To Sāriputta

Identity

Segment range snp4.16:1.1–21.4. Norman vv. 955–975 (21 verses). The AV's closing sutta.

Four attested titles:

• Sāriputtasutta — the standard received Pāli title, after the sutta's named questioner. Used in modern editions and the segment-file header.
• Therapañhasutta ("Elder's-Question Discourse") — the alternative title preserved by Pj II 569.26.
• Therapuṭṭha ("Asked-by-the-Elder") — the name used by the AV's own uddāna at snp4.16:23.4 when listing the sixteen suttas of the collection.
• Upatisa-pasine ("Upatissa's Question") — the name under which the sutta appears in Aśoka's Bhabru edict (c. 250 BCE), per Kosambi's identification (see Reception, below). The fact that three of the four titles centre on Sāriputta's asking rather than on Sāriputta himself is the load-bearing piece here: the oldest naming-tradition identifies the sutta by its pucchā structure and by Sāriputta's personal rather than ordinary name.

Text and form

Two-metre architecture — unique in the AV. Vv. 955–962 (Sāriputta's pucchā, 8 verses) are in śloka; vv. 963–975 (the Buddha's visajjanā, 13 verses) are in Triṣṭubh. No other AV sutta has a deliberate Q/A metre-partition (Norman 2001 pp. 386–87). The metre-boundary at v. 962/963 is a compositional seam. Lee 2024 reads the sutta as a two-part compilation.

The metrical registers place 4.16 at the intersection of the AV's two main strata: the śloka section shares metre with the moral-admonitory cluster (4.1, 4.7, 4.10, 4.15); the Triṣṭubh section shares metre with the philosophical-polemical cluster (4.3–4.5, 4.8–4.9, 4.11–4.13). The closing sutta combines both AV metres in a single text.

Speaker-tags. Extrametrical reciter's marks at two points: (iccāyasmā sāriputto) at v. 955, line 2, identifying the opening speaker; (sāriputtāti bhagavā) at v. 963, line 2, marking the shift to the Buddha's voice. Both are standard non-original interpolations of the kind Norman flags through the AV; both are commented on by the Mahāniddesa and therefore pre-Niddesa.

Content

Sāriputta praises the Buddha in three opening verses, frames his question as asked "on behalf of the many bound here," and poses a practical question: how should a bhikkhu who loathes (vijigucchamāna) attachment and frequents remote lodgings — tree-roots, charnel-grounds, mountain-caves — conduct himself? What fears should he not tremble at; what adversities should he overcome; what ways-of-speech, alms-resort, and precepts should he have; what training makes him unified, alert, and mindful, such that he purges his own stains "as a smith [purges] silver"?

The Buddha's answer is a bhikkhu-practice catalog: five environmental fears not to fear (flies, mosquitoes, snakes, human-contacts, four-legged beasts); endurance under illness, hunger, cold, heat; no theft, no lying; mettā for "trembling-and-firm" beings; dispelling of mental muddiness as belonging to kaṇhassa pakkha ("Māra's faction"); uprooting of anger and super-pride; the four laments about food and sleep to be disciplined; moderation in food and lodging; downcast eyes, engaged meditation, equanimity; welcoming reproach from fellow bhikkhus; and training to subdue desire for the five sense-fields. The catalog is the AV's most systematic prescription of monastic practice, parallel in register to Snp 4.14 Round 2 and Snp 4.15 Round 2 but more comprehensive.

The collection closes on a thematic inclusio: at v. 956 the Buddha dispels all darkness (sabbaṁ tamaṁ vinodetvā); at v. 975 the trained bhikkhu is told to banish the darkness (vihane tamaṁ). The AV ends on the trainee's self-dispelling.

Key passages

v. 955 — the opening.

Na me diṭṭho ito pubbe, (iccāyasmā sāriputto) / Na suto uda kassaci; / Evaṁ vagguvado satthā, / Tusitā gaṇimāgato.

Never before has there been seen by me — so said the venerable Sāriputta — nor heard from anyone: such a teacher of graceful speech, come from Tusita as leader-of-a-community.

Opens the sutta in devotional-praise register unusual in the AV. The Buddha-epithets vagguvada ("graceful-speaker") and gaṇin ("community-leader") are AV hapaxes; Tusita is also an AV hapax and carries the verse's main choice-point (below).

v. 962 — the silver-smith simile, closing the pucchā.

Kaṁ so sikkhaṁ samādāya, / ekodi nipako sato; / Kammāro rajatasseva, / niddhame malam attano.

Undertaking what training — unified, alert, mindful — would he purge his own stains as a smith [purges] silver?

Pāda cd (Kammāro rajatasseva, niddhame malam attano) is verbatim with Dhp 239 cd. This is the AV's strongest direct cross-Khuddaka parallel; direction of borrowing is indeterminable but the philological identity is secure.

v. 970 — the four laments.

Kiṁsū asissāmi kuva vā asissaṁ, / Dukkhaṁ vata settha kvajja sessaṁ; / Ete vitakke paridevaneyye, / Vinayetha sekho aniketacārī.

"What will I eat? Where will I eat? Truly, I slept badly — where shall I sleep today?" — these lamentable thoughts let the trainee, the homeless-wanderer, discipline.

Four first-person-direct-speech laments. Dramatization of this kind is rare in the AV (compare only Snp 4.9's Māgaṇḍiya-quotation). The food/sleep binary — two laments on each — is tightly structured and turns on whether asissaṁ and sessaṁ parse as future of as- ("I eat") and setī ("I sleep") respectively; see Choice-points.

v. 973 — welcoming reproach.

Cudito vacībhi satimā'bhinande, / Sabrahmacārīsu khilaṁ pabhinde; / Vācaṁ pamuñce kusalaṁ nātivelaṁ, / Janavādadhammāya na cetayeyya.

Reproached in speech, the mindful one should rejoice; let him break hardness toward fellow holy-life-walkers. Let him utter wholesome speech, not excessive; let him not think in ways that provoke public blame.

The AV's most striking disciplinary injunction: the sage welcomes reproach rather than parries or resents it. Canonically-congruent with MN 15 Anumāna-sutta (bhikkhus should invite reproof) and AN 5.98.

v. 975 — the closing verse.

Etesu dhammesu vineyya chandaṁ, / Bhikkhu satimā suvimuttacitto; / Kālena so sammā dhammaṁ parivīmaṁsamāno, / Ekodibhūto vihane tamaṁ so.

Having disciplined desire in these things — the mindful bhikkhu, well-released in mind, in due course rightly investigating dhamma, unified — let him banish the darkness.

Two echoes tie the verse to the sutta's own opening: ekodi-bhūta picks up ekodi in Sāriputta's closing pucchā at v. 962; vihane tamaṁ picks up sabbaṁ tamaṁ vinodetvā at v. 956. The Buddha dispelled his darkness; the bhikkhu is trained to banish his own. The AV ends here.

Choice-points

v. 955, Tusitā gaṇim-āgato. The Pāli says the Buddha has come from Tusita. Pj II frames the sutta's occasion as the Buddha's descent from Tāvatiṁsa after the rains-retreat teaching Abhidhamma to his mother — a different heaven. The Yizujing YZJ-14 explicitly names Trāyastriṃśa (忉利天 = Tāvatiṁsa), aligning with Pj II against the Pāli verse. Three resolutions are live: (a) the Pāli's Tusita is a textual anomaly, (b) Tusita is used generically for "heavenly-realm," (c) the Pāli preserves an older, pre-Pj-II occasion that the commentary later reworked to match the Tāvatiṁsa-tradition. The Niddesa (I 446.29–447.1) preserves both Tusita-kāya (the heaven) and an etymologizing reading Tusita = "the content ones" = arahants, making no final decision — the commentarial tradition inherited the puzzle.

v. 960, gacchato agataṁ disaṁ. The phrase is nibbāna-imagery — either "the direction not yet gone to" (a-gata), "the direction without gati," or, with a variant reading, "the deathless direction" (amata). Pj II and the Niddesa read amataṁ; Norman prefers agataṁ for the wordplay with gacchato ("going"). Translators scatter: Norman "the direction-not-yet-gone-to"; Bodhi "unreached"; Sujato "untrodden place"; Thanissaro interpretively "undeclined"; Fronsdal paraphrastically "a new direction."

v. 968, addhā bhavanto abhisambhaveyya. Norman (2001 pp. 391–92) reads not addhā bhavanto ("certainly being") but addhabhavanto — a non-palatalized form of adhyu-bhavati ("overcomes"), with -dhy- assimilated to -ddh- rather than palatalized to -jjh-. The parallel construction is Jain-canonical. If Norman is right, 4.16 preserves a rare archaic dialectal form; the Niddesa reads the line instead as ekaṁsen' eva abhibhaveyya ("thoroughly he would overcome"), flattening the archaism.

v. 970, sessaṁ. Norman, Bodhi, Thanissaro, and Sujato all parse sessaṁ as future of setī (sleeps), preserving the food/sleep binary: asissāmi / asissaṁ = "I will eat / where shall I eat"; settha / sessaṁ = "I slept / where shall I sleep." Fronsdal 2016 reads sessaṁ as future of as- (eats), giving "Where shall I eat today?" and flattening the binary. This is the one clear isolated misreading in Fronsdal's 4.16.

Vocabulary and commentary

Lexical profile. Snp 4.16 has the AV's highest surface-form hapax rate at 74.2% (201 AV-hapaxes / 271 distinct types, per data/analysis/av-vocab/). The closing cluster ranks 4.16 (74.2%) > 4.14 (66.0%) > 4.15 (57.0%) — all three sit well above the AV core. Caveat: surface-form, not lexeme-normalized.

The AV-hapax set in 4.16 draws heavily on pan-canonical monastic-practice vocabulary barely present elsewhere in the AV: vagguvada, gaṇin, vyappatha, gocara, ekodi, kammāra, ātaṅka-phassa, khudā, tasa-thāvara, kaṇhassa pakkha, palikhañña, addhā-bhavanto, sekha, okkhitta-cakkhu, jhānānuyutta, sabrahmacārī, and rajāni in the sense of "defilements." The vocabulary is Pātimokkha-adjacent, dhutaṅga-related, and samādhi-technical — markedly different from the diṭṭhi-polemic lexicon of 4.3–4.13.

Sīlabbatāni at v. 961 is also notable: it is the only AV occurrence where the compound is used affirmatively (what precepts-and-vows should the bhikkhu have?) rather than polemically (the AV's critique-target at 4.4, 4.5, 4.12, 4.13). The positive usage points to the non-polemical register of the Q-section.

Mahāniddesa (Mnd 16). The AV's closing Niddesa-chapter, and a showcase for the commentary's taxonomic mode. Characteristic moves: (i) a three-fold classification of question-types at v. 957 (clearing the unseen, confirming the seen, eliminating uncertainty); (ii) the four-fold pariyanta framework at v. 964 (sīla-saṁvara, indriya-saṁvara, bhojane-mattaññutā, jāgariyānuyoga); (iii) three separate readings of anoka at v. 966 (no dwelling-space for conditioned consciousness / for misconduct / = anālaya, "without attachment"); (iv) importing the Vibh 379 psychological five-fears list (ājīvika-bhaya, asiloka-bhaya, etc.) alongside the sutta's own environmental five (flies, mosquitoes, snakes, humans, beasts) — acknowledging two distinct canonical five-fears taxonomies. At v. 955 it preserves both the Tusita-heaven and the etymologizing content-ones readings without deciding between them. The classification-heavy style matches the Niddesa's opening-and-closing-chapter register and reads as a deliberate summation of the AV's doctrinal apparatus.

Cross-recensional witnesses

Pāli: full; the 21-verse Q/A structure in two metres.

Chinese Yizujing YZJ-14 蓮花色比丘尼經 ("Nun Utpalavarṇā Sūtra") at [T0198_p0184c24]–[T0198_p0186c28]: 39 verses total = 20 parallel + 19 added (Lee 2024 Table 2). The 19 added concentrate in a 29-verse Uppalavaṇṇā praise-sequence prepended to Sāriputta's Pāli-parallel opening. Primary-text verification: the Chinese opens with an assembly-reception prose frame; then at [T0198_p0186a03] a female "venerable one" (賢者) rises and praises the Buddha in ~29 verses; at [T0198_p0186b12] Sāriputta (舍利弗) rises and at [T0198_p0186b14] delivers 未嘗見有是者 — the Pāli Sn 955 Na me diṭṭho ito pubbe. The Chinese title honours Uppalavaṇṇā's priority of speech, not Sāriputta's canonical role; the structure is prepended first-voice, not gender-replacement. The Yizujing also situates the sutta explicitly at the Tāvatiṁsa (忉利天) descent, aligning with Pj II against the Pāli verse's Tusita. Positional displacement: Pāli 16 → Chinese 14 — the single largest repositioning in the Yizujing ordering.

Uppalavaṇṇā is the bhikkhunī foremost in iddhi (AN 1.14 §5.2); Sāriputta is foremost in wisdom (AN 1.14 §4.1). The Chinese pairs the two foremost disciples — one of each gender — as the sutta's opening framers. Whether the prepended Uppalavaṇṇā material originates in a lost pre-Chinese Indic source or reflects specifically Chinese-Buddhist traditions of bhikkhunī-praise is not decidable from the Chinese text alone.

Sanskrit: not attested. Hoernle 1916 covers Snp 4.7–4.10 only.

Gāndhārī: not attested in our in-repo editions. The Split Arthapada scroll (Allon 2024's single-find consolidation with British Library + University of Washington + Islamabad Museum fragments) runs from Sn 841 (Snp 4.9) through the end of the collection, but the Snp 4.16 verses fall in the scroll's damaged portion; Lee 2024 does not transcribe them.

Jain Isibhāsiyāiṁ 38.6 ~ Sn 974 (SC parallels graph: snp4.16#vns981 ~ Isibhāsiyāiṁ 38.6). The Isibhāsiyāiṁ (Ṛṣibhāṣitāni, "Sayings of the Rishis") is a late-Śvetāmbara prakīrṇaka preserving teachings attributed to non-Jain sages; chapter 38 is explicitly attributed to Buddhassa Sakkaputtassa — "Buddha the Śākyan-son." One AV verse is thus preserved in the Jain canonical tradition as an explicitly-marked saying of the Buddha. This is the rarest of the AV's cross-religious parallels: a case of one tradition preserving another's verse under explicit authorial attribution. Primary text not in-repo; flag for acquisition (Schubring 1942 edition). The SC parallel is a "~" partial match; verification of pāda-level identity with Sn 974 awaits the source.

Coverage note. Snp 4.16 is effectively 2-recension at verse-level (Pāli + Yizujing) but carries the richest extra-canonical attestation of any AV sutta via the Bhabru edict and Isibhāsiyāiṁ.

Internal cross-references

Within the AV. The verbal echoes ekodi (v. 962) ↔ ekodibhūto (v. 975) and tamaṁ vinodetvā (v. 956) ↔ vihane tamaṁ (v. 975) form an inclusio across the sutta. Cross-sutta identities: v. 971 rusitopi vācaṁ pharusaṁ na vajjā ~ 4.14 v. 932 pharusena ne na paṭivajjā (non-retaliation formula); v. 969 paññaṁ purakkhatvā = 4.3 v. 785; v. 972 okkhitta-cakkhu / pāda-lola ↔ 4.14 vv. 922, 925; v. 968 kodhā-atimāna ↔ 4.10 v. 855, 4.14 v. 928. The late-AV cluster (4.14, 4.15, 4.16) shares more formulaic stock with each other than with the AV core.

Within the Khuddaka. The strongest verbatim parallel in the AV: v. 962 cd = Dhp 239 cd (kammāro rajatasseva niddhame malamattano). Snp 1.3 Khaggavisāṇa is the closest genre-match for the practice-catalog register of Round 2; Snp 3.11 Nālaka (= Aśoka's Moneya-sūte) shares the canonical-anthology Bhabru-cluster; Snp 1.12 Muni-sutta (= Aśoka's Muni-gāthā) shares the muni-description genre.

Prose-nikāya uptake. SN 12.31 Bhūtamidaṁ has the Buddha asking Sāriputta to expound a Pārāyaṇa verse — an explicit canonical witness to Sāriputta as a Sutta-Nipāta exegetical authority, counterpart to the AV-side role he occupies in 4.16. MN 15 Anumāna-sutta has Sāriputta himself teaching bhikkhus to invite reproof, a direct thematic parallel to v. 973. AN 1.189 canonically establishes Sāriputta as foremost in wisdom; the bahūnam idha baddhānaṁ atthi pañhena āgamaṁ frame ("I have come with a question on behalf of the many bound here") is characterologically consistent with that role.

Reception and external attestation

Mahāniddesa: Mnd 16 treats all 21 verses; see Vocabulary and commentary, above. The Niddesa does not name the sutta as Aśoka's Upatisa-pasine; that identification is modern.

Paramatthajotikā II: Pj II 569.26 supplies the alternative title Therapañhasutta and frames the occasion as the Buddha's descent from Tāvatiṁsa after the rains-retreat teaching of Abhidhamma to his mother (per Dhp-a III 216 on Dhp 181). Pj II's frame is shared with the Yizujing against the Pāli verse.

The AV's own uddāna (at snp4.16:23.4): names the sixteenth sutta therapuṭṭha ("Asked-by-the-Elder"), not Sāriputta. The internal naming-tradition preserved in the collection-summary matches Pj II's alternative title and aligns with the Bhabru edict's pasine (pañha / praśna = "question") — a three-way confluence that strengthens Kosambi's identification.

Aśoka's Bhabru edict (c. 250 BCE): the fifth of seven dhamma-paliyāyāni Aśoka recommends to the Saṅgha for repeated recitation is Upatisa-pasine. The identification with Snp 4.16 rests on: Upatisa = Upatissa (Sāriputta's personal name; Sāriputta = "son of Sārī," a matronymic) and pasine = praśne / pucchā. No other extant Pāli text matches as cleanly. The identification is accepted by Kosambi 1912 (editio princeps), Law 1933 vol. 1 p. 6 and vol. 2 pp. 665–66, Jayawickrama 1976 pp. 136–42, Norman 2001 Introduction, and the consensus of modern Aśoka scholarship. This is the only AV sutta with direct extra-canonical dating; composition must predate c. 250 BCE by a margin sufficient for canonical-recitation status.

Three of the seven Bhabru texts come from the Sutta-Nipāta (1.12, 3.11, 4.16) — strong evidence that the Snp's core was already a recognized canonical collection by c. 250 BCE.

Peṭakopadesa: Pe chapter 1's AV-citation density concentrates on Snp 4.2 (×2), 4.6 (×2), and 4.7 (×1) — the extended Mahākaccāna-AV web discussed in Part I Ch 8. No verse of Snp 4.16 is cited in the extant Pe apparatus.

Reading

Snp 4.16 closes the AV in a register unlike anything else in the collection: a named canonical disciple poses a specific practical question "on behalf of the many bound here," and the Buddha answers with the most systematic bhikkhu-practice catalog in the AV. The form is formal-pedagogical; the tone is devotional-then-prescriptive; the kāma and diṭṭhi concerns that drive the collection's core are not in view. In a collection that opens with anonymous observation on sense-pleasures and moves through philosophical polemic against view-supremacy, it closes with the rules of a forest-bhikkhu's day.

The two-metre architecture makes the sutta a miniature of the whole AV — śloka-register for the pupillary-frame, Triṣṭubh-register for the instruction — and the inclusio from the Buddha's darkness-dispelling to the bhikkhu's own self-dispelling locates the AV's final lesson in the trainee. The three title-traditions that name the sutta by its asking (Therapañha, Therapuṭṭha, Upatisa-pasine) rather than by Sāriputta himself point in the same direction: what the oldest tradition recognized here was a pucchā-instruction unit, the sutta as a teaching-occasion.

Snp 4.16 carries the heaviest historical load in the AV. Aśoka's recommendation places its composition before c. 250 BCE; the Isibhāsiyāiṁ 38.6 parallel places at least one of its verses in the Jain canon as a saying of the Buddha; the uddāna's therapuṭṭha naming and Pj II's Therapañha confirm an internal tradition that matches the Aśokan name. The Yizujing's Uppalavaṇṇā-prepended opening shows that the sutta's frame was recensionally elastic — the Pāli's minimal vatthugāthā is one possibility among several — but the verse-core, in both Q and A, is preserved closely across the two recensions that attest it.