The Sūtra on Transmigration Through Existences

The Sūtra on Transmigration Through Existences (Bhava­saṅkrānti­sūtra) is set in the Kalandakanivāpa, at the Veṇuvana near Rājagṛha, where King Śreṇya Bimbisāra of Magadha approaches the Buddha and asks him how, given the doctrine that formations are empty, an action that has long ceased can appear before the mind at the moment of death.1 The Buddha presents the analogy of a beautiful woman in a dream; the sleeping person dreams of cavorting with her, and even after waking foolishly longs to find her. The analogy illustrates how a karmic deed of the distant past, which arose from perception and ensuing afflictive emotions and then ceased, manifests in the mind of someone on the verge of death. The Buddha goes on to explain the transition from the final moment of one life to the first moment of the next, according to the ripening of that karmic deed, without any phenomena actually being transferred. Thus, in this sūtra, the Buddha provides a fundamental explanation for how transmigration between lives occurs in conformity with the view that there is no self‍—as an immutable, incomposite entity‍—that goes from this life to the next. The Buddha concludes with a set of seven verses that do not summarize his prose teachings or even mention transmigration at all, but rather offer a succinct teaching on emptiness, focusing in particular on the two truths and the fictitious nature of all nominal designations.

It is notable that this sūtra can be divided into two distinct parts‍—the prose and the verse sections‍—which are thematically quite different. The second has little relation to the title of the sūtra and seems to be more a deep reading of the preceding prose than the kind of verse summary commonly found in other Mahāyāna sūtras. This divide is one key to navigating the sūtra’s complex history, which is explained below. In essence, this early Mahāyāna sūtra holds an important place in the tradition as one of the earliest statements on the two truths2 and was used by both the Madhyamaka and early Yogācāra schools as a scriptural authority on the ultimate truth.

The Bhava­saṅkrānti­sūtra was translated and edited by the Indian scholars Jinamitra and Dānaśīla and the Tibetan translator and editor Yeshé Dé, who were active from the eight–ninth centuries ᴄᴇ. The Denkarma and Phangthangma imperial catalogs, dated to the early ninth century, both include it in their lists of translated sūtras.3 It is also listed in the Mahāvyutpatti.4

There was no known Sanskrit witness of this sūtra until recently, when a manuscript containing twenty texts, all of them sūtras, was found in the Potala Palace in Lhasa. Bhikṣuṇī Vinītā published a critical edition and English translation of this collection in the series Sanskrit Texts from the Autonomous Region (2010). Unfortunately, due to the inaccessibility of the manuscript collection and because it was missing a final colophon, its origin and date are currently unknown.5 There seems to be a thematic connection among the twenty sūtras. Vinītā gives the example of moral discipline (śīla) as a recurrent theme running through the manuscript,6 and we can likewise note the recurrence of themes of karmic cause and effect and the hierarchy of merit. Interestingly, this sūtra is quoted among others, including several from the Potala manuscript, by Kawa Paltsek (ska ba dpal brtsegs) in a text contained in the Tengyur called the *Pravacana­ratnākhyāna­śākya­vaṃśāvalī.7 Here we can identify the same recurrent themes among the quotations.

Other fragments from the Sanskrit text have been found extant in quotation in other works, with a great number of variations.8 The Pitā­putra­samāgama­sūtra (The Sūtra of the Meeting of Father and Son) contains a passage that closely parallels the prose passage of the Bhava­saṅkrānti­sūtra in both the Tibetan and Sanskrit. The extant Sanskrit is found in quotation in Śāntideva’s Śikṣāsamuccaya.9 In addition to being quoted extensively among treatises, the verse section of the Bhava­saṅkrānti­sūtra has some passages in common with or closely resembling verses found in other sūtras and treatises. There are three similarly titled treatises found in the Tengyur, all attributed to Nāgārjuna, called the Bhavasañcara (Toh 2277), the Bhavasaṅkrānti (Toh 3840), and the Bhava­saṅkrānti­parikathā (Toh 4162), and for the second of these there is a word-by-word commentary called the Bhava­saṅkrānti­tīka (Toh 3841), attributed to a “Paṇḍita Maitreyanātha.”10 All three of these treatises contain within them verses that loosely match the verse section of the Bhava­saṅkrānti­sūtra. Ostensibly these treatises are themselves a commentary on the sūtra; however, the relationships are ambiguous, and some scholars have suggested the possibility that it was the treatises, in some form, that had a later influence on the verse section of the sūtra.11 Since the whole set of the canonical quotations and parallel passages to the verse section is complex and extensive, a comprehensive list of these instances is found in the appendix.

There are three translations of the Bhava­saṅkrānti­sūtra found in the Chinese Tripiṭaka: one (Taishō 575) translated by Bodhiruci (菩提流支, sixth century ᴄᴇ), another (Taishō 576) translated by Buddhaśānta (佛陀扇多, sixth century), and a third (Taishō 577) translated by Yijing (義淨, seventh to eighth century).

There are several other Western-language translations of this sūtra available. In 1936, Giuliana Stramigioli translated both the Tibetan and Chinese (Taishō 557) versions, separately, into Italian. The Tibetan translated was derived from a thirteenth- or fourteenth-century manuscript found in Tholing Monastery.12 N. Aiyaswami Sastri published a translation in 1931 from the Narthang Kangyur and another in 1938 from the Tibetan and all three Chinese translations. Fernando Tola and Carmen Dragonetti translated the sūtra into Spanish based on Stramigioli’s edition in 1977, again in 1980 based on Sastri’s edition, and into English in 1986. As is mentioned above, there is also Vinītā’s 2010 English translation based the Sanskrit manuscript from the Potala. Most recently of all, Peter Skilling has included a fine translation of the sūtra and some helpful notes on it in his 2021 collection, Questioning the Buddha: A Selection of Twenty-Five Sutras.13

Noriaki Hakamaya, in his detailed analysis of Bhava­saṅkrānti­sūtra, theorizes that the prose section of the Bhava­saṅkrānti­sūtra and its parallel in the Pitā­putra­samāgama­sūtra originated as an explanation on the topic of transmigration that later developed under the influence of the Prajñā­pāramitā tradition, at which time elements of the emptiness doctrine would have been introduced. He further suggests that the verse section may have been an addition of the Yogācāra school.14

Hakamaya’s thesis is supported by several contextual points in the sūtra. To begin with, the prose section presents an explanation of rebirth as resulting from the causal force of an action (karma) without the actual transmigration of any self or permanent phenomenon. This accords with the standard Buddhist nominalist view, in which what seems to be a coherent “self” is merely a collection of factors known as the five aggregates, which include consciousness, and thus rebirth occurs through causality and without a “self” transferring from this life to the next. In general, this explanation of rebirth would be accepted by any Buddhist school. One exception would be those schools described by their critics as Pudgalavādins, or “proponents of a person,” who assert that a person (pudgala) transmigrates from one life to the next. However, two such schools, the Vātsiputrīya and Saṃmitīya, make this claim while maintaining the Buddhist tenet of no-self (anātman): this “person” is defined as neither permanent nor impermanent and is neither the same nor different from the five aggregates.15

Interestingly, the Tibetan, Sanskrit, and Chinese versions of the sūtra unanimously state that “no phenomenon whatsoever transmigrates from this world to another world,” with one notable exception in Bodhiruci’s Chinese translation (Taishō 575), which states in the same passage, “there is one phenomenon that transmigrates from this life into the future life.”16 Intriguingly, the Bodhiruci version does not indicate what exactly this one phenomenon is or how it functions in the process of transmigration. Discussing this passage, N. Aiyaswami Sastri speculates that this version, as the earliest translation, may indicate that the original narrative of the sūtra originated from schools, such as the Vātsiputrīya and Saṃmitīya, that assert the transmigration of the “person” or some similar phenomenon. In this case, the title of the sūtra, Transmigration Through Existences, would indeed accord with those that say transmigration occurs.17 However, our research has not unearthed any other evidence that traces the sūtra specifically to these schools, and all the later editions clearly do not reflect this view.

Regardless of this disparity, the prose section of the sūtra provides a standard early Buddhist explanation of rebirth. As in the analogy of the dream, even something we know so well to be illusory can have consequences: the dreamer remembers the dream woman or even goes out searching for her.18 In the same way, although nothing passes from one life to another, the effects of the actions of one life can shape the next.

The prose section of the Bhava­saṅkrānti­sūtra is concordant with the Sautrāntika view that consciousness, which itself seems to be a continuity, consists only of distinct momentary consciousnesses arising and ceasing. Thus, it is said in this sūtra that when one reaches the last moment of consciousness of one life, in the next moment the first consciousness of the next life arises in turn, and these are connected only by the potential of the past karma that shapes each moment as it arises.

The Sautrāntikas maintained that karma, too, is momentary, and thus an action leads to an evolving stream of germinal potentialities (bīja or “seed”) leading to the fruition. The difference in the Bhava­saṅkrānti­sūtra and other Mahāyāna sūtras is that all the elements of the transmigration‍—consciousness, karma, and so forth‍—are said not to arise, remain, or cease; they are devoid of essential nature, so any seeming functioning of karma is an illusion.

A particularly interesting feature of the prose is that it makes this doctrinal assumption that the karma of a single action arises in the mind at the moment of death and is instrumental in determining a being’s next state of existence. While the Buddha accepts this principle implicit in King Bimbisāra’s question, he does not explain the reasoning behind it, which could be interpreted in different ways according to various Buddhist views. It is worth noting that this sūtra does not implicate an intermediate state between lives and seems to imply that the first moment of consciousness of the next life arises immediately following the cessation of the final moment of consciousness in the previous one.19 In light of this, without forcing a specific interpretation of how this causality functions, we can say that the arising of the karma in the mind in the final moment of death provides an explanation for how the continuum of consciousness transmigrating from one existence to the next occurs in lieu of a self or any perpetuating phenomena to be carried through that transmigration. This explanation also maintains the ethical principle of the law of karma, indispensable to Buddhism, that one’s ethical conduct determines one’s next existence and that pleasure and suffering arise from one’s past actions.

It is also up for interpretation exactly how the analogy of the dream should be understood. The Buddha gives this analogy to explain the principle of how the karma appears at the moment of death, stating that the woman in the dream is analogous to that karma, and that her arising in the mind of the man waking from the dream is to be understood as analogous to the karma appearing before the mind at the moment of death. Certainly, he is not implying that the phenomena of karma itself appears, since it has explicitly long since passed, but its potentiality or causal force is present somehow at that crucial moment of death. Perhaps the analogy also emphasizes that the karma itself is nonexistent, like the woman in a dream. But on the whole, the precise implications of this analogy might be explicated differently according to the theory of karma and transmigration held by any particular Buddhist school.

In the Buddha’s final remarks of the prose section we find some passages that align the narrative closely with the Prajñāpāramitā teachings on how phenomena lack any essential nature and on emptiness, in particular at 1.­14, where the phenomena including final consciousness, transmigration, karma, first consciousness, and birth are declared to be empty of themselves.

Moving on to the verses, the first four pertain to the Buddhist nominalist position that all phenomena are only nominally existent (prajñaptisat), having no independent substantial existence (dravyasat), a position that was used by Madhyamaka thinkers to support their school’s interpretation of emptiness.20

The final three verses present a śūnyavāda interpretation of the two truths. The relative truth refers to conventional appearances, as in verse five when “one who observes correctly” says “The eye sees forms.” From the perspective of the ultimate truth there is only the nature of emptiness, all phenomena existing dependently, or, according to the statement in verse six, “seeing comes from a conjunction.”21 Finally, verse seven points out the “supreme truth” that “The eye does not see form / And the mind does not know phenomena,” meaning that the conceptual notion of the two truths is itself provisional.

Although the verse section presents a Madhyamaka position, and despite having parallel treatises attributed to Nāgārjuna, Hakamaya intriguingly suggests the possibility that these verses were an addition of the Yogācāra school. The evidence he provides is that the Bhava­saṅkrānti­sūtra is cited in numerous Yogācāra texts, including the Bodhi­sattva­bhūmi and Sāgara­megha’s commentary on its treatment of the second verse, in which it is interpreted according to the Yogācāra trisvabhāva (three nature) theory,22 and the fact that the sūtra is cited in dialectical treatises to refute the Yogācāra position as the basis of their scriptural authority.23 Regardless of the question of the sūtra’s history, the verse section is quoted or echoed in many treatises as a definitive statement on the two truths and the nominalist position of phenomena.

All of this is merely an overview, food for further research, and is intended to draw attention to some facets of the fascinating history of this brief sūtra.

We have based our translation primarily on the Tibetan version in the Degé Kangyur, but we consulted the Sanskrit and versions in other Kangyurs in the case of questionable terms or passages, in order to establish the most plausible and accurate readings of the text. The citations of the Sanskrit we provide in the notes are from Vinītā’s emendations of the handwritten Potala manuscript. Instances where our translation diverges from the Degé have been noted, and any significant differences found in the various versions of the sūtra are recorded and explained in the notes. In general, all these versions, along with the Sanskrit and Chinese sources, are unanimous in terms of their general structure and meaning (except for the phrase found in Bodhiruci’s translation, Taishō 575, mentioned above). However, there are many minute variations to be found among the various versions, especially in comparison to the parallel passages found in the Pitāputra­samāgama­sūtra and quotations of both the Bhava­saṅkrānti­sūtra and Pitāputra­samāgama­sūtra found in many other texts. These numerous variations are both the likely result and evidence of the sūtra’s complex development.

Since the minor variations found in comparing the different versions are particularly numerous, we have chosen to provide annotations only for differences that change the meaning in a significant way or that we otherwise deemed to be interesting or noteworthy.24 To note all variant readings would require the preparation of a diplomatic edition of the multiple texts, which lies beyond the scope of this translation. As mentioned above, a comprehensive list of all the Kangyur and Tengyur quotations and congruent passages relating to the verse section of the Bhava­saṅkrānti­sūtra is given in the appendix.