1–31 ISSN: 2313-2000 e-ISSN: 2313-2019
On the Ekottarik gama T 125 as a Work of Zhu Fonian
Michael Radich
Senior Lecturer, Religious Studies, Victoria University of Wellington, New Zealand
Abstract
On the basis of a large set of diverse stylistic markers, this paper argues that the Ekottarik gama T 125 was translated by Zhu Fonian, and not by Sa ghadeva. The paper also considers implications of its findings for the broader corpus of texts ascribed to Zhu Fonian, and for methods in assessing ascriptions of Chinese Buddhist texts on the basis of internal evidence.
Keywords:
Ekottarik gama, Zhu Fonian, Sa ghadeva, Chinese Buddhist translations, ascription
Michael Radich
Introduction
The Chinese tradition contains mixed testimony about the translatorship of the Ekottarik gama. It is likely that multiple versions once circulated, and the extant canonical text, the Zengyi ahan jing T 125, is ascribed variously to *Dharmanandin and (Gautama) Sa ghadeva
in the various versions of the text recorded in the apparatus to the Taish Tripi aka.1 Consequently, modern scholarship has long been riven over the question of whether it is correct to ascribe the extant text to Sa ghadeva or to Dharmanandin’s team, in which the main work of translation would have been done by Zhu Fonian (d.u.).2 In a companion study to the present work,3 Ven. An layo Bhikkhu and I argued that there are too many stylistic differences between T 125 and Sa ghadeva’s benchmark work, the Madhyam gama T 26, for the two to be due to the same hand. This argument was based upon a very wide range of stylistic markers of many types, including transcription terms, translation terms, formulae for the opening and closing of s tras, formulae for various phases of practice and moments on the path, other Buddhist technical terms, proper names for persons and places common in Buddhist texts, and ordinary nouns for realia. Many of these markers, moreover, occurred many times in their respective texts, suggesting that they were reliable and frequent features of the respective authors’ style. The internal evidence we adduced in that argument was already far more copious and varied than that adduced by any previous studies of the same problem.4
1 T 125, 2: 549b11 and n. 11.
2 For detailed references on these various dimensions of the problem, see Radich and An layo, “Were the Ekottarika- gama.”
3 Radich and An layo, “Were the Ekottarika- gama.”
4 Previous studies have relied on a fairly small number of stylistic markers. For instance, Nattier, “‘One Vehicle’,” 195–196 n. 48, discusses about six markers distinguishing the Ekottarik gama from the Madhyam gama; Unebe, “Jiku Butsunen” is usually referred to as the most detailed study to date of Zhu Fonian’s style, but builds its argument on a single set of terms (for the members of the eightfold path), which, moreover, are weaker as criteria for the translatorship of the Ekottarik gama than many of the terms we examine below;
Unebe, “T shin” (the main aim of which is not to examine our present question) discusses five terms or sets of terms; Lin, “Xiancun” uses approximately nine sets of markers (albeit strong ones). Mizuno, “Kan’yaku...yakushutsu,” 88–89, ironically studies the largest number of markers of any of these authors (36 markers), only to arrive at an incorrect result (see n. 38).
However, in that study, we contented ourselves with the negative conclusion that T 125 is highly unlikely to be by Sa ghadeva. The present study aims, as far as possible, to identify the most likely author of T 125, again by studying the terminology and phraseology of the text in much greater detail than has previously been attempted. I will adduce terminological and phraseological evidence even more copious than that adduced in the companion study, to argue that it is overwhelmingly probable that T 125 was translated by Zhu Fonian.
Preliminaries: Selection of the Benchmark Zhu Fonian Corpus
External evidence, as discussed above, presents us with two main candidates for translatorship (or authorship5) of the Ekottarik gama: Zhu Fonian, and Sa ghadeva.6 In this study, I will therefore focus on stylistic markers capable of distinguishing between these two figures. That is to say, I will examine the Ekottarik gama T 125 for markers characteristic of Zhu Fonian’s idiom, in contrast with that of Sa ghadeva.
To render my treatment as robust as possible, I will take as my benchmark for Zhu Fonian’s style a conservative pool of texts. The principal criteria in assembling this corpus are these: 1) the ascription to Zhu Fonian on the basis of external evidence should be unproblematic; and 2) the texts should be genuine translation texts. This leads me immediately to exclude a number of texts from my benchmark corpus:
5 By this equivocation, I mean to mark the possibility that some portions of the collection may have been added in China, and that Zhu Fonian could himself be responsible for some such additions. I will return to this problem below.
6 As always in studies of such corpora, “Zhu Fonian” should be regarded as a shorthand label of convenience for “(the) translation group(s) centering on Zhu Fonian,” and similarly “Sa ghadeva” for “(the) translation group(s) centering on Sa ghadeva.” It may ultimately be possible to distinguish between markers of the contributions of individual members of such translation groups to the collective style, but the present study makes no pretense of being so powerful. It is nonetheless important to think in terms of the group(s) centering on real translators such as Zhu Fonian, rather than groups centering on Indic reciters (or even figureheads) such as *Dharmanandin, because the latter focus can lead us on a wild goose chase in search of supposed “authorial signatures” that may not exist.
the Mohe banre chao jing T 226, which is probably not due to Zhu Fonian;7
the Shi zhu duan jie jing T 309,8 which Nattier has shown was composed in China, and is not a genuine translation;9 the Pusa chu tai jing T 384, which may also be a
Chinese composition;10
7 Kajiyoshi, Genshi, 68–76. On the basis of a complex discussion, Kajiyoshi concludes that the extant T 226 is most likely to be an alternate translation of the
as hasrik prajñ ramit by Dharmarak a, which is noted variously in the catalogues as the “New Daoxing jing” and the “retranslation of the shorter version [of the Prajñ ramit - tra]” . For our present purposes, however, it suffices to note Kajiyoshi’s grounds for believing that the text has nothing to do with Zhu Fonian. Kajiyoshi notes that the Chu sanzang ji ji
T 2145 preserves a preface by Dao’an (312/314–385?) to a text with a title very similar to our extant T 226: Mohe boluore boluomi jing chao ; T 2145, 55: 52b8–52c26. However, the details of this notice show that it could not have referred to the present T 226. Rather, it describes a text which was a partial retranslation of the larger Prajñ ramit , which incorporated parts of the earlier translations by Mok ala (Fang guang banre jing T 221) and Dharmarak a (Guang zan jing
T 222). Moreover, the notice actually states that this translation was made by Fohu (and Dharmamitra , who “held the [foreign] text”), not by Zhu Fonian. Finally, Kajiyoshi also argues that Fohu (Fotuluocha ,
*Buddharak a?) cannot be identified with Zhu Fonian, since a number of notices in the tradition clearly refer to these two figures in distinct capacities, e.g. as producing successive translations of the same text, or working together on a single translation (see T 2145, 55: 99a25–b5; 99b7–9; 73c3–8; 64c11–15). It seems clear, then, that the ascription of the extant T 226 to Zhu Fonian is based upon a double error: the conflation of Fohu with Zhu Fonian, and the conflation of the extant T 226 with a different, lost text that originally bore a title very similar to our extant T 226.
8 In using this shorter title, which is the mode in which the text refers to itself internally, I follow Nattier, “Re-evaluating,” 231 and n. 2. In the Taish , the text is given the more fulsome title Zuishengwen pusa shi zhu chu gou duan jie jing
, which seems to date from Zhisheng (writing in 730).
9 Nattier, “Re-evaluating.” Nattier shows that the Chinese sources of the text include at least *Mok ala’s Fang guang banre jing T 221; the Chengju guangming dingyi jing T 630 ascribed to Zhi Yao
, but itself probably composed in China (Nattier, “Re-evaluating,” 241–242 n.
26; Nattier, Guide, 96–102; some of the connections between T 309 and T 630 had already been partially studied before Nattier by Pu, “Notes”); and the
*Ak amayati-nirde a T 403 of Dharmarak a.
the *Antar bhava-sutra T 385, which may also be a Chinese composition;11
the Da yun wuxiang jing juan dijiu T 388, which I have shown is almost certainly not due to Zhu Fonian;12 and
the Pusa yingluo benye jing T 1485, which has long been regarded as an “apocryphon.”
We also exclude the Jñ naprasth na T 1543, since it is ascribed to Zhu Fonian and Sa ghadeva as a team, and as such, should in principle not help us distinguish between the styles of our two candidates.
Finally, I also exclude four other texts. These texts were possibly actually translated by Zhu Fonian, but the information the tradition gives us about their translation is conflicted, and all four bear the names of other “translators” in the Taish canon:
The Zun Poxumi pusa suoji lun T 1549: In the Taish , this text is credited to *Sa ghabh ti/Sa ghabhadra[?]
. However, Zhu Fonian’s role as a translator is affirmed by an anonymous preface preserved in the Chu sanzang ji ji
T 2145, which states that [Sa gha]bhadra/~bh ti?), [Dharma]nandin , and [Sa gha]deva / merely “held [read] the Western text” .13
Ayu wangxi huai mu yinyuan jing T 2045: In the Taish , this translation is ascribed to *Dharmanandin . However, the Chu sanzang ji ji preserves a preface by Zhu Fonian himself. Although the evidence is circumstantial, this preface seems to indicate that the actual work of translation was done by Zhu Fonian. Not only is it the only extant preface in Zhu Fonian’s name, but the preface also concludes with his reflections upon his
10 = Pusa cong doushutian xiang shen mu tai shuo guangpu jing
. Legittimo, “Synoptic,” “Analysis;” Nattier, “Re-evaluating,”
234, 256.
11 Nattier, “Re-evaluating,” 256. Nattier notes that several of the problematic works here (T 309, T 384, T 385) share with T 656 certain common features: they are
“sole exemplars” (i.e. they have no attested parallels in any language); and they date to a later period in Zhu Fonian’s career, when the circumstances under which he was working had changed, and he seems to have been working alone;
Nattier, “Re-evaluating,” 234–35, 256.
12 Radich, “Problems.”
13 T 2145, 55: 71c8–72a8; repeated at T 1549, 28: 721a3–b4.
translation practice, which would be odd, if the preface were not attached to one of his translations.14 Elsewhere, the Chu sanzang ji ji simply ascribes this same text directly to Zhu Fonian, and states that when Dharmanandin “issued” the text, Zhu Fonian served as interpreter as well as writing the preface.15
The Si ahanmu chao jie T 1505: In the Taish , this text is ascribed to “Kum rabuddhi et al.” . However, an anonymous preface transmitted in Chu sanzang ji ji states that the role played by Kum rabuddhi was to “hold [read out] the Western text” … … , whereas the actual translation was by Zhu Fonian and Fohu (*Buddharak a?).16 The Sengqieluocha suoji jing T 194: In the
Taish , this text is ascribed to “Sa ghabh ti/Sa ghabhadra[?]
[“et al.” , Korean only]. However, an anonymous preface to the text preserved in the Chu sanzang ji ji states only that the text had been brought to Chang’an by Sa ghabh ti/Sa ghabhadra(?) . The preface does not directly identify his role in the translation process (on the pattern of other prefaces, it seems likely that he merely “held”, i.e. read, the
“Western text” ). On the other hand, the preface does explicitly state that Zhu Fonian translated, and Huisong acted as amanuensis.17 Our information for this text is complicated further by the fact that we also have a conflicting postface, which holds that Sa ghabh ti/Sa ghabhadra(?) recited the text, and the translation was done by *Vibh (sic) [and?]
*Buddharak a (or could this mean something like “the Vaibh ika *Buddharak a”?).18
Excluding these texts19 yields the following conservative corpus as a benchmark. As I will show, this corpus certainly suffices as a reference point
14 T 2145, 55: 51b14–c16.
15 T 2145, 55: 10c4–6, 111b16–18.
16 T 2145, 55: 10b13–16 (repeated at T 1505, 25: 1a3–24; see also T 2145, 55:
64c3–23).
17 T 2145, 55: 71b2–23.
18 T 2145, 55: 71b24–c7; Kamata, Ch goku, 107–108, argues that this postface is mistaken.
19 Sources also associate the following lost texts with Zhu Fonian:
1) The Biqiuni da jie , for which a note preserved in Chu sanzang ji ji says Zhu Fonian “held [in his hands, i.e. read/recited] the Western [text]”
to establish a very rich set of markers of Zhu Fonian’s regular style, as it contrasts with that of Sa ghadeva.20
rgh gama T 1 (D );
Ud navarga T 212;21
Pusa yingluo jing T 656;
. However, as Kamata suggests, this does not seem to indicate that Zhu Fonian was responsible for the actual translation of this text; CSZJJ T 2145:
81b21–24 (also 10a26–29), Kamata, Ch goku, 97–98, 102.
2) The Biqiu [shi song] da jie [ ] : One note in the Chu sanzang ji ji says that Zhu Fonian translated this text; T 2145, 55: 10a23–25. However, according to Dao’an’s preface, Zhu Fonian’s role was only to write down (or copy?) the Indic (Sanskrit?) text , whereas the actual translation was done by Daoxian ; T 2145, 55: 80b5.
3) A Madhyam gama , which appears in Sengyou’s
alongside the notice of Zhu Fonian’s Ekottarik gama, with the information that the text was recited orally by Dharmanandin, and translated by Zhu Fonian
…… ; T 2145, 55: 10b21–26. Mizuno
“Kan’yaku...yakushutsu,” “Kan’yaku,” suggested that fragments of this collection were preserved in a range of individual texts extant in the present Taish canon. However, Hung, Bingenheimer, and Wiles, “Quantitative,”
problematizes the ascription of these texts to Zhu Fonian and his collaborators.
They found that the group of 24 Madhyam gama texts studied by Mizuno are indeed united by a common stylistic signature, but that “on the basis of the present research we are not able to prove that these 24 texts were translated, as Mizuno holds, by Dharmanandin and Zhu Fonian specifically,” 122.
4) Huan wang jing , mentioned in the anonymous postface to T 194 preserved in the Chu sanzang ji ji, T 2145, 55: 71c1–2.
20 Su, “Terms” has examined instances in three texts of this core Zhu Fonian corpus ( rgh gama, T 212 and T 1428) in which the same Indic term is translated more than one way. As Su’s examples show, the terminology and style of this corpus is not entirely uniform or stable. However, the evidence surveyed below shows amply that even if variety exists, there is also sufficient regularity across these texts to identify reliable markers of Zhu Fonian’s translatorship/authorship—
some of them exclusive (see Table 7).
21 Note, however, that Hiraoka, “Shutsu y ky ,” has argued that the prose portions of the Ud navarga T 212 contain narrative material aligning with multiple Mainstream transmission traditions, and the text is therefore unlikely to have been composed in its present form in an Indic context. Rather, he suggests that it was probably expanded and modified at the point of translation into Chinese.
Hiraoka also mentions, following Watanabe Kazuko , that T 212 borrows a number of stories from the Faju piyu jing T 211, which Mizuno and Enomoto have argued was itself composed in China;
Hiraoka 843, 844 n. 7.
the Dharmaguptaka-vinaya T 1428; and the Binaiye T 1464.
Our only solid reference point for the style of Sa ghadeva is the Madhyam gama T 26, and I will therefore take that text as our benchmark for his style.
Examination of Stylistic Markers
The Ekottarik gama shares a large amount of phraseology and terminology with the Zhu Fonian texts in this benchmark corpus, while those same markers are never found in the Madhyam gama.
Not only are phrases matching our criteria copious in this Zhu Fonian reference corpus, but they are also of various types, and this fact, in itself, makes the likelihood stronger that we are seeing a global difference between two authorial styles. It will therefore be convenient to list the evidence by type.
First, we find various proper names—the names of common persons, places, gods etc.—shared by the Ekottarik gama and the Zhu Fonian reference corpus, but never found in the Madhyam gama.
Conventions for all Tables
None of the terms listed ever appears in the Madhyam gama.
Counts are approximate.22
Items already discussed in Radich and An layo, “Were the Ekottarika- gama,” are listed in [square brackets].
Bold indicates especially strong markers. Markers for which counts are bold (middle column) appear over 100 times in the Ekottarik gama. Markers for which benchmark corpus texts are bold (rightmost column) appear in all five benchmark texts (as well as the Ekottarik gama).
Indic equivalents and/or English translations for Chinese terms and phrases are given only to aid readers in assessing the type of language at issue. They
22 All counts based upon electronic searching, as these are, should be regarded as provisional, since the digitization process may be subject to error; since most such searches (e.g. with the CB Reader) do not take into account textual variants, even as witnessed in the Taish apparatus; and because the Taish editing process, including the apparatus, was itself subject to error.
do not represent a claim that the Chinese in every instance in the texts corresponds to exactly the Indic term or meaning given.
Table 1: Names of persons etc. found in the Ekottarik gama and the Zhu Fonian corpus, but never in the Madhyam gama (see above for conventions)
term instances in
Ekottarik gama
other Zhu Fonian texts
[ ]
riputra
411x , T 212, T 656, T 1428, T 1464
akrodev Indra
150x , T 212, T 656, T 1428, T 1464
[ ]
Jetavana
132x , T 212, T 1428, T 1464 (copious)
[ ]
jag ha
108x , T 212, T 1428, T 1464
[ ]
Aj ta atru
91x , T 212, T 1428, T 1464
kyamuni
63x , T 212, T 656, T 1428
Vai
48x , T 212, T 1428
dhak a
23x , T 212, T 1428, T 1464
[ ]
Jambudv pa
19x , T 656, T 1428, T 1464
Up li
16x T 1428 (copious), T 1464
Next, we also find transcription terms with the same pattern of distribution.
Table 2: Transcription terms found in the Ekottarik gama and the Zhu Fonian corpus, but never in the Madhyam gama (see above for conventions)
term instances in
Ekottarik gama
other Zhu Fonian texts
rama abr hmana
112x , T 212, T 1428, T 1464
yojana
78x , T 212, T 656, T 1428, T 1464
[ ]
asura
74x , T 212, T 656, T 1464
23 nirv a
17x , T 212, T 656, T 1464
sa yaksa buddha
13x , T 212, T 656, T 1464
asa khyeya
9x , T 212, T 656, T 1428
The same pattern of distribution marks numerous Buddhist terms which are translated into Chinese (or translated in part), rather than transcribed.
Table 3: Buddhist technical terms found in the Ekottarik gama and the Zhu Fonian corpus, but never in the Madhyam gama (see above for conventions)
term instances in
Ekottarik gama
other Zhu Fonian texts
[ ]
“Tath gata, Arhat, Sa yaksa buddha”
74x , T 212, T 656, T 1428
[( ……)
] *spar a,
*spra avya
54x , T 212, T 656, T 1428, T 1464
“classes of sentient beings”
51x , T 212, T 656
[ ]
*durgati[traya]
48x , T 212, T 656, T 1428
[ ]
yatana
45x , T 212, T 656, T 1464
“supernatural power”
44x , T 212, T 656, T 1428, T 1464
“three monastic robes”
38x , T 212, T 656, T 1464
“precepts and monastic rule”
37x , T 212, T 656, T 1428
“the Sa gha”
36x , T 212, T 656, T 1428, T 1464
tra
31x , T 212, T 656, T 1428
“...cessation, and the path”24
27x , T 212, T 656, T 1428, T 1464
23 See Palumbo, Early Chinese Commentary, 123 and n. 62, 91 n. 186, 216.
24 Nirodha and rga, viz., the third and fourth noble truths.
[ ] ki cany yatana
26x , T 212, T 1428
du khasatya
20x , T 212, T 1428, T 1464
25 paranirmitava avartin
20x , T 1428
bh svara
20x , T 212
“emptiness sam dhi”
18x , T 212, T 656, T 1428
nirodhasatya
18x , T 212, T 1428
“countless expedients”
17x , T 212, T 656, T 1428, T 1464
samudayasatya
15x , T 1428
“minds of sentient beings”
12x , T 212, T 656, T 1428
[ ]
rgasatya
10x
[ ]
pra rabdhi- sa bodhya ga
8x , T 656, T 1428
dharmakath
6x , T 1428, T 1464
25 Cf. also , which has a similar distribution. Jan Nattier points out that the combination of two supposed heavens named + is apparently an error distinctive of Zhu Fonian; for paranirmitava avartin makes sense, but appears to name a heaven that does not exist in Indian Buddhism (*nirmitava avartin); Nattier (personal communication). (without ) first appears in Dharmarak a for paranirmitava avartin. In those contexts it is clearly distinct from *nirm arati, which appears next to it in lists and is translated or
(presumably by an etymologization from na “pride” = ); e.g. T 186, 3:
489c13–16. (without ) is otherwise very restricted in distribution (thereby incidentally forming part of a widespread pattern in which Zhu Fonian’s idiom shows unusually heavy debts to Dharmarak a). However, in lists in Zhu Fonian texts, appears alongside , and any other more usual or comprehensible equivalent for nirm arati is missing (cf. Nattier, “Re- evaluating,” 252). This could mean that the name has somehow been reinterpreted as nirm arati (as an anonymous JCBR reviewer suggests), but if so, the logic by which this equivalence is arrived at is opaque to me.
“*asa khyeyakalpa things (sic?)”
6x , T 212, T 656
*uttaro manu yadharma
5x , T 212, T 656, T 1428, T 1464
( )26
“attachment to precepts”
5x , T 212
“cut off the fetters”
4x , T 212, T 656, T 1428
“joined palms” = añjali
3x , T 212, T 1428 (copious), T 1464
“receive full precepts”
2x T 1428 (copious)
In addition, we also find common nouns, verbs, and adjectives matching the same pattern (though in classical Chinese, the boundaries between these parts of speech can in some cases be fuzzy).
Table 4: Common nouns, verbs and adjectives found in the Ekottarik gama and the Zhu Fonian corpus, but never in the Madhyam gama (see above for
conventions)
Where the boundary between technical terms and ordinary nouns is fuzzy, the term is listed here, rather than in Table 3.
term instances in
Ekottarik gama
other Zhu Fonian texts
27 “appear” 174x , T 212, T 656, T 1428
“...and suchlike”
125x , T 212, T 656, T 1428, T 1464
“deluded notions”
98x , T 212, T 656, T 1428
“benefit” 80x , T 212, T 656, T 1428
26 This term refers to the second of three types of “fetters” (sa yojana), namely, rigid attachment to the letter of precepts. It is quite restricted in its distribution in the canon as a whole. In texts supposedly before Zhu Fonian, it appears only in T 1557 (7x), ascribed to An Shigao. It never appears in Kum raj va. Among other texts contemporary to Zhu Fonian, it appears in T 1506 (4x), T 1547 (47x), and T 1550 (9x). Outside our “reference” corpus, it also appears in other texts ascribed to Zhu Fonian: T 309 (1x), T 1543 (86x), T 1549(8x).
27 Cf. also , Ekottarik gama (96), D , T 212, T 656, T 1428—also, naturally, never in the Madhyam gama.
“medicine”
81x , T 212, T 656, T 1428, T 1464
28
“secluded and quiet”
80x , T 212, T 656, T 1428
“spirit(s)” 63x , T 212, T 656, T 1428
“calamity” 58x , T 212, T 656, T 1428
29
“suspicion”
53x , T 212, T 656, T 1428, T 1464
“beginning and end, primary and ancillary”
41x , T 212, T 656, T 1428, T 1464
“foolishness” 39x , T 212, T 656, T 1428
[ ]
“village”
38x , T 212, T 656, T 1428, T 1464
“benevolence, favour”
38x , T 212, T 656, T 1428
“morning” 37x , T 212, T 1428, T 1464
“teachings” 26x , T 212, T 656, T 1428
“length and breadth”
24x , T 212, T 656, T 1428, T 1464
“receive”
24x , T 212, T 656, T 1428, T 1464
“flow” 22x , T 212, T 656
“transformation”
17x , T 212, T 656, T 1428, T 1464
“anger” 10x , T 212, T 656
“heavenly palace”
11x , T 212, T 1428, T 1464
“steadily mindful”
6x , T 212, T 1428
“past and present”
2x , T 212, T 656
“refined cane sugar”
2x , T 212, T 1428 (copious), T 1464
We also find numerous formulaic phrases, regular collocations, and recurring combinations of ordinary words distributed in the same manner.
28 The more specific phrase appears in Ekottarik gama (52x) and T 212 (5x).
29 Often in phrases like , .
Table 5: Recurring phrases found in the Ekottarik gama and the Zhu Fonian corpus, but never in the Madhyam gama (see above for conventions)
term instances in
Ekottarik gama
other Zhu Fonian texts
[ ]
“At one time, the Buddha was at...”
440x , T 212, T 656, T 1428
“because/the reason being that...”
284x , T 212, T 656, T 1428, T 1464
“said to the Buddha”
240x , T 212, T 656, T 1428, T 1464
[ ]
“sat to one side”
174x , T 212, T 656, T 1428, T 1464
……
“replied, ‘Indeed...’”
112x , T 212, T 656, T 1428
“the monk replied”
104x , T 212
30 “could not control himself”
89x , T 212, T 656, T 1428, T 1464
“know it as it really is”
83x , T 212, T 656, T 1428
“at that time, nanda”
77x , T 212, T 1428, T 1464
“‘took’ extinction”
76x , T 212, T 656
“attained purity of the Dharma eye”
62x , T 212, T 656, T 1428, T 1464
“for this reason”
58x , T 212, T 1428, T 1464 31 “in the air” 54x , T 212, T 656, T 1428
“as you say”
54x , T 212, T 656, T 1428, T 1464
[ ]
“birth and death are exhausted, and brahmac rya is
complete”
52x , T 212
30 Cf. also , which accounts for many (but not all) of these instances of , and has a similar distribution.
31 Cf. also .
“end his life” 45x , T 212
“felt terror” 42x , T 212, T 1428
32
“entered into a hell”
42x , T 212, T 656, T 1428, T 1464
[ ]
“become an arhat”
41x , T 212, T 1428
“for this reason”
39x , T 212, T 1428, T 1464
[ ]
“attain arhatship”
34x , T 212, T 656, T 1428, T 1464
X “arises from X meditation...”
19x T 1428, T 1464
33 “took a seat” 18x , T 1428 (copious), T 1464 “go and tell” 12x , T 212, T 1428, T 1464
“appeared before [him...]”
10x , T 212, T 656, T 1428
X 34
“in addition, there are X dharmas”
6x , T 656, T 1428
[ ] “bared the right shoulder”
5x , T 1428 (copious)
“at another time”
4x T 1428 (copious)
“receive and preserve [the text] and
recite it...”
3x T 656 (copious)
“the dharma of which one thinks”
3x , T 212, T 656
“knows the minds of other people”
2x , T 212, T 656
“travelling among humankind”
2x , T 1428 (107x)
35 “...treasure [is] complete”
2x , T 656
“was in [his] mother’s womb”
1x , T 212, T 656
32 Cf. Lin, “Xiancun,” 135–36.
33 Cf. also , which has a similar distribution.
34 “X” represents a number, e.g. .
35 Variously, in reference to the saptaratna, or one or another thereof.
We even find conjunctions, adverbs, and verbs of speech and thought matching the same pattern.
Table 6: Conjunctions, adverbs, and verbs of speech and thought found in the Ekottarik gama and the Zhu Fonian corpus, but never in the Madhyam gama
(see above for conventions)
term instances in
Ekottarik gama
other Zhu Fonian texts
“never again”
130x , T 212, T 656, T 1428, T 1464
“had this thought...”
87x , T 212, T 656, T 1428, T 1464
“reply”
83x , T 212, T 656, T 1428, T 1464
“not un-”
(double negative)
44x , T 212, T 656
“call” 11x , T 212, T 656, T 1428
[ ] “first” 9x , T 212, T 656, T 1428
“ask” 8x , T 1428
“think”
6x , T 212, T 656, T 1428, T 1464
“say to oneself”
3x , T 212, T 1428 (copious), T 1464
In sum, Tables 1–6 show that the Ekottarik gama contains a very large number of markers that are common to the texts most reliably ascribed to Zhu Fonian, none of which ever appear in the Madhyam gama. Some of these markers (counts shown in bold in the tables, central column) appear over 100 times in the Ekottarik gama, and thus constitute very strong evidence that the markers concerned are regular, recurring habits of the translator, and therefore, more likely to be significant for the identification of translatorship based upon style. Similarly, other markers (text names/numbers shown in bold in the tables, rightmost column) appear in all five of our benchmark texts, which are most reliably ascribable to Zhu Fonian (as well as in the Ekottarik gama).
These markers are thus overwhelmingly likely to constitute reliable, regular features of Zhu Fonian’s idiom, especially as distinguished from that of Sa ghadeva. In addition, these markers cover a wide range of types of language—the names of many of the most central persons and places in the Buddhist tradition; technical Buddhist terms, in both transcription and translation; recurring formulaic phrases, collocations and stereotyped phrases;
and “nuts and bolts” functional parts of speech like conjunctions, verbs of speech and thought, and adverbs.
The sheer quantity of such markers found in the Ekottarik gama—
thousands in total—is also powerful evidence that it is vastly more probable that Zhu Fonian translated the text, than that Sa ghadeva did. Readers should also note that even the copious evidence of the above tables is merely a sampling of such markers, rather than an attempt to list them exhaustively. As observant readers may have noted, in a somewhat artificial attempt to render more manageable the sheer quantity of evidence, while at the same time ensuring that all markers listed were relevant not just to Zhu Fonian but to his
gama idiom, the tables above were largely restricted to terms that also appear in the rgh gama. If this artificial restriction were removed, even more markers of this type could certainly be found in quantity.
Finally, we can also find even stronger internal evidence that Zhu Fonian was probably the translator of our extant Ekottarik gama. Table 7 lists terms and phrases that appear in the Ekottarik gama, but are otherwise only ever found in (other) texts by Zhu Fonian—that is, never in any other translation texts in the entire Taish Tripi aka; this means, naturally, that such terms are also never found in the Madhyam gama.
Table 7: Terms and phrases in the Ekottarik gama unique to the Zhu Fonian corpus
Note that some of the terms listed also appear in Zhu Fonian works outside the reference corpus (i.e. T 309, T 384).
term/phrase Ekottarik gama elsewhere
[X] ……
“What are the X dharmas?
Namely...”
35x T 212 (2x),
T 309 (2x), T 656 (1x)
“in a secluded and quiet spot”
35x T 212 (3x)
“Prince Long Life”36 34x T 212 (8x)
36 Cf. : , T 161, 3: 386a8; : ,
T 152, 3: 5a20–21. All instances in the Ekottarik gama are found in 24.8 = MN 31, MN 128. The Indic name underlying this Chinese is uncertain. In the Theravada Vinaya, the name is D gh vu; for further details, see An layo, Comparative Study, 2: 732 n. 230.
37 38
“in the city of R jag ha, in the Kara a-Venuvana”
33x T 212 (4x),
T 1464 (2x)
“The monks received a teaching from the Buddha”
27x T 1428 (1x)
( ) 39
“will never [again] take up [residence in] a womb, and knows it
as it is”
22x T 212 (3x)
“At that time, the World-Honoured One then spoke this th ”
19x T 212 (5x), T 309, T 656
40 “[when]
the body is destroyed and life comes to an end, is reborn in a good
destiny or a heaven”
19x T 212 (3x)
“the culprit replied...” 13x (1x), T 212 (8x)
Vajj 12x (1x)
( / ) ( / ) 41
“[don] the three monastic robes, and leave the home [to practice] the
path”
12x (5x), T 212 (1x)
[e.g. discard...] “the defiled and perfect the undefiled mind, liberation of mind, liberation
through wisdom”
12x (3x), T 212 (1x)
X Y ……
“X knew what Y was thinking, and then...”
10x (1x), T 212 (2x), T 309 (2x), T 1428 (4x)
37 Sometimes also (T 212, T 1464); once T 212.
38 Mizuno, “Kan’yaku...yakushutsu,” 89 (see n. 4).
39 Lin, “Xiancun,” 131, 133.
40 Lin, “Xiancun,” 131, 133, 135.
41 The specific variant is found only in the Ekottarik gama and T 212. is found only in the rgh gama (also
, ).
The (copiously attested) Madhyam gama version of this formula is rather
[ ] .
( )
“[sees] with the heavenly eye, [which is] pure and without taint or
blemish”
6x T 212 (19x)
“then, right where he sat, eliminated all dirt and impurity, and attained
purity of the Dharma eye”
6x T 212 (1x),
T 309 (1x), T 1428 (18x)
“the same is also the case with classes of sentient beings”
5x T 212 (19x),
T 384 (1x)
( / )
“having taught the Dharma, he rose immediately from his seat”
5x (1x), T 1428 (2x), T 1464 (3x)
( ) ( )(
/ )42 “god[s], men of the world, ras, ra-like deities
[and humans/ rama as]”
5x (9x), T 309 (1x), T 1428 (1x)
43 “...impermanent, constituting a mutable dharma”
4x (12x), T 212(1x)
(var. , ) ……
“went to the World-Honoured One...did obeisance at his feet, sat to one side, and related these events
to the World-Honoured One”
3x T 1428 (174x)
44 ……
“said to one another, ‘This rama a...”
2x T 1428 (3x),
T 1464 (32x)
“...should say to that monk”
2x (3x), T 1428 (12x)
“to bear direct witness oneself in the here and now”
1x (13x), T 1428 (1x)
42 alone is in some senses an even stronger marker, with few exceptions outside of the Zhu Fonian corpus: rgh gama (13x), Ekottarik gama (17x), T 212 (2x), T 226 (1x), T 309 (8x), T 384 (1x), T 656 (1x), T 1428 (7x), with scattered instances in T 223, T 224, T 227, T 650, T 657, T 816, and T 1509.
43 Even is unique to Zhu Fonian, and it is also found in T 1464 (1x).
44 is also a strong Zhu Fonian marker, though not unique to him; it never appears in the Madhyam gama.
“to the end of his life not kill, not steal, not engage in sexual misconduct, not cheat, and
not booze”
1x (14x)
“Or there might be a sentient being with a certain type of body and a
certain type of mind”
1x (5x)
Conclusions
The above analyses should demonstrate that in our extant Ekottarik gama, markers characteristic of Zhu Fonian are extremely numerous. None of the above terms appear in the Madhyam gama. In combination with the external evidence, the internal evidence surveyed above provides extremely strong support for Zhu Fonian’s translatorship (or perhaps partial authorship) of the text.
The evidence presented in this paper is only one half of a more complex case against the ascription of T 125 to Sa ghadeva, and in favour of ascription to Zhu Fonian. In our companion study to the present paper, Ven.
An layo Bhikkhu and I showed that in numerous cases, the same underlying meaning in the Indic source text is systematically translated differently into Chinese in the Ekottarik gama and the Madhyam gama.45 In the present study, I have shown that conversely, there exists in the Ekottarik gama an overwhelming number of terms and phrases, together used a total of thousands of times, which are characteristic of the other works most reliably ascribed to Zhu Fonian, but never appear in the Madhyam gama. Furthermore, these characteristic Zhu Fonian markers cover a wide range of types of phraseology, including proper names; transcription terms; technical Buddhist terms;
common nouns, verbs and adjectives; and more functional parts of speech like conjunctions, adverbs and verbs of speech (Tables 1–6 above). In addition, it is possible to discover a number of very strong markers—usually longer phrases, like narrative formulae and recurring characteristic combinations of words—which are found in the Ekottarik gama, and are otherwise unique to the Zhu Fonian corpus (Table 7).
45 Radich and An layo, “Were the Ekottarika- gama.”
The evidence discussed in this paper alone comprises a total of 137 markers, occurring between them more than 6,200 times in the Ekottarik gama, and approximately 15,520 times further in the core Zhu Fonian corpus of D , T212, T656, T1428, and T1464.46 To reiterate, none of these markers occur even once in the Madhyam gama. By contrast, using the same methods, I have only been able to find in the Ekottarik gama around two dozen markers possibly more characteristic of Sa ghadeva’s style, as defined by his benchmark text, the Madhyam gama, occurring between them less than 100 times in the Ekottarik gama.47 Numerical calculations based upon the
46 Counts are necessarily approximate for a combination of reasons: variant readings in various witnesses to the texts; difficulties accounting for patterns featuring non-contiguous strings with some of the search methods employed; and so on.
47 Using the same techniques used to find my other evidence, I was able only to find a few markers occurring in both the Ekottarik gama and the Madhyam gama, but never elsewhere in the Zhu Fonian corpus (the pattern we would expect if the Ekottarik gama were by Sa ghadeva). For example:
(M 154x/E 1x); (M 89x/E 1x); (M 82x/E 7x);
(M 73x/E 2x); (M 67x/E 4x); (M 57x/E 1x); (M 57x/E 3x); (M 54x/E 26x); (M 52x/E 1x); (M 50x/E 1x); …… (M 44x/E 6x);
(M 43x/E 1x); (M 41x/E 12x); (M 40x/E 1x); (M 40x/E 9x); (M 36x/E 1x); (M 35x/E 1x); (M 34x/E 1x); (M 30x/E 3x);
(M 27x/E 2x); (M 27x/E 1x); (M 22x/E 1x);
(M 22x/E 1x); (M 22x/E 1x); (M 21x/E 3x); […] (M 21x/E 2x).
I do not claim that this list is exhaustive, but it does contain all the markers that I could find, matching the search conditions, that are most copious in the Ekottarik gama. The list features only 26 markers, which between them occur only 93 times in the Ekottarik gama; the five most frequent markers alone (
, , …… , , ) occur between them
a total of 60 times, leaving only 33 instances in total of the remaining 21 markers.
This is not a pattern characteristic of a translation style. In fact, as we see above, for some reason, most of these markers occur dozens of times in the Madhyama-
gama, but only once or twice in the Ekottarik gama. Moreover, it is also noticeable that some discourses feature more than one of these markers, and at least some discourses featuring such markers appear anomalous within the Ekottarik gama for other reasons as well (e.g. they feature apparently Mah na elements, or they are “merged” discourses). The problems raised by such discourses, and the presence within them of small quantities of markers anomalous to the Zhu Fonian corpus, are complex, so much so that I cannot even
qualitative methods used here will always be at best suggestive, but these orders of magnitude strongly suggest that the Zhu Fonian markers identified here and in the companion study are vastly more numerous in the Ekottarik gama than markers characteristic of Sa ghadeva. At the very least, this should mean that the burden of proof should now lie with any scholar who wishes to argue, despite the evidence presented here, that the Ekottarik gama was in fact translated by Sa ghadeva.
In conclusion, the overall pattern of evidence suggests overwhelmingly that our extant Ekottarik gama is the work of Zhu Fonian.
Directions for Future Research
The primary focus of this paper has been to investigate the translatorship of the Ekottarik gama only. Since external evidence presents us with two main candidates, Zhu Fonian and Sa ghadeva, as possible translators of this text, the appropriate method was to proceed primarily by comparing the Zhu Fonian corpus with only one point of comparison, namely, the Madhyam gama, which is our main point of reference for Sa ghadeva’s style.
A full and rigorous study of problems of attribution in Zhu Fonian’s entire corpus would require the identification of markers which distinguish Zhu Fonian’s work from a much larger set of points of comparison—ideally, at least all translation works prior to his time for which ascription is secure, and all secure ascriptions among his contemporaries. I intend to undertake such a study of the full corpus in the near future.
However, even though it is restricted to markers that distinguish Zhu Fonian’s style from that of Sa ghadeva, the present study, in combination with the earlier paper by Ven. An layo and myself, is still the most thorough study to date of the distinguishing features of the style of Zhu Fonian. It is worth noting, therefore, that the present study has already suggested in passing that some texts much more regularly feature characteristic Zhu Fonian markers than others. These passing observations provide us with a set of hypotheses that we can further evaluate in future investigations.
substantiate here my claim that such broader patterns exist, let alone treat them satisfactorily. I hope to address this problem in future work. For the present, I can only suggest that the occurrence of such anomalous markers in small quantities in the Ekottarik gama is not a problem unique to apparent Sa ghadeva markers, nor does it seem to me sufficient reason to doubt the significance of the overwhelmingly more copious evidence studied above for the question of translatorship.
Very few of the Zhu Fonian markers studied here appear in T 226, T 1485, and T 388. This observation provides further circumstantial support for suggestions in prior scholarship that these works are probably not by Zhu Fonian (and that T 226 and T 1485 are therefore incorrectly ascribed in the present Taish canon).
By contrast, relatively many of our markers appear in T 309, circumstantially supporting Nattier’s conclusion that although it was composed in China, the text is nonetheless by Zhu Fonian.48 The same is true of T 384 and T 385, which, as we saw above, scholars have also suggested may have been composed in China.
The Jñ naprasth na T 1543 is ascribed in the present Taish canon to Zhu Fonian and Sa ghadeva working together. However, the markers identified in this study do not seem to indicate that both figures had an equal hand in producing the wording of our received text. The text features very few distinctive markers of Sa ghadeva’s style, as represented by the Madhyam gama. By contrast, it does feature several dozen of our Zhu Fonian markers—none of which ever appear in the Madhyam gama. Provisionally, then, it appears likely that Zhu Fonian had a much greater hand in producing the wording of our present Jñ naprasth na than Sa ghadeva.
However, as already mentioned, the evidence presented in this paper (with the exception of the small number of “strong” markers presented in Table 7) is primarily of use in distinguishing Zhu Fonian’s works from the works of Sa ghadeva, and therefore, is at best of only provisional value in distinguishing Zhu Fonian from other translators (or authors). Full corroboration of these hypotheses, therefore, must await future work.
Finally, I must also mention that the findings of the present paper cannot pretend to exhaustively resolve all problems of translatorship/authorship and style within the Ekottarik gama itself.
We have seen above that overwhelming evidence shows that the style of the Ekottarik gama is to be associated far more closely with Zhu Fonian than Sa ghadeva. Moreover, in future work on the Zhu Fonian corpus in general, I will present still more evidence of very strong (i.e. often, for all practical intents and purposes, exclusive) Zhu Fonian markers, which are also found in great quantity in the Ekottarik gama, supporting this ascription even further.
At the same time, however, the Ekottarik gama is clearly a complex, even polyvocal, text. To begin with, An layo has already presented considerable
48 Nattier, “Re-evaluating.”
evidence (including summaries of prior scholarship) showing that it contains material which is unusual, by comparison with other Nik ya/ gama collections: discourses which are “mixed”, “merged” or in Lamotte’s terms,
“composite”49 (that is to say, single discourses which comprise material that is found in more than one discourse in other Nik ya/ gama collections);50 discourses featuring possible Mah na influence;51 discourses which double, with telling differences, discourses found elsewhere within the same Ekottarik gama;52 and discourses containing material which seems late in other transmission traditions (for example, only being found in commentarial literature in P li).53 Of course, many of these features of the text could have been produced during its history outside China, prior to translation.54 At the same time, such features also require us to keep an open mind about possible complexities in the Chinese text proper, as distinct from complexities deriving from the underlying Indic “original(s).”55
In this light, it may be significant that in the process of preparing the present study, it was possible to observe in passing some marginal features of the Ekottarik gama that do not appear to fit with the overall pattern of stylistic evidence surveyed above.
For example, Ekottarik gama 43.2 (which An layo has identified as containing “late” material only paralleled in P li commentaries, and showing possible Mah na influence56) contains the phrase, “From [his = the Buddha’s] mouth, five-coloured light issued forth, and shone throughout a
billion world-systems .”57 The sequence
, which appears at the end of this phrase, is otherwise unknown in any Zhu Fonian text, and indeed, in any text prior to Zhu Fonian;
49 Lamotte, “S tra composite.”
50 See e.g. An layo, “Reflections,” 9–10; An layo, “Two Versions,” 52 n. 145;
An layo, “Discourse Merger.”
51 See e.g. An layo, “Mah na;” An layo, “Two Versions,” 16–17.
52 See e.g. An layo, “Influence,” 7 n. 45; An layo, “Two Versions.”
53 See e.g. An layo, “Influence,” 2, 7; An layo, “Buddha’s Past Life,” esp. 105–
106.
54 Note, however, that An layo, “Two Versions” (see esp. 25–43, 49–50, 56) has argued on stylistic grounds that at least Ekottarik gama 50.4 is probably by a different hand than the remainder of the Ekottarik gama, and was probably added to the collection in China.
55 Cf. An layo, “Influence,” 7–8 and n. 45.
56 An layo, “Mah na,” 14 n. 32.
57 T 125, 2: 758b14–15.
but it appears a number of times in texts soon after 400, including seven instances in Kum raj va’s *Mah prajñ ramitopade a T 1509, and other instances in *Dharmak ema (T 157, T 397), the “S tra of the Wise Man and the Fool” T 202, and in Faxian’s Mah parinirv a- mah tra T 376. Of course, it is possible that these other texts got this locution, directly or indirectly, from the sole Ekottarik gama instance as their ultimate source, but it is also possible that such evidence might betray revision of the text, or interpolation into it, later than the initial translation by the team including Zhu Fonian in 383.
Problematic for different reasons is the archaic transcription
for “Tath gata, Arhat, Sa yaksa buddha,” which appears in Ekottarik gama 8.2, 8.3, 8.4, 8.5, 8.6, 8.7, 8.8, 8.10, 12.3, and 42.3.58 This transcription otherwise looks characteristic of *Lokak ema (with a single variant);59 outside these Ekottarik gama discourses, however, the precise form seen here only ever occurs in Zhu Fonian’s T 384 (which is suspected of being an original composition), and in the Sarv stiv da-vinaya
T 1435 ascribed to *Punyatara and Kum raj va.
As a final brief example, Ekottarik gama 27.5 somewhat famously features Maitreya, and has been identified as a locus of possible Mah na influence.60 Among several stylistically anomalous features exhibited by this discourse, we find (four times) the transcription for ramit , which dates back to *Lokak ema. In texts supposedly by Zhu Fonian, this transcription otherwise only appears in T 226, T 388, and T 1485—i.e. texts which, as we have seen, are almost certainly not by Zhu Fonian; and twice in T 384. (It also, incidentally, never appears in the Madhyam gama.)
It must be emphasized that marginal examples like these cannot undermine the central conclusions of this study. In light of the evidence presented in the body of this paper, our conclusion must still be that at its core, the Ekottarik gama is overwhelmingly more characteristic of Zhu Fonian than the supposed translator whose name it bears in the present Taish Tripi aka, Sa ghadeva. At the same time, such anomalies also show that there is still more to the text than we have been able to fully study here, and such features of the text are an important topic for future research.
58 I am grateful to Ven. An layo for first drawing my attention to the occurrence of these terms in the text. Cf. for the same set of epithets, listed in Table 3 above.
59 In *Lokak ema, .
60 An layo, “Mah na,” 17 and n. 42, citing Harrison, “Ekottarik gama.”
