Part II, Chapter Six　Verification: Unknowable Appropriation and the Five-Stage Checkpoints¶

Part II, Chapter Six, of the Madhyamaka essay traced the historical drift of criteria for realization, from the cessation of suffering toward skill in debate. Yogācāra's verification problem has a different shape. Its criteria have scarcely drifted: the characteristics of stages are in black and white, unchanged for a thousand years. Its problem is a structural defect present from the outset: the path with the densest checkpoints is installed on the substrate with the poorest observability. The milestones are perfectly clear; the roadbed lies where no one can see it. This chapter takes three steps: first record the density of checkpoints, a real contribution to verification; then examine how unknowability undermines those checkpoints from the root; finally ask into whose hands the authority to recognize attainment falls within this structure.

I. Credit due: the densest checkpoints among the five systems

As Part I, Chapter One, described, each of the five stages and ten grounds has its characteristics: arousing the aspiration through four powers at accumulation; warmth, summit, acceptance, and the highest worldly state at preparation, each with a concentration and content of contemplation; sixteen moments of mind at seeing; ten grounds of cultivation, each eliminating a portion of innate obstacles and realizing a portion of thusness; and finally vajra-like concentration and complete transformation of the basis. Every station has a name, something eliminated, and something realized. Compared with Madhyamaka, where the content of seeing remains disputed between direct realization and inference (Madhyamaka essay, Part II, Chapter Six), Yogācāra offers Mahāyāna's richest public descriptive vocabulary for the progress of practice. In principle, a Yogācāra practitioner can answer “which stage are you at, and on what characteristics do you judge yourself to be there?” This is a real asset to a culture of verification.

II. Undermining the checkpoints: three sites of unknowability

Checkpoints need verifiable content to be checkpoints. Examining the crucial nodes of this table reveals three sites of systemic unknowability.

First, the substrate is unknowable. Every checkpoint ultimately checks the waxing and waning of defilement and purity in seeds, yet seeds are stored in ālaya with its “unknowable appropriation” (Part I, Chapter One). Milestones claim to mark changes in seeds inside the substrate, while the substrate itself is open to no one's inspection. Each checkpoint's actual operation thus degenerates into using observable characteristics—states of concentration, pliancy, forms of contemplation—as signs from which to infer unobservable seed changes. The correspondence between signs and seeds is precisely what the theory stipulates. In principle, verification therefore cannot escape the circle of theoretical self-confirmation, a replay in verification of the preceding chapter's map problem.

Second, the pivot is unknowable. At seeing, thusness is directly realized and nondiscriminative wisdom becomes present. By definition, this is beyond words and characteristics, known for oneself as warmth and cold are. It is precisely the table's most important station: the two obstacles arising from discrimination are suddenly eliminated here, and noble status is established here. Yet whether it has occurred has no public criterion beyond the person's own report and subsequent behavior as corroboration. This shares a structure with Chan's seeing one's nature and esoteric Buddhism's realization, the tendency toward elitism anticipated in Introduction, Chapter One. Yogācāra's tables have not altered it; they merely give public names to an event that cannot be publicly verified.

Third, eligibility is unknowable. The doctrine of five distinct lineages stipulates inherent differences among śrāvaka, pratyekabuddha, bodhisattva, indeterminate, and no lineage. Sentient beings without lineage can never attain Buddhahood.¹ Lineage is stored in ālaya as the presence, absence, and type of originally present uncontaminated seeds; the person cannot know it for themselves, nor can others directly observe it. This claim must be made somewhat narrower. The treatise's entry requirement is possession of both kinds of lineage. First is the naturally abiding lineage: causes of uncontaminated dharmas naturally acquired, attached to root consciousness from beginningless time. Second is the lineage formed through cultivation: formed by perfuming through hearing and so forth after hearing the Dharma flowing from the dharmadhātu. Only with both can one enter. The latter is in principle half open to checking: whether one has heard the Dharma flowing from the dharmadhātu, and to what extent hearing has perfumed one, are publicly visible matters. The treatise also records a position that does not connect lineage to naturally present uncontaminated seeds at all, establishing it instead through the presence or absence of obstacles (Volume One, Part II, Essay Twenty, Chapters Three and Six). The criticism still holds: if one item in a conjunction is unknowable, the conjunction is unknowable. Yet the wording should not describe every part as unknowable. Thus practitioners cannot themselves confirm “whether this road is in principle open to me.” In practice, deciding lineage can only appeal to teachers' observation and inference; the basis for that judgment, naturally present uncontaminated seeds, is the system's most deeply concealed posit. From the outset, theory places verification's authority in the hands of authority figures, rather than reaching them through the drift seen in the Madhyamaka essay.

Beyond these three sites of unknowability, one retrospective blow must be added. The baseline's standard declaration of verification—“My birth is exhausted, the holy life established, what had to be done is done; I know for myself that I will receive no further existence” (Madhyamaka essay, Part II, Chapter Six)—is the practitioner's declaration after personally checking at the observable level. Once transformational birth and death is established (Part II, Chapter Three), this declaration contracts: birth is exhausted only in the sense of birth and death in bounded lifetimes. The Śrīmālā Sūtra explicitly says that the arhat “has birth, because residual dharmas of birth are not exhausted.” Another, deeper birth that the system stipulates he cannot observe is entered under his name: a mind-made body produced with the abiding ground of ignorance as condition and uncontaminated karma as cause.² Consciousness-only teachers leave room through the five distinct lineages: the parinirvāṇa of a śrāvaka of fixed lineage counts, while one of indeterminate lineage turns toward bodhi, receives a transformational body, and continues. Yet this room welds retrospective contraction to the third site of unknowability. Whether an arhat's “birth is exhausted” is a terminus or a transfer depends on lineage, which cannot be known for oneself. Throughout, what overturns or suspends the verification is a new theoretical entity, not any new observation. Adding the unknowable level thus reveals a second power: it not only undermines present checkpoints but can retrospectively revise completed verification. The finish line itself becomes an adjustable parameter of theory.

III. Two registers: epistemological and feeling-based criteria

The three sites of unknowability concern “what cannot be checked.” A deeper problem remains: even if checking succeeded, the entries would belong to different books. The baseline's verification criterion is grounded in feeling: one is not struck by the second arrow; painful feeling no longer produces grief, sorrow, vexation, and suffering; greed, hatred, and delusion no longer arise. The criterion lies in the long-term direction of feelings and afflictions, open to personal checking and others' corroboration (Introduction, Chapter Two). Yet the pivotal verification throughout Yogācāra is epistemological: whether one has seen the path depends on whether a wordless direct realization has occurred. The two criteria differ by more than fineness of detail. They are two registers: one records the cessation of suffering itself, the other a cognitive event stipulated by doctrine as “equivalent to liberation.”

The difference appears in one question: how can one distinguish someone who has attained nondiscriminative wisdom from someone who has gone dangerously astray in meditation? Under the feeling-based criterion there is an answer, and a slow one: examine the course of greed, hatred, delusion, and suffering over the years and decades that follow. Cessation can be checked over the long term. Under the epistemological criterion there is no answer. The event is wordless, privately known, momentary; its content is by definition inexpressible. An actual direct realization and an intense abnormal experience with no referent can produce identical self-reports at the time and afterward. The tradition is aware of this problem. Meditation texts contain many sections distinguishing demonic states, yet their criteria remain the fit of characteristics and conduct with doctrine, still gathering evidence from the periphery.³ The system can therefore only appeal outward: subsequent behavior as corroboration, stage characteristics as comparison, and ultimately a teaching lineage versed in doctrine as judge. This converges with the next section. The authority to recognize attainment belongs to the scholastic lineage not only because of three sites of unknowability, but because verification is entered in a register only theory can read. Anyone can read the baseline's register, but slowly. Yogācāra's is read quickly, settling matters in one moment, at the cost that only the system itself is qualified to announce what was read.

“Nonconceptual wisdom” renders the same term as nondiscriminating wisdom (无分别智) in the text. The final appeal to 师门 is to the teaching lineage that judges the practitioner's attainment.

IV. Where authority to recognize attainment ends up, and the form of the scholastic lineage

Together, the three sites of unknowability determine the form of Yogācāra's culture of verification: the tables are public; interpreting them is the experts' domain. Who can interpret characteristics, recognize stages, and distinguish lineage? Those who thoroughly understand the treatise. Verification's authority thus naturally gathers at the center of doctrinal transmission: treatise authors and their scholastic lineages. Its historical form matches this. Yogācāra is Mahāyāna's most scholastic system. The disputation halls of Nālandā, the Xuanzang line's translation workshops, and Faxiang's tradition of doctrinal interpretation all organize authority around “mastery of the system” (Volume One 6.5, 10.7). The drift recorded in the Madhyamaka essay, from cessation to skill in debate as criterion, need not occur here, because Yogācāra's criterion presupposes theoretical mastery by design. The endpoint is the same: Madhyamaka drifts there; Yogācāra builds a road straight to it.

One point must be made for Yogācāra. Its normative texts do not confuse “understanding treatises” with “realizing fruition.” Their authors know that stages must be attained through practice and realization, not discourse. The detail of their descriptions also gives sincere practitioners more reference points for self-examination than any other system. This chapter criticizes no one's honesty, but the structure. When substrate, pivot, and eligibility are all unknowable, even the densest checkpoints can gather evidence only in circles within theory. What “verification” names has changed from “checking claims against experience” to “checking experiences against the system.”

V. Conclusion

The baseline's verification is observable in the present life and does not outsource its criterion (Introduction, Chapter Two). Its measure is to verify progress along the road, not completion at the endpoint, as the section on “the metaphysical bill” explains. Yogācāra reverses this measure altogether. It makes tables for both endpoint and whole course, yet locates their ultimate referents—seeds, thusness, and lineage—entirely beyond observation. The criterion changes registers too, from cessation of suffering to occurrence of a cognitive event. The fuller the tables, the more complete theory's monopoly on verification. Madhyamaka fails for lack of a table to consult; Yogācāra fails because theory issues and checks its own table. Of the two failures, the latter is less visible, precisely because it looks like an exemplary answer to the problem of verification. Before assessing attachment, one last question remains: what exactly does such a system crave?

References¶