Saṃdhinirmocana Sūtra

I applied various methods and followed a series of steps during the process of translating the Saṃdhi­nirmocana­sūtra from the Tibetan:

I first collated all the available Tibetan editions of Saṃdhi­nirmocana: Bd, C, D, Dunhuang (Hakayama 1984–87), Do, F, H, L, N, S, Kǫ, U, VD, X, and Z, as well as the various extant Sanskrit fragments found in Buescher (2007), Levi (1925), Matsuda (1995, 2013), Nagao (1964), and Tucci (1971). For the Chinese, we used Xuanzang’s translation.¹⁰ I then produced a critical edition of the text prologue (nidāna) to get a sense of the textual variations across major available editions of the Tshalpa (tshal pa), Thempangma (them spangs ma), mixed Kangyurs, and independent Kangyur groups. In addition, Dr. Kojirō Katō (Tokyo University), who is editing the Saṃdhi­nirmocana­sūtra, kindly sent me his critical edition of the seventh chapter.¹¹ His work has proved invaluable to confirming the findings of my own work on the prologue. The editions belonging to the Thempangma differ significantly from those included in the Tshalpa line of transmission. As an independent Kangyur close to the Thempangma line, the Phukdrak (phug brag) edition offers very interesting readings on the level of syntax and lexicography compared to the editions of the Tshalpa group. It also diverges from the Thempangma witnesses in many locations. In the absence of colophons mentioning the translators’ and editors’ names across the available editions, it remains difficult to understand the history of these witnesses from the perspective of the underlying translation and editing process. As a consequence of its palatable variant readings compared to the Tshalpa and Thempangma editions, I used the Phukdrak witness quite extensively while translating the Degé edition, as well as the Stok edition and the Degé version of the Viniścaya­saṃ­grahaṇī, to examine more thoroughly difficult passages. The available Sanskrit fragments were on occasion also useful to determine the Sanskrit equivalent of a Tibetan technical term. They, however, did not reveal major variations from the Tibetan texts. I referred to the Dunhuang recension sporadically, as Schmithausen warned us not to follow it blindly.¹² I also referred to Xuanzang’s translation regarding a few difficult passages of the text. This translation is similar to those of the Tshalpa group and might have been carried out on the basis of a Sanskrit manuscript similar to the one (or those) used for the translation upon which the Tibetan Tshalpa editions are based.

The Saṃdhi­nirmocana­sūtra is a major text of Indian Buddhism that has early on attracted the attention of Buddhist Studies scholars. It has been translated into French, English, and German (partially). Lamotte (1935) provided a critical edition of Saṃdhi­nirmocana and a translation from Tibetan (N) and Chinese (Taishō 676) into French. He also attempted to reconstruct or identify technical terms in Sanskrit, which have for the most part been confirmed by subsequent finds of Sanskrit fragments.¹³ Lamotte’s work is a major resource for the study of Saṃdhi­nirmocana. It has been until now the standard edition and translation of this text on account of its accuracy and methodological academic approach. As a side note, I followed Lamotte’s segmentation of the text into paragraphs. His French translation is generally reliable, although some technical passages can be significantly improved, particularly in the case of chapter 8 on meditation and chapter 10 on the result of the path. Frauwallner (1969) gives a partial translation of Saṃdhi­nirmocana (i.e., chapters 6 and 7) from Tibetan into German. As one would expect, Frauwallner’s academic translation of these two chapters aims at accuracy over readability. Kawasaki 1976 is also a partial translation of chapter 8 into English (§6.1–9). This translation does not improve Lamotte’s. Powers’ (1995) translation from Tibetan (D) into English, in spite of its merit, could be widely improved upon in terms of methodology, accuracy, and readability.¹⁴ Brunnhölzl 2018 offers a partial translation of chapter 7, as well as a few key passages from chapter 5.

Translations by Cleary (1999) and Keenan (2000) are from the Chinese (Taishō 676) into English. I used Keenan’s work to get a sense of the Chinese text while translating the Degé edition but only referred to Cleary occasionally. Keenan’s work seems to me more accurate than Cleary’s, although the latter was useful for unraveling difficult passages, since his style is more free and primarily intends to communicate the meaning of the text. Cornu (2005) has provided a translation of the text from Tibetan (D) into French that mainly follows Powers 1995, a somewhat regrettable fact since Lamotte’s (1935) is more accurate. Schmithausen’s (2014) work contains numerous difficult passages of Saṃdhi­nirmocana translated from various Tibetan and Chinese editions into English. It is an invaluable resource for the study of Saṃdhi­nirmocana. In addition, it offers useful Sanskrit reconstructions of important technical terms. Together with Lamotte’s translation, it has been a constant companion while translating the text.

The Saṃdhi­nirmocana­sūtra is part of a larger network of texts, both in the Kangyur and the Tengyur:

1. The nidāna of Saṃdhi­nirmocana almost exactly matches those of the Buddha­bhūmi­sūtra (D275) and the Tathāgata­guṇa­jñānā­cintyaviṣayāva­tāra­nirdeśa­sūtra (D185). The Buddha­bhūmi­sūtra is a very short text that was also translated into Chinese by Xuanzang in 646 (see Keenan 1980, p. 336ff.). Textual parallelisms of this kind are useful to double checking some passages or gathering more background information about the source text.

2. As mentioned above, Saṃdhi­nirmocana is also found in extenso in the Viniścaya­saṃ­grahaṇī of the Yogācāra­bhūmi and is therefore part of a tradition of texts sharing common ideas.¹⁵ This point should be kept in mind while translating, particularly when one has to evaluate the potential impact of terminological choices from the perspective of a more philosophical approach to the text, which, in the case of Saṃdhi­nirmocana, should be a major concern. For example, one should pay attention to the fact that interpreting “representation-only” (vijñaptimātra) as a strong form of idealism essentializing mind could be misleading from the perspective of a cultural translation of the worldview propounded in Saṃdhi­nirmocana since mind, just as much as the external object, is explicitly declared to be empty of any own-being, essence, or intrinsic nature (see chapter 8 on the three kinds of niḥsva­bhāvatā) in this text.¹⁶

3. Another important point is the presence of the aforementioned five commentaries on Saṃdhi­nirmocana found in the Tengyur (D).¹⁷ I occasionally referred to these works while finalizing the final draft of the translation. However, I first focused on the available editions of Saṃdhi­nirmocana itself as I did not want to be influenced by the interpretations of later authors. Instead, I attempted to go through all possible logically meaningful readings according to the Tibetan and Sanskrit sources without any preconceptions resulting from my reading of later commentarial traditions.

4. In the same vein, one should note that Saṃdhi­nirmocana has played a major role in Tibetan hermeneutical debates. For centuries, it has been considered a central scripture referred to extensively in the writings of Tibet’s great luminaries, such as Jé Tsongkhapa (rje tsong kha pa, 1357–1419) or Jamgön Mipham Gyatso (’jam mgon mi pham rgya mtsho, 1846–1912). While it would certainly be fascinating to study the impact of Saṃdhi­nirmocana in the context of Tibetan Madhyamaka,¹⁸ I chose not to take into account Indian or Tibetan commentaries or exegeses of the sūtra in order to focus on the source text itself.

I proceeded to search all articles and monographs referring to Saṃdhi­nirmocana I could find at the very beginning of this translation project. In this quest for relevant academic research, I benefited from the excellent bibliography found in Delhey 2013 regarding research done on the Viniścaya­saṃ­grahaṇī of the Yogācāra­bhūmi, which I expanded with a list of complementary reference works (see the bibliography). Among the existing academic literature on Saṃdhi­nirmocana, Schmithausen 2014 stands out and, unsurprisingly, proved to be a major resource for this translation project.

1. The translation of the title of the text became the object of several discussions among scholars regarding the meaning of the Sanskrit words saṃdhi and nirmocana as a consequence of Lamotte’s first complete translation of the text.¹⁹ Among the various available options, I opted for simplicity and translated the Sanskrit Saṃdhi­nirmocana­sūtra with “The Sūtra Unraveling the Intent,” which I believe renders accurately the meaning and structure of the text. Various interlocutors indeed ask the Buddha repeatedly to explain difficult points in order to clarify the purpose of his seemingly contradictory or complex doctrines on the nature of reality.

2. Regarding the content of the sūtra itself, I proceeded to organize secondary sources by chapter and referenced this research in the notes accompanying my translation.²⁰ The last chapter of Saṃdhi­nirmocana includes a very technical passage on valid cognition (pramāṇa) whose definitions predate Dignāga’s system of logic. Translating Trisong Detsen’s *Samyagvāk­pramāṇoddhṛta­sūtra, which is a commentary on the teaching on the four principles of reason (rigs pa bzhi), would help us better understand pre-Dignāgean Buddhist logic.²¹