Notes on T255

Taishō text No. 255 is a version of the extended Heart Sutra. It is not found in the ancient catalogues. The Taishō editors noted: Dùnhuáng shí shì běn 燉煌石室本 i.e. "based on a manuscript from Dunhuang" meaning it was unknown before the 20th century. T 255 is attributed by Taishō to Fǎchéng 法成, in 856 CE. In turn Fǎchéng has been identified with the Tibetan translator Chos grub (pronounced like Chodrup; fl. early 9th century).

In perusing the literature of the Heart Sutra, one often encounters the assertion that T255 was translated from Tibetan. This claim is usually unattributed, however I recently noticed that Channa Li (2021: 13, n.19) attributes the claim to the Japanese scholar Ueyama Daishun 上山大峻 (1934 - 2022). Li references Ueyama (1968, 1990) and notes that two texts, including T255, "were presumed by Ueyama to be possibly translated from Tibetan, which should be verified by more concrete studies." Li (2024: 24) adds:

Ueyama claims that this version may have been translated from the Tibetan version of the longer version of the Prajñāpāramitāhṛdaya [Heart Sūtra]. However, he does not elaborate on the concrete evidence.

Li ultimately leaves the question open, but notes (2024: 25):

One piece of supporting evidence lies in Chödrup’s translation of the term mingliao (明了, literally meaning ‘illumination’), which was more likely translated directly from the Tibetan term snang ba (‘illumination,’appearance’) than from Sanskrit avabhāsa (‘appearance’).

What follows are my edited notes of my attempt at a more concrete study. Since this post is long, I should warn readers up front that, in the end, there is only one example of such a difference, and it is countered by one which seems to point to a Sanskrit original. But neither is entirely unequivocal. So the end result of this study is no result. I cannot see anything in T255 that forces me to conclude it was translated from Tibetan. It's entirely possible that someone with a better eye and more language skills will find an example, however (a) no one else seems interested in carrying out these kinds of detailed studies, so don't hold your breath, and (b) if there is an example, it's likely to be a very subtle thing and, thus, ambiguous. If you are only interested in the conclusion, you can stop now. If you are interested in the process, please read on.

At the outset we need some idea of how we could tell if a Chinese text was translated from Tibetan or Sanskrit. How could tell, for example, if an expression such as 明了 is better attributed to a Tibetan or Sanskrit source. My basic method is to do a close reading of the four texts outlined below, looking for two things:

1. An expression found in T255, but not found in T253.
2. A lexical or grammatical mismatch between the Sanskrit and Tibetan texts that explains the difference between T255 and T253.

SOURCES

I will mainly compare four texts:

• T 253, as representative of the Chinese extended text.*
• T 255.
• The canonical Tibetan extended Heart Sutra, especially as reflected in Silk's (1994) Recension A (hereafter TibA).
• The extended Prajñāpāramitāhṛdaya in Conze's edition (taking into account my published corrections: Attwood 2024).

* T252 is also extended, but has added a very different introduction and conclusion. T254 is merely a lightly edited version of T253. T257 was only translated ca 1000 CE.

For convenience, I will use Silk's divisions of the text into paragraphs, focussing on paragraphs D-W, i.e. ignoring the preliminaries, the title, and any colophons, and focusing only on the text of the Heart Sutra per se. I parsed the Tibetan with help from the Tibetan and Himalayan Library translation tool in conjunction with Silk's (1994) texts and translations.

Li notes that T255 is more like Silk's TibA, however we cannot assume that Chos grub had the canonical Tibetan text to hand, if indeed he did translate from Tibetan.

Something to keep in mind is that all the extended Heart Sutra texts, including the odd ones like T252, largely reproduce the standard text where possible (with only minor variations). In other words, the middle part of the extended text follows the standard text in Chinese, Sanskrit, and Tibetan. I'm not sure that this amounts to anything more than being a curious fact.

So now let's dive into a close reading, noting variants. I use colour coding for comparative purposes. I haven't been very consistent in doing this.

COMPARISON

Paragraph D

• 253: 如是我聞：一時佛 在王舍城 耆闍崛山中，與大比丘眾及菩薩眾俱。時佛世尊即入三昧，名廣大甚深。
• 255: 如是我聞：一時薄伽梵住王舍城鷲峯山中，與大苾蒭眾及諸菩薩摩訶薩俱。爾時，世尊等入甚深明了三摩地法之異門。
• TibA. 'di skad bdag gis thos pa dus gcig na / bcom ldan 'das rgyal po'i khab bya rgod phung po'i ri la dge slong gi dge 'dun chen po dang / byang chub sems dpa'i dge 'dun chen po dang thabs gcig tu bzhugs te / de'i tshe bcom ldan 'das zab mo snang ba zhes bya ba chos kyi rnam grangs kyi ting nge 'dzin la snyoms par zhugs so //
• Skt. evaṃ mayā śrutam | ekasmin samaye bhagavān rājagṛhe viharati sma gṛdhrakūṭe parvate mahatā bhikṣusaṃghena sārdhaṃ mahatā ca bodhisattvasaṃghena | tena khalu samayena bhagavān gambhīrāvabhāsaṃ nāma dharmaparyāyaṃ bhāṣitvā samādhiṃ samāpannaḥ.
• Thus have I heard. One time the Bhagavan was dwelling on Vulture's Peak in Rājagṛha, together with a large assembly of bhikṣu-s and a large assembly of bodhisatvas. At that time, the Bhagavan entered a samādhi named "appearance of the profound".

Notes:

• T253 fó 佛 "Buddha" versus T255 bógāfàn 薄伽梵 “Bhagavān”
• zài 在 "at" vs zhù 住 "residing" (= viharati sma). Very similar characters, easily mistaken.
• Different spellings of Gṛdhakūṭa 耆闍崛山 vs 鷲峯山
• T253 sānmèi 三昧; T255 sānmódì 三摩地 = Skt samādhi. Tib ting nge 'dzin.
• T255 adds 明了 “clear” = snang ba and 法之異門 “distinct mode”

Both T253 and T255 refer to a great assembly 大 ... 眾 (mahatā ... -saṇghena) of bhikṣu-s 比丘, but drop "great" for the assembly of bodhisatvas. T253 has plain "bodhisatva-assembly" 菩薩眾, but T255 has bodhisatva-mahāsatva. Neither Tibetan nor Sanskrit have mahāsatva here though both have it elsewhere in the text. Conze notes no variants with mahāsatva here and I'm not aware of any.

Silk (1994: 172) "...the Blessed One was entered into the concentration of the Preaching of the Dharma called "profound illumination". However, on gambhīra-avabhāsa, compare Han (2020: 398):

In the same way, as many as there are the appearance of thoughts, the appearance of forms (rūpāvabhāsa), or the appearance of sounds (śabdāvabhāsa), all those reflect in a single appearance of the bodhisatva who maintains the ocean-seal samādhi, thus it is called the ocean-seal samādhi.

Also note the following Pāli passage from the Mahānidāna Sutta (DN 15)

Gambhīro cāyaṃ, ānanda, paṭiccasamuppādo gambhīrāvabhāso ca. (DN II 55)

Ānanda, this dependent arising is profound and appears profound.

Also:

Siyā nu kho, bhante, esevattho vitthārena vuccamāno gambhīro ceva assa gambhīrāvabhāso cā’’ti? (SN 12.24)

Gambhīro cāyaṁ, ānanda, paṭiccasamuppādo gambhīrāvabhāso ca. (SN 12.60)

Note that in Aṣṭa no bodhisatvas are identified as being present. Also bodhisatvas are discussed in the abstract, no one is ever directly addressed as, or referred to as being, "a bodhisatva".

明了

Since Li (2024) identified 明了 as a possible indicator of Chodrup having a Tibetan source text lets examine this closely, beginning by parsing each version of the passage separately so we can compare them.

253: 時 / 佛世尊 / 即 / 入 / 三昧 / 名/ 廣大甚深。
At that time / tathāgata / then / entered / a samādhi / named / vast and very profound

Where 廣大甚深 is 廣大 "broad vast" and 甚 "very" 深 "deep"

255: 爾時 / 世尊 / 等入 / 甚深 / 明了 / 三摩地 / 法之異門。
At that time / bhagavan / fully entered / a very deep / lucid / samādhi / dharma teaching.

TibA: de'i tshe / bcom ldan 'das / zab mo snang ba / zhes bya ba / chos kyi rnam grangs / kyi / ting nge 'dzin / la / snyoms par zhugs so.
At that time / Bhagavan / Profound Illumination / named / dharmaparyāya / (of) / samādhi / (into) / entered

Note that in TibB, the phrase chos kyi rnam grangs is transposed into the phrase zab mo snang ba giving zab mo chos kyi rnam grangs snang ba, substantially changing the meaning of this passage. Although Silk does not discuss this problem, see the different translations (Silk 1994: 172-173)

Skt: tena khalu samayena bhagavān gambhīrāvabhāsaṃ nāma dharmaparyāyaṃ bhāṣitvā samādhiṃ samāpannaḥ.
At that time / the Bhagavan / profound appearance / named / dharma teaching / having spoken / samādhi / entered.

So the terms we are discussing are

• Chinese: shèn shēn míng liǎo 甚深明了 "very deep and lucid"
• Sanskrit: gambhīrāvabhāsa "profound appearance"
• Tibetan: zab mo snang ba = gambhīra-avabhāsa

(Note: Hopkins Dictionary s.v. snang ba "appear; perceive; light; illuminate; appearance; illumination")

As far as I can see 明了 in T255 suggests someone has misread avabhāsa due to the etymological fallacy. Semantically, the root is √bhā "shine" but pragmatically, the prefix ava- changes the sense to "appear". Compare this with lokayati "look" and avalokayati "examine" (not, as Conze mistakenly asserts, "looks down"). The standardised Tibetan translation of avabhāsa is snang ba. The ambiguity between the semantic "shine" and pragmatic "appear" senses occurs in both Sanskrit and Tibetan. So this does not seem to be diagnostic as Li (2024) suggested. As far as I can see.

Paragraph E

This para mirrors the standard text, in the sense that it shows Guanyin in his characteristic role (in this context), i.e. observing the skandhas.

• T253: 爾時眾中有菩薩摩訶薩，名觀自在，行深般若波羅蜜多時，照見五蘊皆空，離諸苦厄。
• T255: 復於爾時，觀自在菩薩摩訶薩行深般若波羅蜜多時，觀察照見五蘊體性悉皆是空。
• Skt: tena ca samayena āryāvalokiteśvaro bodhisattvo mahāsattvo gambhīrāyāṃ prajñāpāramitāyāṃ caryāṃ caramāṇaḥ evaṃ vyavalokayati sma pañca skandhāṃs tāṃś ca svabhāvaśūnyaṃ vyavalokayati.
• TibA: yang de'i tshe byang chub sems dpa' sems dpa' chen po 'phags pa spyan ras gzigs dbang phyug shes rab kyi pha rol tu phyin pa zab mo spyod pa nyid la rnam par blta zhing / phung po lnga po de dag la yang rang bzhin gyis stong par rnam par blta'o //

Here Guanyin is practising the practice of profound perfection of prajñā. He observes (vyavalokayati sma) the five skandhas. However, the text goes awry after this, at least in Sanskrit and Tibetan.

In English, Sanskrit, and Tibetan we divide visual actions into looking and seeing, which are analogous to seeking and finding. In the standard Sanskrit text, the two verbs are both in the periphrastic past (a third person singular present indicative verb with the periphrastic particle sma): vayavalokayati sma "he observed" and paśyati sma "he saw". These are derived from the roots vyava√lok and √paś respectively. In Tibetan, we might expect to contrast blta "looking" with mthong "seeing". The latter occurs for example, in TibB, para I.

In the Sanskrit and Tibetan texts, where we expect the second verb to mean "seeing", we find a repetition of vyavalokayati "observing" but without sma and thus in the present tense. So instead of "he observed" and "he saw", we have "he observed" (past) and "he observes, he is observing" (present tense).

This is a previously unnoticed grammatical error called a constructional mismatch. What we expect, per the standard text is a combination of vyava√lok and √paś in the same person, number, tense, mood, etc.

However, since the error occurs in both Sanskrit and Tibetan, this uncorrected error is not diagnostic for our purposes. Moreover, we cannot tell whether the mistake occurred in composition or copying. However since it occurs in both Sanskrit and Tibetan, we can infer that the canonical Tibetan text was translated from an already defective Sanskrit text.

Paragraphs F and H

In Para F, TibB spells Śāriputra as shā ra dwā ti'i bu, i.e. Śāradvātiputra. This spelling is also found in TibA in para H.

This spelling is not reflected anywhere in T255, or any other Chinese Heart Sutra text.

Paragraph I

This paragraph is the end of the extended introduction. As with Para E, something goes wrong here.

In T251 the sentence has four clauses: (1) Guanyin practiced (行) Prajñāpāramitā; (2) he observed (照見) the five skandhas, and (3) [saw] they are absent; (4) and [as a result] he transcended (度) suffering. The verb for "see" is omitted (and generally no one notices this). The Sanskrit standard text only has three clauses, with (4) being omitted in all known witnesses; with the three verbs being (1) caramāṇo*, (2) vyavalokayati sma, and (3) paśyati sma.

* Actually a present participle, used to indicate an action simultaneous with the main verb.

In T251, Guanyin is the agent of all four verbs. In the extended texts, the agent has become the kulaputra/kuladuhitṛ (hereafter kula°), but there is another constructional mismatch.

Compare:

• T253: 「舍利子！若善男子、善女人行甚深般若波羅蜜多行時，應觀五蘊性空。
• T254: 「若善男子及善女人，欲修行甚深般若波羅蜜多者，彼應如是觀察，五蘊體性皆空。
• Skt: yaḥ kaścic chāriputra kulaputro va kuladuhitā vā asyāṃ gambhīrāyāṃ prajñāpāramitāyāṃ caryāṃ cartukāmas tenaivaṃ vyavalokitavyam pañca skandhāṃs tāṃś ca svabhāvaśūnyān paśyati sma.
• TibA: shā ri'i bu rigs kyi bu 'am rigs kyi bu mo gang la la shes rab kyi pha rol tu phyin pa zab mo spyod pa spyad par 'dod pa des 'di ltar rnam par blta bar bya ste / phung po lnga po de dag kyang rang bzhin gyis stong par rnam par yang dag par rjes su blta'o //
• TibB (2): ... / phung po lnga po de dag ngo bo nyid kyis stong par yang dag par rjes su mthong ba de ltar blta bar bya ste//

In Paragraph G, Śāriputra asks the question, How should a kula° practice (kathaṃ śikṣitavyaṃ)? Since the verbal form here is a future passive participle (or gerundive), we expect the answer to be in the same mode, i.e. "The kulaputra should [do something]."

Para I begins the reply as expected using the same verbal mode, i.e. vyavalokitavya "[the kula°] should observe" , in the same future passive mode. And what they should observe is the five skandhas (a classic Buddhist meditation practice). Having stated this, we have a complete and comprehensible sentence.

However, in Para I, all the extended texts now add some variation on "and he sees them [i.e. the skandhas] as lacking svabhāva".

However, the counterpart "see", uses the periphrastic past paśyati sma "[the kulaputra] saw", when we expect a future tense verb here: i.e. "The kula° should observe the skandhas and they [will see something]." The shift in tense is a constructional mismatch.

While the look/see structure must be explicit in Sanskrit and English (and Tibetan?), it may be implied in Chinese. Hence, in the Chinese versions, we see verbs meaning "look, observe, etc" (i.e. yīngguān* 應觀, guān 觀) but no verb meaning "see". For comparison, T251 has zhàojiàn 照見 "observe, inspect".

* note that yīng 應 is also transcribed yìng in Buddhist contexts, cf. DDB s.v. 應

What Guanyin sees when he looks at the five skandhas (zhàojiàn wǔ yùn 照見五蘊) in T251 is jiē kōng 皆空, literally "all empty" or "everything is absent" (Note: T250 omits jiē 皆). We might not have noticed this without the Sanskrit translations, which have to specify both look and see.

My interpretation of this is that we are describing a samādhi in which dharmas have stopped arising and thus the branches of experience (skandhāḥ) are absent (śūnya). Where absent is an epistemic term meaning "cannot be perceived" or "is not found" and not, as usually assumed, a metaphysical term meaning "does not exist". Treating this an epistemic denial is far less paradoxical and far more interesting.

In the standard text (T251, 250) Guanyin is the agent of both looking and seeing.

Note that where T251 asserts that the skandhas are 皆空, the Sanskrit standard text and all the extended texts append some reference to svabhāva, i.e. T253 性空 "absent-natured" (omitting 皆) and T255 體性皆空. "intrinsic natures are all absent".

Tibetan: rang bzhin gyis "intrinsically" stong par "empty" = Skt svabhāvaśūnyān

I can find no diagnostic differences between Sanskrit and Tibetan in this paragraph.

Paragraph J

This is the famous passage usually translated "form is emptiness..." etc. Silk translates the Tibetan as "Matter is empty", which seems to be an uncharacteristic blunder.

My approach to Prajñāpāramitā sets aside the usual metaphysical overlay. Following the way that Sue Hamilton approached the Pāli suttas, I read the Heart Sutra as concerned with experience rather than reality. That is to say, without any implied reference to Husserl, that the Heart Sutra is making phenomenological points rather than metaphysical points. We can call this the Hamiltonian hermeneutic.

Similarly, the negations that follow (na rūpa etc) reflect the absence of sensory experience following the "cessation of recognition and experience" (saṃjñāvedayitanirodha), which is brought about by yǐwúsuǒdégù 以無所得故. Huifeng (2014) showed that Kumārajīva coined this term to translate anupalambhayogena "by means of the practice of nonapprehension", which refers to withdrawing attention from sense experience using meditative (or, better, self-hypnotic) techniques.

When no sense experience arises, due to nonapprehension one cannot apply ontologies of sense experience such as skandha, dhātu, or āyatana to the resulting absence. An ontology of absence would be an oxymoron.

As I have repeatedly said in my writing: rūpa is to the eye as sound is to the ear. This fundamental observation has to be kept in mind because it tells us that the one thing that rūpa-skandha absolutely cannot be is "matter".

Here, we encounter for the first time a difference that might be diagnostic. The highlighted term occurs in Tibetan but not Sanskrit, and in T255 but not T253.

• T253: 舍利子！色不異空，空不異色。色即是空，空即是色。受、想、行、識亦復如是。
• T255: 色即是空，空即是色。色不異空，空不異色。如是受、想、行、識亦復皆空。
• Skt: rūpaṃ śūnyatā, śūnyataiva rūpam; rūpān na pṛthak śūnyatā, śūnyatāyā na pṛthag rūpam. Evaṃ vedanā-saṃjñā-saṃskāra-vijñānaṃ.
• TibA: gzugs stong pa’o // stong pa nyid kyang gzugs so // gzugs las stong pa nyid gzhan ma yin no // stong pa nyid las kyang gzugs gzhan ma yin no // de bzhin du tshor ba dang / ‘du shes dang / ‘du byed dang / rnam par shes pa rnams stong pa’o //

Having compared rūpa and śūnyatā, T251 summarises for the other skandhas: 受想行識亦復如是, i.e. "Vedanā, saṃjñā, saṃskāra, and vijñāna are the same as this". And T253 follows T251. All of these Chinese texts are consistent with the Sanskrit: evaṃ vedanā-saṃjñā-saṃskāra-vijñānaṃ.

However, T255 phrases this differently:

如是受、想、行、識亦復皆空。i.e. "Similarly, vedanā, saṃjñā, saṃskāra, [and] vijñāna [are] likewise all insubstantial".

The redundant "likewise" is for emphasis. The Tibetan reads:

de bzhin du tshor ba dang / ‘du shes dang / ‘du byed dang / rnam par shes pa rnams stong pa’o //
Likewise, vedanā and saṃjñā and samskāra and vijñāna [are] empty.

The phrase rnams stong pa corresponds to jiē kōng 皆空. There is no expression in the Sanskrit text that corresponds to this. This suggests that the translator may have been working from a Tibetan source similar to TibA (and not similar to TibB which has a major transposition error in this passage).

Paragraph K

There are no significant differences between Sanskrit and Tibetan in this paragraph.

In K, where T253 has zhū fǎ kōng xiàng 諸法空相 "all phenomena are marked with absence". T255 has the synonymous expression yī qiè fǎ kōng xìng 一切法空性. 諸 and 一切 both effectively mean "all". 相 is perhaps the most common translation of lakṣana "characteristic"; while 性 means "nature" (it's used about to translate svabhāva in Para I).

Paragraph L.

Here there is a different in the opening clause:

• 253: 是故空中無色...
• 255: 舍利子！是故爾時空性之中，無色...
• Skt: Tasmāt tarhi śāriputra śūnyatāyāṃ na rūpam...
• TibA: shā ri'i bu de lta bas na stong pa nyid la gzugs med /

T253 does not include the name Śāriputra, T255 does. However, both Skt and TibA include the name.

T253 has 是故 "therefore". T255 has 是故 爾時 "therefore, at that time". However, in this case, T255 follows the Sanskrit and not the Tibetan.

• Skt. Therefore (tasmāt) at that that time (tarhi), in absence (śūnyatāyām) no appearance (na rūpaṃ)
• TibA: de lta bas na (therefore) stong pa nyid la (in absence) gzugs med (no form).

Paragraphs M, N, O

No significant differences.

Paragraph P

• T253: 以無所得故，菩提薩埵依般若波羅蜜多故心無罣礙。無罣礙故，無有恐怖，遠離顛倒夢想，究竟涅槃。
• T255: 是故舍利子！以無所得故，諸菩薩眾依止般若波羅蜜多，心無障礙，無有恐怖，超過顛倒，究竟涅槃
• Tasmāc Chāriputra aprāptitvena bodhisattvānāṃ prajñāpāramitāmāśritya viharati cittāvaraṇaḥ| cittāvaraṇanāstitvādatrasto viparyāsātikrānto niṣṭhanirvāṇaḥ|
• shā ri'i bu de lta bas na byang chub sems dpa' rnams thob pa med pa'i phyir / shes rab kyi pha rol tu phyin pa la brten cing gnas te / sems la sgrib pa med pas skrag pa med de / phyin ci log las shin tu 'das nas mya ngan las 'das pa'i mthar phyin to //

T253 opens with 以無所得故 which Huifeng (2014) identified as Kumārajīva's unique translation of (tacca) anupalambhayogena "(and that) by means of practising non-apprehension", coined for the purpose of translating the Pañcaviṃśatisāhasrikā Prajñāpāramitā (Pañc). The term recurs throughout Pañc and T223. As such, we have to see aprāptitvāt as a mistranslation of 以無所得故, based on confusion between 得 (pra√āp) and 所得 (upa√labh), fostered by the previous phrase 無得 "no attainment" (Skt na prāptiḥ). Huifeng also showed that, in Pañc, the final negation is given as "no attainment and no realisation" na prāpti nābhisamayaṃ (followed by examples of each that are omitted from the Heart Sutra). And as such Kumārajīva either mistranslated or had a defective manuscript.

Huifeng noted this term ought to go at the end of Para O, since it qualifies all the negations from L-O. And I confirmed that the term almost always occurs in the sentence or paragraph final position, where it qualified what comes before.

And keep in mind that for this para there is a serious mismatch between Chinese and Sanskrit, with the Sanskrit translation being particularly garbled at this point. This makes comparison difficult if not impossible.

The main difference occurs here:

• T253 心無罣礙, where 罣礙 means “being caught, entangled, impeded from within.”
• T255 心無障礙, where 障礙 means “being blocked, obstructed more generally.” This phrase may be influenced by the Sanskrit term citta-āvaraṇa, which Huifeng (2014) noted is a mistranslation of 心無罣礙. While no exact Sanskrit counter part exists, we can show indirectly that it likely corresponds to na kvacit sajjati "not stuck on anything".

However, there is no corresponding difference in Skt or TibA, since cittāvaraṇaḥ = sems la sgrib pa

Paragraphs R, S, T

No significant differences.

Paragraph U

• T253: 如是說已。即時，世尊從廣大甚深三摩地起，讚觀自在菩薩摩訶薩言：「善哉，善哉！善男子！如是，如是！如汝所說。
• T255: 爾時，世尊從彼定起，告聖者觀自在菩薩摩訶薩曰：「善哉，善哉！善男子！如是，如是！如汝所說。
• Skt: atha khalu bhagavān tasmāt samādher vyutthāya āryāvalokiteśvarasya bodhisattvasya mahāsattvasya sādhukāram adāt. sādhu sādhu kulaputra | evam etat kulaputra, evam etad.
• TibA: de nas bcom ldan 'das ting nge 'dzin de las bzhengs te / byang chub sems dpa' sems dpa' chen po 'phags pa spyan ras gzigs dbang phyug la legs so zhes bya ba byin nas / legs so legs so // rigs kyi bu de de bzhin no //

There is a minor difference between T253 and T255 in the first clause. Both use the construction: 從 "from" …. 起 "arose". T253 repeats the name of the samādhi and uses the same transcription, i.e. 三摩地. However, T255 abbreviates this to 從彼定起 "... arose from that samādhi, this time translating samādhi 定 rather than using the transcription in Para D above.

However, there is no corresponding difference between Skt samādher vyutthāya and Tibetan ting nge 'dzin de las bzhengs te, which both mean "arose from samādhi"

Paragraph V

• T 253: 甚深般若波羅蜜多行，應如是行。如是行時，一切如來皆悉隨喜。
• T 255: 彼當如是修學般若波羅蜜多。一切如來亦當隨喜。
• Skt: gambhīrāyāṃ prajñāpāramitāyāṃ caryāṃ cartavyaṃ yathā tvayā nirdiṣṭam anumodyate sarva-tathāgatair arhadbhiḥ.
• TibA: rigs kyi bu de de bzhin te / ji ltar khyod kyis bstan pa de bzhin du shes rab kyi pha rol tu phyin pa zab mo la spyad par bya ste / de bzhin gshegs pa rnams kyang rjes su yi rang ngo //

rigs kyi bu / de de bzhin te /
kulaputra / in the same way

ji ltar / khyod kyis / bstan pa / de bzhin du / shes rab kyi pha rol tu phyin pa / zab mo la / spyad par bya ste /
just as / by you / was taught / likewise / Prajñāpāramitā / profound / should engage in

de bzhin gshegs pa rnams / kyang rjes su yi rang ngo //
Tathagatas (all) / rejoice //

For reasons that are not clear to me, Silk (1994: 184-185) translates de bzhin gshegs pa as "the Sugatas", rather than the more conventional tathāgata (per the THL translation tool).

The two Chinese texts are obviously quite different here. Apart from the fact that T255 omits the middle two clauses, present in both Skt and Tibetan, the sentence construction is different.

T253: "The practice of the profound Prajñāpāramitā should be practiced this way. When practising this way, all (一切) the tathāgatas entirely (皆悉) rejoice.

T225: Prajñāpāramitā should be (當) cultivated and practiced (修學) in this way (如是) by him (彼).

Neither Chinese text has a parallel to "just as you have taught it" (yathā tvayā nirdiṣṭam) or "worthy" (arhadbhiḥ).

However, the differences here are not diagnostic, since Skt and TibA are more or less same, except that TibA has no parallel arhadbhiḥ either. Arguments from absence are weak. In this case, an omission might occur for any number of reasons, including scribal error.

Paragraph W.

• T253: 爾時世尊說是語已，具壽舍利弗大喜充遍，觀自在菩薩摩訶薩亦大歡喜。時彼眾會天、人、阿修羅、乾闥婆等，聞佛所說，皆大歡喜，信受奉行。
• T255: 時薄伽梵說是語已。具壽舍利子， 聖者觀自在菩薩摩訶薩，一切世間天、人、阿蘇羅、乾闥婆等，聞佛所說，皆大歡喜，信受奉行。
• Skt: Idam avocad bhagavān. āttamanā āyuṣmānc chāriputraḥ āryāvalokiteśvaro bodhisattvo mahāsatvo sā ca sarvāvatī pariṣat sadeva-mānuṣāsura-gandharvaś ca loko bhagavato bhāṣitam abhyanandann iti.
• TibA: bcom ldan 'das kyis de skad ces bka' stsal nas / tshe dang ldan pa shā ri'i bu dang / byang chub sems dpa' sems dpa' chen po 'phags pa spyan ras gzigs dbang phyug dang / thams cad dang ldan pa'i 'khor de dag dang / lha dang / mi dang / lha ma yin dang / dri zar bcas pa'i 'jig rten yi rangs te / bcom ldan 'das kyis gsungs pa la mngon par bstod do //

Note that here the punctuation added to the Chinese by CBETA is inconsistent. And there is a slight difference in the construction, with T253 has Guanyin and Śāriputra rejoicing individually, before everyone else, while T255 has everyone rejoicing at the same time.

Otherwise, there are no significant differences.

Conclusion

Given that every manuscript copy of the Heart Sutra is different, when we compare any two versions of the text, we expect to find differences. Moreover, we also expect different translators to use different expressions for the same passage. And indeed, there are many such differences between T253 and T255.

The goal was to identify distinctive features of T255 that could be explained by a similar distinction in the Tibetan texts, and absent from Sanskrit. To achieve this goal, I carefully parsed each of the four texts, one sentence at a time, looking at lexicon, syntax, and grammar.

I did note two previously unreported instances of construction mismatch, i.e. cases of sentences where the verb unexpected changes tense or mood. These occur in Para's E and I. This is further evidence that the extended text was also a Chinese production. I've already noted that the extended text exists in two distinct recensions: T252 and the rest. Ben Nourse's unpublished conference presentation makes it seem likely to me that the extended texts were composed in or around Dunhuang.

I found only one example of a significant difference in which T255 followed Tibetan rather than Sanskrit. In Paragraph J, T253 and the Sanskrit text both follow T251 in concluding the discussion of appearance and absence by noting that the other skandhas are the same.

• T251: 受想行識亦復如是
• Sanskrit: evaṃ vedanā-saṃjñā-saṃskāra-vijñānaṃ.

T255 alone adds "and likewise all are absent" 亦復皆空

The Tibetan, in both TibA and TibB (Silk 1994: 120-121) also adds a similar qualification:

de bzhin du tshor ba dang / ‘du shes dang / ‘du byed dang / rnam par shes pa rnams stong pa’o //< br/> Likewise, vedanā and saṃjñā and saṃskāra and vijñāna [are] empty.

And contrarily, we find one example that appears to point the other way in Paragraph L:

• T253: 是故... "Therefore".
• T255 是故 爾時... "Therefore, at that time".
• Skt. tasmāt tarhi... "Therefore at that that time."
• TibA: de lta bas na... "Therefore..."

Unfortunately, while this exercise took many hours, there are no unequivocal examples that force us to adopt Ueyama's conclusion that Chos grub translated T255 from Tibetan. While I don't rule out someone else finding such evidence, I cannot find any. This result doesn't disprove the assertion by Ueyama. What it does is call into doubt the rationale for making the assertion in the first place. There seems to be no reason to believe that Chos grub was working from Tibetan, and thus no reason to assert this as a possibility.

~~Φ~~

Bibliography

Attwood Jayarava (2024). "Revised Editions of the Prajñāpāramitāhṛdaya and Bānrěbōluómìduō xīn jīng «般若波羅蜜多心經»." Asian Literature and Translation. 11(1). 52-92.

Han, Jaehee. 2020. The Sky as a Mahāyāna Symbol of Emptiness and Generous Fullness: A Study and Translation of the Gaganagañjaparipṛcchā. Volume 2: Edition and Translation. Dissertation for Dr Philos, University of Oslo.

Li, Channa. 2022. "Toward A Typology of Chödrup’s (Tib. Chos Grub, Chin. Facheng 法成) Cursive Handwriting: A Palaeographical Perspective." BuddhistRoad Paper 1.2. Ruhr Universität Bochum.

Li, Channa. 2024. “Toward a History of Chödrup’s (fl. First Half of 9th C., Tib. Chos grub, Chin, Facheng 法成)Monastic Activities: An Introduction and a Working Chronology.” BuddhistRoad Paper 1.3. Ruhr Universität Bochum.

Ueyama Daishūn上山大峻. 1968. “大蕃國大徳三藏法師沙門法成の研究(下). 東方學報 39: 119–222. [Daibankoku daitoku sanzōhōshi shamonhōjō no kenkyū (ge) .” [Tōhō gakuhō Studies on the Great Monk of Tufan Empire, Tripiṭakācārya, Śramaṇa ChödrupJournal of Oriental Studies, Kyoto]

Ueyama Daishūn 上山大峻. 1990. 敦煌仏教の研究 [Tonkō bukkyō no kenkyū Studies on Dunhuang Buddhism]. Kyoto: Hōzōkan, .