How Buddhist Rebirth Changes Over Time

ONE OF THE FACTS about the foundation texts of Buddhism that most people don't seem to have taken in is that rebirth is an idea with a history. The idea did not spring into being fully formed. And what's more we can discern this history in the Pāli texts themselves. It has been traced in detail by Gananath Obeyesekere in his book Imagining Karma. In this post I want to review the development of rebirth from its primitive form to the full blown received version, basing myself on Obeyesekere, along with some observations and diagrams of my own. The received tradition tends to obscure the variations in the texts, but they can be (at least partially) reconstructed. So this is a kind of archaeology in the spirit of Foucault. A caveat here is that we don't know the absolute chronology of these changes, we only know that they were all preserved, somewhat unevenly, with the fixing of the Canon.

The most basic form of rebirth eschatology is binary. It involves 'this world' (ayaṃ loko) and 'the other world' (paro loko) a way of referring to rebirth that one finds scattered throughout the Canon, and which may have been retained as an idiom long after the binary model had been augmented. In this simple model of rebirth one lives on earth; then after death one rises up to the other world (always up), where one lives for a long time; then one falls back to be reborn on earth again. For example in M 49 the movement is described by this sequence of verbs: jāyati jīyati mīyati cavati upapajjati--being born, living, dying, falling, being rebirth. Rebirth is automatic, and human.

Brahmins also began with a binary cyclic eschatology. Indeed it seems as though rebirth eschatologies were indigenous, or at least endemic, in India. The Brahmin ancestors (or fathers) live in the other world. This cycle is what is referred to as saṃsāra - which means 'going through; course; passage' (from saṃ- 'with, together, complete' √sṛ 'flow, run, move'). The cycle is believed to be endless and beginningless. At this early stage rebirth is not problematised; its just a description of the how the world is. However for the Brahmins going to the next world, like all significant life moments, required the performance of certain rituals. There is no sense of morality being a factor here, but the need for the rituals to be performed correctly had a similar effect. The arrival of morality is the next thing to discuss.

What morality does to any afterlife is divide it. If one has lived well the other world is a place of reward, and if one has not lived well the other world is a place of punishment. In Buddhist texts we find the distinction in the pair of terms 'good destination' (sugati) and and 'bad destination' (duggati. Skt durgati). Another pair of terms are 'heaven' sagga (Skt svarga) and 'hell' (niraya). The word svarga 'shining place' has a long history in the Vedic tradition. It was where the gods lived, but also where the ancestors lived, so in simple terms the other world was svarga. It was situated beyond the sky. However initially there is no clear reference to hell in Indian texts, it's not really until Buddhism that hell plays any definite role in Indian cosmology or eschatology. The word niraya means 'going down'. Because the idea of a subterranean hell appears to be absent from earlier Vedic texts, some scholars have speculated that the idea of hell comes Zoroastrianism (via the Iranian Śākya tribe - see Possible History). Like heaven, the early hell is a place where you go to live out the consequences of the actions done in life, but not a place where one does actions with consequences. We see this explicitly in the Devadūta Sutta (M 130) where one is tortured in hell, but does not die, and therefore cannot be reborn elsewhere until the wicked actions have exhausted their force. Actions carried out in hell appear to have no bearing on this fate.

Note that liberation is outside of space and time and described as "dhuva, sassata, nicca, etc." by both Brahmins and Buddhists. Because the Brahmanical diagram would look just the same I say the two are topologically identical.

At the same time a third option appears, which is liberation (mokṣa, vimokṣa) from going around the cycles. The idea is first seen in literature in the Bṛhadāraṇyaka Upaniṣad (BU). By re-jigging the dates of the ancient India texts and placing BU after the Buddhist texts, Johannes Bronkhorst manages to argue that this idea must have come from the śramaṇa milieu. However it's doubtful whether his revised chronology will stand up to scrutiny, and I know of no other scholar who has adopted it yet.  Even so, my work on the Iranian origins of the Śākya tribe makes it seem possible that the idea of liberation (i.e. a single destination eschatology) might have been introduced into both milieus around the same time (ca. 850 BCE) from Iran; leaving the current consensus on chronology intact. However it arose, the option of liberation from saṃsāra becomes the major preoccupation of Indian religion from about the middle of the first millennium BCE down to the present. And given how it spread in various guises it must be seen as one of the most influential ideas in the whole history of ideas.

It seems as though these early versions of rebirth eschatology are similar to Brahmanical views, but they might have been more widespread. Rebirth eschatologies are not common amongst the Indo-European speaking peoples (with some ancient Greeks as a debatable exception) but they are ubiquitous in India. So, like linguistic features such as retroflex consonants, rebirth might have been a regional feature. In any case what happens next is the incorporation of some explicitly Brahmanical elements into the Buddhist model. These are not taken on their own terms, in fact presented in distorted, rather mocking ways.

For the Brahmins we meet in the Canon going to Brahmā's realm (brahmaloka) is synonymous with mokṣa or liberation from saṃsāra. Richard Gombrich has argued that the Buddha used brahmasahāvyatā as a synonym for nibbāṇa; which in turn explains the brahmavihāra (literally "dwelling with/on/like God") meditations. Buddhists denied Brahmanical soteriology, and did two things: they brought Brahmā's realm back into saṃsāra, but placed it over the god realm (devaloka) creating a new refined level of saṃsāra (also called ārupaloka); and they multiplied the Creator God into a whole class of very refined beings called Brahmās (plural). On one hand the Brahmās are the highest beings in saṃsāra and people in the texts are very impressed when one of them visits the Buddha, and one of them, Brahmasahampati, is responsible for convincing the Buddha to teach; and on the other hand they are depicted as being deluded about their own nature, trapped in saṃsāra and therefore subject to death. The other thing that happens at this stage is the separation of the spirits of the dead from the gods. The word peta (Skt. preta) has two possible etymologies one which derives it from the word for father (pitṛ) and the other which derives it (as an action noun) from a verb meaning 'gone before' or 'departed' (pra-√ī). In any case this common word for the spirits of the dead who are in the other world becomes a pejorative. Perhaps because the Brahmins made sacrifices to the gods and to their fathers, in Buddhism the preta came to stand for a class of ghosts who were constantly hungry, but unable to ever satisfy that hunger.

At the same time, or perhaps a little later, the idea arose that one could be reborn as an animal. This idea is first seen in the Bṛhadāranyaka Upaniṣad where the fate of those who do not carry out the rituals is to be reborn as an invertebrate. So at first it appears to be a somewhat chauvinistic Brahmanical idea, but it catches on and is incorporated into the Buddhist eschatology.

Click to enlarge

The final stage involves the emergence of the full-blown version of the Buddhist cosmology with the brahmaloka, devaloka and hell realms being divided into many different layers, and the layers of the first two being related to states of meditation. The devas and their counterparts the asuras undergo their separation and the asuras are sometimes (but not always) given their own realm. In some older parts of the Ṛgveda the two terms deva and asura are synonyms. Varuṇa for example is referred to as both deva and asura. However the contest between them required a winner and loser, and the asuras lost. (In Iran they won and the devas are seen as demons.) Some remnants of the early stories are preserved, often with little alteration, in the sakkasaṃyutta of the Saṃyutta Nikāya (the 11th chapter, beginning on p.317 in Bodhi's translation). For the purposes of diagramming the brahmaloka and devaloka are often treated as aspects of a single domain, though Brahmā is never referred to a deva. This gives us the traditional six domains of rebirth: human, deva, asura, preta, hell, animal, as seen, for example on the bhavacakra or 'Wheel of Becoming'. It is possible to go to any realm from any other realm, but liberation is only possible from the human realm.

One of the major changes from beginning to end is the likelihood of a human birth. Initially it is 100% certain. Even in a morality influenced eschatology one always returns to this world as a human being eventually. However, by the end of the process the likelihood of being born human is vanishingly small. The chance compares unfavourably with the probability that a blind turtle raising its head from the great ocean just once a century might put its head through the hole in a plough harness (yoke not yolk!) which is floating about at random on the ocean. While this is not impossible, the chances are vanishingly small. If we take this on face value we have almost 0% chance of being born human. Related to this is the possibility of multiple rebirths in hell or heaven, particularly the former. This suggests a growing concern over the waywardness of human beings and a greater desire to curb behaviour with the threat of exile from humanity in the afterlife. In other words it looks like a hint that rebirth theory changed in response to social change. This should not be surprising as a huge number of Vinaya rules, including the pāṭimokkha ceremony itself, are made in response to public pressure.

In this essay I've been looking at the development of the idea of Rebirth in the Pāli texts. Given the way that kamma changed after the Pāli Canon was closed, it is only reasonable to assume that ideas about rebirth also continued to change. I will briefly mention one other major development in rebirth theory which was the invention of the so-called Pure Land: a parallel universe with a living Buddha. The Pure Land was not simply another level in this universe, not another level of heaven, but an entirely separate and complete universe (though usually lacking the durgati). The parallel universe was not invented because the ancients had insights into the nature of the multiverse or M Theory, it was a theological necessity for those who had begun to believe that the presence a living Buddha was necessary for liberation (the same theological anxiety can be see in the Suvarṇabhāsottama Sūtra; and in Peter Masefield's Theravāda oriented book Divine Revelation in Pali Buddhism.). The Pure Land is a place where liberation is guaranteed by the constant living presence of a Buddha (I would argue that at this point the Buddha has become a god, theos; and that the term theology is entirely appropriate). The resident Buddha in fact creates this parallel universe through their practice of the perfections, emphasising the importance of hard work. Fantastically rococo in many other respects, each Pure Land is entirely flat for some reason. I mentioned Pure lands last week, and it is a fascinating area, but for another essay. Those interesting in how Pure Land theory developed should read this article by one of my favourite authors:

Nattier, Jan. (2000) 'The Realm of Aksobhya: A Missing Piece in the History of Pure Land Buddhism.' Journal of the International Association of Buddhist Studies. 23 (1): 71–102. Online: http://archiv.ub.uni-heidelberg.de/ojs/index.php/jiabs/article/view/9167

Those who oppose the idea that rebirth is implausible often fall back on simplistic arguments like: rebirth has always been accepted by Buddhists, it's been analysed and accepted as true many times. However this argument seldom takes in the subtleties of the history of the idea. Rebirth clearly changes during the period between of the inception of Buddhism and the closing of the canon. Several different versions of rebirth are, as it were, trapped in the amber of the Pāli texts. But rebirth continued to change. The received tradition, as is usual, never acknowledges the variety of the models, nor the subtle contradictions in the collection of texts. Received traditions are all about presenting an internally coherent narrative, and ironing out difficulties. So inconsistent aspects of the textual tradition are reinterpreted or simply bracketed out. This is not a new process. And confirmation bias is not a new problem.

Contrarily those who seek to deny that rebirth was part of the original teaching don't have a leg to stand on. Rebirth is prominent in the older hagiographical accounts like the Ariyapariyesanā Sutta, and in the older parts of the Sutta Nipāta. Rebirth is quite obviously an important part of Buddhism in the earliest records we have. The idea that rebirth is somehow in the background, or was added later, is insupportable based on current evidence. That rebirth no longer seems plausible is an entirely different proposition. And one that creates a dilemma that I have no wish to underplay. We have yet to really work out the implications of this news, though it is the news. Understanding that our doctrines have always been quite changeable and responsive to social change, seems to me to be an important factor in loosening our grip on traditional doctrines with a view to letting them go. Everything changes. Resisting changes causes suffering. The only way forward for Buddhism is, well, forward.

~~oOo~~

How the Doctrine of Kamma Changes Over Time

Anubis weighing the soul of Ani the Scribe against the Law to decide his fate in the afterlife. Egyptian Book of the Dead.

EVER SINCE I FIRST published my article on the confession of King Ajātasattu in the Samaññaphala Sutta (Journal of Buddhist Ethics), I have wanted to write a follow-up essay which showed how the idea of karma changed after the early Buddhist period. I'm using this post as a sketch to be filled out later. More recently, I outlined a possible pre-history of the idea of karma, linking it to Egyptian ideas about how morality affected the afterlife transmitted to India via Iran, which then interacted with local Indian beliefs. In this essay I will look at some milestones along the route of a major change that happened to Buddhist moral theory as it moved out of its 'early' phase, in particular, the early ideas of the Pure Land (first centuries CE), the Mahāyāna version of the Samaññaphala Sutta (ca. 5th century CE) , and a special mantra which appears for the first time in the Sarvatathāgata-tattvasaṃgraha (ca. 7th century).

In the early Buddhist theory the results of actions are inescapable; there is nothing that stands between us and the consequences of our actions, not death, not god or the Buddha, and no form of praxis. The Buddhist commentators came to see this belief as epitomised in Dhammapada Verse 127:

Not in the sky, nor the middle of the ocean,
Nor in a mountain cave;
Though terrified, there is nowhere on earth
Where one might escape from an evil action.

Buddhaghosa, cites this verse, for example, in his commentary on the Mahāpadāna Sutta while explaining the term dhammatā, 'naturalness'. He uses it to explain the inevitability of karma (kamma-niyāma), which is one of the five niyāmas. In the Ajātasattu article I cite a couple of texts which suggest ways in which one might lessen the impact of the consequences of our previous actions, but there is no way to avoid them entirely. This is a distinctive moral teaching of the Early Buddhists, and yet precisely this aspect of Buddhist morality changes.

In the Samaññaphala Sutta, King Ajātasattu is troubled by his conscience and goes to meet the Buddha. During the meeting, he confesses that he has killed his father (and the Buddha's friend and patron), King Bimbisāra. The Buddha accepts this news, and acknowledges that the King wishes to return to lawfulness. However, when Ajātasattu leaves, the Buddha says to the bhikkhus "the king is wounded (khatāya), and done for (upahatāya)" (D i.86). Had Ajātasattu not killed his father, he would have attained the dhammacakkhu after hearing the Dhamma discourse. What’s more, patricide is one the five actions which result in immediate rebirth in hell after death. The patricide is said to be atekiccha, 'incurable' or 'unpardonable' (see A iii.146). Buddhaghosa's commentary records that on death Ajātasattu goes straight to the Hell of Copper Kettles. Note that his comeuppance comes only after death, and compare my conjecture that the notion of being judged on your actions after death was introduced to India from Iran. In M 130 we find that a trip to hell lasts for as long as it takes for the consequences to play out; i.e., one was not repeatedly reborn in hell, but more on the development of rebirth theory next week.

In my article, I showed that when Ajātasattu is told the Buddha of having killed his own father he cannot be considered to be 'making amends' (as modern translators suggest), nor does the Buddha 'forgive' him, since such a thing is not in his power. Incidentally, I showed that everyone (including both translators and lexicographers) had previously misinterpreted the word paṭikaroti and the phrase yathādhammā paṭikaroti (returning to lawfulness). Ajātasattu confesses and makes a resolution to return to moral behaviour. The Buddha simply acknowledges the confession and resolution, and does not intervene in any way, because in this system of morality he cannot.

In this worldview we are each the heirs of our own actions (c.f. the five facts to reflect on) and literally nothing can change that. Indeed, I would argue that karma can hardly be expected to work as a moral deterrent unless this is true. But it seems that Buddhists found this limitation too onerous, and introduced ways to first limit and then eliminate the damage done to themselves by evil actions.

The Samaññaphala Sutta has survived in several versions preserved in the Chinese Canon, and a fragment in Sanskrit (Sanskrit title: Śrāmaṇyaphala Sūtra), which have been studied by MacQueen (1988). In the Pāli sutta, the Ajātasattu story is relatively unimportant and merely frames the important middle section of the sutta dealing with heterodox teachings. MacQueen shows that the later versions make the meeting of the Buddha and Ajātasattu the most important feature, and finds the middle section relatively unimportant. One Chinese version (C1 in MacQueen’s notation) says of Ajātasattu “his transgression is diminished; he has removed a weighty offence” (p.48-49). C2 has by contrast “he has completely done away with imperfections and impurities and is free from the Outflows [āsravas]” (p.69) which is to say that just meeting the Buddha completely purifies Ajātasattu, releases him from the consequences of killing his father, and elevates him to awakening. On this change MacQueen comments: “In the 5th century A.D. this religious event [Ajātasattu’s conversion] was of far more interest than the issue of whether or not there were immediate fruits to the life of a monk” and that “the more depraved the person is who is saved, the more the Buddha’s divine power is demonstrated” (p.215).

A major change has taken place in Buddhist doctrine somewhere between the closing of the Pāli Canon and ca. 5th century AD (when the Chinese copies of the Śrāmaṇyaphala Sūtra were made).

A related development occurred in the Akṣobhyavyūha and Sukhāvatīvyūha Sūtras (ca 1st or 2nd centuries BCE) which heralded the possibility of a new eschatology. It seems that the original idea was that with a lot of dedication and practice one might attain the Pure Land of Akṣobhya, and from there liberation was certain. However, Akṣobhya's Pure Land was soon eclipsed by Amitābha's. The Sukhāvatīvyūha Sūtras set forth a soteriology in which if we simply call the name of Amitābha he will meet us at death and guide us to Sukhāvatī from where we are guaranteed awakening. Indeed, the idea of Akṣobhya's Pure Land was so deeply buried that it was only rediscovered in the 20th century (see Nattier 2000) We have only to reflect on the enormous and wide ranging influence of Pure Land Buddhism throughout Asia, found in every branch of Mahāyana Buddhism, to see what a very appealing idea this possibility of being saved was, especially amongst the ordinary population of Buddhist countries. 'Calling the name' of the Buddha as a practice seemed to come under the wing of the old practice of buddhānusati/buddhānusmṛti (recollecting the superior qualities of the Buddha), the two together constituting key prototypes for the Tantric visualisation of a Buddha accompanied by chanting their mantra. There are some precedents in the Mahāvastu, which probably date from a century or two earlier than the Sukhāvatīvyūha Sūtras, and the 'calling the name' practice is linked to mantra chanting in the Kāraṇḍavyūha Sūtra (ca. 4th century).

However, the change expounded in the Śrāmaṇyaphala Sūtra is still not as revolutionary as what was to come. The Śrāmaṇyaphala Sūtra allows that in the past someone who met the Buddha might be saved from the evil consequences of their actions, which reinforced the specialness of the Buddha, but was not immediately relevant to present day practitioners who lived in a time which was many centuries removed from the Buddha. However, the Sukhāvatīvyūha Sūtras show that the idea of the Buddha as an eternally living presence was beginning to take hold. Indeed, this is a key metaphysical problem in another popular Mahāyāna text: the Suvarṇabhāsottama Sūtra (or Golden Light Sūtra). Here the puzzle is that though merit making extends life, and the Buddha is said to have infinite merit, yet he appeared to live a normal human lifespan. In response to this question the Bodhisattva Ruciraketu has a vision in which it is revealed to him (by a maṇḍala of Buddhas) that, in fact, the lifespan of Śākyamuni is infinite. And, of course, this is good news to those who believe that one must be in the presence of a living Buddha to attain awakening because now the Buddha is always living, and always accessible.

The metaphysical problems introduced by an eternal Buddha (which is a form of sassatavāda; Skt śāśvatavāda) seem to have been outweighed by the soteriological possibilities.

The process of change that I have been tracking with respect to the doctrine of karma reaches its apotheosis in a Tantric text composed in the late 7th or early 8th century: the Sarvatathāgata-tattvasaṃgraha. This text survives in two Sanskrit manuscripts (though they are copies from a much later date) as well as in Chinese and Tibetan versions). It is in this text, perhaps the first to expound a mature Tantric Buddhism, that we find the Hundred Syllable Vajrasattva Mantra or hṛdaya as the text calls it. (See, also, my article in the Western Buddhist Review. 5). This hṛdaya mantra stands apart from the surrounding text, by which I mean it seems to be a distinct idea and is not integrated into a sādhana or other practice. However, it is accompanied by a few lines before and after which give it some context. This context is the one which will be familiar, I imagine, to all tantrikas, since it explicitly says that the mantra will purify a person, and make liberation possible for them when their vow keeping has become lax, no matter what evil acts they have committed.

Tantric Buddhism, then, appears to admit no impediment to liberation, no action so heinous that it will make liberation impossible in this life. Whatever evil one has committed, one simply chants the Vajrasattva Mantra and one is released from the consequences of wicked actions. It does not require grace or intercession from a god; one's sins are simply set aside through the chanting of the mantra. I find this extraordinary, but I know from first hand experience that some tantrikas take this quite literally.

This is surely one of the most dramatic and far-reaching changes in the history of Buddhist ideas. Our doctrine is completely turned on its head over the course of several centuries. And these are not the only variations. S 36.21 outlines why we cannot consider karma to be responsible for everything that happens to us. Some scholars have seen the post-Canonical development of the five-fold niyāma as a continuation of this idea, though this, in large part, seems to stem from innovations introduced by Mrs Rhys Davids, and is not really supported by the texts which discuss the niyāmas. However, in present day Tibetan Buddhism the doctrine is that everything that happens to us is because of our karma.

There is no single unified Theory of Karma in Buddhism, either synchronically (in our time) or diachronically (across time). Instead, there are multiple theories, and very many exegetes explaining the "Truth" of karma. Some of these 'truths' are mutually exclusive. Sectarians tend not to be conversant with the details of the different theories, since sectarian teachers present their version of karma as the Truth. Those who are conversant with a range of karma theories find them difficult to reconcile. 'Actions have consequences' is what it boils down to, but it's hard to see this as a great revelation from the Buddha, since everyone knows this platitude already. The how and when of actions having consequences are Buddhism's specific contribution to moral theory, but unfortunately Buddhists themselves disagree on precisely these points.

~~oOo~~

Bibliography

• Attwood, Jayarava Michael. (2008) Did King Ajātasattu Confess to the Buddha, and did the Buddha Forgive Him? Journal of Buddhist Ethics. 15. Online: http://blogs.dickinson.edu/buddhistethics/files/2010/05/attwood-article.pdf
• Jayarava. 
  • (2010) The Hundred Syllable Vajrasattva Mantra. Western Buddhist Review. 5 Online: http://westernbuddhistreview.com/vol5/vajrasattva-mantra.pdf
  • (2011) Five Facts to Continuously Reflect On. JR
  • (2012) Possible History for the Buddhist Idea of Karma. JR  
• MacQueen, G. A Study of the Śrāmaṇyaphala-sūtra. Wiesbaden: Otto Harrassowitz, 1988.
• Nattier, Jan. (2000) 'The Realm of Aksobhya: A Missing Piece in the History of Pure Land Buddhism.' Journal of the International Association of Buddhist Studies. 23 (1): 71–102. Online: http://archiv.ub.uni-heidelberg.de/ojs/index.php/jiabs/article/view/9167.
   

Chinese Versions of the Śramaṇyaphla Sūtra studied by MacQueen:

C1. 沙門果經 Shāmén guǒ jīng (Śrāmaṇyaphala sūtra) in T01n0001:長阿含經 (Dīrghāgama). Translated 413CE Buddhayaśas (佛陀耶舍) and Zhú Fúniàn (竺佛念). CBETA T01n0001_p0107a16-114b02: http://www.cbeta.org/result/normal/T01/0001_017.htm

C2. 寂志果經 Jì zhì guǒ jīng (Śrāmaṇyaphala sūtra). T01n0022. Not part of a collection. Trans. 381-395 CE by 竺曇無蘭Zhú Tánwúlán. CBETA: http://www.cbeta.org/result/normal/T01/0022_001.htm

C3. Untitled. but referred to as 無根信 Wúgēn xìn ('Faith Without Roots' = Skt. amūlakā śraddhā) T02n125p762a07 ff. (Śrāmaṇyaphala sūtra) in T02n0125: 增壹阿含經 (Ekottarikāgama) T 2.124: 762-764. 7th sūtra, 39th fascicle, 43rd section. Trans either 384 CE or 397 CE uncertainty depends on who translated the text. CBETA http://www.cbeta.org/result/normal/T02/0125_039.htm

C4. Untitled e (T24.1450.205a09) (=partial version of Śrāmaṇyaphala sūtra) in T24.1450根本說一切有部毘奈耶破僧事 (Mūlasarvāstivāda Vinaya Saṃghabheda-vastu). Translated 710 CE by 義淨Yìjìng. CBETA http://www.cbeta.org/result/normal/T24/1450_020.htm

Note that each of the four versions is in a different place in the Canon: Dīrghāgama (= Pāli Dīgha Nikāya); stand-alone; Ekottarikāgama (= Pāli Aṅguttara Nikāya); and Vinaya.