14 November 2014

Arguments For and Against Antarābhava.

One of the features of Buddhist rebirth beliefs under the microscope, is a great deal of disagreement and dissent between various Buddhist schools of thought and even internally to each school. This disagreement is seldom given sufficient attention. There is no single agreed account of rebirth or karma and I've already used this blog to highlight a number of disputes that in some cases are unresolved after more than 2000 years. In this essay I want to return to the subject of the antarābhava or interim state. I previously tackled in The Antarābhava or Interim State as a Vitalist Concept (11 July 2014) which critiqued the views of Sujato and Piya Tan. In this essay I will note some findings from an article by Qian Lin (2011) and another by Robert Kritzer (2000).

Lin points out that many of the traditional arguments for or against the existence of the antarābhava rely on lists of people who are called anāgāmin. Since this did play a central role in Piya Tan's apologetic for antarābhava and I glossed over it in my previous essay I will go into it in a lot more detail here. Lin surveys the relevant literature in Pali, Sanskrit, Gandhari, and Chinese and summarises the various lists of types of anāgāmin, giving information about the sectarian affiliations of the lists and discussing the discrepancies. He points out that even under close scrutiny, the history of the idea of antarābhava is unclear. We cannot tell which version of the antarābhava (or even no antarābhava) came first. I will make a comment on this at the end of this essay.

The word anāgāmin means "one who does not come [back]" (from ā√gam 'come') and is usually translated as "non-returner". In early Buddhist texts there are four types of noble disciples (P ariyapuggala): stream-entrants (P. sotāpanna), once-returners (sakadāgamin), non-returners (anāgāmin) and arahants. The various types are defined by which of the 10 fetters they have broken or weakened; and by how many rebirths they have yet to suffer in the kāmadhātu or sphere of sensual desire. The anāgāmin, having broken all of the five lower fetters, attains nibbāna without further rebirth in the kāmadhātu (hence they do not 'come back').

One thing to be aware of here is the Buddhist habit of working out permutations. If we have the unawakened and the awakened, the Buddhist exegetes had a penchant for listing all the possible states and treating each as if it were a real category. Another example is the paccekabuddha. It's unlikely that this category of awakened who did not teach has any basis in history (though compare Vinay Gupta), but if one is working through the possibilities, then this is one situation that can hypothetically exist. In all likelihood the anāgāmin is merely hypothetical (indeed the category is impossible to test). Thus although a lot of ink has been spilt over the interim realm based on the interpretation of this category, whatever the conclusion is, it has to be taken with a grain of salt. The discussion only makes sense within the religious parameters of Buddhism, and only follows the internal logic of Buddhism. It tells us nothing whatever about the world. That said, my task is to essay the various forms of afterlife believe held by Buddhists, so clarifying this aspect of Buddhist belief is important for a complete history of the idea.

To complicate matters there are canonical and post-canonical lists of subtypes of anāgāmin which vary in unpredictable ways: for example they may have the same list items but in a different order, and some philological problems remain with the texts, so that some terms are unclear in meaning. In these lists there are five sub-types of anāgāmin, of which one called antarāparinirvāyin which must mean something like "one who is liberated in-between". In other languages:
  • Pāli antarāparinibbāyin
  • Gāndhāri aṃtarapariṇivaï
  • Chinese 中般涅槃
Texts grouped by list type with school affiliation
(see Lin p.165)
The crux of the subsequent argument rests precisely on the question, "Between what?" The situation becomes more complicated as even the subtypes are sub-divided so that there are three kinds of antarāparinirvāyin. There are various approaches to explaining a total of seven sub-types of anāgāmin and there are three different lists of seven (the texts the different lists appear in along with their sectarian affiliation are represented in the table, right). The most prominent is the Pali Purisagati (Destination of Men) Sutta (AN 7.55; iv.70-4). This describes each type in terms of their practice, their level of realisation and uses a simile to illustrate the differences. Of the various lists all have the antarāparinirvāyin as the first member, but they are spread over a number of texts related to a range of different schools.

The Case Against Antarābhava

Lin surveys two main interpretations of the lists of anāgāmin types. The first occurs in the Aṅguttara Nikāya and the Chinese Madhyāgama and utilises the Iron Bowl Simile. In this simile an iron bowl (ayokapāle) is heated all day and struck with a hammer (Lin may have based his discussion on the Chinese counterpart in the Madhyāgama, as he discusses the simile in terms of an iron "slab": 159-60). The fate of the anāgāmin is likened to a chip or spark which flies off. For the sake of brevity, we'll stick to the similes for the antarāparinirvāyin anāgāmin. Struck by the hammer the chip...
  1. arises and is extinguished (nibbattitvā nibbāyeyya)
  2. arises, flies up, and is extinguished (nibbattitvā uppatitvā nibbāyeyya)
  3. arises, flies up, strikes the floor, and is extinguished (nibbattitvā uppatitvā anupahacca talaṃ nibbāyeyya).
The traditional Theravāda interpretation of the antarāparinirvāyin anāgāmin found in the Puggalapaññatti is that the practitioner is reborn as a deva in the rūpadhātu and achieves liberation there before mid-life. This is consistent with the Theravāda view outlined above. "In-between" here is literally taken to mean the mid-point of life (in the rūpadhātu) i.e. between deaths. The 舍利億䰓誾曇論 = *Śāriptrābhidharma (T 1548) associated with the Dharmaguptaka Sect has a similar interpretation. Note that here that nibbattitvā is from nir√vṛt and nibbāyeyya is from nir√vā, and thus despite superficial similarities (rv > bb in Pali) the two words are not etymologically related.

Theravāda exegesis, particularly the Abhidhamma text, Kathāvatthu, explicitly denies the possibility of an antarābhava (Kv 361-5; Aung & Rhys Davids 1960: 212-213). A major problem with antarābhava from the Theravāda point of view is that the word is not found in the suttas. The whole idea of an antarābhava is in conflict with models such as the khandhas and the possible destinations for rebirth (gati). It is never mentioned as a gati. There is also the huge problem of continuity. For the Theravādin Ābhidhammikas the continuity of the viññānasota or stream of consciousness can only be maintained if rebirth is instantaneous: the last moment of consciousness in the dying person (cuticitta) must be the direct condition for the arising of the first moment of consciousness (paṭisandhicitta) in the new person. The more so because the cuticitta and the paṭisandhicitta have the same object (ālambana), as does any subsequent moment of bhavaṅgacitta (resting-state mental activity). If this series is interrupted the whole Theravāda model of how karma produces rebirth, including their solution to Action at a Temporal Distance, breaks down. So, historically, Theravādins reject the antarābhava on both scriptural and logical grounds.

Even so, in practice many modern day Theravādins accept the existence of an antarābhava, as noted in my previous essay. Lin cites the study by Rita Langer (2007: 82-84) which records that in Sri Lanka most lay people and many bhikkhus, against Theravāda orthodoxy, believe in an antarābhava. This ties in with local folk beliefs about the afterlife. Prolific translator Bodhi also seems to accept the idea of an antarābhava in his Aṅguttara Nikāya translation (see 2012: 1782 n.1536). Blogger and writer, Sujato also seems to accept it. Sujato (2010) glosses the Theravāda arguments against antarābhava and concludes:
"These argu­ments sound sus­pi­ciously post hoc. The real reason for the oppos­i­tion to the in-between state would seem rather that it sounds sus­pi­ciously like an anim­ist or Self the­ory."
While he is correct to be suspicious of vitalist or animist theories, he does not consider impact of discontinuity between beings on viññānasota (i.e. the destruction of the whole mechanism for karma carefully worked out by the Theravāda Ābhidhammikas). For Sujato the clinching argument comes from a single reference in the Kutuhalasāla Sutta (SN 44.9)
‘And further, master Got­ama, when a being has laid down this body, but has not yet been reborn in another body, what does the mas­ter Got­ama declare to be the fuel?’ 
‘Vac­cha, when a being has laid down this body, but has not yet been reborn in another body, it is fuelled by crav­ing, I say. For, Vac­cha, at that time, crav­ing is the fuel.’ [Sujato's translation]
His note shows that at least one of the Chinese counterparts to this text does not imply any gap. They also show that this passage is overlooked by the Kathāvatthu discussion. The question, then, is did this text even exist at that time? Sujato concludes that:
"the Buddha, following ideas current in his time – for Vac­chag­otta was a non-Buddhist wanderer (parib­bā­jaka) – accepted that there was some kind of interval between one life and the next."
Apart from general caveats about what the Buddha might or might not have believed being entirely obscured by history, we must concede that this sutta is phrased in such a way as to allow for the idea that the author might have accepted a gap between death and rebirth. However note that Buddhaghosa glosses this by saying it refers to the moment (khaṇa) when between death (cuti) and arising of the paṭisandhicitta (SNA 3.114), i.e. Buddhaghosa is concerned to preserve the integrity of the viññāṇasota. 

The context here resists the interpretation of antarābhava. Vacchagotta is involved in speculation about where famous people have been reborn or even if they have been reborn at all. The question raised is about rebirth generally, about how rebirth can occur at all. Vacchagotta's doubt is specifically related to not being reborn, he is perplexed about how someone is not reborn. In the metaphor "Fire burns with fuel, not without fuel" (aggi saupādāno jalati, no anupādāno). The metaphorical distance between one fire and the next is spatial not temporal. In answer to the question, what causes fire to spread across space and ignite new fires, the answer is wind (vāto), the archetype of physical movement. The wind element causes fires to spread. To then read the question about rebirth in temporal terms, as explaining a time gap between bodies (kāya) is to misunderstand the metaphor. The question, really, is about what drives a person (satta) from body to body (note the metaphysics of the question are still not orthodox Buddhism).

On the other hand it is de rigueur for Buddhists to allow the beliefs of their interlocutors to stand in an argument without disputing them, but to turn the conversation away from the content of beliefs towards practice. Thus when in the Tevijja Sutta the Buddha declares to the two Brahmin students that, unlike their own teachers, he definitely does know Brahmā, Brahmā's world and the way to Brahmā's world, we need not take the author literally. He is using the language of the theistic Brahmins without contention because his purpose is not to dispute metaphysics, but to direct attention to experience. Now, when the author of the Kutuhalasāla Sutta puts these words in Gotama's mouth he does not waste time having Gotama refute the metaphysics of rebirth, but simply gives the standard answer as to the condition for all kinds of rebirth: if one has any kind of existence the primary condition for that is craving. It's not, as Sujato seems to imply, that craving (taṇha) is a special kind of fuel (upādāna) for existence in the antarābhava. Craving is what keeps the rounds of rebirth turning. Taṇha is always the upādāna for bhava.

So if we see the Buddha answering a general question about rebirth in terms of an otherwise absent idea of antarābhava it really doesn't make sense. We cannot from such obscure and difficult passages claim to know the mind of the Buddha. In terms of Theravāda metaphysics, another kind of being in a previously unmentioned interim state is a philosophical disaster: the whole Abhidhamma model of karma collapses (which effectively means that Theravāda Buddhism collapses because answers to so many other questions ride on the model of karma). This means that even if some Theravādins believe in an antarābhava they are left with the task of reconstructing the whole of Theravāda metaphysics to account for it. In the process they abandon Buddhaghosa. Though we can see that antarābhava is attractive, it's clear that the implications of the belief have not been thought through.

The Case for Antarābhava

The literature which argues the case for the antarābhava is more extensive than the contrary. Lin highlights the Saṅgītiparyāya as containing an important argument in favour of antarābhava. This text (T 1536) is a Sarvāstivāda commentary on the Saṅgītisūtra (= P Saṅgīti Sutta DN ) included in their Abhidharma. In this reading the antarāparinirvāyin dies in the kāmadhātu, arises in the antarābhava and attains nibbāna before being reborn in the rūpadhātu. Other types of anāgāmin are reborn in the rūpadhātu and attain nibbāna from there, slowly or quickly. This pattern is also followed in the Vibhāṣā and the *Saṃyuktābhidharmahṛdaya. The Abhidharmakośa mostly agrees and confirms the reading of antarāparinirvāyin.

Like the Theravādins, the Sarvāstivādin Ābhidharmikas had been developing Buddhist doctrine in order to solve problems in the received teachings, particularly the problem of Continuity and the problem of Action at a Temporal Distance (See Sarvāstivāda Approach to the Problem of Action at a Temporal Distance). As a result of the solution they adopted, the Sarvāstivādins ended up with the opposite problem to the Theravādins. Where the Theravādin model of continuity breaks down with an antarābhava, the Sarvāstivādins reasoned that there would be no way to maintain continuity through death without an antarābhava.

The Sāṃmitīyanikāyasśāstra (associated with the Sāṃmitīya Sect) argues that vijñāna without rūpa (i.e. a body) is not possible and that some kind of body is required to carry vijñāna from one rebirth to the next (Kritzer 2000: 241). This is significant, because wrapped up with antarābhava is the idea of the manomayakāya the so-called "mind-made body". Although neither Lin nor Kritzer mention this entity it is crucial in some accounts of the afterlife and thus at some point we will need to consider what it is and how it functions (I'll return to this idea in a forthcoming essay).

For a further detail of the Yogacāra arguments for antarābhava we can turn to Kritzer (2000). His article examined the views of Vasubandhu, especially as found in the Abhidharmakośabhāṣya (Vasubandhu's auto-commentary on the Abhidharmakośa) but also crucially the Viniścayasaṃgrahaṇi. The Bhāṣya is both the most systematic and one of the most influential accounts of the subject, as much for its portrayal of Vasubandhu's opponents as for his own views. Much of the contemporary scholarly writing on antarābhava is based on the Bhāṣya, and in many ways it has been over used as a source text on schools whose own literature is lost, fragmentary or only preserved in Chinese (especially the Sarvāstivādins). Kritzer points out that despite commonalities with the Sarvāstivāda account, the two should not be equated as he shows by examining arguments in the Vibhāṣā, one of the foundation texts of the Sarvāstivāda.

I want to write a separate summary of Kritzer (2000), since it will be quite long, but for now will try to give a flavour of the arguments. The crux seems to be a development of the idea vijñāna supported by rūpa mentioned above. Vasubandhu returns to an agricultural metaphor for the life-cycle of humans comparing us to rice plants (cf. comments on the fivefold-niyāma in Experience and Free Will in Early Buddhism). Vasubandhu's interpreters have read this different ways, but what he seems to be getting at is that the rice seed provides continuity between rice plants. What we do not see is one rice plant becoming another rice plant with no interval. Vasubandhu imagines that humans produce "seeds" when they die (though here he seems not to be referring to the karmic seeds stored in the ālayavijñāna). These seeds provide us with an interim body of a sort that sustains vijñāna until it can connect with rūpa again in rebirth (it's here that the idea of a mind-made body is both relevant and paradoxical because it suggests that a manokayakāya is the manas playing the role of rūpa in order to be a condition for the arising of vijñāna - i.e. it involves circularity that is disallowed by other doctrines of how conditionality works). The Viniścayasaṃgrahaṇi contains a series of questions and answers including this one:
Question: how does one know that there is an intermediate existence? Answer: because [when a being] dies here, there is no way for his citta and caittas to go without support to another place. It is not like an echo because [an echo] is merely an illusion. It is not like a reflected image because that [object] does not perish. And it is not like grasping an object because there is no movement [of consciousness in the case of perception]. Because these similes are inappropriate, the intermediate existence must be understood to exist. Thus, one must contemplate the arising of rūpaskandha in accordance with this”. (Kritzer 247)
This ties in with another image related to rice. Vasubandhu uses the example of a load of rice being transported from one village to another. It does not simply disappear from one village and appear in another, but goes on a journey through a series of stages. In other words Vasubandhu is, unlike many of his predecessors, thinking explicitly and abstractly about causation. Change or movement, as Vasubandhu observes it, is not instantaneous but gradual and thus rebirth cannot be instantaneous either. This may well hark back to Nāgārjuna's abstruse discussions of change in the first chapter of the Mūlamadhyamakakārikā.

Sujato may well argue here that this metaphor is analogous to the fire metaphor in the Kutuhalasāla Sutta: that the transportation through space is the root metaphor for rebirth, and that as transit through space is not instantaneous then rebirth cannot be instantaneous. Something must effect the transit between bodies. In response we might question whether reifying the metaphor is helpful. Is the transmission of certain crucial (moral) information about the actions of the previous life onto the next life so as to determine the realm and circumstances of rebirth simply a physical process, like spreading fire or transporting rice grain? Or does the metaphor allow for differences? For advocates of substance dualism the mind is clearly a different stuff to the body and cannot be subject to physical laws or it would not work. One of the features of ESP, a feature of many Buddhist discourse, is that it works with no regard for physical distance: in clairvoyance for example, one knows the thoughts of others as they think them. For advocates of substance monism the idea of an afterlife is so unlikely that it is hardly worth thinking about, but presumably a substance monist would insist that information transfer must take an appreciable time: like downloading a file from the internet. However, no Buddhist metaphysics excludes miracles, magic or ESP.

Vasubandhu is clearly trying to avoid the charge of eternalism by making the antarābhava analogous to other states of being: vijñāna arises in dependence on the manifestation of rūpaskandha in the antarābhava. The scriptural argument against this is simple and was stated in the Kathāvatthu more than 2000 years ago: if there is a an interim state of being, then why is it not included in traditional lists of such states? If there is rūpa then this is (effectively) a rebirth. Why is it not listed as a rebirth destination (gati)?

Vasubandhu's main argument is similar in form to Xeno's paradox. The counter argument is that if some interim state between rebirths (even transition from kāmadhātu to the rūpadhātu) is definitely required, then the same argument holds for the transition from the kāmadhātu to the antarābhava. By Vasubandhu's reasoning we are forced to postulate an antarā-antarābhava and along with it some even more subtle form of being. And so on ab absurdum. Every transition requires an interim state between the original state and the changed state with infinite regress. So the idea of an antarābhava does not solve Vasubandhu's observed problem with causality.


The logic of the arguments outlined is entirely bound up with versions of the Buddhist worldview. As with all afterlife beliefs, there is no way to argue about the antarābhava from first principles. How we view the antarābhava is entirely dependent on what we stipulate at the outset. On traditional arguments, it is either required or forbidden depending on our starting assumptions about how karma and rebirth work. For religious Buddhists this has meant, essentially that religious arguments (based on scripture) carried considerable weight and that reasoned arguments were always constrained by religious arguments.

And thus it is all the more curious that contemporary religious figures such as Theravāda bhikkhus and scriptural commentators reject the religious arguments of their own tradition and adopt the antarābhava, even though it invalidates their own model of karma and rebirth. Such doctrinal conflicts have clearly never bothered the religious lay people very much. Lay Buddhism has always been a religion of faith and propitiation rather than intellect and theology.

My earlier essay pointed out some of the philosophical problems that the antarābhava entails: it seems to involve a form of eternalism. This is something that the Continuity problem cannot ever avoid: either there is discontinuity or there is continuity. In the former the problem of how to transmit information karma is unsolved, in the latter the solution is inevitably eternalistic. The idea of dependent arising doesn't actually solve this dilemma, it only disguises it. There are any number of problems with using pratītyasamutpāda as a Theory of Everything. One cannot take a description of the phenomenology of mental activity presenting itself to awareness and turn that into a general metaphysics and especially not into a physics without creating problems.

So despite the fact that Theravādins settled on their explanation (until recently) and Māhāyānikas settled on Vasubandhu's explanation, in fact neither the problem of Continuity, nor the problem of Action at a Temporal Distance, were definitively solved by either party. The problems were simply shelved as original intellectual contributions dried up. In India, Buddhist exegesis became a competition with non-Buddhists traditions on matters previously considered inconsequential to the Buddhist project; while in Sri Lanka and Burma it turned into increasingly elaborate restatements of old ideas. I'm not well enough informed about Buddhism outside Indian to form definition opinions, but my impression is that the problems of assimilating Buddhism into a culture like China presented such massive problems that Buddhist theology went in entirely different directions. The Chinese seem to have deified the Buddha, whereas the Tibetans were constantly occupied with managing the massive proliferation of teachings. Modern Buddhism largely ignores discontinuities and is mainly concerned with presenting Buddhism as a transcendent truth with no visible flaws, a panacea that applied to everything, results in Utopia, emerging fully formed from a singularity we call Buddha. Could we be any further from the historical nature of our own religion?

At the outset I mentioned that it was unclear from Lin's account whether antarābhava was part of the original narrative of Buddhism or not. I now think it is clear that it is a late addition. Awareness of problems like Continuity and Action at a Temporal Distance only emerge in the post-sutta literature of the Abhidharma. Antarābhava simply doesn't occur in any early text, even when the concept of punabbhava is prominent. The single reference which seems to point to a poorly defined belief in at least a spatial distance between lives, hardly changes the picture. The fundamental disagreement about antarābhava means it can only have emerged once Buddhism had began to fragment into sects. The arguments evinced by the various sides rely on mature Abhidharma theories. The Theravādins only consider it as a reaction to the Abhidharma theories of other schools. So antarābhava was not part of the original Buddhist narrative about the afterlife. That said, the problems which led to antarābhava being proposed as a solution were in place early on.



Aung, Shwe Zan & Rhys Davids, C. A. F. (1960) The Points of Controversy: or, Subjects of discourse being a translation of the Kathāvatthu from the Abhidhammapiṭaka. Pali Text Society. First published 1915.
Bodhi. (2012) The Numerical Discourses of the Buddha. Wisdom.
Kritzer, Robert. (2000) 'Rūpa and The Antarābhava.' Journal of Indian Philosophy 28: 235–272.
Langer, Rita. (2007). Buddhist Rituals of Death and Rebirth: Contemporary Sri Lankan Practice and its origin. Routledge.
Lin, Qian. (2011) 'The antarābhava Dispute Among Abhidharma Traditions and the List of anāgāmins.' Journal of the International Association of Buddhist Studies 34(1-2): 149–186.
Sujato (2010) Rebirth and the In-Between State in Early Buddhism. http://santifm.org/santipada/2010/rebirth-and-the-in-between-state-in-early-buddhism
Related Posts with Thumbnails