Disputed Emptiness: Vimalamitra’s Mādhyamika Interpretation of the Heart Sutra in the Light of His Criticism on Other Schools

Abstract

The *Āryaprajñāpāramitāhṛdayaṭīkā (hereafter PHT), Vimalamitra’s (ca. the 8th- to 9th cent.) commentary on the long Heart Sutra (hereafter HS), is not merely a commentary on words and phrases of the HS, but it also refers to and criticizes non-Buddhist schools and other schools within Buddhism. However, due to its textual situation, namely, the original Sanskrit being lost and it remains only as the Tibetan translation, the discussion there has not always been well understood. In particular, it has been suggested in recent years that Vimalamitra endorsed the Yogācāra perspective in the PHT. In this paper, I will primarily examine Vimalamitra’s interpretation of the famous four sets of phrases, such as “rūpa (form) is emptiness/empty” in the PHT, and clarify Vimalamitra’s understanding of key Buddhist concepts such as “emptiness” or “dependent origination”. In doing so, I argue that the PHT is written sorely from the Madhyamaka perspective, and Vimalamitra criticizes other schools, including the Yogācāra. Furthermore, by comparison with the writings of Kamalaśīla, an older contemporary of Vimalamitra, I further clarify his position as a Mādhyamika adherent. Finally, I will discuss the significance of those philosophical arguments in the PHT to the modern readers of the HS.

1. Introduction: The HS and Vimalamitra’s Commentary

1.1. “Commentaries on the Long HS Preserved in the Tibetan Canon”

First, I provide a brief account of the Heart Sutra (hereafter HS) insofar as it is relevant to this paper, based on the recent overview given by Zacchetti (2015).

“The gist of this scripture … is based on passages found in Larger Prajñāpāramitā texts … forming an elaboration of the notion of emptiness which is not found in the Aṣṭasāhasrikā”.

(p. 194)

“The Sanskrit text exists in two versions, a shorter and a longer one. The main difference between the two is that the longer version includes the canonical initial and closing formulas which are typical of sutras … Watanabe (2009, p. 80) maintains that the Heart Sūtra took shape in the 5th–6th century and was then expanded into the long version in the 7th century”.

(p. 195)

“The canonical Tibetan translation represents the long version of the Heart Sūtra. It is ascribed by a colophon (only found in one witness) to Vimalamitra and Dge slong Rin chen sde (for a detailed discussion of this document, (Silk 1994, pp. 47–56)”.

(p. 196)

Second, I provide some basic information on the “Indian” commentaries on the HS.

The Tibetan canon includes eight Indian commentaries on the HS. These were composed by: 1. Vimalamitra (D[erge] no. 3818), 2. Jñānamitra (D no. 3819), 3. Vajrapāṇi (D no. 3820), 4. Praśāstrasena (D no. 3821), 5. Kamalaśīla (D lacks. P no. 5221), 6. Dīpaṃkaraśrījñāna/Atiśa (D no. 3823), 7. Śrīmahājana (D no. 3822), 8. Śrīsiṃha-Vairocana (D no. 4353).1

These are all commentaries on the longer version or the long HS, not on the short version with which we are more familiar through Xuanzang’s HS.2 This is in contrast to the Chinese and Japanese commentaries on the HS, which are almost exclusively commentaries on the short HS.3 Moreover, all eight of the above commentaries only remain in Tibetan translations or versions, and none survive in Sanskrit or Chinese translation. In this connection, I have argued that two of the eight (those by Śrīsiṃha-Vairocana and Jñānamitra) were in fact originally written in Tibetan, not translated into it (Horiuchi 2021b). Furthermore, it is appropriate to call them not “Indian commentaries on the HS”,4 as previously suggested, but rather “commentaries on the long HS preserved in the Tibetan canon”. Previously, I eclectically placed the word Indian into parentheses (thus, “Indian”); however, in this paper, I would like to propose the above terminology for the first time. Furthermore, this term can include both the eight commentaries and also the ones by such as rNgog blo ldan shes rab (hereafter rNgog) and Tāranātha, which were originally written in Tibetan, and will contribute to clarifying these texts from a broader perspective.

1.2. Vimalamitra’s Commentary

Among the commentaries on the long HS preserved in the Tibetan canon, I here focus on Vimalamitra’s (ca. the 8th- to 9th cent.) commentary *Āryaprajñāpāramitāhṛdayaṭīkā (hereafter PHT) for the following four reasons. First, it is the most detailed and carefully annotates the HS word for word. In particular, the commentary criticizes other schools in commenting on the HS. Second, it is certain that the text was not written originally in Tibetan but was annotated by an Indian scholar; suggestions on how to read the HS in an Indian context can be derived. Third, Vimalamitra’s thought, as shown in his PHT, needs further elucidation, despite the existence of the previous studies. Lopez (1996), Tan and Liu (2005), and Ōyagi (2016) translated the PHT, identified most of the texts cited therein, and presented its contents. Furthermore, Mathes (2021) offered his own opinion on the ideological standpoint of the PHT. However, translated Tibetan literature is challenging to handle and requires rigorous philological investigation, which has been lacking in previous studies. Often, important points have been misunderstood. Fourth, the PHT was considered such an important commentary on the HS that two later Tibetan commentaries (by Atiśa and rNgog blo ldan shes rab (hereafter rNgog)) can be regarded as subcommentaries on the PHT.5

Incidentally, there are studies on Vimalamitra’s other works, but they are not directly relevant to this paper. However, since the commentary on the Saptaśatikā Prajñāpāramitā (Perfection of Wisdom in 700 Verses: ’Phags pa shes rab kyi pha rol tu phyin pa bdun brgya pa’i rgya cher ’grel pa *Ārya-Saptaśatikāprajñāpāramitāṭīkā, hereafter SPT), is closely related to the PHT in that it contains several parallel discussions and citations (Lopez 1996, p. 22), which will be examined with the PHT in detail.

Here, I will describe the methodology I have adopted in analyzing the PHT. The first is to read the PHT by thoroughly consulting Vimalamitra’s other work: the SPT. The second is to refer to the rNgog, which is a subcommentary to the PHT. The third is to refer to the P[eking] and D[erge] editions and also another edition of the PHT that I named the T edition.6 The fourth is to clarify Vimalamitra’s position in the history of Madhyamaka thought by comparing it with similar aspects found in the writings of Kamalaśīla, an older contemporary of Vimalamitra.

On this last point, I note that Śāntarakṣita (ca. 725–788) entered Tibet in 779, and Kamalaśīla (ca. 740–795) arrived in 794. On the other hand, Vimalamitra is said to have entered Tibet around the time of the death of King Khri srong lde btsan, that is, around 797, and remained there, engaging in the translation of Sanskrit texts into Tibetan and composition of texts, including the PHT, for 13 years (Akahane 2004, pp. 49, 62). Therefore, to clarify Vimalamitra’s ideological background, it would be helpful to consult Kamalaśīla’s works, such as the Bhāvanākrama (BhK), the Tattvasaṃgrahapañjikā (hereafter TSP) and the Madhyamakāloka (hereafter MA).7 Incidentally, it is also noteworthy that, although this work is extremely brief (P330b6-333a6), Kamalaśīla also wrote a commentary on the HS before Vimalamitra, as noted in Section 1.1.

1.3. The Scope of This Article

With regard to Vimalamitra’s philosophical position in the PHT, Mathes (2021) stated “Vimalamitra leaves no doubt that he shares Nāgārjuna’s view of things” (p. 665). However, he also considered some passages in the PHT to show that “Vimalamitra also endorses the Yogācāra model of reality”. (p. 665)

However, I disagree with this. A detailed reading of the relevant passages, including their context, rather clearly suggests that Vimalamitra is criticizing the Yogācāra’s standpoint or model of reality and that the PHT is written sorely from the Madhyamaka perspective. To prove this, I primarily examine Vimalamitra’s criticism on other schools, as expressed in his interpretation of the famous four sets of phrases, such as “rūpa (form)8 is emptiness/empty”. This part is important because Vimalamitra refers to and criticizes other schools within and outside of Buddhism in interpreting those HS passages. Based on a detailed structural analysis of the relevant section, I argue that Vimalamitra regards that the HS criticizes those other schools, including the Yogācāra. Furthermore, I will clarify Vimalamitra’s understanding of the key Buddhist concepts such as “emptiness” or “dependent origination” as a Mādhyamika, who advocates the ideas of the Madhyamaka school, in a way that highlights the differences between his understanding and that of other schools (Section 2). Finally, I discuss the significance of the PHT to modern readers of the HS.

The last point concerns how the enigmatic HS is to be read. The HS’s intent is not always clear, and it seems that commentators, both ancient and modern, have read their own ideological positions into the HS. In this regard, I will cite a comment on the “Indian” commentaries on the HS by Eckel (1987), quoted in Attwood (2017, p. 54).

[T]o approach the Indian commentaries in the hope that they will somehow yield the ‘original’ meaning of the text is to invite disappointment... What they thought it meant was shaped as much by the preoccupations of their own time as it was by the words of the sūtra itself.

(Eckel 1987, pp. 69–70)

Indeed, the PHT is no exception here. However, a new reading of the passage in the PHT, which I discuss in Section 3, especially in Section 3.2, shows that Vimalamitra’s understanding of the HS is in keeping with the basic lines of Mahāyāna Buddhism and has a religious significance that can appeal to modern readers of the HS and that is also in keeping with the new basic lines presented by some modern scholars.

2. The PHT’s Understanding of the HS: In Controversy with Other Schools

2.1. Works Cited in the PHT

The PHT cites a number of sūtras and śāstras. As the Prajñāpāramitā literature, it cites the Aṣṭasāhasrikā, Pañcaviṃśatisāhasrikā, Ratnaguṇasaṃcayagāthā, and the Abhisamayālaṃkāra. As Mahāyāna sūtras, it cites the Candrapradīpa (Samādhirājasūtra), Saṃdhinirmocanasūtra (hereafter SNS), Lalitavistara, and Daśabhūmikasūtra, etc. As śāstras, it cites Maitreya’s Abhisamayālaṃkāra and Nāgārjuna’s Yuktiṣaṣṭikā and Ratnāvalī. It also cites works by Dignāga and Dharmakīrti (Prajñāpāramitāpiṇḍārthasaṃgraha and Pramāṇavārttika, respectively).

Although it cites a tantric text Vairocanābhisaṃbodhi, as previous studies have noted, no particularly esoteric elements are found in the PHT (Ōyagi 2016, p. 73). However, the fact that he cites much of the text of the PHT does not necessarily clarify his philosophical position.

2.2. Basic Tenet of the PHT

Regarding the ideological position of Vimalamitra in the PHT, Mathes (2021) succinctly and appropriately highlights:

Vimalamitra leaves no doubt that he shares Nāgārjuna’s view of things, which becomes most clear by the fact that he adduces verse 19 of the Yuktiṣaṣṭikākārikā:

That which has arisen dependently

Has not arisen in terms of an own nature

How can that, which has not arisen in terms of an own nature,

Truly be called ‘arisen’?

(p. 665)

A more in-depth and detailed discussion of Vimalamitra’s ideological position in relation to Mādhyamika is given in Section 3.1, but since he criticizes other schools based on this basic position, it will be sufficient to review these above points for discussion in the succeeding sections.

2.3. “Four Formulas of Emptiness” and Critique of Other Schools

In this section,9 I investigate the PHT’s interpretation on the famous formulas of emptiness in the HS: (A) rūpaṃ śūnyatā/śūnyam (rūpa is emptiness/empty), (B) śūnyataiva rūpam (precisely the emptiness is the rūpa), (C) rūpān na pṛthak śūnyatā (emptiness is not different from the rūpa), and (D) śūnyatāyā na pṛthag rūpam (rūpa is not different from the emptiness), which I will name the “four formulas of emptiness”. In interpreting these formulas, Vimalamitra refers to other schools and presents an interesting dispute. Hence, this section provides important material on the fundamental Buddhist ideas of emptiness and dependent origination. In addition, it discusses the broader Indian philosophical issue of continuity and change.

I divide this section of the PHT into five paragraphs.

PHT: [1] D272a6, [2] D272a6-, [3] D272b1-, [4] ([4.1]) D272b3-5; [5] D273a7ff. (Lopez 1996, pp. 56–58)

There is a further discussion after [4.1] above in D272b5–273a7. Of these, I will analyze only two parts as [4.2] and [4.3] in Section 3.1. They are just one part of the PHT that ranges D267b1–280b7.

• [1] rūpaṃ śūnyam (PHT, D272a6, P291b, T13):

[The HS/Avalokiteśvara] said: “(A) rūpa is empty”. It (i.e., this sentence) is syntactically connected (sbyar) to [the phrase] “with regard to [its] own nature (svabhāva-)”, since it is the subject [here].

((A) gzugs stong pa’o zhes bka’ stsal te/ skabs yin pa’i phyir rang gi ngo bo nyid kyis zhes ’byung ba dang sbyar ro//)

Although there is no problem with the preceding translations of this passage, I will further examine it in relation to the structure of the HS.

First, as Mathes (2021) highlights:

The first part of our formula (‘form is emptiness’) is quoted in its variant reading ‘form is empty’, and Vimalamitra adds “that because of the context one has to add ‘in terms of own nature (svabhāva)’”.

(pp. 664–65)

The phrase that appears in the short HS that “rūpa is emptiness (rūpaṃ śūnyatā)” (Conze 1967, p. 150) has been interpreted in various ways. However, according to Vimalamitra, it means nothing more than that “rūpa lacks intrinsic/its own nature (rūpaṃ (Nominative) svabhāvena (Instrumental) śūnyam)”. Since the basic syntax of emptiness is “something (Nominative) is devoid of (śūnyam) something (Instrumental)”, this is understood as following this grammatical pattern. This may be considered mundane; it also means that Vimalamitra interprets this expression in the HS from the basic line of thought of emptiness.

Next, I will clarify the terms “skabs, topic” and “sbyar, syntactical connection”. This passage corresponds to paragraph J in the division of the long HS by Silk (1994). In paragraph I immediately preceding it, it has: pañca skandhās tāṃś ca svabhāvaśūnyān samanupaśyati sma (he saw that there are five aggregates, and [he saw] them to be empty of their own nature). In addition, in paragraph E, it has: pañca skandhās tāṃś ca svabhāvaśūnyān vyavalokayati (he (i.e., Avalokiteśvara) observes that there are five aggregates, and [he observes] them to be empty of their own nature). The expression skabs, topic, refers to these statements, and sbyar, meaning “syntactical connection or syntactically connected” means that the word svabhāva should be supplied or connected to the phrase in paragraph J because of the topic that is already mentioned in E and I.

Furthermore, as paragraph J says, “evam eva vedanāsaṃjñāsaṃskāravijñānam (the very same is true of sensation, representation/image, predispositions, and consciousness)” Vimalamitra interprets the above phrase as follows:

PHT, D274, P294b, T18, Horiuchi 2021b, pp. 66–67:

“In the same manner (as above)”, namely, as in the case with (*iva) rūpa, [five] aggregates beginning with “sensation” and ending in “consciousness” are, in sum, taught to be “empty of their own nature”.

Therefore, it should be understood in the following way (*evam): (A’) sensation (vedanā) is empty, (B’) emptiness is sensation, (C’) emptiness is not different from sensation, and (D’) sensation is not different from emptiness. Similarly, the [same phrase] should be applied to recognition (saṃjñā) and so on.

(de bzhin du gzugs dang ’dra bar tshor ba la sogs pa nas (nas] PT; na D) tha ma rnam par shes pa la thug pa phung po dag mdor (mdor] DP; don mdor T) na rang gi ngo bos (bos] DP; bo T) stong pa nyid du bstan to//

des na ’di ltar rig par bya ste/ tshor ba stong pa’o// stong pa nyid tshor ba’o// tshor ba las stong pa nyid gzhan ma yin no// stong pa nyid las tshor ba gzhan ma yin no// de bzhin du ’du shes la sogs pa (pa] DP; pa la T) yang sbyar bar bya’o//)

Thus, this paragraph J is an illustration or further elaboration on the statement in paragraphs E and I.

Paragraph E: pañca skandhās tāṃś ca svabhāva-śūnyān vyavalokayati.

Paragraph I: pañca skandhās tāṃś ca svabhāva-śūnyān samanupaśyati sma.

Paragraph J: (A) rūpaṃ [svabhāva-]śūnyam … (B–D; The same applies to the rest of the five skandhas)

• [2] Two assumptions of rūpa (PHT (continuance from [1]))

These two types are assumed10 to relate to an appearance (*ābhāsa) such as rūpa: (1) there are those who advocate that [they are] real entities (*vastu), and (2) there are those who advocate that [they are] designations (*prajñapti).11

(1’) Of these, Vaibhāṣika, Mīmāṃsaka, and so on advocate that an appearance such as rūpa (*rūpādy-ābhāsa) is an entity (*vastu).

(2’) The Sautrāntika[s] are those who advocate that the appearance of an object (*artha-ābhāsa) is a designation (*prajñapti). They say that “an object such as [a] blue (*nīlādyartha)12 precisely (*eva) exists externally. Experience (*anubhava), however, is an image (*ākāra), which is perfumed/impregnated (*vāsita) in this consciousness (*vijñāna)”.13

(’di ltar gzugs la sogs par snang ba la (1) dngos po yin pa dang/ (2) btags pa yin par smra (smra] DT; snang P) ba rnams rnam pa de gnyis su rtog (rtog] PT; rtogs D) par byed do//

(1’) de la gzugs la sogs pa snang ba dngos po yin par smra ba ni bye ^[P292a] brag tu smra ba dang/ dpyod (dpyod] DT; spyod P) pa pa la sogs pa’o//

(2’) don snang ba btags pa yin par smra ba ni mdo sde pa ste/ de dag ’di skad du don {ni} sngon po la sogs pa phyi rol ^[D272b] na yod pa nyid (nyid] DT; φ P) yin la/ myong ba ni rnam par shes pa ’di la rnam pa gang bsgos (bsgos] PT; bsgom D) pa yin no//)

In (1), the Vaibhāṣika and Mīmāṃsaka are mentioned, and in (2), the Sautrāntika is. The first two are here labeled nirākārajñānavāda or nirākāravāda and the latter is called sākārajñānavāda or sākāravāda. Let us take an overview of the difference, following Kajiyama’s (1965) explanation.

Sākāravāda, or the theory that knowledge is endowed with the image of its object, is maintained by the Sāṃkhya, Vedānta as well as the Sautrāntikabauddha. The theory, in Mookerjee’s words, “holds that knowledge of external reality is made possible by virtue of the objective reality leaving an impress of its likeness on the mirror of consciousness. The nirākāravāda is held by the Nyāyavaiśeṣika, Mīmaṃsaka, Jaina, and the Vaibhāṣikabauddha, and the theory maintains that our consciousness is like a clean slate and does not depart an inch from its intrinsic purity even when it apprehends the external reality. Consciousness is an amorphous substance and remains so in all its activities. It is like light and reveals the object with its form and qualities without undergoing any morphological articulation in its constitution.

(p. 26)

In brief, according to the nirākārajñānavāda, consciousness has no form and perceives the object itself. According to the sākārajñānavāda, what we recognize is an image. However, an outside object can be assumed to exist by inference (Kajiyama 1983, p. 6ff.).

• [3] On (A) rūpaṃ śūnyam, (B) śūnyataiva rūpam (PHT (continuance from [2])):

(1’’) Many imaginations of own nature with regard to blue and so on, namely, bad nets that are imagined in various modes (yongs su gyur pa, lit. transformation) such as atom (aṇu), the whole (avayavin, lit. having the parts) that is constituted by that (i.e., aṇu), consciousness (vijñāna), the primary germ (pradhāna), and the Brahman that is word (śabdabrahman), all of them will be destroyed just by denying intrinsic/the own nature (svabhāva). Therefore, precisely that [i.e., the phrase: (A) rūpaṃ [svabhāva]śūnyam] should be stated.

Precisely (*eva) the “rūpa” that is clearly appearing as an entity that is well known—even among cowherds and women14—is “empty of intrinsic nature” like the city of the Gandharva (*gandharvanagara). Namely, there is no intrinsic nature at all in this appearance of rūpa. Therefore, this “emptiness of intrinsic nature” is called “rūpa” (i.e., (B) śūnyataiva rūpam). By this [phrase, (B)] the emptiness of intrinsic nature of rūpa that Vaibhāṣika Mīmāṃsaka, and others claim is explained.

((1’’) sngon po la sogs pa la rang gi ngo bor rnam par rtog pa mang po rdul phra rab dang/ des brtsams pa cha shas can (*avayavin) dang/ rnam par shes pa dang/gtso bo (*pradhāna) dang/ sgra’i tshangs pa (*śabdabrahman) la sogs pa yongs su gyur pa tha dad par rtog pa’i dra ba ngan pa de dag thams cad ni rang gi ngo bo bkag pa nyid kyis bcom par ’gyur bas de nyid (i.e., (A) rūpaṃ śūnyam) brjod par bya’o//

gang de dag (dag] DT; dag la P) ba lang rdzi dang bud med yan chad la grags pa dngos su gsal bar snang ba’i gzugs de nyid dri za’i grong khyer ltar ngo bo nyid kyis stong pa yin te/ gzugs snang ba ’di la rang gi ngo bo ’ga’ yang med do// de bas na (B) rang gi ngo bo stong pa nyid ’di la gzugs so zhes bya ste/ ’dis ni bye brag tu smra ba dang/ dpyod (dpyod] DT; spyod P) pa pa la sogs pa ’dod pa’i gzugs rang gi ngo bo stong pa nyid du bstan pa yin no//)

Here, the theories of both the Buddhist and non-Buddhist schools are enumerated. Still, Vimalamitra’s refutation of the first position (1) is simple enough; he mentions five non-Buddhist doctrines and claims only that they are all rejected “just by denying intrinsic/the own nature (svabhāva)”, and there is no further explanation given of any theory or any detailed critique. The preceding translations of the PHT also present only translations. In this connection, we recall that the Tattvasaṃgraha (hereafter TS) of Śāntarakṣita, an older contemporary of Vimalamitra, and its commentary by his pupil, Kamalaśīla’s TSP, criticize other schools in detail. Therefore, we review these statements on the basis of previous studies and clarify the doctrines referred to in the PHT.

The terms “atom (aṇu)” and “the whole (avayavin)” refer to the Nyāya and Vaiśeṣika theories. This is also criticized in Kamalaśīla’s MA.15 Keira (2004) notes:

The Nyāya-Vaiśeṣikas say that whole (avayavin) is another substance (dravya) than its parts (avayava). They hold that the simple collection of parts, i.e., atoms (paramāṇu), cannot be the object of perception, because atoms are imperceptible things.

(p. 186)

The atoms and the whole are mentioned in the MA in the context of a discussion of the object of perception. The PTH also refers to them in the same context (about the nature of blue and so on).

The next term, consciousness (rnam par shes pa, vijñāna), recalls the Yogācāra’s doctrine. However, a range of theories on consciousness or cognition are criticized in the MA; Keira (2004) provides a subject header, “Cognition (vijñāna) is not one in nature”, in his translation and edition of a part of the MA at this point and shows that the Nirākāravādins, Sautrāntikas, Vaiśeṣikas, Jainas and Mīmāṃsakas, Cārvākas, Sāṃkhyas, and Yogācāras are criticized in the text here (pp. 189–96 (translation); pp. 255–57 (text)). The discussion concludes as follows.

4.2.1.1.2.9 No cognition is many in nature

§23. Therefore, cognition is not established as being one in nature in any doctrine at all. When that [i.e., oneness] is not established, manyness is not established either. This has been explained earlier.

(Keira 2004, p. 196)

Incidentally, the above is an application of one of the five reasons for the things to be lacking intrinsic nature (see Section 3.1), named “ekānekaviyogahetu, the neither one nor many reason”.16 Vimalamitra does not use the same argument, but he criticizes those theories by denying intrinsic nature as well.

The next term “the primary germ (pradhāna)” refers to the prakṛti of Sāṃkhya. This is one of the two main principles of the school, along with puruṣa (soul or spirit). Prakṛti=pradhāna is the original source of the material world from which the material world evolves according to the Sāṃkhya school. In the MA, this is introduced with the following introductory phrase.

§7. Time (kāla), Spirit (puruṣa), primordial matter (pradhāna), Brahman and so forth, which are imagined by some people to be the causes establishing the various worlds.

(Keira 2004, p. 184)

This is also criticized by the same “neither one nor many reason” as before (Keira 2004, p. 185).

The last term, sgra’i tshangs pa in the PHT has not been understood correctly by the previous studies,17 but it refers to the śabdabrahman, a well-known doctrine of Bhartṛhari.18 The point of this doctrine is to advocate the identification of language or word (śabda) with the Absolute as Brahman (Bilimoria 1995, p. 138). The MA criticizes Brahman together with pradhāna as we have seen above. However, there is no mention of śabdabrahman in the MA. The TS, on the other hand, has a chapter called śabdabrahmaparīkṣā (examination of śabdabrahman), which suggests that his theory was considered an important opponent at the time. Since the TS, in criticizing it, presents the gist of the doctrine in the form of a verse, I quote the English translation by Jha (1937).

The whole lot of things is recognised as evolved out of that Brahman which is of the essence of Word-Sound, the Highest—unaffected by destruction and origination.

(p. 118)

(TS: nāśotpādāsamālīḍhaṃ brahma śabdamayaṃ param |

yat tasya pariṇāmo ’yaṃ bhāvagrāmaḥ pratīyate || 128 ||)

The above is a supplemental explanation of the theories criticized by Vimalamitra from the writings of an older contemporary. Compared to the detailed criticisms provide for by Śāntarakṣita and Kamalaśīla, Vimalamitra’s criticism is quite simple. However, criticizing them in relation to the phrase of the HS is something that Kamalaśīla’s HS commentary does not do either and is unique to Vimalamitra.

What is important in relation to the following argument is that Vimalamitra is interpreting the first two sūtra phrases ([A] and [B]) as antidotes to the incorrect view by other schools such as Vaibhāṣika.

In summary:

(A) rūpaṃ [svabhāva]śūnyam, (B) śūnyataiva rūpam

Refute the Vaibhāṣika, Mīmāmsaka, etc.

Incidentally, regarding the difference between (A) and (B), the following statement in the rNgog, which is a subcommentary to the PHT (see Section 2.4 for details), is worth noting.

rNgog (Horiuchi 2019, p. 130):

“(A) Rūpa is empty” negates ultimate rūpa, and “(B) emptiness is rūpa” establishes conventional rūpa.

(gzugs stong pa’o zhes (zhes] P; φ C) bya bas don dam pa’i gzugs dgag ^[P366] pa dang/ stong pa nyid gzugs so zhes bya bas kun rdzob kyi gzugs bsgrubs pa(s))

• [4] On (C) rūpān na pṛthak śūnyatā, (D) śūnyatāyā na pṛthag rūpam (PHT (continuance from [3])):

[4.1]

(2’’) If the Sautrāntikas think that (snyam du sems na) “Appearance, such as blue (*nīlādy-ābhāsa), is empty of [its] own nature of [the outer] object (*arthasvabhāvena) inasmuch as appearance has the nature of consciousness (*vijñānātmakatvāt). However (ni, *tu), inasmuch as [an] outer object (*bāhyārtha) surely exists (*asty eva), rūpa (form, form and color) is different from emptiness”,19 they (Sautrāntikas) are also refuted by this [phrase/these phrases]: namely, “(C) emptiness is not different from rūpa (rūpān na pṛthak śūnyatā) and (D) rūpa is not different from the emptiness (śūnyatāyā na pṛthag rūpam)”.

([4.1]

(2’’) mdo sde pa rnams snang ba rnam par shes pa’i ngo bo yin pa’i phyir sngon po la sogs par snang ba ni don gyi rang gi ngo bos stong pa yin no// phyi rol gyi don ni yod pa ^[T14] kho na yin pas stong pa nyid las gzugs gzhan yin no snyam du sems na/ de dag kyang (C) gzugs las stong pa nyid gzhan ma yin no// (D) stong pa nyid las kyang (kyang] DP; φ T) gzugs gzhan ma yin no zhes ’byung ba ’dis ’gog par byed do//)

The Sautrāntikas provide an understanding of the emptiness of rūpa (rūpaṃ [svabhāva]śūnyam) in that they are understanding the object (≒ rūpa) as appearance that has the nature of consciousness (see Kajiyama’s (1965) explanation that I cited in [2]). However, it is only with regard to the emptiness of the object, which is appearance in consciousness. They advocate the existence of outside objects. Therefore, to refute this position, according to Vimalamitra, the two phrases were spoken in the HS.

(C) rūpān na pṛthak śūnyatā, (D) śūnyatāyā na pṛthag rūpam

Refute the Sautrāntika

• [5] On the Yogācāra view (PHT, D273a7ff., P293a5ff., T15ff.):

[5.1]

Another perspective: the discriminated form (*vikalpita-rūpa), that is, the dependent characteristic (*paratantra-lakṣaṇa) is constantly and eternally (*nityaṃ nityakālaṃ dhruvaṃ dhruvakālaṃ) deprived of (rahita) [, namely/and is] empty (*śūnya)20 of the imagined/conceptualized form (*parikalpita-rūpa21), that is, the two characteristics of object and subject (*grāhya-grāhaka-lakṣaṇa). That (devoidness or emptiness) is the form of reality (*dharmatā-rūpa) that is perfected nature (pariniṣpannasvabhāva).22

[5.2]

[5.2.1] By this [phrase] “(B) the very emptiness is rūpa”, [the HS/Avalokiteśvara] shows the identity (*ekatva) also (yang) of the two, namely, of dependent (*paratantra) and perfected (*pariniṣpanna). [Namely, the] very [phrase] “(B) the very emptiness is rūpa” is stated to determine the identity of the emptiness, namely the perfected nature (*pariniṣpanna-svabhāva), and rūpa, namely, the dependent characteristic (*paratantra-lakṣaṇa).23

[5.2.2] Thus, after stating the identity through the establishment (i.e., positively), through negation (i.e., negatively) too (kyang), [the HS/Avalokiteśvara] denies the distinction [of emptiness = pariniṣpanna and of rūpa = paratantra] through [the phrases] “(C) emptiness is not different from rūpa” and so on (i.e., [C] and [D]).24

([5.1]

rnam grangs gzhan yang rnam par brtags pa’i gzugs gzhan gyi dbang gi mtshan nyid ni kun tu brtags pa’i ngo bo gzung (gzung] DP; bzung T) ba dang ’dzin pa’i ^[D273b] mtshan nyid gnyis kyis (kyis] DP; dang/ T) rtag pa rtag pa’i dus dang ther zug ther zug gi dus su bral ba stong pa de ni chos nyid kyi gzugs te/ yongs su grub pa’i mtshan nyid do//

[5.2]

[5.2.1] gzhan gyi dbang dang yongs su grub pa de gnyis ka (gnyis ka] DP; gnyi ga T) yang (B) stong pa nyid gzugs so zhes ’byung ba ’dis (’dis] PT; ’di D) gcig pa nyid du ston to// stong pa nyid yongs su grub pa’i ngo bo dang/ gzugs gzhan gyi dbang gi mtshan nyid gcig pa nyid du nges par gzung ba’i phyir/ (B) stong ^[T16]pa nyid gzugs so zhes bya ba (ba] DP; ba la T) nyid smos pa yin no//

[5.2.2] de ltar gcig pa nyid sgrub pa’i sgo nas brjod nas dgag pa’i sgo nas kyang/ (C) gzugs las stong pa nyid gzhan ma yin no (no] DP; φ T) zhes ’byung ba la sogs pas ^[P293b] tha dad pa nyid ’gog par byed do//)

2.4. Vimalamitra as a Mādhyamika: A Critique on the Yogācāra

As noted in Section 1.3, Mathes (2021) offered his own opinion on the ideological position of the PHT. He analyzed passages [5.1] to [5.2.1] in my division of paragraph, showing the Tibetan text and English translations. In doing so, Mathes prefaces the text with the statement that “Vimalamitra also endorses the Yogācāra model of reality”. Subsequently, he presents his translation of [5.1] and [5.2.1] that begins with “An alternative way [to explain form and emptiness] is as follows:”. (p. 665). Consequently, he understands [5.1] and [5.2.1] as Vimalamitra’s position.

However, I disagree with this opinion: I understand [5.1] as a presentation of the theories of the Yogācāra school and [5.2.1] (and [5.2.2]) as criticisms of them by Vimalamitra.

This is evident from the structure of the PHT discussion so far: in [2], for example, the theories of other schools were presented, and in [3], Vimalamitra presented (A) and (B) of the “four formulas of emptiness” as criticisms of the other schools’ theories. In [4] too, the Sautrāntika’s theory was presented initially, and then Vimalamitra presented the formula (C) and (D) in the HS as criticisms of it. If so, it is reasonable to assume that a similar structure is used here.

In [5.1], the three-natures (trisvabhāva) theory is presented. However, in [5.2], paratantra and pariniṣpanna are considered identical. However, in the Yogācāra, pariniṣpanna and paratantra are not stated as identical but “not different [and] not identical”. I will cite Vasubandhu’s Triṃśikā on this point:

Triṃśikā: For precisely this reason, it (i.e., pariniṣpannasvabhāva or perfected nature) is not different [and] not identical with dependent (paratantra) [nature].

(ata eva sa naivānyo nānanyaḥ paratantrataḥ | 22ab)

This also indicates that [5.2] is not the Yogācāra’s theory, but that Vimalamitra as a Mādhyamika criticizes the Yogācāra’s model of reality: three-natures (trisvabhāva) theory.

Moreover, for this PHT, the rNgog is useful for reading the PHT, as it presents a synopsis and briefly annotates the words. There is a continuation of this argument, excluded from this paper, but which can be divided into [5.3] and [5.4]. The rNgog takes the series of arguments beginning with [5.1] as the denial of the other’s system, and further subdivides it into four subdivisions. Among them, [5.1]–[5.3] are given the heading as “(i) the setting up of the qualities of the emptiness of/by others and (ii) the refutation on it”.

rNgog (Horiuchi 2019, p. 131):

(2) The denial of the other’s system, which is the second [explanation on the particularity of emptiness], is fourfold: (i) the setting up of the qualities of the emptiness of/by others, (ii) the refutation on it …

(i’) The first part is [5.1] “taking it from another point of view (interpretation)”, up to “grag go [thus the Yogācāra] says”. (ii’) The second [5.4] continues up to “Nirvāṇa also is like a dream, like an illusion”.

((2) gnyis pa gzhan gyi lugs dgag pa la bzhi ste/ (i) gzhan gyi stong pa’i mtshan nyid rnam par gzhag pa dang/ (ii) de sun dbyung ba dang/ ...

(i’) dang po ni [5.1] rnam grang gzhan yang nas grag go zhes bya ba’i bar ro// (ii’) gnyis pa ([5.4]) ni mya ngan las ’das pa yang sgyu ma lta bu rmi lam lta bu’o zhes bya ba’i bar ro// ...)

The grag go appearing in (i’) refers to the last word of [5.3] in the PHT, although it differs from the readings in the editions (PHT, D273b [grags te], P293b, T16 [grag ste]). Since the following phrase de la bshad pa (Reply to that: …) clearly refers to the beginning of a reply from Vimalamitra, the reading of grag go, which indicates the ending of the range of the other’s opinion, is more appropriate for the PHT.

In summary, the rNgog understands [5.1–5.3] the opinion of the Yogācāra. I agree that [5.1] and [5.3] is the Yogācāra’s theory; however, as for [5.2], I understand it to be a criticism by Vimalamitra. In summary, the following

Table 1. Affiliations of statements in the PHT.

PHT’s Passage	Mathes	rNgog	Horiuchi
[5.1] D273a7-: rnam grangs gzhan yang …	V	Y	Y
[5.2] D273b1-: gzhan gyi dbang dang …	V	Y	V
[5.3] D273b3-4: de yang mi ’gyur ba … grag ste/grags te (grag go in the rNgog)	(not mentioned)	Y	Y
[5.4] D273b4-: de la bshad pa …	(Do.)	V	V

Note: “V” stands for Vimalamitra; “Y” for the Yogācāra.

is the case:

The rNgog and I agree that Vimalamitra’s position is only that of a Mādhyamika and he did not endorse the Yogācāra view.

Incidentally, the reason that “(A) rūpaṃ [svabhāva]śūnyam” is not applied here must be that Vimalamitra believes that the Yogācāras, as the Sautrāntikas, understand the emptiness of rūpa in a certain way.

(B) śūnyataiva rūpam: pariniṣpanna and paratantra are identical (a statement in a positive expression)

(C) rūpān na pṛthak śūnyatā, (D) śūnyatāyā na pṛthag rūpam: do. (a statement in a negative expression)

Refute the Yogācāra

3. Vimalamitra’s Madhyamaka Position and the Importance of the PHT to Understand the HS

In this section, we will first discuss a part of the continuation of the [4] of the PHT in Section 2.3 and examine further discussions on the idea of the dependent origination in the PHT (3.1) and investigate Vimalamitra’s Madhyamaka position further. Next, I will highlight that the philosophical argument is backed by meditation and practice (3.2). Finally, I will overview Vimalamitra’s interpretation as a Mādhyamika master regarding the state of being reached after passing through the meditative practice (3.3).

3.1. Vimalamitra’s Madhyamaka Position

The PHT, after [4.2] in Section 2.3, continues as follows:

PHT, D272b, P292a, T14:

[4.2] [Q] Furthermore, how should it be known that there is none of one’s own nature of skandhas and so on (*skandhādi-svabhāva-abhāva)?

[A] Answer: Since skandhas (aggregates), āyatanas (spheres), and dhātus (elements)25 are originated dependently (pratītyasamutpanna), the position that [things are] without cause and that [things are] permanent is denied.

[4.3] The position that a thing arises from a cause also entails two possible considerations (brtag pa; or rather read rtog pa [options]). When a thing (i.e., an effect) arises, it can either (i) arise at the same time as the cause or (ii) arise at a different time.

(i’) In the case of the first position (simultaneous arising), [there are three absurdities:] (a) cause and effect, i.e., all of the preceding and following positions,26 would be perceived simultaneously, (b) cause and effect would be indistinguishable, and (c) an eon (*kalpa) would be a moment.

(ii’) As for the second position (arising at different times), since the two, namely, causes and effects do not unite simultaneously, there is no capacity (nus pa med) [for the cause]. Therefore, it follows that there are no causes. If there are no causes, it follows that [things] always exist or [always] do not exist. For there is no other thing on which [they] rely. If they rely on [something], then things would be temporary.27

[However,] there can be no third position. The two [positions] of simultaneity and at different times abide, mutually excluding each other.

[Conclusion:] Therefore, an appearance such as rūpa is “empty of intrinsic nature” like a mirage.

Regarding them (appearances) as well, the assumption that they originate being dependent on this or that is convention (*saṃvṛti).28

As I first highlighted in 2018, there is a parallel argument to that given above [4.3] in the SPT by the same Vimalamitra and in Kamalaśīla’s BhK I, 201. Since I presented only the text of the SPT therein, here I will present a translation and further discussion of it.

SPT, D ma 14a5-7, P ma 17b8–18a2:

When an effect arises, (i) it may occur simultaneously to the cause or (ii) at a different time.

If [it is] (i’) simultaneous, then [there are three absurdities:] (a) everything would be perceived at the same time, (b) cause and effect would be indistinguishable, and (c) an eon would be a moment.

(ii’) Even if the time were different, there would be no cause because there would be no capacity [for the cause]. If there is no cause, then there is nothing to rely on, so it will always exist, or not exist.

[Conclusion:] Therefore, ultimately (*paramārthatas), there is never any such thing as arising.29

[4.2] [A] of the PHT says that since things are originated dependently (pratītyasamutpanna), the position that things are without cause and that things are permanent is denied. If this is the case, then there is a cause for the arising of phenomena. In [4.3], the PHT further examines the cause as follows.

Arising means that an effect is produced from a cause. Therefore, it is either the cause and effect exist (i) simultaneously or (ii) at different times. According to the PHT, first, if the former, three absurdities arise. I would like to address one small but important textual issue. Here, the earlier translations of the PHT that are based solely on the DP and have not examined the relevant literature translated the third absurdity (c) as “an eon would be just one year”.30 However, there is an interesting variant in the T edition that previous studies did not consult. According to this, (c) means “an eon (kalpa) will be one moment (kṣaṇa)”. Moreover, there is a similar discussion of this back and forth in Kamalaśīla’s BhK, where there is an expression “kalpasya kṣaṇamātratāprasaṅgaḥ” (there is an undesirable consequence that an eon [kalpa] becomes one moment; BhK I, 201.22–23). In addition, the SPT also has “(c) bskal ba yang skad cig gcig (*ekakṣaṇa) tu ’gyur ro//”. I understood it that way because this is a reasonable way to highlight the undesirable consequence (prasaṅga).

In addition, the phrase “nus pa med (there is no capability)” of the option that cause and effect differ in time (in (ii’)) would require explanation. If the time in which the cause exists is different from the time in which the result exists, then there is a disconnect between the two. This implies the irrationality that the cause cannot act on the result, namely, the cause has no capability to act on the result. Furthermore, what is incapable of producing an effect cannot be called a cause.

Finally, the PHT says that arising exists conventionally, while the SPT says it on the flip side of the coin.

Vimalamitra’s position can be said to be that of a Madhyamaka, but I further investigate to which extent this represents the Madhyamaka position by comparing it with Kamalaśīla’s BhK.

As is well known, Kamalaśīla enumerated five reasons for the things to be lacking intrinsic nature (Cf. Ejima 1980, p. 232ff; Keira 2004, p. 9ff). Among his many works, the BhK provides a summary description of the first and the fifth reason from the Madhyamakāloka (Ichigō 2011, p. 23ff.). The first reason is termed the diamond-splinters (rdo rje gzegs ma’i gtan tshigs; vajrakaṇahetu) and is summarized as follows in the Madhyamakāloka (Keira 2004).

[Vyāpti:] What is ultimately void of being produced by itself, by others and by both and of being produced without any cause, is really without intrinsic nature. It is just like, for example, sky-lotuses and so forth.

(gang dag don dam par rang dang/ gzhan dang/ gnyi ga las skye ba dang/ rgyu med pa las skye ba dang bral ba de dag ni yang dag par na ngo bo nyid med pa yin te / dper na nam mkha’i padma la sogs pa Ita bu’o//).

(p. 10)

As Ejima (1980, p. 233) points out, this is based on Nāgārjuna’s Mūlamadhyamakakārikā (hereafter MMK), I.1, and is based on the background of discussion by the MMK commentators.

Not from itself, not from another, not from both, nor without cause:

Never in any way is there any existing thing that has arisen.

(na svato nāpi parato na dvābhyāṃ nāpy ahetutaḥ |

utpannā jātu vidyante bhāvāḥ kva cana ke cana || I.1 ||)

(Siderits and Katsura 2013)

The argument developed in the BhK as a variation on this first reason is very similar to that of Vimalamitra’s argument, above. Let me present the synopsis of the passages in question by Tucci, the editor of the BhK I.

The non-origination of all things is ascertained by (a) āgama and (b) yukti (p. 199)

(a) āgama, authority (p. 199)

(b) yukti, arguing: b-1 origination is not uncaused (p. 200)

b-2 origination is not caused; criticism of God as creator (p. 200)

b-3 a noneternal cause is also illogical (p. 201)

The argument in [4.2][A] of the PHT echoes BhK’s b-1 and b-2, as b-2 denies arising from an eternal (nitya) cause, God (īśvara). Moreover, b-3 (BhK I, 201.13–202.8) is closely related to the PHT and the SPT, so let me give a detailed synopsis of my own.

“A noneternal cause is also illogical” or criticism of arising from impermanent causes.

(I) Since the past and future do not exist, there is no arising from them (Cf. b-1 above).

(II) There is no arising from the present cause either. Arising can be either (i) at the same time with the cause or (ii) at a different time. However, neither is appropriate.

(i’) If a cause is existent at the same time with an effect, the latter is considered completed regardless of the former.

(ii’) If a cause is existent at a different time with an effect, (ii’-1) the time is either disconnected or (ii’-2) not disconnected.

(ii’-1) If the time is disconnected, there is an undesirable consequence of arising from the past.

(ii’-2-1) If the time is not completely disconnected, there is an undesirable consequence that a kalpa (eon) is one moment (kalpasya kṣaṇamātratāprasaṅgaḥ), because all moments enter one moment.

(ii’-2-2) If only a part is disconnected, there is an undesirable consequence that a moment has a part.

(III) Criticism of arising from oneself (svatas) and both (ubhayatas, from oneself and others)31

[Conclusion]

Therefore, ultimately (paramārthatas), these things are precisely not arisen. However, since the arising is existent by convention (saṃvṛtyā), there is no contradiction with scripture, etc.

(BhK I, 202.2-4: tasmāt paramārthato ’nutpannā evāmī bhāvāḥ | saṃvṛtyā tūtpādasya vidyamānatvān nāgamādivirodhaḥ|)

The similarity between the above description of BhK, SPT, and PHT should be obvious.

Moreover, to the best of my knowledge, at least with reference to the expression “an eon (kalpa) will be one moment (kṣaṇa)”, this discussion first appears in Kamalaśīla. Specifically, it appears in Kamalaśīla’s Madhyamakālaṃkārapañjikā (Ichigō 1985, p. 172), but not in Śāntarakṣita’s Madhyamakālaṃkāra, which it annotates (Ichigō 1985, p. 173). It also appears in other writings of Kamalaśīla, namely, in the *Sarvadharmāniḥsvabhāvasiddhi (D no. 3889, 282a: bskal pa yang skad cig ma nyid du thal ba) and the *Āryaprajñāpāramitāvajracchedikāṭīkā (D no. 3817, 261a: bskal ba yang skad cig tsam du ’gyur ba). However, this expression or argument is only mentioned in later literature in the manner that follows Kamalaśīla’s writings: Abhayākaragupta’s (d. 1125) *Āryāṣṭasāhasrikāprajñāpāramitāvṛttimarmakaumudī (D no. 3805, 30b: bskal pa yang skad cig tsam du ’gyur te), Jñānavajra’s (ca. the 11th to 12th cent.) commentary on the Laṅkāvatārasūtra (D no. 4019, 113a: bskal pa yang skad cig gi tshad tsam nyid du ’gyur ba), and *Mādhyamikasiṃha’s (??) *Saṃkṣiptanānādṛṣtivibhāga (D no. 3898, 4b: bskal pa yang skad cig ma tsam nyid du thal bar ’gyur bas).

Based on the above evidence, Vimalamitra can be evaluated as a Mādhyamika in the vein of Kamalaśīla.

3.2. Meditative Context of the Emptiness

As Watanabe (2009, p. 252) highlights, the Indo-Tibetan commentaries on the HS allot the stages of practice such as five paths (i.e., path [mārga] of accumulation [saṃbhāra], preparation [prayoga], vision/seeing [darśana], meditation/cultivation [bhāvanā], and no further learning [aśaikṣa]) described by the Yogācāra and the Abhisamayālaṃkāra to the text of the HS. For example, Kamalaśīla’s commentary, which precedes the PHT, says:

With respect to that, beginning with Thus, even those five aggregates are empty of intrinsic existence through consciousness is empty sets forth the paths of (1) accumulation and (2) preparation. The eight terms from Śāriputra, it is thus through unfilled, set forth (3) the path of vision, which has a nature of the uninterrupted path and the path of liberation …32

Atiśa’s commentary, which is a subcommentary on the PHT, also says:

How one should practice is the path of accumulation, the path of preparation, the path of vision, the path of meditation, and the path of no more learning.33

The PHT quotes at length from the SNS, referring to the 11 stages of the bodhisattva (from the first stage to the Buddha’s stage).34

Accordingly, the following statement by Attwood (2017), which attempts to broadly read the HS in the context of the Prajñāpāramitā literature, is also worth noting.

If we read the Prajñāpāramitā literature as expounding an epistemology rather than an ontology, some of the apparently paradoxical statements become clearer, especially if we keep in mind the context of meditations in which experiences cease (at least temporarily) without the cessation of consciousness per se. (p. 71)

Here, I will further highlight that the above philosophical argument as expounded by the PHT is supported by the context of meditation.

First, the PHT associates various aspects of dharma (phenomena) with the three gates of liberation (*vimokṣamukha) or three concentrations (samādhis): śūnyatā (emptiness), ānimitta (without sign), and apraṇihita (without wish). The following is taken from Horiuchi (2021b), p. 65.

In the Chinese Prajñāpāramitāhṛdaya, there is the sentence: 諸法空相 zhufa kongxiang (1) 不生busheng (2) 不滅 bumie (3) 不垢 bugou (4) 不淨 bujing (5) 不増 buzeng (6) 不減 bujian. These six items are negative phrases possessing 不 (bu, without) at the beginning. The Sanskrit equivalent of 空相 is śūnyatālakṣaṇā. As is well known, the Tibetan translation divides this into two words: śūnyatā and alakṣaṇā. Indian and Tibetan commentators adopted this understanding. Thus, it follows that they read the Sanskrit as sarvadharmāḥ (1) śūnyatā (2) alakṣaṇā (3) anutpannā (4) aniruddhā (5) amalā (6) avimalā (7) anūnā (8) asaṃpūrṇāḥ |. These eight items are predicates of the subject “all phenomena (諸法)”, and since Vimalamitra calls them rnam pa brgyad [po], I too shall refer to them as the “eight aspects”.

The PHT, after providing a detailed interpretation of each aspect, says (Horiuchi 2021a, pp. 196–98; Horiuchi 2021b, pp. 75–76):

[Q] Why [were] only eight aspects [taught], and why [were they] taught in this order [in the Prajñāpāramitāhṛdaya]?

[A] Answer (’dir bshad pa, *atrocyate): This (’di, i.e., the HS) is the heart of the perfection of wisdom; “heart” further means something that is best and principal (pradhāna) over other branches.

The principal meaning of the perfection of wisdom is the three gates of liberation (*vimokṣamukha), such as emptiness. These (three gates of liberation) are included in those eight aspects. And precisely those [eight] aspects have the order here [in the three samādhis].

([Q] ci’i phyir rnam pa brgyad kho na dang go rims kyang ’di ltar gsungs she na/

[A] ’dir bshad pa</> ’di ni shes rab kyi pha rol tu phyin pa’i snying po yin la snying po yang gzhan (gzhan] DP; φ T) yan lag thams cad las dam pa gtso bo yin no// shes rab kyi pha rol tu phyin pa’i gtso bo’i don ni stong pa nyid la sogs pa rnam par ^[D275b] thar pa’i sgo gsum yin no// de ni rnam pa brgyad po de dag gis bsdus pa yin la/ rnam pa de dag nyid ’dir go rims yin te (yin te] DT; te P)/)

I will omit the explanation of the specific allocation. In summary, the relationship between “the three gates of liberation (*vimokṣamukha)” = three samādhis and the eight aspects is as follows:

(I) śūnyatā: (1; 2)

(II) ānimitta: (3–6)

(III) apraṇihita: (7; 8)

The above confirms that the PHT views the observation of the phenomena (dharmas) in the context of the meditation, i.e., three samādhis. In addition, the PHT’s interpretation of the clause of paragraph L of the HS: tasmāt tarhi (śāriputra) śūnyatāyāṃ na rūpam (lit. therefore, O Śāriputra, in emptiness, no rūpa …) is also noteworthy.

This is another enigmatic clause because it seems to be denying the teachings or elements of existence that have been preached since early Buddhism with the negative particle na (no, not). Leaving aside the variants in the Sanskrit text, let us draw on Conze’s translation (Conze 1958).

Therefore, O Śāriputra, in emptiness there is no form, nor feeling, nor perception, nor impulse, nor consciousness; No eye, ear, nose, tongue, body, mind; No forms, sounds, smells, tastes, touchables or objects of mind; No sight-organ element, and so forth, until we come to: No mind-consciousness element; There is no ignorance, no extinction of ignorance, and so forth, until we come to: there is no decay and death, no extinction of decay and death. There is no suffering, no origination, no stopping, no path. There is no cognition, no attainment and no nonattainment.

(p. 97)

Conze adds an explanation for the words “in” and “is” in this text, as follows:

The in can obviously have no spatial meaning. Nor can is―implied by the Sanskrit, and stated in the English―be an ordinary “is”, because it is equivalent to “is not”, the duality of “is” and “is not” having been abolished or transcended.

(p. 98)

This is a well-thought-out explanation, but it also seems to read too much into the text. On the other hand, Yonezawa (2009), in his detailed annotation of the HS, points out that this section seems to be a denial of ideas that have existed since early Buddhism, but Mahāyāna Buddhism recognizes these and inherits them as a method of practice. Furthermore, he draws attention to the earlier statement in the HS that Avalokiteśvara observed that all the five aggregates (skandhas) are empty while practicing the perfection of wisdom (paragraph E of the HS, see Section 2.3 [1]), presenting the following interpretation.

It is said na (no, not) here. But this is considered to be under the premise of the understanding that cognizes that everything is empty, after having perfected the practice of the perfection of wisdom and perfected wisdom.

(p. 188, translation is by the citator from Japanese)

This point is interesting in two ways: first, in that it focuses on the observation by Avalokiteśvara mentioned earlier in the HS, i.e., that it was a reading based on the structure of the HS, and second, it echoes the understanding of an epistemological reading of the HS, as asserted by Attwood (2017), mentioned earlier. Furthermore, according to the author’s reading of the PHT, much further back than this, it, was presenting an epistemological perspective on the relevant phrase in the HS.

PHT, D275b, P296a, T20-21 (Commentary on paragraph L: tasmāt tarhi (śāriputra) śūnyatāyāṃ na rūpa):

Now, in order to explain the results of “seeing” according to the [eight] aspects described [in the HS above], it is said, “Therefore, then (tasmāt tarhi)”, etc. “Therefore” means, because one has seen in accordance with the aspects explained above. “Then” means at that time.

The syntactical connection (sbyar) [of the HS] is “When emptiness (śūnyatāyāṃ)” is observed, “rūpa” is “not (na)” to be observed. Namely, precisely the nonseeing of all dharmas is the seeing of the emptiness.

(da ni bstan pa’i rnam pas rnam par bltas pa’i (bltas pa’i] DP; lta ba’i T) ’bras bu bstan pa’i phyir/ de lta bas na de’i tshe zhes bya ba la sogs pa gsungs te/ de lta bas na zhes bya ba ni gsungs pa’i rnam pas rnam par lta bas na’o// de’i tshe ni (ni] DT; φ P) de’i dus na’o//

stong pa nyid yang dag par rjes su mthong ba la gzugs yang dag par rjes su bltar med ces ’byung (’byung] DT; byung P) bar sbyar te/ chos ^[T21] thams cad mi mthong ba nyid ni stong pa nyid mthong ba yin no//)

The Tibetan in the first half is “stong pa nyid yang dag par rjes su mthong ba la gzugs yang dag par rjes su bltar med ces ’byung bar sbyar te”. While this may be challenging to understand, if we consider the correspondence with the HS, we can perceive that, unlike what is given in the preceding translation,35 the PHT is annotatively supplementing the “śūnyatāyāṃ na rūpaṃ (stong pa nyid la gzugs med)” in the HS, as underlined. Recall that “sbyar” means a syntactic connection of the sūtra passage in the commentarial texts (see [1] in Section 2.3). Thus, my assumption of the Sanskrit behind it is *śūnyatāyāṃ samanupaśyamānāyāṃ na rūpaṃ samanupaśyitavyam. Although my assumption of the Sanskrit is only one idea, Vimalamitra is making a supplementary commentary on the passage of the HS. According to this interpretation, the phrase “na rūpaṃ (no/not rūpa)” in the HS is understood by Vimalamitra not in the sense of “there is no rūpa”, but rather that “the rūpa cannot be observed [when the emptiness is observed]”.

3.3. Nonseeing

Incidentally, in the previous section, there was an important sentence “precisely the nonseeing of all dharmas is the seeing of the emptiness (chos thams cad mi mthong ba nyid ni stong pa nyid mthong ba yin no//)”. This too is an indication of Vimalamitra’s position as a Mādhyamika. Moreover, PHT, D274b says “nonseeing of what exists ultimately is the seeing of the reality (*tattva) (don dam par yod pa mi mthong ba nyid de kho na mthong ba yin la)”, and a little later the PHT quotes the following as scriptural source (āgama).

PHT, D274b, P294b, T18:

Likewise, it is taught such as

Precisely the nonseeing of rūpa is precisely the seeing of rūpa.36

In the Samādhirāja[sūtra] too, it is said:

Nonseeing of anything is the seeing of all dharmas.37

(de skad du

gzugs ma mthong ba nyid gzugs mthong ba nyid do

zhes bya ba la sogs pa gsungs pa dang/ ting nge ’dzin gyi rgyal po las kyang/

gang yang ma mthong ba ni chos thams cad mthong ba yin no

zhes gsungs so//)

The position that nonseeing of dharmas at all is the right view is found in much of the literature of the Madhyamaka school. For example, Śāntideva (ca. 690–750), Śāntarakṣita, and Kamalaśīla endorse this standpoint by citing or relying on the Dharmasaṃgītisūtra as Ichigō (2015, pp. 68–70) illustrates:

Ārya-dharmasaṃgīti-nāma-mahāyāna-sūtra (P no. 904, 74b): bcom ldan ’das chos thams cad ma mthong ba ni yang dag par mthong ba’o//

Śāntideva’s Śikṣāsamuccaya (Bendall ed., 264): adarśanaṃ bhagavan sarvadharmāṇāṃ {darśanaṃ} samyagdarśanam.38 (D no. 3940, 146b: bcom ldan ’das chos thams cad ma mthong ba ni yang dag pa (read par?) mthong ba’o//)

Śāntarakṣita’s Madhyamakālaṃkāravṛtti (Ichigō 1985, p. 286): bcom ldan ’das chos thams cad mi mthong ba ni yang dag par mthong ba’o//

O Bhagavat, nonseeing of all dharmas is the right seeing.

Kamalaśīla’s BhK I, 212: katamaṃ paramārthadarśanam? sarvadharmāṇām adarśanam.

What is the seeing of the ultimate? It is the nonseeing of all dharmas.

This Mādhyamika position also opposes the Yogācāra who advocates that nondual gnosis (gnyis su med pa’i ye shes, *advayaṃ jñānam or advayajñānam) ultimately (don dam par, *paramārthatas) exists. The PHT has a detailed discussion of this in D274a3ff. and there is a similar discussion in SPT, D60bff. too. Furthermore, the PHT can certainly be positioned as a work of the Madhyamaka school.

4. Conclusions

This paper elucidates Vimalamitra’s interpretation of the HS in his PHT as a Mādhyamika master, by clarifying its criticism on other schools.

Considering the structure of the HS, Vimalamitra understands the phrase “rūpa (form or form-and-color) is empty” to mean that rūpa is empty of intrinsic/its own nature (svabhāva). This is true of rūpa and of all the five aggregates (skandhas). In short, the five skandhas lack their own nature. This is the basic premise of the thought of emptiness, and we can estimate that Vimalamitra interprets the HS in accordance with the basics of the tenet of the thought of emptiness in Mahāyāna.

Vimalamitra’s ideological position, as developed in the PHT, is solely that of the Madhyamaka school; he did not endorse the Yogācāra. Vimalamitra understands the “four formulas of emptiness” such as “rūpa is empty” as the HS’s criticism of other schools inside and outside of Buddhism. For him, the HS is the book of the Madhyamaka.

Based on the unique similarity of the discussion in Vimalamitra’s PHT and SPT and his older contemporary Kamalaśīla’s BhK, Vimalamitra can be evaluated as a Mādhyamika in the vein of Kamalaśīla.

The arguments developed there are highly philosophical, but they are grounded in the meditative context of Mahāyāna practice and the observation of emptiness. More specifically, PHT understands the observation of phenomena (dharmas) in the context of meditation.

In this connection, in recent years, scholars have focused on an epistemological reading of the HS. A new reading of the phrase of the PHT in this paper suggests that Vimalamitra interpreted the phrase in the HS: “in emptiness, no/not rūpa” epistemologically. Namely, he understood it as follows: “the rūpa cannot [be observed] when the emptiness [is observed]”.

The PHT’s understanding of the HS as literature to be read along the basic lines of Mahāyāna Buddhism and the basic Mahāyāna theory of practice could be the distinctive feature and significance of the PHT’s interpretation of the HS for the contemporary readers of the HS.

However, other sections of the PHT require further detailed re-examination. In this case, the framework of “Commentaries on the long HS preserved in the Tibetan canon”, presented for the first time in this paper, will be useful for further comprehensive and precise elucidation of this field in the future. We hope that this framework will help to further clarify the field in a more comprehensive and precise manner.