru to say, that the Vedas and the Upanishads are shruti. Still for a scholar the evident gap between them is evident, as well as the gap even between different suktas (compare a standard praise from a family mandala and a ‘philosophical’ hymn from the 10th mandala of RV).
        Similarly, there is a gap between narrative parts of epics on the one hand and Gita and Mokshadharma on the other. Most probably (as Y. Vassilkov claims), didactic parts of the MBH were added in the period, when the brahmins in the tirthas started reciting the epics, instead of the kshatriyas.
        Of course, one should not totally ignore these genre classifications, but they should not be overestimated.

  3. Hi Elisa, Edwin has been making this case for some time. You can see in it argued in his translation of and commentary on the YS (North Point Press, 2009).

    In a forthcoming article (“Hindu Disproofs of God: Refuting Vedāntic Theism in the Sāṃkhya Sūtra,” Oxford Handbook of Indian Philosophy) I deal with the problem of the god of the YS somewhat tangentially. There I suggest that terms like “creator god” and “omnipotence” may not be categories generalizable across cultures. For instance, Brahma Sūtras 2.1.35-37, when read carefully, suggest not a “creator god” but perhaps more of an “organizer god.” There Īśvara does not create the individual selves, as they are themselves beginningless; the BS is quite explicit about this (in its response to a pūrvapakṣa making a Sāṃkhya-like argument against god’s existence). He is also not higher than karma, whose dictates he must follow strictly (or otherwise be guilty of the Sāṃkhya charge of playing favorites among individuals).

    If you read the bhāṣya on YS 3.45, you’ll see that the bhāṣyakāra, whoever he was, describes Īśvara as using his extraordinary powers to put the world in order at the beginning of each world-cycle. But, on my reading, he is not a “creator” per se here either. He is more of an adhiṣṭhātṛ (governor, organizer). So, perhaps not quite the same as Kṛṣṇa in the BhG.

    As food for thought, here’s what H. Nakamura wrote about Īśvara in the Brahma Sūtra: “. . . the attempt to use the law of Karma to solve the problem of individual sufferings in a Brahman-created world is a special point of the philosophy of Brahma-sūtra. At the same time the Highest God was not an absolutely free personal god, because he is dependent upon external factors for world creation. Since he merely allocates the karmic effect appropriate to the individual self, his function was that of an automaton. He is a stern god and not a god of grace; he is a god who makes possible individual action, bondage, and liberation and is the basis of all things, but merely acts as a mechanism and does not positively encourage either good or bad acts on the part of individual selves. This god merely abides (sthiti) without doing anything in particular (1.3.7), for the spiritual liberation of individual self is dependent upon the religious discipline and practice of the individual. The burning bhakti worship of later Hindu sects is not seen in the Brahma-sūtra” (History of Early Vedanta, Vol. 1, p. 496).

  4. If the YŚ’s īśvara could, at the very least, be incontrovertibly established as the “first organizer” (as Andrew mentioned), that would be a huge improvement on some modern readings. The fact that īśvara is both “the teacher of the ancients” and the “previously perfected one” who has ordered the world, demonstrates (along with other intra-textual support) that He is an actual, active entity. Thus, the YŚ’s īśvara is not a mere idea or archetypal guru (Eliade), not one who only teaches “by sheer being” (Feuerstein), and is not a poetic or pedagogic description of the ordinary puruṣa in the liberated state (Pflueger). How īśvara is actually does this, i.e. by taking on a body of pure sattva, is an altogether different discussion.

    Arguing that īśvara creates (or is) the pradhāna and/or the individual puruṣas is a taller task indeed as there is scant direct evidence in the YŚ. Edwin takes this step largely based on similar descriptions of īśvara found in the Gītā. I generally side with his approach because I do not think that Patañjali is founding an entirely new definition for the the term in the YŚ, and hence one has to look at common understandings of the word in similar śāstra-s of the time period (hence my initial question to Elisa).

    *On a side note, Hariharānanda’s ideas about how the individual puruṣa can become a world creator/organizer are interesting, even if they diverge from the traditional commentators.

    • Granted that creation e nihilo is not in question in classical India, I wonder what we would do of the connection between Yoga and Sāṅkhya in the case of an organiser God… YBh on 3.45 seems to be a good argument (thanks, Andrew!) in favour of a deeper break between (classical, atheist) Sāṅkhya and Yoga, especially if we agree that YS and YBh are a single text.
      As for the argument about the continuity with the background, let me play the role of the pūrvapakṣinī again and ask why emphasizing the BhG at the expenses of Sāṅkhya, the influence of which is apparent in the PYŚ.

      • Recognizing the Gītā and Mokṣadharmaparva in this discussion is necessary because, following Dasgupta, Larson, and others, “Pātañjala Sāṃkhya” resembles a form of Epic sāṃkhya more than Classical Sāṃkhya. Thus our reliance on the Sāṃkhyakārikā to illuminate the PYŚ has come as both blessing and curse. Hence, your statement, “the influence of which is apparent in the PYŚ”, is true, but the question is, “Which sāṃkhya?” There is little reason to assume that the PYŚ even knows of a sāṃkhya like Īśvarakṛṣṇa’s.

  5. Dear colleagues,
    As we can see from the previous discussion, Yoga-Theology is tricky, because philosophy and philology are (again) closely related. Taking the Yoga Sutra as a separate work and trying to extract from this the theological views of its author is extremely difficult, due to the brevity of this (part of a) work. In my estimation, the sūtra-s 1.23-27, 2.1, 2.32 und 2.45, which are the only ones containing the word īśvara (or a pronoun substituting a form of īśvara) do not allow for a reconstruction of the theology of yoga in any detail.

    The material becomes more comprehensive if one including the so-called bhāṣya-part of the PYŚ as a source of information. A further source that (not only) in my estimation is useful for studying Pātañjala-Yoga is the Pātañjalayogaśāstravivaraṇa. All these sources are difficult to interpret in their own right, and complete critical editions do not yet exist.

    The īśvara-section of the Vivaraṇa is the topic of a book by Kengo Harimoto entitled “God, Reason, and Yoga” (Hamburg 2014), which contains a first critical edition, and an annotated translation of this this section.

    If we address the sources with the question of whether the īśvara of Yoga is a special kind of transcendental subject (puruṣa) and whether he is omniscient and almighty, the answer is: He is both. More specifically, that he is a subject (puruṣa) is established in YS 1.24 (kleśakarmavipākāśayair aparāmṛṣṭaḥ puruṣaviśeṣa īśvaraḥ), whereas his being almighty and all-knowing is mentioned in the bhāṣya-part of the same section (PYŚ 1.24,15 in Maas 2006): tac ca tasyaiśvaryaṃ sāmyātiśayamuktam (or -vimuktam).

    As Andrew correctly mentioned before, PYŚ 3.45 is also relevant for this discussion, because the term aiśvarya reoccurs here to designate eight superhuman powers. The last of these is the power to achieve whatever one wants. The passage reads in the edition of Āgāśe (Poona 1904, ĀĀSS 47):

    yatrakāmāvasāyitvaṃ satyasaṃkalpatā yathā saṃkalpas tathā bhūtaprakṛtīnām avasthānam | na ca śakto ’pi padārthaviparyāsaṃ karoti | kasmāt | anyasya yatrakāmāvasāyinaḥ pūrvasiddhasya tathā bhūteṣu saṃkalpād iti | etāny aṣṭāv aiśvaryāṇi.

    On the backdrop of the Vivaraṇa and of three important manuscripts (from Ahmedabad, Jaisalmer and Trivandrum), I would suggest changing the reading satyasaṃkalpatā to satyasaṃkalpatvāt, and bhūteṣu to teṣu, which then refers to padārtha- (meaning just “thing”). The Vivaraṇa identifies the pūrvasiddha indeed as parameśvara, and provides the turning of fire into being cold as an example of padārthaviparyāsa (p. 302).

    This material does not, however, say much if anything about the role of īśvara in the “creation” of the universe. I therefore suspect that Andrew interpreted the passage cited above as describing “Īśvara as using his extraordinary powers to put the world in order at the beginning of each world-cycle.“

    Whatever solution may by found by looking at the material more closely than I can do at the moment, it is unlikely that the general views of Sāṅkhya, according to which the world as we see it is a transformation of primordial matter (pradhāna) caused by the presence of puruṣa-s, who are completely unchanging and inactive, will have to be modified. But one never knows in advance …

    • many thanks for joining the discussion, Philipp, and for your well-thought comments. Notwithstanding my admiration for B.K. Matilal, I think that in-depth philological scrutiny is an unsurpassable way to unearth treasures which went unnoticed to previous editors and scholars who might have been smoothing edges.

    • Thank you very much Philipp. I light of your response, I am still curious about your opinion on my original question. Given the usage of the term “īśvara” (with parameśvara and maheśvara) during this epoch, if we agree that the PYŚ’s īśvara is “almighty and omnipotent”, omniscient, a world-organizer, the first teacher, and designated by the praṇava, is not this īśvara most plausibly understood in the sense of KṛṣṇaNārāyaṇa/Śiva? Around the 4-5th cent. CE, do we have any evidence in mokṣa śāstras for an “īśvara” with these grandiose characteristics who is not understood as such?

      One may argue that Patañjali conceived an entirely new definition for the term in the PYŚ, but this claim is both tenuous and speculative.

  6. Elisa, I have reasons to believe that the epics were composed sometime between 3rd century BCE and 2nd century CE. The earliest epigraphic evidence of writing is from c. 260 BCE (both Brahmi & Kharosthi, as evidenced by the Ashokan edicts, with the Kharosthi being the older script of the two and may have been in existence for about 100 years by then). This was a time when these scripts were new and were not suitable for a phonetic representation of late vedic (the spoken language), hence between c. 350 BCE and 100 CE, you only find an imperfect & non-standardized epigraphy in Kharosthi & Brahmi of what must have been late vedic (and which is now incorrectly called Gandhari & middle-Indic similar to Pali). The Buddha died around 400-375 BCE. Panini is from 350-300 BCE, and probably wrote his grammar to standardize the spoken language that had been badly epigraphically mutilated by the Buddhists trying to write their canon in the yet imperfect Brahmi & Kharosthi.

    All Classical Sanskrit literature is posterior to Panini . Hence I believe that the epics can only be from the period after 300 BCE.

    This is why I also think that Sanskrit regained importance later on, for there was no such thing as spoken middle-Indic in the BCE era.

    • Thanks for this comment. I agree only in part (Sanskrit surely regained importance later on, but as for the writing systems I would rather use von Hinüber). More importantly, I am not sure I understand in which sense what you say is relevant for the understanding of “Īśvara” in the YS/PYŚ. Could you elaborate?

  7. The word īśvara in the Nyāya-sūtras and Nyāya-bhāṣya seems not to be used to refer to a creator of the universe. See the attached pages by Nagin Shah.

    The conception of īśvara there is rather that of a teacher (upadeṣṭā) and jīvanmukta. God was considered to have been previously a bound soul. Pakṣilasvāmin gives a very detailed list of the characteristics of īśvara, and being the creator of the universe is not one of them. (I’m of course using ‘creator of the universe’ in the Indian theistic sense, where it doesn’t refer to a creator of souls or matter – both of these being eternal – but rather to a kind of moulder of this pre-existing matter at the beginning of each cosmic cycle.)

    If it’s not īśvara who joins together the atoms after pralaya, who/what is it? According to the Vaiśeṣika-sūtras this joining is caused by adṛṣṭa.

    It seems that it is Praśastapāda who introduces the idea of īśvara as the creator of the universe into Nyāya and Vaiśeṣika.

    So I find the argument that the word īśvara in the Yoga-sūtras must refer to the creator or the universe because the same word does so in the Bhagavad Gītā weak.

    I don’t think I can add attachments to a comment after all. If anyone would like the Nagin Shah piece about Īśvara in early Nyāya and Vaiśeṣika, mail me and I can send it by email. It is an appendix to volume 2 of his: ‘A Study of Jayanta Bhaṭṭa’s Nyāyamañjarī, A Mature Sanskrit Work on Indian Logic’.

    • Alex,
      Thanks for the file!
      A couple thoughts:

      First, if it is admitted that the NS/NBh claim that īśvara was a previously bound soul who has “destroyed” impurities and achieved liberation in a former life, this would rub against the PYŚ’ conception of an īśvara who has never been samsarically bound. The only notable commentator I can think of who claims otherwise is Hariharānanda. Hence there would be reason to argue that the two conceptions of īśvara are dissimilar. Not to mention that in the PYŚ there is no explicit connection of īśvara with the eightfold yogic powers, as these are only mentioned in the context of the aspiring yogin.

      >So I find the argument that the word īśvara in the Yoga-sūtras must refer to the creator or the universe because the same word does so in the Bhagavad Gītā weak.

      I don’t see how this directly relates to the NS/NBh issue. I would also say that the argument for creatorship in the PYŚ (or at least my own argument) does not stem from the word “īśvara”‘s understanding in the BG. I find YBh 3.45 sufficient to substantiate a type of governorship, if not creatorship, to īśvara in the PYŚ.

  8. Thank you for the text, Alex.

    Certainly, Vatsyayana’s Izvara sounds like more of a perfected Yogin than the God of Uddyotakara, Vacaspati, and the rest who is the efficient cause of the world. But it’s unclear to me if V. is just using Yogic vocabulary to express the nature of God, or if he is making the more substantial claim that Izvara is just a special sort of Yogin. (Bulcke pointed out connections with the bhASya’s description of Izvara and the Yoga school in his book on nyAya/vaizesika theism from the 40’s).

    Shah (p. 213-219) has made a few translational choices on the nyAyabhASya that are not necessarily given in the text (like saying “mundane soul”; where is “mundane” in the text?). I think that Shah also emphasizes “process” here more than is given in the text. But there are some hints of it, to be sure. For comparison, here’s how Phillips and I have translated the bhASya on the IzvarasUtras.

    Vātsyāyana [228.3-16]: God grants causal power to what has been done by human beings. That is, for an individual who is acting for a certain karmic fruit, God makes that fruit come forth. If God did not, human action would be karmically fruitless. Thus, what humans do is not sufficient for karmic fruition, since fruition is actuated by God, and yet in the absence of human action there is no fruition.

    God is distinct from other selves owing to his special qualities. God is most akin to a self; it doesn’t make sense to think of him as something else. But he is different from other selves in that demerit (adharma), mistaken cognition, and delusion are absent while he is perfect in merit (dharma), knowledge, and yogic concentration and accomplishment (samādhi). As a consequence of his merit and concentration, he possesses lordliness in the form of the eightfold yogic powers (siddhi) such as the ability to make oneself as small as an atom. His merit (dharma), which conforms to his intention, activates the merit and demerit collected in each individual self as well as in gross elements such as earth. And so God’s irresistible will in creating should be understood as enacting what individuals have themselves done; that is to say, it is not insulated from the influence of what individuals themselves have done.

    God is a trustworthy authority. His relationship to creatures is like that of a father to his progeny. His nature is not to be thought of as something other than that of a self. None of his properties aside from knowledge could serve as an inferential mark proving his existence, and from sacred tradition we know that God is a perceiver, a knower, and omniscient. And if the Lord were entirely beyond the range of perception, inference, and scripture, who could ever show that he was not knowable through a psychological characteristic such as knowledge?

    If one proposed that God’s creative actions occur without being influenced by what individuals have themselves done, this proposition would be refuted for all the reasons we have given against the view that bodies are born independently of karmic merit (see NyS 3.2.60-72).


    • Many thanks for this anticipation, Matthew, and congratulations for your fluent translation. To be honest, more than “mundane”, what strikes me in Shah’s choices is what you also render differently, namely pratyātma- (in pratyātmavṛttīn dharmādharmasañcayān […] pravartayati) and nirmāṇaprākāmyam. For readers who did not read Shah: Shah thinks that praty- in the first compound is not distributive, but only means “relative to”, as in pratyakṣa. Thus, the whole expression would mean something like: ‘he impels dharma and adharma present in his own soul’. As for nirmāṇaprākāmyam, Shah interprets it as referring to the will to create yogic bodies (nirmāṇa*kāya*), which are needed to consume the accumulated karman. He cites further passages by Vātsyāyana and Jayanta to substantiate the claim that jīvanmuktas ‘souls liberated while alive’ can create yogic bodies for this purpose. What do you think of it? Did you check the coeval usage of nirmāṇa in philosophical texts?

      A further question: I tend to avoid the verb “create”, since I am afraid it evokes in a Western-educated audience the idea of a creatio e nihilo, which is unknown in Indian sources of that time (as far as I know). What do you think about it?

      • Hi Elisa. Thanks so much for your reply and kind words.

        I don’t mind “create.” First of all, it has nice resonances with kR. And we use “create” all the time to describe ordinary acts of making that depend on existing ingredients. Also, I think that most readers are not really literate in or informed by a distinct Christian inheritance in the way that they may have been a generation or two ago. Finally, the word “God” has such baggage, too, and many other words besides. But we use them, and we explain or allow context to make clear the shades of meaning here.

        I didn’t look up nirmāṇa, as I’ve seen it used to mean something like “making” or “fashioning” in this context elsewhere. (As far as I know, Stephen didn’t either.) We used a “thin” reading of the term.

        I hope you are doing well!

Leave a Reply

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>