Chapter 5. The debate between Consciousness-Only and Madhyamaka¶

Essays 5 and 6 explained Madhyamaka and Consciousness-Only respectively from within their own philosophical systems. This chapter brings them together and presents their debate directly. It is the longest-lasting and philosophically deepest internal disagreement in the history of Indian Mahāyāna philosophy. At its center is a question still disputed today: after a thoroughgoing analysis of emptiness, can "the dependently arisen flow of consciousness" still be granted a reality different from "the fiction of the imagined nature"?

1. The origins of the debate¶

Madhyamaka and Consciousness-Only share a common source. Both are Mahāyāna, accept Prajñāpāramitā emptiness, and criticize Abhidharma's realism of intrinsic nature. In this sense, they are not wholly alien opponents. They start from the same ground and take different paths.

The point of divergence was identified in 6.1, Section 1. After saying "everything is empty," Madhyamaka largely stops at negative statements. Consciousness-Only considers this stopping insufficient. Emptiness dismantles the old ontology, but does not tell practitioners how experience actually occurs, how consciousness operates, or how afflictions accumulate and are purified. Consciousness-Only intervenes precisely to supply this positive theory of cognition and practice on the premise of emptiness.

From Bhāvaviveka onward, the Madhyamaka response has consistently been that Consciousness-Only quietly introduces a new realism in supplying this theory. It substitutes consciousness for external objects, moving real existence from one side to the other without curing the root ailment of attachment to intrinsic nature.

The debate between the schools can be arranged along an approximate timeline:

The Chinese diagram distinguishes two strands: the Consciousness-Only (Yogācāra) tradition and the Madhyamaka tradition. The two converge in the synthesis associated here with Śāntarakṣita and continued by Kamalaśīla.

2. Bhāvaviveka's refutation and the two levels of dispute¶

The first Madhyamaka scholar to undertake a systematic refutation of Consciousness-Only was in fact Bhāvaviveka (c. sixth century), not Candrakīrti. In the Lamp of Wisdom (Prajñāpradīpa) and Heart of Madhyamaka (Madhyamakahṛdaya), he directly criticized the Consciousness-Only position of Asaṅga's lineage. This is the earliest documented direct exchange between the schools.¹ The Mahāyāna Treatise on the Jewel in the Hand, translated by Xuanzang, preserves a fuller exchange for comparison. Each of its two fascicles records a passage introduced by "Some Yogācāra scholars say as follows," first reproducing a lengthy argument from the Yogācāra side and then answering it sentence by sentence. The words of the Consciousness-Only side are thus preserved in a book from the Madhyamaka side.²

The two fascicles do not dispute the same thing. This distinction must be established first. The first disputes the "existence" of the dependent nature; the second disputes the "real existence" of suchness. Consciousness-Only defends different things in the two passages: a functional position in the first, a real existent in the second.

First fascicle: the dispute over the dependent. The Consciousness-Only argument proceeds in five steps. First, it incorporates Madhyamaka's position, saying that "what arises through conditions lacks intrinsic nature and is therefore empty" precisely accords with its own teaching: "This thus affirms the meaning of the Yogācāra scholars and accords with right reasoning." Second, it establishes a formula of remainder: "That of which it is empty really does not exist; this on the basis of which it is empty really does exist." Third, it identifies "that" and "this" with the imagined and dependent natures. Fourth, it states the fear of annihilation: if what bears emptiness is itself emptied away, emptiness has nowhere to rest. "If this does not exist, that is annihilation." Fifth, it concludes with a warning: to say that the dependent too is nonexistent is to fall into a deep pit of error.

Bhāvaviveka replies in two ways. One is a dilemma. If the dependent is posited as conventionally existent, "what you establish is already established": our side has already accepted it. If it is posited as ultimately existent, "there is no example possessing the same property." At the ultimate level, no mutually accepted existent can serve as the basis of an example, so an inference cannot be established. The other reply is positive acknowledgment: "If in this sense you say that the intrinsic nature of the dependent exists, that is well said, for I too allow such an intrinsic nature." Madhyamaka already grants intrinsic nature at the conventional level. The fear of annihilation therefore fails: once conventional intrinsic nature is allowed, provisionally designated dharmas have a basis, and nothing has been emptied away.

The interpretation of this level becomes clear. The dispute in the first fascicle is largely a confusion of levels. What Consciousness-Only is really defending is the functional position of "a place for emptiness to rest." Bhāvaviveka fully grants this function, conventional existence, and a basis for provisional dharmas. Consciousness-Only still refuses, fearing that conventional "existence" is not firm enough. But firmness is a question of levels, not of existence versus nonexistence. The two sides' uses of "existence" do not fall at the same level, and the dispute is correspondingly narrowed.

Second fascicle: the dispute over suchness. This level carries a different weight. Consciousness-Only first makes a concession, then defends one point: "There is no further ultimate above the ultimate; suchness is the ultimate of all dharmas. Thus, to say in ultimate terms that suchness is empty accords with reason, but to say that suchness is not really existent does not accord with reason." It concedes that "suchness can be called empty," but defends the inadmissibility of "suchness is not really existent." Bhāvaviveka answers in four ways: two inferences, such as "That suchness is not the true ultimate, because it is an object of cognition, like visible form and the like"; four scriptural proofs, of the kind "the eye of wisdom sees nothing at all"; one reductio (shepherds too know that there is no silk upon silk, so all shepherds would become seers of truth); and, finally, the objection of the non-Buddhist self.

The objection of the non-Buddhist self is the strongest blow in this passage of the second fascicle. Bhāvaviveka's words are: "If you say that suchness is beyond language yet really existent, then it is the non-Buddhist self, called 'suchness' through a difference of name and conception." The self posited by non-Buddhists is likewise said to be really existent, all-pervading, permanent, and free from conceptual discrimination. It too is said to lie beyond the reach of language and beyond the objects of discriminating understanding. Bhāvaviveka therefore judges that no difference between them can be stated. His own positive position follows: suchness "is only the permanent cessation of all conceptual discrimination, not a really existent nature, neither apart nor nonexistent."²

This objection targets the form "beyond language yet really existent." Ineffability grants no exemption. To call something free from all conceptual discrimination while still really existent merely moves the commitment outside language; it does not remove the commitment itself. This is the only point at which Bhāvaviveka and Consciousness-Only directly contradict each other at the same level.

The status of the reproduced testimony must also be identified. The passage in the first fascicle accords with the wording of the Cheng weishi lun and can be used as testimony both sides accept. The passage in the second is reproduced by an opponent and is out of step with such relational expressions as "neither different nor nondifferent" and "neither apart nor nonexistent" assigned to the perfected in the base treatise. It is recorded only as evidence that "this reading does occur within the Consciousness-Only lineage," not as the school's settled doctrine.

Bhāvaviveka makes a further methodological criticism. Asaṅga's arguments against defenders of external objects lack rigor and do not follow the three-part form of Dignāga's logic. His supposed "refutation of external objects" merely supplies an alternative, without formally proving that external objects do not exist. This agrees with this book's criticism of the argument against external objects in 6.3, Section 4.

Bhāvaviveka himself adopts a Svātantrika position on "conventional particular characteristics," allowing "nominal particular characteristics" at the conventional level. He therefore does not completely reject a certain limited reality. He insists only that Consciousness-Only should not establish it on the side of consciousness. There is thus a subtle symmetry in their debate. Both accept some limited reality conventionally, but Consciousness-Only locates its basis in consciousness, while Bhāvaviveka locates it in the nominal particular characteristics of dharmas. Candrakīrti considers both concessions excessive: even the opening for nominal particular characteristics should never have been made.

3. Candrakīrti's systematic attack¶

The refutation of Consciousness-Only by Candrakīrti (c. seventh century) was presented in detail in 5.5, Section 4. Here we collect its principal points and add what 5.5 did not develop.

Candrakīrti makes three attacks (see 5.5). First, self-cognition (svasaṃvedana) has an untenable logical structure: a knife cannot cut itself, and consciousness cannot know itself. Second, consciousness's "establishment by its own characteristic" stands in exactly the same position before the analysis of emptiness as that of external objects. Consciousness has no exemption unavailable to external objects. Third, Consciousness-Only explains everyday common sense more circuitously than Madhyamaka, without superior explanatory power.

Beyond these three attacks, Candrakīrti has a more fundamental criticism that deserves separate explanation here.

Candrakīrti's fundamental criticism: dependently arisen consciousness is still a real existent

Consciousness-Only connects with emptiness through the three natures and the three absences of intrinsic nature. It says that "the dependent is not the imagined, and the dependent itself also lacks intrinsic nature: absence of intrinsic nature in arising." Candrakīrti considers the connection superficial. The crucial point is that Consciousness-Only always insists that "the dependent exists," expressed as "false discrimination exists." Although it arises in dependence on others and lacks intrinsic nature, it exists in some way.

For Candrakīrti, this "existing in some way" is already the problem. Prāsaṅgika Madhyamaka holds that even "dependently arisen existence," if it has a basis of reality beyond nominal convention, leaves consciousness with a substantial anchor to some degree, however thin and however dependent that reality may be. Candrakīrti defines the legitimacy of the conventional through "worldly consensus," without needing "the existence of the dependent itself" as its foundation. Consciousness-Only, by contrast, needs the dependent as the basis for the manifestation of the perfected, suchness. This function of "serving as a basis" is precisely what Candrakīrti thinks Consciousness-Only has not thoroughly emptied.³

Put more plainly, Candrakīrti requires that the dependent too be established only within nominal convention. Consciousness-Only preserves the dependent as a certain effective existence. Candrakīrti holds that retaining this existence is precisely what leaves Consciousness-Only with a residue of realism.

4. The Consciousness-Only position and its countercritique of Madhyamaka¶

Candrakīrti's attack touches a real tension within the Consciousness-Only system, already noted at the end of 6.4. But Consciousness-Only does not silently accept it. Asaṅga, Vasubandhu, and Dharmapāla all offer arguments in response to Madhyamaka criticism.

Countercritique one: Madhyamaka's "emptiness" cannot support a system of practice on its own

The strongest Consciousness-Only countercritique is not a direct logical refutation, but a broader philosophical judgment: a system that only refutes and establishes nothing cannot supply operational coordinates for practice.

Madhyamaka says that everything is empty, speaks of nominal designation, and appeals to worldly consensus. But when practitioners sit down to observe their own minds, they need to know what is arising, what is operating, what practice transforms, and what realization sees. Madhyamaka's framework cannot answer these questions. Its "conventional life operates as usual" is broad reassurance, not concrete guidance. The storehouse consciousness, manas, three natures, and transformation of the basis supplied by Consciousness-Only form a map that practitioners can actually use. However much its metaphysical foundation is questioned, this map functionally fills the gap Madhyamaka leaves.

This countercritique does not say that "Consciousness-Only metaphysics is correct." It says that "Madhyamaka's negative formulations are incomplete in their function for liberation." This is a fundamental difference of concern between the schools: Madhyamaka is more concerned with "not grasping," Consciousness-Only with "how to transform."⁴

Countercritique two: self-cognition is not analogous to "external cognition"

Candrakīrti criticizes self-cognition through the analogy "a knife cannot cut itself." Cognition is an outward relation directed at another and cannot turn back upon itself. Consciousness-Only answers that the analogy rests on the wrong basis. A knife cutting itself is a physical relation and cannot serve as an analogy for cognition. The self-illumination of consciousness, its self-cognizing aspect, does not mean that "consciousness as one object knows consciousness as another object." It means that, when consciousness is active, that activity is itself manifestly known. Its existence is its being known: a nonrelational feature of experience itself.

More clearly, the self-cognition of consciousness resembles the immediacy of pain. When you are in pain, you know you are in pain. This "knowing" is not a further cognition arising outside pain to know it; it is pain's own felt character, its sentience. What Consciousness-Only seeks to describe through self-cognition is this intrinsic luminosity of experience, not an external cognitive structure in which "a second consciousness knows the first."⁵ Candrakīrti's knife analogy mismatches a physical relation with experience's intrinsic luminosity.

This book remains doubtful whether the reply succeeds. If self-cognition describes "the felt character of experience itself," it remains theoretically open whether it must be established as an independent "aspect." Sthiramati's position of dispensing with the self-cognizing aspect is precisely that establishing it independently is unnecessary.

Countercritique three: the "existence" of the dependent is not the existence of the imagined

In response to Candrakīrti's criticism that "the dependent is still a real existent," Consciousness-Only distinguishes two kinds of "existence." The "existence" of the imagined is the substantial existence of what is independent, self-sufficient, and established by intrinsic nature. Consciousness-Only fully agrees that there is no "existence" in this sense. The "existence" of the dependent means that "activity arising through conditions is itself occurring." This does not refer to a fixed substance, but to a flowing process, arising and ceasing at every moment without any fixed, unchanging being of its own.

Pursuing the Consciousness-Only position, one could say that Candrakīrti conflates these two kinds of "existence," criticizing "the dependent exists" as though it meant "the dependent has intrinsic nature." This step is this book's reconstruction. The cited work does not address whether Consciousness-Only scholars explicitly made this objection. "The dependent exists" says only that there exists a process occurring under conditions. The "exists" in this proposition carries no implication of "establishment by intrinsic nature."⁶

Candrakīrti's reply is that, as soon as you say "there is a dependently arisen process there," however much you emphasize its dependence, you have left some reality beyond nominal convention at that "there." Otherwise, what does "there is a process there" say? This is the schools' real disagreement, not merely a difference of wording.

5. Distinguishing the levels of dispute¶

Bringing the arguments together, we must first distinguish two levels of dispute. Reckoning them together would mistake disagreements at different levels for a single contest.

First level: the "existence" of the dependent. The question is whether, after the imagined has been denied, the activity of dependently arisen consciousness has a basis of reality independent of nominal convention.

Consciousness-Only answers yes. Although the dependent lacks intrinsic nature, it exists in some way. It is the basis upon which the perfected can appear and the actual object transformed by practice. This "existence" differs from the real existence of the imagined, yet remains a dependent reality irreducible to nominal convention.

At this level, the two Madhyamaka scholars give different answers. Bhāvaviveka says, "I too allow such an intrinsic nature." Madhyamaka already grants intrinsic nature conventionally; what you seek to establish has already been accepted on our side. The difference lies only in the level at which this intrinsic nature is established. For Bhāvaviveka, therefore, this dispute largely dissolves into a confusion of levels. Candrakīrti's answer is the direct denial: the "existence" of the dependent is also nominally designated. "There is a dependently arisen process" has meaning only within the framework of worldly consensus. Nothing has existence beyond nominal designation.

Second level: the "real existence" of suchness and the perfected. The question is whether suchness is a form of notation or a real existent. Here the Consciousness-Only side will not yield: "To say that suchness is not really existent does not accord with reason." Bhāvaviveka answers that suchness "is only the permanent cessation of all conceptual discrimination, not a really existent nature." Through the objection of the non-Buddhist self, he points out that the form "beyond language yet really existent" is indistinguishable from the self posited by non-Buddhists. At this level, the two sides oppose each other within the same coordinates, as Section 2 set out reply by reply.

Once the levels are distinguished, the assessment changes accordingly. At the first level, Bhāvaviveka's difference with Consciousness-Only is mainly about levels, while Candrakīrti's is substantive. The latter is ultimately a question about "the basis of conventional truth." The validity of conventional truth rests either on a dependently arisen existence or purely on nominal convention itself. Consciousness-Only chooses the former; Candrakīrti the latter. At the second level, no distinction of levels removes the conflict. This is the actual point of direct dispute between Bhāvaviveka and Consciousness-Only.

This book holds that the disagreement at the first level is consistent with each school's own system. Neither side can easily overwhelm the other by logic alone. Consciousness-Only has grounds for criticizing Madhyamaka for "leaving practitioners with nothing to work on." Candrakīrti has grounds for criticizing Consciousness-Only for "retaining residual real existence." Pursued to the end, each path can form an internally coherent system, but neither can supply a final decision from within the other's framework.

6. The limits of connecting through the three absences of intrinsic nature¶

Consciousness-Only uses the three absences of intrinsic nature to connect with emptiness (6.4, Section 5), seeking to show that "consciousness-only" and "everything is empty" are two expressions of the same truth. The logic of the connection is ingenious. Here we explain further why it is incomplete.

The crucial step is "the dependent nature's absence of intrinsic nature in arising." The dependent has no intrinsic nature of "arising by itself"; it depends on conditions. Both Consciousness-Only and Madhyamaka accept this. But Candrakīrti asks further: accepting that "the dependent has no independent capacity to arise" is not the same as accepting that "the dependent itself is nominally designated." There is a leap between these propositions:

Consciousness-Only accepts that the dependent has no independent intrinsic nature (absence of intrinsic nature in arising). It does not accept the step from this premise to the conclusion that the dependent itself is therefore nominally designated and has no basis of reality beyond designation.

Consciousness-Only is willing to accept that "the dependent lacks intrinsic nature," but retains the claim that "the dependent has existence in some sense." This "existence in some sense" is the residue that the connection through the three absences of intrinsic nature cannot dissolve. The connection explains "lack of intrinsic nature," but does not dissolve "having existence." Candrakīrti's emptiness requires that neither have independence.³

7. An attempt at synthesis: Yogācāra-Madhyamaka¶

Indian Buddhism did not stop at the opposition between the schools. The eighth-century Śāntarakṣita (c. 725–788) and his disciple Kamalaśīla proposed a philosophical synthesis, generally called "Yogācāra-Madhyamaka" (Yogācāra-Mādhyamika), seeking to combine their strengths.⁷

Śāntarakṣita's two-step synthesis

In the Ornament of Madhyamaka (Madhyamakālaṃkāra), Śāntarakṣita proposes a philosophical approach in stages:

First, use the epistemological arguments of Consciousness-Only to deny the independent existence of external objects. Following the path of Dignāga and Dharmapāla, show that external objects are only manifestations of consciousness, with no material entities existing independently of it. Consciousness-Only handles this step.

Second, use Madhyamaka's analysis of emptiness to deny the independent existence of consciousness itself. Following the path of Nāgārjuna and Candrakīrti, show that consciousness likewise lacks intrinsic nature. Its existence too is dependently arisen and nominally designated; it cannot become the foundation of any ultimate reality. Madhyamaka handles this step.

External objects are thus denied first, the contribution of Consciousness-Only; consciousness is then denied as well, the contribution of Madhyamaka. The final conclusion of thoroughgoing emptiness is not that "consciousness is real," but that "all dharmas are empty: external objects are empty, and consciousness too is empty."⁷

The advantage of this strategy is that it gives the epistemological arguments of Consciousness-Only a place in practice, as the first step of dismantling the real existence of external objects. At the same time, Madhyamaka's emptiness ensures the final conclusion that "consciousness too is empty," preventing Consciousness-Only from leaving a residue of realism on the side of consciousness.

The cost is that it effectively demotes Consciousness-Only to an auxiliary step on the path of practice (refuting external objects) rather than a final philosophical position. Consciousness-Only thinkers may not accept this demotion. They hold that the three natures already suffice to show that consciousness is dependently arisen and lacks intrinsic nature, without borrowing Madhyamaka emptiness to deny it once again. More fundamentally, after both steps are completed and "dependently arisen consciousness has been denied by the analysis of emptiness," what does practice act upon? This returns us to the challenge that Madhyamaka faces on its own, the need for something to work on in practice under emptiness. Śāntarakṣita supplies levels, but does not fully resolve this problem.

Śāntarakṣita's historical influence

Śāntarakṣita and Kamalaśīla entered Tibet in the late eighth century, bringing this Yogācāra-Madhyamaka system and profoundly influencing the philosophical foundations of Tibetan Buddhism. Tibetan Madhyamaka education thereafter developed between "Yogācāra-Madhyamaka" and "Prāsaṅgika Madhyamaka," Candrakīrti's lineage, with successive scholars repeatedly comparing and ranking them. In particular, Tsongkhapa (fourteenth–fifteenth centuries) of the Geluk school vigorously upheld Candrakīrti's Prāsaṅgika as the highest view and devoted specific discussion to Śāntarakṣita's synthetic approach (see Essay 11).

8. This book's assessment¶

Candrakīrti's criticism is powerful, but not final.

Candrakīrti's criticism of self-cognition touches a weak point in the epistemological structure of Consciousness-Only, but its reply concerning "the intrinsic luminosity of experience" is not ineffective. Versions of this dispute remain active in modern philosophy of mind, in discussions of the phenomenal character and reflexivity of consciousness. His criticism that "the dependent is still a real existent" exposes the limits of the connection through the three absences of intrinsic nature. This is a substantive attack. Yet Candrakīrti's own position, "conventional consensus as the basis," likewise faces the question of why conventional consensus itself is valid, as this book noted in 5.5, Section 3. Both schools leave unanswered questions at their respective limits. On Bhāvaviveka's side, the issues must be recorded by fascicle: the first largely dissolves into a confusion of levels; only the second, the dispute over suchness as really existent, sets the parties against each other at the same level.

Madhyamaka cannot replace the positive account supplied by Consciousness-Only.

What Consciousness-Only supplies has real philosophical value in history: detailed descriptions of the accumulation and purification of afflictions; systematic analyses of mental states at each stage of practice; and an explanation, through manas, of "why the feeling of self-grasping persists after no-self has been accepted intellectually." These correspond to phenomena actually observable in meditation. At the level of practice, Madhyamaka's "dismantling of attachment" and Consciousness-Only's "transformation" are complementary, not simply substitutes for one another.

Volume 2 will develop the following point. Consciousness-Only theory touches a real pressure point in Madhyamaka's account of liberation: how epistemological and philosophical emptiness can actually loosen greed, hatred, and delusion. Yet the very scale and subtlety of Yogācāra theory produce a kind of doctrinal self-reference within the path of cultivation.

References¶