Against a Mahāyāna Absolute: Why Absolutism Need Not Be a
Conclusion of Mahāyāna Philosophy
Thesis submitted in accordance with the requirements of the University
of Liverpool for the degree of Doctor in Philosophy by Gary Joseph
Donnelly.
September 2018
Against a Mahāyāna Absolute: Why Absolutism Need Not Be a Conclusion of
Mahāyāna Philosophy
Gary Joseph Donnelly
This work will argue that Mahāyāna philosophy need not result in endorsement of some cosmic
Absolute in the vein of the Advaitin ātman-Brahman. Scholars such as Bhattacharya, Albahari
and Murti argue that the Buddha at no point denied the existence of a cosmic ātman, and instead
only denied a localised, individual ātman (what amounts to a jīva). The idea behind this, then, is
that the Buddha was in effect an Advaitin, analysing experience and advocating liberation in an
Advaitin sense: through a rejection of the individual ātman and knowledge (jñāna) of and
immersion into the universal ātman-Brahman.
I will explore how different religious traditions define and shape the Absolute according to their
own religious convictions, illustrating a divergence in conception from the very start, before
exploring key differences between the Advaitin conception of the Absolute as put forth by
Śaṅkara and as defended by Bhattacharya in The Ātman-Brahman in Ancient Buddhism. I then
challenge Bhattacharya’s claims that prajñāpāramitā literature necessarily endorses the ātmanBrahman and that Mahāyāna philosophies reorientate Buddhists towards the truth of the ātmanBrahman.
I do this by arguing that there are viable interpretations of Vasubandhu’s Yogācāra and
Nāgārjuna’s Madhyamaka that do not advocate such a belief, that prajñāpāramitā literature can
be viewed as a project in episteme rather than ontology, and that we need not find a ground of
the same sort as the ātman-Brahman in the Buddhist flux of experience. I conclude by showing
that whilst Absolutism is a theme in some schools of Buddhism, it need not be – contra
Bhattacharya – the conclusion of two major Mahāyāna philosophies.
विद्यातुराणाां न सुखां न वनद्रा
For those in pursuit of knowledge, there is neither comfort nor sleep
Acknowledgements
My gratitude must first go the Arts and Humanities Research Council, without whose generous
support this work would never have been completed. Their stipend meant that, unlike some of
my peers, I did not have to worry too much about money. For this, I am very grateful. Secondly,
thanks are owed to my examiners: Prof. Richard Gaskin and Dr. Mik Burley. It is thanks to my
examiners that my viva experience was thoroughly pleasant and afforded me the opportunity
not only to talk at length about my research (something that does not occur very often), but to
gain some insightful, precise feedback.
Next, to my supervisors. First, Prof. Simon Hailwood, who (somehow) managed to keep me on
top of all things admin and all things procedural – including my initial funding proposal back when
I was wrapping up my master’s degree. Work-related admin is something with which I am
traditionally awful, and so I am very grateful for this help over the past five or so years. Second,
to Dr. Chris Bartley, who has put up with my ramblings since 2010, and without whom I would no
doubt still be sat fruitlessly labouring over a ‘Teach Yourself Sanskrit’ book. There have been but
a few real defining influences on my life, but it is no exaggeration to say that Chris is one of them.
I cannot quantify just how much I have learned in the time I have known him. Needless to say
that my gratitude is boundless.
Now to my family. My Mum, Dad, and brother have never faltered in their support of what must
have looked from the outside as something of a strange career choice. They held the faith even
when mine had vacated, put up with my mood swings, my whinges and rants, and – particularly
on the final stretch of the write-up – my ridiculous working hours (often writing until 5 or 6 am).
It is true to say that I could not have done this without any of them. I shall remain forever grateful
for their unwavering faith and support. Finally, I would like dedicate this completed work to my
grandmothers: two proud, working class Scouse women, who encouraged me to take this route
and ‘get an -ology’. Both were excited to see me begin this project and it remains a matter of
great regret that neither would live to see it completed. I only hope that they would be proud.
CONTENTS
ABBREVIATIONS
I
INTRODUCTION
1
§1: DEFINING THE ABSOLUTE
16
§2: THE QUESTION OF ĀTMAN IN EARLY BUDDHISM
34
2.1 BHATTACHARYA’S ARGUMENT
2.2 ABHIDHARMIC DHARMA THEORY
2.3 AN EARLY BUDDHIST REJECTION OF BRAHMAN?
§3: THE ĀTMAN-BRAHMAN
3.1 KNOWING THE ĀTMAN-BRAHMAN
3.2 ULTIMATE KNOWLEDGE
§4: ADVAITA AND BUDDHISM
4.1 THE VAIBHĀṢIKA ACCOUNT OF POSSESSION
4.2 VASUBANDHU ON SVABHĀVA, CHANGE, AND DENIAL OF THE ĀTMAN
4.3 BRAHMAN, ACTION, AND MĀYĀ
4.4 YOGĀCĀRA IDEALISM?
§5: MADHYAMAKA AND ABSOLUTISM
5.1 UNDERSTANDING DHARMAKĀYA
5.2 BRAHMAKĀYA, DHAMMABHŪTA, BRAHMABHŪTA
5.3 THE TATHĀGATA AND ENSUING IMPLICATIONS
§6: NĀGĀRJUNA’S APPROACH
6.1 MADHYAMAKA METAPHYSICS
6.2 WHY POSIT CONVENTIONALITY AND ULTIMACY?
6.3 THE ULTIMATE AS A CONVENTIONAL DESIGNATION
51
70
81
92
101
111
143
155
166
178
182
194
195
211
214
229
239
246
251
CONCLUDING REMARKS
263
BIBLIOGRAPHY
267
Page |i
ABBREVIATIONS
AKBh
Abhidharmakośabhāsya of Vasubandhu
AsPP
Aṣṭasāhasrikāprajñāpāramitāsūtra
BG
Bhagavad Gītā
DN
Dīgha Nikāya
G
Guide to the Perplexed of Maimonides
MHK
Madhyamakahṛdayakārikās of Bhāvaviveka
MMK
Mūlamadhyamakakārikā of Nāgārjuna
MT
Mystical Theology of Pseudo-Dionysius the Areopagite
NKJV
Holy Bible, New King James Version
SN
Saṃyutta Nikāya
TSK/TSN
Trisvabhāvakārikā/Trisvabhāvanirdeśa of Vasubandhu
Ud.
Udāna of the Khuddaka Nikāya
VV
Vigrahavyāvartani of Nāgārjuna
Page |1
Introduction
This work is primarily an attempt to defend Mahāyāna philosophy broadly construed
against the claim that it ought to result in absolutism, understood here and throughout this thesis
as the belief that there is an Absolute or unconditioned Being that is the foundation of
conditioned existences. There are, of course no shortage of examples of absolutist thinking
within the wide range of Mahāyāna thought, and scholars such as Kamaleswar Bhattacharya have
seized upon such examples in an attempt to argue that Buddhist thought is, at its heart, focused
on bringing its practitioners to knowledge of the absolute ātman-Brahman; the ultimate identity
of souls and the foundational principle, as understood by adherents of the Advaita or non-dualist
form of Vedānta. Indeed, Bhattacharya’s Ātman-Brahman in Ancient Buddhism caused quite a
stir for advancing precisely this claim. Its central thesis is that the Buddha did not deny the cosmic
ātman-Brahman, instead only denying the ultimate reality of the individual ātman, or jīva (2015:
7; 13; 23; 32). This would make the anātman doctrine that has become synonymous with
Buddhism as understood today either a non-original feature added later by Buddhist scholastics,
or an original feature that has been misinterpreted and misunderstood.1 It would also mean that
Buddhism is at its heart a variant of Vedānta.
I will not spend much time discussing the ways in which some Buddhist schools of thought
do indeed endorse absolutism: Bhattacharya already makes clear that this is sometimes the case.
Instead, I will focus my efforts on explaining why it does not need to be the case. This work is
1
Bhattacharya seems to think that the second applies, though there is some disagreement in recent scholarship as
to whether the Buddha advocated anātman at all. See Wynne (2009) for more.
Page |2
thus a response to a subtle claim made in Bhattacharya’s book, namely that it is the Mahāyāna
schools ‘which put things right’ (2015: 39) in returning the Buddhist focus away from a doctrine
of selflessness (anātmavāda) to the absolutism of the ātman-Brahman. I will defend the thesis
that absolutism need not be a conclusion of Mahāyāna philosophy, and I intend to argue for this
thesis as follows.
First, I assess some conceptions of the Absolute according to different religious and
philosophical traditions, arguing that there is no single account of an Absolute that fully coheres
across multiple religious traditions, and so we ought to be clear precisely what sort of absolute
we are advocating (or not!) when discussing the Absolute. I argue that whilst ineffability is
common to mystical accounts of the Absolute (insofar as a mystic can give an account of an
ineffable Absolute), there are nevertheless many metaphysical assumptions and assignations
that vary wildly dependent on the tradition under examination. Consequently, it is not enough
to assume that it is the very same Absolute that is being experienced in each case simply because
mystics report some sort of ineffable experience. It could be the case, for example, that mystics
have simply achieved a psychological state that needs no attribution of metaphysical priority, or
indeed any metaphysical import at all. My overarching argument in this section is twofold: first,
that there are so many differences that are brought up so frequently when discussing the
Absolute according to different traditions that it barely makes sense to assume that they are all
talking about the same phenomenon or set of phenomena. Second, that the Buddha would deny
all such ‘accounts’ of the Absolute. Consequently, we ought to exercise caution when attempting
to draw parallels across traditions.
Page |3
Section 2 sees me address the question of ātman-Brahman in Buddhism in three ways.
First, I give a survey of Bhattacharya’s argument, offering some alternative readings of his sources
and referring to such Mahāyāna titans as Nāgārjuna and Vasubandhu in an attempt to delineate
what Bhattacharya sees as a necessarily monistic interpretation from what could be, if read with
some degree of nuance and awareness of the wider context, simple examples of upāya. I argue
that Bhattacharya overstates his case and that in choosing to leave ātman untranslated in some
key passages, he obscures the meaning therein. It is here that I introduce Vasubandhu for the
first time and use his conception of the three natures (trisvabhāva) to explain an apparently
absolutist extract. I further argue that more instances interpreted by Bhattacharya to support
an Absolute can be interpreted alternatively to support the Buddhist doctrine of anātman. I end
by outlining Bhattacharya’s position that anātman is actually synonymous with paramātman, and
that because the Buddha knows the nature of all dharmas (phenomena) to be anātman, he has
in reality arrived at the pinnacle of understanding: anātman is the ātman-Brahman.
From here, I give an outline of Abhidharmic dharma theory and explain that I think that
Bhattacharya is mistaken to equate the emptiness of dharmas with the ātman-Brahman.
Bhattacharya’s position seems to be that if dharmas are impersonal, then it is precisely this
impersonality that constitutes their true intrinsic nature. In such an instance, emptiness
(śūnyatā) would effectively be the svabhāva of all phenomena. I then ask if this is a feasible
position to take and conclude it might be, depending on how Bhattacharya understands a dharma
to function. I argue that he must follow Nāgārjuna in asserting their emptiness in order to avoid
issues around there being many svabhāvas instead of just one: the impersonal ātman-Brahman.
The problems with this position will provide the basis for sections 5 and 6.
Page |4
Section 2 ends with a discussion around a potential early Buddhist rejection of the
Brahman. Reading the Brahmanimantanika Sutta in a certain way sees some strong parallels
appear between the Upaniṣadic Brahman and the world of Baka-Brahmā that is visited by the
Buddha. I argue that the Buddha purposely eschews all ideas of immersion into a state of being
where existence is permanent and there is no aging, no death or rebirth.
The Buddha
demonstrates that his mind is more powerful than that of Baka-Brahmā, and employs his superior
mental powers to disappear from the powerful deity, demonstrating that the realm of BakaBrahmā – which relies on a specific mental state to enter – is actually transient and subject to all
of the mundane dissatisfactions that Baka-Brahmā thinks do not apply. This reading ties in, I
argue, with a sermon detailed in the Mūlapariyāya Sutta, where the Buddha explicitly denies that
anything comes forth from ’unbinding’, which can be understood as the liberated state. If this
liberated state is to be the ātman-Brahman, we would be at a loss to account for the existence
of the world if, as the Buddha argues, nothing comes from it or it is not immanent in the way that
the ātman-Brahman is required to be.
Section 3 addresses the ways in which it might be claimed that we can know this elusive
ātman-Brahman. I begin by examining how the Upaniṣads talk about the transcendent ātmanBrahman and why scriptural descriptions of it are valuable despite being ultimately inapplicable.
The position of Śaṅkara is that scripture is unique in its ability to assert the existence of the
ātman-Brahman; without this, we would not know to search for it. This is because, thinks
Śaṅkara, the ātman-Brahman is not an object to be discovered by empirical means. This raises
the question, then, how we might ‘know’ the ātman-Brahman at all given that its existence is so
radically other: we might even say that it is ‘radically inaccessible’ (Tillemans, 1999: 29), and so
Page |5
cannot ever be discovered through empirical experience alone or indeed through reason alone.
Śaṅkara attempts to deal with this issue by prioritising the authority of the scriptures themselves
(śruti) and their verbalisation (by a guru/teacher; śabda). I then compare the Buddhist recourse
to scripture, which I conclude is much weaker than that of the Advaitin, with thinkers such as
Dharmakīrti acknowledging that scriptural inferences always carry some risk of uncertainty and
should be avoided wherever possible: this is not a position that appears to be shared by Śaṅkara.
The focus then moves onto the significance of the pramāṇas to both Śaṅkara and
Nāgārjuna, discussing some superficial similarities and accounting for the significant differences
in the role of pramāṇa and scriptural authority (a type of pramāṇa for Śaṅkara). I go on to
conclude that Nāgārjuna has no use at all for any pramāṇas, developing this argument with
recourse to the Vigrahavyāvartani, explaining why I believe it to be the case that for Nāgārjuna,
anti-essentialism and anti-foundationalism are paramount: he believes that such things render
liberation impossible. This logic should extend to any notions of Brahman. I then argue that
whilst both Śaṅkara and Nāgārjuna ought to agree that if the pramāṇas have any use at all, it is
only in terms of the conventional, the fundamental Advaitin reliance on scriptural establishment
of the ātman-Brahman means that they in effect do make some sort of ultimate point: the ātmanBrahman is ultimately existent. It is the only existent. This is something that Nāgārjuna believes
to be impossible.
The section concludes with the introduction of a discussion of emptiness (śūnyatā). I
argue that the point of this concept is to remove all conceptual constructions/proliferations
(prapañca) and metaphysical views/positions (dṛṣṭi) – including those of ‘ultimate’ realities or
Page |6
truths. The aim here is purely soteriological: Nāgārjuna is not particularly concerned with
providing some account of the makeup of the world. He is instead concerned with adapting the
way in which we look upon the world so that we may stem our various dissatisfactions (duḥkha).
In section 4, some similarities are introduced between Advaitin doctrine and that of the
Mādhyamikas, leaning on Frank Whaling’s (1979) account of Śaṅkara’s relationship with the
Buddhist doctrines of his time. I spend some time examining the close similarities between both
Śaṅkara and Nāgārjuna’s conceptions of conventional and ultimate, a distinction that seems to
have been absent from Vedānta until the time of Śaṅkara (Nicholson, 2007:531), before moving
on to a survey of Śaṅkara’s understanding of some other Buddhist doctrines. I note that there is
either a wilful or an accidental conflation of divergent doctrines belonging to separate Buddhist
schools. I suggest that this conflation may well have been an intentional attempt to portray the
Buddha as confused and inconsistent by showing him to propound three separate, incompatible
doctrines (Śaṅkara, 2011: 428).
The section continues with the introduction of some Buddhist doctrinal positions, a move
that I hope will demonstrate why Śaṅkara thought these Buddhists to propound theories
incompatible with his own (a point with which Bhattacharya must disagree). In order to do this,
I discuss the Vaibhāṣika doctrine of possession, noting that their belief in ultimately existent
atoms (dharmas), each possessing a svabhāva, precludes Śaṅkara from identifying with them in
any way. It also precludes the Vaibhāṣikas themselves from endorsing an ātman-Brahman, for
they already have an atomic account of the fundamentals of existence. The next step is to discuss
the Sautrāntika response to the doctrine of possession, which hinges upon a rejection of the
Page |7
Vaibhāṣika belief that a svabhāva is ultimately existent in the past, present, and future. The
Sautrāntika response – devised by Vasubandhu – posits radical momentariness of dharmas,
whereby their existence is wholly constituted by their causal activity (svalakṣaṇa). In this case,
the dharma is no longer basic, but rather its causal efficacy is basic. A svalakṣaṇa can only
manifest in the present, and so this in theory dispenses with the problematic endorsement of
dharmas that exist eternally. I explain why Vasubandhu thinks that this account solves some
fundamental Vaibhāṣika problems and provides a deliberate rejection of a fundamental ground
of experience, accounting for ‘direct and indirect causal efficacy in the face of momentariness
and the absence of a unifying substratum’ (Cox, 1995: 96). I go on to argue that Vasubandhu –
like Nāgārjuna – is concerned with denying that there are immutable ultimate entities, explaining
how he recharacterises svabhāva from immutable essence denoting ultimacy to a useful fiction
that is ultimately unreal: a svabhāva is no more than an appearance or mental construction
(parikalpita), and so lacks ultimacy.
Next, I extrapolate Vasubandhu’s use of svabhāva and introduce his ideas surrounding
change and the denial of ātman. Here I continue the claim made earlier that although
Vasubandhu uses the term ‘svabhāva’, he does so in a specific way, viz. in a manner that does
not bestow it with ultimate reality. Thus, his trisvabhāva doctrine does not endorse a trilogy of
ultimately real entities, but accounts for a tripartite way of seeing phenomena, each of which can
be understood to be conventional rather than ultimate. I proceed to give accounts of the
similarities in method of the Yogācāra and the Madhyamaka, noting that whilst some differences
can be noted in the ways in which words such as ‘śūnyatā’ are used, both traditions aim at an
epistemological realisation; the realisation that there are no permanent, immutable, ultimate
Page |8
entities. The argument hinges on my interpretation of Nāgārjuna as utterly unconcerned with
propagating a metaphysics by which the world can be explained. I argue, then, that both
Yogācāra philosophy and Madhyamaka philosophy deny that there can be a super-real ultimate
substratum, despite consciousness being a type of ground in Yogācāra. Importantly, this ground
is not of the same sort as the ātman-Brahman, viz. as the material and efficient cause of the
universe, immanent in all things, as a ground of Being. Instead, it is a ground of experience, which
is to say that it provides the basis for all of our experience and cognition. Our consciousness
does this without recourse to one single ultimate substance. Consciousness is intersubjective,
and each consciousness is to be understood as distinct, not as a misapprehended plurality
imposed over a single entity. In other words, consciousness ought not to be understood in the
way that the jīva is understood by Bhattacharya. I specifically refer to Lusthaus’ (2002: 489)
argument that the Yogācāra acknowledges multiple distinct consciousnesses, not simply the
misapprehension of consciousnesses that all have their nexus in a Brahman-like substratum. This
is a claim quite different to that advanced by Bhattacharya in support of the ātman-Brahman!
Further, I argue that as a ‘real’ entity as understood by Vasubandhu requires an entity be
momentary and not eternal, the ātman-Brahman cannot be considered a candidate for ‘reality’
in the Yogācāra tradition, much less be considered the only reality.
Following this, the discussion moves to the Advaitin doctrine of māyā and its
(im)plausibility as an account of the appearance of plurality and action in the world (we
misapprehend the true nature of ātman-Brahman owing to māyā). I then measure this against
the Buddhist account of delusion, which has its roots in measured analysis and reason. I conclude
that any Buddhist that takes seriously dependent origination, reason, and analysis simply cannot
Page |9
accept the doctrine of māyā as the answer to problematic questions regarding Brahman, action
and difference.
The final part of section 4 focuses on a discussion of Yogācāra as idealism. If it is the case
that Yogācāra is idealistic and endorses only one consciousness, then the claim that this
Mahāyāna school outright endorses absolutism might well be warranted. I borrow heavily from
Lusthaus and argue that should his characterisation of Yogācāra be correct, then we cannot
reasonably state that Yogācāra is idealism, or at least not as we commonly understand the term.
I contrast Lusthaus’ argument that Yogācāra is not idealism with divergent views, including the
position of Trivedi (2005), who argues that Yogācāra might best be understood as a type of
‘epistemic idealism’, whereby it would not make any ontological claims. It would then instead
‘claim that we know things not as they really are, as claim epistemic realists, but rather as they
are given to us by our ideas, our concepts, and categories’ (2005: 232). That is to say that we
cannot know things in themselves, only as they are represented to us. Trivedi, like Lusthaus, does
not think that consciousness somehow manufactures a mind-dependent world (2005: 236), and
so eschews the idea that Vasubandhu’s thought and the Yogācāra more generally can be
classified as metaphysical idealism. Following this line of thought, I argue that Vasubandhu is,
like Nāgārjuna, concerned not with building a metaphysical system, but instead with providing
some sort of roadmap to deconstructing our mistaken views. I suggest that the point of the
Yogācārin method is nirvikalpaka-jn͂āna, or non-discriminating cognition, a type of awareness
that does not impose onto the world or grasp out at it. The section is concluded with a brief
discussion of an excerpt from the Ch'eng wei-shih lun (translated by Lusthaus) which appears to
P a g e | 10
argue that to assert a paramātman in the name of the Yogācāra is to spectacularly misunderstand
both the aim of Yogācāra and the method it employs.
Section 5 involves some examinations of tricky Pāli/Sanskrit terminology, beginning with
dhammakāya/dharmakāya. This is a word that has many connotations in the Buddhist corpus,
and in my experience, the favoured meaning imparted to it generally depends on who is asked.
There is a trend in some Buddhist literature and some Buddhist schools to interpret the
dharmakāya as an absolute; it is on this view a realm in which the Buddha always exists, an
unconditioned, pure, metaphysically privileged realm not unlike that of the ātman-Brahman.
Entry to this realm, it is thought, is liberation. Surely, though, this would be unsatisfactory to
somebody like Nāgārjuna, who expended so much time and effort denying the possibility of
ultimate entities (and so realms)? I suggest that Nāgārjuna might opt to interpret dharmakāya
in an alternative manner, perhaps as an attribute of the Buddha rather than the Buddha’s Being.
I then cite Harrison’s (1992) argument that (limited) uses of dharmakāya in the
Aṣṭasāhasrikāprajñāpāramitāsūtra ought to be interpreted adjectivally, and so as some attribute
possessed by the Buddha rather than as some external thing in itself.
There follows a contextual discussion of the potential absolutist interpretations of
buddha-body and saṅgha-body, which are both mentioned in the same context as the dharmabody in the AsPP. It is difficult to understand how the saṅgha might be thought of as a
transcendental entity. It makes more sense to think of the saṅgha-body as the ‘collected qualities
of the saṅgha’, or the totality of qualities possessed by monastics – the principles of the wider
religious family, if you will. Similarly, if buddha-body is to be understood transcendentally, then
P a g e | 11
what separates it from the dharmakāya? Thus buddha-body might simply refer to the collected
qualities of all buddhas as ‘revealed’ in the Perfection of Wisdom. These are not necessarily
equivalent with the Dharma, but exemplify the personal qualities of those enlightened minds
that have fully understood and eventually come to ‘live’ the Dharma.
I then spend some time with Xing (2005), who argues that if we claim that the
dharmakāya of all buddhas is identical, all we are really claiming is that every buddha manifests
or embodies the Dharma in the same way, presumably because they have all acquired the same
qualities and the same insight into the world. I conclude by surmising that we can avoid
absolutism if we interpret such tricky words with some nuance and an awareness of their uses in
early Buddhist literature. For a Mādhyamika seeking to avoid absolutism, then, my contention is
that the dharmakāya refers to little more than the qualities exemplified, taught, and lived out by
the Buddha. There is no transcendental aspect required.
Next, I spend a short amount of time discussing how we might understand words like
brahmakāya in a similar vein to dharmakāya. I follow Xing in arguing that brahmakāya can be
understood as a synonym for dharmakāya (Xing, 2005: 71), likely for pedagogical reasons. I argue
that Bhattacharya’s assertion that brahmakāya, brahmabhūta and so on are always intended to
be taken transcendentally does not stand up to scrutiny because each word can be understood
in simpler, more mundane ways that still cohere with wider Buddhist doctrine.
There is then a short discussion about how a Buddhist concerned with denying ultimacy
and absolutism might want to interpret ‘tathāgata’ and tathāgatagarbha. The thrust of this
subsection is to provide an alternative interpretation of the tathāgata that does not rely on the
P a g e | 12
idea of an essential principle underpinning all existence (akin to the ātman-Brahman). I discuss
pertinent extracts in the MMK that appear to refute any ultimate difference between saṃsāra
and nirvāṇa, and advance the argument that Nāgārjuna cannot and does not endorse the
existence of an ultimate substratum like the ātman-Brahman. The tathāgatagarbha literature
should not then be taken at face value; we need to read carefully and closely in order to
determine if there are any absolutist connotations, or if the literature is simply aimed at swaying
those who might already hold a substance-view over to the Buddhist cause. It is well known that
the Buddha ‘knew his audience’ and often talked in ways appropriate to them rather than fully
appropriate to his teachings – that part would come later. It is my contention that it is at least
plausible that the tathāgatagarbha literature occupies this sort of space within the Mahāyāna
corpus.
The section continues with the argument that the point of Madhyamaka and Yogācāra
praxis is not to cling to ideas of existence or nonexistence, but rather to jettison these thought
process to begin with. The idea is not to arrive at knowledge of some ultimately existent
substratum akin to the ātman-Brahman, but to arrive at a point where we no longer impose any
conceptualisations whatsoever on our experience. Finally, I add that the tathāgatagarbha
literature might be understood in the same way as the prajñāpāramitā literature; as an episteme.
This is to say that it does not want to build a prescriptive metaphysics, but instead aims at a
‘gradual’, progressional epistemic unfolding of Buddhist teachings. For essentialists, it seems
that liberation is ontological. It involves entry into some ultimate realm or knowledge of some
ready-made ultimate substratum (like the ātman-Brahman). For progressionalists, liberation is
epistemological. Prajñā is that which brings about a mental change, and this change informs how
P a g e | 13
we interact with the world.
If we interact without grasping, without imposition and
appropriation, we might be said to be awakened. Taken this way, the tathāgatagarbha is not an
Absolute, but is simply the active unfolding of this potential in a progressive manner. To talk of
a person having tathāgatagarbha is then simply to talk of their having the potential to awaken.
The final section of this work focuses exclusively on how Nāgārjuna deals with the
problem of absolute existence. I discuss those positions that take śūnyatā to be either an
ultimate truth or an ultimate reality, arguing that Nāgārjuna is not attempting to give an account
of reality – ultimate or otherwise – but rather an account of how we experience reality. He is
thus developing a soteriological method that changes how we interact with the world in order to
quell duḥkha.
Consequently, I argue that the criticisms advanced by Burton (1999) are
wrongheaded insofar as they attempt to deal metaphysically with a method that seeks to eschew
all metaphysics. This theme continues in the next subsection, where I give an account of
Nāgārjuna’s ‘use’ of metaphysics. I go on to argue that Nāgārjuna toys around with the
metaphysical positions of his opponents only to illustrate their futility in the soteriological
scheme that he advocates. He is simply talking to his audience in terms they understand before
imploring them to abandon these concepts for the sake of their liberation.
If I am correct, then we ought to understand the ‘ultimate’ referred to by Nāgārjuna in
conventional terms. Ultimacy then becomes a concept under which nothing really falls. It has a
pragmatic use, allowing Nāgārjuna to delineate the highest conventional teachings from the
mundane. The highest teachings are ‘ultimate’ insofar as they are the highest set of a wholly
contingent bunch of teachings. It is with this in mind that I go on to argue that all Buddhist
P a g e | 14
teachings are contingent, and Nāgārjuna is, after all, putting forth Buddhist teachings relating to
soteriology, not teachings relating to the nature of the universe. Even if he were, these teachings
too would be contingent! I then demonstrate why it is the case that the Four Noble Truths are
contingent, arguing that they could not arise without deluded minds and dissatisfaction. I borrow
from Candrakīrti, Siderits and Katsura in accounting for conventionality as normal, everyday
relations between semantic and cognitive aspects, an uncritical and mistaken way of interacting
with the world. The next step is accounting for the value of referring to an ultimate that is itself
not ultimately existent. On this, I argue that the ultimate truth so conceived by Nāgārjuna is a
type of upāya: a useful way to think about things, but one that is ultimately to be abandoned as
mere concept. The result of this path is, I suggest, the abandonment of metaphysics in favour of
the experience of liberation: Nāgārjuna writes at MMK 25.24 that ‘the extinguishing of all
cognition, the extinguishing of reification, is blissful’. This extinguishing of cognition must include
notions of ultimacy.
In the final subsection, I conclude that given what has gone before, an ultimate
designation is for Nāgārjuna actually a conventional designation. It is communicated within a
necessarily conventional linguistic system and has no reality of its own. Its use is also distinctly
conventional, viz. to designate dependent concepts. I cite and discuss various passages from the
MMK to demonstrate that for Nāgārjuna, the denial of ultimacy holds from the top down and in
every circumstance. There is simply no room for ultimate entities, ultimate realities or ultimate
truths. In the majority of cases, even asking the question regarding ultimacy is mistaken and
serves only to distract from the Buddhist path. The notion of ultimacy is then narrowly useful to
Buddhists because it allows stark lines to be drawn regarding which teachings are useful and
P a g e | 15
which teachings are very useful. What is declared to be ultimate is rather the highest of the
conventional teachings, but it cannot be ultimate in any sense that Nāgārjuna would recognise.
Simply put, there is no place for ultimacy when we reach the summit of Madhyamaka praxis. It
is yet another prapañca; another concept reified and to which we wrongly cling. Liberation, then,
must consist in the abandonment of ultimates and the halting of conceptual cognition. It cannot
consist in knowing the reality of the ātman-Brahman. As Nāgārjuna writes, ‘those who proclaim
the real nature of the ātman and separate entities we do not consider experts in the Buddha’s
teaching’ (MMK 10.16). It is my contention, then, that absolutism need not be a conclusion of
Mahāyāna philosophy.
P a g e | 16
§1: Defining the Absolute
The notion of an Absolute is common to all of the world’s major religions in one way or
another, be it Islam’s tawḥīd, the Hindu Brahman, mystical Christianity’s transcendent God,2 the
Jewish elohút, or the Buddhist dharmakāya.3 Most scholars agree that we in the West have the
influence of Neoplatonism to thank for this development; such was its focus on mysticism, gnosis
and the One (i.e. the Absolute). It is with the advent of Neoplatonism that we begin to see the
primacy of a unitary principle underpinning all of reality. This unitary principle (the Good, the
One, the First, or in a more modern vernacular, the Absolute) is assumed to be ontologically prior
to the world which depends on it, and so it is a higher reality than that which is immediately
available to us. Little wonder, then, that this idea crept into religion and divinity!4 The immediate
2
Take, for example, Pseudo-Dionysius the Areopagite’s via negativa characterisation of the transcendent God:
We therefore maintain that the universal and transcendent Cause of all things is
neither without being nor without life, nor without reason or intelligence; nor is
He a body, nor has He form or shape, quality, quantity or weight; nor has He
any localized, visible or tangible existence; He is not sensible or perceptible; nor
is He subject to any disorder or inordination nor influenced by any earthly
passion; neither is He rendered impotent through the effects of material causes
and events; He needs no light; He suffers no change, corruption, division,
privation or flux; none of these things can either be identified with or attributed
unto Him.
(MT 4: 15)
The exact meaning of dharmakāya is contentious, and will be covered in more detail later (§5.1). However, one
popular and enduring way to interpret it is as a necessary and primary underlying substrate that has been defiled
and can be uncovered through diligent practice and meditation. Such an interpretation sees the dharmakāya as –
if not technically a godhead – then as a similarly transcendent ultimate reality that can be accessed through prajñā
(wisdom; insight into reality).
3
4
The Stanford Encylopedia of Philosophy provides an overview of the development of Neoplatonism and its
ensuing influence on the theological and philosophical tapestries of various religions (Wildberg, 2016).
P a g e | 17
question is whether all these characterisations refer to the exact same thing, or whether there is
some genuine variation. If there is difference and variation, how might this be accounted for? If
it is the case that every religious tradition has the same conception of the Absolute, then why are
there differing religious doctrines in the first place? The overarching question I want to address,
though, is ‘what type of absolute did the Buddha deny?’, and so this section will examine some
ways in which an absolute has been conceived across religious traditions.
In one sense, it seems relatively easy to answer at least parts of these questions: there
are strong traditions of apophasis in relation to the Absolute (ultimate truth, the godhead, and
so on) within each of the world’s major religions.5 This simply means that each religious tradition
holds within itself schools of thought that hold the Absolute (godhead) to be ineffable; we cannot
know the essence of it, or its ‘substance’ (Maimonides, G 1.58, p83) and so instead speak only of
what the Absolute is not.6 This can be taken further still with the claim that the Absolute does
not ‘exist’ in a meaningful sense. Whilst we exist and the objects that populate the world exist,
the Absolute is so radically other that to speak of its ‘existence’ would be mistaken. As Fideler
5
Negative theology (lahoot salbi) has historically been a feature of the Islamic theological tapestry. Though not
mystics (they were, in fact, rationalists), the al-muʿtazilah are one example, believing that God ‘is not merely
numerically one but also that he is a simple essence’, and so argue against the notion ‘that he has a body or any of
the characteristics of bodies such as colour, form, movement and localization in space; hence he cannot be seen, in
this world or the next’ (Robinson, 1998). Similarly, Schimel refers to Bāyezīd, a Sufi mystic that adhered to ‘austere
via negationis and constant mortification, by emptying himself of himself, until he had reached, at least for a
moment, the world of absolute unity where, as he said, lover, beloved, and love are one’ (1975: 49). This is a clear
example of the use of the negative method to achieve knowledge (gnosis) of or communion with the godhead.
Maimonides argues that ‘[i]n the contemplation of His essence, our comprehension and knowledge prove
insufficient; in the examination of His works, how they necessarily result from His will, our knowledge proves to be
ignorance, and in the endeavour to extol Him in words, all our efforts in speech are mere weakness and failure’ (G
58, p83).
6
P a g e | 18
puts it, ‘God is above entity’ (2002: 119, note 57). In this sense, we can say that most of the
thinkers that utilise the negative method believe that there are transcendent truths, but that
they lie beyond both existence (as we know it) and the intellect. This outlook typically qualifies
such adherents as ‘mystics’.7 We can find examples of the via negativa approach in Judaism
through thinkers such as Maimonides, Christianity through such adherents as Pseudo-Dionysius
the Areopagite,8 Islam via sects like the Muʿtazilites, in the Kabbalah,9 Advaita Vedānta,10 and
The OED defines ‘mystic’ as ‘any person who seeks by contemplation and self-surrender to obtain union with or
absorption into God, or who believes in the spiritual apprehension of truths which are beyond the intellect; a
person who has or seeks mystical experiences’ (‘mystic, n. and adj.’ OED Online. Oxford University Press). This
does not strictly apply to the Muʿtazilites, who tactically deployed the negative method in some circumstances, but
were nevertheless fully committed to rationalism (and not mysticism).
7
8
Pseudo-Dionysius the Areopagite differs from other mystics insofar as where some mystics call for gnosis of the
Absolute (jñāna is cognate with gnosis in the Indian religious traditions, but usually has a more limited scope), he
refers to the ‘Divine Darkness’, an agnosia. The Shrine of Wisdom editors define as ‘a transcendent unknowing —
a super-knowledge not obtained by means of the discursive reason’ (MT 1, p9). On this, Dionysius the Areopagite
writes:
…in the diligent exercise of mystical contemplation, leave behind the senses and
the operations of the intellect, and all things sensible and intellectual, and all
things in the world of being and non-being, that thou mayest arise by
unknowing towards the union, as far as is attainable, "with Him who transcends
all being and all knowledge. For by the unceasing and absolute renunqatiop [sic:
renunciation?] of thyself and of all things thou mayest be borne on high,
through pure and entire self-abnegation, into the superessential Radiance of
the Divine Darkness.
(MT 1: 9)
9
Wolfson (1994: VII) cites Azriel, a famed Jewish mystic, on the nature of Ein Sof (eternity, the unending), the
Kabbalistic conception of God prior to any self-manifestation. Although using some positive predicates to describe
Ein Sof, Azriel goes on to say that any comprehension of Ein Sof is achieved purely ‘through the negative way’. He
would likely argue, as do Śaṅkara and countless other mystics, that these positive predicate do not ultimately say
anything of the Absolute, but instead serve to simply orient our minds toward it so that we may be placed on the
right path, so to speak. Once on this path, it is the negative method declaring that which the Absolute is not that
can (perhaps counter-intuitively) furnish us with knowledge of the Absolute.
In his commentary to Brahmasūtra 1.1.3, Swami Vireswarananda writes that ‘Brahman has no form etc. and so
cannot be cognised by direct perception. Again in the absence of inseparable characteristics, as smoke is of fire, it
cannot be established by inference or analogy (Upamāṇa)’ (Vireswarananda, 2014: 24). He also cites Śaṅkara as
10
P a g e | 19
indeed Buddhism. 11 Though of vastly different religious traditions with different methods of
worship, different metaphysical conceptions and assumptions, and even different conceptions of
‘God’, each tradition nevertheless contains adherents that – to some degree or another –
advocate the via negativa as a means by which to ‘know’ God. I will spend some time looking at
how thinkers from each tradition do this in the hope that we can flesh out some similarities and
indeed some differences in how conceptualisations of the godhead are dealt with.
Maimonides puts the ineffability of the Absolute in no uncertain terms:
I would observe that,--as has already been shown--God's existence
is absolute, that it includes no composition, as will be proved, and
that we comprehend only the fact that He exists, not His essence.
Consequently it is a false assumption to hold that He has any
referring to the Kena Upaniṣad (either 1.3 or 1.4 depending on which version is used) when arguing that Brahman
cannot be an object of thought or indeed of the act of knowing (Vireswarananda, 2014: 28). So Kena Upaniṣad
1.3/1.4 (my translation):
na tatra cakṣur gacchati na vā gacchati no mano /
na vidmo na vijānimo yathaitadanuśiṣyāt //
anyadeva tad viditādatho aviditādadhi /
iti śuśruma purveṣāṃ ye nastad vyācacakṣire //
Not there goes the eye, nor goes speech, nor the mind; we do not know It, we
do not know how one might teach This.
It is distinct from the known; It is above the unknown. Thus we have heard
from the ancients who explained It to us.
Vireswarananda goes further still, arguing in the commentary to Brahmasūtra 1.1.4 that knowledge (specifically
‘knowledge of Brahman’) cannot be said to depend on an activity of the mind because ‘the result of action is either
creation, modification, purification or attainment. None of these is applicable to the knowledge of Brahman,
which is the same thing as Liberation’ (Vireswarananda, 2014: 29). For Śaṅkara, if liberation were something that
could be created or modified, it could not be permanent (Vireswarananda, 2014: 29). As liberation is synonymous
with knowledge of Brahman, this is an unacceptable conclusion for Śaṅkara as it first undermines the very point of
Advaitin praxis (why spend time and effort reaching liberation only for it to have the same result as saṃsāra?), and
second undermines the idea of Brahman as the permanent, unchanging efficient cause of the universe.
Murti writes of the dharmakāya: ‘The Cosmical body is [the Buddha’s] essential nature; it is one with the
Absolute’ before asserting that ‘[a]s the Dharmakāya, Buddha fully realises his identity with the Absolute
(dharmatā, śūnyatā) and unity (samatā) with all beings’ (2016: 284). This is important and will be developed in
detail later.
11
P a g e | 20
positive attribute: for He does not possess existence in addition to
His essence.
(G 1.58: 82)
This ineffability of the Absolute seems to be a consistent feature occurring at various
points across different religious traditions, and it is no different here for Maimonides. We can
see that he spends considerable effort detailing why it must be the case that God has no
attributes. Simply put, for a being to possess an attribute of any kind, Maimonides thinks that
there must be a duality involved, viz. that of the attribute itself and the attribute bearer. In order
to speak of attributes, we need to speak of something that has those attributes rather than
something that simply is. Maimonides’ opposition to this is simple: if it is the case that God has
attributes, then God has parts; having parts means that God is not a simple unity.12 The question
of attributes and their relation to the divine (godhead) is a theme that is a consistent feature of
many religious systems, most of which posit their own answers. Not content with denying
positive attributes to God, Maimonides goes further still, explaining why it is the case that we
must not predicate any attribute whatsoever of God. Arguing against such predications,
Maimonides (G 1.58, p83) writes that
Those who read the present work are aware that, notwithstanding
all the efforts of the mind, we can obtain no knowledge of the
essence of the heavens--a revolving substance which has been
measured by us in spans and cubits, and examined even as regards
the proportions of the several spheres to each other and respecting
most of their motions--although we know that they must consist of
matter and form; but the matter not being the same as sublunary
matter, we can only describe the heavens in terms expressing
negative properties, but not in terms denoting positive qualities.
12
‘God’ here – and in the following discussion – is synonymous with ‘Absolute’.
P a g e | 21
In this instance, it is clear that like other mystics, Maimonides thinks that ultimately, to
assign any attribute to God – positive or negative – is to get it wrong. This is very simply ‘because
we do not know [God’s] substance’ (G 1.58), and we cannot make true statements about that
which we do not know. However, given that we somehow need to orientate ourselves toward
God, using negative attributes to do this is generally better than our using (and inevitably reifying)
positive attributes. After all, how might one orientate himself or herself toward the divine if
nobody could ever describe that which they are orientating towards? Communication of divinity
is vital to religious praxis even if we cannot ultimately rely on the things said in such
communications. The fact remains however, that in the final analysis nothing at all can be
attributed to God: God transcends language and linguistic designation. Indeed, ‘in the endeavour
to extol Him in words, all our efforts in speech are mere weakness and failure’ (G 58, p83). This
is a position that exists elsewhere in the religious landscape – it is, in fact, common across
traditions, though its significance and impact varies in each different religious framework.
In his commentary to the Brahmasūtra (also known as the Vedāntasūtra), Swami
Vireswarananda13 – an Advaitin and interpreter of Śaṅkara – writes:
The scriptures, therefore, never describe Brahman as this or that,
but only negate manifoldness which is false, in texts like, “There is
no manifoldness in It” (Ka. 2.4.11), and “He who sees manifoldness
in It goes from death to death” (Ibid. 2.4.10).
(Vireswarananda, 2014: 28)
13
Vireswarananda was the tenth president of the Ramakrishna Math. A student of Brahmananda, who was himself
a direct disciple of Ramakrishna, Vireswarananda was a prolific writer and translator of Sanskrit texts, including the
commentaries of both Śaṅkara and Rāmānuja on the Brahmasūtra.
P a g e | 22
We can understand Brahman as the godhead or the ultimately existent Absolute that
underpins all of reality (and so to this degree we can say that it is analogous to the idea of God
held by Maimonides). Vireswarananda is very clearly mirroring the thoughts of Maimonides:
ultimately, ascribing any sort of attribute onto the Absolute is mistaken. Rather, to know the
Absolute is to negate ‘manifoldness which is false’, viz. to realise a unity and simplicity in the
Brahman.
Śaṅkara also denies the appropriateness of ascribing properties to the Absolute, but
acknowledges the requirement to speak around it in order to orientate people inwards as
opposed to outwards. In his commentary to Brahmasūtra 1.1.4, he writes:
But what then, it will be asked, is the purport of these sentences
which, at any rate have the appearance of injunctions; such as, ‘The
Self is to be seen, to be heard about?’– They have the purport, we
reply, of diverting (men) from the objects of natural activity. For
when a man acts intent on external things, and only anxious to
attain the objects of his desire and eschew the objects of his
aversion, and does not thereby reach the highest aim of man
although desirous of attaining it; such texts as the one quoted divert
him from the objects of natural activity and turn the stream of his
thoughts on the inward (the highest) Self.
(Śaṅkara, 2011: 35-36)
Here we can see that for Śaṅkara, the significance of the scriptures when referring to
Brahman’s attributes is not so much to describe the Brahman, but to elicit enough of an interest
from the reader or listener that they decide to eschew material things and pursue their own
knowledge of Brahman. This is primarily because knowledge of the material world does not
translate to knowledge of Brahman, and like Maimonides, Śaṅkara recognises that despite the
godhead being ultimately beyond attribution and description, people still need to be able to say
something about it in order to orient themselves toward it in a meaningful (and hopefully
P a g e | 23
successful) manner. The point is really that knowing the empirical world is not sufficient to know
the Absolute. It is nevertheless the case that knowing the Absolute endows one with the
knowledge that the empirical world that relies upon it is an illusory manifestation 14 (māyā),
clouding the true nature of the Absolute (in this case, Brahman). Śaṅkara elaborates on this point
in his commentary to Brahmasūtra 1.1.2, writing that
Brahman is not an object of the senses, it has no connection with
those other means of knowledge. For the senses have, according
to their nature, only external things for their objects, not Brahman.
If Brahman were an object of the senses, we might perceive that
the world is connected with Brahman as its effect; but as the effect
only (i.e. the world) is perceived, it is impossible to decide (through
perception) whether it is connected with Brahman or something
else. Therefore the Sūtra under discussion is not meant to
propound inference (as the means of knowing Brahman), but rather
to set forth a Vedānta-text.
(Śaṅkara, 2011: 19)
I hope that some recurrent themes are becoming obvious. Despite their drastically
different religious frameworks and traditions, both Maimonides and Śaṅkara believe that a
negative or apophatic approach to the divine is the only one that can make sense in the created
world. Some knowledge of the Absolute is possible, but not through the usual means of
interaction with the empirical world or the concepts it engenders. For Śaṅkara, it is in a complete
knowledge of the absolute that liberation consists: we fundamentally are the Absolute, and once
we uncover this latent knowledge – a form of self-knowledge – we gain knowledge of our
This is not to say that the empirical world is ‘unreal’ as such, but rather that its true essential character, the
Brahman, is obscured within it.
14
P a g e | 24
fundamental identity as Brahman.15 This experience of a transcendental (ultimate) intuition of
Brahman – brahmānubhava – can be categorised as a mystical experience precisely because it is
transcendental knowledge of (and so experience of) divinity that is said to be incomprehensible
when viewed through the empirical lens of the conventional world of māyā (Preti, 2014: 724).
But what of a thinker like Maimonides? Can Maimonides ‘know’ the divine in the way that
Śaṅkara can (at least in principle)? Such questions are relevant to the wider question here,
because it is not enough for each religious discipline to conceive of a godhead, or an absolute
divinity: their characters might be radically different; we might not be able to know their
characters at all.
These differentiations might then suggest that despite each religious tradition having
some idea of an absolute, the ideas are not identical. Even if every tradition agrees that there is
some ineffable Absolute, we are not at all justified in arguing that this shared quality (of
ineffability) means that each conception of the Absolute is identical in nature. More to the point,
for the Buddhist, we are not at all justified in taking as a matter of faith that there is an Absolute.
This is very simply because we do not know that this is the case; indeed, we cannot know that
this is the case. In the case of Maimonides, there is at least some disagreement over whether or
For Śaṅkara we always were the Brahman, we just did not quite know it. Upon realising the transcendent
intuitive self-knowledge that we are Brahman (brahmānubhava), we become liberated because we finally know –
really know – that we are one with Brahman. There are no more delusions, illusions and so on; we settle in the
‘bliss’ (ānanda) of the Brahman. In this sense, it is a process of becoming.
15
P a g e | 25
not his philosophy allows us to ‘know’ the Absolute. 16 On the face of things, Maimonides is
committed to a position that must always deny that the essence of the Absolute is in any sense
knowable. This position is prima facie contrary to somebody like Śaṅkara that thinks that we can
know the Absolute (because, ultimately, we are the Absolute); we just cannot speak of it
accurately in normal linguistic terms. Maimonides, then, thinks that whilst we can see the effects
of God’s existence – the ‘back’ of God17 – we cannot know God’s essence. In the exchange at
Exodus 33, God does not deign to tell us why we cannot know his essence (‘face’), only that
nobody may do so and live.
Maimonides stands in agreement with Śaṅkara on the one hand: both agree that there is
an Absolute from which all existence stems. This is a significant point of agreement. The points
of disagreement, though, are so stark as to seem insurmountable. First, Maimonides appears to
Blumenthal (2009: VII) laments that Maimonides has been codified in Jewish thought as a sort of ‘Jewish Kant’, a
rationalist that in many ways pre-empted modern rationalism and ‘modern scientific spirit’. This is mistaken,
argues Blumenthal, because Maimonides expends considerable time and effort utilising mystical language. On
this, Blumenthal writes that ‘[i]t cannot be happenstance that Maimonides uses them; rather, he clearly intends to
allude to a spiritual experience and reality which, though rooted in previous intellectual activity, transcends that
realm’ (2009: XII). Further to this, Blumenthal opines that Maimonides advocates three stages of ‘true spirituality’,
the third of which culminates in ‘a condition in which a person is in extended bliss’ that is ‘not a fleeting moment in
human spiritual life but an ongoing state of mystical consciousness’ (2009: XII-XIII). I do not presume to address
the merits of Blumenthal’s arguments here, rather I mention this only to illustrate that there is some deviation
from the consensus that Maimonides was primarily concerned with reconciling rationalism with Jewish philosophy.
16
17
Maimonides takes this view from Exodus 33.19-23 (NKJV):
Then He said, “I will make all My goodness pass before you, and I will proclaim
the name of the Lord before you. I will be gracious to whom I will be gracious,
and I will have compassion on whom I will have compassion.” But He said, “You
cannot see My face; for no man shall see Me, and live.” And the Lord said,
“Here is a place by Me, and you shall stand on the rock. So it shall be, while My
glory passes by, that I will put you in the cleft of the rock, and will cover you
with My hand while I pass by. Then I will take away My hand, and you shall see
My back; but My face shall not be seen.”
P a g e | 26
be committed to the maxim that we cannot know the Absolute – we can only know its effects.
Perhaps this speaks to his characterisation as the ‘Jewish Kant’. Second, Maimonides certainly
does not think that we – that is, humanity – are microcosms of the Absolute. We have already
seen that for Śaṅkara, we are all the Brahman and our individual existence as jīva is nothing but
illusion. It would be a sacrilege to Maimonides to suggest that we are God. Clearly, then, we
have a situation where both Śaṅkara and Maimonides agree that there is an Absolute. And yet
this is not sufficient for both to agree that they are talking about the very same thing: each one’s
version of this Absolute is fundamentally different.
I am inclined to think, then, that whilst variant religious traditions might agree in abstracto
that there is an Absolute Being, they fundamentally differ on the character of said Absolute. An
ever-present conception seems to be that of grounding (i.e. the Absolute Being is, in some way,
the grounds of reality), but this still tells us relatively little. How is this Absolute Being a grounding
principle? What does it mean for this principle to be a grounds for reality, and in what way does
it exist? I expect that in terms of Maimonides and Śaṅkara, neither would acknowledge the
other’s ‘Absolute’ as a true representation of what the Absolute is: Śaṅkara because he thinks we
can know exactly what the Absolute is (with a bit of work), and Maimonides because he thinks it
is simply impossible for us to ever know what is so radically other than us.18 I think that these
fundamental distinctions would also hold in some significant ways between Islamic, Christian and
Kabbalistic conceptions of the Absolute. The common thread seems to be that even when
18
At least, he is generally understood to think this way.
P a g e | 27
traditions agree that we can have some type of mystical gnosis of the Absolute, we still cannot
say what the Absolute is or what knowledge of it specifically involves. A quirk of the necessary
inadequacy of language this might be, but the consequence is that we cannot actually know
whether mystics of all religious bents are referring to the same thing.
What is it, then, that mystics claim to experience? Is every mystic – even when of
divergent traditions – experiencing the same thing? It might be (as the religious pluralists claim)
that mystics talk in the same way about the Absolute because they are experiencing the same
existing thing in similar ways. Might it be the case, though, that there is another reason for their
reporting of similar experiences of the same thing? William James hints at this in Varieties of
Religious Experience, suggesting that mystical experiences appear to have some pathological
context, but quickly dismisses this citing lack of reliable evidence (1917: 387,note 230). James
spends some time talking around physical causes whilst simultaneously failing to acknowledge
them. He argues that substance use (specifically alcohol and nitrous oxide, but we can easily
apply his reasoning to psychoactive substances and so on) can ‘stimulate the mystical faculties of
human nature’ allowing ‘genuine metaphysical revelation’ (1917: 387). It might turn out to be
the case that such substances do indeed unlock some hidden level of mystical consciousness, but
it might equally be the case that such substances simply increase our propensity to reify concepts
that do not capture anything real. Indeed, this is the very reason why the Buddhist monastic
code cautions against ingesting any mind-altering substances. That is not to say that a person
cannot come by such ecstasy without the help of mind-altering substances: such realisations
might be spontaneous and entirely ‘natural’. James notes, though, that they generally assume a
P a g e | 28
certain character and that this character can be actively cultivated: the most concerted efforts to
do this have been made by the Indian meditative traditions (1917: 400).
Yet I think it the case that there remains too much divergence between mystics of
different religious loyalties for it to be the case that each of them is reporting an experience of
the same Absolute. I have already outlined some distinctive differences between the Advaitin
mysticism of Śaṅkara and the type of mysticism alluded to by Maimonides, for example. For his
part, James (1917: 424-425) eventually concedes that
[t]he classic religious mysticism, it now must be confessed, is only a
“privileged case.” It is an extract, kept true to type by the selection
of the fittest specimens and their preservation in “schools.” It is
carved out from a much larger mass; and if we take the larger mass
as seriously as religious mysticism has historically taken itself, we
find that the supposed unanimity largely disappears.
James goes on to illustrate some of these stark differences, noting that for some
traditions, the Absolute is dualistic, for others it is a monism; for some it is pantheistic, for others
monotheistic and for still others, beyond any such categorisation (1917: 425). As we have seen,
the cases of mystical experience held up by each tradition are, according to James, the ones that
best fit their worldview. Should an Advaitin mystic have some experience that contradicts key
Advaitin doctrines (by, for example, gaining gnosis of a dualism and not a monism), it is a sure
bet that it would not be recorded and recounted by subsequent Advaitins as proof of the efficacy
of their method. Furthermore, there are such ‘mystical states’ which are pathological – they are,
as James puts it, ‘characteristic symptoms of enfeebled or deluded states of mind’ (1917: 426).
Such a ‘mysticism’ is destructive, causing real issues both for the person experiencing whatever
delusion plagues them and for the people that must deal with the effects of the behaviour that
P a g e | 29
such delusions might facilitate. It thus appears that ‘mysticism’ has two sides: a positive and a
negative. James, it seems to me, regards the former as mysticism proper, and the latter as not
really mysticism at all. It is not immediately clear to me why this distinction ought to be
maintained with any fervour: if one is possible, then so, presumably, is the other. If one is the
sign of an ‘enfeebled mind’, then why not the other? James agrees that we non-mystics are not
obliged to ‘acknowledge in mystical states a superior authority conferred on them by their
intrinsic nature’ (1917: 427).19 Instead, he writes that we ought to consider such states to be
‘inroads from the subconscious life, of the cerebral activity correlative to which we as yet know
nothing’ (1917: 427).
In other words, such states are indicative of something in the
‘subconscious’, but of precisely what, we do not know.20
We see, then, that the concept of the Absolute can be shaped according to whichever
religious or mystical tradition espouses it. In short, it seems to be a concept deployed by thinkers
James’ use of ‘intrinsic nature’ is interesting. As we will soon see, for Buddhists such as Nāgārjuna, ‘intrinsic
nature’ denotes a very specific sort of existence, viz. that which is ultimately real. Given that James clearly believes
that mystical experiences transfer to the experiencer a noetic quality, I suspect that he also believes that such a
knowledge is indeed ‘ultimately true’. On noesis, he has the following to say (1917: 380-381):
19
Although so similar to states of feeling, mystical states seem to those who
experience them to be also states of knowledge. They are states of insight into
depths of truth unplumbed by the discursive intellect. They are illuminations,
revelations, full of significance and importance, all inarticulate though they
remain; and as a rule they carry with them a curious sense of authority for
after-time.
It is important to note that this ‘authority’ carried by the noetic experience holds only for the experiencer and not
for anybody outside of that experience.
20
This has since been discussed at some length in philosophical, psychological and biological literature. Newberg
& D’Aquili provide a useful account of how brain processes and manipulation of them through ritual and so on
might give rise to a ‘mystical’ state of consciousness (2000: 260). Miller (2009) offers a thoughtful position relating
to Newberg & D’Aquili’s stance regarding the discussion of transcendental reality (namely that they are not
entitled to discuss it at all).
P a g e | 30
of certain biases and loyalties that has then been reified according to those very biases and
loyalties. Miller conceives of a situation whereby the ‘Absolute’ is actually numerically identical
to ‘baseline reality’ (viz. reality as experienced normally, outside of mystical or enlightened
experience). This is a position to which I am immediately drawn precisely because I think that it
might have some degree of correlation with the position of Nāgārjuna to be discussed in later
sections. Miller’s thesis is that baseline reality and the Absolute are the same thing but
experienced in a different manner (2009: 48). He expands thus:
For example, there may be one sense in which the world is
fundamentally united (as appears to be the case in experiences of
AUB [Absolute Unitary Being]), and another sense in which the
objects of the world can be differentiated. To state just one
example of the way in which this could be so: perhaps the world is
united in the sense that it is a single physical system (with
conservation of energy and so on), but that it is not united in the
sense that there are local variations in the kinds or amounts of
energy in different parts of the system. These parts, though, might
not be fundamentally metaphysically distinct, if they are all causally
interconnected with one another. If a person had this kind of belief,
then the experience of AUB could conform to the criterion of
intersubjective coherence. If a person believed that either the
experience of AUB or the experience of baseline reality is an
illusion, however, then the experience of AUB would fail the
criterion of intersubjective coherence.
(2009: 48)
There are some parallels to be drawn with Buddhist thought here. Miller suggests that
the parts of a whole need not be fundamentally distinct if they are all causally interconnected:
this is the principal tenet of the Buddhist conception of dependent origination
(pratītyasamutpāda). According to this doctrine, there are no fundamental, metaphysically
distinct ‘simples’; rather, there is simply an infinite web of interconnections between all psycho-
P a g e | 31
physical phenomena.21 A potential issue comes when we think about Miller’s final criterion in
relation to Buddhism broadly construed. If Buddhists think that ‘the experience of baseline
reality is an illusion’, then they cannot be said to think of baseline reality and the Absolute (AUB)
as the same thing experienced in a different way (viz. upon an epistemological shift). There are
indeed Buddhists that believe this to be the case, and so Miller’s ‘intersubjective experience’ is
not an explanation open to those particular Buddhists. The majority of this work will, though, be
assessing such things through a Madhyamaka lens, and it is clear to me that Nāgārjuna’s
insistence on denying all difference between saṃsāra (in Miller’s terms, baseline reality) and
nirvāṇa (in Miller’s terms, Absolute Unitary Being) – coupled with his insistence that we should
jettison notions of ‘ultimacy’ (metaphysical primacy) – means that his view might well fit Miller’s
criteria.22 Of course, Nāgārjuna would not denounce Buddhist sūtras that speak of the ‘delusion’
of sentient beings, and so there would need to be some account of this apparent incongruence.
This need not be a major issue, for I think that we can interpret the Buddhist focus on ‘delusion’
in a specific way, viz. one in which to be ‘deluded’ is to hold only a partial view of reality. In
21
The details of dependent origination are further discussed in the next section.
22
MMK 25.19-20 (my translation):
na saṃsārasya nirvāṇāt kiṃ cid asti viśeṣaṇam /
na nirvāṇasya saṃsārāt kiṃ cid asti viśeṣaṇam //19//
nirvāṇasya ca yā koṭiḥ koṭiḥ saṃsaraṇasya ca /
na tayor antaraṃ kiṃ cit susūkṣmam api vidyante //20//
There exists no difference whatsoever between saṃsāra and nirvāṇa /
There exists no difference whatsoever between nirvāṇa and saṃsāra //19//
The limit of nirvāṇa [is the] limit of saṃsāra /
There is not even the subtlest difference between them //20//
P a g e | 32
Miller’s terms, I think that this could work in two ways: first, one delusion would be to hold that
‘baseline reality’ is all there is to the world. Second, the other delusion would be to hold that
AUB is all there is to the world (and thus impart the metaphysical priority that Miller warns
against (2009: 42)). Those without delusion would simply see the totality of reality.
There is also a potential parallel with Advaita Vedānta: Bhattacharya – following the lead
of Śaṅkara – asserts that Advaitins too see the change as epistemic rather than metaphysical. For
example, Bhattacharya writes that one who knows Brahman ‘looks upon the world with new
eyes’ (2015: 13); ‘[w]e consider as entirely authentic the Mahāyānic doctrine according to which
there is, in the transcendent sense, no distinction between saṃsāra and nirvāṇa’ (2015: 137);
‘the difference between them is our way of looking at them; it is epistemic not metaphysical’
(2015: 137). Such a position would, on the face of things, seem to tally with what Miller calls the
‘intersubjective experience’. Miller notices the immediate problem with this position, however,
writing that some Hindu mystics do not, in fact, accept that both the baseline reality and AUB are
real. Instead, they hold that the AUB is real whereas the baseline reality is not: ‘[o]ne example
of this view is the claim of Hindu mystics that the world of sense experience is an illusion (māyā)
and that only the transcendent reality Brahman is real’ (Miller, 2009: 48). We might think that
such a position is incompatible with the above comments from Bhattacharya, but this is not quite
the case. Bhattacharya is right that liberation for the Advaitin consists in an epistemological
change – in seeing the world differently – but this does not change the fact that all Advaitins think
Brahman to be metaphysically prior to this world of māyā. It is in fact fundamental to the
Advaitin system that this is the case. Bhattacharya endorses Murti’s view that the ‘empirical’
(baseline reality) is the ‘veiled form or false appearance of the Absolute’ (Bhattacharya, 2015:
P a g e | 33
137-138; Murti, 2016: 232). Further, Bhattacharya presumably endorses Murti’s contention that
if both the empirical and the Absolute were real, then the Absolute ‘lacking determinations and
without any recognisable content, would even be less real than the empirical’ (Bhattacharya,
2015: 138; Murti, 2016: 232-233). This is unambiguous and clearly assigns metaphysical priority
to the AUB (Brahman). As we will see, I do not believe that Nāgārjuna makes this same sort of
jump.
All this considered, it seems that there is no consensus – indeed there can be no
consensus – on what constitutes the Absolute; if all religions are ultimately talking about the
same thing, we might expect a stronger consensus than ‘there is a unifying principle, we can know
it’. This maxim is really quite vague. Madhyamaka Buddhism – as we have briefly seen – has a
still different approach to the Absolute.
We find within Buddhism’s rich tapestry both
affirmations of an Absolute (sometimes very close to that of the Advaitins) and outright rejections
of it. It is with the latter that I am concerned, as I intend to demonstrate that contrary to
Bhattacharya’s thesis, Buddhists of the ilk of Nāgārjuna and his Mādhyamika followers ought not
to support any conception of a permanent Absolute. This includes – as I shall later demonstrate
– the conception of the Absolute as some fundamentum absolutum inconcussum veritatis, as the
unshakeable ground of truth.
P a g e | 34
§2: The Question of Ātman in Early Buddhism
Buddhism in its richly developed modern form has a famously deep-seated aversion to
ātmavāda (self-theory or doctrine of self), and so by extension, an aversion to endorsements of
an Absolute. This is for good reason: Buddhists – overall – think that any sort of belief in
permanence serves only to prolong and propagate dissatisfaction (duḥkha), namely via a belief
in an enduring immutable ‘self’ (ātman). The reasoning behind this is relatively straightforward:
Buddhism is at its heart a doctrine and praxis of change; change from deluded and unawakened
to awakened and liberated. The idea of an immutable essence is then anathema to this doctrine:
how might a person that is essentially (and thus permanently) unawakened ever have a chance
at dissatisfaction’s cessation? The Buddhist line is simply that an immutable essence precludes
– by definition – any hope of liberation. It is the clinging to the effects entailed by a belief in a
permanent self that constitute the most insidious forms of dissatisfaction. This in turn makes the
very idea of ātman the most insidious cause of dissatisfaction. The general idea is that selfinterest and attachment to both ‘I’ and things that we surmise as belonging to the ‘I’ drive anxiety
and so dissatisfaction.
Strong notions of ‘I’ mean that we become attached to transient things, but more
significantly, it means that we have strong attachment to ourselves. A strong ‘I’ notion means
that we suffer anxieties relating to ourselves; we are concerned about the possibility of being
sick, of ageing, and of the possibility (and eventual certainty) of death. More trivially, we have
anxieties about those things that we construe as ‘ours’, and even nice, initially pleasurable things
end up as dissatisfying. A pleasurable experience (hearing one’s favourite record, for example)
P a g e | 35
gives us a sense of satisfaction (sukha) for a characteristically short period of time (are any of us
in a permanent state of satisfaction after enjoying only once something that pleases us?).
Therefore, when I take great pleasure in hearing my favourite record, I am cultivating attachment
and grasping on multiple levels. I initially enjoy the temporary feeling that hearing the record
elicits within me. I can then be said to grasp at three things in regards to pleasure: the record
itself (as deliverer of pleasure), the feeling of pleasure, and the ability to feel the pleasure. More
specifically, I suppose we become attached to the ability to sate a desire as and when it arises,
thus reinforcing the idea of self – of me both doing the action required and feeling the benefit of
its result.23 I then feel the urge to replicate this feeling of satisfaction/contentedness and so
spend time and effort on reproducing the pleasurable effect by whichever means we came by it
(in this case, hearing our favourite record). The result is a temporary sense of satisfaction, but a
stronger attachment to the ‘I’ that we think is being satisfied. It is this latter point against which
the Buddhist project attempts to work.
Of course, the counterpart to cultivation of attachment to pleasure is varying degrees of
dissatisfaction. On the Buddhist worldview, such dissatisfaction arises in virtue of our attachment
to the specific phenomenon (the record) and the circumstances that might prevent us from
actualising the feeling of satisfaction to which we are attached (playing the record). This only
It seems as though any attachment to a feeling must contain some symmetrical attachment (perhaps ‘grasping’
might be a more suitable term here) to the ‘I’ that experiences the feeling. In other words, the feeling serves to
reinforce what Buddhists ultimately take to be a mistaken belief in this immutable ‘I’. If we were told that we were
to be in a position whereby we could not physically feel this (or, more starkly, any) pleasure, it would cause some
anxiety.
23
P a g e | 36
ceases – and then only temporarily – when we find ourselves in a situation whereby we can once
again play our favourite record. We might find ourselves in the unfortunate situation whereby
we can no longer feel a given (or indeed any) pleasure whatsoever – this would, I expect, cause
great anxiety: nobody wants to be in a position whereby they never again feel pleasure (or
perhaps more accurately, where they cannot sate a desire and feel the (temporary) result).24 We
find ourselves, then, trapped in an endless circle of temporary, fleeting desire-satiation, all of
which is ultimately doomed to inadequacy, and all of which serves to simply reinforce the very ‘I’
notions that fuel the cycle to begin with.
Interestingly (and undoubtedly factoring into Bhattacharya’s thought), both Advaita and Buddhism talk of
liberation in some very similar ways. Advaita talks of the ‘bliss’ (ānanda) of Brahman, so called because all desire is
removed and so there is wanting for nothing. Buddhist sources generally stop short of using words analogous to
ānanda (though not always!) and instead speak of the ‘extinction’ of desire. This amounts, I think, to broadly the
same thing. The difference is then a matter of what this extinction of desire consists in, if anything at all. The
Buddhists would usually claim that there is a sort of unconditioned awareness but that this does not imply an
immersion into a further immutable substantial whole. The Advaitins would claim that this unconditioned
awareness is itself a substantial Reality (Brahman), and our awareness of it is a result of us finally knowing our true
nature (i.e. we are no more than the ātman-Brahman).
24
To this end, the Saṃyutta Nikāya (II Book IX.1.2; PTS p359) has the following to say (my translation):
Kataman͂ca bikkhave asaṅkhataṃ / Yo bikkhave rāgakkhayo dosakkhayo
mohakkhayo idaṃ vuccati bikkhave asaṅkhataṃ //
O bhikkhus, what is the unconditioned [Absolute] (asaṅkhatam)? It is, O
bhikkhus, the extinction of desire [lust; greed] (rāgakkhaya) the extinction of
hatred (dosakkhaya), the extinction of confusion [illusion] (mohakkhaya). This,
O bhikkhus, is called the unconditioned.
Here we have a clear account of the liberated experience: the removal (extinction) of desire, hatred, and illusion.
This experience must be blissful precisely because all desires are not just temporarily sated, but fully dissolved,
much like the accounts of ānanda for the Advaitin. There is, however, an important privation in the account from
the Saṃyutta Nikāya – there is no mention of this being the substantial, eternal reality that Brahman is purported
to be.
P a g e | 37
This being the case, the Buddhists argue, we ought not to ascribe any permanent nature
to either the circumstances, the feelings or the objects involved; all of which come about as a
result of pratītyasamutpāda (dependent origination). This is a complex inter-dependent web of
connections between every experienced phenomenon, all of which combine in innumerable
ways in order to affect everything else. Insofar as Buddhism has an ontological principle, it is this;
there are no phenomena which are unconditioned by the process of dependent origination, and
so this causal principle dictates that everything in the experienced world is intimately and
inextricably related to everything else.25 There is, on the face of things, no first cause and no
permanent substratum.26
Buddhists are usually understood to deny any underlying unchanging identity to the
experiencer (me; you), the circumstances that present themselves in the world, or to the feelings
that present themselves to us. In fact, it is usually taken to be the case that Buddhists deny
underlying, unchanging identity in the world in toto.27 So, for many, the defining characteristic
25
Some Buddhist schools have a nuanced take on this, which I shall here sketch briefly. It is true that a number of
Buddhist schools hold that the state of nirvāṇa is unconditioned by dependent origination. Indeed, it is the only
state that is untouched by dependent origination. On this understanding, we might be tempted to accept such a
nirvāṇa as an Absolute. Nāgārjuna and his followers in the Madhyamaka school eschew this idea, explicitly arguing
that there is (and can be) no difference at all between saṃsāra and nirvāṇa (MMK 25.19). The reasons for this
conclusion will be covered in detail later.
26
It must be noted that if some Buddhist schools and interpreters do indeed follow the analysis that we will shortly
see put forth by Bhattacharya, then they might reasonably be said to endorse some sort of first cause (namely the
ātman-Brahman).
27
There is a nuance here, namely that for some Buddhists, the basic constituents which combine via
pratītyasamutpāda to manifest as all psycho-physical phenomena do indeed have permanent, unchanging
identities (svabhāva). Such Buddhists still hold that the world as experienced lack such a grounding because all
experience is a conglomeration of dharmas combining in different ways via dependent origination. According to
such a position, the role of dependent origination is preserved insofar as all experienced phenomena are indeed
P a g e | 38
of Buddhism – in contradistinction to other Indian religions such as Jainism and the various
schools of Hinduism – is precisely that it advances the position that there is no enduring
substantial self (anātman) and largely bases its soteriological methodology on the realisation of
this truth. Speaking of truth, it would be appropriate to interject briefly with a note on ‘truth’ in
Buddhism and in Advaita. In both Buddhist and Advaitin literature, we see reference to
saṃvṛtisatya (conventional truth/reality) and paramārthasatya (ultimate truth/reality) in
relation to both their respective teachings and end goals. It looks as though the two truths were
devised as a means by which to navigate seemingly contradictory or inconsistent scripture. Such
a process would allow a commentator or adherent to make sense of things that initially look to
be in conflict, for example the Buddha’s referring to the ‘self as refuge’28 whilst also appearing to
conditioned by these intrinsically existent dharmas. The Madhyamaka school of Nāgārjuna holds, however, that
anātman should apply from the top down, and so even dharmas cannot hold an immutable identity of any sort.
For the Mādhyamika, then, for dependent origination to hold at all, it must hold in all circumstances, and this
necessarily means that dharmas could not possess svabhāva.
28
Pérez-Remón (1980: 20) translates the following section of the Mahāparinibbāṇa Sutta:
Tasmātihānanda, attadīpā viharatha attasaraṇā anaññasaraṇā, dhammadīpā dhammasaraṇā
anaññasaraṇā
Therefore, Ānanda, stay as those who have the self as island, as those who have the self as
refuge, as those who have no other refuge; as those who have dhamma as island, as those who
have dhamma as refuge, as those who have no other refuge.
P a g e | 39
deny the self.29,30 Conventional truths are, according to Siderits and Katsura truths for which
‘acceptance reliably leads to successful practice. Our commonsense convictions concerning
29
Thanissaro Bhikkhu (1997) translates the following passage from the Samanupassanā Sutta:
Sāvatthinidānaṃ. “Ye hi keci, bhikkhave, samaṇā vā brāhmaṇā vā anekavihitaṃ
attānaṃ samanupassamānā samanupassanti, sabbete pañcupādānakkhandhe
samanupassanti, etesaṃ vā aññataraṃ. Katame pañca? Idha, bhikkhave,
assutavā puthujjano ariyānaṃ adassāvī ariyadhammassa akovido ariyadhamme
avinīto, sappurisānaṃ adassāvī sappurisadhammassa akovido sappurisadhamme avinīto rūpaṃ attato samanupassati, rūpavantaṃ vā attānaṃ; attani
vā rūpaṃ, rūpasmiṃ vā attānaṃ. Vedanaṃ … saññaṃ … saṅkhāre … viññāṇaṃ
attato samanupassati, viññāṇavantaṃ vā attānaṃ; attani vā viññāṇaṃ,
viññāṇasmiṃ vā attānaṃ.
At Savatthi. There the Blessed One said, "Monks, whatever contemplatives or
brahmans who assume in various ways when assuming a self, all assume the
five clinging-aggregates, or a certain one of them. Which five? There is the case
where an uninstructed, run-of-the-mill person — who has no regard for noble
ones, is not well-versed or disciplined in their Dhamma; who has no regard for
men of integrity, is not well-versed or disciplined in their Dhamma — assumes
form (the body) to be the self, or the self as possessing form, or form as in the
self, or the self as in form.
30
It can be argued that these two positions are not actually incompatible with each other when understood a
certain way. Bhattacharya references these sorts of canonical examples (see 2015: 31, 38 for examples of this) to
illustrate that the Buddha was at all times endorsing a transcendental self, viz. the ātman-Brahman. This reading,
though, depends on a very specific reading of ‘ātman’- one that always translates it as a substantial, ultimate self.
However, in some cases, we need not make this jump, and in exercising this self-control, we change the force of
the sentence quite significantly. It should always be borne in mind that some uses of ātman in Sanskrit literature
and attā in the Pāli literature – be they Buddhist sources or otherwise – are simply as a reflexive pronoun. In the
example given in note 98, this realisation changes things in an important way. Instead of encouraging the
practitioner to ‘take refuge in the Self’, where ‘Self’ has an initial capital letter and means the sort of subtle
essence present inside of us that Bhattacharya endorses, we have ‘take oneself as refuge’ in a much weaker sense,
meaning something more along the lines of ‘do not look to anybody else for your liberation’.
This line from the Mahāparinibbāṇa Sutta then no longer endorses the idea that we ought to delve inside our
essential ātman to realise the ultimate truth and become one with Brahman. It very simply means that nobody
else is responsible for our liberation, and we should take refuge in ourselves and take personal responsibility for
living the dhamma. This is a huge change in meaning. In the case given above (and other cases like it), it looks as
though Bhattacharya and his fellow Advaitins have, from a Buddhist perspective, fallen foul of a reification: taking
a simple reflexive pronoun to represent an enduring, ultimately existent entity!
This interpretation can be applied with equal force, I think, when Bhattacharya cites Dhammapada 160 (2015: 31):
‘attā hi attanō nāthō kō hi nāthō parō siyā /attanā hi sudanténa nāthaṃ labhati dullabhaṃ’. We can interpret this
as something like ‘one is indeed one’s own refuge. How can another be a refuge to you [one]? With oneself
properly controlled, one attains a refuge that is difficult to attain’. Does this read as though we ought to be
P a g e | 40
ourselves and the world are for the most part conventionally true, since they reflect conventions
that have been found to be useful in everyday practice’ (2013: 4). We can add to this definition
all linguistic designations naming objects and events.
Names represent entities that are
conventionally real, which is to say objects that do not possess an intrinsic nature (svabhāva) and
so are not basic, substantial and immutable. We will see that for Nāgārjuna, this characterisation
will come to account for all entities. For the Advaitins, it accounts for entities mistakenly assumed
to be intrinsically real when one has not realised the unity of the world in Brahman.
In
comparison, ‘[t]o say of a statement that it is ultimately true is to say that it corresponds to the
nature of reality and neither asserts nor presupposes the existence of any mere conceptual
fiction’ (2013: 4).31 This is true for both Buddhists and Advaitins. For somebody like Śaṅkara, the
Brahman is the only ultimate truth and ultimate reality. It is the only reality that exists, and
everything is a manifestation of it. For many Buddhists, it is an ultimate truth that there is no
substantial self: this conforms to how things ultimately are from an enlightened Olympian
meditating on the permanence of our essential ātman, or as though we ought to take responsibility for our own
liberation? I am inclined to read this in the same way as the line from the Mahāparinibbāṇa Sutta: reference to
atta (ātman) is as a simple reflexive pronoun, used in a conventional manner. That is to say that there are no
essential or ultimate connotations to its use. In both cases, then, we need not appeal to some essential feature or
principle called ātman, we simply have two distinct examples of the Buddha’s imploring practitioners to take
responsibility for their own spiritual journey instead of investing heavily in other people (and, ultimately, in him!).
A ‘conceptual fiction’ is something that is ‘thought to exist only because of facts about us as concept-users and
the concepts that we happen to employ’ (2013: 4). In other words, a conceptual fiction is something that we
impose onto the world, not something that exists independently of us as concept-users. The stock example is that
of a chariot: a chariot is nothing over and above the collection of part assembled in a specific way (‘chariot-wise’, if
you will). The chariot does not exist as an entity over and above or outside of this collection of so-arranged parts,
but it is a convenient shorthand for us to refer to ‘the chariot’ (or ‘the car’ or ‘the house’, for example).
31
P a g e | 41
perspective. 32 What then can we say about the Buddhist ultimate truth that there is no
permanent ātman?
Kamaleswar Bhattacharya (1928-2014) 33 argued in his Ātman-Brahman in Ancient
Buddhism that Buddhism’s famed ultimate truth has, in fact, been routinely mischaracterised
over the course of a rich Buddhist history. For Bhattacharya, the ātman denied by Buddhism is
that mistakenly associated with the jīva34 and not the ātman-Brahman of Advaita. Bhattacharya
contends that the Buddha was concerned with continuing the Upaniṣadic tradition and leading
his followers to knowledge of the transcendent ātman-Brahman: a theory that stands at stark
odds with the modern understanding of Buddhism as a denial of a permanent self. To this end,
Bhattacharya’s principal argument is that ‘[t]he Buddha certainly denied the ātman’, but with the
caveat that the ātman denied ’is not the Upaniṣadic ātman’ (2015: 207). What Bhattacharya
means here is as yet obscured by his apparent unwillingness to translate ātman. He elsewhere
writes that ‘[b]efore stating that Buddhism has denied the ātman, modern authors should,
therefore, have been precise as to which ātman is meant’ (2015: 34). This strikes me as a good
idea, and so I will briefly outline what ātman could mean.
32
There will be much more to say on this as we continue, and in much more detail, but for now, this rudimentary
understanding of conventional and ultimate should suffice for navigation of this work until such a point that more
detail is required.
Among Bhattacharya’s many publications are Brahmanic religions in ancient Cambodia, from epigraphy and
iconography; The Dialectical Method of Nāgārjuna (Vigrahavyāvartani); Some thoughts on Early Buddhism with
special reference to its Relation to the Upaniṣads.
33
Jīva is here taken to mean ‘individual self’ in the Advaitin sense rather than ‘immortal individual essence’ in a
Jaina sense: the Jaina understanding would have jīva equivalent with the ātman-Brahman without any monistic
import, whereas it is clear in Advaitin literature that this is not what is meant.
34
P a g e | 42
Lipner characterises the ātman as the ‘innermost reality [identity] of the individual, the
subtle essence’ (2010: 53), and the Chāndogya Upaniṣad (VIII.7.1) describes a nature (ātman)
that ‘is free from sin, free from old age, free from death, free from grief, free from hunger, free
from thirst, whose desires come true and whose thoughts come true’. Bhattacharya writes that
‘it is not outside of ourselves that we grasp the Real’ (2015: 10), but what does this actually mean?
Bhattacharya distinguishes between two senses of ātman – one which is real, and one which is
mistakenly taken by the unenlightened to be real, but which actually is not. The two senses of
ātman to which Bhattacharya is referring are the ultimately real paramātman of the ātmanBrahman, and the reified vijñāna-ātman (commonly translated as individual consciousness;
analogous to the jīva). Bhattacharya thinks that Buddhist doctrine of anātman explicitly negates
the latter in order to promote knowledge of the former. This distinction will prove to be the
bedrock for his thesis that Buddhism does not deny the spiritual ātman-Brahman, and instead
only negates a sort of personal ātman that is analogous with the human ego. He writes that
‘[t]he ātman is not the individual ego, but rather “the super-reality of the jīva, the individual ego”’
(2015: 5); ‘neither the Upaniṣads nor Buddhism deny the empirical reality of the individual. They
only deny its ontological substantiality’ (2015: 17)35; ‘when they [the Upaniṣads] state that the
ātman-brahman is the sole Reality, [they] are in fact denying that psycho-social being which men,
too often, consider as the ātman’ (2015: 34). It is clear that Bhattacharya thinks that people
I suppose that ‘empirical reality of the individual’ might mean a sense of self or the existence of this thing ‘an
individual human’ walking, talking and interacting with the world. Here, Bhattacharya’s point is that whilst both
the Upaniṣads and Buddhism acknowledge this sort of ‘individual’ conventionally, they do not acknowledge its
ultimate existence. A ‘conventional’ acknowledgement is simply to allow use of the term ‘individual’ as a type of
useful fiction with some instrumental value insofar as we have to navigate the ‘empirical’ (read: conventional, not
ultimately real) world. A conventional truth, entity etc. cannot ever have ‘ultimate’ status.
35
P a g e | 43
simply misapply the label ātman to that which it is not. In other words, the ‘psycho-social being’
referred to is actually anātman in the simplest sense – it is not the ātman. What, then, is the
ātman to which Bhattacharya is referring? How might we access it? In words not dissimilar from
those religious thinkers covered in §1, Bhattacharya (2015: 6) writes that
The ātman is the ‘inner ruler’ (antaryāmin) which resides in the
Universe but is distinct from and Unknown to the Universe. All our
activities derive from it; there is no other seer than it, no other
hearer, no other thinker and no other knower. Even so, it remains
invisible itself, inaudible, unthinkable, unknowable. The ātman is
the “inner light” (antarjyotis) of man.
The corporeal (saśarīra) ātman is mortal; it experiences pleasure
and pain. But the incorporeal (aśarīra) ātman, the authentic ātman,
is immortal; it is exempt from all pleasure and from all suffering.
We can see here that Bhattacharya thinks that the embodied (saśarīra) ātman, that
ātman which we directly experience, that we take as the ‘I’, is a mortal entity. It is the superreality behind this individual ātman, the ātman-Brahman, which is the true Reality. It certainly
sounds as though Bhattacharya is operating on a similar plane to those of the mystics discussed
in the previous section. He talks of an ultimate principle ātman (Brahman) in very similar terms
to Maimonides’ discussion of God: unknowable, unthinkable, somehow beyond this world. There
is a caveat, however, that separates Bhattacharya from thinkers such as Maimonides – this state
of ignorance need not be the case. We can come to know the ātman-Brahman, just not by
conventional means.
In the Ātmabodha, attributed to Śaṅkara, we read
dṛśyate śrūyate yadya brahmaṇonyanna tadbhavet /
tattvajn͂ānācca tadbrahma saccidānandamadvayam //64//
P a g e | 44
Whatever is seen, [whatever] is heard, is not other than Brahman.
[Through] clear knowledge of reality, [one sees] the Brahman;
Being, pure thought, bliss, unity.36
Knowing the ātman-Brahman, then, is a matter of ‘clear knowledge of reality’. For
Śaṅkara, it seems we can know ‘God’ (the Absolute, the Brahman), but we have to gain this
knowledge in a very specific way, alien to our usual means of knowledge. Swami Nikhilānanda
asserts in the commentary to the above śloka that it is via the attainment of ‘Right Knowledge’
that one knows the Brahman, and even the state of unknowing is necessarily Brahman; ‘[f]rom
the standpoint of Brahman even ignorance and its products, names and forms, are nothing but
Brahman’ (1962: 227-228).
Nikhilānanda also reiterates the stock Advaitin view that ‘what
appears as the manifold universe to the ignorant is realised by the illumined to be indivisible and
non-dual Brahman’ (1962: 226). The implications are, I think, clear. Indeed, Bhattacharya (2015:
5) asks ‘[w]hat is the Upaniṣadic ātman?’ He writes that it is
neither the body nor the totality of the psycho-physical elements
which make up the empirical individual. The ‘body’ is no more than
a ‘support’ (adhiṣṭhāna) of the incorporeal (aśarīra) ātman – of the
ātman ‘without ātman’ (anātmya, nirātman, nirātmaka). “Just as
the best is yoked to the cart, so the prāṇa (that is, the ātman) is
yoked to this body.” The ātman is the ‘inner ruler’ (antaryāmin)
which resides in the Universe but is distinct from and unknown to
the Universe. All our activities derive from it; there is no other seer
than it, no other hearer, no other thinker and no other knower.
Even so, it remains invisible itself, inaudible, unthinkable,
unknowable.
(Bhattacharya, 2015: 5-6)
36
Unless otherwise indicated, translations from Pāli and Sanskrit sources are my own.
P a g e | 45
He continues that the ātman actually is Brahman: ‘[o]ur essence is the ātman which is
beyond all these elements [psycho-physical components]’ (2015: 7), and it is ‘beyond the relation
implied by all thought’ (2015: 8). It is this Absolute which ‘opens to us on completion of our
evolution, comprises life, will and consciousness, although it goes beyond them’ (2015: 10). Thus
the ātman-Brahman is the cause of and substratum of everything despite being radically other
than all that it causes and supports.
All of this is orthodox Advaitin doctrine. For Śaṅkara, the world as it seems to us is not to
be taken at face value: true reality inheres only in the ‘infinite, eternal, unchanging, pure bliss
consciousness that is Brahman, or Paramātman’ (Betty, 2010: 216).37 Śaṅkara is here explicitly
equating the ātman with the Brahman: the two share an intimate connection that cannot be
overstated; Brahman is the entirety of the cosmos, and this has direct influence over how we are
to understand ātman as an essential principle. To this end, Katha Upaniṣad 6:2-4 says:
yadidaṃ kiṃca jagatsarvaṃ prāṇa ejati niḥsṛtam /
mahadbhayaṃ vajramudyataṃ ya etadviduramṛtāste bhavanti
//2//
bhayādasyāgnistapati bhayāttapati sūryaḥ /
bhayādindraśca vāyuśca mṛtyurdhāvati pañcamaḥ //3//
iha cedaśakadboddhuṃ prākśarīrasya visrasaḥ /
tataḥ sargeṣu lokeṣu śarīratvāya kalpate //4//
All the universe emanates from this breath [prāṇa; breath of life,
Brahman] and moves [in Brahman].
[That Brahman] causes great fear, like a poised thunderbolt. Those
that know this [Brahman] become immortal //2//
Paramātman is perhaps best rendered as ‘supreme reality’. The idea for Śaṅkara is that the Brahman is the
‘supreme reality’, and realising it – becoming it – is the highest possible attainment.
37
P a g e | 46
From fear of Him, fire burns; from fear [of Him], the sun gives
heat.
From fear [of him], Indra [lord of the gods] and Vāya [god of the
winds], and Death, the fifth, speed [on their way]. //3//
If, here and now, one is able to know [the Brahman] before the
disintegration of the body,
Then one is fit for embodiment in the created world. //4//38
Here, the Brahman is shown to be the universal essential principle from which all things
emanate, including individual selves. We can also clearly see that the principal spiritual project
for humans is to realise or know Brahman.
38
Readers might notice that there is an incongruity in the Sanskrit text of 6.4. My translation above illustrates that
the Sanskrit reads ambiguously, for it seems to be suggesting that if a person does gain knowledge of the Brahman,
they will nevertheless be reborn (embodied; śarīratva) into the created worlds (sargeṣu lokeṣu). This, though,
seems to fly in the face of established Advaitin soteriological doctrine, viz. that knowledge of the Brahman is the
key to liberation. Indeed, it seems to be contradicted only two verses earlier, where it is stated that ‘[t]hose that
know this [Brahman] become immortal’.
How, then, to solve this quandary? One possible route would be to amend sargeṣu to svargeṣu. In so doing, the
locative plurals ‘svargeṣu lokeṣu’ would then translate as ‘in the heavenly realms’ rather than the previous ‘in the
created world/realms’. This is a position that has been suggested by numerous scholars across the years, not least
by Robert Ernest Hume in translation of the Upaniṣads. Here, Hume provides a brief but useful discussion of the
issue at hand when he writes that
[t]he reading svargeṣu instead of sargeṣu would yield the more suitable
meaning ‘in the heavenly worlds.’ At best, the stanza contradicts the general
theory that perception of the Ātman produces release from reincarnation
immediately after death. Consequently Śaṅkara supplies an ellipsis which
changes the meaning entirely, and Max Muller hesitatingly inserts a ‘not' in the
first line. The present translation interprets the meaning that the degree of
perception of the Ātman in the present world determines one's reincarnate
status.
(Hume, 1921: 359)
We might remain confused as to how one might become ‘embodied’ in the ‘heavenly realms’ (realms which, I
think, we need to understand as Brahman in its totality as understood from a liberated viewpoint) when the thrust
of Advaitin philosophy is – as we shall see – focussed on turning inward. Advaitins following Śaṅkara’s lead think
that we should concern ourselves with knowing the Brahman rather than concerning ourselves with material
things in the material world. To this end, Vasu suggests that we should understand ‘embodiment In the heavenly
realms’ as meaning something like residing in a spiritual body which ‘is immaterial and consists of the bliss and
intelligence’ rather than thinking that it means we are ‘reborn’ in the conventional saṃsāric sense (Vasu, 1905:
175). .
P a g e | 47
What does this mean in terms of ātman and Brahman? Chris Bartley puts it succinctly
when he says that ‘[t]he cosmos is thought of as a single whole that has essence and this is what
is called the Brahman. Individual selves, microcosm[ic] versions of the cosmos, too have essence
and this is what is called ātman. If essence is indivisible, the Brahman equates to ātman’ (Bartley,
2011: 10). This reading is further strengthened by the Upaniṣad’s assertion that those that fail
to reach liberation through knowledge of Brahman are doomed to return and ‘put on’ a body –
further driving home the notion that we are not identical with our bodies, but with the essential
principle contained inside them.
We might think that this is far removed from the ‘traditional’ views of Buddhism, and on
the face of things, we would be right. One might reasonably assume that Buddhism must at least
have been sufficiently different to its Brahminical contemporaries; else, it would not have
developed as a separate tradition to the Vedic schools at all. Thomas Wood has previously stated
that ’a full reconciliation of the Vedānta and Buddhism was manifestly impossible’, and that ‘no
orthodox Buddhist could have upheld the doctrine of an Absolute which is unchanging, pure
consciousness’ (1992: 73).
It is surprising, then, that several scholars find textual justification to support the thesis
that the Buddha tacitly endorsed the ātman. According to Bhattacharya, the Buddha ‘simply said,
in speaking of the skandha/khandhas, ephemeral and painful, which constitute the psychophysical being of a man: n’ etaṃ mama, n’ eso ‘ham asmi, na m’ eso attā, “This is not mine, I am
not this, this is not my ātman”’ (Bhattacharya, 2015: 6). In other words, whilst the Buddha did
indeed deny that the interaction of the skandhas amounted to the ātman, by negating what is
P a g e | 48
not ātman, he tacitly endorsed – in virtue of this negation – a type of ultimate, spiritual ātman:
the Brahman. The upshot of this is that Buddhism is in fact some sort of revised Brahmanism, or
more specifically, as Bhattacharya eventually implies, a type of early Advaita Vedānta later
inherited by Śaṅkara. This reading is anathema to the developed Buddhist understanding of
anātman with which we are familiar today, which is usually taken as denying any and all forms of
‘self’, whether the self is understood as the reification of interactions between the skandhas, or
a more permanent transcendent ‘Brahman’ as attested in the Upaniṣads. It is again worth
mentioning that Bhattacharya here leaves ātman untranslated. It is possible that this is because
it is an ambiguous word that derives its precise meaning (insofar as one can be determined) from
the context in which it is used – it can be tricky to translate accurately. It might also be the case,
though, that Bhattacharya leaves ātman untranslated because when he uses the word he has a
very specific context in mind, namely the blissful, ultimate nature described in the Upaniṣads.
However, this context need not be shared by the utterances of the Buddha. A possibility that
Bhattacharya is loath to consider is that the Buddha is simply stating ‘this is not my nature’
without implying that there actually is a different, blissful nature.
For Bhattacharya, then, it is specifically the transcendent Brahman of the Upaniṣads that
the Buddha is indirectly affirming, and achieving nirvāṇa thus amounts to realising this Brahman.
His arguments for this are numerous and include some choice readings of Mahāyāna sūtras; a
specific use and understanding of svabhāva in Madhyamaka thought; presupposing the ātman
as (although transcendental!) an existent thing, and a very specific interpretation of the negative
method. All of these factors and more will be addressed in what follows, but first I will give some
preliminary remarks on Bhattacharya’s opening gambit: the thesis that the Buddha, through
P a g e | 49
denying the ātman in one sense, actually endorses a more fundamental sort of ātman in another
sense.
The starting point of Bhattacharya’s argument here is the idea that the negative method
necessarily implies a positive existent. In discussing the negative method that is employed by the
Buddha, Bhattacharya specifies that for something to be negated – in this instance, things to be
identified with the ātman – there must be an ultimately real ground underpinning that which is
negated. This sort of objection is particularly familiar to the Mādhyamika, and has enjoyed a
resurgence in popularity via such philosophers as Giuseppe Ferraro in recent years (Ferraro,
2013). Throughout, I will assess the merits of Bhattacharya’s claim in light of his various
arguments supporting it, taking a specifically Mādhyamika tack throughout. There are several
reasons for this. First, though Bhattacharya is at pains to point out that various interpretations
of parts of Pāli Buddhism (and so the canonical Tripiṭaka (Pāli: Tipiṭaka)) support his conclusion
that the Buddha affirms by negation what he calls the ātman-Brahman (effectively making the
Buddha an early Advaitin - more on this later). Consequently, he thinks that the established
schools of the Abhidharma are mistaken in how they address essence (and thus ātman) (2015:
38-39). Second, Bhattacharya thinks that it is the Mahāyāna schools ‘which put things right’
(2015: 39), and so argues that it is the Mahāyāna that provides the real scope for his
interpretation to flourish. I intend to examine this contention and show why I think there is at
least one Mahāyāna school – Madhyamaka – which should not, if it is to remain true to its own
foundations, take such a view. Third, Bhattacharya endorses Ramanan’s position equating
svabhāva (own-being), svarūpa (own-form; own-nature) and ātman as ‘the essential nature…of
the individual as well as of all things’ (Bhattacharya, 2015: 144, note 249). It is at this same point
P a g e | 50
that Bhattacharya claims (via Venkata Ramanan) that Madhyamaka accepts ātman as svabhāva
on the proviso that this svabhāva is not a ‘separate, substantial entity inhabiting the body of each
individual’ (2015: 144, note 249). There are, prima facie, some problems with this stance when
examined from a Buddhist perspective, especially in light of the anātman doctrine, and especially
from the Madhyamaka viewpoint. It appears as though Bhattacharya has reified the notion of
svabhāva (as a ‘separate substantial entity’ which somehow inhabits a body) only to claim that
this reified sort of svabhāva is not what is accepted by Mādhyamikas. It is not immediately
obvious to me that this is at all what anybody claims – entities either have or lack a svabhāva,
they do not have or lack types of svabhāva, and so the distinction is redundant. In any case, how
these notions are dealt with by the relevant Buddhist and Advaitin traditions will have great
influence over how we should understand notions of truth; notions of truth that are integral to
understanding how both the soteriological methods and goals of each school operate.
For these reasons then, I will emphasise why I think that Madhyamaka Buddhists would
not – or at the very least should not – endorse an ātman-Brahman of the same sort as the
Upaniṣads. This inevitably means that I will dispute Bhattacharya’s argument that for both
Advaitins and Buddhists, the ultimate truth is identical (viz. the ultimate truth is the ātmanBrahman). I will also examine how and why the Mahāyāna leaves itself open to such comparisons
by paying some attention to the various ideas connected to essence or the Absolute at play in
the background to its philosophy (svarūpa; svabhāva; dharmakāya and tathāgatagarbha;
śūnyatā). The sum of these efforts will be, I hope, a refutation of the thesis that the Buddha was
a proto-Advaitin who tacitly endorsed the ātman-Brahman and further a refutation that the
ultimate truth (and thus the ultimate destination) is identical in both Advaita Vedānta and
P a g e | 51
Buddhism. I will, then, take the Mādhyamika approach to this issue and argue that Nāgārjuna
could not have (and would not have) under any circumstances endorsed a notion of an Absolute,
ultimate reality. I will do this by extrapolating key Madhyamaka principles and illustrating that
accepting an Absolute of any sort is necessarily contrary to the Middle Way as Nāgārjuna
designed it, and that śūnyatā as devised and elaborated by Nāgārjuna cannot provide the
ultimate grounding of existence that commentators such as Bhattacharya suppose it can.
2.1 Bhattacharya’s Argument
Let us start, then, at the beginning. Bhattacharya opens by citing a Cambodian inscription
that curiously appears to both acknowledge that impersonality (given here as nairātmya) is
incompatible with supreme-selfhood (paramātman), but nevertheless claim that the Buddha
taught insubstantiality as a means to achieve realisation of supreme-selfhood (Bhattacharya,
2015: 1).39 The idea presented here is the linchpin that holds together Bhattacharya’s entire
thesis: using no-self to deny what the ātman is not necessarily leaves to one side that which the
ātman really is. The inscription, it is worth bearing in mind, is attributed to the reign of
Rājendravarman, and so is thought to be dated somewhere between 944 and 968 CE: a
considerable amount of time after the advent of the Mahāyāna, and, importantly, centuries after
I think that interpretations such as Bhattacharya’s are to a large degree a result of the tendency to translate all
instances and variations of ātman as ‘self’. Being a multi-faceted term, we can sometimes translate ātman as ‘I’ in
the sense of a simple reflexive pronoun, as ‘self’, or as ‘essence’ or ‘nature’. If, for example, we use ‘nature’
instead of ‘self’, we have something along the lines of insubstantiality is taught as the supreme nature of
phenomena. Another way to put this would be to say that the nature of things is to lack svabhāva. The Buddhist
might still be uncomfortable with talk of a ‘supreme nature’, but its occurrence can be accounted for by
distinguishing between conventional and ultimate truths and the limitations of, aims of and usefulness of language
in relation to this distinction. It might refer only to a concept intended to prove a point but itself empty of any
substantial existence.
39
P a g e | 52
the
dissemination
of
important
Mahāyānist
texts
such
as
Nāgārjuna’s
seminal
Mūlamadhyamakakārikā (MMK). 40 Supporting Bhattacharya’s reading is a citation from the
Mahāyāna-Sūtrālaṃkāra translated by S. Lévi (and slightly modified by Bhattacharya) as
In utterly pure Emptiness, the Buddhas have attained to the summit
of the ātman, which consists in Impersonality. Since they have
found, thus, the pure ātman, they have reached the heights of
ātman.
(Bhattacharya, 2015: 2)
The translation goes on to include a commentary, 41 specifying that the ātman of the
Buddhas ‘consists in the essential Impersonality’, which is in turn ‘absolute Thus-ness’, and is also
‘ātman in the sense of the own-nature of the Buddhas’ (2015: 2).42
The Mahāyāna-Sūtrālaṃkāra itself is a Yogācāra text attributed to Asaṅga (himself a
Yogācārin), and so is a later development of (or deviation from) earlier Mahāyāna thought. As
Conze writes, despite the Mādhyamikas and Yogācārins being ‘quite distinct in their interests and
intentions’, the Yogācārins nevertheless ‘regarded the Mādhyamika doctrine as a preliminary
stage of their own, which however missed the true and esoteric core of the Buddha’s teachings’
40
There are notorious difficulties with dating ancient texts and their authors. I do not intend to cover these
controversies here, but will rather adhere to the generally accepted – if speculative – timeframes generally used by
scholars of Buddhism.
Precisely whose commentary this is, we do not know. Bhattacharya states in his notes that ‘We do not touch
here on the much discussed question of the author or authors of this text and of its commentary’. Other
translations, like the recent 2014 Ornament of the Great Vehicle Sutras, claim to make use of Vasubandhu’s
commentary, but in this case, do so from the Tibetan rather than the Sanskrit. There are disagreements as to the
authorship, and it is by no means clear that Vasubandhu did write the Mahāyānasūtrālaṃkārabhasya. For the
sake of argument, I will assume that he did write it, and that we can interpret it in the same vein as his other texts.
41
Following the spirit of this translation, Bhattacharya, in note 249 (via n.7), endorses K. Venkata Ramanan’s bold
claim that not only do Mādhyamikas accept ātman as an ‘essential nature’, but they also equate it with the
svabhāva and svarūpa ‘of the individual as well as of all things’ (2015: 114). .
42
P a g e | 53
(Conze, 1993: 38-39). Conze also notes that, for their part, the Mādhyamikas regarded the
Yogācārin project as ‘a quite incomprehensible perversity’ (1993: 39), presumably owing to the
supposedly idealist position occupied by later Yogācārins. This also speaks to the variation of
thought between schools loosely united under the ‘Mahāyāna’ banner, and serves as a
preliminary warning against prematurely lumping together doctrinally divergent Mahāyāna
schools. There are just as many points of disagreement as there are points of agreement! I here
tentatively suggest that this could prove to be a reason why Bhattacharya’s contention that it
was the Mahāyāna which ‘put things right’ regarding the nature of reality might be too broad a
stroke.
Regarding the translation given by Bhattacharya, I note with some interest that ātman
again remains untranslated. This is most likely because Bhattacharya wants ātman to be taken
in the Advaitin sense, viz. as an instantiation of the Brahman. The Brahman is, of course,
necessarily impersonal: it is not embodied. Ultimately, Brahman is beyond substance-attribute
distinctions (Nirguṇa Brahman). On the other hand, Brahman is the only substance in the world:
the cause of the universe, and the very reality that underpins it (Saguṇa Brahman). The apparent
distinction (Nirguṇa-Saguṇa) is not a real distinction within Brahman, it is simply illustrative of
the ‘limits of conventional language in describing brahman’ (Rambachan, 2006: 89). Rambachan
elsewhere adds regarding the Brahman that ‘[i]ts nature transcends all definitions that are based
on distinctions’ (2006: 89). Nevertheless, insofar as Brahman is the sole reality responsible for
all existence, we might describe it as ‘substantial’. Indeed, as Rambachan points out, despite
Brahman transcending all definitions and so on, ‘there is no object that enjoys a separate
P a g e | 54
ontological existence and nature from brahman’ (2006: 88).43 This speaks to the tension inherent
in discussions of entities (though Brahman is technically beyond the concept ‘entity’) that
allegedly transcend linguistic designations: nothing we say will ever capture anything about them
from an ultimate perspective (and consequently, we cannot communicate any ‘truths’).
However, if Brahman is pure Being, then it must necessarily exist, and so whilst an Advaitin might
be perturbed by the idea of imposing any such label onto the Brahman, I think it relatively
uncontroversial to assert that according to Advaita, the Brahman is necessarily existent, and so
is in an important respect, substantial.44
I think that because Bhattacharya is beginning from an Advaitin position, he is always
inclined to understand ātman in the Advaitin context. It is clear enough from is argument that
he wants his readers to understand ātman in this way, too. It is possible, however, to understand
this specific ātman in a weaker sense. Bearing in mind that the Mahāyāna-Sūtrālaṃkāra is, after
all, a Buddhist text, I feel vindicated in taking this small step. If we choose to translate ātman as
‘nature’ in a weak sense rather than ‘self’ in the strong sense of a permanent, immutable essence,
then the meaning of the extract changes a little. We now have a meaning more like
In utterly pure Emptiness, the Buddhas have attained to the summit
of their nature, which consists in impersonality. Since they have
To this end, Anderson also writes that ‘it should be noted that Advaita Vedanta takes it for granted that there is
‘being’. Its project, ontologically, is to clarify what this ‘being’ entails. By definition ‘being’ is not only that which
cannot be subrated [sublated], but also cannot come into or go out of existence’ (2012: 276). ‘Being’ is
synonymous with ‘Brahman’, and thus is substantial insofar as it is the only existing thing. All experienced
phenomena enjoy reality only because they are Brahman.
43
44
I feel it important to note that an incorporeal entity can still be substantial.
P a g e | 55
found, thus, their pure nature, they have reached the heights of
their nature.
Or:
In pure emptiness, the Buddhas have achieved the pinnacle of their
nature: a state of impersonality. Having discovered their pure
nature, they have arrived at their supreme nature.45
Leaving ātman untranslated leaves the door open to understanding it in the Advaitin
sense advanced and preferred by Bhattacharya. Replacing ātman with nature, though, adds
another possibility. No doubt an Advaitin would simply object that the ātman-Brahman simply
is the nature of the world (and so of all that are within it). My slight change to the extract can,
after all, still be read in such a manner. It can, however, also be read differently. In line with the
orthodox Buddhist aversion to asserting ‘self’, we might understand the extract as using ‘nature’
in a conventional sense (and thus not necessarily implying any ultimate existence), then what we
have is the Buddhist text simply using an empty concept (that is empty of svabhāva) to
demonstrate a point.46 What the Buddhists have realised in emptiness (śūnyatā) is thus that the
nature of things is insubstantiality/impersonality, and this is the highest realisation regarding the
nature of things. In other words, the greatest realisation about the nature of things is that all
things lack intrinsic existence (svabhāva). This need not, and indeed does not imply that the
‘nature’ realised has a substantial existence behind it (and so can avoid recourse to something
like the Brahman). Instead, it can simply be a concept, referred to for the sake of ease and
45
46
śūnyatāyāṃ viśuddhāyāṃ nairātmyātmāgralābhataḥ / buddhāḥ suddhātmalābhitvād gatā ātmamahātmatām //
Again, something is conventionally existent according to Buddhist terminology when it lacks an intrinsic
existence (i.e. svabhāva). Something is ultimately existent when it possesses intrinsic existence (svabhāva).
P a g e | 56
communicability, without really referring to anything substantial. This could, I think, begin to
form the basis of a coherent Madhyamaka interpretation.
I also note with some interest that translations of the commentary to the MahāyānaSūtrālaṃkāra do not tend to specify whether or not their talk of paramātman – the ‘great self’ –
is in a saṃvṛti or paramārtha sense. I think that we can deduce from Vasubandhu’s explication
of the trisvabhāva doctrine that both the verse and commentary are saṃvṛti and so nothing but
conventional designations. 47 This means that any talk of own-nature of the buddhas is a
conceptual construct that is ultimately unreal. The phrasing is the way it is to allow an
unenlightened mind to comprehend the splendour of the awakened in a sort of metaphor. The
buddhas have not actually climbed to the summit of the great ātman, because in the final analysis
– the empty perfected aspect – such a thing is unreal. We can explain the use of ‘own-nature’
and ‘supreme self’ as used in both Thurman’s translation and the translation of the
Dharmachakra Translation Committee in a more nuanced way. The own-nature referred to is –
according to Vasubandhu’s trisvabhāva – subject to the interplay of the three natures. The first
47
Explaining the trisvabhāva, Gold (2015: 149) writes that
[t]he first nature is the fabricated nature, which is the thing as it appears to be,
as it is erroneously fabricated. Of course, to use this term (“fabricated”) is to
indicate the acceptance that things do not really exist the way they appear.
This is a thing’s nature as it might be defined and explained in ordinary
Abhidharma philosophy – its traditional svabhāva, but with the added proviso
that we all know that this is not really how things work. . . The second nature is
the dependent nature, which Vasubandhu defines as the causal process of the
thing’s fabrication, the causal story that brings about the thing’s apparent
nature. The third nature, finally, is the emptiness of the first nature – the fact
that it is unreal, that the appearance does not exist as it appears.
P a g e | 57
aspect of the three natures, parikalpita, is the fabricated nature or fabricated aspect. This is the
thing as it appears to exist and is usually associated with svabhāva. The appearance is not real,
as the Yogācārin knows. Therefore, the own-nature referred to in the commentary to IX.23 of
the Mahāyāna-Sūtrālaṃkāra is an appearance, but it is not ultimately real. The second aspect is
paratantra, or dependent nature or aspect. This is, as Gold (2015: 149) writes, the causal story
that brings about the parikalpita nature. In the case of the own-nature of buddhas, we might say
that the paratantra is the culmination of events, actions and personality traits that lead us to
view them as having achieved the ‘great ātman’. It is the causal story that leads us to impose
conceptual constructions onto the world. The final aspect, the pariniṣpanna, or created nature
or aspect, is the emptiness (śūnyatā) of the fabricated nature: the knowledge that it is not
intrinsically real and is an imposition resulting from a causal flux. In the case of the own-nature
of the buddhas described at both Mahāyāna-Sūtrālaṃkāra IX.23 and in the corresponding
commentary, we must now see that such an own-nature is unreal.
What we are left with, then, is a description that reifies emptiness as some sort of
substantial entity. Asaṅga writes that it is within śūnyatā that the buddhas climb to the summit
of the great self, implying that emptiness and selflessness (anātman) are paramātman. But by
Vasubandhu’s own methodology, this conclusion cannot be left standing. The very fact that the
own nature described is empty of intrinsic existence means that it is mischaracterised: a
svabhāva is for Vasubandhu always parikalpita. When we apply Vasubandhu’s analysis, the
passage actually alludes not to an Upaniṣadic ātman-Brahman as asserted by Bhattacharya above
and
suggested
by Thurman
in his footnotes to the
2004 translation
of the
Mahāyānasūtrālaṃkārabhāṣya (2004: 82, note 36), but to a denial of the ultimate existence of
P a g e | 58
that nature. The passage is saṃvṛti and so discusses and lauds the attributes of the buddhas in
conventional, worldly language. What it does not do is give an ultimate account, at least if we
use Vasubandhu’s own arguments as a yardstick by which to measure the nature of phenomena.
Reifying the emptiness of phenomena as a self, then, is anathema to Vasubandhu.48 Applying his
own account of trisvabhāva, we ought to come to the conclusion that the buddhas have achieved
a great feat in realising that the nature of things is that there is not a nature of things. That is to
say that they have seen that all phenomena lack svabhāva and no longer view the world in
svabhāvic terms.
This is supported, I contend, by the subsequent verse’s commentary, which provides
arguments as to why Buddhahood is said to neither exist nor not exist. This is precisely because
a buddha is selfless given that ‘suchness’ (thusness, relating to the tathāgata) is characterised
‘by the (ultimate) nonexistence of persons and things’ (Thurman, 2004: 82). A nature (svabhāva)
is a thing, and so ultimately does not exist. It is also worth noting that part of the Sanskrit
compound pudgaladharmābhāvalakṣaṇatvāt (‘given that it is characterised by the nonexistence
of dharmas and persons’) uses the Sanskrit word for person, pudgala. This is not an interesting
observation until we realise that for Vasubandhu, ‘pudgala’ is a synonym for ātman. In the
Abhidharmakośabhāṣya (1991: 1327), Vasubandhu spends considerable time attacking Brahmin
monks for contemplating the ātman. He writes that there is no ātman, no pudgala, over and
above the skandhas. Talk of ātman and pudgala, thinks Vasubandhu, is simply a reification of
impermanent impersonal interactions between the five skandhas. It is a mistake of
48
Vasubandhu’s attitudes to ‘self’, svabhāva, will be discussed in §4.
P a g e | 59
consciousness, which is itself causally conditioned by interactions between the mind (manas) and
the dharmas that constitute the psycho-physical world (1991: 1326).
Like all Buddhists,
Vasubandhu thinks that ultimately, this mistaken reification can be addressed.
For Vasubandhu, then, buddhahood is characterised by the absence of ultimate entities,
namely persons (substantial selves) and ‘things’ (bhāva: also ‘being’, ‘existing’).
Natures
(svabhāva) fall under the latter categorisation, and so ultimately, they cannot exist. There can
be conventional ‘characteristics’ (lakṣaṇa) that describe a lack of something, and so there is no
obvious problem here in arguing that a characteristic of Buddhahood is that there are no natures.
We know that a characteristic can describe a privation because Vasubandhu uses the example of
fading shadows in vision, claiming that a fading shadow cannot be said to be (ultimately) existent
because ‘its characteristic is the nonexistence of heat or shadows’ (Thurman, 2004: 83). 49
Similarly, it cannot be said to be (ultimately) nonexistent either, because we experience the
fading of shadows. The point is that a fading shadow does not exist as it appears (2004: 83).
With all that said, it is relatively easy to see how Bhattacharya might reach his conclusion
that the Mahāyāna-Sūtrālaṃkāra supports his thesis that Buddhism aims at knowledge of the
true ātman-Brahman. If we were to read the quotation only slightly differently, we could replace
the idea of ātman with that of a specific understanding of dharmakāya and have next to no
practical difference in meaning. Of course, dharmakāya is itself a troublesome notion that is
often interpreted in radically different ways, and to some degree, this sort of difficulty is par for
49
We might also say that a vegetarian foodstuff is characterised by a lack of meat. Anarchy is characterised by a
lack of government, and so on. In the above cases, the designation ‘thusness’ and ‘buddhahood’ are characterised
by a lack of permanent immutable phenomena.
P a g e | 60
the course when translating from Sanskrit.50 The term itself loosely translates as truth-body (and
was in its earliest forms most likely rendered as something like the body of teachings [of the
Buddha]) but is understood differently depending on where we care to look. It is not unusual,
for example, for dharmakāya to be understood as reality-body; the basis of reality, the
underpinning substrate from which all must stem, the knowledge of which brings liberation. This
is a normal, common theme in some Mahāyāna literature. It is not some fringe idea that operates
outside of accepted Buddhist doctrine; it is accepted Buddhist doctrine for numerous Mahāyāna
schools.
Indeed, Tillemans writes that dharmakāya can be understood as ‘the Buddha's
omniscient mind or the buddhas' omniscient minds (=jñānātmakadharmakāya, ye shes chos sku)
or the absolute and unitary nature of those minds (=svābhāvikakāya, ngo bo nyid sku)’ (2019:
641).
This talk of an essential, unconditioned substratum is the starting point for Bhattacharya’s
thesis, and we can see that it is not without precedent. A featureless ultimate substratum from
which the totality of the experienced world emanates could just as easily be the ātman-Brahman
as it could be the Mahāyāna dharmakāya. This will be covered in detail in section 5. Before I get
to any of that, though, I will proceed in fleshing out Bhattacharya’s own reasoning for his
conclusion that the Buddha actually tacitly endorsed a sort of supreme ātman. Bhattacharya
cites the commentary to the Ratnagotravibhāga – another Yogācāra text – at some length in both
Sanskrit and in translation, in what I assume is an attempt to prove two things: first, that the
I discuss in some detail how a Mādhyamika might opt to interpret a word and concept as troublesome as
dharmakāya in §5.
50
P a g e | 61
‘fourfold misapprehension’ illustrates that there is an ātman to be discovered, and second that
the concept of nairātmya is not, in fact, incompatible with that of paramātman. It is worth
looking at both of these two points in some more detail. The fourfold misapprehension in
particular is given significant attention. The translation given by Bhattacharya states that
The idea of the permanent in what is impermanent, of happiness in
what is sorrowful, of the ātman in what is non-ātman, of the pure
in what is impure, that is to say, in such things as corporeal form,
etc., that is what is called the fourfold misapprehension.
(Bhattacharya, 2015: 3)
Again, Bhattacharya leaves ātman untranslated. If we were to translate it in the spirit of
the paragraph, the relevant misapprehension would look something like ‘nature in what is not
nature’, which is to say that it is a misapprehension to impose the idea of an essential nature
onto something that lacks an essential nature. This is not necessarily to say that there are other
things that possess or constitute an essential nature, but might simply mean that it is mistaken
to reify natures when no natures can exist. This interpretation would be in line with the wider
Buddhist doctrine of anātman and particularly in line with the Madhyamaka doctrine of śūnyatā.
In the latter case, to seek nature in what is not nature is analogous to reifying svabhāva when
none can be found: for the Mādhyamika, what is empty (śūnya) of svabhāva is necessarily
without a nature or a self (anātman). As Bhattacharya elsewhere notes (2015: 144), possession
of svabhāva is possession of an ātman, and so to lack svabhāva is to lack an ātman. In other
words, it is not about seeking ātman in the wrong place with the implication that there is a right
place to look, but about reifying ātman where there is no possibility of its existing.
P a g e | 62
Nevertheless, for Bhattacharya, that the ātman is being sought in what is non-ātman
necessarily means that there must be something that is ātman. In other words, any negated
object requires a real existent that is to be negated. On this point, Murti has this to say:
Negation itself is significant because there is an underlying reality –
the subjacent ground. If there were no transcendent ground, how
could any view be condemned as false? A view is false because it
falsifies the real, makes the thing appear other than what it is in
itself. Falsity implies the real that is falsified.
(Murti, 2016: 234-235)51
This view supports Bhattacharya’s contention that in order for the Buddha to say what is
not ātman, there must be something else that actually is ātman. It is an oft-repeated line,
especially when talking about the conventional world of conditioned phenomena in Buddhism.52
In 2013, Giuseppe Ferraro, in criticism of the semantic interpretation of emptiness put forth by
Mark Siderits, wrote that
A first evident logical weakness of semantic interpretation is that
concepts such as ‘conventionality’ or ‘conceptual being’ are
inconceivable without admitting some idea of reality or
independent being. Therefore, the phrase “ultimate reality (PO)
does not exist and everything is only conceptual reality (SO)” is
51
Also cited in Bhattacharya (2015: 36).
The significance of the difference between the two truths of ‘conventional’ and ‘ultimate’ is mentioned in MMK
24.8-9:
52
dve satye samupāśritya buddhānāṃ dharmadeśanā /
lokasaṃvṛtisatyaṃ ca satyaṃ ca paramārthataḥ //8//
ye ‘nayor vijānti vibhāgaṃ satyayor dvayoḥ /
te tattvaṃ na vijānanti gambhīre buddhaśāsane //9//
The Dharma-instruction of the Buddha rests on two truths:
conventional truth and ultimate truth.
[Those] who do not know the distinction between the two truths,
they do not understand the Buddha’s profound teaching.
P a g e | 63
inconsistent from a logical point of view. Indeed, if we exclude that
‘real’ might exist beyond the conceptual, we are not ‘eliminating
reality’ but are rather saying that the conceptual is the only ‘real’.
(Ferraro, 2013: 211)
Ferraro is really disputing the efficiency of dealing with the two-truths through the lens
of the semantic interpretation and so his immediate focus is slightly different to that of
Bhattacharya, but the underlying point is the same: we can only negate something in virtue of its
actual existence. Where Bhattacharya thinks that we can only deny what is not ātman in virtue
of there being something else that is ātman, Ferraro thinks that we can only make sense of
conventional (conditioned) existents in virtue of there being ultimate (unconditioned) existents
that transcend the conventional: the ātman-Brahman is one such existent. The ātman-Brahman
is, according to Bhattacharya, entirely compatible with the Buddhist doctrine of insubstantiality
(2015: 11) in virtue of its own incorporeality (2015: 5-6). On such a reading, scholars of Buddhism
like Ferraro appear to support Bhattacharya’s overarching thesis: there simply must be some
ultimately real hyper-reality that grounds existing beings.
We have, though, compelling reason to question the way in which ‘real’ is used by both
Ferraro and Bhattacharya.53 Contained within such uses of ‘real’ is – unsurprisingly – the idea of
a deeper reality with a privileged ontology, and this is, at least on the surface of things,
particularly un-Buddhist. Indeed, whilst Ferraro manages to stop short of claiming that the real
is incorporeal (instead arguing that the real is simply a mind-independent existence (Ferraro,
‘Thus, it is not outside of ourselves that we grasp the Real: through all eternity the Real is present in us, but we
do not see it, blinded as we are by our false conceptions’ (Bhattacharya, 2015: 10).
53
P a g e | 64
2013: 211)), Bhattacharya does not, claiming that the real is simply the ātman-Brahman. This is
in turn our essential nature (further, it is the essential nature of everything), our true incorporeal
being.
We now have a situation where not only are negations assumed to require an ultimate
reality or existent in order to make any sense, but the ultimate reality so assumed is treated as
something existent that is there to be discovered or known. In making this move, both Ferraro
and Bhattacharya fail to recognise that a discussion of – and negation of – a concept like the
ātman or the ultimately real (I make this distinction because Ferraro does not equate his idea of
ultimate reality with the ātman) does not actually depend on the independent existence of an
ātman or ultimate reality. A Buddhist thinker need not accept that the ātman exists in order to
state that the skandhas do not constitute it. All that is required is the recognition that an idea or
concept is under discussion, and regardless of whether or not this concept has a corresponding
existent entity, simply knowing what the concept means and entails is enough to facilitate a
discussion of it. In other words, all the Buddhist really need acknowledge is the concept of the
ātman.
It thus strikes me that in saying ‘these things are not ātman’, the Buddhist need not
implicitly affirm a belief in the ātman, but need only be speaking to their opponent in mutually
understood terms.
When a Madhyamaka thinker reduces all empirical things to the
‘conventional’ level, they are not necessarily admitting to the existence of two distinct levels of
existence with ‘conventional’ somehow beneath the privileged ontology afforded to existents on
the ‘ultimate’ level. In fact, all they really need admit to is either their understanding of the
P a g e | 65
Buddhist proclivity to explain experience via ‘ultimate’ existents, or their recognition that there
appears to be two-levels of truth (and thus existence), and are simply talking in terms that other
Buddhists – and indeed non-Buddhists – would understand; presumably before jettisoning such
notions.54 They need not believe an ultimate plane actually exists in order to dispute how it
might theoretically work, just as I need not believe that God exists in order to make sense of a
conversation about the attributes of God with an atheist. I can talk of Harry Potter-style
Dementors with my young cousin, but I think it a stretch to say that either of us need accept their
actual existence in the empirical world (though maybe my cousin would!).
Another convenient way to think of this is as follows: it is not necessary for a unicorn to
exist independently in the world for me to dispute with a friend that it is not the same as or
equivalent with the Minotaur. If my friend says ‘this unicorn is the Minotaur’, I can reasonably
respond with ‘that unicorn is not the Minotaur’ despite my knowledge that neither the unicorn
nor the Minotaur empirically exist in the world. In so doing, I cannot reasonably be accused of
tacitly affirming the ultimate existence of both unicorns and Minotaurs. All that is really
demonstrated is that both my friend and I understand what the idea of a unicorn involves (its
status as a horse with a horn growing from its head) and what the idea of the Minotaur involves
(its status as a creature with the body of a man and the head of a bull).55 We can say that as long
as we understand the idea or concept of something, it does not much matter if it exists
54
This jettisoning of conceptual thought is, I believe, the crux of Madhyamaka thought. Throughout the course of
this paper, we will see some similarities with Advaita in this regard, too.
55
Obviously, we need to understand what a horse, horn, man and bull are, too.
P a g e | 66
independently of the mind or if it does not. Being able to categorise ideas is what really matters,
not whether there ever existed entities or objects falling under those ideas.
The same
understanding can be applied to Siderits’ argument regarding the Madhyamaka use of
‘conventional’ as critiqued by Ferraro. The Mādhyamika need not accept the existence of an
ultimate plane to ‘ground’ the reality of empirical objects or of anything else. Instead, all they
really need is an understanding of the theoretical implications of the ‘ultimate’, so that when
they refer to the two-truths, they do so with a specific goal in mind: eventually dispelling any and
all notions of ultimacy because notions of ultimacy qualify as metaphysical views (and
metaphysical views lead to attachment). 56 It is a simple fact that an entity need not exist
independently of mind in order to be talked about or entertained in thought.
Given the Buddha’s famed propensity for avoiding extremes and advocating a ‘middle
way’, it is at least possible that his apparent hesitance to deny ātman outright (instead choosing
only to say what does not constitute the ātman, namely identifying the ātman as corporeal, as
resulting from the skandhas) stems from a reluctance to commit to either a permanently existent
or a nonexistent ātman. This would, after all, present as either an eternalist or annihilationist
point of view: both are to be avoided according to the principle of the middle way.
Bhattacharya, we have seen, also believes that the ātman is ‘ātman in the sense of the
own-nature of the Buddhas’, and in his footnotes, he endorses the position of K. Venkata
56
This relies on reading Madhyamaka philosophy according to the semantic interpretation of emptiness, a position
to which I am generally sympathetic. I also acknowledge, however, that this is not the only reading, or indeed the
dominant reading of Madhyamaka.
P a g e | 67
Ramanan, which equates svabhāva, svarūpa and the ātman with ‘the essential nature…of the
individual as well as of all things’ (Bhattacharya, 2015: 114, note 249). When translating a section
of the commentary to the Ratnagotravibhāga in support of his position, Bhattacharya writes that
the Buddha, because of his ‘perfect knowledge’ of the natures of things (yathābhūtajñānena),
has
achieved
perfect
intuition
of
the
(sarvadharmanairātmyaparapāramiprāptaḥ).
impersonality/selflessness
of
all
dharmas
The translation continues ‘This impersonality
accords, from every point of view, with the characteristics of the ātman. It is thus always regarded
as ātman, because it is Impersonality which is ātman’ (2015: 4-5). The point being driven home
here is that the ātman-Brahman of the Upaniṣads is something of which nothing can ultimately
be said – a notion we covered in some detail earlier. On this reading, it seems to me that it is not
just Buddhist impersonality or anātman which is thus identical with the Brahman, but also
Maimonides’ conception of God, and, indeed, any mystical conception of a godhead or Absolute
which admits of an attribute-less transcendent reality.
The Sanskrit relating to this point as presented in Bhattacharya’s book: ‘tac cāsya
nairātmyam anātmalakṣaṇena yathādarśanam avisaṃvāditatvāt sarvakalam ātmābhipretaḥ
nairātmyam evātmeti kṛtvā’ (2015: 3). This, using Bhattacharya’s own translation as a basis,
translates as something like ‘this impersonality accords, in every way, with the characteristics of
anātman (anātmalakṣaṇena), it is thus always accepted as ātman, it is impersonality/no-self
which is ātman.’ In note 19 on page 41, Bhattacharya makes the point – albeit in a laboured way
– that he wants to convey that on the one hand, the views of adepts of other doctrines are
‘contradictory to the characteristics of ātman’; on the other hand, despite this, the view of the
Buddha in some ways nevertheless accords with these seemingly contradictory views.
P a g e | 68
Let us dwell on this apparent paradox for a moment. Does the Buddha’s view of ātman
coincide with that of the anyatīrthyāḥ or not? Word substitutions aside, Bhattacharya says of
this that ‘on the one hand, the view of the anyatīrthyāḥ is contradictory to the characteristics of
the ātman (ātmalakṣaṇena visaṃvāditatvāt); on the other hand, the view of the Tathāgatas
accords with them’ (2015: 41, note 19). What exactly is at stake here? We might think of it like
this: the view of the anyatīrthyāḥ (an adept of another doctrine, viz. a non-Buddhist) is to seek
ātman where ātman is not, and so in virtue of their looking in the wrong places and associating
the wrong things with the ātman, the non-Buddhist’s idea of ātman is necessarily always
anātman, or not-ātman. It is, as Bhattacharya puts it, ‘at variance with the characteristics of the
ātman’ (2015: 4) simply in virtue of it not being the ātman. Bhattacharya’s point here, though
subtle, is that first, the anyatīrthyāḥ and the Buddha are in accord insofar as they are both seeking
ātman in the first instance, but are in opposition regarding both the means by which and the
places in which it is to be sought. For his part, the Buddha also arrives at anātman, but does so
in a different, more deliberate way, the end goal of which is – paradoxically – to gain knowledge
of what is the ātman by filtering out those phenomena that are not the ātman. In such a case,
the impersonality of dharmas is the essence of anātman when applied to mundane things (a
reified ‘I’ as conceived via the skandhas and so on), and realisation of this impersonality is
similarly the true state of the Absolute (or, as Bhattacharya would say, of the spiritual ātmanBrahman).57
This is essentially identical to the position in some Mahāyānist schools that sees emptiness as the essence of all
phenomena and thus the true character of the Absolute. It is often argued that some Mādhyamikas also subscribe
to this viewpoint. Murti, for his part advances the view that Madhyamaka has at its heart a conception of an
57
P a g e | 69
To put it in other words, the Buddha’s perfect insight allows him to perceive the
impersonality of worldly phenomena, and it is this impersonality that constitutes the character
of the true ātman (viz. the ātman-Brahman). Now we can say that impersonality accords with
anātman because anātman is the impersonality of dharmas that are mistakenly reified as ātman.
Bhattacharya thinks that when the Buddha gained perfect knowledge of the impersonality of
dharmas (sarvadharmanairātmyaparapāramiprāptaḥ), he actually gained knowledge of the
Upaniṣadic ātman-Brahman (on the Śaṅkaran reading).58 That is to say that, counter-intuitive as
it seems, the realisation of anātman in dharmas is also the realisation of a unified, transcendent
absolute ātman underpinning the universe (though it is beyond the empirical reality of the
universe): the real ātman – as opposed to those things mistaken for ātman – is the impersonality
of the transcendent Absolute. Impersonality accords in every way with the principles of anātman
(because realising anātman through meditative practice leads one to the realisation of the
impersonality of all phenomena), and so ‘it [impersonality] is always accepted as ātman’ (2015:
5) precisely because the nature of the true ātman is the nondual impersonality of Brahman. The
anyatīrthyāḥ would then, according to Bhattacharya, be in agreement with the Buddha that there
is an ātman of some description, the difference comes in the methodology involved in gaining
Absolute (2016: 234), but whilst I acknowledge that this position endures in some Madhyamaka sects today, there
are good reasons to question whether or not it is representative of Nāgārjuna’s position as expounded in the
Mūlamadhyamakakārikā.
58
This turn of phrase might sound clumsy, but it is rather difficult to express in language how a person comes to
know the ātman-Brahman. Merely speaking of ‘realisation’ of, or ‘attainment’ of, the ātman-Brahman necessarily
makes the ātman-Brahman an object or result of some action (Suthren Hirst 2005: 39-40). This cannot actually be
the case, as the ātman-Brahman simply is, regardless of a subject-object relationship. As such, we need to bear in
mind that for Śaṅkara ‘knowledge of brahman is not knowledge of an object but the state in which all objectivising
superimpositions have been removed’ (Suthren Hirst, 2005: 40). Whenever I refer to ‘realising’ the Brahman, then,
it is very much with this difficulty borne in mind, and I am not seeking to objectify that which cannot be objectified
(viz. the ātman-Brahman), but simply communicating a change in state on behalf of the practitioner.
P a g e | 70
knowledge of it. The anyatīrthyāḥ grasps at things that are emphatically not ātman in a
misguided pursuit of what is ātman, whereas the Buddha eschews this type of grasping and
realises that impersonality is the true nature of the ātman, and the means by which this is realised
is by successfully determining that which is anātman.
2.2 Abhidharmic Dharma Theory
Bhattacharya supposes that if dharmas are the building blocks of experience and are, as
argued by Nāgārjuna (among others), impersonal, then it is precisely this impersonality that
constitutes the true intrinsic nature of all phenomena. In such an instance, emptiness would be
the svabhāva of all phenomena. Is this a satisfactory account? In order to answer this, we will
need to give some account of what dharma theory actually entails. Y. Karunadasa (1996: 2) writes
that
The dhamma [dharma] theory was not peculiar to any one school
of Buddhism but penetrated all the early schools, stimulating the
growth of their different versions of the Abhidhamma… There are
sound reasons for believing that the Pāli Abhidhamma Piṭaka
contains one of the earliest forms of dhamma theory, perhaps even
the oldest version.
Karunadasa argues that whilst dharma theory is, strictly speaking, an Abhidharma
innovation, the method of analysis present in early Buddhist scriptures demonstrates a clear link
between the formative Buddhist investigations into the empirical world and the more developed
dharma theories of the various Abhidharmas. How then can this relation be traced? Karunadasa
thinks that the early Buddhist modes of investigation are mutually-dependent – a reasonable
assumption given the Buddha’s famed emphasis on pratītyasamutpāda (dependent
P a g e | 71
origination). 59 Further, Karunadasa holds that this interdependence can be divided into five
modes of analysis that work something like this: first, analysis of nāmarūpa (1996: 3-4). This is a
dvandva compound that translates as name and form.60 It designates the two basic aspects of
the empirical person: mental aspects (nāma) and physical aspects (rūpa) (1996: 3). The second
mode of analysis is that of the five skandhas,61 the third that of the six elements (dhatus),62 the
fourth that of the twelve āyatanas (the six sense faculties and their corresponding objects),63 and
the fifth that of the eighteen dhatus. These are ‘an elaboration of the immediately preceding
mode obtained by the addition of the six kinds of consciousness which arise from the contact
between the sense organs and their objects’ (Karunadasa 1996: 4). This laborious journey
through the five modes of analysis is not in vain. Karunadasa argues that the reasoning behind
each mode of analysis varies, each preceding mode is further analysable by its succeeding mode
(1996: 5). This result, whether by design or by coincidence, is an endorsement of the principle
of pratītyasamutpāda: ‘It is in fact with reference to these five kinds of analysis that Buddhism
frames its fundamental doctrines. The very fact that there are at least five kinds of analysis shows
59
Dependent origination (pratītyasamutpāda) is the thesis that all things are affected by innumerable causes and
conditions and lack an ultimate, singular grounding start point. This especially applies to the twelve nidānas that
account for the causal relationship that produces (from the start point of avidyā (ignorance)) saṃsāra and so
duḥkha. In this sense, it is a fundamental Buddhist doctrine.
A dvandva compound has the same meaning as a series of nouns followed by ca (‘and’), and so in this case,
nāma rūpa ca; name form and; name and form.
60
The five skandhas (Pāli: khandhas) are rūpa (body), vedanā (sensation), saṃjñā (conceptual thought), saṃskāra
(mental formations; dispositions of character), and vijñāna (discernment; sense-based perception).
61
62
These elements are said to be earth, water, temperature, air, space, and consciousness (Karunadasa 1996: 3).
Predictably, the āyatanas are eyes-visible form; ears-sound; nose-smells; tongue-tastes; body-touch; mindmental objects.
63
P a g e | 72
that none of them can be taken as final or absolute’ [my emphasis] (1996: 4). This is an important
point. Whereas Bhattacharya thinks that there is a final mode of analysis (one that results in
knowledge of the impersonal ātman-Brahman), Karunadasa writes that the whole purpose of
these modes of analysis in early Buddhism is to ‘prevent the intrusion of the notions of “mine,”
“I,” and “my self” into what is otherwise an impersonal and egoless congeries of mental and
physical phenomena’ (1996: 4). This changes slightly, he notes, with the advent of the various
Abhidharmas, at which point ‘the Abhidhammic doctrine of dhammas developed from an
attempt to draw out the full implications of these five types of analysis’ (1996: 4-5). At this point,
the analysis of the world into dharmas is indeed seen as final, though there is still a doctrinal
aversion to imputing a unitary transempirical reality.
So far we can see both a difference and a similarity to Bhattacharya’s position. First,
Karunadasa forces the point that the analyses present in Buddhist literature prior to dharma
theory cannot be final or absolute, for reasons that will be explained shortly. Further, he
eventually argues that ‘the Pāli Abhidhamma Piṭaka did not succumb to this error of conceiving
the dhammas [dharmas] as ultimate entities or discrete entities’ (1996: 8), instead portraying
dharmas as simple epistemic tools, the utility of which is in their being used to give accounts of
specific instantiations of experience. This point is strengthened by Noa Ronkin (and at some
length) when she argues that for the early Abhidharmas at least, the svabhāva of dharmas was
not thought of as essential in the way that Bhattacharya is using the word (Ronkin, 2005: 93).64
64
On this, Noa Ronkin (2005: 94) writes that
P a g e | 73
The consequence of this is that for the early Abhidharma at least, neither dharmas nor their
svabhāva can reasonably be equated with ātman as essential nature. Bhattacharya, as we have
seen, characterises this ultimate impersonality as the svarūpa and svabhāva of all things. If this
impersonality is the svabhāva of dharmas, then the dharmas are essentially characterised by
something that is ultimately existent. Their reality is nothing else but the ātman-Brahman, which
pervades and is the support of all existence. Whilst Bhattacharya holds that the ātman-Brahman
is ultimately beyond conceptualisation, then, he must nevertheless always hold that it is Real: in
fact, it is the only Reality. We can ‘know’ that the ātman-Brahman is the fundamental reality –
the Absolute – intellectually through scriptural direction and argument, but it takes something
extra to know it through experience. Śaṅkara calls this latter form of knowledge of the ātmanBrahman ‘anubhava’ (self-experience). It is, of course, the latter form of ‘knowledge’ that really
counts – true self-knowledge (ātmabodha) is attained only through experience. This same
experience has to be directed by scripture in accordance with Śaṅkara’s emphasis on śruti – at
least in the initial stages.
The difference is that the Buddhists, it seems, do not want to make this a commitment to
an unconditioned monistic first cause, and for good reason. Karunadasa sketches the way in
[t]he Paṭisambhidāmagga endorses a broad notion of sabhāva as the nature
that the dhammas essentially share, but it is by no means clear that this nature
necessarily defines what a dhamma is, or that a dhamma exists by virtue of this
nature that it possesses… Nowhere is it stated that a dhamma is defined,
determined or exists by its sabhāva…
By ‘essentially’ here, Ronkin is not referring to an ultimate, unconditioned essence. Instead, she is referring to an
individuating feature, something that serves to determine x entity from y entity. This is clearly much more
mundane than the sense in which Bhattacharya uses ‘essential’ when equating ātman, svabhāva, and svarūpa: in
this sense, ‘essential’ points to the unconditioned ultimate principle that is responsible for all existence. Such an
account, thinks Ronkin, makes far too strong an ontological claim.
P a g e | 74
which svabhāva (Pāli: sabhāva) became associated with ultimacy (or ‘the highest’ (level of
analysis); Pāli: paramattha, Skt.: paramārtha), writing that dharmas eventually came to be
understood as ‘the final limits into which empirical existence can be analysed’ (1996: 19). As the
‘highest’ level of analysis, dharmas became associated with paramārtha, the implication being
that ‘dhammas [Skt.: dharmas] are ultimate existents with no possibility of further reduction’
(1996: 19). Karunadasa adds that it was from this point that ‘own-nature (sabhāva) [Skt.:
svabhāva] came to be further defined as ultimate nature (paramattha-sabhāva)’ (1996: 19). Can
we equate this type of ultimate nature with that endorsed by Bhattacharya? It appears as though
there is some degree of convergence between the position of Bhattacharya and the Abhidharma
position put forth by Karunadasa. Bhattacharya wants to claim that the ultimate nature (what
Karunadasa refers to in Pāli as the paramattha-sabhāva) of all dharmas is the impersonal
unconditioned Absolute. The ultimate nature of dharmas is the ātman-Brahman. But how does
this sit with Karunadasa’s account? At first glance, there might be a similarity: Karunadasa
explains that
the mental as well as the material dhammas are not actually
separable one from another. In the case of the mental dhammas,
the term used is saṁsaṭṭha (conjoined); in the case of the material
dhammas, the term used is avinibbhoga (inseparable). This raises
the question why the dhammas are presented as a plurality.
(1996: 24)
There is more to this than meets the eye, however. Karunadasa is quick to add that
despite not being strictly separable, dharmas are nevertheless distinguishable. The claim is not
that dharmas are all essentially the same thing (viz. the impersonal ātman-Brahman), or even
that they all originate from the same thing (again the ātman-Brahman). It is simply that we can
P a g e | 75
tell what they are in relation to each other if we correctly analyse them. Indeed, Karunadasa
writes that ‘[i]t is this distinguishability that serves as the foundation of the dhamma theory’
(1996: 24).65 Also significant is that for the early Abhidharma systems, the dharmas are not
strictly unconditioned, regardless of their ‘ultimate’ status. They are viewed as ultimate very
simply because the Ābhidharmikas thought that all analysis of experience stopped with them.
They are foundational only insofar as analysis of experience bottoms out with them – this does
not necessarily mean that they have to be assigned some sort of ontological primacy of the type
that Bhattacharya is keen to impose. In fact, the dharmas are indeed said to have conditioned
origination (Pāli: sappaccatatā), and Karunadasa helpfully describes five axiomatic reasons why
this is accepted to be the case. All five of these reasons are relevant in one way or another to
the current discussion, and so I shall quote Karunadasa’s explanations verbatim:
(i) It is not empirically possible to identify an absolute original cause
of the “dhammic” process. Such a metaphysical conception is not
in accord with Buddhism’s empirical doctrine of causality, the
purpose of which is not to explain how the world began but to
describe the uninterrupted continuity of the saṁsāric process
whose absolute beginning is not conceivable. In this connection it
must also be remembered that as a system of philosophy the
Abhidhamma is descriptive and not speculative.
(ii) Nothing arises without the appropriate conditions necessary for
its origination. This rules out the theory of fortuitous origination
(adhiccasamuppannavāda).
(iii) Nothing arises from a single cause. This rules out theories of a
single cause (ekakāraṇavāda).
Their rejection is of great
significance, showing that the Abhidhammic view of existence
rejects all monistic theories which seek to explain the origin of the
65
It is important for us to remember that the various dharma/dhamma theories are not prescriptive or
speculative, but descriptive. They do not aim to prescribe an account of the world, but to give a full analysis of the
world as experienced.
P a g e | 76
world from a single cause, whether this single cause is conceived as
a personal God or an impersonal Godhead. It also serves as a
critique of those metaphysical theories which attempt to reduce
the world of experience to an underlying transempirical principle.
(iv) Nothing arises singly, as a solitary phenomenon. Thus on the
basis of a single cause or on the basis of a plurality of causes, a single
effect does not arise. The invariable situation is that there is always
a plurality of effects. It is on the rejection of the four views referred
to above that the Abhidhammic doctrine of conditionality is
founded.
(v) From a plurality of conditions, a plurality of effects takes place.
Applied to the dhamma theory, this means that a multiplicity of
dhammas brings about a multiplicity of other dhammas.
(Karunadasa, 1996: 25-26)
It is easy to see why Bhattacharya chose not to rely on the intricacies of Abhidharmic
Buddhism to bolster his theory. On the above evidence, the Abhidharma system of analysis flatly
denies that any monistic understanding is possible. Despite dharmas being ultimately real (‘the
ultimate, irreducible data of empirical existence’ (Karunadasa, 1996: 20)) they do not, according
to the early Ābhidharmikas, share the same intrinsic nature in the way that Bhattacharya would
need them to. Bhattacharya claims that the true nature of all dharmas is their insubstantial
impersonality – he equates this with what Nāgārjuna calls śūnyatā (emptiness). But we need to
be careful about how we characterise this impersonality. For the Ābhidharmikas, each dharma
simply is its intrinsic nature. This means in turn that each dharma – in virtue of being
distinguishable – has a distinct nature that serves to distinguish it as x dharma as opposed to y
dharma rather than simply having the ātman-Brahman as their nature. For somebody like
Nāgārjuna, the difference is even starker. He does not agree with the Ābhidharmikas that each
dharma has an (or is its) intrinsic nature precisely because he believes that nothing possesses (or
P a g e | 77
‘is’) an intrinsic nature to begin with. Further, he does not think that ultimately real dharmas can
be conditioned or arise from causes and conditions in the way that Karunadasa outlines above.
Indeed, the very first chapter of the MMK is an attack on the notion of dharmas
possessing (or being) intrinsic natures (Skt.: svabhāva, Pāli: sabhāva) whilst also being subject to
causes and conditions. The basic idea is that dharmas (and so svabhāva) cannot be ultimate
entities and also be reliant on causes and conditions for their existence because this would
necessitate some sort of change in that dharma. If dharmas are equivalent with their intrinsic
natures, then change is impossible.66 Karunadasa writes, ‘to claim that [a dharma’s] intrinsic
nature undergoes modification is to deny its very existence’ (1996: 21): an obvious issue.
66
See all of MMK 1 for the totality of this argument. This particular point is perhaps made most forcefully between
MMK 1.6-10:
naivāsato naiva sataḥ pratyayo ‘rthasya yujyate /
asataḥ pratyayaḥ kasya sataś ca pratyayena kim //6//
na san nāsan na sadasan dharmo nirvatate yadā /
kathaṃ nirvatako betur evaṃ sati hi yujyate //7//
anārambaṇa evāyaṃ san dharma upadiśyate /
anthānārambaṇe dharme kuta ārambaṇaṃ punaḥ //8//
anutpanneṣu dharmeṣu nirodho nopapadyate /
nānantaram ato yuktaṃ niruddhe pratyayaś ca kaḥ //9//
bhāvānāṃ niḥsvabhāvānāṃ na sattā vidyante yataḥ /
satīdam asmin bhavatīty etan naivopapadyate //10//
A condition of an effect that is existent or non-existent is not accepted /
Of what [use] are conditions for nonexistents? And for whom [is there use in]
conditions for existents? //6//
When no dharma operates that is existent, nonexistent or both existent and nonexistent /
How in this case can [something be called] an operative cause (nirvatakahetu)? //7//
P a g e | 78
Karunadasa tells us that dharmas appear in clusters: ‘a psychic instance can never occur
with less than eight constituents, i.e. consciousness and its seven invariable concomitants’ (1996:
26). He adds that the relation between the different dharmas in a cluster is ‘one of necessary
conascence’ (Pāli: niyata-sahajāta; being born or originated together/at the same time) as part
of ‘a complex correlational system’: the implication of this is that there are no singular, solitary
phenomena (1996: 26).
What Karunadasa has outlined is an awkward though not strictly paradoxical position –
the claim appears to be that dharmas are at once both inseparable (and so unitary) and also
distinguishable (and so a plurality). This simply does not work for Bhattacharya if dharmas are
ultimate existents sharing the same ultimate nature (viz. the ātman-Brahman), and it is likely that
this tension influenced his dismissal of Abhidharma doctrine; he instead claims that it was the
Mahāyāna that eventually ‘put things right’ (2015: 39) with the doctrine of emptiness. This is a
position that he interprets as stating that the world is composed of entities that share the same
impersonal nature: that of a transcendent ātman-Brahman.
Bhattacharya refers to the Buddha’s seeing all dharmas and recognising their
impersonality as true knowledge of ātman-Brahman. Bhattacharya agrees in principle with the
It has been taught that dharmas [existents] have no objective support /
But [where there is] no objective support, again, why [posit] an objective support? //8//
When dharmas are unproduced [by conditions], cessation does not occur /
When [a dharma has] ceased, what [is a] condition? Thus, a direct condition is not suitable. //9//
Since things without intrinsic nature are not [ultimately] existent /
‘This existing, that comes to be’ does not obtain. //10//
P a g e | 79
Buddha
that
all
dharmas
are
insubstantial/impersonal:
he
cites
the
Buddha’s
sarvadharmanairātmyaparapāramiprāptaḥ – his perfect intuition of the insubstantiality of all
dharmas – as evidence of the Buddha’s enlightenment. For Bhattacharya, this enlightenment is,
of course, in full agreement with Advaitin doctrine, at least where the ultimate impersonality of
all (conventionally) existent things is concerned (Bhattacharya, 2015: 4-5).
Second, there is some similarity between Karunadasa’s account and Bhattacharya’s
argument: ultimately, the ātman-Brahman is impersonal and is devoid of dualistic notions such
as ‘self’ and ‘other’, of ‘me’ or of ‘you’. I suspect, then, that Bhattacharya would take
Karunadasa’s argument about the Buddhist modes of analysis and drive the point that whilst they
go about things in a slightly different way (viz. in aiming to buttress introspection against ideas
of self rather than examining and gaining knowledge of the self), the final result is identical. The
culmination of both Buddhist and Advaitin efforts is, for Bhattacharya, arrival at knowledge of
and experience of an egoless, pure (lacking defilements), unconditioned Absolute.
Bhattacharya endorses the position that this incorporeal ātman is at once impersonal and
Absolute, and holds that ‘to know the ātman-Brahman is, in effect, to become it…as long as we
do nothing but conceive of it, we are far from knowing it’ (2015: 7-8). It is here that we really see
how realising impersonality/no-self can aid our achievement of realising the true self, the ātmanBrahman: the point is to become the ātman-Brahman and thus remove the subject-object
dualism that blights our day-to-day life. He writes that failing to become the ātman-Brahman,
and persisting to merely conceive of it, or assign attributes to the concept of it means that ‘…the
ātman, the Self, remains an object to us, and, therefore, a non-Self’ (2015: 8). The use of ‘non-
P a g e | 80
self’ here is slightly subtler than previous usages, and it is determining the instance in which the
actual ātman can be misapprehended rather than determining instances in which entirely the
wrong things are taken as ātman. It is as thought the practitioner is on the right tracks, but has,
in holding onto a preconceived notion of the ātman-Brahman, scuppered their chances of
actually realising it. The very act of loading the ātman-Brahman with conceptual constructions
detracts from its actual status as Absolute, corrupts the mind pursuing it and makes it anātman,
a non-self in the most basic form of the expression. The ātman-Brahman is supposed to be
necessarily inexpressible in positive terms because it is beyond linguistic designation. This is, at
least, the view of Śaṅkara’s Advaita. It is in one sense useful to conceive of the ātman-Brahman
in the early stages of spiritual development – it is probably even necessary if the diligent
practitioner wishes to arrive at it in the proper way (why would practitioners bother to devote
themselves to realising something that has no positive impact?) – but it is simply not enough on
its own, and continuing to ‘see’ it in this manner is to miss out on it altogether.
Bhattacharya explains that ‘[a]ll truths as can be formulated are, in fact, but
approximations of Truth, which is inexpressible; none of them can be identified with Truth itself’
(2015: 9). ‘Truth’ in this context is synonymous with ‘Absolute’ or ‘ultimate’, or, unsurprisingly,
ātman: only the ātman-Brahman is ‘Truth’ per se, as it is the totality of Being. Bhattacharya, in a
passage that would not be out of place in any Madhyamaka textbook, says of approximate truths
P a g e | 81
that ‘[t]hey aid us in reaching [ultimate Truth], they guide our progress towards it; but they must
be transcended if it is to be reached’ (2015: 9).67
In any case, successfully reaching this state of Being and becoming the ātman-Brahman
is, for Bhattacharya, synonymous with the Buddhist goal of reaching nirvāṇa (2015: 11).
Liberation is to see the emptiness of phenomena; it is their impersonality.
This is why
Bhattacharya must reject Abhidharmic dharma theory and try to embrace (a version of)
Nāgārjuna’s vision of śūnyatā. If many svabhāvas exist, then Bhattacharya’s thesis fails, for all
existence has but one intrinsic nature: the impersonal ātman-Brahman. This is fundamental. It
is the ‘most profound spiritual reality’ in which one can dwell, as the ‘ātman is the ultimate
Reality upon which the empirical world is founded’ (2015: 12), or in other words, it is the ultimate
that provides the basis for the conventional. For Śaṅkara, ‘the entire expanse of [empirical]
things is mere illusion’ and not ultimately real: only Brahman is ultimately real (Śaṅkara, 2009:
138).
2.3 An Early Buddhist Rejection of Brahman?
We know that the concept of anātman (Pāli: anattā) is now synonymous with Buddhism
and that it in principle rejects the idea of any persistent ātman. We also know that Bhattacharya
thinks that the real purpose of the Buddhist doctrine of anātman is to deny a specific type of
ātman and not to deny the ātman in toto: Bhattacharya, as we have seen, makes the argument
The use of ‘truth’ (small ‘t’) and ‘Truth (capital ‘T’) here is significant and is reminiscent of (though not identical
with) Madhyamaka discussion of conventional and ultimate (truth; reality). Bhattacharya is splitting the world into
two levels here: that of ‘truth’, a conventional designation that is but an inferior approximation of the ultimate
‘Truth’, which is inexpressible yet able to be experienced.
67
P a g e | 82
that the type of ātman rejected is only the empirical jīva and not the reality behind it, the ātmanBrahman. Thus, anātman only applies to things that a person mistakenly takes to be ātman: the
things that we wrongly invest in as parts of ourselves are anātman, whereas the true ātman is
the ātman-Brahman. Such an understanding trades on the early Buddhist literature specifying
that which cannot be ātman (namely anything that comes under the remit of the skandhas)
rather than denying ātman outright. Alexander Wynne writes of this general approach that ‘the
five aggregates are impermanent, subject to change and so unsuitable to be regarded as one’s
ātman’ (2009: 61).68 Wynne agrees with Bhattacharya when he notes that much of the earliest
Buddhist discourse regarding ātman does, in fact, tell us where not to look rather than proscribing
searching per se. Citing from numerous Buddhist sources including the Catuṣpariṣat Sūtra of the
Mūlasarvāstivāda (2009: 60), the Māhavastu of the Mahāsāṃghika (2009: 60-61), and the
Mahāsatipaṭṭhāna Sutta (2009: 63) to illustrate the focus of the early Buddhist texts, Wynne
shows that it does indeed look to be the case that the very earliest Buddhist scripture concerns
itself only with refuting ātman in specific circumstances, viz. a denial of ātman ‘focused on the
lack of ‘self’ in the five aggregates’ (2009: 63).
The idea lurking in the background is then, I suppose, that something that is suitable to be regarded as one’s
ātman is eternal and unchanging – Bhattacharya capitalises on this possibility to argue that the ātman-Brahman of
Advaita thus fulfils the requirement to be considered as one’s true ātman.
68
P a g e | 83
There is at least one sutta that can be read as a denial of the notion of an ātman-Brahman,
however. Take the following extracts from Majjhima Nikāya 49 (Brahmanimantanika Sutta):69
The Blessed One said: "On one occasion recently I was staying in
Ukkattha in the Subhaga forest at the root of a royal sala tree. Now
on that occasion an evil [pernicious] viewpoint70 had arisen to BakaBrahma: 'This is constant. This is permanent. This is eternal. This is
total. This is not subject to falling away — for this does not take
birth, does not age, does not die, does not fall away, does not
reappear. And there is no other, higher escape.'
The Buddha continues:
"When this was said, Baka Brahma told me, 'But, good sir, what is
actually constant I call "constant." What is actually permanent I call
"permanent." What is actually eternal I call "eternal." What is
actually total I call "total." What is actually not subject to falling
away I call "not subject to falling away." Where one does not take
birth, age, die, fall away, or reappear, I say, "For this does not take
birth, does not age, does not die, does not fall away, does not
reappear." And there being no other, higher escape, I say, "There is
no other, higher escape."
Some unpacking is required here. First, we see the Buddha outline the ‘pernicious
viewpoint’ held by Baka-Brahmā, a powerful deity that believes that he has seen the universe
how it really is. It is important to note that Baka-Brahmā does not seem to be accounting for an
empirical world in purely physical terms. He is, in fact, referring to the ‘world’ (read: cosmos) as
instantiated in and by himself as heavenly king, on both a material and psychological level. Baka-
69
For the sake of convenience, I here use Thanissaro Bhikkhu’s (2007) translation of this sutta from the Pāli.
I prefer ‘pernicious viewpoint’ to ‘evil viewpoint’ because it adequately communicates that the viewpoint has
damaging consequences for those that hold it without passing any undue moral judgement. I think calling such a
viewpoint ‘evil’ is rather to overstate the case. One can hold a mistaken, damaging viewpoint without it being
‘evil’, after all!
70
P a g e | 84
Brahmā is effectively claiming that he and/or the world that he inhabits is the eternal true reality.
This is explained by Ajahn Brahm (2006) a couple of minutes into his audio commentary to the
sutta, when he details that Baka-Brahmā thinks of himself as a creator of an eternal, unchanging
(and so ultimate) reality.71
There is a very significant line contained in the above extract; namely where Baka-Brahmā
talks about ‘[w]here one does not take birth, age, die, fall away, or reappear’. A realm or entity
where one is not born, does not age, does not die, fall away or suffer rebirth: this sounds
suspiciously like the Brahman as discussed in the Upaniṣads and as later expounded in the ātmanBrahman doctrine of the Advaitins. I think that the presence of these few words is evidence
enough of the targets that the Buddha had in mind: Brahmins. Consequently, I think that a case
can be made that this sutta provides early evidence of a Buddhist rejection of the ātmanBrahman, a case that is significant in terms of Bhattacharya’s argument because it is found in the
Pāli literature traditionally ascribed to the Buddha. As we have seen, Bhattacharya claims that
the Buddha did not ever explicitly comment on the ātman-Brahman. Perhaps this is an allegorical
comment on precisely that.
Nevertheless, Baka-Brahmā’s most crucial claim is that liberation cannot be found outside
of the definition of the world that he has offered, that there is nothing higher than his account
of the totality of existence. This is a claim that would be advanced at a later time by Śaṅkara, and
is emphatically advanced by Bhattacharya when he writes that ‘[o]ne who has realized the ātman
Thanissaro Bhikkhu (2007: note 1) also writes in a footnote that ‘Baka Brahma here appears to be referring both
to his Brahma world and to the state of mind that enables one to inhabit his Brahma world.’
71
P a g e | 85
is not outside the world, but he looks upon the world with new eyes’ (2015: 13). Bhattacharya is
making Baka-Brahmā’s point for him – liberation consists in looking upon the permanent, the
eternal with ‘new eyes’. Liberation is to be realised in this world, but also in and as a specific
vision of this world: in Advaitin terms, in a world supported by the ātman-Brahman. I contend
that Baka-Brahmā’s words here place him in such a world, and more, they show that he is actively
advancing this thesis. Baka-Brahmā’s arrogance and certainty lead him to dismiss the Buddha as
just another mediocre ascetic, doomed to spend life under the influence and command of BakaBrahmā. 72 However, the Buddha quickly establishes himself as fully awakened and thus not
simply equal to, but actually superior to Baka-Brahmā as regards knowledge of liberation. The
Buddha informs Baka-Brahmā that he is ignorant of some spheres of existence outside of his own
realm – a consequence of having spent too long dwelling in one place, an instance of avidyā made
manifest by Baka-Brahmā’s reluctance to think beyond himself. Such realms are not physical
realms, they are deva realms and are thus made by mental activity (manomaya), which need not
necessarily imply anything ontologically, viz. that they enjoy some substantial existence. 73 Such
72
From the Brahmanimantanika Sutta translated by Thanissaro Bhikkhu (2007):
'There were, monk, before your time, brahmans & contemplatives in the world
whose ascetic practice lasted as long as your entire life span. They knew, when
there was another, higher escape, that there was another, higher escape; or,
when there was no other, higher escape, that there was no other, higher
escape. So I tell you, monk, both that you will not find another, higher escape,
and that, to that extent, you will reap your share of trouble & weariness. Monk,
if you relish earth, you will lie close to me, lie within my domain, for me to
banish and to do with as I like. If you relish liquid ... fire ... wind ... beings ...
devas ... Pajapati ... brahma, you will lie close to me, lie within my domain, for
me to banish and to do with as I like.’
Jayarava Attwood (2014) writes convincingly both on why the favoured interpretation ought to be ‘made by
mental activity’ as opposed to something like ‘made of mind’ (implying that ‘mind’ is some sort of substance from
73
P a g e | 86
realms are, then, meditative realms, made by mental activity (they are attained via what are
called jhānas in Pāli (Skt.: dhyānas) – meditative states). The implication is that the Buddha has
a more powerful mind, more thoroughgoing insight than Baka-Brahmā. Despite Baka-Brahmā’s
protestations, it becomes evident that the Buddha knows more than Baka-Brahmā, and so is his
superior:
There is, brahma, the body named Subhakinha (Beautiful
Black/Refulgent Glory) ... the body named Vehapphala (Skyfruit/Great Fruit), {the body named Abhibhu (Conqueror)} which
you don't know, don't see, but that I know, I see. Thus I am not your
which things can be built), and also on why we ought not to take such discussions to be an assertion of substantial
existence. This position is not shared by scholars such as Donald Swearer (1973: 448), who instead opines that
the ethical and the ontic are definitely related in term mind, that is, the mind
appears as the center point. It has, as it were, the power to create the "self."
The ethical dimension stems from this fact. If the mind is ignorant and impure,
one will suffer; if, on the other hand, the mind is enlightened and pure, one will
attain happiness.
It is not immediately clear, however, that a construction of ‘self’ does indeed carry any serious ontological weight.
We can say that the mind can (and does) ‘create’ narratives – even if these narratives are ultimately false – and yet
it is a jump to assert any ontological significance to this creation of mind. Such creations do not appear to
necessitate any substantial existence at all. Instead, we simply have a concept entertained in mind under which no
real entity falls. In the same way that a deranged person might think that they are able to float through the air, a
deluded person thinks they have (or are) an ātman. Both are mistaken designations, false attributions. Indeed,
the entire Buddhist project rests on the rejection of ātman! It strikes me that Swearer’s view runs closely to that
of the ancient pudgalavādins – Buddhists that accepted the substantial reality of a ‘person’ or form of ‘self’. His
reasons for his train of thought are, he says, from the Dhammapada. He cites it thus (1973: 448):
Mind is the forerunner of (all evil) states. Mind is chief; mind-made are
If one speaks or acts with wicked mind, because of that, suffering follows
even as the wheel follows the hoof of the draught-ox.
Mind is the forerunner of (all good) states. Mind is chief; mind-made are
they. If one speaks or acts with pure mind, because of that, happiness follows
one, even as one's shadow that never leaves.
It is easy enough to see from where Swearer draws his conclusions. Good states are ‘mind made’, and acting with
‘pure mind’ brings about these good states. Is the claim here really that our mind changes the makeup of the
world, or is it instead that our mindset changes how we engage with the world? In light of my argumentation
throughout this work, I am more inclined to think that the latter applies.
P a g e | 87
mere equal in terms of direct knowing, so how could I be your
inferior? I am actually superior to you.
Having directly known earth as earth, and having directly known the
extent of what has not been experienced through the earthness of
earth, I wasn't earth, I wasn't in earth, I wasn't coming from earth,
I wasn't "Earth is mine." I didn't affirm earth. Thus I am not your
mere equal in terms of direct knowing, so how could I be inferior? I
am actually superior to you.
Again, there are a few things to unpack here. First, the Buddha lists the ‘bodies’ that are
experienced via the jhānas: a jhāna being, of course, a meditative state and thus implying no
ontological substance. The premise here is, as I have said, simply that the Buddha knows more
than Baka-Brahmā, a supposed great, powerful deity. The Buddha is illustrating that despite
Baka-Brahmā’s claims to the contrary, there is some release or liberation different from and
outside of – higher than – this Brahmā realm (the ātman-Brahman?). Such release is to be found
via meditative insight, and it is outside of this realm not insofar as it is an existent place to which
we can go, but rather in that it is a meditative state beyond Baka-Brahmā’s comprehension or
ability. In virtue of being existent as meditative states beyond Baka-Brahmā’s comprehension,
such realms are ‘higher’ than the account of existence offered by Baka-Brahmā. This in turn
means that there is at least the possibility of some ‘higher escape’ provided that sufficient insight
is developed.
Swearer (1973: 447) writes that ‘[t]hrough attaining the four jhānas the
consciousness or mind (citta) is made pure (parisuddha), freed from blemish, devoid of evil
(kilesa), stable and immovable. The citta is thereby freed to direct itself toward the "insight that
comes from knowledge."’ In other words, the mind is directed not onto worldly concerns or
reifications, but onto and into itself to develop an insight not to be found in worldly reifications.
If I am right in linking the cosmic world occupied by Baka-Brahmā to the Brahman, then it strikes
P a g e | 88
me that the Buddha’s insight cannot be related to some sort of cosmic self (because he has
denied all that Baka-Brahmā has asserted of it). Nor can it be related to the individual ātman,
which even Bhattacharya acknowledges was denied by the Buddha. Maybe it is the case, then,
that the higher insight alluded to by the Buddha is actually anātman, tying in with the Buddha’s
denial that what I take to be a metaphor for the Brahman is permanent:
"When this was said, I told Baka Brahma, 'How immersed in
ignorance is Baka Brahma! How immersed in ignorance is Baka
Brahma! — in that what is actually inconstant he calls "constant."
What is actually impermanent he calls "permanent." What is
actually non-eternal he calls "eternal." What is actually partial he
calls "total." What is actually subject to falling away he calls "not
subject to falling away." Where one takes birth, ages, dies, falls
away, and reappears, he says, "For here one does not take birth,
does not age, does not die, does not fall away, does not reappear."
And there being another, higher escape, he says, "There is no other,
higher escape."'
Even the great creator deity Baka-Brahmā is so tied up in his own sense of self (and selfimportance!) that he has effectively led himself to delusion and ignorance regarding the world.
There are mental realms potentially open to him that he is ignorant of and that he lacks the
discipline or insight to access. There is thus a ‘higher escape’, but he is too self-absorbed to
recognise it. There is no permanence to be found in that which he asserts to be the basis of
reality, nor is there any refuge to be found in the reality in which Baka-Brahmā claims does not
age, die, ‘fall away’ and so on. He dwells in this reality as the jīva might be said to dwell within
the universal ātman-Brahman, that is to say labouring under delusion. Far from being liberated
by his knowledge of and living within what he assumes to be the totality of existence, BakaBrahmā has in fact contributed to his own trapping in duḥkha by attributing to his reality all of
the things that it is not!
P a g e | 89
If this sutta is not intended to have parallels with the Vedic conception of Brahman, I
would be very surprised indeed. Should my interpretation hold some water, then this is an early
example of the Buddhist rejection of a cosmic Brahman: the Buddha rebukes Baka-Brahmā for
his assertion that his realm (which I understand to be analogous to the ‘realm’ of Brahman) is
permanent, the highest, and the full totality of existence. The most telling aspect here is that the
Buddha does not make explicit reference to an ātman-Brahman type realm that is superior to
that realm of Baka-Brahmā. He instead talks of imposing permanence and so on to what is not
permanent, and by extension must refute the idea that the mindset required to ‘enter’ BakaBrahmā’s realm is the highest mindset that one might achieve. This is, I contend, not because
the Buddha believes that there exists some ātman-Brahman that accounts for the totality of
existence, but because the Buddha believes the converse. I point briefly to an extract from the
Mūlapariyāya Sutta as translated by Thanissaro Bhikkhu (1998):
He directly knows Unbinding as Unbinding. Directly knowing
Unbinding as Unbinding, he does not conceive things about
Unbinding, does not conceive things in Unbinding, does not
conceive things coming out of Unbinding, does not conceive
Unbinding as 'mine,' does not delight in Unbinding. Why is that?
Because the Tathagata has comprehended it to the end, I tell you.
‘Unbinding’ can here be understood as ‘nirvāṇa’ and so ‘liberation’: the Buddha denies
that anything at all comes from the liberated state. This is interesting because for Bhattacharya
as an Advaitin, ‘liberation’ comes in knowing the ātman-Brahman, and that same ātmanBrahman is, as we have seen, both the material and efficient cause of the universe. In other
words, things appear to ‘come out of’ the ātman-Brahman (despite this appearance being
ultimately false), and the ātman-Brahman is immanent in all things as their essential reality. The
P a g e | 90
Buddha here appears to reject the idea that being is tied up in or owes its existence to some
other cosmic being. Although the sutta in question is strictly an argument against Sāṃkhya
philosophy, the general thrust can be deployed against Advaita. Thanissaro Bhikkhu (1998)
writes that
there has long been — and still is — a common tendency to create
a "Buddhist" metaphysics in which the experience of emptiness, the
Unconditioned, the Dharma-body, Buddha-nature, rigpa, etc., is
said to function as the ground of being from which the "All" — the
entirety of our sensory & mental experience — is said to spring and
to which we return when we meditate.
This does indeed seem to sum up Bhattacharya’s project! In equating śūnyatā with the
‘ground of being from which the All. . . is said to spring’,74 Bhattacharya, Murti et al. are building
a Buddhist metaphysics, one that leads to the ātman-Brahman. Thanissaro Bhikkhu (1998) signs
off his commentary by simply noting that ‘[a]ny teaching that follows these lines would be subject
to the same criticism that the Buddha directed against the monks who first heard this discourse.’
This is simply to say that any attempts at reification of experience into some sort of cosmic self
are at odds with the positions defended by the Buddha in the above extracts. I think that we can
reasonably read into the above an early rejection of Brahman and so a preliminary rejection of
the thesis that the Buddha actually endorsed the ātman-Brahman. If my readings are a fair
reflection of the content, then we have two early examples of the Buddhist aversion to
absolutism in any sense, not just in the narrow sense of the jīva: to equate the Buddhist liberation
with that of the Advaitins is to further impose ideas onto reality. Even if we want to say that
74
The ‘All’ simply being the totality of experience and experienced phenomena.
P a g e | 91
these ideas are true only conventionally (and thus not true in an ultimate sense), there remains
one idea of an ultimate ground of being. Such an idea should, on the accounts given above, be
absent from any Buddhist understanding of the world.
P a g e | 92
§3: The Ātman-Brahman
In order to assess whether or not the Upaniṣadic ātman-Brahman is compatible with
Buddhist thought, we need first to situate it, in general terms, in its ‘natural’ place. For his part,
Bhattacharya provides a plethora of citations relating to what the Absolute ātman-Brahman is
from his Śaṅkaran perspective, though continually acknowledging that no positive statement can
ultimately be predicated of it. As it forms the entirety of Bhattacharya’s argument, I will first
focus on the ātman within Advaitic interpretations of the Upaniṣads.75 From there, I will examine
how knowledge of the ātman-Brahman – and knowledge of the ultimate more generally – might
work. From Bhattacharya’s perspective, then, the ātman-Brahman is
…not the chariot (ratha), but the “master of the chariot” (rathin),76
its “inciter” (pracodayitṛ). Even though it moves in all bodies
(pratiśarīreṣu carati), it “rests in its own greatness” (sve mahimni
tiṣṭhati), above phenomena (uparistha), not subject to their contact
“like a drop of water on a lotus petal” (bindur iva puṣkare). It is
eternally pure (śuddha), peaceful (śānta), without individuality
(nirātman)…, “empty” (śūnya). It is what makes us act (kārayitṛ),
but does not act itself (akartṛ). Being “in-itself” (svastha), it is “as a
spectator” (prekṣakavat) of our acts, good and bad, which do not
affect it at all (sitāsitaiḥ karmaphalair anabhibhūtaḥ).
(Bhattacharya, 2015: 29)
It is obvious to any reader that Bhattacharya favours the Śaṅkaran rendering of Vedānta over and above that of
other Vedāntin schools, and consequently, he tends to rely upon the Śaṅkaran rendering of the Upaniṣads, too.
75
The point here is that whilst the body is the chariot, the ātman sits beyond even the intellect, which is the
charioteer piloting the chariot, and the mind, which is the reins. The ātman is thus the ‘master’ or ‘lord’ of the
chariot but does not – in theory – have a direct hand in the control of the chariot.
76
P a g e | 93
Considering that the ātman-Brahman is beyond both conceptual thought and linguistic
designation, both Bhattacharya and the Upaniṣads that he quotes certainly have a lot to say
about it! Nevertheless, his descriptions do appear to accord with how the ātman-Brahman is
represented in the Upaniṣads, particularly in part III of the Aitareya Upaniṣad. There is some
incongruence, however, between the ‘spectator’ that is ‘without individuality’ and the notion
that, as Bhattacharya claims, the ātman-Brahman is the lord of the chariot but simultaneously is
a step removed from action. In fact, Bhattacharya claims that the ātman is ‘what makes us act
(kārayitṛ), but does not act itself (akartṛ)’ (2015: 29). In the case of the chariot as described in
the Kaṭha Upaniṣad, then, the ātman is present, and makes the charioteer – the intellect – act,
but it does this without acting itself. Precisely how a lord might impel his charioteer to act
without first acting himself is, it seems to me, something of a mystery. Perhaps the comparison
is simply a poor one. Even if the ātman simply bestows a propensity upon the intellect to act in
a certain way, I think that we would have to concede that this is a causal action, albeit in a
relatively weak sense.
Such examples also presuppose a duality between the acting agent (the charioteer) and
the lord that causes the agent to act (the ātman-Brahman). According to Advaitin doctrine, in
ultimate terms this duality simply cannot stand. The example is, then, necessarily flawed and
paradoxical, though the Advaitin would argue that this is because we do not have access to the
Olympian point of view that we occupy subsequent to the enjoyment of the knowledge of the
ātman-Brahman. Anything we say about the ātman-Brahman is problematic because its reality
transcends all that can be said of it. We cannot capture it in words; we can only offer
unsatisfactory approximations. This notion is, of course, not new to anybody that is familiar with
P a g e | 94
apophaticism and mysticism more generally. We saw in §1 that this type of thought dominates
religious mysticism of every kind.
It can be claimed then that, assuming the Śaṅkaran
interpretation of the Upaniṣads is correct, there is no real problem with our being unable to
accurately account for the ātman-Brahman through language: ultimately, language is irrelevant
to the experience, and it is after all the experience that is important.
This is relatively
controversial.77
Another compounding factor arises later in the same Upaniṣad when it is claimed that
this same ātman is ‘beyond time, space, and causality, eternal, Immutable’ (Kaṭha Upaniṣad
I:3:15). We might question how anything can be attributed to the ātman if it is beyond causality
and beyond linguistic attributions, much less how we can thank it for the creation of the cosmos!
This of course is the point for the Advaitin: we cannot really say anything about the Brahman. It
is ultimately beyond all linguistic designations, all worldly attributes, and all characterisation. It
seems unsatisfying to claim that conventionally, we can talk about the ātman-Brahman in such
terms, but then say that ultimately they do not apply. Nevertheless, this sort of tension is
common throughout the Upaniṣads. It is a point that we will revisit repeatedly as we progress
through this work.
77
David Burton (2001: 64) rails against this sort of understanding of mystical experiences, writing that it seems to
him to be incoherent to claim ‘knowledge’ of something that is inexpressible or non-conceptual. He argues that if
there truly were non-conceptual experiences (during which one cannot discriminate or discern), then they should
be contentless. The upshot of their being contentless would be that the mystic should not be able to claim that
they recall the experience whatsoever, let alone have ‘knowledge’ of it. That mystics (including Śaṅkara) do claim
to have a knowledge of and recalled experience of such mystical episodes suggests, claims Burton, that the
experience itself involved some degree of discrimination and thus some sort of conceptualisation.
P a g e | 95
The Kaṭha Upaniṣad continues:
yad idaṃ kin͂ca jagat sarvaṃ prāṇa ejati niḥsṛtam /
mahad bhayaṃ vajram udyataṃ ya etad vidur amṛtās te bhavanti //2//
All the universe emanates from Brahman [prāṇa; breath of life]
and moves [in Brahman] /
[That Brahman] causes great fear, like a poised thunderbolt. Those
that know this [Brahman] become immortal //2//
(Kaṭha Upaniṣad, II:3:2)
Again, we see here that the ātman-Brahman is assigned some causal power. The universe
could not emanate from the Brahman’s ‘breath of life’ if there were no means by which the
ātman-Brahman could exercise this power of creation. Of course, we must take into account the
use of figurative language to explain these apparent incongruities. Personification of the
Absolute and the assignation of attributes and abilities to it is something that we encounter in
religious texts of all descriptions. Bhattacharya would likely argue that this does not detract from
the overarching message that ultimately, the ātman-Brahman is impersonal, ultimately, the
ātman-Brahman is inactive, and ultimately, the ātman-Brahman is beyond categorisation: the
language referring to the ātman-Brahman is then simply a conventional designation; a
storytelling tool designed to make focused introspection seem attractive and worthwhile.
All that the assignation of positive attributes and personification really serve to do, then,
is remind practitioners of the splendour of their religious path and provide a sort of figurehead
to which they may relate. Such devices are specifically designed to illustrate why meditating to
discover the ātman-Brahman is worth the effort. Indeed, Albahari is at pains to explain that
attaching such (other) worldly qualities to something that is ultimately beyond such designations
P a g e | 96
is, in fact, an exercise designed to ‘function pragmatically to orient the mind towards the Real,
by “affirming essential qualities that are really only denials of their opposites”’ (2002: 9).
Albahari’s suggestion finds endorsement from both Anantanand Rambachan when he writes that
‘[t]he translation of ānanda as “bliss” is useful for emphasizing the desirability of brahman and
the celebrative and joyful meaning of liberation’ (Rambachan, 2006: 21), and by Bhattacharya
when he equates the empirical (in distinction to the transcendent) pursuit of Brahman as the
attempt to attain a ‘vision of Plenitude’ that ‘is called ānanda “Bliss”’ (2015: 10).78
One might also ask why Śaṅkara – having realised that the ātman-Brahman is ultimately
beyond attribution and transcendent of all notions of deity – still deigned to build temples to
popular deities (as well as compose hymns to these same deities) and advocate the worship of a
personal God.79 Unsurprisingly, says Biderman, the primary reason for Śaṅkara’s even referring
to a ‘God’ or ‘Lord’ (Īśvara) in this sense is to see God as directly relating to the world ‘both as
the first cause of the world and as an object of devotion’ (1982: 246). If the ātman-Brahman
(nirguṇa Brahman) is transcendent and ultimately ineffable, then by contrast, Śaṅkara’s
conception of Īśvara is – as Deutsch remarks – ‘that about which something can be said’ (Deutsch,
1969: 12). What is the significance of this? We have noted that on the surface of things, there
78
Rambachan, it should be noted, explicitly acknowledges that the ātman-Brahman cannot actually be
characterised as ‘bliss’ because this contradicts the fundamental Upaniṣadic teaching that the ātman-Brahman is
‘timeless and present in all states and mental conditions’ (2006: 21). He suggests equating the term ānanda with
‘limitlessness’ and reading it as a description of the nature of ātman rather than a description of an attribute. Of
course, one might wonder whether a description of a nature is not simply a description of an attribute (or a set of
attributes) and so be tempted to dismiss this semantic wrangling as little more than wordplay.
As Biderman helpfully notes, Śaṅkara’s repeated contrasting of Īśvara (saguṇa brahman: Brahman with qualities)
with nirguṇa brahman (Brahman without qualities) ensures a notion of a personal ‘God’ persists – to a point, at
least – even in nondual Advaita (1982: 426).
79
P a g e | 97
appears to be some sort of tension between the idea – that Bhattacharya, following Śaṅkara,
endorses – that the ātman-Brahman (or nirguṇa brahman) is ultimately beyond all attribution
and conception, and the codification of a path to liberation that hinges on the very thing about
which nothing can be spoken. This is of little practical help when discussing what liberation is
and why one should opt to pursue it (often through great difficulty!). What Albahari, Rambachan,
Biderman and Deutsch have in common, then, is the idea that Śaṅkara recognises that we need
to know some sort of lower truth before we can begin to realise a higher one, and, in fact, such
a contention also forms the backdrop to much of Bhattacharya’s own thesis. Swami Nikhilānanda
writes in his translation of Śaṅkara’s Ātmabodha that the reasons for Śaṅkara’s referring to a
personal God, writing hymns to popular deities, and building temples that encourage deity
worship can be readily explained. For Nikhilānanda as for Albahari et al., there is no contradiction
in Śaṅkara’s extolling the virtues of these personal deities and still ultimately holding that the
ātman-Brahman is nondual, absolute and ineffable. On this, he writes that
The devotee catches a glimpse of the Absolute through the form of
the Personal God, who is the highest manifestation of the Infinite
that a finite mind can comprehend on the relative plane. Śaṅkara
reiterates this principle in his philosophy. The beginner learns the
art of concentration through worship of the Personal God (saguṇa
Brahman) and acquires purity of heart through performance of
unselfish duties. Endowed then with concentration and purity, he
sets himself to the task of acquiring Knowledge of Brahman and
realises, in the end, the Impersonal Absolute.
(Nikhilānanda, 1962: xv)
Śaṅkara (and, for the Advaitin, the Upaniṣads) are consequently following a method
familiar to their Buddhist opponents.
That is to say that they are pointing to the
Ultimate/Absolute without actually speaking directly of it: it is, at its core, simply a negative
P a g e | 98
method. To do otherwise would be speaking of what is, in effect, unspeakable. It is precisely
because the ātman-Brahman is so difficult to know that Śaṅkara is tailoring his teachings to the
level of attainment that each practitioner occupies, opting to reveal Truth by the gradual
disclosure of lesser truths. That Īśvara is connected to the world and can be recognised as such
is significant, for it can then provide the necessary foothold for the practitioner to begin to ascend
the transcendental heights that result in realisation of the ātman-Brahman. Even though Īśvara
itself is not the highest realisation per se, it is the highest realisation available on the ‘relative
plane’ or conventional level. The idea is then that once a practitioner has sufficiently developed
their mind through devotion to brahman with qualities, they will be better furnished to realise
the brahman without qualities.80 This journey takes place via a mental process summed up by
the Sanskrit word bādha, usually translated as ‘contradiction’, but which is frequently translated
in Advaitin contexts as ‘cancellation’ or ‘sublation’.81
What is sublation and what role does it play? Deutsch explains that it is ‘the mental
process whereby one disvalues some previously appraised object or content of consciousness
80
It should be mentioned that whilst Buddhism traditionally has little to say on the notion of a personal creator
god, various Buddhist traditions do indeed make use of devas, deities and demons in their narratives (Vajrayāna
Buddhism makes extensive use of deities as meditative tools, for example). Devas are, of course, subject to karma
and so their existence in the heavens is of a finite timeframe. Further, the realms of existence in which these
devas dwell are completely detached from that of our human existence, and so they generally cannot intervene in
human physical or spiritual affairs. In Vajrayāna traditions, deities are generally not supposed to be thought as
actually existent, but are rather representations of archetypes that are used to guide practice. Though
undoubtedly interesting, this wide-ranging discussion is not directly relevant to my current project, and so I will
not dwell too much on this particular facet of Buddhist tradition. It is enough to say that where deities do feature
in Buddhist traditions, they generally are not used in the same way as the Abrahamic religions or indeed in the
same way as the devotional Hindu schools.
Those referring back to Deutsch’s work will note that Deutsch himself prefers to reconstruct bādha in his
writings as ‘subration’ (Deutsch, 1969: 15). This is an idiosyncrasy of Deutsch’s work. I will stick with ‘sublation’ as
there appears to be no discernible difference in meaning and so no meaningful reason for Deutsch’s choice.
81
P a g e | 99
because of its being contradicted by a new experience’ (Deutsch, 1969: 15). To use the clichéd
example oft cited in Advaitin texts as well as Buddhist, we might mistakenly take the coiled rope
on the floor of a darkened room to be a coiled snake. In shock, we reach for a light, and it turns
out that upon closer inspection, when illuminated by our light, this ‘snake’ happily turns out to
be no more than a coiled rope. In this instance, we would say that our initial assessment of the
object of consciousness (the rope) was confused or mistaken, and this mistaken assessment has
since been sublated by the new information gleaned after we switched on the light (viz. that the
snake is actually simply a rope!). It is, as Deutsch writes, ‘a mental process through which one
rectifies errors’, but importantly, it also requires ‘a turning away from, or rejection of, an object
or content of consciousness as initially appraised in the light of a new judgement or experience
which takes the place of the earlier judgement and to which “belief” is attached’ (1969: 16). This
means that in some circumstances (Deutsch gives the example of a mathematical concept which
does not work in one situation but does for another), despite an error of judgement having been
made, sublation does not occur. The significance of sublation for both Advaita and Buddhism is
twofold.82 First, it accounts for the different levels of reality afforded to things and concepts
within the respective philosophies, and second, it illustrates the means by which one can make
spiritual progress. Let us examine why this is the case.
Ascription of ‘different levels of reality’ within the Advaita project is actually relatively
straightforward: the more something is capable of being sublated, the less reality it has. The
82
The process of sublation is not, in fact, unique to Advaita; it is a necessary process for any system operating
within the framework of the two truths. This means that sublation is as much a part of the Buddhist process as it is
a part of the Advaitin process: the need to ‘awaken’ assumes the need to sublate worldly experience!
P a g e | 100
distinction really hinges on what the Buddhists, predating Śaṅkara with their theory of two levels
of truth (and so in some cases, reality), designated as conventional and ultimate truth, a
distinction that, as many scholars – including Whaling (1979: 15) and Suthren Hirst (2005: 90) –
recognise, very likely influenced Śaṅkara via the work of Gauḍapāda. Things that are experienced
on the conventional level are capable of first, being mistaken, and second, being sublated once
other information becomes known that contradicts this prior mistaken judgement or experience.
Even more than this, experiences and judgements on the conventional level are not only able to
be sublated, they are necessarily sublated. This is – as we have already seen – because for the
Buddhists, Śaṅkara, and Bhattacharya, everything on the conventional level is but an imperfect
approximation of the Truth of the ultimate level. 83 Owing to this inherent imperfection,
everything experienced on the conventional level must be sublated in order for us to make
spiritual progress. Put simply, as we begin to realise the ātman-Brahman, ultimate truth
supersedes conventional truth: every conventional judgement or experience is necessarily
replaced by experience of the ultimate; the ātman-Brahman. Suthren Hirst captures the process
nicely when she writes that ‘[w]hen their true nature is realised, these [conventional experiences]
all act as analogies for the process of sublation… one realises the provisional nature of the
conventional world on “waking” to the ultimate truth’ (2005: 92).
83
There is a caveat here in relation to Indian Madhyamaka Buddhism (as opposed to its current-day Tibetan
incarnation) which will be fully explained later. Put simply, it is not obvious that Nāgārjuna actually believed in two
levels of existence or in a privileged ontology of any kind.
P a g e | 101
3.1 Knowing the Ātman-Brahman
There is a more elementary question behind all this talk of Brahman. How do we know
that the Brahman exists prior to embarking upon this intense journey of reflective discovery?
Swami Nikhilānanda asserts that the Vedāntins believe that ‘Brahman is neither a dogma of
religion nor a private mystical experience, but a metaphysical truth based upon universal reason
and experience’ (1962: xvii). In other words, the Brahman exists whether we choose to believe
it as a matter of religious adherence or not. More than that, its existence can be deduced ‘based
upon universal reason and experience’, and so a religious bent of one flavour or another is
presumably all but irrelevant when it comes to knowing that the Brahman exists! This conflicts,
though, with Śaṅkara’s commentary to Brahmasūtra 1.1.2, which deals with the definition(s) of
Brahman. Despite the ātman-Brahman being beyond definition and conceptual thought, Śaṅkara
still allows something to be said about it. Given first is the definition of Brahman as the ‘cause of
the world’, which is the ‘Taṭastha Lakṣaṇa’, or ‘that characteristic of a thing which is distinct from
its nature and yet serves to make it known’ (Swami Vireswarananda, 2014: 21). Śaṅkara writes
in his Brahmasūtrabhāṣya 1.1.2:
That omniscient omnipotent cause from which proceeds the origin,
subsistence, and dissolution of this world–which world is
differentiated by names and forms, contains many agents and
enjoyers, is the abode of the fruits of actions, these fruits having
their definite places, times and causes, and the nature of whose
arrangement cannot even be conceived by the mind,–that cause,
we say, is Brahman.
(Śaṅkara, 2011: 16)
There is a subtlety to tease out here. It is problematic to say in a passage like this that the
Brahman is the cause of the world and all within it whilst simultaneously holding that the
P a g e | 102
Brahman does not act. Bhattacharya, citing the Maitri Upaniṣad says of the ātman-Brahman;
‘[i]t is what makes us act (kārayitṛ), but it does not itself act (akartṛ)’, and that it (the Brahman)
is ‘”as a spectator” (prekṣakavat) of our acts, good and bad, which do not affect it at all (sitāsitaiḥ
karmaphalair anabhibhūtaḥ)’ (2015: 29). According to the Advaitin Swami Vireswarananda, this
apparent causal attribute does not form any part of the Brahman’s nature, which is ‘eternal and
changeless’ (Swami Vireswarananda, 2014: 21). Vireswarananda further writes that Śaṅkara also
uses a scriptural definition to assert that ‘Truth, Knowledge, Infinity is Brahman’, and this
definition he refers to as the ‘Svarūpa Lakṣaṇa’, or ‘that which defines Brahman in Its true
essence’ (2014: 21). How does any of this conflict with Swami Nikhilānanda’s assertion that the
Brahman can be known through ‘universal reason’? Śaṅkara also maintains that the Brahman
cannot be known via reason alone: ‘Brahman is not an object of the senses, it has no connection
with those other means of knowledge. For the senses have, according to their nature, only
external things as their objects, not Brahman’ (2011: 19). This means that if the Brahman cannot
be known via the means of knowledge (pramāṇas), then it cannot strictly be reasoned toward.
Instead, it must be experienced. To this end, reason is only useful insofar as it improves one’s
(conventional) understanding of Brahman: it can never account for Brahman. It would look as
though Nikhilānanda’s assertion that the Brahman’s existence can be known ‘through universal
reason’ is, then, at odds with the position of Śaṅkara.
Whilst Śaṅkara obviously places a high value on the experience of the Brahman, it seems
clear that for him, the śruti is the authoritative source of knowledge of the Brahman. To see this,
we need only look at his Brahmasūtrabhāṣya 1.1.2, where on the one hand he lauds the
P a g e | 103
significance of ‘intuition’ (anubhava) of the Brahman when paired with śruti revelation (2011:
18), yet later in the same text (2.1.6), asserts the following:
For Brahman, as being devoid of form and so on, cannot become an
object of perception; and as there are in its case no characteristic
marks (on which conclusions, &c. might be based), inference also
and the other means of proof do not apply to it; but like religious
duty, it is to be known solely on the ground of holy tradition.
(Śaṅkara 2011: 306-307)
Śaṅkara maintains, according to Swami Vireswarananda, that despite the veracity of the
pramāṇas (means of knowledge) on a conventional level, ‘Brahman cannot be so established
independently of the scriptures (śruti)’ very simply because ‘Brahman is not an object of the
senses’ (2014: 21).84 This is an assertion that we see borne out in the extracts given above.
84
Buddhism – principally following the work of Dignāga and Dharmakīrti – allows two pramāṇas: perception
(pratyakṣa) and inference (anumāṇa). There is also an emphasis on scripture, but this is largely viewed as a kind of
derivative of perception and inference, and so there is a marked difference in the approach of Buddhist schools to
that of the Advaitins. ‘Śruti’ literally translates as ‘what is heard’, but should be understood in the Advaitin context
as something like ‘revelation’ (specifically, scripture that was ‘spoken to’ – revealed to – the ṛṣis). Thus, the key
point to take away from Śaṅkara’s position is that when it comes to knowledge of the Brahman, scripture (śabda;
śruti) is the authoritative pramāṇa.
For clarity, the six pramāṇas generally allowed in Advaita are pratyakṣa (perception); anumāṇa (inference); śabda
(verbalisation of the testimony of the scriptures; verbalisation of śruti; the revealed word); upamāṇa
(comparison/analogy); arthāpatti (postulation/presumption); anupalabdhi (non-recognition/non-perception)).
There is some debate over whether or not Śaṅkara actually endorsed six pramāṇas, with some scholars claiming
that Śaṅkara only ever endorsed three (Radhakrishnan, 1962: 488), and others, such as Sharma, acknowledging
that there is reference to ‘at least five’ pramāṇas in Śaṅkara’s collected works (1992: 520).
Nevertheless, this is a significant point of departure from Buddhism and might serve to illustrate part of a decisive
split not just between Advaita and Buddhism’s methods of teaching and liberation, but also in the principles
grounding these very traditions, viz. how belief in an ātman-Brahman manifests in the first place.
P a g e | 104
Nikhilānanda and Śaṅkara agree, though, that Brahman can be experienced, or to put it
in a way more palatable to the staunch Advaitin, known.85 As we might expect, there is a caveat
on Śaṅkara’s part. Brahma-Sūtra 1.1.3, ‘śāstrayonitvāt’ (literally ‘because scripture is the
source’), proclaims that scripture is the source of self-knowledge (and so because ‘self’ is
brahman, by extension, scripture is the source of knowledge of brahman) – as ‘Brahman has no
form etc. and so cannot be cognised by direct perception’ (Vireswarananda, 2014: 24). For
Śaṅkara, the omniscience represented within the scriptures serve as a sort of signpost that should
impel the practitioner to seek the origin of that omniscience, which consists not in the scripture,
but in omniscience itself (Śaṅkara, 2011: 20). In other words, whilst the scripture manifests the
omniscience and orients us towards it, the scripture is not itself the source of that omniscience.
There is an obvious implication here for both the nature of Brahman and the nature of
knowledge: we have already seen that despite Nikhilānanda’s assertion to the contrary, for the
Advaitin, the Brahman cannot be discovered, found or known through reason alone. In this
sense, we might borrow a phrase from Tillemans (1999: 29) and say that the ātman-Brahman is
‘radically inaccessible’ (Sanskrit: atyantaparokṣa). Indeed, it seems to me that this is a driving
reason for Advaita’s wide range of (arguably) six pramāṇas as opposed to the comparably austere
two pramāṇas of Buddhism. Śruti and śabda necessarily occupy a primary position, for it is only
via the scriptures themselves (śruti) and their verbalisation (by a guru/teacher; śabda) that the
ātman-Brahman can be established at all. This is a fundamental reliance upon the authority of
Again, ‘experienced’ implies a subject-object relationship that the Advaitin would be at pains to point out cannot
exist in a unified, undifferentiated Brahman. It is simply a case of knowing what one is rather than ‘experiencing’
something new or different.
85
P a g e | 105
scripture. Of course, reliance on scripture is – to differing degrees – also present in Buddhism,
but the very simple fact is that, as Bhattacharya must concede, the Buddha does not ever claim
in any surviving literature that his own method is established in virtue of the authority of
revelation. The Buddha’s path to liberation was established through experience, that is to say
that the Buddha’s awakening relied only on insight gained through meditation rather than on any
scriptural direction. The Buddha does not claim that any Buddhist teaching is true simply because
it is written, or because it has been said. Still, the Buddhist traditions that followed the Buddha
have, to some degree, also relied upon ‘scripture’, at least once they finally got around to writing
things down. It would be strange to claim that scripture (āgama) does not direct Buddhists when
every Buddhist tradition bases its core beliefs on the Buddhist method as espoused in scriptures.
Nevertheless, even this reliance is, as Tillemans observes, presented in a novel way. To this end,
Tillemans writes that
[t]his tension between scripture and reason, which is a recurrent
one amongst religious philosophers, was however approached in a
novel way by the Buddhists, a way which allowed them to accept
certain “propositions of faith” but nonetheless retain a rationalistic
orientation and extreme parsimony with regard to acceptable
means of knowledge.
(Tillemans, 1999: 27)
Interestingly, Dharmakīrti – despite his usual proclivity toward vastubalapravṛtta
inferences – allows some recourse to scripture.86 The means by which he justifies this step is in
86
Tillemans (1999: 28-29) writes that an inference which functions according to vastubalapravṛtta functions
‘objectively, or ‘by the force of real entities’, which is to say that such inferences ‘should be evaluated purely on
the basis of facts and states of affairs, and not in any way because of belief, acceptance or faith in someone or his
words’. In terms of a transcendental ātman-Brahman, the implication is clear – a Buddhist is simply not justified to
assert it in virtue of religious dogma; he needs to be able to see it, experience it, and point to it. An Advaitin,
though, is, as we have seen, fully entitled to assert based on religious tradition and scriptural authority.
P a g e | 106
some ways analogous to the means by which Śaṅkara justifies his reliance on scripture for
knowledge of the existence of the ātman-Brahman, and so is of relevance to the current
discussion. Let us see how Dharmakīrti allows such recourse to scriptural authority. First, the
‘epistemological school’ to which Dignāga and Dharmakīrti are said to belong allows three types
of objects: the perceptible (pratyakṣa), the imperceptible (parokṣa), and the ‘radically
inaccessible’ (atyantaparokṣa). As Tillemans points out, the perceptible includes things with
form (rūpa) – such as everyday objects – that are accessible to direct perception; the
imperceptible includes things – impermanence, selflessness – that can be proven via the ‘usual
vastubala kind of inference’. The radically inaccessible includes objects ‘such as the different
heavens (svarga) or the details of the operation of the law of karma, which are, of course,
inaccessible to direct perception’, and we might say that these things are then ‘beyond the limits
of normal rationality’ (Tillemans, 1999: 29). Dharmakīrti intentionally restricts the import of
scriptural authority to instances of the radically inaccessible, which means that he preserves the
integrity of his arguments in favour of inferences being ‘objectively’ grounded more generally (in
the case of pratyakṣa and parokṣa).
There are, however, some more limitations on the use of scripture. Tillemans points out
that Tibetan scholars following Dharmakīrti detail a ‘threefold analysis’ which can ascertain when
it is appropriate to deploy scriptural authority. The point of this analysis is to test whether or not
scripture can be a sound basis for inferential reasoning. In order to be suitable, the relevant
scriptural passage must be ‘(i) unrefuted by direct perception, (ii) unrefuted by
vastubalapravṛttānumāna, and (iii) free from contradiction with other propositions whose truth
is scripturally inferred’ (Tillemans, 1999: 30). Tillemans recognises an inductive argument here,
P a g e | 107
writing that ‘the scripture’s assertions concerning pratyakṣa and parokṣa are seen to be
trustworthy, and so, similarly, its assertions about atyantaparokṣa, if not internally inconsistent,
should also be judged trustworthy’ (1999: 30). How does this compare to the types of knowledge
advanced by Śaṅkara? There is some crossover – both clearly allow some recourse to scripture
to ‘prove’ some points, namely regarding the radically inaccessible. It seems that Dharmakīrti
allows much less recourse, however. Śaṅkara reckons that scripture serves to orientate us
towards the ātman-Brahman and that we could never know the ātman-Brahman if not for the
written words of the divinely inspired ṛṣis.
To this end, the Vedas are self-sufficient and independent of other pramāṇas. As Śaṅkara
writes in his bhāṣya to Brahmasūtra 2.1.1: ‘[t]he authoritativeness of the Veda with regard to the
matters stated [viz. that there is ‘one universal self’: ātman-Brahman] by it is independent and
direct, just as the light of the sun is the direct means of our knowledge of form and colour’
(Śaṅkara, 2011: 295). This is strikingly unambiguous. Just as our knowledge of colour and form
is made possible by the light of the sun, it is the light of the scriptures that make possible
knowledge of the ātman-Brahman. This, it seems to me, places a heavy emphasis on the role
scripture to the detriment of the rest of the pramāṇas allowed by Śaṅkara. Like Dharmakīrti,
though, Śaṅkara is quick to place some restrictions on the scope of scripture, most strikingly is
his insistence that scriptural knowledge claims cannot be valid if they contradict another
P a g e | 108
pramāṇa. 87 This is identical to the restriction placed by Dharmakīrti! What, then, is the
difference?
The short answer is that Śaṅkara appears to see scriptural authority as central: śruti as
the most significant of the pramāṇas.88 The reason for this is that it makes the ātman-Brahman
known to us. Śaṅkara does not appear to foresee a situation in which scriptural inference would
be inaccurate. Dharmakīrti does not share this view, going so far as to deny that scriptural
inference is really a fully-fledged inference at all.89 Tillemans illustrates the crux of the argument
as follows. Whilst scriptural inference is preferable to guesswork or randomness, it is not a
precise science and is not preferable in toto. This is partly because words have no necessary
relation to the objects that they signify (1999: 42), and partly because given that scriptural
inference is an instance of inductive reasoning, incorrect assumptions can be made and so
incorrect conclusions drawn. This sort of inference relies upon probabilities, and so there is an
Indeed, Śaṅkara writes the following in his bhāṣya to the Bṛhadāraṇyaka Upaniṣad (trans. Mādhavānanda, 1950:
301-302):
87
Things in the world are known to possess certain fixed characteristics such as
grossness or fineness. By citing them as examples the scriptures seek to tell us
about some other thing which does not contradict them. They would not cite an
example from life if they wanted to convey an idea of something contradictory
to it. . . You cannot prove that fire is cold, or that the sun does not give heat,
even by citing a hundred examples, for the facts would already be known to be
otherwise through another means of knowledge. And one means of knowledge
does not contradict another, for it only tells us about those things that cannot
be known by any other means. Nor can the scriptures speak about an unknown
thing without having recourse to conventional words and their meanings.
88
89
When śruti refers to the revealed scriptures about which śabda is authoritative.
Tillemans outlines the reasons why scriptural inference cannot be certain (whereas a normal inference ought to
be) in significant detail (1999: 41-47).
P a g e | 109
inherent risk of error. Where scriptural inferences are used, it is, says Tillemans, always subject
to a proviso. Thus, ‘if we make the move of accepting a scripture’s statements on radically
inaccessible matters, it is because we are not, as far as we can judge, precluded from doing so,
and because we want to or need to do so for our spiritual goals’ (1999: 45). The Sanskrit cited by
Tillemans is ‘varam āgamat pravṛttāv evaṃ pravṛttir’, or ‘if [one is] engaged with scripture, better
engage [with scripture] like this’. The turn of phrase indicates, I think, that Dharmakīrti knew
that scriptural inferences were uncertain (or a least carried a risk of uncertainty), but that we
could perhaps minimise the risk of straying too far into prospective uncertainty. The way in which
we might do this is by both analysing them according to the threefold analysis, and bearing in
mind the pragmatic reasons that we have for using them in the first place.
The difference then is that for Śaṅkara, scripture is authoritative when it comes to
determining the existence of and how to orientate oneself towards the ātman-Brahman. This is
an important way in which the scripture is authoritative, especially when the related soteriology
entirely hinges on knowing the ātman-Brahman. It is to effectively say that the only way one can
reach liberation is to know the ātman-Brahman, and the only way in which one can know the
ātman-Brahman is via the requisite scriptures (namely the Upaniṣads), presumably with some
added help from Śaṅkara’s commentaries on such epics as the Gītā. Dharmakīrti, on the other
hand, does not place any such burden on the Buddhist scriptures, because its most central of
claims (minus the operations of karma) are backed via vastubala inferences. Whereas Śaṅkara
thinks that we need recourse to scripture to know the ātman-Brahman, Dharmakīrti thinks that
we can know the truth of anātman more directly, through perception and inference. The main
aspect of Advaita soteriology is, when it comes down to it, solely reliant on scriptural authority.
P a g e | 110
The main aspects of Buddhist soteriology are, in contrast, grounded in reason: the scope of
scriptural authority for the Buddhist is much narrower.
This amounts to a real difference in how we gain important types of knowledge. For
Advaita we know that knowledge of the ātman-Brahman is primary. This is directed by scripture
alone, for we would not even know of the ātman-Brahman’s existence were it not for the
scriptures. For Dharmakīrti as a Buddhist, the proof of the pudding is really in the eating. We do
not in principle need scripture to discover the Buddhist path; we could reason our way there
regardless of scripture. With all this in mind, we might ask of what use is scripture for someone
like Dharmakīrti? Dunne (2004: 243) translates this from Pramāṇavārttika 1.218cd:
Every judicious person who wishes to act analyses statements to
determine what is and what is not scripture (āgama); he does so as
one who wishes to act [effectively], and not because of some
pernicious habit. Learning what should be put in practice from the
scripture, he thinks, “Having acted accordingly, I might realise my
goal.” On the basis of the trustworthiness of that scripture with
regard to things that can be experienced [through perception or
empirical inference], that person acts with regard to other things
[i.e. the supersensible objects described in that scripture] because
such is the case for most practical action in the world.90
From this, it is obvious that for Dharmakīrti, scripture does have some instrumental value.
It is to be measured with some caution, however, and only in relation to those aspects of the
scripture which we can verify ourselves: if these aspects are verifiable through perception and
inference, then we might provisionally trust in the more outlandish, supersensible aspects of the
90
All annotations are Dunne’s.
P a g e | 111
scripture. This is, as it turns out, very different indeed from the emphasis placed on scripture by
the Advaitins.
3.2 Ultimate Knowledge
We have established that for Śaṅkara, the ātman-Brahman falls into the category of
‘radically inaccessible’ in the same way that the Absolute does for thinkers such as PseudoDionysius and Maimonides; after all, we have already seen in the Brahmasūtrabhāṣya the role of
scripture in revealing the ātman-Brahman’s existence. It is here that Śaṅkara explicitly denies
that the ātman-Brahman can be known in any way other than through the revelation catalogued
in the scriptures.91 We must remind ourselves, though, that for Advaita, there are potentially six
pramāṇas and not only two: we have seen how Śaṅkara allows one pramāṇa specifically in order
to safeguard the validity of both scripture and the verbal communication of scripture, a move not
permitted by Buddhists of any tradition.
Further, in his translation of Śaṅkara’s commentary to Brahmasūtra 1.1.4,
Vireswarananda writes that
[t]he uniqueness of Brahman is quite apparent, as It cannot be
realised either by direct perception or inference in the absence of
form etc. and characteristics respectively. Reasoning also has been
adopted by the scriptures here by citing the example of clay to
elucidate their point. As different objects are made out of clay, so
are all things created from this Brahman.
(Swami Vireswarananda, 2014: 26)
That is to say that we would not know of the ātman-Brahman’s existence – or even to bother looking for it – if it
were not for the scriptures detailing as much. It is not something that can be discovered out in the world by some
accident.
91
P a g e | 112
This passage is telling on two fronts. First, Vireswarananda specifically denies that
knowledge of the ātman-Brahman is any sort of inference – a distinction that will prove relevant
shortly. 92 Second, we see an instantiation of another pramāṇa, analogy (upamāṇa), when
Vireswarananda elaborates on the example of one lump of clay making many things. It is the
scriptures that give this example (specifically Chāndogya Upaniṣad 6:1:2-6), but the ātmanBrahman discussed in these same scriptures is also supposed to be ultimately non-dual, inactive
and so without attributes. It must be the case, then, that this example of action is given not as
an ultimate, literal account of the Brahman, but rather as an analogy to allow the unenlightened
practitioner to begin to gain knowledge of the Brahman on a conventional, basic level. The same
applies to the pramāṇa of inference (anumāṇa): on an ultimate level, no inference gives insight
to the nature of the ātman-Brahman, but on the conventional level that we occupy, some
inferences serve to orientate the practitioner towards Brahman. How might this work? The most
obvious instance is when Śaṅkara argues that it follows from the fact that there is a body of
scripture ‘possessing the quality of omniscience’ that the Brahman ‘is the source, i.e. the cause
of the great body of Scripture’ (2011: 20). He further states that ‘Scripture consisting of the Rigveda, &c., as described above, is the source or cause, i.e. the means of right knowledge through
which we understand the nature of Brahman’ (2011: 20), and goes on to write that it is via
scripture alone that we come to know Brahman as the cause and origin of the world. How does
scripture tell us this? Quite simply because Brahman inspired the scriptures! There are, of
I think it worth mentioning that ‘knowledge of’ here means something very specific, viz. knowledge of the
essence of the Brahman. It refers to an intuitive mystical gnosis.
92
P a g e | 113
course, some inferences here. But the point is that such inferences are conventional and do not
stand alone as Truths: the point of the scriptures talking in such lowly conventional terms – which
are ultimately inapplicable to the Brahman – is to inspire action on the part of the reader. The
scripture should spur the reader to introspection and meditation upon the ātman-Brahman. It is
then through that action that one might come to know Brahman through intuitive experience.
It is still the case, however, that the pramāṇas are doing the work here. Though the
Brahman is ultimately formless and without attributes, at least two pramāṇas are at play when
we speak about ‘knowledge’ of it: first, scriptural authority, second, analogy (which builds on the
ideas revealed through scripture). There might even be a third pramāṇa at play – anubhava:
direct experience, intuition – though whether or not this strictly counts as a pramāṇa is
somewhat controversial (Sharma, 1992).93 However, there is a caveat. Insofar as the pramāṇas
are useful to the Advaitin practitioner, they are only useful on this conventional level. It is here
that we begin to see some strong parallels with the Madhyamaka of Nāgārjuna. Śaṅkara
recognises that there is a tension between the ultimate existence of the ātman-Brahman and the
process of directing people toward this ultimate existence using words (śabda, śruti). Like
Nāgārjuna, Śaṅkara believes that there are two levels of truth and reality: conventional and
Sharma argues that anubhava should be classed as a pramāṇa, but in a slightly different way than Śaṅkara would
usually admit. By Sharma’s reading (1992: 522), this would involve classing anubhava as a ‘means of valid
knowledge’ rather than a ‘valid means of knowledge’ (which is how a pramāṇa is usually defined). The difference
is subtle but significant. Sharma writes that ‘Śaṅkara seems to equivocate, for śruti is a valid means of knowledge
about Brahman in the vyāvahārika realm, and anubhava is a means of valid pāramārthika knowledge. In the case
of śruti by itself, one 'knows' about Brahman because one cites the scriptures; in the case of anubhava, one cites
the scriptures because one knows [Brahman]’ (1992: 522). The idea is that at the time anubhava is used as a
pramāṇa, the need for pramāṇas passes: once one has direct perception of the Brahman, the need for language
describing the knowledge (be it valid means of knowledge or means of valid knowledge) disappears.
93
P a g e | 114
ultimate.
For Śaṅkara, Brahman is the ultimate truth/reality, and everything else is, in
Bhattacharya’s words, an imperfect approximation: ‘[a]ll truths as can be formulated are, in fact,
but approximations of Truth, which is inexpressible; none of them can be identified with Truth
itself’ (2015: 9).
This means that every account of reality that does not culminate in
intuition/knowledge/experience of the ātman-Brahman is necessarily inferior. It is lacking in a
fundamental way – it is avidyā.
Nāgārjuna thinks that in order for the pramāṇas to be established in the way that his
Naiyāyika opponents assume they are, the pramāṇas must themselves be ultimate in nature.
That is to say that the pramāṇas must establish their own existence, and as such, cannot be
subject to causes, conditions or influences. According to such criteria, a pramāṇa must then
possess a svabhāva. It is a matter of fact that Nāgārjuna has no time for svabhāva, and so it
should come as no surprise that his starting position will be that a pramāṇa cannot exist with
svabhāva. Viz., a pramāṇa cannot be self-caused/self-established.
Of course, if we accept svabhāva, then we have quite a different view. A proponent of
svabhāva might well claim that we do, in fact, see intrinsic natures in the world’s phenomena.
Where Nāgārjuna sees a mistake, such a person might see insight into the true nature of reality
(i.e. that everything exists with svabhāva). Accordingly, such an objector might want to argue
that the pramāṇas are indeed self-established in virtue of their respective intrinsic natures.
P a g e | 115
Indeed, at Vigrahavyāvartani (VV)94 8, Nāgārjuna offers one such objection to his arguments that
all entities are empty of svabhāva:
nairyāṇikasvabhāvo dharma nairyāṇikās̒ca ye teṣām /
dharmāvasthoktānāmevamanairyāṇikādīnām //8//
And those things which are conducive to liberation [possess] intrinsic natures
[which are] conducive to liberation /
Thus, [the same holds for] things spoken of [in relation to] the state of things
[and for things that] are not conducive to liberation //8//
Nāgārjuna further details his opponent’s stance in the commentary:
iha ca dharmāvasthoktānāṃ nairyāṇikānāṃ dharmānāṃ
nairyāṇikaḥ svabhāvaḥ anairyāṇikānāmanairyāṇikaḥ
(svabhāvaḥ)95 bodhyaṅgikānāṃ bodhyaṅgikaḥ
abodhyaṅgikānāmabodhyaṅgikaḥ bodhipakṣikāṇāṃ
bodhipakṣikaḥ abodhipakṣikāṇāmabodhipakṣikaḥ / evamapi
s̒eṣānām / tadyasmādevamanekaprakāro dharmāṇāṃ svabhāva
dṛṣṭastasmādyadyuktaṃ niḥsvabhāvāḥ sarvabhāvā
niḥsvabhāvatvāccūnyā iti tanna /
Now, things mentioned in connection with the state of things, and
things [that are] conducive to liberation, have an intrinsic nature
[that is] conducive to liberation. [Things that are] not conducive to
liberations [have an] intrinsic nature [that is] not conducive to
liberation, the limbs of liberation [have an intrinsic nature that is]
the limbs of liberation, [those that are] not the limbs of liberation
[have an intrinsic nature that is] not the limbs of liberation. [The
things that] belong to perfect wisdom (bodhipakṣika) [have an
intrinsic nature that] belongs to perfect wisdom, [those things
that] do not belong to perfect wisdom [have an intrinsic nature
that] do not belong to perfect wisdom. Thus for the remaining
Sanskrit text for the Vigrahavyāvartani taken from Bhattacharya, Johnston, Kunst, The Dialectical Method of
Nāgārjuna: Vigrahavyāvartanī (2002).
94
95
This is added in the Tibetan translation once, and the Chinese translation after each item in the list, though it is
apparently not present in the original Sanskrit (Bhattacharya, Johnston, Kunst, 2002: 49). The meaning remains
largely the same whether svabhāva is added or not, with the first instance of svabhāvaḥ enough (strictly speaking)
to establish that we are referring to the intrinsic natures of each consecutive item mentioned. I have added it here
simply because I think it reads a little easier with the second ‘svabhāvaḥ’ included.
P a g e | 116
[things]. Because in this way different kinds of intrinsic natures
are seen, the statement ‘all things lack intrinsic nature, [and
because of this] lack of intrinsic nature [all things are] empty
(śūnya)’ is not valid (tan na).
The opponent is stating that because we can readily distinguish intrinsic natures, the
Mādhyamika contention that all entities are empty of intrinsic natures cannot be established.
This serves two purposes in relation to the pramāṇas. First, because the pramāṇas are obviously
supposed to be conducive to liberation (insofar as they allow one to determine what is Real and
disregard that which is not), nairyāṇikasvabhāvo dharma nairyāṇikās̒ca ye teṣām is simply a
statement of fact: things that are useful to liberation have intrinsic natures that make them useful
to liberation. Second, if the pramāṇas do indeed have an intrinsic nature, then they are
established from their own side and so require no further metaphysical or epistemic justification.
This means that by extension, the pramāṇas are ultimately existent. In fact, even if it were
somehow the case that the pramāṇas were not conducive to liberation, the opponent holds that
they would still have an intrinsic nature. In such a case, this nature would simply be one that is
not conducive to liberation! It is obvious that the objector holds an essentialist view of the world
– the existence of svabhāva is not in question for them, instead, the only question is around how
each intrinsic nature is to be categorised (conducive/not conducive to liberation, etc.).
We might say that so far, this looks like less an argument from the opponent and more of
an assertion. The reasons for their holding this position are not developed until VV 9, where the
opponent makes the familiar argument that without intrinsic natures grounding things in the
world, the statement purporting to negate intrinsic natures could not possibly exist. The
statement ‘there are no intrinsic natures’ requires that there be an existent referent, and for the
P a g e | 117
opponent, the only way such a referent could possibly exist is through possession of an intrinsic
nature. Thus at VV 9 we read:
yadi ca na bhavetsvabhāvo dharmānāṃ niḥsvabhāva ityeva /
nāmāpi bhavennaivaṃ nāma hi nirvastukaṃ nāsti //9//
If things [had] no intrinsic nature, even the name ‘no intrinsic
nature’ [would] not exist /
For [there] is no name without [an] object [to which it refers] //9//
This is a simple reassertion of the view – common then as it is now – that there must be
some grounding principle behind reality. Of course, we know that we do indeed have names for
things that do not exist in reality: we would be hard pushed to find an existent banshee to which
the name ‘banshee’ refers, for example. Nāgārjuna further extrapolates his opponent’s position
in the commentary to VV 9, writing in the Sanskrit:
yadi sarvadharmāṇāṃ svabhāva na bhavettatrāpi niḥsvabhāva
bhavet / tatra niḥsvabhāva ityevaṃ nāmāpi na bhavet / kasmāt /
nāma hi nirvastukāṃ kiṃcidapi nāsti /
tasmānnāmasadbhāvātsvabhāvo bhāvānāmasti
svabhāvasadbhāvāccāśūnyāḥ sarvabhāvāḥ / tasmādyaduktaṃ
niḥsvabhāvāḥ sarvabhāvā niḥsvabhāvatvācchūnyā iti tanna /
If all things [have] no intrinsic nature, there would be an absence
of intrinsic nature. Then, even the name ‘absence of intrinsic
nature’ [would] not exist. Why? Because a name without an
object [to which it refers] does not exist. Thus, because [the]
name exists, [so too does] intrinsic nature. Intrinsic nature being
true (sadbhāva), all things are non-empty (aśūnyāḥ sarvabhāvāḥ).
Therefore, the statement ‘all things lack intrinsic nature, [and
because of this] lack of intrinsic nature [all things are] empty
(śūnya)’ is not established (tan na).
The argument from the opponent is simple enough to grasp – we cannot say that an entity
lacks svabhāva if we want to speak meaningfully about the world. Intrinsic nature is the
grounding principle of all existent things, and without this grounding principle – intrinsic nature
P a g e | 118
– there is nothing to be negated. We can only make sense of the statement ‘there is no intrinsic
nature’ if there is actually an intrinsic nature to deny - something can only be negated in virtue
of its existence. The opponent’s view assumes that intrinsic nature exists and that it grounds
everything. This means that every object, statement, entity etc. is grounded by its svabhāva
(which by extension means that every entity has what we might loosely call a ‘self’).
It is not obvious, however, that Nāgārjuna need believe an intrinsic nature actually exists
in order to dispute how it might theoretically work. Just as I need not believe that God exists in
order to make sense of a conversation about the attributes of God with a theist, so too may
Nāgārjuna discuss with an opponent principles and ideas that do not actually exist. As I have
already said, I can talk of Harry Potter-style Dementors with my young cousins (sort of – I am far
from an expert!), but I think it a stretch to say that either of us need accept their actual existence
in the world. I can write a story about vampires without accepting their actual existence. Safe
to say then, that negating the idea of intrinsic existence does not necessarily require that the
negating statement be grounded in reality by an intrinsic nature. This is to in some ways place
the cart before the horse – ideas do not need to exist as ‘real’ entities in order to be discussed:
pramāṇas do not need to exist in order to be negated.
As mentioned earlier, a convenient way to think of this is as follows: it is not necessary
for a unicorn or minotaur to exist independently in the world for me to deny their existence in
the world. I cannot reasonably be accused of tacitly affirming the ultimate existence of both
unicorns and minotaurs by writing that they are mythical creatures that do not exist. All that is
really demonstrated is that I understand what both the idea of a unicorn involves (its status as a
P a g e | 119
horse with a horn growing from its head) and what the idea of the Minotaur involves (its status
as a creature with the body of a man and the head of a bull) and so on. We can say that as long
as we understand the idea or concept of something, it does not much matter if it exists
independently of the mind. Being able to categorise ideas is what really matters, not whether
there ever existed entities or objects falling under those ideas. This is, I hazard, Nāgārjuna’s
starting point. Whilst he understands what intrinsic nature entails conceptually, he thinks that
he can still deny its existence (much as he denies the establishment of the pramāṇas) without
the baggage of a tacit affirmation of the ultimate existence of intrinsic nature (or the pramāṇas).
Such an affirmation would, after all, be an insidious trap that binds us to a view. As Huntington
(2017: 18) so eloquently writes:
One thing is clear: Nāgārjuna’s soteriological aim entails the
cessation of all clinging to ideas (vikalpa), views (dṛṣṭi), assertions
(pakṣa), and propositions (pratijñā), and the consequent immersion
in a groundless state of non-abiding—what is referred to in the
Mūlamadhyamakakārikā as the “extinction of conceptual diffusion
in emptiness”.
The point of the objection is that the opponent thinks that all of reality must be reducible
to some sort of ultimate basic. Obviously, the issue for the Mādhyamika is that this state of affairs
is both deeply unattractive as well as untenable in practical terms. Nāgārjuna expends a great
deal of effort in refuting the very notion of objects and ideas being ‘established’ (by which he
means real, or ultimately existent rather than simply conventionally existent), and so even the
pramāṇas by which thinkers such as Śaṅkara justify their knowledge of the ātman-Brahman
cannot be ultimately existent. This is significant to us not simply in terms of the rejection of
Advaitin epistemology, but also more broadly. Nāgārjuna clearly and consistently demonstrates
P a g e | 120
a deep-seated aversion to accepting any ‘ultimate’ entity, rejecting the possibility of svabhāva at
every opportunity. This means that whilst Nāgārjuna does not offer one large, concerted
argument specifically against the ātman-Brahman, the consistent application of his philosophy
must necessarily preclude one from equating śūnyatā with the ātman-Brahman.
Let us then look at Nāgārjuna’s particular arguments against the establishment of
pramāṇas as presented in the VV. To be clear, my aim here is not to refute or put forth a theory
of knowledge per se. It is strictly to illustrate the far-reaching nature of Nāgārjuna’s staunch antiessentialism and to illustrate the lengths to which he is prepared to go in order to refute any kind
of permanence, reification, or theory-building. I hope that it is by now clear that these lengths
are crucial in distinguishing Nāgārjuna’s version of Madhyamaka from Advaitin doctrine. The
most important arguments for my purposes begin around VV 30, and so it is here that I will start.
The verse on its own is relatively terse and difficult, and so I will quote both the verse and the
commentary from VV 30-33 and extrapolate from there. It is worth bearing in mind, I think, that
the majority of the arguments presented here are, though aimed primarily at the Naiyāyikas,
effective toward anybody holding this sort of pramāṇa viewpoint: this includes other Buddhists.
To reiterate, the issue is not with a pramāṇa viewpoint per se. It is with the essentialism that the
viewpoint – especially when taken on any terms that resemble the Naiyāyika position – seems to
entail.
VV 30 reads:
yadi kiṃcidupalabheyaṃ pravartayeyaṃ nivartayeyaṃ vā /
pratyakṣādibhirarthaistadabhāvānme ‘nupālambhaḥ //30//
P a g e | 121
yadyahaṃ kiṃcidarthamupalabheyaṃ96
pratyakṣānumānopamānāgamaiścaturbhiḥ pramāṇaiścaturṇāṃ
vā pramāṇānāmanyatamena ata eva pravartayeyaṃ vā
nivartayeyaṃ vā / yathārthamevāhaṃ kaṃcinnophalabhe
tasmānna pravartayāmi na nirvartayāmi / tatraivaṃ sati yo
bhavatopālambha ukto yadi pratyakṣādīnāṃ
pramāṇānāmanyatamenopalabhya bhāvānvinivartayasi nanu tāni
pramāṇāni na santi taiśca pramāṇairapi gamyā arthā santīti sa me
bhavatyevānupālambhaḥ /
If I apprehended something [via means of] perception etc., then I
would affirm or deny [something about the object apprehended] /
[But since such] an object does not exist, I am not to blame //30//
If I apprehended some object [via the] four pramāṇas, by
perception, inference, comparison, and authority (āgamaiḥ), or
through one of these four pramāṇas, then I would affirm or deny.
Since, in fact (eva), I do not ever perceive an object [via the
pramāṇas], I neither affirm nor deny. In this context, your
criticism is ‘Though you deny objects after having apprehended
them by one of the pramāṇas [such as] perception, [you claim
that] those pramāṇas do not exist and the objects apprehended
by [those] pramāṇas do not exist.’ That does not concern me at
all.
There is an anti-essentialist undercurrent throughout the verses cited that is worth
considering in some more detail. First, to VV 30: here we see Nāgārjuna contend that he neither
asserts nor denies anything about objects perceived via the pramāṇas precisely because there is
no object in existence which is perceived via the pramāṇas. What this means is not that objects
that are perceived do not exist (or at least appear to exist!), but rather that objects cannot be
perceived by pramāṇas. This subtle difference is meant to highlight that perception does not
96
The first member of this compound is given in the BJK version of the VV as kaṃcid. I follow the amendment by
Rāhula Sāṅkṛtyāyana in substituting this by kiṃcid instead. There is no discernible change in meaning with this
substitution, I simply think that it reads better; the particle kam being principally Vedic in terms of origin and era of
use (Whitney, 1879: 408).
P a g e | 122
occur via some things called pramāṇas – a controversial notion even in Buddhist circles! We will
see that Nāgārjuna thinks that this is the case because he does not think that pramāṇas can be
satisfactorily explained or established. Nāgārjuna goes on to write that any criticism based
around the idea that he somehow utilises the pramāṇas to perceive objects which he then
roundly denies simply ‘does not concern’ him – this is because it falls foul of an elementary
misunderstanding and misrepresents his actual argument. But how can this be the case? The
objector seems to be arguing that the pramāṇas are what we perceive the world through
whether or not we believe that they exist. To them, the very fact that Nāgārjuna perceives at all
means that he is utilising at least some pramāṇas – he might as well be sitting on a chair whilst
claiming that the chair does not exist!
Of course, Nāgārjuna would indeed sit on a chair and claim that it does not exist, because
when he uses the word ‘existent’ (bhāva), he generally means ‘ultimately existent’.97 This is
apparent throughout his various writings, but is perhaps most neatly summarised at MMK 15.4
where Nāgārjuna argues that ‘existence’ is a result of intrinsic or extrinsic natures.98 For a thing
Siderits and Katsura (2013: 158) note that Nāgārjuna might sometimes be employing wordplay when he uses
bhāva; it can after all mean ‘existent’ or ‘nature’. Owing to such ambiguities inherent in Sanskrit, it is largely down
to the interpreter to try to make sense of such quirks in style, relying on, above all, context. It is owing to the
overall context of this chapter of the MMK that I feel comfortable in discerning that bhāva here stands for
‘existent’, which is then shorthand for ‘ultimate existent’ (viz. an existent with svabhāva).
97
98
MMK 15.4:
svabhāvaparabhāvābhyām ṛte bhāvaḥ kutaḥ punaḥ /
svabhāve parabhāve ca sati bhāvo hi sidhyati //4//
Without intrinsic and extrinsic natures, whence [comes an] existent? /
[For an] existent is established on account of intrinsic or extrinsic nature.
P a g e | 123
to be ‘established’, it must be ‘ultimately established’ and so ultimately exist. In other words, it
must have svabhāva. It is crucial to remember that for Nāgārjuna, to possess svabhāva is to be
ultimately existent, immutable and permanent.99 This is the central theme underpinning all of
his work – svabhāva sabotages the soteriological project of Buddhism as its possession precludes
change. It is for this reason that Nāgārjuna spends so much time railing against what he considers
essentialist philosophical and religious positions – it is precisely because of their essentialism (viz.
their endorsement of svabhāva) that they remove the means to liberation. 100 Nāgārjuna’s
problem is not with perception – he knows very well that we perceive and experience. This is, in
Nāgārjuna elsewhere argues that neither intrinsic nor extrinsic natures are possible, for example at MMK 15.1-2
where Nāgārjuna argues that svabhāva cannot be produced by causes and conditions because it would then be a
product and thus not intrinsic at all. This is unacceptable to even the Ābhidharmikas, for as Karunadasa writes,
‘’[s]ince a dhamma and its intrinsic nature are the same (for the duality is only posited for purposes of
explanation), to claim that its intrinsic nature undergoes modification is to deny its very existence’ (1996: 21).
Nāgārjuna thinks that something that is a product is modified, namely by its causes and conditions. Consequently,
a produced existent (something that exists with svabhāva) cannot exist.
By the time of Nāgārjuna’s writing, this was, I think, the standard understanding of svabhāva across many
traditions. It was not always this way, however. For more on the development of svabhāva and dharma as
ultimately existent entities, see Ronkin (2005). For more on the cognitive and ontological aspects of svabhāva, see
Westerhoff (2007).
99
The route to liberation is here hindered in two ways. First, if svabhāva exists (remember that it would
necessarily be an ultimately existent thing), then this would be enough to physically halt any change at any place at
any time. The idea is that a thing with svabhāva simply cannot be otherwise, and so it cannot stop being the way
that it is. In terms of liberation and removal of defilements and the like, this would entail that every defilement
exists the way it exists permanently, and so accounting for liberation becomes extremely problematic. The upshot
would be that nobody could ever become liberated. Nāgārjuna does not accept this – the Buddha is proof that we
can reach liberation! For Nāgārjuna, then, svabhāva cannot exist and must instead be a mistaken imputation
forced onto objects by deluded minds. This is what forms the basis of every argument deployed by Nāgārjuna in
both the MMK and the VV.
100
Second, the very idea of svabhāva halts progress along the soteriological journey. Why? Very simply because it is
something to which practitioners cling. This also goes for Buddhist practitioners, who sometimes think that
śūnyatā is the essential principle of everything, making it equivalent to svabhāva in all but name (indeed,
Bhattacharya makes this very claim (2015:13)). Nāgārjuna wants to avoid this sort of conflation and remove any
form of clinging: to do this is to reify that which is supposed to remove the habit of reification! There will be more
to say on this a little later.
P a g e | 124
fact, a basic maxim of Buddhism.101 Rather, Nāgārjuna could be classified as what Mohanta calls
a ‘cognitive sceptic’: elaborating on what this position entails, Mohanta (1997: 53) writes that
Cognitive scepticism may be taken thus for a philosophical attitude
which suspends the possibility of making conclusive statements
concerning valid cognition (prāma) for want of sufficiently
warranted grounds or pramāṇas. A cognitive sceptic does not go
for theory-making.
Nāgārjuna is concerned, then, with three things. First with casting doubt on the
assumption that the pramāṇas establish the (ultimate) reality of their objects (prameyas), second
with casting doubt on the assumption that the pramāṇas themselves are established as ‘real’ (in
Nāgārjunian terms, ultimately real), and third, with avoiding advancing any theory of his own.102
What is the misunderstanding that ‘does not concern’ Nāgārjuna? It is the idea that he
perceives via the same pramāṇas that he argues are not established. In light of the above, this
is simply not a problem for Nāgārjuna. He is not claiming that our experience of the world is false
per se, nor is he claiming that there is no perception or even that we cannot make some sorts of
claims about the world that are conventionally grounded in experience. Rather, the point is that
ultimately, such claims are not grounded in anything. This is significant in terms of the current
work, because it means that by extension, there is no ground. Put another way, there can be no
101
Insofar as Buddhism aims to analyse, reduce and then remove all experiences of dissatisfaction (duḥkha).
As Ayer (1956: 40) notes, ‘All that [the cognitive sceptic] requires, is that errors should be possible, not that
they should actually occur. For his charge against our standards of proof is not that they work badly; he does not
suggest that there are others which would work better. The ground on which he attacks them is that they are
logically defective; or if not defective, at any rate logically questionable.’ The ‘no-thesis’ approach alluded to by
Ayer is ubiquitous in Madhyamaka philosophy.
102
P a g e | 125
ātman-Brahman. Thus for Nāgārjuna it is the case that ultimately, the pramāṇas are not
established, and nor are their objects (prameyas):
yadi ca pramāṇataste teṣāṃ teṣāṃ prasiddhirarthānām /
teṣāṃ punaḥ prasiddhiṃ brūhī kathaṃ te pramāṇānām //31//
yadi ca pramāṇatasteṣāṃ teṣāmarthānāṃ prameyāṇāṃ
prasiddhiṃ manyase yathā manairmeyānām teṣāmidānīṃ
pratyakṣānumānopamānāgamānāṃ caturṇāṃ pramāṇānāṃ
kutaḥ prasiddhiḥ / yadi tāvanniṣpramāṇānāṃ pramāṇānāṃ
syātprasiddhiḥ pramāṇato ‘rthānāṃ prasiddhiriti hīyate pratijñā /
And if proof of these objects is established based on pramāṇas /
tell [me] how these pramāṇas are [in turn] established //31//
And if you think that such objects of true cognition are established
via the pramāṇas as a measuring instrument [establishes that
which is to be] measured, how are the four pramāṇas –
perception, inference, comparison, authority (āgama) –
established? If the pramāṇas may be established without
[recourse to] pramāṇas, [then your] proposition [that] ‘objects are
established via the pramāṇas’ is abandoned.
Here at VV 31, Nāgārjuna notes that his opponent argues for the existence of objects of
valid knowledge because they can be perceived through the pramāṇas, which are the only true
means of knowledge. According to his opponent, these pramāṇas thus give insight into the true
nature of reality. However, the question is once again raised as to what it is that establishes the
pramāṇas themselves (and so designates them as the tools by which true knowledge might be
gained). Nāgārjuna is quick to bring in the relational characters of pramāṇas as the means of
valid knowledge, and prameyas as the objects of valid knowledge. One does not make sense
without the other: how can we have instruments of valid knowledge if we do not have objects of
knowledge for the instruments to help us ‘know’? For the Naiyāyika opponent it is clearly the
P a g e | 126
case, then, that the objects of knowledge are ‘established’ by the pramāṇas, but what of the
pramāṇas themselves?
Nāgārjuna asks how it might be the case that the pramāṇas establish something without
first being established themselves. The problem here is twofold. First, accepting that the
pramāṇas have a causal power to establish objects of knowledge (i.e. that the pramāṇas are the
means through which we perceive reality) without asking what grounds the pramāṇas
themselves is, at best, an example of unjustified dogmatism. It is at worst indicative of a shortsighted essentialism.103 The significance of this type of accountability cannot be overstated –
Nāgārjuna is, after all, a loyal Buddhist. His Buddhist praxis and philosophy revolves around the
acceptance of radical change and radical interconnectivity through pratītyasamutpāda, but this
commitment would be jeopardised by a dogmatic assertion of a permanent ‘grounds’ of
knowledge. Second, if we do ask how the pramāṇas are established and conclude that they
‘prove themselves’, then we effectively say that they have svabhāva. The end result for
Nāgārjuna is that mindlessly accepting the pramāṇas as ‘established’ leads to undesirable and
untenable positions. We can assert their essential establishment without due consideration of
the facts (based, for example, on religious tradition); we can explicitly make the argument that
they are somehow self-established and so exist intrinsically with svabhāva; we can make some
recourse to an infinite regress of causes, where each pramāṇa establishes another, which
establishes another, which establishes another, and so on. At some point, one of these infinitely
Nāgārjuna thinks that essentialism is impossible. More than that, he thinks that it is demonstrably impossible.
With this in mind, holding that the pramāṇas ‘just are’ existent is, for Nāgārjuna, a lazy sidestep into an
essentialism that makes little intellectual or practical sense.
103
P a g e | 127
regressed pramāṇas – established by innumerable other pramāṇas – then somehow establishes
its prameyas as well.
Thus we read at VV 32:
anyairyadi pramāṇaiḥ pramāṇasiddhirbhavettadanavasthā /
nādeḥ siddhistatrāsti naiva madhyasya nāntasya //32//
yadi punarmanyase pramāṇaiḥ prameyāṇāṃ prasiddhiteṣāṃ
pramāṇānāmanyaiḥ pramāṇaiḥ
prasiddhirevamanavasthāprasaṅgaḥ / anavasthāprasaṅge ko
doṣaḥ / anavasthāprasaṅga ādeḥ siddhirnāsti / kiṃ kāraṇam /
teṣāmapi hi pramāṇānāmanyaiḥ pramāṇaiḥ
prasiddhiteṣāmanyairiti nāstyādiḥ / āderasadbhāvāt kuto
madhyaṃ kuto ‘ntaḥ / tasmātteṣāṃ pramāṇānāmanyaiḥ
pramāṇaiḥ prasiddhiriti yaduktaṃ tannopapadyata iti /
If the pramāṇas are established by [other] pramāṇas, [then] there
is an infinite series /
[This being the case,] neither the beginning, nor the middle, nor
the end can be established //32//
If you think [that the] objects of true cognition (prameya) are
established as such by the means of true cognition (pramāṇas),
then there follows an infinite series (anavasthāprasaṅga). What
fault [is there] in an infinite series? Because you think that these
pramāṇas are established by other pramāṇas, and those other
pramāṇas [through other pramāṇas], no beginning exists. Without
a beginning, how [can there be] a middle and end? Consequently,
your statement ‘the pramāṇas are established by [other]
pramāṇas’ is not valid.
Toward the very end of VV 31, Nāgārjuna discusses the case for other pramāṇas
establishing some other pramāṇas. But does this explanation work for us? At VV 32, Nāgārjuna
argues that this is not an acceptable resolution to the issue of the pramāṇas’ establishment.
Why?
Simply because from such a position, ‘there follows an infinite series
[anavasthāprasaṅga]’, and this particular sort of infinite regress is undesirable. We already know
P a g e | 128
that Nāgārjuna is not afraid of accepting the occasional infinite regress – he fully accepts the
infinite web of causation that is necessitated by pratītyasamutpāda, for example. The difference
is that the regress currently under discussion is vicious rather than virtuous: Westerhoff explains
that the regression is vicious ‘since the burden of proof is transferred in its entirety to the
preceding stage, as a [single] epistemic instrument would have to establish all the succeeding
ones’ (Westerhoff, 2017). The major difference between this type of infinite regress and the type
of infinite regress present in discussion around pratītyasamutpāda is that in the latter case, one
object or process is not 100% responsible for another. The very idea of pratītyasamutpāda is
that everything is linked in inextricable, intimate ways. For every one object, there are huge
numbers of intertwined causes and conditions. The Buddhist position is that this stands in each
and every causal relationship to have ever come to pass, it is one of the most fundamental of
conventional truths. Compare this to the infinite regress described at VV 31, where it is the case
that every pramāṇa must be 100% established by a previous one, ad infinitum.
This point is elaborated upon in VV 32, where Nāgārjuna specifies that the nub of the
problem with an infinite regress of this type is that ‘neither the beginning, nor the middle, nor
the end can be established’. Beginnings, middles and ends are important in Buddhism. They are
especially important in Madhyamaka, which is, after all, claimed to be the ‘middle way’ between
the extremes of absolutism and annihilationism. Without an end, for example, liberation would
be impossible – how might we end suffering if beginnings, middles and ends are not to be found?
In our immediate context, if a beginning, middle and end of an entity such as a pramāṇa cannot
be determined, then we run the risk of characterising it as eternal: how can an entity that has no
P a g e | 129
determinable beginning, middle or end possibly cause or be caused?104 If the pramāṇas have an
infinite regress to other pramāṇas (one instrument establishes another instrument, which
establishes another instrument, which… etc.), then they must have a start point and an end point,
lest any notion of causation be rendered incoherent. It seems that there needs to be an infinite
trail of pramāṇas in order to validate even one single object of true cognition (prameya). Further,
at which point would the infinite trail of pramāṇas terminate in order make the step to establish
the object? It is difficult on these criteria to account for a process by which an infinite series of
pramāṇas serving only to justify other pramāṇas then breaks from itself to establish not another
pramāṇa, but a prameya instead. The upshot here is that if the pramāṇas establish other
pramāṇas, then it seems that this is all that they could do. We would end up, for example, with
the establishment of means of knowledge (a pramāṇa), but not of the thing known (a prameya).
There is also an extra issue for the Naiyāyika if they wish to claim that the pramāṇas
somehow possess svabhāva (and we have seen that they do indeed want to make this claim) and
yet still have this reliance on other pramāṇas. Discounting for a moment Nāgārjuna’s argument
against infinite regressions, if pramāṇas self-establish in virtue of their svabhāva (recall from VV
9 that the objectors do indeed think that existent objects possess svabhāva), they cannot possibly
establish other objects (namely the prameyas) that also presumably possess a svabhāva.
Nāgārjuna argues at MMK 15.1-2:
na saṃbhavaḥ svabhāvasya yuktaḥ pratyayahetubhiḥ /
hetupratyayasaṃbhūtaḥ svabhāvaḥ kṛtako bhavet //1//
If the pramāṇas are established in virtue of their svabhāva, then they are – when svabhāva is understood in
ontological terms – necessarily eternal.
104
P a g e | 130
svabhāvaḥ kṛtako nāma bhaviṣyati punaḥ katham /
aḳrtrimaḥ svabhāva hi nirakpekṣaḥ paratra ca //2//
[An] intrinsic nature dependent upon causes and conditions is not
possible /
[An] intrinsic nature [that is] produced from causes and conditions
would be a product //1//
But how [can there] exist an intrinsic nature [that is] produced? /
For an unproduced intrinsic nature is independent of anything else
//2//
Taken in context, the result would be that if each pramāṇa has an intrinsic nature, and if
each prameya has an intrinsic nature, each pramāṇa could not possibly establish each (or indeed
any) prameya. The reasoning is very simple: for one thing to establish another is for one thing to
cause another. For Nāgārjuna, such causation can only be accounted for in conventional terms,
viz. in virtue of emptiness (śūnyatā), and so an intrinsically existent entity causing another
intrinsically existent entity is a non-starter. The very definition of being intrinsically existent (and
so possessing svabhāva) is – according to Nāgārjuna – to be uncaused by anything else. It is for
this reason that emptiness takes primacy in his philosophy, for it is only because of the emptiness
of all entities that any change can occur, and thus it is via emptiness that liberation is possible.
This is the central thesis of the MMK. For Nāgārjuna then, if both pramāṇas and prameyas have
svabhāva, then it is simply impossible for one to be established by the other. Indeed, it is
impossible for the pramāṇas to be established whatsoever:
teṣāmatha pramāṇairvinā prasiddhirvihīyate vādaḥ /
vaiṣamikatvaṃ tasminviśeṣahetuśca vaktavyaḥ //33//
atha manyase teṣām pramāṇānāṃ vinā pramāṇaiḥ prasiddiḥ
prameyāṇāṃ punararthānāṃ pramāṇaiḥ prasiddhiriti evam sati
yaste vādaḥ pramāṇaiḥ prasiddhirarthānāṃ itis a hīyate /
vaiṣamikatvaṃ ca bhavati keṣāṃcidarthānāṃ pramāṇaiḥ
prasiddhiḥ keṣāṃcinneti / viṣeśahetuśca vaktavyo yena hetunā
P a g e | 131
keṣāṃcidarhtānāṃ pramāṇaih prasiddhiḥ keṣāṃcinneti / sa ca
nopadiṣṭaḥ / tasmādiyamapi kalpanā nopapanneti /
[OR:] atrāha / pramāṇānyeva svātmānaṃ parātmānaṃ ca
prasādhayanti / yathoktaṃ /
dyotayati svātmānaṃ yathā hutāśastathā parātmānaṃ /
svāparātmānāvevaṃ prasādhayanti pramāṇānīti //
[NR:] yathāgniḥ svātmānaṃ parātmānaṃ ca prakāśayati tathaiva
pramāṇāni prasādhayanti svātmānaṃ parātmānaṃ ceti /
atrocyate /
[If you think that] these [pramāṇas] are established without the
pramāṇas, [then] your doctrine is abandoned /
[There is an] inequality, and you should state the special reason
[for this inequality] //33//
Now, [if] you think that those pramāṇas are established without
[recourse to other] pramāṇas, [but that] the objects of true
cognition are, however, established through the pramāṇas, [then]
your doctrine that [all] objects are established through the
pramāṇas is abandoned. There is an inequality: some objects are
established through the pramāṇas, and some are not. The reason
for this inequality – why some objects are established by the
pramāṇas and others are not – should be stated, but it is not
specified. Thus, this hypothesis too is not justified.
[Opponent’s reply:]
The pramāṇas prove themselves as well as others. As it is said:
‘As fire illuminates itself and other objects, so too do the
pramāṇas prove themselves and others.’
[Nāgārjuna’s reply:]
As fire illuminates itself and others, so do the pramāṇas prove
themselves and other objects. [To this] we say…105
Nāgārjuna leaves open the last sentence – he goes on in subsequent verses (I do not intend to address the
ensuing verses in detail here) to attack the idea that fire is self-illuminating and argues that if fire illuminates itself,
it must also be said to consume itself; an absurd position to hold.
105
P a g e | 132
Having established that the Naiyāyikas cannot feasibly hold that the pramāṇas are
established by other pramāṇas, Nāgārjuna now pits this conclusion against the Naiyāyika
assumption that the pramāṇas establish the objects of cognition (prameyas). Thus at VV 33, he
asks how it can be the case that the pramāṇas establish prameyas but cannot establish
themselves:
Now, [if] you think that those pramāṇas are established without
[recourse to other] pramāṇas [but that] the objects of true
cognition, are, however, established through the pramāṇas, [then]
your doctrine that [all] objects are established through the
pramāṇas is abandoned.
This argument is relatively straightforward. If the pramāṇas are not established via other
pramāṇas, then the thesis that everything knowable is established via the pramāṇas fails. This
is because the pramāṇas themselves are apparently not established via themselves or by other
things of the same type. Why is this a problem? Simply because the pramāṇas are supposed to
be known to us. The idea is that all psycho-physical objects can be known via the pramāṇas, but
according to Nāgārjuna’s refutations laid out here, the pramāṇas themselves cannot be known –
an obvious pitfall. We need to know what our means of valid knowledge are before we know if
they have been deployed in such a way that we have gained valid knowledge. The idea of valid
knowledge is useless to us unless we know that there are definite ways by which we can acquire
it. The second problem to be explained is, according to Nāgārjuna, that the Naiyāyika apparently
holds the position that the pramāṇas establish the prameyas, but the pramāṇas themselves
remain unestablished. Thus,
[t]here is an inequality: some objects are established through the
pramāṇas, and some are not. The reason for this inequality – why
P a g e | 133
some objects are established by the pramāṇas and others are not –
should be stated, but it is not specified. Thus, this hypothesis too is
not justified.
The Naiyāyikas do not specify why this seemingly arbitrary distinction is made – why is it
that some things are established by the pramāṇas and other things are not? The response from
the Nyāya adherent is that ‘[a]s fire illuminates itself and other objects, so too do the pramāṇas
prove themselves and others.’ This is, to Nāgārjuna, simply absurd: he holds that if fire
illuminates itself, it must also be said to consume itself. Given that this analogy is absurd, so too
is the idea that pramāṇas establish themselves and other objects in the same way. For
Nāgārjuna, much as fire cannot illuminate itself unless it also consumes and destroys itself, the
pramāṇas cannot establish themselves without falling into the same trap of absurdity.
So what is left of the pramāṇas? It would, I think, be to place the cart before the horse
to argue that because Nāgārjuna thinks that there are no established pramāṇas he also thinks
that there is no knowledge. If this were the case, his own Buddhist project would be fatally
undermined. 106 If we cannot know anything, then the means to liberation is compromised.
Further, even knowing that there can be any liberation would be on this account, it seems,
impossible. We could not know about pratītyasamutpāda, we could not know about the Four
Noble Truths, we could not know that the Buddha’s methods work, and so on for all aspects of
Buddhist praxis. We would not even be able to know about dissatisfaction (duḥkha) broadly
construed, and so would not know that we are dissatisfied! It should go without saying that
Nāgārjuna does not want to completely undermine the Buddhist path – he is a Buddhist, and he
106
There are those that think that it is! Burton (2001) is one such example.
P a g e | 134
thinks that the Buddha’s methods for liberation are efficacious. Clearly, then, Nāgārjuna
acknowledges that we can know things about the world.
The issue is clearly one of
’establishment’, which is to say it is a problem of proving intrinsic existence. Pramāṇas establish
the existence of prameyas and it is by using the framework that results from this interaction that
Naiyāyikas build their definitions of correct knowledge. It is important to bear in mind that these
endeavours are not fundamentally concerned with explaining the world as it is (as is for example,
the modern scientific/empirical method), but rather with explaining how we might reach
liberation. The same is true for Nāgārjuna. It is soteriological aims which underpin all of this
wrangling, not a scientific understanding of the world. To put it in starker terms, Nāgārjuna is
concerned with a prescriptive account of liberation, not with a descriptive account of reality.
Instead, I propose a more nuanced reading – one that is in principle acceptable to Śaṅkara
when it comes to knowledge of the world around us. Nāgārjuna clearly thinks that the pramāṇas
cannot exist ultimately, and I think that for the Mādhyamika, a knock-on effect of this stance is
that they cannot give an account of the ultimate, either. We are faced then with a situation in
which the pramāṇas offer – at best – a conventional explanation of the world around us. That is
to say that whilst the pramāṇas might be useful tools in navigating the conventional world, they
offer no input as regards the status of the ultimate. It seems to me that Śaṅkara must also
broadly subscribe to this view, for although he allows recourse to pramāṇas in order to facilitate
liberating insight, he also holds that ultimately, such means of knowledge are formulaic, largely
reliant on language and consequently ineffective. It seems to me that for Śaṅkara, the pramāṇas
are conventional insofar as they allow a practitioner to discern truth within the world of māyā.
The end goal is to transcend the world of māyā. Although liberation does not consist in physically
P a g e | 135
moving beyond the world of illusion, it does require an intuitive gnosis of the Absolute ātmanBrahman, and in the end, it is only this gnosis that matters: all attempts to capture it, describe it
and so on are, as we have seen, doomed to failure. As such, even in a situation where there is
belief in an Absolute and the end goal is some gnosis of it, the pramāṇas must be ultimately
inadequate in providing any definitive knowledge of it. The pramāṇas can orientate, but they
cannot strictly be used as a yardstick against which the ultimate is to be measured. This would,
I suggest, need to be Śaṅkara’s final position.
It would be tempting to argue that this is also necessarily true of Nāgārjuna’s position.
Indeed, I suspect that Bhattacharya would argue this, and I am convinced that a thinker like Murti
would, too.107 It cannot, however, be the case that Nāgārjuna holds an identical view to Śaṅkara
on this matter. It might be the case that Nāgārjuna could in principle assign some value to the
pramāṇas in much the same way that Śaṅkara does.108 That is to say that Nāgārjuna could in
principle view the pramāṇas as useful conventional designations that provide practitioners with
a yardstick by which to measure what is conventionally true. As a Buddhist, he would of course
have recourse to fewer pramāṇas than Śaṅkara (Buddhists generally allow only two (sometimes
three, as we saw earlier): perception and inference), but these two can in principle be useful in
navigating the world as it is experienced conventionally. As is the case for Śaṅkara, Nāgārjuna
R. Ninian Smart pulls no punches when he writes that ‘Murti, who disliked having foreign and homegrown
untouchables studying works flowing from a sacred revelation (he was a Brahmin, of course), thought that his
version of Madhyamaka was a "poor man's Advaita”’ (McCagney, 1997: xi). This attitude manifests in Murti’s
writing and – owing to his citing of Murti’s ideas – risks being represented in the writing of Bhattacharya.
107
108
It at least seems to me that Śaṅkara should view the pramāṇas in this way.
P a g e | 136
too would need to jettison the notion that the pramāṇas can provide some sort of ultimate
knowledge. We have seen in the discussion of the VV above that Nāgārjuna’s real issue with the
pramāṇas is twofold: first, he argues against the idea that the pramāṇas are or can be ‘ultimately’
established. That is simply to say that he does not think that the pramāṇas could have svabhāva.
With this being the case, they are not in a position to ‘establish’ that which exists with svabhāva
(for the Advaitin, this would be the ātman-Brahman). This is, of course, the second issue with
which Nāgārjuna grapples: the scope of the pramāṇas. How can that which is not ‘established’
itself (i.e. that which does not have svabhāva) then ‘establish’ something else (with a svabhāva)?
In other words, how might an unestablished pramāṇa establish a prameya (that to be known;
the ātman-Brahman)? Interestingly, the issue of scope also presents itself for the Advaitin, albeit
in a different way than it presents for the Mādhyamika. I shall explain how this is the case.
Although the Advaitin must concede the inadequacy of the pramāṇas in capturing the
ultimate character of the ātman-Brahman, they do still have a use in terms of the ultimate: they
confirm that there is an ultimate/Absolute. Indeed, we have seen at the beginning of this
subsection that for Śaṅkara, it is the authority of scripture that proves the existence of the
Brahman. These same scriptures then use analogy to build on the assertion of Brahman’s
existence in order that we might orientate ourselves towards it. The difference between the role
of pramāṇas for the Advaitin and the role of pramāṇas for the Mādhyamika is then that for the
Advaitin, even if they do not (indeed they cannot) provide any true account of the ultimate’s real
nature (which is ineffable), they do at least establish something as ultimately immanent. For the
Mādhyamika, the pramāṇas as accounted for in the literature of the Advaitins – much like that
of the Naiyāyikas – cannot account for anything in an ultimate sense. Nāgārjuna thus rejects the
P a g e | 137
pramāṇas so conceived by his opponents, though I do not think that he would strictly need to
disregard them in a conventional sense if it was accepted that they have no svabhāva.109 In such
a scenario, the pramāṇas might provide some expedient means by which we can navigate the
world of convention. But this is all that the pramāṇas could do, and this is a very limited
application. If the pramāṇas do not account for anything in an ultimate sense, then why bother
with them at all? They are, after all, designed to protect the practitioner against error and
mistaken positions by providing a metric through which they can ascertain ‘true’ statements and
beliefs about the world.
McCagney writes that whilst Nāgārjuna and his followers did not explicitly answer
questions regarding how the mind objectifies concepts or about the nature of error, neither did
they simply disregard them (1997: 42). The point for Nāgārjuna is, I am at pains to repeat, not to
give an account of the world, but to give an account of the means to liberation. McCagney writes
about the nature of error that ‘such questions were dismissed for philosophical and soteriological
reasons. The Madhyamaka effort concerned the radical purification of mental fictions rather
than an explanation of error’ (1997: 42). Consequently, I think that we can assume that
Nāgārjuna’s reluctance to speak about pramāṇas outside of his argument with the Naiyāyikas
stems from his seeing both the pramāṇas and further discussion of them as a waste of time.
After all, the Madhyamaka project is not to system build, but to tear down systems
This would stand in contrast to Buddhists like Dharmakīrti who have no qualms about integrating svabhāva into
theoretical frameworks. For Dharmakīrti, svabhāva and ‘essential connections’ are crucial to his account of
inference. Of course, Dharmakīrti also argues for the existence of ultimately real dharmas, something that
Nāgārjuna vociferously disputes.
109
P a g e | 138
systematically! Nāgārjuna takes seriously prapañca (conceptual proliferation), and it seems to
me that he would likely consider the pramāṇas to be an instance of conceptual proliferation; a
pointless reification that serves no real soteriological purpose.
With what, then, are we left? Nāgārjuna does not view the pramāṇas as useful expedients
by which to remove error, false views and so on. Instead, they are at worst a manifestation of
those false views (especially if as the Naiyāyikas claim, they have svabhāva) and do more harm
than good. At best, they are needless reifications that do no real work. Even if they could tell us
something about the conventional world, this is for Nāgārjuna inadequate, for we have already
noted that he is not at all interested in giving an account of the world. Instead, Nāgārjuna thinks
that we can simply undercut the work allegedly done by the pramāṇas. As McCagney puts it,
‘[t]he point, for the Mādhyamikas, is not to understand the source of error, but to eliminate it’
(1997: 43), and Nāgārjuna’s preferred means of achieving this is not by talking about concepts
and phenomena and judging them according to pramāṇas, but by simply halting the reification
and proliferation of concepts to begin with. Obviously, this is a radical, fundamental aim. The
pramāṇas cannot exist with svabhāva and as such are as transient and reified as the next concept.
Treating them as some peculiarly established thing by which we can attain an established
‘knowledge’ is a problem – put simply, nothing is established because nothing exists
independently of another (viz. nothing exists with svabhāva).
McCagney discusses the ‘establishment’ of entities at some length. She sums up the
Nāgārjunian notion thus (1997: 60):
[N]o event [or entity] can exist independently of another since it
would depend on there being another from which it could differ. If
P a g e | 139
the event [or entity] did have independent existence (svabhāva), it
could not exist in relation (dependent on conditions) and so it could
not exist at all. And if svabhāva (self-nature) causes another event
[or entity] to arise, that event [or entity] would be "other-nature"
and the original condition would not exist independently or
essentially.
All of this metaphysical deconstruction relies on śūnyatā – the emptiness of svabhāva –
of all entities. McCagney instead translates śūnyatā as ‘openness’, which is probably not as much
of a stretch as it first seems. If we think of a vast, open space, we might say that it is ‘empty’.
McCagney repeatedly forces the point that in various places in the Aṣṭasāhasrikāprajñāpāramitā,
śūnyatā is used in conjunction with metaphors of space (ākāśa), writing that ‘[t]he symbols in
early Prajñāpāramitā texts show that the Mahayana notion of ākāśa derives from meditation
(dhyāna) on the sky, which is experienced as vast, luminous and without boundaries’ (1997: XX).
This would mean that Nāgārjuna’s equation of śūnyatā and pratītyasamutpāda amounts to an
endorsement of ‘the openness or limitlessness of events’ (1997: XX), which does not in turn imply
anything permanent (nitya), unmade (akriyate), independent (nirapekṣaḥ paratra), or without
causes and conditions (ahetu pratyaya). In other words, Nāgārjuna is equating the openness or
limitlessness of events with a lack of svabhāva, which is indeed permanent, unmade,
independent, and without causes and conditions.
I want to insist here that Nāgārjuna (and so the Madhyamaka in general) repeatedly
denies that anything can be substantial, permanent, uncaused and independent. It should then
be obvious that this includes any substantial entity such as the ātman-Brahman. Even when the
diligent Mādhyamika moves beyond such dualisms as existence and nonexistence – asserting
neither for fear of inadvertently attaching oneself to a view (dṛṣṭi), the question, such as it is,
P a g e | 140
remains open-ended. Nāgārjuna does not appear to be denying that things appear to exist, only
that things exist ultimately; permanently, uncaused, self-sufficiently, with svabhāva. They do not
exist in the manner in which we uncritically take them to exist. As nonexistence is defined in
opposition to existence, he does not claim that any entity does not exist, either. Buddhapālita
(translated by Paul Williams) writes that
(57) Therefore the meaning of dependent designation is precisely
that an entity which is dependently designated cannot be said to be
existent or nonexistent because it is completely empty of intrinsic
nature. [But] there is no fault in a conventional statement (tha
snyad kyi tshig, probably vyavahāra-vacana or -vākya).
He continues further, clarifying why this is the case:
(58) How is it logically possible to say that the Tathāgata, who is
dependently designated, either exists or does not exist? For if a
Tathāgata existed, he would just exist, even without an
appropriation, but he does not exist without an appropriation. How
can one who does not exist without an appropriation be said to
exist? How, too, can a Tathāgata who is dependently designated be
said not to exist? For a nonexisting uḍumbara flower cannot be
designated.
(Williams, 2005: 33-34)
The idea here is that the Madhyamaka position that anything that does not exist with
svabhāva is ‘dependently designated’. It is because the Mādhyamika rejects the claim that
anything exists with svabhāva that they must also stop short of asserting existence or
nonexistence. For the Mādhyamika, existence and nonexistence are similarly problematic
trappings: real ‘existence’ (with svabhāva) denotes entities that are immutable and ahetu
pratyaya; real ‘nonexistence’ (without svabhāva) would thus denote entities that did not appear
in the first place. This is very clearly not what Nāgārjuna and his followers think is going on in the
P a g e | 141
world. If existence by/with svabhāva obtained in the world, then every existent entity would be
ultimately grounded in virtue of its ultimately existent intrinsic nature (svabhāva). Any existent
entity must have svabhāva (and so must necessarily exist). It cannot do otherwise. This means
it cannot come into existence, change, or go out of existence (because for Nāgārjuna, to have
svabhāva is to be permanently, necessarily existent). The significance of this understanding of
the world cannot be overstated. According to that same svabhāva doctrine, then, it is impossible
for there to be an entity without svabhāva because all existent entities necessarily have
svabhāva. 110 To speak of a nonexistent entity is then nonsensical. Hence Buddhapālita’s
insistence that the Mādhyamika assert neither existence nor nonexistence: neither is
appropriate, both result in states of affairs that are demonstrably false. Instead, Buddhapālita
follows Nāgārjuna in claiming that entities cannot exist by svabhāva, but can and do exist
conventionally as designations.111
The point of refusing to subscribe to a dṛṣṭi that asserts either existence or nonexistence
is purely soteriological. As we have already seen, the Mādhyamika method aims not to give an
account of the world, but instead to expedite our liberation from duḥkha (dissatisfaction). To
assert a monism in an entity akin to the ātman-Brahman is for Nāgārjuna a manifestation of
avidyā (ignorance; mischaracterisation) in much the same way that asserting the ultimacy of a
This might seem a peculiar way to use a phrase such as ‘existent entity’. The idea is that an entity with
svabhāva is basic, like the dharmas with which Nāgārjuna is so concerned. For such an entity to be ‘really
existent’, it must exist with svabhāva. If it does not, then it is derivatively existent, not ultimate and ‘not really
existent’.
110
111
Even ‘exists’ is here used only conventionally, viz. without any substance (svabhāva) underpinning it.
P a g e | 142
pramāṇa or a dharma is avidyā. It is to reify a concept under which nothing really falls and assign
it a permanence that is according to Nāgārjuna demonstrably impossible. This stands in stark
opposition to Śaṅkara and subsequent Advaitins, for whom – as we have seen – the permanent
(nitya) uncaused (ahetu) ātman-Brahman is the solution to avidyā (ignorance; misconception)
and all that avidyā entails.
It should now be clear why Nāgārjuna jettisons the very notion of pramāṇas after
attacking them in the VV: pramāṇas are nothing but instantiations of prapañca, and such
instantiations ought to be removed at the root. Each pramāṇa is a dṛṣṭi: it is a view regarding
what can or cannot be considered a means to valid knowledge and in turn establishes the
prameya (thing to be known). They – as we have seen – usually apply to metaphysical concerns
such as the nature of Reality (knowledge of the ātman-Brahman). Part of Nāgārjuna’s issue with
pramāṇas then is that he does not think that they can actually do anything given that svabhāva
cannot exist – this has been demonstrated above. The other part of the issue is that Nāgārjuna
does not think that there is an ultimate reality to be known via these means of knowledge. How
can an epistemic instrument provide a means to knowledge about an entity that cannot be
established as existent? We cannot know about what is not there to be known! By extension,
the ātman-Brahman is not established primarily because Nāgārjuna thinks that all phenomena
are ‘selfless’ (lacking svabhāva) and insubstantial (nairātmya). If the ātman-Brahman were
existent in the way that Advaitins require it to be, it would exist with svabhāva and so have
substantiality: indeed, it could be the only substantial entity and the only svabhāva.
P a g e | 143
§4: Advaita and Buddhism
In this section, I will give a brief account of the relationship between Advaita Vedānta and
Buddhism, illustrating some doctrinal developments from Gauḍapāda and Śaṅkara and their
relation to Buddhist doctrines. From there, I will attempt to demarcate how Śaṅkara viewed
contemporary Buddhist schools, and then detail some important ways in which both the
Madhyamaka and Yogācāra schools differ from Śaṅkaran Advaita in relation to seeking liberation
within a substantial Absolute. Given that out of the Madhyamaka and Yogācāra schools it is
generally thought that it is the Yogācāra that veers most closely to an Advaitin form of Buddhism,
this section will primarily be concerned with Vasubandhu’s conception of Yogācāra. My intention
in focussing on Vasubandhu is primarily to demonstrate that despite claims from scholars such
as Bhattacharya and Murti, absolutism need not be a necessary consequence of Yogācāra
philosophy.
First, to the Upaniṣads. Frank Whaling (1979: 21) attests that taking the Upaniṣads
literally when they make positive attributions to the ātman-Brahman was traditionally the norm:
Hindu predecessors to Śaṅkara’s forerunner, Gauḍapāda, had indeed ‘taken the Upaniṣadic
creation texts literally’, and though they taught the sole reality of Brahman, they had also
‘allowed the possibility of modifications in Brahman, and even of parts to Brahman’ (1979: 21).
It was Gauḍapāda that rejected the apparent dualism of unity and diversity within Brahman and
‘insisted that duality was unreal and that advaita [non-duality] was ultimate’ (Whaling, 1979: 21),
and so it was really with Gauḍapāda that the big doctrinal switch was made in Advaita.
P a g e | 144
Gauḍapāda, writes Whaling, was then the first Advaitin to systematically reform Vedānta:
he rejected the traditional view of ritual as a means to liberation; stressed ‘the identity of the jīva
and Brahman, and insisted that their difference was only apparent’; rejected the early Vedas as
authoritative, and rejected the contemporary Vedāntin view that ‘there was a genuine
transformation of Brahman into the world and the world into Brahman’ (1979: 22). Gauḍapāda
was undoubtedly a reformer.
Whilst Whaling concedes that it is at least possible that such a Vedāntin revolution could
have been entirely internal, he concludes that on the balance of probabilities, it is much more
likely that Gauḍapāda was either directly or indirectly influenced by Buddhism than it is that a
newly-reformed Vedānta influenced the comparably strong and well-established Mahāyāna
schools (Whaling, 1979: 22-23). Importantly, Whaling argues that this ‘does not mean that
[Gauḍapāda] has become a Buddhist or that he has forsaken Vedānta’, but only that he has
‘merely reinterpreted the message of the Upaniṣads in the language and thought forms of his
day, and in so doing, he paved the way for Śaṅkara to carry on his work in a more systematic way’
(1979: 21). It was with Gauḍapāda that Advaita began to take its form most recognisable to us
today, and it is from Gauḍapāda, says Whaling, that Śaṅkara inherited the means by which he
would expound Advaita in contrast to Buddhism.
It is in consideration of contact between the newly-reformed Advaita and the established
Buddhist schools that Whaling (1979: 5) asks ‘[d]id Śaṅkara understand Buddhism?’ This is of
significance to this work, because Śaṅkara’s understanding of Buddhism (or, potentially, lack
thereof) is of some significance when attempting to eke out points of conflict between Buddhism
P a g e | 145
and Advaita. Furthermore, determining – as best we can, at any rate – what Śaṅkara actually
thought about his Buddhist cotemporaries is perhaps the best way of establishing if Śaṅkara saw
for himself the similarities that Bhattacharya apparently does between his doctrine and those of
the Buddhists. This is especially relevant if – as I am inclined to think – Buddhist schools
contemporary to Śaṅkara (such as Madhyamaka) do not actually share or endorse his type of
absolutism.
Obviously, this question is different to the one initially posed by Bhattacharya, which
effectively asks if the Buddha is an Advaitin,112 but I hope that the pertinence that the question
of Śaṅkara’s relationship to Buddhism has to the wider question asked by Bhattacharya is so
obvious as to negate the need for any further justification. In examining the relationship between
Advaita and the Upaniṣads, I hope to glean some sort of understanding relating to Bhattacharya’s
own obvious predilection for an Advaitic reading of the Upaniṣads and the means by which this
might affect his view of both precanonical and canonical Buddhism.
It might be thought that if the Buddha had effectively been espousing the same doctrine
as Śaṅkara, then Śaṅkara would himself would have recognised this and been aware of portions
of – if not most of – the contemporary Buddhist corpus. Whether or not he would admit this is
a different matter – it is possible that such an acknowledgement by Śaṅkara was prevented owing
112
Ostensibly, Bhattacharya presents a thesis that the Buddha re-hashes pre-existing Upaniṣadic ideas in a novel
way, viz. that he placed less emphasis on the philosophy, and more emphasis on the path. The implication seems
to be that this novelty is the reason why Buddhism survived as a path distinct from the Advaita that it somehow
duplicates. Nevertheless, Bhattacharya thinks that the Buddha’s path affirms in every significant way the existence
of the Upaniṣadic ātman-Brahman (2015: 209-210), and further, that this advocates the same sort of liberation as
popularised by Śaṅkara. If true, this would make the Buddha – at least in principle – an Advaitin of the same sort
as Śaṅkara. (Technically, I suppose, it would make Śaṅkara an Advaitin of the same sort as the Buddha!)
P a g e | 146
to sectarian reasons. I suspect that for Bhattacharya, this issue might be no more than peripheral,
for if, as he argues in his book, the Buddha simply laid a soteriological path based upon Upaniṣadic
doctrine, Śaṅkara’s doing the same centuries later is simply testament to the veracity and validity
of those very same Upaniṣads. There is nothing strange in and of itself in the idea that two people
restated the same ideas at two separate points in time, especially if, as argued, their starting
points were the same. Thus, the real point of interest is in the Buddha’s alleged adherence to
Upaniṣadic principles. How we understand what these Upaniṣadic principles actually are,
however, relies on numerous factors, not least of which are the significantly different
interpretations of the same scriptures by different Hindu schools, none of which I have the space
or time to cover in any great detail. Additionally, it is nevertheless a point of historical fact that
Buddhism developed as a tradition outside of and not cognate with, Hindu thought in general. If
the Buddha really did expound the same doctrine as both the Upaniṣads and Śaṅkara, then we
might expect there to be an explanation for this quirk of history, and it would not be
unreasonable to expect this reason to be already well established.
On balance, then,
investigating the links between Śaṅkaran Advaita and its contemporary Buddhist schools might
prove fruitful in determining how best to read both the Upaniṣadic conception of ātmanBrahman and its relation to the Buddhism of the Nikāyas and the Mahāyāna.
On the question of Śaṅkara’s understanding of Buddhism, Whaling has a lot to say. First,
he argues that Śaṅkara himself recognised – and thus, presumably engaged with – three Buddhist
schools: the Sarvāstivādins, Mādhyamikas, and the Yogācārins. We immediately notice that
Śaṅkara does not appear to engage with the ‘precanonical’ Buddhism with which Lindtner,
Bhattacharya and Albahari are all concerned. This is at least partly understandable given that it
P a g e | 147
was not the Theravāda that was present in India during Śaṅkara’s lifetime, but the Mahāyāna,
and so, even with its waning influence, it is natural that it is the Mahāyāna that bore the full
extent of Śaṅkara’s criticism. In narrower terms relevant to the current project, it is worth
recalling that Bhattacharya claims that it is the Mahāyāna schools of Buddhism ‘which put things
right’ in terms of how ātman is treated doctrinally (2015: 39), and so Śaṅkara’s focus on these
three Mahāyāna schools is for these reasons unproblematic. His attitudes to these schools, in
fact, should allow us to gain insight into Śaṅkara’s wider views relating to Buddhist doctrine, and
thus allow us to begin to assess the extent to which he himself – and consequently, his system of
Advaita – was influenced by Buddhism. This should in turn allow us to make some sort of
judgement regarding the precise extent to which Advaita and the Mahāyāna accord; a crucial
point if there is any weight to Bhattacharya’s assertion that it was the Mahāyāna ‘which put
things right’ in terms of reverting back to the Upaniṣadic conception of ātman-Brahman (2015:
39).
Whaling (1979: 4) argues that as Śaṅkara’s arguments against Sarvāstivādin doctrine
coincide with those of Bhāskara, we can probably adduce that these criticisms are not original to
Śaṅkara, but are both drawn ‘from the same common source’, viz. these arguments are the stock
Vedāntin refutations against the Sarvāstivādins.
The generic Vedāntin response to the
Sarvāstivādin dharma theory is that if ‘basic elements did exist and act independently, then there
[is] no reason why they should ever cease to do so, thus the conditions for cessation of activity
in the sense of nirvāṇa would be jeopardised’ (1979: 4). The idea here being that if there were
independently existent atomic particles making up the entirety of what we perceive as reality,
then there could be no hope of liberation. Why is this? Well, because if reality really does consist
P a g e | 148
of independent elements that express their reality in spite of the other elements around them,
there is no reason to suppose that we can affect these independent elements and stop their
expression in order to bring around liberation. In other words, independently existent particles
of this sort would necessarily be such that we could not affect their existence (because they are
independently existent and so are unaffected by the operations of existent particles or forces
outside of themselves). This takes a particularly problematic turn when we think of defilements
and negative traits. If suffering is the manifestation of independently-existent atomic particles
(as it necessarily has to be if the whole of reality consists of these particles), then there is no
reason to suppose that we can ever stop these particles from expressing their realities (viz. we
could not stop the relevant dharmas from expressing defilements and so propagating suffering).
Very simply, we could not bring about the cessation of activity that supposedly characterises
nirvāṇa. For the Advaitin, this particular problem is conveniently bypassed: reality consists of
ātman-Brahman and the empirical world is no more than an ultimately unreal expression of
Brahman: only Brahman is real, and everything in the empirical world derives a sort of
conventional, contingent reality from the ultimate reality of Brahman (Rambachan, 2006: 77).
Interestingly, the Vedāntin argument against such essentially existent atomic elements is
very similar to that offered by Nāgārjuna in the Mūlamadhyamakakārikā and by Vasubandhu in
the Abhidharmakośabhāṣya. Against Sarvāstivādin dharma theory, Nāgārjuna claimed that
nirvāṇa could not be existent without being characterised by old age and death, which would in
P a g e | 149
turn mean that nirvāṇa, is conditioned rather than ultimate.113 In a similar vein to the Advaitins,
Nāgārjuna argues throughout the Mūlamadhyamakakārikā that if the basic elements of empirical
reality that combine to constitute all psycho-physical phenomena (the Sarvāstivādin dharmas)
were independently existent, immutable and permanent, there could be no hope for liberation.
Why not? For precisely the same reason put forth by Śaṅkara: their immutability means that
they will never stop being the way they are; nothing we could possibly do would ever affect such
independent, self-sufficient entities. This in turn means that the change in conditions required
for liberation can never (and could never in the future) be achieved. This, think both Śaṅkara and
Nāgārjuna, negates the possibility of independently existent, permanent elements existing in a
world where liberation is possible. As liberation is the highest goal and – significantly – regarded
as an accurate scriptural inference (as well as being demonstrably possible), the existence of
permanent, independently existent atomic elements is rejected.
113
MMK 25.4-5:
bhāvas tāvan na nirvāṇaṃ jarāmaraṇalakṣaṇam /
prasajyetāsti bhāvo hi na jarāmaraṇaṃ vinā //4//
bhāvaś ca yadi nirvāṇaṃ nirvāṇaṃ saṃskṛtaṃ bhavet /
nāsaṃskṛto hi vidyate bhāvaḥ kva cana kaś cana //5//
Nirvāṇa is not an existent; [if it were, its having] the characteristics of decay and death would
follow /
No existent is without [the characteristics of] decay and death //4//
And if nirvāṇa were an existent, nirvāṇa would be conditioned /
An unconditioned existent is never found anywhere //5//
P a g e | 150
To elaborate, for the Madhyamaka Buddhists as well as the Advaitins, anything that is
empirically existent is not – and what is more, cannot be – ultimately existent.114 On the face of
things, though, this looks problematic for Bhattacharya’s claim that the Buddhist nirvāṇa is
identical with the Upaniṣadic ātman-Brahman. Śaṅkara (and, we may assume, by extension,
Bhattacharya) characterise the Brahman as being – in the words of Rambachan – ‘timeless and
present in all states and mental conditions’ (2006: 21). There looks to be a degree of tension
here if the Brahman is timeless and ever-present (eternal and thus unconditioned) but the
nirvāṇa with which this Brahman is supposed to be identical is nonexistent (read: not ultimately
existent) and thus conditioned (impermanent, unreal). I do not think it controversial to postulate
that Bhattacharya would respond that from the ultimate perspective, there is only an apparent
tension and that Nāgārjuna is advocating the transcendence of descriptions like ‘existent’,
‘nonexistent’, ‘conditioned’ and ‘unconditioned’. On this I concede that he would be – in part, at
114
Hugh Nicholson points out that this sort of distinction between conditioned and unconditioned was a distinctly
Śaṅkaran progression in Advaita, writing that Śaṅkara ‘introduced a hitherto unknown distinction between the
conditioned and unconditioned forms of brahman’ that was ‘perhaps borrow[ed] from some version of the
Buddhist Two Truths doctrine’ (Nicholson, 2007: 531). We find superficial parallels regarding the conditioned and
unconditioned in Nāgārjuna’s Mūlamadhyamakakārikā. Some examples can be found at 1.10, 5.2 and 7.16 where
Nāgārjuna argues that whatever exists in dependence (viz. is conditioned) is necessarily without an intrinsic nature,
and whatever entity is without an intrinsic nature or defining characteristic (svabhāva) is unreal (not ultimately
real). At 7.33, Nāgārjuna goes one further and shows that this is only part of the story: the unconditioned cannot
be ultimately real, either. This is why the similarities are only superficial: Nāgārjuna thinks that because we cannot
ultimately explain how conditioned things exist, then we cannot ultimately explain how unconditioned things exist,
either. For this reason, both ‘conditioned’ and ‘unconditioned’ entities and states are ultimately unreal, and
Madhyamaka philosophy undergoes a divergence from, rather than a convergence with, the sort of liberation
attested by Śaṅkaran Advaita. For this reason, it cannot be claimed that the Mahāyānists ‘put things right’ in
relation to the Upaniṣadic ātman-Brahman. Very simply, the troubles associated with clinging to any sort of
permanent entity whilst simultaneously seeking liberation should preclude the Mādhyamika from positing any
entity of a similar sort to the ātman-Brahman. For the Advaitin, even though the Brahman is ultimately nondual
and beyond conception, it is at least Real (indeed, it is the only Reality). Nāgārjuna would surely reject any such
assertion as an example of reification.
P a g e | 151
least – correct. What, then, might be said about nirvāṇa from a Madhyamaka viewpoint? Siderits
and Katsura read Nāgārjuna as wrestling with whether nirvāṇa might be an existent (viz. a
positive being, or bhāva), a nonexistent (viz. a negative being, abhāva), both, or neither (2013:
292). The conclusion reached at the end of the 25th chapter of the Mūlamadhyamakakārikā is
that nirvāṇa cannot possibly be any of these things: Nāgārjuna explicitly equates nirvāṇa to
saṃsāra at MMK 25.19 and then goes on to refute the notion that even nirvāṇa can exist
ultimately! At this point, it is sufficient to say that the support for his thesis that Bhattacharya
expects from Madhyamaka philosophy looks to be at risk from the very outset.115
Another interesting caveat that provides some illumination into Śaṅkara’s attitudes to
first, other doctrines, and second, matters of history is that he appears to conflate two
Sarvāstivādin schools and criticise them as one entity. As we know, there can be significant
doctrinal and practical differences between sects that are ostensibly subsumed under the same
school of thought: Śaṅkara himself has stark doctrinal differences from the other Vedāntin
schools! Nevertheless, in an attitude that Whaling opines is ‘typical of his indifference to history’
(1979: 6), Śaṅkara conflates the doctrines of the Sautrāntikas and the Vaibhāṣikas. I will spend a
little time illustrating how this is the case because I think that it sheds some light on two things:
Added to which, Śaṅkara seems to classify the Mādhyamikas as ‘nihilists’ (Śaṅkara, 2011: 401); a familiar charge
against the school that persists to this day. This is odd if, as Bhattacharya suggests when he cites K. Venkata
Ramanan, both the Mādhyamikas and Śaṅkara are in agreement regarding the ātman-Brahman:
115
…one can say that the one accepts or denies the ātman as much as the other;
both [the Mādhyamika and the Advaita Vedānta] deny ātman as a separate
substantial entity inhabiting the body of each individual, and both accept ātman
in the sense of the essential nature, the svarūpa or the svabhāva, of the
individual as well as of all things. (2015: 114, note 249)
P a g e | 152
first, on Śaṅkara’s apparent ignorance regarding the intricacies of Buddhist doctrine, and so his
ignorance of the Buddhist path to liberation. Second, on the diversity of opinion in early
Buddhism as to how liberation might be reached and by extension, as to what constituted an
accurate portrayal of the Buddha’s teachings. Both of these factors interrelate with my
investigation into Bhattacharya’s reading of both the Upaniṣads (and thus Advaita) and early
Buddhism. Given that Śaṅkara was at least partly concerned with refuting contemporary schools
of Buddhism (in what seems to have been a successful attempt to consolidate Advaita in India),
we might think ourselves justified in assuming that he was indeed aware of (at least some of) the
doctrinal differences between the rival schools that he was attacking. That Śaṅkara was at least
superficially aware of some differences (viz. that he was aware that there were fundamental
differences between the three schools he attacks) is made clear in Vedānta Sūtra 2.II.32 (Śaṅkara,
2011: 428). Here, he writes that the ‘Buddha by propounding the three mutually contradictory
systems. . . [has made it clear] that he was a man given to make incoherent assertions.’116 From
this I find it relatively unproblematic to suggest that Śaṅkara was indeed aware – at the very least
– of the basic differences in doctrine between different Buddhist schools. This in mind, I suggest
that he either really believed that the Buddha was the confused, single source of all of these
schools and their apparently divergent incompatible doctrines, or that he glossed over these
differences in a deliberate attempt to lump together – and so dismiss together – the various
Buddhist traditions.
An extended version of this quotation adds that if the Buddha is not deluded and confused, then a ‘hatred of all
beings induced him to propound absurd doctrines by accepting which they would become thoroughly confused’
(1890: 428). The ‘three mutually contradictory systems’ referred to are the respective schools of the Vaibhāṣikas,
Sautrāntikas, and Mādhyamikas.
116
P a g e | 153
Whaling briefly surveys the rest of Śaṅkara’s objections to Buddhism and outlines them
as follows: Śaṅkara argues against the twelve links (nidānas) of dependent origination
(pratītyasamutpāda) on the very simple basis that each nidāna is sufficient only to explain the
cause of the succeeding nidāna and not to explain the causal chain as a whole (Whaling, 1979: 45). There is rejection of the Buddhist doctrine of momentariness based on the idea that each
momentary particle’s action would stop prior to the rise of the next particle, thus leaving no link
between cause and effect (1979: 5). This leads to an attack on the Buddhist idea that
unconditioned (asaṃskṛta) dharmas are non-momentary and are thus eternal.117 Śaṅkara saw a
tension insofar as ‘there is a self-contradiction in the Bauddha statements regarding all the three
kinds of negative entities, it being said, on the one hand, that they are not positively definable,
and, on the other hand, that they are eternal’ (1890: 413). Whaling then gives a brief account of
Śaṅkara’s attack against the Buddhist argument from similarity, which is, he points out, an
original argument not found in prior Advaitin literature (1979: 5). This argument is relatively
simple and hinges on the supposition that if there is no enduring subject that is able to mentally
grasp two similar things, then recognition cannot be based on similarity. Again, the reasoning is
simple: Śaṅkara thinks that if a mind is also subject to momentariness (as it presumably must be
according to Buddhist doctrine), then it necessarily cannot endure. If the same mind does not
endure, then each thing grasped is done so by what is effectively a different mind. From this
117
These asaṃskṛta dharmas are space (ākāśa), cessation through discernment (pratisaṃkhyānirodha), and what
Whaling calls ‘cessation through the absence of a productive cause’, but might be better rendered as something
like ‘cessation without discernment’ or ‘cessation not dependent upon discernment’ (apratisaṃkhyānirodha)
(1979: 5). Thibaut translates apratisaṃkhyānirodha as ‘cessation not dependent on such an act’, where the ‘act’
referred to is a ‘sublative act of the mind’, viz. discernment (Śaṅkara, 2011: 410).
P a g e | 154
premise, Śaṅkara argues that if the Buddhist were to admit ‘that there is one mind grasping the
similarity of two successive momentary existences, he would thereby admit that one entity
endures for two moments and thus contradict the tenet of universal momentariness’ (Śaṅkara,
2011: 414). Further, Śaṅkara argues that the act of saying ‘this is similar to that’ demarcates two
distinct existent things, viz. two separate cognitions that are then linked by a ‘judgment of
similarity’ (2011: 414). He continues ‘[i]f the mental act of which similarity is the object were an
altogether new act (not concerned with the two separate similar entities), the expression “this is
similar to that” would be devoid of meaning’, which is to say that insofar as we recognise ‘this’
as similar to ‘that’, we must also recognise that this judgement of similarity is predicated on two
existent things separated temporally. The recognition (as similar objects) of these existents
depends on their separation both temporally and conceptually, and for Śaṅkara, this in turn has
to at least partly rely on a permanence not admitted by the Buddhists.
In terms of the ātman, Śaṅkara argues that we do not rely on a notion of similarity, but
that we are instead directly conscious of it ‘being that which we were formerly conscious of, not
of it being merely similar to that’ and that ‘the conscious subject never has any doubt whether it
is itself or only similar to itself; it rather is distinctly conscious that it is one and the same subject
which yesterday had a certain sensation and to-day remembers that sensation’ (1890: 415). This
is, to Śaṅkara’s mind, the single most effective argument against the Mādhyamikas, and is, to my
knowledge, the only direct attack he makes against them.118
118
One might be tempted to ask whether, if the empirical world is illusory (as it is according to the Advaitin
māyāvāda), then is it not possible that the conscious subject is somehow deluded in its thinking of itself as
P a g e | 155
In what follows, I will discuss some apparent points of doctrinal divergence between the
three Buddhist schools mentioned by Śaṅkara. This will serve a dual purpose – first, it will, I hope,
illustrate why Śaṅkara thought these schools so different from his own (viz. because they deny
ultimate entities such as the ātman). Second, I hope that in so doing it will become clear why
Bhattacharya’s thesis regarding the Mahāyāna and the endorsement of an Absolute that creates
and underpins reality should fail. I will then detail why I think Yogācāra Buddhism should reject
any suggestion of an ultimately existent ātman-Brahman by discussing Vasubandhu’s own
philosophy. My final position will see the claim staked that whilst absolutism might be one
possible outcome for a Yogācārin (indeed, such variations exist), it need not be the default
Yogācāra position. This claim would, I think, be shared by Śaṅkara, who we have seen berate
Buddhist philosophy in contrast to his own.
4.1 The Vaibhāṣika Account of Possession
Now to the details. A significant point of departure between the two of the three
Buddhist schools mentioned by Śaṅkara is the ‘doctrine of possession’. When discussing
Vasubandhu’s account of the Vaibhāṣika doctrine of possession in his Abhidharmakośabhāsya
(AKBh), Collett Cox writes that
[p]ossession is used here with regard to the first moment in which
a factor is attained; accompaniment refers to one's state of being
endowed with that factor, or being endowed with the acquisition
permanent in much the same way that it is deluded in thinking of anything in the outside world as real and
permanent? I suspect that this idea occurred to Śaṅkara, and that he likely thought two things. First, that the
conscious subject (ātman) knows itself more intuitively, immediately and intimately than it is possible for it to
know anything external to it, and second, that the argument to the contrary is perilously close to that of the
Buddhists: they would argue that the very idea of the ātman is simply a mistaken reification. Despite criticisms
attesting to his ‘crypto-Buddhism’, it is clear (as shown above) that Śaṅkara at least regarded his own philosophy as
distinct from that of the Buddhist schools.
P a g e | 156
of that factor, in the second and subsequent moments until it is
discarded.
(1995: 82)
That there is an idea of ‘one’s state of being endowed’ with something is subtly
problematic for the Buddhist that wants to refute the idea of an enduring subject. On this, Conze
writes ‘[t]he term prāpti [acquisition; obtaining] obviously sails very near the concept of a
“person” or “self”. “Possession” is a relation which keeps together the elements of one stream
of thought, or which binds a dharma to one “stream of consciousness”, which is just an evasive
term for an underlying “person”’ (1983: 141). In other words, if there is an underlying entity or
reality that is capable of ‘acquisition’ and ‘possession’, then this would seem to have the
undesirable consequence of affirming some sort of enduring ‘I’ or ātman. Conze elaborates when
he says that ‘possession’ must imply ‘a support which is more than the momentary state from
moment to moment, and [is] in fact a kind of lasting personality, i.e. the stream as identical with
itself, in a personal identity, which is here interpreted as “continuity”’ (1983: 141). The
Vaibhāṣikas believe that possession and its opposite, non-possession, must be discrete dharmas
(and so existent as immutable factors across the three times) that are detached from thought in
order to successfully account for the abandonment of defilements (and thus liberation). Cox
helpfully contextualises why this needs to be the case when she writes that according to
Vaibhāṣika ontology, ‘[d]efilements, like all factors, exist as real entities in the three time periods
but manifest their activity of defiling only in the present’ (1995: 89).119
Of course, as ‘real entities’, these dharmas have svabhāva. The Vaibhāṣikas assert that each dharma is
characterised by this svabhāva and that the momentariness of each dharma refers to its activity rather than to the
119
P a g e | 157
Cox then goes on to give a comprehensive overview of how the process of possession
works. According to the Vaibhāṣikas,
[e]ach defilement is said to arise in the present in relation to a
particular object-support through certain causes and conditions
and to be connected to a given life-stream by a simultaneously
arising possession. Even when the present activity of that
defilement and its possession cease and become past, they both
continue to be connected to that life-stream through subsequent
present possessions that arise successively dependent upon that
original possession. These successive possessions form a stream of
effects of uniform outflow (niṣyandaphala) that not only connects
a life-stream to that past defilement but also serves as a cause for
the arising of the possession of future defilements. Within the lifestream of each individual, these streams of possession connecting
one to past defilements continue regardless of whether or not
defilements are presently active; the streams can only be
interrupted or terminated through religious praxis.
(Cox, 1995: 89-90)
As Cox notes, these defilements exist as entities across all three times (past, present and
future) and so because of this they cannot be ‘abandoned’ by simply destroying them. How then
can the Vaibhāṣikas remove these defilements to reach a liberated state? The answer comes in
the sense of abandoning through separation. This is accomplished by severing the possession of
a defilement, and thus nullifying the connection of a given defilement to a given stream (Cox,
1995: 90), and is the reason why the doctrine of possession is so important for the Vaibhāṣikas
to begin with: their ontological commitments require it. As defilements exist as real dharmas
across the three times and are thus permanent and immutable, they cannot be changed or
dharma itself, which is categorically not momentary: it exists permanently across all three times! There is a
distinction, then, between a dharma’s svabhāva and its activity; the svabhāva (and thus the dharma itself) is
permanent, whereas the causal activity of the dharma is what is impermanent.
P a g e | 158
destroyed. Liberation can only be achieved, then, if one manages to sever the connection that
the defilements have to the stream. This, Cox continues, happens in two stages. First, there is
the cessation of the ‘possession of connection’ to a given defilement, and this serves as the cause
for the arising of the second stage, the ‘possession of disconnection’ (or the possession of the
non-possession of) the defilement in question (1995: 90).
The second step is the
pratisaṃkhyānirodha, or ‘cessation through discernment’. La Vallée-Poussin notes that ‘[t]here
are as many pratisaṃkhyānirodhas as there are "objects of attachment", past, present or future’
(1930: 39), and it is when a practitioner ‘is in possession of all; he is perfectly disconnected from
all impure things; the universal detachment is his own; he possesses the Nirvāṇa, the Nirvāṇa of
all’ (1930: 40). Cox, too, notes that it is possession of all possible pratisaṃkhyānirodhas that the
Sarvāstivādins equate with nirvāṇa (1995: 90).120 Note that this is not, then, a transcendental
realm or subsumption into an essential principle of the sort defended by Bhattacharya, but
instead a case of permanently severing the attachment of defilements to one’s stream of
consciousness. Instead, it is a meditative process that is designed to help the practitioner see
things as they really are, and things really are built of dharmas. This brings us back full-circle
Cox (1995: 91-92) details the means by which this ‘change’ takes place according to Vaibhāṣika doctrine. She
writes that though continuing to exist as discrete dharmas across the three times, the defilements (and their
streams) are interrupted by ‘counteragents’ (and streams of counteragents). These counteragents – and their
streams – are sufficient to block the arising of any subsequent defilements, and they do this via the same process
of ‘possession’ and ‘non-possession’: ‘[a]s factors dissociated from thought, possessions or non-possessions of
factors of differing moral qualities can arise simultaneously in one moment of thought. The possession of a
particular defilement can arise in the same moment as the counteragent to that defilement, or, strictly speaking, in
the same moment as the possession of that counteragent. In this way, the uniform outflow of successive
possessions of past and present defilements can be interrupted and the arising of future defilements can be
obstructed through the presence of yet other possessions, specifically, the possessions of the counteragents
(pratipakṣa) to those defilements’ (1995: 91-92). This is how the Vaibhāṣikas argue that spiritual progress can be
made despite the permanence of dharmas.
120
P a g e | 159
regarding Śaṅkara: it is clear to see why he would not endorse such a metaphysics. The
Vaibhāṣikas still endorse ultimately real atoms – each with their own independent, self-sufficient
natures – that converge and combine in order to account for the world: this is not in accordance
with a doctrine of ātman-Brahman unless the actual svabhāva of all these different dharmas is
in fact the ātman-Brahman. For the Vaibhāṣikas at least, it is clear that this is not the case.
What then, of the Sautrāntika response to this doctrine of possession? The difference, as
we will see, is, despite their shared heritage with the Vaibhāṣikas (both ostensibly belong to the
Sarvāstivādin school of thought), stark. Vasubandhu rejected the differentiation between a
dharma’s svabhāva and its activity and instead argues that a dharma’s existence is constituted
by its activity. Cox (1995: 94) explains that this means that a dharma does not exist as a discrete
entity across the three times, which in turn means that ‘momentariness refers to the
transitoriness of the factor [dharma] as a whole’, and presents an ontological principle that
‘implies that only the present moment exists’.
The reasons for this are surprisingly
straightforward. We have already touched on an objection to the permanence of dharmas on
the basis that there is no reason why, if a dharma is permanent, that it should ever stop being
the way that it is, or more pertinently, ever stop being active or manifesting its activity. I think
that Vasubandhu probably surmised that this provided a stark challenge to the Vaibhāṣika
soteriological project that their doctrine of possession could not satisfactorily overcome. Gold
(2015: 36) explains that the Vaibhāṣika cannot claim that the conditions necessary for the
dharmas to be continually active in all three times do not yet exist. This is because the belief that
dharmas exist across all three times necessarily means that all that has existed in the past, exists
P a g e | 160
in the present, and will exist in the future, is already necessarily existent. Thus, it should be the
case that a dharma’s activity is always active.121
Gold gives the example of an eye: the future conditions that, ‘together with an eye, may
some day produce a visual sensation, must already exist, to the same degree that the future eye
does’, meaning that consequently, ‘a future eye should be able to see its future visual objects,
While this is true, we have seen via Cox (1995: 90) that the Vaibhāṣika thought that these activities could be
blocked by the relevant counteragents: this does not mean that defiling dharmas stop existing or stop trying to
manifest their activity, only that they are prevented from doing so. Consequently, we might think that this
bypasses the problem before it gets started. Nevertheless, there is still a tension here in terms of causality and
discrete dharmas that exist permanently: if possession and non-possession are discretely existent dharmas as
upheld by the Vaibhāṣika, then it becomes increasingly difficult to account for their change in circumstances or the
changes in circumstance that they facilitate, viz. possession or non-possession should be permanently existent
states that continue to manifest circularly (as all the requisite conditions already exist across all times), as
according to Vasubandhu’s objection.
121
In other words, if possession and non-possession are both discrete dharmas rather than two stages of the same
dharmic activity, there is some tension in accounting for how, for example, ‘possession’ arises and attaches itself
to the relevant lifestream. One would assume that if possession manifests its activity at all then it always
manifests this same activity in the same place both in virtue of its inability to change and in virtue of the conditions
required for the activity to manifest at all (being permanently existent across all three times). There is a similar
issue with non-possession. To even begin to try and solve this problem, there would presumably need to be a
circular possession of possession; possession of possession of possession; possession of possession of possession
of possession and so on ad infinitum, which is undesirable for numerous reasons, not least because it would
greatly overpopulate Vaibhāṣika ontology.
Cox (1995: 86) demonstrates that the Vaibhāṣikas were aware of this issue and so claimed that
[t]hough this would appear to incur the fault of infinite regress, the original
possession and the secondary possession of possession function reciprocally;
the original possession possesses both the factor and the secondary possession
of possession, and the secondary possession of possession, in turn, possesses
the original possession.
It strikes me that this solution might bypass the problem of regression, but the problem of explaining how entities
that exist (and so manifest their natures permanently) across the past, present and future could possibly change or
effect change remains. As I have said, it seems that dharmas that have as their nature the action of possession
ought to be always possessing or being possessed. If that is the case, then positing them at all is superfluous as
they are simply static entities unable to perform any practical, worthwhile role. They are, always have been, and
always will be possessed or possessing that which it is within their nature to possess or be possessed by.
P a g e | 161
because they exist, along with it, as future entities; together, they should then be able to produce
the causal result of the contact of eye organ’ (2015: 36). To Vasubandhu, this makes little sense,
and so as both Gold (2015: 37) and Cox (1995: 94) recognise, he rejected the Vaibhāṣika
ontological model and instead advocated a model of a different sort, denying the immutability
of svabhāva but not denying the presence of it (in conventional terms, at least!). What
Vasubandhu does reject is the equating of a dharma’s causal activity and its svabhāva, instead
arguing that a dharma’s activity is constituted by its svalakṣaṇa (own-characteristic), its causal
power or efficacy. It is not the dharmas themselves that are basic, it is their svalakṣaṇa, and a
given svalakṣaṇa can only manifest in the present. This obviously removes the focus upon
dharmas (and so svabhāva) existing across three times and means that dharmas are conditioned
and thus not ultimately existent. Cox (1995: 94) sums it up well when she writes that ‘[c]ausal
interaction then becomes meaningful only as a relation between the present and its immediately
preceding moment, and all present arising can be explained only through a stream of contiguous
conditioning’.
This stream of conditioning is affected by ‘seeds’, which, Cox writes, Vasubandhu equates
with the five skandhas: ‘the very mental and material aggregates of which the life-stream
consists’ (1995: 95). Cox explains that these seeds have the potential to bring about effects inside
a lifestream: this is what is known as their ‘seed-state’. Given that they are not actualised events
but instead potentialities, multiple seed-states can exist in any lifestream.122 Seed-states are
Cox (1995: 95): ‘Since this seed-state is a potentiality and not an actualized event manifesting definite qualities,
seed-states of any moral quality can coexist in one life-stream.’
122
P a g e | 162
conditioned (dependently originated) and momentary, and as such bypass any recourse to
ultimately existent svabhāva.123
As far as Vasubandhu is concerned, this solves multiple problems with the Vaibhāṣika
account. First, it is closer to pratītyasamutpāda in that there are no discrete dharmas with
svabhāva posited as existing ultimately outside of dependent origination. Second, it is closer
than the Vaibhāṣika model to an account of momentariness that, whilst not strictly a feature of
early Buddhism (von Rospatt, 1995: 15-17), can tentatively be traced back to first century CE
canonical literature and can likely be assumed to have been a topic of serious discussion among
Buddhist circles even earlier than that (1995: 17).124 Third, it disposes with the difficulties that
manifest because of the belief that all three times exist. Fourth, this account of momentariness
123
Cox (1995: 5) gives a succinct and detailed account of how seed-states operate. I shall reproduce it here:
Like all conditioned factors, these aggregates and their potential capability as
seed-states are momentary, and this potentiality is passed along through the
contiguous conditioning by which aggregates are produced in each successive
moment. Thus, the actualization of a seed's potential at a later time is not the
direct result of the original factor or action by which the stream of that seedstate was initially implanted. Rather, the later actualization is conditioned
indirectly through the successive reproduction of the efficacy of the original
action in each consecutive moment in the form of a seed-state. At a certain
moment, when the appropriate causes and conditions coalesce, the seed's
potential is actualized.
124
Von Rospatt (1995: 18) ultimately concludes that
on the whole the examination of the Nikāyas/Āgamas and of alleged quotations
of the Buddha yields little concrete information on the development of the
theory of momentariness. Besides the vague possibility that already by the first
century B.C. the theory may have been current (possibly even as the teaching of
the Buddha), it reveals only that if the theory had existed at all by the time of
the final redaction of the canon, then only without acquiring a canonical status
– possibly because it would have been confined to certain circles of Buddhists.
P a g e | 163
and continuity (especially in terms of moral accountability and spiritual progress) is, it seems,
designed to function in the absence of a unifying principle or ground of experience. As Cox (1995:
96) puts it, it is designed to ‘account for direct and indirect causal efficacy in the face of
momentariness and the absence of a unifying substratum’.125 I agree with Gold when he argues
that this can be linked convincingly to Vasubandhu’s denial of ātman.126 Gold’s point is that there
is no need for Vasubandhu to ‘posit continuity between past experiences and present memories
for the same reason that there is no need to posit a real nonexistent to take up the space when
an existent entity passes away’ (2015: 110).127
Although each of the reasons given above is significant in their own different way, I am
primarily concerned with the first and fourth of these reasons insofar as they interrelate: the
denial of ultimate intrinsic natures and unifying principles. Vasubandhu had several reasons to
deny the ultimate, immutable existence of svabhāva, but paramount among them is his
reluctance to veer from doctrinal orthodoxy.
Of course, Vasubandhu was not always a Yogācārin (and thus not always a Mahāyānist), but the lack of a single
unifying substratum is for him – as for Nāgārjuna – a constant. It seems odd that Vasubandhu would expend so
much effort denying svabhāva and a unified substratum if he intended his philosophy to endorse the existence of
the ātman-Brahman.
125
126
About which, more will follow (p167-176).
127
Vasubandhu specifically thinks that it is nonsensical to infer from the absence of an existent entity a
nonexistent cause that brings about the nonperception of that same entity. Such a ‘nonexistent’ is a conceptual
construction, a parikalpita, brought about by an inference and which cannot cause anything real. Recall that for
Vasubandhu, a ‘real’ entity is causally engaged with other entities; mere inference is not sufficient for a causal
connection. The destruction of an entity cannot be caused by any other entity because for Vasubandhu,
destruction/absence is not a real thing. It is, in fact, a conceptual construction that is imposed on the world when
an entity is no longer present: destruction is, as Gold puts it ‘simply the nonexistence of a recently past thing’
(2015: 109).
For a more detailed analysis of this position, see Gold (1995: 107-110).
P a g e | 164
At AKBh 298.21-22, Vasubandhu cites the Buddha:
svabhāvaḥ sarvadā cāsti bhāvo nityaśca neṣyate /
na ca svabhāvād bhāvo 'nyo vyaktamīśvaraceṣṭitam //
Intrinsic nature (svabhāva) exists at all times, [but its] existence is
not accepted (neṣyate) as eternal (nitya) /
Nor is there existence different to intrinsic nature – [this is] clearly
stated by the Lord //
Verses like this can be difficult to interpret because they are both frustratingly terse and
look on the face of things to be contradictory. Such appearances are typical of Buddhist
literature. On the one hand, Vasubandhu appears to be endorsing an immutable svabhāva and
simultaneously equating it with all existence. More than that, it could even be interpreted that
the Buddha (and so Vasubandhu in virtue of the quotation) is claiming that svabhāva exists across
all three times – just like the Vaibhāṣikas! Yet on the other hand he is clearly stipulating that
intrinsic nature be ‘not accepted as eternal’. We have seen how dharma theory leaves open the
reification of dharmas (which ‘are’ their svabhāva) as immutable, permanent entities, eternal
and uncaused. Vasubandhu is aware of this problem and so denies that svabhāva is this
immutable thing present across all three times (past, present and future), instead eventually
arriving at the doctrine of trisvabhāva, or three natures. 128 How, then, to account for the
Buddha’s apparent assertion that svabhāva ‘exists at all times’ (sarvadā: always, forever)? Gold
contends that Vasubandhu holds that the Buddha means not that intrinsic natures exist eternally
128
Recall from §2 that according to this theory, entities do not have one immutable, eternal, really existent
svabhāva; instead, what we usually understand to be svabhāva is best understood as three interconnected
perspectives.
P a g e | 165
(as held by the Vaibhāṣikas), but rather that a given entity can have no nature but its intrinsic
nature (2015: 38). In other words, for an entity in the world to ‘exist’ (conventionally), it must
have a svabhāva, but this svabhāva is resolutely not eternal. Indeed, Vasubandhu’s later
trisvabhāva doctrine states that the second nature of an entity (it can be understood as the
second perspective on an entity) is its causal story, that the svabhāva commonly understood is
actually causally conditioned according to dependent origination (pratītyasamutpāda) (Gold,
2015: 149). We experience things in such a way, but at the same time, a person with sufficient
insight knows that this is an appearance and not the truth of the matter. We have not a singular
svabhāva, but a svabhāva that is ‘inherently threefold’ (Gold, 2015: 148). This, of course, negates
the Vaibhāṣika understanding of immutable, independent svabhāvas – the very fact that the
nature of an entity is in three distinct but dependent parts means that the nature is not a unitary,
self-sufficient, immutable thing. As Gold notes, the fact that svabhāva is now ‘not accepted as
eternal’, not thought of as a unitary, self-sufficient thing, means that svabhāva must be a mental
construction, a parikalpita (2015: 148).
We now know why Vasubandhu rejects the doctrine of possession: it relies upon the
permanent existence of the three times (past, present, future), and the dharmas that provide the
mechanics of the theory require intrinsic existence across those three times. It is for these
reasons that Nāgārjuna would also reject a doctrine of possession: if there are no svabhāvas,
then there can be no existence across three times. Nāgārjuna is keen to point out that change is
constantly occurring and permanence is an impossibility. We will see more on this in a later
section. This being the case, how could a doctrine of possession be endorsed by a Mādhyamika?
If there are no svabhāvic dharmas, there is nothing to be possessed across time. If the three
P a g e | 166
times lack svabhāva, they too are not simultaneously existent. A Madhyamaka argument against
Vaibhāṣika possession, then, is as simple as that. It is not just Nāgārjuna for whom impermanence
is of central importance – it appears that Vasubandhu took Nāgārjuna’s arguments on this front
very seriously indeed. As we will see in the next section, permanent existence is a problem for
Vasubandhu because he thinks that entities can only be ‘real’ (and so can only have any causal
efficacy) if they are momentary, have a causal effect, and are immediately cognisable (i.e.
knowable). In what follows, I will elaborate on why Vasubandhu thought this way and the extent
to which this philosophy precludes belief in permanent unchanging entities such as svabhāva and
the ātman-Brahman.
4.2 Vasubandhu on Svabhāva, Change, and Denial of the Ātman
For Vasubandhu, then, intrinsic natures might appear to be real, but they cannot be
ultimately real if we are to account sufficiently for change and thus for the Buddhist soteriological
process: they must all be mental constructs.129 This must apply, of course, to a universal intrinsic
nature such as the transcendent ātman-Brahman just as it applies to the personal ātman (what
Bhattacharya (2015: 5-6) refers to as the jīva). Buddhists – including Vasubandhu – would
129
On Vasubandhu’s conception of liberation, Trivedi (2005: 234) writes:
To see things as they really are, claims Vasubandhu, is to see them in
meditation without the distorting dualistic mentations and imputations of our
ordinary consciousness. Instead, they are seen in an ineffable meditative
experience as being dependent and always changing, as being part of a flow of
things that have no essences or fixed natures or own-beings (niḥsvabhāva).
This, I suggest, is what Vasubandhu means by the claim that the perfected or
fulfilled aspect (pariniṣpannasvabhāva) of things is their dependent aspect
(paratantrasvabhāva) without the imagined or constructed aspect
(parikalpitasvabhāva).
P a g e | 167
generally make no distinction between ‘jīva’ and ātman (talk of the jīva is in my experience either
entirely absent from Buddhist discussions of self, or simply taken to be equivalent with ātman),
but as we have seen, Bhattacharya contends that the ātman – synonymous with ātman-Brahman
– is the hyper-reality behind the jīva. This distinction is, I think, largely irrelevant to Vasubandhu’s
denial of the ātman because denial of one ought to necessitate denial of the other. I will first
give an account of Vasubandhu’s denial of the ātman and then assess how useful this denial is in
separating his type of Buddhism from the Advaita of Śaṅkara. From there, it should be clear
whether Vasubandhu would or indeed could endorse an ātman-Brahman.
So far, the defining mark of permanence, of an ātman, has been ‘svabhāva’.
A
permanent, fixed, immutable essence. Nāgārjuna is probably the most famous Buddhist to take
issue with the conception of svabhāva as a fixed essence, arguing that it precludes change,
cannot be found under even the deepest analysis, and is little more than a reified idea that serves
only to frustrate the Buddhist soteriological project.
The Buddhist aversion to ascribing
permanent existence or ātman, though, is well attested even in the canonical literature.130 It
looks as though Vasubandhu took Nāgārjuna’s criticisms of svabhāva seriously. We know that
Wynne (2009: 77) dates the first instance of a codified anātman doctrine to around the time of the Second
Sermon, adding that a ‘‘no self’ doctrine cannot be taken back to the Buddha, but was of such influence that it
came to define the Buddhist mainstream for more than two thousand years.’ Wynne agrees with Bhattacharya
insofar as they both claim that the Buddha explained only what was not self rather than that there is no self.
However, such distinctions might prove to be toothless. As Thanissaro Bhikkhu writes, ‘[s]ome writers try to
qualify the no-self interpretation by saying that the Buddha denied the existence of an eternal self or a separate
self, but this is to give an analytical answer to a question that the Buddha showed should be put aside’ (1996). We
ought to take the same tack with arguments such as Bhattacharya’s. The Buddha thought such questions should
be set aside because they are mere distractions, pointless at best and damaging at worst to the soteriology that he
was laying out. The point is liberation, and at that point, writes Thanissaro Bhikkhu (1996), ‘questions of self, noself, and not-self fall aside. Once there's the experience of such total freedom, where [or why] would there be any
concern about what's experiencing it, or whether or not it's a self?’
130
P a g e | 168
he divided the concept up into three distinct but interrelated aspects, the trisvabhāva. Far from
asserting the ultimate existence of such svabhāvas, Vasubandhu is at once accounting for the
experiences that deluded minds have of entities whilst also redefining the reified svabhāva as
something that is necessarily conditioned and constructed. This means that by definition, the
trisvabhāva could not be ultimately existent, immutable, or eternal.
Indeed, a ‘real’ svabhāva of the sort endorsed by the Vaibhāṣikas would, thinks
Vasubandhu, necessarily preclude change – a position shared with Nāgārjuna. This should in
theory rule out the idea that Vasubandhu’s version of Yogācāra is absolutist, though there is some
degree of disagreement on this.131 Interpreters such as Tola and Dragonetti point to the opening
kārikā of the Trisvabhāvakārikā (also Trisvabhāvanirdeśa) of Vasubandhu in support of the idea
of a transcendent Absolute. They translate pariniṣpanna as ‘absolute’, citing convention (1983:
234), whereas Trivedi and Williams prefer to translate it as ‘perfected’ (this is also a legitimate
Trivedi (2005: 233) notes that Paul Williams appears to characterise Yogācāra as ‘absolute idealism’. By this,
Trivedi means that Williams characterises Yogācāra in such a way that ‘what exists ultimately (and, in some
versions of absolute idealism, creates all that exists) is one overarching mentalistic or spiritual thing or principle or
force, whether the Absolute or Mind or Brahman’ (2005: 202-233). Williams does indeed appear to support such a
view when he writes that ‘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’ (2000: 157). However, whilst this does
endorse a substratum, it does not necessarily lead us to some endorsement of ātman-Brahman, which is
fundamentally different to ‘mind’ so considered in Yogācāra. Williams (2000: 157) clarifies his position when he
writes that
131
[w]hat 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 aspect.
What we actually have as substratum then is a flux, the events in which are (conventionally) subject to dependent
origination and karma. Ultimate reality simply is this flux. It is the basis of our experience, and it is the final
analysis of our condition. Williams is quick to point out that this does not make the flux ‘some immutable
Absolute’ (2000: 159).
P a g e | 169
option: see Monier-Williams, 1960: 596). I shall illustrate why how we understand this word
influences how we make sense of the concept more generally and of Vasubandhu’s project when
it comes to talk of svabhāva.
In the Sanskrit, we read at TSK/TSN 1:
kalpitaḥ paratantraśca pariniṣpanna eva ca /
trayaḥ svabhāvā dhīrāṇāṃ gambhīrajñeyamiṣyate
Tola and Dragonetti (1983: 251) render this into English thus:
It is admitted that the three natures, the imaginary, the dependent
and the absolute one, are the profound object of the wise men's
knowledge.
Talk of an ‘absolute nature’ sounds perilously close to talking about a real svabhāva or a
nature that is somehow ‘beyond’ the world but underpinning it, much like the Advaitin
conception of the ātman-Brahman. Could this really be what Vasubandhu meant? Trivedi and
Williams are perhaps more measured with their translation of pariniṣpanna as ‘perfected’, and
this is likely due to their awareness that the word can be understood as ‘reality’, ‘existing’ and
even ‘real being’ (Monier-Williams, 1960: 596). These possible translations certainly seems to
lend themselves to the concept of a Real nature hidden among those that we mistakenly assume
to be ‘real’ natures. In other words, it leaves open the possibility that Bhattacharya is correct: all
entities have a Real nature, and that is the ātman-Brahman. We simply look in the wrong places
when we mistakenly assume something other than the ātman-Brahman to be the ground of all
existence or ‘real being’. Paul Williams thinks that even for the Yogācārin, ‘[i]n order for there to
be absence of subject-object duality there has actually to exist something which is erroneously
divided into subjects and objects’ (2000: 158). He goes further, writing that in the Yogācāra, ‘it
P a g e | 170
is very much not the case that there is universal absence of own-existence (svabhāva)’ (2000:
158). In other words, Yogācārins endorse the existence of svabhāva – potentially of the same
sort as the ātman-Brahman – as the hyper-reality or ground that we bifurcate because of our
deluded thought process and deluded engagements with concepts. This is ostensibly true for
some Yogācārins (there is, like any Buddhist school, divergence over technicalities and details!),
but I do not think it is the case for Vasubandhu, as I will demonstrate.
What exactly did Vasubandhu mean? Williams disagrees with both Gold and Trivedi in
his discussion of the absolute or perfected aspect (perfected nature, absolute nature;
pariniṣpannasvabhāva) insofar as he concludes that this perfected aspect does not imply the
nonexistence of svabhāva. Instead, he argues that the emptiness (śūnyatā) referred to in
Yogācāra 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’ (2000: 157). There is still emptiness; it is just
not used in the same way that Nāgārjuna used it. It is worth noting, however, that Nāgārjuna
certainly thought that emptiness understood as the lack of svabhāva necessitated a lack of
subject-object duality. Indeed, McCagney writes that for Nāgārjuna, ‘[t]he term “śūnyatā”
functions by pointing to the incoherence of assuming that events are determinate or definable’
(1997:95). Thus, McCagney (1997: 93) opines that events and entities are beyond categorisations
of ‘existence’ ‘not because they have svabhāva, independent eternal existence or nonexistence,
but because they are niḥsvabhāva, open-ended (śūnya) and indeterminate (animitta). Therefore
they can occur.’ This is to say that a diligent Mādhyamika ought not to commit to either position:
I ought not to affirm my laptop’s existence, but nor should I affirm its nonexistence. Getting
tangled up in either concept is the trap into which people generally fall, and this trap only
P a g e | 171
compounds our dissatisfaction.
‘Existence’ is used in a nuanced way: my laptop exists
conventionally, a designation that illustrates the product of innumerable causes and conditions
that have combined in various ways to produce that on which I type. My laptop does not exist
ultimately, by or with svabhāva; it does not enjoy a privileged status according to which it has an
immutable, supra-mundane nature. It does not have a permanent, eternal identity. This is what
it means to be ‘ultimate’ in the Buddhist literature succeeding the Abhidharma.
Whilst the distinction between śūnyatā in a Madhyamaka context and śūnyatā in a
Yogācāra context can be forced in the above manner, then, it strikes me that they need not
necessarily be considered as fundamentally different in scope or result. Taking away the
ontological aspect for a moment, both aim at the epistemological realisation that there are no
immutable, eternal entities. There is of course disagreement regarding the scope of śūnyatā in
terms of ontology. I am not at all convinced that Nāgārjuna is actually concerned with providing
a metaphysic of ontology for reasons that will become clear as we progress. He certainly appears
to be concerned with removing other people’s ontological positions, but he does so with a
specific goal in mind. It is uncontroversial to say that Nāgārjuna is mainly concerned with
providing an account of the means to liberation; this is the raison d'être of all Buddhist praxis and
of all the Buddhist philosophy that aims to make some sort of sense of this praxis. This account
of liberation from dissatisfaction (duḥkha) mainly focusses on how we understand and interact
with the world. People impose ontological views and these views then corrupt or detract from
their soteriological aim. Nāgārjuna wants to recapture this central objective. He thus reduces
ontological claims down to absurdities, as in the case of svabhāva (and, I think by extension,
dharmas as characterised by Ābhidharmikas). But this is not to say that he replaces these
P a g e | 172
ontological principles with śūnyatā. Śūnyatā looks to me to function as a tool of understanding,
namely of understanding change and causality with the terminal aim of explaining how we might
change from deluded to awakened. What it does not look to be – though it has been taken as
such by several subsequent Buddhist schools – is some sort of substantial substratum or
Absolute. This would be the most egregious of reifications and an affront to Nāgārjuna’s entire
project. In other words, use of śūnyatā strikes me primarily as an epistemic endeavour. This
would certainly be in keeping with the Buddha’s famed reluctance to answer what he saw as
irrelevant metaphysical questions; ruminating on the ontological makeup of the world would for
Nāgārjuna be as pointless and distracting as it was for the Buddha. Consequently, we should
focus on our understanding of the key principles of liberation, none of which involve speculation
about a ground or Absolute principle in the vein of the ātman-Brahman.
Despite the apparent disregard by Nāgārjuna for any sort of grounding principle, I do not
think that he wants to claim that entities do not ‘exist’ conventionally, for we engage with things
all of the time, and he recognises that we must navigate this world of objects. Nor is it to say
that events do not occur conventionally, for we experience them all of the time. Instead, we can
see that Nāgārjuna wants to say that they do not enjoy a privileged existence; they are not
endowed with svabhāva. They are not ultimately existent. King points out that for Nāgārjuna,
there is no possibility of substantial existence (dravya sat), and all dharmas are, therefore
nominal (prajn͂apti sat) (1995: 120). There are, then, no substantial entities – how could there
be a substantial ātman-Brahman? There is a caveat in terms of Yogācāra, however. The
Yogācārins claim that consciousness (translated by Lusthaus as vijn͂apti, synonymous with citta)
is a dravya – it is a real thing. Lusthaus writes that to be real in a Yogācārin sense is to be
P a g e | 173
momentary, causally efficacious, and able to be cognised (2002: 453). Such a thing is still saṃvṛti.
The idea at play here is that whilst consciousness does not enjoy a privileged existence (and so
whilst ‘real’, it is not ‘ultimately real’), it ought not to be rejected as nonexistent (Lusthaus, 2002:
462).
There is some concern in the Yogācārin texts that a Madhyamaka approach might
unwittingly fall foul of this problem. Mādhyamikas do indeed tend to negate everything. They
would have no problem in arguing ‘neither consciousness nor not consciousness’, for example,
placing the role of what Yogācārins consider to be the primary means by which we analyse our
existence in some sort of existential limbo. 132 For the Yogācārins, this is a problem precisely
because to deny consciousness, or to remain ambiguous about consciousness, is to deny
everything. Lusthaus (2002: 463) elaborates:
Without some acceptance of the facticity which is never anything
or anywhere other than consciousness, nothing whatsoever can be
affirmed or denied, nothing can be known or understood.
Knowability, by definition, requires consciousness, i.e., an
amenability to awareness. Without some basis for knowledge, not
a single determination can be made about the form or content of
one's experience.
Consciousness is then a ground of sorts – it grounds our knowledge of the world insofar
as it allows both experience of and reflection upon objects, events and so on. A ground of this
sort, though, is not quite equivalent to a ground of the sort that Bhattacharya concludes that
Mahāyānists endorse. It is not a ground of all being. It is not the material and efficient cause of
I think that this concern is somewhat misplaced in the wider scheme of things. In negating or ‘denying’
consciousness, all a Mādhyamika is really doing is denying its ultimate ontological status one way or the other.
132
P a g e | 174
existence as Brahman is purported to be. Nor is consciousness a fundamental Absolute:
Vasubandhu acknowledges other minds without supposing that they all share the same causal
nexus, viz. as microcosms of the ātman-Brahman. In order to avoid the charge of solipsism, the
Yogācārins need to account for more than one mind. Lusthaus (2002: 489) demonstrates that
Yogācāra necessitates multiple consciousnesses, arguing that such intersubjectivity is essential
to the Yogācāra account of the world:
Yogācāra does not advocate solipsism. Consciousness is
intersubjective; karma is communal as well as personal. Therefore,
the existence of other minds is affirmed. . . Not only is no attempt
made to reduce `other minds' to mere projections of one's own, but
the very core of Buddhism – the teaching of Dharma by one sentient
being to another – is made absolutely contingent on there being
consciousnesses external to and yet perceptible by other
consciousnesses. In other words, the entire point of Yogācāra
phenomenology rests on both the necessity and possibility that
there be communication between distinct minds.
Thus the final realisation, insofar as there is one, looks to me to be identical in the case of
both Madhyamaka and Yogācāra: there are no immutable, intrinsic natures. For Nāgārjuna, the
realisation of śūnyatā provides this insight. For Vasubandhu, the realisation of the perfected
aspect provides this insight. In both cases, there is a common factor: the insight remains the
same.
This truth is to be realised in meditation, a feature that Nāgārjuna’s use of śūnyatā
appears to share with the Yogācārins.133 The Yogācāra of Vasubandhu thus shares a couple of
Williams writes that ‘[t]hat emptiness, that very absence itself, is the perfected aspect, and it has to be known
directly on the deepest possible level, in meditation’ (2000: 158).
133
P a g e | 175
important features with Madhyamaka. Śūnyatā still necessitates a lack of svabhāva, and this
truth is to be realised in meditation. Both appear to disregard any sort of substantial substratum
existing through the past, present and future, although it can certainly be argued that
Nāgārjuna’s means of doing so is more thoroughgoing than that of Vasubandhu. We have seen
that Nāgārjuna categorically rejects any substance at all, equating substantial existence with
possession of svabhāva. But what of Vasubandhu? Is not Yogācāra concerned with mind-only,
and this being the case, is the mind a substance? Williams writes that the ‘mind’ is the primary
substratum for Yogācāra; it ‘is the one primary existent that serves as the substratum for
everything else’ (2000: 160). This phraseology is awkward, for my own previous talk of substrata
has been in relation to supra-mundane, immutable Absolutes. It is true that for Yogācārins –
including Vasubandhu – the mind is the primary consideration. This does not, however, mean
that the mind is some sort of ultimate reality or basic cosmic principle.
Lusthaus writes that the ‘real’ for a Yogācārin is saṃvṛti (conventional) and not
paramārtha (ultimate) (2002: 453), which sheds some light on Williams’ contention. Though the
mind is that with which the Yogācārins are concerned, and though it is the lens through which all
experience must be analysed, it is not an Absolute. If it were, it would be thought of in terms of
the ultimate; as paramārtha. This is precisely how Bhattacharya, Murti et al. think of the ātmanBrahman and śūnyatā. We have seen above that śūnyatā cannot be equated with the ātmanBrahman. There is again a caveat – whilst the mind/consciousness is not a substantial absolute,
there are for Vasubandhu some substantial entities: this is a marked point of departure from
Nāgārjuna’s Madhyamaka. Gold writes that for Vasubandhu, it is present entities and only
present entities that exist substantially (2015: 40). It is tempting to think that this opens up a
P a g e | 176
route for the possibility of an ātman-Brahman, but we have already seen that the ātmanBrahman is thought of as eternal, unchanging and so on. It is not clear to me that we could
recharacterise a sort of radically momentary, radically present ātman-Brahman because such a
characterisation necessitates – as Vasubandhu is at pains to point out - a chain of causal efficacy.
An entity is only substantial insofar as it is present (present phenomena can only be momentary)
and insofar as it has causal efficacy. Gold explains in the concluding remarks to Paving the Great
Way that for Vasubandhu, a self (ātman) cannot be substantially real because ‘[f]rom a causal
point of view, there is no agent and no experiencer who plays an indispensable role in the causal
story of the aggregated elements that make up the apparent self’ (2015: 216). In other words,
the personal ‘self’ so construed in Hindu literature (what Bhattacharya refers to as the jīva) is a
reification precisely because it has no bearing on the causal account of experience. The skandhas
account for experience perfectly well, and so positing a ‘self’ serves no verifiable purpose. The
causal story remains intact without the ātman.
But what about a transcendent ātman-Brahman? Bhattacharya – as we have seen – was
clear in his assertion that whilst the Buddha rejects the reality of a jīva, he does not reject the
reality of the ātman-Brahman. Further, recall that he contends that the Mahāyāna schools were
the ones to remedy this oversight by actively endorsing a doctrine of ātman-Brahman. This does
not seem like an avenue that is open to Bhattacharya in relation to Vasubandhu’s version of
Yogācāra. According to Vasubandhu, something real (and substantial) requires a causal story.
Further to this, it is required to have some causally efficacious role in that causal story. The
ātman-Brahman satisfies these criteria. Recall that for the Advaitins, it is both the material and
efficient cause of the world, and so everything that exists is really Brahman. At this point, we
P a g e | 177
need not be too concerned with the intricacies of accounting for causality within a substance that
is ultimately without attributes and immutable (such a discussion could fill a book of its own!),
we need only acknowledge that Advaitins do believe that Brahman is the cause of and support
of the world. Also recall that Brahman is thought of as eternal and unchanging. It is its eternality
and immutability that precludes the ātman-Brahman from being real and substantial according
to Vasubandhu’s formulations: an eternal substance cannot be momentary in the true meaning
of the word. The best we could say would be that the eternal ātman-Brahman somehow
manifests objects/events momentarily, but this still leaves open two problems. First, the ātmanBrahman itself is still eternal, and so unreal. Second, the ātman-Brahman does not on this
account have the ability to effect change in a temporal causal series. This is simply because an
eternal ātman-Brahman could not change, and acting to bring about change necessitates some
change in the state of the thing doing the act.134 Of course, we know that for Śaṅkara and the
Advaitins, ‘there is no object that enjoys a separate ontological existence and nature from
brahman’ (Rambachan, 2006: 88). But for Vasubandhu, we have seen that there is no way that
an eternal ātman-Brahman could really exist: talk of eternality and so on would make the
Brahman some static thing, not acting, not causing and not responsible for anything. Whilst this
might accord with the Advaitin neti neti approach, it does nothing to account for a world that the
This is a mundane point. If I want to do something, there are several changes in ‘state’. First, I must decide to
do something, which is a change in my mental state. Then I must act, which changes my physical state; I need to
get up and go into the kitchen if I want to make a cup of tea. Then there are results, which might again change
both of these states in some way. We can see that the act of effecting change also necessitates some degree of
change – no matter how minor – in the thing doing the acting. Śaṅkara denies this in relation to the Brahman,
using the doctrine of māyā to explain away how the Brahman only appears to act. This is deeply unsatisfying and
is, to my mind, not satisfactorily explained by any Advaitin literature.
134
P a g e | 178
Advaitins still want to claim is Brahman. More to the point, by Vasubandhu’s criteria, the
Brahman simply cannot exist. Brahman does not act, is eternal and yet somehow created the
world. None of this tallies with Vasubandhu’s account of a real entity, which, as we know, is to
be causally efficacious, cognisable and momentary. To be causally efficacious takes on an added
dimension in terms of Vasubandhu’s philosophy. It is not enough to cause some effect; the entity
must itself be a momentary effect of a prior, past cause. Causal efficacy is reliant on a chain, the
chain of pratītyasamutpāda, which has no discernible beginning. The Brahman is by definition
uncaused. It cannot be part of this causal chain, just as it cannot be truly momentary in the sense
that it is caused, manifests, passes. It seems clear to me then that no Yogācārin following
Vasubandhu’s lead ought to endorse the ātman-Brahman as the immutable ground of all
existence. To do so is to surely miss the point, and so I think that Bhattacharya’s thesis that
Mahāyānists believe in an equivalent to the ātman-Brahman fails if that Mahāyānist is a
Yogācārin (or at least a Yogācārin in the tradition of Vasubandhu). This demonstrates that it is
not necessarily the case that the Mahāyāna directs practitioners toward an ātman-Brahman.
4.3 Brahman, Action, and Māyā
Bhattacharya, we saw earlier, writes that Brahman ‘is what makes us act (kārayitṛ), but it
does not itself act (akartṛ)’ (2015: 29). Such language is typical of Advaitin literature and typical
of the Upaniṣads. Bearing in mind what has been said above, Vasubandhu’s retort is simple
enough – how can that which does not change cause something else to change? I previously
wrote that every act requires a change in the agent, be it mental or physiological. Is it the case
that the ātman-Brahman can somehow cause actions without itself acting? Would not an action
require some sort of motive? Is Brahman just so radically other to us that it can cause acts
P a g e | 179
without acting, without motive, and we are simply unable to understand how? This is not a
defence offered by Śaṅkara, nor is it a defence offered by Bhattacharya. It is deeply dissatisfying
to claim that a cause of an effect does not itself have a cause: this is the Buddhist account of
dependent origination (pratītyasamutpāda). For the Buddhists, it is a matter of experience and
analysis that every cause has a myriad of causes and conditions behind it; never at any time can
we discern there to be a single start point of causes. There is instead an infinite web of
innumerable causes and conditions interacting to bring about all psycho-physical phenomena.
How then can the Advaitin be justified in claiming that the ātman-Brahman, that which does not
act but causes action is responsible for the world? How can Brahman create if it is eternal and
unaffected by motives? Fost cites Bādarāyana’s account of creative action, claiming that
Brahman ‘is moved not by need or necessity but rather by a free, spontaneous, and joyous
creativity, a release of energy for its own sake’ (1998: 393), which relays the stock Advaitin
response, but does not get us any closer to an adequate explanation. According to this account,
Brahman is still moved by a motive; it is simply that the motive is playfulness. On this, Fost writes
that ‘notion of "sport" or "play" (līlā) represents a third sort of activity, one that is neither
purposive nor purposeless’ (1998: 393). I understand that play might be said to be ‘mindless’
insofar as we can engage in it in an uncritical, unthinking way, but I think that to claim that play
is ‘neither purposive or purposeless’ is somewhat mistaken. Play might not have a purpose
outside of itself, that is to say that play can be done for its own sake and so on. But there is still
a motivation to play, even if that motivation is simply to play for playing’s sake. Brahman cannot
be motivated. Indeed, Brahman cannot be affected by anything. Consequently, we are left in
P a g e | 180
the dark as to why there is anything at all. For this reason, I find neither Bādarāyana nor Fost’s
account particularly satisfying.
Buried in all of this wrangling is a significant point: if Brahman creates according to mere
sport or playfulness, the Brahman has still acted. Fost responds to this by arguing that the
explanation of playfulness is not meant to be taken literally as an ultimate account of Brahman.
I have written elsewhere that for the Advaitin, nothing at all can be predicated of Brahman on an
ultimate level. Instead, Fost writes, we should understand līlā only on the conventional level, as
a helpful but ultimately sublatable designation (1998: 396). He then equates līlā with māyā,
writing that ‘[t]he metaphysical category is māyā not līlā, which is used only as a metaphor to
defend the absolute freedom of Brahman’ (1998: 396). Māyā is another tricky topic for the
Advaitins. Translated literally, it means ‘illusion’ or ‘magic’, and the Advaitins generally consider
it to be the cosmic force that obfuscates the Brahman in the eyes of the unenlightened. As
Brahman is everything (and indeed everything is Brahman), māyā must be an aspect of Brahman.
The Advaitin position is that māyā is a veil of ignorance (avidyā) that conceals our true nature,
the ātman-Brahman. The precise nature of māyā and its relation to the Brahman is not
developed in any detail by Śaṅkara. There are the usual metaphors of ropes mistaken for snakes
until knowledge sublates the misapprehension of reality and defeats the delusion, as it were, but
this does not go far enough. It simply tells us how we might operate within māyā, not how māyā
came to be a thing in the first place. Śaṅkara also talks about māyā as an illusion performed by
a magician: it affects the audience, but not the performer. Fost cites Śaṅkara thus:
One of the favorite analogies used by Advaitins to depict the world
of māyā is that of the magician and his deceptive ploys. Just as a
P a g e | 181
magician, who with his conjuring tricks "plays" with our perceptual
faculties in order to create the illusion that something has come
from nothing or that one thing has changed into another, so
Brahman by his mysterious, creative power deludes us into
believing that the phenomenal world is real. Śaṅkara writes: "As
the magician is not at any time affected by the magical effect
produced by himself, because it is unreal, so the highest Self is not
affected by the world-[effects (or appearances)].
Again, we have an explanation – plausible or not – of how māyā might manifest as part
of Brahman without actually affecting Brahman, but we are again left with a question of agency.
A magician chooses to perform an illusion, cast a spell, and ‘play with our perceptual faculties’,
does the Brahman? The answer needs to be a resounding ‘no’, and so the metaphor is
unsatisfying. I expect that the response from an Advaitin would very likely be that the metaphor
is not meant to give a literal account of the relationship between Brahman, māyā, and the jīva.
All of this stuff is, after all, ultimately beyond linguistic designation. Brahman is beyond agency,
beyond attributes, beyond everything. Nevertheless, there seems to be no causal explanation
for our delusion and the apparently random obfuscation of the ātman-Brahman even on the
conventional, worldly level. This has to be a source of some dissatisfaction even for an Advaitin.
In contrast, a Buddhist account of delusion has at its core a causal chain to account for how we
get to this point of delusion and avidyā. We reify because we mistake the causal interactions of
skandhas and the causal interactions of all other psycho-physical phenomena as indicative of
intrinsic existence. This mistake means that we think of the world around us as immutable and
independent rather than mutable and dependent – we then reify both the existence of the ‘I’ and
the existence of day-to-day entities and invest into them all sorts of meanings and feelings,
cultivating different levels of attachment and resulting in different levels of dissatisfaction.
P a g e | 182
The primary cause of this dissatisfaction is, famously, the ‘self’ (ātman). The Advaitin
account, or so it seems to me, has no such causal account by which to explain how or why we –
being identical to the ātman-Brahman – confuse or mistake both the world and ourselves. Where
one is offered, it is incomplete and dissatisfying, having recourse to metaphors that do not quite
address the real questions and taking as a given the existence of a Brahman from which māyā
has shrouded us, and about which we are perennially confused. Māyā is itself a mystery, its
existence appearing to be taken on faith. Māyā is, as Lusthaus says, ‘an inexplicable cosmic flaw’
(2002: 484). We have the assertion of Brahman as a basic substantial ground of everything, and
we have māyā inexplicably clouding our self-knowledge (which is in the final analysis equivalent
to obscuring our knowledge of the Brahman). This is a clear point of departure between Advaita
and Buddhism. How can Buddhist schools that conceive of the world in the ways in which I have
outlined above be said to endorse or believe that there is a transcendent ātman-Brahman as put
forth by Śaṅkara? It is my contention that if they take seriously pratītyasamutpāda, reason, and
analysis, they simply cannot.
4.4 Yogācāra Idealism?
Finally, can the argument be made that Yogācāra as a type of idealism necessitates that
there is only one consciousness, of which we are all parts or manifestations? If this is the case,
then Yogācāra would indeed be a Mahāyāna school which endorses a permanent, immutable,
single substance akin to the ātman-Brahman. This would, of course, mean that Bhattacharya
would be right – at least in part! Recall that Bhattacharya cited the Mahāyāna-Sūtrālaṃkāra to
illustrate that Asaṅga equates emptiness and impersonality with the ātman-Brahman (2015: 2) –
I offered some arguments against Bhattacharya’s specific interpretations in §1; I also advanced
P a g e | 183
my own alternative understanding of the passage in question based upon Vasubandhu’s wider
philosophical method, and so I will not rehash those same arguments here. It is important to
note, however, that in some contexts, Yogācāra is indeed interpreted as endorsing a non-dual
substance like the ātman-Brahman.
Dan Lusthaus has much to say on the question of idealism as applied to Yogācāra. The
general thrust of the arguments presented throughout Lusthaus’ Buddhist Phenomenology is that
‘idealism’ is an inappropriate term to apply to Yogācāra. This is because for Lusthaus, despite the
name Yogācāra strictly meaning ‘mind-only’ or ‘consciousness-only’, no one form of idealism
really captures the Yogācārin project. To this end, he outlines (and rejects) three main sorts of
idealism, which I now give (Lusthaus, 2002: 5):135
1) The mind or some supermental, non-material entity or force
creates all that exists. This is metaphysical idealism.
2) The ultimate ground of all that is or can be conceived is the
cognizing subject, such that the subjective self is the one
epistemological nonreducible factor. This is a different form of
metaphysical idealism, closer to epistemological idealism.
3) Critical epistemological idealism, as opposed to metaphysical
idealism, need not insist on metaphysical or ontological
implications, but merely claims that the cognizer shapes his/her
Lusthaus is at odds with Trivedi (2005: 232), who think that Yogācāra might best thought of as a type of
epistemic idealism:
135
Epistemic idealism, in contrast, makes not an ontological claim but rather the
claim that we know things not as they really are, as claim epistemic realists, but
rather as they are given to us by our ideas, our concepts, and categories. . .a
metaphysically agnostic reluctance to make ontological pronouncements,
combined with something like epistemic idealism, just might be Vasubandhu’s
position if he doubts not external objects themselves but externality, that is if
he doubts not external objects but instead whether our ordinary consciousness
can say anything about objects outside its acts of cognizing them.
P a g e | 184
experience to such an extent that s/he will never be able to
extricate what s/he brings to an experience from what is other to
the cognizer. Like can only know like, so what is truly other is
essentially and decisively unknowable precisely because it is other,
foreign, alien, inscrutable.
It is the first of these with which I am most concerned: ‘metaphysical idealism’. This
sounds, on the face of things, very close to what Bhattacharya et al. propose when they argue
that Mahāyāna Buddhism necessarily entails an ātman-Brahman. On Lusthaus’ definition, the
‘non-material entity or force [which] creates all that exists’ would simply be the ātman-Brahman.
That which we understand to be our consciousness would be no more than a microcosmic
instantiation of this ātman-Brahman, somehow deluded into thinking that we are other than it.
Reality would be one single consciousness creating the world for itself. This, says Lusthaus, is to
mischaracterise Yogācāra. Lusthaus writes that rather ‘than claiming that a cosmic mind creates
the universe, they assert, on the contrary, that one only comes to see things as they actually
become by 'abandoning' or destroying (vyāvṛti,) the mind’ (2002: 5).
We have seen in Vasubandhu’s philosophical method that to reify the mind would be as
mistaken as denying it outright. As Trivedi puts it, ‘consciousness is not the ultimate reality or
solution, but rather the root problem’ (2005: 233). This is not histrionics or a simple instance of
overstating the case. Vasubandhu is, it seems to me, primarily interested in accounting for our
experience of the world and how we manage to mistakenly reify things; an act that will then go
on to cause us dissatisfaction. In this regard, I think his way of approaching the Buddhist soteric
method is analogous to that of Nāgārjuna. Neither seem too concerned with ontological systembuilding, but are instead trying to give a roadmap by which we can deconstruct our worldviews
in order to release ourselves from the existential anxieties that we face. Against the charge of
P a g e | 185
metaphysical idealism, Trivedi counters ‘that all phenomena as they appear to us, as we know
and experience them, are due to the representations and constructions of our
consciousness alone (vijn͂apti-mātra). It does not mean, as is sometimes thought, that
external objects themselves are created by the mind or are somehow mind-dependent’ (2005:
236). He bases this assessment on the opening verse to Vasubandhu’s Viṃśatikā, which I
translate as:
vijñaptimātramevaitadasadarthāvabhāsanāt /
yathā taimirikasyāsatkeśacandrādidarśanam //1//
[All of] this is consciousness-only (vijn͂apti-mātra), because of the
appearance of unreal things, like [a person] with dimmed-eyes136
sees unreal hairs, moons and so on.
It is tempting to read a metaphysical idealist position into this. We could interpret
Vasubandhu as arguing that only consciousness is real because sometimes we see things exterior
to our consciousness that are not real. We might even interpret it to mean that consciousness
creates the objects that appear to us. I do not think that this is what Vasubandhu is really driving
at. Instead, I concur with Lusthaus (2002: 463) and Trivedi (2005: 236) when they argue that
Vasubandhu is simply forcing the point that our frame of reference is exclusively tied to our
consciousness. Our experiences as we see, feel, and know them are ‘consciousness-only’
because our subjective consciousness of them is the only way in which they can be known.
Obviously, this is a quite different claim to that which says Vasubandhu is espousing the minddependence of ‘external’ objects. Just because somebody with an eye disorder sees unreal hairs
136
I think this odd turn of phrase might relate to cataracts, or some such similar eye condition.
P a g e | 186
and unreal moons does not mean that every moon or hair they see is unreal. As Trivedi points
out, that a person with an eye disorder sees – owing to their faulty perceptual apparatus – some
unreal hairs and moons does not mean that there are no hairs or moons whatsoever outside of
their perceptual apparatus (2005: 236). Thus, for Trivedi, Yogācāra does not necessarily deny
events and phenomena external to us: ‘Vasubandhu is not denying that reality exists, as an
idealist might do, but instead is only denying that our conceptual constructions, as presented to
us, correspond to something out there’ (2005: 237-238). The idea is that for Vasubandhu, reality
exists outside of us, it is just the case that our experience of it – as reified – is constructed in
consciousness.
Lusthaus has a slightly different take, though it is, like Trivedi, in no way absolutist. For
Lusthaus, it is externality itself that is problematic for Yogācārins, and not external objects. This
is because we do not experience ‘externality’ in immediate perception – to the Yogācārin, it is
another construct that we impose onto an experience after the fact. Externality is unreal insofar
as it cannot be directly cognised (it is instead inferred retrospectively), and has no momentary
causal efficacy: recall that these are two of the three criteria that Yogācārins use to distinguish
the ‘reality’ of phenomena. The idea is that when we retroactively push externality onto
phenomena, we do so to appropriate them, which is another way of saying that we do it to forge
some degree of attachment. Lusthaus writes that ‘externality is the necessary condition for
appropriation’ (2002: 484). I do not think it much of a stretch to add that appropriation is a
sufficient condition for attachment!
Of course, attachment to phenomena is generally
problematic for Buddhists, as I covered in a previous section. Here, we find that attachment is
linked with appropriation, which in turn means that we have a degree of attachment to
P a g e | 187
everything that we cognise. Lusthaus writes that Yogācārins explain this by arguing that because
consciousness is an activity that usually requires consciousness of (that is to say that
‘consciousness’ detached from cognition of something makes no sense), 137 when we are
conscious of something taken to be external, we appropriate it into our cognition (2002: 486).
This appropriation is an act that imposes a duality and generally imposes reifications: we
appropriate objects or events and think of them as ‘real’, investing in them attributes, properties
and attachments that are not there. The mind, however, is not something that can be
appropriated.138
Liberation must come, then, as non-discriminating cognition (nirvikalpaka-jn͂āna): a type
of cognition that does not appropriate anything, does not impose anything, and requires the
absence of any sort of discursive, discriminative thought. Sponberg (1979: 52) adds that there is
a positive aspect, ‘the direct and intuitive cognition of the Absolute’, but from what has been
discussed, we know that ‘absolute’ in a Yogācāra sense is not equivalent with ‘absolute’ as we
would understand it, for example, in Advaita. We saw previously that the ‘absolute’ or ‘perfected
137
The following extract really gets to the crux of what Lusthaus (2002: 492) is trying to demonstrate:
All I know directly is what happens immediately within my own consciousness.
In other words, consciousness is always and everywhere a case of cognitive
closure. But the closure can never be absolute. What is not of my consciousness
in the genitive and generative sense (i.e., what does not exist simply in virtue of
my consciousness either possessing or creating it) may still exert an influence
on me, it may still be perceived remotely, i.e., filtered through my cognitive
apparatus.
Lusthaus forces this point when he writes that ‘[y]ou can't move someone else's mind (or hand) the way you
can move your own, since it has an independent cetanä. Secondly, hands grasp tangible things. But consciousness
does not 'grasp' (chih, appropriate) other minds, implying that minds are intangible’ (2002: 490).
138
P a g e | 188
nature’ was the understanding that all experienced phenomena are empty (śūnyatā), and so
‘direct and intuitive cognition of the absolute’ simply amounts to what the Mādhyamika might
call ‘the realisation of emptiness’. 139 This cognition is still of something, but it is so without
appropriation. In this way, the enlightened mind can still operate within the world, but it does
not grasp at either the world or itself.
Sponberg argues that nirvikalpaka-jn͂āna is thus a mode that allows the Yogācārin to
operate ‘in both nirvāṇa and saṃsāra, in the supramundane Absolute and in the mundane realm
of discrimination’ (1979: 52), but I think that this is a step too far for reasons outlined above. It
seems to me that this is the only real account of nirvāṇa open to a Yogācārin: they cannot claim
it a permanent place or state for reasons already discussed, and so I wonder in what sense the
‘absolute’ can really be said to be supramundane. It is not a substratum that supports reality, or
a substance that creates reality, such as the ātman-Brahman. Much like the Madhyamaka
project, Yogācārin liberation lacks an ‘essence (niḥsvabhāva) to clearly distinguish it from
anything else. There is, however, no final and all encompassing essence or svabhāva which might
function as some sort of essential self or paramātman to all things’ (King, 1998: 69). It is simply
a way of seeing things, a way of interacting with and thinking about the world. Liberation comes
when consciousness just is; it is conscious of itself in its purest form, without appropriation or
grasping. This does not suggest to me a hyper-real substratum. Intuition of the perfected aspect
is how we reach apratiṣṭhita-nirvāṇa, or liberation not permanently established. We know, of
139
Whether or not this is actually possible is a debate for another day. My main concern here is not to prove that
these conceptions of nirvāṇa are realistically feasible, only that the conception of them need not result in some
absolute like the ātman-Brahman.
P a g e | 189
course, that it is vital to both the Yogācārin and Mādhyamika accounts of nirvāṇa that it is not
permanent, for permanence of nirvāṇa would make it either a separate metaphysical realm, or
a permanent mental state that is either eternal and thus not escapable, or eternal and so not
attainable.140
So is the denial of externality a form of metaphysical idealism? Lusthaus claims that the
Yogācārins deny externality is real on both doctrinal and pragmatic grounds, and I add that
externality is not something immediately cognisable or causally efficacious in the way required
for Yogācārins like Vasubandhu to call phenomena ‘real’.
There is, however, a catch.
Metaphysical idealism is an ontological concern. It looks to explain the way the world is via some
means or other. Lusthaus contends, though, that Yogācāra does not care to build an ontology.
There is no concern at all for ‘ontological regions, but rather psychosophical regions’ (2002:
484).141 All this means is that – like Nāgārjuna – Vasubandhu and the Yogācārins are primarily
interested in how we interact with the world, why we construct reifications, our karma, and how
we might end duḥkha (Lusthaus, 2002: 484). Ontological theorising is a distraction that at best
distracts us from and at worst actively works against the Buddhist soteriological method.
Metaphysical idealism is not simply an epistemic concern, though perhaps Bhattacharya
and other Advaitins would prefer to characterise it that way. At its heart is the ontological idea
This aside from other problems, such as nirvāṇa’s being divorced from pratītyasamutpāda, which I have
covered elsewhere.
140
‘Psychosophy’ as a term is popular in some theosophical circles and has its origins in the Greek words ‘psychē’
and ‘sophiā’. In this context, I think Lusthaus would translate these as ‘mind’ (as opposed to ‘soul’ or ‘spirit’) and
‘wisdom’.
141
P a g e | 190
that there is a single source of existence. It is an epistemic shift that allows one to see this truth,
but the ontological aspect remains throughout. It is the highest truth that all is ātman-Brahman.
Yogācāra, as I hope to have demonstrated does not endorse such a position. Indeed, if it is to be
philosophically consistent, it cannot endorse such a position. Lusthaus cites directly from the
Ch'eng wei-shih lun to argue that Vasubandhu’s Yogācāra recognises the existence of other
minds.
142
The idea is that Yogācāra both acknowledges other minds but maintains
consciousness-only. Lusthaus’ translation is as follows:
You should examine and alertly listen. If there were only a single
consciousness how could the ten directions, the sages and ordinary
folk, causes and effects, and so on, be distinguished? Who would
look for [the teachings] and who would espouse them? What
[would differentiate] the Dharma from its seeker? Thus, the words
'wei-shih' have a deep meaning. The word shih (consciousness,
vijn͂apti, vijñāna) in general reveals that all sentient beings each
have [their own] eight consciousnesses, six types of caittas, altered
[consciousness] (so-pien) qua nimitta- and darśana- [bhāgas],
distinguishing divisions, and tathatā which is disclosed through the
principle of emptiness. Since the self-characteristics (svalakṣaṇa) of
the consciousnesses, [the dharmas] associated with consciousness,
the two altered [bhāgas], the three divisions, and the four real
natures [of the preceding categories], as well as all other dharmas,
are never separate from consciousness, we have established the
[sense in which we use the] term shih.143
(Lusthaus, 2002: 487)
The Ch'eng wei-shih lun is a Chinese text attributed to Xuanzang. Its Sanskrit name is Vijñaptimātratāsiddhi, or
the Discourse on the Perfection of Consciousness-only. It is based around Vasubandhu’s Triṃśikāvijñaptikārikā and
is an important text in Chinese Yogācāra.
142
143
All annotations Lusthaus’.
P a g e | 191
The central thesis at play is, we can see, that there are of course other minds. We are not
simply microcosms of one single supermind, superself, or paramātman. The final point made is
that the ‘only’ in consciousness-only is used to force the point that nothing in experience can
ever be separated from consciousness, and so it is with consciousness that one must work in
order to quell dissatisfying experiences. Lusthaus (2002:491) translates an extract from the
Ch'eng wei-shih lun which gives an account of how other minds are independent but not separate
to our own:
It is only like a mirror, which 'perceives' what appears [within it as]
external objects. [This kind of perception is the type we] term
'discerning (liao) other minds,' though they can't be immediately directly discerned. What is discerned immediately-directly is [one's
consciousness'] own alterations (so-pien). Hence the
[Saṅdhinirmocana] Sūtra says: There is not the slightest dharma
which can grasp the remaining dharmas; only when consciousness
arises does one project/perceive the appearance of that, which is
called 'grasping that thing.
The point is that there are other consciousnesses, but that we are only aware of them in
the first place because of changes in our own consciousness: it is of these changes in our own
consciousness that we are immediately aware.
This means that other minds are both
independent of our own consciousness insofar as they exist, but not separate from our own
consciousness insofar as we only know of this existence via the changes elicited in our own sphere
of consciousness. In this way, then, other minds exist independently of our own mind, but they
can never be known outside of our own mind. Can this apply to objects and so on that populate
the world? If so, it would banish any spectre of metaphysical idealism within the Yogācāra.
Lusthaus (2002: 491) translates the following line in conclusion of the above argument: ‘[o]ther
mind is this sort of condition; rūpa, etc. are the same case.’ This is an incredible claim if Yogācāra
P a g e | 192
is actually idealist in nature. Other minds are real; they exist independently of our own
consciousness of them, but it is only in our own consciousness that we can know them. The same
is true for rūpa, commonly translated as ‘form’, denoting objects, stuff, entities. Rūpa is, as
Lusthaus writes, ‘a remote ālambana’ (2002: 491), or a remote sense-object. It is independent
of our consciousness, but not wholly separate from it because our knowledge of it in the first
place relies on its causing some effect or change within our consciousness. If this bold claim is
true (and there are no reasons that jump out at me to suppose that it is not), then Yogācāra
simply cannot be called ‘idealist’ in any of the senses earlier outlined by Lusthaus.
Lusthaus is thus vindicated in his claim that to try to prove that only mind exists, or to
take the ‘position’ or view (dṛṣṭi) that only mind exists is to miss the point by some distance.
Consciousness is that lens through which the Yogācārins analyse experience because it is the
means through which we ‘do’ our experiencing. Lusthaus is right to state that consciousness so
analysed by the Yogācārins – not as a permanent immutable entity, but as an ever-responding,
ever-changing, intersubjective flux consisting of consecutive moments of consciousness – is
difficult to cling and attach to. In Lusthaus’ words, ‘[c]onsciousness itself is in and as itself
impossible to grasp, rendering it less susceptible to the psychosophic abuses that an external,
physical, possessible world is prone to, or even encourages. One can cling to ideas, but not a
fleeting moment of consciousness’ (2002: 488). This is the crux of the Yogācārin method. This
understanding does not resemble any of the forms of idealism outlined at the beginning of this
subsection, and it certainly does not resemble a monistic absolutist doctrine comparable to that
of the Advaitins. Consequently, it is my contention that Yogācāra philosophy understood as it is
P a g e | 193
in this work is not absolutist and cannot lead to the conclusion that all of reality is the ātmanBrahman.
P a g e | 194
§5: Madhyamaka and Absolutism
I have already said that Bhattacharya is not alone in claiming that Buddhism teaches some
sort of absolutism. It would be a misrepresentation of the Buddhist literature – at least from a
Mahāyāna point of view – to claim otherwise. I follow Tillemans (2019) in acknowledging that
insofar as Mahāyāna Buddhism speaks of the dharmakāya, it tends to do so in an absolute sense;
that is to say that discussions around dharmakāya usually tend to hold it as some sort of ultimate
reality, the essential nature from which the entirety of reality emanates. Tillemans argues –
wisely, in my view – that we should not be so quick to write off absolutistic notions in Mahāyāna
literature as some sort of ‘wrong reading’ imputed by a modern bias. To do so, he argues would
be to ignore a major trend in Indian Buddhist thought and to throw out the proverbial baby with
the bathwater.
He writes that ‘there were important ways Prajñāpāramitā and
Abhisamayālaṃkāra commentators took dharmakāya as a substantive and accorded it an
absolute sense. This is not a theological flight of fancy; it is a major philosophical idea in Buddhist
scholasticism’ (Tillemans, 2019: 641, note 8). It is clear then that despite what we know of the
traditional Buddhist preoccupation with insubstantiality and avoidance of absolutes of any kind
(this is, after all, what the famed ‘middle path’ is all about!), there is a relatively strong tradition
within the Mahāyāna schools of asserting an absolute reality. I do not contest this – it is a matter
of historical fact. Instead, I want to argue that this assertion of an absolute reality is not a
necessary part of the Mahāyāna. Further, I want to argue that Nāgārjuna – the Mahāyāna’s
founding father – does not appear to support any such assertion in his seminal writing, the
Mūlamadhyamakakārikā (MMK), and as we have seen, no absolutism is endorsed in the
Vigrahavyāvartani (VV). With this in mind, I will discuss some of the terms pointed to by
P a g e | 195
Bhattacharya as equivalent (2015: 13) and analysing them – as far as possible – in a Madhyamaka
context. In so doing, I hope to show that Nāgārjuna did not – and so we need not – impute from
such terms any idea of an Absolute like the Upaniṣadic ātman-Brahman.
5.1 Understanding Dharmakāya
Despite the Mahāyāna tendency to speak of dharmakāya in positive, absolutist terms, it
would be disingenuous to state that the use of dharmakāya (or its Pāli equivalent, dhammakāya)
in wider Buddhist literature and traditions is always absolute. There is a strong, demonstrable
Theravāda tradition that views dharmakāya/dhammakāya as the sum of the Buddha’s verbal
teachings or even the sum of the Buddha’s mental qualities (and so correctly discerning all
dharmas is the dhammakāya). I will return to this reading shortly. Chanida attributes the
difference in interpretation to Buddhaghoṣa, who at various points in his works tends to use
dhammakāya in these different ways (Chanida, 2008: 6). Following Buddhaghoṣa, says Chanida,
Dhammapāla144 interpreted dhammakāya ‘as bodies of those extraordinary qualities connected
with the Buddha’s mental purity’ (Chanida, 2008: 7, note 36).145 It seems obvious to me that
‘bodies’ can here be understood in terms of a collection, viz. the collection of qualities connected
with the Buddha’s mental purity, or the collection of qualities possessed by tathāgatas. It is this
Though there are multiple Theravāda writers of this name, I here refer to that Dhammapāla famed for
commentaries on seven early Theravāda texts in or around the fifth century.
144
I think it relatively uncontroversial to assume that ‘body’ here refers to a collection of something – in this case,
qualities or teachings – in the same vein as ‘aggregate’. It need not refer to a physical body, but simply a collection
of content that is in some way linked or alike.
145
P a g e | 196
interpretation (as opposed to a metaphysically substantialist absolute interpretation) that I
contend is necessitated by Nāgārjuna’s wider philosophical outlook.
The principal passage cited in regards to the appearance of dhammakāya’s appearance
in the canonical literature is Dīgha Nikāya (DN) 27.9 (Aggaññasutta), the rear end of which states
(in Pāli):
Taṃ kissa hetu? Tathāgatassa h'etaṃ Vāseṭṭhā, adhivacanaṃ
dhammakāyo iti pi, brahmakāyo iti pi, dhammabhūto iti pi,
brahmabhūto iti iti pi.
Why is this? Because, Vāseṭṭha, this designates the Tathāgata
dhammakāya, brahmakāya, dhammabhūta, brahmabhūta.
How we choose to render these words is of crucial importance. Let us begin with
dhammakāya: is the Tathāgata designated as the collection of their teachings? Are they simply
the sum of their mental qualities? Is, as some Mahāyānists would have us think, the Tathāgata
to be designated as the manifestation of an unchanging, ever-present ultimate reality? There is
a lot of ground to cover here, and I do not intend to tread all of it. Instead. I will give a brief
overview of the different ways that dhammakāya/dharmakāya might be interpreted and then
assess these interpretations in the light of Madhyamaka philosophy more broadly construed.
It is from this starting point then that I will contest Bhattacharya’s claim that the
Mahāyāna ‘put things right’ (2015: 39) in terms of developing an absolutist theory, and that it is
the Mahāyāna texts that classify the ultimate as ‘śūnyatā, tathatā, bhūtakoṭi, dharmadhātu,
dharma-kāya . . . and also ātman in the Upaniṣadic sense’ (2015: 13). I will argue that for
Nāgārjuna at least (and so for Indian Madhyamaka) this equivalence is mistaken, and that
śūnyatā, dharmakāya are, whilst related to the same soteriological aims, nevertheless distinct.
P a g e | 197
I start, then, with dharmakāya. Echoing what has been said above, Paul Harrison explains
that
the dharma-kaya ("Dharma-body," "Body of Truth," "Cosmic Body,"
"Absolute Body," etc.) is both formless and imperishable,
representing the identification of the Buddha with the truth which
he revealed, or with reality itself. As such the dharma-kaya is often
linked with various terms for reality, such as dharmata, dharmadhatu, and so on, and has even been regarded as a kind of Buddhist
absolute, or at least at one with it.
(Harrison, 1992: 44)
We can see how dharmakāya might be interpreted, but so far we are no better off in
terms of understanding why this is the case. It is later in the article that Harrison really sets out
his stall. Analysing the use of dharmakāya in the whole Aṣṭasāhasrikāprajñāpāramitāsūtra
(AsPP), Harrison lists five instances in which the compound occurs and argues that translators
sometimes get their interpretations wrong. This is, he argues, because there is a tendency to
assume that dharmakāya is to be always interpreted as a noun (thus as a tatpuruṣa or
karmadhāraya substantive) rather than as a bahuvrīhi adjective (1992: 50). His reading of
dharmakāya at chapters 4, 17, and 31 of the AsPP as an adjectival compound is, he argues, in line
with the Pāli equivalent as used at DN 27.9, where it is, he says, meant to be understood as an
attribute of the Buddha rather than as a thing in itself. Thus he writes that ‘[t]o put it in more
elegant English, the Buddha is truly “embodied” in the dhamma, rather than his physical person’
(1992: 51). Eckel (1992: 97) has a slightly different take on things, citing passages from AsPP III.146
These passages show the Buddha determining that a person who is devoted to the Perfection of
146
The translation used by Eckel is that of Conze, 1973 (specifically, the passages contained between pages 105107).
P a g e | 198
Wisdom (reading, reflecting on, reciting, honouring, and reproducing the text) will receive greater
merit than a person that builds and worships stūpas containing relics of the Buddha. In this sense,
says Eckel, the ‘Dharma body’ is the prajñāpāramitā text, and it is, in a real sense, a relic of the
Buddha. It is not then the Buddha himself that is embodied in the text, but a relic of his thought
(Eckel, 1992: 98).147
Also of relevance to AsPP III are the occurrences of śarīra. Eckel translates śarīra as ‘relic’
(such as the remains of the Buddha, a monk, a historic religious item); it can also simply mean
‘body’, usually (though – importantly – not always) in the physical sense. It is, then, arguably
147
Eckel (1992: 97) also makes the point that the extract given below is not drawing any sort of distinction
between worship of a relic and study of a text; nor is it drawing a distinction between relics that are physical or
intellectual. Instead, writes Eckel, ‘[i]t is simply drawing contrast between the worship of one physical object and
another.’ He later elaborates on this, writing that ‘[c]omparisons between different kinds of worship are common
in Buddhist literature’, before giving some examples (1992: 213, note 4). This is a fair point to raise (it is indeed
true of many Buddhist writings!), and it is one that has some merit here – a great deal of time is spent extolling the
virtue of honouring the Perfection of Wisdom by covering it in cloth, garlands of flowers, surrounding it with bells
and flags, and so on.
However, it is also worth noting that the Buddha spends a lot of time detailing in the extract to
Aṣṭasāhasrikāprajñāpāramitā 3 exactly how ‘cognizing the all-knowing’ is so important to other beings’ (viz. to
people that are not the Buddha) spiritual journey. Similarly, the Buddha emphasises that worship of the
Prajñāpāramitā is identical to worshipping the cognition of the omniscient (and so worshipping the Buddha’s own
thoughts and teachings), and he does not say this in isolation. When the Buddha speaks of the ‘cognizance of
omniscience’, it appears that his meaning is that it is the practitioner that does the cognising of the Buddha’s
omniscience – in so doing, the practitioner actually takes part in the Perfection of Wisdom, even if only at an
elementary, basic level. This, it seems to me, suggests that the practitioner is indeed being urged to engage with
the text: to study the text is to cognise the (words: thoughts: teachings of) the omniscient. To worship the book
that the text is written in without studying (reading) it is, it seems obvious to me, to cognise none of these things.
Honouring it with flowers, oils, holy robes and so on is but one aspect of the process – yes, it will gain merit
according to Buddhist orthodoxy, but is that really enough? Can these ritualistic acts really be called ‘cognizing the
all-knowing’? Does ritual (which can be performed pretty mindlessly) really amount to the ‘cognizance of
omniscience’? I think not. It strikes me that cognising the thoughts/words of the omniscient entails reading or
hearing them, especially if we are to have the dharma-body, saṅgha-body , and Buddha-body ‘revealed’ to us!
Simply placing a book on a pedestal and prostrating before it and so on is, whilst useful in dedicatory and karmic
terms, surely not enough to ‘cognize the omniscience of the all-knowing’. Of course, a short but significant feature
of the AsPP extract given below is the idea that the person ‘copies out’ the book – I suspect it would be very
difficult to copy out a book in a language one understands without engaging with its contents!
P a g e | 199
technically distinct from kāya, which can refer to a physical body or a collective ‘body’ (aggregate;
group: of texts, for example), though any distinction that I have found seems to be imposed by
later writers in the commentarial traditions. It is, though, when instances of the compound
dharmaśarīra occur that things become a little more blurry. The compound dharmaśarīra can,
according to the Monier-Williams Sanskrit Dictionary, mean either a body or collection of virtues,
or a body or collection of Buddhist relics (Monier-Williams, 1960 512) – recall that dharmakāya
can mean a collection of something, too. With this in mind, it looks as though śarīra might – in
the context of the compound dharmaśarīra – provide some clue as to how the Buddha’s ‘truth
body’ is to be understood after all.
Indeed, at AsPP III, it looks as though either understanding of dharmaśarīra can apply: the
prajñāpāramitā texts referred to by the Buddha can be said to be a ‘relic’ of his thought, but they
can also be said to contain – documentarily, at least – the body of virtues possessed by or the
body of teachings conveyed by the Buddha. 148 Some scholars understand dharmaśarīra still
differently, arguing that it can be equated with dharmakāya – I have Mitomo (1983) in mind,
I tentatively suggest that ‘relic’ would be here understood in one of two ways. First is to see the book
containing the Prajñāpāramitā as a historical relic in the way that an ancient Bible might be (and so an object of
historical or cultural significance that has survived from a much earlier time). The second way would be to treat
the book itself as a religious relic that elevates the capturing of the Buddha’s thoughts to the same sort of level as
a physical relic of the same sort as a piece of his body that has been preserved after death. Either one can be
argued for in the current context, though I do not find either particularly convincing in the wider context of the
passage cited. That a book outlining the Prajñāpāramitā can be a ‘collection of virtues’ in an abstract way is, I
suggest, relatively uncontroversial. Insofar as the book faithfully documents the teachings of the Buddha –
teachings based on and stemming from the Buddha’s virtues as an enlightened being – it can be claimed to be a
‘collection of virtues’ at least in an abstracted documentary sense: a documented collection detailing the Buddha’s
virtues. Indeed, there are those that claim that shrines to books containing prajñāpāramitā teachings are identical
to dharmakāya in the transcendental sense (Mitomo, 1983: 1119), and so they are physical manifestations of the
Absolute. Needless to say that I do not think that this is the appropriate conclusion to draw here.
148
P a g e | 200
here (though he is not alone on this). Whilst I will eventually argue that dharmaśarīra and
dharmakāya can be thought of in analogous terms (in at least some significant ways), I do not
think that we need necessarily subscribe to any absolutist or transcendental interpretations. The
details of all these interpretations will be discussed shortly, looking at a couple of excerpts from
the AsPP. I will then give what I think should be a typical Madhyamaka response to ideas of
absolutism and transcendentalism, drawing primarily on the work of Nāgārjuna. In the excerpt
of the AsPP that I have chosen, we find the following:149
Thus spake Śakra Devānāmindra (Lord of the Gods): ‘The Lord has
said that it is through learning the perfection of wisdom that the
Tathāgata has obtained supreme enlightenment [and] perfect
understanding.’
149
Extract from Aṣṭasāhasrikāprajñāpāramitā 3:
evamukte śakro devānāmindro bhagavantametadavocatihaiva bhagavan bhagavatā prajñāpāramitāyāṃ
śikṣamāṇena tathāgatenārhatā samyaksaṃbuddhena anuttarā samyaksaṃbodhiḥ sarvajñatā pratilabdhā
abhisaṃbuddhā / bhagavānāhatasmāttarhi kauśika nānenātmabhāvaśarīrapratilambhena tathāgatastathāgata iti
saṃkhyāṃ gacchati / sarvajñatāyāṃ tu pratilabdhāyāṃ tathāgatastathāgata iti saṃkhyāṃ gacchati / yeyaṃ
kauśika sarvajñatā tathāgatasyārhataḥ samyaksaṃbuddhasya prajñāpāramitānirjātaiṣā / eṣa ca kauśika
tathāgatasyātmabhāvaśarīrapratilambhaḥ prajñāpāramitopāyakauśalyanirjātaḥ san sarvajñajñānāśrayabhūto
bhavati / enaṃ hyāśrayaṃ niśritya sarvajñajñānasya prabhāvanā bhavati, buddhaśarīraprabhāvanā bhavati
dharmaśarīraprabhāvanā bhavati saṃghaśarīraprabhāvanā bhavati / ityevaṃ
sarvajñajñānahetuko'yamātmabhāvaśarīrapratilambhaḥ sarvajñajñānāśrayabhūtatvātsarvasattvānāṃ
caityabhūto vandanīyaḥ satkaraṇīyo gurukaraṇīyo mānanīyaḥ pūjanīyo'rcanīyo'pacāyanīyaḥ saṃvṛtto bhavati /
evaṃ ca mama parinirvṛtasyāpi sataḥ eṣāṃ śarīrāṇāṃ pūjā bhaviṣyati / tasmāttarhi kauśika yaḥ kaścitkulaputro
vā kuladuhitā vā imāṃ prajñāpāramitāṃ likhitvā pustakagatāṃ vā kṛtvā sthāpayet enāṃ ca divyābhiḥ
puṣpadhūpagandhamālyavilepanacūrṇacīvaracchatradhvajaghaṇṭāpatākābhiḥ satkuryāt gurukuryāt mānayet
pūjayet arcayet apacāyet ayameva kauśika tayoḥ kulaputrayoḥ kuladuhitrorvā bahutaraṃ puṇyaṃ prasavet /
tatkasya hetoḥ sarvajñajñānasya hi kauśika tena kulaputreṇa vā kuladuhitrā vā pūjā kṛtā bhaviṣyati yaḥ kulaputro
vā kuladuhitā vā iha prajñāpāramitāyāṃ likhyamānāyāṃ pustakagatāyāṃ vā satkāraṃ gurukāraṃ mānanāṃ
pūjanāmarcanāmapacāyanāṃ pūjāṃ ca vividhāṃ kuryāt ayameva tato bahutaraṃ puṇyaṃ prasavet / tatkasya
hetoḥ sarvajñajñānasya hi kauśika tena pūjā kṛtā bhaviṣyati yaḥ prajñāpāramitāyai pūjāṃ kariṣyati //
P a g e | 201
The Lord [responds]: ‘Therefore, Kauśika, [it is] not by obtaining this
physical personality 150 that the tathāgata comes to be called
(saṃkhyāṃ gacchati) the tathāgata, [it is] by obtaining omniscience
[that] the tathāgata comes to be called the tathāgata.’
‘Kauśika, the omniscience of the tathāgata, of one who has attained
complete enlightenment, comes forth from the perfection of
wisdom.’
‘Kauśika, the physical personality of the tathāgata, [on which the]
cognizance [of] omniscience is dependent, is the result of skill-inmeans [of] understanding the perfection of wisdom.’151
‘Supported by this cognizance [of] omniscience, the Buddha-body,
Dharma-body, and Saṅgha-body are revealed.’152
‘As the [acquisition of the] physical personality causes [the]
cognizance [of] omniscience, the cognition [that] causes
omniscience becomes a refuge [and a] shrine for all beings. [It is]
worthy of being honoured, of being revered and of being
worshipped.’
‘And after my parinirvāṇa, [my] relics will [also] be worshipped.’
I follow Conze (1973: 105) in here translating ātmabhāvaśarīra as ‘physical personality’ rather than ‘physical
body’ or variations on that theme. This is because in the context of this extract, talk of a physical body without the
connotations of the connected personality seems redundant: the Buddha has acquired a certain personality, a set
of personal dispositions, in virtue of his enlightenment and omniscience. It would be bizarre to say that the
Buddha had acquired his physical body in virtue of his enlightenment – he was necessarily embodied before he
was enlightened! Harvey (1995: 234) writes of the Buddha’s ‘personality’ that ‘when the Buddha is in a nonnibbanic state, he manifests his nature by a 'body', or personality, which is redolent with factors of the Path’,
illustrating that what technically means ‘body’ can here be translated in this manner.
150
I think that the Buddha must mean that the ‘cognizance of omniscience’ by other people is dependent upon the
tathāgata’s personality, viz. the Buddha’s personality – modified upon becoming awakened (thanks to the
Perfection of Wisdom) – is what is responsible for conveying his omniscience to others so that they might cognize
it. I suppose that this would occur when the practitioner dedicates themselves to copying out the book.
151
I have here translated buddhaśarīra, dharmaśarīra and saṁghaśarīra as ‘Buddha-body’, ‘Dharma-body’, and
‘saṅgha-body’ rather than ‘Buddha-relic’, ‘dharma-relic’, and ‘saṅgha-relic’. It could be argued, however, that
each are ‘relics’ that are revealed through awareness of the omniscient if we understand ‘relic’ to be the historic
record of the Buddha’s thought/teaching, in which case, awareness of or cognizance of omniscience might be said
to ‘reveal’ the Buddhas thoughts on each aspect in turn. We can also choose to interpret these things as a body of
ideas related to each jewel (I do not think it a coincidence that the choice of ‘bodies’ marries to the three jewels
prescribed by the Buddha!). The significance of each of these bodies is thus made clear upon our cognizance of
omniscience, viz. upon our awareness of the thoughts/teachings of the omniscient (the Buddha).
152
P a g e | 202
‘Therefore, Kauśika, whoever – son or daughter of noble birth –
having copied [out] the Perfection of Wisdom, having made [it into]
a book, establishing it, worshipping it, honouring it with pūjā, with
heavenly flags, banners, perfume, flowers, bells [et cetera], this
being the case Kauśika, those two will produce greater merit.’
‘In so doing, Kauśika, the son or daughter of noble birth [that has]
made a copy of the Perfection of Wisdom, who worships, honours
and respects it with pūjās of many kinds; they produce the greatest
merit.’
Why so? Because, Kauśika, in worshipping the Perfection of
Wisdom, he worships the cognition [of the] omniscient.’153
As touched on earlier, we have in this passage two central ideas at play. First, we have
the idea that adherence to the Perfection of Wisdom is the only way to achieve enlightenment
(and so omniscience). Second, we have the idea that a person that writes down (and reproduces
– an arduous task requiring concentration and dedication) the Perfection of Wisdom and honours
it (via worship, and, presumably, by living according to it) will be performing a deed more
meritorious than the person that simply worships the relics of the Buddha’s physical body after
his death. It is the distinction between the significance of worshipping and honouring the words
(and so thoughts) of the Buddha over and above his physical body that primarily interests me,
but both aspects are relevant to the current discussion.
The Buddha is very clearly attempting to impress upon Śakra (also referred to as Kauśika)
the importance of the Perfection of Wisdom to both accrual of merit (karman) and to liberation.
153
I chose this particular passage from the AsPP because I think that for my purposes, it gives the most
comprehensive overview in the shortest amount of space. That is to say that it almost makes the point for me that
the ‘body of the Dharma’ can be contained in a book chronicling the Buddha’s thoughts and teachings. Indeed, the
Buddha himself says that ‘in worshipping the Perfection of Wisdom, [the practitioner] worships the omniscient’.
P a g e | 203
The Buddha says that enlightenment ‘comes forth from the Perfection of Wisdom’ and that
omniscience is dependent upon ‘understanding the Perfection of Wisdom’, emphasising the
significance of meditative insight into reality when it comes to liberation.154 The Buddha makes
a distinction between his personality (and so his dispositions, his methods of teaching and so on)
and his status as a tathāgata, and this is simply because he wants to force the point that he is
not awakened because of his personality. He is awakened because he has completely and
perfectly understood the Perfection of Wisdom, which itself facilitates omniscience. In fact, the
Buddha’s personality is, he claims, simply a result of skill-in-means: ‘Kauśika, the physical
personality of the tathāgata is the result of skill-in-means and [of] understanding the perfection
of wisdom [on which] omniscience is dependent.’ This seems to suggest that the Buddha’s
personality is a manifestation of the Perfection of Wisdom that comes about as a result of the
Buddha’s skill-in-means of the same Perfection of Wisdom. In other words, owing to the
Buddha’s skill in both understanding the Perfection of Wisdom and in teaching it, his personality
comes to manifest that same Perfection of Wisdom (i.e. he ‘lives’ what he teaches; he is an
embodiment of the Perfection of Wisdom). This reading is, I believe, supported elsewhere in the
154
There are different ways to understand exactly what such insight involves. A common way to understand the
purposes of the Prajñāpāramitā texts is to see them as guides that aim to direct the practitioner to the nature of
ultimate reality (which is generally characterised as non-dual, non-conceptual, and transcendental). In this sense,
then, it is easy to see how such a transcendental focus on non-duality would appeal to Bhattacharya, especially
given that, as Harrison (1992: 48) points out, this non-dual transcendental vision of reality is also non-conceptual:
it sounds precisely like the ātman-Brahman described by Bhattacharya! It is not clear, however, that this is the
only way to understand the Prajñāpāramitā: somebody like Nāgārjuna, for example, would in my opinion object to
the idea of immersion into a true reality and instead argue that the quelling of conceptual thought need not
necessitate a oneness with some sort of essential principle. On this reading, the general thrust of the
Prajñāpāramitā texts would be preserved, viz. their direction toward a meditative state that precludes conceptual
thought (and so halts reification of entities).
P a g e | 204
Buddhist canonical and post-canonical texts. For example, at Saṃyutta Nikāya (SN) 22.87 we see
the Buddha referring to himself as the embodiment of the dhamma (dharma):
Alaṃ vakkali kiṃ te iminā pūtikāyena diṭṭhena? Yo kho, vakkali,
dhammaṃ passati so maṃ passati; yo maṃ passati so dhammaṃ
passati. Dhammañhi, vakkali, passanto maṃ passati; maṃ
passanto dhammaṃ passati.
Stop, Vakkali! Why [do] you [want to] see this foul body? He who
sees [the] Dhamma, Vakkali, sees me; he who sees me, sees [the]
Dhamma. In seeing [the] Dhamma, Vakkali, [one] sees me; in
seeing me, [one] sees [the] Dhamma.155
The Buddha says at Aṣṭasāhasrikāprajñāpāramitā 3 that ‘supported by this omniscience,
the Buddha-body, Dharma-body, and Saṅgha-body are revealed’, which can be understood in a
couple of different ways. First, in the obvious transcendentalist manner. On this view, the
omniscient tathāgata sees the transcendental truths of the dharma-body, saṅgha-body, and
buddha-body – presumably as ultimate essential principles in the same vein as the
transcendental dharmakāya as explained by Paul Harrison (1992: 44). We already know that for
a significant number of Buddhists, the dharmakāya is to be understood in such a manner. This
being the case, then why should we not also interpret dharmaśarīra – a word which can in
principle be a synonym of dharmakāya – in the same way? It occurs to me that if we are to take
dharmaśarīra in this sense, then we must also for the sake of consistency take saṁghaśarīra and
One might also advance the argument that when the Buddha says that he ‘is’ the dhamma, he is speaking
figuratively in accordance with upāya (expedient means, skilful means). This would mean that his ‘being’ the
dhamma is not necessarily an ultimate truth (and so does not reflect the way that things really are – he is not
literally the dhamma), but is simply a helpful pedagogical device that allows the practitioner to better orientate
themselves toward the path.
155
P a g e | 205
buddhaśarīra in the same sense. Indeed, it looks as though they are intended to be taken in the
same sense when we read them in context in the Sanskrit.
This is, I think, where things become a little problematic for a transcendentalist approach
– at least if we want to view the transcendentalist view as one of metaphysical ultimate realities
in the vein of some popular understandings of dharmakāya. First, it is not obvious to me how we
might take the ‘saṅgha-body’ in a strongly transcendental sense. By this I mean I am unsure how
a metaphysical transcendence might be accounted for in the same way that, for example, there
is sometimes a metaphysical import to dharmakāya as ultimate reality. If we understand
saṁghaśarīra
as
‘saṅgha-relic’,
we
might
understand
the
book
(that
is
the
Aṣṭasāhasrikāprajñāpāramitā) as containing documentary evidence of the saṅgha in a similar
way that we can understand it as providing documentary evidence of the Buddha’s teachings and
documentary evidence of the collection of virtues necessarily possessed by any buddha. In this
sense, it is a ‘relic’ of thought and praxis. It can, it seems to me, also be simultaneously
understood as a ‘body’ of thought and praxis, and so if we move away from the notion of ‘body’
as a transcendental metaphysically Real ‘thing’, the precise translation seems to matter less.156
In any case, the transcendentalism issue is, I think, more of an issue here than it is for
Dharma-body and Buddha-body. Both Dharma-body and Buddha-body can be understood
transcendentally as essentially the same thing – Bhāvaviveka certainly thought that there was no
I have intentionally neglected to comment on the sense of śarīra as physical relics left behind after a monk or
buddha has died and been cremated. Whilst these might be interesting in a wider Buddhological context, it is clear
enough that this is not an application referred to in the extract of Aṣṭasāhasrikāprajñāpāramitā 3.
156
P a g e | 206
difference at all between the Dharma-body and the Buddha-body (which is for all intents and
purposes identical to the Tathāgata-body that he references (Eckel, 1992: 42)).157 I suppose that
in some loose sense, we might be able to claim that the spiritual endeavours of each member of
the saṅgha (be they layperson or monastic) ‘transcends’ the physical borders of specific Buddhist
groups, communities, countries and so on. This sort of understanding would simply mean that
there is some ‘thing’ that binds members of the saṅgha outside of their cultural or communal
bounds: a Buddhist in Sri Lanka shares something with a Buddhist in North America, who in turn
shares something with a Buddhist in South Africa, for example. This shared aspect is Buddhism
– to this extent, then, we might say that the saṅgha-body transcends borders and so on. And yet
this sense of ‘transcend’ seems fundamentally different to the sense advanced in discussions of
the dharmakāya – instead of some strong sense of metaphysical transcendence, of
transcendence into a metaphysical Reality, we have a weaker transcendence; one that is linked
by ideas. The saṅgha-body is a transcendental entity then insofar as it is a set of shared ideals
that extends further than individuals, small communities, and even nations. Of course, Nāgārjuna
Eckel also translates significant portions of Bhāvaviveka’s Madhyamakahṛdayakārikās (MHK) and the
accompanying commentary, the Tarkajvālā, and it is here that we can really see the transcendental import around
the ideas of Dharma-body, tathāgata-body, and Buddha-body:
157
idaṃ tat paramaṃ brahma brahmādyairyanna gṛhyate /
idaṃ tat paramaṃ satyaṃ satyavādī jagau muniḥ //289//
Now, that Supreme Brahman is not grasped by [those (gods)] attached to and connected with
Brahma /
That sage [who] spoke the truth taught that this [Dharma-body] is the ultimate truth //
What is interesting here is the manner in which Bhāvaviveka attempts to distinguish the ‘supreme Brahman’ (MHK
III.289) of the Hindus (and so the Advaitins), and that of the Buddhists - especially Mādhyamikas. It is the Dharmabody however it is conceived) which is the ultimate truth, not the Supreme Brahman that eludes those associated
with Brahma.
P a g e | 207
would acknowledge that any such identity is not ultimately real and does not have svabhāva.
This would mean that the saṅgha-body is a conceptual construct, which in turn makes it
conventional. The same applies to the ideas, rules, and beliefs shared by ‘members’ of this body.
Indeed, the context of dharmakāya in the Sarvāstivādin interpretation appears to be
different still! Guang Xing writes that on the Sarvāstivādin view, Buddhists take refuge in the
dharmakāya of the Buddha because the rupakāya158 is impure. What precisely does this mean?
Xing interprets a section of the Mahāvibhāṣa:
Some people say that to take refuge in the Buddha is to take refuge
in the body of the Tathāgata, which comprises head, neck, stomach,
back, hands and feet. It is explained that the body, born of father
and mother, is composed of the defiled dharmas, and therefore is
not a source of refuge. The refuge is the Buddha’s fully
accomplished qualities (aśaikṣadharma) 159 which comprise bodhi
and the dharmakāya.
(2005: 49)
The last sentence is the interesting part of this extract – the Sarvāstivādins recognised
that the Buddha’s physical body is necessarily impure given that it is made up of defiled dharmas.
This is the case whether the Buddha has ‘nothing left to learn’ or not – in virtue of his body being
This is perhaps best translated as the ‘physical body’ or ‘physical form’ of an enlightened being, where rūpa is
‘form’ and kāya is ‘body’. This is usually described as the physical form of an enlightened being, viewed as a living
embodiment of wisdom or as the embodiment of qualities.
158
We can translate aśaikṣadharma as something like ‘dharmas of those with nothing left to learn.’ There are said
to be ten such dharmas and so the implication is relatively clear: the Buddha has acquired the ten qualities that
constitute liberation. It is in these qualities that we must take refuge, which is to say that these qualities can also
liberate us provided that we work on their cultivation.
159
P a g e | 208
physical (even if it is the physical manifestation of wisdom!), it is impure and conditioned.
Precisely because this is the case, Xing further writes that
[t]he emphasis of the Sarvāstivāda is on the attainment of
Buddhahood, the dharmakāya, while the physical body is
considered secondary. The dharmakāya of all Buddhas is the same.
Therefore, according to the Sarvāstivāda, taking refuge in the
Buddha is to take refuge in all Buddhas because the term ‘Buddha’
includes all the Tathāgatas since they are of the same kind. Thus
another question arises: since the dharmakāya or the Dharma is so
important to the Sarvāstivādins, why do they first take refuge in the
Buddha and not the Dharma? The Sarvāstivādins explained that the
Buddha was the founder. If the founder had not taught, then the
Dharma would not have been manifested. Thus the Buddha is seen
as the first refuge, just as a patient first seeks a good doctor before
asking him for medicine. The patient then seeks a nurse to prepare
the medicine. The Buddha is like the doctor, the Dharma like the
medicine, and the Saṃgha is the nurse. Such is the order of the
three refuges.
(2005: 50)
I think that this helps clarify things a little. The practitioner first seeks refuge in the
Buddha, specifically in his aśaikṣadharma. This is, we can see, equivalent to taking refuge in all
buddhas because every buddha is of the same sort: they have the same aśaikṣadharma in virtue
of following the same path and attaining the same insight and wisdom. In saying that the
dharmakāya of all buddhas is identical, Xing appears to be claiming that every buddha manifests
or embodies the dharma in the same way, presumably because they have attained the same
insights and wisdom. He adds elsewhere that the earliest Buddhists probably understood the
dharmakāya simply as the teachings of the Buddha, and then later as the scripture containing
these teachings (2005: 36). It was from this latter position, Xing suggests, that disagreements
P a g e | 209
between schools regarding how to quantify the dharmakāya began to spring forth.160 It is not
too much of a step to see how from discussions of quantification (in terms of possession of
dharmas) there could be reification of the dharmakāya as a transcendental Absolute, especially
given how Karunadasa (1996: 20) accounts for the supposed ultimacy of dharmas. Nevertheless,
the suggestion at this point is that the dharmakāya is still being thought of as a collection of
qualities rather than as an Absolute of some kind. The practitioner then takes refuge in the
saṅgha (Buddhist community, usually of monastics), which Xing describes as a nurse that
prepares medicine (and presumably gives support – monks and nuns generally live, chant, and
meditate together). The Dharma itself is then the medicine, prescribed by the doctor (the
Buddha), and administered by the saṅgha who then support you while it takes effect. Refuge is
to be taken in all three because all three are vital to curing our dissatisfaction (duḥkha).161
We can see then that in principle, dharmakāya and dharmaśarīra have a close
relationship with each other. In dharmakāya, we have reference to the collection of Dharma:
the totality of the Buddha’s teachings, or we have reference to the dharmas possessed by the
Buddha that allows him to manifest his wisdom and that account for the character of his
160
Xing (2005: 36-44) provides a very thorough account of how the dharmas that constitute various conceptions of
dharmakāya might be interpreted according to the schools from which each list originated. It is outside of the
scope of my current aims to go into much more detail on this.
There could be a case made that the prajñāpāramitā texts constitute a dharmakāya in a slightly different – but
no less absolutist – manner. The Sikhs have their Guru Granth Sahib; a collection of holy scriptures that form the
basis of the Sikh religion but which are also taken in a very literal sense to be the living leader of the Sikh faith. The
collection of texts is thus seen as an eternal, living, real manifestation of God’s word and will. It is in principle
possible to interpret Buddhist scriptures in a similar vein, as an eternal, ‘living’ manifestation of the Absolute (I am
not aware of any sect which does this). On such a reading, the Tathāgata is designated ‘dharmakāya’ insofar as
they are a fleeting physical manifestation of this eternal Absolute.
161
P a g e | 210
enlightened state: the difference here seems to be negligible. In dharmaśarīra, we have either
reference to the collection of virtues acquired by the Buddha (and so extolled in Buddhist
literature), or the ‘relic’ documenting the Buddha’s cognition (and so communicated in the
ensuing literature). Let us not forget that the Buddha’s teachings are supposed to lead to a state
of affairs where the practitioner can too acquire the virtues of a tathāgata (and thus awaken): it
seems to me that it is the case that dharmakāya and dharmaśarīra are two sides of the very same
soteriological coin. To put it another way, a practitioner cannot awaken without acquiring the
collection of virtues that all tathāgatas have possessed (possessing these virtues necessarily
means that any defilements are removed). The general idea is that practitioners would struggle
to acquire these virtues without first being told and shown how to by the Buddha’s (collective)
teachings.162 Further, to understand and adhere to the teachings is to acquire the virtues, whilst
to acquire the virtues is to know the ‘essence’ of the body of the Buddha’s teachings. It seems
clear to me, then, that the dharmakāya need not be characterised as an eternal metaphysical
absolute that is somehow ‘entered into’ by advanced practitioners. It can instead be understood
as the collected ‘body’ of teachings and virtues; it can be understood as a mere concept. As we
know, for Nāgārjuna, all concepts are constructed and conventional, and this happily bypasses
any idea of ultimate existence. It is my contention, then, that this interpretation of dharmakāya
might look like something that Nāgārjuna would endorse.
It is not impossible to reach liberation without hearing and understanding the Buddha’s teachings, at least
according to the Mahāyāna; it is simply rare.
162
P a g e | 211
5.2 Brahmakāya, Dhammabhūta, Brahmabhūta
The next of the four words that we need to place in context is brahmakāya.163 This can
be rendered as brahma-body, which Walshe (1995: 604, note 823) thinks should be understood
as something like ‘the highest’. This is to be understood with the latter part of the sentence:
‘become dhamma, become brahma’ (dhammabhūta and brahmabhūta respectively).
On
Walshe’s interpretation, then, this segment should be taken to mean that the tathāgata
(synonymous with buddha more generally: any enlightened being) has ‘become the highest’ due
to achievement of liberation via their own efforts. On this reading, the use of ‘brahma’ is
uncontroversial. Ergardt (1977: 96-97) claims that rather than stating that the awakened has
become one with the Brahman (see, for example, the usage of brahmabhūta at BG 18.54)164, one
that is brahmabhūta has become so in virtue of their adherence to the religious lifestyle (Pāli:
brahmacariya; Skt.: brahmacarya). This means that the Buddha and the arahants 165 are
brahmabhūta not because they have melded with the ātman-Brahman, but because they fulfil
the requirements dictated by the religious lifestyle they lead: they perform their religious duties
DN 27.9: Taṃ kissa hetu? Tathāgatassa h'etaṃ Vāseṭṭhā, adhivacanaṃ dhammakāyo iti pi, brahmakāyo iti pi,
dhammabhūto iti pi,brahmabhūto iti iti pi.
163
164
BG 18.54:
brahmabhūtaḥ prasannātmā na śocati na kāṅkṣati / samaḥ sarveṣu bhūteṣu
madbhaktiṃ labhate parām //
[He who has] become one with the Absolute (brahmabhūtaḥ), [who has
attained] the serene-nature (prasannātmā), [does] not lament nor [does he]
desire: [he] is equanimous [towards] all living beings; [he] attains highest
devotion to Me.
165
Skt: arhat; in the Tipiṭaka, an arahant is one that is free from defilements, that is enlightened. This definition
gradually evolved to mean different things to different Buddhist schools (particularly upon the advent of the
Mahāyāna), but such later developments are peripheral to the point here.
P a g e | 212
with diligence. Analysing this, Ergardt writes that ‘the obvious qualities of a life according to
dhamma must be the foundation for the meaning of the word brahmabhūta’ (1977: 97).166
Xing takes a different view, writing that the Buddha appears to be comparing himself to
Brahma for effect, writing that in such instances where brahmakāya appears ‘[a] parallel is clearly
being drawn between a brāhmaṇa and a Sakyaputtiyasamaṇa, and to indicate that dhammakāya
is equated to brahmakāya’ (2005: 71).167 Xing further cites Buddhaghoṣa to advance the claim
that according to the Theravādins, dharmakāya/dhammakāya ‘simply means the teachings of
the Buddha’ (2005: 71). This being so, brahmakāya is usually understood as ‘the body of Brahma’
or the ‘divine body’. We can also understand it as ‘the highest body’. This is, I think, subject to
the same sorts of nuance as dhammakāya in the contexts of the Nikāyas: ‘body’ as a collection
or aggregate. It is not obvious to me, then, that Bhattacharya is justified in assuming that the
Buddha is equating himself with Brahmā for any other purpose than to reason with Brahmins in
the terminology that they are used to. I contend that at DN 29.9, the Buddha simply means that
he has become the highest (brahmabhūta), become the dhamma (dhammabhūta), and that the
wisdom of the Tathāgata constitutes the collected dharma (dhammakāya), which is the highest
collection (brahmakāya). It is straightforward enough to see why the Buddha would, when
At least in these Buddhist contexts, for we have already said that brahmabhūta has different connotations in
Hindu contexts.
166
In this instance, a brāhmaṇa is a Brahmin priest and a Sakyaputtiyasamaṇa is a follower of the Buddha.
Sakyaputtiyasamaṇa translates from the Pāli as something like ‘ascetic [that] belongs to the son of the Śākyas’, or
‘monk that follows the Buddha’ (note in Sanskrit that the Buddha is referred to as Śākyamuni, ‘sage of the Śākyas’).
167
P a g e | 213
talking to a Brahmin, use this terminology: he equates himself with Brahmā in order to soften the
reluctance of Brahmā’s followers to take on board the dharma.
Unsurprisingly, Bhattacharya disagrees with both of the positions outlined above. Whilst
he concedes that the aim of the relevant passage is to ‘establish the superiority of Dhamma in
relation to brahmins’ (2015: 124), he goes on to claim that the phrasing at DN 27.9 is supposed
to be taken in a transcendental sense. That is to say that for Bhattacharya, talk of the Buddha
being dhammakāya and dhammabhūta is to be taken literally – the Buddha is the dharma in the
same way that everything in the world is the ātman-Brahman. He has become the dharma and
the Brahman in the same way that the liberated Advaitin has become Brahman. I think that this
is mistaken: brahmabhūta can, for example, be translated as ‘become pious’, ‘become highest’,
‘become the best’. Similarly, brahmakāya can, if analysed in a manner similar to dharmakāya,
simply mean ‘the greatest collection (of qualities)’. This is because as I have already said, kāya
can simply mean ‘body’ (as in the collection of things constituting a ‘body’), whilst brahma is
frequently used in Buddhist literature as a synonym for ‘supreme’ (Harvey, 1995: 234).
We are thus left with some relatively uncontroversial and decidedly non-absolutist
accounts of words that might otherwise have prima facie lent themselves to the Advaitin project.
These interpretations are, I feel, more in line with the general understanding of Buddhism at the
time of the Nikāyas. I have to wonder that if the Buddha had meant to preach the dharma as an
addendum to the Vedānta, would he not have just said so? Interpreting key terms in this way is
entirely plausible, and read in the ways that I have outlined, I contend that these terms need not
be imputed with a substantialist interpretation, even if that has clearly been the case across some
P a g e | 214
Mahāyāna traditions. I reiterate that my argument is not that Bhattacharya is mistaken in noting
that some Mahāyāna schools propagate substantialist ideas of an Absolute, it is simply that he is
mistaken in asserting that the Mahāyāna does this in toto. Madhyamaka when understood by
the writings of its founder, Nāgārjuna, should not do so. It is Bhattacharya’s position that
Madhyamaka should (and indeed does!) endorse such a position, and it is true that parts of the
Madhyamaka school do hold such a view: it is present in parts of tathāgatagarbha and
prajñāpāramitā literature, however we choose to interpret it.168 It is not, however, a necessary
conclusion.
5.3 The Tathāgata and Ensuing Implications
We can make further sense of the contexts afforded to dharmakāya (and, by extension,
dharmaśarīra) if we place them in context with the other attributes assigned to the Tathāgata in
the Dīgha Nikāya. We have already seen that for Bhāvaviveka, the dharma-body, Buddha-body
and Tathāgata-body seem to be equivalent. In terms of etymology, Buddhagoṣa outlines eight
ways that tathāgata might be understood,169 but it is generally thought that only two of these
interpretations are tenable. As such, these two etymologies are the most commonplace in
scholarship today. It is to these two interpretations that I now turn. The first compound generally
accepted by modern scholarship is that of tathā (adverb: thus) + gata (past participle of √gam:
Be that as endorsing an Absolute, as graduated teachings according to upāya (which would involve teaching
things only to later sublate them with advancing levels of ‘truth’), or in some other manner. I suppose that the
ambiguity is part of the point: if one interpretation is favoured or intended, we might expect it to be explicit.
168
169
Chalmers (1898: 104-105) gives a comprehensive overview of Buddhagoṣa’s etymologies. It is not, however, all
relevant to my point here, and so I forego the arduous task of replicating and assessing all of what Buddhaghoṣa
(and, for that matter, Chalmers!) has to say on this.
P a g e | 215
gone). This interpretation then resolves as thus-gone, which sounds pretty but doesn’t on the
face of things help much. Thus-gone from what? Suggestions such as this by John Makransky
(2003: 1) state that ‘tatha-gato, meaning “one who has gone thus,” who has attained nirvana like
all prior buddhas, freed from the conditioned, distorted mentalities and sufferings of mundane
existence.’ This at least gives us something to work with, but it is cumbersome for a couple of
reasons. First, it runs the risk of reifying nirvāṇa: the implication here is that the Buddha has
‘gone’ to his awakening or ‘gone’ to nirvāṇa, which implies that nirvāṇa is an ultimate place or
plane that is somehow outside of the practitioner and is indeed somewhere to which we are able
to ‘go’. Lindtner (1997: 116) acknowledges that the notion of nirvāṇa as a place to which we can
go does seem to have a place in the canonical literature, namely at DN 11.88 [PTS D i223] and
Udāna 8.1 and 8.3 [PTS Ud. 80].170 This would be unacceptable to a great many Buddhists, but
170
DN 11.88:
kattha āpo ca paṭhavī tejo vāyo na gādhati.
kattha dīghañca rassañca aṇuṃ thūlaṃ subhāsubhaṃ,
katta nāmañca rūpañca asesaṃ uparujjhatīti.
Tatra veyyākaraṇa bhavatī:
viññāṇaṃ anidassanaṃ anantaṃ sabbato pahaṃ
ettha āpo ca paṭhavī tejo vāyo na gādhati
ettha dīghañca rassañca aṇuṃ thūlaṃ subhāsubhaṃ
ettha nāmañca rūpañca asesaṃ uparujjhati.
viññāṇassa nirodhena etthetaṃ uparujjhatīti //
Ud. 8.1:
atti bhikkave, tadāyatanaṃ, yattha neva paṭhavi, na āpo, na tejo, na vāyo, na
ākāsānañcāyatanaṃ, na viññānañcāyatanaṃ, na ākiñcaññāyatanaṃ, na
nevasaññānāsaññāyatanaṃ, nāyaṃ loko, na paraloko, na ubho candimasuriyā. Tatrāpāhaṃ
bhikkhave, neva āgatiṃ vadāmi, na gatiṃ, na ṭhitiṃ, na cutiṃ, na upapattiṃ. Appatiṭṭhaṃ
appavattaṃ anārammaṇamevetaṃ. Esevanto dukkhassāti //
P a g e | 216
Ud. 8.3:
evaṃ me sutaṃ: ekaṃ samayaṃ bhagavā sāvatthiyaṃ viharati jetavane anāthapiṇḍikassa
ārāme. Tena kho pana samayena bhagavā bhikkhū nibbānapaṭisaṃyuttāya dhammiyā kathāya
sandasseti samādapeti samuttejeti sampahaṃseti. Te ca bhikkhū aṭṭhi katvā manasi
katvāsabbaṃ cetaso samannāharitva ohitasotā dhammaṃ suṇanti.
atha ko bhagavā etamatthaṃ viditvā tāyaṃ velāyaṃ imaṃ udānaṃ udānesi: atthi bhikkhave,
ajātaṃ abhūtaṃ akataṃ asaṅkhataṃ. No ce taṃ bhikkhave, abhavissā ajātaṃ abūtaṃ akataṃ
asaṅkhataṃ, nayidha jātassa bhūtassa katassa saṅkhatassa nissaraṇaṃ paññāyetha //
At DN 11.88, I assume this conclusion has been reached owing to the repeated presence of ‘ettha’ (there/here)
and ‘kattha’ (where) denoting places. In context, they question where the four elements, name and form, and
eventually object-based, dualistic consciousness are brought to an end. When this happens, the practitioner
experiences viññāṇaṃ anidassanaṃ, a difficult term that is perhaps best translated as ‘non-partaking
consciousness’ (non-manifest is also relatively popular). Bhikkhu Sujato (2011) writes in a blog post that the
context of this particular extract of the DN is often neglected and that far from trying to sneak in some idea of a
‘cosmic consciousness’ (as Bhattacharya would no doubt claim!) or trying to point to a ‘place’ where such
discernments cease, the Buddha is instead reformulating the questions asked in order to serve a specific purpose,
i.e. to refute the Brahminical positions. On this, Sujato (2011) writes that
the reason for the Buddha’s reformulation of the original question becomes
clear. The errant monk had asked where the ending of the four elements was –
which is of course the formless attainments. But the Buddha said the question
was wrongly put, as this would merely lead beyond the form realm of Brahma
to the formless realms. The real question is what lies beyond that, with the
cessation of consciousness. It is not enough for matter to be transcended, one
must also transcend mind as well. If not, one ends up, apart from all the other
philosophical problems, with a mind/body dualism.
If Sujato is right (and it seems to me that his interpretation is entirely plausible), then the Buddha’s rhetoric at DN
11.88 can be sufficiently explained not as endorsing the view that liberation exists somewhere else (i.e. that we
literally travel to it), but rather in terms of his replying to specific query regarding Brahma. As Sujato writes, ‘the
four material elements cease temporarily in the formless attainments, which is the highest reach of the
Brahmanical teachings – even this much Brahma, being a deity of the form realm, did not know’ (2011), and
further still, ‘the Buddha’s real teaching is not to temporarily escape materiality, but to reach an ending of
suffering. And since all forms of viññāṇa (yaṁ kiñci viññāṇaṁ…) are said countless times to be suffering, even the
infinite consciousness has to go’.
In the case of Udāna 8.1, the Buddha speaks of the state, place or dimension (tadāyatanam) in which
conceptualisation stops. It is, I think, reaching to argue that either of these instances (DN 11.88 or Ud. 8.1) really
illustrate that the Buddha points to nirvāṇa as a place or as a substantial thing whatsoever, especially in the case of
Ud. 8.1 where the Buddha seems to me to be more concerned with negating lists of things that can be clung to
than he is with pointing towards an Absolute. If we choose to translate tadāyatanam as (mental) ‘state’ rather
than ‘region’ or ‘dimension’, for example, the meaning of the whole passage dramatically changes. I thus follow
Ireland, who when listing the possible translations for tadāyatanam writes that, ‘Here it [tadāyatanam] is not
meant in any directional (or temporal) sense’ (1977: 161). Adopting this position allows us the luxury of a more
nuanced understanding of the passage, where the Buddha simply wants to illustrate that there is a meditative
P a g e | 217
especially to Nāgārjuna, with whom my wider argument is primarily concerned. Perhaps we
could say that the tathāgata has ‘thus-gone’ from the trappings of saṃsāra, from the
conventional to the ultimate, but this again suggests that upon awakening, there is movement
to some place that is fundamentally different to the place occupied by the unawakened, and it
involves a claim to the existence of an ‘ultimate’ that Nāgārjuna would find problematic.
But is this a necessary conclusion? We might be able to understand ‘go’ and ‘gone’ in a
weaker sense, one that bypasses connotations of ultimacy or of a super-reality into which we are
subsumed. This understanding might see us view the ‘thus-gone’ as a person that has ‘gone’
from one understanding of reality to another, more nuanced understanding. This would tally
with Ireland’s (1977: 161) argument that usually directional or temporal words such as
tadāyatanam (translated in numerous ways, usually as ‘region’ or dimension’) – which are
present in such places as Ud. 8.1 – can be understood as ‘state’. If we translate ‘tadāyatanam’
in this manner, then we can say in a weak sense that a Buddhist practitioner can ‘go’ from one
understanding of the world to another. In such an instance, a tathāgata is somebody that has
‘thus-gone’ from delusion to awakening, but in a very specific way. They have not ‘gone’ to an
ultimate plane, nor have they been subsumed into the ātman-Brahman. They have simply
changed their understanding of the phenomena of the world. Gombrich regards all of the above
as ‘fanciful’ (2009: 151), and instead suggests that we understand tathāgata as ‘thus-being’, or
state in which conceptualisation ceases without grasping at or endorsing the view that there is a specific ‘place’ in
which this occurs. I hope it is becoming obvious by this point that in speaking of some ultimate ‘place’ where
something happens, we add discernments, and discernments inevitably lead to reification and so dissatisfaction
(sentiments echoed above by Sujato).
P a g e | 218
‘one who is like that’, which is to say that the tathāgata is a certain way, but that it is unable to
be sufficiently described (2009: 151).
This is an idea that I think is mirrored in Nāgārjuna’s philosophy. According to his work in
the MMK, it cannot possibly be the case that we ‘go’ to liberation in any absolutist sense. Instead,
it is implied at MMK 25.19-20 that the concepts of saṃsāra and nirvāṇa are epistemic rather that
physical, metaphysical, or otherwise hyperphysical.
The interpretation of tathāgata as
somebody that has ‘gone’ elsewhere (be it a mental plane or somewhere else) appears to tally
with the interpretations that characterise dharmakāya as a ‘cosmic body’ or as an otherwise
existent Absolute to be entered into. Bhattacharya does note that he agrees with Nāgārjuna that
the basis of liberation is epistemic and not ontological. He writes that ‘[o]ne who has realized
the ātman is not outside the world, but he looks upon the world with new eyes’ (2015: 13), and
so agrees in principle with the idea that liberation is a matter of outlook and understanding rather
than a shift in ontological status. There is still an ontological claim being made, though.
Fundamental to Bhattacharya’s outlook is that there is a Real thing that we eventually see clearly,
namely the ātman-Brahman. We saw in §2 that there are good reasons to believe that early
Abhidharma Buddhists would outright reject any notion of a unitary super-reality underpinning
existence. We have similarly seen that some schools of thought in the Mahāyāna would indeed
interpret dharmakāya in absolutist terms – something that Tillemans calls ‘a major philosophical
idea in Buddhist scholasticism’ (Tillemans, 2019: 641, note 8).’ However, my point is that not all
Mahāyānists would or should subscribe to this position, despite Bhattacharya’s implications to
the contrary (2015: 39). I contend that Nāgārjuna does not (indeed, cannot) hold such a view.
P a g e | 219
The MMK is replete with examples that make monistic, ultimate or Absolute
interpretations of key Buddhist terms unlikely if not impossible, but perhaps the most salient in
the context of nirvāṇa are at MMK 25.19-20 and 25.24 given below:
na saṃsārasya nirvāṇāt kiṃ cid asti viśeṣaṇam /
na nirvāṇasya saṃsārāt kiṃ cid asti viśeṣaṇam //19//
nirvāṇasya ca yā koṭiḥ koṭiḥ saṃsaraṇasya ca /
na tayor antaraṃ kiṃ cit susūkṣmam api vidyante //20//
There exists no difference whatsoever between saṃsāra and
nirvāṇa /
There exists no difference whatsoever between nirvāṇa and
saṃsāra //19//
The limit of nirvāṇa [is the] limit of saṃsāra /
There is not even the subtlest difference between them //20//
sarvopalambhopaśamaḥ prapan͂copaśamaḥ śivaḥ /
na kva cit kasyacit dharmo buddhena deśitaḥ //24//
The extinguishing of all cognition, the extinguishing of reification,
is blissful /
No Dharma [was] ever taught by the Buddha to anyone //24//
If – as it looks here – it is the halting of cognition that characterises nirvāṇa, then it does
indeed appear to be the case that the tathāgata is somebody that has made some sort of mental
shift: they have halted reification and thus halted duḥkha. Indeed, this point is initially laid out
in the MMK’s maṅgalaśloka, where Nāgārjuna writes that the Buddha taught the benevolent
pacification of prapañca (prapañcopaśamaṃ śivam). This halting of reification, the pacification
of conceptual proliferation, is, it seems, the liberated state according to the Mādhyamika: it is a
way of being. We might be tempted to construe an Advaitic slant to this. We have already said
that Bhattacharya supports the idea that liberation consists in a mental shift, in seeing the world
differently rather than in going to a new plane of existence (or other such fantastical notions).
P a g e | 220
The difference between the position of Nāgārjuna and the position of Bhattacharya is then that
Bhattacharya argues that this mental shift occurs in the light of recognising the absolute monism
of Reality, in recognising and, in fact becoming the ātman-Brahman (2015: 21). Nāgārjuna’s
starting premise is to deny that any such monistic or pluralistic ultimate existence is possible:
such an entity requires an intrinsic, immutable nature that Nāgārjuna thinks impossible. The
difference in starting points could not be starker! We shall come to see that whilst Bhattacharya
declares that ‘[t]he belief in the existence of an eternal I is folly, but the belief in the nonexistence of an I is even greater folly’ (2015: 25), Nāgārjuna (MMK 18.6) declares that the
question itself is folly:
ātmety api prajn͂apitam anātmety api deśitam /
buddhair nātmā na cānātmā kaścid ity api deśitam //6//
‘Self’ is disclosed [and] ‘not-self’ is taught
by buddhas: neither ‘self’ nor ‘not-self’ is also taught //6//
This seems paradoxical, but there are good reasons for this position. It is not some sort
of linguistic trick on behalf of the Mādhyamika. It is a terse statement intended to impress upon
the reader that from an ultimate perspective, neither self nor not-self is an appropriate
discernment. The key point that overarches even this position is that there can be no ultimate
perspective! This might initially sound confusing, but the pedagogical intent is easily analysed.
There are many examples of the Buddha conveying his teachings in a manner that could
reasonably be said to imply a substantial self (ātman) – Siderits and Katsura accurately opine that
this is a fact which is ‘generally acknowledged’ by most Buddhist scholars (2013: 199). This is, we
can say, to tailor the Buddha’s eventual position toward listeners that might not even have
believed in karma – if they did not believe in karma then it is likely that they had an
P a g e | 221
underdeveloped sense of moral responsibility. In such cases, it is helpful for the Buddha to foster
a provisional belief in ātman in order to orientate the listener towards developing their sense of
personal responsibility. This would be classed as an example of the Buddha’s upāya, of his skillin-means when teaching. He is simply taking the long view: he nurtures an understanding of
ātman early on so that the listener becomes practitioner. Once listeners become practitioners,
they begin their journey towards understanding not-self: a view that aims to cut through the
fetters of attachment associated with a strong belief in a substantial self or I.
The basic idea here is that at each step, the Buddha is replacing damaging beliefs with
less damaging ones: each view is amended via a process of sublation that is facilitated by the
Buddha’s graduated or progressive teachings. A belief in ātman – though misguided – is better
than a belief in nothing and an abandonment of virtue. A person that believes in ātman will
generally believe that their good deeds and bad deeds do not die with them, but will condition
their future lives. This, in theory, impels the listener to take responsibility and reorientate their
life from a degenerate towards a virtuous existence. Later, the Buddha can teach that this belief
is itself misguided, and that belief in a substantial self – although better than degeneracy and
belief in nothing at all – is to be abandoned if the practitioner is to make spiritual progress.
Then there are passages that either disregard or outright negate the substantial self. We
have seen that Bhattacharya claims at numerous points that the Buddha’s apparent disregard of
the ātman (he never seems to speak about it directly) might be because ‘the Upaniṣads had
already spoken enough about it’ (2015: 37). He claims that the negations of ātman present in
the canonical Buddhist literature provide a tacit affirmation that ‘the Absolute is the sole Reality’
P a g e | 222
(2015: 35).
We might find some support for this position within some Buddhist literature:
Bhattacharya specifically appeals to the Ratnagotravibhāga.
The Ratnagotravibhāga is, as Richard King writes, ‘unique as the only Indian śāstra
devoted specifically to an exposition of the tathāgatagarbha doctrine’ (1995: 217). The usual
translation of tathāgatagarbha is buddha-nature or sometimes buddha-seed, which should give
some insight into what the concept alludes to, viz. the potentiality to become awakened. That
Bhattacharya would appeal to such a text is unsurprising given the controversy surrounding the
Buddhist literature dealing with buddha-nature – there is precedent even within Buddhist schools
for the idea to be taken in an absolutist way. As implied in the translation, there is a tendency to
associate buddha-nature with one of two things. First, with the notion of an embryonic seed
contained somewhere within us (or, perhaps, identical with us), a real, actual potential to
become enlightened. Secondly, the idea that the buddha-nature reflects our natural state, which
has simply become obscured by various imperfections and reifications – this is in a similar vein
to the luminous mind (Skt.: prakṛtiprabhāsvaracitta; Pāli: pabhassaracitta) referred to by the
Buddha in the Aṅguttara Nikāya (1.49-52). This would mean that the dharmakāya (which is to
be obtained via the removal of defilements) is present all along and is simply uncovered once
defilements are removed – this certainly looks to have parallels to the sort of ātman that
Bhattacharya thinks that Buddhism endorses. Both of these concepts are on these terms
amenable to his agenda, though to differing degrees. First, we might think that the idea that we
have within us the potential to awaken is an uncontroversial point that any Vedāntin should
endorse, for what is the point in Vedāntic practice if one does not gain knowledge of the ātmanBrahman? The potential to gain this knowledge must presuppose the act of actually gaining the
P a g e | 223
knowledge required for liberation: it is there in all of us whether we take the opportunity to
realise it or not. This gains additional traction if we understand buddha-nature to be identical
with sentient beings.171 Whilst Bhattacharya’s interpretation appears to have some basis, there
is at least one scriptural explanation for the apparent absolutist bent of tathāgatagarbha
literature such as the Ratnagotravibhāga. Eltschinger (2013: 46) translates an extract from the
Laṅkāvatāra Sūtra (78.5-79.9) where the Buddha explains that absolutist teachings in Buddhist
literature are not meant to be taken at face value. 172 They are, in fact, different from the
teachings of the ‘outsiders’ (Hindu teachings relating to ātman and Brahman) because
tathāgatagarbha is selflessness. Here is part the extract as translated by Eltschinger:
For this reason, O Mahātami, the teaching of the tathāgatagarbha
differs from the outsiders’ teaching of self. Thus by teaching the
tathāgatagarbha, [the tathāgatas] teach (°upadeśena nirdiśanti)
the tathāgatagarbha in order to attract the outsiders who adhere
to the doctrine(s) of the self. How indeed could they quickly
awaken to the supreme perfect awakening, [those persons] whose
thought (āśaya) has fallen into the false view [consisting in] the
erroneous concept of the self [and] who possess a thought that has
fallen from the domain of the three [doors] to liberation? It is for
this purpose, O Mahātami, that the tathāgatas, arhats [and]
perfectly awakened ones teach the tathāgatagarbha. Therefore,
King cites Grosnick (1979: 52): Grosnick argues that ‘”all beings are the garbhas of the Tathagāta,” where the
tathāgatagarbha is a tatpuruṣa compound that is the original meaning of the term’ (King, 1995: 305). This
illustrates that in practice as well as theory, the precise definition of tathāgatagarbha is open to debate, though it
must also be noted that Grosnick himself strongly resisted the tendency to classify tathāgatagarbha as some sort
of monistic absolute (King, 1995: 311).
171
172
Eltschinger also notes that there is a curious lack of both Buddhist and Hindu sources referring to the
tathāgatagarbha doctrine as endorsing a self (2013: 44). One might expect such a thing to be seized upon by
opponents both in rival religious movements and from opponents in rival Buddhist sects: this occurred with the
infamous pudgalavādins, their doctrine known to other Buddhists as a heresy because it endorsed a sort of ātman.
It is perhaps telling that there does not seem to be any arguments made from either side that the
tathāgatagarbha texts are heretical on account of endorsing some sort of ātman.
P a g e | 224
this teaching [of theirs] differs from the outsiders’ doctrine(s) of the
self. And hence, O Mahātami, you ought to follow selflessness173
which is [nothing but] the tathāgatagarbha in order to go beyond
the outsiders’ false view(s).
We can see here that there is more to the tathāgatagarbha literature than immediately
meets the eye. The claim is that the tathāgatagarbha teachings are little more than a device for
proselytisation: they are taught to appeal not to Buddhists, but to other religious people for
whom inquiry into the nature of ātman is liberating.
Such a person might read the
tathāgatagarbha literature and see links between the doctrine they currently adhere to and this
form of Buddhism; this might then make it easier for them to convert to Buddhism! Eltschinger
summarises thus: ‘[i]n other words, if the tathāgatagarbha teachings are meant as an expedient
device aimed at proselytizing the substantialist non-Buddhists, emptiness and selflessness aim at
diverting the “converted” from false conceptions of, and excessive attachment to the self’ (2013:
51). It can be characterised, then, as an attempt to enhance the middle way.
Such a view can be further supported by an interesting extract from the Mahāparinirvāṇa
Sūtra, which states that despite appearances, tathāgatagarbha teachings do not advance a belief
in a permanent ātman. Eltschinger (2013: 60) translates from the Tibetan:
If what is called “self” were an eternally permanent (kūṭasthanitya)
dharma, there would be no freedom from suffering (duḥkha). And
if what is called “self” did not exist, pure religious conduct
(brahmacarya) would be of no avail […] It is to be known that the
buddha-nature is the middle way (madhyamā pratipat) altogether
free from the two extremes (antadvaya) […] Non-duality is reality:
by nature self and not-self are without duality (gn͂is su med ma).
The Lord Buddha has thus affirmed that the meaning of the
173
nairātmya; selflessness, impersonality, insubstantiality.
P a g e | 225
tathāgata is unfathomable […] In the Prajñāpāramitā-Sūtra also I
have already taught that self and not-self are without duality by
characteristic.
The general thrust of this extract as I read it is that we ought not to cling to notions of
self, not-self, both self and not-self, neither self nor not-self. The point is not to cling to such
notions at all, not to endorse one or the other over and above one or the other. This clearly ties
into Nāgārjuna’s whole method, and is, I contend directed not toward the personal ātman (jīva)
or even the ātman-Brahman, but simply with the ending of conceptual proliferation (prapañca).
The idea is not to arrive at knowledge of some ultimately existent substratum akin to the ātmanBrahman, but to arrive at a point where we no longer impose conceptualisations on our
experience. Non-duality is, I think, to be experienced in meditation; we saw in other sections
that this is also the case for Vasubandhu. The whole endeavour then strikes me not as an exercise
in ontology, but in epistemology: how ought we to understand the world around us – how should
we think about and engage with the world around us – if we are to do away with dissatisfaction?
I think that this is the crux of the method, and to this end, it is cognate with the other Buddhist
methods discussed thus far.
Is, as the above sūtra claims, this also the case for the tathāgatagarbha and
prajñāpāramitā literature? Dan Lusthaus offers a short but significant discussion regarding how
best to understand the prajñāpāramitā and tathāgatagarbha literature. The problem as he sees
it amounts to two different conceptions of liberation: sudden and gradual. For Lusthaus, the
Buddhist traditions that tend toward essentialism (and so the traditions that Bhattacharya would
point to in defence of his thesis) do so because their understanding of liberation is one that sees
jñāna as ‘the means or agent for attaining some-thing which in itself is impervious to or
P a g e | 226
indifferent to the vicissitudes of epistemological approaches, though made accessible through
such approaches’ (2002: 256). This is the ‘sudden’ view of liberation, so called because if ‘a readymade transcendental realm already exists, then what is essential about Awakening remains
entirely separate from temporal considerations’ (2002: 257). There is no necessary unfolding of
the process within a temporal timeframe.
The jñāna required for entry into such a
transcendental realm can simply manifest, causing immediate entry into the liberated state. In
other words, the right type of intuitive knowledge makes available to us some sort of ‘thing’ that
constitutes the ultimate reality or ultimate truth (or both). Such a thing would, in virtue of its
ultimacy, be considered eternal, immutable, and intrinsically existent. This obviously would apply
to the ātman-Brahman, but it can as we have seen also apply – though not necessarily – to
Buddhist concepts like the dharmakāya. Particular knowledge would then unlock access to this
ultimately existent thing and provide liberation.
This stands in contrast to a gradual process, which sees jñāna acquired over the course of
what amounts to the Bodhisattva path. In such an account, Lusthaus argues that there occurs ‘a
progressional unfolding that never posits anything apart from the process itself’ (2002: 257). In
other words, there is not any ‘realm’ of the enlightened, nor is there an essential, permanent
feature of awakening. Instead, there is simply a temporal process of Buddhist praxis; we practice,
gradually ‘awaken’, and then act in the world according to the jñāna that we have acquired via
this same gradual process. I think of the difference in very simple terms: for essentialists,
liberation is ontological. It involves ‘entry’ to some eternal state, a ready-made realm. For
progressionalists, liberation is epistemological. Prajñā is that which brings about a mental
change, and this change informs how we interact with the world. If we interact without grasping,
P a g e | 227
without imposition and appropriation, we might be said to be awakened. As Lusthaus writes, in
such a state, ‘[i]mpermanence (e.g., in terms of rūpa) continues unabated, but it is now upekṣa,
i.e., no longer experienced as loss. Amata [deathless, deathlessness] is thus an epistemic change,
not ontological’ (2002: 265).174 Understood in this manner, prajñāpāramitā literature does not
endorse an ātman-Brahman, because such an endorsement requires not only an epistemological
aspect, but also an ontological aspect: everything is Brahman! Similarly, tathāgatagarbha
literature can simply be understood as endorsing the mental potential to gain the relevant insight
to overcome duḥkha. It need not necessitate an Absolute into which we meld, or about which
we gain gnosis. It is, as Lusthaus (2005: 255) writes, the ‘active functioning of one's potential for
Buddhahood.’ It can be no more and no less than this. It would seem very odd to view our
potential as residing in an Absolute, or indeed to view it as the Absolute itself.
It seems clear to me, then, that whilst some Buddhist sects have indeed interpreted both
tathāgatagarbha and prajñāpāramitā literature to have essentialist, absolutist ends, such
conclusions need not be a necessary outcome of the literature’s analysis. In the preceding
Lusthaus (2002: 265) makes a compelling argument that the early Buddhist goal of becoming ‘amata’
(deathless) be understood in the following way:
174
To be undying means to not arise, to not arise means 'unborn' (anutpāda) in the
sense of not constrained by conditioning, not condemned to habitually repeat
previous experience, to have one's experience determined by the moment by
moment 'arising of conditions' (utpāda). "I don't die" means (1) there is no self
which undergoes death (but this is strictly formulaic, not existential), and (2) it
connotes asaṃskṛta, in the sense that Chinese translated that term, viz. wu-wei,
i.e., the non-conditioned spontaneity and freedom of tzu-jan. Further, it meant
to not suffer loss due to impermanence, to remain unaffected by loss and gain.
Impermanence (e.g., in terms of rūpa) continues unabated, but it is now
upekṣa, i.e., no longer experienced as loss. Amata is thus an epistemic change,
not ontological.
P a g e | 228
section, I have outlined some alternative ways of understanding key terms that lend themselves
to reification as synonyms for a Buddhist Absolute. The crux of the matter involves eschewing
ontological concerns in favour of epistemological ones. It is by doing this that I think that we can
make sense of what are some challenging, difficult themes in challenging, seemingly inconsistent
texts. If we interpret the more controversial aspects of the literature in these ways, it seems
clear to me that we must deny that the resulting conclusions endorse any sort of ontological
Absolute.
Bhattacharya’s interpretations that contend that Madhyamaka, Yogācāra,
tathāgatagarbha, and prajñāpāramitā literature lead to a Buddhist endorsement of the ātmanBrahman are then rejected.
P a g e | 229
§6: Nāgārjuna’s Approach
We have seen thus far that if we interpret aspects of the Mahāyāna in certain ways, there
is no need to arrive at a conception of an Absolute underpinning and running through our
experienced reality. That is to say, interpreted in the ways in which I have outlined, there are
good reasons to doubt Bhattacharya’s thesis that Mahāyāna philosophies result in endorsement
of the ātman-Brahman as the material and efficient cause of the world. Nāgārjuna has played a
role throughout the previous discussion, but only partially and relatively fleetingly. My next task,
then, is to assess whether or not Nāgārjuna’s Mūlamadhyamakakārikā can be read in such a way
as to endorse the ātman-Brahman.
The claim by Bhattacharya is, as we have seen, that śūnyatā – Nāgārjuna’s principal
philosophical tool – is the ātman-Brahman (2015: 13). The argument for this is principally that
because the ātman-Brahman is ultimately beyond characterisation, beyond attributes, and
impersonal, then Nāgārjuna’s arguments in the MMK that conclude with the impersonality and
emptiness (of intrinsic nature) of all phenomena is, in effect, an argument for the ātmanBrahman. Mādhyamikas disagree, often claiming that this is to misconstrue emptiness, and also
to overstate the case: Nāgārjuna simply doesn’t make any claims of his own!
David Burton, however, takes the view that Nāgārjuna does indeed make claims of his
own. His point is that Nāgārjuna must make a knowledge claim regarding how things really are
whenever he talks about śūnyatā (1999: 37). In other words, if Nāgārjuna is claiming that all
phenomena are empty of intrinsic existence (svabhāva), then he is making some sort of claim
about the workings of the world. The interesting thing here is the way Burton characterises
P a g e | 230
Nāgārjuna’s claims. For Burton, the knowledge-claim made by Nāgārjuna can be characterised
in two distinct ways. First, we can understand śūnyatā as an ultimate truth that accurately
corresponds to an ultimate reality; in other words, it is the true nature of phenomena to lack
intrinsic nature. If śūnyatā is an ultimate reality, we can make a link with the claim advanced by
Bhattacharya that it is the nature of the world to be impersonal.
Second, Burton refers to a different sort of understanding that dictates that ‘one knows
that one does not know and cannot know how things actually are’ (1999: 37), which we might in
turn characterise as viewing śūnyatā as an inexpressible thing that is ultimately beyond our
understanding. On either of these views, the claimant is obviously making some sort of
assessment regarding how the world actually is – they are either claiming that there is an
underlying nature, śūnyatā, or they are claiming that we cannot know if there is an underlying
nature.
The focus of this point for Burton is Nāgārjuna’s relatively controversial claim at MMK
13.8 that those for whom emptiness is a ‘view’ are incurable.175 Burton contends that contrary
to some prominent interpretations (see Siderits & Katsura, 2013: 145), a person is not ‘incurable’
simply in virtue of taking śūnyatā as some sort of metaphysical view. On this reading, the
positions outlined above – mere knowledge claims, taking emptiness as a view in a weakened
MMK 13.8: śūnyatā sarvadṛṣṭīnāṃ proktā niḥsaraṇaṃ jinaiḥ / yeṣāṃ tu śūnyatādṛṣṭis tān asādhyān babhāṣire //
175
We can understand this as something like ‘emptiness is declared by the conquerors [as the] remedy to remove all
views. But those for whom emptiness is a view are said [to be] incurable.’ It is with what actually constitutes
taking emptiness as a view that Burton is here concerned.
P a g e | 231
sense – are not the issue at hand for Nāgārjuna. Instead, Burton argues that to become
‘incurable’ is to specifically fall ‘into either of the two extremes of nihilism or essentialism’ (1999:
37). The nihilist would simply take emptiness to mean that no entities exist. Burton writes that
according to his view, the nihilist would believe that no entities whatsoever exist rather than the
more subtle claim that no entities with svabhāva exist (1999: 37).
On this first account, Nāgārjuna is a nihilist if he denies outright that any entities exist; a
position that he ostensibly does not appear to advance. 176 Nevertheless, Burton will indeed
conclude that the logical conclusion of Nāgārjuna’s philosophy is nihilism, whether he intended
it or not.177 Of course, if Nāgārjuna is indeed a nihilist, then Bhattacharya’s argument fails – at
least if we adhere specifically to Nāgārjuna’s Madhyamaka as found in his writings rather than to
later adaptations and evolutions of the school as found elsewhere. After all, how might
Nāgārjuna’s Madhyamaka assert or endorse an Absolute if nothing at all exists?
For Burton, the emptiness-eternalist has two possible roads that they might traverse:
gross eternalism and subtle eternalism. The gross eternalist misunderstands śūnyatā as an
unchanging, ultimately existent Absolute. Burton surmises that given Nāgārjuna’s assertion at
But also a conclusion that would preclude him from advancing any sort of support for or belief in the ātmanBrahman!
176
This conclusion is first outlined in Burton’s introduction, where he writes that ‘I believe that the knowledgeclaim that all entities lack svabhāva entails nihilism’ (1999: 4). Giuseppe Ferraro also advanced this thesis in his
2013 paper, A Criticism of M. Siderits and J. L. Garfield's 'Semantic Interpretation' of Nāgārjuna's Theory of Two
Truths. Ferraro later adapted his position to argue that Nāgārjuna is not a nihilist, but rather occupies a realist
anti-metaphysical position. This position, thinks Ferraro, ‘seems to hinder and prevent the possibility of any
nihilistic interpretation of Nāgārjuna’ (2017: 73).
177
P a g e | 232
MMK 25.19 that there is no difference between saṃsāra and nirvāṇa, this grossly Absolutist view
is obviously false. It could be the case, though, that ‘nirvāṇa is the true nature of cyclic existence,
viz. its absence of svabhāva, rather than being a non-dependently originating Absolute Reality
which is separate from the dependently originating world’ (1999: 38). 178 Subtle eternalism,
opines Burton, could mean that whilst ‘emptiness means entities dependently originate, it might
not be understood that emptiness also means that dependently originated entities are one and
all conceptual constructs’ (1999: 38).
Burton’s concern looks to be that the ‘incurable’
practitioner fails to realise that all dependently originated phenomena are in fact conceptual
constructs. He elsewhere claims that conceptual constructs have no foundational existence
(dravyasat) (1999: 35), and so I think his point is that a Buddhist could feasibly hold the opinion
that śūnyatā is equivalent with pratītyasamutpāda but simultaneously assign some sort of real
existence to the dependently originated entities that they encounter.179 This would not be too
dissimilar to the Sarvāstivādin dharma theory insofar as it allows for pratītyasamutpāda but only
if the entities that dependently originate are reducible down to immutable, ultimately existent
dharmas.180
178
Thus, it seems that for Burton, the idea that a Madhyamaka Absolute might be equivalent with the Advaitin
ātman-Brahman is fundamentally flawed: one would need to be characterised by a lack of any svabhāva, whereas
the other is characterised by a substantial existence (thus a svabhāva).
Though Nāgārjuna does initially appear to equate śūnyatā with pratītyasamutpāda at MMK 24.18, the following
line adds a caveat, namely that śūnyatā is a dependent concept (prajn͂apti). This is to say that emptiness is not
equivalent with dependent origination because they are the same thing, but to say that emptiness follows from
dependent origination: it is an example of its operation (Siderits & Katsura, 2013: 278).
179
This reduction is said to preserve the principle of pratītyasamutpāda because all phenomena dependently
originate based upon the basic dharmas, i.e. all experienced things dependently originate as a result of interactions
of dharmas; all phenomena dependently originate apart from the dharmas.
180
P a g e | 233
Talk of a ‘true nature of cyclic existence’ is interesting, because it implies that the idea
that we have of said nature is currently false. Burton’s claim is that the misunderstanding that a
gross emptiness-eternalist might hold could be one that acknowledges that the svabhāva view is
false and contends that the impersonal niḥsvabhāva (lacking intrinsic existence) conception of
nirvāṇa is actually the true nature of reality. This is problematic for Nāgārjuna because what is
in effect occurring is that the practitioner is, in their zeal to deny svabhāva, simply replacing it
with another reified concept under which nothing falls (śūnyatā). The effect is a straightforward
substitution of one ‘real’ nature with another sort of ‘real’ nature, which is by Nāgārjuna’s
analysis both undesirable and impossible. Nāgārjuna would, it seems to me, prefer to do away
with the idea of ‘real’ natures in toto, viewing them as nothing more than reifications. It is in this
sense that scholars often interpret Nāgārjuna as having ‘no thesis’ – it is not that he wants to
state how things ultimately are; he wants to dispel any such idea. In order to try to achieve this,
Nāgārjuna then states instead how things are not. It is clear how the objections brought earlier
by Bhattacharya might also be brought by Burton: in saying how things are not, Nāgārjuna is at
the very least implying how things are. We have seen that the idea is that there can be no
negation without that which is to be negated. I think that this point would carry more weight if
it could be demonstrated that Nāgārjuna is actually interested in defending any sort of
metaphysic, but it seems to me that Nāgārjuna is in the business of removing metaphysics from
the discourse (owing to its potential for reification). In such a scenario, we might want to claim
that Nāgārjuna is less concerned with providing an account of the world as it is, and more
concerned with analysing the world as it appears to us and as we interact with it. After all,
P a g e | 234
Nāgārjuna at no point denies that we experience and interact with a ‘reality’, he only denies that
we can think of it in essentialist, substantial terms (i.e. in terms of svabhāva).
Nevertheless, Burton is suspicious of the ‘no-thesis’ position attributed to Madhyamaka.
It is the case, for example, that Nāgārjuna gives what we are to assume is an exhaustive account
of (ultimate) origination at MMK 1.1:
na svato nāpi parato na dvābhyāṃ nāpy ahetutaḥ /
utpannā jātu vidyante bhāvāḥ kva cana ke cana //1//
Not from itself, not from another, not from both, nor without
cause:
Never in any way is there any existing thing that has arisen //1//
I think that Burton is certainly correct when he says that this account of origination has a
specific aim in mind, namely to illustrate that ‘entities with svabhāva do not originate in any of
these four ways’ (1999: 40). The way in which Nāgārjuna makes these claims, thinks Burton,
means that Nāgārjuna is not a sceptic. Nāgārjuna is emphatically arguing that ultimate entities
do not originate in this manner; he thinks that the evidence available to us when we analyse the
world logically proves to us that this is the case. For Burton, this is not the position of a sceptic.181
I make the point, however, that Nāgārjuna’s claims about how things are not (that entities with
svabhāva do no originate in this way; that entities with svabhāva could not originate whatsoever
(MMK 1.13-14)) might not fall under the scope of ultimate truth. Indeed there are those that
hold the position that we are not entitled to infer anything at all from Nāgārjuna’s negations. So
I mentioned earlier that according to Mohanta (1997: 53), and by drawing on Ayer (1956: 40), Nāgārjuna can be
characterised as a specific type of sceptic: a cognitive sceptic.
181
P a g e | 235
much of Nāgārjuna’s writings depend on a nuanced understanding of the differences between
the scopes of conventional and ultimate truths that it would be folly merely to assume that any
apparent knowledge claims – whatever they might be – ought to be automatically treated as
ultimate claims describing how things really are. Ferraro recently published a novel take on this,
arguing that Nāgārjuna’s philosophy as presented in the MMK and VV is realistic ‘insofar as it
acknowledges – in line with the Buddha’s position – the existence of a reality in itself’ (2017: 79).
This reality could in principle be the object of an ‘ultimate’ cognitive experience, or to phrase it
slightly differently, this reality could be the object of the epistemic transformation that
Nāgārjuna’s writings direct us towards.
Ferraro’s reading here seems uncontroversial – given that Nāgārjuna spends so much
time arguing against immutable intrinsic natures in favour of radical change (to facilitate
pratītyasamutpāda and thus the Buddhist path generally), I think it relatively straightforward that
Nāgārjuna acknowledges some sort of reality (and so is not a nihilist in the way argued by scholars
such as Burton (2001)). His problem then is not with the fact that we experience a reality and
operate within it, but rather with how we interact, analyse and deal with the reality in front of
us. It is not that he denies that we operate in this world, but rather that he denies that we
operate in a specific conception of this world, viz. the world consisting of immutable, permanent,
intrinsically existent (with svabhāva) dharmas. Such a vision of the world is, on Nāgārjuna’s
interpretation, imposed onto the world by us as a result of uncritical cognition. 182 It is then an
Throughout the MMK and at certain points in the VV (some translated and discussed in this work), Nāgārjuna
laments that we impose onto the world the views that we experience various types of existence or nonexistence.
182
P a g e | 236
issue of epistemology rather than one of ontology – Nāgārjuna’s anti-metaphysical bent suggests,
I think, that he is really in the business of preventing us imposing a metaphysic onto the world
that we experience, and this is very simply because doing so breeds attachment (to dṛṣṭi; views,
specifically metaphysical views). Thus, counter-intuitive as it might seem, Nāgārjuna is arguing
that we ought not to impose any notions that might claim to account for how the world ultimately
is (viz. svabhāva). Burton would say that this is still a knowledge-claim, because it at least implies
how the world ultimately is not (without svabhāva). It is, but it is a knowledge-claim of specific
and limited scope. Nāgārjuna – if my reading is correct – wants to deny both the possibility and
utility of describing an ultimate state of affairs. The knock-on effect of this is that in the end, it is
pointless to discuss an ultimate state of affairs, and so whilst talk of an ultimate might in some
instances be a simple pedagogical device (upāya), it is to be eventually abandoned as another
example of dṛṣṭi that insidiously breeds attachment.
Indeed, Ferraro opines that Nāgārjuna’s philosophy ‘is rigorously antimetaphysical to the
extent that it opposes (again in tune with at least one dimension of the Buddha’s teaching) any
attempt to define reality, that is, to project any dṛṣṭi on it or, in other words, to construct any
possible metaphysics’ (2017: 80). This is the case despite his realist view of the world, i.e. his
view that there is some reality within which we operate, and that we can change our outlook
For example, imposing a svabhāva view is an example of holding a dṛṣṭi. Similarly, imposing śūnyatā onto to world
as an existent thing (bhava) is also a dṛṣṭi. The point, then, is to remove and avoid any further dṛṣṭi!
P a g e | 237
within this reality to become awakened/liberated.183 The point then is that we do not impose
anything onto the world. We instead ‘empty’ it of definitions in order to avoid constructing and
imposing dṛṣṭi. Nevertheless, Ferraro has a nuanced take on what constitutes ‘ultimate truth’.
He writes that the process of ‘emptying’ has no definitional content in relation to ‘ultimate truth’,
and this is precisely because the concept emptiness (śūnyatā) and the process that it entails are
intended not to give an account of reality, but to remove any dṛṣṭi (2017: 80). Ferraro qualifies
this in his footnotes, stating that
[according to this reading] the word satya in the locution
paramārthasatya does not qualify the characteristic of a true
statement (or belief), that is, a statement that corresponds to a real
state of things. Indeed, paramārthasatya is not a specific
representation of some particular state of things. Rather, it is the
(general) way buddhas see reality: a way that certainly corresponds
to how reality is in itself, but that does not equate to a specific
statement or a series of statements.
(2017: 80, note 33)
In other words, Ferraro’s argument is that Nāgārjuna understands paramārthasatya
(absolute truth; the highest truth) in a specific way – a way that differs in a radical way from that
understanding to which other Buddhist schools adhere. On this reading, Nāgārjuna interprets
Ferraro specifies that in his view, Nāgārjuna presupposes that reality exists and that we operate within it. He
does not, however, go any further than that by, for example, aiming to give an ultimate account of the way that
reality actually is:
183
Accepting a broad definition of ‘metaphysics’ (as the one GS probably have in
mind) we should therefore distinguish between a metaphysics1, which simply
presupposes that reality exists, and a metaphysics 2, which offers some kind of
definition of reality itself. On this basis, my interpretation is that in Nāgārjuna
we only find a metaphysics1 and a rejection of all forms of metaphysics2.
(2014: 453)
P a g e | 238
paramārthasatya as a general way of being, the result of the shift in cognition that occurs once
a practitioner has successfully employed the concept of śūnyatā to empty out their conventional
experience (this necessarily includes ‘emptying’ emptiness so as not to reify it as a dṛṣṭi!).184 This
reading also equates paramārthasatya with nirvāṇa (2014: 459; 2017: 79-80); we know that
Nāgārjuna asserted that there is no difference between saṃsāra and nirvāṇa at MMK 25.19, and
so I think that Ferraro is correct in arguing that the distinction – insofar as there is one to be made
– is one of epistemology and not ontology. By this, I mean that if I read Ferraro correctly, then
his understanding of Nāgārjuna tallies with my own understanding of MMK 25.19: there is no
ontological difference between saṃsāra and nirvāṇa precisely because Nāgārjuna believes the
following two things. First, that we should not think of the world in terms of metaphysics because
it actively hinders our soteriological aims – this includes taking emptiness as a view (reification of
a dṛṣṭi). Second, liberation consists in thinking differently about the world, not in there being a
(meta)physical change in the world. The world is how the world is; Nāgārjuna sees no change in
it from when we are in the throes of delusion to when we are awakened, nor does he feel any
urge to impose upon the world some sort of metaphysic. The change must consequently be
internal and cognitive.
It is also the case that saṃvṛtisatya (conventional truth) is not a ‘specific true expression’ of reality; it is instead
a ‘cognitive approach or an epistemic level’ that is common to unenlightened people: put very simply, the
difference between conventional truth and ultimate truth is ‘a general way of seeing things’ (Ferraro, 2017: 80,
note 33). This sort of understanding allows for a similarity between the Advaitin method and that of the
Mādhyamika: liberation consists in a change of outlook. There is still a significant point of divergence, however,
and that is where the Mādhyamika has ‘emptied’ any dṛṣṭi, the Advaitin still holds at least one: the ultimate is a
substantial entity; it is the ātman-Brahman.
184
P a g e | 239
On Ferraro’s reading, then, emptiness is not to be reified as a metaphysical view
whatsoever – a position slightly different from Burton’s, who as we have already said thinks that
to take emptiness as a view is to treat emptiness specifically as a form of eternalism or nihilism.
For Burton it looks as though treating emptiness ‘metaphysically’ is not in itself an issue. Ferraro,
it seems, makes the more fundamental claim that emptiness is not to be imposed as a view (dṛṣṭi)
at all and in any circumstance.185 It is to be utilised and then disregarded. I do not think that
Ferraro’s reading necessitates that Nāgārjuna violate his ‘no-thesis’ maxim in any significant way.
Whilst Nāgārjuna must invariably hold that there is some reality and that there is some way of
thinking about and engaging with it, he emphatically rejects the idea that we should hold any
dṛṣṭi about that reality or the ways in which we interact with it. I submit that this is where the
no-thesis maxim really applies: it is a soteriological tool aimed at removing impositions onto
reality. When Nāgārjuna claims that he has no thesis to pursue, he really means that he has no
horse in the metaphysical race; that he has no metaphysical thesis outside of that which states
we ought not to have a metaphysical thesis! He thinks, then, that we can be in the world and
navigate the world without the imposition of any metaphysic: this is the point of Nāgārjuna’s
method.
6.1 Madhyamaka Metaphysics
Burton thinks that one problematic mystical interpretation of Nāgārjuna sees him trying
to explain the inexpressible through verbal expression. The idea is that while we cannot say
To wit, whereas Burton supposes that there is a certain type of metaphysical view that Nāgārjuna warns
against, Ferraro takes the more foundational stance that Nāgārjuna opposes any metaphysical view (dṛṣṭi) at all.
185
P a g e | 240
anything of the ultimate, we can speak in conventional terms to try to orient ourselves toward
the ultimate: recall that this is the role of scripture according to Śaṅkara and is a tactic employed
by a great many mystics as outlined in §1. Burton further pushes this point when he claims that
Nāgārjuna ‘must be saying, at least, that it is necessary to rely on these [conventional] teachings
to come to know ultimate truth’ (1999: 80). This is not too controversial a position: indeed it is
shared by esteemed Madhyamaka scholar Jay L. Garfield.186 As regards Burton’s first point, it is
not clear to me that Nāgārjuna does attempt to give an account of the inexpressible. Further, it
is not clear to me that he really wants to. I think that Ferraro is, on balance, right when he writes
that despite the ultimate being described by Nāgārjuna at MMK 18.9 as not proliferated by
proliferations (prapañcair aprapañcitam), non-conceptual (nirvikalpam), and undifferentiated
(anānā-artham), this does not mean that the ultimate is ineffable – instead, it just means that it
is different from the conventional. Further, the descriptions of the ultimate here all hinge on
cognitive awareness: not proliferated by proliferations, not cognised via concepts, 187 not
differentiated and so on. Following Ferraro’s argument, these negations (not x, not y) merely
state that the awakened see things not in terms of the conventional (which is populated by
proliferations, which is full of reification, which does rely on conceptual thought). We have
already said that we are not entitled to infer from these negations anything about what the
Garfield writes that ‘the understanding of ultimate truth is in an important sense the understanding of the
nature of the conventional truth’ (Garfield, 1995: 299), hinting at the contingency of the ultimate on the
conventional – to understand the ultimate, we must understand the conventional.
186
187
The term vikalpa can be problematic insofar as there are questions around how we might cognise something
sans-concepts. This is an area of debate to which no justice can be done in the space available here, and so I shall
not attempt to answer this most complex of questions.
P a g e | 241
ultimate is really like. Consequently, Ferraro sums up Nāgārjuna’s approach thus: ‘[o]nce again,
the negations should be interpreted in the prasajya fashion, and do not suggest a metaphysics
counterposed to the one that is rejected’ (2017: 54). The negations present in the writing of
Nāgārjuna, then, should not be taken to imply anything further. This is contrary to the
interpretations of both Bhattacharya (2015: 36) and Murti (2016: 234-235), and is contra the
position offered above by Burton. If Ferraro is correct about Nāgārjuna’s intention (and it seems
to me that his interpretation of Nāgārjuna is at least plausible), then Burton’s concern is
unwarranted. Nāgārjuna simply does not offer any sort of description of an ineffable Absolute –
he is in fact chiefly concerned with the removal of descriptive content. Madhyamaka philosophy
then is decidedly unconcerned with any account of a mystical metaphysic, and there can
consequently be no Madhyamaka account of a mystical experience. If the raison d'être of
Madhyamaka philosophy is indeed to remove any dṛṣṭi then there is simply no reason to think
that Nāgārjuna is ever concerning himself with characterising some ultimate reality, and because
of this, Nāgārjuna is not a mystic.
Now to the second point. I remain unconvinced that Nāgārjuna must commit himself to
using conventional truths to understand ultimate truths. It seems to me that Nāgārjuna simply
recognises that in terms of a soteriology (and let us not forget that Buddhism is at top and bottom
a soteriology before all else), imposing the idea of a metaphysical ‘ultimate’ onto the world is
counter-productive. He thinks that getting rid of essentialist ideas is what equates to liberation
from suffering. This is an epistemic process rather than an ontological one. So, is Nāgārjuna
really saying that understanding conventional truth x will lead to understanding ultimate truth y?
I think that the question itself is mistaken, as I shall now explain.
P a g e | 242
If we accept the claims that ultimate and conventional truths do not provide a ‘specific
true expression’ of the content of ‘reality’ (whatever that might turn out to be), we are left with
the somewhat surprising notion that Nāgārjuna does not really want to describe reality at all.
What he actually wants is to describe the ways in which we tend to engage with reality as-it-is,
and then illustrate how one way is more expedient than the others are if we hope to achieve
liberation from dissatisfaction and suffering. This means that Nāgārjuna is not much concerned
with describing the world, but with describing how we respond to it psychologically. To this end,
Nāgārjuna is making some sort of knowledge-claim, but it is not about reality per se (which is
assumed but never really detailed); it is instead about how we think about reality. Nāgārjuna
thinks that the knowledge he has is in relation to enlightenment, not in relation to the nature of
reality. If liberation is a cognitive shift that allows a certain way of seeing things, then it is enough
for Nāgārjuna to account for this psychologically over and above accounting for ‘reality’, which is
simply assumed. Emptiness is thus a tool for ridding our minds of these insidious essentialist
thought processes that ascribe svabhāva to phenomena and thus impose a metaphysics onto the
world. And yet we would be just as wrong to suppose that emptiness facilitates a lack of being
in a given phenomenon. Ascribing svabhāva is to fall into eternalism, for we have already said
that svabhāva actually ‘is’ that which possesses it. It is the fundamental nature of a given
phenomenon, immutable and permanent. Denying it outright is useful for a person caught up in
essentialist traps, but it is not the full story. A denial of svabhāva is also undesirable – we can
easily fall into the trappings of nihilism, as many Madhyamaka detractors have continually
pointed out. What then should our final position be?
P a g e | 243
Ferraro writes that
the mere position (or presupposition) of an ultimate reality
(accessible to the buddhas) cannot be considered a form of
substantialism or reificationism. Indeed, according to this reading,
phenomena are just conventionally real, without any ontic referent
or substantiality; the very reality in itself could also not be
considered a substance or something endowed with the attribute
of existence. In fact, the non-implicative (prasajya) nature of
Nāgārjuna’s negations prevents us from inferring the substantiality
of ultimate reality from the negation of substantiality of
conventional reality. Actually, the category of being does not
qualify the tattva-paramārtha any more than the category of nonbeing (or the combination or the lack of being and non-being).
(2017: 81)
Ferraro seems to be arguing that on Nāgārjuna’s account, the reality we experience does
not require any substantiality. That is to say that we do not need that reality to exist via svabhāva
or some other substance that is svabhāva in all but name – this is what Nāgārjuna spends so
much time denying.
It is Nāgārjuna’s opinion that imposing such essentialist ideas onto
phenomena is simply to propagate attachment via dṛṣṭi. In turn, this hinders our progress
towards liberation. With such a notion in mind, we might wonder what use talking of the
conventional and ultimate really is. If Nāgārjuna’s philosophical method should be understood
as a concerted effort to remove any possible dṛṣṭi (and I think that it should), then speaking of a
conventional and ultimate way of seeing things might contradict this aim. Ferraro makes the
point that the difference between the conventional and ultimate is ‘a general way of seeing
things’ (Ferraro, 2017: 80, note 33), and so I hazard that for Ferraro, this general way of seeing
things need not have any particular metaphysical import.
P a g e | 244
Nevertheless, there is a question to be answered here. Another part of the semantic
interpretation is concerned with reification of positions, specifically with reification of the idea
of an ‘ultimate truth’ (and so ultimate reality). Indeed, Siderits’ famous (and much contested)
slogan is ‘the ultimate truth is that there is no ultimate truth’ (Siderits, 2016: 24). This requires
a little unpacking, and I will do this as we go along. First, an obvious point: if the ultimate truth
is that there is no ultimate truth, then this thesis appears to run counter to that of Ferraro. We
have seen that Ferraro thinks that there is some sort of ultimate, and it is characterised by a
change in the way we exist in the world. Can these two positions be reconciled? I think there
might be a ‘middle way’.
The scope of the semantic interpretation of emptiness can be summarised as follows:
śūnyatā is, when interpreted semantically, a principle concerned not with the nature of the
world, but with the nature of truth. Siderits writes that ‘[s]pecifically, it takes the claim that all
things are empty to mean that the ultimate truth is that there is no ultimate truth – there is only
conventional truth’ (2003: 11). The strongest defence of this position is, I think, at MMK 25.24,
a short but significant line that would be easily missed if we did not recognise its context:
sarvopalambhopaśamaḥ prapan͂copaśamaḥ śivaḥ /
na kva cit kasyacit dharmo buddhena deśitaḥ //24//
The extinguishing of all cognition, the extinguishing of reification,
is blissful /
No Dharma [was] ever taught by the Buddha to anyone //24//
This is taken from the chapter entitled ‘Analysis of Nirvāṇa’, which Nāgārjuna thinks is
itself neither existent nor ‘an absence’ (abhāva; nonexistent). We know that the Buddha did
indeed teach the Dharma, and so does Nāgārjuna: he ends his magnum opus by offering
P a g e | 245
salutations to the Buddha, whom Nāgārjuna thinks taught the ‘true Dharma’ (saddharma) to
terminate all views (sarvadṛṣṭiprahāṇāya)! There must then be something else going on in this
seemingly innocuous kārikā. I think that Nāgārjuna is thinking here in terms of the ultimate; if
there is no possibility of svabhāva and thus no possibility of an eternal, immutable character or
identity (or, indeed, ātman), then the Dharma of the Buddha is contingent. It is contingent in the
sense that its arising relies on causes and conditions, and that its continued ‘existence’ is reliant
on yet more causes and conditions. 188 These causes and conditions are simple enough to
conceive; the Dharma is taught because there are unenlightened minds attaching themselves to
and grasping at things; they labour under existential dissatisfactions. The continued existence of
these unenlightened minds means that the Dharma endures (conventionally), and once a mind
becomes enlightened, there is no need for the Dharma to exist: owing to its dependent origins,
should every mind become enlightened, the Dharma – like all other phenomena – would perish.
In these ways, the Dharma is dependently originated and exists only conventionally. Accordingly,
the truth of the Dharma as a tool to reach liberation is also thoroughly dependent and so
conventional. It has no innate character, no eternal, intrinsic existence, and so ‘no [permanent,
ultimately existent] Dharma was ever taught by the Buddha to anyone’. In other words, the
Dharma is not ultimately existent and so not ultimately true. It is conventionally existent and so
conventionally true, and it is useful only as long as there are deluded minds to liberate; its truth
is contingent on these minds and their deluded states.
I am here using ‘existence’ in a conventional manner. That is to say that I do not mean to imply that the
Dharma is ultimately existent.
188
P a g e | 246
6.2 Why Posit Conventionality and Ultimacy?
We might again wonder what the point of distinguishing between the conventional and
ultimate is to Nāgārjuna. It holds some significance, because he claims at MMK 24.9 that anybody
that does not understand the distinction between the two truths does not understand the
Dharma as taught by the Buddha! 189 We have seen some wrangling by Ferraro and Burton
around this question, but neither provide a fully satisfactory account. Siderits and Katsura (2013:
272-273) suggest two ways in which the distinction between conventional and ultimate might be
understood. The first is borrowed from Candrakīrti, and understands ‘conventional’ to mean
something along the lines of ‘customary practices of the world’ (2013: 272). Saṃvṛti is thus ‘of
the nature of (the relation between) term and referent, cognition and the cognized, and the like’
(2013: 272), which means that conventionality is simply normal, everyday relations between
semantic and cognitive aspects; between words and their associated cognitive ‘objects’.190 The
189
MMK 24.9 thus reads:
dve satye samupāśritya buddhānāṃ dharmadeśanā /
lokasaṃvṛtisatyaṃ ca satyaṃ ca paramārthataḥ //8//
ye ‘nayor vijānti vibhāgaṃ satyayor dvayoḥ /
te tattvaṃ na vijānanti gambhīre buddhaśāsane //9//
The Dharma-instruction of the Buddha rests on two truths:
conventional truth and ultimate truth //8//
[Those] who do not know the distinction between the two truths,
they do not understand the Buddha’s profound teaching //9//
They are ‘cognitive objects’ because names and so on are prapañca: the old example of a chariot being nothing
above a name assigned to parts arranged chariot-wise illustrates this point. Entities are reifications, they are
semantic constructions assigned to conceptual constructions that occur through cognitive processes. The claim
here is not that things do not exist. Instead, the claim is that we invest things with a sort of existence that they do
not actually have. Things do not exist the way we take them to exist. Every part of every entity has innumerable
causes and conditions – we do not see this reflected in language or the usual ways of thinking about the world.
We invest them with permanence: a chariot is just a chariot, with some svabhāva indicating as much. For thinkers
like Nāgārjuna, it is fine to refer to a chariot, but we must recognise that this is a conceptual construction referring
190
P a g e | 247
underlying point is then that conventional thought is everyday thought, and everyday thought is
in some way mistaken: namely insofar as it tends to reification and conceptual proliferation
(prapañca). At MMK 24.16, Nāgārjuna specifies exactly how this conventional thought manifests
itself: as taking existents to have svabhāva and thus be intrinsically real. This in turn, says
Nāgārjuna, means that conventional thought – thought that reifies – takes objects to be without
causes and conditions (ahetupratyaya). It is an uncritical mindset, a way of thinking about and
engaging with the world; a way that is that is mistaken.
It is here that we see the import of referring to an ultimate truth: it allows a clear
demarcation between thought processes, an old way of seeing things (conventional) and a new
way of seeing things (ultimate). It might help for us to think of śūnyatā as an ultimate truth when
we are caught up in conventionalities and are ascribing svabhāva to that which lacks svabhāva:
it stands as a ‘higher’ principle on the road to liberation precisely because clinging to notions of
‘self’ via intrinsic nature is the nexus of all dissatisfaction. This being the case, we can likely
characterise use of the two truths as part of the Buddha’s skilful means (upāya), as a raft that
helps us cross the proverbial river, but is of no further use when we get across the river and stand
at the foot of the mountain. Nāgārjuna, I think, would have seen the two truths in a similar
manner – he cannot uphold an ultimate truth if he is to remain consistent. It seems to me that
to understand the two truths is to see them as handy ways of demarcating higher and lower
levels of teachings that lead to better or worse ways of engaging with the world. In this sense,
to a collection of dependent causes, conditions and parts. In this sense, it is conventional. It is also in this sense
that it is a ‘cognitive object’. The way we think a chariot exists is owing to a mistaken way of thinking about the
world and the phenomena that interact within it. Precisely because there are no svabhāvas, there are no
independent entities. The thought that there are independent entities is exactly that: a thought.
P a g e | 248
we might think of things termed ‘ultimate truths’ as penultimate teachings! They are the best of
conventional understanding, but are still themselves somehow inadequate – they are not the
final step, but they are about all we can say before taking a final step. Nāgārjuna is quick to point
out that we must reach liberation via recourse to the conventional (MMK 24.10),191 the worldly
means that surround us and that we make use of, the most basic of which is language. This
means that we might class conventional truths as the ‘normal’ way of engaging with the world,
and ultimate truth as a type of upāya that is useful to us, but which is in the final analysis – insofar
as there is one – to be abandoned as mere concept.
Recall that Buddhists from the Abhidharma onwards account for experience via dharmas:
I have elsewhere covered the intricacies of how dharmas were thought by the Ābhidharmikas to
operate, and so will not retrace those steps here. Instead, I offer two final kārikās from Nāgārjuna
providing the bedrock to the argument that all dharmas are empty and thus impermanent and
so there cannot be any ‘ultimate’ entities. At MMK 25.22-23 we find:
śūnyeṣu sarvadharmeṣu kim anantaṃ kim antavat /
kim anantam ca nānantaṃ nāntavac ca kim //22//
kiṃ tad eva kim anyat kiṃ śāśvataṃ kim aśāśvatam /
aśāśvataṃ śāśvataṃ ca kim vā nobhayam apy atha //23//
All dharmas [being] empty, what [is] without end? What has an
end? /
What [is both] with end and without end; what [is neither]
without end nor [having an] end? //22//
On this, Nāgārjuna writes ‘vyavahāram anāśritya paramātho na deśyate’, which we can understand as
something like ‘the ultimate truth is not taught detached from the world of appearances’, viz. liberation must be
described, taught and reached in the conventional world. I think that this lends credence to the notion that for
Mādhyamikas, conventional-ultimate distinctions are simply epistemological tools to aid us on the path rather
than ontologically significant realms.
191
P a g e | 249
What is identical? What is different? What is eternal? What is
noneternal? /
What [is both] eternal and noneternal? What is neither (na
ubhaya)? //23//
The point that Nāgārjuna is making is that given that all psycho-physical dharmas are
empty of intrinsic nature (svabhāva), what sense does it make to talk about entities that are
eternal, noneternal, both or neither? The question is itself mistaken; it is an instance of
imposition, of conceptual proliferation (prapañca), and is ultimately a form of grasping that
complicates the path to liberation. It seems clear then that Nāgārjuna does not see any value in
positing an existent ‘ultimate’. ‘Ultimate’ is used to mark some distinctions when discussing a
difficult subject, but when it comes down to it, there are no entities or states of affairs which fall
under the concept ‘ultimate’. The ‘ultimate’ is a useful construction, designating the precipice at
which linguistic designations and discursive cognition must end. It signifies the edge of the
linguistic world. As with minotaurs and unicorns, it is important only that we understand what
the concept means in relation to other concepts, in this case the ‘conventional’ and all that
entails. It is not important that some entity or state of affairs actually fall under this concept.
This still sounds like a metaphysical claim, and to some degree, it is. But we are by this
point at an advanced stage along the path, and this is important. At this stage, Nāgārjuna is
toying with metaphysics in order to demonstrate the futility of doing metaphysics. When he asks
‘what is identical, what is different? What is eternal, what is non-eternal?’ and so on, Nāgārjuna
is doing much more than engaging in mere sophistry. These questions ought to apply to
everything, and the resulting reductio so advocated by commentators like Candrakīrti should lead
us to stop asking metaphysical questions in toto. This is a controversial thesis, with many scholars
P a g e | 250
arguing that Nāgārjuna advocates nihilism, or that he ends up at some metaphysical viewpoint
somewhere down the line, but I think that both of these positions are wrongheaded. We can see
very clearly that Nāgārjuna does engage in some metaphysics, but he only engages with
metaphysics insofar as he is required to in order to prove the silliness of metaphysical systems.
Accordingly, the metaphysical requirement upon Nāgārjuna really extends only as far as
jettisoning certain commitments and ideas that get in the way of our liberation. This can be seen
with Nāgārjuna’s insistence that emptiness itself must be empty, a tactic clearly aimed at halting
any possible reification of śūnyatā as an ultimate entity. Further than this, Nāgārjuna wants to
halt all conceptual thought. We see at MMK 25.24 that ‘the extinguishing of all cognition, the
extinguishing of reification, is blissful’, and at MMK 27.30 that the Buddha taught the ‘true
Dharma (saddharma) for the abandonment of all views’ (sarvadṛṣṭiprahāṇāya). The ultimate is
a view; it is descriptive not insofar as it designates some ultimately real state of affairs, but insofar
as it describes the means by which the practitioner might reach the penultimate stage to
liberation. Counter-intuitively, the final stage must be the disregarding of concepts of ‘ultimate’
and so on, so that no concepts at all remain.
Even the Four Noble Truths are only conventionally true, being relevant to us at all only
as long as there is origination, cessation and so on (MMK 24.40).192 No Noble Truth is true in and
of itself, but instead emerges in reliance upon those things that cause it and otherwise affect it
in some way. Should dissatisfaction not occur, then we do not need to account for its origination,
192
The truths being 1) dissatisfaction, 2) the origination of dissatisfaction, 3) the cessation of dissatisfaction, and 4)
the means to the cessation of dissatisfaction
P a g e | 251
cessation or the means by which we might expedite its cessation. Once dissatisfaction dissolves,
the Noble Truths no longer hold – how could they? Those that do not understand this type of
intimate connection between śūnyatā and pratītyasamutpāda – a connection that precludes the
existence of any and all ‘ultimate’ entities –
are called ‘dull-witted’ (MMK 24.11;
mandamedhasam) and ‘slow’ (MMK 24.12; manda).
Those of sufficiently sharp mind know, then, that there are no ultimate entities or states
of affairs, and as such, no corresponding ultimate truths. This is how Nāgārjuna characterises his
Middle Path: his position is between the extremes of ultimate existence and ultimate
nonexistence. It is thus clear to me that Nāgārjuna could not possibly have endorsed any entity
or principle analogous to the ātman-Brahman. This would be to reify śūnyatā and, as discussed
earlier, to propagate one more insidious form of self-attachment, which has the unthinkable
consequence of effectively negating Madhyamaka’s raison d’être. As we see at MMK 24.11,
Nāgārjuna thinks that such a position would be advanced only by the dull-witted!
6.3 The Ultimate as a Conventional Designation
Ultimate truth, then, should be thought of as a type of prapañca, although a less
damaging one than those that traditionally fall under the scope of the conventional. It is a
concept under which nothing can fall if we are to account for change via emptiness, and it is a
concept about which speculation is pointless once we reach the final stage of the path to
liberation. It is an imposition with no reality of its own: like everything else, it must be
dependently originated and relies for its communication upon conventional means of
communication. This is of significance if we want to claim that Madhyamaka endorses a
P a g e | 252
permanent, immutable Absolute like the ātman-Brahman. Nāgārjuna is clear that we must
jettison notions of ultimacy at MMK 25.19-24 and 27.29-30. Let us examine these kārikās.193
At MMK 25.19-24, we find:
na saṃsārasya nirvāṇāt kiṃ cid asti viśeṣaṇam /
na nirvāṇasya saṃsārāt kiṃ cid asti viśeṣaṇam //19//
nirvāṇasya ca yā koṭiḥ koṭiḥ saṃsaraṇasya ca /
na tayor antaraṃ kiṃ cit susūkṣmam api vidyate //20//
paraṃ nirodhād antādyāḥ śāśvatādyāś ca dṛṣṭayaḥ /
nirvāṇam aparāntaṃ ca pūrvāntaṃ ca samāśritāḥ //21//
śūnyeṣu sarvadharmeṣu kim anantaṃ antavat /
kim anantam antavac ca nānantaṃ nāntavac ca kim //22//
kiṃ tad eva kim anyat kiṃ śāśvataṃ kim aśāśvatam /
aśāśvataṃ śāśvataṃ ca kim vā nobhayam apy atha //23//
sarvopalambhopaśamaḥ prapan͂copaśamaḥ śivaḥ /
na kva cit kasyacit dharmo buddhena deśitaḥ //24//
There exists no difference whatsoever between saṃsāra and
nirvāṇa /
There exists no difference whatsoever between nirvāṇa and
saṃsāra //19//
The limit of nirvāṇa [is the] limit of saṃsāra /
There is not even the subtlest difference between them //20//
Views concerning [what is] beyond cessation, the end of the
world, and the perpetuity of the world /
depend on [views concerning] nirvāṇa, death, and the future
//21//194
A couple of these kārikās have been mentioned elsewhere in this work. I hope that their repetition will be
forgiven: the overall context is, I think, important.
193
Siderits and Katsura (2013:303) translate ‘nirvāṇa, the future life, and the past life’, stating that this kārikā
relates to the Buddha’s reluctance to answer questions regarding states of being following ‘the cessation of such
194
P a g e | 253
All dharmas [being] empty, what [is] without end? What has an
end? /
What [is both] with end and without end; what [is neither]
without end nor [having an] end? //22//
What is identical? What is different? What is eternal? What is
noneternal? /
What [is both] eternal and noneternal? What is neither (na
ubhaya)? //23//
The extinguishing of all cognition, the extinguishing of reification,
is blissful /
No Dharma [was] ever taught by the Buddha to anyone //24//
Nāgārjuna’s style here can seem cryptic or confusing, but I think we can interpret it in a
relatively straightforward way. First, Nāgārjuna denies the difference between saṃsāra and
nirvāṇa, which is not without controversy if you are a Buddhist that believes in some ultimate
nirvāṇa realm to which entry is gained upon awakening! However, given that Nāgārjuna has by
this point in the text spent so long arguing against the possibility of svabhāva (and so svabhāvic
realms or entities), it is my contention that the best interpretation here is to take Nāgārjuna as
denying the existence of an immutable, ultimately existent nirvāṇa realm. As svabhāva is the
only way by which entities might exist ultimately, it makes no sense for us to say that ultimately,
there is either saṃsāra or nirvāṇa. Their shared ‘limit’ is that they are both conventional: neither
exists with svabhāva. In this sense then there is no difference between the two.
Next, we see that views regarding what is beyond nirvāṇa, about the perpetuity or
finitude of the world and so on are thought by Nāgārjuna to be directly reliant upon views about
composite things as persons, whether the world is limited in space, and whether the world has limits in time.’
Such questions presuppose and answer to whether or not nirvāṇa has a beginning and end.
P a g e | 254
nirvāṇa. Siderits and Katsura are in my opinion correct when they link this kārikā to the
indeterminate questions that the Buddha famously refused to answer (2013: 303). If nirvāṇa has
a beginning or end, then it follows that the other questions must also have discernible answers.
We find Nāgārjuna’s answer to this underlying question at MMK 11.1:
pūrvā prajn͂āyate koṭir nety uvāca mahāmuniḥ /
saṃsāro ‘navarāgro hi nāsyādir nāpi paścimam //1//
The Great Sage [has] said [that] the first point [of saṃsāra] cannot
be known /
Saṃsāra [is] without first and last, [it is] without beginning or end
//1//
The underlying point to all of this is that such questions – even if they could be answered
– are mere distractions. The act of asking them in the first place is misguided: views regarding
one thing depend on views regarding something else, depend on views regarding something else,
and so the chain continues in a circle of dissatisfaction, never finding one single, ultimate source.
It is pointless to look for a beginning or end within saṃsāra, and doing so causes nothing but
anxieties related to wrong views. It is a waste of our energies to focus on these distractions
precisely because even if they could be answered, the resulting knowledge would not aid our
quest for liberation. Nāgārjuna spends a lot of time arguing against the possibility of real
(intrinsically existent) stages of existence. Chapter 11 of the MMK is dedicated to demonstrating
that there are no ‘real’ prior or posterior stages of anything. This is because for there to be real,
distinct stages, we need to account for succession in time, which as Siderits and Katsura explain,
cannot be explained ‘without positing an absolute beginning, a posit which would be irrational’
(2013: 127). Nāgārjuna thinks this to be the case first because his denial of svabhāva also means
P a g e | 255
a denial of absolutes, and second because positing such a thing would violate the laws of
pratītyasamutpāda.195
We see this point really forced in the following lines. Given that svabhāva is impossible
and all dharmas are thus empty, it is nonsensical to talk about beginnings, ends and so on. All
dharmas being empty, there can be no possibility of ultimately existent entities. Nāgārjuna had
hinted at this in Chapter 7 when he wrote that given that conditioned (dharmas) are
unestablished, unconditioned (dharmas) cannot be established either.196 Thus, an underlying
195
MMK 11.3-6 really spells this out. Briefly, the argument is that if there were a first birth that was uncaused by
old age and death, then the traditional account of rebirth would be invalidated: birth would be causeless.
Similarly, if old age and/or death were the first point in the series (of rebirths), then they too would be uncaused
by birth and so ultimately uncaused. Nāgārjuna, like all Buddhists, thinks it nonsensical to talk of uncaused events
or entities. Further, pratītyasamutpāda itself rules out the possibility of absolute uncaused beginning. This same
argument must hold, thinks Nāgārjuna, for all phenomena and entities in all possible circumstances.
196
MMK 7.33-34:
utpādasthitibhaṅgānām asiddher nāsti saṃskṛtam /
saṃskṛtasyāprasiddham ca kathaṃ setsyaty asaṃskṛtam //33//
yathā māyā yathā svapno gandharvanagaraṃ yathā /
tathotpādas tathā sthānaṃ tathā bhaṅga udāhṛtam //34//
[With] origination, duration, and dissolution not established, the conditioned
[does] not exist /
And [with] the conditioned not established, what unconditioned [thing can be]
proven? //33//
Like an illusion, like a dream, like the city of the Gandharvas,
thus origination, duration, dissolution [have been] declared //34//
P a g e | 256
substratum, permanent, immutable, pure and so on cannot exist. The very fact that there is
change in the world precludes its existence. Indeed, at MMK 7.17, Nāgārjuna asks how any nonarisen entity (think of the ātman-Brahman here) can be said to come into being:
yadi kaścid anutpanno bhāvaḥ saṃvidyate kvacit /
utpadyeta sa kiṃ tasmin bhāva utpadyate ‘sati //17//
If any non-arisen entity somewhere exists, [that would] arise /
But [when such an entity] does not exist, how then [can an]
existent arise?
Siderits and Katsura point out that the idea at play here is that entities that have not yet
arisen have some type of proto-existence where it ‘exists’ as an ‘as-yet-unoriginated entity’
(2013: 82); this is, of course, the sort of thing that the Sarvāstivādins believed as a result of their
dharma theory. Nāgārjuna wants to know how an entity in a state of existential limbo can truly
be said to arise given that origination is the coming into existence of that which did not exist – in
any state – before. An action requires an existent thing, and so for something to arise in a real
sense, it needs to already exist. But we have seen that only something new that did not exist in
any state previously can arise, and so Nāgārjuna thinks that we cannot conclude that any dharma
really arises (he makes this clear at MMK 7.29). It is important to realise that Nāgārjuna is not
ever claiming that we do not experience origination, cessation and so on: he is instead disputing
their ultimate existence.
The counter-intuitive conclusion is that the fact that change,
origination, cessation and so on are all experienced in the world means that they cannot exist
ultimately; they cannot exist with svabhāva.
The Advaitin would respond that the ātman-Brahman does not come into existence. It
simply is existence.
I think that Nāgārjuna would find this sort of reasoning entirely
P a g e | 257
unsatisfactory: pratītyasamutpāda dictates that everything is subject to birth, decay, aging, and
destruction, and we have seen that Nāgārjuna thinks that this holds from the top down and in
every conceivable circumstance. This is why he has such a problem with dharmas as conceived
as having svabhāva. We can deploy the argument from MMK 25.23 against the Advaitin Brahman
view: given that Nāgārjuna has demonstrated that entities cannot exist ultimately, how can
anything be described as eternal, non-eternal, both eternal and non-eternal, or neither? The
ātman-Brahman cannot be conceived of in any sense – it is incoherent. Additionally, we saw
Śaṅkara earlier claim that the existence of the ātman-Brahman is in the first instance attested by
the Upaniṣads. I think it uncontroversial to say that Nāgārjuna would not find this compelling:
although there is recourse to Buddhist scripture in every Buddhist sect including Madhyamaka,
we can see that the MMK is concerned not only with repeating Buddhist doctrine, but with
reconciling it with experience. The metaphysics that Nāgārjuna does engage in are based on
experience: we see change, we can analyse causes and conditions and so on. This is a shared
factor with Yogācāra. Nāgārjuna simply uses our experience of the world to demonstrate that
clinging to metaphysical views (dṛṣṭi) is a source of duḥkha: this is the world, we are engaging
with it in the wrong way when we impose all these metaphysical dogmas onto it.
Śūnyatā is not an entity that exists ultimately: it is itself empty of intrinsic existence, it
should not be reified, and it describes a privation (of svabhāva), not an existent thing: the lack of
svabhāva is not some inverse type of permanent character, for this would be to reify emptiness
as a svabhāva in all but name. Reifying an empty concept is for Nāgārjuna still problematic insofar
as doing so has the propensity to bring about insidious types of clinging (and thus dissatisfaction).
It is clear to me that followed to its logical ends, Nāgārjuna’s Madhyamaka cannot in terms of the
P a g e | 258
conventional countenance the assertion of an immutable, necessary entity that is both the
material and efficient cause of the world: it is simply an incoherence that tallies neither with
experience nor with Buddhist doctrine. In terms of the ultimate, such a question cannot arise
because all reification – indeed all conceptualisation – has ended. In such an instance, talk of an
ultimately existent ātman-Brahman is obviously impossible. Talk of an ultimate anything is
impossible.
Now to the final two kārikās of Chapter 27 of the MMK. The context of this chapter
generally is ‘An Analysis of Views’, the views in question being those concerning past and future
existence. Siderits and Katsura write that the thrust of the various arguments contained in the
chapter are ‘meant to refute a wide variety of theories about the ultimate nature of reality. In
each case a key assumption of the theory under attack [is] that there are things with intrinsic
natures [svabhāva]’ (2013: 335). The outcome of the chapter appears to be the conclusion that
emptiness precludes any ‘ultimate’ phenomena from existing. What does Nāgārjuna have to say
on this? At MMK 27.29-30 we find:
atha vā sarvabhāvānāṃ śūnyatvāc chāśvatādayaḥ /
kva kasya katamāḥ kasmāt saṃbhaviṣyanti dṛṣṭayaḥ //29//
sarvadṛṣṭiprahāṇāya yaḥ saddharmam adeśayat /
anukampām upādāya taṃ namasyāmi gautamam //30//
So, since all existents are empty, where, to whom, which, for what
reason would views such as ‘the eternal’ occur?
I pay homage to Gautama, who by means of compassion taught
the true Dharma for the abandonment of all views.
Given that all phenomena are empty of svabhāva, there can from an ultimate perspective
be no place, person, or reason in which, to whom, or according to which views such as eternalism
P a g e | 259
arise. Indeed, the views themselves are also impossible from an ultimate perspective. Of course,
the ‘ultimate perspective’ is itself a mere designation, indicative of an advanced position on the
Buddhist path, but not of the final destination. Holding that some sort of ultimate truth or
ultimate reality is the outcome of the path would itself be a view, and as we see at 27.30, the
Dharma necessitates the removal of all views. This total removal must include the view that
there is an ultimate truth or an ultimate reality, even if this ultimate view is the only view held
by a given person. Nāgārjuna’s aversion to asserting any sort of ultimate is, I think, well founded.
If there is an ultimate mode of existence or some ultimate truth that liberates, it would need to
be permanent. It would, according to Nāgārjuna’s account of ultimacy, need to exist with
svabhāva; an immutable, unoriginated thing detached from dependent origination and
untouched by the very causes and conditions which account for all Buddhist causation. This is
precisely the sort of existence that is proposed for the ātman-Brahman. Despite its ultimately
being beyond conceptualisation and beyond attribution, it is in every circumstance
acknowledged as the totality of existence. It is literally responsible for everything, ‘ultimate’ in
every sense of the word, and so despite being beyond conceptualisation and so on, it remains
the ultimate ‘object’ of liberation. Even in the final analysis, it is there: it is, in fact, all that is
there. I am not at all convinced that this is what Nāgārjuna was working toward when he wrote
the MMK. I earlier discussed how in the Vigrahavyāvartani, Nāgārjuna refuted the notion that
pramāṇas can be ultimately established (and in turn establish some sort of ultimate knowledge)
because nothing at all can be ultimately established: I think that the same thread is woven
throughout the Mūlamadhyamakakārikā, too. There is one particularly telling kārikā at the end
of MMK 10:
P a g e | 260
ātmanaś ca satattvaṃ ye bhāvānāṃ ca pṛthak pṛthak /
nirdiśanti na tān manye śāsanasyārthakovidān //16//
[Those who] proclaim the real nature of the ātman and separate
entities /
[we] do not consider experts in [the Buddha’s] teaching //16//
This criticism is aimed in the MMK at the pudgalavādins, but I think that it can apply with
equal force to those that would advance an ātmavāda (doctrine of ātman): is this not what
Bhattacharya, Murti et al. are trying to do? The Advaitin project is to give an account of the
ātman and its ultimate identity as a microcosmic instantiation of the Brahman. The vast majority
of Advaitin literature is dedicated to orientating practitioners toward this truth or towards
fleshing out the details regarding how they might best try to orientate practitioners towards this
truth.197 Intuitive knowledge of this ultimate identity is the liberating factor. Bhattacharya, as
we have seen, thinks that a śūnyavāda (doctrine of emptiness) is equivalent in all but name to
the ātmavāda of the Advaitins. I hope that my arguments throughout this work illustrate why
this need not be the case.
What is left after all this? For Nāgārjuna we have seen that the end result needs to be
the extinguishing of all cognition; the extinguishing of all reification. This, as I have said, must
include any notions around what is ‘ultimate’, be that in truth or in existence. Once we have
reached this point, talk of ultimate truths or ultimate existence are seen for what they are, mere
prapañca, hindering the pursuit of liberation by providing an idea to cling to (vikalpa). We must
remove this final obstacle so that we might experience what Huntington (2018: 18) calls the
Even the dry philosophical discussions about pramāṇas and epistemological principles is geared toward
providing reliable means by which a dedicated practitioner can analyse the world to come to know Brahman.
197
P a g e | 261
‘consequent immersion in a groundless state of non-abiding—what is referred to in the
Mūlamadhyamakakārikā as the “extinction of conceptual diffusion in emptiness”.’ I think that
this is for Nāgārjuna as for Vasubandhu a state to be experienced in meditation. The halting of
cognition (which constitutes liberation) necessitates the abandonment of views regarding the
ultimate precisely because we could feasibly hold no attachment at all other than the attachment
to the ultimate truth (as we conceive of it). This is still an attachment, and it still has the potential
(or, I suggest, its inevitable consequence is) to breed the usual anxieties and duḥkha in the usual
manners.
Given what has been said above, we eventually need to jettison our concept of ultimacy,
lest we find ourselves clinging to it, surreptitiously sating a reified enquiring ‘I’, the ‘I’ that we
think holds this truth to be ultimate in the first place, and so we end up prolonging our own
duḥkha. For this reason, ‘ultimate truth’ must be a conventional construction. It is undoubtedly
a useful concept, but only up to the point where we no longer need it to distinguish better ideas
from worse ideas. When we reach that summit of practice where all ideas are to be eschewed
and all discursive cognition is to be discarded, we have no place for any idea at all, not even of
an ultimate truth. To keep such an idea is to attach oneself to it, to identify with it, and thus to
allow a more insidious dissatisfaction. Nāgārjuna’s method must culminate in a disregarding of
the concept of ultimacy, and with it, a disregarding of any notion of ātman-Brahman as our
reality. We should not concern ourselves with such pointless questions as ‘what is reality?’, we
should instead just experience without grasping and without reification. As such, ‘ultimate’ is a
conventional designation; a useful fiction which can help our soteriological aim, but which is not
sufficient to provide that same salvation. It is a tool that, like the fabled raft, helps us across the
P a g e | 262
river, but prevents us from traversing the mountain. It is my belief that Nāgārjuna recognised
this, and that the method presented in the MMK is intended to deal with this most insidious of
problems. It is in this sense that I think we can stake the claim that – conventionally! – ‘the
ultimate truth is that there is no ultimate truth’.
P a g e | 263
Concluding Remarks
I have covered a lot of ground in this work – I nevertheless hope that I have maintained a
common thread throughout. I stated at the beginning that my intention was modest; simply to
demonstrate that absolutism need not be a conclusion of Mahāyāna philosophy. My arguments
have been mostly small, but aimed at several smaller targets, each of them different aspects
within the Buddhist corpus. I think this reflects the nature of diversity within the field that we
might broadly call ‘Buddhist philosophy’. All of my small arguments nevertheless add up to one
overarching point: Mahāyāna philosophy – or at least some of the greatest examples of it – need
not result in some type of absolutism. If I am wrong, then it is not an exaggeration to claim that
2000 years of Buddhist praxis and philosophy has been significantly misguided. The stakes are –
in principle– rather high.
I have not covered in detail the ways in which some Buddhists do indeed endorse and
propagate absolutism. There are a couple of reasons for this, the first being that I think we can
take it for granted that this does happen, but not that it should happen. Second, I fear that such
a gargantuan task is just too large for me to have attempted here. Instead, I have tried to
maintain a relatively narrow focus on how we might avoid absolutism in the vein of Madhyamaka,
Yogācāra, and even according to some prajñāpāramitā philosophy. It is in any case true that
some Buddhists do indeed buy into forms of absolutism, and if Bhattacharya had simply stated
this and stopped there, then I would have been in full agreement with him. The issue as I see it
is that going by the arguments presented in the 2015 translation, Bhattacharya thinks that all
Buddhists should endorse absolutism, and those that do not simply misunderstand the point of
Buddhism! I hope to have demonstrated within this work that Bhattacharya (and scholars that
P a g e | 264
follow his lead) are mistaken on some fundamental points. Linguistic interpretations of awkward
Sanskrit terms aside, I think that the bases of both Yogācāra and Madhyamaka are fundamentally
opposed to the endorsement of an Absolute substratum. This is what I would say constitutes the
real backbone of this work.
My main point, then, is that neither Vasubandhu nor Nāgārjuna would endorse the
ultimate existence of the ātman-Brahman. But there is more! They would not argue for its
ultimate non-existence, either. This is because, I have argued, that questioning non-existence,
existence, possible existence and so on is little more than an exercise in dissatisfaction. For
Nāgārjuna, our problems are – in the vast majority of cases – analysable down to belief in entities
that are self-existent, or that have intrinsic nature. It is a predilection to thinking that objects
have svabhāva that is the nub of the problem. This not only applies to Hindu thinkers like
Śaṅkara, who would hold that the ātman-Brahman is the svabhāva of the universe and all it
contains, but also to other Buddhists, who think of entities as conditioned, but conditioned by
unconditioned, intrinsically existent atoms or dharmas. To be intrinsically existent is to have a
svabhāva, and to have a svabhāva is to exist ultimately. Nāgārjuna thinks that all entities are
empty (śūnyatā) of svabhāva, and so Nāgārjuna is thus concerned with denying the possibility of
ultimacy from the very beginning. We saw in the VV that contrary to Śaṅkara, Nāgārjuna does
not think that any ultimate entity whatsoever can possibly be established.
I have also interpreted the Yogācāra of Vasubandhu in a similar manner, arguing that
despite formulating a trisvabhāva doctrine, he too views svabhāva as parikalpita. This is, I argue,
because the way in which Vasubandhu formulated the trisvabhāva doctrine precludes us from
P a g e | 265
taking svabhāva to be an ultimately existent thing. That we must now understand svabhāva to
be a tripartite, dependent concept proves that it cannot be the unitary, self-sufficient entity that
it is sometimes taken to be. It is necessarily dependent and is a simple conceptual imposition.
Far from endorsing some permanent existent, then, I argue that Vasubandhu seeks to liberate
the mind from conceptual proliferation in much the same way as Nāgārjuna: by providing a
method that will allow the practitioner to see that there are no permanent, immutable selfsufficient entities. This clearly rules out the prospect of an ātman-Brahman, transcendent or not.
Having disposed of the problematic issue of svabhāva (and with it, permanent, ultimate
existence), I turned to non-existence. For Nāgārjuna and for Vasubandhu, it is the case that
staunchly preaching that x and y do not ultimately exist is equally as problematic as claiming that
x and y do ultimately exist. I propose that both must then jettison ultimate non-existence in the
same way that they jettison ultimate existence. What we are left with is a denial of ultimacy on
either side, which I think constitutes Nāgārjuna’s ‘middle way’. More than this, we have a total
rejection of metaphysics, a point made most forcefully on the part of Nāgārjuna. I suggest that
both Nāgārjuna and Vasubandhu share a distaste for ultimacy: Nāgārjuna to the point that he
will not even talk about it, and wants to halt all conceptual proliferation, reification and grasping
at the world. For Vasubandhu, the point is realising the emptiness of the flux. He is more
comfortable speaking of metaphysics and metaphysical concepts, just as long as we realise that
they are – when all is said and done – empty of intrinsic existence. What is important is removing
conceptualisation from consciousness, so that we might experience emptiness. That is to say
that Vasubandhu believes that in following the Yogācāra, we realise that our experiences are
empty of intrinsic nature and of ultimate existence, non-existence and so on. For the Yogācārin,
P a g e | 266
we ought to experience without grasping, without system building, and without imposing a
metaphysic. This is, I think, largely the same for the Mādhyamika. The result then is a realisation
that is all but identical in both cases. The difference is the method used to achieve it. What is
clear in both cases is that there is no room for an ātman-Brahman and as such, that
Bhattacharya’s thesis is – in these two cases – mistaken.
Given that the initial claim made by Bhattacharya was that the Mahāyāna put things right
in terms of reorientating Buddhist praxis back toward the ātman-Brahman, it is, I think, sufficient
for me to demonstrate not that the entire Mahāyāna corpus runs contrary to Bhattacharya’s
thesis, but only that parts of it do. I hope that I have shown in the preceding chapters that the
two significant Mahāyāna schools under discussion are those parts. I have supplemented my
arguments for this with some peripheral challenges to Bhattacharya’s choice of citations (the
prajñāpāramitā literature, for example). Whilst I believe that these aid my overall argument,
they are but small additions. My claim is then that Madhyamaka should not endorse absolutism,
and Yogācāra need not endorse absolutism.
According to my interpretations of both
Madhyamaka and Yogācāra, then, absolutism need not be a conclusion of Mahāyāna philosophy.
P a g e | 267
Bibliography
Ajahn Brahm, 2006. MN49 Brahmanimantanika Sutta – Invitation to Brahma. [Online] Available
at: https://bswa.org/teaching/mn49-brahmanimantanika-sutta-invitation-to-brahma-withajahn-brahmavamso/ [Accessed July 2018].
Albahari, M., 2002. Against no-Ātman Theories of Anattā. Asian Philosophy, 12(1), pp.7-20.
Anderson, J., 2012. An Investigation of Moksha in the Advaita Vedanta of Shankara and
Gaudapada. Asian Philosophy, 22(3), pp.275-87.
Asaṅga, 1907. Mahāyānasūtrālaṃkārabhāṣya: Exposé de la doctrine du Grand Véhicule, vol. I.
Paris.
Asaṅga, 1911. Mahāyānasūtrālaṃkārabhāṣya: Exposé de la doctrine du Grand Véhicule, vol. II.
Paris.
Aṣṭasāhasrikā Prajñāpāramitā, 1960. Darbhanga: The Mithila Institute of Post-Graduate Studies
and Research in Sanskrit Learning. Available at: http://www.dsbcproject.org/canontext/book/68.
Attwood, J., 2014. Manomaya: Background to Mind-Made Bodies. [Online] Available at:
http://jayarava.blogspot.com/2014/11/manomaya-background-to-mind-made-bodies.html
[Accessed July 2018].
Ayer, A., 1956. The Problem of Knowledge. Edinburgh: Penguin Books.
Bartley, C., 2011. An Introduction to Indian Philosophy. 1st ed. London: Continuum.
Betty, S., 2010. Dvaita, Advaita, and Viśiṣṭadvaita: Contrasting Views of Mokṣa. Asian
Philosophy, 20(2), pp.215-24.
Bhagavad-Gītā, 2007. Translated by A. C. Bhaktivedanta Swami Prabhupāda. Bhaktivedanta
Book Trust.
Bhattacharya, K., 2015. The Ātman-Brahman in Ancient Buddhism. Cotopaxi, Colorado: Canon
Publications.
Bhattacharya, K., Johnston, E.H. & Kunst, A., 1978. The Dialectical Method of Nāgārjuna:
Vigrahavyāvartanī. Delhi: Motilal Banarsidass.
Bhikkhu Sujato, 2011. Sujato's Blog. [Online] Available at:
https://sujato.wordpress.com/2011/05/13/vinna%E1%B9%87a-is-not-nibbana-really-it-justisn%E2%80%99t/ [Accessed March 2018].
P a g e | 268
Biderman, S., 1982. A 'Constitutive' God: An Indian Suggestion. Philosophy East and West, 32(4),
pp.425-37.
Blumenthal, D.R., 2009. Maimonides' Philosophic Mysticism. Daat: A Journal of Jewish
Philosophy & Kabbalah / דעת: כתב-וקבלה יהודית לפילוסופיה עת, (64/66), pp.V-XXV.
Burton, D., 2001. Emptiness Appraised: A Critical Study of Nāgārjuna's Philosophy. Delhi: Motilal
Banarsidass.
Candrakīrti, [1903-13] 1970. Prasannapadā. Madhyamakavṛtti. Ed. L. de La Vallée Poussin.
Mūlamadhyamakakārikās (Mādhyamikasūtras) de Nāgārjuna, avec le commentaire de
Candrakīrti. St. Petersburg: Bibliotheca Buddhica IV, 1903-1913. Reprinted Osnabrück: Biblio
Verlag, 1970.
Chalmers, R., 1898. Art. XII.—Tathāgata. Journal of the Royal Asiatic Society of Great Britain &
Ireland, 30(1), pp.103-15.
Conze, E., 1973. The Perfection of Wisdom in Eight Thousand Lines & its Verse Summary.
Bolinas, California: Four Seasons Foundation.
Conze, E., 1983. Buddhist Thought in India: Three Phases of Buddhist Philosophy. 2nd ed.
London: George Allen & Unwin.
Conze, E., 1993. Buddhism: A Short History. Oxford: Oneworld.
Cox, C., 1995. Disputed Dharmas: Early Buddhist Theories on Existence. Tokyo: The International
Institute for Buddhist Studies.
Deutsch, E., 1969. Advaita Vedānta: A Philosophical Reconstruction. Honolulu: East-West Center
Press.
Dharmachakra Translation Committee, 2014. Ornament of the Great Vehicle Sūtras: Maitreya's
Mahāyānasūtrālaṃkāra. Boston: Snow Lion.
Dīgha Nikāya, 2016. Aggaññasutta. [Online] Available at: https://suttacentral.net/pi/dn27
[Accessed February 2018].
Dīgha Nikāya, 2016. Kevaṭṭasutta. [Online] Available at: https://suttacentral.net/pi/dn11
[Accessed February 2018].
Dionysius the Areopagite, 1923. The Mystical Theology. Fintry, Brook nr. Godalming: The Shrine
of Wisdom. Available at: http://www.esotericarchives.com/oracle/dionys1.htm# [accessed 8
July 2017].
Dunne, J.D., 2004. Foundations of Dharmakīrti's Philosophy. Somerville: Wisdom Publications.
P a g e | 269
Eckel, M.D., 1992. To See the Buddha: A Philosopher's Quest for the Meaning of Emptiness.
Princeton: Princeton University Press.
Eltschinger, V. & Raité, I., 2013. Self, No-Self, and Salvation: Dharmkīrti's Critique of the Notions
of Self and Person. Budapest: OAW.
Ergardt, J.T., 1977. Faith and Knowledge in Early Buddhism: An Analysis of the Contextual
Structures of an Arahant-Formula in the Majjhima-Nikāya. Leiden: EJ Brill.
Ferraro, G., 2013. A Criticism of M. Siderits and J. L. Garfield’s ‘Semantic Interpretation’ of
Nāgārjuna's Theory of Two Truths. Journal of Indian Philosophy, 41(2), pp.195-219.
Ferraro, G., 2014. Grasping Snakes and Touching Elephants: A Rejoinder to Garfield and Siderits.
Journal of Indian Philosophy, 42(4), pp.451–62.
Ferraro, G., 2017. Realistic-Antimetaphysical Reading Vs Any Nihilistic Interpretation of
Madhyamaka. Journal of Indian Philosophy, 45(1), pp.73-98.
Fideler, D., 2002. Neoplatonism and the Cosmological Revolution: Holism, Fractal Geometry and
Mind-in-Nature. In B.H. R, ed. Neoplatonism and Contemporary Thought, Part Two. New York:
State University of New York Press. pp.103-19.
Fost, F.F., 1998. Playful Illusion: The Making of Worlds in Advaita Vedānta. Philosophy East and
West, 48(3), pp.387-405.
Garfield, J.L., 1995. The Fundamental Wisdom of the Middle Way: Nāgārjuna’s
Mūlamadhyamakakārikā. New York: Oxford University Press.
Gold, J.C., 2015. Paving the Great Way; Vasubandhu's Unifying Buddhist Philosophy. New York:
Columbia University Press.
Gombrich, R., 2009. What the Buddha Thought. London: Equinox.
Grosnick, W.H., 1979. The Zen Master Dōgen's understanding of the Buddha-nature in light of
the historical development of the Buddha-nature concept in India, China, and Japan.
Unpublished PhD Thesis. University of Wisconsin.
Harrison, P., 1992. Is the Dharma-kaya the Real "Phantom Body" of the Buddha? The Journal of
the International Association of Buddhist Studies, 15(1), pp.44-94.
Harvey, P., 1995. The Selfless Mind: Personality, Consciousness and Nirvāṇa. Curzon Press.
Hume, R.E., 1921. The Thirteen Principal Upanishads: Translated from the Sanskrit with an
Outline of the Philosophy of the Upanishads and an Annotated Bibliography. London: Oxford
University Press.
P a g e | 270
Huntington, C.W.J., March 2018. Nāgārjuna’s Fictional World. Journal of Indian Philosophy,
46(1), pp.153–77.
Ireland, J.D., 1977. The Buddha's Advice to Bāhiya. Pāli Buddhist Review, 2(3), pp.159-61.
James, W., 1917. The Varieties of Religious Experience: A Study in Human Nature. London:
Longmans, Green, And Co.
Jantrasrisalai, C., 2008. Early Buddhist Dhammakāya: Its Philosophical and Soteriological
Significance. Ph.D Thesis. Sydney: Department of Studies in Religion, University of Sydney.
Jayatilleke, K.N., 1963. Eary Buddhist Theory of Knowledge. London: George Allen & Unwin.
Karunadasa, Y., 1996. The Dhamma Theory: Philosophical Cornerstone of the Abhidhamma.
Kandy: Buddhist Publication Society.
Khuddaka Nikāya, 2013. Udāna: Paṭhamanibbānasuttaṃ. [Online] Available at:
https://www.accesstoinsight.org/tipitaka/sltp/Ud_utf8.html#pts.080 [Accessed February 2018].
Khuddaka Nikāya, 2013. Udāna: Tatiyanibbānasuttaṃ. [Online] Available at:
https://www.accesstoinsight.org/tipitaka/sltp/Ud_utf8.html#pts.080 [Accessed February 2018].
King, R., 1995. Early Advaita Vedānta and Buddhism: The Mahāyāna context of the
Gauḍapādīya-kārikā. New York: State University of New York Press.
King, R., 1998. Early Advaita and Madhyamaka Buddhism: The Case of the "Gauḍapādīyakārikā".
International Journal of Hindu Studies, 2(1), pp.67-83.
Lindtner, C., 1997. The Problem of Precanonical Buddhism. Buddhist Studies Review, 14(2),
pp.109-39.
Lindtner, C., 1999. From Brahmanism to Buddhism. Asian Philosophy, 9(1), pp.5-37.
Lipner, J., 2010. Hindus: their religious beliefs and practices. Oxon: Routledge.
Lusthaus, D., 2002. Buddhist Phenomenology: A Philosophical Investigation of Yogācāra
Buddhism and the Chʼeng Wei-Shih Lun. New York: RoutledgeCurzon.
Madhyamakahṛdayaḥ, 2001. Chennai: The Adyar Library and Research Center.
Maimonides, M., 1904. The Guide for the Perplexed. Translated by M. Friedländer. London:
Routledge & Kegan Paul Ltd. Available at: http://www.sacred-texts.com/jud/gfp/gfp000.htm
[accessed 24 July 2017].
Maitreyanātha/Āryāsaṅga, 2004. The Universal Vehicle Discourse Literature
(Mahāyānasütrālamkāra), Together with Its Commentary (Bhāṣya) by Vasubandhu. New York:
American Institute of Buddhist Studies.
P a g e | 271
Majjhima Nikāya, 2005. Brahmanimantanikasutta. [Online] Available at:
https://legacy.suttacentral.net/pi/mn49 [Accessed June 2018].
Makransky, J., 2003. Tathagata. [Article for Encyclopedia of Religion, 2nd edition, 2003]
Available at: http://www.johnmakransky.org/downloads/Tathagata.pdf [Accessed 16 January
2018].
McCagney, N., 1997. Nāgārjuna and the Philosophy of Openness. Lanham, Maryland: Rowman
& Littlefield.
Miller, J.S., Summer 2009. Are Mystical Experiences Evidence for the Existence of a
Transcendent Reality? Evaluating Eugene d’Aquili and Andrew Newberg’s Argument for
Absolute Unitary Being. Florida Philosophical Review, IX(1), pp.40-55.
Mitomo, R., 1983-1984. An aspect of Dharma-śarīra. Journal of Indian and Buddhist Studies
(Indogaku Bukkyogaku Kenkyu), 32(2), pp.1120-1115.
Mohanta, D., 1997. The Concept of Pramāṇa and the Sceptical Arguments of Nāgārjuna. Indian
Philosophical Quarterly, 24(1), pp.53-72.
Monier-Williams, M., 1960. A Sanskrit-English Dictionary. Oxford: Clarendon Press. Available at:
https://archive.org/stream/sanskritenglishd00moniuoft#page/n4/mode/1up.
Murti, T.R.V., 2016. The Central Philosophy of Buddhism: A Study of the Mādhyamika System.
Delhi: Motilal Banarsidass.
Nāgārjuna, 2007. In Praise of Dharmadhātu. Translated by K. Brunhölzl. Ithaca, New York: Snow
Lion Publications.
Newberg, A.B. & D'Aquili, E.G., 2000. The Neuropsychology of Religious and Spiritual
Experience. Journal of Consciousness Studies, 7(11-12), pp.251-66.
Nicholson, H., 2007. Two Apologetic Moments in Sankara's Concept of Brahman. Journal of
Religion, 87(4), pp.528–55.
Nikhilānanda, S., 1962. Ātmabodhaḥ: Self-Knowledge of Sri Sankaracharya. Madras: Sri
Ramakrishna Math.
OED Online, 2017. mystic, n. and adj. Oxford: Oxford University Press. Available at:
http://www.oed.com.liverpool.idm.oclc.org/view/Entry/124647?redirectedFrom=mystic
[accessed 24 July 2017].
Olivelle, P., 1998. The Early Upanisads: Annotated Text and Translation. Oxford: Oxford
University Press.
P a g e | 272
Pérez-Remón, J., 1980. Self and Non-Self in Early Buddhism. The Hague, The Netherlands:
Mouton Publishers.
Preti, A.A., July 2014. Brahmānubhava as Überpramāṇa in Advaita Vedānta: Revisiting an Old
Debate. Philosophy East and West, 64(3), pp.718-39.
Radhakrishnan, S., 1962. Indian Philosophy. London: George Allen and Unwin Ltd.
Rambachan, A., 2006. The Advaita Worldview: God, World, and Humanity. Albany, New York:
SUNY Press.
Robinson, N., 1998. The unity of God In: Ash‘ariyya and Mu‘tazila. In Routledge Encyclopedia of
Philosophy. Taylor and Francis.
Ronkin, N., 2005. Early Buddhist Metaphysics: The Making of a Philosophical Tradition. Oxon:
Routledge.
Saṃyutta Nikāya, 1894. Part IV: Salāyatana-vagga. London: Pāli Text Society.
Saṃyutta Nikāya, 2016. Vakkalisutta. [Online] Available at: https://suttacentral.net/pi/sn22.87
[Accessed February 2018].
Śaṅkara, 1950. The Bṛihadāraṇyaka Upaniṣad with Commentary by Śaṅkarācārya. Translated by
Swami Mādhavānanda. Almora, India: Advaita Ashrama.
Śaṅkara, 2009. The Vedānta sūtras with the commentary of Śaṅkara (Vol. 2). Translated by G.
Thibaut. Delhi: Motilal Banarsidass.
Śaṅkara, 2011. The Vedānta sūtras with the commentary of Śaṅkara (Vol. 1). Translated by G.
Thibaut. Delhi: Motilal Banarsidass.
Schimmel, A., 1975. Mystical Dimensions of Islam. Chapel Hill: University of North Carolina
Press.
Sharma, A., 1992. Is Anubhava a Pramāṅa According to Śaṇkara? Philosophy East and West,
42(3), pp.517-26.
Siderits, M., 2003. On the Soteriological Significance of Emptiness. Contemporary Buddhism,
4(1), pp.9-23.
Siderits, M., 2016. Thinking on Empty: Madhyamaka Anti-realism and Canons of Rationality. In
M. Siderits & J. Westerhoff, eds. Studies in Buddhist Philosophy. Oxford: Oxford University
Press. pp.24-38.
Siderits, M. & Katsura, S., 2013. Nāgārjuna's Middle Way: Mūlamadhyamakakārikā. Boston:
Wisdom Publications.
P a g e | 273
Siderits, M., Tillemans, T. & Charkabarti, A., eds., 2011. Apoha: Buddhist Nominalism and
Human Cognition. New York: Colombia University Press.
Sponberg, A., 1979. Dynamic Liberation in Yogācāra Buddhism. The Journal of the International
Association of Buddhist Studies, 2(1), pp.44-64.
Sprung, M., 1979. Lucid Expression of the Middle Way: The Essential Chapters from the
Prasannapadā of Candrakīrti. London: Routledge & Kegan Paul.
Suthren Hirst, J.G., 2005. Śaṃkara's Advaita Vedānta: A Way of Teaching. New York:
RoutledgeCurzon.
Swearer, D.K., 1973. Control and Freedom: The Structure of Buddhist Meditation in the Pāli
Suttas. Philosophy East and West, 23(4 (Oct. 1973)), pp.435-55.
Thanissaro Bhikkhu, 1996. No-self or Not-self? [Online] Available at:
https://www.accesstoinsight.org/lib/authors/thanissaro/notself2.html [Accessed July 2018].
Thanissaro Bhikkhu, 1997. Kevatta (Kevaddha) Sutta. [Online] (English Translation) Available at:
https://www.accesstoinsight.org/tipitaka/dn/dn.11.0.than.html#fn-1 [Accessed February
2018].
Thanissaro Bhikkhu, 1997. Samanupassana Sutta: Assumptions. [Online] Available at:
http://www.accesstoinsight.org/tipitaka/sn/sn22/sn22.047.than.html [Accessed March 2017].
Thanissaro Bhikkhu, 1998. Mūlapariyāya Sutta. [Online] Available at:
https://www.accesstoinsight.org/tipitaka/mn/mn.001.than.html [Accessed August 2018].
Thanissaro Bhikkhu, 2007. Brahmanimantanika Sutta: The Brahma Invitation. [Online] Available
at: https://www.accesstoinsight.org/tipitaka/mn/mn.049.than.html#fn-1 [Accessed July 2018].
Thanissaro Bhikkhu, 2017. Aṅguttara Nikāya. [Online] Available at:
http://www.accesstoinsight.org/tipitaka/an/ [Accessed 30 January 2017].
The Holy Bible: New King James Version, 1982. Nashville, TN: Thomas Nelson.
Tillemans, T.J.F., 2019. Mādhyamikas Playing Bad Hands: The Case of Customary Truth. Journal
of Indian Philosophy, 47(4), pp.635–44.
Tillemans, T.J.F., 1999. Scripture, Logic, Language: Essays on Dharmakīrti and His Tibetan
Successors. Boston: Wisdom Publications.
Triṃśikāvijñaptikārikāḥ, 2004. [Online] Available at: http://www.dsbcproject.org/canontext/content/337/1390 [Accessed August 2018].
Trivedi, S., 2005. Idealism and Yogacara Buddhism. Asian Philosophy, 15(3), pp.231-46.
P a g e | 274
Vallée-Poussin, L.d.L., 1930. The Two Nirvāṇadhātus According to the Vibhāṣa. The Indian
Historical Quarterly, March. pp.39-45.
Vasu, S.C., 1905. Kathopanishad with the Sanskrit Text, Anvaya, Vritti, Word Meaning,
Translation, Notes and Index. Allahbad: The Panini Office.
Vasubandhu, 1991. Abhidharmakośabhāṣyam: Volume IV. Berkeley: Asian Humanities Press.
Viṃśatikākārikā, 2004. [Online] Available at: http://www.dsbcproject.org/canontext/content/340/1393 [Accessed August 2018].
Vireswarananda, S., 2014. Brahma Sūtras According to Śri Śaṅkara. Translated by S.
Vireswarananda. Kolkata: Advaita Ashrama.
von Rospatt, A., 1995. The Buddhist Doctrine of Momentariness: A Survey of the Origins and
Early Phase of This Doctrine up to Vasubandhu. Stuttgart: Franz Steiner Verlag.
Walshe, M., 1995. The Long Discourses of the Buddha: A Translation of the Dīgha Nikāya.
Boston: Wisdom Publications.
Wensinck, A.J., 1932. The Muslim Creed: Its Genesis and Historical Development. New York:
Barnes & Noble.
Westerhoff, J.C., 2017. 'Nāgārjuna'. In E.N. Zalta, ed. The Stanford Encyclopedia of Philosophy
(Spring 2017 Edition).
Whaling, F., 1979. Śaṅkara and Buddhism. Journal of Indian Philosophy, 7(1), pp.1-42.
Whitney, W.D., 1879. A Sanskrit Grammar. London: Trübner & Co.
Wildberg, C., Spring 2016. Neoplatomnism. In E.N. Zalta, ed. Stanford Encyclopedia of
Philosophy.
WIlliams, P., 2000. Buddhist Thought. London: Routledge.
Williams, P., ed., 2005. Buddhism: Critical Concepts in Religious Studies. Volume IV: Abhidharma
and Madhyamaka. Abingdon, Oxon: Routledge.
Wolfson, E.R. & וולפסון, א.ר., 1994. בראשיתה בקבלה ושלילה חיוב תארי/ Negative Theology and
Positive Assertion in the Early Kabbalah. Daat: A Journal of Jewish Philosophy & Kabbalah / דעת:
כתב-וקבלה יהודית לפילוסופיה עת, (32/33), pp.V-XXII.
Wood, T.E., 1995. Nagarjunian Disputations: A Philosophical Journey through an Indian Looking
Glass. Delhi: Sri Satguru Publications.
P a g e | 275
Wynne, A., 2009. Early evidence for the 'no self' doctrine? A note on the Second anātman
Teaching of the Second Sermon. [Online] Oxford Centre for Buddhist Studies Available at:
https://ocbs.org/wp-content/uploads/2015/09/awynne2009atijbs.pdf [Accessed May 2018].
Xing, G., 2005. The Concept of the Buddha: Its evolution from early Buddhism to the trikāya
theory. Oxon: RoutledgeCurzon.