Philosophical Reflections on the Sahopalambhaniyama

Introduction

In 1991 the Japanese scholar T. Iwata published a masterful study of the sahopalambhaniyama argument. In his work he carefully traces the development and interpretation of the argument throughout the entire Buddhist epistemological tradition, from Dignāga to Jñānaśrīmitra, and also surveys responses to the argument from a variety of Buddhist and non-Buddhist opponents. He is not, however, primarily concerned with evaluating the argument philosophically–though questions of logic inevitably arise when interpreting the argument–and, it would seem, not at all with assessing it as a piece of philosophical reasoning independently of its historical context. That is what I would like to do in this essay. In what follows I shall construct a version of the sahopalambhaniyama (henceforth SUN) argument and attempt to judge its strengths and weaknesses as an argument, in particular, its formal validity and the defensibility of what I shall represent as its premises. Although I shall indeed allow myself to be prompted by the statements of Dharmakīrti (as well as other classical authors), and will even at the beginning touch on the question how Dharmakīrti might have used the argument–what it was supposed to prove–I make no claim that the version of the SUN argument I will be considering was Dharmakīrti’s. Aside from the fact that other scholars are much more qualified to ascertain that than I,^[1] that is, tell us how the SUN argument as presented by Dharmakīrti should be understood, I wish to be relieved of the burden of historical accuracy in this essay in order better to focus on the logical and philosophical features of my reconstruction of it.

I believe it is worthwhile extracting the SUN argument from its historical and textual context and considering it in isolation just as an argument for at least two reasons. First, I believe that the version of the argument I shall devise is an argument for idealism unknown to Western philosophers. One could call it a ‘new’ argument for idealism if there were any traffic these days in arguments for idealism!^[2] Although it is not the case that idealism is a hotly contested view now and a ‘new’ argument might just push it over the top, so to speak–that is, bring it within the range of plausible or at least defensible philosophical positions–any argument for it that is not obviously fallacious is of inherent interest. For philosophers remain interested in the possibilities of reason. An argument that purports to prove so unexpected a thesis as that nothing can exist outside of consciousness represents, in the first instance, a bold claim about what philosophical reasoning can accomplish.

Second, although I would hesitate to say that I consider the argument a viable one, it does not, as far as I can tell, have any obvious flaws. It is valid, and its premises appear to be at least tenable. I believe it is a strong argument–which is not to say, of course, that I think it is conclusive.

I should say at the outset that I believe that the usual prejudice against idealism, the opinion that there could not possibly be any good arguments for it, is wholly unphilosophical. For what is it based on? Merely, as far as I can tell, the unshakeable conviction that we live in a world populated by mind-independent, physical objects. The very thrust of any argument for idealism, however, is to call this very conviction into question on the basis of common observations about the nature of our experience. Typically–we will consider an exception later–those who would see any argument for idealism as a non-starter, as already refuted by perception, are simply begging the question.

Perhaps one final remark is in order to reinforce the value of investigating the SUN philosophically as well as historically. Despite the interest in aspects of Indian philosophical thought that are contiguous with contemporary analytic philosophy, one cannot deny that many of the greatest Indian thinkers–such as Nāgārjuna, Dharmakīrti, and Śaṅkara–were attempting to demonstrate views about the nature of reality that are radically opposed to those of contemporary philosophers. In particular, they held that the empirical world is an illusion. As some scholars now seem to recognize, in order to give a balanced picture of Indian thought we should take Indian arguments for anti-realism just as seriously as arguments concerning topics in philosophy of logic, epistemology, and philosophy of language.^[3]

The SUN argument as an argument for idealism

Thus far I have referred to the SUN argument as an argument for idealism. But was it really meant to prove that? Here, I would like tentatively to explore this admittedly historical question–though still in a philosophical way–which will set up the examination of the argument itself. Iwata 1991 (henceforth “Iwata”) I, 18–20 suggests that Dharmakīrti, at least, used it to prove only dvirūpatā: the thesis that a cognition includes within itself two aspects, the object-form (arthākāra) and the subject-form (grāhakākāra, i.e., the cognition itself as the “grasper”), hence that the immediate object of experience is the form within cognition. Iwata cites as his main evidence the fact that, when presenting the SUN argument both in his Pramāṇavārttika (PV 3.387–397) and his Pramāṇaviniścaya (PVin 1.54–58), Dharmakīrti either directly states or implies that the conclusion of the argument–namely, that object or object-form and cognition are not different–holds even if there is an external object.^[4] Thus, evidently, he believed that the argument would be acceptable even to a Sautrāntika, who postulates something outside of cognition as the cause of the object-form. The SUN argument is thus intended to refute the bāhyārthavādin who is also a nirākāravādin, that is, in modern terminology, the ‘naive’ or ‘direct’ realist who holds that we directly perceive physical objects. It does not rule out the existence of an external object by itself.

Nevertheless, while some of Dharmakīrti’s successors followed him in viewing the argument as a proof of dvirūpatā, hence in the first instance as a refutation of the nirākā-rajñānavāda, others, Buddhists and Brahmins alike, saw it as a proof of vijñaptimātratā, mere-cognition: the thesis that there are no uncognized objects or that nothing exists outside of cognition. Iwata notes how the argument might have been seen by Dharmakīrti himself as practically amounting to a proof of mere-cognition. If what we are directly experiencing are only forms in cognition, that is to say, if our experience of a world that seemingly contains physical objects can be accounted for solely in terms of changes taking place within cognition itself, then it becomes superfluous to postulate physical objects at all. As Iwata puts it, “Insofar as the external object, in the sense in which it is postulated in Sautrāntika … is fundamentally imperceptible, that is, it can only be represented by means of inference, there is no longer any reason for it to be presupposed as the object of knowledge” (Iwata I, 6). Thus, the external object becomes “epistemologically completely superfluous” (Iwata I, 6). But perhaps that meant for Dharmakīrti that it does not exist at all? Although Dharmakīrti has moved beyond the direct refutation of the possibility of physical objects that Vasubandhu appears to undertake in his Viṃśikā –when arguing that the Buddha could not have literally meant there are such things when he referred to the āyatanas^[5] –he might still have been in sympathy with the overall strategy for proving vijñaptimātratā in that text, as Kellner/Taber 2014 have now explained it. Namely, there are no objects outside of cognition because there is no evidence for them. This strategy tacitly appeals to the epistemological principle, employed in Indian philosophy prior to Vasubandhu and still widely used for centuries after him (and after Dharmakīrti), that something exists if there is a pramāṇa for it, and something does not exist if there is not.^[6] Kellner (2017: 29–31) is quite right to emphasize (contra Ratié 2014) that the constraints of his logical theory would have prevented Dharmakīrti from devising a formal anumāna based on an anupalabdhihetu demonstrating that objects outside of cognition do not exist. Yet he surely must have known that when asking questions like, “If the cognition has the form of blue, etc., what is the pramāṇa for an external object?”^[7] –implying that in light of arguments like the SUN which undermine perception as evidence for external objects and the lack of conclusive inferential evidence, there is no reason to postulate them–he was coming as close as he could to asserting that they do not exist, and that is what most of his readers would probably take away.^[8]

Iwata generally presupposes that when arguing for the two-fold nature of cognition (dvirūpatā) Dharmakīrti is in dialogue with the Sautrāntika or Buddhist ‘externalist’ (as some prefer). There is good evidence for this. The framework of the initial discussion of dvirūpatā in both PV 3 and PVin 1 is the identification of self-awareness as the result (phala) of the pramāṇa perception, and Dharmakīrti wants to show that his solution is acceptable to both the Sautrāntika and the Yogācāra. If the immediate result of the SUN argument, however, is the demonstration of dvirūpatā, then the logical opponent against whom it is directed would be, as I’ve already suggested, a bāhyārthavādin who is also a nirākāravādin, rather than a Sautrāntika.^[9] Indeed, as is well known, much of the time in his works Dharmakīrti is arguing against a Mīmāṃsaka, an incorrigible direct realist who thinks that perception by itself establishes the existence of external objects. Especially in this case–if the SUN targets a direct realist–it would be tantamount to a proof of vijñaptimātratā. For the direct realist, unlike the representationalist (the Sautrāntika), does not think we can infer the existence of physical objects; we just see them. Once it has been shown to such a person, by proving dvirūpatā, that what he is seeing is really just the form in cognition, then what he thought were external objects are eliminated and he has nothing else, no inferred entities, with which to replace them.

Thus, the SUN argument, even when seen as a proof of dvirūpatā, could also be seen as amounting to a proof of vijñaptimātratā if supplemented by a further step–whether or not this was Dharmakīrti’s actual intention–namely, the claim that if there is no reason to believe in objects outside of cognition, then it is reasonable to assume that there are none. Or else it could be seen as amounting to a proof of vijñaptimātratā if one stipulates that it is meant to be deployed in a specific dialectical situation: for the direct realist, if what we are perceiving are not mind-independent objects, then there are none at all. On neither of these readings would the proof of mere-cognition be a rigorously deductive one. On the former, the conclusion that there are no uncognized objects is, in the final stage of the argument, in effect offered as the best explanation for the lack of any evidence for them. Thus, it is strictly something that could only be presumed, not asserted. On the latter, insofar as the argument is context-bound, it will not be persuasive for all interlocutors (for instance, non-Mīmāṃsaka representationalists).

What I would like to do, then, is develop a third reading of the SUN argument accord-ing to which it is rigorously deductive: that objects do not exist outside cognition will follow immediately from the conclusion of the argument, that object and cognition are not different. Some of Dharmakīrti’s opponents apparently perceived the argument in this way, as sufficient by itself for proving cognition-only.

Consider, for instance, Śaṅkara’s presentation of it in his Brahmasūtrabhāṣya:

Moreover, a cognition arises having mere experience as its common nature. The inclination (pakṣapāta) it has in regard to a particular object, so that it is the cognition of a pillar, the cognition of a wall, or the cognition of a pot–that is not possible without some difference that belongs to the cognition [itself], so that necessarily the cognition’s similarity with the object must be assumed. And that being assumed, since the object-form is obtained (avaruddha) just by the cognition, the postulation of an external object is pointless. Moreover, the non-difference of object and cognition results from their invariable co-apprehension; for it is not the case that there is the apprehension of the one when there is no apprehension of the other. And this is not possible if there is a difference of [their] nature, since there is no [other] cause [of their co-apprehension] consisting in a connection. For that [reason] as well, there is no [external] object.^[10]

It is clear from this passage that Śaṅkara understands the proof of the fact that cognition bears the form of the object, which is essentially dvirūpatā, as distinct from the proof of the non-difference of object and cognition from their invariable co-apprehension. From the cognition’s bearing the object-form within itself it follows only that an external object is ‘pointless’–that is, there is no reason to postulate one; but from the invariable co-apprehension of object and cognition it follows that there is no external object.^[11]

Therefore, one may even be historically justified in seeing the SUN argument as proving vijñaptimātratā directly, without any additional, non-deductive step or specification of the dialectical environment–hence as rigorously deductive. I shall now attempt to reconstruct the argument explicitly along such lines.

The strongest formulation of the argument

We may take as our point of departure Dharmakīrti’s classic statement of the argument at PVin 1.54ab:

Because they are invariably apprehended^[12] together, blue and the cognition of it are not different.

sahopalambhaniyamād abhedo nīlataddhiyoḥ.

I shall not be following Iwata’s method of examining the argument according to the terms of an Indian anumāna: What is the dharmin, the hetu,^[13] and the sādhya, and so forth? I shall rather be formulating the argument in the modern way, as a set of statements consisting of premises and a conclusion following from them that could be translated into predicate calculus if one wanted (though I shall not do that); for I believe that putting the argument in this form makes it more accessible to philosophers. Nevertheless, what is normally referred to as the subject of the argument (dharmin), that about which non-difference is being demonstrated, should be clarified at the outset. Normally, blue and the cognition of blue or, more generally, an object and the cognition of the object, are taken as the subject of the argument, and what is being proved about them by the argument is that they are not different.^[14] What is important to keep in mind when investigating the argument is that, in order for it to be able to establish vijñaptimātratā, the status of blue, or the object, must be undecided at the outset: it could be either not different from cognition or different from it. That is to say, it must be left undecided whether blue is something occurring inside or outside cognition. If one assumed at the outset that “blue” refers to an aspect of the cognition itself, that is, to the so-called object-form (arthākāra or grāhyākāra), then one would not prove anything very interesting or significant by demonstrating that it is not different from the cognition.^[15] In fact, one might even in that case object that, if the subject of the inference is interpreted in that way, and the purpose of the argument is to prove vijñaptimātratā, then the argument would be assuming what it is supposed to prove. If, on the other hand, the status of blue is left open, the argument immediately becomes much more interesting. For it then establishes (if it is successful!) that what one might very well have taken to be something external to cognition–a mind-independent object–turns out in fact to be something that is not different from it. The achievement of the argument would be, in effect, to bring everything that one is naively inclined to suppose exists outside of cognition inside cognition itself, and that would be a rather spectacular result indeed.

Our initial task in evaluating the SUN argument is to formulate it in the best possible way as a series of premises and a conclusion so as to avoid obvious prima facie objections.

Now some statements of the argument by Dharmakīrti himself suggest that what in Western logic would be taken as the minor premise of the argument should be formulated as follows:

A. Whenever there is an apprehension of an object, there is an apprehension of cognition and vice versa.

Thus, PV 3.387:

In what way can an object, which is invariably experienced at the same time together with cognition, be shown to be different from that?^[16]

Here, Dharmakīrti seems not to be talking about the cognition specifically of blue, as he seems to be in PVin 1.54ab, but of cognition in general. That is, he is saying: whatever object we happen to be aware of, that is not different from cognition.

Consider, however, a specific object, say blue, and the statement:

B. Whenever there is an apprehension of blue, there is an apprehension of cognition and vice versa.

This statement, especially the conjunct implied by the phrase “and vice versa,” is clearly false.^[17] We are conscious of many other things besides blue, namely, red, yellow, brown, sweet, bitter, rough, smooth, and so on. Awareness in general is not invariably accompanied by blue.

This point was very clearly stated by G. W. Moore in his famous essay “The Refutation of Idealism:”

We all know that the sensation of blue differs from the sensation of green. But it is plain that if both are sensations they also have some point in common. What is it that they have in common? And how is this common element related to the points in which they differ? I will call the common element ‘consciousness’ without yet attempting to say what the thing I so call is. We have then in every sensation two distinct terms, (1) ‘consciousness,’ in respect of what all sensations are alike; and (2) something else, in respect of which one sensation differs from another. It will be convenient if I may be allowed to call this second term the ‘object’ of a sensation: this also without attempting to say what I mean by the word. (Moore 1922: 17)

Precisely because the object in consciousness varies while consciousness remains constant, Moore goes on to say, we believe that consciousness is one thing, the object another.

In fact, we needn’t rely on a modern Western author to appreciate this point. It was already effectively stated by Śaṅkara in his critique of the SUN argument:

Moreover, when there is a cognition of a pot [then] a cognition of a cloth, there is a difference of the two qualifiers pot and cloth but not of the thing qualified, the cognition; just as when there is a white cow [then] a black cow, there is a difference of the whiteness and the blackness, but not of cowness. And a difference of the one from the two is [thereby] established, and of the two from the one. Hence, there is a difference of cognition and object.^[18]

Thus, with B as a counterexample–the statement has now been shown to be false–it seems, on the face of things, that A cannot be true.

On the other hand, it does seem plausible to suggest that blue and the cognition, specifically, of blue are invariably apprehended together, as the wording of PVin 1.54ab implies.

Now, if we take together all (presumably true) statements such as,

Blue is invariably apprehended together with the cognition of blue.

Red is invariably apprehended together with the cognition of red.

Etc.

Then one can derive the universal generalization:

C. For every object X and every cognition of the object Cx, one apprehends X if and only if one apprehends Cx,

which could also be stated as,

C^[19] . Any object and the awareness of that object are invariably apprehended together.

And C^[20] looks very much like A above.^[21] Dharmakīrti does not make this clarification himself, but it is not inconsistent with his statements. Indeed, when he says things like, “For even though they have separate appearances, there is no nature of blue that is a different thing from the experience,”^[22] (in his prose explanation of PVin 1.54ab), he could very well mean by “experience” (anubhava) that particular type of experience, namely, an experience of blue.

Thus, a general formulation of the minor premise of the SUN argument, with the terms “object” and “cognition,” seems appropriate after all, such as PV 3.389:

Neither an object without awareness nor awareness without an object is ever observed being experienced. Therefore, they are not distinct.^[23]

Taking C^[24] as the minor premise of the argument, then, the following suggests itself as the major premise:

D. Two things that are invariably apprehended together are not different.

Putting together D and C^[25] as the major and minor premises, we get the following formulation of the argument.

Premise 1 (major premise): Two things that are invariably apprehended together are not different.

Premise 2 (minor premise): Any object and the cognition of that object are invariably apprehended together.

Conclusion: Therefore, any object and the cognition of that object are not different.

I maintain that the following statements are obvious corollaries of the argument:

In general, object and awareness are in some sense the same thing.^[26] There is no object external to/without cognition.

(For an object cannot exist unless there is awareness of it; hence, esse est percipi.)

The world of objects that are experienced is “nothing but cognition.”

Thus, with the SUN argument formulated as suggested above, vijñaptimātratā follows from it immediately.

Examination of premise 1

The SUN argument as formulated above is a logically valid argument: the conclusion follows necessarily from the premises. To evaluate the argument further we must determine whether the premises are true.

Now, one of the classic objections raised against the SUN argument is that what is construed in Indian logic as the “reason” or hetu, namely, “because they are invariably apprehended together,” is a “contradictory reason” (viruddhahetu); that is to say, it actually proves the opposite of the desired conclusion.

For when we say that things are apprehended “together,” we obviously imply that they are different things. Thus,

I saw the ācārya walking together with his student,

and

I see the stars together in the sky,

prima facie suggest that one has to do, not with one thing, but with more than one thing.^[27] A related objection is that the reason “because they are invariably apprehended together” is inconclusive (anaikāntika); it does not reliably indicate that the things in question are not different. However, it does not mean that they are different, either. An inconclusive reason is not the same as a contradictory one. Thus, although the stars of a constellation are invariably seen together, they are different.^[28]

In my formulation of the argument these objections constitute challenges to the truth of the first, major premise–which expresses what in Indian logic is the “pervasion” (vyāpti) of the reason by the property-to-be-proved (sādhya). Thus, the question arises, How should one interpret the first premise so that it is not vulnerable to these sorts of objections? Here, there are at least two strategies one could follow. (1) One could stipulate a specific meaning of the word “together.” (2) One could stipulate a specific meaning of the word “invariably.” I shall discuss these suggestions in order.

Suggestion 1: The first strategy for defeating the sorts of objections raised above is to suggest that one should understand “together” correctly. It does not mean simply juxtaposed or simultaneous, but completely together, that is, inseparable like water and milk. Proposals quite similar to this were in fact made by different Buddhist authors in response to the objection that sahopalambhaniyamāt is a contradictory or inconclusive reason.^[29]

This proposal is not without problems, however.^[30] For one thing, it would seem to disqualify the example Dharmakīrti himself gives in support of his reason, “like two moons” seen by a person afflicted by an eye disease. For the two moons, presumably, can be clearly distinguished.^[31] Another problem with interpreting Premise 1 in this way, namely, as saying that two things that are invariably apprehended as inseparable are not different, is that it now sounds an awful lot like: two things that are invariably apprehended as not different are not different. Thus, the non-difference of entities would already be given with their invariable co-apprehension, so that by assuming, with Premise 2, the invariable co-apprehension of object and cognition, one is assuming what one is supposed to be proving–their non-difference. The argument becomes a petitio principii.^[32]

More generally, however, it is difficult to come up with examples of two things invariably seen together, as inseparable, that are the same thing.

Consider:

the Commander-in Chief of the U.S. Armed Forces and the President of the United States an Indian-Head Nickel and a Buffalo Nickel^[33]

[Figure: a duck-rabbit]

It seems that the latter two cases in particular, where it is a question of a thing with different visual aspects, are akin to the cognition of blue and blue. But are they seen together as inseparable? Indeed, both aspects are not seen at the same time–seeing the Indian head, one must turn the nickel over to see the buffalo and vice versa, and one can only shift from seeing the duck to seeing the rabbit and back–, so it is difficult to understand how they could be seen as inseparable or as one.

Thus, it becomes doubtful that Premise 1, even with a stronger construal of “together” as “inseparable,” would have any basis in observation–unlike the major premise of most syllogisms. We are confident in saying, “Where there is smoke, there is fire,” because we have observed many times that something that is smoking is on fire and have never observed that something that is smoking is not on fire, but that would not be the case here. Suggestion 2: The second strategy for removing obvious challenges to the truth of the first premise is to suggest that one has to understand “invariably,” which translates niyama, correctly.^[34] Namely, one should take it modally, as meaning necessarily.^[35] Thus, sahopalambhaniyamāt would mean “because they are necessarily apprehended together.”

One way of interpreting this would be that one cannot imagine apprehending one and not apprehending the other.

I am not aware of any Buddhist authors making this move in response to the objections of viruddhahetutva and anaikāntikahetutva.^[36] Dharmakīrti himself does not formulate the relation between the apprehension of an object and the apprehension of a cognition modally. Rather, he typically says things like, “For it is not the case that, when there is not the apprehension of one form, there is apprehension of the other” (PVin 1 40,2–3). He does not use modal language like “there could not be apprehension of the other.”

We do, however, find something like a modal formulation in the Tattvasaṅgraha:

That (X) whose awareness would necessarily (dhruvam) be the awareness of that (Y), is non-different from that (Y) or does not differ from that.^[37]

We also have the statement of Dharmottara:

Apprehension of the knowable [arises] only having the nature of the appre-hension of the cognition, not otherwise; and apprehension of the cognition has only the nature of the apprehension of the knowable.^[38]

Let us, then, replace Premise 1 with

Premise 1’: Two things that are necessarily apprehended together are not different.

Premise 1’ seems a priori justified. That is to say, the generalization expressed by Premise 1’–the vyāpti of the traditional SUN argument–is not based on observation but is a truth of reason, if you will.

In support of this claim I refer to a recent article by Galen Strawson, “What is the Relation Between an Experience, the Subject of the Experience, and the Content of the Experience?” from his collection Real Materialism.^[39] In this essay Strawson argues that for any occurrent experience e, e cannot exist without a subject of the experience s; the subject of the experience s cannot exist without an experiential content it is experiencing, c; and the experience e also cannot exist without an experiential content c and vice versa.

Thus we get

e ⇔ s ⇔ c,

where “x ⇔ y” means “if x then necessarily y, and if y then necessarily x.”

Strawson does not appeal to invariable co-apprehension to support the equivalences e ⇔ s ⇔ c; rather, he notes certain conceptual relationships. For instance, in arguing for e ⇔ s he writes, “There cannot be an experience without a subject of experience simply because experience is necessarily experience for –for someone-or-something. Experience involves experiential ‘what-it-is-likeness,’ and experiential what-it-is-likeness is necessarily what-it-is-likeness for someone-or-something.”^[40] In arguing for the equivalence e ⇔ c Strawson offers observations such as, “Evidently there can’t be concretely occurring experiential content without there being an experience of some sort which the content is the content of,”^[41] and, “For surely this very experience couldn’t have had a different content and still be the experience it is?”^[42] Thus, Strawson’s method for establishing equivalences for e, s, and c is different from that of the SUN argument–which of course is only concerned with the necessary equivalence e ⇔ c. Moreover, Strawson takes c to be the content of an experience, an “experiential content;” it is “internal,” “phenomenological.” Thus, it belongs to the experience at the outset. Its status is not initially undetermined, as merely the knowable that might or might not belong to the cognition, as it is on my interpretation of the SUN argument.

Nevertheless, Strawson’s reflections become particularly useful for us when he suggests that one can move from

e ⇔ s ⇔ c

to

e = s = c

by force of the following principle, which he borrows from Descartes:

There is a real distinction between two phenomena … if and only if they can possibly “exist apart,” and a merely conceptual distinction between them if and only if they are conceptually distinct, like trilaterality and triangularity, but cannot possibly exist apart.^[43]

Strawson (and Descartes) accept this as an a priori principle. Following them, then, it seems a priori justified–that is, not based on any observational data–that

Two things that are necessarily apprehended together are not different,

which is our Premise 1’. Of course, to say that two things are necessarily apprehended as co-occurring is slightly different from saying that two things necessarily co-occur, but it seems the difference is so slight as to be negligible. The only way to know that things are a certain way is by apprehending them as being so. I shall, however, return to this point below.

Before moving on to consider Premise 2, it may be helpful to clarify how Strawson is not proving the same thing as the SUN argument in the way I am interpreting it. Indeed, it would seem that by arguing for e = c Strawson is arguing for what the Buddhists referred to as dvirūpatā, essentially, the non-difference of the object-form from the cognition (cognition, that is, has two aspects: the cognition and the object), which I have distinguished from vijñaptimātratā above. Strawson of course is not interested at all in the thesis of idealism; however, he believes that e = s = c, hence e = c, is “a deep truth.”^[44] He believes it is particularly significant that s and c cannot exist apart, hence that they must be identical in some sense–the content of the experience is the subject. Why? The main reason appears to be that Strawson, as a physicalist or materialist, believes that all of e, s, and c are not just correlated with, but are physical entities: portions of “process-stuff” in the nervous system, which will eventually be pinpointed by neuroscience. The equivalence e = s = c, then, predicts that the neuroscientists will someday discover that the neurological process-stuff that is e ‘just is’ the neurological process-stuff that is independently identified as s, and that s ‘just is’ the neurological process-stuff that is independently identified as c.^[45] Here we have–though Strawson does not exactly say this–a priori philosophy proposing a specific research program in brain science! Be that as it may, in the end Strawson’s essay offers nothing for the Buddhist; for materialism is just as much a non-starter for the Buddhist–at least for Dharmakīrti–as idealism is for the contemporary analytic philosopher.^[46]

It seems that in light of this examination of Premise 1’ it is safe to say that the first premise of the argument, for now, looks quite solid. Thus, having established that the argument is valid, that is, commits no logical mistake, and that its first premise appears true, it would seem that the best hope for defeating it is to show somehow that the second premise is false.

Examination of premise 2

Is it true that object and cognition are necessarily apprehended together? Well, it seems uncontroversial that one cannot possibly apprehend a cognition of blue without apprehend-ing what it is a cognition of–blue. What seems, and was for Indian philosophers, much more controversial is the other half of the equivalence: that one could not apprehend blue without apprehending the cognition of it.

Dharmakīrti is clearly attempting to establish this with the second half of PVin 1.54:

The seeing of an object is not established for someone for whom the appre-hension is not evident.^[47]

The idea is that a cognition that does not reveal itself cannot reveal an object. Dharmakīrti continues,

For there is not an apprehending of an object due to the presence of the object, but rather due to the presence of the apprehension. And if that [presence] is not known by a pramāṇa it does not attach itself to/conform to [verbal, conceptual, and corporal forms of] behaviour that presuppose existence. If that [apprehension] were not established, there would be no establishing of the object either, so everything would disappear; for if it were not established, even though it exists, it cannot be treated as existing. Therefore, someone who does not apprehend the awareness of something is not aware of anything.^[48]

Dharmakīrti seems to be saying that apprehending something, but not being aware that one is apprehending it, is as good as not apprehending it at all.

Despite the strong prima facie appeal of this view it was aggressively attacked by opponents of Yogācāra anti-realism. One of the most forceful critiques is by Kumārila, who in the Śūnyavāda chapter of his Ślokavārttika is taking on an argument that anticipates the SUN, which was already known to the Vṛttikāra: if the form (ākāra) that appears in a cognition can be shown to belong to the cognition itself, then there is no reason to postulate an object.^[49] In arguing for the possession of the form by the cognition the pūrvapakṣin offers, among other reasons, that for something to illumine or manifest anything it has to be apprehended itself: “… the object which is to be illumed is not ascertained when the appearance of the cognition is not apprehended, because its illumination is dependent on that, like a pot when there is the light of a lamp.”^[50] The opponent even says that the cognition must reveal itself. We fail to be aware of something only because either a source of illumination is lacking or there is something obstructing it. When a cognition arises, however, neither is the case: nothing obstructs it, and it itself is the source of illumination.^[51] Moreover, if a cognition necessarily apprehends itself, it must apprehend itself as having some form. Thus, the ākāra must belong to the cognition.^[52]

Kumārila’s refutation of this argument in his siddhānta is extensive and complex. A better understanding of it awaits a new edition and translation of the Śūnyavāda. Never-theless, it is clear that it involves at least the following three elements. (1) In fact we are not aware of both object and cognition at the same time. Usually we are only aware of the object. “…When one cognizes the object (grāhya), the form blue and so forth, one does not at that time ever observe an awareness having the form of the subject (grāhaka).”^[53] And sometimes one is just aware of the subject without the object, for instance when one remembers experiencing something but cannot remember what it was.^[54] (2) It is not necessary for the cognition itself to be apprehended in order for it to cognize an object. The senses provide a counterexample: we apprehend things by means of them without apprehending them themselves.^[55] (3) There is no example of something that illumines itself as it illumines other things. Fire and so forth illumine other objects without their own natures being revealed. When they are apprehended, it is the sense that apprehends them.^[56] In general, Kumārila seems to subscribe to the principle that something cannot act on itself: an axe cannot cut itself, a finger cannot touch itself, and so on.

Obviously, we cannot adjudicate this dispute here–essentially, the debate about the truth of the second premise of the SUN argument–which continued for centuries after Kumārila and Dharmakīrti. We should note, however, that there is widespread support in modern philosophy, in both the analytic and continental traditions, for the thesis that some kind of self-awareness or reflexivity is essential to consciousness.

Consider the following précis of Sartre’s position by the contemporary phenomenologist Dan Zahavi.

Sartre, probably the best-known defender of a phenomenological theory of self-consciousness, considered consciousness to be essentially characterized by intentionality. He also claimed, however, that each intentional experience is characterized by self-consciousness. Thus, Sartre took self-consciousness to constitute a necessary condition for being conscious of something. To perceive a withering oak, a dance performance, or a red pillow consciously without being aware of it, that is, without having access to or being acquainted with the experience in question was, for Sartre, a manifest absurdity. This line of thought is elaborated in the important introduction to L’être et le néant, where he claimed that an ontological analysis of intentionality leads to self-consciousness since the mode of being of intentional consciousness is to be for-itself (pour soi), that is, self-conscious. An experience does not simply exist; it exists for itself, that is, it is given for itself, and this self-givenness is not simply a quality added to the experience, a mere varnish, but it rather constitutes the very mode of being of the experience. As Sartre wrote: “This self-consciousness we ought to consider not as a new consciousness, but as the only mode of existence which is possible for a consciousness of something.”^[57]

Thus, it seems obvious that Premise 2, if not actually true, is at least defensible–many philosophers of different traditions have held and defended it. The main point to be made here in favor of the SUN argument is this, that if one accepts Premise 2, as it appears many do, then, given that Premise 1 already looks quite strong, the argument will go through.

Further objections?

With the strategy of refuting the SUN argument decisively by contesting its premises not looking very hopeful, one is challenged to come up with other ways to attack the argument.

(1) Perhaps one can ask, Does the argument really establish vijñaptimātratā as its conclusion? Is the non-difference of object and cognition really tantamount to the idealist thesis that esse est percipi –an object cannot exist unless there is an awareness of it–as I maintained above?

Consider in this connection the one-liner the Vṛttikāra throws out at the end of his refutation of the Buddhist argument that cognition is ‘empty,’ that is, without any objective support, because the form (ākāra) we are aware of belongs to the cognition, hence object (artha) and cognition (buddhi) are one:

Moreover, if one likes, if there is sameness of nature [of the object and the cognition], then it is the cognition that doesn’t exist, not the object, which is [after all] perceptible.^[58]

That is to say, if one proves the non-difference of object and cognition, then what is the true nature of the (one) thing they both are? Is it the object or the cognition? It seems that there is no way to decide this question. Hence, it would be illegitimate to draw the idealist conclusion, that it is only the cognition that is real, and that the object, although it appears to be something distinct from the cognition, is really just the cognition!^[59]

The Vṛttikāra, in posing this question, seems to be presupposing that the object has some determinate nature that the cognition could also turn out to have; presumably, it is physical in nature. In doing so, however, he ignores the stipulation I made at the beginning of my exposition of the SUN argument–but which should hold for any argument for idealism–that the status of the object must be left undecided at the outset in regard to whether it is an aspect of the cognition or not. For, once again, if one assumes that it is merely the object-form, then the argument proves very little; but equally, one cannot assume that it is, say, a physical object that could exist unperceived–the very opposite of what the argument is supposed to prove. Given that the SUN argument establishes that object and cognition are not different, the independent existence of the object–as, say, something physical–is ruled out by the argument itself and the only remaining possibility is that it belongs to the cognition. In other words, were it shown (by some other argument) that the object does exist independently of the cognition, then one might conduct an investigation into its nature, whether for instance it is something physical. Having proven, on the other hand, that object and cognition are not different, and the nature of the cognition as something internal and phenomenal being beyond doubt, the object acquires that nature as well.

(2) A different approach–one historically attempted by certain Brahmin authors–is to claim that the very thesis (pratijñā) of the argument is contradicted by perception. This objection probably originated with Kumārila but it is very forcefully stated, once again, by Śaṅkara:

The non-existence of an external object certainly cannot be ascertained. Why? Because it is apprehended (upalabdheḥ).^[60] For, for each cognition an external object is apprehended, as [for instance] a post, a wall, a cloth, [or] a pot. And something that is apprehended cannot not exist. Just as if someone who is eating, experiencing the satisfaction produced by the food for himself, were to say, “I am not eating,” or “I am not satisfied,” so if someone apprehending for himself an external object by means of contact with his sense faculty were to say, “I do not apprehend it, and it does not exist”–how could we take his statements seriously?^[61]

In other words, perception itself reveals to us objects outside consciousness. The externality of the object, the fact that it exists independently of being perceived, is given as part of the content of perception itself. As Kumārila puts it, the idea that arises, upon perceiving something, that one’s cognition has an external support, is never overturned–unlike, say, dreaming cognitions.

Therefore, by implicit appeal to the principle of intrinsic validity that a cognition is true unless and until it is overturned by another, that idea is correct.

Moreover, the notion that a cognition has an external object is true, for it is a notion devoid of any contradiction, like the cognition [one has upon waking] that contradicts a dream.^[62]

This argument is more sophisticated than the mere prejudice against idealism that I dis-missed as question-begging at the beginning of this essay, for it is an actual argument that provides us with a reason for believing that our conviction that we are perceiving things outside us is true.^[63] Nevertheless–and very briefly–I do not think this is a fatal objection to the SUN argument. Even if our conviction in the externality and independence of the objects we experience is firm and never overturned, there is still no explanation for how such a conviction could be true. We would, somehow, have to be able to step outside our minds–adopt a “God’s-eye view”–and observe that our perceptions are really being caused by physical objects. Needless to say, that is impossible.^[64]

(3) Having exhausted two initially promising avenues of attack against the SUN argu-ment we are compelled to return to its premises and scrutinize them one last time. I do think that Premise 2, though not proven, is certainly defensible, and so I shall not attempt to investigate it further. There was, however, a point that came up at the end of my discussion of Premise 1’ that I did not pursue.

It certainly seems indisputable that if one thing can possibly be apprehended without the other, they are different. This is equivalent to the statement

If two things are not different, then they are necessarily apprehended together.

But is the converse, which is Premise 1’, true?

If two things are necessarily apprehended together, then they are not different.

I have offered in support of this principle the examples of the President of the United States and Commander-in-Chief of the U.S. Armed Forces and an Indian Head Nickel and a Buffalo Nickel, but ended up doubting that it could be established a posteriori.

I proposed that it be accepted as a priori justified instead, and appealed to the principle Galen Strawson (relying on Descartes) articulates in asserting the identity of experience, subject, and content of experience, namely (in one of its formulations):

If two things cannot possibly exist apart, so that there is only a conceptual distinction between them, then they are one.

The wording of this principle, however, is slightly different from Premise 1’ above. Premise 1’ refers to the apprehension of two things, whereas Strawson’s principle refers to the existence (or occurrence) of two things. Now, the question arises: Does this difference in wording constitute an important difference?

It seems that an opponent of the SUN argument could claim that it does. It may be the case that one never apprehends an object without apprehending the cognition that apprehends it, and that it is even inconceivable that one could. But why, exactly? It could indeed be because one can only apprehend an object by means of a cognition, and–as we have seen Sartre argue–all cognitions are self-aware, so that whenever one is aware of an object one is also aware of one’s cognition. Similarly, one can see an object by means of light, and light also illumines itself; when one sees an object one is aware of both the object and the illumination. Yet the object is different from the light that illumines it.

The word “apprehension” in Premise 1’, then, introduces a condition that guarantees that object and cognition will seem invariably to co-occur. Precisely because it qualifies the co-occurrence of two things as their apprehended co-occurrence, then, Premise 1’ does not rule out categorically that an object could occur without a cognition. Indeed, that is precisely the possibility that the realist thinks she envisions.

Again, it is Śaṅkara who states the problem about as succinctly as one could: “For this reason also,^[65] it is to be assumed that invariable co-apprehension, too, is caused by cognition and object being means and what is assisted (upāyopeyabhāvahetuka), not by their non-difference.”^[66] Here he is taking advantage of an opening left by Dharmakīrti himself. In his prose discussion of PVin 54ab Dharmakīrti clarifies that the invariable co-apprehension of object and cognition establishes the non-difference of their nature, “because there is no [other] cause [of their co-apprehension] consisting in a connection” (pratibandhakāraṇābhāvāt).^[67] In the case of visible form and light, on the other hand, there is a connection consisting of the object’s acquiring the capacity to produce a cognition of visible form, or else the arising of a capacity of the sense to cognize it.^[68] It is this connection, and not their non-difference, that accounts for their co-apprehension–the light enables the apprehension of visible form.

What is most interesting is that Dharmakīrti does not think that there is any other factor that determines the invariable co-apprehension of object and cognition other than their non-difference. The question is, Why not? Why couldn’t, as Śaṅkara suggests, their co-apprehension be due to the fact that the cognition is what enables us to apprehend the object?

I believe that this is the most serious objection that can be leveled against the SUN argument. Unfortunately, I do not have space to develop an adequate response to it, which would take us deep into difficult philosophical issues (especially in philosophy of mind)–and also, of necessity, back to the texts. For the Buddhists themselves, of course, were aware of this problem, which may have been first formulated by Śubhagupta: the invariable co-apprehension of object and cognition may be due, not to their non-difference, but to the fact that the preceding causal complex causes them to arise at the same time as grāhya and grāhaka, respectively.^[69] And Dharmakīrti already indicates the line of response later Buddhist epistemologists will take: “The other object, however, which arises simultaneously [with the cognition] does not shape [the cognition], because it does not influence it; how would [the object] be manifest, which the cognition [in turn] does not shape with its own form?”^[70] That is to say, simultaneously arising entities cannot stand in a causal relationship to each other, any more than, say, “the right and left horns of a cow.”^[71] But more generally, and independently of the historical development of the arguments, the question is: How can there be any causal relation between cognition and physical object at all? That is, How can consciousness illumine an object? Even if the (questionable) assumption that object and cognition arise simultaneously is abandoned, the realist has a lot of explaining to do! The defender of the SUN argument can defend the first premise simply by shifting the burden of proof: Explain to me how cognition, which is experienced as being non-material and non-spatial, can influence a physical thing to bring about the quality of its being manifest to us, or how could a physical object impose its form upon cognition?^[72]

In summary, the SUN argument appears to be a strong one. It is not invalid; there is evidence in support of its premises; and initial, and even secondary, objections to the premises can be answered. Even though one may not be convinced by it, it is not easy to refute.^[73]

References and abbreviations

Arnold 2008 D. Arnold, Buddhist Idealism, Epistemic and Otherwise: Thoughts on the Alternating Perspectives of Dharmakīrti. Sophia 47 (2008) 3–28.

BS Brahmasūtra. See BSBh.

BSBh Brahmasūtrabhāṣya of Śaṅkara: Brahmasūtra Śāṅkara Bhāṣya, with the Commen-taries Bhāmatī, Kalpataru and Parimala, ed. A. Śāstri. Varanasi 1982.

BSKā Bāhyārthasiddhikārikā of Śubhagupta, ed. N. A. Sastri. Bulletin of Tibetology 4/2 (1967) 1–96.

Chakrabarti 1990 A. Chakrabarti, On the Purported Inseparability of Blue and the Aware-ness of Blue: An Examination of sahopalambhaniyama. In: Mind Only School and Buddhist Logic, ed. D. Tulku. Delhi 1990, 17–36.

Descartes 1985 R. Descartes, The Philosophical Writings of Descartes, trans. J. Cotting-ham et al. 2 vols. Cambridge 1985.

Foster 2008 J. Foster, A World for Us: the Case for Phenomenalistic Idealism. Oxford 2008.

Frauwallner 1968 E. Frauwallner, Materialien zur ältesten Erkenntnislehre der Karma-mīmāṃsā. Vienna 1968.

Iwata 1991 T. Iwata, Sahopalambhaniyama: Struktur und Entwicklungen des Schlusses von der Tatsache, daß Erkenntnis und Gegenstand ausschließlich wahrgenommen werden, auf deren Nichtverschiedenheit. 2 vols. Stuttgart 1991.

Kellner 2011 B. Kellner, Self-Awareness (svasaṃvedana) and Infinite Regresses: A Com-parison of Arguments by Dignāga and Dharmakīrti. Journal of Indian Philosophy 39 (2011) 411–426.

Kellner 2017 B. Kellner, Proofs of Idealism in Buddhist Epistemology: Dharmakīrti’s Refutation of External Objects. In: Indian Epistemology and Metaphysics, ed. J. Tuske. London 2017, 102–128.

Kellner/Taber 2014 B. Kellner and J. Taber, Studies in Yogācāra-Vijñānavāda Idealism I: the Interpretation of Vasubandhu’s Viṃśikā. Asiatische Studien/Études Asiatiques 68/3 (2014) 709–756.

Matsumoto 1980 S. Matsumoto, Sahopalambha-niyama. Sōtōshū Kenkyūin Kenyūsei Kenkyū Kiyō 12 (1980) 298–265.

Mesquita 1990 R. Mesquita, Yāmunācāryas Philosophie der Erkenntnis: eine Studie zu seiner Saṃvitsiddhi. Vienna 1990.

Moore 1922 G. E. Moore, The Refutation of Idealism. In: Philosophical Studies. London 1922, 1–30.

NBhū Nyāyabhūṣaṇa of Bhāsarvajña: Nyāyabhūṣaṇam, ed. Yogīndrānanda. Varanasi 1968.

PV 3 Pramāṇavārttika of Dharmakīrti, Chapter Three, ed. H. Tosaki. Tōkyō 1985.

PVin 1 Dharmakīrti’s Pramāṇaviniścaya, Chapters 1 and 2, ed. E. Steinkellner. Beijing/ Vienna 2007.

Ratié 2010 I. Ratié, The Dreamer and the Yogin: On the relationship between Buddhist and Śaiva idealisms. Bulletin of School of Oriental and African Studies 73/3 (2010) 437–478.

Ratié 2011 I. Ratié, Can One Prove that Something Exists Beyond Consciousness? A Śaiva Criticism of the Sautrāntika Inference of External Objects. Journal of Indian Philosophy 39 (2011) 479–501.

Ratié 2014 I. Ratié, On the Distinction Between Epistemic and Metaphysical Buddhist Idealisms: A Śaiva Perspective. Journal of Indian Philosophy 42 (2014) 353–375.

Rickless 2013 S. Rickless, Berkeley’s Argument for Idealism. Oxford 2013.

Saccone 2018 M. S. Saccone, On the Nature of Things: A Buddhist Debate on Cognitions and Their Object. Vienna 2018.

Strawson 2008 G. Strawson, What is the Relation Between an Experience, the Subject of the Experience, and the Content of the Experience? In: Real Materialism, and Other Essays. Oxford 2008, 151–188.

ŚVM Ślokavārttika of Kumārila. In: Ślokavārtikavyākhyā Tātparyaṭīkā of Uṃveka Bhaṭṭa, ed. S. K. Ramanatha Sastri. Madras 1971.

ŚVT Ślokavārttika of Kumārila: Mīmāmsā Slokavārttika, with the Commentary Kasika of Sucaritamiśra, ed. K. Sāmbaśiva Sāstrī. Parts 1 and 2. Trivandrum 1990.

Taber 2003 J. Taber, Dharmakīrti Against Physicalism. Journal of Indian Philosophy 31 (2003) 479–503.

Taber 2010 J. Taber, Kumārila’s Buddhist. Journal of Indian Philosophy 38 (2010) 279–298.

TS(P) Tattvasaṅgraha of Śāntarakṣita and Tattvasaṅgrahapañjikā of Kamalaśīla: Tattvasaṅ-graha of Ācārya Shāntarakṣita, with the Commentary Pañjikā of Shri Kamalashīla, ed. D. Shastri. Varanasi 1982.

Westerhoff 2009 J. Westerhoff, Nāgārjuna’s Madhyamaka: a Philosophical Introduction. Oxford 2009.

Westerhoff 2010 J. Westerhoff, Twelve Examples of Illusion. Oxford 2010.

Zahavi 2005 D. Zahavi, Subjectivity and Selfhood: Investigating the First-Person Perspec-tive. Cambridge, Mass. 2005.

Footnotes and references:

[1]:

In addition to Iwata’s study see Arnold 2008, Ratié 2014, and Kellner 2017. Kellner’s study is, to date, the most penetrating.

[2]:

I know of only two defenders of idealism among contemporary philosophers: S. Rickless (2013) and J. Foster (2008).

[3]:

I have in mind J. Westerhoff in particular. See, e.g., Westerhoff 2009 and 2010.

[4]:

PVin 1 43,4–5: saty api bāhye’rthe sahopalambhavedanābhyāṃ bhāsamānasya nīlades tatsaṃvidaś cāvivekaḥ siddhaḥ. In PV 3.387–397 this is implied when Dharmakīrti, after developing the argument in vv. 387–390c, concedes apparently to the Sautrāntika that one could infer an external object as the cause of the arising of a perceptual cognition were it not possible to explain it as the effect of the immediately preceding cognition (390d–391).

[5]:

See Arnold 2008. Although Manorathanandin, in one passage, recommends that one who demands a proof that there is no external object even after it has been shown that the object is never experienced independently of cognition should consult Vasubandhu’s mereological arguments, I don’t think one can conclude that Dharmakīrti, who nowhere mentions them, endorsed them too. See Ratié 2014: 358–362.

[6]:

Or, something does not exist anupalabdheḥ, “because it is not apprehended.” See Kellner/Taber 2014:

[7]:

dhiyo nīlādirūpatve bāhyo’rthaḥ kiṃpramāṇakaḥ / PV 3.432ab. Or PV 3.333: “[Objection:] If an external object were experienced, what mistake would there be? [Reply:] None at all! Only this [remains to be asked:] Why would it be said that an external object is perceived?” Kellner 2017: 24.

[8]:

Thus, Ratié 2014 shows that Abhinavagupta understood the SUN in this way while adapting it to his own purposes: “…For the Śaivas the epistemic argument is not confined to a demonstration that there is no argument proving the existence of an external object; or rather, the very demonstration that there is no sādhaka whatsoever in favour of the external object constitutes a bādhaka, an argument refuting its existence” (364). Kumārila seems to understand the strategy of the argument he is refuting similarly. See ŚV, Śūnyavāda 7–9. If the ākāra we perceive belongs to the cognition, “then there is not the supposition of an [external] object” (8d). In the context of the problematic of the Śūnyavāda that means that it does not exist. See Taber 2010: 283.

[9]:

Śāntarakṣita and Kamalaśīla would agree with this assessment, as Iwata I, 25–27 interprets them.

[10]:

BSBh 544,1–6: api cānubhavamātreṇa sādhāraṇātmano jñānasya yo’yaṃ prativiṣayaṃ pakṣapātaḥ stambhajñānaṃ kuḍyajñānaṃ paṭajñānam iti nāsau jñānagataviśeṣam antareṇopapapdyata ity avaśyaṃ viṣayasārūpyaṃ jñānasyāṅgīkartavyam. aṅgīkṛte ca tasmin viṣayākārasya jñānenaivāvaruddhatvād apārthikā bāhyārthasadbhāvakalpanā. api ca sahopalambhaniyamād abhedo viṣayavijñānayor āpatati. na hy anayor ekasyānupalambhe’nyasyopalambho’sti. na caitat svabhāvaviveke yuktaṃ pratibandha-kāraṇābhāvāt. tasmād apy arthābhāvaḥ. The first sentence of this passage closely follows PV 3.302; the penultimate sentence almost verbatim corresponds to PVin 1 40,2–4. Thanks to B. Kellner for pointing this out.

[11]:

But the final sentence of the passage, “For that [reason] as well (api), there is no external object,” may suggest that Śaṅkara also thought that showing it is ‘pointless’ amounted to proving that it does not exist. Such a proof, however, would involve an additional step. To say that something is pointless is obviously not the same as saying it is nonexistent.

[12]:

Although upalabdhi often means specifically perception, it also has the broader meaning of an apprehension by any pramāṇa, which I prefer to follow in my treatment because I think it is more suitable for what I shall reconstruct as the first or major premise of the argument.

[13]:

He also asks, more specifically of the hetu, sahopalambhaniyamāt, what is the meaning of saha and what is the meaning of niyama.

[14]:

See Iwata I, 41–44.

[15]:

Though Galen Strawson might disagree. See below. Precisely this point is made by Śubhagupta BSKā 87.

[16]:

sakṛt saṃvedyamānasya niyamena dhiyā saha / viṣayasya tato’nyatvaṃ kenākārareṇa sidhyati // Unlike for Pramāṇaviniścaya 1, we do not have a critical edition of Pramāṇavārttika 3. I’ve used the edition of Tosaki.

[17]:

That is to say, B is really the conjunction, “Whenever there is an apprehension of blue there is an apprehension of cognition, and whenever there is an apprehension of cognition there is an apprehension of blue.”

[18]:

BSBh 550,1–3.

[19]:

In addition to Iwata’s study see Arnold 2008, Ratié 2014, and Kellner 2017. Kellner’s study is, to date, the most penetrating.

[20]:

In addition to Iwata’s study see Arnold 2008, Ratié 2014, and Kellner 2017. Kellner’s study is, to date, the most penetrating.

[21]:

Chakrabarti 1990: 32–33 comes up with virtually the same solution to this problem.

[22]:

na hi bhinnāvabhāsitve’py arthāntaram eva rūpaṃ nīlasyānubhavāt…, PVin 1 40,1. Cf. Iwata I, 110–111.

[23]:

nārtho’saṃvedanaḥ kaścit anarthaṃ vāpi vedanam / dṛṣṭaṃ saṃvedyamānaṃ tat tayor nāsti vivekitā //

[24]:

In addition to Iwata’s study see Arnold 2008, Ratié 2014, and Kellner 2017. Kellner’s study is, to date, the most penetrating.

[25]:

In addition to Iwata’s study see Arnold 2008, Ratié 2014, and Kellner 2017. Kellner’s study is, to date, the most penetrating.

[26]:

For the purposes of my reconstruction of the SUN argument it is not necessary to specify whether “not different” means merely not different or, taken as “non-different,” implies sameness. As far as I can tell, either way the argument would go through and would amount to a proof of idealism. For Buddhist defenders of the argument the debate about the precise meaning of abheda was driven by other doctrinal issues, especially the problems of the unity and reality of the object-form. See Iwata I, 110–215.

[27]:

See Iwata I, 93–94, referencing Jaina authors, e.g., Prabhācandra, and others (Iwata II, 84, n. 95). The objection may have been raised first by the Vaibhāṣika (?) Śubhagupta; see BSKā 71, cited by Kamalaśīla TSP 692,2–3. The example of the teacher and his student is from Yāmuna’s Ātmasiddhi; see Mesquita 1990: 38. I regret that Saccone 2018 was not yet available for me to take into consideration in this essay; to do so would have required substantial revisions.

[28]:

See Iwata II, 84, n. 98; BSKā 68. The anaikāntikatā objection is generally thought to apply if saha is understood specifically to mean “at the same time.”

[29]:

Devendrabuddhi and Śākyabuddhi, for instance, interpret saha as eka, so that sahopalambha = ekopalambha or abhedopalambha, an apprehension as united, or an apprehension as not different (Iwata I, 67–71). Śāntarakṣita, Kamalaśīla, and Dharmottara interpret sahopalambha as eka evopalam-bhaḥ, a single apprehension [of both]. (See Iwata I, 84–88 on Dharmottara and 88–89 on Kamalaśīla. Cf. Matsumoto 1980: 284–283. See n. 33 below.) For a schematic presentation of the different options for construing saha see Iwata I, 66–67, and the entire section, 66–103, for a comprehensive discussion.

[30]:

Bhāsarvajña neatly refutes most of the Buddhist maneuvers to evade these objections, NBhū 132–135.

[31]:

Dharmakīrti himself says, PVin 1 40,1–2, that blue and its experience have separate appearances “like two moons, etc.:” na hi bhinnāvabhāsitve’py arthāntaram eva rūpaṃ nīlasyānubhavāt tayoḥ sahopalambhaniyamād dvicandrādivat. I am not aware of this criticism being raised by a classical author, but the literature is vast.

[32]:

This is essentially the objection, raised by some classical opponents against the interpretation of sahopalambhanaḥ as ekatvenopalambhanaḥ, that in that case the hetu would not be different from the sādhya, resulting in siddhasādhanatā, “proof of what is already established.” See Iwata I, 94; BSKā 76.

[33]:

An American Indian-Head nickel was a five-cent coin minted in the United States from 1913 to 1933, now a collector’s item. By definition, the nickel shows an Indian on the ‘heads’ side and a buffalo on the ‘tails’ side.

[34]:

Iwata translates niyama as “ausschließlich.”

[35]:

Mesquita 1990 for instance translates it as “notwendigerweise,” Matsumoto 1980 as “necessarily.”

[36]:

Iwata’s discussion of niyama is relatively short. It would seem that the interpretation of niyama was not a matter of major concern in the debate over the SUN argument.

[37]:

TS 2029: yatsaṃvedanam eva syād yasya saṃvedanaṃ dhruvam / tasmād avyatiriktaṃ tat tato vā na vibhidyate // See Matsuoka, this volume, for a different translation. This verse, however, represents the view that sahopalambha implies that there is just one apprehension of X and Y: eka evopalambhaḥ. See TSP 692,23–24: na hy atraikenaivopalambha ekopalambha ity ayam artho’bhipretaḥ. kiṃ tarhi? jñānajñeyayoḥ parasparam eka evopalambhaḥ, na pṛthag iti. This is different from what I am suggesting, and I think is attended by the sorts of problems brought out by Bhāsarvajña. See above, nn. 25 and 26.

[38]:

jñeyopalambho jñānopalambhātmaka eva nānyathā / jñānopalambho’pi jñeyopalambhātmaka eva //, from Dharmottara’s Pramāṇaviniścayaṭīkā. See Iwata I, 107–108 and II, 89, n. 15.

[39]:

Strawson 2008. Thanks to Mark Siderits for bringing this article to my attention.

[40]:

Strawson 2008: 153. It should be noted that the type of subject Strawson conceives of as necessary for experience is a ‘thin’ subject, which is not ontologically distinct from the experience but only its subjective pole and which is different from one experience to the next. Thus, it is not incompatible with a Buddhist view of consciousness which involves no self distinct from cognition but includes a subjective factor, namely, the grāhakākāra or subject form. See esp. Strawson 2008: 182–183, also 167–168.

[41]:

Strawson 2008: 153.

[42]:

Strawson 2008: 177.

[43]:

Strawson 2008: 168. Strawson cites Descartes 1985: I, 213–215 (Principles of Philosophy). See also Strawson 2008: 181: “…Let me now formally endorse the principle that if there is at most a conceptual distinction between two apparently distinct (concrete) particulars, if they cannot possible exist apart, then they are not really two but only one…”

[44]:

Strawson 2008: 180.

[45]:

Strawson 2008: 183–185.

[46]:

See Taber 2003.

[47]:

apratyakṣopalambhasya nārthadṛṣṭiḥ prasidhyati // Iwata I, 106.

[48]:

PVin 1 41,1–5. na hi viṣayasattayā viṣayopalambhaḥ kiṃ tarhi tadupalambhasattayā. sā cāpramāṇikā na sattānibandhanān vyavahārān anuruṇaddhi. tadaprasiddhau viṣayasyāpy aprasiddhir ity astaṅga-taṃ viśvaṃ syāt. sato’py asiddhau sattāvyavahārāyogyatvāt. tasmān nānupalabhamānaḥ kasyacit saṃvedanaṃ vedayate nāma kiñcit. My translation follows Kellner 2011: 420 in part.

[49]:

See Taber 2010.

[50]:

ŚV, Śūnyavāda 22.

[51]:

ŚV Śūnyavāda 23–24.

[52]:

Taber 2010: 284–287.

[53]:

ŚV Śūnyavāda 74.

[54]:

ŚV Śūnyavāda 82cd–83.

[55]:

ŚV Śūnyavāda 179cd–180ab.

[56]:

ŚV Śūnyavāda 64–66.

[57]:

Zahavi 2005: 12. Note Zahavi’s claim about the ubiquitousness of this view: “One should not over-estimate the homogeneity of the phenomenological tradition; like any other tradition, it spans many differences. Although phenomenologists might disagree on important questions concerning method and focus, and even about the status and existence of a self, they are in nearly unanimous agreement when it comes to the relation between consciousness and self-consciousness. Literally all the major figures in phenomenology defend the view that the experiential dimension is characterized by a tacit self-consciousness” (11). Kellner 2017: 21, interestingly enough, also refers to Sartre’s view of con-sciousness when discussing the so-called saṃvedana argument and cites Zahavi’s treatment of it, but she warns that the conclusion Dharmakīrti draws, that cognition cannot be of an external object, “is one that phenomenologists shun.” My point is simply that the SUN argument would force them to accept it.

[58]:

Frauwallner 1968: 30,10–11: api ca kāmam ekarūpatve buddher evābhāvo, na tv arthasya pratyakṣasya sataḥ.

[59]:

Ernst Steinkellner presented this objection to me after I delivered a preliminary version of this paper at the Fifth International Dharmakīrti Conference, in Heidelberg, in August, 2015. He did not refer to the passage in the Śābarabhāṣya, and I suspect he came up with it on his own. He went on to say, with a mischievous twinkle in his eye, that whenever he tries to think about the philosophical problems the Buddhist epistemologists are concerned with he finds himself faced with similar conundrums–and that is why he prefers not to think about them. The Buddha, perhaps, would have said the same thing. It was in order to avoid a similar problem that some Buddhist interpreters of the argument, e.g., Dharmottara, suggested that “non-difference” (abheda) should be understood to connote just the negation of difference and not also the affirmation of identity, that is, as a prasajyapratiṣedha rather than a paryudāsa. Otherwise, if cognition and object were the same, then the object-form being unreal (as some believed), the cognition would become unreal. See Matsumoto 1980: 281–278; Iwata I, 173–183.

[60]:

This glosses BS 2.2.28: nābhāva upalabdheḥ. This by itself suggests that the Brahmanical understanding of the purport of Buddhist arguments against the existence of external objects was that they do not exist because they are not apprehended (anupalabdheḥ). They understood the Buddhist arguments against the existence of a self in the same way. See Kellner/Taber 2014: 726–727; 731–732.

[61]:

katham upādeyavacanaḥ syāt. BSBh 547,4–548,4. See the continuation of the passage. Śaṅkara points out, among other things, that when the Buddhist says the object-form appears as if it were an external object, he tacitly admits that perception itself reveals to us that the objects we experience are external.

[62]:

bāhyārthālambanā buddhir iti saṃyak ca dhīr iyam // bādhakāpetabuddhitvād yathā svapnādibādhadhīḥ / ŚVT Nirālambanavāda 79cd–80ab. ŚVM reads svapnādibāhyadhīḥ in 80b.

[63]:

Kumārila, ŚV Nirālambanavāda 79ab, introduces this as a counter-proof (pratisādhana) that renders the reason or thesis of the idealist’s anumāna defective (he is considering a different one in the Nirāla-mbanavāda). The reason would become, technically, a viruddhāvyabhicārihetu, or else the argument’s thesis would be disproven by another inference (78cd).

[64]:

Ratié 2010: 444–445, n. 23 and 2011: 484–486 shows how Utpaladeva and Abhinavagupta, and apparently certain Vijñānavādins, used this type of argument to refute the Sautrāntika: we cannot infer the existence of external objects as the cause of our perceptual experiences, because we cannot perceive them independently of cognizing them, in order to establish a causal relationship between them and our experiences. As Abhinavagupta puts it, “…Since the external object is not an object of perception, one cannot establish any causal relation between this [external object] and anything else; as a consequence, the cognition [of this or that object] is not a kāryahetu in regard to this [object] to be inferred” (Ratié 2011: 485, slightly adjusted).

[65]:

The argument that precedes this was that the fact that the cognition is similar in form to the object does not eliminate the object; to the contrary, it presupposes it, moreover the object is perceived externally (bahirupalabdheś ca viṣayasya) (BSBh 549,4–5).

[66]:

ata eva sahopalambhaniyamo’pi pratyayaviṣayayor upāyopeyabhāvahetuko nābhedahetuka ity abhyupagantavyam, BSBh 549,6–550,1.

[67]:

Iwata I, 221: “denn es gibt keine in einer Verbindung bestehende Ursache [für die Ausschließlichkeit des Zusammenwahrgenommenwerdens].”

[68]:

PVin 1 40,3–7. See Iwata I, 221–226. His presentation is rather dense and I am not sure how he understands the passage.

[69]:

See BSKā 66; Kamalaśīla presents Śubhagupta’s position at TSP 694,9–20 (ad TS 2031), citing BSKā 66 and 81.

[70]:

nākārayati cānyo’rtho’nupakārāt sahoditaḥ / vyakto’nākārayañ jñānaṃ svākāreṇa kathaṃ bhavet // PV 3.420. Cf. Iwata I, 231.

[71]:

Cf. TSP 695,4–5: nāpi tadutpatteḥ sahavedanam, sahabhūtayoḥ kāryakāraṇābhāvāt; Iwata I, 228–233. The analogy is from Nyāyakumudacandra, cited by Iwata I, 232.

[72]:

This is essentially what Jñānaśrīmitra does in his Advaitabinduprakaraṇa. His line of investigation strikes one as an elaboration of the thought behind what is known as the saṃvedana argument: awareness is by nature just the “appearing in a certain way,” and like the awareness of awareness, can only be of awareness. See Kellner 2017: 20–22.

[73]:

I believe that Berkeley’s arguments for idealism, for instance, are easier to refute.