March 29, 2020

On the Tibetan Rendering of the Name Patañjali/Pātañjali

The name “Patañjali” (v. l. “Pātañjali”) has been translated into Tibetan as “Chur lhung,” and it has been recorded in the Mahāvyutpatti (Sakaki, no. 3498; Fukuda & Ishihama, no. 3496). To be sure, the name has also been rendered as “Thal mo lhung” (’Jam dbyangs mkhyen brtse’i dbang po, Legs par sbyar ba’i bstan bcos kyi byung tshul cung zad bshad pa ngo mtshar zla zhun gsar pa’i ’dzum phreng. In Sa skya’i chos ’byung gces bsdus. 6 vols. Beijing: Krung go’i bod rig pa dpe skrun khang, 2009, vol. 6, p. 184.7). The rendering “Thal mo lhung” is acceptable and it does not have to be “Thal mor lhung.” It is comparable to the rendering of “Devadatta.” Both “lHa s/byin” and “lHas s/byin” are acceptable. Ācārya Sems dpa’ rdo rje points out that the Tibetan translation “Chur lhung,” or rather “Chu lhung” (as he has it), is actually a “misunderstanding” (go nor), and he translates the name as “Thal mo sbyar ba can.” See Sems dpa’ rdo rje, mKhan chen zhi ba ’tshos mdzad pa de nyid bsdu ba’i ’grel ba (sic) dpal de kho na nyid gsal bar byed pa’i sgron ma. The Red and Black Crown Karmapa Series 41. [Kalimpong]: Shri Diwakar Publications, 2017 [Tibetan Commentary on Chapters 1–6 of Śāntarakṣita’s Tattvasaṃgraha]. But is the Tibetan rendering really based on a misunderstanding of the meaning of the name “Patañjali”?
            Not being a Sanskritist, it is not easy to dig Sanskrit sources that explain the name  “Patañjali.” But a quick online search reveals some interesting information associated with the etiology of “Patañjali.” According to one, Ādiśeṣa, the Nāga king, who is the bearer (or even emanation) of Viṣṇu, who was seeking a mother, fell into the handful of water that Goṇikā, a woman who was praying to the sun god to bestow her a son, had scooped up to offer, as the handful of water was the only gift she could find. The veracity of Patañjali’s etiology is secondary, but the Tibetan translators and their Indian collaborators who translated “Patañjali” most probably knew Patañjali’s etiology and hence translated the name as “One Who Fell into the [Handful of] Water” (Chur lhung) or “One Who Fell on the Palms [Filled with Water] (Thal mor lhung).” I, for one, do not think that the Tibetan translators misunderstood the name “Patañjali.” This case is very much comparable to the Tibetan rendering of the name “Umā” as “dKa’ b/zlog ma.” Ācārya Sems dpa’ rdo rje’s translation does not seem to account for pat or pata. With regard to the etymology of the name, Mayrhofer, after considering some speculations, states: “Alles recht ungelaubhaft.” See Manfred Mayrhofer, Etymologisches Wörterbuch des Altindoarischen. 3 volumes. Heidelberg: C. Winter 1992–2001. Cf. MW (s.v. patañjali): “fr. pata añj°?” This speculation or suggestion is not found in the PW.
            In short, we repeatedly realize that Tibetan translators, when they did translate proper names, did not always follow what seemed to them their literal meanings but considered their etiological backgrounds. So it seems it is necessary to do some background study before we conclude that the Tibetan translators misunderstood and mistranslated certain things. By the way, as Ācārya Sems dpa’ rdo rje rightly points out, the Tibetan translation of the title Tattvasaṃgraha should not really be De kho na nyid bsdus pa as if we have Tattvsaṃgṛhīta but should rather be De kho na nyid bsdu pa.


March 19, 2020

A Tale of “Śāṇakavāsin” or “Śāṇavāsin”

Here comes a tale of “Śāṇakavāsin” or “Śāṇavāsin.” I mean the name “Śāṇakavāsin,” not the person behind it. The story of the person, who is said to be one of the seven/eight persons to whom the genealogy of the Buddha’s doctrine, primarily the Vinaya as a lived or applied tradition, is said to be successively entrusted, is well known. In Tibetan, these individuals are collectively called bstan pa’i gtad rabs bdun/brgyad. I wish we can trace the Sanskrit expression for it, if there exists one. This tale is actually a tale of error.
In Monier-Williams’ Sanskrit-English dictionary (MW), we find an entry: Ślanavāsa. It is supposed to be “m. N. of an Arhat, Buddh.” If a Sanskritist raises his or her eyebrows by seeing the name Ślanavāsa, it would be understandable. Not being a Sanskritist, I became curious if it refers to one of the bstan pa’i gtad rabs bdun/brgyad or sixteen sthaviras (gnas brtan bcu drug).
By the way, these sixteen persons may be all arhants, but unlike what we find in secondary sources, they are, at least according to the Tibetan tradition, called gnas brtan bcu drug, and not dgra bcom pa bcu drug (“sixteen arhats/arhants”). My instinctive question was what its source could be. The abbreviation “Buddh.” means “Buddhist literature.” A quick online search for the name “Śāṇakavāsin” seems to invariably bring me back to MW. My suspicion grew.  Because we know that MW is based on PW, I looked it up. It states: “m. N. pr. eines Arhant Tāran. 4. 51; vgl. die Anm. auf  S. 4.” Aha, at least, the PW provides its source. And “Tāran.” here is Tāranātha’s rGya gar chos ’byung. So I had to look up Schiefner’s German translation of the Tāranātha’s rGya gar chos ’byung. To be sure, I looked at Schiefner 1869: 4, and particularly, n. 1.  The note states: “Im Tibetischen yul bslan pa oder yul slan pa; es unterliegt keinem Zweifel, das hier eine Corruption des Namens Śāṇavāsika oder Śāṇakavāsa vorliegt; es findet sich dieselbe schon in der tib. Uebersetzung des Āryamahākāruṇāpuṇḍarīkasūtra in B. 50 des Kandjur Blatt. 81; es ist aber slan pa aus sā na pa corrumpirt und sā na aus Śāṇa; vergl. Mahāvanso pag. 19 und Köppen, Religion des Buddha S. 148 und 391.” Note that I updated some of the conventions in this citation. This note is actually very valuable. I did not verify the Tibetan translation of the Mahākāruṇāpuṇḍarīkasūtra but straight away wanted to use the TBRC. Indeed, we do find a couple of occurrences in the bKa’ ’gyur.  One occurrence is from Tāranātha’s rGya gar chos ’byung. Both spellings can be found: yul bslan pa and yul slan pa. To be sure, bslan pa and slan pa were not understood by these sources to be names of some place but explicitly as a name of a dgra bcom pa or dge slong. Those scholars had no easy access to all the various sources like we do now but we repeatedly find out that they were often very cautious scholars. They deserve our respect and gratitude. What Schiefner’s note above is essentially saying is that bslan pa (or slan pa) is a corruption of śāṇaTo be sure, śaṇa is said to be a “kind of hemp.” I think this is the same as what is called in Tshang lha language zangs/bzang ru. Before cotton or silk became popular in Bhutan, people seem to have used śaṇa fabric to make clothes and other articles such as bags, sacks, and so on. I understand śaṇa to be hemp or flex plant and śāṇa to be fabric made of it or things such as garments made of this fabric. It is thus interesting to see that Tibetans also rendered śāṇa as “śaṇa cotton” (sha na’i sprin bal). See, for example, TSD. As we can see in the Tibetan rendering so ma’i gos can (for śāṇakavāsin or perhaps śāṇavāsin), śana was also understood to be the somā plant. This is a bit strange. It is also clear that Tibetans generally understood vāsa to be “garment” (gos) and vāsin to be “possessing garment” (gos can).
Coming back to the name of our arhant in question, we should first dismiss (a) Ślanavāsa. Also (b) Śāṇavāsa (PWMW) and (c) Śāṇavāsika (PWMW) seem not to be based on primary Sanskrit sources. We also find (d) Śanakavāsin (Strong 1992: 303, n. 2) but its attestation in primary Sanskrit sources or its correctness seems to be questionable. (e) Śāṇavāsin (Strong 1992: 303, n. 2) seems to be found, for example, in the Guṇakāraṇḍavyūhasūtra and Ratnamālāvadāna (GRETIL). But these deal with the same topic? It needs to be checked. It appears to be used there in the sense of those wearing śāna-garments, that is, in the plural. But possibly Chinese and Tibetan had Śāṇavāsin as the name of our arhant. See Köppen 1857: 391, n. 2, which reports three different phonetic renderings of the name in Chinese sources: Schang ho sieu, Sche na po sse, and Schang no kia po scha (sic). The third is considered to be more exact. Also forms such as Śaṇavāsa, Śaṇika, Śoṇavāsī (nominative?), Śānavāsika, and Śāṇavāsika have been recorded in Chattopadhyaya & Chimpa 1970: 24, n. 24. But what seems, however, certain is the name (f) Śāṇakavāsin, attested in the Divyāvadāna and recorded in the BHSD, can be treated as the best choice of all.
But why did we or how did we come to Ślanavāsa, which found its way into the PWMW, and so on? As we saw above, some Tibetan translators translated the name of our arhant as yul bslan pa. Obviously Tāranātha’s rGya rgyar chos ’byung has two instances where our arhant is mentioned. In one, he is called Sha na’i gos can and in another, he is called dGra bcom pa Yul bslan pa, and the latter was reconstructed by Schiefner as Ślanavāsa (Chattopadhyaya & Chimpa 1970: 13, n. 75). It does not seem to be the case that Otto von Böhtlingk and Rudolph Roth preferred Ślanavāsa over Śāṇavāsa and Śāṇavāsika. They simply recorded all three. This was followed by the MW, and the like.
Regarding the Tibetan translation “Yul slan pa” or “Yul bslan pa,” I prefer the theory that “Yul bslan pa” was simply a corruption of “Yul sha ṇa.” Two kinds of error seem to be involved here. First, the corruption of Sanskrit śaṇa (phonetically rendered in Tibetan) as slan pa (thereby assuming the form of a Tibetan word) seems to be quite an early “transmissional error.” It appears that Śāṇavāsin or Śāṇakavāsin was only partly translated. Of the two components, śana or śāṇa was not translated but merely phonetically rendered. And vāsin (or something close to it) was translated into Tibetan as yul and interpreting it as a “place” or “dwelling.” This may be called “translational error” because the name has nothing to do with “place” but certainly “garment” (as suggested by many other Tibetan translators who correctly translated the name as Sha na’i gos can). In short, the name, at least in a couple of Tibetan translations, seems to have been initially “Yul sha ṇa,” which was corrupted to “Yul slan pa” and “Yul bslan pa.” By way of conclusion, we can learn one lesson from the tale of “Śāṇakavāsin” or “Śāṇakavāsin.” While we cannot uncritically and absolutely trust any source, no matter how much aura of authoritative they might emit, we can always learn something from any source, no matter how insignificant it might appear to be.



March 11, 2020

A Sanskrit Word for rDzogs chen?

Apropos rDzogs chen or rDzogs pa chen po. In one of my earlier blog remarks, I noted that one of the four garuḍendras in the Saddharmapuṇḍarīkasūtra (Dutt 1986: 3.7) is called “Mahāpūrṇa” (rDzogs chen). But it is, of course, practically of no use for the quest of a possible Sanskrit name, if there was at all one, for what came to be known in Tibet as rDzogs chen or rDzogs pa chen po and as the “Great Perfection.” Since the great work of Samten Karmay, the English name “Great Perfection,” too, seems to have become standard. While flipping through the pages of the Tibetan translation of Vilāsavajra’s Nāmamantrārthāvalokinī, I stumbled upon this line (B, vol. 32, p. 120.7): de la mkhas pa chen po ni blo rdzogs pa chen po’o ||. It made me wonder about the Sanskrit word for rdzogs pa chen po here. Fortunately, the Sanskrit text of the Nāmamantrārthāvalokinī is extant and the pertinent line is from the portion of the text edited and studied by Anthony Tribe (i.e. Skt. Tribe 2016: 283.1–3). It turns out that the Sanskrit text here reads mahākṛtin equated with mahāpaṇḍita. That mahākṛti is not used as an abtract feminine noun but as a concrete masculine noun is confirmed by mahākṛtin’s equation with mahāpaṇḍita. Thus blo rdzogs pa chen po here is to be understood as “a great one whose intellect has become perfect/matured.” The component kṛtin is related with kṛta as in kṛtayuga, rendered into Tibetan as rdzogs ldan gyi dus (“the Golden Age”). If, and that is with a big IF, there had been a Sanskrit word for rDzogs chen or rDzogs pa chen po (“Great Perfection”), one wonders if it could have been something like “Mahākṛti.”

March 06, 2020

ཡུལ་ཆེན་པོ་བཅུ་དྲུག་གི་ནང་གསེས་སྲེག་པ་ཞེས་པའི་སྐོར།

There seem to two regions or countries with the name Aśvaka, namely, one in the north-west, and the other in the south. Aśvaka, in the sense of one of the sixteen mahājanapadas, is said to be the one in the south. In Buddhist Pāli sources, it is called Assaka (CPD), and in Sanskrit sources it is called Aśmaka as well as Aśvaka (Gupta 1989: 17–19). See also MW (s.v. “N. of a people, …  (cf. aśmaka).” The CPD does record the Sanskrit name Aśmaka but not Aśvaka (in the sense of the name of mahājanapada). A list of the sixteen mahājanapadas transmitted via the Tibetan translation of the *Vasiṣṭhasūtra (gNas ’jog gi mdo), translated by sKa ba dPal brtsegs, has already been noted by Dan Martin. The list has also been cited in dPal brtsegs’s gSung rab rin po che’i gtam. The difficulty with the list found in Tibetan sources is identifying the proper Sanskrit names behind those names translated into Tibetan or left un-translated or in partial phonetic transcription. One such name that poses a difficulty is “Sreg pa.” This is clearly a translation and not a phonetic transcription. In the *Vasiṣṭhasūtra (gNas ’jog gi mdo), it is spelled “Sreg pa,” whereas in the gSung rab rin po che’i gtam, it is spelled “Srag pa.” Of course, one should check all versions. I personally presume that the former is the correct reading and the latter corrupt. But the question it what could have been the Sanskrit name behind the Tibetan name and why has it been rendered thus? If we consult Negi’s dictionary, we see that “Aśmaka” has been rendered into Tibetan as “rDo can” (clearly as a name of country or mahājanapada). I surmise that “Sreg pa” is a rendering of Aśvaka. But is not Aśvaka supposed to mean something like “a small or bad horse” or “a toy-horse” (MW)? And what does the Tibetan sreg pa (here) mean? It turns out that sreg pa (according to Jäschke 1881) means a “partridge.” In the Tshig mzod chen mo (s.v. gong ma sreg), the bird is called bya sreg pa, a bird which is said to similar to a bya gong mo (also called lha bya gong mo or spang bya sreg pa). Some sources identify it to be a “grouse” (Jäschke 1881: s.v. gong mo). But can Aśvaka mean a certain kind of a bird? MW does state that Aśvaka also means a “sparrow” but as attested only in lexicons (L). But at least there seems to have been some basis why Tibetan translators translated Aśvaka here in the sense of a certain bird known in Tibetan as sreg pa. Old Tibetan documents (e.g. Pelliot tibétain 1285) do allude to bya’ sreg pa. In conclusion, I believe that I was able to plausibly explain that the Sanskrit name behind “Sreg pa,” as one of the sixteen mahājanapadas, could have well been Aśvaka.

February 20, 2020

སྲིན་པོའི་ཡུལ།

Padmasambhava can be regarded a historical figure, who then became shrouded in legendary and mythical accounts, and then finally became idealized, iconized, metaphysicized, and transcendentalized. The idea of Padmasambhava and what he stands for thus become more interesting and significant. Those who are skeptical of “Padmaism” but do not even have an iota of doubt about Padmapāṇism or Avalokiteśvaraism, should pause for a moment, and ask if Padmapāṇi was born at all, and if so, when and where. If not, what does the whole thing mean to Mahāyāna? But this is not my concern here.
            I recall dGe-’dun-chos-’phel stating that one could trace most of the continents and subcontinents mentioned in the Buddhist meta-geography in Jambudvīpa, so to speak, not mythical places but concrete locations in Jambudvīpa. I, too, have a feeling that one is so used to thinking in terms of mythical places that actual places on earth seem implausible. Let us take, for examples, Klu’i-yul (Nāga country), gNod-sbyin-gyi-yul (Yakṣa country), Dri-za’i-yul (Gāndhāra country), and Srin-po’i-yul (Rākṣasa country). Nāgārjuna “invited” the Śatasahasrikā from the Nāga country and Padmasambhava departed for the Rākṣasa country. These places are inaccessible to ordinary human beings as far as the assumption goes. But it turns out, for example, that the Gāndhāra country is a real or actual place on earth. So, my question is: How about the Rākṣasa country? Is it a real/actual place on earth?
            The textual sources of Padmaism state: “I am [now] going to the Rākṣasa country in the South-West” (nga ni lho nub srin po’i yul du ’gro). Furthermore, “I am going; I am going to the Rākṣasa country” (nga ’gro srin po’i yul du ’gro). Recently, some Tibetan scholars have written something online about the location of the Rākṣasa country. The issue is that there is the idea of Rākṣasa country and of Rākṣasa continent/island (Srin-po’i-gling), and the latter is identified with Cāmara/Ḍāmara continent (rNga-yab). The two directions that we find in these contexts are South-West (lho nub) and North-West (nub byang). One should not forget that directions are dependent on one’s perspective. It is true that Rākṣasadvīpa is associated with Rāvaṇa and Laṅkā. It would seem that if we identify the Rākṣasa country with Rākṣasadvīpa, then the direction South-West would be more reasonable. The Tibetan traditions, not just those associated with Padmaism, seem to have two different identifications of the Rākṣasa country, one Rāvaṇa’s Rākṣasadvīpa (i.e. Laṅkā) and the other Oḍḍiyāna (or Oḍiyāna, Oḍyāna, Oḍḍayana, Uḍiyāna, Uḍyāna, Uḍḍayana) in the Swat valley. These two places seem to have been conflated and confused. Kaḥ-thog Tshe-dbang-nor-bu (TBRC) lists several places in that area including Kashmir (Kha-che), Gilgit-Baltistan (Bru-sha), Bactria/Tocharistan (Tho-gar), and Rākṣasa country (Srin-po’i-yul). There are some more places listed. Perhaps Rākṣasa country is to be seen as a specific region within the broader domain of or in the precinct of Oḍḍiyāna. This is, in my view, exactly what F. W. Thomas does and places Rākṣasa country in the South-West of Oḍḍiyāna (Thomas 1935: 291). I personally am inclined to think that when Padmasambhava said “I am [now] going to the Rākṣasa country in the South-West” (nga ni lho nub srin po’i yul du ’gro), he meant that he just wanted to go back home!




February 11, 2020

འགྲེས་རྩ། འགྲེས་རིང་། འགྲེས་ཐུང་། འགྲེས་རྐང་།

In the Tibetan accounts of the Śatasāhasrikā, we find terms such as ’gres rtsa’gres ring, and ’gres thung. One can notice from the context that ’gres is certainly used as a technical term: “basic ’gres, longer ’gres, and shorter ’gres.” But what exactly is ’gres in our context? The closest plausible word seems to be ’gres rkang. See the Tshig mdzod chen mo (s.v. ’gres rkang): tshigs bcad dum bu re’i mtshams su gzud rgyu’i rigs ’gres gtong rung ba’i tshig rkang. According to this, a ’gres rkang is a pāda or a line which occurs repeatedly and hence can be inserted in between blocks of verses or passages. But this does not seem to be only meaning. Some sources (see TBRC or BDRC) simply seem to use ’gres rkang in the sense of “items” of phenomena (i.e. “subject of the property of emptiness” (stong gzhi chos can) usually divided to those that belong to saṃkleśa (e.g. 5 skandhas) and those that belong to vyavadāna (e.g. 6 pāramitās). Even in this sense, what would exactly be ’gres rtsa’gres ring, and ’gres thung? This needs further investigation. The following expressions seem to shed some light the use of the word ’gres rkang: ’gres rkang gi mtshams la ’jug par bya; gzhan rnams ’gres rkang du bzhag ’dug pa; ’gres rkang phyes ma phyes kyi khyad par tsam; ’gres rkang ded par bya; ’gres rkang sbyor ba; de bzhin du ’gres rkang ded pas legs po ’ong mchis so ||.





January 20, 2020

Trying To Solve a Philological Mystery


A master philologist is a like a master detective. I cannot claim to be a master philologist but like any philologist or a detective, I believe to have a case and I believe to have solved it. I wonder if and how convincing my solution may seem to other philologists. At least, let us say, I have a hypothesis. 
Of the four or five Indian commentaries in Tibetan translation on the Bhadracarī, the *Bhadracarīpraṇidhānarājaṭīkā is ascribed to one Indian author whose Sanskrit name has been rendered into Tibetan as “rGyan bzang po.” Whether he and Śākyamitra each composed a separate ṭīkā or whether they co-authored one ṭīkā seems worth investigating. But this issue is not significant for my present case. The issue that concerns me here is what could have been the Sanskrit name of “rGyan bzang po.” To be sure, the name in Tibetan is well attested in some of the earliest catalogues and other textual witnesses and hence it is not an issue here. Already in 1912, in his German dissertation on the Bhadracarī, Kaikioku Watanabe (Watanabe 1912: 23), reconstructed the name as “(Vi)bhūṣaṇabhadra” or Alaṃkārabhadra.” But like any good philologist, he was cautious. He conceded: “The exact form of the name of the composer remains unsettled because it is otherwise unknown” (Die genaue Form des Verfassernamens bleibt unbestimmt, weil er sonst unbekannt ist). Dan Martin in his Tibskrit has reconstructed “rGyan bzang po” as *Subhūṣita. However, in the same Tibskrit, Martin records “Bhadrapaṇa” to which he adds “Author of a commentary on the Bzang po spyod pa’i smon lam. It’s listed as no. 5515 in the Peking Tanjur. Schaik, Prayer, p. 185” (i.e. van Schaik 2009: 185). But actually in Herrmann-Pfandt 2008: 319, the bZang po spyod pa’i smon lam (P 5515) in question here has already been identified as the *Bhadracarīpraṇidhānarājaṭīkā by “rGyan bzang po.” The Sanskrit name of the author is given as Bhadrāpaṇa in Herrmann-Pfandt 2008: 319. There is a remark in German (“Identifikation eindeutig”). But there is no remark to the name of the author. The bilingual name has been indexed (pp. 466, 472) and the orthography of the name in Sanskrit is consistently “Bhadrāpaṇa.”
The question now is how can we explain the Sanskrit–Tibetan correspondence of the name. There is no difficulty with bhadra and bzang po. It has been presupposed that bhadrāpaṇa = bhadra + āpaṇa. The second component of the name cannot possibly be apaṇa. But can āpaṇa mean rgyan? According to MW (s.v. āpaṇa), it can mean “a market, a shop,” “waves” or “commerce, trade” (only in lexicons with no real usage). But there seems to be no element of Tibetan rgyan in this word. And more importantly, does the Tibetan word really mean “ornament” or “ornamented” as the earlier attempts of reconstruction presupposed? But before we answer this, let us consider the orthography “Bhadrapaṇa” that Martin gives based on van Schaik 2009: 185. And what does paṇa mean? According to MW (s.v. paṇa), it can mean “paṇa m. (ifc. f(ā).) play, gaming, playing for a stake, a bet or a wager (with gen.; loc. or ifc.; paṇaṃ√kṛ, to make a bet; paṇe ni√as, to stake at play),” “the thing staked or the sum played for, wages, hire, reward.” But would such a meaning agree with the meaning of the Tibetan word rgyan? Yes, it would. According to Jäschke 1881 (s.v. rgyan, II1–2), rgyan can, as a noun, mean “a stake or pledge at play” and “lot” and in the verbal construction (as in rgyan btsugs/rgyab) it could mean “to bet” or “to wedger.” This is also confirmed by the Tshig mdzod chen mo (s.v. rgyan 2): rgyal dang skugs, ... sho sbag brgyab nas rgyan rgyal ba, ... dngul zong ci rigs kyi rgyan ’gyed pa, ... shing bu dang, rde’u sogs bsgrugs te rgyan rgyag pa, ... rgyal pham thob shor gyi ched du rgyan ’phen pa, ... rtsed mo thob shor gyi ched du dngul ’bor chen rgyan du btsugs pa. Thus the word rgyan in the name “rGyan bzang po” must mean something like “bet” or “lot,” and the Sanskrit name, with great probability, had been “Bhadrapaṇa,” and meant something like “Good Bet” or “Good Lot.”