The term Mile 彌勒 appears to be a simple problem and has actually been treated as such Von Gabain, for example, maintains that it was Serindian culture that influenced the development of the concept of 彌勒 Mile Metrak, as the Chinese transliteration 彌勒 Metrak did not derive from Sanskrit but from mittra in Khotansaka and metrak in Tocharian, hence the term miroku in Japanese.[i]  It seems to me that von Gabain has oversimplified this problem, which is, in fact, considerably more complicated. Involved here are at least three questions: (1) Did Metrak derive directly from Maitreya? (2) Was the term Metrak formed independently? And finally, (3) Why did 彌勒 Metrak and not 梅呾利耶 Maitreya appear in the earliest Chinese versions of Buddhist texts? There are the problems that I shall deal with in this article.



Ⅰ. Did Metrak Derive Directly from Maitreya?

There is undoubtedly some connection between Maitreya in Sanskrit and Metteya in Pāli on the one hand, and Metrak in Tocharian on the other, as all three words refer to the “Future Buddha”. That all three are related to Maitri in Sanskrit is also without doubt. Morphologically speaking, however, the Tocharian word is somewhat different from either the Sanskrit or Pāli word. How is this to be interpreted? With regard to Indian Buddhist history, the most rational interpretation seems to be that Metrak derived from Maitreya and Metteya, for Tocharian came into being later than Sanskrit and Pāli.

There are indeed scholars who have arrived at this conclusion, one of whom is Franz Bernhard. In his “Gāndhārī and the Buddhist Mission in Central Asia,”  Bernhard confirms the part that Gāndhārī played in the spread of Buddhism to Central Asia and China.[ii] His argument is certainly very convincing in that Gāndhārī was a northwestern dialect of ancient India, in which The Dhammapada discovered in Xinjiang, China, was written. Moreover, in geographical terms, it is reasonable to argue that Gāndhārī served as a medium of crucial importance in the Buddhist mission in Central Asia.

In passing, Bernhard touched on Metrak and 彌勒 Mile. After quoting von Gabain who traces the origin of the Chinese transliteration of 彌勒 Mile to Tocharian, he argues that “this idea, however does not solve the problem but simply transfers it, as in Tocharian a phonetic change from ‘y’ to ‘k’ is quite unknown.”  Bernhard suggests that Tocharian should be replaced by Gāndhārī, where such suffix doublets as

-aya-aga

-eya-ega

are often seen,“g” being a voiced fricative. In Gāndhārī are found “metraya/metreya: metraga,” which sometimes become “metrae/metre” when the ending consonants are deleted. It is to these forms that the East Tocharian term Metrak and the Chinese terms 彌勒 Mile are traced. Furthermore, other related word forms in Khotan-saka, Parthian, Uighur, Mongolian, Kalmyk and Manchu can either directly or indirectly be traced back to those forms in Gāndhārī.

In West Tocharian (Kuchean) are found two distinct forms. Maitreye and Maitrak, which, in Bernhard’s view, register the paired-off suffixes and two distinct stages of linguistic borrowing as well.

Here Bernhard cites “On the Earliest Chinese Transliterations of the Name of Buddha,” an essay of mine written some forty years ago.[iii] In it I argue that while 佛陀 Fuotuo, can be traced to “Buddha” etymologically, 佛 Fuo is not of Sanskrit but of Tocharian origin. 佛 Fuo is not a shortening of 佛陀 as is popularly believed; rather the latter is a lengthening of the former. Bernhard admits that 佛 did not derive from “Buddha”, yet he reconstructs a socalled Old Tocharian form *but, asserting that this was where the Chinese transliteration 佛 came from. He concludes that 佛 can testify to the role of Tocharian as a medium if 彌勒 fails to do so.

Now let’s come back to the ending “ega.”“g” is voiced while “k” is voiceless. Granted that voiced fricatives were not used as endings in Tocharian, the phonological transition from “g” to “k” cannot have come about smoothly. Therefore, the problem needs further inquiry.

It seems that something might be worked out if we switch to the relations between “-eya” and “-eka.” The interchangeability of “ya” and “ka” is evident in Old Indian Dialects. According to Pischel,[iv] Amg. daga＝udaka＝udaya, posakiya＝upavasathika, tumbavīniya＝tumbhavīnaka, etc. Lüders[v] cites several instances of “-ika” in place of “-iya”: ekavāciya and ekavācika, rathiyā and rathikā, ghātikā and ghātiyā, jāgariyā and jāgarikā, etc, As K.R. Norman[vi] points out, in the Upkt of the Asokan Inscriptions, the phonological transitions such as -k/-g-＞-y- and -y-＞-k- do occur. For instance, janiyo, nom. plur. of janī, varies to janiko in some particular inscriptions.

Even as we take up Gāndhārī, we find instances of -y-＞-k- and -k-＞-y-. John Brough[vii] renders mrga ﹥muya and kāka ﹥kaya as instances of the transition from -g-/-k- to -y-, and udaya ﹥udaka as one of the reverse transition -y-﹥-k-. In a note,[viii] Brough discusses the problem of Maitreya[ix]﹥Metrak. After enumerating Metrak in Agnean (Tocharian A or East Tocharian), Maitràk in Kuchean (Tocharian B or West Tocharian) and m’ytr’k in Sogdian, he observes that a variant of m’ytr’k  also existed in Sogdian i.e. m’ytr’y, a quite ambiguous form (refer to Gershevitch, Grammar of Manichean Sagdian,§960). He points out that the Chinese form 彌勒 Mile may have come from Metraka. These might be purely literary borrowings mechanically transcribed; but it is of course possible that even within Gāndhārī the written forms may have subsequently induced pronunciations with [k], if only as a reading-style adopted by the less literate monks for use in the recitation of sacred texts.” The meaning of the cited passage is clear in spite of its obscure wording. Brough demonstrates in great detail the subtle involved relations between -y- and -k-. What is meant here is that sometimes monks were induced to pronounce -y- as -k-. In short, Brough’s interpretation of Metrak is entirely different from Bernhard’s.

Different as their opinions seem to be, both Brough and Bernhard hold that the term Metrak derived directly from Maitreya. As to Gāndhārī, Brough noted its important role in the Buddhist transmission in Central Asia and China even before Bernhard. According to Brough, a great number of early Chinese transliterations of Buddhist terminology were based on Gāndhārī as a source language.[x] However, it does not follow that 彌勒 Mile must likewise have come from Gāndhārī. Brough and Bernhard have good grounds for taking Maitreya as the origin of Metrak, and they are well supported in their arguments. Nevertheless, there may be other and perhaps more reasonable approaches to the problem of 彌勒 Mike.

Ⅱ. Was Metrak Formed Independently?

“Independent formation” refers to the process in which Metrak was formed in accordance with the Tocharian rules of word-formation, independent of Gāndhārī or any other intermediary language.

As a matter of fact, in their Tocharische Grammatik,[xi] Sieg, Siegling and Schulze devote a whole chapter to the discussion of some related problems. According to the three authors, the ending “-ik-” functions in the following ways.

1. It may turn an abstract noun into a person with the characteristic represented by that noun. For example, kākmart (prestige)＋-ik﹥ kākmärtik (B kamartike), meaning a person having prestige, or a prestigious person; kāruṃ (mercy)﹥kārunik (cf. Sanskrit: “kārunika”) meaning a merciful person; kritāṃ (the meaning of which is not clear)＋-ik﹥kritānik, i.e. a person doing kritāṃ; spaktāṃ (A＝B, service, devotion) ﹥spaktānik (B spaktanīke), i.e., a devotee, a donator.

It may turn an adjective into a noun. For example, A āsāṃ/B asāṃ (respectable)＋-ik﹥āsānik, meaning an arhat, a person worthy of being worshipped.

3. Where loan words are concerned, the situation becomes more complicated. The Sanskrit word niraya＋pāla (ka) is varied to ñarepālik in Tocharian, meaning a guard of hell; pindapāta(Skt.) ﹥piṃnwātik (Toch.) from which comes the Chinese term 分衛 (alms-begging) Fenwei. The function of -ik in these cases is not clear. Another example: ṛṣi (Skt, seer, sage)＞riṣak (B rṣāke).

居士(Jushi) is gṛhin, gṛhastha in Sanskrit and kātäke (B kattāke) in Tocharian, yet the latter did not emerge from Sanskrit, but very probably from Iranian languages. Awesta kad means a house, and in New Persian kat-x-udā refers to layman（居士）. Therefore, it is the ending “ä” that causes kātäk to mean a person living at home.

As to Metrak, the three authors observe without further explanation that “Maitreya” frequently occurred in the form of Metrak. In Tocharian B there existed both forms, Maitrāk and Maitreye. Then, in a note, the authors make it clear that they have been informed by Andreas of the fact that Metrak also occurred in Middle Persian manuscript remains found in Turfan. Here I venture to offer a different interpretation of Metrak or Maitrāk. This form, in my opinion, is not to be traced back to the Sanskrit form Maitreya; rather it came directly from another Sanskrit form Maitri meaning “kindness”. Maitrī＋-k﹥Metrak/Maitrāk, i.e., a kind person—the original for the earliest free translation 慈氏 Cishi. In my opinion this explanation is well grounded and reasonable. Thus, my view is that Metrak was formed independently and that it is not necessary to derive it from “Maitreya.”[xii]



Ⅲ. Why Was It That 彌勒 and Not 梅呾利耶
Appeared in the Earliest Chinese
Versions of Buddhist Scripture?

Now that the problem of Maitreya and Metrak is clarified, we can set about examining how these terms occurred in the Chinese Buddhist texts, as this is a question of pivotal importance to the studies of the Buddhist transmission from India to Serindia and China, and of the History of Chinese Buddhism.

Maitreya and Metrak occurred in the Chinese versions of Buddhist texts in three different forms.

1. 彌勒―the phonetic transliteration of Metrak

彌勒 appears many hundreds of times in the Chinese Tripiṭaka and it is therefore not possible nor necessary to enumerate all of them. What is of significance to my inquiry is: (1) when did the earliest translations appear; (2)in which sūtras they occurred; and (3) who translated them. By “the earliest” I mean the periods of the Later Han and the Three Kingdoms. The occurrences of 彌勒 from the Two Jins, the Northern and Southern Dynasties through the Sui and Tang down to the Song and Ming Dynasties will not be listed here, as they are irrelevant to my discussion. Recorded here are the earliest texts that contain彌勒 in order of their appearance in 大正新修大藏經 (Taishōsinshūdaizōkyō, hereafter abridged to T.)

六度集經 (Liu Du Ji Jing, ṣ aṭpāramitā-saṅgraha-sūtra, trans. Kang Seng Hui of the Kingdom of Wu 吳康僧會譯) Vol. 1︰彼國王者彌勒是That king was Metrak. T. 3, 3b;

大方便佛報恩經 (Da Fang Bian Fo Bao En Jing) in the Later Han Catalogue, Sūtra of the Great Upāya [Mahopāya] by which Buddha recompenses the Favour [of his Parents]. Trans, anonymous失譯人名在後漢錄Vol.1︰彌勒菩薩“Metrak Bodhisattva.”T. 3, 124b;

佛說月明菩薩經 (Fo Shuo Yue Ming Pu Sa Jing, Buddhabhāṣita-Candraprabhābodhisattva-sūtra, trans. Zhi Qian of Wu吳支謙譯)︰爾時智力王者，今彌勒菩薩是。“The king jñānabala of that time is now the Metrak Bodhisattva,”T. 3, 411c;

佛說興起行經 (Fo Shuo Xing Qi Jing, Buddhabhāṣita-nidānacaryā-sūtra, trans. Kang Meng Xiang of the Later Han Dynasty後漢康孟祥譯)，Vol.2:爾時病比丘彌勒者，則今彌勒菩薩是。“The sick Bhikṣu Metrak is now Metrak Bodhisattva.”T.4 172b;

撰集百緣經 (Zhuan Ji Bai Yuan Jing, Śatanidāna—saṅgraha—sūtra, trans. Zhi Qing of Wu吳支謙譯)，Vol. 10︰我及彌勒俱為菩薩。“Both Metrak and I were Bodhisattva.”T. 4, 253c;

雜譬喻經 (Za Pi Yu Jing, Saṃyuktāvadāna-sūtra, trans. Zhi Lou Jia Chen of Later Han後漢支婁迦讖譯)︰欲睹彌勒佛時三會二百八十幾人得真人時.“I would like to see the occasion when at the 3 assemblies of Metrak Buddha 280 koṭis of people attain Arhathood.”T. 4, 499b;

舊雜譬喻經 (Jiu Za Pi Yu Jing, An Old Saṃyuktāvadāna-sūtra, trans. Kang Seng Hui of Wu吳康僧會譯)，Vol.1︰彌勒作佛時，當得應真度脫。“When Metrak turns Buddha, he will attain Arhathood and get deliverance.”T. 4, 512b;

道行般若經 （Dao Xing Ban Ruo Jing, Dasasāhasrikāprajñāpāramitā-sūtra, trans. Zhi Lou Jia Chen of Later Han後漢支婁迦讖譯），Vol.1︰彌勒菩薩。“Metrak Bodhisattva.”T. 8, 425c;

Ibid. Vol. 3︰彌勒菩薩。“Metrak Bodhisattva.”T. 8, 438a;

Ibid. Vol. 5︰彌勒菩薩。“Metrak Bodhisattva.”T. 8, 451b;

大明度經（Da Ming Du Jing, Daśasāhasrikā-prajñāpāramitā-sūtra, trans. Zhi Qian of Wu吳支謙譯），Vol. 4:彌勒闓士.“Metrak Bodhisattva.”T. 8, 496c;

佛說八吉祥神咒經 (Fo Shuo Ba Ji Xiang Shen Xhou Jing, Aṣṭabuddhaka, trans. Zhi Qian of Wu吳支謙譯)︰菩薩千人皆彌勒等.“One thousand Bodhisattvas and all equal to Metrak.”T. 14, 72b;

Ibid.彌勒菩薩“Metrak Bodhisattva.”T. 14, 73a;

佛說維摩詰經 (Fo Shuo Wei Mo Jie Jing, Vimalakīrtinirdeśa, trans. Zhi Qian of Wu吳支謙譯),Vol.1︰於是佛告彌勒菩薩。“Then the Buddha told Metrak Bodhisattva.…” T. 14, 523c;

佛說長者子制經 (Fo shuo Zhang Zhe Zi Zhi Jing, Śreṣṭhi-pūtrajeta-sūtra. trans. An Shi Gao of Later Han後漢安世高譯)︰我持慈心皆付彌勒佛.“I maintain kindness in my mind totally devoted to Metrak Buddha.”T. 14, 801b;

佛說純真陀羅所問如來三昧經 (Fo Shuo Chun Zhen Tuo Lou Sho Wen Ru Lai San Mei Jing , Druma-kinnara-paripṛcchā-ratnatathāgata-samādhi-sūtra, trans. Zhi Lou Jia Chen of Later Han後漢支婁迦讖譯)，Vol. 1︰復有菩薩，名曰彌勒。“Further there was a Bodhisattva named Metrak.”T. 15, 349a;

Ibid, Vol.2︰佛語彌勒。“The Buddha told Metrak.”T. 15, 367a;

The above citations bring to light the basic situation regarding the occurrences of 彌勒 Metrak in the Chinese Buddhist texts during the Later Han and the Three Kingdoms. It was not a long period, but there nevertheless appeared a considerable number of translators (康僧會，支謙，康孟詳，支婁迦讖，安世高，etc.) who, moreover, came from quite different places such as月氏，Indoscythia，康居Sogdiana, and 安息 Parthia. It is worth noticing that whoever the translator was and wherever he came from translators unanimously employed the term Metrak 彌勒。 The phonetic transliteration of Metrak 彌勒 was an accepted term for the Future Buddha during that period.

2. 梅呾利耶 or the like—the phonetic transliterations of Maitreya 梅呾利耶unrelated to Metrak, is the phonetic transliteration of the Sanskrit term Maitreya.  Its occurrences in the Chinese Buddhist texts are as follows:

阿毗達磨大毗婆沙論 (A Pi Da Mo Da Pi Po Sha Lun, Abhidharma-mahāvibhāsā-śāstra, trans. Xuan Zang of the Tang Dynasty，唐玄奘譯)，Vol. 177；二名梅呾儷藥。“The second one was called Maitreya.”T. 27, 890b;

阿毗達磨順正理論 (A Pi Da Mo Shun Zheng Li Lun, Abhidharma-nyāyānusāra-śāstra, trans. Xuan Zang Tang唐玄奘譯)，Vol. 44︰一名梅怛儷 ( 麗 ) 藥。“The first one was called Maitreya.”T. 29 591c;

妙法蓮華經玄贊 ( Miao Fa Lian Hua Jing Xuan Zan, Profound Eulogy of addharmapundarīka-sūtra, written by Kui Ji of Tang唐窺基撰 ) end of Vol. 10:梵云梅呾利末那故。“Because the Sanskrit word is Maitrimanas.”T. 34, 849c;

三彌勒經疏 ( San Mi le Jing Shu, Commentary on Three Maitreya-sūtras, written by Jing Xing of Korea新羅憬興撰 )︰今正梵音云梅呾利耶.“The right wording in Sanskrit now is Maitreya.”T. 38, 305b;

說無垢稱經疏 ( Shou Wu Gou Chen Jing Shu, Commentary on Vimalakīrtinirdeśa, written by Kui Ji of Tang唐窺基撰 )，end of Vol. 3:梵云梅呾利耶此云慈.“The Sanskrit word is Maitreya, which means here in Chinese Kind-hearted.”T.38, 1048b;

Ibid Vol. 4:梵云梅呾利耶，翻為慈氏。古云彌帝麗，或云彌勒，皆訛略也。“The Sanskrit word is Maitreya, which is translated into a ‘Kind-hearted person’.” In ancient times it was called Maitri or Metrak which are both wrong and abridged.”T.38 1057b;

金光明最勝王經疏 ( Jin Guang Ming Zui Sheng Wang Jing Shu, Commentary on Suvarṇaprabhāsa-sūtra, written by Hui Zhao慧詔撰，trans. Yi Jing of Tang唐義淨譯 )， Vol. 1︰慈氏，梵音梅呾麗曳，此云慈氏。或云那，即慈姓中生，或本性行慈。或云曳尼，即女聲中。“Kind-hearted person, in Sanskrit wording Maitreya, here in China he is called a Kind-hearted person, or Maitreyāna, which means Born in the Maitreya Caste, or Kind by Nature; it is also called Maitreyāni, being feminine gender”T. 39, 188b;

首楞嚴義疏注經 ( Shou Leng Yan Yi Shu Zhu Jing, Commentary on Sūraṅgama, ed. Zi Jun of Song宋子璿集 )，Vol. 5,Ⅱ︰具云梅咀 ( wrongly written for “呾”) 利曳那，此云慈氏“The full complete name is Maitreyāna, which means here in China Kind-hearted person.”T. 39, 902a;

俱舍論記 ( Ju She Lun Ji, Description of Abhidharmakośa-śāstra, compiled by Pu Guang of Tang唐普光述 )，Vol. 18︰二名梅怛儷藥，夾注︰梅怛，此云慈；儷藥，此云氏。菩薩在慈姓中生，從姓立名，故名慈氏。舊云彌勒，訛也。“The second one was called Maitreya. Interlinear notes: Mait means here in Chinese kindness; reya means here a person. The Bodhisattva was born in the caste Kindness (Maitrī), hence he was called the Kind-hearted. The old transliteration Metrak is wrong.”T. 41, 282b;

俱舍論疏 ( Ju She Lun Shu, Commentary on Abhidharma-kośa-śāstra, written by Fa Bao of Tang唐法寶撰 )，Vol. 18:二名梅怛儷藥.“The second one was called Maitreya.”T. 41, 682a;

成唯識論述記 ( Cheng Wei Shi Lun Shu Ji, Description of Vidyāmātrasiddhi-śāstra, written by Kui Ji of Tang唐窺基撰 )，Vol. 4:梵言梅呾利耶，此翻言慈氏。梅呾羅曳尼，此聲轉之異。婆羅門十八姓中，慈為一姓。氏謂氏族，皆當皆生此種姓家，故以為號。“The Sanskrit word Maitreya is translated into ‘Kind-hearted’ here in China. Maitreyāni is a different morphological form. The caste Maitreya is one of the 18 Brahman castes. “氏” means clan. All who are born in this caste should be called Maitreya.”T. 43, 352b;

法苑珠林 ( Fa Yuan Zhu Lin, Pearl Forest in Dharma Garden written by Dao Shi of Tang唐道世撰 )，Vol. 29︰又大寺中有刻木梅呾麗耶 ( 舊云彌勒 ) 菩薩像.“Further, in the great temple there is a statue of Maitreya (formerly Metrak) Bodhisattva carved in wood. T. 53, 498b;

一切經音義 ( Yi Qie Jing Yin Yi, Commentary on All Sūtras, written by Hui Lin of Tang唐慧琳撰 )，Vol. 27:彌勒，梅怛利曳，此云慈氏，慈為本姓，或以心行為姓也·“Metrak Maitreya means a ‘Kind hearted person’ here in Chinese. Or the clan name Kind-hearted is deduced from [kind] mentality.”T. 54, 482c;

翻譯名義集 ( Fan Yi Ming Yi Ji, Mahāvyutpatti, ed. Fa Yun of Song 宋法雲編 )，Vol.1︰彌勒，《西域記》云梅哩 ( wrongly written for “呾”) 麗耶，唐云慈氏，即姓也。舊曰彌勒，訛也。什曰︰‘姓也。阿逸多，字也。南天竺婆羅門子。’《淨名疏》云︰‘有言從姓立名。今謂非姓，恐是名也。何者？彌 勒此翻慈氏。過去為王，名曇摩流支，慈育國人。自爾至今。常名慈氏。始阿逸多，此云無能勝。有言阿逸多是名。既不親見正文，未可定執。觀《下生經》云，時 修梵摩，即與子立字，名曰彌勒。“Metrak, in Records of the Western World, is called Maitreya 〔哩wronglywritten for “哩”〕. In the Tang language it means kind hearted person which becomes a clan name. Formerly it was called Mi-Le [Metrak], which is wrong. Kumārajiva said it was his clan name. Ajita was his given name. He was a son of a Brahman in South India. The Vimalakîrti Comentary states: Somebody says, his personal name comes from his clan family caste name. Now in my opinion it is probably a personal name. Why? Metrak is translated here into Chinese “Kind hearted Person.” He was in a former birth a king called Dharmaruci, who took care of his subjects kind heartedly. Hence he has from that time until now always been called the “Kind-hearted One.” From the beginning(?) Ajita has always mean invincible. It was said that, Ajita was his personal name. Because I haven’t seen the text personally, I cannot say anything definite. We see in 下生經 (Maittreya-Vyākaraṇa): at that time Subrahman gave his son the name Metrak. T. 54, 1058b;

These examples, in my view lead us the following conclusions:

(a)  梅呾利耶Meitanliya and other phonetic transliterations of Maitreya appeared       as late as the Tang Dynasty.

(b)  These phonetic transliterations occurred predominantly in annotations, commentaries and dictionaries, and only rarely in the Sūtras proper.

(c)  Diversity exists within a context of roughly identical Chinese characters employed in the transliterations. Master Xuan Zang himself was not consistent.  He used 梅怛儷藥 in the two instances cited above while in Records of the Western World of the Tang Dynasty ( 大唐西域記 )，Vol. 7, the transliteration appears as 梅呾麗耶。

(d)  Some of the explications by Chinese monks are simply nonsensical.梅呾，此云慈；儷藥，此云氏。 That Mait means here in Chinese Kind-hearted, reya means here in Chinese for instances is wrong because in Sanskrit “Maitreya” cannot in any circumstances be divided into “Mait” and “reya”. Such is also true of 梅呾利曳那 Meitanliyana as there is no such word as Maitreyann in Sanskrit. As to或云那，即慈姓中生 (huo yun na, ji ci xing zhong sheng) the Sanskrit equivalent should be “Maitreyaja”; however, “ja” means “生”(being born) while “na” ( 那 ) has nothing to do with “生”.

(e)  The relationship between 彌勒 Metrak and 阿逸多 Ajita has not been clarified.

(f)  In Records of the Western World of the Tang Dynasty ( 大唐西域記 ) as well as in the above citations 彌勒 Mile was held to be a wrong and an abridged translation ( 訛略也 Elüeye ), which fact hints that the origin of 彌勒 had long been inaccessible to Chinese monks, including the Great Master Xuan Zang.

3. 慈氏Cishi—the free translation

In the Chinese Buddhist texts, the term 慈氏 Cishi is used with remarkable frequency comparable to that of 彌勒 Mile/metrak or only slightly lower. Examples given below will be restricted to the earliest periods, that is, the Later Han and the Three Kingdoms, so that the use of 慈氏 Cishi can be studied along with that of 彌勒 Metrak. Citations after those periods are omitted, as 慈氏 on a parallel with 彌勒 appear in great abundance in the translated Buddhist texts of every dynasty.

大明度經 ( Da Ming Du Jing, Daśasāhasrikā-prajñāpāramitā-sūtra, trans. Zhi Qian Wu吳支謙譯 ), Vol. 2:爾時慈氏闓士語善業︰有闓士大士代歡喜最尊分德法。“At that time, Metrak Bodhisattva told Subhūti: Bodhisattva Mahāsattva rejoiced in the Prajñāpāramitā.”T. 8, 486a;

Now let’s compare this passage with those in different versions of the same sūtra.

道行般若經︰摩訶般若波羅蜜漚惒拘舍 羅勸助品第四 ( Dai Xing Ban Ruo Jing, Daśasāhasrikā-prajñāpāramitā, trans. Zhi Lou Jia Chen Later Han後漢支婁迦讖譯 )，§4:爾時彌勒菩薩謂須菩提︰若有菩薩摩訶薩勸助為福。“At that time, Metrak Bodhisattva told Subhūti: if a Bodhisattva Mahāsattva causes the people to take upon themselves the religious duty to accumulate puṇya [Merit]…” T.8, 438a;

摩訶般若鈔經︰善權品第四 ( Mo He Ban Ruo Chao Jing, Daśasāhasrikāprajñāparamitā Shan Quan Pin, §4 trans. Dharmapriya and Zhu Fo-Nian符秦曇摩蜱共竺佛念譯 )，Vol. 3:爾時彌勒菩薩謂須菩提︰若有菩薩摩訶薩勸助為福。“At that time, Metrak Bodhisattva told Subhūti: if a Bodhisattva Mahāsattva causes the people to take upon themselves the religious duty to accumulate punya [Merit]…”T. 8, 519c.

It can be seen clearly that what was translated as 慈氏闓士 and 闓士大士 in the first version was rendered into 彌勒菩薩 and 菩薩摩訶薩 in the last two, the former being free translations and the latter, phonetic transliterations. The term 闓士 kaishi occurs frequently in Zhi Qian’s version and sometimes varies to 開士 kaishi or even 闡士 chanshi in other sūtras.[xiii]

Other examples follow:

大明度經 ( Da Ming Du Jing, Daśasāhasrikā-prajñāpāramitā-sūtra, trans. Zhi Qian Wu吳支謙譯 ) Vol. 2:慈氏闓士作無上正真平等覺時。

“When Maitreya Bodhisattva turns to Anuttarasamyaksambuddha.” T. 8,492c;

Ibid.若於兜術天上從慈氏闓士問 慧。“If he asks Maitreya Bodhisattva about Prajñā in Tuṣita heaven.…” T. 8, 492c;

法鏡經 ( Fa Jing Jing, Ugra-paripṛcchā, trans. An Xuan Later Han 後漢安玄譯 )︰便使慈氏開士及一切行淨開士聽。“Bhagavān: You just let Metrak Bodhisattva and all Sila-pāramitā-practising Bodhisattvas hear it.”T. 12, 22c;

佛說無量壽經 ( Fo Shuo Wu Liang Shou Jing, Aparimitāyus-sūtra, trans. Kang Seng Kai of Cao Wei,曹魏康僧鎧譯，Vol.1︰慈氏菩薩。“Metrak Bodhisattva,” T. 12,265c;

Ibid. Vol.2︰慈氏菩薩 (Metrak Bodhisattva), T. 12, 278a. It is noticeable that慈氏菩薩becomes彌勒 in b and c on the same page of Vol.2 of that sūtra.

佛說出無量門微密持經 ( Fo Shuo Chu Wu Men Wei Mi Chi Jing, Buddhadhāsita-amitamukhaguhyadhara-sūtra, trans. Zhi Qian of Wu 吳支謙譯 )︰慈氏菩薩“Metrak Bodhisattva,”T. 19, 680b.

Here again, I would like to point to the fact that parallel versions of this sūtra have different renderings of the same term.慈氏菩薩Cishi Pusa is found in some versions, like Zhi Qian’s, and 彌勒菩薩 Metrak Pusa, in others. Here are two such examples:

佛說出生無量門持經 ( Fo Shuo Chu Sheng Wu Linag Men Chi Jing, Buddhadhāṣita-jātāmita-mukhabhara-sūtra, trans. Fo Tuo Ba Tuo Luo East Jin Buddhabhadra東晉佛陀跋陀羅 )︰彌勒菩薩。“Metrak Bodhisattva,”T. 19, 682b;

阿難陀目佉尼呵離陀經 ( A Nan Tuo Mu Qu Ni He Li Tuo Jing, Anantamukha-sādhaka-dhāranī, trans. Qiu Na Ba Tuo Luo Liu Song, Gunabhadra,劉宋求那跋陀羅譯 )︰慈氏菩薩。“Metrak Bodhisattva,”T. 19, 685a.

With this, I end citations which furnish us with a general idea of the occurrence of 慈氏 Cishi in the Chinese versions of Buddhist texts during the Later Han Dynasty and the Three Kingdoms.

From the study of the occurrences of the three terms 彌勒 Metrak 梅呾利耶Meitanliya and 慈氏 Cishi we can inductively arrive at the following points:

(a)  彌勒 Metrak and 慈氏 Cishi appeared simultaneously in the earliest periods, i.e. the Later Han and the Three Kingdoms, while 梅呾利耶 Meitanliya and the like appeared later.

(b)  In the earliest periods, the translator could freely use either 彌勒 Metrak or 慈氏 Cishi even within the same sūtra. Translators treated phonetic transliteration and free translation in an indiscriminate way.

(c)  From the Later Han and the Three Kingdoms through to the Ming Dynasty, 彌勒 Metrak(Mile) and 慈氏 Cishi were used as parallel and equivalent to each other.

Now that a great many examples have been given, I can finally furnish a definite answer to the question put forward in the title of this part: “Why was it 彌勒 and not 梅呾利耶 that appeared in the earliest Chinese versions of the Buddhist Scripture?” Actually those versions were not based on Sanskrit or Pāli texts. Among their sources there may have been a small amount of Gāndhārī texts, but Ancient Serindian languages, including those of Xinjiang. And most probably, Tocharian, constituted the major part. In one of my articles written about thirty years ago,[xiv] I said: “This fact shows us that China and India did not have much direct contact with each other in the early phase of their cultural exchange. It was the Tocharian speaking people that served as a link between the two countries.” I based that conclusion on an analysis of two terms: 恆河 Henghe (the Ganges) and 須彌 ( 山 ) (Sumeru), to which 彌勒 Metrak can be added now. My theory is that it was the term Metrak and not Maitreya that appeared before the earliest translators, to whom the meaning of Metrak was associated with Maitri meaning “kindness,” the ending -ak only turning the abstract noun into “a kind person,” which, in turn, was rendered into 慈氏 Cishi. Things were indeed quite obvious.[xv]

Finally, I want to touch on why Metrak was transcribed 彌勒. Almost no one has raised such a question perhaps because no scholars have ever thought of it as aproblem. However, close scrutiny does pose a difficulty. I.e.‚ -rek was quite correctly transcribed as 勒, since 勒 Lek is pronounced in the entering tone ending in -k. Mi 彌 on the other hand, is not an entering tone ending in -t, and so 彌 is not an appropriate transcription of Met.

An interpretation has been offered by H.W. Bailey[xvi] who applies the rule “tr＞dr＞1” to the phenomenon “tr＞l” in Metrak. He cites several examples, among which are the Sanskrit word trayastrimsa transcribed as 忉利 Taoli; the Sanskrit term “kṣudra＞kṣulla＞culla”; the phonetic transcription of 剎利 from the Sanskrit term “kṣatriya,” and so on. In my opinion, Bailey’s thesis is quite satisfactory and can at least serve as a justifiable interpretation.



　　 NOTES