Chapter 4 Further Thoughts on Rāmakaṇṭha’s Relationship to Earlier Positions in the Buddhist-Brāhmaṇical Ātman Debate

In: Śaivism and the Tantric Traditions
Alex Watson
Search for other papers by Alex Watson in
Current site
Google Scholar
Open Access

The present article is a continuation of my previous work on where precisely to place Rāmakaṇṭha’s self-theory (ātmavāda) in the nexus of other rival positions. I am delighted to have been included in this volume and in the conference which led to it, in honour of my former DPhil supervisor, Professor Alexis Sanderson, with whom I spent many hours reading Rāmakaṇṭha’s Nareśvaraparīkṣāprakāśa (as well as Kumārila’s Ślokavārttika, ātmavāda, and Vasubandhu’s Abhidharmakośabhāṣya, pudgalaviniścaya)—and indeed at whose suggestion I began working on Rāmakaṇṭha’s philosophical texts.

A previous article of mine (Watson 2014) places Rāmakaṇṭha in the middle ground between Nyāya and Buddhism. What I would like to do here is present some considerations that run counter to that. I do not think they invalidate my earlier contentions, but they do reveal them to be one-sided and incomplete.

In section 1 I introduce key issues in the self debate between Nyāya and Buddhism, in order to then be able to locate Rāmakaṇṭha in relation to these two. In section 2 I briefly explain my “middle ground” idea that was put forward in the 2014 article.1 In section 3 I present evidence for seeing Rāmakaṇṭha as just as extreme as Nyāya. In section 4 I present evidence for seeing him as being even more extreme than Nyāya, with Nyāya being the moderate position. In section 5 I present evidence for Nyāya not being so moderate after all. In the concluding section I ask where all of this leaves us.

1 The Naiyāyika-Buddhist Debate about the Existence of the Self

As a means of identifying what precisely separated Nyāya and Buddhism2 on the question of the self (ātman), I like to use a triple distinction, one that I had not seen in secondary literature prior to my 2014 article, but which I think arises naturally from the primary sources. This is a distinction between three ways in which the self was envisaged by the Naiyāyikas, and three corresponding ways in which the self was argued against by the Buddhists. To elaborate the idea, I will use the same set of diagrams that I used in my 2014 article.


Figure 4.1 is intended to illustrate the difference between the Buddhist doctrine of momentariness, according to which we are something different in every single moment, and the Brāhmaṇical idea of the self as the unchanging essence of a sentient being. The self is something that remains the same—both numerically and qualitatively—over time; a Buddhist individual is both numerically and qualitatively different in each moment.


Whereas in figure 4.1 we were focusing on the conception of the self as unchanging essence, in figure 4.2 we are dealing with the self as substance, represented by the large circle on the left—a substance that is the owner of certain qualities (guṇa), represented by the small circles. For Naiyāyikas qualities cannot exist without some substance in which they inhere; hence we can infer the existence of the self from the existence of qualities such as cognition, pleasure and pain.3 For Buddhists cognitions, feelings and the other constituents (skandha) of an individual exist bundled together, without belonging to some greater whole.

In figure 4.3 we are dealing with the conception of the self as agent; and we have to distinguish two senses of “agent”—the doer (kartṛ) of actions, and the subject of cognitions (jñātṛ). Actions and cognitions (the circles can represent either) all share a common agent for Nyāya, namely the self (represented by the line). For Buddhism they do not: the agent of one particular action or cognition will have ceased to exist by the time another action or cognition comes into being.

The Nyāya and Vaiśeṣika concept of the agent of physical actions was sometimes elaborated by comparison to a puppeteer: as a puppeteer brings into being movements of the body of the puppet, so the self brings bodily movements into being.4 And the Naiyāyikas and Vaiśeṣikas both borrowed the Grammarians’ (Vaiyākaraṇas) definition of the agent as that which is autonomous, i.e. that cause of action which does not depend on anything else.5 Buddhism opposes this concept of an autonomous agent standing over the psycho-physical stream of events, manipulating it from above. Any physical action will depend on the previous moment of consciousness, this will depend on what causes it, and the latter will depend on what causes it, etc. There is nothing here that is autonomous; and if there were it would not be able to be a cause as it would stand outside the causal chain. So for Buddhism the agent of an action is simply its principal cause, and this will be an event within the psycho-physical stream.6 Buddhism thus replaces a two-tier model with a one-tier one.

If we are talking not of a physical action but specifically of a cognition, its agent will not be a previous moment of consciousness but rather the cognition itself. The agent of a perception, i.e. the thing doing the perceiving, is the perception itself. The subject of any act of awareness, i.e. that which is aware, is the stream of consciousness at that particular moment. So here too we have a contrast between a Naiyāyika two-tier model, in which the cognizer (jñātṛ), i.e. the self, is ontologically quite distinct from the cognition (jñāna), versus a Buddhist one-tier model in which the cognizer is pluralized and dispersed into the stream of cognitions.

2 Rāmakaṇṭha in the Middle Ground

Where does Rāmakaṇṭha fit in all of this? On two counts he falls with Nyāya and on two counts he falls with Buddhism. He falls with Nyāya (1) in opposing the Buddhist doctrine of momentariness and in upholding the existence of a self that is the unchanging essence of the individual; and (2) in maintaining that there is an autonomous, unconditioned agent of our physical actions. But he agrees with Buddhism (1) that cognition does not inhere in something other than itself: there is no self-substance over and above cognition; and (2) that there is no cognizer over and above cognition, no agent of awareness over and above awareness. For him, as for Buddhism, that which does the cognizing or perceiving is just cognition/awareness/consciousness (jñātṛ/grāhaka = jñāna).

How does he manage to reconcile all of these positions? I.e., how does he manage to preserve the existence of an unchanging, autonomous self when he denies that there is anything—any substance or agent—over and above cognition/consciousness? The answer is that he equates the self and cognition/consciousness (jñāna/saṃvit/cit/caitanya). This means that he has to maintain that cognition/consciousness is permanent and unchanging, not plural and momentary as it is for both Buddhism and Nyāya. Thus the three views can be laid out on a continuum.

For Nyāya there is both cognition, and a self over and above that. For Buddhism there is no self, there is just cognition. For Rāmakaṇṭha there is a self, but it is just cognition. Rāmakaṇṭha falls in the middle in the sense that, like Buddhism, he as it were crosses out the line under Nyāya—he reduces the self to cognition; but he does not go all the way down the path of Buddhist argumentation, as he joins up the dotted line into one unbroken one: he argues that cognition is permanent.


Figure 4.4

Rāmakaṇṭha’s view in the middle ground

I will not here dwell on how Rāmakaṇṭha further elaborates and defends his idea of unchanging cognition, as I have done so elsewhere.7

3 Rāmakaṇṭha as Extreme as Nyāya

It is at this point that I would like to question and supplement my earlier line of thinking. What precisely is this continuum intended to capture? What determines the particular location on it that a position will occupy?

One thing it might be intended to capture is the amount of change that a position posits in the subject of experience. On the left we have the Naiyāyika position according to which the self is eternally unchanging. On the right we have the Buddhist position according to which the subject is changing, both qualitatively and numerically, in every moment. But for Rāmakaṇṭha the self is completely unmodifiable (avikārya), so on this measure his position should fall with Nyāya at the left hand extreme.

One thing it was certainly intended to capture is heaviness or lightness of postulation. On the right we have the Buddhist position that postulates no more than all the disputants agree exists: a sequence of cognitions. On the left we have the Naiyāyika position that postulates an extra entity over and above that: an eternal self that is the owner and knower of those cognitions. On this measure it looks as though the Naiyāyika position should indeed fall to the left of Rāmakaṇṭha’s: the former seems to postulate two things where Rāmakaṇṭha postulates only one.

But even here there is a consideration that seems to negate this difference. Rāmakaṇṭha actually accepts two very different kinds of “cognition” (jñāna). In Rāmakaṇṭha’s discussions with Buddhism, a common objection put by the Buddhist pūrvapakṣin is: if cognition (jñāna) is unchanging and single,8 as you Saiddhāntikas claim, how can we account for the difference between a cognition of blue, a cognition of yellow, an awareness of pain, an awareness of pleasure, etc.? Rāmakaṇṭha’s usual answer is to maintain that throughout such a sequence cognition itself is indeed single and unchanging.9 All change happens on the object side of the subject-object (grāhaka-grāhya) divide. The subject/perceiver, i.e. cognition or awareness, is like a light shining out always in the same form; the red, blue, pleasure and pain are different objects that are illuminated in turn by the light, but they do not affect the nature of the light (= cognition) at all.

But sometimes he gives a different answer, namely that there are two different kinds of cognition, i.e. that the word jñāna can be used in two different senses. There is cognition proper, i.e. the self (ātman), the perceiver (grāhaka/jñātṛ), and there is the cognition that is located not in the self but in the intellect (buddhi). The first is termed pauruṣaṃ jñānam,10 the cognition of the self, or grāhakātmasaṃvit,11 cognition whose nature is the perceiver; the second is termed adhyavasāyātmakaṃ jñānam12 / adhyavasāyātmikā saṃvit,13 cognition that is of the nature of determination, or parāmarśātma jñānam,14 cognition that is of the nature of verbal determination.15 The second kind is not unitary and unchanging, but divided into discrete instances, verbalizable as, for example, “this is blue,” “this is yellow,” “this is a pain.” If these are discrete instances of cognition (jñāna), how is this answer compatible with the first answer that all change occurs on the side of objects, not on the side of the subject? The answer is that these discrete instances of determination are actually objects of cognition in relation to the perceiver.16 If we are using “cognition” to refer to the perceiver, they are objects of cognition; but they themselves can be referred to as “cognitions” if we understand “cognitions” to mean determinations (adhyavasāya, parāmarśa). These products of the intellect (buddhi) are transient, i.e. plural over time; the nature of the self is unchanging and single over time.

This is no innovation of Rāmakaṇṭha’s: both the earlier exegetes in his Saiddhāntika tradition, and the Saiddhāntika scriptures themselves, locate jñāna at two different places on the scale of tattvas that is partially inherited from Sāṅkhya: within the material world (i.e. that which evolves out of māyā) at the level of buddhi-tattva, where jñana features as one of the eight properties (referred to in Śaiva literature with such terms as bhāva, dharma, guṇa, aṅga) of the buddhi,17 and as the immaterial nature of selves (which are qualitatively, but not numerically, identical to Śiva).

Bearing in mind this distinction between two different kinds of cognition, let us revisit the situation that is illustrated in figure 4.4. That diagram, since it prints the line under Rāmakaṇṭha on the same level as the dotted line under Nyāya and the dotted line under Buddhism, makes the following assumptions: (1) Cognition for Rāmakaṇṭha refers to (more or less) the same thing as cognition for Nyāya. (2) Cognition for Rāmakaṇṭha refers to (more or less) the same thing as cognition for Buddhism. These were not unreasonable assumptions. They reflect the way that Rāmakaṇṭha himself presents the situation. When arguing against Nyāya, he effectively says: unlike you Naiyāyikas, we accept only cognition, not some extra self over and above that in which it inheres.18 And when arguing against Buddhism he assures us that his self/cognition is just the jñāna that Buddhists talk about, not some container of it, some self-substance to which it belongs.19

But now that we have seen that Rāmakaṇṭha actually recognizes two kinds of cognition, different possible assumptions, and hence different possibilities for the drawing of the diagram, emerge. We could take Buddhist jñāna as corresponding more to Rāmakaṇṭha’s plural cognitions. And if we do, then the fact that Rāmakaṇṭha adds an extra layer above that makes him very much akin to Nyāya—at least on this point of how many things they are postulating.

Similarly, we could take Rāmakaṇṭha’s first kind of cognition as comparable to Nyāya’s self, and the second kind as comparable to Nyāya’s cognitions (jñāna). When we adopt this perspective, the difference between Nyāya and Rāmakaṇṭha regarding the extent of their postulation disappears. We arrive at the situation depicted in figure 4.5.


Figure 4.5

Rāmakaṇṭha’s view as equivalent to Nyāya

To what extent is it reasonable to take the first kind of cognition as equivalent to Nyāya’s self? It is reasonable to the extent that we are aligning Nyāya’s self and Rāmakaṇṭha’s self, though the former is not of the nature of jñāna whereas the latter is. To what extent is it reasonable to take the second kind of cognition as equivalent to Nyāya’s cognitions and Buddhism’s cognitions? The fact that the second kind is plural makes it a more natural equivalent than the unitary first kind; but the fact that the second kind of cognition is unconscious (a product of primal matter) makes it less equivalent to Buddhist and Naiyāyika cognitions, which are sparks of awareness, than the first kind, which is equivalent to awareness.

As well as the amount of change in the subject, and the amount of postulation, there is a third possibility for what the continuum could capture: the amount of dynamism in the subject. On this measure does Rāmakaṇṭha’s self fall closer to a Buddhist stream of consciousness than a Naiyāyika self does? Yes. Nyāya’s rock-like20 self is in its own nature devoid of consciousness and agency; Rāmakaṇṭha’s is of the nature of an outpouring—the outpouring of the light of consciousness—and during liberation it remains not just conscious, but also an agent. But Rāmakaṇṭha’s self is not as dynamic as that of the non-dualistic Śaivas. For them vimarśa, representative cognition, is a power (śakti) of the self.21 Rāmakaṇṭha resists this move, since for him (as for them) there is no difference between a power and the thing that has the power (śakti and śaktimat), so to accept that vimarśa, which is changing, is a power of the self would have been to accept some change in the self. Vimarśa for him belongs in the buddhi, not the self;22 so it does not affect the unchanging nature of the self. Rāmakaṇṭha was more concerned than the non-dualist Śaivas to protect the self from any change, though his self does have more dynamism than a Naiyāyika one. There is a tension in Rāmakaṇṭha’s self between its lack of change and its dynamism, one to which we will return in the final paragraphs of the article.

4 Rāmakaṇṭha More Extreme Than Nyāya

We have seen that by two out of these three measures, Rāmakaṇṭha’s self looks to be just as “extreme” as Nyāya’s. In fact there is one consideration that arguably makes it more extreme. Rāmakaṇṭha’s self is perhaps even further removed from change than Nyāya’s, given that for Nyāya changing cognitions inhere in the self. For Rāmakaṇṭha, by contrast, changing cognitions, as we saw above, are objects of awareness (grāhya) for the self; they occur outside of it, in the intellect (buddhi). So we arrive at a different continuum, illustrated in figure 4.6, where the large circle is the self and the dotted lines are cognitions. The position that postulates the most amount of change in the subject is the Buddhist position on the right; the position that postulates the least amount of change in the subject is Rāmakaṇṭha’s position on the left; Nyāya falls in the middle.


Figure 4.6

Rāmakaṇṭha’s view as more extreme than Nyāya

And what is asserted here of cognitions is also true of latent impressions (saṃskāras), pleasures and pains. Whereas for Nyāya these all reside in the self, Rāmakaṇṭha and his fellow Saiddhāntikas, arguing that this would entail the unwanted consequence that the self is subject to change, locate latent impressions, pleasures and pains outside of the self in the buddhi.23

The fact that we are dealing here with two different conceptions of the relationship between self and cognitions—for Nyāya inherence (samavāya), and for Rāmakaṇṭha a subject-object relation (grāhyagrāhakabhāva)—means that certain objections that are put to Rāmakaṇṭha are not so applicable to Nyāya.

First, a common objection that is articulated in Buddhist texts against a Sāṅkhya self,24 and in Rāmakaṇṭha’s texts as a Buddhist objection to his own view,25 is as follows. If the self were completely unmodifiable, it would not be able to perceive objects, for to perceive a pot, say, followed by a cloth requires being able to register the change in the object-sphere from the pot to the cloth. And there is no way to register a change without being affected by the change. The way in which a perceiver perceives an object is by being modified by that object in some way. If the perceiver is permanently unaffected and unmodified, it is mysterious how it can be a perceiver.

Here the Naiyāyika will respond that the self is modified to the extent that separate cognitions inhere in it, the first caused by the pot and the second caused by the cloth.

Second, Rāmakaṇṭha combines the claim that the self is unmodifiable with the claim that it is an agent. This lays him open to an objection that, unlike the last one, is not applicable to Sāṅkhya: how can something that is beyond all change be an agent? There are two aspects to this objection. (1) How can something that does not itself “move” cause movement? Here Rāmakaṇṭha’s response is to adduce the example of a magnet. A magnet causes movement in iron filings without itself moving; so similarly, the self causes bodily movement without itself moving.26 But even if it were conceded that the magnet example renders plausible a self as “unmoving mover,” there is a related but different problem. (2) If the self is beyond all change, how can we explain that at one moment it brings about one action and at another moment it brings about a completely different action? Surely an agent that is always in the same state would have to either always be doing nothing or always be doing exactly the same thing?

Here again, the objection is not so applicable to Naiyāyikas, for they accept the existence of different impulses (prayatnas) that inhere in the self. A prior bodily movement will be brought about by a particular impulse, and a subsequent, different kind of bodily movement will be brought about by a different impulse.

5 Nyāya Not so Moderate after All

Section 3 gave reasons for judging Rāmakaṇṭha’s position not to fall in the middle ground between Nyāya and Buddhism, but to be just as extreme as Nyāya.

Section 4 gave reasons for judging Rāmakaṇṭha’s position to be more extreme than Nyāya’s, with the latter as the moderate one, closer to Buddhism. But this section introduces a consideration that reveals Nyāya to be not so moderate after all.

It is true that Nyāya holds cognitions, impulses (prayatna), latent impressions (saṃskāra), pleasures, pains, etc., to inhere in the self, whereas Rāmakaṇṭha (in accord with his Sāṅkhya inheritance)27 locates all of these outside of the self, indeed outside of the realm of the immaterial altogether, within the world derived from māyā and prakṛti. We thus seem to be left with the picture given in figure 4.6, where the small lines can represent any of the things just listed: they fall within the self for Nyāya, but outside of it for Rāmakaṇṭha. But the statement that they fall within the self for Nyāya has to be qualified. For Nyāya actually regards them as part of the “not-self.” How can that be, given that they are qualities of the self? Because of Nyāya’s view of the firm difference between a thing and its qualities (guṇaguṇibheda). Any substance for Nyāya is a completely different thing from its qualities. Change in the qualities of a substance will not necessarily result in any change or modification of the substance. And there is even more distance between the self and its qualities than there is between a physical substance and its qualities. For the self is eternal, its qualities temporary; it is omnipresent, while they are restricted to a particular location.28 This Naiyāyika idea that the self in particular, and substances in general, are unaffected by changes in their qualities was one of the things that separated their view from that of Kumārila and the Bhāṭṭa Mīmāṃsakas. For the latter held the relation between a thing and its qualities to be difference-cum-nondifference (bhedābheda); they maintained that changes in the qualities of a thing do affect the thing itself. The self, for them, although never ceasing to exist, is subject to modification when its qualities change.29 The fact that the Naiyāyikas rejected this, protecting the self from any effect of change in its qualities, means that their view cannot be so starkly differentiated from Rāmakaṇṭha’s view as figure 4.6 suggests.

This impression is strengthened by Naiyāyika passages dealing with liberation (apavarga, mokṣa). Here Naiyāyika authors assert that the self’s nature is, and always has been, free of all its particular qualities (sakalaguṇāpoḍha).30 These qualities are thus irrelevant to its nature. They are described as extrinsic to it, not innate (na naisargikaḥ).31 They are part of the “not-self” with which it confuses itself while in saṃsāra, and which is to be abandoned (heya).32

This thoroughgoing difference and separateness of the self’s qualities from the self—indicated by all of the considerations above—means that the Nyāya position can arguably be represented as it is in figure 4.7, and hence regarded as just as extreme as Rāmakaṇṭha’s position.


Figure 4.7

Nyāya as equivalent to Rāmakaṇṭha’s view

6 Concluding Remarks

Where do all of these divergent analyses leave us? What is cancelled out by what? It may be better to see each perspective as having its own validity. We have four perspectives:

  1. Rāmakaṇṭha in the middle ground (section 2 and Watson 2014).

  2. Rāmakaṇṭha as just as extreme as Nyāya (section 3).

  3. Rāmakaṇṭha as more extreme than Nyāya (section 4).

  4. Rāmakaṇṭha as just as extreme as Nyāya (section 5).

Or rather, three perspectives, since 2 and 4 are the same. In what sense is 1 valid? The Śaiva equating of self and consciousness/cognition and its view of the self as the shining forth of prakāśa, the light of consciousness, is indeed something that differentiates Rāmakaṇṭha’s view from Nyāya in substantial and significant ways. Related to this is the Śaiva rejection of the existence of substances (dravya) over and above qualities (guṇa), or property-possessors (dharmin) over and above properties (dharma). And just as, for Rāmakaṇṭha, there is no self-substance over and above consciousness/cognition, so too there is no self as agent (kartṛ) or knower (jñātṛ/grāhaka) over and above consciousness/cognition. From the point of view of this equating of self and consciousness, perspective 1 is valid, and figure 4.4 captures a certain relationship that Rāmakaṇṭha’s position bears to Nyāya and Buddhism. This perspective and this figure also accord with the way that Rāmakaṇṭha’s self is more dynamic than Nyāya’s.

But it is not valid to see things only in this way, because when we add the consideration that Rāmakaṇṭha accepts two kinds of cognition, this allows for the possibility of aligning not his first kind, but his second kind (that which is plural and is located in the buddhi) with Nyāya’s and Buddhism’s cognition. When we do that, and we add Rāmakaṇṭha’s self (i.e. his first kind of cognition) above that, his position becomes equivalent to Nyāya’s (both in terms of the number of things postulated and the amount of change in the self)—and figure 4.5 becomes appropriate.

But this does not exhaust the number of ways of seeing the relationship between Rāmakaṇṭha and Nyāya, because there is a further significant consideration. While it is true that the selves of both Nyāya and Śaivasiddhānta are eternally unchanging (which supports the idea that they are both as extreme as each other: perspective 2), it is also the case that cognitions, impulses (prayatna), latent impressions (saṃskāra), pleasures, pains, etc., inhere in the self for Nyāya, but fall completely outside it for Rāmakaṇṭha. Emphasizing this makes it appear that Rāmakaṇṭha’s self, like a Sāṅkhya self, is even more removed from change than Nyāya’s. This perspective 3 (illustrated in figure 4.6) carries some weight, but is no final resting place, for if we concentrate on those places where Nyāya emphasizes that qualities are extrinsic to substances, and that consciousness/cognitions, etc., are part of the not-self, not affecting its nature in any way, then it appears that there is in fact no more change in a Naiyāyika self than a Saiddhāntika one. So to adopt perspective 4 is to revert to perspective 2, which this time can be illustrated with a slightly different diagram: figure 4.7.

The explanation for the difference between perspectives 3 and 4 is a certain tension within Nyāya, one that was already noted by Frauwallner in 1956.33 Frauwallner argues that the Naiyāyika self is the product of two separate influences. On the one hand, it derives from the self as characterized in the old philosophy of nature; on the other hand it was subject to continual attraction to the self of the Sāṅkhyas. According to the former, selves were of limited size, and hence capable of movement; they were that which transmigrates from body to body; and they were all qualitatively distinct from each other, as a result of being characterized by their own qualities (svaguṇaviśiṣṭa). There was thus a huge difference between these selves and those of Sāṅkhya: the latter were all-pervading and hence incapable of movement; it was not they that transmigrate (but rather a subtle body); and they were devoid of all qualities. But over time the difference of the selves of the Naiyāyikas and Vaiśeṣikas from those of the Sāṅkhyas decreased. The former became all-pervading and hence incapable of movement; they ceased to be that which transmigrates from body to body (the manas took on this role); and their relationship to their qualities became reassessed. They were no longer characterized by their own qualities (svaguṇaviśiṣṭa); rather those qualities came to be regarded as extrinsic to them, as part of the not-self, and as needing to be abandoned if the self is to rest in its own nature. Previously selves were affected by changes in their qualities; subsequently they were not. Previously selves were qualitatively different from each other; subsequently they were all qualitatively identical as a result of being, in their true nature, devoid of particular qualities.

The residues in a Naiyāyika self of earlier philosophy of nature weigh in favour of perspective 3; the influence of the Sāṅkhya model of a self pulls it towards perspective 4.

Just as the difference between perspectives 3 and 4 results from focusing on different aspects of Nyāya, so the difference between perspectives 1 and 2 results from focusing on different aspects of Śaivasiddhānta. Just as we have offered an explanation of the first difference in terms of a tension within Nyāya, is there also an explanation of the second difference in terms of a tension within Śaivasiddhānta? In fact there is. Rāmakaṇṭha’s self is one whose nature derives to a large extent from Sāṅkhya: passing cognitions do not inhere in it, saṃskāras do not reside in it, pleasures and pains fall outside of it. It is elevated above all change; it sits above and beyond the various tattvas that comprise the psycho-physical universe. This is the kind of self that Rāmakaṇṭha inherits from his scriptural tradition. But his concept of self is also the result of a second influence, one which goes back to some of the Śaiva scriptures34 and which had been increasingly making itself felt among the Śaivas of Kashmir (both Saiddhāntikas such as Nārayaṇakaṇṭha and non-dualists such as Utpaladeva) during the previous century. This second influence sees the self not as a static Sāṅkhya-like entity, but as dynamic and as an outpouring of light (prakāśa)—the light of cognition/consciousness. The contrast between the two influences is strong. (1) According to the first influence it is a passive non-agent; according to the second it is a fully-fledged agent. I.e., not only is it an agent while unliberated, as in Nyāya, but it is also an agent in liberation; indeed at that time its agency expands into omnipotence. (2) According to the first, jñāna falls completely outside of the self; according to the second, the self is of the nature of jñāna.

Perspective 2 (which was presented in section 3) results from focusing on the first influence, the Sāṅkhya inheritance. Perspective 1 (presented in section 2) results from focusing on the new Śaiva insights.


I would like to thank Michael Allen for a conversation over lunch in Harvard Square in 2013 that planted the seeds of many of the new directions of thought in this article.


In sections 1 and 2 some material is reused from the 2014 article.


I use “Buddhist” and “Buddhism” in this article to refer specifically to the Buddhism that was engaged with in debate by Naiyāyikas and Śaivas, that is to say the Abhidharma of Vasubandhu’s Abhidharmakośabhāṣya, and the Yogācāra and Sautrāntika of Dignāga, Dharmakīrti and their followers. What is said of Buddhism in this article does not, therefore, hold true of Madhyamaka, for example.


The argument involves three contentions, each of which had their own supporting arguments: (1) Qualities cannot exist without substances to which they belong; (2) Cognition, desire, aversion, pleasure, pain, and volition are qualities; (3) The self is the only possible substance to which these qualities could belong. See Nyāyavārttika ad 1.1.10, p. 62,12–18 (that is to say, p. 62, lines 12–18—a convention used throughout this essay), and Praśastapādabhāṣya p. 16,3–7. For the second stage of the argument in particular, see Nyāyavārttika ad 3.2.18, Nyāyamañjarī Vol. 2, p. 278,14–15 and Candrānanda ad Vaiśeṣikasūtra 2.2.28. For the third stage of the argument, see e.g. Nyāyamañjarī, vol. 2, pp. 284,6–293,2 and Nyāyasūtra 3.2.47 with the commentaries ad loc. See also Chakrabarti 1982; Oetke 1988, 255–256, 258–260, 280, 286–300, 359–360, 464; Matilal 1989, 74, 77; Matilal 1994, 286; Preisendanz 1994, 187, 209, 278–281; Kano 2001; and Watson 2006, 174–184.


See for example Praśastapādabhāṣya, p. 15,12 and Candrānanda ad Vaiśeṣikasūtra 3.2.4, p. 28,18–19.


Aṣṭādhyāyī 1.4.54.


Abhidharmakośabhāṣya, p. 1228,1–1229,1.


Watson 2006, 333–382; 2010; and 2014, 186.


“Single” here means single in each individual; the Saiddhāntikas, unlike the non-dualistic Śaivas, maintained that each being’s self/consciousness was eternally separate from every other’s.


See for example Nareśvaraparīkṣāprakāśa ad 1.6ab, pp. 26,19 ff., passage beginning kas tarhi nīlaprakāśāt pītaprakāśasya bhedaḥ? na kaś cit … (translated and discussed in Watson 2006, 335–348); Mataṅgavṛtti, vidyāpāda introducing 6:34c–35a, p. 172,7 ff. (translated and discussed in Watson 2006, 349–382); Nareśvaraparīkṣāprakāśa ad 1.5, p. 14,2–18 (translated and discussed in Watson 2006, 220 ff.), which has parallel passages at Mataṅgavṛtti, vidyāpāda pp. 158,5–10, 172,16–21, 173,1–7 and Paramokṣanirāsakārikāvṛtti ad 43, p. 168,50–169,62 (following the line numbers as printed on those pages); Nareśvaraparīkṣāprakāśa ad 1.6ab, p. 26,4–13 (translated and discussed at Watson 2010, 111–112 and 2014, 186, note 26).


Kiraṇavṛtti ad 2.25ab, p. 54,3.


Mataṅgavṛtti, vidyāpāda p. 174,1–2 and 174,8.


Kiraṇavṛtti ad 2.25ab, p. 54,2.


Mataṅgavṛtti, vidyāpāda p. 174,4.


Mataṅgavṛtti, vidyāpāda p. 175,4.


For an elaboration of Rāmakaṇṭha’s distinction between these two kinds of cognition, see Watson 2006, 360–382.


See Kiraṇavṛtti ad 1:15, p. 18,33–35, Parākhyatantra 4:93ab, Tattvasaṅgraha of Sadyojyotis 13ab, Mṛgendratantra, vidyāpāda 10:23, Boccio 2002, and Watson 2006, 376 ff.


See for example Parākhyatantra 4:74 ff., Mataṅgapārameśvara 17.2, Mṛgendratantra, vidyāpāda 10:24 (cited at Goodall 2004, 254, note 383), Bhogakārikā 55, and Sadyojyotis’ commentary on Svāyambhuvasūtrasaṅgraha 2:12. This feature of Śaivism is inherited from Sāṅkhya; see Sāṅkhyakārikā 23.


In the Nareśvaraparīkṣāprakāśa ad 1.4cd, pp. 8,17–12,19 (on which see chapter 1 of Watson 2006, and Watson 2010, 86–89), Rāmakaṇṭha aligns himself with Buddhism against not only the Naiyāyikas, but also the Vaiśeṣikas and Sāṅkhyas, on the grounds that they all accept a self over and above cognition, whereas he accepts only cognition.


See Nareśvaraparīkṣāprakāśa ad 1.5, pp. 13,5–14,2 (discussed and translated in Watson 2006, 213–220); and Kiraṇavṛtti ad 2:25ab, p. 53,4–8 (discussed and translated in Watson 2010, 87–89).


See Nyāyamañjarī vol. 2 p. 432,4: jaḍaḥ pāṣāṇanirviśeṣa eva tasyām avasthāyām ātmā bhavet; “The self in that state [of liberation] would be unconscious, just like a rock.” These words are put by Jayanta into the mouth of an opponent; but the Naiyāyika siddhāntin does not deny the similarity of the liberated soul to a rock. I thank Harunaga Isaacson for this reference.


See Sanderson 1992, 288–289.


See (1) Mataṅgavṛtti, vidyāpāda ad 6:35b–d, p. 175,4–5 (and Watson 2006, 366), where Rāmakaṇtha asserts that parāmarśa, despite rising and passing away, does not indicate the non-eternality of the self, as it occurs not in the self but in the buddhi (and the ahaṅkāra): that vimarśa is synonymous with parāmarśa for Rāmakaṇṭha is suggested by Nareśvaraparīkṣāprakāśa ad 1:17, pp. 41,5–44,3 (translated and discussed in Watson 2006, 313–332), where they are used interchangeably; (2) Kiraṇavṛtti ad 2:25ab, pp. 53,10–54,17, where Rāmakaṇṭha distinguishes adhyavasāya, which is transient and a property (dharma) of the buddhi, from saṃvedanātmakaṃ jñānam, which belongs to the soul and is always occurrent; that Rāmakaṇtha uses adhyavasāya and parāmarśa (and niścaya) as synonyms is suggested by Mataṅgavṛtti, vidyāpāda ad 6:35b–d, p. 174,4–175,9; (3) Mataṅgavṛtti, vidyāpāda ad 17:2, p. 382,12–21, which distinguishes adhyavasāya from saṃvedana and identifies the former as the svabhāva of the buddhi; (4) Nareśvaraparīkṣāprakāśa ad 1:6ab, p. 28,12–19, where adhyavasāya is said to be a guṇa of the buddhi.


See Mataṅgavṛtti, vidyāpāda ad 17:2, p. 383,1–8 and ad 6:34c–35a, pp. 173,11–174,1.


E.g. Tattvasaṅgraha of Śāntarakṣita 287 and 288.


See Mataṅgavṛtti, vidyāpāda introducing 6:34c–35a, p. 172,7–10; Nareśvaraparīkṣāprakāśa introducing 1:6ab, p. 17,3–4; and Nareśvaraparīkṣāprakāśa ad 1.6ab, p. 26,19–20.


Kiraṇavṛtti ad 3:1; Paramokṣanirāsakārikāvṛtti ad 3c, pp. 115,99–116,106; and Watson, Goodall and Sarma 2013, 258–259, especially note 196.


For the Sāṅkhya stance on these matters and its difference from Nyāya, see Dasti 2013, 121–135.


For an account of the evolution of the increasing distance that developed between the self and its qualities in Nyāya and Vaiśeṣika, see Frauwallner 1956, 91–104; 1984, 61–71.


Ślokavārttika, ātmavāda 21–23.


Nyāyamañjarī Vol. 2, p. 359,6: sakalaguṇāpoḍham evāsya rūpam.


Nyāyamañjarī Vol. 2, p. 359,5.


Nyāyamañjarī Vol. 2, p. 265,10–12 and p. 430,3–4. Nyāyabhāṣya p. 6,9–11.


Frauwallner (1956, 91–104), (1984, 61–71).


See Watson (2010, 80).


Primary Sources

  • Abhidharmakośabhāṣya. Swāmi Dwārikā Dās Śāstrī, ed. Abhidharmakośa and Bhāṣya of Ācārya Vasubandhu with Sphuṭārthā Commentary of Ācārya Yaśomitra. Bauddha Bhāratī Series, nos. 5–7, 9. Vārāṇasī: Bauddha Bhāratī, 1970–1973.

  • Aṣṭaprakaraṇa. Brajavallabha Dvivedī, ed. Tattvaprakāśa, Tattvasaṅgraha, Tattvatrayanirṇaya, Ratnatraya, Bhogakārikā, Nādakārikā, Mokṣakārikā, Paramokṣanirāsakārikā. Yogatantra-granthamālā, no. 12. Vārāṇasī: Sampurnananda Sanskrit University, 1988.

  • Aṣṭādhyāyī. S.M. Katre, ed. Aṣṭādhyāyī of Pāṇini. Delhi: Motilal Banarsidass, 1989.

  • Kiraṇavṛtti. Dominic Goodall, ed. Kiraṇavṛttiḥ. Bhaṭṭa Rāmakaṇṭha’s Commentary on the Kiraṇatantra. Volume 1: Chapters 1–6. Collection Indologie, no. 86.1. Pondichéry: Institut français de Pondichéry/École française d’ Extrême-Orient, 1998. (In references to this edition, line numbers are not counted from the top of the page; they rather follow the line numbers printed in the edition, which start counting from the previous verse-segment.)

  • Tattvasaṃgraha of Śāntarakṣita. Dwarikadas Shastri, ed. Tattvasangraha of Ācārya Shāntarakṣita with the commentary “Pañjikā” of Shri Kamalashīla. 2 vols. Bauddha Bhāratī Series, nos. 1–2. Vārāṇasī: Bauddha Bhāratī, 1981–1982.

  • Tattvasaṃgraha of Sadyojyotis. See Aṣṭaprakaraṇa.

  • Nareśvaraparīkṣāprakāśa. Madhusūdan Kaul Śāstrī, ed. Nareśvaraparīkṣā of Sadyojyotis with the Prakāśa Commentary of Rāmakaṇṭha. Kashmir Series of Texts and Studies, no. 45. Reprint, Delhi, 1989 [Śrīnagar: Research Department Jammu and Kashmir State, 1926].

  • Nyāyabhāṣya. Anantalal Thakur, ed. Gautamīyanyāyadarśana with Bhāṣya of Vātsyāyana. Nyāyacaturgranthikā, no. 1. New Delhi: Indian Council of Philosophical Research, 1997.

  • Nyāyamañjarī. K.S. Varadacharya, ed. Nyāyamañjarī of Jayantabhaṭṭa, with Ṭippaṇi—Nyāyasaurabha by the Editor. 2 vols. Oriental Research Institute Series nos. 116 and 139. Mysore: Oriental Research Institute, 1969 and 1983.

  • Nyāyavārttika. Anantalal Thakur, ed. Nyāyabhāṣyavārttika of Bhāradvāja Uddyotakara. Nyāyacaturgranthikā, no. 2. New Delhi: Indian Council of Philosophical Research, 1997.

  • Nyāyasūtra. See Nyāyabhāṣya.

  • Paramokṣanirāsakārikāvṛtti. See Watson, Goodall and Sarma 2013.

  • Parākhyatantra. Dominic Goodall, ed. The Parākhyatantra. A Scripture of the Śaiva Siddhānta. Collection Indologie, no. 98. Pondichéry: Institut Français de Pondichéry/École française d’ Extrême-Orient, 2004.

  • Praśastapādabhāṣya. Johannes Bronkhorst and Yves Ramseier, ed. Word Index to the Praśastapādabhāṣya: a Complete Word Index to the Printed Editions of the Praśastapādabhāṣya. Delhi: Motilal Banarsidass, 1994.

  • Bhogakārikā of Sadyojyotis. See Aṣṭaprakaraṇa.

  • Mataṅgavṛtti. See Mataṅgapārameśvara.

  • Mataṅgapārameśvara. N.R. Bhatt, ed. Mataṅgapārameśvarāgama, Vidyāpāda, avec le commentaire de Bhaṭṭa Rāmakaṇṭha. Collection Indologie, no. 56. Pondichéry: Institut Français d’ Indologie, 1977.

  • Mṛgendratantra. Madhusūdan Kaul Śāstrī, ed. The Śrī Mṛgendra Tantram: Vidyāpāda & Yogapāda, with the Commentary (-vṛtti) of Nārāyaṇakaṇṭha. Kashmir Series of Texts and Studies, no. 50. Reprint New Delhi: Panini, 1982 [Bombay; Srinagar: Jammu and Kashmir State Archaeological and Research Department, 1930.]

  • Yuktidīpikā. Albrecht Wezler and Shujun Motegi, eds. Yuktidīpikā, the Most Significant Commentary on the Sāṃkhyakārikā. Alt- und Neu-Indische Studien, no. 44. Stuttgart: Franz Steiner Verlag, 1998.

  • Vaiśeṣikasūtra. Muni Sri Jambuvijayaji. Vaiśeṣikasūtra of Kaṇāda with the Commentary of Candrānanda. Gaekwad’s Oriental Series, no. 136. Baroda: Oriental Institute, 1961.

  • Ślokavārttika. Dvārikādāsa Śāstrī, ed. Ślokavārttika of Śrī Kumārila Bhaṭṭa with the Commentary Nyāyaratnākara of Śrī Pārthasārathi Miśra. Prāchyabhārati Series, no. 10. Varanasi: Tara Publications, 1978.

  • Sāṅkhyakārikā. See Yuktidīpikā (Appendix II).

  • Svāyambhuvasūtrasaṅgraha. Pierre-Sylvain Filliozat, ed. Svāyambhuvasūtrasaṅgraha: The Tantra of Svayaṃbhū, Vidyāpāda, with the Commentary of Sadyojyoti. Kalāmūlaśāstra Series, no. 13. Delhi: Indira Gandhi National Centre for the Arts/Motilal Banarsidass, 1994.

Secondary Sources

  • Boccio, Fabio. 2002. “Die Konzeption der buddhi als ‘Genußobject’ in Sadyojyotis’ Bhogakārikā.” In Śikhisamuccayaḥ. Indian and Tibetan Studies (Collectanea Marpurgensia Indologica et Tibetica), edited by D. Dimitrov, U. Roesler and R. Steiner, 11–26. Wiener Studien zur Tibetologie und Buddhismuskunde 53. Wien: Arbeitskreis für Tibetische und Buddhistische Studien, Universität Wien.

    • Crossref
    • Export Citation
  • Chakrabarti, Arindam. 1982. “The Nyāya Proofs for the Existence of the Soul.” Journal of Indian Philosophy 10: 211–238.

    • Crossref
    • Export Citation
  • Dasti, Matthew. 2013. “Nyāya’s Self as Agent and Knower.” In Free Will, Agency, and Selfhood in Indian Philosophy, edited by M. Dasti and E.F. Bryant, 112–136. New York: Oxford University Press.

  • Frauwallner, Erich. 1956. Geschichte der indischen Philosophie. Volume 2. Salzburg: Otto Muller Verlag.

  • Frauwallner, Erich. 1984. History of Indian Philosophy. Volume 2. Translated from original German into English by V.M. Bedekar. Delhi: Motilal Banarsidass.

  • Goodall, Dominic. 2004. The Parākhyatantra. A Scripture of the Śaiva Siddhānta. A Critical Edition and Annotated Translation. Collection Indologie, no. 98. Pondichéry: Institut Français de Pondichéry/École française d’ Extrême-Orient

    • Crossref
    • Export Citation
  • Halbfass, Wilhelm. 1976. “Zum Begriff der Substanz (dravya) in Vaiśeṣika.” Wiener Zeitschrift für die Kunde Südasiens 20: 141–166.

    • Crossref
    • Export Citation
  • Hulin, Michel. 1980. Mṛgendrāgama: sections de la doctrine et du yoga avec la Vṛtti de Bhaṭṭanārāyaṇakaṇṭha et la Dīpikā d’ Aghoraśivācārya. Traduction, Introduction et Notes. Collection Indologie, no. 63. Pondichéry: Institut français d’ Indologie.

  • Kano, Kyo. 2001. “Pariśeṣa, Prasaṅga, and Kevalavyatirekin—the Logical Structure of the Proof of Ātman.” Journal of Indian Philosophy 29: 405–422.

  • Matilal, Bimal Krishna. 1989. “Nyāya Critique of the Buddhist Doctrine of Non-soul.” Journal of Indian Philosophy 17 (1): 61–79.

  • Matilal, Bimal Krishna. 1994. “The Perception of the Self in Indian Tradition.” In Self as Person in Asian Theory and Practice, edited by R. Ames, et al., 279–295. New York: State University of New York Press.

  • Oetke, Claus. 1988. “Ichund das Ich. Alt- und Neu-Indische Studien 33. Stuttgart: Franz Steiner Verlag.

  • Preisendanz, Karin. 1994. Studien zu Nyāyasūtra III.1 mit dem Nyāyatattvāloka Vācaspati Miśras II. Alt- und Neu-Indische Studien, no. 46. Stuttgart: Franz Steiner Verlag.

  • Sanderson, Alexis. 1992. “The Doctrine of the Mālinīvijayottaratantra.” In Ritual and Speculation in Early Tantrism. Studies in Honour of André Padoux, edited by T. Goudriaan, 281–312. Albany: State University of New York Press.

  • Watson, Alex. 2006. The Self’s Awareness of Itself. Bhaṭṭa Rāmakaṇṭha’s Arguments Against the Buddhist Doctrine of No-Self. Publications of the De Nobili Research Library, no. 32. Wien: De Nobili Research Library.

  • Watson, Alex. 2010. “Rāmakaṇṭha’s Concept of Unchanging Cognition (nityajñāna): Influence from Buddhism, Sāṃkhya, Vedānta.” In From Vasubandhu to Caitanya. Studies in Indian Philosophy and Its Textual History. Papers of the 12th World Sanskrit Conference. Volume 10.1, edited by J. Bronkhorst and K. Preisendanz, 79–120. Delhi: Motilal Banarsidass.

  • Watson, Alex. 2014. “The Self as a Dynamic Constant. Rāmakaṇṭha’s Middle Ground Between a Naiyāyika Eternal Self-Substance and a Buddhist Stream of Consciousness-Moments.” Journal of Indian Philosophy 42 (1): 173–193. doi: 10.1007/s10781-013-9215-2.

  • Watson, Alex, Dominic Goodall and S.L.P. Anjaneya Sarma. 2013. An Enquiry into the Nature of Liberation. Bhaṭṭa Rāmakaṇṭha’s Paramokṣanirāsakārikāvṛtti, a commentary on Sadyojyotiḥ’s refutation of twenty conceptions of the liberated state (mokṣa). Collection Indologie, no. 122. Pondichéry: Institut Français de Pondichéry/École française d’ Extrême-Orient.

  • Collapse
  • Expand

Śaivism and the Tantric Traditions

Essays in Honour of Alexis G.J.S. Sanderson

Series:  Gonda Indological Studies, Volume: 22


All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 82 35 5
PDF Views & Downloads 88 37 4