Chapter Three: Vasubandhu—The Twenty Verses and Thirty Verses on Consciousness-Only¶

As the previous chapter (6.1) explained, the Vasubandhu who first mastered Sarvāstivāda and Sautrāntika Abhidharma and wrote the Abhidharmakośa was the same person who later turned to Mahāyāna consciousness-only thought under Asaṅga’s influence. His two central treatises clearly preserve the course of this transformation. The Twenty Verses on the Demonstration of Consciousness-Only (Viṃśatikā-vijñaptimātratā-siddhi) is polemical: it seeks to prove “consciousness-only” philosophically by denying the independent existence of external objects. The Thirty Verses on Consciousness-Only (Triṃśikā-vijñaptimātratā) is systematic: it presents the entire structure of consciousness in the most economical form possible. The two works complement one another—the former dismantles, and the latter establishes.

I. The Legend: How Asaṅga Persuaded Vasubandhu¶

As usual, we will first record the legend and then turn to what can be historically established.

Traditional accounts describe Asaṅga and Vasubandhu as half-brothers through their mother—or full brothers, according to other versions. Both entered the Sarvāstivāda order in their youth, and both became outstanding scholars of their age. Asaṅga turned to Mahāyāna first, while Vasubandhu remained within the Abhidharma tradition, wrote the Abhidharmakośa, and composed works criticizing Mahāyāna scriptures as not being the Buddha’s word. Concerned by this, Asaṅga arranged for Vasubandhu to visit him.

When they met, Asaṅga did not debate him directly. After Vasubandhu had fallen asleep, Asaṅga had an attendant recite the Daśabhūmika Sūtra (the “Chapter on the Ten Grounds” of the Avataṃsaka) within his hearing. Vasubandhu reportedly heard it while half asleep, was deeply moved on the spot, and awoke regretting that he had used his tongue to slander Mahāyāna. He believed that the tongue should be cut out as penance. Asaṅga stopped him: why cut out the tongue when he could instead use it to spread Mahāyāna? Vasubandhu then turned to it completely and became the most important theoretician of consciousness-only philosophy.¹

As usual, the historical value of this narrative is limited. It depicts a conversion through emotional influence rather than a process of philosophical persuasion. In one respect, this is refreshingly honest: if hearing a passage of scripture was enough to effect the change, its motive was emotional and devotional rather than philosophical argument. The candor of the traditional narrative is itself worth noting.

Another common episode says that Asaṅga repeatedly ascended to Tuṣita Heaven to receive teachings from the bodhisattva Maitreya, while Vasubandhu appeared there through supernatural power and listened to their exchanges. This detail is still harder to verify. We likewise preserve it as legend without treating it as historical evidence.

II. A Timeline of Intellectual Development¶

The works of Asaṅga and Vasubandhu, together with the Five Treatises of Maitreya, form the core literature of Yogācāra’s first two centuries. Their approximate chronology and intellectual development can be arranged as follows:

Several qualifications are needed. The dates of the Five Treatises of Maitreya are highly disputed; the table gives only broad ranges. Philosophically, Dignāga and Dharmakīrti belong to the Yogācāra current, but their central contributions concern epistemology and logic (8.1 and 8.2); they are listed here simply to display the line of transmission. “Early” and “post-conversion” Vasubandhu distinguish intellectual stages, not two people. This book’s position is that there was one Vasubandhu. All the dates in the table are conjectural and are given differently by different authorities: the Five Treatises of Maitreya are the most disputed of them; no settled account of Sthiramati’s dates exists; the range for Dignāga follows the one adopted in 8.1; and Dharmapāla’s dates follow the older East Asian reckoning and agree with the record in 10.4.

III. The Trajectory of the Philosophical Transformation¶

Vasubandhu’s movement from Abhidharma to consciousness-only was not a sudden leap; it extended an internal logic already present in his earlier thought. In the Abhidharmakośa, he had already inclined toward Sautrāntika, questioning Sarvāstivāda entities such as “acquisition” and “unmanifest form” and using “seeds” as a more economical explanation. Sautrāntika representationalism was the crucial intermediate station: if what we cognize directly is always only a cognitive image, and cognition never encounters anything but representations within consciousness, must we still assume the existence of the external object that is merely inferred?

The Sautrāntika answer was yes. An independent external object was still needed to explain why different people in the same place at the same time see the same thing. Vasubandhu argues in the Twenty Verses that this reason is insufficient: all such phenomena can be explained without positing external objects.²

A second path originates in the Abhidharmakośa’s analysis of the “self.” Vasubandhu was already familiar with an argumentative structure that analyzes a seemingly natural entity, the “self” or “person,” as a functional description of a stream of consciousness, without positing a bearer outside that stream. Applied to external objects, this method becomes the basic posture of consciousness-only: it analyzes the apparently natural “external world” as a manifestation of consciousness, without positing an independent material entity outside consciousness.

IV. The Twenty Verses: Four Arguments against External Objects¶

The central thesis of the Twenty Verses can be summarized in one sentence: external objects are not real; everything called an “external object” is only a transformation of consciousness (vijñapti-pariṇāma). Vasubandhu knew that denying external objects would immediately invite four standard objections. His replies to them form the body of the treatise.³

The four objections and the four replies are each gathered into a single verse. The verse of objection lists four standards of acceptance: “If consciousness had no real object, then the fixity of place and of time, the nonfixity of the continuum, and the performance of function should not hold.” The verse of reply assigns one sample to each: “Place and time are fixed, as in a dream; bodies are not fixed, as with the ghosts who alike see a river of pus and the like; and as in a dream there is loss, so function holds.”⁴ The distribution of the samples follows the text as it stands: fixity of place, fixity of time, and causal function all take the dream, while the nonfixity of the continuum takes the hungry ghosts who see a river of pus. A further simile of the wardens of hell appears below. It does not belong to this distribution; the treatise sets it in a separate passage, where a single simile discharges all four points at once.

Objection One: Spatial Restriction

If no external object exists, where does the content or image aspect of consciousness come from? Moreover, why does a certain kind of cognition occur only in a certain place? I see a cup because the cup is “there.” Without an external object, why would cognition of a cup not occur everywhere?

Vasubandhu replies that dreams also have spatial restrictions. In a dream, one sees particular things only in particular places, not everywhere at once, although no external objects exist there. Spatial restriction in cognition therefore does not require the existence of external objects. It can be determined entirely by the structure of consciousness itself—its seeds, karma, and conditions of perfuming.

Objection Two: Temporal Restriction

The parallel question is: why do I see the cup now but not at another time? Does temporal restriction not show that an objectively existing external thing appears there at a particular time?

Vasubandhu again appeals to dreams. Dream time is likewise restricted: one does not see everything simultaneously throughout the night, yet there are no external objects in a dream. Temporal restriction can also be explained by the internal structure of consciousness, without invoking an independent object.

Objection Three: Shared Perception among Multiple Subjects

This is the strongest objection. If external objects do not exist, why do different people in the same place at the same time have the same perception? Is this intersubjective agreement not the strongest evidence that external objects exist independently? If everything is a manifestation of each individual’s consciousness, why does each person not see something entirely different?

The sample Vasubandhu assigns to this objection is the hungry ghosts, not hell. The verse of reply reads “bodies are not fixed, as with the ghosts who alike see a river of pus,” and the gloss reads “through one class of karma and one ripening, many bodies gather together and all alike see a river of pus”: several continua that have performed the same class of action and undergo the same class of ripening converge in one place and see one and the same river of pus. The river is not outside consciousness, and the seeing in common occurs all the same.⁴ What this step supplies is an alternative mechanism for intersubjective agreement. Several people seeing the same thing can be explained by similar seeds maturing simultaneously under similar conditions; the explanatory work is done inside consciousness, and logically it is no weaker than an appeal to the independent existence of external objects.³

Objection Four: Causal Efficacy

A knife in a dream cannot actually cut the skin, whereas a waking knife can. Does this difference in causal efficacy not prove that external objects are real?

Vasubandhu replies that differences in causal efficacy can likewise be explained within consciousness. Waking cognition operates in conjunction with the corresponding karmic conditioning and a complex coordination of tactile consciousness and the body; dreams lack this conjunction of conditions. He also adduces nocturnal emission: sexual stimulus in a dream can produce a real physiological effect. This shows that content internal to consciousness can possess genuine causal efficacy and does not become ineffectual merely because there is no external object.³

The Simile of Hell: All Four Points at Once

After the four replies have been given severally, the treatise sets out one further simile that discharges all four points with a single example: “Everything is as in hell, where the wardens and the rest are seen in common and can perform acts of torment; therefore all four points hold.”⁵ The wardens, dogs, crows, and iron mountains of hell, the treatise says plainly, are not “things numbered among real sentient beings”; rather, the beings there, “through the ascendant force of one class of karma and one ripening, in the same place and at the same time and in many continua, all alike see wardens, dogs, crows, iron mountains, and the like come to them and perform acts of torment.” Place is fixed, time is fixed, many continua see in common, and the acts of torment take real effect: the four standards of acceptance are met together in one example. Why the wardens do not count as real sentient beings the treatise settles on four further grounds. The wardens do not undergo the sufferings of hell and so are not denizens of hell. If the two sides tormented each other, neither party could be counted a denizen of hell. Being equal in size and in strength, they ought not to be extremely afraid of one another. And beings who cannot themselves endure the flames of the iron ground are in no position to torment anyone else.⁵

The Dilemma of Karmic Force: The Burden of Proof Moves

Here the opponent falls back to a final position: the wardens are “different great elements” produced by the ascendant force of karma, which undergo transformations of this kind, and the name “warden” is established upon that set of elements. Vasubandhu’s answer is the pivot of the whole treatise: “If you allow that by the force of karma different great elements arise and undergo transformations of this kind, why not allow as much of consciousness?”⁶ To explain the shared seeing and the shared undergoing, the opponent has already conceded that karmic force can produce an entire apparatus of torment. Vasubandhu merely changes what is produced from great elements into consciousness, adding not a single step to the mechanism. The verse that follows immediately makes the inconsistency explicit: “The perfuming of karma is in one place, and the result is held to be in another; the consciousness that is perfumed has the result, and for refusing this what reason is there?”⁶ By the treatise’s own admission the karmic perfuming of the denizens of hell lies only in the continuum of consciousness and nowhere else. To refuse the result where the perfuming is, and to insist upon the result where the perfuming is not, is an asymmetry for which the opponent has offered no reason.

The form of this move deserves to be recorded plainly, because the burden of proof shifts here. Vasubandhu adds no new evidence for consciousness-only; he points to the step the opponent has already surrendered. The question at issue changes from “can consciousness-only be made to work?” into “on what ground do you stop halfway along a mechanism you have already accepted?”

The four objections having been resolved one by one, and the simile of hell having discharged all four together, Vasubandhu concludes that the independent existence of external objects is not a hypothesis required by the phenomena we see and cognize. The world is not nothing; its mode of existence is simply a manifestation of consciousness rather than a material entity independent of consciousness.

From the standpoint of modern philosophy, this conclusion crosses a gap in the argument that needs to be marked. Vasubandhu’s strategy is to provide, for each of the four objections, an alternative explanation “within consciousness,” thereby showing that the independent existence of external objects is not a necessary hypothesis. But “the phenomena can be explained without external objects” and “external objects do not exist” are propositions of two different kinds. The former establishes only that consciousness-only is a logically possible theory; it cannot prove that external objects in fact do not exist. Frameworks in which external objects exist and frameworks in which they do not can each explain the same phenomenal data in their own terms. This is a classic case of the underdetermination of theory by evidence: observable phenomena are compatible with two competing theories, and the phenomena alone cannot decide which theory is correct.

More directly, Vasubandhu proposes an internally coherent speculative theory that demonstrates how a consciousness-only framework can explain phenomena. But that is not the same as proving from those phenomena that external objects do not exist. Moving from “my theory does not require X” to “X does not exist” is a logical leap. A critic may reasonably say: your theory is one possible explanation, and mine, which includes external objects, is another. You must supply a further argument showing either why yours is superior or why the competing explanation fails in principle; it is not enough merely to show that your alternative works. The argumentative structure of the Twenty Verses does not close this gap. It only neutralizes four objections and never positively establishes the proposition that “external objects do not exist.” A modern observation may be added: the very explanatory omnipotence displayed in these four rounds of reply should raise suspicion. In modern philosophy of science, the ability to accommodate every possible experience is a defect rather than a virtue. A theory that rules out no possible observation is constrained by none; “confirmation everywhere” is a symptom of untestability.⁷ The replies concerning place, time, shared perception, and causal efficacy each incorporate a whole class of potential counterexamples into an explanation internal to consciousness. This comprehensive reach prevents consciousness-only from being refuted, but it also prevents it from being tested.

The section of the Twenty Verses on the hidden intent leaves three sentences of self-limitation, the most important act of self-restraint in the whole treatise. The three stand together beneath a question the author puts to himself: “If one enters the selflessness of dharmas by knowing that dharmas in every mode do not exist, then consciousness-only too is utterly nonexistent, and what is there to establish?”⁸ First, consciousness-only does not assert that dharmas in every mode do not exist. The selflessness of dharmas means “the selflessness of those dharmas whose intrinsic natures and distinctions are imagined by the foolish,” and the sentence closes with “it is not a repudiation of existence in every mode.” Second, the nature of consciousness-only is itself on the list of what may not be grasped: the text reads “the nature of consciousness-only that another consciousness grasps is likewise without substance, and this is called the selflessness of dharmas.” Third, the purpose of establishing the teaching of consciousness-only is to bring beings to enter the selflessness of all dharmas, not to establish one further existent.

These three run in the same direction as the words that state the thesis at the outset: the treatise “denies only the external object and does not dismiss the associated factors.” What is denied is the object outside consciousness; what is retained are the mental factors and the mechanism; and the general name “consciousness-only” is itself not permitted to be raised into a real existent.⁹ The closing verse agrees in its wording: “According to my own capacity I have briefly established the meaning of consciousness-only.” It says briefly established, and does not claim to have demonstrated. This self-limitation is Vasubandhu’s posture of contact with Madhyamaka, and it becomes a focal point in later commentarial disputes: Sthiramati’s lineage takes it as warrant for pressing the emptiness of consciousness further, while Dharmapāla’s lineage holds the self-sufficient structure of consciousness at the self-cognizing aspect.

V. The Thirty Verses: The Three Transformations and the System of Consciousness¶

The Thirty Verses is Vasubandhu’s systematic outline. Unlike the Twenty Verses, it has no prose commentary by Vasubandhu himself, only thirty brief verses. This helped make it one of the most extensively commented works in Buddhist history. After Vasubandhu’s death, ten major Indian commentators (the “Ten Great Masters”) filled the interpretive space left by the verses in accordance with their own theoretical tendencies.¹⁰

The Three Transformations

The opening verse goes directly to the heart of the matter: “Through the provisional designation of self and dharmas, manifold appearances unfold; these depend upon transformations of consciousness.” In other words, the “self” and “dharmas” of ordinary discourse are various appearances manifested by consciousness and possess no independent substance. The “transformation” (pariṇāma) of consciousness has three forms:

The framework of three transformations distills Asaṅga’s system of storehouse consciousness into verse, while explicitly expressing the central consciousness-only position that every conscious activity is a “transformation.”¹⁰

The Meaning of “Transformation of Consciousness”

“Transformation” (pariṇāma) is the key term. It does not mean that “consciousness creates the world.” It means that when cognition occurs, consciousness differentiates into a perceiving aspect (the side that cognizes) and an image aspect—the side that is cognized. What is called the “external object” is really the image aspect manifested by consciousness itself, not a thing independently existing outside it. The two sides are aspects of the movement of consciousness itself, not consciousness illuminating an independent external world.

A comparison with modern philosophy may help clarify the doctrine, although it implies no historical influence. It somewhat resembles Fichte’s “I positing the not-I.” In its activity, the self or consciousness posits an opposing term, the image aspect, which is the product of that activity rather than an independent external thing. Indian consciousness-only and German idealism differ greatly in their modes of argument and their concerns; this analogy is only an aid to understanding.

VI. The Four-Aspect Theory: How Consciousness Cognizes Itself¶

The Thirty Verses does not explain self-cognition in detail, but this became one of the most important controversies in the commentarial tradition. Since it bears on the entire epistemological architecture of consciousness-only, it deserves separate treatment.

The problem begins with a question: while consciousness cognizes an object, how does it know that it is cognizing? If every cognition requires another cognition to cognize it, an infinite regress follows. Consciousness-only theorists therefore need a form of “self-cognition” (svasaṃvedana) to stop the regress.

Where the Cheng weishi lun records this matter, it sets down two aspects, three aspects, and four aspects as three ascending positions, introducing them with “one interpretation holds” and with two impersonal formulas of the type “those who maintain,” and it names no one from beginning to end. Later East Asian commentary attached the last of the three to Dharmapāla (c. 530–561) and called it the “four-aspect theory”:

Dharmapāla’s logic is that the self-cognizing aspect and cognition of self-cognition mutually verify one another, forming a closed loop. The former knows the perceiving aspect; the latter knows the former; and the former can in turn know the latter. This mutual verification does not generate an infinite regress, while ensuring that every level of conscious activity is known.¹⁰

Sthiramati (c. fifth–sixth century) took a different position. He regarded the perceiving and image aspects as manifestations produced by false discrimination: they belong to the imagined nature and are not truly existent; what can be said to subsist is consciousness itself. The matter of attribution needs a chronological qualification. The Cheng weishi lun names no one at all in setting down these several positions, and the classification of Sthiramati, Nanda, Dignāga, and Dharmapāla under one, two, three, and four aspects is the language of later East Asian commentary rather than the self-description of any of the four; the dates of all four are likewise conjectural and are given differently by different authorities. The criticism that “accepting self-cognition covertly introduces substantiality” instead comes from the Madhyamaka line, including Candrakīrti; Section 6.5 returns to it.¹¹

The division between Dharmapāla’s and Sthiramati’s lineages directly determined which version of consciousness-only Xuanzang brought back to China. At Nālandā, Xuanzang studied in the lineage of Śīlabhadra, Dharmapāla’s disciple. The East Asian Faxiang school therefore treated Dharmapāla’s lineage as authoritative. Candrakīrti’s critique of “self-cognition” in the fourth section of 5.5 likewise targets this position.

VII. The Ten Great Masters and the Commentarial Landscape¶

Vasubandhu did not write a commentary on the Thirty Verses. By the time Xuanzang reached Nālandā in the seventh century, ten major theoreticians had each composed one. Xuanzang brought the material of all ten back to China and blended it into a single work, taking Dharmapāla’s lineage as primary while drawing on the other nine, entitling it the Cheng weishi lun and attributing it to “Dharmapāla and others.” The blended translation is Xuanzang’s own doing; what Kuiji wrote is the Cheng weishi lun shuji, a commentary upon that blended translation, which is a separate thing from the blending itself.¹² The Cheng weishi lun is the foundational text of the Faxiang school and the most important document for understanding East Asian consciousness-only thought.¹⁰

Dharmapāla and Sthiramati represent the clearest poles among the Ten Great Masters. Dharmapāla affirmed the four-aspect theory and the real existence of self-cognition, tending to preserve a self-sufficient structure of consciousness. Sthiramati pressed the emptiness of consciousness more thoroughly and treated self-cognition as an inferential construction rather than an independently real existent, tending toward Madhyamaka. Their different degrees of emphasis on emptiness reflect a persistent tension within consciousness-only: how thorough must the claim of emptiness be before it ceases covertly to introduce a new realism?

VIII. Is Consciousness-Only a Form of Idealism?¶

Modern scholars frequently ask this question, and it deserves a direct answer.

The superficial resemblance: Consciousness-only says that all external objects are manifestations of consciousness and that no material entity exists independently of consciousness. This sounds much like Berkeley’s subjective idealism: “to be is to be perceived,” and the material world is only the content of mind.

The crucial difference: Berkeley’s idealism rests upon an omniscient divine mind. The material world does not depend upon any individual human mind, but it exists in dependence upon the mind of God, which guarantees its objectivity. Consciousness-only has no such role for a divine mind. It explains intersubjective agreement by means of karma and shared karma: several streams of consciousness present similar contents because they have undergone similar karmic perfuming. No external guarantor is required.¹³

The more fundamental difference is that consciousness-only does not say “consciousness is the ultimately real substance.” Particularly in Vasubandhu, Asaṅga, and theoreticians influenced by Madhyamaka, consciousness itself is empty (vijñāna-śūnyatā) and dependently arisen. Its mode of existence is likewise “arising in dependence upon others and lacking intrinsic nature”; it is not an ultimate foundational entity. This distinguishes consciousness-only from idealism in general. Idealism ordinarily treats mind or consciousness as the final basis requiring no further explanation. Consciousness-only says that consciousness, too, is conditioned and empty: it is the starting point for dismantling the illusion that external objects exist independently, not the terminus of explanation.¹³

A more accurate characterization might therefore be that consciousness-only is a form of “methodological idealism.” It gives analytical priority to consciousness in order to remove attachment to the real existence of external objects, but it does not ontologically establish consciousness as an ultimately real ground; consciousness itself remains subject to the analysis of emptiness. This complex position is the deep source both of internal consciousness-only disputes (Sthiramati against Dharmapāla) and of the controversy between consciousness-only and Madhyamaka—Asaṅga against Candrakīrti.

IX. Tensions and Problems¶

1. The Tension between “External Objects Are Not Real” and Common Sense

Consciousness-only denies the independent existence of external objects, yet accepts an effective and shared experiential world on the conventional level. The question is: if external objects are only manifestations of consciousness, what difference in ordinary practice is there between “the table is there waiting for me to cognize it” and “the table is merely the appearance manifested by my consciousness”? The consciousness-only answer is that the difference lies not on the level of ordinary practice (conventional truth), but on the theoretical level—ultimate truth. But does this distinction dilute the consciousness-only claim until it becomes difficult to identify? This resembles the predicament faced by Candrakīrti’s establishment through ordinary convention in the third section of 5.5: the more one protects conventional efficacy, the harder it is to say what additional content “consciousness-only” contributes.

2. The Problem of Other Minds

If external objects are not real, what about other minds? Within the consciousness-only framework, “other people” are also image aspects manifested by my consciousness. Does this imply solipsism—the conclusion that only my own consciousness exists? Vasubandhu recognized the problem in the Twenty Verses and gave a similar explanation in terms of karmic perfuming: my stream of consciousness has been perfumed so as to display the “existence of other minds.” But this answer appears weak. It explains why I have a cognition that “other minds exist,” without explaining whether other minds actually exist.¹³

3. Self-Cognition and Infinite Regress

Dharmapāla’s four-aspect theory uses cognition of self-cognition to stop the infinite regress. But does a structure in which “self-cognition and cognition of self-cognition mutually verify each other” genuinely stop the regress? Or does it merely push the regress inward and change its form? Candrakīrti’s criticism, that a knife cannot cut itself, targets precisely the logical difficulty internal to this closed loop.

Some of these tensions remained incompletely resolved within the consciousness-only tradition; others continued through later debates between consciousness-only and Madhyamaka. That is the subject of this part’s fifth chapter (6.5). The next chapter first examines the three natures and transformation of the basis, consciousness-only’s most complete formulation of liberation.

References¶