On Interpretations of the anusmṛti of the Three Jewels
- Buddhānusmṛtivṛtti, Vyākhyāyukti, and Related Texts
by HORIUCHI Toshio
2) *Dharmānusmṛti-vṛtti (DAV), D No. 3983 (15a7–15b7), P No. 5483, tr. Φ (undescribed);
4) *Buddhānusmṛti-ṭīkā (-rgya cher ’grel pa) (BAT), D No. 3987 (55b4–63b4), P No. 5487, tr. Dānaśīla, dPal brtsegs rakṣita In addition to those, there are several other texts: 5) Vasubandhu’s Vyākhyāyukti (VyY), D No. 4061, P No. 5562. Ch II (Interpretations to Sūtrakhaṇḍa (SK) 1 (D 40b2–41b1), SK 5 (D 43a7–44a2), SK 8 (D 44b4–45a4), SK 74 (D 69b7–70a4) (Chapter II of this text, which deals with the interpretation of the Sūtrakhaṇḍaśata, includes interpretations of buddha,
5') Guṇamati’s Vyākhyāyuktiṭīkā (VyYT), D No. 4069, P No. 5570;
6) Vīryaśrīdatta’s Arthaviniścayasūtranibandhana (AVSN, The Arthaviniścaya Sūtra and Its Commentary (Nibandhana). ed. N. H. Samtani. Patna: Kashi Prasad Jayaswal Research Institute, 1971), pp. 241.3–248.1 (This text includes an interpretation of four avetyaprasādas); 6') Anonymous commentary to AVSN:
Arthaviniścayaṭīkā, D No. 4365, P No. 5852;
7) Haribhadra’s [[Abhisamayālaṃkārālokā] (AAA, Abhisamayālaṃkārālokā Prajñāpāramitāvyākhyā: The Work of Haribhadra. ed. Unrai Wogihara, Tokyo: Sankibo Buddhist Book Store, 1932, 1973), pp. 183.4–185.5 (This text includes an interpretation of the epithets of the tathāgata/ Buddha). As previous studies have pointed out,1) the interpretations of these items, namely, three jewels or epithets of the Buddha, in those texts bear such similarities that we can examine
On Interpretations of the anusmṛti of the Three Jewels (HoriucHi) the close relationship between them. However, a detailed investigation of these texts has yet to be conducted. Accordingly, in this paper, I investigate some of the relationships between these texts.
1. Relationship between Texts
Previous studies have revealed some relationships among these texts, highlighting their similarities. In my examination, I employ AVSN as an axis from which these similarities ﬂow through the four following avetyaprasādas:
1) Buddha (buddhe ’vetyaprasāda, AVSN, pp. 241.3–248.1)=/≒2) BAV (almost as a whole)=/ ≒ BAT (ﬁrst half of this text: namely, D 55b4–58b13))≒VyY, Ch II, SK 1, [Interpretation I];4)
2) Dharma (pp. 248.2–249.6)=*Dharmānusmṛtivṛtti (as a whole)≒VyY Ch II, SK 5 [Interpretations I, II]; 3) Saṅgha (pp. 249.7–256.3)=*Saṅghānusmṛtivyākhyā (ﬁrst half of the text: namely, D 16a1–16b6)=VyY Ch II, SK 8 [Interpretation I]; 4) Śīla (pp. 256.4– 259.1)=VyY Ch II, SK 74, VyYT, Abhidharmakośabhāṣya, and Abhidharmakośavyākhyā． 2. Relationship between VyY and the “three texts by Asaṅga”
One should, of course, bear in mind that commentaries written about the same text, even if they are written by diﬀerent authors, may plausibly make similar points and gives similar interpretations. However, as far as BAV, DAV, SAVy (hereinafter referred to jointly as the “three texts by Asaṅga”), and VyY are concerned, these texts, in many instances, correspond word for word. Therefore, if one does not assume they are founded on some common third source, which has not come down to us now, it follows that either VyY or the “three texts by Asaṅga” cited the other. 2.1. VyY, DAV, and SAVy If Asaṅga is truly the
author of the three texts, and because VyY was written by Vasubandhu, it is natural to assume that Vasubandhu cited Asaṅga’s texts. However, “Asaṅga’s” DAV and SAVy include commentarial phrases that are also seen in VyYT, a commentary to the VyY. Below are examples of these similarities. 2.1.1. Interpretation
of aihipaśyika, an Adjective to the dharma VyY, D 43b3, P 49b3: thun mong ma yin pa’i phyir (6) ’di mthong ba la yod pa ste/ ’di mthong ba rnams la yod pa’i phyir ro// (“Since it (Buddha’s dharma/teaching) is uncommon [to the dharma of heretics], it is “that which exists in those who sees here (*aihipaśyika).”
However, it is a rare word because it is deduced from aihipaśyika. Therefore, Guṇamati commented that ipahaśyas refers to chos ’di pa rnams, *ihadhārmikas, which is a common term used to refer to Buddhists. VyYT, D 161b7, P 27a7: ’di mthong ba rnams la yod pa’i phyir ro zhes bya ba ’di mthong ba ni chos ’di pa rnams so// Surprisingly, this kind of commentarial phrase is found in DAV (and AVSN cites DAV). AVSN, 249.1–3: asādhāraṇatvād (6) aihipaśyikaḥ; ihapaśyeṣu *bhāvād/ ihapaśyā ihadhārmikāḥ/ ihaiva tattvaṃ* paśyantīti kṛtvā/5) DAV, D 15b4–5, P 19a1–2: thun mong ma yin pa (6) 'dir ltos shig pa ni dngos po la ste/
chos dang ldan pa ’di la ’dir ltos la ’di nyid du de kho na mthong bar gyis shig pa'o// The Tibetan translation of DAV is rather imperfect, as in the case of BAV and SAVy. Moreover, there are several textual problems here as well.6) However, focusing only on the double-underlined parts, a close comparison to AVSN reveals that chos dang ldan pa ’di la corresponds to ihadhārmikāḥ, whereas ’dir ltos la corresponds to ihapaśyā. DAV here comments that ihapaśyā
2.1.2. Interpretation of nyāyapratipanna (practicing for the sake of nyāya=nirvāṇa), an Adjective to the saṅgha
21) to illustrate that nyāya in the nyāyapratipanna refers to nirvāṇa. VyY, D 44b6, P 51a2–3: de yis (de yis] P; de'i D) rigs pa’i chos ni bla na med pa kun tu ’thob par ’gyur ba nyid yin la zhes bya ba gsungs pa’i phyir ro// However, the sūtra that he cited itself just says that nyāya is anuttara-dharma (cf. AVSN below). Therefore, VyYT
comments as follows: VyYT, D 164b5, P 30b6:  rigs pa’i chos ni bla na med pa zhes bya ba ni mya ngan las ’das pa’i chos (*nirvāṇadharma) bla na med pa zhes bya ba’i tha tshig go// Here, we ﬁnd the term *nirvāṇadharma. By this commentary, one can easily understand that nyāya=anuttara-dharma=nirvāṇadharma. A similar commentarial phrase is found in SAVy (and AVSN).
AVSN, 250.7–251.1: . . . nyāyaṃ dharmam anuttaram7) iti vacanāt/ nyāyaṃ nirvāṇadharmam ity arthaḥ/ SAVy, D 16b2, P 20a2–3: de’i rigs pa ni chos bla na med pa brnyes par byed par ’gyur ro zhes gsungs so// rigs pa (*nyāya) ni mya ngan las 'das pa'i chos (*nirvāṇadharma) so zhes bya ba'i don to// 2.1.3.
Interpretation of sāmīcīpratipanna, an Adjective to saṅgha. VyY states as follows: VyY, D 45a1–2, P 51a6:
(b) longs spyod dang tshul khrims dang lta ba mthun pa nyid kyis so// VyYT states as follows. The double-underlined part is a commentary to Vasubandhu’s above sentence. VyYT, D 164b7, P 30b8–31a1: (b)
(ii) tshul khrims mthun pa nyid ni tshul khrims so// (iii) lta ba mthun pa nyid ni lta ba’o// Similar (in the case of SAVy) or the exact (in the case of AVSN) commentarial phrase is found in the following two texts. AVSN, 251.6–7: (b) tatparibhogaśīladṛṣṭisamānatayā ca/ tatra (i) paribhogasamānatā āmiṣeṇa
(i) longs spyod mnyam pa ni zang zing gis so// (ii) bslab pa mnyam pa ni tshul khrims kyis so// (iii) lta ba mnyam pa ni yang dag pa'i lta bas so// In the preceding subsections, I have provided three examples that show that DAV and SAVy, which are ascribed to Asaṅga, contain commentarial phrases found in Guṇamati’s VyYT, a commentary to VyY. One can, of course, assume that VyY cited DAV and SAVy in a concise way, omitting the double-underlined parts.8)
However, because of the nature of the VyY, such supposition is unlikely, since the purpose of composing VyY is to guide those who wish to write a commentary to sūtra, as the title vyākhyā-yukti, method of [[[sūtra]]] exegesis, indicates. Further, Chapter II of the VyY, which includes these passages, is intended to show examples to help interpret the meaning of words (padārtha). Therefore, there is no reason for VyY to eliminate such a complementary commentarial phrase as its absence makes the sentences more enigmatic. Based on these facts, I assume that VyY is cited by DAV and SAVy. Moreover, I assume
that DAV (and SAVy too, if one does not assume that Vasubandhu omitted the passage that is found in SAVy and Guṇamati supplied it again)9) was written more recently than VyYT. 2.2. Vasubandhu and āgama In order to reinforce the above assumption, let us point out the improbability of assuming that DAV, SAVy, and BAV preceded VyY. VyY, when interpreting dharma and saṅgha, refers to and cites *āgama, lung. The nature of this āgama, which is mentioned 18 times in Chapter II of the VyY, should be examined further; however, the point here is that he cites some authoritative interpretation of dharma and saṅgha. If DAV
and SAVy were the writings of Asaṅga that preceded Vasubandhu, he could have cited these very two writings by Asaṅga as āgamas. However, what he cites involves a diﬀerent interpretation (I cannot trace the interpretation in these āgamas in the existing texts), and the interpretations similar to DAV and SAVy are actually provided in the body of the VyY as if they were his own interpretations (What I mean here is that there is no introductory mark or sign
to indicate that these are citations). Moreover, one should remember that Vasubandhu is an author who usually distinguishes his citations clearly from the rest of his text. This theory also holds up with regard to BAV. Vasubandhu’s interpretation of buddha is very close to that of BAV. However, Vasubandhu does not refer to BAV (in this section, he does not refer to lung, *āgama) but his interpretation, which is similar to BAV, is shown without any indication
that it is a cited from other text. For this reason, I assume that VyY precedes “the three texts by Asaṅga.” 2.3. Other doubts about the “three texts by Asaṅga” There are other oddities in the three texts ascribed to Asaṅga.
1) The title of the three texts diﬀers: *Buddhānusmṛti-vṛtti (-’grel pa), *Dharmānusmṛtivṛtti (-’grel pa), *Saṅghānusmṛti-vyākhyā (-bshad pa). 2) There are no Indian sources that cite these three texts by name (which is why I
used the “*” in the full title of the three texts like *Buddhānusmṛtivṛtti), nor are there any Indian sources that cited or gave commentary to these texts as those by Asaṅga. These oddities (1 & 2) reveal that it is only the title and the colophon of the Tibetan translations of these three texts that suggest their Sanskrit titles and ascribe them as Asaṅga. 3) DAV and SAVy, at the beginning of their text, refer to *abhedyaprasāda as if the whole texts are interpretations of *dharme abhedyaprasāda and *saṅghe abhedyaprasāda.
However, BAV does not refer to buddhe abhedya/avetya-prasāda at all. This fact also illustrates inconsistency in the reasons BAV, DAV, and SAVy were composed.
The results of my examination of these various texts have led to the following conclusions: 1) VyY preceded BAV, DAV, and SAVy; 2) VyYT preceded DAV and SAVy;
b) authorship, and c) the reason they were composed, remain. I will show in another article that AAA cited and developed the interpretation of the epithets of tathāgata by BAV and that BAV, ascribed to Asaṅga, cited BAT, ascribed to Vasubandhu, or, more precisely, that BAV extracted interpretations of the epithets of tathāgata in BAT.
1) Regarding the previous studies of these texts, see Horiuchi forthcoming.
2) =/≒ signiﬁes “equal or nearly equal.” AVSN and BAV are equal in certain places, although, at a glance, it does not seem so because of the poor Tibetan translation of BAV. However, in some places, they deﬁnitely diﬀer. That is why I used both=and≒. For additional details, see Horiuchi forthcoming.
3) The latter half of BAT (D 58b1–63b4) is unique and has no correspondence in BAV, VyY, or AVSN.
4) Vasubandhu gives two interpretations of this SK. [Interpretation I] shows that it is the ﬁrst interpretation that corresponds to AVSN, BAV, and BAT. 5) *bhāvād/ ihapaśyā ihadhārmikāḥ/ ihaiva tattvaṃ] emended; bhāvād ihapaśyā ihadhārmikāḥ/ ihaiva tattvaṃ Ms, G, N; bhāvād ihapaśyaḥ, ihadhārmikaḥ, ihaiva tathātvaṃ (tathātvaṃ] tatvaṃ N) AVSN, 249.fn.3.
Honjō (1989), a Japanese translation of AVSN, already pointed out the relationship between the AVSN above and VyYT (Peking ed.) and also suggested almost perfect emendation to the text as follows: bhāvād/ ihapaśyāḥ ihadhārmikāḥ/ ihaiva tattvaṃ (VYT I 27a7). 6) Here, there is an incorrect translation of case endings as I have shown them in wave and an eye-skip of a word. I do not comment in detail here. 7) I skipped some words as indicated by “. . .” because there is a textual problem here.
1, if one assumes that DAV precedes VyY, it follows that Vasubandhu just cited one sentence from DAV. Guṇamati, on the other hand, acknowledging the two additional sentences, the double-underlined and bold ones, only supplied the double-underlined part. DAV, the oldest one, possesses all three sentences! 9) VyY and VyYT seem to have been utilized as a set. As Honjō (1989) has shown, AVSN, for example, cites VyY by supplying some passages from VyYT. Bibliography Honjō Yoshifumi 本庄良文．1989. Bonbun wayaku Ketsujō Gikyō, Chū 梵文和譯 決定義經・註．Kyoto.
Horiuchi Toshio 堀内俊郎．2016. Seshin no agon-kyō kaishaku: Shakkiron dainishō yakuchū 世親の阿含 経解釈：釈軌論第2章訳註．Tokyo: The Sankibo Press. ―. forthcoming. “Butsuzuinen Chū Butsuzuinen Kōchū ni taisuru bunkengakuteki kenkyū: Arthaviniścayasūtranibandhana tono taihi de (1)” 仏随念注 仏随念広注に対する文献学的研究： Arthaviniścayasūtranibandhanaとの対比で(1). Tōyōgaku Kenkyū 東洋学研究 55 . (This work is supported by JSPS KAKENHI 17K02224.)