5  Mahāyāna Philosophy

The Perfection of Wisdom (Prajñāpāramitā)

As far as we can tell at the moment, the earliest specifically Mahāyāna literature consists of sūtras of the Prajñāpāramitā-type. Since these are Mahāyāna sūtras they thus claim a disputed status as the word of the Buddha. Within India itself the status of Mahāyāna sūtras was always disputed. The circulation of such sūtras was likely to have been much more a matter of individual and small-group activity (carried as treasures by individual wanderers, for example) than the activity of the Sangha of a Vinaya tradition as a whole. Moreover while Indian travellers wandered into Central Asia and thence to China, the wandering was not all one-way. Monks and nuns did indeed leave India for Central and Southeast Asia and China by sea and land routes, no doubt sometimes carrying with them their precious scriptures. These may have included certain scriptures central to the life of the monk or nun concerned but not considered authentic by the wider Sangha community. But of course we know that monks and nuns also came into the Holy Land of India from abroad. Jan Nattier has argued with considerable plausibility that perhaps the most popular Mahāyāna Prajñāpāramitā text of all, the short Heart (Hṛdaya) Sūtra, was actually in the form we have it now as a sūtra, originally a Chinese work abstracted and compiled from a Chinese translation of a much larger Prajñāpāramitā text. It may then have subsequently and successfully been introduced into India itself, probably by the Chinese pilgrim Xuanzang, and translated into Sanskrit (Nattier 1992). The same introduction from outside India has been claimed for the main sūtra of Bhaiṣajyaguru, the Medicine Buddha (Birnbaum 1980: 52 ff.). This sūtra was well known enough in India to be quoted by the great Indian scholar and poet Śāntideva in the early eighth century but could easily have been introduced into India at an earlier date.

The Prajñāpāramitā literature is large and repetitive. Sūtras are commonly named by the number of verses (the ‘Eight-thousand Verse Perfection of Wisdom’, Aṣṭasāhasrikā Prajñāpāramitā, for example). There is some agreement among scholars that from an original core the basic text (or idea) was expanded as far as the Hundred-thousand Verse, and then contracted into shorter ‘summary sūtras’. The earliest version does indeed appear to be the Aṣṭasāhasrikā, although what Edward Conze (1960) sees as its verse summary, the Ratnaguṇasaṃcayagāthā, may be the very earliest form of the text. The Pañcaviṃśatisāhasrikā (‘Twenty-five Thousand Verse’), also extremely important, is probably a little later. There is some disagreement as regards the dating of what is one of the most popular Prajñāpāramitā texts, the famous Vajracchedikā Sūtra (the ‘Diamond’, or ‘Diamond Cutter’, otherwise known as the ‘Threehundred Verse’). Edward Conze would take as an example of a later summary sūtra (see Williams 2009: 47–9).¹

One has to be very careful in outlining the core message of the Prajñāpāramitā literature not to explain it too fully. These are sūtras not systematic philosophical or doctrinal (‘Buddhalogical’) treatises. They are very clear that they have a message – a message that is repeated again and again – and that message is one of criticism and exhortation. It is a message of inspiration, perhaps the message of a Dharma-preacher (the dharmabhāṇaka), rather than the message of a philosopher or doctrinal theorist. The message is a message of exhortation to their fellows in the ‘non-Mahāyāna’ world. But when we seek to explain more fully, in a more systematically rational manner, the sermon of the Perfection of Wisdom we almost inevitably find ourselves employing the language and perspective of Mādhyamika philosophy or, in explaining the path of the bodhisattva, language of much later path-structures. This systematic explanation we shall come to very soon, but for the moment let us just heed the sermon itself. It has three principal themes, repeated and illustrated again and again as if to seep into the deepest recesses of the Good Buddhist’s mind. The first theme is that of the very peak, the perfection (pāramitā), of wisdom (prajñā). Its content is emptiness (śūnyatā). And its context is the path and practices of a bodhisattva, one whose aim is not just enlightenment (obtained by Arhats), but Perfect Buddhahood for the benefit of all sentient beings.

The Perfection of Wisdom (Prajñā)

Broadly speaking, prajñā is the state of mind that comes from properly understanding something. In Buddhism, as a technical term, it is used primarily for that understanding which sees how it really is in contrast to the way things appear to be. Just as that seeing can both be a matter of understanding that things are really like this, and also actually being in a state of mind where one sees directly and immediately how it is, so we can refer to different levels of prajñā, from understanding to nonconceptual insight. We have seen already that Buddhist thought was from the beginning marked by a distinction between the way things appear to be and the way they actually are. It should thus come as no surprise to find that within the framework of Abhidharma prajñā is used to refer to discernment of the ultimate primary existents. As we have seen, they are to be distinguished from conceptual constructs. Therefore prajñā refers to the discrimination of dharmas.² The Prajñāpāramitā literature refers frequently to not discriminating dharmas, but its message is nevertheless encapsulated within the specific Abhidharma project and it is within the Abhidharma framework that we must understand the expression ‘perfection’ of prajñā. The Perfection of Wisdom speaks not of the wrongness of what had previously been considered to be wisdom, but rather of its perfection. Previous wisdom is indeed wisdom, but it is imperfect. Mahāyāna texts will treat a series of ‘perfections’ (pāramitā) mastered by the bodhisattva. The common list is six – the perfections of giving (dāna), morality (or ‘precepts’; śīla), endurance (ksānti), exertion (vīrya), meditative concentration (dhyāna), and finally wisdom (prajñā) – but the perfection of wisdom is primary, and is said to lead the others as a man with eyes leads those who are blind.

The perfection of prajñā is the final prajñā, the final proper understanding of the way things truly are. But it seems to me that the perfection of prajñā – although it is stated with such missionary zeal in the Prajñāpāramitā texts – is in so many ways an affirmation of what has gone before. If, as we shall see, all dharmas are empty, lacking intrinsic nature, then that is merely to confirm their own status as ontologically no more than tables, persons, or forests. There is no question that tables, persons, and forests are conceptual existents, empty, just as the previous Abhidharma thinkers had shown. All are agreed on that.

Emptiness (Śūnyatā)

What is immediately apparent to anyone who glances at a Perfection of Wisdom text is the endless list of things that are said to be ‘empty, like a magical illusion’. This is indeed the principal philosophical teaching of the Prajñāpāramitā literature, and it was only possible because of the Abhidharma framework that we have examined previously. Buddhism from the very beginning had used the terms ‘empty’ (śūnya) and ‘emptiness’ (śūnyatā) to apply to the truth discovered by the eye of proper understanding (prajñā), the eye of the Buddha. First this was with reference to the five aggregates empty of Self or anything pertaining to a Self. Then it was applied to the whole list of 75, 82 or however many classes of fundamental constituents (dharmas) discovered by the different Abhidharmas likewise to be empty of Self or anything pertaining to Self. In addition, the term śūnya was also used in the Abhidharma (perhaps by an almost imperceptible shift in meaning) to refer to the nature of secondary, conceptual, existents empty of any status other than conceptual existents, empty of their own ‘intrinsic nature’, empty of primary, irreducible, existence. Persons, tables, forests and so on are empty of Self, but they are also empty of irreducible primary existence. Here these amount to the same thing. But while there is no Self at all, and all things are empty of Self, for the Abhidharma there must exist some things which have primary existence, and secondary conceptual existents are themselves empty of that primary existence which is, of course, possessed by primary existents, dharmas. The absence of Self cannot mean there are actually no primary existents at all.

But in the Prajñāpāramitā literature the same term śūnya is used to refer to absolutely everything, and it entails that absolutely everything is ‘like a magical illusion’.³ We need to be quite clear about the range of this claim, for there are scholars (such as Conze) who would want to limit it and argue for some sort of monistic Absolute – a primary existent, an Ultimate Reality par excellence – behind the Prajñāpāramitā negations. But the Aṣṭasāhasrikā (Eight-thousand Verse) is quite unequivocal:

Even Nirvana, I say, is like a magical illusion, is like a dream. How much more so anything else! … Even if perchance there could be anything more distinguished, of that too I would say that it is like an illusion, like a dream.

(Aṣṭasāhasrikāprajñāpāramitā, trans. Conze 99)

In other words absolutely all things have the same status as persons, tables, and forests.⁴ They are all conceptual constructs and therefore cannot be vested with intrinsic nature. Crucially, they therefore cannot be grasped. One cannot substitute grasping after tables and so on with grasping after dharmas as the refuge, the fixed point in a world of disappointment and suffering. Thus the classical earlier Prajñāpāramitā literature constantly asks what is referred to by the term x, what dharma this is, with the response that nothing can be found, nothing can be grasped, and yet the bodhisattva should heroically resist all fear. To see otherwise is to grasp, and to grasp is to miss enlightenment. Thus enlightenment comes from ceasing to grasp even the most subtle sources of attachment, and this ceasing to grasp requires seeing those things which could serve as sources of attachment as empty, mere conceptual constructs. Nothing must be grasped. All things are empty. On the level of what is an ultimate, primary existent there is nothing, for there are simply no ultimate, primary existents. On such a level therefore there is an endless absence, an endless emptiness. Thus to think that dharmas have primary existence is to grasp. As an exhortation this is an appeal to complete letting-go. For both philosophical reasons and also perhaps existential reasons this teaching of emptiness may for some have been terrifying. It certainly looks like nihilism, and it encourages a deep letting-go in meditation that could indeed be the true spiritual equivalent of the going-forth that the monk underwent in leaving family, friends, and village. At least, that is the impression one gets from the texts. Yet emptiness is also the antidote to fear, a fear which in its frequent mention must have been some problem for Buddhists at this time.⁵ For if all is empty, what is there left to fear? And who is there left to do the fearing?

The Bodhisattva

The Perfection of Wisdom literature itself does not claim that to see all things without exception as empty is some special teaching for followers of the Mahāyāna. Since any other perspective would involve grasping, it follows that it is necessary to see emptiness in order to attain cessation of grasping and therefore in order to attain any state that could be called one of enlightenment. As the Aṣṭasāhasrikā puts it,

whether one wants to train on the level of Disciple, or Pratyekabuddha, or Bodhisattva, – one should listen to this perfection of wisdom, … and in this very perfection of wisdom should one be trained and exert oneself.

(Aṣṭasāhasrikāprajñāpāramitā, trans. Conze 84)⁶

Thus as far as the Aṣṭasāhasrikā is concerned it is not possible finally to attain complete cessation of grasping as long as one sees some sort of contrast between the ontological status of dharmas and that of tables, persons, and forests. That is, one cannot cease craving as long as one sees an actual, real, ontological distinction between primary existents and secondary conceptual existents. Under such circumstances it would therefore not be possible to attain any degree of enlightenment. Note, however, that it does not follow from this that there is anything wrong as such with analysis into dharmas. To know that tables are conceptual constructs, as is taught by those who would analyse into dharmas, would I think have been taken as knowing what is correct, and as indeed contributing to letting-go of attachment for tables. The problem however lies in considering that there is some fundamental ontological contrast with dharmas themselves, which might make dharmas suitable objects of attachment.

Although insight into complete emptiness is necessary for any enlightenment, Prajñāpāramitā sūtras constantly contrast the aspiration of their hero or heroine, the bodhisattva, with that of one who aims for a lower goal, which is to say one aiming to become an Arhat. The idea of a bodhisattva is not new to the Mahāyāna. The bodhisattva is accepted by all Buddhist traditions as one who has seriously taken a vow, properly speaking in the presence of a previous Buddha, to follow the path to Buddhahood, and who is striving to live in accordance with that vow. Thus the person who was to become Śākyamuni Buddha also took a vow to become a Perfect Buddha many aeons ago in the presence of a previous Buddha. There is however a problem here. Presumably Śākyamuni actually could have attained enlightenment for himself (the state of an Arhat) in the presence of that previous Buddha. Why did he undergo the many, many rebirths necessary in order to follow the path to Buddhahood if the eventual goal of Buddhahood is not qualitatively different to – not in some significant way very much superior to – the state of an Arhat? We are told that he undertook the long path to Buddhahood out of compassion, in order to be able to help others more effectively – but why? Clearly it could not have been for some additional quality that would be of benefit to himself, since there is held to be no greater fulfilment for oneself than freedom from suffering, nirvāṇa, the state of an Arhat. Thus Śākyamuni in his previous life must have taken the vow of a bodhisattva solely out of altruism. This is, however, absurd if there is nothing about being a Buddha which is qualitatively superior, indeed spiritually superior, to being an ordinary Arhat. But if there is something qualitatively superior, it can only be described in terms of altruism, since there is nothing left for the Buddha to gain for himself beyond becoming an Arhat. And if this Buddhahood is qualitatively superior, then those who do not attain an altruistic Buddhahood must be missing out on the highest spiritual goal.

Thus it seems to follow that Śākyamuni Buddha must have taken the vow to become a Perfect Buddha out of altruism, and the state of a Perfect Buddha must be qualitatively superior to that of nirvāṇa, the state of an Arhat. That superiority must lie in its altruism. Therefore the bodhisattva vows, in the words of the Aṣṭasāhasrikā:

My own self I will place in Suchness [the true way of things], and, so that all the world might be helped, I will place all beings in Suchness, and I will lead to Nirvana the whole immeasurable world of beings.

(Aṣṭasāhasrikāprajñāpāramitā, trans. Conze 163)

And yet the Perfection of Wisdom literature repeatedly states that the bodhisattva takes such great vows without perceiving that there is any actual being who is saved, for all is really empty. This is truly the Perfection of Wisdom.

Perhaps anyway it was some such reasoning as this that led to the exaltation of the way of the bodhisattva in Mahāyāna literature as a final spiritual career for all who can aspire to it. Whereas in non-Mahāyāna sources accounts of the bodhisattva career are descriptive – this is what great ones have done in the past and perhaps some (like a bodhisattva known as Maitreya) will do in the future – in Mahāyāna sources they eventually become prescriptive. All who can should finally take upon themselves the vow of the bodhisattva, to attain Perfect Buddhahood for the benefit of all sentient beings. Any other goal is relative to this great goal inferior, i.e. that of a Hīnayāna. Indeed in certain later strata of the Perfection of Wisdom and other Mahāyāna literature we find accounts of the great abilities of those well advanced on this bodhisattva path to help others. Practitioners of this path are encouraged to gain what appear to be miraculous powers in order to benefit those who petition them. Very advanced, perhaps even nearly enlightened, bodhisattvas have immense abilities to help. Thus they become what have been called ‘celestial bodhisattvas’, such as Avalokiteśvara – particularly associated with compassion – or Mañjuśrī, the bodhisattva of wisdom, who have the ability to help, the wish to help, and great amounts of merit due to their immense good deeds done in the past. This is merit that they are happy to give away to others who are poor in their ‘roots of merit’ and are thus suffering the many pains of saṃsāra.⁷

Note – and this is important – that in the light of all of this it is far too simplistic to speak merely of ‘nirvāṇa’ within the context of Mahāyāna Buddhism. There is the nirvāṇa associated with the goal of becoming an Arhat; there is the state of a pratyekabuddha; and there is also the full enlightenment associated in Mahāyāna with Perfect Buddhahood. It is common in Mahāyāna texts to refer to nirvāṇa contrasted with saṃsāra and indicate that the bodhisattva, and eventually a Buddha, in attaining freedom from suffering but not abandoning those in saṃsāra, is beyond this duality of nirvāṇa versus saṃsāra. The state of enlightenment attained by a Buddha is thus called a ‘non-abiding’ or ‘unrestricted’, or ‘not-fixed’ nirvāṇa (apra-tiṣṭhitanirvāṇa). It therefore becomes very problematic indeed to portray the bodhisattva, as do so many books available in the West, as postponing nirvāṇa. What nirvāṇa is he or she supposed to postpone? Clearly not that of an Arhat or a pratyekabuddha, since these are said to be inferior nirvāṇas. And it is misleading to think that the bodhisattva, who is portrayed as wishing to attain Buddhahood for the benefit of all sentient beings, literally postpones Buddhahood. There is no reason for a bodhisattva to postpone the apratiṣṭhi-tanirvāṇa. If this is what bodhisattvas do of course then any Buddha would have broken his or her vows. The Buddha would have to be taken absurdly as being in some sense deficient in compassion! We saw from the quote above taken from the Aṣṭasāhasrikā that the bodhisattva wishes to attain Buddhahood precisely in order better to help others. This is not to say that a text may not exhort the bodhisattva even to abandon Buddhahood as a selfish personal goal, in order to concentrate more fully on helping others. But the result of this is that one attains Buddhahood all the more quickly. This is indeed the true way to Buddhahood. Any alternative would be paradoxical.⁸

Mādhyamika

By ‘Buddhist philosophy’ here I mean in the broadest sense discussions, speculations and arguments concerning ‘seeing things the way they really are’ (yathābhūtadarśana). Just as Mahāyāna Buddhism as a whole can best be seen as a vision, an aspiration, within a Buddhism which therefore in itself is non-Mahāyāna, mainstream Buddhism, so I think Mahāyāna philosophy should be understood as a particular expression of and response to Buddhist philosophy as a whole. The name for Buddhist philosophy as a whole, it seems to me, is ‘Abhidharma’, in the sense that Abhidharma sets the agenda, the presuppositions and the framework for Buddhist philosophical thought, understood in the way in which I have delineated it. Philosophy here developed within Abhidharma discussions among what were probably a minority of elite monks. Mahāyāna philosophy, far from representing a negation of the approach of the Abhidharma, is best seen as a series of strategies within the Abhidharma enterprise.

The use of ‘Mādhyamika’ (‘Middling’) as the name for this Mahāyāna school of Buddhist philosophy and its characteristic philosophical position appears to be lacking in our earliest Mādhyamika sources. It cannot be found at all in the philosophical works of Nāgārjuna (probably second-third centuries CE), for example, usually thought of as the founder of Mādhyamika. Nāgārjuna does indicate nevertheless that he considers himself to hold to a distinctive position by employing the expression śūnyatāvādin, one who holds to the position of emptiness Vigrahavyāvartanī v. 69). Of course, Buddhism from the beginning had referred to itself as a ‘Middle Path’. For a Mādhyamika, however, the principal significance of claiming to be the follower of the Middle Path par excellence is an understanding of emptiness (śūnyatā) as the middle between eternalism and annihilationism. This understanding undoubtedly can be traced back to universalisation of the idea of dependent origination (pratītyasamutpāda) as the middle between holding to the eternal existence of an unchanging Self, and holding to annihilation at death. Thus it should come as no surprise to find in Mādhyamika sources emptiness being equated with or derived from dependent origination (Madhyamakakārikā (MMK) 24:18). Because even dharmas originate due to causes and conditions they too must be empty of primary, substantial existence.

Therefore the way by which Nāgārjuna and his followers sought rationally to demonstrate the Perfection of Wisdom claim that all things without exception are akin to illusions was through showing that all things are without intrinsic nature (all things are niḥsvabhāva, they are secondary existents, conceptual constructs). They are this way because they are the results of causes and conditions, they are dependently-originated. In general terms a Mādhyamika in Buddhist philosophy is a śūnyatāvādin, one who holds to the absolute universality of emptiness (śūnyatā), i.e. absence of intrinsic nature (svabhāva). He or she is therefore one who holds and sets out to demonstrate that absolutely everything is nothing more than a conceptual construct.⁹ The founder of this approach is always said to be Nāgārjuna, although there is some reason to think that arguments associated with Nāgārjuna may have been extant before his time.

I do not wish to say very much here about the legendary life of Nāgārjuna, or that of his great disciple Āryadeva. Modern scholars do not accept the traditional Tibetan account which would have Nāgārjuna living for some 600 years and becoming both the great Mādhyamika philosopher as well as a tantric yogin and wonderworker (siddha). They prefer to speak of at least two Nāgārjunas. There is also great debate over which works can be attributed authentically to the philosopher Nāgārjuna.¹⁰ As far as Mādhyamika philosophy is concerned, however, there is some agreement that the following complete works might reasonably be attributed to the Master:

i)  The Madhyamakakārikā (‘Verses on Mādhyamika’, perhaps originally simply called ‘Wisdom’ (Prajñā), commonly abbreviated as MMK) – Nāgārjuna’s main work, still extant in Sanskrit;

ii)  The Vigrahavyāvartanī; verses extant in Sanskrit together with an autocommentary – a reply by Nāgārjuna to his critics.

Save for a few fragments the following works survive only in Tibetan and sometimes Chinese translation:

iii)  Yuktiṣaṣṭikā (‘Sixty Verses on Reasoning (yukti)’);

iv)  Śūnyatāsaptati (‘Seventy Verses on Emptiness’);

v)  Vaidalyaprakaraṇa – attacking the categories of the Hindu Nyāya epistemologists;

vi)  Ratnāvalī (‘The Jewel Garland’, a long epistle apparently to a king. A shorter royal epistle also attributed to Nāgārjuna is the Suhṛllekha);

vii)  The Catuḥstava – four hymns also reasonably reliably attributed to Nāgārjuna.

As far as Nāgārjuna’s disciple Āryadeva is concerned, by far his most important work is the Catuḥśatakakārikā (‘Four Hundred Verses’), extant in Tibetan and Chinese translation.

It is not obvious what it means to talk of sub-schools of Mādhyamika in India. Indeed it is not clear how far Mādhyamika thought in India was very influential or taken very seriously. Possible exceptions are the so-called Yogācāra-Svātantrika Mādhyamika from the eighth century onwards, and perhaps Candrakīrti’s tradition of Mādhyamika in the eleventh century. This was long after his own death in the seventh century.¹¹ All Mādhyamikas hold to complete absence of primary, substantial existence, and thus to the universality of conceptual construction. Tibetan scholars, recognising that there were nevertheless differences and debates in India on various issues between Mādhyamikas, subsequently divided Mādhyamikas into Svātantrika and Prāsaṅgika Mādhyamika, dividing again Svātantrika into Sautrāntika-Svātantrika and Yogācāra-Svātantrika. Even so, there was some dispute as to exactly what distinguished thinkers of each sub-school. To take just one influential perspective, if we could follow the view of the Tibetan Tsong kha pa (fourteenth-fifteenth centuries CE) then a Svātantrika Mādhyamika such as Bhāvaviveka (c. 500–70) holds that ultimately all things lack intrinsic nature (svabhāva). According to Tsong kha pa he appears not to hold this to be the case conventionally, from the pragmatic everyday point of view. For Candrakīrti the Prāsaṅgika, on the other hand, the very notion of intrinsic nature, a svabhāva, is contradictory on either ultimate or conventional level. There is sometimes said to be a further difference between Svātantrika and Prāsaṅgika Mādhyamika as regards whether it is necessary to employ the proper logical form or ‘syllogistic’ structure of reasoning derived from the canons of Indian logic in order to refute the primary existents held by the opponent. The Svātantrikas are said to have argued that one should. For the rival Prāsaṅgikas it is sufficient to employ any reasoning at all which indicates a contradiction in the primary existents of the opponent and which he or she finds convincing, such as the use of favoured prasaṅga arguments, a kind of reductio ad absurdum, showing the opponent on grounds that need only be acceptable to the opponent himself that his position must entail for the opponent himself absurdities. Moreover a Yogācāra-Svātantrika like Śāntarakṣita (seventh—eighth century) is held to differ from the Sautrāntika-Svātantrika Bhāvaviveka (not to mention the Prāsaṅgika Candrakīrti) particularly on the issue of the status of objects of experience. While all agree as Mādhyamikas that all things are merely conceptual constructs, Śāntarakṣita seems to have held that those conventional conceptual constructs are all of the nature of mind. In other words subject and object in all experience are of the same fundamental nature, which is in some sense mentalistic. In this respect Śāntarakṣita and his followers are like the Yogācāra (q.v.) school (cittamātra – ‘mind-only’), although they are held to differ in denying that this mentalistic stuff (‘consciousness’, ‘mind’) is itself a primary existent, with its own intrinsic nature. Bhāvaviveka does not hold that subject and object are of the same stuff, and in this respect he is said to be similar to the Sautrāntika approach to these issues.¹²

It seems to me that the key to understanding Mādhyamika lies in a proper appreciation of the significance identifying emptiness and dependent origination had within the context of the Abhidharma distinction between substantial primary existence and secondary conceptual existence.¹³ Thus I want to argue that Mādhyamika represents a strategy within an Abhidharma debate, an affirmation of the Abhidharma analysis as far as it goes combined with a claim to detect a contradiction in any ontological distinction – any distinction regarding the way things really exist – between primary and secondary existence. Certain followers of Abhidharma had not thought through far enough the full ontological implications of the Abhidharma project. We have seen that to be a conceptual existent is to be capable of being dissolved away under a particular sort of critical analytic investigation. That investigation is an investigation that searches to find if X is the sort of thing that has existence in its own right. In other words it searches to find whether X can or cannot be dissolved into component parts that, as it were, bestow the existence of X upon it when conceptualised in a particular sort of way. Later Tibetan thinkers would refer to this sort of search as an ‘ultimate investigation’, a search to find out if X has ultimate (i.e. primary) existence or not. The existence of a table is a particular way (for particular practical purposes) of conceptualising the top, legs and so on. Thus a conceptual existent does not have its existence contained within itself. It does not have its very own unique intrinsic nature, svabhāva. Its existence as such is given to it by conceptual construction. Thus it is niḥsvabhāva, lacking intrinsic nature. Therefore in Mādhyamika philosophy a particular sort of analysis is carried out, an analysis which investigates each of the categories held by opponents to contain entities possessing intrinsic nature in order to see whether those entities can be dissolved under this ultimate analysis.

Note also that in previous Abhidharma terms to have a svabhāva is to have a particular type of existence, an existence that cannot be dissolved away into component parts. It is therefore to have an existence that is not thought to be the result of the conventional conceptualising process. Not to have a svabhāva is still to exist, but that existence is contingent upon a conventional practical conceptualising activity, an actual existence nevertheless guaranteed by reducibility into primary existents. The contradiction considered within this model and at the heart of Nāgārjuna’s demonstration of the truth of the ontology of the Prajñāpāramitā sūtras lies in the implication relationship he draws between dependent origination and emptiness. For a previous follower of the Abhidharma, to be a conceptual existent is clearly to be the result of causes and conditions, and notably the conceptualising process. But to be a primary existent is also, in the main, to be the result of the causal process as well.¹⁴ What justifies the primary existent is that it is an irreducible into which the secondary existent can be analysed. It is that which must be. Thus to be a primary existent is in most cases by no means to be unrelated to causes and conditions. Yet, Nāgārjuna wants to say, that is absurd. A secondary existent cannot be found to have existence in its own right because it can be reduced to primary existents. If that is so then it’s very being as a secondary conceptual existent is granted to it through existence being bestowed upon it by nothing other than its causes and conditions. The concept of the svabhāva must therefore reduce (slide?) from that of intrinsic nature when contrasted with constructed nature to that of inherent, or intrinsic, existence, i.e. selfcontained existence, existence that is not bestowed upon it at all from outside. But inherent or intrinsic existence is an equivalent of existing from its own side, and hence quite independently of any causal and conceptualising process. On the other hand it should be quite clear that anything that is the result of causes and conditions must of course lack its own intrinsic existence.

Thus while there may still be a relative distinction between ‘primary’ and ‘secondary’ existents (a table can still be analysed into parts), anything which is the result of causes and conditions must be niḥsvabhāva, that is, lacking svabhāva and hence empty (śūnya) of it. If X, whatever it is, cannot be found when searched for under the sort of analysis that is investigating the (ultimate) existence of X, then X is empty. If X is the result of causes and conditions – particularly if it can be shown to be the result of conceptualisation – then X is empty. Hence Nāgārjuna applies analytic investigation to some of the principal categories of Buddhist thought as well as those of non-Buddhists, such as causation itself, movement, time, the Buddha, nirvāṇa, as well as the Self. He declares emptiness whenever anything is found to be the result of some sort of causal process. In this way the assertions of complete emptiness ‘like an illusion’ in the Prajñāpāramitā sūtras can be demonstrated through analytic investigation. It is shown through using dependent origination in the most impeccable Buddhist, even Abhidharma, manner.¹⁵ As Nāgārjuna puts it:

Whatever comes about conditioned by something else is quiescent from the point of view of inherent [intrinsic] existence. Therefore both the process of origination and the act of production itself are quiescent. … Like an illusion, a dream or a castle in the air are production, duration and cessation declared to be.

(MMK 7:16/34)

Nāgārjuna’s approach, therefore, is to take a category held by others to be capable of withstanding analysis – such as causation itself (in the famous MMK Ch. 1) – and analyse it. This is the sole concern of the Mādhyamika, to analyse the positions of the opponent, not to put forward counter-positions which might entail something of their own capable of resisting analysis. Hence Nāgārjuna’s famous statement on Vigrahavyāvartanī 29 that he has no thesis (pratijñā) to prove. How can there be causation, since it cannot be rationally explained between a cause and effect that are same. This is because such causation would be pointless, and also it would lead to an infinite regress because when the effect came into existence the cause would reoccur and hence the effect would also arise once more, and so on. Causation also cannot be demonstrated where cause and effect are intrinsically different, since then there would then be no actual connection between the two and hence no causal connection. Cause and effect could not be both same and different, for such a position would be subject to both sets of problems. Nor could there be an effect coming from no cause at all, because then production would be random, and without a proper causal connection intentional action would become pointless.¹⁶ The Self cannot be the same as the five aggregates, taken separately or together, nor could it be different from them. And so on.

Note that in Abhidharma terms to find that X does not resist ultimate analysis (i.e. it is empty of intrinsic nature, niḥsvabhāva) should not be taken as entailing that it does not exist at all. No follower of the Abhidharma maintains that tables and chairs do not exist. They can still be used for everyday purposes, and that usage is quite correct. Thus Nāgārjuna can argue in his Vigrahavyāvartanī that the fact that his own words lack intrinsic existence does not entail that they cannot carry out their function of refutation. His position is not contradictory in that sense. Once it is appreciated that emptiness is an implication of dependent origination and is by no means identical with non-existence it can be seen that for something to be empty implies that such a thing must in some sense exist, since it must have originated through some sort of causal dependence.

Emptiness here is not some kind of Absolute Reality approached perhaps through a ‘way of negation’ (via negativa). It is the very absence (as such, a pure non-existence) of intrinsic existence in the case of X, whatever that X may be, which is the result of X’s arising due to causes and conditions. If a table is empty it is because it has come into existence in the dependent way that tables come into existence. The table is empty of intrinsic existence, and that quality, that complete absence of intrinsic existence, possessed by the table is its emptiness. In a famous discussion in MMK Ch. 24 an opponent accuses Nāgārjuna with having destroyed the whole of Buddhism with his teaching of emptiness. Nāgārjuna replies that his opponent has misunderstood emptiness and its purpose, and his commentator Candrakīrti strongly reiterates the relationship between emptiness and dependent origination. It is necessary to understand the two truths taught by the Buddha. Without relying on everyday practice (vyavahāra) the ultimate is not taught, while without resorting to the ultimate there is no nirvāṇa. The ultimate truth here is emptiness, in that it is what is ultimately true about things. Things themselves as empty of intrinsic existence are the conventional. Without reference to things there could be no teaching of emptiness. Moreover, Nāgārjuna continues, where emptiness is seen to be rational and acceptable all things are seen to be rational and acceptable. This is because since emptiness is an implication of dependent origination, the alternative to emptiness would be intrinsic existence and therefore an unchanging block-universe (or of course literally nothing at all). If X exists but is not empty, X would be intrinsically or inherently existent and thus would never go out of existence. And in a wonderful reversal, Nāgārjuna accuses his opponent who denies emptiness with destroying the teaching of the Buddha. Who could become enlightened if their state of unenlightenment were intrinsically existent, and thus not the result of causes and conditions?¹⁷

Note also that there is a very real sense in which emptiness is dependent on things. Emptiness is the absence of intrinsic existence in the case of X. If there were no X then there could not be an emptiness of X. In a hypothetical case in which absolutely nothing existed – there simply were nothing – there could also be no emptiness. Thus in a sense emptiness exists in dependence upon that which is empty. It follows that being dependently originated in this way, emptiness is itself therefore empty. While emptiness is the ultimate truth in that it is what is ultimately true about X, it is not an ultimate truth in the sense that it is itself some sort of real primary existent. The ultimate truth is that all things, including any emptiness itself, lack ultimate truth. Therefore Mādhyamika uses ‘ultimate truth’ in two senses:

1. The first is the ultimate truth as an ultimate truth, i.e. something resistant to analysis, a primary existent. In this sense, Mādhyamika is saying that there is no such thing as an ultimate truth.

2. The second is the ultimate truth as the ultimate way of things (the dharmatā), how it ultimately is, what is found to be the case as a result of ultimate analysis, searching for primary existence. This is the lack, the absence, of that primary existence, i.e. emptiness.

Thus it is the ultimate truth in sense ii) that there is absolutely no ultimate truth in sense i).

It is important in studying the Mādhyamika approach to ultimate and conventional not to separate the two and think that Mādhyamika is advocating the ultimate truth as some sort of final goal into which we plunge beyond the conventional, ‘a state of emptiness’. Buddhism is not a move of this sort away from the conventional to an ultimate, but rather is a move of gnosis, an understanding of the conventional as merely conventional rather than bestowing it with a false sense of inherent or intrinsic, and therefore graspable, existence. The whole point is to see things the way they really are, to understand the ultimate way of things. Then the follower of Mahāyāna engages in the world for the benefit of others. The point is not to move (as it were) from this world to another realm of the ultimate, a realm of pure emptiness.¹⁸ Perhaps it was partly to prevent this move away from the world to a supposed ultimate realm that Nāgārjuna made his famous and much misunderstood statement about there being nothing whatsoever which differentiates nirvāṇa from saṃsāra (MMK 25: 19–20). This statement cannot be taken in context as meaning that this world is itself the realm of enlightenment. Nor can it be taken as indicating that enlightenment (or emptiness itself) lies just in a way of looking at the world, let alone that the life of monastic renunciation has somehow missed the point. Emptiness is not a way of looking at something. It is the quality of that thing which is its very absence of intrinsic existence – emptiness is in fact a property possessed by all things without exception as a result of their causally conditioned nature – and it seems that Nāgārjuna wants to equate saṃsāra and nirvāṇa only and simply as regards that quality of emptiness that they have in common. Moreover Nāgārjuna himself appears to have been a monk who fully expected the normal monastic lifestyle (as were all the great Mādhyamikas in India). There is no suggestion that he would have compromised on the supremacy and superiority (in realistic practical terms the necessity) of that monastic lifestyle. It is not the case, Nāgārjuna wants to say, that saṃsāra is empty but nirvāṇa is not. Nirvāṇa (or indeed emptiness itself) and saṃsāra both lack intrinsic existence and the one cannot be taken as an ultimate refuge from the other. According to universal Buddhist tradition Nāgārjuna was a follower of Mahāyāna, and he did not want his teaching of emptiness to entail plunging into the ultimate emptiness and a flight from the welfare of sentient beings. That would be the way to becoming an arhat, and not the way of the Mahāyāna.

In spite of its popularity outside India and among modern scholars the Mādhyamika system actually seems to have been rather neglected in the history of Indian thought. Hayes (1994) has suggested that this may well be because many of Nāgārjuna’s arguments are simply fallacious. But I do not believe that fallacies in Nāgārjuna’s arguments really get to the core of what was felt to be wrong with Mādhyamika. We have seen already that Nāgārjuna appears to slide from intrinsic nature in the sense of the type of existence that is not the result of conceptual construction to intrinsic existence, existence uncaused in any way. Since even dharmas are for Nāgārjuna one way or another the results of causes,¹⁹ this enables Nāgārjuna to maintain that all things are lacking intrinsic existence and hence intrinsic nature (svabhāva), and this entails they must be prajñāptimātra, merely conceptual constructs. This prajñāptimātra status of absolutely all things is a particular feature of Mādhyamika. Yet for the followers of Mādhyamika it is always held that such a position does not reduce to one of nihilism – nothing exists at all – since emptiness is said to be the same as dependent origination, not non-existence. But I suggest that most Buddhists in India, familiar with the Abhidharma, still felt that the Mādhyamika position was tantamount to nihilism for the simple reason that it is incoherent to maintain that all are merely conceptual constructs. It is, we might say, part of the meaning of ‘secondary, conceptual existence’ that there is primary, substantial existence.

It makes no sense to talk of all things being secondary existents. If all things were secondary existents then all things would be constructs with nothing left for them to be constructed out of. This must mean that nothing exists at all. It is not sufficient to reply with Nāgārjuna that this ignores the two truths, since if all is merely a conceptual construct then there could be no foundation for the two truths. Everything is foam which dissolves into nothing.²⁰

Thus through focusing on the Abhidharma background to the Mādhyamika project we can see that those who accused Mādhyamika of nihilism did not misunderstand Mādhyamika. They did not fail to understand the implication between emptiness and dependent origination. Rather they simply failed to be convinced that absolutely everything as merely conceptual constructs could avoid nihilism.²¹ It is not surprising, therefore, that in opposing as nihilism a position of prajñāptimātra clearly that of Mādhyamika, Yogācāra scholars precisely had to put forward the primary existence of something, something which could serve as an alternative to the primary existents (dharmas) of the other Abhidharma approaches.

Yogācāra

The Saṃdhinirmocana Sūtra

Not all Mahāyāna sūtras advocated universal emptiness as did the Perfection of Wisdom literature, at least when it was interpreted as equalling absence of intrinsic nature. Particularly interesting in this context is a text known as the Saṃdhinirmocana Sūtra. Among those Buddhists who knew of the teaching of emptiness many (I suggest most) felt it to be a quite absurd nihilism. Yet for some who claimed to follow the Mahāyāna this gave rise to a dilemma since it seems clear that in the Prajñāpāramitā literature the Buddha did indeed teach emptiness as the very perfection of wisdom. If emptiness as complete lack of primary existence cannot be taken literally, how is it to be understood? In Chapter 7 of the Saṃdhinirmocana Sūtra the Buddha is asked directly what was his intention behind declaring that ‘all dharmas [and therefore all things] lack their own essential nature’ (trans. Powers 1994: 96/7)? In other words, why say such a thing since it could scarcely be taken seriously without entailing that literally everything is a conceptual construct with nothing for it to be constructed out of? In the sūtra the Buddha replies that the correct way of understanding his teaching of emptiness is in terms of the ‘three natures’ (or ‘three aspects’; trisvabhāva, q.v.) which, when properly appreciated, will be seen not to entail nihilism at all. It is a misunderstanding of emptiness to take it as meaning that literally all things are conceptual constructs. As we shall see, it must follow therefore that at least one thing is not a conceptual construct (prajñāptisat), at least one thing must have primary existence (dravyasat), must exist with its intrinsic nature (i.e. in this sense – as a substratum to construction – have a svabhāva). In the Saṃdhinirmocana Sūtra the Buddha gives a vision of the history of Buddhism which pinpoints both how he has been misunderstood and also the position of his present teaching as its final explanatory clarification.

The sūtra recollects that at the very beginning of Buddhism, at his first discourse in the Deer Park at Sarnath called ‘setting in motion the Wheel of Dharma’, the Buddha taught such topics as the Four Noble Truths and so on. Nevertheless, this teaching was not intended to be a philosophically definitive teaching reflecting exactly the ultimate way of things. It is interpretable, neyārtha. For example really there are no persons as ultimate realities, they are, of course, conceptual constructs. Eventually this first teaching became a topic of controversy and dispute. Likewise he taught what was in fact a second ‘setting in motion the Wheel of Dharma’, a more advanced teaching, in the Prajñāpāramitā literature. This was a teaching of dharmas (therefore all things) lacking intrinsic nature. This teaching too, however – and this is important – was taught for a practical spiritual purpose and was not intended to be philosophically definitive. In time this teaching too was in fact misunderstood and became a basis for controversy and dispute. Note, therefore, that the teaching of universal emptiness as understood by Mādhyamika is itself declared to be not a philosophically definitive teaching reflecting exactly the ultimate way of things. Thus it is not the final ultimate truth. Once again we see that the logic of denying that all things lack svabhāva (i.e. are empty) must be to claim that at least one thing has a svabhāva. The final, definitive (nītārtha) teaching, which can be no basis at all for controversy and dispute, is contained within this Saṃdhinirmocana Sūtra itself. It is the third and final ‘setting in motion the Wheel of Dharma’.²² Those who unfortunately took the teachings of the second turning as definitive, on the other hand, either over negated as a consequence and completely destroyed everything, or they decided that this nihilism could not really be the teaching of the Buddha. Thus they unfortunately committed the great mistake of denying the Dharma of the Buddha (trans. Powers 1994: 118/9 ff.).²³

Yogācāra Teachers and Texts

The Saṃdhinirmocana Sūtra maintains that a literal understanding of the universality of conceptual construction (i.e. emptiness) must be wrong. Therefore we can see that at least one thing must have the status of being a primary existent. A Mahāyāna school of thought that holds such a position must obviously differ in at least this respect from Mādhyamika.²⁴ This school of thought is called Yogācāra – an expression that may indicate originally a particular interest in the data of meditation experience (yoga) – or sometimes Vijñānavāda, Vijñaptimātra or Cittamātra. The terms vijñāna (‘consciousness’), vijñapti (‘cognitive representation’) and citta (‘mind’), all indicate the orientation of this school towards what we might call the mentalistic side of our being. The addition of the term mātra – ‘only’, or ‘merely’, or ‘nothing but’ – also suggests that this school not only accepts at least one thing as a primary existent, but indeed it accepts only one thing. That thing is variously termed, but clearly in some sense it must be mentalistic. Thus this is the school of ‘Mind-only’.

It seems to me that Yogācāra was probably one way or another the most popular and influential of philosophical schools in ancient India associated with Mahāyāna. Within the Yogācāra tradition we find extensive discussions of Mahāyāna religious ideas and ideals – the status of the Buddha or the bodhisattva path, and meditation practice, for example – as well as issues relating to philosophical ontology. There is also – I would argue not surprisingly – a whole Yogācāra Abhidharma system. However at this point it is the Yogācāra approach to issues of ontology and the mind that interests us. The suggestion that ‘all that belongs to the triple world is mind only’ is found in a number of sūtras that are quite early such as the Pratyutpanna Sūtra and Daśabhūmika Sūtra, as well as the Saṃdhinirmocana Sūtra. Other sūtras that contain important Yogācāra material include the Avataṃsaka Sūtra and the Laṅkāvatāra sūtra. Note that commonly in the sūtras the introduction of this cittamātra (‘mind-only’) point occurs within the context of a discussion of visions seen in meditation experience. If certain Mahāyāna Buddhist thinkers, disturbed by Mādhyamika, wanted to find a primary existent to serve as a substratum for all things, saṃsāra and nirvāṇa alike, the fact that they chose something that was mentalistic might well reflect the general mentalistic orientation of Buddhism through meditation practice. In particular it may well show an interest in the implications of alternative and yet seemingly real worlds of meditation visions.

It is possible that the earliest named Yogācāra teacher was the shadowy Maitreyanātha. Tibetan tradition however holds that works attributed to him were in fact delivered to his pupil Asaṅga (c. fourth century CE) by none other than the great ‘celestial’ bodhisattva – to be the very next Buddha here on earth – Maitreya himself. Five works are sometimes attributed to Maitreya, whoever he was:

1. Abhisamayālaṃkāra (‘Ornament for the Realisations’) – a text on the Prajñāpāramitā path;

2. Madhyāntavibhāga (‘The Discrimination of the Middle from the Extremes’);

3. Dharmadharmatāvibhāga (‘The Discrimination of dharmas and their True Nature (dharmatā)’);

4. Mahāyānasūtrālaṃkāra (‘Ornament for the Mahāyāna Sūtras’); and the

5. Ratnagotravibhāga, otherwise known as the Uttaratantra, on the tathāgatagarbha or Buddha-nature teachings (q.v.).

It is quite likely that not all these five texts stem from the same hand. Actually it is the middle three with their commentaries that form some of our most important sources for classical Yogācāra. Asaṅga himself wrote some commentaries, as well as an important compendium of Yogācāra Mahāyāna, the Mahāyānasaṃgraha (‘Compendium on Mahāyāna’) and a work specifically establishing Yogācāra Abhidharma, the Abhidharmasamuccaya (‘Collection on Abhidharma’). One of the very earliest Yogācāra texts, however, is a large work probably of multiple authorship but sometimes attributed to Asaṅga, the Yogācārabhūmi (‘Stages of Yogācāra’). There is also a story – which has in the past been doubted by some modern scholars – that Asaṅga converted his brother Vasubandhu to Mahāyāna. There may have been two (or more) Vasubandhus. Attributed to a Mahāyāna Vasubandhu are the Triṃśikā (‘Thirty Verses’), the Viṃśatikā (‘Twenty Verses’), the Trisvabhāvanirdeśa (‘Teaching on the Three Natures’) and other texts including commentaries on some of the above works of ‘Maitreya’.²⁵

Mind and the ‘Three Natures’ (Trisvabhāva)

We have seen that in Yogācāra a mentalistic factor – let us call it Mind – is the one primary existent that serves as the substratum for everything else, both enlightenment and unenlightenment.²⁶ In the Saṃdhinirmocana Sūtra the antidote to nihilism is said to be the ‘three natures’, and this teaching of the three natures explains what Mind is, and the relationship of Mind to phenomenal illusion. How are these explained in the classical Indian Yogācāra texts?

The first of the three natures is the ‘constructed nature’ (parikalpitasvabhāva). What this amounts to is the aspect of our life which is a polarisation into separate subjects (called the ‘grasper’, grāhaka) confronting objects (the ‘grasped’, grāhya). This is the realm of subject-object duality, the world as seen by the unenlightened and also the realm of linguistic operation. Since as we know Yogācāra thinks in terms of just one primary existent, substratum for delusion as well as enlightenment, clearly duality cannot actually be correct. Duality is a wrenching apart of what is actually a unity, one basic ‘substance’ (ekadravya). This polarisation is erroneous.

The second nature is the ‘dependent nature’ (paratantrasvabhāva). It is the flow, a dependently originated continuum, of cognitive experiences (vijñapti), the substratum, that which is erroneously polarised into subjects and objects. If we settle down and examine carefully we will see that all the world of objects, and ourselves who confront those objects, are nothing more than a series of experiences. Actually there is vijñaptimātra, merely cognitive experiences, or merely representations. As a flow of experiences this flow must, of course, be mentalistic in nature. Note this. It could not be physical, and in Abhidharma terms there is realistically nothing else for it to be other than something mentalistic. Also it could not be the case that this flow does not exist at all – there is really no substratum – or there would be no experiences at all and therefore there would be nothing. Thus in Mādhyamika terms a flow of experiences must be found under ultimate analysis.²⁷ But note also that since the constructed nature is the realm of language, the other two natures including this flow of experiences as it actually is are strictly beyond language and can only be indicated obliquely through linguistic usage.

The ‘perfected nature’ (pariniṣpannasvabhāva) is the true way of things, which has to be seen in meditation. It is also said to be emptiness. But in Yogācāra texts emptiness is redefined to mean that the substratum which must exist in order for there to be anything at all is empty of subject-object duality. Thus emptiness must indeed be known for liberation, but this emptiness which is the perfected nature – the highest nature inasmuch as it that which is to be known by those avid for liberation – is defined as the very absence of the constructed nature in the dependent nature.

Students often get very confused about this, but it is not as difficult as it might appear. What we have to know in order to let go of the grasping which is unenlightenment is that the flow of experiences which we erroneously understand in terms of subjects and objects is actually, finally, all there is. It is therefore empty of those subjects and objects as separate polarised realities. That emptiness, the quality of ‘being empty of ’ is the perfected nature, and it has to be seen directly in the deepest possible way in order to cut the deep causes of unenlightenment and become enlightened. As in Mādhyamika, emptiness is an absence, a pure negation. This time, however, it is not absence of own- or intrinsic existence but rather absence of subject-object duality. For Yogācāra it is very much not the case that there is universal absence of intrinsic nature (svabhāva). In order for there to be absence of subject-object duality there has actually, really, to exist something which is erroneously divided into subjects and objects.

It is crucial in understanding Yogācāra philosophy to understand properly the three natures. Note that while the perfected nature is the highest nature in the sense that it is the highest thing to be known, it is not in classical early Indian Yogācāra the highest nature in the sense that it is itself the one reality. One should be very careful not to confuse the one reality in mainstream classical Indian Yogācāra with the perfected nature. Being the ultimate reality in an ontological sense, and being the highest thing to be known, the ultimate way of things, are in Yogācāra different. If Yogācāra teaches one mentalistic primary existent as substratum, in terms of the three natures that is the dependent nature. But what has to be known for liberation, the supreme in that sense, is that the flow of experiences which makes up our life is empty of polarised subjects and objects (empty of the constructed nature). That emptiness, that very absence itself, is the perfected nature, and it has to be known directly on the deepest possible level, in meditation.

Thus for Yogācāra the teaching of the three natures is the final teaching, the antidote to Mādhyamika ‘nihilism’, inasmuch as it denies what is to be denied – the constructed nature – but does not deny what should not be denied – the dependent nature. This dependent nature which once served as the substratum for saṃsāra is thus potentially still there to serve as the substratum for enlightenment. Texts therefore talk about the tainted dependent nature, and the purified dependent nature. This teaching is held to be the true Middle Path between over- and under-negation.

Certain Yogācāra texts, notably Vasubandhu’s Viṃśatikā, also give arguments to counter any objections to the idea that all is simply experience, that there are no (subjects and) objects external to consciousness, all is one primary existent (ekadravya). For example, the existence of spatial and temporal distinctions is no problem since that can be explained on the model of dream experiences where we experience spatio-temporal difference. The fact that unlike in the case of an hallucination many experience the same thing can be explained on the model of the Buddhist hells, where it is accepted that many undergo a common collective hallucination. Also it is difficult for opponents to explain the actual existence of ‘external’ objects, since things cannot be wholes in their own right because they are not experienced that way. They are experienced as having parts, and the division into parts threatens to extend to infinity. Objects also cannot be constructs out of fundamental minute ‘atoms’. This is because an atom, that is not further indivisible and therefore without spatial extension, capable of aggregation into spatially extended gross objects, is simply incoherent. Thus there can be no explanation of the world of matter. Yet clearly there is something. Since there is no matter (rūpa), but nevertheless there must be something, that something must itself be mentalistic (citta).²⁸

The dependent nature as a flow of experiences is the base, the substratum, ‘ultimate reality’ in Yogācāra. But note that this is not some immutable Absolute, and Yogācāra thinkers seem to differ over whether it is ever in itself really tainted. Inasmuch as we can look at consciousness from the point of view of the working-out of the phenomenal illusion of everyday saṃsāric experience, however, the Yogācāra tradition speaks of eight types of consciousness. There are the normal five sensory consciousnesses, the mental consciousness (manovijñāna) that, among other things, experiences mental events and also synthesises the data from the senses, the ‘tainted mind’ (kliṣṭamanas) and the ‘substratum consciousness’ (ālayavijñāna). The substratum consciousness is the flow, texts say the torrent, of underlying consciousness. While changing from moment to moment it serves to provide a necessary substratum for individual experience and also individual identity not just throughout one life but over the infinite series of lifetimes.²⁹ The ‘tainted mind’ observes the substratum consciousness and mistakenly conceives it to be a Self. Clearly such a changing flow could never be an actual Self in the sense in which Buddhists deny the Self. One of the main functions of the substratum consciousness is to serve as the ‘seedbed’, the repository for seeds (bīja) which result from karmically determinative deeds and which therefore issue in future experiences, including from particular sorts of seeds the very experiences of the ‘inter-subjective’ world itself. The seeds, which are momentary, form a series within the substratum until their fruition, and the substratum consciousness is ‘perfumed’ by their presence. Yogācāra thinkers disagreed over whether all seeds were the results of karmically determinative acts, or whether perhaps some were primevally latent in the substratum, and differed also over whether the substratum consciousness continued even at enlightenment. In China the Indian missionary Paramārtha (490–569) seemed to think that the substratum consciousness would cease, to be replaced by an ninth ‘immaculate consciousness’ (amalavijñāna). But notwithstanding this, the general view in Indian Yogācāra appears to be that at enlightenment the consciousness that was the substratum consciousness would continue to exist forever in a completely radiant and purified form.³⁰

The Buddha-Nature (Tathāgatagarbha) in India

Broadly speaking, the teaching of the tathāgatagarbha in Indian Mahāyāna is concerned with that factor possessed by each sentient being which enables him or her to become a fully enlightened Buddha. It is, as the leading contemporary scholar of the tathāgatagarbha David Seyfort Ruegg has put it, ‘the “buddhamorphic” Base or Support for practice of the Path, and hence the motivating “cause” (hetu: dhātu) for attainment of the Fruit (phala) of buddhahood’ (1989a: 18–19). The earliest sources strongly advocating the possession of the tathāgatagarbha appear to be sūtras such as the short and appropriately named Tathāgatagarbha Sūtra. It may well have been composed in the mid- or second half of the third century CE. This century corresponds with one estimate of the date of another crucial tathāgatagarbha sūtra, the Śrīmālādevīsiṃhanāda Sūtra (‘Discourse that is the Lion’s Roar of Queen Śrīmālā’, Wayman and Wayman 1974: 1–4). This latter sūtra, however, seems to show a much more elaborate doctrinal understanding of what the tathāgatagarbha might be, and how exactly it relates to Buddhahood. Also vitally important for understanding the tathāgatagarbha doctrine in the Indic cultural world – and important in the transmission of these ideas to East Asia where they have become an integral part of the Buddhist vision – is the [Mahāyāna] Mahāparinirvāṇa Sūtra. There are also some other significant tathāgatagarbha sūtras such as the Aṅgulimālīya Sūtra as well as interesting and important references to the tathāgatagarbha and related ideas in such sūtras as the Laṅkāvatāra and Avataṃsaka Sūtras (see Gomez 1995). Given the importance of these teachings particularly in Far Eastern Buddhism, it is striking that the only developed attempt in India to understand them within the systematic context of a philosophical treatise (śāstra) is in the Ratnagotravibhāga (Uttaratantra) and its commentary (Vyākhyā), both works of disputed authorship.³¹

So far we have followed through in particular one crucial strand in the development of Buddhist philosophical thought – the ontological issue of construction and substratum, prajñaptisat and dravyasat. Such an issue is indeed soteriological, related to a letting-go, and the range of that letting-go and the possibilities of its occurrence. Issues of ontology in Buddhist thought take place it seems to me, within the context of debates which are first and foremost broadly those of Abhidharma. I strongly suspect, however, that notwithstanding issues of the ontology of the tathāgatagarbha that developed later, particularly outside India, the topic of the tathāgatagarbha did not originate in India within this broadly Abhidharma ontological context. In other words the very context within which the issue of the tathāgatagarbha emerged was conceptually not one which was concerned with relating it to questions of ontology.

We can willingly understand, starting with the distinction between conceptual construct and primary existent, how some might argue that complete letting-go and philosophical consistency required that literally all things are conceptual constructs (Mādhyamika). We can also understand how some others (Yogācāra) might feel that this in fact collapses into nihilism and there must actually be a primary existent that is the non-dual flow of experience itself (which after all, cannot be denied even if we can deny certain things about it). While not exclusively so, nevertheless these are very much ontological issues. It is clear from our earliest sources such as the Tathāgatagarbha Sūtra, on the other hand, that the topic of the tathāgatagarbha is more to do with specifically religious issues of realising one’s spiritual potential, exhortation and encouragement, not ontology. We are not as such in an Abhidharma world. Our context, our immediate conceptual world, is quite different. It is perhaps rather the world of advocating the supremacy of the Mahāyāna against rival ‘lower’ paths, for if the tathāgatagarbha – the Buddha-nature – is in all sentient beings, all sentient beings should, and quite possibly in the end will, follow the path to a supreme Buddhahood. This path will leave the Arhats and pratyekabuddhas far behind. Issues of the ontological status of the tathāgatagarbha developed later. In China Fazang (Fa-tsang) in the seventh century claimed that the tathāgatagarbha tradition represents a fourth turning of the Dharma-wheel. In other words the tathāgatagarbha tradition represents a different philosophical and ontological position from Mādhyamika and Yogācāra. In spite of this, however, I do not believe that in its Indian origin this was the intention of the teaching of the tathāgatagarbha.

The Tathāgatagarbha Sūtra consists mainly of a series of examples showing that even though one is in the midst of defilements there dwells within all sentient beings a tathāgata, a Buddha, or (perhaps a later explanation), a tathāgatagarbha. This tathāgatagarbha is a tathāgata-womb, or a tathāgata-embryo, or a tathāgata-calyx, or a tathāgata-inner sanctum, or a tathāgata-husk, or a tathāgata-seed, or a tathāgata-interior (trans. Grosnick 1995: 92–3; cf. Zimmermann trans. 2002). Either way, something supremely valuable is contained within all this dross. The sūtra even goes so far as to have the Buddha state that hidden within the defilements is ‘the tathāgata’s wisdom, the tathāgata’s vision, and the tathāgata’s body … eternally unsullied, and … replete with virtues no different from my own … the tathāgatagarbhas of all beings are eternal and unchanging’ (Grosnick 1995: 96). The Buddha exhorts people, ‘do not consider yourself inferior or base. You all personally possess the buddha nature’ (op. cit.: p. 101). This short sūtra is a cry of encouragement, not a philosophical treatise. Some examples used suggest the tathāgatagarbha as a potentiality, some – perhaps wishing greater encouragement – use examples that speak of an actuality already achieved. Either way (and this ambiguity gave rise to endless doctrinal debates later), the message of the sūtra is that we all have a tremendous and probably unrealised spiritual potential.³²

Problems might start to arise, however, with the actual choice of terms used to refer to this tathāgatagarbha in some other sūtras, notably perhaps the Mahāparinirvāṇa Sūtra, a sūtra which is boldly prepared to use the term ātman, Self, for the tathāgatagarbha.³³ Perhaps it was this that began the attempt to clarify or explain the nature and ontological status of the tathāgatagarbha given the difficulties that would inevitably arise through using such a problematic term in a Buddhist context. The Mahāparinirvāṇa Sūtra is a long sūtra with a complex textual history. It does not always appear to be very consistent. But it is obvious that the Mahāparinirvāṇa Sūtra does not consider it impossible for a Buddhist to affirm an ātman providing it is clear what the correct understanding of this concept is, and indeed the sūtra clearly sees certain advantages in doing so. For example, since non-Buddhists are portrayed as considering the Buddha a nihilist due to his teaching of not-Self, providing there is no compromise of Buddhist tenets – and there does not have to be such a compromise – portraying the tathāgatagarbha as ātman might help convert non-Buddhists to Buddhism. It might thus help them to realise that Buddhism is not a form of spiritual nihilism. After all, if there is a tathāgatagarbha that serves as the very foundation for attaining Buddhahood then Buddhism could not be nihilism. No questions have to be begged on what actually corresponds to the term tathāgatagarbha. Moreover adherents of the tathāgatagarbha argued that by structural opposition, if saṃsāra is, as Buddhists say, impermanent, not-Self, suffering and impure then Buddhahood (i.e. the tathāgatagarbha) as the negation of saṃsāra can indeed be portrayed without further commitment as permanent, Self, bliss and purity. Furthermore the tathāgatagarbha is by definition that very thing within sentient beings which enables them to become Buddhas, which means that the spiritual path is not impossible, and which shines forth in Buddhahood. Thus it does indeed fit some of the characteristics associated with a Self. Nevertheless the sūra as it stands is quite clear that while for these reasons we can speak of it as Self, actually it is not at all a Self, and those who have such Self-notions cannot perceive the tathāgatagarbha and thus become enlightened (see Ruegg 1989a: 21–6).

Problems in the teaching of the tathāgatagarbha can thus be neutralised by claiming among other things that it was just a skilful strategy to bring Buddhism to non-Buddhists who might otherwise be frightened by the truth. Alongside this approach, providing we are clear what the tathāgatagarbha is, calling it a Self need not be seen as in any way compromising Buddhism. But what, therefore, is the tathāgatagarbha? There is some suggestion in the sūtra that it could be taken as actually the very absence of Self itself. After all, all Buddhist traditions agree that this is what one has to know directly in order to become enlightened (‘the Tathāgata has spoken of not-self (bdag med pa) as self, in reality there is no self ’; trans. Ruegg 1989a: 23). Or, as the Laṅkāvatāra Sūtra (trans. Suzuki 1973: 68 ff., 190 ff.) suggests, the tathāgatagarbha could actually be another name for the Yogācāra substratum consciousness. It is the answer to the question what it is about sentient beings qua sentient beings that enables them to become Buddhas. It must therefore be something permanent in sentient beings. It is thus also possible to link the tathāgatagarbha with an old Buddhist concept of the ‘natural luminosity of the mind’ (prakṛtiprabhāsvaracitta), the idea that the mind is in its own nature never defiled. Defilements are simply adventitious to it. Therefore it is the primeval innate purity of the mind – taints are not essential to it – which enables Buddhahood to occur. We accordingly find the Avataṃsaka Sūtra referring to the wisdom, the mind, the gnosis of a Buddha which is present although unrealised in each sentient being (Gomez 1995: 109–11). Or the tathāgatagarbha could be, in Yogācāra terms, as the Mahāyānasūtrālaṃkāra suggests, the pure dependent nature (Griffiths 1990: 62–3). Or, again, if one is a Mādhyamika then that which enables sentient beings to become Buddhas must be the very factor that enables the minds of sentient beings to change into the minds of Buddhas. That which enables things to change is their simple absence of intrinsic existence, their emptiness. Thus the tathāgatagarbha becomes emptiness itself, but specifically emptiness when applied to the mental continuum. None of this, even the fact that the tathāgatagarbha is permanent, need entail any compromise with the Buddhist teaching of not-Self, since while speaking of an ātman none of these things need be thought of as an unchanging, inherently existing, ontologically real and independent, eternally enlightened True Self.

Perhaps it was the Śrīmālādevīsiṃhanāda Sūtra that first introduced the explicit association of the tathāgatagarbha with the dharmakāya (q.v.), the Buddha’s highest ‘body’, what a Buddha finally is in him- or herself. The term tasthāgatagarbha is said to be actually the name we give in the case of unenlightened beings to what in the case of Buddhas is called the dharmakāya. The dharmakāya is said to be ‘beginningless, uncreate, unborn, undying, free from death; permanent, steadfast, calm, eternal; intrinsically pure … This Dharmakāya of the Tathāgata when not free from the store of defilement is referred to as the Tathāgatagarbha’ (trans. Wayman and Wayman 1974: 98; see also Williams 2009: 106). The tathāgatagarbha/dharmakāya is also explicitly said to be empty (śūnya). But as we have seen with Yogācāra, the mere presence of an attribution of śūnya/śūnyatā to something in Buddhism does not in itself entail that these expressions are being used with the Mādhyamika sense of ‘absence of intrinsic existence’, or ‘merely having conceptual existence’. If the question of the actual ontological status of the tathāgatagarbha arises, it has not yet been settled simply by the use of śūnya/śūnyatā. In tathāgatagarbha texts, as with the Śrīmālā Sūtra, the tathāgatagarbha is said to be empty inasmuch as it is intrinsically free of defilements, but also not empty inasmuch as it truly and intrinsically possesses all the qualities of the Buddha (op. cit.: 99). The tathāgatagarbha is moreover explicitly said here not to be a Self (op. cit.: 106), although the dharmakāya for its part is nevertheless said to have the ‘perfection of self ’ (op. cit.: 102).

The Ratnagotravibhāga (Uttaratantra) appears to be the only systematic treatise on the tathāgatagarbha tradition composed in India. It is not clear just how influential it was, although its influence might have been relatively slight.³⁴ Anyway, for this text the true way of things (‘Suchness’, ‘Thusness’, tathatā) as tainted is called the tathāgatagarbha. As immaculate, on the other hand, that same tathatā is called the dharmakāya (trans. Takasaki 1966: 186–7). In itself, this tathatā is said to be ‘unchangeable by nature, sublime and perfectly pure’ (op. cit.: 287). Crucial to understanding the Ratnagotravibhāga is the idea of the intrinsic purity of consciousness. Buddhahood precisely is permanent and unconditioned because it does not involve bringing anything about – adding or removing anything – but rather realising what has always been the case.³⁵ The impurities that taint the mind and entail the state of unenlightenment (saṃsāra) are completely adventitious. From the point of view of the mind itself, in terms of its essential nature, they are simply not present. That is how they are able to be (as it were) ‘removed’, never to return. On the other hand from the point of view of the mind’s pure radiant intrinsic nature, because it is like this it is possessed of all the many qualities of a Buddha’s mind. These do not need actually to be brought about but merely need to be allowed to shine forth. Because they are intrinsic to the very nature of consciousness itself they, and the very state of Buddhahood, will never cease. How it is possible for consciousness to be intrinsically pure and yet defiled is one of a number of mysteries said by the Ratnagotravibhāga to be understandable by Buddhas alone (op. cit.: 188 ff.), for issues concerning the tathāgatagarbha are precisely the deepest issues accessible only to Tathāgatas themselves. For the rest of us and for the time being there can be only faith (op. cit.: 296).

Key Points to Chapter Five

• The Prajñāpāramitā (Perfection of Wisdom) literature consists of sūtras, commonly named by the number of verses. The literature has three key doctrines. (i) Prajñā (wisdom): this is the state of mind that comes from properly understanding how things really are in contrast to the way they appear to be. The Perfection of Wisdom speaks not of the wrongness of what had previously been considered to be wisdom, but rather of its perfection, the final proper understanding of the way things truly are. (ii) Emptiness (śūnyatā): this entails that absolutely everything is ‘like a magical illusion’. Nothing can be found, nothing can be grasped. Enlightenment comes from ceasing to grasp even the most subtle sources of attachment, and this ceasing to grasp requires seeing those things which could serve as sources of attachment as empty, mere conceptual constructs. Nothing must be grasped. All things are empty. (iii) The bodhisattva: all who can should finally take upon themselves the vow of the bodhisattva, to attain Perfect Buddhahood for the benefit of all sentient beings. Any other goal is inferior, i.e. Hīnayāna. Very advanced bodhisattvas have immense abilities to help. Thus they become ‘celestial bodhisattvas’, such as Avalokiteśvara – particularly associated with compassion – or Mañjuśrī, the bodhisattva of wisdom. The state of enlightenment attained when one is a Buddha is called a ‘non-abiding’ nirvāṇa. It is hence problematic to portray the bodhisattva as simply postponing nirvāṇa.

• Mādhyamika (‘Middling’): Mahāyāna philosophy, far from representing a negation of the approach of the Abhidharma, is best seen as a series of strategies within the Abhidharma enterprise. In Mādhyamika emptiness is derived from dependent origination. All things without exception are akin to illusions because all things are without intrinsic nature. All things are only secondary existents, conceptual constructs. They are so because they are the results of causes and conditions, they are dependently-originated. Anything that is the result of causes and conditions must of course lack its own intrinsic existence and hence its own intrinsic nature. Because even dharmas originate due to causes and conditions they too must be empty of primary, substantial existence. Nāgārjuna applies analytic investigation to some of the principal categories of Buddhist thought as well as those of non-Buddhists, such as causation itself, movement, time, the Buddha, nirvāṇa, as well as the Self. He declares emptiness whenever anything is found to be the result of some sort of causal process. In this way the assertions of complete emptiness ‘like an illusion’ in the Prajñāpāramitā sūtras can be demonstrated through reasoning. However, this is not the same as saying everything simply does not exist at all. Since emptiness is an implication of dependent origination for something to be empty entails that such a thing must in some sense exist.

• Yogācāra: Yogācāra was probably the most popular and influential of philosophical schools in ancient India associated with Mahāyāna. Among those Buddhists who knew of the teaching of emptiness many felt that on the Mādhyamika interpretation it nevertheless must entail nihilism, since it cannot be that literally everything is a conceptual construct since there would then be nothing left for things to be constructed out of. According to Yogācāra it is a misunderstanding of emptiness to take it as meaning that literally all things are conceptual constructs. In Yogācāra a mentalistic factor (‘Mind’) is the one primary existent that serves as the substratum for everything else. The antidote to nihilism is said to be the ‘three natures’. The first is the ‘constructed nature’. It is polarisation into separate subjects confronting separate objects, the realm of subject-object duality. This cannot actually be correct. Duality is a wrenching apart of what is in reality a unity, one basic ‘substance’. The second nature is the ‘dependent nature’. It is the flow of cognitive experiences, the substratum, that which is erroneously polarised into subjects and objects. It could not be the case that this flow does not exist at all – there is really no substratum – or there would be no experiences at all and therefore there would be nothing. The ‘perfected nature’ is the true way of things, which has to be seen in meditation. It is also said to be emptiness. But in Yogācāra texts emptiness is redefined to mean that the substratum which must exist in order for there to be anything at all is empty of subject-object duality. The perfected nature is hence defined as the very absence of the constructed nature in the dependent nature.

• The Buddha-nature (tathāgatagarbha): this teaching in Indian Mahāyāna is concerned with that factor possessed by each sentient being (i.e. the tathāgatagarbha) which enables him or her to become a fully enlightened Buddha. This topic concerns religious issues of realising one’s spiritual potential, and exhortation and encouragement, rather than ontology. In the Tathāgatagarbha Sūtra the Buddha teaches that hidden within the defilements of each sentient being is ‘the tathāgata’s wisdom, the tathāgata’s vision, and the Tathāgata’s body … eternally unsullied, and … replete with virtues no different from my own’. Some other sources are prūpared even to use the term ‘Self’ (ātman) for this. Problems of how to interpret what this factor might be, and in what ways the expression ātman when properly understood might be acceptable to Buddhists, have continued to the present day.