Introduction

Content

i.1The Dharma Scripture “Transformation of Karma” (short title: Transformation of Karma), which is presented here for the first time in an English translation, takes places in Śrāvastī, in Prince Jeta’s Grove, where the Buddha is staying with a large following of 1,250 monks. The Buddha is visited by a brahmin youth named Śuka, who requests the Buddha to explain how karmic ripening can lead to such a great diversity of beings. In response, the Buddha explains that individual karmic results are the reason for the diversity of beings, and he teaches in detail fifty-one categories of negative and positive consequences together with the specific actions that function as their causes. Next, the Buddha lists ten negative consequences for transgressing each of the five precepts: abstaining from taking life, abstaining from stealing, abstaining from sexual misconduct (here transgressing the vow of celibacy), abstaining from lying, and abstaining from consuming intoxicants. Lastly, the Buddha explains the karmic advantages that result from building and worshiping a caitya of the Tathāgata. Also enumerated are the advantages gained through making offerings and leading a religious life, living a secluded life, becoming a Buddhist mendicant, and developing confidence.

Textual Classification

i.2Transformation of Karma belongs to a group of Buddhist scriptures that scholars of Buddhism have called the Karma­vibhaṅga or Śukasūtra class. The texts in this group deal extensively with the topic of karma and rebirth according to individuals’ actions. Characteristic of the treatment of karma and rebirth in these texts is their detailed, catalog-like listing of specific karmic consequences and their causative actions. It is clear, however, that this is not meant to be understood as determinism. For example, in several places the presence or absence of intention and remorse are mentioned as factors that can determine whether a particular karmic result will be experienced, whether its outcome can be influenced, and whether it can be avoided altogether.

i.3In the Tibetan canon, Transformation of Karma is classified as a sūtra belonging to the Śrāvakayāna, and in the Degé edition of the Kangyur it is listed in the sūtra section under “various sūtras belonging to the Hīnayāna.” The Japanese scholar Kudo Noriyuki and others proposed that the related Karma­vibhaṅga (Toh 338) belonged to the canon of the early Buddhist ordination lineage of the Sāṃmitīyas (or Sammatīyas). The Sāṃmitīyas were a branch of the Vātsīputrīya sect, whose adherents were also known as the “Personalists” (Skt. Pudgalavādin) because they posited the existence of the individual or person (Skt. pudgala) that transmigrates from one existence to the next and is neither identical with nor different from the five aggregates. Whether Transformation of Karma, too, can be considered to have belonged to the canon of the Sāṃmitīyas, however, is difficult to say with certainty without further research. However, the text contains some clues that may point in that direction. The first indication is a curious statement found in 1.­25, which lists as one of the karmic causes that lead to rebirth in the realm of the asuras “conceiving of nonexistence with regard to the self (Skt. ātman).” If our translation of this difficult passage is correct, the view expressed therein may betray the doctrinal position of the Pudgalavādins. A second clue is found in 1.­88, which contains the following statement: “one will not become lost in the intermediate state between death and rebirth.” We know that not all the early Buddhist sects accepted the concept of an intermediate state between death and rebirth. Those that did, according to André Bareau, included the Sarvāstivādins (Vaibhāṣikas), the Sammatīyas, the Pūrvaśailas, the later Mahīśāsakas, and the Dārṣṭāntikas.

i.4The Karma­vibhaṅga group of texts appears to have been particularly popular in Central Asia and in China, where the continued interest in it is shown by the existence in the Chinese Tripitaka of no fewer than six different translations spanning a period of eight centuries from the late third or early fourth century ᴄᴇ to the eleventh. Four of these are listed in the Tōhoku catalog of the Tibetan canon of the Degé edition as corresponding to the text translated here (particularly the Fen bie shan e bao ying jing 分別善惡報應經, Taishō no. 81). However, none of them seems to be an exact match.

The Title

i.5Sanskrit versions of the titles of translations of Indic texts in the Kangyur and Tengyur are traditionally given in transliteration at the beginning of the texts‍—but for Toh 339, the Sanskrit title differs from its Tibetan title. The Tibetan phrase rnam par ’gyur ba would suggest an original Sanskrit vikāraḥ, vikṛtiḥ, vikriyā, or vipariṇata, etc., all of which belong to the English semantic field of change. The text’s transliterated Sanskrit title, on the other hand, is given as Karma­vibhaṅga­nāma­dharma­grantha, for which in Tibetan one would rather expect something like rnam par ’byed pa. The word vibhaṅga occurs in Buddhist literature as the title of certain works, for instance, as part of the title of one of the three major parts of the Pāli Vinaya (Suttavibhaṅga); as the title of the second, and oldest, book of the Pāli Abhidhamma (Vibhaṅga); as the Division of the Expositions (Vibhaṅgavagga), which is a section of the Majjhima Nikāya of the Pāli Canon that contains the Pāli “relatives” of the Karma­vibhaṅga-group of texts (the Cūḷa- and Mahākammavibhaṅga-suttas, MN 135, 136); or as part of the title of a short work on the twelve links of dependent origination (Skt. pratītya­samutpāda) composed in Sanskrit (the Pratītya­samutpāda­vibhaṅga­nirdeśa­sūtra) that was found at the site of the great Indian university Nālandā, inscribed on bricks. In those contexts, Sanskrit vibhaṅga means “explanation,” in the sense of offering a more detailed explanation or elucidation‍—occasionally illustrated by examples, stories, or word-for-word explanations‍—of individual items in a list, such as the list of Vinaya rules or, in this case, a list of categories of karmically relevant actions and results. The derived meanings “exposition” and “commentary” can be regarded as generic terms applied to texts exhibiting certain stylistic features like the ones mentioned.

i.6The final part of the Tibetan title, chos kyi gzhung (Skt. Dharmagrantha), seems to be very rare as part of the title of canonical works. A search for the expression chos kyi gzhung in the online database Resources for Kanjur and Tenjur Studies of the Tibetan Manuscripts Project of the University of Vienna yielded only the text translated here. Instead of translating the Tibetan rnam par ’gyur ba literally as “transformation,” “change,” or “ripening,” we have accepted the transliterated Sanskrit given in the text itself as the original title of the work, and have translated it consistent with the common Buddhist usage of the Sanskrit word vibhaṅga when applied to texts that exhibit certain characteristics. This choice furthermore highlights its kinship with the longer Karma­vibhaṅga (Toh 338).

i.7Mention must be made here, however, of the fact that the Kangyur editions not belonging to the Tshalpa (tshal pa) line preserve variant titles: the versions in the Stok Palace manuscript, the Shey Palace manuscript, and the Shelkar manuscript Kangyurs‍—all belonging to the Thempangma (them spangs ma) line‍—give the Sanskrit name of the text as Karma­vibhaṅga­nāma­dharma­paryāya.

i.8Furthermore, the Shelkar Kangyur records as the Tibetan title las kyis rnam par ’gyur ba zhes bya ba’i chos kyi gzhung: “The Dharma Scripture Called Transformation through Karma.” If this is not simply a scribal error (the Stok and Shey Kangyurs read las kyi rnam par ’gyur ba), this title may support Sanskrit pariṇāma (“transformation,” “change,” “ripening”) as part of a possible original Sanskrit title of the work.

i.9Yet another variant of the title should be noted. The Denkarma (Tib. ldan/lhan dkar ma), one of the three imperial catalogs of Tibetan translations made during the early transmission of the Dharma in Tibet during the eighth to ninth centuries, gives the title as las kyi rnam par ’gyur ba bstan pa’i gzhung.

On Karma

i.10Apart from its occurrence in the title of the sūtra, we have largely avoided the lexicalized English word karma as a rendition of the Tibetan las. Instead, we translate las as “action(s).” We feel that contemporary common usage of the loanword karma is often fuzzy or unclear in that it seemingly blends together the senses “the totality of a person’s actions and conduct during successive incarnations,” a resultant state as “fate or destiny,” which is perceived to be causally influenced by the totality of a person’s actions, the “law or principle through which such influence is believed to operate,” and the woolly but popular sense of “a distinctive aura, atmosphere, or feeling.” We do, however, use the lexicalized adjective karmic (as, for example, in karmic result , or karmic cause and effect) in the sense of “relating to one’s action(s)” in the translation. We have furthermore used the expression karma and rebirth in this introduction knowing full well that these terms are equally difficult to define abstractly, but with the understanding that in Indic and Buddhist contexts the terms karma and rebirth are amalgamated in the term saṃsāra .

i.11As for the “mechanics” of karmic cause and effect, the Buddha explains it in Transformation of Karma, 1.­6:

“Therefore, again, the multiple varieties of sentient beings, due to their being bad, good, or middling, are karmically connected with many kinds of actions, many kinds of sufferings, and many kinds of views, in accordance with what I have taught as the black and white fruits of karmic ripening. Young brahmin, it is like this: Through carrying out black actions, sentient beings are reborn in the unfortunate rebirth destinies; they are reborn as hell beings, animals, ghosts (that is, those who have gone to the afterlife), and asuras. Through carrying out white actions, beings are reborn as devas or as humans.”

i.13Two things are noteworthy in this passage regarding Buddhist doctrine. First is the work’s assumption of six realms of rebirth (Skt. gati). Some early Buddhist schools accepted only five realms, including the asuras in the god realm. By contrast, here the unfortunate realms or destinations of rebirth (Skt. durgati) are constituted by hell beings, animals, ghosts, and asuras; the fortunate realms or destinations of rebirth (Skt. sugati) include the devas and humans, omitting the asuras.

i.14The second observation concerns the classification of actions into “black” and “white.” One frequently encounters this classification in Abhidharma-style works where, in an extension of the basic color metaphor, karmically efficacious action is divided into four kinds: (1) wholesome (Skt. kuśala, i.e., morally or ethically good), (2) unwholesome (Skt. akuśala, i.e., morally or ethically bad), (3) mixed (i.e., actions and results possessing both morally or ethically good and bad aspects and corresponding resultant experiences), and (4) neutral (Skt. avyākṛta or āniñjya, “undetermined” or “immovable, immobile,” e.g., the actions of an enlightened being, an arhat, etc.). For example, the Saṅgītisutta of the Dīgha Nikāya (DN 33) explains as follows: “Four kinds of kamma: There is (a) black kamma with black result, (b) bright [white] kamma with bright result, (c) black-and-bright kamma with black-and-bright result, (d) kamma that is neither black nor bright, with neither black nor bright result, leading to the destruction of kamma.”

i.15The same explanation, but much more technical and elaborate in its interpretive details, is found in Vasubandhu’s Abhidharma­kośabhāṣya: “Action is of four types, white, black, etc. The sūtra teaches that action is of four types: black, of black retribution; white, of white retribution; black-white, of black-white retribution; neither black nor white, without black or white retribution; and that which destroys the other actions.”

i.16The third type, of black-white action, is further explained: “Good action of the sphere of Kāmadhātu, being mixed with the bad, is black-white; its retribution is mixed, so it is thus black-white. This definition is to be understood as applying not to the nature of the action itself, but to the ‘series’ or the person; in one and the same mental series, good action is mixed with bad action. There is no action which is black-white, nor any retribution which is black-white, which would be a contradiction.”

i.17The third kind of action as black-white is not specifically mentioned in Transformation of Karma’s explanations of karmic categories, but the possibility seems to be implied. For instance, when a person first experiences happiness, which later changes to the experience of unhappiness due to certain black-white actions as causes (1.­44–1.­48). Actions of the fourth kind, which are actions after enlightenment and which have no effect in saṃsāra, are more difficult to identify with certainty in Transformation of Karma. The explanations in 1.­55, 1.­57, and 1.­58 seem to state that arhats are still subject to karmic punishment and reward. But they do not make any statement about the karmic efficacy (or inefficacy) of acts carried out by arhats, which was a point of contention between, for instance, the Sarvāstivādins and the Theravādins.

Tibetan and Sanskrit Versions

i.18Three different recensions or versions of the Karma­vibhaṅga are preserved in different Kangyur editions, which Kudo calls “Tib-1” (Toh 338), “Tib-2” (Toh 339, the text translated here), and “Tib-3,” respectively. The content of the third version not included in the Degé Kangyur (“Tib-3”) but found in other mixed-lineage Kangyurs and in the Themphangma Kangyurs has been explored by Walter Simon. Tib-3 is distinguished by the inclusion, among other things, of an introductory story that is found only in one Sanskrit manuscript, and not in Toh 338 or Toh 339. This introduction to the discourse proper tells the story of the conversion of Śuka, the son of the brahmin Taudeya, who after his death was reborn as a dog in his son’s house; he is identified by the Buddha on one of his visits to Śuka’s house during his rounds to collect alms. Śuka does not believe the Buddha’s assertion that the dog is his deceased father and at first becomes angry; but he is later convinced by the Buddha’s demonstration of the identity of the dog as his late father. After being the addressee of the Buddha’s teachings on karmic cause and effect on several occasions, Śuka eventually becomes a lay follower.

i.19Walter Simon noted three distinguishing features of Tib-3 in comparison to Toh 338 and 339: (1) “the preservation … of the ‘introductory tale,’ ” (2) the absence of the “illustrative stories,” and (3) “the inclusion of … a table of contents.” These features, he wrote, “point to an earlier Sanskrit original.” Transformation of Karma, too, contains a a table of contents, and it lacks stories illustrating the various karmic categories. But, more importantly, it also lacks the introductory tale characteristic of the (Mahā-)­Karma­vibhaṅga (Lévi 1932, 21–29) and its Tibetan version in the manuscript Kangyur of the British Museum.

i.20Like the Tib-3 version of the Karma­vibhaṅga, the summary of the fifty-one karmic categories at the beginning of Transformation of Karma (1.­8–1.­63) does not include the items contained in 1.­64–1.­68 (a list of ten negative consequences each of the actions of killing, stealing, lying, unchastity, and drinking alcohol) or 1.­69–1.­90 (the advantages gained by making various offerings at caityas ). The Tib-3 version contains the advantages of venerating and making offerings at caityas , but it does not contain the list of five negative actions and their consequences included in 1.­64–1.­68 of Transformation of Karma. It is reasonable to argue, therefore, that 1.­64–1.­68 of Transformation of Karma, and perhaps a portion of 1.­69–1.­90, are later additions to the text.

i.21As we have already mentioned, no complete Sanskrit version of Transformation of Karma is extant. In his study of the Khotanese fragments of the Karma­vibhaṅga, Mauro Maggi remarks that Transformation of Karma “corresponds” to “S2” (that is, a manuscript fragment of the Śukasūtra from Eastern Turkestan, in the British Library, Oriental and India Office Collections, London, which was edited by A. F. R. Hoernle in 1916 and reproduced in Lévi 1932, 235–36). However, Sylvain Lévi had already observed that although the two texts‍—the Central Asian Sanskrit fragment and the Tibetan text of Transformation of Karma Toh 339 (i.e., Lévi’s “T2”)‍—exhibit a family likeness, they are different. A careful comparison of Lévi’s edition of the extant Sanskrit portions of the Śukasūtra fragment from Central Asia (paragraphs VIII–XII) with the corresponding paragraphs in the Tibetan text of Transformation of Karma confirms Lévi’s observation.

i.22To sum up, no exactly corresponding Sanskrit source text of the Tibetan translation of Transformation of Karma seems to be extant. Nevertheless, due to their similarity, the Sanskrit fragment of the Central Asian recension allowed us to throw some light on certain obscure terms in the Tibetan text.

i.23Regrettably, we possess no information about who translated the text or when. The Tibetan text in the Degé edition (as well as in the other editions) has no colophon, which usually provides this information, and according to the Tōhoku catalog, the identity of the translator(s) is unknown. It is not possible at this point to establish the relationship between these different extant versions of the text, and Transformation of Karma thus seems to be the sole textual witness of yet another recension or a different version in the Karma­vibhaṅga or Śukasūtra group.

i.24Nevertheless, we have frequently consulted (in a rather unsystematic way and as indicated in the footnotes) the extant Sanskrit witnesses: the two Sanskrit manuscripts in the National Archives of Nepal, Kathmandu, published by Lévi in 1932 and Kudo in 2004; the two-folio fragment of a different recension, published by Fukita in 1990 and Kudo in 2004; a Kuchean Sanskrit fragment of the Śukasūtra, published by Lévi in 1932; and the Pāli (MN 135, 136), Tibetan (Toh 338), and Khotanese (in English translation by Maggi in 1995) versions of the related (Mahā-)Karma­vibhaṅga (Lévi 1932) to aid our understanding of this often obscure Tibetan text.

The Tibetan Translation

i.25The Tibetan text of Toh 339 is notable for its idiosyncrasies. It contains several nonstandard expressions and names of Buddhist concepts that are worthy of mention, such as ngan pa, ring du ’khyams pa, and mya ngan bsring(s). Furthermore, it contains an idiosyncratic rendering of the four formless meditative absorptions (Skt. ārūpya­samāpatti) or, as they are called in Transformation of Karma, “the four states of imperturbability.” These observations lead us to believe that the translation of Transformation of Karma predates the third Tibetan language revision or language reform of 814 ᴄᴇ, as it was codified in the Mahāvyutpatti and the sgra sbyor bam po gnyis pa. On the other hand, a more detailed analysis of the wording describing the four states of imperturbability, as well as the frequent use of an epithet for the Buddha of yang dag par gshegs pa, instead of the more standard de bzhin gshegs pa, has led us to also consider the possibility that Toh 339 may have been translated from the Chinese. A comparison with several of the Chinese versions of our text by Mr. Yi Ding, however, could not confirm this hypothesis.

i.26What we can safely say, though, is that Transformation of Karma represents a translation that does not strictly adhere to the reformed, standardized language that was prescribed by decree for the translations from Indic originals in the ninth-century Tibetan manuals for translators, the Mahāvyutpatti and the sgra sbyor bam po gnyis pa.

i.27It is generally assumed that the Tibetan translation activity of the first dissemination of the Dharma began around the middle of the eighth century ᴄᴇ and continued until the demise of the Tibetan empire, that is, by 840 ᴄᴇ. Since the identity of Transformation of Karma with the las kyi rnam par ’gyur ba bstan pa’i gzhung (no. [282]) in the Denkarma (Tib. lhan kar ma) catalog seems unambiguous, these two dates (i.e., 763 ᴄᴇ and 840 ᴄᴇ) are likely the earliest and latest dates, respectively, of the translation of Transformation of Karma into Tibetan. If we additionally consider the fact that Toh 339 contains pre-reform terms and expressions, we may be able to push the date for the latest date closer to the beginning of the ninth century. In the absence of a colophon or other pertinent information about the text, however, it is impossible to be more precise.

The English Translation

i.28This translation into English was made directly from the Tibetan text of the Degé edition with the help of its parallel versions. However, mention must be made of the pioneering French translation by Léon Feer in 1883. This French Sanskritist and Orientalist, who was the first to translate Kangyur materials into a European language, gave us the first translation of Transformation of Karma in a European language in his Fragments: Extraits du Kandjour. However, Feer did not have at his disposal the wealth of material on the Karma­vibhaṅga group of texts that we have today, and he was thus unable to consult similar or related versions of the text to assist him in his efforts to make sense of obscure and difficult passages in the Tibetan text. Our translation refers frequently to Feer’s translation in the notes, especially where Feer’s interpretation of the Tibetan deviates from ours. However, our translation should not be considered definitive or final. Rather, it is an attempt to improve upon Feer’s where possible.

i.29Since the text itself is not very long, and the Tibetan text in the Kangyur does not have any chapter subdivisions, we have likewise refrained from inserting any such divisions. We have, however, referenced Feer’s sensible general outline of the text in the notes to the translation. Feer divided the text into three main parts, each with subdivisions and individually numbered paragraphs, and in the notes to our translation we have noted his general division of the text.

i.30In producing the translation, we have used the Degé edition (siglum D) as the basis. We have also perused the Comparative Edition of the Kangyur (Tib. dpe bsdur ma) (A), together with its list of variant readings. Where the Degé text seemed ambiguous or was not readily comprehensible, we have directly consulted two versions of the text belonging to the Thempangma line, namely, the Stok Palace manuscript Kangyur (S) and the Shey Palace manuscript Kangyur from Ladakh (Z). For variant readings that affect the meaning of a passage, we have done our best to provide an alternative English translation in the notes. Where we were unable to provide an alternative translation for a variant (e.g., because we didn’t understand it) that nevertheless seemed significant enough to be recorded, the variant is followed by a question mark. Except for the two witnesses (S and Z) belonging to the Thempangma line, all other variant readings (of the Yongle, Lithang, Kangxi, Choné, Narthang, Urga, and Lhasa editions of the Kangyur; see Abbreviations for their sigla) are taken from list of variants in the Comparative Edition edition.