Introduction

I live. I perform actions and have experiences. I die. I am reborn. The actions that I perform affect the quality of my rebirth. I can escape from this cycle.

These statements, while highly simplistic, are a broad reflection of the teachings of both the ancient Indian Brahmanical religion – the forerunner of what we now call Hinduism – and of Buddhism, as well as of other ancient Indian faiths. Indeed, the doctrine of karma and rebirth is “of near universal pervasiveness” in Indian religious thought, accepted, with some variations, by Hindus, Buddhists and Jains alike (Tull 2004: 318). Where these faiths differ radically, however, is over the question of who or what it is which has these experiences, who or what it is which is eventually reborn, and who or what it is which is able to escape from the cycle. This essay will explore some of the challenges which the Buddhists faced when answering these questions, and will look briefly at some of the ways in which different schools of Buddhist thought dealt with those challenges. I will begin by briefly outlining the Brahmanical viewpoint, before exploring the Buddhist rejection of that viewpoint, and explaining how various schools of Buddhist thought sought to answer the questions which arise from that rejection. I will conclude by considering the extent to which any one or more of those answers succeeds in satisfactorily dealing with what I refer to in this essay as “the conundrum of continuity”.

The Brahmanical Tradition

I use the term “the Brahmanical tradition” to refer to the various schools of religious thought which claim to derive authority from the Vedas. The Vedic canon itself comprises the four Vedic Samhitas – the Rg Veda, Sāma Veda, Yajur Veda and Atharva Veda – together with a number of other texts (Brāhmanas, Āranyakas and Upanishads), all generally thought to have been composed in the period between around 1500 BCE and the beginning of the Common Era[note] See, e.g., Flood 1996: 37; Roebuck 2000: xxvi. It is worth noting at this point thatcertain of the Vedic Upanishads were almost certainly composed after the life of the Buddha, and several are often thought to reflect Buddhist influences, though that is of marginal relevance to this essay. [/note]

A substantial part of the teachings of the Upanishads comprises teachings about the nature of the self. Broadly speaking, the Upanishads teach the existence of an essential self – ātman[note]Literally “oneself”, the reflexive pronoun[/note] – beyond and separate from the mind, senses and intellect, described as “dearer than a son, dearer than wealth,… dearer than everything else” (Brhadāranyaka Upanishad 1.4.8, translation Olivelle). For the sages of the Upanishads, knowledge of the nature of the ātman was the key to liberation. Later Upanishads[note]E.g. Svetāsvatara and Maitrī[/note] added to the requirement of knowledge of ātman yogic practices of controlling the mind and senses and/or the need for the intervention of some form of divine grace, but throughout the Upanishads it is emphasised that knowledge of ātman, and perhaps of its identity with brahman, the invisible essence of all things, ultimately leads to liberation from the cycle of birth and rebirth (samsara) conditioned by the results of one’s actions (karma).

It is in the early Upanishads that we see the earliest expositions of the doctrines of karma and rebirth in Indian religious thought. One of the earliest classic statements appears in the Brhadāranyaka Upanishad:

“What a man turns out to be depends on how he acts and on how he conducts himself. If his actions are good, he will turn into something good. If his actions are bad, he will turn into something bad.” (Brhadāranyaka Upanishad 4.4.5, translation Olivelle)

This is expanded on in the Cāndogya Upanishad:

“… people here whose behaviour is pleasant can expect to enter a pleasant womb, like that of a woman of the Brahmin, the Kshatriya or the Vaishya class. But people of foul behaviour can expect to enter a foul womb, like that of a dog, a pig or an outcaste woman”. (Candogya Upanishad 5.10.7, translation Olivelle)

For the Upanishadic sages, the question of who or what is reborn is answered simply: what is reborn is the ātman. Again, it is one of the earliest Upanishads, the Brhadāranyaka, which states this succinctly:

“It is like this. As a caterpillar, when it comes to the tip of a blade of grass, reaches out to a new foothold and draws itself onto it, so the self (ātman), after it has knocked down this body and rendered it unconscious, reaches out to a new foothold and draws itself onto it.” (Brhadāranyaka Upanishad 4.4.3, translation Olivelle)

Philosophically, the doctrine of the ātman – moving from birth to rebirth until liberated from the cycle – is compellingly simple. Clearly, the physical body cannot have this quality of continuity, as we know that from moment to moment it decays and is recreated on a cellular level. Nor can the mind have the quality of continuity, as we know that from moment to moment our thoughts change, and we know that in one life we cannot recall events from a previous life. If, therefore, there is to be rebirth, there must – says the Brahmanical tradition – be an entity with a degree of permanence beyond the physical body and the mind which is reborn. That entity is the ātman. While describing the ātman presented the Upanishadic sages with a number of challenges – the sage Yājñavalkya, for example, famously describes it as “neti, neti” (“not this, not that”) (Brhadāranyaka Upanishad 3.9.26, translation Olivelle) – it is clear that, in the Brahmanical tradition, “this ultimate metaphysical self is the unchanging constant underlying all our various and unstable experiences. As such it is indestructible and ultimately unaffected by any specific experience and quite beyond suffering.” (Gethin 1998: 134)

The Buddhist rejection of ātman

The denial of the existence of the ātman or any other form of self – the doctrine of anātman – is universally within Buddhism presented as a key feature of the early teachings of the Buddha. Yet, as Paul Williams points out, the denial of the self does not appear in the Dhammacakkappavattana Sutta, which purports to contain the Buddha’s first sermon, in which he sets out what he discovered on reaching enlightenment. Tradition holds that the teaching on anātman came immediately afterwards in a discourse known as the Anattalakkhana Sutta (Williams 2000: 57). The latter text contains the Buddha’s teaching of the five skandhas: the five “aggregates” which combine to form the sentient being: rūpa (physical form); vedanā (sensation); samjñā (perception); samskāra (mental formations) and vijñāna (consciousness). In this teaching, the Buddha points out that no one of these five aggregates can itself constitute the ātman for, if it did, it would not lead to affliction and would obey the person of whom it is indeed the self. As we know that each of these five can, and does, cause affliction and as we know that each of the five is to some degree beyond our control, it follows that no one of them can have the qualities of an unchanging, indestructible self.

The teaching of non-self ties in closely with another essential element of the Buddha’s early teachings, namely that of the impermanence of all things. As we have already seen, each of the five skandhas contains within it the quality of impermanence – each of them is in a constant state of change from moment to moment. For that reason also, none of them can of itself constitute the self.

It is important to note here that the Buddha does not deny the existence, at least in conventional terms, of the individual person. To do so would make no sense, as it is apparent to all that the world is populated by a vast number of beings, all of whom have some form of distinct identity and can, if we wish, be given a name. The Buddha’s point in these core teachings is rather that each of those beings is comprised of a combination of the five skandhas, a combination which is in a constant state of flux. The name which is given to that combination of skandhas is no more nor less than a conventional and convenient way of referring to that particular combination. The most frequently cited teaching to illustrate this is from the Milindapanha, a dialogue between King Milinda and the sage Nāgasena. The relevant part of the dialogue begins with Milinda asking Nāgasena his name, and receiving this reply:

“I am known as Nāgasena; fellow monks address me… as Nāgasena. But although (my) parents gave the name of Nāgasena … yet it is but a general term, designation, and a common usage. For there is no permanent person present here.” (Olson 2005: 36)

Questioned further, Nāgasena goes on to deny – in accordance with the Buddha’s teachings on anātman – that he is any one of material form, feelings, perceptions, tendencies or consciousness, leading Milinda to the apparently nihilistic conclusion that “there is no Nāgasena”. Nāgasena, however, pointed out to the king that:

“If you… walk… on the dry ground and hot sand, trampling on sharp gravel and pebbles and sand, your feet would hurt you, your body would experience pain, your thought would be impaired, and tactile consciousness would arise, accompanied by anguish.” (Olson 2005: 36)

In other words, the conclusion that there is no person cannot be correct. To complete the teaching, Nāgasena provides the king with the analogy of the king’s chariot, pointing out that, while none of the individual parts of the chariot can be said to be the chariot, it would be wrong to deny the existence of the chariot, but that “chariot” exists “as a denotation, designation, as a common usage, as a mere name”. In the same way, while none of the individual skandhas can be said to be Nāgasena, Nāgasena exists “as a denotation, designation, as a common term, as merely a name. But according to the highest meaning the person is not present.” (Olson 2005: 37)[note]The use of the chariot image here may well not have been accidental. In a number of  places the Upanishads use the image of the chariot when discussing the self.[/note]

This distinction between “conventional” and “higher” or “ultimate” meaning or truth is an important one to keep in mind when analysing existence in Buddhist terms. While – like Nāgasena or the chariot – beings and things may apparently exist and have a degree of continuity, that existence and continuity only operate at the conventional level – in other words, as convenient shorthand – and do not necessarily reflect the actual, ultimate reality. As we shall see, Buddhist thinkers developed a number of ways of analysing that ultimate reality, all within – or at least ostensibly within – the parameters of the core teachings of (a) impermanence; (b) that the quality of saüsāric existence is suffering and (c) anātman.

Classification into the five skandhas is just one of many ways in which Buddhists sought to classify sentient existence. Early Buddhists, for example, introduced the idea of the 12 āyatanas (the five senses and the mind, together with their respective objects) and the 18 dhātus (the 12 āyatanas together with the five kinds of sense consciousness and the mind consciousness). A little later, the Abhidharma schools came to classify existence by reference to dharmas, a word perhaps best translated as “elements”: in other words, the basic building blocks (both physical and mental) of existence incapable of further sub-division. For the Abhidharma schools, these dharmas are either “conditioned” – in other words within the realms of samsāra – or “unconditioned” – within the realm of liberation or nirvāna. Within the Abhidharma schools, the Sarvāstivādins identified 72 dharmas and the Theravādins 81. Each sentient being is composed at any given time of a combination of these dharmas, but no one dharma is capable of existing alone, so no single dharma can have the qualities of ātman. Similarly, to avoid any suggestion of permanence, conditioned dharmas flow in a constantly changing stream of momentary phases. As we shall also see, the nature of the dharmas themselves and of the way they flow eventually became a matter of debate between different schools trying to resolve the conundrum of continuity.

The question of continuity

The combination of the three basic characteristics of existence (impermanence, suffering and non-self) presents an immediate dilemma. If there is no permanence and no self, how does the conventional person continue in existence without a constant and apparently random continual change of physical and mental identity – in simple terms why am I short, blond, male and relatively placid every day and not tall, dark, female and mentally highly agitated some days? If there is no permanence and no self, how does the law of karma operate to determine the quality of my rebirth? What is it within me that stores the results of my actions and what is it within me that transmigrates to the next life? If there is no permanence and no self, how can basic norms of ethics and morality operate (or, as Rupert Gethin puts it, “If I am not the same person who robbed the bank yesterday, how can I be held responsible?” (Gethin 1998: 140))? Ultimately, what is it within me which reaches nirvāna? In other words, if Buddhist teaching denies eternalism (the existence of an essential self), does it by definition become nihilistic or is there indeed a “middle way”?

The basic key to finding this middle way lies in the law of dependent origination, or pratītyasamutpāda. Put briefly, the law of dependent origination teaches the causal interdependence of things in a chain of 12 links[note]The 12 links are ignorance, karmic formations, consciousness, mind and body, the six senses (including the mind), sense contact (birth), feeling, craving, grasping, life, rebirth, old age and death.[/note]. We have seen that each individual being is a conglomerate of components – whether we analyse those components as the five skandhas, the 12 āyatanas or some combination of dharmas. It is the law of dependent origination, or the causal interdependence of one link to another, which explains how those components change in a way which is not entirely random: in other words, why I remain short, blond, male and placid, and why I am older now than I was last year. It is also the law of dependent origination which helps explain how my next life is determined and brought about, though the question of exactly which part of me stores the results of my actions and transmigrates into the next life remained one which continued to vex Buddhists.

The law of dependent origination, however, begs the question of exactly what it is that I remember, and how I learn and retain skills, if all things are impermanent. Put in Abhidharma terms, if the combination of dharmas that existed when I first learned to ride a bicycle no longer exists, how is it that, many years later, I can remember how to ride, even if I have not done so for a long time? Given impermanence and non-self, what is it that stores the results of my actions and how is it that I am reborn?

Continuity in different schools of Buddhism

The generic answer to these two questions is “consciousness”. But what is “consciousness” for these purposes? For the Buddha and the Buddhists, consciousness is a process, not a thing, incapable of existence in isolation, without both an object and its retinue of concomitants (cetasikas), whether ethically good, ethically bad or undetermined[note]The Pāli sources identify 52 cetasikas, a minimum of 7 of which exist at any given moment.[/note]. We have already seen that the consciousness which continues cannot be the consciousness which forms one of the five skandhas, or that one skandha would begin to look uncomfortably like an ātman. And this is really the ultimate dilemma: how can the Buddhist accept karma and samsāra (and the possibility of liberation from saüsāra) if there is nothing to be reborn and therefore nothing to be liberated? So far as rebirth is concerned, the basic early Buddhist position is that cognitive consciousness cognises the external world and interacts with it through the media of the senses and mind, performing volitional activities which generate karma “in the form of potentialities for future rebirths” (Skorupski 2008a: 6). At the time of death, latent (as opposed to cognitive) consciousness performs the three functions of dying, continuing and being reborn in accordance with the karmic potentialities which have somehow been stored. In early Buddhism, the projection of consciousness from one life to the next was considered to happen instantaneously. But, as Skorupski points out, “the continuity of the stream of consciousness and of karmic potentialities has been a matter of speculations” (Skorupski 2008a: 6). Different schools of Buddhist teaching have wrestled with these problems and have come up with a number of creative answers, some of which are analysed below.

1. Sarvāstivāda

For the Sarvāstivādins, the key to analysing existence lay in dharma theory. As we have already seen, for the Abhidharma schools (of which the Sarvāstivādins are one), existence can be analysed into a number (in this case 72) of building blocks, none of which is capable of further division and none of which is capable of independent existence. For the Sarvāstivādins, not only do combinations of dharmas exist, and not only do individual dharmas arise and cease to exist in a very short space of time, but individual dharmas also have their own ultimate existence, and can exist in the past, present or future, with each dharma in its present mode arising from a karmic connection to its past mode and each dharma in its present mode establishing a karmic connection to its future consequences. However, if a dharma can exist in the past, present or future, are we getting close to the suggestion that a dharma has some degree of permanence? The Sarvāstivādin answer was to isolate the notion that each present dharma has its own form of existence, or svabhāva, separate from its past and future existence.

The question of how the results of actions might be stored was dealt with by the Sarvāstivādins by the notion of the dharma called prāpti. The question to be addressed is quite simply that of how to ensure that the karmic result of my volitional action is actually experienced by me, given that, by the time of my rebirth, I will be formed of a totally new combination of dharmas. The Sarvāstivādin analysis of this situation was to see each of my volitional actions as not only having the dharmas associated with the action itself – in other words, the dharmas associated with the physical and mental qualities of the action – but also the dharma of prāpti which, in essence, stored my action and made it “mine”. Of course, as a dharma, prāpti itself must be impermanent and momentary, but this was easily dealt with by the device of a continuous stream of similar prāptis relating to the same volitional action and ensuring that “the karmic result will occur to me because I am the one who has the prāpti series” (Williams 2000: 117).

It was the Sarvāstivāda school which also introduced the ideas of an intermediate existence between life and death, and of some form of intermediate being, or antarābhava, projected from one life to the next by the force of karma. The duration of the intermediate existence has also been a matter for speculation and debate, but “the prevailing theory maintains that it persists for seven days, and that it can be repeated seven times” (Skorupski 2008a: 7), giving a total of 49 days. The idea of the antarābhava derives perhaps from the Buddha’s teaching that conception required the three components of sexual union, fertility and a gandharva, a term later interpreted as meaning some form of consciousness arising from a previous existence. The term gandharva eventually became one of the principal terms used by the Sarvāstivādins for their antarābhava.[note]The term gandharva was also found in earlier Indian religious teaching as a form of celestial spirit musician, and might have had origins in Indian folk religion as well as in Vedic teaching.[/note]

2. Sautrāntika

The defining quality of the Sautrāntika school was its rejection of the Sarvāstivādin notion that dharmas could exist in the past, present or future. For the Sautrāntikas, this implied an unacceptable degree of permanence. For them, each dharma only exists in the moment (kshana) in which it is active, which moment must itself be indivisible into smaller moments, in other words each dharma comes into and out of existence instantaneously. By the time the result of an activity arises – e.g. by the time I hear a sound – the dharmas which created my perception of that sound have long ceased to exist. Or, put another way, it is impossible to step into the same river twice, as, even if it looks like the same river, the water into which I step the second time will not be the water into which I stepped the first time.

As a natural concomitant to this notion of instantaneity, the Sautrāntikas rejected the Sarvāstivādin notion of prāpti. For them, the storing of the karmic effects of action was explained by a constant modification from moment to moment of the saütāna (or stream of consciousness) of the individual. Each volitional action constitutes a seed (bīja) which “initiates a gradual transformation (parināma) of subsequent mind-events which gives rise in the next moment to the fruition of that action” (Sanderson 2004: 42). This transformation is often explained by the metaphor of “perfuming”: each of my volitional actions “perfumes” the whole series of modifications. At death, the final moment of this series of modifications produces the karmic effects in the next rebirth. For some Sautrāntikas, these modifications occurred at the level of the sūkshmacitta, a form of subtle consciousness continuing from one lifetime to the next. (Skorupski 2008b: 158)

Arguably, the Sautrāntika notion of seeds and perfuming evades the issue. While providing interesting metaphors for the process of continuity, the Sautrāntikas did not see the bījas as separate dharmas. That being the case, what was their ontological status? Did they have real existence and, if so, how could one explain their continuity in the stream of consciousness?

3. Theravāda

The Theravādins too rejected the idea that dharmas could exist other than in the present, also seeing in that notion a hint of eternalism. For the Theravādins, the key to continuity lay in the bhavanga, an inactive state of consciousness to which the mind returns, however momentarily, between each active process of consciousness. The bhavanga is the form of consciousness which first arises in the embryo, and is coloured by the karmic results of previous lifetimes. It is also the bhavanga which makes the link between a dying person and his or her next rebirth.

The Theravādins differentiated the stream of consciousness into two related series: the cognitive (vīthicitta) series and the latent (vīthimuttacitta) series (Skorupski 2008b: 160), each of which performs a number of different functions. The cognitive series arises from the bhavanga and lapses into it once the process of perception is complete (Skorupski 2008b: 160). Gethin argues that “for a given being, the bhavaïga is something of a constant throughout a being’s life” (Gethin 1991-3: 28); “a being’s bhavanga is of the same type throughout his or her life” (Gethin 1991-3: 20); and “bhavanga does not simply define what one is, it defines precisely who one is” (Gethin 1991-3: 19). That being the case, how close is the bhavanga to the self?

The important point to remember here is that the bhavanga is a form of consciousness, or citta[note]And, although latent, is not unconscious.[/note]. As we have already seen, consciousness cannot exist without an object or without its concomitants and, as a single dharma, citta must be momentary. The bhavanga therefore cannot be static, but arises and vanishes every moment during the periods when the cognitive forms of consciousness do not occur. Nor can the bhavanga exist in isolation from other dharmas or cetasikas. Even in deep sleep, when the cognitive consciousness does not function, the bhavanga must continue to arise and fall momentarily (Gethin 1991-3: 15).

The Theravādins also rejected the Sarvāstivādin doctrine of the antarābhava, or intermediate state between death and rebirth. For the Theravādins, all that was needed to explain the continuity between lifetimes was the bhavanga which, on death of a person, arose instantaneously in the new conception with all the karmic imprints of the previous life.

4. Pudgalavāda

The Pudgalavādins were a highly controversial school, of which little original material survives, so that most of our knowledge comes from the attacks on their doctrines by other schools (Williams 2000: 124; Gethin 1998: 223), though Cousins describes them as “widely influential and accepted as part of the Buddhist tradition” (Cousins 1994: 17). According to Cousins, the Theravādin text the Kathavattu makes clear that the Pudgalavādins held there to be a “person” (pudgala) with real existence, with the qualities of an additional, indivisible dharma (Cousins 1994: 22). For the Pudgalavādins, the pudgala is that which experiences, which acts, which stores the karmic results of those actions, which transmigrates and which attains nirvāna. It follows, however, that either the pudgala has separate identity as a matter of ultimate truth – in which case, how is it different from the Brahmanical ātman? – or it merely exists conventionally as a conglomeration of the five skandhas – in which case how is the Pudgalavādins’ teaching different from more “orthodox” Abhidharma schools of Buddhism?

The Pudgalavādins’ eventual answer (as represented by Vasubandhu in chapter 9 of the Abhidharmakosha, though as to whether this represented their original position see Williams 2000: 125) was that the pudgala exists neither at the level of ultimate nor conventional truth, and that it is neither the same as, nor different from, the skandhas. It is said to be “undefinable” (avaktavya). In other words, it has the ontological status of a dharma which is neither conditioned (and so destroyed at death) nor unconditioned (for, if it were unconditioned, it would certainly fall into the trap of eternalism).

Williams 2000: 125-7 points out that the Pudgalavādins were “wrestling with genuine philosophical problems”: for ethical and moral teachings to make any sense, there has to be some way in which the same “person” who acted receives the moral and karmic fruits of that action. If that person is simply a conglomeration of the skandhas, why does he or she not simply disappear when the skandhas are destroyed at death? As Williams says: “The Pudgalavādins were constrained almost to the point of absurdity by the language of Buddhist scholasticism… [they] found puzzlement and problems where their fellow Buddhists found clarity and simplicity.”

5. Madhyamaka

When we come to the Mahāyāna schools, we find less emphasis on the analytical approach of the Abhidharma traditions outlined above. For the Madhyamikas, the key concept was the emptiness (shūnyatā) of all dharmas. Every existent thing is merely a conceptual construct existing at the level of conventional truth (samvrti-satya) without any ultimate self-existence.

“Even nirvāna… is like a magical illusion, is like a dream. How much more so anything else” (Ashtasāhasrikāprajñāpāramitā, translation Conze).

It follows, therefore, that if, as the Madhyamikas contend, all is empty at the level of ultimate truth (paramārtha-satya), the question of what continues and how is largely irrelevant. For the Madhyamikas, as for others of the Mahāyāna tradition, the more important focus was the cultivation of the wisdom (prajñā) to understand the distinction between ultimate and conventional truth through the Perfection of Wisdom (prajñāpāramitā) and the bodhisattva path. Adherence to theories – even dharma theory – is itself a source of suffering deriving from ignorance: knowledge of the ultimate truth is only accessible through yogic experience, and not by empirical analysis.

6. Yogācāra

However, the Yogācāra school of Mahāyāna, while also professing the illusory nature of all existent things, nevertheless felt that the question of continuity needed to be addressed. For the Yogācārins, the Madhyamaka teaching of emptiness of all things came too close to nihilism. The Samdhinirmocana Sūtra shows the Buddha teaching that “it is a misunderstanding of emptiness to take it as meaning that literally all things are conceptual constructs” (Williams 2000: 153). It follows, therefore, that at least one thing must be something other than a conceptual construct, and, for the Yogācārins, that one thing was to be found in their analysis of the mind, as attempts to analyse material form (rūpa) always foundered on the inability of objects to exist in their own right, other than as a conglomeration of different components.

For the Yogācārins, each existent thing has three natures (svabhāva): dependent nature (paratantra) or conditioned, or conventional, reality; imaginary nature (parikalpita) or the self-existence which is wrongly attributed to the dependent nature; and absolute nature (parinishpanna) or the complete absence of self-existence of the dependent nature (Jackson 1997: 339). In other words, each existent thing lacks what Jackson describes as a “self-sufficient externality” (Jackson 1997: 339), and so is effectively illusory and no different from the consciousness which perceives it.

The early Buddhist analysis of consciousness had focussed on the six-fold classification of the five senses and the mind itself, together with their respective objects and, sometimes, their respective forms of consciousness (e.g. eye-consciousness, ear-consciousness etc)[note]See page 6.[/note]. The Yogācārins broadly adopted this classification, but analysed the consciousness much more deeply. For them, the active workings of the mind continued to comprise the awareness of the data presented to the mind by the five senses and the conscious thoughts. However, underlying these six forms of consciousness lie two more layers: the first (klishta-manas) being the “defiled mind”, afflicted with the defilements of thoug