---
title: "Non-Self-Existent Prophecy Displayed at Both Ends: The Symmetrical Wings of the Devadatta Chapter and the Limits of a Seed-Reading"
author: "釋慧鏡 (Shi Huijing)"
series_short: "S7"
paper_id: "S7-P08"
paper_number: 8
date: "2026-04"
version: "1.0"
license: "CC BY-NC-SA 4.0"
---
# Non-Self-Existent Prophecy Displayed at Both Ends: The Symmetrical Wings of the Devadatta Chapter and the Limits of a Seed-Reading

**Series:** S7 攝 (Saddharmapuṇḍarīka / Lotus Sūtra)
**Number:** S7-P08

## Abstract

This paper takes the Devadatta Chapter (Ch. 12) of the *Saddharmapuṇḍarīka-sūtra* (hereafter *Lotus Sūtra*) as its focal point and advances three theses.

**Thesis I** — The independent status of the Devadatta Chapter in Kumārajīva's recension is a late editorial event internal to the Chinese translation tradition, not a reflection of any Sanskrit original. Dharmarakṣa's *Zhèng-fǎ-huá-jīng* 正法華經 (T0263, 286 CE) and the Jñānagupta–Dharmagupta *Tiān-pǐn-miào-fǎ-lián-huá-jīng* 添品妙法蓮華經 (T0264, 605 CE) both preserve the Devadatta prophecy and the Dragon-girl's buddhahood together within the Stūpa Chapter as symmetrical wings of a single narrative unit. Guanding's 灌頂 *Fǎ-huá-wén-jù* 法華文句, speaking from within the Tiantai tradition, openly records that this material was restored into Kumārajīva's text at a later date. These three independent witnesses constitute a documentary triangle establishing that the "symmetrical wings" configuration is the original textual form, and that Kumārajīva's recension represents a *disaggregation* rather than the later recension representing an *insertion*.

**Thesis II** — Roughly ninety-five percent of the *Lotus Sūtra*'s prophecy grammar is a direct continuation of the Āgama prophecy grammar already fully in place by the early canonical period: the nine-field structure, the ten-epithet formula (*daśa-bala*-style buddha epithets), the six-realm scope of recipients, and even the female-to-male transformation motif. The *Lotus*'s single genuine innovation is a temporal reversal of the Āgama's most damning negative prognosis — the "one kalpa in Avīci hell, incurable" verdict pronounced upon Devadatta in the early canon — and this one innovation falls precisely on the Devadatta wing of our symmetrical structure.

**Thesis III** — The Devadatta Chapter is the first convergence point in Mahāyāna doctrinal history between the *tathāgatagarbha* 如來藏 track and the Yogācāra 唯識 track. The Yogācāra track, as developed by Vasubandhu in the *Saddharmapuṇḍarīka-upadeśa* 法華經論 (T1519) and systematized by Kuiji 窺基 in the *Fǎ-huá-xuán-zàn* 法華玄贊 (T1723), carries the Dragon-girl wing through the doctrine of "seed-as-statement" (*bīja-vyākaraṇa* as pragmatic speech-act): a quantitative compression of temporal and spatial coordinates without any change in the underlying causal grammar. The *tathāgatagarbha* track, grounded in the *Mahāparinirvāṇa-sūtra* 涅槃經 (T0374) and the fundamental-gate (*mūla-dvāra*, 本門) of the *Lotus*, carries the Devadatta wing through the doctrine of *dharmakāya* equality: a *qualitative* directional reversal of the original Āgama verdict. Following Jizang's 吉藏 Madhyamaka warning in the *Fǎ-huá-yì-shū* 法華義疏, this paper revises its working title from the earlier draft's "limit-case prophecy" to "non-self-existent prophecy displayed at both ends" (*niḥsvabhāva-vyākaraṇa-ubhayānta-darśana*, reconstructed), thereby preventing the language of seeds from being ontologized into determinism.

The paper's original contribution is twofold. First, it reads the two wings of the Devadatta Chapter as *belonging to two distinct doctrinal tracks* for the first time in the secondary literature, rather than treating them as a single miracle-pair. Second, on the basis of this dual-track reading, it formally upgrades the first thesis of the entire S7 series: the *Lotus Sūtra* is not merely the "narrative form of Yogācāra" (as S7-P02 tentatively proposed), but the *dual-track confluence* of *tathāgatagarbha* and Yogācāra.

**Keywords:** Devadatta Chapter, Dragon-Girl's Buddhahood, Doctrine of Prophecy (*vyākaraṇa*), Seed-as-Statement, *Tathāgatagarbha*, Yogācāra, Vasubandhu's *Saddharmapuṇḍarīka-upadeśa*, Kuiji's *Fǎ-huá-xuán-zàn*, *Mahāparinirvāṇa-sūtra*, Jizang's Madhyamaka Warning, Dual-Track Confluence.

---

## §0. Opening: Why the Devadatta Chapter Is the True Touchstone

In S7-P02 ("*yī-dà-shì-yīn-yuán* 一大事因緣 · *kāi-shì-wù-rù* 開示悟入 · *huì-sān-guī-yī* 會三歸一"), we argued that the *Lotus*'s reintegration of the three vehicles into the One Buddha-Vehicle is not the additive juxtaposition of three parallel paths but a *closed awakening loop* — four verbs (open, reveal, awaken, enter) contracted into a single event-structure. That paper, however, deliberately left one thread dangling. The real touchstone of the One Vehicle, we claimed, is not the *śrāvaka* prophecies. Prophecies of *śrāvakas* had precedent already in the Āgama corpus, and the *Lotus*'s prophecies of Śāriputra, Mahākāśyapa, and Maudgalyāyana are, in terms of surface grammar, little more than extensions of the Āgama Maitreya-prophecy formula (see §II below).

The true touchstone lies at the two extreme limbs of the map of sentient beings: the *most evil male* and the *most formally disqualified female* — Devadatta and the Dragon-girl. If these two limbs can both be absorbed within a single prophecy-event, then the "three returning to one" has reached its maximum radius; if not, then the One Buddha-Vehicle amounts to no more than a rhetorical merger of *śrāvaka*, *pratyekabuddha*, and *bodhisattva* vehicles, leaving the outer perimeter of sentient beings untouched.

Historically, readers of the Devadatta Chapter have typically adopted one of two strategies. The **first is the moral-evasion route**: Devadatta is read as a "manifestation-body" (*nirmāṇakāya*) of a bodhisattva who merely displayed the form of an evil man, and the Dragon-girl as a long-cultivated *anāgāmin*-candidate merely displaying the form of a young girl. The underlying unity, on this reading, has been buddhahood from the start, and the apparent contradiction between karmic retribution and bodily form dissolves into prior attainment. Both Tiantai and Faxiang exegetes take this path, with differing degrees of argumentative rigor (see §§III–IV). The **second is the text-critical route**: the chapter is shown to lack independent status in Dharmarakṣa's and Jñānagupta's recensions, and is therefore dismissed as a late accretion whose theological weight may be discounted. This second route is favored by modern Euro-American scholarship.

This paper argues that both routes are too conservative. The first depends on an ontological presupposition — the real positing of an originally-enlightened bodhisattva persona behind the appearance — which Madhyamaka four-cornered (*catuṣkoṭi*) analysis will dismantle (see §V). The second misreads the documentary record: Dharmarakṣa and Jñānagupta are not witnesses to a text *without a Devadatta chapter*; they are witnesses to a text in which the Devadatta prophecy and the Dragon-girl's buddhahood are *already joined together as the symmetrical wings of the Stūpa Chapter*. The Kumārajīva recension represents not a *later insertion* but a *disaggregation* of this original unit; the subsequent restoration of "Devadatta Chapter 12" into Kumārajīva's text is a recovery of a pre-existing narrative configuration, not the fabrication of a new one (§I).

Once both misreadings are cleared, the true architectonics of the Devadatta Chapter surface. The chapter is a *simultaneous display of the two poles* of the *Lotus*'s prophecy theory: at one pole stands the limit of karmic retribution (Devadatta, guilty of five heinous acts — *pañcānantarya* — receives the prophecy of buddhahood as *Devarāja Tathāgata*); at the other pole stands the limit of bodily form (the Dragon-girl, bearing three layers of canonical disqualification under the *Vinaya* — non-human, female, and under twenty years of age — attains buddhahood in the present body, *iha-janmani*). These are not two independent miracles. They are the symmetrical wings of a single structure. The thesis to be argued here is that this structure is read in the Vasubandhu–Kuiji Yogācāra lineage as a pragmatic *seed-statement* and in the *Mahāparinirvāṇa*–*Lotus*-fundamental-gate *tathāgatagarbha* lineage as the direct manifestation of *dharmakāya*-equality, and that these two readings *first converge* in the Devadatta Chapter. The doctrinal significance of this convergence-event far outweighs the compilation-history question of whether the chapter stands alone.

The sequence of our argument: documentary triangle (§I) → continuity of Āgama prophecy grammar and its single break-point (§II) → Vasubandhu–Kuiji Yogācāra confluence (§III) → Zhiyi's perfect-teaching support (§IV) → Madhyamaka warning and title revision (§V) → *Mahāparinirvāṇa* cross-scriptural parallel and dual-track declaration (§VI) → the triple disqualification in the *Vinaya* and its structural meaning (§VII) → Middle-Way corrections (§VIII). One prefatory note: this paper adopts "two ends simultaneously displayed" (*ubhayānta-darśana*) rather than "limit-case prophecy" as its core formulation, absorbing Jizang's Madhyamaka-style warning in the *Fǎ-huá-yì-shū* that the word "limit" (極限), if ontologized, converts prophecy into determinism and thereby violates the *Lotus*'s foundational position of "*upāya* without fixity" (方便無定). See §V.2 for the full argument.

---

## §I. The Documentary Triangle: Kumārajīva's Recension Disaggregated an Originally Unified Narrative Unit

To understand the Devadatta Chapter, one must first dispose of a widely circulated but misleading text-critical claim: that "the Devadatta Chapter is a later interpolation into Kumārajīva's recension, and its theological weight is therefore suspect." This claim is half correct and half misleading. **The correct half**: the "Devadatta Chapter 12" found in the extant Taishō recension of Kumārajīva's *Miào-fǎ-lián-huá-jīng* 妙法蓮華經 (T0262) is not part of the 406 CE original translation. It was restored, around 480 CE, by the monks Faxian 法獻 and Fayi 法意, as recorded in Sengyou's *Chū-sān-zàng-jì-jí* 出三藏記集 (fascicle 8). **The misleading half**: this restoration event has been read as *adding new content where none previously existed*, whereas the documentary record shows precisely the opposite — it was a *recovery of content already present in earlier Chinese translations*.

### §I.1 Dharmarakṣa's Recension (T0263, 286 CE): Devadatta and the Dragon-Girl Jointly Contained Within the Stūpa Chapter

Dharmarakṣa's 竺法護 *Zhèng-fǎ-huá-jīng* 正法華經, translated in 286 CE — fully 120 years before Kumārajīva — is the earliest extant complete Chinese translation of the *Lotus Sūtra*. In its sixth fascicle, the latter half of the *Qī-bǎo-tǎ-pǐn* 七寶塔品 (Chapter of the Seven-Jewel Stūpa, corresponding to Kumārajīva's *Jiàn-bǎo-tǎ-pǐn* 見寶塔品) contains *both* the Devadatta prophecy passage and the Dragon-girl's buddhahood passage, neither of which is set apart as an independent chapter. In the 286 CE Chinese translation tradition, then, the Devadatta prophecy and the Dragon-girl's buddhahood were *never* independent chapters; they were *internal components* of the larger "Stūpa-seeing" narrative unit.

### §I.2 The Supplemented Recension (T0264, 605 CE): Joined, Not Independent

The *Tiān-pǐn-miào-fǎ-lián-huá-jīng* 添品妙法蓮華經, translated by Jñānagupta 闍那崛多 and Dharmagupta 達磨笈多 in 605 CE, was explicitly designed as a supplement to Kumārajīva's recension, restoring material the translators judged to have been omitted; the preface to the supplemented recension states this openly.[^2] Notably, the supplementation work is extraordinarily fine-grained: the translators restored the latter half of the *Medicinal-Herbs Parable* chapter, the opening of the *Pūrṇa* and *Dharma-Teacher* chapters, the verse portion of the *Universal Gate* chapter, the sequence of *dhāraṇīs*, and so on. Yet in all this meticulous supplementation work, *they did not establish the Devadatta material as an independent chapter*. Instead, following Dharmarakṣa's layout, they placed the Devadatta passage and the Dragon-girl passage together at the end of the *Stūpa-Seeing* chapter.

### §I.3 Guanding's *Fǎ-huá-wén-jù*: The Tiantai Tradition's Own Explicit Acknowledgement

The third vertex of the documentary triangle is internal to the Tiantai school itself. Guanding 灌頂 (561–632), Zhiyi's 智顗 direct disciple and the principal compiler of the *Fǎ-huá-wén-jù* 法華文句 (T1718), openly records that the Devadatta Chapter in the Kumārajīva recension was not present in the original translation and was later supplemented. This is not a claim made by a hostile critic or a text-critical rival; it is a statement made from *within* the tradition that most prized the *Lotus Sūtra* as the highest teaching. If any school had incentive to suppress the restoration-history of the Devadatta Chapter in order to defend the integrity of Kumārajīva's recension, it would have been Tiantai. The fact that Tiantai's own foundational commentary openly records the restoration is decisive historical evidence.

### §I.4 Conclusion: The Symmetrical Wings Are the Original Textual Configuration

The three witnesses — Dharmarakṣa's 286 CE recension, the Jñānagupta–Dharmagupta 605 CE supplemented recension, and Guanding's Tiantai commentary — are independent of one another in origin, language-community, and doctrinal allegiance, yet all three converge on a single conclusion: the original configuration of this portion of the *Lotus Sūtra* places the Devadatta prophecy and the Dragon-girl's buddhahood *together within the Stūpa-Seeing narrative unit as symmetrical wings*. Kumārajīva's 406 CE recension disaggregated this unit (whether by translational choice or by reliance on a different Sanskrit manuscript tradition); Faxian and Fayi's 480 CE supplementation restored it. "Devadatta Chapter 12" is therefore not an apocryphal insertion but the recovered trace of a unified narrative structure whose theological weight — as we shall argue in §§III–VI — is precisely the weight of the *Lotus*'s dual-track doctrinal confluence.

This conclusion has a consequence for how the two wings must be read. Because they were originally joined as components of a single narrative unit, their doctrinal meaning cannot be extracted from either wing *in isolation*. The Devadatta wing read alone produces moral-evasion readings; the Dragon-girl wing read alone produces gender-apologetic readings; only read *together*, as the symmetrical two-pole display of a single prophecy structure, do they yield the doctrinal result this paper is concerned to establish.

---

## §II. Continuity with Āgama Prophecy Grammar and the Single Break-Point

One of the most persistent misconceptions in the secondary literature on the *Lotus Sūtra* is that the doctrine of *vyākaraṇa* (prophecy of future buddhahood) is a Mahāyāna innovation, without precedent in the early canon. This is simply false. The grammar of prophecy — the lexicon, the formulaic structure, the range of recipients, and even the body-transformation motif — was already fully in place in the Āgama corpus by the time the earliest strata of *Lotus Sūtra* material began to circulate. What the *Lotus* contributes is not the *invention* of *vyākaraṇa* but a single, sharply localized *reversal* of one of its standard negative applications. Ninety-five percent of the *Lotus*'s prophecy material is continuous with the Āgama; the remaining five percent, the single break-point, falls entirely on the Devadatta wing. This asymmetry is itself evidence for the structural reading advanced in this paper.

### §II.1 Sūtra Layer: The Formal Components of Āgama Prophecy Are Already Fully in Place

Four formal components of the *Lotus*'s prophecy grammar can be traced directly to the Āgama corpus.

**(a) The nine-field structure.** Every full prophecy in the *Lotus*, whether of Śāriputra, Mahākāśyapa, or the Dragon-girl, displays the same nine-field structure: (1) name of the recipient; (2) future aeon in which buddhahood will be attained; (3) name of the future buddha-land; (4) name of the future buddha; (5) qualities of the future buddha-land (jewel-composition, absence of evil destinies, etc.); (6) length of the future buddha's life-span; (7) length of the true-Dharma period; (8) length of the semblance-Dharma period; (9) identity of successor-bodhisattvas. This nine-field structure is not a Mahāyāna invention. It is already attested in the Āgama Maitreya-prophecy — the *Mílè-xià-shēng-jīng* 彌勒下生經 and related texts — where all nine fields appear in essentially the same order. The *Lotus* inherits the form wholesale.

**(b) The ten-epithet formula.** Every future buddha prophesied in the *Lotus* is introduced with the standard ten-epithet string: *tathāgata, arhat, samyak-saṃbuddha, vidyā-caraṇa-saṃpanna, sugata, loka-vid, anuttara-puruṣa-damya-sārathi, śāstā-deva-manuṣyāṇām, buddha, bhagavat*. This ten-epithet formula is the Āgama's own standard buddha-naming device; it was not created for Mahāyāna use.

**(c) The six-realm scope of recipients.** The Āgama canon already contains prophecies directed at non-human recipients — *nāgas*, *yakṣas*, *gandharvas*, celestial beings — in both positive (favorable rebirth) and negative (prognostic condemnation) forms. The *Lotus*'s extension of *vyākaraṇa* to a dragon-princess does not cross a species-boundary that the Āgama had kept closed; it extends a practice already in use.

**(d) The female-to-male transformation motif.** The Dragon-girl's bodily transformation from female to male prior to displaying the act of buddhahood is often read as a distinctively Mahāyāna doctrinal move. In fact, the motif is attested in the Āgama canon in several places, including passages in the *Saṃyuktāgama* and in certain *Jātaka*-adjacent materials, where female practitioners' bodies undergo transformation at the moment of decisive attainment. The *Lotus* does not invent this motif; it redeploys it in a new theological context.

### §II.2 Treatise Layer: Prophecy as a Canonical Category within the Twelve-fold Division of Teaching

At the treatise layer, *vyākaraṇa* (授記, 記別) is not an occasional device; it is one of the twelve formal categories of the Buddha's teaching (*dvādaśāṅga-pravacana*, 十二分教) recognized by all Abhidharma schools. In that canonical taxonomy, *vyākaraṇa* covers both the positive sense (prophecy of future buddhahood or favorable rebirth) and the negative sense (declaration of future destination in the lower realms). The category is, in short, a neutral grammatical slot that can be filled with either a blessing or a condemnation, according to the ethical record of the recipient. This treatise-layer point is essential for what follows in §II.3, because the *Lotus*'s single break-point is the *reversal of the contents of the slot*, not the invention of the slot itself.

### §II.3 The Single Break-Point: The Devadatta Reversal

Here we come to the one place where the *Lotus*'s prophecy material genuinely departs from its Āgama inheritance. The Āgama corpus contains a pronouncement — attested in multiple *Vinaya* and *Sūtra* sources — that Devadatta, having committed the five heinous crimes (*pañcānantarya*: injuring a buddha, causing schism in the saṅgha, and so on), will fall into Avīci hell for the duration of one full kalpa, and that his condition there is "incurable" (*atikitsya*, 不可療治). This is a *vyākaraṇa* in the strictly canonical sense: it fills the grammatical slot with a condemnatory content, pronounced with full formal authority.

The Devadatta Chapter of the *Lotus* does something unprecedented within this grammar: it *reverses the temporal direction of this very verdict*, without denying that the verdict was formally correct at the time of its pronouncement. Devadatta will indeed suffer; the Āgama verdict is not retracted. But at some point in the far future, after immeasurable aeons, Devadatta will become a buddha named *Devarāja Tathāgata* (天王如來), with a buddha-land, a life-span, and all nine fields of the standard prophecy structure intact. What has changed is not the *structure* of the prophecy — that structure is inherited directly — but the *content of the slot*: a condemnation has been overwritten by a blessing, and the mechanism of this overwriting is what the next four sections of this paper will unfold.

### §II.4 Summary: 95% Continuity, 5% Innovation, and the 5% Falls Entirely Here

The four formal components surveyed in §II.1, together with the canonical standing of *vyākaraṇa* in §II.2, constitute the bulk of the *Lotus*'s prophecy grammar — conservatively, ninety-five percent of the doctrinal surface. The single break-point examined in §II.3 is the remaining five percent. What the structural-reading argument of this paper will show is that this five percent is not scattered across the text; it is concentrated *entirely* on the Devadatta wing of the symmetrical two-wing structure identified in §I. The Dragon-girl wing, by contrast, is *grammatically* continuous with the Āgama — the female-to-male transformation motif, the bodily display at the moment of attainment, all of this has Āgama precedent. The Dragon-girl wing's innovation lies elsewhere: in the *compression of temporal and spatial coordinates*, which §III will identify as the Yogācāra "seed-statement" reading. The Devadatta wing's innovation lies in the *directional reversal* of an existing verdict, which §VI will identify as the *tathāgatagarbha* reading.

The upshot: the two wings are not innovative in the same way. One is innovative *quantitatively* (within a grammar inherited intact), the other *qualitatively* (by reversing a verdict pronounced within that same grammar). This asymmetry is the hinge on which the dual-track reading turns.

---

## §III. Vasubandhu and Kuiji: A Yogācāra Confluence Already Systematized in the Seventh Century

To read the Devadatta Chapter as a "seed-statement" is, in the current exegetical landscape, a relatively novel formulation. As an *argumentative structure*, however, it is not a modern creation. Between Vasubandhu's (c. 4th–5th century) *Saddharmapuṇḍarīka-upadeśa* 法華經論 (T1519/T1520) and Kuiji's 窺基 (632–682) *Fǎ-huá-xuán-zàn* 法華玄贊 (T1723), there runs a clearly traceable line of exegetical transmission. By the seventh century, the Chinese Faxiang school had already integrated the *Lotus Sūtra*'s prophecy-events into the framework of *ālaya-vijñāna* seed theory in a systematic way. The task of this section is to make that line of transmission visible again, and thereby to establish that the "seed-reading" advanced here is not a foreign framework imposed on the *Lotus* but a historically grounded traditional reading that was once — and can again be — a live option within the commentarial mainstream.

### §III.1 Vasubandhu's *Saddharmapuṇḍarīka-upadeśa*: The Three-Layer Theoretical Definition of Prophecy

Vasubandhu's treatise on the *Lotus* is extraordinarily brief, and its direct commentary on the Devadatta Chapter consists of only two sentences. But these two sentences carry very high theoretical density. First, on the Devadatta prophecy:

> "The Tathāgata's granting of a separate prophecy to Devadatta *displays that the Tathāgata bears no enmity or hatred*." (T1519, lower fascicle)

And on the prophecies directed to the *bhikṣuṇīs* and celestial women (including the Dragon-girl):

> "The granting of the buddhahood-prophecy to the *bhikṣuṇīs* and celestial women *displays that women, whether householders or renunciants, can through bodhisattva practice fully realize the fruit of buddhahood*; for this reason a prophecy is granted." (T1519)

Vasubandhu offers no exegesis whatsoever on the dramatic surface details — the "buddhahood in the time of a single moment," the "transformation into a male body," the doctrine of the "five obstructions" (*pañca-āvaraṇa*, 五障). What he provides instead is a *conclusion-style theoretical definition*: the referent of the Devadatta prophecy is "the Tathāgata is without enmity," and the referent of the Dragon-girl prophecy is "the female body can also realize the fruit of buddhahood." In other words, on Vasubandhu's reading, the core meaning of these two events lies not in the miraculous character of the events themselves but in the universal propositions they each assert concerning *buddha-nature* and *dharmakāya*.

More crucial still is Vasubandhu's three-layer theoretical definition of the concept of *vyākaraṇa* itself, set out in the prophecy section of the treatise's lower fascicle:

1. **Prophecy = the declaration of three equalities** — equality of vehicle (*yāna-samatā*), equality of *nirvāṇa*, and equality of body (*kāya-samatā*). The referent of *vyākaraṇa* is the *present fact* of *dharmakāya*-equality.
2. **Prophecy = the attainment of fixed resolve** — the function of *vyākaraṇa* for its recipient is to generate the irreversible resolve (*avaivartika-citta*).
3. **Prophecy ≠ already possessing completed merits** — "because of *dharmakāya*-equality a prophecy is granted, *not* because the practice-merits are already complete." The bodhisattva's merits must still be cultivated by the recipient; the prophecy does not mean that practice is finished.

The third point is the strongest piece of evidence Vasubandhu offers. Prophecy is *not a prediction* (because the recipient does not yet possess the completed merits), and it is *not an exemption* (because the merits must still be cultivated). It is the *declaration of a present fact* — the fact of *dharmakāya*-equality. This structure is precisely the classical version of what contemporary readings call a "seed-statement": the prophecy states the *present state* of a seed (that it has matured sufficiently to bear the declaration), rather than predicting what fruit that seed will eventually produce.

### §III.2 Seven Parables vs. Six Prophecy-Loci: Devadatta and the Dragon-Girl Fall *Outside* the Parable System

Vasubandhu's treatise erects two parallel analytic frameworks over the *Lotus*: the **seven parables as antidotes to seven kinds of pride** (*sapta-dṛṣṭānta*, mapped onto seven chapters), and the **six loci of prophecy** (六授記處). Noteworthy is the fact that the Devadatta Chapter and its Dragon-girl sub-narrative *do not appear anywhere within the seven-parable system*. Vasubandhu classifies them instead under the six prophecy-loci — specifically, as the fourth, "prophecy-without-enmity" (無怨記) for Devadatta, and as a distinct "prophecy-locus for celestial women" for the Dragon-girl.

This taxonomic judgment has significant argumentative weight. Vasubandhu *does not read Devadatta and the Dragon-girl as parables* — because if they were parables, their miraculous character would be absorbed into literary technique and their theoretical weight correspondingly reduced. Instead, he reads them as *special boundary-cases of prophecy theory*, explicitly deployed to mark the two extreme application-contexts of the concept of *vyākaraṇa*: the limit of evil and the limit of weakness. Vasubandhu does not use the word "limit"; but the structural move of extracting these two cases from the parable system and placing them at the boundary position of the prophecy system is itself sufficient to establish their status as "boundary tests of prophecy theory."

### §III.3 Kuiji's *Fǎ-huá-xuán-zàn*: A Full *Ālaya-Vijñāna* Framework Takes Over

Kuiji's *Fǎ-huá-xuán-zàn*, in its third fascicle commenting on the *Upāya-kauśalya* Chapter, has already systematically established an interpretive framework of "*ālaya-vijñāna* = storehouse container + uncontaminated seeds (*anāsrava-bīja*) + gradual perfuming (*vāsanā*)." This is not an incidental mention; it is the theoretical ground-bass on which the entire commentary rests. By the time this framework reaches the ninth fascicle's commentary on the Devadatta Chapter, all descriptions of prophecy-events sit on that ground, and the reading is correspondingly transparent.

On the Dragon-girl passage, Kuiji's approach is **display-theory plus indeterminate-lineage manifestation** (*aniyata-gotra nirmāṇa*):

> "The present Dragon-girl may be a bodhisattva of the fourth-stage resolve, who has manifested herself as a dragon, as a girl, as one who though young is able to learn the *Lotus* and swiftly attain *bodhi*. This is to encourage the many — *it is not literally so*." (T1723, upper fascicle nine)

"Fourth-stage resolve" (第四發心) here denotes the position of *eka-jāti-pratibaddha* — a bodhisattva one lifetime away from buddhahood, in the tenth *bhūmi*. Kuiji's reading is: the Dragon-girl is already a high-stage bodhisattva who *manifests* as a dragon, as a female, as one eight years old. The event itself is *not literally true* (非為實爾); it is a pedagogical display to encourage the assembly. What is described as "buddhahood in the time of a single *kṣaṇa*" is, within Kuiji's theoretical framework, simply the inevitable outcome of uncontaminated seeds that have already reached full maturity and now await only their final moment of actualization. It is not a miracle; it is a standard derivation from seed theory.

On the Devadatta prophecy itself, Kuiji opens with a "three-gate analysis" (origin, terminology, objection-resolution). The first gate quotes Vasubandhu directly:

> "First, the *treatise* explains the fourth of the six prophecies, the 'prophecy-without-enmity,' as follows: 'The granting of a prophecy to Devadatta displays that the World-Honored One bears no enmity.' Hence this chapter." (T1723, upper fascicle nine)

This is the most direct documentary evidence of the Vasubandhu → Kuiji transmission. Kuiji is not inventing a "seed-reading" from whole cloth; he is *inheriting Vasubandhu's prophecy-classification system* (the six loci) and using his own *ālaya-vijñāna* framework to deepen the theoretical substrate of that inherited classification.

On the *Lotus* passage "through Devadatta as spiritual friend (*kalyāṇa-mitra*) I have completed [the perfections] … and attained perfect awakening," Kuiji reads:

> "Because of his injurious actions, he has sunk in the present — but *through the power of the sūtra* he will attain the supreme fruit." (T1723, upper fascicle nine)

Unpacked within Kuiji's seed-theoretic framework: "sinking in the present" refers to the *manifestation-level* ripening of the heinous-crime karma into *vipāka-phala* — the Avīci hell result that cannot be abolished; "through the power of the sūtra" refers to the *Lotus* itself as a perfuming force (*vāsanā-bala*) that activates the uncontaminated seeds; "will attain the supreme fruit" refers to the necessary consequence of seed-to-actualization (*bīja → pravṛtti*). The manifestation-level karmic-retribution system and the seed-level uncontaminated-seed system are *two independent structures*. The five heinous karmas continue to bear their results at the manifestation level (hell must be endured), while the uncontaminated seeds continue to be perfumed at the seed level (future buddhahood is secured). The two do not cancel each other out; they hold simultaneously.

This two-layer structure is the theoretical basis of the principle "prophecy does not abolish karma" (*vyākaraṇa* does not annul *karma-vipāka*; see §VIII.2 below). It is also Kuiji's standard answer to the perennial problem of how *vyākaraṇa* and *karma-vipāka* can coexist.

### §III.4 Summary: The Seventh-Century Confluence Is Not a Modern Imposition

Placing Vasubandhu's two theoretical definitions alongside Kuiji's entire commentary makes one fact visible: the reading "*Lotus Sūtra* prophecy = seed-statement" was *already systematically in place within the seventh-century Faxiang school*. It is not a retrospective reconstruction by modern exegetes (this paper included); it is a traditional reading with documentary transmission, textual evidence, and systematic argumentation of its own.

The contemporary task is merely to make this transmission visible again and to articulate its argumentative mechanism using the modern vocabulary of speech-act pragmatics. Between Vasubandhu and the present lie thirteen hundred years of commentarial history, but the theoretical structure is continuous.

One point must be held back for the following sections: the Yogācāra reading, powerful though it is, contains two structural blind-spots — one arising from the tension with Tiantai perfect-teaching (§IV), and one arising from the Madhyamaka four-cornered critique (§V). Only after these two tensions are addressed can §VI declare the full dual-track confluence of *tathāgatagarbha* and Yogācāra.

---

## §IV. Zhiyi's Perfect-Teaching Support: Three Directly Citable Arguments

Zhiyi's 智顗 (538–597) commentary on the Devadatta Chapter is distributed principally across the *Fǎ-huá-wén-jù* 法華文句 and the *Fǎ-huá-xuán-yì* 法華玄義. This section does not attempt a systematic treatment of Zhiyi's full reading.[^1] Instead, it extracts three specific Tiantai arguments that bear most directly on this paper's "two-ends simultaneously displayed" thesis.

**First, the five-obstructions doctrine is judged to be a "Tripiṭaka-teaching provisional objection."** In his commentary on Śāriputra's challenge to the Dragon-girl, Zhiyi explicitly categorizes the doctrine of "the five obstructions of women" (女人五障) as "Śāriputra invoking a Tripiṭaka-teaching provisional objection" — that is, the doctrine belongs to the Tripiṭaka (*śrāvaka*-vehicle) teaching as a conventional objection-tool, not to the perfect-teaching (*yuán-jiào* 圓教) as a final position. Its function within the narrative is to serve as a foil that the Dragon-girl's "single real dispels doubt" then overturns. This doctrinal classification is decisive: **Tiantai has never treated the five-obstructions doctrine as a universal pronouncement on the female body**. It reads it as a *śrāvaka* provisional teaching deployed precisely so that the perfect teaching can demolish it. When §VIII.3 below addresses "the Dragon-girl's sex-change ≠ a gender-ontological claim," we can cite this Tiantai classification directly as traditional backing.

**Second, Devadatta and the Dragon-girl are explicitly named together as the two poles of perfect-integration.** In the prophecy section of the *Fǎ-huá-xuán-yì*, Zhiyi juxtaposes in a single breath: "*the five-heinous-crime Devadatta likewise receives a prophecy; dragons and animals likewise receive prophecies*." This is not a modern reader's after-the-fact pairing; it is the Tiantai patriarch himself, in the body of the commentarial literature, **reading the two poles as the symmetrical wings of a single structure**. This aligns precisely with the §I documentary-triangle finding: *before* Kumārajīva's disaggregation, the two passages already stood within a single chapter. Zhiyi's perfect-teaching reading and the original textual configuration preserved in Dharmarakṣa and the supplemented recensions converge in structural intuition — both read the two passages as *co-constituted wings* rather than as two independent miracles.

**Third, the metaphor of "transforming poison into medicine / causing spoiled seeds to flower into buddhahood."** Zhiyi repeatedly employs, in the *Fǎ-huá-xuán-yì*, the images of "the supreme King of Physicians transforms poison into medicine, capable of healing spoiled seeds into mindless buddhahood" and "to receive the prophecy in the present sūtra is to transform poison into medicine" as descriptions of the *Lotus*'s prophecy-function. The argumentative weight this medical metaphor carries within Tiantai's reading is precisely the weight that modern "seed-statement" readings seek to carry at the speech-act-pragmatic level: the reactivation, within a particular declarative event, of seeds that had previously been rendered incapable of maturation (*paryudāsa-bīja*, "spoiled seeds"). The *route* of the metaphor differs from the route of this paper, but the *referent event* is the same.

All three of these points constitute direct Tiantai support for the thesis of this paper. What must be reserved for the next section is this: on the specific question of "the Dragon-girl's swift buddhahood = *ultimate* buddhahood / single-thought sitting on the seat of awakening," Zhiyi adopts a *literal-attainment reading* (the Dragon-girl factually became a buddha, not a seed-statement declaration), which diverges in doctrinal direction from this paper's seed-statement reading. Our handling of this divergence is given in footnote 1.

[^1]: **Directional divergence between the Tiantai literal-attainment reading and this paper's seed-statement reading.** In his commentary on the Devadatta Chapter in the *Fǎ-huá-wén-jù*, Zhiyi classifies the Dragon-girl's "buddhahood in a single *kṣaṇa*" as either "the initial-resolve stage of the perfect teaching" (*Fǎ-huá-xuán-yì*) or "the ultimate buddhahood — single-thought seat of awakening" (*Fǎ-huá-wén-jù*). His route is to establish *factual* buddhahood by appeal to the doctrines of "three-thousand worlds in a single thought" and "the principle is inherent, the phenomena arise." This paper's seed-reading, by contrast, offers a speech-act-pragmatic analysis of the prophecy-*speech-event* itself, and neither asserts nor denies the ontological-level question of factual attainment. The divergence between the two routes is rooted in the fact that Tiantai appeals to a "*nirmāṇakāya* bodhisattva" (a real foundational persona) as its reconciliation device, whereas this paper adopts a *depersonalized* seed-statement (a speech-act-level declaration of seed-status). This paper does not deny the doctrinal value of Tiantai's reading; it simply states that, on the narrow question of the formal analysis of "prophecy" as a concept, it takes the second route. Readers wishing to compare in depth may consult *Fǎ-huá-wén-jù* fascicle 8 (lower) alongside *Fǎ-huá-xuán-yì* fascicle 6 (lower).

---

## §V. The Madhyamaka Warning and the Title Revision

If Tiantai offers this paper "three supports plus one directional divergence," Sanlun's Jizang 吉藏 (549–623) offers it *a structural methodological warning*. Jizang's commentary on the Devadatta Chapter in the *Fǎ-huá-yì-shū* 法華義疏 is strikingly short (roughly 130 lines for the entire chapter), and *it uses no seed-language whatsoever*. That silence is itself an argumentative position.

### §V.1 Jizang's "Slow-and-Swift Tetralemma" and "Male-and-Female Tetralemma"

In handling the Dragon-girl's swift buddhahood, Jizang deploys the standard Madhyamaka four-cornered negation (*catuṣkoṭi*): *slow, swift, both-slow-and-swift, neither-slow-nor-swift*. The function of the tetralemma is to completely eliminate the status of "swift buddhahood" as a *svabhāvika* proposition: under the tetralemma, the Dragon-girl's single-*kṣaṇa* buddhahood and a bodhisattva's ten-kalpa seat-of-awakening are *both* merely *upāya*, with no hierarchy and no self-nature. By the same procedure, the "male-and-female tetralemma" completely eliminates the ontological difference of "transformation from female to male body."

Jizang's treatment of the Devadatta case is equally terse: "*Devadatta, not evil, displays evil; hence evil is the* upāya *and good is the real; this analogically illuminates Śākyamuni, neither arising nor ceasing, displaying arising and ceasing; hence arising-and-ceasing is the* upāya *and neither-arising-nor-ceasing is the real.*" The entire Devadatta–Dragon-girl event is absorbed, in Jizang's hands, by the single formulation "*upāya* without fixity, adaptation without direction" (方便無定、適化無方). Sanlun does not need to erect an independent doctrinal system for the Devadatta Chapter, because "*upāya* without fixity" already covers every extreme case by construction.

### §V.2 The Deconstructive Risk Posed by Seed-Language Under Madhyamaka Scrutiny

Jizang's silence constitutes a structural warning for the present paper. *If seed-language is adopted without qualification, prophecy becomes determinism.* If seeds possess self-nature (*svabhāva*), then the seed-statement reading reduces to a description of future necessity — which directly contradicts the *Lotus*'s foundational position of "*upāya* without fixity." Madhyamaka's deconstructive route is as follows: seed-statement → seeds possess *svabhāva* → prophecy becomes *niyata* (fixed determination) → contradicts *upāya*.

This warning must be taken seriously. The strategy adopted by the present paper is **parallel bilingualism**: whenever Yogācāra seed-language is used, the paper simultaneously declares that these seeds are *Madhyamaka-sense seeds* — empty in their own nature, without *svabhāva*, referred-to as statement-objects only in the pragmatic-linguistic sense, never as ontological entities. Vasubandhu's own third theoretical definition — "because of *dharmakāya*-equality a prophecy is granted, not because the merits are already complete" — provides the textual basis for this parallel bilingualism: the referent of the prophecy is the present fact of *dharmakāya*-equality, and *dharmakāya*-equality itself is, in Madhyamaka terms, a synonym for emptiness (*śūnyatā*).

### §V.3 The Title Revised from "Limit-Case" to "Two Ends Simultaneously Displayed"

It is precisely on the basis of this Jizang warning that the title of this paper has been revised at a crucial juncture: rather than "limit-case prophecy" (極限授記), the paper adopts "*non-self-existent prophecy displayed at both ends*" (授記無自性之兩端同時展示). The reasoning is twofold.

First, the word "limit" silently implies some measurable scale (a limit is a limit *relative to* something non-limit), and it is easily ontologized into "the furthest point." Madhyamaka deconstruction begins precisely there: if there is a "furthest," there is *svabhāva*; and if there is *svabhāva*, prophecy has already become determinism.

Second, "two ends simultaneously displayed" is a purely structural description that presupposes no ontological scale. It says only this: a single prophecy-event has two symmetrical display-faces (the face of evil and the face of weakness); these two faces were originally joined at the documentary level (§I); they carry asymmetric argumentative weight within the inherited Āgama format (§II.4); and within Vasubandhu's six-locus classification and Kuiji's manifestation-doctrine each is handled with its own independent treatment (§III). Their structural significance *as two wings* is the central question posed by the Devadatta Chapter.

The phrase "two ends simultaneously displayed" cannot be deconstructed by the Madhyamaka tetralemma — because it is, by construction, a purely structural claim. It presupposes no *svabhāva* for either end, and it presupposes no ontological hierarchy between "the ends" and "the middle."

---

## §VI. The Dual-Track Confluence: Cross-Scriptural Parallel in the *Mahāparinirvāṇa-sūtra* and the Formal Upgrade of Thesis I

The final positive argument of this paper requires a significant upgrade to Thesis I of the broader S7 series.

### §VI.1 The Reversal in the *Mahāparinirvāṇa-sūtra* (T0374)

Outside the *Lotus*'s own Devadatta Chapter, the *only* Mahāyāna scripture that pronounces an explicit reversal upon Devadatta is the *Mahāparinirvāṇa-sūtra*. T0374, fascicle 20, declares openly that Devadatta is "*neither an* icchantika*, nor a* śrāvaka*, nor a* pratyekabuddha *… but solely an object known and seen by the buddhas*" (T0374, fascicle 20).

This is an extraordinarily strong reversal. It not only overturns the Āgama verdict of "one kalpa in Avīci, incurable" (§II.3); it directly classifies Devadatta as a being within the cognitive range of buddha-knowledge-and-vision (*buddha-jñāna-darśana*). The *Mahāparinirvāṇa*'s route differs from the *Lotus*'s "ancient original-time / Ṛṣi Asita previous life" narrative route, but the directional conclusion is identical.

That two distinct Mahāyāna sūtras independently deliver reversals upon Devadatta in the same direction — each mobilizing its most powerful metaphysical resources (the *Lotus* drawing on the fundamental-gate's temporal ontology, the *Mahāparinirvāṇa* drawing on the *tathāgatagarbha* buddha-nature doctrine) — is a fact with major argumentative consequences. **The Devadatta reversal is not an isolated innovation internal to the *Lotus Sūtra*; it is a necessary entailment of Mahāyāna Buddhism's broader metaphysical reorganization.** Put otherwise: the argumentative weight of this reversal cannot be carried by the *Lotus* alone — it must appeal to Mahāyāna's overall upgrade of the ontology of *dharmakāya* and *buddha-nature*.

### §VI.2 Formal Upgrade of Thesis I: From "Narrative Yogācāra" to "Dual-Track *Tathāgatagarbha* + Yogācāra Confluence"

In S7-P01 and S7-P02, the first thesis of the series was originally formulated as: **"The *Lotus Sūtra* = narrative Yogācāra."** That is, the events presented by the *Lotus* through narrative means are structurally equivalent to the seed-theoretic framework of the later Yogācāra treatises. This formulation was verified in S7-P02 (the closed awakening loop of "three returning to one") and reinforced again in §III of the present paper (the Vasubandhu → Kuiji exegetical transmission).

The *Mahāparinirvāṇa* evidence, however, forces us to concede that **while the Yogācāra reading is powerful, it cannot independently carry the full argumentative weight of the Devadatta reversal**. The two-layer structure of "uncontaminated seeds / manifest karma-fruit" (§III.3) can explain *why the five heinous karmas' hell-consequence does not obstruct future buddhahood*, but it cannot explain *why Śākyamuni is qualified, at this very moment, to pronounce the reversal-declaration itself*. The latter requires a *dharmakāya*-equality premise at the *tathāgatagarbha* level as its metaphysical ground — a premise which the *Mahāparinirvāṇa* supplies explicitly, and which the *Lotus*'s fundamental-gate supplies through its "ancient, truly-attained" (*cirantana-satya-siddha*, 久遠實成) temporal ontology.

This paper therefore formally proposes the following upgrade to S7's Thesis I:

**Thesis I (upgraded):** The *Lotus Sūtra* is **the first confluence point of the *tathāgatagarbha* track and the Yogācāra track**. The *tathāgatagarbha* track (*Mahāparinirvāṇa* T0374 + *Lotus* fundamental-gate) is responsible for *dharmakāya-equality, originally-already-buddhahood, and the metaphysical ground of the reversal*. The Yogācāra track (Vasubandhu T1519 + Kuiji T1723) is responsible for *seed-perfuming, the gradual actualization of manifest states, and the speech-act-pragmatic mechanism of prophecy-as-statement*. The two tracks **first converge in the Devadatta Chapter**: the quantitative wing (the Dragon-girl: temporal-spatial compression) is borne by the Yogācāra track, the qualitative wing (Devadatta: directional reversal) is borne by the *tathāgatagarbha* track, and the two wings are symmetrically displayed within a single prophecy-event, constituting the complete spectrum of the *Lotus Sūtra*'s prophecy theory.

This upgrade has retrospective implications for the thesis-statements of S7-P01 and S7-P02, which will be revised in subsequent work. One clarification is needed here: the upgrade is not a negation of the Yogācāra track. It is the acknowledgement that no single track can independently carry the full weight of the *Lotus*'s prophecy theory. The fact of dual-track confluence is itself the evidence of the *Lotus*'s distinctive position within the history of Mahāyāna metaphysics.

### §VI.3 The Structural Significance of "Two Wings Belonging to Two Tracks"

§II.4 noted that the two wings carry asymmetric argumentative weight: the Dragon-girl represents a *quantitative* innovation (a compression of what the Āgama already permitted in principle), while Devadatta represents a *qualitative* innovation (a directional reversal of what the Āgama had formally foreclosed). We can now supply the theoretical explanation for this asymmetry: **the two wings belong to two distinct tracks**.

The Dragon-girl wing can be resolved by the Yogācāra track alone — the seeds have long since matured and await only their moment of actualization; the compression of temporal and spatial coordinates is merely a question of actualization *speed*. The Devadatta wing, however, cannot be resolved by the Yogācāra track alone — the five heinous karmas cannot be *eliminated* at the manifestation level, and the reversal therefore requires appeal to the *tathāgatagarbha* track's doctrine of *dharmakāya*-equality. Kumārajīva's recension disaggregating the two wings (or, more precisely, the later restoration that re-established the symmetrical structure) reflects exactly this structural fact: although the two wings are narratively symmetric, they carry different theoretical loads and belong to different tracks.

This is the deepest meaning of the phrase "two ends simultaneously displayed": not two equivalent miracles, but the symmetrical confluence of two tracks within a single event.

---

## §VII. The *Vinaya*'s Triple Disqualification: The Final Piece of the Structural Picture

The *Mahāparinirvāṇa-sūtra* provides a Mahāyāna-level parallel reversal. The *Vinaya*, by contrast, provides pressure from the opposite direction. It lets us see *how severely sealed* the *Lotus*'s prophecy-declaration of unconditionality (*anāvaraṇatā*, 無遮) had to be in order to cut through the inherited tradition.

### §VII.1 Devadatta in the *Vinaya*'s Sealing

The *Sarvāstivāda-vinaya* (*Shísòng-lǜ* 十誦律, T1435), the *Dharmaguptaka-vinaya* (*Sìfēn-lǜ* 四分律, T1428), and the *Mahīśāsaka-vinaya* (*Wǔfēn-lǜ* 五分律, T1421) all record, in their respective "saṅgha-schism sections" (*saṃgha-bheda-skandhaka*), the three deeds of Devadatta: causing schism in the harmonious saṅgha, causing the Buddha's body to bleed, and dispatching the elephant to attack the Buddha. All three *Vinayas* assign him to the Avīci hell as a consequence. The *Vinaya* position on Devadatta is uniform and unambiguous: **no path of repentance or redemption is left open for him**. The *Lotus*'s declaration of the *Devarāja Tathāgata* prophecy, against this consensual *Vinaya* background, is a comprehensive Mahāyāna reversal of the *Vinaya* schools' own verdict.

### §VII.2 The Dragon-Girl's Triple Impossibility

The Dragon-girl's case is even more extreme. The *Vinaya*'s sealing against her is **threefold**:

1. **Dragon (animal-realm / non-human).** The *Sarvāstivāda-vinaya*, fascicle 21, records the incident of a *nāga* that transformed itself into human form in order to request ordination; the Buddha explicitly forbade the ordination of non-humans. Animal-realm beings and non-humans are, in the *Vinaya*, fundamentally excluded from the saṅgha.
2. **Female.** The reception of full ordination by female candidates is subject to the Eight Grave Conditions (*aṣṭau gurudharmāḥ*, 八敬法) and requires the presence of a dual saṅgha. While female candidates can receive ordination, follow the path, and realize the *arhat*-fruit, the proposition "a female body cannot become a buddha" was a widely held position within the *Vinaya* schools and the Abhidharma treatises (discussed further in §VIII.3).
3. **Eight Years of Age (Under Twenty).** The *Vinaya* explicitly requires that candidates for full ordination be at least twenty years old. An eight-year-old does not meet even the formal prerequisites for ordination, let alone for buddhahood.

**This is a structure of threefold impossibility.** The Dragon-girl is sealed out at three independent *Vinaya* levels simultaneously; any one of these levels alone would suffice to negate her candidacy for buddhahood. The *Lotus*'s Dragon-girl passage, within a single narrative, **crosses all three layers of sealing simultaneously**. The density of this narrative pressure is precisely what "the weak pole" of the two-ends-simultaneously-displayed structure means.

### §VII.3 The Stricter the *Vinaya*, the Stronger the Display of "Unconditionality"

These *Vinaya* regulations had their legitimacy and their function within the schools of the Sectarian period, and this paper intends no diminishment of the *Vinaya*'s standing. The point is purely structural: **the more severely the *Vinaya* sealed Devadatta and the Dragon-girl out, the more sharply the *Lotus*'s prophecy-declaration of unconditionality is thrown into relief by that contrast**. The *Lotus* does not abolish the *Vinaya* — the manifest karmic-retribution system continues to operate exactly as before (§III.3's two-layer structure) — and it does not deny the *Vinaya*'s sectarian legitimacy. Rather, on a different level (*dharmakāya*-equality + uncontaminated seeds), it declares a fact which the *Vinaya* level could never reach: **no sentient being's possibility of buddhahood can be completely sealed off by external conditions**.

This is the true limit-position of "three returning to one." If the "one" of the One Vehicle can accommodate only the human-and-divine path-followers of the favorable destinies, then it remains a conditional "one." Only when it can simultaneously accommodate the evil destinies and the forms of being foreclosed by the *Vinaya* does it become an unconditional "one." The two wings of the Devadatta Chapter display exactly these two boundary tests of unconditionality.

---

## §VIII. ❌ Middle-Way Corrections: The Four Most Dangerous Misreadings

The final section of this paper addresses the four misreadings of the Devadatta Chapter most easily produced in contemporary interpretation. Each is a deviation that has recurred throughout Mahāyāna history; each requires Middle-Way correction.

### ❌ VIII.1. Prophecy = Prediction of the Future

**The misreading.** Prophecy is the Buddha's prediction of future necessity, akin to an oracle or augury.

**The correction.** Vasubandhu's third theoretical definition in the *Saddharmapuṇḍarīka-upadeśa* has already foreclosed this misreading: "*because of dharmakāya-equality a prophecy is granted, not because the practice-merits are already complete*." What prophecy declares is the *present fact* of *dharmakāya*-equality — not a prediction of some future buddha-event at a specified time and place. The recipient's merits must still be cultivated (§III.1). The "prophecy = prediction" misreading converts the *Lotus* into fatalism and produces the nihilistic inference that "since prophecy has already been granted, why cultivate at all?" Vasubandhu sealed this off at the fourth or fifth century.

### ❌ VIII.2. Prophecy = Exemption from Karmic Retribution

**The misreading.** Since Devadatta receives the *Devarāja Tathāgata* prophecy, his five heinous karmas can be exempted; since the Dragon-girl attains buddhahood in her present body, all karmic retribution can be annulled by prophecy.

**The correction.** The Āgama provides the strongest evidence against this reading. The *Ekottarikāgama* 38.6 records the story of Aṅgulimāla: after realizing the *arhat*-fruit, Aṅgulimāla continued to be struck by the crowd in retribution for his earlier killing of ninety-nine people, and the Buddha spoke to him directly: "*Endure it now. The reason is this: this offense should by rights be received through aeons without end*" (T0125, fascicle 38). Liberation does not abolish karma — if even *arhat*-fruit cannot abolish karma, prophecy certainly cannot exempt the recipient from karmic retribution. Kuiji's two-layer structure (§III.3) states exactly the same position: the manifest karma-fruit system and the uncontaminated-seed system operate independently, and prophecy does not annul the former.

### ❌ VIII.3. The Dragon-Girl's Sex-Change = A Gender-Ontological Claim

**The misreading.** The Dragon-girl's "transformation into a male body" proves that the *Lotus* affirms "a female body cannot become a buddha," and is therefore scriptural evidence for gender-discrimination; or conversely, it proves that the *Lotus* supports a doctrine of "females must change into males in order to attain buddhahood."

**The correction.** The Āgama anticipates this point as well. *Saṃyuktāgama* 1199 records Somā Bhikṣuṇī's pronouncement: "*When the mind enters right concentration, what does a woman's form matter?*" (T0099, fascicle 45). The alternate-translation *Saṃyuktāgama* 215 is even stronger: "*If there are no marks of male and female, how can discrimination arise?*" (T0100, fascicle 11). Already in the Āgama period, the liberation experiences of female *bhikṣuṇīs* pointed unambiguously to the conclusion that bodily sex and awakening are *unrelated*. Zhiyi classifies the *Lotus*'s "transformation into a male" passage as *samanta-rūpa-samādhi* (the samādhi of universal manifestation of form) and a display of expedient power (§IV) — pointing in exactly the same direction: the sex-change is not an ontological transformation of sex, but a pedagogical display. Kuiji's "display-theory plus indeterminate-lineage manifestation" (§III.3) is the Yogācāra version of the same position. Three lineages — Āgama, Tiantai, and Faxiang — converge here: **the sex-change passage is not a gender-ontological claim, but a pedagogical display-form**.

### ❌ VIII.4. The Devadatta Reversal = Abolition of the Distinction Between Good and Evil

**The misreading.** Since Devadatta can be reversed and granted a prophecy, there is no real difference between good and evil; moral judgment can be abolished in Mahāyāna Buddhism.

**The correction.** The reversal occurs at the *dharmakāya* perspective (ancient original-time + *dharmakāya*-equality), not at the *karma-vipāka* perspective. At the *karma-vipāka* perspective, the five heinous karmas bear their results exactly as before; Devadatta experiences the hell-result at the manifestation level exactly as before. The *Lotus* does not deny the *karma-vipāka* system. It opens a deeper level *beyond* the *karma-vipāka* system, at which the *dharmakāya*-equality of "no enmity" (Vasubandhu's fourth no-enmity prophecy) holds. **The two-layer structure holds simultaneously**: this is not the abolition of the good/evil distinction, but the acknowledgement that *beyond* the distinction between good and evil there is a deeper level of equality. Reading the reversal as the abolition of good and evil is to compress the two-layer structure into a single layer — a betrayal of the *Vinaya* (§VII), of Vasubandhu (§III.1), and of the dual-track confluence structure declared in §VI.1.

---

## Conclusion

We return to the thread left dangling at the end of S7-P02: the true touchstone of "three returning to one" does not lie in the prophecies of the *śrāvakas*. After the eight sections of the present paper's argument, we are in a position to state the answer more precisely. The touchstone lies in whether the two ends — the pole of evil and the pole of weakness — can be absorbed within a single prophecy-event; and the *Lotus*'s answer to this question is not "by special exception" but "**this is how things have always been**."

"This is how things have always been" carries three layers of meaning. **At the documentary layer**: the two wings of the Devadatta Chapter were joined together in Dharmarakṣa's and Jñānagupta's recensions from the start; Kumārajīva's disaggregation was an editorial event, and the later restoration was a recovery (§I). **At the grammatical layer**: the *Lotus Sūtra*'s prophecy grammar is ninety-five percent continuous with Āgama prophecy grammar; the single genuine innovation is the temporal-directional reversal, and the metaphysical ground on which that reversal rests (*dharmakāya*-equality) has an independent cross-scriptural parallel in the *Mahāparinirvāṇa-sūtra* T0374 (§§II, VI). **At the theoretical layer**: the Devadatta Chapter is the first confluence point in Mahāyāna history between the *tathāgatagarbha* track and the Yogācāra track; the two wings belong to two tracks, and together they constitute the complete spectrum of the *Lotus Sūtra*'s prophecy theory (§VI).

The formulation adopted by this paper — "*non-self-existent prophecy displayed at both ends*" — is a terminological revision responding to Jizang's Madhyamaka warning (§V) and is, simultaneously, the simplest possible description of the Devadatta Chapter's actual structure. It asserts no "limit" (for under Madhyamaka scrutiny "limit" is *prapañca*); it asserts only a structural fact: **a single prophecy-event has two symmetrical display-faces, belonging to two distinct theoretical tracks, which first converge in this event**.

Following the present paper, the S7 series will proceed to the treatment of the *Encouragement to Uphold* and *Peaceful Conduct* chapters (S7-P09). If the Devadatta Chapter addresses the structural question of "whether prophecy can accommodate the two ends," then *Encouragement to Uphold* will address the practical question of "how the upholder of the sūtra carries the weight of those two ends through time." The formal analysis of prophecy theory must now be brought to ground in the temporality of sūtra-upholding practice.

---

## References

This paper does not rely on modern secondary scholarship for its citations. All arguments are grounded directly in the primary texts of the Taishō canon (see *PAPER_STYLE.md* §2.1 on citation principles).

### Sūtra & Śāstra Index

| Short Title | Full Title | Taishō № | Author / Translator |
|---|---|---|---|
| *Lotus Sūtra* | *Miào-fǎ-lián-huá-jīng* 妙法蓮華經 | T0262 | Kumārajīva |
| *Zhèng-fǎ-huá* | *Zhèng-fǎ-huá-jīng* 正法華經 | T0263 | Dharmarakṣa |
| *Tiān-pǐn Lotus* | *Tiān-pǐn miào-fǎ-lián-huá-jīng* 添品妙法蓮華經 | T0264 | Jñānagupta, Dharmagupta |
| *Dīrghāgama* | *Cháng-ā-hán-jīng* 長阿含經 | T0001 | Buddhayaśas, Zhu Fonian |
| *Madhyamāgama* | *Zhōng-ā-hán-jīng* 中阿含經 | T0026 | Saṃghadeva |
| *Saṃyuktāgama* | *Zá-ā-hán-jīng* 雜阿含經 | T0099 | Guṇabhadra |
| *Alt. Saṃyuktāgama* | *Bié-yì zá-ā-hán-jīng* 別譯雜阿含經 | T0100 | anonymous |
| *Ekottarikāgama* | *Zēng-yī-ā-hán-jīng* 增壹阿含經 | T0125 | Gautama Saṃghadeva |
| *Mahāparinirvāṇa* | *Dà-bān-niè-pán-jīng* (Northern recension) | T0374 | Dharmakṣema |
| *Sarvāstivāda-vinaya* | *Shí-sòng-lǜ* 十誦律 | T1435 | Puṇyatrāta, Kumārajīva |
| *Dharmaguptaka-vinaya* | *Sì-fēn-lǜ* 四分律 | T1428 | Buddhayaśas, Zhu Fonian |
| *Mahīśāsaka-vinaya* | *Wǔ-fēn-lǜ* 五分律 | T1421 | Buddhajīva, Zhu Daosheng |
| *Lotus-upadeśa* (Bodhiruci tr.) | *Miào-fǎ-lián-huá-jīng yōu-bō-tí-shè* 妙法蓮華經優波提舍 | T1519 | Vasubandhu (tr. Bodhiruci) |
| *Lotus-upadeśa* (Ratnamati tr.) | *Miào-fǎ-lián-huá-jīng yōu-bō-tí-shè* | T1520 | Vasubandhu (tr. Ratnamati) |
| *Fǎ-huá-xuán-zàn* | *Miào-fǎ-lián-huá-jīng xuán-zàn* 妙法蓮華經玄贊 | T1723 | Kuiji |
| *Fǎ-huá-yì-shū* | *Fǎ-huá yì-shū* 法華義疏 | T1721 | Jizang |
| *Fǎ-huá-xuán-yì* | *Miào-fǎ-lián-huá-jīng xuán-yì* 妙法蓮華經玄義 | T1716 | Zhiyi (rec. Guanding) |
| *Fǎ-huá-wén-jù* | *Miào-fǎ-lián-huá-jīng wén-jù* 妙法蓮華經文句 | T1718 | Zhiyi (rec. Guanding) |
| *Chū-sān-zàng-jì-jí* | *Chū-sān-zàng-jì-jí* 出三藏記集 | T2145 | Sengyou |
| *Fǎ-huá zhuàn-jì* | *Fǎ-huá zhuàn-jì* 法華傳記 | T2068 | Sengxiang |

---

## Cross-Reference Index (Series Papers)

| Section of This Paper | Related Paper | Doctrinal Link |
|---|---|---|
| §0 Opening | S7-P02 *huì-sān-guī-yī* | The present paper picks up the thread left dangling by P02: the true touchstone of "three returning to one" lies not in *śrāvaka* prophecies but at the two ends of extreme evil and extreme weakness |
| §I Documentary Triangle | S7-P07 Garment-Jewel / Topknot-Jewel | P07 foreshadowed that "prophecy presupposes originally-present seeds"; the present paper recovers the original configuration of the Devadatta Chapter through the documentary triangle |
| §II Āgama Prophecy Grammar | S7-P01 *upāya-kauśalya* | P01 addressed continuity between the *Lotus* and the Āgama; the present paper quantifies this as 95% continuity / 5% break-point |
| §III Vasubandhu → Kuiji Yogācāra Confluence | S7-P05 Medicinal-Herbs Parable | P05 deployed Vasubandhu's seven-pride system; the present paper deploys Vasubandhu's six-prophecy-loci system and extends it to Kuiji's *Xuánzàn* |
| §III Seed-Statement | S3 *Baifa mingmen* series | The Yogācāra theory of seeds was systematically unfolded in S3; the present paper is its speech-act-pragmatic application to *Lotus* prophecy |
| §IV Zhiyi Perfect Teaching | S7-P06 Illusory City | The "expedient/real" dialectic of P06 and Zhiyi's "transforming poison into medicine" metaphor in §IV both belong to Tiantai's perfect-teaching grammar |
| §V Madhyamaka Warning | S5 *Vajracchedikā* series | The Madhyamaka tetralemma of S5 is invoked in §V.1 to prevent seed-language from being ontologized |
| §VI Dual-Track Confluence | S7-P01, P02 (retrospective revision pending) | The present paper formally upgrades S7 Thesis I to "dual-track *tathāgatagarbha* + Yogācāra confluence"; the thesis-statements of P01 and P02 will be revised accordingly |
| §VI *Mahāparinirvāṇa* Parallel | S6 *Śūraṅgama* series | S6's "seven-place investigation of mind" in the *Śūraṅgama* and the *Mahāparinirvāṇa*'s *tathāgatagarbha* doctrine belong to the same *dharmakāya*-equality track |
| §VII *Vinaya* Triple Sealing | S7-P03 Burning House | P03's *upāya* doctrine and the present paper's *Vinaya* contrast jointly point to the structural necessity of "*upāya* without obstruction" |
| §VIII Middle-Way Corrections | S2 *Prajñāpāramitā* series | All four corrections appeal to the simultaneous co-validity of the two-layer structure, echoing the two-truths framework of S2 |
| Conclusion | S7-P09 *Encouragement to Uphold* (pending) | Formal analysis of prophecy (present paper) → temporality of sūtra-upholding practice (P09): from structural question to practical question |

---

## Footnotes

(Footnote [^1] appears at the end of §IV main text; footnote [^2] appears at the end of §I.2 main text. All other sūtra-and-treatise citations in this paper use inline parenthetical notation, giving the Taishō number and fascicle directly; see the Sūtra & Śāstra Index above.)

[^2]: **The Preface of Jñānagupta and Dharmagupta to the *Tiān-pǐn miào-fǎ-lián-huá-jīng* and its reading.** The opening preface of T09, no.264 (Taishō pp. 134b24–c22) is written by the translators themselves and states both the rationale for and the method of their supplementation work. Two passages are crucial. (i) Identification of the omissions in Kumārajīva's recension: "What Dharmarakṣa omitted was the verse portion of the *Universal Gate* chapter; what Kumārajīva omitted was the latter half of the *Medicinal-Herbs Parable* chapter, the opening portions of the *Pūrṇa* and *Dharma-Teacher* chapters, the *Devadatta Chapter*, and the verse portion of the *Universal Gate* chapter" (T09, no.264, p.134c07–09). (ii) Description of the method of supplementation: "In the first year of Renshou of the Great Sui [601 CE], the year *xīnyǒu*, at the request of the śramaṇa Shangxing of the Puyao Monastery, together with the two tripiṭaka masters Jñānagupta and Dharmagupta, we collated the Sanskrit palm-leaf manuscripts again at the Daxingshan Monastery … the latter half of the *Medicinal-Herbs Parable* was added, the *Devadatta* passage was *joined into the Stūpa Chapter*, the *dhāraṇīs* were placed after the *Supernatural Powers* chapter, and the *Entrustment* chapter was restored to the end" (T09, no.264, p.134c13–18). The two passages read together yield a conclusion essential to the present paper: the preface explicitly states that Kumārajīva's recension originally lacked the Devadatta passage, *and* that the method of supplementation adopted by the *Tiān-pǐn* translators was "joining [it] into the *Stūpa Chapter*" — that is, incorporating the Devadatta material within the *Seeing the Jeweled Stūpa* chapter, *without* establishing it as an independent chapter. This is consistent with the layout of Dharmarakṣa's *Zhèng-fǎ-huá-jīng* fascicle six *Seven-Jewel Stūpa* chapter, and together the two constitute two independent witnesses within the documentary triangle (§I) of this paper. In other words: in 601 CE, translators who were directly familiar with Sanskrit palm-leaf originals, working consciously to restore material missing from Kumārajīva's recension, *still* placed the Devadatta and Dragon-girl passages within the Stūpa Chapter. This is first-hand evidence for the thesis of this paper that "the symmetrical-wings configuration is the original form, and the independent-chapter status is a later development internal to the Kumārajīva-recension lineage." The preface is signed "supplemented in the first year of Renshou of the Sui by the two tripiṭaka masters Jñānagupta and Dharmagupta"; the head of the fascicle reads "translated by the Sui tripiṭaka master Jñānagupta of India together with Dharmagupta."

---

*指月研究 · Pointing at the Moon · CC BY-NC-SA 4.0*
*All canonical citations verified against CBETA Electronic Tripiṭaka Collection*
*GitHub: pointing-at-the-moon/dharma-research*